From 6250a3724618bbb13a47e16fe6a2e5ad860b8433 Mon Sep 17 00:00:00 2001 From: Waldemar Brodkorb Date: Fri, 23 Jun 2017 16:33:53 +0200 Subject: linux: update rt-patch for 4.4/4.9/rpi-git, update 4.11.x kernel --- target/config/Config.in.kernelversion | 2 +- target/linux/config/Config.in.kernel | 6 +- target/linux/patches/4.4.72/patch-realtime | 13014 ++++----- target/linux/patches/4.9.32/patch-realtime | 9018 +++---- .../patch-realtime | 26047 ------------------ .../patch-realtime | 26986 +++++++++++++++++++ 6 files changed, 38225 insertions(+), 36848 deletions(-) delete mode 100644 target/linux/patches/72134397d72079a533c8fc742701fdc7f5ae7c5b/patch-realtime create mode 100644 target/linux/patches/8327f281c8d5083b13b3a4e6e2a36ee7c3d67913/patch-realtime (limited to 'target') diff --git a/target/config/Config.in.kernelversion b/target/config/Config.in.kernelversion index 4b6cfd467..7293d392b 100644 --- a/target/config/Config.in.kernelversion +++ b/target/config/Config.in.kernelversion @@ -28,7 +28,7 @@ config ADK_TARGET_KERNEL_VERSION_GIT select ADK_TARGET_KERNEL_IMAGE if ADK_TARGET_SYSTEM_BEAGLEBONE_BLACK config ADK_TARGET_KERNEL_VERSION_4_11 - bool "4.11.5" + bool "4.11.6" depends on !ADK_TARGET_ARCH_AVR32 depends on !ADK_TARGET_ARCH_CRIS depends on !ADK_TARGET_ARCH_METAG diff --git a/target/linux/config/Config.in.kernel b/target/linux/config/Config.in.kernel index b53b5f025..0400fbd1e 100644 --- a/target/linux/config/Config.in.kernel +++ b/target/linux/config/Config.in.kernel @@ -166,8 +166,7 @@ config ADK_KERNEL_PREEMPT_RTB select ADK_KERNEL_HIGH_RES_TIMERS select ADK_KERNEL_PATCH_REALTIME depends on !ADK_KERNEL_OPROFILE - depends on ADK_TARGET_KERNEL_VERSION_4_1 \ - || ADK_TARGET_KERNEL_VERSION_4_4 \ + depends on ADK_TARGET_KERNEL_VERSION_4_4 \ || ADK_TARGET_KERNEL_VERSION_4_9 \ || (ADK_TARGET_KERNEL_VERSION_GIT && ADK_TARGET_BOARD_BCM28XX) help @@ -178,8 +177,7 @@ config ADK_KERNEL_PREEMPT_RT_FULL select ADK_KERNEL_HIGH_RES_TIMERS select ADK_KERNEL_PATCH_REALTIME depends on !ADK_KERNEL_OPROFILE - depends on ADK_TARGET_KERNEL_VERSION_4_1 \ - || ADK_TARGET_KERNEL_VERSION_4_4 \ + depends on ADK_TARGET_KERNEL_VERSION_4_4 \ || ADK_TARGET_KERNEL_VERSION_4_9 \ || (ADK_TARGET_KERNEL_VERSION_GIT && ADK_TARGET_BOARD_BCM28XX) help diff --git a/target/linux/patches/4.4.72/patch-realtime b/target/linux/patches/4.4.72/patch-realtime index 1d870326c..b49cd6f03 100644 --- a/target/linux/patches/4.4.72/patch-realtime +++ b/target/linux/patches/4.4.72/patch-realtime @@ -1,1153 +1,1002 @@ -diff -Nur linux-4.4.62.orig/arch/arm/include/asm/switch_to.h linux-4.4.62/arch/arm/include/asm/switch_to.h ---- linux-4.4.62.orig/arch/arm/include/asm/switch_to.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/include/asm/switch_to.h 2017-04-18 17:38:07.902637922 +0200 -@@ -3,6 +3,13 @@ - - #include - -+#if defined CONFIG_PREEMPT_RT_FULL && defined CONFIG_HIGHMEM -+void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p); -+#else -+static inline void -+switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) { } -+#endif +diff --git a/Documentation/hwlat_detector.txt b/Documentation/hwlat_detector.txt +new file mode 100644 +index 000000000000..cb61516483d3 +--- /dev/null ++++ b/Documentation/hwlat_detector.txt +@@ -0,0 +1,64 @@ ++Introduction: ++------------- + - /* - * For v7 SMP cores running a preemptible kernel we may be pre-empted - * during a TLB maintenance operation, so execute an inner-shareable dsb -@@ -25,6 +32,7 @@ - #define switch_to(prev,next,last) \ - do { \ - __complete_pending_tlbi(); \ -+ switch_kmaps(prev, next); \ - last = __switch_to(prev,task_thread_info(prev), task_thread_info(next)); \ - } while (0) - -diff -Nur linux-4.4.62.orig/arch/arm/include/asm/thread_info.h linux-4.4.62/arch/arm/include/asm/thread_info.h ---- linux-4.4.62.orig/arch/arm/include/asm/thread_info.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/include/asm/thread_info.h 2017-04-18 17:38:07.902637922 +0200 -@@ -49,6 +49,7 @@ - struct thread_info { - unsigned long flags; /* low level flags */ - int preempt_count; /* 0 => preemptable, <0 => bug */ -+ int preempt_lazy_count; /* 0 => preemptable, <0 => bug */ - mm_segment_t addr_limit; /* address limit */ - struct task_struct *task; /* main task structure */ - __u32 cpu; /* cpu */ -@@ -142,7 +143,8 @@ - #define TIF_SYSCALL_TRACE 4 /* syscall trace active */ - #define TIF_SYSCALL_AUDIT 5 /* syscall auditing active */ - #define TIF_SYSCALL_TRACEPOINT 6 /* syscall tracepoint instrumentation */ --#define TIF_SECCOMP 7 /* seccomp syscall filtering active */ -+#define TIF_SECCOMP 8 /* seccomp syscall filtering active */ -+#define TIF_NEED_RESCHED_LAZY 7 - - #define TIF_NOHZ 12 /* in adaptive nohz mode */ - #define TIF_USING_IWMMXT 17 -@@ -152,6 +154,7 @@ - #define _TIF_SIGPENDING (1 << TIF_SIGPENDING) - #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) - #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) -+#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) - #define _TIF_UPROBE (1 << TIF_UPROBE) - #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) - #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) -@@ -167,7 +170,8 @@ - * Change these and you break ASM code in entry-common.S - */ - #define _TIF_WORK_MASK (_TIF_NEED_RESCHED | _TIF_SIGPENDING | \ -- _TIF_NOTIFY_RESUME | _TIF_UPROBE) -+ _TIF_NOTIFY_RESUME | _TIF_UPROBE | \ -+ _TIF_NEED_RESCHED_LAZY) - - #endif /* __KERNEL__ */ - #endif /* __ASM_ARM_THREAD_INFO_H */ -diff -Nur linux-4.4.62.orig/arch/arm/Kconfig linux-4.4.62/arch/arm/Kconfig ---- linux-4.4.62.orig/arch/arm/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/Kconfig 2017-04-18 17:38:07.902637922 +0200 -@@ -33,7 +33,7 @@ - select HARDIRQS_SW_RESEND - select HAVE_ARCH_AUDITSYSCALL if (AEABI && !OABI_COMPAT) - select HAVE_ARCH_BITREVERSE if (CPU_32v7M || CPU_32v7) && !CPU_32v6 -- select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL && !CPU_ENDIAN_BE32 -+ select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL && !CPU_ENDIAN_BE32 && !PREEMPT_RT_BASE - select HAVE_ARCH_KGDB if !CPU_ENDIAN_BE32 - select HAVE_ARCH_SECCOMP_FILTER if (AEABI && !OABI_COMPAT) - select HAVE_ARCH_TRACEHOOK -@@ -68,6 +68,7 @@ - select HAVE_PERF_EVENTS - select HAVE_PERF_REGS - select HAVE_PERF_USER_STACK_DUMP -+ select HAVE_PREEMPT_LAZY - select HAVE_RCU_TABLE_FREE if (SMP && ARM_LPAE) - select HAVE_REGS_AND_STACK_ACCESS_API - select HAVE_SYSCALL_TRACEPOINTS -diff -Nur linux-4.4.62.orig/arch/arm/kernel/asm-offsets.c linux-4.4.62/arch/arm/kernel/asm-offsets.c ---- linux-4.4.62.orig/arch/arm/kernel/asm-offsets.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/kernel/asm-offsets.c 2017-04-18 17:38:07.902637922 +0200 -@@ -65,6 +65,7 @@ - BLANK(); - DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); - DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); -+ DEFINE(TI_PREEMPT_LAZY, offsetof(struct thread_info, preempt_lazy_count)); - DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit)); - DEFINE(TI_TASK, offsetof(struct thread_info, task)); - DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); -diff -Nur linux-4.4.62.orig/arch/arm/kernel/entry-armv.S linux-4.4.62/arch/arm/kernel/entry-armv.S ---- linux-4.4.62.orig/arch/arm/kernel/entry-armv.S 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/kernel/entry-armv.S 2017-04-18 17:38:07.902637922 +0200 -@@ -215,11 +215,18 @@ - #ifdef CONFIG_PREEMPT - get_thread_info tsk - ldr r8, [tsk, #TI_PREEMPT] @ get preempt count -- ldr r0, [tsk, #TI_FLAGS] @ get flags - teq r8, #0 @ if preempt count != 0 -+ bne 1f @ return from exeption -+ ldr r0, [tsk, #TI_FLAGS] @ get flags -+ tst r0, #_TIF_NEED_RESCHED @ if NEED_RESCHED is set -+ blne svc_preempt @ preempt! ++The module hwlat_detector is a special purpose kernel module that is used to ++detect large system latencies induced by the behavior of certain underlying ++hardware or firmware, independent of Linux itself. The code was developed ++originally to detect SMIs (System Management Interrupts) on x86 systems, ++however there is nothing x86 specific about this patchset. It was ++originally written for use by the "RT" patch since the Real Time ++kernel is highly latency sensitive. + -+ ldr r8, [tsk, #TI_PREEMPT_LAZY] @ get preempt lazy count -+ teq r8, #0 @ if preempt lazy count != 0 - movne r0, #0 @ force flags to 0 -- tst r0, #_TIF_NEED_RESCHED -+ tst r0, #_TIF_NEED_RESCHED_LAZY - blne svc_preempt -+1: - #endif - - svc_exit r5, irq = 1 @ return from exception -@@ -234,8 +241,14 @@ - 1: bl preempt_schedule_irq @ irq en/disable is done inside - ldr r0, [tsk, #TI_FLAGS] @ get new tasks TI_FLAGS - tst r0, #_TIF_NEED_RESCHED -+ bne 1b -+ tst r0, #_TIF_NEED_RESCHED_LAZY - reteq r8 @ go again -- b 1b -+ ldr r0, [tsk, #TI_PREEMPT_LAZY] @ get preempt lazy count -+ teq r0, #0 @ if preempt lazy count != 0 -+ beq 1b -+ ret r8 @ go again ++SMIs are usually not serviced by the Linux kernel, which typically does not ++even know that they are occuring. SMIs are instead are set up by BIOS code ++and are serviced by BIOS code, usually for "critical" events such as ++management of thermal sensors and fans. Sometimes though, SMIs are used for ++other tasks and those tasks can spend an inordinate amount of time in the ++handler (sometimes measured in milliseconds). Obviously this is a problem if ++you are trying to keep event service latencies down in the microsecond range. + - #endif - - __und_fault: -diff -Nur linux-4.4.62.orig/arch/arm/kernel/entry-common.S linux-4.4.62/arch/arm/kernel/entry-common.S ---- linux-4.4.62.orig/arch/arm/kernel/entry-common.S 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/kernel/entry-common.S 2017-04-18 17:38:07.906638076 +0200 -@@ -36,7 +36,9 @@ - UNWIND(.cantunwind ) - disable_irq_notrace @ disable interrupts - ldr r1, [tsk, #TI_FLAGS] @ re-check for syscall tracing -- tst r1, #_TIF_SYSCALL_WORK | _TIF_WORK_MASK -+ tst r1, #((_TIF_SYSCALL_WORK | _TIF_WORK_MASK) & ~_TIF_SECCOMP) -+ bne fast_work_pending -+ tst r1, #_TIF_SECCOMP - bne fast_work_pending - - /* perform architecture specific actions before user return */ -@@ -62,8 +64,11 @@ - str r0, [sp, #S_R0 + S_OFF]! @ save returned r0 - disable_irq_notrace @ disable interrupts - ldr r1, [tsk, #TI_FLAGS] @ re-check for syscall tracing -- tst r1, #_TIF_SYSCALL_WORK | _TIF_WORK_MASK -+ tst r1, #((_TIF_SYSCALL_WORK | _TIF_WORK_MASK) & ~_TIF_SECCOMP) -+ bne do_slower_path -+ tst r1, #_TIF_SECCOMP - beq no_work_pending -+do_slower_path: - UNWIND(.fnend ) - ENDPROC(ret_fast_syscall) - -diff -Nur linux-4.4.62.orig/arch/arm/kernel/patch.c linux-4.4.62/arch/arm/kernel/patch.c ---- linux-4.4.62.orig/arch/arm/kernel/patch.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/kernel/patch.c 2017-04-18 17:38:07.906638076 +0200 -@@ -15,7 +15,7 @@ - unsigned int insn; - }; - --static DEFINE_SPINLOCK(patch_lock); -+static DEFINE_RAW_SPINLOCK(patch_lock); - - static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags) - __acquires(&patch_lock) -@@ -32,7 +32,7 @@ - return addr; - - if (flags) -- spin_lock_irqsave(&patch_lock, *flags); -+ raw_spin_lock_irqsave(&patch_lock, *flags); - else - __acquire(&patch_lock); - -@@ -47,7 +47,7 @@ - clear_fixmap(fixmap); - - if (flags) -- spin_unlock_irqrestore(&patch_lock, *flags); -+ raw_spin_unlock_irqrestore(&patch_lock, *flags); - else - __release(&patch_lock); - } -diff -Nur linux-4.4.62.orig/arch/arm/kernel/process.c linux-4.4.62/arch/arm/kernel/process.c ---- linux-4.4.62.orig/arch/arm/kernel/process.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/kernel/process.c 2017-04-18 17:38:07.906638076 +0200 -@@ -319,6 +319,30 @@ - } - - #ifdef CONFIG_MMU -+/* -+ * CONFIG_SPLIT_PTLOCK_CPUS results in a page->ptl lock. If the lock is not -+ * initialized by pgtable_page_ctor() then a coredump of the vector page will -+ * fail. -+ */ -+static int __init vectors_user_mapping_init_page(void) -+{ -+ struct page *page; -+ unsigned long addr = 0xffff0000; -+ pgd_t *pgd; -+ pud_t *pud; -+ pmd_t *pmd; ++The hardware latency detector works by hogging all of the cpus for configurable ++amounts of time (by calling stop_machine()), polling the CPU Time Stamp Counter ++for some period, then looking for gaps in the TSC data. Any gap indicates a ++time when the polling was interrupted and since the machine is stopped and ++interrupts turned off the only thing that could do that would be an SMI. + -+ pgd = pgd_offset_k(addr); -+ pud = pud_offset(pgd, addr); -+ pmd = pmd_offset(pud, addr); -+ page = pmd_page(*(pmd)); ++Note that the SMI detector should *NEVER* be used in a production environment. ++It is intended to be run manually to determine if the hardware platform has a ++problem with long system firmware service routines. + -+ pgtable_page_ctor(page); ++Usage: ++------ + -+ return 0; -+} -+late_initcall(vectors_user_mapping_init_page); ++Loading the module hwlat_detector passing the parameter "enabled=1" (or by ++setting the "enable" entry in "hwlat_detector" debugfs toggled on) is the only ++step required to start the hwlat_detector. It is possible to redefine the ++threshold in microseconds (us) above which latency spikes will be taken ++into account (parameter "threshold="). + - #ifdef CONFIG_KUSER_HELPERS - /* - * The vectors page is always readable from user space for the -diff -Nur linux-4.4.62.orig/arch/arm/kernel/signal.c linux-4.4.62/arch/arm/kernel/signal.c ---- linux-4.4.62.orig/arch/arm/kernel/signal.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/kernel/signal.c 2017-04-18 17:38:07.906638076 +0200 -@@ -572,7 +572,8 @@ - */ - trace_hardirqs_off(); - do { -- if (likely(thread_flags & _TIF_NEED_RESCHED)) { -+ if (likely(thread_flags & (_TIF_NEED_RESCHED | -+ _TIF_NEED_RESCHED_LAZY))) { - schedule(); - } else { - if (unlikely(!user_mode(regs))) -diff -Nur linux-4.4.62.orig/arch/arm/kernel/smp.c linux-4.4.62/arch/arm/kernel/smp.c ---- linux-4.4.62.orig/arch/arm/kernel/smp.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/kernel/smp.c 2017-04-18 17:38:07.906638076 +0200 -@@ -230,8 +230,6 @@ - flush_cache_louis(); - local_flush_tlb_all(); - -- clear_tasks_mm_cpumask(cpu); -- - return 0; - } - -@@ -247,6 +245,9 @@ - pr_err("CPU%u: cpu didn't die\n", cpu); - return; - } ++Example: + -+ clear_tasks_mm_cpumask(cpu); ++ # modprobe hwlat_detector enabled=1 threshold=100 + - pr_notice("CPU%u: shutdown\n", cpu); - - /* -diff -Nur linux-4.4.62.orig/arch/arm/kernel/unwind.c linux-4.4.62/arch/arm/kernel/unwind.c ---- linux-4.4.62.orig/arch/arm/kernel/unwind.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/kernel/unwind.c 2017-04-18 17:38:07.906638076 +0200 -@@ -93,7 +93,7 @@ - static const struct unwind_idx *__origin_unwind_idx; - extern const struct unwind_idx __stop_unwind_idx[]; - --static DEFINE_SPINLOCK(unwind_lock); -+static DEFINE_RAW_SPINLOCK(unwind_lock); - static LIST_HEAD(unwind_tables); - - /* Convert a prel31 symbol to an absolute address */ -@@ -201,7 +201,7 @@ - /* module unwind tables */ - struct unwind_table *table; - -- spin_lock_irqsave(&unwind_lock, flags); -+ raw_spin_lock_irqsave(&unwind_lock, flags); - list_for_each_entry(table, &unwind_tables, list) { - if (addr >= table->begin_addr && - addr < table->end_addr) { -@@ -213,7 +213,7 @@ - break; - } - } -- spin_unlock_irqrestore(&unwind_lock, flags); -+ raw_spin_unlock_irqrestore(&unwind_lock, flags); - } - - pr_debug("%s: idx = %p\n", __func__, idx); -@@ -529,9 +529,9 @@ - tab->begin_addr = text_addr; - tab->end_addr = text_addr + text_size; - -- spin_lock_irqsave(&unwind_lock, flags); -+ raw_spin_lock_irqsave(&unwind_lock, flags); - list_add_tail(&tab->list, &unwind_tables); -- spin_unlock_irqrestore(&unwind_lock, flags); -+ raw_spin_unlock_irqrestore(&unwind_lock, flags); - - return tab; - } -@@ -543,9 +543,9 @@ - if (!tab) - return; - -- spin_lock_irqsave(&unwind_lock, flags); -+ raw_spin_lock_irqsave(&unwind_lock, flags); - list_del(&tab->list); -- spin_unlock_irqrestore(&unwind_lock, flags); -+ raw_spin_unlock_irqrestore(&unwind_lock, flags); - - kfree(tab); - } -diff -Nur linux-4.4.62.orig/arch/arm/kvm/arm.c linux-4.4.62/arch/arm/kvm/arm.c ---- linux-4.4.62.orig/arch/arm/kvm/arm.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/kvm/arm.c 2017-04-18 17:38:07.906638076 +0200 -@@ -496,18 +496,18 @@ - struct kvm_vcpu *vcpu; - - kvm_for_each_vcpu(i, vcpu, kvm) { -- wait_queue_head_t *wq = kvm_arch_vcpu_wq(vcpu); -+ struct swait_queue_head *wq = kvm_arch_vcpu_wq(vcpu); - - vcpu->arch.pause = false; -- wake_up_interruptible(wq); -+ swake_up(wq); - } - } - - static void vcpu_sleep(struct kvm_vcpu *vcpu) - { -- wait_queue_head_t *wq = kvm_arch_vcpu_wq(vcpu); -+ struct swait_queue_head *wq = kvm_arch_vcpu_wq(vcpu); - -- wait_event_interruptible(*wq, ((!vcpu->arch.power_off) && -+ swait_event_interruptible(*wq, ((!vcpu->arch.power_off) && - (!vcpu->arch.pause))); - } - -@@ -566,7 +566,7 @@ - * involves poking the GIC, which must be done in a - * non-preemptible context. - */ -- preempt_disable(); -+ migrate_disable(); - kvm_timer_flush_hwstate(vcpu); - kvm_vgic_flush_hwstate(vcpu); - -@@ -585,7 +585,7 @@ - local_irq_enable(); - kvm_timer_sync_hwstate(vcpu); - kvm_vgic_sync_hwstate(vcpu); -- preempt_enable(); -+ migrate_enable(); - continue; - } - -@@ -639,7 +639,7 @@ - - kvm_vgic_sync_hwstate(vcpu); - -- preempt_enable(); -+ migrate_enable(); - - ret = handle_exit(vcpu, run, ret); - } -diff -Nur linux-4.4.62.orig/arch/arm/kvm/psci.c linux-4.4.62/arch/arm/kvm/psci.c ---- linux-4.4.62.orig/arch/arm/kvm/psci.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/kvm/psci.c 2017-04-18 17:38:07.906638076 +0200 -@@ -70,7 +70,7 @@ - { - struct kvm *kvm = source_vcpu->kvm; - struct kvm_vcpu *vcpu = NULL; -- wait_queue_head_t *wq; -+ struct swait_queue_head *wq; - unsigned long cpu_id; - unsigned long context_id; - phys_addr_t target_pc; -@@ -119,7 +119,7 @@ - smp_mb(); /* Make sure the above is visible */ - - wq = kvm_arch_vcpu_wq(vcpu); -- wake_up_interruptible(wq); -+ swake_up(wq); - - return PSCI_RET_SUCCESS; - } -diff -Nur linux-4.4.62.orig/arch/arm/mach-at91/at91rm9200.c linux-4.4.62/arch/arm/mach-at91/at91rm9200.c ---- linux-4.4.62.orig/arch/arm/mach-at91/at91rm9200.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-at91/at91rm9200.c 2017-04-18 17:38:07.906638076 +0200 -@@ -12,7 +12,6 @@ - #include - - #include --#include - - #include "generic.h" - #include "soc.h" -@@ -33,7 +32,6 @@ - - of_platform_populate(NULL, of_default_bus_match_table, NULL, soc_dev); - -- arm_pm_idle = at91rm9200_idle; - at91rm9200_pm_init(); - } - -diff -Nur linux-4.4.62.orig/arch/arm/mach-at91/at91sam9.c linux-4.4.62/arch/arm/mach-at91/at91sam9.c ---- linux-4.4.62.orig/arch/arm/mach-at91/at91sam9.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-at91/at91sam9.c 2017-04-18 17:38:07.906638076 +0200 -@@ -62,8 +62,6 @@ - soc_dev = soc_device_to_device(soc); - - of_platform_populate(NULL, of_default_bus_match_table, NULL, soc_dev); -- -- arm_pm_idle = at91sam9_idle; - } - - static void __init at91sam9_dt_device_init(void) -diff -Nur linux-4.4.62.orig/arch/arm/mach-at91/generic.h linux-4.4.62/arch/arm/mach-at91/generic.h ---- linux-4.4.62.orig/arch/arm/mach-at91/generic.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-at91/generic.h 2017-04-18 17:38:07.906638076 +0200 -@@ -11,27 +11,18 @@ - #ifndef _AT91_GENERIC_H - #define _AT91_GENERIC_H - --#include --#include -- -- /* Map io */ --extern void __init at91_map_io(void); --extern void __init at91_alt_map_io(void); -- --/* idle */ --extern void at91rm9200_idle(void); --extern void at91sam9_idle(void); -- - #ifdef CONFIG_PM - extern void __init at91rm9200_pm_init(void); - extern void __init at91sam9260_pm_init(void); - extern void __init at91sam9g45_pm_init(void); - extern void __init at91sam9x5_pm_init(void); -+extern void __init sama5_pm_init(void); - #else - static inline void __init at91rm9200_pm_init(void) { } - static inline void __init at91sam9260_pm_init(void) { } - static inline void __init at91sam9g45_pm_init(void) { } - static inline void __init at91sam9x5_pm_init(void) { } -+static inline void __init sama5_pm_init(void) { } - #endif - - #endif /* _AT91_GENERIC_H */ -diff -Nur linux-4.4.62.orig/arch/arm/mach-at91/Kconfig linux-4.4.62/arch/arm/mach-at91/Kconfig ---- linux-4.4.62.orig/arch/arm/mach-at91/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-at91/Kconfig 2017-04-18 17:38:07.906638076 +0200 -@@ -99,6 +99,7 @@ - config COMMON_CLK_AT91 - bool - select COMMON_CLK -+ select MFD_SYSCON - - config HAVE_AT91_SMD - bool -diff -Nur linux-4.4.62.orig/arch/arm/mach-at91/pm.c linux-4.4.62/arch/arm/mach-at91/pm.c ---- linux-4.4.62.orig/arch/arm/mach-at91/pm.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-at91/pm.c 2017-04-18 17:38:07.906638076 +0200 -@@ -31,10 +31,13 @@ - #include - #include - #include -+#include - - #include "generic.h" - #include "pm.h" - -+static void __iomem *pmc; ++After the module is loaded, it creates a directory named "hwlat_detector" under ++the debugfs mountpoint, "/debug/hwlat_detector" for this text. It is necessary ++to have debugfs mounted, which might be on /sys/debug on your system. + - /* - * FIXME: this is needed to communicate between the pinctrl driver and - * the PM implementation in the machine. Possibly part of the PM -@@ -87,7 +90,7 @@ - unsigned long scsr; - int i; - -- scsr = at91_pmc_read(AT91_PMC_SCSR); -+ scsr = readl(pmc + AT91_PMC_SCSR); ++The /debug/hwlat_detector interface contains the following files: ++ ++count - number of latency spikes observed since last reset ++enable - a global enable/disable toggle (0/1), resets count ++max - maximum hardware latency actually observed (usecs) ++sample - a pipe from which to read current raw sample data ++ in the format ++ (can be opened O_NONBLOCK for a single sample) ++threshold - minimum latency value to be considered (usecs) ++width - time period to sample with CPUs held (usecs) ++ must be less than the total window size (enforced) ++window - total period of sampling, width being inside (usecs) ++ ++By default we will set width to 500,000 and window to 1,000,000, meaning that ++we will sample every 1,000,000 usecs (1s) for 500,000 usecs (0.5s). If we ++observe any latencies that exceed the threshold (initially 100 usecs), ++then we write to a global sample ring buffer of 8K samples, which is ++consumed by reading from the "sample" (pipe) debugfs file interface. +diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt +index ca64ca566099..19bef799c902 100644 +--- a/Documentation/kernel-parameters.txt ++++ b/Documentation/kernel-parameters.txt +@@ -1640,6 +1640,15 @@ bytes respectively. Such letter suffixes can also be entirely omitted. + ip= [IP_PNP] + See Documentation/filesystems/nfs/nfsroot.txt. - /* USB must not be using PLLB */ - if ((scsr & at91_pm_data.uhp_udp_mask) != 0) { -@@ -101,8 +104,7 @@ ++ irqaffinity= [SMP] Set the default irq affinity mask ++ Format: ++ ,..., ++ or ++ - ++ (must be a positive range in ascending order) ++ or a mixture ++ ,...,- ++ + irqfixup [HW] + When an interrupt is not handled search all handlers + for it. Intended to get systems with badly broken +diff --git a/Documentation/sysrq.txt b/Documentation/sysrq.txt +index 13f5619b2203..f64d075ba647 100644 +--- a/Documentation/sysrq.txt ++++ b/Documentation/sysrq.txt +@@ -59,10 +59,17 @@ On PowerPC - Press 'ALT - Print Screen (or F13) - , + On other - If you know of the key combos for other architectures, please + let me know so I can add them to this section. - if ((scsr & (AT91_PMC_PCK0 << i)) == 0) - continue; +-On all - write a character to /proc/sysrq-trigger. e.g.: - -- css = at91_pmc_read(AT91_PMC_PCKR(i)) & AT91_PMC_CSS; -+ css = readl(pmc + AT91_PMC_PCKR(i)) & AT91_PMC_CSS; - if (css != AT91_PMC_CSS_SLOW) { - pr_err("AT91: PM - Suspend-to-RAM with PCK%d src %d\n", i, css); - return 0; -@@ -145,8 +147,8 @@ - flush_cache_all(); - outer_disable(); - -- at91_suspend_sram_fn(at91_pmc_base, at91_ramc_base[0], -- at91_ramc_base[1], pm_data); -+ at91_suspend_sram_fn(pmc, at91_ramc_base[0], -+ at91_ramc_base[1], pm_data); - - outer_resume(); - } -@@ -369,6 +371,21 @@ - at91_pm_set_standby(standby); - } ++On all - write a character to /proc/sysrq-trigger, e.g.: + echo t > /proc/sysrq-trigger -+void at91rm9200_idle(void) -+{ -+ /* -+ * Disable the processor clock. The processor will be automatically -+ * re-enabled by an interrupt or by a reset. -+ */ -+ writel(AT91_PMC_PCK, pmc + AT91_PMC_SCDR); -+} ++On all - Enable network SysRq by writing a cookie to icmp_echo_sysrq, e.g. ++ echo 0x01020304 >/proc/sys/net/ipv4/icmp_echo_sysrq ++ Send an ICMP echo request with this pattern plus the particular ++ SysRq command key. Example: ++ # ping -c1 -s57 -p0102030468 ++ will trigger the SysRq-H (help) command. + -+void at91sam9_idle(void) -+{ -+ writel(AT91_PMC_PCK, pmc + AT91_PMC_SCDR); -+ cpu_do_idle(); -+} + - static void __init at91_pm_sram_init(void) - { - struct gen_pool *sram_pool; -@@ -415,13 +432,36 @@ - &at91_pm_suspend_in_sram, at91_pm_suspend_in_sram_sz); - } - --static void __init at91_pm_init(void) -+static const struct of_device_id atmel_pmc_ids[] __initconst = { -+ { .compatible = "atmel,at91rm9200-pmc" }, -+ { .compatible = "atmel,at91sam9260-pmc" }, -+ { .compatible = "atmel,at91sam9g45-pmc" }, -+ { .compatible = "atmel,at91sam9n12-pmc" }, -+ { .compatible = "atmel,at91sam9x5-pmc" }, -+ { .compatible = "atmel,sama5d3-pmc" }, -+ { .compatible = "atmel,sama5d2-pmc" }, -+ { /* sentinel */ }, -+}; + * What are the 'command' keys? + ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + 'b' - Will immediately reboot the system without syncing or unmounting +diff --git a/Documentation/trace/histograms.txt b/Documentation/trace/histograms.txt +new file mode 100644 +index 000000000000..6f2aeabf7faa +--- /dev/null ++++ b/Documentation/trace/histograms.txt +@@ -0,0 +1,186 @@ ++ Using the Linux Kernel Latency Histograms + -+static void __init at91_pm_init(void (*pm_idle)(void)) - { -- at91_pm_sram_init(); -+ struct device_node *pmc_np; - - if (at91_cpuidle_device.dev.platform_data) - platform_device_register(&at91_cpuidle_device); - -+ pmc_np = of_find_matching_node(NULL, atmel_pmc_ids); -+ pmc = of_iomap(pmc_np, 0); -+ if (!pmc) { -+ pr_err("AT91: PM not supported, PMC not found\n"); -+ return; -+ } + -+ if (pm_idle) -+ arm_pm_idle = pm_idle; ++This document gives a short explanation how to enable, configure and use ++latency histograms. Latency histograms are primarily relevant in the ++context of real-time enabled kernels (CONFIG_PREEMPT/CONFIG_PREEMPT_RT) ++and are used in the quality management of the Linux real-time ++capabilities. + -+ at91_pm_sram_init(); + - if (at91_suspend_sram_fn) - suspend_set_ops(&at91_pm_ops); - else -@@ -440,7 +480,7 @@ - at91_pm_data.uhp_udp_mask = AT91RM9200_PMC_UHP | AT91RM9200_PMC_UDP; - at91_pm_data.memctrl = AT91_MEMCTRL_MC; - -- at91_pm_init(); -+ at91_pm_init(at91rm9200_idle); - } - - void __init at91sam9260_pm_init(void) -@@ -448,7 +488,7 @@ - at91_dt_ramc(); - at91_pm_data.memctrl = AT91_MEMCTRL_SDRAMC; - at91_pm_data.uhp_udp_mask = AT91SAM926x_PMC_UHP | AT91SAM926x_PMC_UDP; -- return at91_pm_init(); -+ at91_pm_init(at91sam9_idle); - } - - void __init at91sam9g45_pm_init(void) -@@ -456,7 +496,7 @@ - at91_dt_ramc(); - at91_pm_data.uhp_udp_mask = AT91SAM926x_PMC_UHP; - at91_pm_data.memctrl = AT91_MEMCTRL_DDRSDR; -- return at91_pm_init(); -+ at91_pm_init(at91sam9_idle); - } - - void __init at91sam9x5_pm_init(void) -@@ -464,5 +504,13 @@ - at91_dt_ramc(); - at91_pm_data.uhp_udp_mask = AT91SAM926x_PMC_UHP | AT91SAM926x_PMC_UDP; - at91_pm_data.memctrl = AT91_MEMCTRL_DDRSDR; -- return at91_pm_init(); -+ at91_pm_init(at91sam9_idle); -+} ++* Purpose of latency histograms + -+void __init sama5_pm_init(void) -+{ -+ at91_dt_ramc(); -+ at91_pm_data.uhp_udp_mask = AT91SAM926x_PMC_UHP | AT91SAM926x_PMC_UDP; -+ at91_pm_data.memctrl = AT91_MEMCTRL_DDRSDR; -+ at91_pm_init(NULL); - } -diff -Nur linux-4.4.62.orig/arch/arm/mach-at91/sama5.c linux-4.4.62/arch/arm/mach-at91/sama5.c ---- linux-4.4.62.orig/arch/arm/mach-at91/sama5.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-at91/sama5.c 2017-04-18 17:38:07.906638076 +0200 -@@ -51,7 +51,7 @@ - soc_dev = soc_device_to_device(soc); ++A latency histogram continuously accumulates the frequencies of latency ++data. There are two types of histograms ++- potential sources of latencies ++- effective latencies ++ ++ ++* Potential sources of latencies ++ ++Potential sources of latencies are code segments where interrupts, ++preemption or both are disabled (aka critical sections). To create ++histograms of potential sources of latency, the kernel stores the time ++stamp at the start of a critical section, determines the time elapsed ++when the end of the section is reached, and increments the frequency ++counter of that latency value - irrespective of whether any concurrently ++running process is affected by latency or not. ++- Configuration items (in the Kernel hacking/Tracers submenu) ++ CONFIG_INTERRUPT_OFF_LATENCY ++ CONFIG_PREEMPT_OFF_LATENCY ++ ++ ++* Effective latencies ++ ++Effective latencies are actually occuring during wakeup of a process. To ++determine effective latencies, the kernel stores the time stamp when a ++process is scheduled to be woken up, and determines the duration of the ++wakeup time shortly before control is passed over to this process. Note ++that the apparent latency in user space may be somewhat longer, since the ++process may be interrupted after control is passed over to it but before ++the execution in user space takes place. Simply measuring the interval ++between enqueuing and wakeup may also not appropriate in cases when a ++process is scheduled as a result of a timer expiration. The timer may have ++missed its deadline, e.g. due to disabled interrupts, but this latency ++would not be registered. Therefore, the offsets of missed timers are ++recorded in a separate histogram. If both wakeup latency and missed timer ++offsets are configured and enabled, a third histogram may be enabled that ++records the overall latency as a sum of the timer latency, if any, and the ++wakeup latency. This histogram is called "timerandwakeup". ++- Configuration items (in the Kernel hacking/Tracers submenu) ++ CONFIG_WAKEUP_LATENCY ++ CONFIG_MISSED_TIMER_OFSETS ++ ++ ++* Usage ++ ++The interface to the administration of the latency histograms is located ++in the debugfs file system. To mount it, either enter ++ ++mount -t sysfs nodev /sys ++mount -t debugfs nodev /sys/kernel/debug ++ ++from shell command line level, or add ++ ++nodev /sys sysfs defaults 0 0 ++nodev /sys/kernel/debug debugfs defaults 0 0 ++ ++to the file /etc/fstab. All latency histogram related files are then ++available in the directory /sys/kernel/debug/tracing/latency_hist. A ++particular histogram type is enabled by writing non-zero to the related ++variable in the /sys/kernel/debug/tracing/latency_hist/enable directory. ++Select "preemptirqsoff" for the histograms of potential sources of ++latencies and "wakeup" for histograms of effective latencies etc. The ++histogram data - one per CPU - are available in the files ++ ++/sys/kernel/debug/tracing/latency_hist/preemptoff/CPUx ++/sys/kernel/debug/tracing/latency_hist/irqsoff/CPUx ++/sys/kernel/debug/tracing/latency_hist/preemptirqsoff/CPUx ++/sys/kernel/debug/tracing/latency_hist/wakeup/CPUx ++/sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio/CPUx ++/sys/kernel/debug/tracing/latency_hist/missed_timer_offsets/CPUx ++/sys/kernel/debug/tracing/latency_hist/timerandwakeup/CPUx ++ ++The histograms are reset by writing non-zero to the file "reset" in a ++particular latency directory. To reset all latency data, use ++ ++#!/bin/sh ++ ++TRACINGDIR=/sys/kernel/debug/tracing ++HISTDIR=$TRACINGDIR/latency_hist ++ ++if test -d $HISTDIR ++then ++ cd $HISTDIR ++ for i in `find . | grep /reset$` ++ do ++ echo 1 >$i ++ done ++fi ++ ++ ++* Data format ++ ++Latency data are stored with a resolution of one microsecond. The ++maximum latency is 10,240 microseconds. The data are only valid, if the ++overflow register is empty. Every output line contains the latency in ++microseconds in the first row and the number of samples in the second ++row. To display only lines with a positive latency count, use, for ++example, ++ ++grep -v " 0$" /sys/kernel/debug/tracing/latency_hist/preemptoff/CPU0 ++ ++#Minimum latency: 0 microseconds. ++#Average latency: 0 microseconds. ++#Maximum latency: 25 microseconds. ++#Total samples: 3104770694 ++#There are 0 samples greater or equal than 10240 microseconds ++#usecs samples ++ 0 2984486876 ++ 1 49843506 ++ 2 58219047 ++ 3 5348126 ++ 4 2187960 ++ 5 3388262 ++ 6 959289 ++ 7 208294 ++ 8 40420 ++ 9 4485 ++ 10 14918 ++ 11 18340 ++ 12 25052 ++ 13 19455 ++ 14 5602 ++ 15 969 ++ 16 47 ++ 17 18 ++ 18 14 ++ 19 1 ++ 20 3 ++ 21 2 ++ 22 5 ++ 23 2 ++ 25 1 ++ ++ ++* Wakeup latency of a selected process ++ ++To only collect wakeup latency data of a particular process, write the ++PID of the requested process to ++ ++/sys/kernel/debug/tracing/latency_hist/wakeup/pid ++ ++PIDs are not considered, if this variable is set to 0. ++ ++ ++* Details of the process with the highest wakeup latency so far ++ ++Selected data of the process that suffered from the highest wakeup ++latency that occurred in a particular CPU are available in the file ++ ++/sys/kernel/debug/tracing/latency_hist/wakeup/max_latency-CPUx. ++ ++In addition, other relevant system data at the time when the ++latency occurred are given. ++ ++The format of the data is (all in one line): ++ () \ ++<- ++ ++The value of is only relevant in the combined timer ++and wakeup latency recording. In the wakeup recording, it is ++always 0, in the missed_timer_offsets recording, it is the same ++as . ++ ++When retrospectively searching for the origin of a latency and ++tracing was not enabled, it may be helpful to know the name and ++some basic data of the task that (finally) was switching to the ++late real-tlme task. In addition to the victim's data, also the ++data of the possible culprit are therefore displayed after the ++"<-" symbol. ++ ++Finally, the timestamp of the time when the latency occurred ++in . after the most recent system boot ++is provided. ++ ++These data are also reset when the wakeup histogram is reset. +diff --git a/Makefile b/Makefile +index a5ecb29c6ed3..5521b0c3abf2 100644 +--- a/Makefile ++++ b/Makefile +@@ -785,6 +785,9 @@ KBUILD_CFLAGS += $(call cc-option,-Werror=strict-prototypes) + # Prohibit date/time macros, which would make the build non-deterministic + KBUILD_CFLAGS += $(call cc-option,-Werror=date-time) - of_platform_populate(NULL, of_default_bus_match_table, NULL, soc_dev); -- at91sam9x5_pm_init(); -+ sama5_pm_init(); - } ++# enforce correct pointer usage ++KBUILD_CFLAGS += $(call cc-option,-Werror=incompatible-pointer-types) ++ + # use the deterministic mode of AR if available + KBUILD_ARFLAGS := $(call ar-option,D) - static const char *const sama5_dt_board_compat[] __initconst = { -diff -Nur linux-4.4.62.orig/arch/arm/mach-exynos/platsmp.c linux-4.4.62/arch/arm/mach-exynos/platsmp.c ---- linux-4.4.62.orig/arch/arm/mach-exynos/platsmp.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-exynos/platsmp.c 2017-04-18 17:38:07.906638076 +0200 -@@ -230,7 +230,7 @@ - return (void __iomem *)(S5P_VA_SCU); - } +diff --git a/arch/Kconfig b/arch/Kconfig +index 4e949e58b192..3b26d76933fb 100644 +--- a/arch/Kconfig ++++ b/arch/Kconfig +@@ -9,6 +9,7 @@ config OPROFILE + tristate "OProfile system profiling" + depends on PROFILING + depends on HAVE_OPROFILE ++ depends on !PREEMPT_RT_FULL + select RING_BUFFER + select RING_BUFFER_ALLOW_SWAP + help +@@ -52,6 +53,7 @@ config KPROBES + config JUMP_LABEL + bool "Optimize very unlikely/likely branches" + depends on HAVE_ARCH_JUMP_LABEL ++ depends on (!INTERRUPT_OFF_HIST && !PREEMPT_OFF_HIST && !WAKEUP_LATENCY_HIST && !MISSED_TIMER_OFFSETS_HIST) + help + This option enables a transparent branch optimization that + makes certain almost-always-true or almost-always-false branch +diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig +index 34e1569a11ee..79c4603e9453 100644 +--- a/arch/arm/Kconfig ++++ b/arch/arm/Kconfig +@@ -33,7 +33,7 @@ config ARM + select HARDIRQS_SW_RESEND + select HAVE_ARCH_AUDITSYSCALL if (AEABI && !OABI_COMPAT) + select HAVE_ARCH_BITREVERSE if (CPU_32v7M || CPU_32v7) && !CPU_32v6 +- select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL && !CPU_ENDIAN_BE32 ++ select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL && !CPU_ENDIAN_BE32 && !PREEMPT_RT_BASE + select HAVE_ARCH_KGDB if !CPU_ENDIAN_BE32 + select HAVE_ARCH_SECCOMP_FILTER if (AEABI && !OABI_COMPAT) + select HAVE_ARCH_TRACEHOOK +@@ -68,6 +68,7 @@ config ARM + select HAVE_PERF_EVENTS + select HAVE_PERF_REGS + select HAVE_PERF_USER_STACK_DUMP ++ select HAVE_PREEMPT_LAZY + select HAVE_RCU_TABLE_FREE if (SMP && ARM_LPAE) + select HAVE_REGS_AND_STACK_ACCESS_API + select HAVE_SYSCALL_TRACEPOINTS +diff --git a/arch/arm/include/asm/switch_to.h b/arch/arm/include/asm/switch_to.h +index 12ebfcc1d539..c962084605bc 100644 +--- a/arch/arm/include/asm/switch_to.h ++++ b/arch/arm/include/asm/switch_to.h +@@ -3,6 +3,13 @@ --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); + #include - static void exynos_secondary_init(unsigned int cpu) - { -@@ -243,8 +243,8 @@ - /* - * Synchronise with the boot thread. - */ -- spin_lock(&boot_lock); -- spin_unlock(&boot_lock); -+ raw_spin_lock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - } ++#if defined CONFIG_PREEMPT_RT_FULL && defined CONFIG_HIGHMEM ++void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p); ++#else ++static inline void ++switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) { } ++#endif ++ + /* + * For v7 SMP cores running a preemptible kernel we may be pre-empted + * during a TLB maintenance operation, so execute an inner-shareable dsb +@@ -25,6 +32,7 @@ extern struct task_struct *__switch_to(struct task_struct *, struct thread_info + #define switch_to(prev,next,last) \ + do { \ + __complete_pending_tlbi(); \ ++ switch_kmaps(prev, next); \ + last = __switch_to(prev,task_thread_info(prev), task_thread_info(next)); \ + } while (0) - int exynos_set_boot_addr(u32 core_id, unsigned long boot_addr) -@@ -308,7 +308,7 @@ - * Set synchronisation state between this boot processor - * and the secondary one - */ -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); +diff --git a/arch/arm/include/asm/thread_info.h b/arch/arm/include/asm/thread_info.h +index 776757d1604a..1f36a4eccc72 100644 +--- a/arch/arm/include/asm/thread_info.h ++++ b/arch/arm/include/asm/thread_info.h +@@ -49,6 +49,7 @@ struct cpu_context_save { + struct thread_info { + unsigned long flags; /* low level flags */ + int preempt_count; /* 0 => preemptable, <0 => bug */ ++ int preempt_lazy_count; /* 0 => preemptable, <0 => bug */ + mm_segment_t addr_limit; /* address limit */ + struct task_struct *task; /* main task structure */ + __u32 cpu; /* cpu */ +@@ -142,7 +143,8 @@ extern int vfp_restore_user_hwstate(struct user_vfp __user *, + #define TIF_SYSCALL_TRACE 4 /* syscall trace active */ + #define TIF_SYSCALL_AUDIT 5 /* syscall auditing active */ + #define TIF_SYSCALL_TRACEPOINT 6 /* syscall tracepoint instrumentation */ +-#define TIF_SECCOMP 7 /* seccomp syscall filtering active */ ++#define TIF_SECCOMP 8 /* seccomp syscall filtering active */ ++#define TIF_NEED_RESCHED_LAZY 7 - /* - * The secondary processor is waiting to be released from -@@ -335,7 +335,7 @@ + #define TIF_NOHZ 12 /* in adaptive nohz mode */ + #define TIF_USING_IWMMXT 17 +@@ -152,6 +154,7 @@ extern int vfp_restore_user_hwstate(struct user_vfp __user *, + #define _TIF_SIGPENDING (1 << TIF_SIGPENDING) + #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) + #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) ++#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) + #define _TIF_UPROBE (1 << TIF_UPROBE) + #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) + #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) +@@ -167,7 +170,8 @@ extern int vfp_restore_user_hwstate(struct user_vfp __user *, + * Change these and you break ASM code in entry-common.S + */ + #define _TIF_WORK_MASK (_TIF_NEED_RESCHED | _TIF_SIGPENDING | \ +- _TIF_NOTIFY_RESUME | _TIF_UPROBE) ++ _TIF_NOTIFY_RESUME | _TIF_UPROBE | \ ++ _TIF_NEED_RESCHED_LAZY) - if (timeout == 0) { - printk(KERN_ERR "cpu1 power enable failed"); -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - return -ETIMEDOUT; - } - } -@@ -381,7 +381,7 @@ - * calibrations, then wait for it to finish - */ - fail: -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); + #endif /* __KERNEL__ */ + #endif /* __ASM_ARM_THREAD_INFO_H */ +diff --git a/arch/arm/kernel/asm-offsets.c b/arch/arm/kernel/asm-offsets.c +index 871b8267d211..4dbe70de7318 100644 +--- a/arch/arm/kernel/asm-offsets.c ++++ b/arch/arm/kernel/asm-offsets.c +@@ -65,6 +65,7 @@ int main(void) + BLANK(); + DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); + DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); ++ DEFINE(TI_PREEMPT_LAZY, offsetof(struct thread_info, preempt_lazy_count)); + DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit)); + DEFINE(TI_TASK, offsetof(struct thread_info, task)); + DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); +diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S +index 3ce377f7251f..d044cea59f54 100644 +--- a/arch/arm/kernel/entry-armv.S ++++ b/arch/arm/kernel/entry-armv.S +@@ -215,11 +215,18 @@ __irq_svc: + #ifdef CONFIG_PREEMPT + get_thread_info tsk + ldr r8, [tsk, #TI_PREEMPT] @ get preempt count +- ldr r0, [tsk, #TI_FLAGS] @ get flags + teq r8, #0 @ if preempt count != 0 ++ bne 1f @ return from exeption ++ ldr r0, [tsk, #TI_FLAGS] @ get flags ++ tst r0, #_TIF_NEED_RESCHED @ if NEED_RESCHED is set ++ blne svc_preempt @ preempt! ++ ++ ldr r8, [tsk, #TI_PREEMPT_LAZY] @ get preempt lazy count ++ teq r8, #0 @ if preempt lazy count != 0 + movne r0, #0 @ force flags to 0 +- tst r0, #_TIF_NEED_RESCHED ++ tst r0, #_TIF_NEED_RESCHED_LAZY + blne svc_preempt ++1: + #endif - return pen_release != -1 ? ret : 0; - } -diff -Nur linux-4.4.62.orig/arch/arm/mach-hisi/platmcpm.c linux-4.4.62/arch/arm/mach-hisi/platmcpm.c ---- linux-4.4.62.orig/arch/arm/mach-hisi/platmcpm.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-hisi/platmcpm.c 2017-04-18 17:38:07.906638076 +0200 -@@ -61,7 +61,7 @@ + svc_exit r5, irq = 1 @ return from exception +@@ -234,8 +241,14 @@ svc_preempt: + 1: bl preempt_schedule_irq @ irq en/disable is done inside + ldr r0, [tsk, #TI_FLAGS] @ get new tasks TI_FLAGS + tst r0, #_TIF_NEED_RESCHED ++ bne 1b ++ tst r0, #_TIF_NEED_RESCHED_LAZY + reteq r8 @ go again +- b 1b ++ ldr r0, [tsk, #TI_PREEMPT_LAZY] @ get preempt lazy count ++ teq r0, #0 @ if preempt lazy count != 0 ++ beq 1b ++ ret r8 @ go again ++ + #endif - static void __iomem *sysctrl, *fabric; - static int hip04_cpu_table[HIP04_MAX_CLUSTERS][HIP04_MAX_CPUS_PER_CLUSTER]; --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); - static u32 fabric_phys_addr; - /* - * [0]: bootwrapper physical address -@@ -113,7 +113,7 @@ - if (cluster >= HIP04_MAX_CLUSTERS || cpu >= HIP04_MAX_CPUS_PER_CLUSTER) - return -EINVAL; + __und_fault: +diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S +index 30a7228eaceb..c3bd6cbfce4b 100644 +--- a/arch/arm/kernel/entry-common.S ++++ b/arch/arm/kernel/entry-common.S +@@ -36,7 +36,9 @@ ret_fast_syscall: + UNWIND(.cantunwind ) + disable_irq_notrace @ disable interrupts + ldr r1, [tsk, #TI_FLAGS] @ re-check for syscall tracing +- tst r1, #_TIF_SYSCALL_WORK | _TIF_WORK_MASK ++ tst r1, #((_TIF_SYSCALL_WORK | _TIF_WORK_MASK) & ~_TIF_SECCOMP) ++ bne fast_work_pending ++ tst r1, #_TIF_SECCOMP + bne fast_work_pending -- spin_lock_irq(&boot_lock); -+ raw_spin_lock_irq(&boot_lock); + /* perform architecture specific actions before user return */ +@@ -62,8 +64,11 @@ ret_fast_syscall: + str r0, [sp, #S_R0 + S_OFF]! @ save returned r0 + disable_irq_notrace @ disable interrupts + ldr r1, [tsk, #TI_FLAGS] @ re-check for syscall tracing +- tst r1, #_TIF_SYSCALL_WORK | _TIF_WORK_MASK ++ tst r1, #((_TIF_SYSCALL_WORK | _TIF_WORK_MASK) & ~_TIF_SECCOMP) ++ bne do_slower_path ++ tst r1, #_TIF_SECCOMP + beq no_work_pending ++do_slower_path: + UNWIND(.fnend ) + ENDPROC(ret_fast_syscall) - if (hip04_cpu_table[cluster][cpu]) - goto out; -@@ -147,7 +147,7 @@ +diff --git a/arch/arm/kernel/patch.c b/arch/arm/kernel/patch.c +index 69bda1a5707e..1f665acaa6a9 100644 +--- a/arch/arm/kernel/patch.c ++++ b/arch/arm/kernel/patch.c +@@ -15,7 +15,7 @@ struct patch { + unsigned int insn; + }; - out: - hip04_cpu_table[cluster][cpu]++; -- spin_unlock_irq(&boot_lock); -+ raw_spin_unlock_irq(&boot_lock); +-static DEFINE_SPINLOCK(patch_lock); ++static DEFINE_RAW_SPINLOCK(patch_lock); - return 0; - } -@@ -162,11 +162,11 @@ - cpu = MPIDR_AFFINITY_LEVEL(mpidr, 0); - cluster = MPIDR_AFFINITY_LEVEL(mpidr, 1); + static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags) + __acquires(&patch_lock) +@@ -32,7 +32,7 @@ static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags) + return addr; -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); - hip04_cpu_table[cluster][cpu]--; - if (hip04_cpu_table[cluster][cpu] == 1) { - /* A power_up request went ahead of us. */ -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - return; - } else if (hip04_cpu_table[cluster][cpu] > 1) { - pr_err("Cluster %d CPU%d boots multiple times\n", cluster, cpu); -@@ -174,7 +174,7 @@ - } + if (flags) +- spin_lock_irqsave(&patch_lock, *flags); ++ raw_spin_lock_irqsave(&patch_lock, *flags); + else + __acquire(&patch_lock); - last_man = hip04_cluster_is_down(cluster); -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - if (last_man) { - /* Since it's Cortex A15, disable L2 prefetching. */ - asm volatile( -@@ -203,7 +203,7 @@ - cpu >= HIP04_MAX_CPUS_PER_CLUSTER); +@@ -47,7 +47,7 @@ static void __kprobes patch_unmap(int fixmap, unsigned long *flags) + clear_fixmap(fixmap); - count = TIMEOUT_MSEC / POLL_MSEC; -- spin_lock_irq(&boot_lock); -+ raw_spin_lock_irq(&boot_lock); - for (tries = 0; tries < count; tries++) { - if (hip04_cpu_table[cluster][cpu]) - goto err; -@@ -211,10 +211,10 @@ - data = readl_relaxed(sysctrl + SC_CPU_RESET_STATUS(cluster)); - if (data & CORE_WFI_STATUS(cpu)) - break; -- spin_unlock_irq(&boot_lock); -+ raw_spin_unlock_irq(&boot_lock); - /* Wait for clean L2 when the whole cluster is down. */ - msleep(POLL_MSEC); -- spin_lock_irq(&boot_lock); -+ raw_spin_lock_irq(&boot_lock); - } - if (tries >= count) - goto err; -@@ -231,10 +231,10 @@ - goto err; - if (hip04_cluster_is_down(cluster)) - hip04_set_snoop_filter(cluster, 0); -- spin_unlock_irq(&boot_lock); -+ raw_spin_unlock_irq(&boot_lock); - return 1; - err: -- spin_unlock_irq(&boot_lock); -+ raw_spin_unlock_irq(&boot_lock); - return 0; + if (flags) +- spin_unlock_irqrestore(&patch_lock, *flags); ++ raw_spin_unlock_irqrestore(&patch_lock, *flags); + else + __release(&patch_lock); } - #endif -diff -Nur linux-4.4.62.orig/arch/arm/mach-imx/Kconfig linux-4.4.62/arch/arm/mach-imx/Kconfig ---- linux-4.4.62.orig/arch/arm/mach-imx/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-imx/Kconfig 2017-04-18 17:38:07.906638076 +0200 -@@ -524,7 +524,7 @@ - bool "i.MX6 Quad/DualLite support" - select ARM_ERRATA_764369 if SMP - select HAVE_ARM_SCU if SMP -- select HAVE_ARM_TWD if SMP -+ select HAVE_ARM_TWD - select PCI_DOMAINS if PCI - select PINCTRL_IMX6Q - select SOC_IMX6 -diff -Nur linux-4.4.62.orig/arch/arm/mach-omap2/omap-smp.c linux-4.4.62/arch/arm/mach-omap2/omap-smp.c ---- linux-4.4.62.orig/arch/arm/mach-omap2/omap-smp.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-omap2/omap-smp.c 2017-04-18 17:38:07.906638076 +0200 -@@ -43,7 +43,7 @@ - /* SCU base address */ - static void __iomem *scu_base; - --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); - - void __iomem *omap4_get_scu_base(void) - { -@@ -74,8 +74,8 @@ - /* - * Synchronise with the boot thread. - */ -- spin_lock(&boot_lock); -- spin_unlock(&boot_lock); -+ raw_spin_lock(&boot_lock); -+ raw_spin_unlock(&boot_lock); +diff --git a/arch/arm/kernel/process.c b/arch/arm/kernel/process.c +index 4adfb46e3ee9..15f1d94b47c5 100644 +--- a/arch/arm/kernel/process.c ++++ b/arch/arm/kernel/process.c +@@ -319,6 +319,30 @@ unsigned long arch_randomize_brk(struct mm_struct *mm) } - static int omap4_boot_secondary(unsigned int cpu, struct task_struct *idle) -@@ -89,7 +89,7 @@ - * Set synchronisation state between this boot processor - * and the secondary one - */ -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); - - /* - * Update the AuxCoreBoot0 with boot state for secondary core. -@@ -166,7 +166,7 @@ - * Now the secondary core is starting up let it run its - * calibrations, then wait for it to finish + #ifdef CONFIG_MMU ++/* ++ * CONFIG_SPLIT_PTLOCK_CPUS results in a page->ptl lock. If the lock is not ++ * initialized by pgtable_page_ctor() then a coredump of the vector page will ++ * fail. ++ */ ++static int __init vectors_user_mapping_init_page(void) ++{ ++ struct page *page; ++ unsigned long addr = 0xffff0000; ++ pgd_t *pgd; ++ pud_t *pud; ++ pmd_t *pmd; ++ ++ pgd = pgd_offset_k(addr); ++ pud = pud_offset(pgd, addr); ++ pmd = pmd_offset(pud, addr); ++ page = pmd_page(*(pmd)); ++ ++ pgtable_page_ctor(page); ++ ++ return 0; ++} ++late_initcall(vectors_user_mapping_init_page); ++ + #ifdef CONFIG_KUSER_HELPERS + /* + * The vectors page is always readable from user space for the +diff --git a/arch/arm/kernel/signal.c b/arch/arm/kernel/signal.c +index 7b8f2141427b..96541e00b74a 100644 +--- a/arch/arm/kernel/signal.c ++++ b/arch/arm/kernel/signal.c +@@ -572,7 +572,8 @@ do_work_pending(struct pt_regs *regs, unsigned int thread_flags, int syscall) */ -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); + trace_hardirqs_off(); + do { +- if (likely(thread_flags & _TIF_NEED_RESCHED)) { ++ if (likely(thread_flags & (_TIF_NEED_RESCHED | ++ _TIF_NEED_RESCHED_LAZY))) { + schedule(); + } else { + if (unlikely(!user_mode(regs))) +diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c +index b26361355dae..e5754e3b03c4 100644 +--- a/arch/arm/kernel/smp.c ++++ b/arch/arm/kernel/smp.c +@@ -230,8 +230,6 @@ int __cpu_disable(void) + flush_cache_louis(); + local_flush_tlb_all(); +- clear_tasks_mm_cpumask(cpu); +- return 0; } -diff -Nur linux-4.4.62.orig/arch/arm/mach-prima2/platsmp.c linux-4.4.62/arch/arm/mach-prima2/platsmp.c ---- linux-4.4.62.orig/arch/arm/mach-prima2/platsmp.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-prima2/platsmp.c 2017-04-18 17:38:07.906638076 +0200 -@@ -22,7 +22,7 @@ - - static void __iomem *clk_base; --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); +@@ -247,6 +245,9 @@ void __cpu_die(unsigned int cpu) + pr_err("CPU%u: cpu didn't die\n", cpu); + return; + } ++ ++ clear_tasks_mm_cpumask(cpu); ++ + pr_notice("CPU%u: shutdown\n", cpu); - static void sirfsoc_secondary_init(unsigned int cpu) - { -@@ -36,8 +36,8 @@ /* - * Synchronise with the boot thread. - */ -- spin_lock(&boot_lock); -- spin_unlock(&boot_lock); -+ raw_spin_lock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - } - - static const struct of_device_id clk_ids[] = { -@@ -75,7 +75,7 @@ - /* make sure write buffer is drained */ - mb(); +diff --git a/arch/arm/kernel/unwind.c b/arch/arm/kernel/unwind.c +index 0bee233fef9a..314cfb232a63 100644 +--- a/arch/arm/kernel/unwind.c ++++ b/arch/arm/kernel/unwind.c +@@ -93,7 +93,7 @@ extern const struct unwind_idx __start_unwind_idx[]; + static const struct unwind_idx *__origin_unwind_idx; + extern const struct unwind_idx __stop_unwind_idx[]; -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); +-static DEFINE_SPINLOCK(unwind_lock); ++static DEFINE_RAW_SPINLOCK(unwind_lock); + static LIST_HEAD(unwind_tables); - /* - * The secondary processor is waiting to be released from -@@ -107,7 +107,7 @@ - * now the secondary core is starting up let it run its - * calibrations, then wait for it to finish - */ -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); + /* Convert a prel31 symbol to an absolute address */ +@@ -201,7 +201,7 @@ static const struct unwind_idx *unwind_find_idx(unsigned long addr) + /* module unwind tables */ + struct unwind_table *table; - return pen_release != -1 ? -ENOSYS : 0; - } -diff -Nur linux-4.4.62.orig/arch/arm/mach-qcom/platsmp.c linux-4.4.62/arch/arm/mach-qcom/platsmp.c ---- linux-4.4.62.orig/arch/arm/mach-qcom/platsmp.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-qcom/platsmp.c 2017-04-18 17:38:07.906638076 +0200 -@@ -46,7 +46,7 @@ +- spin_lock_irqsave(&unwind_lock, flags); ++ raw_spin_lock_irqsave(&unwind_lock, flags); + list_for_each_entry(table, &unwind_tables, list) { + if (addr >= table->begin_addr && + addr < table->end_addr) { +@@ -213,7 +213,7 @@ static const struct unwind_idx *unwind_find_idx(unsigned long addr) + break; + } + } +- spin_unlock_irqrestore(&unwind_lock, flags); ++ raw_spin_unlock_irqrestore(&unwind_lock, flags); + } - extern void secondary_startup_arm(void); + pr_debug("%s: idx = %p\n", __func__, idx); +@@ -529,9 +529,9 @@ struct unwind_table *unwind_table_add(unsigned long start, unsigned long size, + tab->begin_addr = text_addr; + tab->end_addr = text_addr + text_size; --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); +- spin_lock_irqsave(&unwind_lock, flags); ++ raw_spin_lock_irqsave(&unwind_lock, flags); + list_add_tail(&tab->list, &unwind_tables); +- spin_unlock_irqrestore(&unwind_lock, flags); ++ raw_spin_unlock_irqrestore(&unwind_lock, flags); - #ifdef CONFIG_HOTPLUG_CPU - static void qcom_cpu_die(unsigned int cpu) -@@ -60,8 +60,8 @@ - /* - * Synchronise with the boot thread. - */ -- spin_lock(&boot_lock); -- spin_unlock(&boot_lock); -+ raw_spin_lock(&boot_lock); -+ raw_spin_unlock(&boot_lock); + return tab; } +@@ -543,9 +543,9 @@ void unwind_table_del(struct unwind_table *tab) + if (!tab) + return; - static int scss_release_secondary(unsigned int cpu) -@@ -284,7 +284,7 @@ - * set synchronisation state between this boot processor - * and the secondary one - */ -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); - - /* - * Send the secondary CPU a soft interrupt, thereby causing -@@ -297,7 +297,7 @@ - * now the secondary core is starting up let it run its - * calibrations, then wait for it to finish - */ -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); +- spin_lock_irqsave(&unwind_lock, flags); ++ raw_spin_lock_irqsave(&unwind_lock, flags); + list_del(&tab->list); +- spin_unlock_irqrestore(&unwind_lock, flags); ++ raw_spin_unlock_irqrestore(&unwind_lock, flags); - return ret; - } -diff -Nur linux-4.4.62.orig/arch/arm/mach-spear/platsmp.c linux-4.4.62/arch/arm/mach-spear/platsmp.c ---- linux-4.4.62.orig/arch/arm/mach-spear/platsmp.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-spear/platsmp.c 2017-04-18 17:38:07.906638076 +0200 -@@ -32,7 +32,7 @@ - sync_cache_w(&pen_release); + kfree(tab); } +diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c +index d7bef2144760..36a3e51492f7 100644 +--- a/arch/arm/kvm/arm.c ++++ b/arch/arm/kvm/arm.c +@@ -496,18 +496,18 @@ static void kvm_arm_resume_guest(struct kvm *kvm) + struct kvm_vcpu *vcpu; --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); + kvm_for_each_vcpu(i, vcpu, kvm) { +- wait_queue_head_t *wq = kvm_arch_vcpu_wq(vcpu); ++ struct swait_queue_head *wq = kvm_arch_vcpu_wq(vcpu); - static void __iomem *scu_base = IOMEM(VA_SCU_BASE); + vcpu->arch.pause = false; +- wake_up_interruptible(wq); ++ swake_up(wq); + } + } -@@ -47,8 +47,8 @@ - /* - * Synchronise with the boot thread. - */ -- spin_lock(&boot_lock); -- spin_unlock(&boot_lock); -+ raw_spin_lock(&boot_lock); -+ raw_spin_unlock(&boot_lock); + static void vcpu_sleep(struct kvm_vcpu *vcpu) + { +- wait_queue_head_t *wq = kvm_arch_vcpu_wq(vcpu); ++ struct swait_queue_head *wq = kvm_arch_vcpu_wq(vcpu); + +- wait_event_interruptible(*wq, ((!vcpu->arch.power_off) && ++ swait_event_interruptible(*wq, ((!vcpu->arch.power_off) && + (!vcpu->arch.pause))); } - static int spear13xx_boot_secondary(unsigned int cpu, struct task_struct *idle) -@@ -59,7 +59,7 @@ - * set synchronisation state between this boot processor - * and the secondary one - */ -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); +@@ -566,7 +566,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) + * involves poking the GIC, which must be done in a + * non-preemptible context. + */ +- preempt_disable(); ++ migrate_disable(); + kvm_timer_flush_hwstate(vcpu); + kvm_vgic_flush_hwstate(vcpu); - /* - * The secondary processor is waiting to be released from -@@ -84,7 +84,7 @@ - * now the secondary core is starting up let it run its - * calibrations, then wait for it to finish - */ -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); +@@ -585,7 +585,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) + local_irq_enable(); + kvm_timer_sync_hwstate(vcpu); + kvm_vgic_sync_hwstate(vcpu); +- preempt_enable(); ++ migrate_enable(); + continue; + } - return pen_release != -1 ? -ENOSYS : 0; - } -diff -Nur linux-4.4.62.orig/arch/arm/mach-sti/platsmp.c linux-4.4.62/arch/arm/mach-sti/platsmp.c ---- linux-4.4.62.orig/arch/arm/mach-sti/platsmp.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mach-sti/platsmp.c 2017-04-18 17:38:07.906638076 +0200 -@@ -35,7 +35,7 @@ - sync_cache_w(&pen_release); - } +@@ -639,7 +639,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); + kvm_vgic_sync_hwstate(vcpu); - static void sti_secondary_init(unsigned int cpu) - { -@@ -48,8 +48,8 @@ - /* - * Synchronise with the boot thread. - */ -- spin_lock(&boot_lock); -- spin_unlock(&boot_lock); -+ raw_spin_lock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - } +- preempt_enable(); ++ migrate_enable(); - static int sti_boot_secondary(unsigned int cpu, struct task_struct *idle) -@@ -60,7 +60,7 @@ - * set synchronisation state between this boot processor - * and the secondary one - */ -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); + ret = handle_exit(vcpu, run, ret); + } +diff --git a/arch/arm/kvm/psci.c b/arch/arm/kvm/psci.c +index 443db0c43d7c..a08d7a93aebb 100644 +--- a/arch/arm/kvm/psci.c ++++ b/arch/arm/kvm/psci.c +@@ -70,7 +70,7 @@ static unsigned long kvm_psci_vcpu_on(struct kvm_vcpu *source_vcpu) + { + struct kvm *kvm = source_vcpu->kvm; + struct kvm_vcpu *vcpu = NULL; +- wait_queue_head_t *wq; ++ struct swait_queue_head *wq; + unsigned long cpu_id; + unsigned long context_id; + phys_addr_t target_pc; +@@ -119,7 +119,7 @@ static unsigned long kvm_psci_vcpu_on(struct kvm_vcpu *source_vcpu) + smp_mb(); /* Make sure the above is visible */ - /* - * The secondary processor is waiting to be released from -@@ -91,7 +91,7 @@ - * now the secondary core is starting up let it run its - * calibrations, then wait for it to finish - */ -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); + wq = kvm_arch_vcpu_wq(vcpu); +- wake_up_interruptible(wq); ++ swake_up(wq); - return pen_release != -1 ? -ENOSYS : 0; + return PSCI_RET_SUCCESS; } -diff -Nur linux-4.4.62.orig/arch/arm/mm/fault.c linux-4.4.62/arch/arm/mm/fault.c ---- linux-4.4.62.orig/arch/arm/mm/fault.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mm/fault.c 2017-04-18 17:38:07.906638076 +0200 -@@ -430,6 +430,9 @@ - if (addr < TASK_SIZE) - return do_page_fault(addr, fsr, regs); - -+ if (interrupts_enabled(regs)) -+ local_irq_enable(); -+ - if (user_mode(regs)) - goto bad_area; +diff --git a/arch/arm/mach-at91/Kconfig b/arch/arm/mach-at91/Kconfig +index 28656c2b54a0..3f501305ca26 100644 +--- a/arch/arm/mach-at91/Kconfig ++++ b/arch/arm/mach-at91/Kconfig +@@ -99,6 +99,7 @@ config HAVE_AT91_USB_CLK + config COMMON_CLK_AT91 + bool + select COMMON_CLK ++ select MFD_SYSCON -@@ -497,6 +500,9 @@ - static int - do_sect_fault(unsigned long addr, unsigned int fsr, struct pt_regs *regs) - { -+ if (interrupts_enabled(regs)) -+ local_irq_enable(); -+ - do_bad_area(addr, fsr, regs); - return 0; - } -diff -Nur linux-4.4.62.orig/arch/arm/mm/highmem.c linux-4.4.62/arch/arm/mm/highmem.c ---- linux-4.4.62.orig/arch/arm/mm/highmem.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/mm/highmem.c 2017-04-18 17:38:07.906638076 +0200 -@@ -34,6 +34,11 @@ - return *ptep; - } + config HAVE_AT91_SMD + bool +diff --git a/arch/arm/mach-at91/at91rm9200.c b/arch/arm/mach-at91/at91rm9200.c +index c1a7c6cc00e1..63b4fa25b48a 100644 +--- a/arch/arm/mach-at91/at91rm9200.c ++++ b/arch/arm/mach-at91/at91rm9200.c +@@ -12,7 +12,6 @@ + #include -+static unsigned int fixmap_idx(int type) -+{ -+ return FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); -+} -+ - void *kmap(struct page *page) - { - might_sleep(); -@@ -54,12 +59,13 @@ + #include +-#include - void *kmap_atomic(struct page *page) - { -+ pte_t pte = mk_pte(page, kmap_prot); - unsigned int idx; - unsigned long vaddr; - void *kmap; - int type; + #include "generic.h" + #include "soc.h" +@@ -33,7 +32,6 @@ static void __init at91rm9200_dt_device_init(void) -- preempt_disable(); -+ preempt_disable_nort(); - pagefault_disable(); - if (!PageHighMem(page)) - return page_address(page); -@@ -79,7 +85,7 @@ + of_platform_populate(NULL, of_default_bus_match_table, NULL, soc_dev); - type = kmap_atomic_idx_push(); +- arm_pm_idle = at91rm9200_idle; + at91rm9200_pm_init(); + } -- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); -+ idx = fixmap_idx(type); - vaddr = __fix_to_virt(idx); - #ifdef CONFIG_DEBUG_HIGHMEM - /* -@@ -93,7 +99,10 @@ - * in place, so the contained TLB flush ensures the TLB is updated - * with the new mapping. - */ -- set_fixmap_pte(idx, mk_pte(page, kmap_prot)); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = pte; -+#endif -+ set_fixmap_pte(idx, pte); +diff --git a/arch/arm/mach-at91/at91sam9.c b/arch/arm/mach-at91/at91sam9.c +index 7eb64f763034..cada2a6412b3 100644 +--- a/arch/arm/mach-at91/at91sam9.c ++++ b/arch/arm/mach-at91/at91sam9.c +@@ -62,8 +62,6 @@ static void __init at91sam9_common_init(void) + soc_dev = soc_device_to_device(soc); - return (void *)vaddr; + of_platform_populate(NULL, of_default_bus_match_table, NULL, soc_dev); +- +- arm_pm_idle = at91sam9_idle; } -@@ -106,44 +115,75 @@ - if (kvaddr >= (void *)FIXADDR_START) { - type = kmap_atomic_idx(); -- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); -+ idx = fixmap_idx(type); + static void __init at91sam9_dt_device_init(void) +diff --git a/arch/arm/mach-at91/generic.h b/arch/arm/mach-at91/generic.h +index b0fa7dc7286d..28ca57a2060f 100644 +--- a/arch/arm/mach-at91/generic.h ++++ b/arch/arm/mach-at91/generic.h +@@ -11,27 +11,18 @@ + #ifndef _AT91_GENERIC_H + #define _AT91_GENERIC_H - if (cache_is_vivt()) - __cpuc_flush_dcache_area((void *)vaddr, PAGE_SIZE); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = __pte(0); -+#endif - #ifdef CONFIG_DEBUG_HIGHMEM - BUG_ON(vaddr != __fix_to_virt(idx)); -- set_fixmap_pte(idx, __pte(0)); +-#include +-#include +- +- /* Map io */ +-extern void __init at91_map_io(void); +-extern void __init at91_alt_map_io(void); +- +-/* idle */ +-extern void at91rm9200_idle(void); +-extern void at91sam9_idle(void); +- + #ifdef CONFIG_PM + extern void __init at91rm9200_pm_init(void); + extern void __init at91sam9260_pm_init(void); + extern void __init at91sam9g45_pm_init(void); + extern void __init at91sam9x5_pm_init(void); ++extern void __init sama5_pm_init(void); #else - (void) idx; /* to kill a warning */ + static inline void __init at91rm9200_pm_init(void) { } + static inline void __init at91sam9260_pm_init(void) { } + static inline void __init at91sam9g45_pm_init(void) { } + static inline void __init at91sam9x5_pm_init(void) { } ++static inline void __init sama5_pm_init(void) { } #endif -+ set_fixmap_pte(idx, __pte(0)); - kmap_atomic_idx_pop(); - } else if (vaddr >= PKMAP_ADDR(0) && vaddr < PKMAP_ADDR(LAST_PKMAP)) { - /* this address was obtained through kmap_high_get() */ - kunmap_high(pte_page(pkmap_page_table[PKMAP_NR(vaddr)])); - } - pagefault_enable(); -- preempt_enable(); -+ preempt_enable_nort(); - } - EXPORT_SYMBOL(__kunmap_atomic); - void *kmap_atomic_pfn(unsigned long pfn) - { -+ pte_t pte = pfn_pte(pfn, kmap_prot); - unsigned long vaddr; - int idx, type; - struct page *page = pfn_to_page(pfn); + #endif /* _AT91_GENERIC_H */ +diff --git a/arch/arm/mach-at91/pm.c b/arch/arm/mach-at91/pm.c +index d687f860a2da..ab53332a9637 100644 +--- a/arch/arm/mach-at91/pm.c ++++ b/arch/arm/mach-at91/pm.c +@@ -31,10 +31,13 @@ + #include + #include + #include ++#include -- preempt_disable(); -+ preempt_disable_nort(); - pagefault_disable(); - if (!PageHighMem(page)) - return page_address(page); + #include "generic.h" + #include "pm.h" - type = kmap_atomic_idx_push(); -- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); -+ idx = fixmap_idx(type); - vaddr = __fix_to_virt(idx); - #ifdef CONFIG_DEBUG_HIGHMEM - BUG_ON(!pte_none(get_fixmap_pte(vaddr))); - #endif -- set_fixmap_pte(idx, pfn_pte(pfn, kmap_prot)); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = pte; -+#endif -+ set_fixmap_pte(idx, pte); ++static void __iomem *pmc; ++ + /* + * FIXME: this is needed to communicate between the pinctrl driver and + * the PM implementation in the machine. Possibly part of the PM +@@ -87,7 +90,7 @@ static int at91_pm_verify_clocks(void) + unsigned long scsr; + int i; - return (void *)vaddr; +- scsr = at91_pmc_read(AT91_PMC_SCSR); ++ scsr = readl(pmc + AT91_PMC_SCSR); + + /* USB must not be using PLLB */ + if ((scsr & at91_pm_data.uhp_udp_mask) != 0) { +@@ -101,8 +104,7 @@ static int at91_pm_verify_clocks(void) + + if ((scsr & (AT91_PMC_PCK0 << i)) == 0) + continue; +- +- css = at91_pmc_read(AT91_PMC_PCKR(i)) & AT91_PMC_CSS; ++ css = readl(pmc + AT91_PMC_PCKR(i)) & AT91_PMC_CSS; + if (css != AT91_PMC_CSS_SLOW) { + pr_err("AT91: PM - Suspend-to-RAM with PCK%d src %d\n", i, css); + return 0; +@@ -145,8 +147,8 @@ static void at91_pm_suspend(suspend_state_t state) + flush_cache_all(); + outer_disable(); + +- at91_suspend_sram_fn(at91_pmc_base, at91_ramc_base[0], +- at91_ramc_base[1], pm_data); ++ at91_suspend_sram_fn(pmc, at91_ramc_base[0], ++ at91_ramc_base[1], pm_data); + + outer_resume(); } -+#if defined CONFIG_PREEMPT_RT_FULL -+void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) +@@ -369,6 +371,21 @@ static __init void at91_dt_ramc(void) + at91_pm_set_standby(standby); + } + ++void at91rm9200_idle(void) +{ -+ int i; -+ -+ /* -+ * Clear @prev's kmap_atomic mappings -+ */ -+ for (i = 0; i < prev_p->kmap_idx; i++) { -+ int idx = fixmap_idx(i); -+ -+ set_fixmap_pte(idx, __pte(0)); -+ } + /* -+ * Restore @next_p's kmap_atomic mappings ++ * Disable the processor clock. The processor will be automatically ++ * re-enabled by an interrupt or by a reset. + */ -+ for (i = 0; i < next_p->kmap_idx; i++) { -+ int idx = fixmap_idx(i); ++ writel(AT91_PMC_PCK, pmc + AT91_PMC_SCDR); ++} + -+ if (!pte_none(next_p->kmap_pte[i])) -+ set_fixmap_pte(idx, next_p->kmap_pte[i]); -+ } ++void at91sam9_idle(void) ++{ ++ writel(AT91_PMC_PCK, pmc + AT91_PMC_SCDR); ++ cpu_do_idle(); +} -+#endif -diff -Nur linux-4.4.62.orig/arch/arm/plat-versatile/platsmp.c linux-4.4.62/arch/arm/plat-versatile/platsmp.c ---- linux-4.4.62.orig/arch/arm/plat-versatile/platsmp.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm/plat-versatile/platsmp.c 2017-04-18 17:38:07.906638076 +0200 -@@ -30,7 +30,7 @@ - sync_cache_w(&pen_release); ++ + static void __init at91_pm_sram_init(void) + { + struct gen_pool *sram_pool; +@@ -415,13 +432,36 @@ static void __init at91_pm_sram_init(void) + &at91_pm_suspend_in_sram, at91_pm_suspend_in_sram_sz); + } + +-static void __init at91_pm_init(void) ++static const struct of_device_id atmel_pmc_ids[] __initconst = { ++ { .compatible = "atmel,at91rm9200-pmc" }, ++ { .compatible = "atmel,at91sam9260-pmc" }, ++ { .compatible = "atmel,at91sam9g45-pmc" }, ++ { .compatible = "atmel,at91sam9n12-pmc" }, ++ { .compatible = "atmel,at91sam9x5-pmc" }, ++ { .compatible = "atmel,sama5d3-pmc" }, ++ { .compatible = "atmel,sama5d2-pmc" }, ++ { /* sentinel */ }, ++}; ++ ++static void __init at91_pm_init(void (*pm_idle)(void)) + { +- at91_pm_sram_init(); ++ struct device_node *pmc_np; + + if (at91_cpuidle_device.dev.platform_data) + platform_device_register(&at91_cpuidle_device); + ++ pmc_np = of_find_matching_node(NULL, atmel_pmc_ids); ++ pmc = of_iomap(pmc_np, 0); ++ if (!pmc) { ++ pr_err("AT91: PM not supported, PMC not found\n"); ++ return; ++ } ++ ++ if (pm_idle) ++ arm_pm_idle = pm_idle; ++ ++ at91_pm_sram_init(); ++ + if (at91_suspend_sram_fn) + suspend_set_ops(&at91_pm_ops); + else +@@ -440,7 +480,7 @@ void __init at91rm9200_pm_init(void) + at91_pm_data.uhp_udp_mask = AT91RM9200_PMC_UHP | AT91RM9200_PMC_UDP; + at91_pm_data.memctrl = AT91_MEMCTRL_MC; + +- at91_pm_init(); ++ at91_pm_init(at91rm9200_idle); + } + + void __init at91sam9260_pm_init(void) +@@ -448,7 +488,7 @@ void __init at91sam9260_pm_init(void) + at91_dt_ramc(); + at91_pm_data.memctrl = AT91_MEMCTRL_SDRAMC; + at91_pm_data.uhp_udp_mask = AT91SAM926x_PMC_UHP | AT91SAM926x_PMC_UDP; +- return at91_pm_init(); ++ at91_pm_init(at91sam9_idle); + } + + void __init at91sam9g45_pm_init(void) +@@ -456,7 +496,7 @@ void __init at91sam9g45_pm_init(void) + at91_dt_ramc(); + at91_pm_data.uhp_udp_mask = AT91SAM926x_PMC_UHP; + at91_pm_data.memctrl = AT91_MEMCTRL_DDRSDR; +- return at91_pm_init(); ++ at91_pm_init(at91sam9_idle); + } + + void __init at91sam9x5_pm_init(void) +@@ -464,5 +504,13 @@ void __init at91sam9x5_pm_init(void) + at91_dt_ramc(); + at91_pm_data.uhp_udp_mask = AT91SAM926x_PMC_UHP | AT91SAM926x_PMC_UDP; + at91_pm_data.memctrl = AT91_MEMCTRL_DDRSDR; +- return at91_pm_init(); ++ at91_pm_init(at91sam9_idle); ++} ++ ++void __init sama5_pm_init(void) ++{ ++ at91_dt_ramc(); ++ at91_pm_data.uhp_udp_mask = AT91SAM926x_PMC_UHP | AT91SAM926x_PMC_UDP; ++ at91_pm_data.memctrl = AT91_MEMCTRL_DDRSDR; ++ at91_pm_init(NULL); + } +diff --git a/arch/arm/mach-at91/sama5.c b/arch/arm/mach-at91/sama5.c +index d9cf6799aec0..df8fdf1cf66d 100644 +--- a/arch/arm/mach-at91/sama5.c ++++ b/arch/arm/mach-at91/sama5.c +@@ -51,7 +51,7 @@ static void __init sama5_dt_device_init(void) + soc_dev = soc_device_to_device(soc); + + of_platform_populate(NULL, of_default_bus_match_table, NULL, soc_dev); +- at91sam9x5_pm_init(); ++ sama5_pm_init(); + } + + static const char *const sama5_dt_board_compat[] __initconst = { +diff --git a/arch/arm/mach-exynos/platsmp.c b/arch/arm/mach-exynos/platsmp.c +index 98a2c0cbb833..310dce500d3e 100644 +--- a/arch/arm/mach-exynos/platsmp.c ++++ b/arch/arm/mach-exynos/platsmp.c +@@ -230,7 +230,7 @@ static void __iomem *scu_base_addr(void) + return (void __iomem *)(S5P_VA_SCU); } -static DEFINE_SPINLOCK(boot_lock); +static DEFINE_RAW_SPINLOCK(boot_lock); - void versatile_secondary_init(unsigned int cpu) + static void exynos_secondary_init(unsigned int cpu) { -@@ -43,8 +43,8 @@ +@@ -243,8 +243,8 @@ static void exynos_secondary_init(unsigned int cpu) /* * Synchronise with the boot thread. */ @@ -1157,8 +1006,8 @@ diff -Nur linux-4.4.62.orig/arch/arm/plat-versatile/platsmp.c linux-4.4.62/arch/ + raw_spin_unlock(&boot_lock); } - int versatile_boot_secondary(unsigned int cpu, struct task_struct *idle) -@@ -55,7 +55,7 @@ + int exynos_set_boot_addr(u32 core_id, unsigned long boot_addr) +@@ -308,7 +308,7 @@ static int exynos_boot_secondary(unsigned int cpu, struct task_struct *idle) * Set synchronisation state between this boot processor * and the secondary one */ @@ -1166,1238 +1015,1181 @@ diff -Nur linux-4.4.62.orig/arch/arm/plat-versatile/platsmp.c linux-4.4.62/arch/ + raw_spin_lock(&boot_lock); /* - * This is really belt and braces; we hold unintended secondary -@@ -85,7 +85,7 @@ - * now the secondary core is starting up let it run its + * The secondary processor is waiting to be released from +@@ -335,7 +335,7 @@ static int exynos_boot_secondary(unsigned int cpu, struct task_struct *idle) + + if (timeout == 0) { + printk(KERN_ERR "cpu1 power enable failed"); +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + return -ETIMEDOUT; + } + } +@@ -381,7 +381,7 @@ static int exynos_boot_secondary(unsigned int cpu, struct task_struct *idle) * calibrations, then wait for it to finish */ + fail: - spin_unlock(&boot_lock); + raw_spin_unlock(&boot_lock); - return pen_release != -1 ? -ENOSYS : 0; + return pen_release != -1 ? ret : 0; } -diff -Nur linux-4.4.62.orig/arch/arm64/include/asm/thread_info.h linux-4.4.62/arch/arm64/include/asm/thread_info.h ---- linux-4.4.62.orig/arch/arm64/include/asm/thread_info.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm64/include/asm/thread_info.h 2017-04-18 17:38:07.906638076 +0200 -@@ -49,6 +49,7 @@ - mm_segment_t addr_limit; /* address limit */ - struct task_struct *task; /* main task structure */ - int preempt_count; /* 0 => preemptable, <0 => bug */ -+ int preempt_lazy_count; /* 0 => preemptable, <0 => bug */ - int cpu; /* cpu */ - }; - -@@ -103,6 +104,7 @@ - #define TIF_NEED_RESCHED 1 - #define TIF_NOTIFY_RESUME 2 /* callback before returning to user */ - #define TIF_FOREIGN_FPSTATE 3 /* CPU's FP state is not current's */ -+#define TIF_NEED_RESCHED_LAZY 4 - #define TIF_NOHZ 7 - #define TIF_SYSCALL_TRACE 8 - #define TIF_SYSCALL_AUDIT 9 -@@ -118,6 +120,7 @@ - #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) - #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) - #define _TIF_FOREIGN_FPSTATE (1 << TIF_FOREIGN_FPSTATE) -+#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) - #define _TIF_NOHZ (1 << TIF_NOHZ) - #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) - #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) -@@ -126,7 +129,8 @@ - #define _TIF_32BIT (1 << TIF_32BIT) - - #define _TIF_WORK_MASK (_TIF_NEED_RESCHED | _TIF_SIGPENDING | \ -- _TIF_NOTIFY_RESUME | _TIF_FOREIGN_FPSTATE) -+ _TIF_NOTIFY_RESUME | _TIF_FOREIGN_FPSTATE | \ -+ _TIF_NEED_RESCHED_LAZY) - - #define _TIF_SYSCALL_WORK (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | \ - _TIF_SYSCALL_TRACEPOINT | _TIF_SECCOMP | \ -diff -Nur linux-4.4.62.orig/arch/arm64/Kconfig linux-4.4.62/arch/arm64/Kconfig ---- linux-4.4.62.orig/arch/arm64/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm64/Kconfig 2017-04-18 17:38:07.906638076 +0200 -@@ -76,6 +76,7 @@ - select HAVE_PERF_REGS - select HAVE_PERF_USER_STACK_DUMP - select HAVE_RCU_TABLE_FREE -+ select HAVE_PREEMPT_LAZY - select HAVE_SYSCALL_TRACEPOINTS - select IOMMU_DMA if IOMMU_SUPPORT - select IRQ_DOMAIN -@@ -582,7 +583,7 @@ - - config XEN - bool "Xen guest support on ARM64" -- depends on ARM64 && OF -+ depends on ARM64 && OF && !PREEMPT_RT_FULL - select SWIOTLB_XEN - help - Say Y if you want to run Linux in a Virtual Machine on Xen on ARM64. -diff -Nur linux-4.4.62.orig/arch/arm64/kernel/asm-offsets.c linux-4.4.62/arch/arm64/kernel/asm-offsets.c ---- linux-4.4.62.orig/arch/arm64/kernel/asm-offsets.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm64/kernel/asm-offsets.c 2017-04-18 17:38:07.906638076 +0200 -@@ -35,6 +35,7 @@ - BLANK(); - DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); - DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); -+ DEFINE(TI_PREEMPT_LAZY, offsetof(struct thread_info, preempt_lazy_count)); - DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit)); - DEFINE(TI_TASK, offsetof(struct thread_info, task)); - DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); -diff -Nur linux-4.4.62.orig/arch/arm64/kernel/entry.S linux-4.4.62/arch/arm64/kernel/entry.S ---- linux-4.4.62.orig/arch/arm64/kernel/entry.S 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/arm64/kernel/entry.S 2017-04-18 17:38:07.910638232 +0200 -@@ -376,11 +376,16 @@ - #ifdef CONFIG_PREEMPT - get_thread_info tsk - ldr w24, [tsk, #TI_PREEMPT] // get preempt count -- cbnz w24, 1f // preempt count != 0 -+ cbnz w24, 2f // preempt count != 0 - ldr x0, [tsk, #TI_FLAGS] // get flags -- tbz x0, #TIF_NEED_RESCHED, 1f // needs rescheduling? -- bl el1_preempt -+ tbnz x0, #TIF_NEED_RESCHED, 1f // needs rescheduling? -+ -+ ldr w24, [tsk, #TI_PREEMPT_LAZY] // get preempt lazy count -+ cbnz w24, 2f // preempt lazy count != 0 -+ tbz x0, #TIF_NEED_RESCHED_LAZY, 2f // needs rescheduling? - 1: -+ bl el1_preempt -+2: - #endif - #ifdef CONFIG_TRACE_IRQFLAGS - bl trace_hardirqs_on -@@ -394,6 +399,7 @@ - 1: bl preempt_schedule_irq // irq en/disable is done inside - ldr x0, [tsk, #TI_FLAGS] // get new tasks TI_FLAGS - tbnz x0, #TIF_NEED_RESCHED, 1b // needs rescheduling? -+ tbnz x0, #TIF_NEED_RESCHED_LAZY, 1b // needs rescheduling? - ret x24 - #endif +diff --git a/arch/arm/mach-hisi/platmcpm.c b/arch/arm/mach-hisi/platmcpm.c +index b5f8f5ffda79..9753a84df9c4 100644 +--- a/arch/arm/mach-hisi/platmcpm.c ++++ b/arch/arm/mach-hisi/platmcpm.c +@@ -61,7 +61,7 @@ -@@ -638,6 +644,7 @@ - */ - work_pending: - tbnz x1, #TIF_NEED_RESCHED, work_resched -+ tbnz x1, #TIF_NEED_RESCHED_LAZY, work_resched - /* TIF_SIGPENDING, TIF_NOTIFY_RESUME or TIF_FOREIGN_FPSTATE case */ - ldr x2, [sp, #S_PSTATE] - mov x0, sp // 'regs' -diff -Nur linux-4.4.62.orig/arch/Kconfig linux-4.4.62/arch/Kconfig ---- linux-4.4.62.orig/arch/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/Kconfig 2017-04-18 17:38:07.902637922 +0200 -@@ -9,6 +9,7 @@ - tristate "OProfile system profiling" - depends on PROFILING - depends on HAVE_OPROFILE -+ depends on !PREEMPT_RT_FULL - select RING_BUFFER - select RING_BUFFER_ALLOW_SWAP - help -@@ -52,6 +53,7 @@ - config JUMP_LABEL - bool "Optimize very unlikely/likely branches" - depends on HAVE_ARCH_JUMP_LABEL -+ depends on (!INTERRUPT_OFF_HIST && !PREEMPT_OFF_HIST && !WAKEUP_LATENCY_HIST && !MISSED_TIMER_OFFSETS_HIST) - help - This option enables a transparent branch optimization that - makes certain almost-always-true or almost-always-false branch -diff -Nur linux-4.4.62.orig/arch/mips/Kconfig linux-4.4.62/arch/mips/Kconfig ---- linux-4.4.62.orig/arch/mips/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/mips/Kconfig 2017-04-18 17:38:08.026642730 +0200 -@@ -2411,7 +2411,7 @@ - # - config HIGHMEM - bool "High Memory Support" -- depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM && !CPU_MIPS32_3_5_EVA -+ depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM && !CPU_MIPS32_3_5_EVA && !PREEMPT_RT_FULL + static void __iomem *sysctrl, *fabric; + static int hip04_cpu_table[HIP04_MAX_CLUSTERS][HIP04_MAX_CPUS_PER_CLUSTER]; +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + static u32 fabric_phys_addr; + /* + * [0]: bootwrapper physical address +@@ -113,7 +113,7 @@ static int hip04_boot_secondary(unsigned int l_cpu, struct task_struct *idle) + if (cluster >= HIP04_MAX_CLUSTERS || cpu >= HIP04_MAX_CPUS_PER_CLUSTER) + return -EINVAL; - config CPU_SUPPORTS_HIGHMEM - bool -diff -Nur linux-4.4.62.orig/arch/mips/kvm/mips.c linux-4.4.62/arch/mips/kvm/mips.c ---- linux-4.4.62.orig/arch/mips/kvm/mips.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/mips/kvm/mips.c 2017-04-18 17:38:08.026642730 +0200 -@@ -454,8 +454,8 @@ +- spin_lock_irq(&boot_lock); ++ raw_spin_lock_irq(&boot_lock); - dvcpu->arch.wait = 0; + if (hip04_cpu_table[cluster][cpu]) + goto out; +@@ -147,7 +147,7 @@ static int hip04_boot_secondary(unsigned int l_cpu, struct task_struct *idle) -- if (waitqueue_active(&dvcpu->wq)) -- wake_up_interruptible(&dvcpu->wq); -+ if (swait_active(&dvcpu->wq)) -+ swake_up(&dvcpu->wq); + out: + hip04_cpu_table[cluster][cpu]++; +- spin_unlock_irq(&boot_lock); ++ raw_spin_unlock_irq(&boot_lock); return 0; } -@@ -1183,8 +1183,8 @@ - kvm_mips_callbacks->queue_timer_int(vcpu); +@@ -162,11 +162,11 @@ static void hip04_cpu_die(unsigned int l_cpu) + cpu = MPIDR_AFFINITY_LEVEL(mpidr, 0); + cluster = MPIDR_AFFINITY_LEVEL(mpidr, 1); - vcpu->arch.wait = 0; -- if (waitqueue_active(&vcpu->wq)) -- wake_up_interruptible(&vcpu->wq); -+ if (swait_active(&vcpu->wq)) -+ swake_up(&vcpu->wq); - } +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + hip04_cpu_table[cluster][cpu]--; + if (hip04_cpu_table[cluster][cpu] == 1) { + /* A power_up request went ahead of us. */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + return; + } else if (hip04_cpu_table[cluster][cpu] > 1) { + pr_err("Cluster %d CPU%d boots multiple times\n", cluster, cpu); +@@ -174,7 +174,7 @@ static void hip04_cpu_die(unsigned int l_cpu) + } - /* low level hrtimer wake routine */ -diff -Nur linux-4.4.62.orig/arch/powerpc/include/asm/kvm_host.h linux-4.4.62/arch/powerpc/include/asm/kvm_host.h ---- linux-4.4.62.orig/arch/powerpc/include/asm/kvm_host.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/powerpc/include/asm/kvm_host.h 2017-04-18 17:38:08.026642730 +0200 -@@ -286,7 +286,7 @@ - struct list_head runnable_threads; - struct list_head preempt_list; - spinlock_t lock; -- wait_queue_head_t wq; -+ struct swait_queue_head wq; - spinlock_t stoltb_lock; /* protects stolen_tb and preempt_tb */ - u64 stolen_tb; - u64 preempt_tb; -@@ -627,7 +627,7 @@ - u8 prodded; - u32 last_inst; - -- wait_queue_head_t *wqp; -+ struct swait_queue_head *wqp; - struct kvmppc_vcore *vcore; - int ret; - int trap; -diff -Nur linux-4.4.62.orig/arch/powerpc/include/asm/thread_info.h linux-4.4.62/arch/powerpc/include/asm/thread_info.h ---- linux-4.4.62.orig/arch/powerpc/include/asm/thread_info.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/powerpc/include/asm/thread_info.h 2017-04-18 17:38:08.026642730 +0200 -@@ -42,6 +42,8 @@ - int cpu; /* cpu we're on */ - int preempt_count; /* 0 => preemptable, - <0 => BUG */ -+ int preempt_lazy_count; /* 0 => preemptable, -+ <0 => BUG */ - unsigned long local_flags; /* private flags for thread */ + last_man = hip04_cluster_is_down(cluster); +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + if (last_man) { + /* Since it's Cortex A15, disable L2 prefetching. */ + asm volatile( +@@ -203,7 +203,7 @@ static int hip04_cpu_kill(unsigned int l_cpu) + cpu >= HIP04_MAX_CPUS_PER_CLUSTER); - /* low level flags - has atomic operations done on it */ -@@ -82,8 +84,7 @@ - #define TIF_SYSCALL_TRACE 0 /* syscall trace active */ - #define TIF_SIGPENDING 1 /* signal pending */ - #define TIF_NEED_RESCHED 2 /* rescheduling necessary */ --#define TIF_POLLING_NRFLAG 3 /* true if poll_idle() is polling -- TIF_NEED_RESCHED */ -+#define TIF_NEED_RESCHED_LAZY 3 /* lazy rescheduling necessary */ - #define TIF_32BIT 4 /* 32 bit binary */ - #define TIF_RESTORE_TM 5 /* need to restore TM FP/VEC/VSX */ - #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ -@@ -101,6 +102,8 @@ - #if defined(CONFIG_PPC64) - #define TIF_ELF2ABI 18 /* function descriptors must die! */ + count = TIMEOUT_MSEC / POLL_MSEC; +- spin_lock_irq(&boot_lock); ++ raw_spin_lock_irq(&boot_lock); + for (tries = 0; tries < count; tries++) { + if (hip04_cpu_table[cluster][cpu]) + goto err; +@@ -211,10 +211,10 @@ static int hip04_cpu_kill(unsigned int l_cpu) + data = readl_relaxed(sysctrl + SC_CPU_RESET_STATUS(cluster)); + if (data & CORE_WFI_STATUS(cpu)) + break; +- spin_unlock_irq(&boot_lock); ++ raw_spin_unlock_irq(&boot_lock); + /* Wait for clean L2 when the whole cluster is down. */ + msleep(POLL_MSEC); +- spin_lock_irq(&boot_lock); ++ raw_spin_lock_irq(&boot_lock); + } + if (tries >= count) + goto err; +@@ -231,10 +231,10 @@ static int hip04_cpu_kill(unsigned int l_cpu) + goto err; + if (hip04_cluster_is_down(cluster)) + hip04_set_snoop_filter(cluster, 0); +- spin_unlock_irq(&boot_lock); ++ raw_spin_unlock_irq(&boot_lock); + return 1; + err: +- spin_unlock_irq(&boot_lock); ++ raw_spin_unlock_irq(&boot_lock); + return 0; + } #endif -+#define TIF_POLLING_NRFLAG 19 /* true if poll_idle() is polling -+ TIF_NEED_RESCHED */ - - /* as above, but as bit values */ - #define _TIF_SYSCALL_TRACE (1<flags) - set_bits(irqtp->flags, &curtp->flags); + * The secondary processor is waiting to be released from +@@ -107,7 +107,7 @@ static int sirfsoc_boot_secondary(unsigned int cpu, struct task_struct *idle) + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return pen_release != -1 ? -ENOSYS : 0; } -+#endif +diff --git a/arch/arm/mach-qcom/platsmp.c b/arch/arm/mach-qcom/platsmp.c +index 9b00123a315d..0a49fe1bc8cf 100644 +--- a/arch/arm/mach-qcom/platsmp.c ++++ b/arch/arm/mach-qcom/platsmp.c +@@ -46,7 +46,7 @@ - irq_hw_number_t virq_to_hw(unsigned int virq) - { -diff -Nur linux-4.4.62.orig/arch/powerpc/kernel/misc_32.S linux-4.4.62/arch/powerpc/kernel/misc_32.S ---- linux-4.4.62.orig/arch/powerpc/kernel/misc_32.S 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/powerpc/kernel/misc_32.S 2017-04-18 17:38:08.030642885 +0200 -@@ -40,6 +40,7 @@ - * We store the saved ksp_limit in the unused part - * of the STACK_FRAME_OVERHEAD - */ -+#ifndef CONFIG_PREEMPT_RT_FULL - _GLOBAL(call_do_softirq) - mflr r0 - stw r0,4(r1) -@@ -56,6 +57,7 @@ - stw r10,THREAD+KSP_LIMIT(r2) - mtlr r0 - blr -+#endif + extern void secondary_startup_arm(void); - /* - * void call_do_irq(struct pt_regs *regs, struct thread_info *irqtp); -diff -Nur linux-4.4.62.orig/arch/powerpc/kernel/misc_64.S linux-4.4.62/arch/powerpc/kernel/misc_64.S ---- linux-4.4.62.orig/arch/powerpc/kernel/misc_64.S 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/powerpc/kernel/misc_64.S 2017-04-18 17:38:08.030642885 +0200 -@@ -30,6 +30,7 @@ +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); - .text + #ifdef CONFIG_HOTPLUG_CPU + static void qcom_cpu_die(unsigned int cpu) +@@ -60,8 +60,8 @@ static void qcom_secondary_init(unsigned int cpu) + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } -+#ifndef CONFIG_PREEMPT_RT_FULL - _GLOBAL(call_do_softirq) - mflr r0 - std r0,16(r1) -@@ -40,6 +41,7 @@ - ld r0,16(r1) - mtlr r0 - blr -+#endif + static int scss_release_secondary(unsigned int cpu) +@@ -284,7 +284,7 @@ static int qcom_boot_secondary(unsigned int cpu, int (*func)(unsigned int)) + * set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); - _GLOBAL(call_do_irq) - mflr r0 -diff -Nur linux-4.4.62.orig/arch/powerpc/kvm/book3s_hv.c linux-4.4.62/arch/powerpc/kvm/book3s_hv.c ---- linux-4.4.62.orig/arch/powerpc/kvm/book3s_hv.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/powerpc/kvm/book3s_hv.c 2017-04-18 17:38:08.030642885 +0200 -@@ -114,11 +114,11 @@ - static void kvmppc_fast_vcpu_kick_hv(struct kvm_vcpu *vcpu) - { - int cpu; -- wait_queue_head_t *wqp; -+ struct swait_queue_head *wqp; + /* + * Send the secondary CPU a soft interrupt, thereby causing +@@ -297,7 +297,7 @@ static int qcom_boot_secondary(unsigned int cpu, int (*func)(unsigned int)) + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); - wqp = kvm_arch_vcpu_wq(vcpu); -- if (waitqueue_active(wqp)) { -- wake_up_interruptible(wqp); -+ if (swait_active(wqp)) { -+ swake_up(wqp); - ++vcpu->stat.halt_wakeup; - } + return ret; + } +diff --git a/arch/arm/mach-spear/platsmp.c b/arch/arm/mach-spear/platsmp.c +index fd4297713d67..b0553b2c2d53 100644 +--- a/arch/arm/mach-spear/platsmp.c ++++ b/arch/arm/mach-spear/platsmp.c +@@ -32,7 +32,7 @@ static void write_pen_release(int val) + sync_cache_w(&pen_release); + } -@@ -707,8 +707,8 @@ - tvcpu->arch.prodded = 1; - smp_mb(); - if (vcpu->arch.ceded) { -- if (waitqueue_active(&vcpu->wq)) { -- wake_up_interruptible(&vcpu->wq); -+ if (swait_active(&vcpu->wq)) { -+ swake_up(&vcpu->wq); - vcpu->stat.halt_wakeup++; - } - } -@@ -1453,7 +1453,7 @@ - INIT_LIST_HEAD(&vcore->runnable_threads); - spin_lock_init(&vcore->lock); - spin_lock_init(&vcore->stoltb_lock); -- init_waitqueue_head(&vcore->wq); -+ init_swait_queue_head(&vcore->wq); - vcore->preempt_tb = TB_NIL; - vcore->lpcr = kvm->arch.lpcr; - vcore->first_vcpuid = core * threads_per_subcore; -@@ -2525,10 +2525,9 @@ - { - struct kvm_vcpu *vcpu; - int do_sleep = 1; -+ DECLARE_SWAITQUEUE(wait); +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); -- DEFINE_WAIT(wait); -- -- prepare_to_wait(&vc->wq, &wait, TASK_INTERRUPTIBLE); -+ prepare_to_swait(&vc->wq, &wait, TASK_INTERRUPTIBLE); + static void __iomem *scu_base = IOMEM(VA_SCU_BASE); +@@ -47,8 +47,8 @@ static void spear13xx_secondary_init(unsigned int cpu) /* - * Check one last time for pending exceptions and ceded state after -@@ -2542,7 +2541,7 @@ - } - - if (!do_sleep) { -- finish_wait(&vc->wq, &wait); -+ finish_swait(&vc->wq, &wait); - return; - } - -@@ -2550,7 +2549,7 @@ - trace_kvmppc_vcore_blocked(vc, 0); - spin_unlock(&vc->lock); - schedule(); -- finish_wait(&vc->wq, &wait); -+ finish_swait(&vc->wq, &wait); - spin_lock(&vc->lock); - vc->vcore_state = VCORE_INACTIVE; - trace_kvmppc_vcore_blocked(vc, 1); -@@ -2606,7 +2605,7 @@ - kvmppc_start_thread(vcpu, vc); - trace_kvm_guest_enter(vcpu); - } else if (vc->vcore_state == VCORE_SLEEPING) { -- wake_up(&vc->wq); -+ swake_up(&vc->wq); - } + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } - } -diff -Nur linux-4.4.62.orig/arch/powerpc/kvm/Kconfig linux-4.4.62/arch/powerpc/kvm/Kconfig ---- linux-4.4.62.orig/arch/powerpc/kvm/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/powerpc/kvm/Kconfig 2017-04-18 17:38:08.030642885 +0200 -@@ -172,6 +172,7 @@ - config KVM_MPIC - bool "KVM in-kernel MPIC emulation" - depends on KVM && E500 -+ depends on !PREEMPT_RT_FULL - select HAVE_KVM_IRQCHIP - select HAVE_KVM_IRQFD - select HAVE_KVM_IRQ_ROUTING -diff -Nur linux-4.4.62.orig/arch/powerpc/platforms/ps3/device-init.c linux-4.4.62/arch/powerpc/platforms/ps3/device-init.c ---- linux-4.4.62.orig/arch/powerpc/platforms/ps3/device-init.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/powerpc/platforms/ps3/device-init.c 2017-04-18 17:38:08.030642885 +0200 -@@ -752,7 +752,7 @@ - } - pr_debug("%s:%u: notification %s issued\n", __func__, __LINE__, op); + static int spear13xx_boot_secondary(unsigned int cpu, struct task_struct *idle) +@@ -59,7 +59,7 @@ static int spear13xx_boot_secondary(unsigned int cpu, struct task_struct *idle) + * set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); -- res = wait_event_interruptible(dev->done.wait, -+ res = swait_event_interruptible(dev->done.wait, - dev->done.done || kthread_should_stop()); - if (kthread_should_stop()) - res = -EINTR; -diff -Nur linux-4.4.62.orig/arch/s390/include/asm/kvm_host.h linux-4.4.62/arch/s390/include/asm/kvm_host.h ---- linux-4.4.62.orig/arch/s390/include/asm/kvm_host.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/s390/include/asm/kvm_host.h 2017-04-18 17:38:08.030642885 +0200 -@@ -427,7 +427,7 @@ - struct kvm_s390_local_interrupt { - spinlock_t lock; - struct kvm_s390_float_interrupt *float_int; -- wait_queue_head_t *wq; -+ struct swait_queue_head *wq; - atomic_t *cpuflags; - DECLARE_BITMAP(sigp_emerg_pending, KVM_MAX_VCPUS); - struct kvm_s390_irq_payload irq; -diff -Nur linux-4.4.62.orig/arch/s390/kvm/interrupt.c linux-4.4.62/arch/s390/kvm/interrupt.c ---- linux-4.4.62.orig/arch/s390/kvm/interrupt.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/s390/kvm/interrupt.c 2017-04-18 17:38:08.030642885 +0200 -@@ -868,13 +868,13 @@ + /* + * The secondary processor is waiting to be released from +@@ -84,7 +84,7 @@ static int spear13xx_boot_secondary(unsigned int cpu, struct task_struct *idle) + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); - void kvm_s390_vcpu_wakeup(struct kvm_vcpu *vcpu) - { -- if (waitqueue_active(&vcpu->wq)) { -+ if (swait_active(&vcpu->wq)) { - /* - * The vcpu gave up the cpu voluntarily, mark it as a good - * yield-candidate. - */ - vcpu->preempted = true; -- wake_up_interruptible(&vcpu->wq); -+ swake_up(&vcpu->wq); - vcpu->stat.halt_wakeup++; - } + return pen_release != -1 ? -ENOSYS : 0; } -diff -Nur linux-4.4.62.orig/arch/sh/kernel/irq.c linux-4.4.62/arch/sh/kernel/irq.c ---- linux-4.4.62.orig/arch/sh/kernel/irq.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/sh/kernel/irq.c 2017-04-18 17:38:08.030642885 +0200 -@@ -147,6 +147,7 @@ - hardirq_ctx[cpu] = NULL; +diff --git a/arch/arm/mach-sti/platsmp.c b/arch/arm/mach-sti/platsmp.c +index c4ad6eae67fa..e830b20b212f 100644 +--- a/arch/arm/mach-sti/platsmp.c ++++ b/arch/arm/mach-sti/platsmp.c +@@ -35,7 +35,7 @@ static void write_pen_release(int val) + sync_cache_w(&pen_release); } -+#ifndef CONFIG_PREEMPT_RT_FULL - void do_softirq_own_stack(void) +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + static void sti_secondary_init(unsigned int cpu) { - struct thread_info *curctx; -@@ -174,6 +175,7 @@ - "r5", "r6", "r7", "r8", "r9", "r15", "t", "pr" - ); +@@ -48,8 +48,8 @@ static void sti_secondary_init(unsigned int cpu) + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); } -+#endif - #else - static inline void handle_one_irq(unsigned int irq) - { -diff -Nur linux-4.4.62.orig/arch/sparc/Kconfig linux-4.4.62/arch/sparc/Kconfig ---- linux-4.4.62.orig/arch/sparc/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/sparc/Kconfig 2017-04-18 17:38:08.030642885 +0200 -@@ -189,12 +189,10 @@ - source kernel/Kconfig.hz - config RWSEM_GENERIC_SPINLOCK -- bool -- default y if SPARC32 -+ def_bool PREEMPT_RT_FULL + static int sti_boot_secondary(unsigned int cpu, struct task_struct *idle) +@@ -60,7 +60,7 @@ static int sti_boot_secondary(unsigned int cpu, struct task_struct *idle) + * set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); - config RWSEM_XCHGADD_ALGORITHM -- bool -- default y if SPARC64 -+ def_bool !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL + /* + * The secondary processor is waiting to be released from +@@ -91,7 +91,7 @@ static int sti_boot_secondary(unsigned int cpu, struct task_struct *idle) + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); - config GENERIC_HWEIGHT - bool -diff -Nur linux-4.4.62.orig/arch/sparc/kernel/irq_64.c linux-4.4.62/arch/sparc/kernel/irq_64.c ---- linux-4.4.62.orig/arch/sparc/kernel/irq_64.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/sparc/kernel/irq_64.c 2017-04-18 17:38:08.030642885 +0200 -@@ -854,6 +854,7 @@ - set_irq_regs(old_regs); + return pen_release != -1 ? -ENOSYS : 0; } +diff --git a/arch/arm/mm/fault.c b/arch/arm/mm/fault.c +index c095455d496e..276ac73f8650 100644 +--- a/arch/arm/mm/fault.c ++++ b/arch/arm/mm/fault.c +@@ -430,6 +430,9 @@ do_translation_fault(unsigned long addr, unsigned int fsr, + if (addr < TASK_SIZE) + return do_page_fault(addr, fsr, regs); -+#ifndef CONFIG_PREEMPT_RT_FULL - void do_softirq_own_stack(void) - { - void *orig_sp, *sp = softirq_stack[smp_processor_id()]; -@@ -868,6 +869,7 @@ - __asm__ __volatile__("mov %0, %%sp" - : : "r" (orig_sp)); - } -+#endif ++ if (interrupts_enabled(regs)) ++ local_irq_enable(); ++ + if (user_mode(regs)) + goto bad_area; - #ifdef CONFIG_HOTPLUG_CPU - void fixup_irqs(void) -diff -Nur linux-4.4.62.orig/arch/x86/crypto/aesni-intel_glue.c linux-4.4.62/arch/x86/crypto/aesni-intel_glue.c ---- linux-4.4.62.orig/arch/x86/crypto/aesni-intel_glue.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/crypto/aesni-intel_glue.c 2017-04-18 17:38:08.030642885 +0200 -@@ -383,14 +383,14 @@ - err = blkcipher_walk_virt(desc, &walk); - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; - -- kernel_fpu_begin(); - while ((nbytes = walk.nbytes)) { -+ kernel_fpu_begin(); - aesni_ecb_enc(ctx, walk.dst.virt.addr, walk.src.virt.addr, -- nbytes & AES_BLOCK_MASK); -+ nbytes & AES_BLOCK_MASK); -+ kernel_fpu_end(); - nbytes &= AES_BLOCK_SIZE - 1; - err = blkcipher_walk_done(desc, &walk, nbytes); - } -- kernel_fpu_end(); - - return err; +@@ -497,6 +500,9 @@ do_translation_fault(unsigned long addr, unsigned int fsr, + static int + do_sect_fault(unsigned long addr, unsigned int fsr, struct pt_regs *regs) + { ++ if (interrupts_enabled(regs)) ++ local_irq_enable(); ++ + do_bad_area(addr, fsr, regs); + return 0; } -@@ -407,14 +407,14 @@ - err = blkcipher_walk_virt(desc, &walk); - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; - -- kernel_fpu_begin(); - while ((nbytes = walk.nbytes)) { -+ kernel_fpu_begin(); - aesni_ecb_dec(ctx, walk.dst.virt.addr, walk.src.virt.addr, - nbytes & AES_BLOCK_MASK); -+ kernel_fpu_end(); - nbytes &= AES_BLOCK_SIZE - 1; - err = blkcipher_walk_done(desc, &walk, nbytes); - } -- kernel_fpu_end(); - - return err; +diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c +index d02f8187b1cc..542692dbd40a 100644 +--- a/arch/arm/mm/highmem.c ++++ b/arch/arm/mm/highmem.c +@@ -34,6 +34,11 @@ static inline pte_t get_fixmap_pte(unsigned long vaddr) + return *ptep; } -@@ -431,14 +431,14 @@ - err = blkcipher_walk_virt(desc, &walk); - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; -- kernel_fpu_begin(); - while ((nbytes = walk.nbytes)) { -+ kernel_fpu_begin(); - aesni_cbc_enc(ctx, walk.dst.virt.addr, walk.src.virt.addr, - nbytes & AES_BLOCK_MASK, walk.iv); -+ kernel_fpu_end(); - nbytes &= AES_BLOCK_SIZE - 1; - err = blkcipher_walk_done(desc, &walk, nbytes); - } -- kernel_fpu_end(); ++static unsigned int fixmap_idx(int type) ++{ ++ return FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); ++} ++ + void *kmap(struct page *page) + { + might_sleep(); +@@ -54,12 +59,13 @@ EXPORT_SYMBOL(kunmap); - return err; - } -@@ -455,14 +455,14 @@ - err = blkcipher_walk_virt(desc, &walk); - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + void *kmap_atomic(struct page *page) + { ++ pte_t pte = mk_pte(page, kmap_prot); + unsigned int idx; + unsigned long vaddr; + void *kmap; + int type; -- kernel_fpu_begin(); - while ((nbytes = walk.nbytes)) { -+ kernel_fpu_begin(); - aesni_cbc_dec(ctx, walk.dst.virt.addr, walk.src.virt.addr, - nbytes & AES_BLOCK_MASK, walk.iv); -+ kernel_fpu_end(); - nbytes &= AES_BLOCK_SIZE - 1; - err = blkcipher_walk_done(desc, &walk, nbytes); - } -- kernel_fpu_end(); +- preempt_disable(); ++ preempt_disable_nort(); + pagefault_disable(); + if (!PageHighMem(page)) + return page_address(page); +@@ -79,7 +85,7 @@ void *kmap_atomic(struct page *page) - return err; - } -@@ -514,18 +514,20 @@ - err = blkcipher_walk_virt_block(desc, &walk, AES_BLOCK_SIZE); - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + type = kmap_atomic_idx_push(); -- kernel_fpu_begin(); - while ((nbytes = walk.nbytes) >= AES_BLOCK_SIZE) { -+ kernel_fpu_begin(); - aesni_ctr_enc_tfm(ctx, walk.dst.virt.addr, walk.src.virt.addr, - nbytes & AES_BLOCK_MASK, walk.iv); -+ kernel_fpu_end(); - nbytes &= AES_BLOCK_SIZE - 1; - err = blkcipher_walk_done(desc, &walk, nbytes); - } - if (walk.nbytes) { -+ kernel_fpu_begin(); - ctr_crypt_final(ctx, &walk); -+ kernel_fpu_end(); - err = blkcipher_walk_done(desc, &walk, 0); - } -- kernel_fpu_end(); +- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); ++ idx = fixmap_idx(type); + vaddr = __fix_to_virt(idx); + #ifdef CONFIG_DEBUG_HIGHMEM + /* +@@ -93,7 +99,10 @@ void *kmap_atomic(struct page *page) + * in place, so the contained TLB flush ensures the TLB is updated + * with the new mapping. + */ +- set_fixmap_pte(idx, mk_pte(page, kmap_prot)); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = pte; ++#endif ++ set_fixmap_pte(idx, pte); - return err; + return (void *)vaddr; } -diff -Nur linux-4.4.62.orig/arch/x86/crypto/cast5_avx_glue.c linux-4.4.62/arch/x86/crypto/cast5_avx_glue.c ---- linux-4.4.62.orig/arch/x86/crypto/cast5_avx_glue.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/crypto/cast5_avx_glue.c 2017-04-18 17:38:08.030642885 +0200 -@@ -59,7 +59,7 @@ - static int ecb_crypt(struct blkcipher_desc *desc, struct blkcipher_walk *walk, - bool enc) - { -- bool fpu_enabled = false; -+ bool fpu_enabled; - struct cast5_ctx *ctx = crypto_blkcipher_ctx(desc->tfm); - const unsigned int bsize = CAST5_BLOCK_SIZE; - unsigned int nbytes; -@@ -75,7 +75,7 @@ - u8 *wsrc = walk->src.virt.addr; - u8 *wdst = walk->dst.virt.addr; - -- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); -+ fpu_enabled = cast5_fpu_begin(false, nbytes); +@@ -106,44 +115,75 @@ void __kunmap_atomic(void *kvaddr) - /* Process multi-block batch */ - if (nbytes >= bsize * CAST5_PARALLEL_BLOCKS) { -@@ -103,10 +103,9 @@ - } while (nbytes >= bsize); + if (kvaddr >= (void *)FIXADDR_START) { + type = kmap_atomic_idx(); +- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); ++ idx = fixmap_idx(type); - done: -+ cast5_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, walk, nbytes); + if (cache_is_vivt()) + __cpuc_flush_dcache_area((void *)vaddr, PAGE_SIZE); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = __pte(0); ++#endif + #ifdef CONFIG_DEBUG_HIGHMEM + BUG_ON(vaddr != __fix_to_virt(idx)); +- set_fixmap_pte(idx, __pte(0)); + #else + (void) idx; /* to kill a warning */ + #endif ++ set_fixmap_pte(idx, __pte(0)); + kmap_atomic_idx_pop(); + } else if (vaddr >= PKMAP_ADDR(0) && vaddr < PKMAP_ADDR(LAST_PKMAP)) { + /* this address was obtained through kmap_high_get() */ + kunmap_high(pte_page(pkmap_page_table[PKMAP_NR(vaddr)])); } -- -- cast5_fpu_end(fpu_enabled); - return err; + pagefault_enable(); +- preempt_enable(); ++ preempt_enable_nort(); } + EXPORT_SYMBOL(__kunmap_atomic); -@@ -227,7 +226,7 @@ - static int cbc_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst, - struct scatterlist *src, unsigned int nbytes) + void *kmap_atomic_pfn(unsigned long pfn) { -- bool fpu_enabled = false; -+ bool fpu_enabled; - struct blkcipher_walk walk; - int err; ++ pte_t pte = pfn_pte(pfn, kmap_prot); + unsigned long vaddr; + int idx, type; + struct page *page = pfn_to_page(pfn); -@@ -236,12 +235,11 @@ - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; +- preempt_disable(); ++ preempt_disable_nort(); + pagefault_disable(); + if (!PageHighMem(page)) + return page_address(page); - while ((nbytes = walk.nbytes)) { -- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); -+ fpu_enabled = cast5_fpu_begin(false, nbytes); - nbytes = __cbc_decrypt(desc, &walk); -+ cast5_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, &walk, nbytes); - } -- -- cast5_fpu_end(fpu_enabled); - return err; + type = kmap_atomic_idx_push(); +- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); ++ idx = fixmap_idx(type); + vaddr = __fix_to_virt(idx); + #ifdef CONFIG_DEBUG_HIGHMEM + BUG_ON(!pte_none(get_fixmap_pte(vaddr))); + #endif +- set_fixmap_pte(idx, pfn_pte(pfn, kmap_prot)); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = pte; ++#endif ++ set_fixmap_pte(idx, pte); + + return (void *)vaddr; + } ++#if defined CONFIG_PREEMPT_RT_FULL ++void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) ++{ ++ int i; ++ ++ /* ++ * Clear @prev's kmap_atomic mappings ++ */ ++ for (i = 0; i < prev_p->kmap_idx; i++) { ++ int idx = fixmap_idx(i); ++ ++ set_fixmap_pte(idx, __pte(0)); ++ } ++ /* ++ * Restore @next_p's kmap_atomic mappings ++ */ ++ for (i = 0; i < next_p->kmap_idx; i++) { ++ int idx = fixmap_idx(i); ++ ++ if (!pte_none(next_p->kmap_pte[i])) ++ set_fixmap_pte(idx, next_p->kmap_pte[i]); ++ } ++} ++#endif +diff --git a/arch/arm/plat-versatile/platsmp.c b/arch/arm/plat-versatile/platsmp.c +index 53feb90c840c..b4a8d54fc3f3 100644 +--- a/arch/arm/plat-versatile/platsmp.c ++++ b/arch/arm/plat-versatile/platsmp.c +@@ -30,7 +30,7 @@ static void write_pen_release(int val) + sync_cache_w(&pen_release); } -@@ -311,7 +309,7 @@ - static int ctr_crypt(struct blkcipher_desc *desc, struct scatterlist *dst, - struct scatterlist *src, unsigned int nbytes) - { -- bool fpu_enabled = false; -+ bool fpu_enabled; - struct blkcipher_walk walk; - int err; +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); -@@ -320,13 +318,12 @@ - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + void versatile_secondary_init(unsigned int cpu) + { +@@ -43,8 +43,8 @@ void versatile_secondary_init(unsigned int cpu) + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } - while ((nbytes = walk.nbytes) >= CAST5_BLOCK_SIZE) { -- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); -+ fpu_enabled = cast5_fpu_begin(false, nbytes); - nbytes = __ctr_crypt(desc, &walk); -+ cast5_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, &walk, nbytes); - } - -- cast5_fpu_end(fpu_enabled); -- - if (walk.nbytes) { - ctr_crypt_final(desc, &walk); - err = blkcipher_walk_done(desc, &walk, 0); -diff -Nur linux-4.4.62.orig/arch/x86/crypto/glue_helper.c linux-4.4.62/arch/x86/crypto/glue_helper.c ---- linux-4.4.62.orig/arch/x86/crypto/glue_helper.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/crypto/glue_helper.c 2017-04-18 17:38:08.030642885 +0200 -@@ -39,7 +39,7 @@ - void *ctx = crypto_blkcipher_ctx(desc->tfm); - const unsigned int bsize = 128 / 8; - unsigned int nbytes, i, func_bytes; -- bool fpu_enabled = false; -+ bool fpu_enabled; - int err; - - err = blkcipher_walk_virt(desc, walk); -@@ -49,7 +49,7 @@ - u8 *wdst = walk->dst.virt.addr; - - fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, -- desc, fpu_enabled, nbytes); -+ desc, false, nbytes); - - for (i = 0; i < gctx->num_funcs; i++) { - func_bytes = bsize * gctx->funcs[i].num_blocks; -@@ -71,10 +71,10 @@ - } + int versatile_boot_secondary(unsigned int cpu, struct task_struct *idle) +@@ -55,7 +55,7 @@ int versatile_boot_secondary(unsigned int cpu, struct task_struct *idle) + * Set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); - done: -+ glue_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, walk, nbytes); - } + /* + * This is really belt and braces; we hold unintended secondary +@@ -85,7 +85,7 @@ int versatile_boot_secondary(unsigned int cpu, struct task_struct *idle) + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); -- glue_fpu_end(fpu_enabled); - return err; + return pen_release != -1 ? -ENOSYS : 0; } +diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig +index 14cdc6dea493..9196cf82f7be 100644 +--- a/arch/arm64/Kconfig ++++ b/arch/arm64/Kconfig +@@ -76,6 +76,7 @@ config ARM64 + select HAVE_PERF_REGS + select HAVE_PERF_USER_STACK_DUMP + select HAVE_RCU_TABLE_FREE ++ select HAVE_PREEMPT_LAZY + select HAVE_SYSCALL_TRACEPOINTS + select IOMMU_DMA if IOMMU_SUPPORT + select IRQ_DOMAIN +@@ -582,7 +583,7 @@ config XEN_DOM0 -@@ -194,7 +194,7 @@ - struct scatterlist *src, unsigned int nbytes) - { - const unsigned int bsize = 128 / 8; -- bool fpu_enabled = false; -+ bool fpu_enabled; - struct blkcipher_walk walk; - int err; - -@@ -203,12 +203,12 @@ + config XEN + bool "Xen guest support on ARM64" +- depends on ARM64 && OF ++ depends on ARM64 && OF && !PREEMPT_RT_FULL + select SWIOTLB_XEN + help + Say Y if you want to run Linux in a Virtual Machine on Xen on ARM64. +diff --git a/arch/arm64/include/asm/thread_info.h b/arch/arm64/include/asm/thread_info.h +index 90c7ff233735..5f4e89fbc290 100644 +--- a/arch/arm64/include/asm/thread_info.h ++++ b/arch/arm64/include/asm/thread_info.h +@@ -49,6 +49,7 @@ struct thread_info { + mm_segment_t addr_limit; /* address limit */ + struct task_struct *task; /* main task structure */ + int preempt_count; /* 0 => preemptable, <0 => bug */ ++ int preempt_lazy_count; /* 0 => preemptable, <0 => bug */ + int cpu; /* cpu */ + }; - while ((nbytes = walk.nbytes)) { - fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, -- desc, fpu_enabled, nbytes); -+ desc, false, nbytes); - nbytes = __glue_cbc_decrypt_128bit(gctx, desc, &walk); -+ glue_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, &walk, nbytes); - } +@@ -103,6 +104,7 @@ static inline struct thread_info *current_thread_info(void) + #define TIF_NEED_RESCHED 1 + #define TIF_NOTIFY_RESUME 2 /* callback before returning to user */ + #define TIF_FOREIGN_FPSTATE 3 /* CPU's FP state is not current's */ ++#define TIF_NEED_RESCHED_LAZY 4 + #define TIF_NOHZ 7 + #define TIF_SYSCALL_TRACE 8 + #define TIF_SYSCALL_AUDIT 9 +@@ -118,6 +120,7 @@ static inline struct thread_info *current_thread_info(void) + #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) + #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) + #define _TIF_FOREIGN_FPSTATE (1 << TIF_FOREIGN_FPSTATE) ++#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) + #define _TIF_NOHZ (1 << TIF_NOHZ) + #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) + #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) +@@ -126,7 +129,8 @@ static inline struct thread_info *current_thread_info(void) + #define _TIF_32BIT (1 << TIF_32BIT) -- glue_fpu_end(fpu_enabled); - return err; - } - EXPORT_SYMBOL_GPL(glue_cbc_decrypt_128bit); -@@ -277,7 +277,7 @@ - struct scatterlist *src, unsigned int nbytes) - { - const unsigned int bsize = 128 / 8; -- bool fpu_enabled = false; -+ bool fpu_enabled; - struct blkcipher_walk walk; - int err; + #define _TIF_WORK_MASK (_TIF_NEED_RESCHED | _TIF_SIGPENDING | \ +- _TIF_NOTIFY_RESUME | _TIF_FOREIGN_FPSTATE) ++ _TIF_NOTIFY_RESUME | _TIF_FOREIGN_FPSTATE | \ ++ _TIF_NEED_RESCHED_LAZY) -@@ -286,13 +286,12 @@ + #define _TIF_SYSCALL_WORK (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | \ + _TIF_SYSCALL_TRACEPOINT | _TIF_SECCOMP | \ +diff --git a/arch/arm64/kernel/asm-offsets.c b/arch/arm64/kernel/asm-offsets.c +index 087cf9a65359..d74475928399 100644 +--- a/arch/arm64/kernel/asm-offsets.c ++++ b/arch/arm64/kernel/asm-offsets.c +@@ -35,6 +35,7 @@ int main(void) + BLANK(); + DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); + DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); ++ DEFINE(TI_PREEMPT_LAZY, offsetof(struct thread_info, preempt_lazy_count)); + DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit)); + DEFINE(TI_TASK, offsetof(struct thread_info, task)); + DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); +diff --git a/arch/arm64/kernel/entry.S b/arch/arm64/kernel/entry.S +index bd14849beb73..cf92d4ee51db 100644 +--- a/arch/arm64/kernel/entry.S ++++ b/arch/arm64/kernel/entry.S +@@ -376,11 +376,16 @@ el1_irq: + #ifdef CONFIG_PREEMPT + get_thread_info tsk + ldr w24, [tsk, #TI_PREEMPT] // get preempt count +- cbnz w24, 1f // preempt count != 0 ++ cbnz w24, 2f // preempt count != 0 + ldr x0, [tsk, #TI_FLAGS] // get flags +- tbz x0, #TIF_NEED_RESCHED, 1f // needs rescheduling? +- bl el1_preempt ++ tbnz x0, #TIF_NEED_RESCHED, 1f // needs rescheduling? ++ ++ ldr w24, [tsk, #TI_PREEMPT_LAZY] // get preempt lazy count ++ cbnz w24, 2f // preempt lazy count != 0 ++ tbz x0, #TIF_NEED_RESCHED_LAZY, 2f // needs rescheduling? + 1: ++ bl el1_preempt ++2: + #endif + #ifdef CONFIG_TRACE_IRQFLAGS + bl trace_hardirqs_on +@@ -394,6 +399,7 @@ el1_preempt: + 1: bl preempt_schedule_irq // irq en/disable is done inside + ldr x0, [tsk, #TI_FLAGS] // get new tasks TI_FLAGS + tbnz x0, #TIF_NEED_RESCHED, 1b // needs rescheduling? ++ tbnz x0, #TIF_NEED_RESCHED_LAZY, 1b // needs rescheduling? + ret x24 + #endif - while ((nbytes = walk.nbytes) >= bsize) { - fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, -- desc, fpu_enabled, nbytes); -+ desc, false, nbytes); - nbytes = __glue_ctr_crypt_128bit(gctx, desc, &walk); -+ glue_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, &walk, nbytes); - } +@@ -638,6 +644,7 @@ ret_fast_syscall_trace: + */ + work_pending: + tbnz x1, #TIF_NEED_RESCHED, work_resched ++ tbnz x1, #TIF_NEED_RESCHED_LAZY, work_resched + /* TIF_SIGPENDING, TIF_NOTIFY_RESUME or TIF_FOREIGN_FPSTATE case */ + ldr x2, [sp, #S_PSTATE] + mov x0, sp // 'regs' +diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig +index 8b0424abc84c..5422d4c0bbdf 100644 +--- a/arch/mips/Kconfig ++++ b/arch/mips/Kconfig +@@ -2411,7 +2411,7 @@ config CPU_R4400_WORKAROUNDS + # + config HIGHMEM + bool "High Memory Support" +- depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM && !CPU_MIPS32_3_5_EVA ++ depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM && !CPU_MIPS32_3_5_EVA && !PREEMPT_RT_FULL -- glue_fpu_end(fpu_enabled); -- - if (walk.nbytes) { - glue_ctr_crypt_final_128bit( - gctx->funcs[gctx->num_funcs - 1].fn_u.ctr, desc, &walk); -@@ -347,7 +346,7 @@ - void *tweak_ctx, void *crypt_ctx) - { - const unsigned int bsize = 128 / 8; -- bool fpu_enabled = false; -+ bool fpu_enabled; - struct blkcipher_walk walk; - int err; + config CPU_SUPPORTS_HIGHMEM + bool +diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c +index a017b23ee4aa..8d4d9270140f 100644 +--- a/arch/mips/kvm/mips.c ++++ b/arch/mips/kvm/mips.c +@@ -454,8 +454,8 @@ int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, -@@ -360,21 +359,21 @@ + dvcpu->arch.wait = 0; - /* set minimum length to bsize, for tweak_fn */ - fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, -- desc, fpu_enabled, -+ desc, false, - nbytes < bsize ? bsize : nbytes); -- - /* calculate first value of T */ - tweak_fn(tweak_ctx, walk.iv, walk.iv); -+ glue_fpu_end(fpu_enabled); +- if (waitqueue_active(&dvcpu->wq)) +- wake_up_interruptible(&dvcpu->wq); ++ if (swait_active(&dvcpu->wq)) ++ swake_up(&dvcpu->wq); - while (nbytes) { -+ fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, -+ desc, false, nbytes); - nbytes = __glue_xts_crypt_128bit(gctx, crypt_ctx, desc, &walk); + return 0; + } +@@ -1183,8 +1183,8 @@ static void kvm_mips_comparecount_func(unsigned long data) + kvm_mips_callbacks->queue_timer_int(vcpu); -+ glue_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, &walk, nbytes); - nbytes = walk.nbytes; - } -- -- glue_fpu_end(fpu_enabled); -- - return err; + vcpu->arch.wait = 0; +- if (waitqueue_active(&vcpu->wq)) +- wake_up_interruptible(&vcpu->wq); ++ if (swait_active(&vcpu->wq)) ++ swake_up(&vcpu->wq); } - EXPORT_SYMBOL_GPL(glue_xts_crypt_128bit); -diff -Nur linux-4.4.62.orig/arch/x86/entry/common.c linux-4.4.62/arch/x86/entry/common.c ---- linux-4.4.62.orig/arch/x86/entry/common.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/entry/common.c 2017-04-18 17:38:08.030642885 +0200 -@@ -220,7 +220,7 @@ - #define EXIT_TO_USERMODE_LOOP_FLAGS \ - (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME | _TIF_UPROBE | \ -- _TIF_NEED_RESCHED | _TIF_USER_RETURN_NOTIFY) -+ _TIF_NEED_RESCHED_MASK | _TIF_USER_RETURN_NOTIFY) + /* low level hrtimer wake routine */ +diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig +index db49e0d796b1..1d2be228661c 100644 +--- a/arch/powerpc/Kconfig ++++ b/arch/powerpc/Kconfig +@@ -60,10 +60,11 @@ config LOCKDEP_SUPPORT - static void exit_to_usermode_loop(struct pt_regs *regs, u32 cached_flags) - { -@@ -236,9 +236,16 @@ - /* We have work to do. */ - local_irq_enable(); + config RWSEM_GENERIC_SPINLOCK + bool ++ default y if PREEMPT_RT_FULL -- if (cached_flags & _TIF_NEED_RESCHED) -+ if (cached_flags & _TIF_NEED_RESCHED_MASK) - schedule(); + config RWSEM_XCHGADD_ALGORITHM + bool +- default y ++ default y if !PREEMPT_RT_FULL -+#ifdef ARCH_RT_DELAYS_SIGNAL_SEND -+ if (unlikely(current->forced_info.si_signo)) { -+ struct task_struct *t = current; -+ force_sig_info(t->forced_info.si_signo, &t->forced_info, t); -+ t->forced_info.si_signo = 0; -+ } -+#endif - if (cached_flags & _TIF_UPROBE) - uprobe_notify_resume(regs); + config GENERIC_LOCKBREAK + bool +@@ -141,6 +142,7 @@ config PPC + select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST + select GENERIC_STRNCPY_FROM_USER + select GENERIC_STRNLEN_USER ++ select HAVE_PREEMPT_LAZY + select HAVE_MOD_ARCH_SPECIFIC + select MODULES_USE_ELF_RELA + select CLONE_BACKWARDS +@@ -319,7 +321,7 @@ menu "Kernel options" -diff -Nur linux-4.4.62.orig/arch/x86/entry/entry_32.S linux-4.4.62/arch/x86/entry/entry_32.S ---- linux-4.4.62.orig/arch/x86/entry/entry_32.S 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/entry/entry_32.S 2017-04-18 17:38:08.030642885 +0200 -@@ -278,8 +278,24 @@ - ENTRY(resume_kernel) - DISABLE_INTERRUPTS(CLBR_ANY) - need_resched: -+ # preempt count == 0 + NEED_RS set? - cmpl $0, PER_CPU_VAR(__preempt_count) -+#ifndef CONFIG_PREEMPT_LAZY - jnz restore_all -+#else -+ jz test_int_off -+ -+ # atleast preempt count == 0 ? -+ cmpl $_PREEMPT_ENABLED,PER_CPU_VAR(__preempt_count) -+ jne restore_all -+ -+ cmpl $0,TI_preempt_lazy_count(%ebp) # non-zero preempt_lazy_count ? -+ jnz restore_all -+ -+ testl $_TIF_NEED_RESCHED_LAZY, TI_flags(%ebp) -+ jz restore_all -+test_int_off: -+#endif - testl $X86_EFLAGS_IF, PT_EFLAGS(%esp) # interrupts off (exception path) ? - jz restore_all - call preempt_schedule_irq -diff -Nur linux-4.4.62.orig/arch/x86/entry/entry_64.S linux-4.4.62/arch/x86/entry/entry_64.S ---- linux-4.4.62.orig/arch/x86/entry/entry_64.S 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/entry/entry_64.S 2017-04-18 17:38:08.034643040 +0200 -@@ -579,7 +579,23 @@ - bt $9, EFLAGS(%rsp) /* were interrupts off? */ - jnc 1f - 0: cmpl $0, PER_CPU_VAR(__preempt_count) -+#ifndef CONFIG_PREEMPT_LAZY - jnz 1f -+#else -+ jz do_preempt_schedule_irq -+ -+ # atleast preempt count == 0 ? -+ cmpl $_PREEMPT_ENABLED,PER_CPU_VAR(__preempt_count) -+ jnz 1f + config HIGHMEM + bool "High memory support" +- depends on PPC32 ++ depends on PPC32 && !PREEMPT_RT_FULL + + source kernel/Kconfig.hz + source kernel/Kconfig.preempt +diff --git a/arch/powerpc/include/asm/kvm_host.h b/arch/powerpc/include/asm/kvm_host.h +index a92d95aee42d..20376580583f 100644 +--- a/arch/powerpc/include/asm/kvm_host.h ++++ b/arch/powerpc/include/asm/kvm_host.h +@@ -286,7 +286,7 @@ struct kvmppc_vcore { + struct list_head runnable_threads; + struct list_head preempt_list; + spinlock_t lock; +- wait_queue_head_t wq; ++ struct swait_queue_head wq; + spinlock_t stoltb_lock; /* protects stolen_tb and preempt_tb */ + u64 stolen_tb; + u64 preempt_tb; +@@ -627,7 +627,7 @@ struct kvm_vcpu_arch { + u8 prodded; + u32 last_inst; + +- wait_queue_head_t *wqp; ++ struct swait_queue_head *wqp; + struct kvmppc_vcore *vcore; + int ret; + int trap; +diff --git a/arch/powerpc/include/asm/thread_info.h b/arch/powerpc/include/asm/thread_info.h +index 7efee4a3240b..40e6fa1b85b2 100644 +--- a/arch/powerpc/include/asm/thread_info.h ++++ b/arch/powerpc/include/asm/thread_info.h +@@ -42,6 +42,8 @@ struct thread_info { + int cpu; /* cpu we're on */ + int preempt_count; /* 0 => preemptable, + <0 => BUG */ ++ int preempt_lazy_count; /* 0 => preemptable, ++ <0 => BUG */ + unsigned long local_flags; /* private flags for thread */ + + /* low level flags - has atomic operations done on it */ +@@ -82,8 +84,7 @@ static inline struct thread_info *current_thread_info(void) + #define TIF_SYSCALL_TRACE 0 /* syscall trace active */ + #define TIF_SIGPENDING 1 /* signal pending */ + #define TIF_NEED_RESCHED 2 /* rescheduling necessary */ +-#define TIF_POLLING_NRFLAG 3 /* true if poll_idle() is polling +- TIF_NEED_RESCHED */ ++#define TIF_NEED_RESCHED_LAZY 3 /* lazy rescheduling necessary */ + #define TIF_32BIT 4 /* 32 bit binary */ + #define TIF_RESTORE_TM 5 /* need to restore TM FP/VEC/VSX */ + #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ +@@ -101,6 +102,8 @@ static inline struct thread_info *current_thread_info(void) + #if defined(CONFIG_PPC64) + #define TIF_ELF2ABI 18 /* function descriptors must die! */ + #endif ++#define TIF_POLLING_NRFLAG 19 /* true if poll_idle() is polling ++ TIF_NEED_RESCHED */ + + /* as above, but as bit values */ + #define _TIF_SYSCALL_TRACE (1<flags) + set_bits(irqtp->flags, &curtp->flags); } ++#endif -+static __always_inline bool __preempt_count_dec_and_test(void) -+{ -+ if (____preempt_count_dec_and_test()) -+ return true; -+#ifdef CONFIG_PREEMPT_LAZY -+ if (current_thread_info()->preempt_lazy_count) -+ return false; -+ return test_thread_flag(TIF_NEED_RESCHED_LAZY); -+#else -+ return false; + irq_hw_number_t virq_to_hw(unsigned int virq) + { +diff --git a/arch/powerpc/kernel/misc_32.S b/arch/powerpc/kernel/misc_32.S +index df4efa304b2c..9cb0c2f6e7ac 100644 +--- a/arch/powerpc/kernel/misc_32.S ++++ b/arch/powerpc/kernel/misc_32.S +@@ -40,6 +40,7 @@ + * We store the saved ksp_limit in the unused part + * of the STACK_FRAME_OVERHEAD + */ ++#ifndef CONFIG_PREEMPT_RT_FULL + _GLOBAL(call_do_softirq) + mflr r0 + stw r0,4(r1) +@@ -56,6 +57,7 @@ _GLOBAL(call_do_softirq) + stw r10,THREAD+KSP_LIMIT(r2) + mtlr r0 + blr +#endif -+} -+ + /* - * Returns true when we need to resched and can (barring IRQ state). - */ - static __always_inline bool should_resched(int preempt_offset) - { -+#ifdef CONFIG_PREEMPT_LAZY -+ u32 tmp; -+ -+ tmp = raw_cpu_read_4(__preempt_count); -+ if (tmp == preempt_offset) -+ return true; -+ -+ /* preempt count == 0 ? */ -+ tmp &= ~PREEMPT_NEED_RESCHED; -+ if (tmp) -+ return false; -+ if (current_thread_info()->preempt_lazy_count) -+ return false; -+ return test_thread_flag(TIF_NEED_RESCHED_LAZY); -+#else - return unlikely(raw_cpu_read_4(__preempt_count) == preempt_offset); -+#endif - } + * void call_do_irq(struct pt_regs *regs, struct thread_info *irqtp); +diff --git a/arch/powerpc/kernel/misc_64.S b/arch/powerpc/kernel/misc_64.S +index db475d41b57a..96b7ef80e05d 100644 +--- a/arch/powerpc/kernel/misc_64.S ++++ b/arch/powerpc/kernel/misc_64.S +@@ -30,6 +30,7 @@ - #ifdef CONFIG_PREEMPT -diff -Nur linux-4.4.62.orig/arch/x86/include/asm/signal.h linux-4.4.62/arch/x86/include/asm/signal.h ---- linux-4.4.62.orig/arch/x86/include/asm/signal.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/include/asm/signal.h 2017-04-18 17:38:08.034643040 +0200 -@@ -23,6 +23,19 @@ - unsigned long sig[_NSIG_WORDS]; - } sigset_t; + .text -+/* -+ * Because some traps use the IST stack, we must keep preemption -+ * disabled while calling do_trap(), but do_trap() may call -+ * force_sig_info() which will grab the signal spin_locks for the -+ * task, which in PREEMPT_RT_FULL are mutexes. By defining -+ * ARCH_RT_DELAYS_SIGNAL_SEND the force_sig_info() will set -+ * TIF_NOTIFY_RESUME and set up the signal to be sent on exit of the -+ * trap. -+ */ -+#if defined(CONFIG_PREEMPT_RT_FULL) -+#define ARCH_RT_DELAYS_SIGNAL_SEND ++#ifndef CONFIG_PREEMPT_RT_FULL + _GLOBAL(call_do_softirq) + mflr r0 + std r0,16(r1) +@@ -40,6 +41,7 @@ _GLOBAL(call_do_softirq) + ld r0,16(r1) + mtlr r0 + blr +#endif -+ - #ifndef CONFIG_COMPAT - typedef sigset_t compat_sigset_t; - #endif -diff -Nur linux-4.4.62.orig/arch/x86/include/asm/stackprotector.h linux-4.4.62/arch/x86/include/asm/stackprotector.h ---- linux-4.4.62.orig/arch/x86/include/asm/stackprotector.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/include/asm/stackprotector.h 2017-04-18 17:38:08.034643040 +0200 -@@ -59,7 +59,7 @@ - */ - static __always_inline void boot_init_stack_canary(void) + + _GLOBAL(call_do_irq) + mflr r0 +diff --git a/arch/powerpc/kvm/Kconfig b/arch/powerpc/kvm/Kconfig +index c2024ac9d4e8..2303788da7e1 100644 +--- a/arch/powerpc/kvm/Kconfig ++++ b/arch/powerpc/kvm/Kconfig +@@ -172,6 +172,7 @@ config KVM_E500MC + config KVM_MPIC + bool "KVM in-kernel MPIC emulation" + depends on KVM && E500 ++ depends on !PREEMPT_RT_FULL + select HAVE_KVM_IRQCHIP + select HAVE_KVM_IRQFD + select HAVE_KVM_IRQ_ROUTING +diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c +index 3c3a367b6e59..2ac7a8b108ac 100644 +--- a/arch/powerpc/kvm/book3s_hv.c ++++ b/arch/powerpc/kvm/book3s_hv.c +@@ -114,11 +114,11 @@ static bool kvmppc_ipi_thread(int cpu) + static void kvmppc_fast_vcpu_kick_hv(struct kvm_vcpu *vcpu) { -- u64 canary; -+ u64 uninitialized_var(canary); - u64 tsc; + int cpu; +- wait_queue_head_t *wqp; ++ struct swait_queue_head *wqp; - #ifdef CONFIG_X86_64 -@@ -70,8 +70,15 @@ - * of randomness. The TSC only matters for very early init, - * there it already has some randomness on most systems. Later - * on during the bootup the random pool has true entropy too. -+ * -+ * For preempt-rt we need to weaken the randomness a bit, as -+ * we can't call into the random generator from atomic context -+ * due to locking constraints. We just leave canary -+ * uninitialized and use the TSC based randomness on top of it. - */ -+#ifndef CONFIG_PREEMPT_RT_FULL - get_random_bytes(&canary, sizeof(canary)); -+#endif - tsc = rdtsc(); - canary += tsc + (tsc << 32UL); + wqp = kvm_arch_vcpu_wq(vcpu); +- if (waitqueue_active(wqp)) { +- wake_up_interruptible(wqp); ++ if (swait_active(wqp)) { ++ swake_up(wqp); + ++vcpu->stat.halt_wakeup; + } -diff -Nur linux-4.4.62.orig/arch/x86/include/asm/thread_info.h linux-4.4.62/arch/x86/include/asm/thread_info.h ---- linux-4.4.62.orig/arch/x86/include/asm/thread_info.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/include/asm/thread_info.h 2017-04-18 17:38:08.034643040 +0200 -@@ -58,6 +58,8 @@ - __u32 status; /* thread synchronous flags */ - __u32 cpu; /* current CPU */ - mm_segment_t addr_limit; -+ int preempt_lazy_count; /* 0 => lazy preemptable -+ <0 => BUG */ - unsigned int sig_on_uaccess_error:1; - unsigned int uaccess_err:1; /* uaccess failed */ - }; -@@ -95,6 +97,7 @@ - #define TIF_SYSCALL_EMU 6 /* syscall emulation active */ - #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ - #define TIF_SECCOMP 8 /* secure computing */ -+#define TIF_NEED_RESCHED_LAZY 9 /* lazy rescheduling necessary */ - #define TIF_USER_RETURN_NOTIFY 11 /* notify kernel of userspace return */ - #define TIF_UPROBE 12 /* breakpointed or singlestepping */ - #define TIF_NOTSC 16 /* TSC is not accessible in userland */ -@@ -119,6 +122,7 @@ - #define _TIF_SYSCALL_EMU (1 << TIF_SYSCALL_EMU) - #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) - #define _TIF_SECCOMP (1 << TIF_SECCOMP) -+#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) - #define _TIF_USER_RETURN_NOTIFY (1 << TIF_USER_RETURN_NOTIFY) - #define _TIF_UPROBE (1 << TIF_UPROBE) - #define _TIF_NOTSC (1 << TIF_NOTSC) -@@ -152,6 +156,8 @@ - #define _TIF_WORK_CTXSW_PREV (_TIF_WORK_CTXSW|_TIF_USER_RETURN_NOTIFY) - #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW) +@@ -707,8 +707,8 @@ int kvmppc_pseries_do_hcall(struct kvm_vcpu *vcpu) + tvcpu->arch.prodded = 1; + smp_mb(); + if (vcpu->arch.ceded) { +- if (waitqueue_active(&vcpu->wq)) { +- wake_up_interruptible(&vcpu->wq); ++ if (swait_active(&vcpu->wq)) { ++ swake_up(&vcpu->wq); + vcpu->stat.halt_wakeup++; + } + } +@@ -1453,7 +1453,7 @@ static struct kvmppc_vcore *kvmppc_vcore_create(struct kvm *kvm, int core) + INIT_LIST_HEAD(&vcore->runnable_threads); + spin_lock_init(&vcore->lock); + spin_lock_init(&vcore->stoltb_lock); +- init_waitqueue_head(&vcore->wq); ++ init_swait_queue_head(&vcore->wq); + vcore->preempt_tb = TB_NIL; + vcore->lpcr = kvm->arch.lpcr; + vcore->first_vcpuid = core * threads_per_subcore; +@@ -2525,10 +2525,9 @@ static void kvmppc_vcore_blocked(struct kvmppc_vcore *vc) + { + struct kvm_vcpu *vcpu; + int do_sleep = 1; ++ DECLARE_SWAITQUEUE(wait); -+#define _TIF_NEED_RESCHED_MASK (_TIF_NEED_RESCHED | _TIF_NEED_RESCHED_LAZY) -+ - #define STACK_WARN (THREAD_SIZE/8) +- DEFINE_WAIT(wait); +- +- prepare_to_wait(&vc->wq, &wait, TASK_INTERRUPTIBLE); ++ prepare_to_swait(&vc->wq, &wait, TASK_INTERRUPTIBLE); - /* -diff -Nur linux-4.4.62.orig/arch/x86/include/asm/uv/uv_bau.h linux-4.4.62/arch/x86/include/asm/uv/uv_bau.h ---- linux-4.4.62.orig/arch/x86/include/asm/uv/uv_bau.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/include/asm/uv/uv_bau.h 2017-04-18 17:38:08.034643040 +0200 -@@ -615,9 +615,9 @@ - cycles_t send_message; - cycles_t period_end; - cycles_t period_time; -- spinlock_t uvhub_lock; -- spinlock_t queue_lock; -- spinlock_t disable_lock; -+ raw_spinlock_t uvhub_lock; -+ raw_spinlock_t queue_lock; -+ raw_spinlock_t disable_lock; - /* tunables */ - int max_concurr; - int max_concurr_const; -@@ -776,15 +776,15 @@ - * to be lowered below the current 'v'. atomic_add_unless can only stop - * on equal. - */ --static inline int atomic_inc_unless_ge(spinlock_t *lock, atomic_t *v, int u) -+static inline int atomic_inc_unless_ge(raw_spinlock_t *lock, atomic_t *v, int u) + /* + * Check one last time for pending exceptions and ceded state after +@@ -2542,7 +2541,7 @@ static void kvmppc_vcore_blocked(struct kvmppc_vcore *vc) + } + + if (!do_sleep) { +- finish_wait(&vc->wq, &wait); ++ finish_swait(&vc->wq, &wait); + return; + } + +@@ -2550,7 +2549,7 @@ static void kvmppc_vcore_blocked(struct kvmppc_vcore *vc) + trace_kvmppc_vcore_blocked(vc, 0); + spin_unlock(&vc->lock); + schedule(); +- finish_wait(&vc->wq, &wait); ++ finish_swait(&vc->wq, &wait); + spin_lock(&vc->lock); + vc->vcore_state = VCORE_INACTIVE; + trace_kvmppc_vcore_blocked(vc, 1); +@@ -2606,7 +2605,7 @@ static int kvmppc_run_vcpu(struct kvm_run *kvm_run, struct kvm_vcpu *vcpu) + kvmppc_start_thread(vcpu, vc); + trace_kvm_guest_enter(vcpu); + } else if (vc->vcore_state == VCORE_SLEEPING) { +- wake_up(&vc->wq); ++ swake_up(&vc->wq); + } + + } +diff --git a/arch/powerpc/platforms/ps3/device-init.c b/arch/powerpc/platforms/ps3/device-init.c +index 3f175e8aedb4..c4c02f91904c 100644 +--- a/arch/powerpc/platforms/ps3/device-init.c ++++ b/arch/powerpc/platforms/ps3/device-init.c +@@ -752,7 +752,7 @@ static int ps3_notification_read_write(struct ps3_notification_device *dev, + } + pr_debug("%s:%u: notification %s issued\n", __func__, __LINE__, op); + +- res = wait_event_interruptible(dev->done.wait, ++ res = swait_event_interruptible(dev->done.wait, + dev->done.done || kthread_should_stop()); + if (kthread_should_stop()) + res = -EINTR; +diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h +index e9a983f40a24..bbdc539fb3c6 100644 +--- a/arch/s390/include/asm/kvm_host.h ++++ b/arch/s390/include/asm/kvm_host.h +@@ -427,7 +427,7 @@ struct kvm_s390_irq_payload { + struct kvm_s390_local_interrupt { + spinlock_t lock; + struct kvm_s390_float_interrupt *float_int; +- wait_queue_head_t *wq; ++ struct swait_queue_head *wq; + atomic_t *cpuflags; + DECLARE_BITMAP(sigp_emerg_pending, KVM_MAX_VCPUS); + struct kvm_s390_irq_payload irq; +diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c +index 6a75352f453c..cc862c486002 100644 +--- a/arch/s390/kvm/interrupt.c ++++ b/arch/s390/kvm/interrupt.c +@@ -868,13 +868,13 @@ no_timer: + + void kvm_s390_vcpu_wakeup(struct kvm_vcpu *vcpu) { -- spin_lock(lock); -+ raw_spin_lock(lock); - if (atomic_read(v) >= u) { -- spin_unlock(lock); -+ raw_spin_unlock(lock); - return 0; +- if (waitqueue_active(&vcpu->wq)) { ++ if (swait_active(&vcpu->wq)) { + /* + * The vcpu gave up the cpu voluntarily, mark it as a good + * yield-candidate. + */ + vcpu->preempted = true; +- wake_up_interruptible(&vcpu->wq); ++ swake_up(&vcpu->wq); + vcpu->stat.halt_wakeup++; } - atomic_inc(v); -- spin_unlock(lock); -+ raw_spin_unlock(lock); - return 1; + } +diff --git a/arch/sh/kernel/irq.c b/arch/sh/kernel/irq.c +index 6c0378c0b8b5..abd58b4dff97 100644 +--- a/arch/sh/kernel/irq.c ++++ b/arch/sh/kernel/irq.c +@@ -147,6 +147,7 @@ void irq_ctx_exit(int cpu) + hardirq_ctx[cpu] = NULL; } -diff -Nur linux-4.4.62.orig/arch/x86/include/asm/uv/uv_hub.h linux-4.4.62/arch/x86/include/asm/uv/uv_hub.h ---- linux-4.4.62.orig/arch/x86/include/asm/uv/uv_hub.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/include/asm/uv/uv_hub.h 2017-04-18 17:38:08.034643040 +0200 -@@ -492,7 +492,7 @@ - unsigned short nr_online_cpus; - unsigned short pnode; - short memory_nid; -- spinlock_t nmi_lock; /* obsolete, see uv_hub_nmi */ -+ raw_spinlock_t nmi_lock; /* obsolete, see uv_hub_nmi */ - unsigned long nmi_count; /* obsolete, see uv_hub_nmi */ - }; - extern struct uv_blade_info *uv_blade_info; -diff -Nur linux-4.4.62.orig/arch/x86/Kconfig linux-4.4.62/arch/x86/Kconfig ---- linux-4.4.62.orig/arch/x86/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/Kconfig 2017-04-18 17:38:08.030642885 +0200 -@@ -17,6 +17,7 @@ - ### Arch settings - config X86 - def_bool y -+ select HAVE_PREEMPT_LAZY - select ACPI_LEGACY_TABLES_LOOKUP if ACPI - select ACPI_SYSTEM_POWER_STATES_SUPPORT if ACPI - select ANON_INODES -@@ -212,8 +213,11 @@ - def_bool y - depends on ISA_DMA_API ++#ifndef CONFIG_PREEMPT_RT_FULL + void do_softirq_own_stack(void) + { + struct thread_info *curctx; +@@ -174,6 +175,7 @@ void do_softirq_own_stack(void) + "r5", "r6", "r7", "r8", "r9", "r15", "t", "pr" + ); + } ++#endif + #else + static inline void handle_one_irq(unsigned int irq) + { +diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig +index 56442d2d7bbc..8c9598f534c9 100644 +--- a/arch/sparc/Kconfig ++++ b/arch/sparc/Kconfig +@@ -189,12 +189,10 @@ config NR_CPUS + source kernel/Kconfig.hz -+config RWSEM_GENERIC_SPINLOCK + config RWSEM_GENERIC_SPINLOCK +- bool +- default y if SPARC32 + def_bool PREEMPT_RT_FULL -+ - config RWSEM_XCHGADD_ALGORITHM -- def_bool y -+ def_bool !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL + + config RWSEM_XCHGADD_ALGORITHM +- bool +- default y if SPARC64 ++ def_bool !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL + + config GENERIC_HWEIGHT + bool +diff --git a/arch/sparc/kernel/irq_64.c b/arch/sparc/kernel/irq_64.c +index e22416ce56ea..d359de71153a 100644 +--- a/arch/sparc/kernel/irq_64.c ++++ b/arch/sparc/kernel/irq_64.c +@@ -854,6 +854,7 @@ void __irq_entry handler_irq(int pil, struct pt_regs *regs) + set_irq_regs(old_regs); + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + void do_softirq_own_stack(void) + { + void *orig_sp, *sp = softirq_stack[smp_processor_id()]; +@@ -868,6 +869,7 @@ void do_softirq_own_stack(void) + __asm__ __volatile__("mov %0, %%sp" + : : "r" (orig_sp)); + } ++#endif + + #ifdef CONFIG_HOTPLUG_CPU + void fixup_irqs(void) +diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig +index 436639a31624..6ee1dd0deadc 100644 +--- a/arch/x86/Kconfig ++++ b/arch/x86/Kconfig +@@ -17,6 +17,7 @@ config X86_64 + ### Arch settings + config X86 + def_bool y ++ select HAVE_PREEMPT_LAZY + select ACPI_LEGACY_TABLES_LOOKUP if ACPI + select ACPI_SYSTEM_POWER_STATES_SUPPORT if ACPI + select ANON_INODES +@@ -212,8 +213,11 @@ config ARCH_MAY_HAVE_PC_FDC + def_bool y + depends on ISA_DMA_API + ++config RWSEM_GENERIC_SPINLOCK ++ def_bool PREEMPT_RT_FULL ++ + config RWSEM_XCHGADD_ALGORITHM +- def_bool y ++ def_bool !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL config GENERIC_CALIBRATE_DELAY def_bool y -@@ -848,7 +852,7 @@ +@@ -848,7 +852,7 @@ config IOMMU_HELPER config MAXSMP bool "Enable Maximum number of SMP Processors and NUMA Nodes" depends on X86_64 && SMP && DEBUG_KERNEL @@ -2406,1840 +2198,2167 @@ diff -Nur linux-4.4.62.orig/arch/x86/Kconfig linux-4.4.62/arch/x86/Kconfig ---help--- Enable maximum number of CPUS and NUMA Nodes for this architecture. If unsure, say N. -diff -Nur linux-4.4.62.orig/arch/x86/kernel/acpi/boot.c linux-4.4.62/arch/x86/kernel/acpi/boot.c ---- linux-4.4.62.orig/arch/x86/kernel/acpi/boot.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kernel/acpi/boot.c 2017-04-18 17:38:08.034643040 +0200 -@@ -87,7 +87,9 @@ - * ->ioapic_mutex - * ->ioapic_lock - */ -+#ifdef CONFIG_X86_IO_APIC - static DEFINE_MUTEX(acpi_ioapic_lock); -+#endif +diff --git a/arch/x86/crypto/aesni-intel_glue.c b/arch/x86/crypto/aesni-intel_glue.c +index 3633ad6145c5..c6d5458ee7f9 100644 +--- a/arch/x86/crypto/aesni-intel_glue.c ++++ b/arch/x86/crypto/aesni-intel_glue.c +@@ -383,14 +383,14 @@ static int ecb_encrypt(struct blkcipher_desc *desc, + err = blkcipher_walk_virt(desc, &walk); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; - /* -------------------------------------------------------------------------- - Boot-time Configuration -diff -Nur linux-4.4.62.orig/arch/x86/kernel/apic/io_apic.c linux-4.4.62/arch/x86/kernel/apic/io_apic.c ---- linux-4.4.62.orig/arch/x86/kernel/apic/io_apic.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kernel/apic/io_apic.c 2017-04-18 17:38:08.034643040 +0200 -@@ -1711,7 +1711,8 @@ - static inline bool ioapic_irqd_mask(struct irq_data *data) - { - /* If we are moving the irq we need to mask it */ -- if (unlikely(irqd_is_setaffinity_pending(data))) { -+ if (unlikely(irqd_is_setaffinity_pending(data) && -+ !irqd_irq_inprogress(data))) { - mask_ioapic_irq(data); - return true; +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes)) { ++ kernel_fpu_begin(); + aesni_ecb_enc(ctx, walk.dst.virt.addr, walk.src.virt.addr, +- nbytes & AES_BLOCK_MASK); ++ nbytes & AES_BLOCK_MASK); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); } -diff -Nur linux-4.4.62.orig/arch/x86/kernel/apic/x2apic_uv_x.c linux-4.4.62/arch/x86/kernel/apic/x2apic_uv_x.c ---- linux-4.4.62.orig/arch/x86/kernel/apic/x2apic_uv_x.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kernel/apic/x2apic_uv_x.c 2017-04-18 17:38:08.034643040 +0200 -@@ -947,7 +947,7 @@ - uv_blade_info[blade].pnode = pnode; - uv_blade_info[blade].nr_possible_cpus = 0; - uv_blade_info[blade].nr_online_cpus = 0; -- spin_lock_init(&uv_blade_info[blade].nmi_lock); -+ raw_spin_lock_init(&uv_blade_info[blade].nmi_lock); - min_pnode = min(pnode, min_pnode); - max_pnode = max(pnode, max_pnode); - blade++; -diff -Nur linux-4.4.62.orig/arch/x86/kernel/asm-offsets.c linux-4.4.62/arch/x86/kernel/asm-offsets.c ---- linux-4.4.62.orig/arch/x86/kernel/asm-offsets.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kernel/asm-offsets.c 2017-04-18 17:38:08.034643040 +0200 -@@ -32,6 +32,7 @@ - OFFSET(TI_flags, thread_info, flags); - OFFSET(TI_status, thread_info, status); - OFFSET(TI_addr_limit, thread_info, addr_limit); -+ OFFSET(TI_preempt_lazy_count, thread_info, preempt_lazy_count); - - BLANK(); - OFFSET(crypto_tfm_ctx_offset, crypto_tfm, __crt_ctx); -@@ -89,4 +90,5 @@ +- kernel_fpu_end(); - BLANK(); - DEFINE(PTREGS_SIZE, sizeof(struct pt_regs)); -+ DEFINE(_PREEMPT_ENABLED, PREEMPT_ENABLED); + return err; } -diff -Nur linux-4.4.62.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.4.62/arch/x86/kernel/cpu/mcheck/mce.c ---- linux-4.4.62.orig/arch/x86/kernel/cpu/mcheck/mce.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kernel/cpu/mcheck/mce.c 2017-04-18 17:38:08.034643040 +0200 -@@ -41,6 +41,8 @@ - #include - #include - #include -+#include -+#include +@@ -407,14 +407,14 @@ static int ecb_decrypt(struct blkcipher_desc *desc, + err = blkcipher_walk_virt(desc, &walk); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; - #include - #include -@@ -1236,7 +1238,7 @@ - static unsigned long check_interval = INITIAL_CHECK_INTERVAL; +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes)) { ++ kernel_fpu_begin(); + aesni_ecb_dec(ctx, walk.dst.virt.addr, walk.src.virt.addr, + nbytes & AES_BLOCK_MASK); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- kernel_fpu_end(); - static DEFINE_PER_CPU(unsigned long, mce_next_interval); /* in jiffies */ --static DEFINE_PER_CPU(struct timer_list, mce_timer); -+static DEFINE_PER_CPU(struct hrtimer, mce_timer); + return err; + } +@@ -431,14 +431,14 @@ static int cbc_encrypt(struct blkcipher_desc *desc, + err = blkcipher_walk_virt(desc, &walk); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; - static unsigned long mce_adjust_timer_default(unsigned long interval) - { -@@ -1245,32 +1247,18 @@ +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes)) { ++ kernel_fpu_begin(); + aesni_cbc_enc(ctx, walk.dst.virt.addr, walk.src.virt.addr, + nbytes & AES_BLOCK_MASK, walk.iv); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- kernel_fpu_end(); - static unsigned long (*mce_adjust_timer)(unsigned long interval) = mce_adjust_timer_default; + return err; + } +@@ -455,14 +455,14 @@ static int cbc_decrypt(struct blkcipher_desc *desc, + err = blkcipher_walk_virt(desc, &walk); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; --static void __restart_timer(struct timer_list *t, unsigned long interval) -+static enum hrtimer_restart __restart_timer(struct hrtimer *timer, unsigned long interval) - { -- unsigned long when = jiffies + interval; -- unsigned long flags; -- -- local_irq_save(flags); -- -- if (timer_pending(t)) { -- if (time_before(when, t->expires)) -- mod_timer_pinned(t, when); -- } else { -- t->expires = round_jiffies(when); -- add_timer_on(t, smp_processor_id()); -- } -- -- local_irq_restore(flags); -+ if (!interval) -+ return HRTIMER_NORESTART; -+ hrtimer_forward_now(timer, ns_to_ktime(jiffies_to_nsecs(interval))); -+ return HRTIMER_RESTART; +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes)) { ++ kernel_fpu_begin(); + aesni_cbc_dec(ctx, walk.dst.virt.addr, walk.src.virt.addr, + nbytes & AES_BLOCK_MASK, walk.iv); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- kernel_fpu_end(); + + return err; } +@@ -514,18 +514,20 @@ static int ctr_crypt(struct blkcipher_desc *desc, + err = blkcipher_walk_virt_block(desc, &walk, AES_BLOCK_SIZE); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; --static void mce_timer_fn(unsigned long data) -+static enum hrtimer_restart mce_timer_fn(struct hrtimer *timer) +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes) >= AES_BLOCK_SIZE) { ++ kernel_fpu_begin(); + aesni_ctr_enc_tfm(ctx, walk.dst.virt.addr, walk.src.virt.addr, + nbytes & AES_BLOCK_MASK, walk.iv); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } + if (walk.nbytes) { ++ kernel_fpu_begin(); + ctr_crypt_final(ctx, &walk); ++ kernel_fpu_end(); + err = blkcipher_walk_done(desc, &walk, 0); + } +- kernel_fpu_end(); + + return err; + } +diff --git a/arch/x86/crypto/cast5_avx_glue.c b/arch/x86/crypto/cast5_avx_glue.c +index 8648158f3916..d7699130ee36 100644 +--- a/arch/x86/crypto/cast5_avx_glue.c ++++ b/arch/x86/crypto/cast5_avx_glue.c +@@ -59,7 +59,7 @@ static inline void cast5_fpu_end(bool fpu_enabled) + static int ecb_crypt(struct blkcipher_desc *desc, struct blkcipher_walk *walk, + bool enc) { -- struct timer_list *t = this_cpu_ptr(&mce_timer); -- int cpu = smp_processor_id(); - unsigned long iv; +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct cast5_ctx *ctx = crypto_blkcipher_ctx(desc->tfm); + const unsigned int bsize = CAST5_BLOCK_SIZE; + unsigned int nbytes; +@@ -75,7 +75,7 @@ static int ecb_crypt(struct blkcipher_desc *desc, struct blkcipher_walk *walk, + u8 *wsrc = walk->src.virt.addr; + u8 *wdst = walk->dst.virt.addr; -- WARN_ON(cpu != data); -- - iv = __this_cpu_read(mce_next_interval); +- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); ++ fpu_enabled = cast5_fpu_begin(false, nbytes); - if (mce_available(this_cpu_ptr(&cpu_info))) { -@@ -1293,7 +1281,7 @@ + /* Process multi-block batch */ + if (nbytes >= bsize * CAST5_PARALLEL_BLOCKS) { +@@ -103,10 +103,9 @@ static int ecb_crypt(struct blkcipher_desc *desc, struct blkcipher_walk *walk, + } while (nbytes >= bsize); done: - __this_cpu_write(mce_next_interval, iv); -- __restart_timer(t, iv); -+ return __restart_timer(timer, iv); ++ cast5_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, walk, nbytes); + } +- +- cast5_fpu_end(fpu_enabled); + return err; } - /* -@@ -1301,7 +1289,7 @@ - */ - void mce_timer_kick(unsigned long interval) +@@ -227,7 +226,7 @@ done: + static int cbc_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst, + struct scatterlist *src, unsigned int nbytes) { -- struct timer_list *t = this_cpu_ptr(&mce_timer); -+ struct hrtimer *t = this_cpu_ptr(&mce_timer); - unsigned long iv = __this_cpu_read(mce_next_interval); +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; - __restart_timer(t, interval); -@@ -1316,7 +1304,7 @@ - int cpu; +@@ -236,12 +235,11 @@ static int cbc_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst, + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; - for_each_online_cpu(cpu) -- del_timer_sync(&per_cpu(mce_timer, cpu)); -+ hrtimer_cancel(&per_cpu(mce_timer, cpu)); + while ((nbytes = walk.nbytes)) { +- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); ++ fpu_enabled = cast5_fpu_begin(false, nbytes); + nbytes = __cbc_decrypt(desc, &walk); ++ cast5_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- +- cast5_fpu_end(fpu_enabled); + return err; } - static void mce_do_trigger(struct work_struct *work) -@@ -1326,6 +1314,56 @@ +@@ -311,7 +309,7 @@ done: + static int ctr_crypt(struct blkcipher_desc *desc, struct scatterlist *dst, + struct scatterlist *src, unsigned int nbytes) + { +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; - static DECLARE_WORK(mce_trigger_work, mce_do_trigger); +@@ -320,13 +318,12 @@ static int ctr_crypt(struct blkcipher_desc *desc, struct scatterlist *dst, + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; -+static void __mce_notify_work(struct swork_event *event) -+{ -+ /* Not more than two messages every minute */ -+ static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2); -+ -+ /* wake processes polling /dev/mcelog */ -+ wake_up_interruptible(&mce_chrdev_wait); -+ -+ /* -+ * There is no risk of missing notifications because -+ * work_pending is always cleared before the function is -+ * executed. -+ */ -+ if (mce_helper[0] && !work_pending(&mce_trigger_work)) -+ schedule_work(&mce_trigger_work); -+ -+ if (__ratelimit(&ratelimit)) -+ pr_info(HW_ERR "Machine check events logged\n"); -+} -+ -+#ifdef CONFIG_PREEMPT_RT_FULL -+static bool notify_work_ready __read_mostly; -+static struct swork_event notify_work; -+ -+static int mce_notify_work_init(void) -+{ -+ int err; -+ -+ err = swork_get(); -+ if (err) -+ return err; -+ -+ INIT_SWORK(¬ify_work, __mce_notify_work); -+ notify_work_ready = true; -+ return 0; -+} -+ -+static void mce_notify_work(void) -+{ -+ if (notify_work_ready) -+ swork_queue(¬ify_work); -+} -+#else -+static void mce_notify_work(void) -+{ -+ __mce_notify_work(NULL); -+} -+static inline int mce_notify_work_init(void) { return 0; } -+#endif -+ - /* - * Notify the user(s) about new machine check events. - * Can be called from interrupt context, but not from machine check/NMI -@@ -1333,19 +1371,8 @@ - */ - int mce_notify_irq(void) - { -- /* Not more than two messages every minute */ -- static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2); -- - if (test_and_clear_bit(0, &mce_need_notify)) { -- /* wake processes polling /dev/mcelog */ -- wake_up_interruptible(&mce_chrdev_wait); -- -- if (mce_helper[0]) -- schedule_work(&mce_trigger_work); -- -- if (__ratelimit(&ratelimit)) -- pr_info(HW_ERR "Machine check events logged\n"); -- -+ mce_notify_work(); - return 1; - } - return 0; -@@ -1639,7 +1666,7 @@ + while ((nbytes = walk.nbytes) >= CAST5_BLOCK_SIZE) { +- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); ++ fpu_enabled = cast5_fpu_begin(false, nbytes); + nbytes = __ctr_crypt(desc, &walk); ++ cast5_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); } - } --static void mce_start_timer(unsigned int cpu, struct timer_list *t) -+static void mce_start_timer(unsigned int cpu, struct hrtimer *t) - { - unsigned long iv = check_interval * HZ; +- cast5_fpu_end(fpu_enabled); +- + if (walk.nbytes) { + ctr_crypt_final(desc, &walk); + err = blkcipher_walk_done(desc, &walk, 0); +diff --git a/arch/x86/crypto/glue_helper.c b/arch/x86/crypto/glue_helper.c +index 6a85598931b5..3a506ce7ed93 100644 +--- a/arch/x86/crypto/glue_helper.c ++++ b/arch/x86/crypto/glue_helper.c +@@ -39,7 +39,7 @@ static int __glue_ecb_crypt_128bit(const struct common_glue_ctx *gctx, + void *ctx = crypto_blkcipher_ctx(desc->tfm); + const unsigned int bsize = 128 / 8; + unsigned int nbytes, i, func_bytes; +- bool fpu_enabled = false; ++ bool fpu_enabled; + int err; -@@ -1648,16 +1675,17 @@ + err = blkcipher_walk_virt(desc, walk); +@@ -49,7 +49,7 @@ static int __glue_ecb_crypt_128bit(const struct common_glue_ctx *gctx, + u8 *wdst = walk->dst.virt.addr; - per_cpu(mce_next_interval, cpu) = iv; + fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, +- desc, fpu_enabled, nbytes); ++ desc, false, nbytes); -- t->expires = round_jiffies(jiffies + iv); -- add_timer_on(t, cpu); -+ hrtimer_start_range_ns(t, ns_to_ktime(jiffies_to_usecs(iv) * 1000ULL), -+ 0, HRTIMER_MODE_REL_PINNED); - } + for (i = 0; i < gctx->num_funcs; i++) { + func_bytes = bsize * gctx->funcs[i].num_blocks; +@@ -71,10 +71,10 @@ static int __glue_ecb_crypt_128bit(const struct common_glue_ctx *gctx, + } - static void __mcheck_cpu_init_timer(void) - { -- struct timer_list *t = this_cpu_ptr(&mce_timer); -+ struct hrtimer *t = this_cpu_ptr(&mce_timer); - unsigned int cpu = smp_processor_id(); + done: ++ glue_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, walk, nbytes); + } -- setup_timer(t, mce_timer_fn, cpu); -+ hrtimer_init(t, CLOCK_MONOTONIC, HRTIMER_MODE_REL); -+ t->function = mce_timer_fn; - mce_start_timer(cpu, t); +- glue_fpu_end(fpu_enabled); + return err; } -@@ -2376,6 +2404,8 @@ - if (!mce_available(raw_cpu_ptr(&cpu_info))) - return; +@@ -194,7 +194,7 @@ int glue_cbc_decrypt_128bit(const struct common_glue_ctx *gctx, + struct scatterlist *src, unsigned int nbytes) + { + const unsigned int bsize = 128 / 8; +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; -+ hrtimer_cancel(this_cpu_ptr(&mce_timer)); -+ - if (!(action & CPU_TASKS_FROZEN)) - cmci_clear(); +@@ -203,12 +203,12 @@ int glue_cbc_decrypt_128bit(const struct common_glue_ctx *gctx, -@@ -2398,6 +2428,7 @@ - if (b->init) - wrmsrl(MSR_IA32_MCx_CTL(i), b->ctl); + while ((nbytes = walk.nbytes)) { + fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, +- desc, fpu_enabled, nbytes); ++ desc, false, nbytes); + nbytes = __glue_cbc_decrypt_128bit(gctx, desc, &walk); ++ glue_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); } -+ __mcheck_cpu_init_timer(); - } - /* Get notified when a cpu comes on/off. Be hotplug friendly. */ -@@ -2405,7 +2436,6 @@ - mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) +- glue_fpu_end(fpu_enabled); + return err; + } + EXPORT_SYMBOL_GPL(glue_cbc_decrypt_128bit); +@@ -277,7 +277,7 @@ int glue_ctr_crypt_128bit(const struct common_glue_ctx *gctx, + struct scatterlist *src, unsigned int nbytes) { - unsigned int cpu = (unsigned long)hcpu; -- struct timer_list *t = &per_cpu(mce_timer, cpu); + const unsigned int bsize = 128 / 8; +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; - switch (action & ~CPU_TASKS_FROZEN) { - case CPU_ONLINE: -@@ -2425,11 +2455,9 @@ - break; - case CPU_DOWN_PREPARE: - smp_call_function_single(cpu, mce_disable_cpu, &action, 1); -- del_timer_sync(t); - break; - case CPU_DOWN_FAILED: - smp_call_function_single(cpu, mce_reenable_cpu, &action, 1); -- mce_start_timer(cpu, t); - break; - } +@@ -286,13 +286,12 @@ int glue_ctr_crypt_128bit(const struct common_glue_ctx *gctx, -@@ -2468,6 +2496,10 @@ - goto err_out; + while ((nbytes = walk.nbytes) >= bsize) { + fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, +- desc, fpu_enabled, nbytes); ++ desc, false, nbytes); + nbytes = __glue_ctr_crypt_128bit(gctx, desc, &walk); ++ glue_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); } -+ err = mce_notify_work_init(); -+ if (err) -+ goto err_out; -+ - if (!zalloc_cpumask_var(&mce_device_initialized, GFP_KERNEL)) { - err = -ENOMEM; - goto err_out; -diff -Nur linux-4.4.62.orig/arch/x86/kernel/cpu/perf_event_intel_rapl.c linux-4.4.62/arch/x86/kernel/cpu/perf_event_intel_rapl.c ---- linux-4.4.62.orig/arch/x86/kernel/cpu/perf_event_intel_rapl.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kernel/cpu/perf_event_intel_rapl.c 2017-04-18 17:38:08.034643040 +0200 -@@ -117,7 +117,7 @@ - }; +- glue_fpu_end(fpu_enabled); +- + if (walk.nbytes) { + glue_ctr_crypt_final_128bit( + gctx->funcs[gctx->num_funcs - 1].fn_u.ctr, desc, &walk); +@@ -347,7 +346,7 @@ int glue_xts_crypt_128bit(const struct common_glue_ctx *gctx, + void *tweak_ctx, void *crypt_ctx) + { + const unsigned int bsize = 128 / 8; +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; - struct rapl_pmu { -- spinlock_t lock; -+ raw_spinlock_t lock; - int n_active; /* number of active events */ - struct list_head active_list; - struct pmu *pmu; /* pointer to rapl_pmu_class */ -@@ -220,13 +220,13 @@ - if (!pmu->n_active) - return HRTIMER_NORESTART; +@@ -360,21 +359,21 @@ int glue_xts_crypt_128bit(const struct common_glue_ctx *gctx, -- spin_lock_irqsave(&pmu->lock, flags); -+ raw_spin_lock_irqsave(&pmu->lock, flags); + /* set minimum length to bsize, for tweak_fn */ + fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, +- desc, fpu_enabled, ++ desc, false, + nbytes < bsize ? bsize : nbytes); +- + /* calculate first value of T */ + tweak_fn(tweak_ctx, walk.iv, walk.iv); ++ glue_fpu_end(fpu_enabled); - list_for_each_entry(event, &pmu->active_list, active_entry) { - rapl_event_update(event); + while (nbytes) { ++ fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, ++ desc, false, nbytes); + nbytes = __glue_xts_crypt_128bit(gctx, crypt_ctx, desc, &walk); + ++ glue_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); + nbytes = walk.nbytes; } +- +- glue_fpu_end(fpu_enabled); +- + return err; + } + EXPORT_SYMBOL_GPL(glue_xts_crypt_128bit); +diff --git a/arch/x86/entry/common.c b/arch/x86/entry/common.c +index 1a4477cedc49..75a301b6a5b6 100644 +--- a/arch/x86/entry/common.c ++++ b/arch/x86/entry/common.c +@@ -220,7 +220,7 @@ long syscall_trace_enter(struct pt_regs *regs) -- spin_unlock_irqrestore(&pmu->lock, flags); -+ raw_spin_unlock_irqrestore(&pmu->lock, flags); + #define EXIT_TO_USERMODE_LOOP_FLAGS \ + (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME | _TIF_UPROBE | \ +- _TIF_NEED_RESCHED | _TIF_USER_RETURN_NOTIFY) ++ _TIF_NEED_RESCHED_MASK | _TIF_USER_RETURN_NOTIFY) - hrtimer_forward_now(hrtimer, pmu->timer_interval); + static void exit_to_usermode_loop(struct pt_regs *regs, u32 cached_flags) + { +@@ -236,9 +236,16 @@ static void exit_to_usermode_loop(struct pt_regs *regs, u32 cached_flags) + /* We have work to do. */ + local_irq_enable(); -@@ -263,9 +263,9 @@ - struct rapl_pmu *pmu = __this_cpu_read(rapl_pmu); - unsigned long flags; +- if (cached_flags & _TIF_NEED_RESCHED) ++ if (cached_flags & _TIF_NEED_RESCHED_MASK) + schedule(); -- spin_lock_irqsave(&pmu->lock, flags); -+ raw_spin_lock_irqsave(&pmu->lock, flags); - __rapl_pmu_event_start(pmu, event); -- spin_unlock_irqrestore(&pmu->lock, flags); -+ raw_spin_unlock_irqrestore(&pmu->lock, flags); - } - - static void rapl_pmu_event_stop(struct perf_event *event, int mode) -@@ -274,7 +274,7 @@ - struct hw_perf_event *hwc = &event->hw; - unsigned long flags; ++#ifdef ARCH_RT_DELAYS_SIGNAL_SEND ++ if (unlikely(current->forced_info.si_signo)) { ++ struct task_struct *t = current; ++ force_sig_info(t->forced_info.si_signo, &t->forced_info, t); ++ t->forced_info.si_signo = 0; ++ } ++#endif + if (cached_flags & _TIF_UPROBE) + uprobe_notify_resume(regs); -- spin_lock_irqsave(&pmu->lock, flags); -+ raw_spin_lock_irqsave(&pmu->lock, flags); +diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S +index ae678ad128a9..3bcef8bdb911 100644 +--- a/arch/x86/entry/entry_32.S ++++ b/arch/x86/entry/entry_32.S +@@ -278,8 +278,24 @@ END(ret_from_exception) + ENTRY(resume_kernel) + DISABLE_INTERRUPTS(CLBR_ANY) + need_resched: ++ # preempt count == 0 + NEED_RS set? + cmpl $0, PER_CPU_VAR(__preempt_count) ++#ifndef CONFIG_PREEMPT_LAZY + jnz restore_all ++#else ++ jz test_int_off ++ ++ # atleast preempt count == 0 ? ++ cmpl $_PREEMPT_ENABLED,PER_CPU_VAR(__preempt_count) ++ jne restore_all ++ ++ cmpl $0,TI_preempt_lazy_count(%ebp) # non-zero preempt_lazy_count ? ++ jnz restore_all ++ ++ testl $_TIF_NEED_RESCHED_LAZY, TI_flags(%ebp) ++ jz restore_all ++test_int_off: ++#endif + testl $X86_EFLAGS_IF, PT_EFLAGS(%esp) # interrupts off (exception path) ? + jz restore_all + call preempt_schedule_irq +diff --git a/arch/x86/entry/entry_64.S b/arch/x86/entry/entry_64.S +index a55697d19824..316081a2ca85 100644 +--- a/arch/x86/entry/entry_64.S ++++ b/arch/x86/entry/entry_64.S +@@ -579,7 +579,23 @@ retint_kernel: + bt $9, EFLAGS(%rsp) /* were interrupts off? */ + jnc 1f + 0: cmpl $0, PER_CPU_VAR(__preempt_count) ++#ifndef CONFIG_PREEMPT_LAZY + jnz 1f ++#else ++ jz do_preempt_schedule_irq ++ ++ # atleast preempt count == 0 ? ++ cmpl $_PREEMPT_ENABLED,PER_CPU_VAR(__preempt_count) ++ jnz 1f ++ ++ GET_THREAD_INFO(%rcx) ++ cmpl $0, TI_preempt_lazy_count(%rcx) ++ jnz 1f ++ ++ bt $TIF_NEED_RESCHED_LAZY,TI_flags(%rcx) ++ jnc 1f ++do_preempt_schedule_irq: ++#endif + call preempt_schedule_irq + jmp 0b + 1: +@@ -867,6 +883,7 @@ bad_gs: + jmp 2b + .previous - /* mark event as deactivated and stopped */ - if (!(hwc->state & PERF_HES_STOPPED)) { -@@ -299,7 +299,7 @@ - hwc->state |= PERF_HES_UPTODATE; - } ++#ifndef CONFIG_PREEMPT_RT_FULL + /* Call softirq on interrupt stack. Interrupts are off. */ + ENTRY(do_softirq_own_stack) + pushq %rbp +@@ -879,6 +896,7 @@ ENTRY(do_softirq_own_stack) + decl PER_CPU_VAR(irq_count) + ret + END(do_softirq_own_stack) ++#endif -- spin_unlock_irqrestore(&pmu->lock, flags); -+ raw_spin_unlock_irqrestore(&pmu->lock, flags); + #ifdef CONFIG_XEN + idtentry xen_hypervisor_callback xen_do_hypervisor_callback has_error_code=0 +diff --git a/arch/x86/include/asm/preempt.h b/arch/x86/include/asm/preempt.h +index 01bcde84d3e4..6f432adc55cd 100644 +--- a/arch/x86/include/asm/preempt.h ++++ b/arch/x86/include/asm/preempt.h +@@ -79,17 +79,46 @@ static __always_inline void __preempt_count_sub(int val) + * a decrement which hits zero means we have no preempt_count and should + * reschedule. + */ +-static __always_inline bool __preempt_count_dec_and_test(void) ++static __always_inline bool ____preempt_count_dec_and_test(void) + { + GEN_UNARY_RMWcc("decl", __preempt_count, __percpu_arg(0), "e"); } - static int rapl_pmu_event_add(struct perf_event *event, int mode) -@@ -308,14 +308,14 @@ - struct hw_perf_event *hwc = &event->hw; - unsigned long flags; - -- spin_lock_irqsave(&pmu->lock, flags); -+ raw_spin_lock_irqsave(&pmu->lock, flags); ++static __always_inline bool __preempt_count_dec_and_test(void) ++{ ++ if (____preempt_count_dec_and_test()) ++ return true; ++#ifdef CONFIG_PREEMPT_LAZY ++ if (current_thread_info()->preempt_lazy_count) ++ return false; ++ return test_thread_flag(TIF_NEED_RESCHED_LAZY); ++#else ++ return false; ++#endif ++} ++ + /* + * Returns true when we need to resched and can (barring IRQ state). + */ + static __always_inline bool should_resched(int preempt_offset) + { ++#ifdef CONFIG_PREEMPT_LAZY ++ u32 tmp; ++ ++ tmp = raw_cpu_read_4(__preempt_count); ++ if (tmp == preempt_offset) ++ return true; ++ ++ /* preempt count == 0 ? */ ++ tmp &= ~PREEMPT_NEED_RESCHED; ++ if (tmp) ++ return false; ++ if (current_thread_info()->preempt_lazy_count) ++ return false; ++ return test_thread_flag(TIF_NEED_RESCHED_LAZY); ++#else + return unlikely(raw_cpu_read_4(__preempt_count) == preempt_offset); ++#endif + } - hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED; + #ifdef CONFIG_PREEMPT +diff --git a/arch/x86/include/asm/signal.h b/arch/x86/include/asm/signal.h +index 2138c9ae19ee..3f5b4ee2e2c1 100644 +--- a/arch/x86/include/asm/signal.h ++++ b/arch/x86/include/asm/signal.h +@@ -23,6 +23,19 @@ typedef struct { + unsigned long sig[_NSIG_WORDS]; + } sigset_t; - if (mode & PERF_EF_START) - __rapl_pmu_event_start(pmu, event); ++/* ++ * Because some traps use the IST stack, we must keep preemption ++ * disabled while calling do_trap(), but do_trap() may call ++ * force_sig_info() which will grab the signal spin_locks for the ++ * task, which in PREEMPT_RT_FULL are mutexes. By defining ++ * ARCH_RT_DELAYS_SIGNAL_SEND the force_sig_info() will set ++ * TIF_NOTIFY_RESUME and set up the signal to be sent on exit of the ++ * trap. ++ */ ++#if defined(CONFIG_PREEMPT_RT_FULL) ++#define ARCH_RT_DELAYS_SIGNAL_SEND ++#endif ++ + #ifndef CONFIG_COMPAT + typedef sigset_t compat_sigset_t; + #endif +diff --git a/arch/x86/include/asm/stackprotector.h b/arch/x86/include/asm/stackprotector.h +index 58505f01962f..02fa39652cd6 100644 +--- a/arch/x86/include/asm/stackprotector.h ++++ b/arch/x86/include/asm/stackprotector.h +@@ -59,7 +59,7 @@ + */ + static __always_inline void boot_init_stack_canary(void) + { +- u64 canary; ++ u64 uninitialized_var(canary); + u64 tsc; -- spin_unlock_irqrestore(&pmu->lock, flags); -+ raw_spin_unlock_irqrestore(&pmu->lock, flags); + #ifdef CONFIG_X86_64 +@@ -70,8 +70,15 @@ static __always_inline void boot_init_stack_canary(void) + * of randomness. The TSC only matters for very early init, + * there it already has some randomness on most systems. Later + * on during the bootup the random pool has true entropy too. ++ * ++ * For preempt-rt we need to weaken the randomness a bit, as ++ * we can't call into the random generator from atomic context ++ * due to locking constraints. We just leave canary ++ * uninitialized and use the TSC based randomness on top of it. + */ ++#ifndef CONFIG_PREEMPT_RT_FULL + get_random_bytes(&canary, sizeof(canary)); ++#endif + tsc = rdtsc(); + canary += tsc + (tsc << 32UL); - return 0; - } -@@ -603,7 +603,7 @@ - pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu)); - if (!pmu) - return -1; -- spin_lock_init(&pmu->lock); -+ raw_spin_lock_init(&pmu->lock); +diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h +index c7b551028740..ddb63bd90e3c 100644 +--- a/arch/x86/include/asm/thread_info.h ++++ b/arch/x86/include/asm/thread_info.h +@@ -58,6 +58,8 @@ struct thread_info { + __u32 status; /* thread synchronous flags */ + __u32 cpu; /* current CPU */ + mm_segment_t addr_limit; ++ int preempt_lazy_count; /* 0 => lazy preemptable ++ <0 => BUG */ + unsigned int sig_on_uaccess_error:1; + unsigned int uaccess_err:1; /* uaccess failed */ + }; +@@ -95,6 +97,7 @@ struct thread_info { + #define TIF_SYSCALL_EMU 6 /* syscall emulation active */ + #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ + #define TIF_SECCOMP 8 /* secure computing */ ++#define TIF_NEED_RESCHED_LAZY 9 /* lazy rescheduling necessary */ + #define TIF_USER_RETURN_NOTIFY 11 /* notify kernel of userspace return */ + #define TIF_UPROBE 12 /* breakpointed or singlestepping */ + #define TIF_NOTSC 16 /* TSC is not accessible in userland */ +@@ -119,6 +122,7 @@ struct thread_info { + #define _TIF_SYSCALL_EMU (1 << TIF_SYSCALL_EMU) + #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) + #define _TIF_SECCOMP (1 << TIF_SECCOMP) ++#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) + #define _TIF_USER_RETURN_NOTIFY (1 << TIF_USER_RETURN_NOTIFY) + #define _TIF_UPROBE (1 << TIF_UPROBE) + #define _TIF_NOTSC (1 << TIF_NOTSC) +@@ -152,6 +156,8 @@ struct thread_info { + #define _TIF_WORK_CTXSW_PREV (_TIF_WORK_CTXSW|_TIF_USER_RETURN_NOTIFY) + #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW) - INIT_LIST_HEAD(&pmu->active_list); ++#define _TIF_NEED_RESCHED_MASK (_TIF_NEED_RESCHED | _TIF_NEED_RESCHED_LAZY) ++ + #define STACK_WARN (THREAD_SIZE/8) -diff -Nur linux-4.4.62.orig/arch/x86/kernel/dumpstack_32.c linux-4.4.62/arch/x86/kernel/dumpstack_32.c ---- linux-4.4.62.orig/arch/x86/kernel/dumpstack_32.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kernel/dumpstack_32.c 2017-04-18 17:38:08.034643040 +0200 -@@ -42,7 +42,7 @@ - unsigned long *stack, unsigned long bp, - const struct stacktrace_ops *ops, void *data) + /* +diff --git a/arch/x86/include/asm/uv/uv_bau.h b/arch/x86/include/asm/uv/uv_bau.h +index fc808b83fccb..ebb40118abf5 100644 +--- a/arch/x86/include/asm/uv/uv_bau.h ++++ b/arch/x86/include/asm/uv/uv_bau.h +@@ -615,9 +615,9 @@ struct bau_control { + cycles_t send_message; + cycles_t period_end; + cycles_t period_time; +- spinlock_t uvhub_lock; +- spinlock_t queue_lock; +- spinlock_t disable_lock; ++ raw_spinlock_t uvhub_lock; ++ raw_spinlock_t queue_lock; ++ raw_spinlock_t disable_lock; + /* tunables */ + int max_concurr; + int max_concurr_const; +@@ -776,15 +776,15 @@ static inline int atom_asr(short i, struct atomic_short *v) + * to be lowered below the current 'v'. atomic_add_unless can only stop + * on equal. + */ +-static inline int atomic_inc_unless_ge(spinlock_t *lock, atomic_t *v, int u) ++static inline int atomic_inc_unless_ge(raw_spinlock_t *lock, atomic_t *v, int u) { -- const unsigned cpu = get_cpu(); -+ const unsigned cpu = get_cpu_light(); - int graph = 0; - u32 *prev_esp; - -@@ -86,7 +86,7 @@ - break; - touch_nmi_watchdog(); +- spin_lock(lock); ++ raw_spin_lock(lock); + if (atomic_read(v) >= u) { +- spin_unlock(lock); ++ raw_spin_unlock(lock); + return 0; } -- put_cpu(); -+ put_cpu_light(); + atomic_inc(v); +- spin_unlock(lock); ++ raw_spin_unlock(lock); + return 1; } - EXPORT_SYMBOL(dump_trace); -diff -Nur linux-4.4.62.orig/arch/x86/kernel/dumpstack_64.c linux-4.4.62/arch/x86/kernel/dumpstack_64.c ---- linux-4.4.62.orig/arch/x86/kernel/dumpstack_64.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kernel/dumpstack_64.c 2017-04-18 17:38:08.034643040 +0200 -@@ -152,7 +152,7 @@ - unsigned long *stack, unsigned long bp, - const struct stacktrace_ops *ops, void *data) - { -- const unsigned cpu = get_cpu(); -+ const unsigned cpu = get_cpu_light(); - struct thread_info *tinfo; - unsigned long *irq_stack = (unsigned long *)per_cpu(irq_stack_ptr, cpu); - unsigned long dummy; -@@ -241,7 +241,7 @@ - * This handles the process stack: - */ - bp = ops->walk_stack(tinfo, stack, bp, ops, data, NULL, &graph); -- put_cpu(); -+ put_cpu_light(); - } - EXPORT_SYMBOL(dump_trace); - -@@ -255,7 +255,7 @@ - int cpu; - int i; - -- preempt_disable(); -+ migrate_disable(); - cpu = smp_processor_id(); +diff --git a/arch/x86/include/asm/uv/uv_hub.h b/arch/x86/include/asm/uv/uv_hub.h +index ea7074784cc4..01ec643ce66e 100644 +--- a/arch/x86/include/asm/uv/uv_hub.h ++++ b/arch/x86/include/asm/uv/uv_hub.h +@@ -492,7 +492,7 @@ struct uv_blade_info { + unsigned short nr_online_cpus; + unsigned short pnode; + short memory_nid; +- spinlock_t nmi_lock; /* obsolete, see uv_hub_nmi */ ++ raw_spinlock_t nmi_lock; /* obsolete, see uv_hub_nmi */ + unsigned long nmi_count; /* obsolete, see uv_hub_nmi */ + }; + extern struct uv_blade_info *uv_blade_info; +diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c +index e75907601a41..a29fc4f84fc4 100644 +--- a/arch/x86/kernel/acpi/boot.c ++++ b/arch/x86/kernel/acpi/boot.c +@@ -87,7 +87,9 @@ static u64 acpi_lapic_addr __initdata = APIC_DEFAULT_PHYS_BASE; + * ->ioapic_mutex + * ->ioapic_lock + */ ++#ifdef CONFIG_X86_IO_APIC + static DEFINE_MUTEX(acpi_ioapic_lock); ++#endif - irq_stack_end = (unsigned long *)(per_cpu(irq_stack_ptr, cpu)); -@@ -291,7 +291,7 @@ - pr_cont(" %016lx", *stack++); - touch_nmi_watchdog(); + /* -------------------------------------------------------------------------- + Boot-time Configuration +diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c +index 1e5d2f07416b..ac0127bb155e 100644 +--- a/arch/x86/kernel/apic/io_apic.c ++++ b/arch/x86/kernel/apic/io_apic.c +@@ -1711,7 +1711,8 @@ static bool io_apic_level_ack_pending(struct mp_chip_data *data) + static inline bool ioapic_irqd_mask(struct irq_data *data) + { + /* If we are moving the irq we need to mask it */ +- if (unlikely(irqd_is_setaffinity_pending(data))) { ++ if (unlikely(irqd_is_setaffinity_pending(data) && ++ !irqd_irq_inprogress(data))) { + mask_ioapic_irq(data); + return true; } -- preempt_enable(); -+ migrate_enable(); - - pr_cont("\n"); - show_trace_log_lvl(task, regs, sp, bp, log_lvl); -diff -Nur linux-4.4.62.orig/arch/x86/kernel/irq_32.c linux-4.4.62/arch/x86/kernel/irq_32.c ---- linux-4.4.62.orig/arch/x86/kernel/irq_32.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kernel/irq_32.c 2017-04-18 17:38:08.034643040 +0200 -@@ -128,6 +128,7 @@ - cpu, per_cpu(hardirq_stack, cpu), per_cpu(softirq_stack, cpu)); - } +diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c +index 4a139465f1d4..ad2afff02b36 100644 +--- a/arch/x86/kernel/apic/x2apic_uv_x.c ++++ b/arch/x86/kernel/apic/x2apic_uv_x.c +@@ -947,7 +947,7 @@ void __init uv_system_init(void) + uv_blade_info[blade].pnode = pnode; + uv_blade_info[blade].nr_possible_cpus = 0; + uv_blade_info[blade].nr_online_cpus = 0; +- spin_lock_init(&uv_blade_info[blade].nmi_lock); ++ raw_spin_lock_init(&uv_blade_info[blade].nmi_lock); + min_pnode = min(pnode, min_pnode); + max_pnode = max(pnode, max_pnode); + blade++; +diff --git a/arch/x86/kernel/asm-offsets.c b/arch/x86/kernel/asm-offsets.c +index 439df975bc7a..b7954ddd6a0a 100644 +--- a/arch/x86/kernel/asm-offsets.c ++++ b/arch/x86/kernel/asm-offsets.c +@@ -32,6 +32,7 @@ void common(void) { + OFFSET(TI_flags, thread_info, flags); + OFFSET(TI_status, thread_info, status); + OFFSET(TI_addr_limit, thread_info, addr_limit); ++ OFFSET(TI_preempt_lazy_count, thread_info, preempt_lazy_count); -+#ifndef CONFIG_PREEMPT_RT_FULL - void do_softirq_own_stack(void) - { - struct thread_info *curstk; -@@ -146,6 +147,7 @@ + BLANK(); + OFFSET(crypto_tfm_ctx_offset, crypto_tfm, __crt_ctx); +@@ -89,4 +90,5 @@ void common(void) { - call_on_stack(__do_softirq, isp); + BLANK(); + DEFINE(PTREGS_SIZE, sizeof(struct pt_regs)); ++ DEFINE(_PREEMPT_ENABLED, PREEMPT_ENABLED); } -+#endif - - bool handle_irq(struct irq_desc *desc, struct pt_regs *regs) - { -diff -Nur linux-4.4.62.orig/arch/x86/kernel/kvm.c linux-4.4.62/arch/x86/kernel/kvm.c ---- linux-4.4.62.orig/arch/x86/kernel/kvm.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kernel/kvm.c 2017-04-18 17:38:08.034643040 +0200 -@@ -36,6 +36,7 @@ - #include +diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c +index 7e8a736d09db..430a4ec07811 100644 +--- a/arch/x86/kernel/cpu/mcheck/mce.c ++++ b/arch/x86/kernel/cpu/mcheck/mce.c +@@ -41,6 +41,8 @@ #include - #include -+#include - #include - #include - #include -@@ -91,14 +92,14 @@ + #include + #include ++#include ++#include - struct kvm_task_sleep_node { - struct hlist_node link; -- wait_queue_head_t wq; -+ struct swait_queue_head wq; - u32 token; - int cpu; - bool halted; - }; + #include + #include +@@ -1236,7 +1238,7 @@ void mce_log_therm_throt_event(__u64 status) + static unsigned long check_interval = INITIAL_CHECK_INTERVAL; - static struct kvm_task_sleep_head { -- spinlock_t lock; -+ raw_spinlock_t lock; - struct hlist_head list; - } async_pf_sleepers[KVM_TASK_SLEEP_HASHSIZE]; + static DEFINE_PER_CPU(unsigned long, mce_next_interval); /* in jiffies */ +-static DEFINE_PER_CPU(struct timer_list, mce_timer); ++static DEFINE_PER_CPU(struct hrtimer, mce_timer); -@@ -122,17 +123,17 @@ - u32 key = hash_32(token, KVM_TASK_SLEEP_HASHBITS); - struct kvm_task_sleep_head *b = &async_pf_sleepers[key]; - struct kvm_task_sleep_node n, *e; -- DEFINE_WAIT(wait); -+ DECLARE_SWAITQUEUE(wait); + static unsigned long mce_adjust_timer_default(unsigned long interval) + { +@@ -1245,32 +1247,18 @@ static unsigned long mce_adjust_timer_default(unsigned long interval) - rcu_irq_enter(); + static unsigned long (*mce_adjust_timer)(unsigned long interval) = mce_adjust_timer_default; -- spin_lock(&b->lock); -+ raw_spin_lock(&b->lock); - e = _find_apf_task(b, token); - if (e) { - /* dummy entry exist -> wake up was delivered ahead of PF */ - hlist_del(&e->link); - kfree(e); -- spin_unlock(&b->lock); -+ raw_spin_unlock(&b->lock); +-static void __restart_timer(struct timer_list *t, unsigned long interval) ++static enum hrtimer_restart __restart_timer(struct hrtimer *timer, unsigned long interval) + { +- unsigned long when = jiffies + interval; +- unsigned long flags; +- +- local_irq_save(flags); +- +- if (timer_pending(t)) { +- if (time_before(when, t->expires)) +- mod_timer_pinned(t, when); +- } else { +- t->expires = round_jiffies(when); +- add_timer_on(t, smp_processor_id()); +- } +- +- local_irq_restore(flags); ++ if (!interval) ++ return HRTIMER_NORESTART; ++ hrtimer_forward_now(timer, ns_to_ktime(jiffies_to_nsecs(interval))); ++ return HRTIMER_RESTART; + } - rcu_irq_exit(); - return; -@@ -141,13 +142,13 @@ - n.token = token; - n.cpu = smp_processor_id(); - n.halted = is_idle_task(current) || preempt_count() > 1; -- init_waitqueue_head(&n.wq); -+ init_swait_queue_head(&n.wq); - hlist_add_head(&n.link, &b->list); -- spin_unlock(&b->lock); -+ raw_spin_unlock(&b->lock); +-static void mce_timer_fn(unsigned long data) ++static enum hrtimer_restart mce_timer_fn(struct hrtimer *timer) + { +- struct timer_list *t = this_cpu_ptr(&mce_timer); +- int cpu = smp_processor_id(); + unsigned long iv; - for (;;) { - if (!n.halted) -- prepare_to_wait(&n.wq, &wait, TASK_UNINTERRUPTIBLE); -+ prepare_to_swait(&n.wq, &wait, TASK_UNINTERRUPTIBLE); - if (hlist_unhashed(&n.link)) - break; +- WARN_ON(cpu != data); +- + iv = __this_cpu_read(mce_next_interval); -@@ -166,7 +167,7 @@ - } - } - if (!n.halted) -- finish_wait(&n.wq, &wait); -+ finish_swait(&n.wq, &wait); + if (mce_available(this_cpu_ptr(&cpu_info))) { +@@ -1293,7 +1281,7 @@ static void mce_timer_fn(unsigned long data) - rcu_irq_exit(); - return; -@@ -178,8 +179,8 @@ - hlist_del_init(&n->link); - if (n->halted) - smp_send_reschedule(n->cpu); -- else if (waitqueue_active(&n->wq)) -- wake_up(&n->wq); -+ else if (swait_active(&n->wq)) -+ swake_up(&n->wq); + done: + __this_cpu_write(mce_next_interval, iv); +- __restart_timer(t, iv); ++ return __restart_timer(timer, iv); } - static void apf_task_wake_all(void) -@@ -189,14 +190,14 @@ - for (i = 0; i < KVM_TASK_SLEEP_HASHSIZE; i++) { - struct hlist_node *p, *next; - struct kvm_task_sleep_head *b = &async_pf_sleepers[i]; -- spin_lock(&b->lock); -+ raw_spin_lock(&b->lock); - hlist_for_each_safe(p, next, &b->list) { - struct kvm_task_sleep_node *n = - hlist_entry(p, typeof(*n), link); - if (n->cpu == smp_processor_id()) - apf_task_wake_one(n); - } -- spin_unlock(&b->lock); -+ raw_spin_unlock(&b->lock); - } - } + /* +@@ -1301,7 +1289,7 @@ done: + */ + void mce_timer_kick(unsigned long interval) + { +- struct timer_list *t = this_cpu_ptr(&mce_timer); ++ struct hrtimer *t = this_cpu_ptr(&mce_timer); + unsigned long iv = __this_cpu_read(mce_next_interval); -@@ -212,7 +213,7 @@ - } + __restart_timer(t, interval); +@@ -1316,7 +1304,7 @@ static void mce_timer_delete_all(void) + int cpu; - again: -- spin_lock(&b->lock); -+ raw_spin_lock(&b->lock); - n = _find_apf_task(b, token); - if (!n) { - /* -@@ -225,17 +226,17 @@ - * Allocation failed! Busy wait while other cpu - * handles async PF. - */ -- spin_unlock(&b->lock); -+ raw_spin_unlock(&b->lock); - cpu_relax(); - goto again; - } - n->token = token; - n->cpu = smp_processor_id(); -- init_waitqueue_head(&n->wq); -+ init_swait_queue_head(&n->wq); - hlist_add_head(&n->link, &b->list); - } else - apf_task_wake_one(n); -- spin_unlock(&b->lock); -+ raw_spin_unlock(&b->lock); - return; + for_each_online_cpu(cpu) +- del_timer_sync(&per_cpu(mce_timer, cpu)); ++ hrtimer_cancel(&per_cpu(mce_timer, cpu)); } - EXPORT_SYMBOL_GPL(kvm_async_pf_task_wake); -@@ -486,7 +487,7 @@ - paravirt_ops_setup(); - register_reboot_notifier(&kvm_pv_reboot_nb); - for (i = 0; i < KVM_TASK_SLEEP_HASHSIZE; i++) -- spin_lock_init(&async_pf_sleepers[i].lock); -+ raw_spin_lock_init(&async_pf_sleepers[i].lock); - if (kvm_para_has_feature(KVM_FEATURE_ASYNC_PF)) - x86_init.irqs.trap_init = kvm_apf_trap_init; -diff -Nur linux-4.4.62.orig/arch/x86/kernel/nmi.c linux-4.4.62/arch/x86/kernel/nmi.c ---- linux-4.4.62.orig/arch/x86/kernel/nmi.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kernel/nmi.c 2017-04-18 17:38:08.034643040 +0200 -@@ -231,7 +231,7 @@ - #endif + static void mce_do_trigger(struct work_struct *work) +@@ -1326,6 +1314,56 @@ static void mce_do_trigger(struct work_struct *work) - if (panic_on_unrecovered_nmi) -- panic("NMI: Not continuing"); -+ nmi_panic(regs, "NMI: Not continuing"); - - pr_emerg("Dazed and confused, but trying to continue\n"); - -@@ -255,8 +255,16 @@ - reason, smp_processor_id()); - show_regs(regs); - -- if (panic_on_io_nmi) -- panic("NMI IOCK error: Not continuing"); -+ if (panic_on_io_nmi) { -+ nmi_panic(regs, "NMI IOCK error: Not continuing"); -+ -+ /* -+ * If we end up here, it means we have received an NMI while -+ * processing panic(). Simply return without delaying and -+ * re-enabling NMIs. -+ */ -+ return; -+ } - - /* Re-enable the IOCK line, wait for a few seconds */ - reason = (reason & NMI_REASON_CLEAR_MASK) | NMI_REASON_CLEAR_IOCHK; -@@ -297,7 +305,7 @@ - - pr_emerg("Do you have a strange power saving mode enabled?\n"); - if (unknown_nmi_panic || panic_on_unrecovered_nmi) -- panic("NMI: Not continuing"); -+ nmi_panic(regs, "NMI: Not continuing"); - - pr_emerg("Dazed and confused, but trying to continue\n"); - } -diff -Nur linux-4.4.62.orig/arch/x86/kernel/process_32.c linux-4.4.62/arch/x86/kernel/process_32.c ---- linux-4.4.62.orig/arch/x86/kernel/process_32.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kernel/process_32.c 2017-04-18 17:38:08.034643040 +0200 -@@ -35,6 +35,7 @@ - #include - #include - #include -+#include - - #include - #include -@@ -210,6 +211,35 @@ - } - EXPORT_SYMBOL_GPL(start_thread); + static DECLARE_WORK(mce_trigger_work, mce_do_trigger); -+#ifdef CONFIG_PREEMPT_RT_FULL -+static void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) ++static void __mce_notify_work(struct swork_event *event) +{ -+ int i; ++ /* Not more than two messages every minute */ ++ static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2); + -+ /* -+ * Clear @prev's kmap_atomic mappings -+ */ -+ for (i = 0; i < prev_p->kmap_idx; i++) { -+ int idx = i + KM_TYPE_NR * smp_processor_id(); -+ pte_t *ptep = kmap_pte - idx; ++ /* wake processes polling /dev/mcelog */ ++ wake_up_interruptible(&mce_chrdev_wait); + -+ kpte_clear_flush(ptep, __fix_to_virt(FIX_KMAP_BEGIN + idx)); -+ } + /* -+ * Restore @next_p's kmap_atomic mappings ++ * There is no risk of missing notifications because ++ * work_pending is always cleared before the function is ++ * executed. + */ -+ for (i = 0; i < next_p->kmap_idx; i++) { -+ int idx = i + KM_TYPE_NR * smp_processor_id(); ++ if (mce_helper[0] && !work_pending(&mce_trigger_work)) ++ schedule_work(&mce_trigger_work); + -+ if (!pte_none(next_p->kmap_pte[i])) -+ set_pte(kmap_pte - idx, next_p->kmap_pte[i]); -+ } ++ if (__ratelimit(&ratelimit)) ++ pr_info(HW_ERR "Machine check events logged\n"); ++} ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++static bool notify_work_ready __read_mostly; ++static struct swork_event notify_work; ++ ++static int mce_notify_work_init(void) ++{ ++ int err; ++ ++ err = swork_get(); ++ if (err) ++ return err; ++ ++ INIT_SWORK(¬ify_work, __mce_notify_work); ++ notify_work_ready = true; ++ return 0; ++} ++ ++static void mce_notify_work(void) ++{ ++ if (notify_work_ready) ++ swork_queue(¬ify_work); +} +#else -+static inline void -+switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) { } ++static void mce_notify_work(void) ++{ ++ __mce_notify_work(NULL); ++} ++static inline int mce_notify_work_init(void) { return 0; } +#endif + - /* - * switch_to(x,y) should switch tasks from x to y. -@@ -286,6 +316,8 @@ - task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT)) - __switch_to_xtra(prev_p, next_p, tss); - -+ switch_kmaps(prev_p, next_p); -+ - /* - * Leave lazy mode, flushing any hypercalls made here. - * This must be done before restoring TLS segments so -diff -Nur linux-4.4.62.orig/arch/x86/kernel/reboot.c linux-4.4.62/arch/x86/kernel/reboot.c ---- linux-4.4.62.orig/arch/x86/kernel/reboot.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kernel/reboot.c 2017-04-18 17:38:08.034643040 +0200 -@@ -726,6 +726,7 @@ - static nmi_shootdown_cb shootdown_callback; - - static atomic_t waiting_for_crash_ipi; -+static int crash_ipi_issued; + * Notify the user(s) about new machine check events. + * Can be called from interrupt context, but not from machine check/NMI +@@ -1333,19 +1371,8 @@ static DECLARE_WORK(mce_trigger_work, mce_do_trigger); + */ + int mce_notify_irq(void) + { +- /* Not more than two messages every minute */ +- static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2); +- + if (test_and_clear_bit(0, &mce_need_notify)) { +- /* wake processes polling /dev/mcelog */ +- wake_up_interruptible(&mce_chrdev_wait); +- +- if (mce_helper[0]) +- schedule_work(&mce_trigger_work); +- +- if (__ratelimit(&ratelimit)) +- pr_info(HW_ERR "Machine check events logged\n"); +- ++ mce_notify_work(); + return 1; + } + return 0; +@@ -1639,7 +1666,7 @@ static void __mcheck_cpu_clear_vendor(struct cpuinfo_x86 *c) + } + } - static int crash_nmi_callback(unsigned int val, struct pt_regs *regs) +-static void mce_start_timer(unsigned int cpu, struct timer_list *t) ++static void mce_start_timer(unsigned int cpu, struct hrtimer *t) { -@@ -788,6 +789,9 @@ + unsigned long iv = check_interval * HZ; - smp_send_nmi_allbutself(); +@@ -1648,16 +1675,17 @@ static void mce_start_timer(unsigned int cpu, struct timer_list *t) -+ /* Kick CPUs looping in NMI context. */ -+ WRITE_ONCE(crash_ipi_issued, 1); -+ - msecs = 1000; /* Wait at most a second for the other cpus to stop */ - while ((atomic_read(&waiting_for_crash_ipi) > 0) && msecs) { - mdelay(1); -@@ -796,6 +800,22 @@ + per_cpu(mce_next_interval, cpu) = iv; - /* Leave the nmi callback set */ +- t->expires = round_jiffies(jiffies + iv); +- add_timer_on(t, cpu); ++ hrtimer_start_range_ns(t, ns_to_ktime(jiffies_to_usecs(iv) * 1000ULL), ++ 0, HRTIMER_MODE_REL_PINNED); } -+ -+/* Override the weak function in kernel/panic.c */ -+void nmi_panic_self_stop(struct pt_regs *regs) -+{ -+ while (1) { -+ /* -+ * Wait for the crash dumping IPI to be issued, and then -+ * call its callback directly. -+ */ -+ if (READ_ONCE(crash_ipi_issued)) -+ crash_nmi_callback(0, regs); /* Don't return */ -+ -+ cpu_relax(); -+ } -+} -+ - #else /* !CONFIG_SMP */ - void nmi_shootdown_cpus(nmi_shootdown_cb callback) - { -diff -Nur linux-4.4.62.orig/arch/x86/kvm/lapic.c linux-4.4.62/arch/x86/kvm/lapic.c ---- linux-4.4.62.orig/arch/x86/kvm/lapic.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kvm/lapic.c 2017-04-18 17:38:08.038643196 +0200 -@@ -1195,7 +1195,7 @@ - static void apic_timer_expired(struct kvm_lapic *apic) + + static void __mcheck_cpu_init_timer(void) { - struct kvm_vcpu *vcpu = apic->vcpu; -- wait_queue_head_t *q = &vcpu->wq; -+ struct swait_queue_head *q = &vcpu->wq; - struct kvm_timer *ktimer = &apic->lapic_timer; +- struct timer_list *t = this_cpu_ptr(&mce_timer); ++ struct hrtimer *t = this_cpu_ptr(&mce_timer); + unsigned int cpu = smp_processor_id(); - if (atomic_read(&apic->lapic_timer.pending)) -@@ -1204,8 +1204,8 @@ - atomic_inc(&apic->lapic_timer.pending); - kvm_set_pending_timer(vcpu); +- setup_timer(t, mce_timer_fn, cpu); ++ hrtimer_init(t, CLOCK_MONOTONIC, HRTIMER_MODE_REL); ++ t->function = mce_timer_fn; + mce_start_timer(cpu, t); + } -- if (waitqueue_active(q)) -- wake_up_interruptible(q); -+ if (swait_active(q)) -+ swake_up(q); +@@ -2376,6 +2404,8 @@ static void mce_disable_cpu(void *h) + if (!mce_available(raw_cpu_ptr(&cpu_info))) + return; - if (apic_lvtt_tscdeadline(apic)) - ktimer->expired_tscdeadline = ktimer->tscdeadline; -@@ -1801,6 +1801,7 @@ - hrtimer_init(&apic->lapic_timer.timer, CLOCK_MONOTONIC, - HRTIMER_MODE_ABS); - apic->lapic_timer.timer.function = apic_timer_fn; -+ apic->lapic_timer.timer.irqsafe = 1; ++ hrtimer_cancel(this_cpu_ptr(&mce_timer)); ++ + if (!(action & CPU_TASKS_FROZEN)) + cmci_clear(); - /* - * APIC is created enabled. This will prevent kvm_lapic_set_base from -diff -Nur linux-4.4.62.orig/arch/x86/kvm/x86.c linux-4.4.62/arch/x86/kvm/x86.c ---- linux-4.4.62.orig/arch/x86/kvm/x86.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/kvm/x86.c 2017-04-18 17:38:08.038643196 +0200 -@@ -5810,6 +5810,13 @@ - goto out; +@@ -2398,6 +2428,7 @@ static void mce_reenable_cpu(void *h) + if (b->init) + wrmsrl(MSR_IA32_MCx_CTL(i), b->ctl); } ++ __mcheck_cpu_init_timer(); + } -+#ifdef CONFIG_PREEMPT_RT_FULL -+ if (!boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) { -+ printk(KERN_ERR "RT requires X86_FEATURE_CONSTANT_TSC\n"); -+ return -EOPNOTSUPP; -+ } -+#endif -+ - r = kvm_mmu_module_init(); - if (r) - goto out_free_percpu; -diff -Nur linux-4.4.62.orig/arch/x86/mm/highmem_32.c linux-4.4.62/arch/x86/mm/highmem_32.c ---- linux-4.4.62.orig/arch/x86/mm/highmem_32.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/mm/highmem_32.c 2017-04-18 17:38:08.038643196 +0200 -@@ -32,10 +32,11 @@ - */ - void *kmap_atomic_prot(struct page *page, pgprot_t prot) + /* Get notified when a cpu comes on/off. Be hotplug friendly. */ +@@ -2405,7 +2436,6 @@ static int + mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) { -+ pte_t pte = mk_pte(page, prot); - unsigned long vaddr; - int idx, type; + unsigned int cpu = (unsigned long)hcpu; +- struct timer_list *t = &per_cpu(mce_timer, cpu); -- preempt_disable(); -+ preempt_disable_nort(); - pagefault_disable(); + switch (action & ~CPU_TASKS_FROZEN) { + case CPU_ONLINE: +@@ -2425,11 +2455,9 @@ mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) + break; + case CPU_DOWN_PREPARE: + smp_call_function_single(cpu, mce_disable_cpu, &action, 1); +- del_timer_sync(t); + break; + case CPU_DOWN_FAILED: + smp_call_function_single(cpu, mce_reenable_cpu, &action, 1); +- mce_start_timer(cpu, t); + break; + } - if (!PageHighMem(page)) -@@ -45,7 +46,10 @@ - idx = type + KM_TYPE_NR*smp_processor_id(); - vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); - BUG_ON(!pte_none(*(kmap_pte-idx))); -- set_pte(kmap_pte-idx, mk_pte(page, prot)); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = pte; -+#endif -+ set_pte(kmap_pte-idx, pte); - arch_flush_lazy_mmu_mode(); +@@ -2468,6 +2496,10 @@ static __init int mcheck_init_device(void) + goto err_out; + } - return (void *)vaddr; -@@ -88,6 +92,9 @@ - * is a bad idea also, in case the page changes cacheability - * attributes or becomes a protected page in a hypervisor. - */ -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = __pte(0); -+#endif - kpte_clear_flush(kmap_pte-idx, vaddr); - kmap_atomic_idx_pop(); - arch_flush_lazy_mmu_mode(); -@@ -100,7 +107,7 @@ - #endif ++ err = mce_notify_work_init(); ++ if (err) ++ goto err_out; ++ + if (!zalloc_cpumask_var(&mce_device_initialized, GFP_KERNEL)) { + err = -ENOMEM; + goto err_out; +diff --git a/arch/x86/kernel/cpu/perf_event_intel_rapl.c b/arch/x86/kernel/cpu/perf_event_intel_rapl.c +index ed446bdcbf31..d2ac364e2118 100644 +--- a/arch/x86/kernel/cpu/perf_event_intel_rapl.c ++++ b/arch/x86/kernel/cpu/perf_event_intel_rapl.c +@@ -117,7 +117,7 @@ static struct perf_pmu_events_attr event_attr_##v = { \ + }; - pagefault_enable(); -- preempt_enable(); -+ preempt_enable_nort(); - } - EXPORT_SYMBOL(__kunmap_atomic); + struct rapl_pmu { +- spinlock_t lock; ++ raw_spinlock_t lock; + int n_active; /* number of active events */ + struct list_head active_list; + struct pmu *pmu; /* pointer to rapl_pmu_class */ +@@ -220,13 +220,13 @@ static enum hrtimer_restart rapl_hrtimer_handle(struct hrtimer *hrtimer) + if (!pmu->n_active) + return HRTIMER_NORESTART; -diff -Nur linux-4.4.62.orig/arch/x86/mm/iomap_32.c linux-4.4.62/arch/x86/mm/iomap_32.c ---- linux-4.4.62.orig/arch/x86/mm/iomap_32.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/mm/iomap_32.c 2017-04-18 17:38:08.038643196 +0200 -@@ -56,6 +56,7 @@ +- spin_lock_irqsave(&pmu->lock, flags); ++ raw_spin_lock_irqsave(&pmu->lock, flags); - void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot) - { -+ pte_t pte = pfn_pte(pfn, prot); - unsigned long vaddr; - int idx, type; + list_for_each_entry(event, &pmu->active_list, active_entry) { + rapl_event_update(event); + } -@@ -65,7 +66,12 @@ - type = kmap_atomic_idx_push(); - idx = type + KM_TYPE_NR * smp_processor_id(); - vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); -- set_pte(kmap_pte - idx, pfn_pte(pfn, prot)); -+ WARN_ON(!pte_none(*(kmap_pte - idx))); -+ -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = pte; -+#endif -+ set_pte(kmap_pte - idx, pte); - arch_flush_lazy_mmu_mode(); +- spin_unlock_irqrestore(&pmu->lock, flags); ++ raw_spin_unlock_irqrestore(&pmu->lock, flags); - return (void *)vaddr; -@@ -113,6 +119,9 @@ - * is a bad idea also, in case the page changes cacheability - * attributes or becomes a protected page in a hypervisor. - */ -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = __pte(0); -+#endif - kpte_clear_flush(kmap_pte-idx, vaddr); - kmap_atomic_idx_pop(); - } -diff -Nur linux-4.4.62.orig/arch/x86/mm/pageattr.c linux-4.4.62/arch/x86/mm/pageattr.c ---- linux-4.4.62.orig/arch/x86/mm/pageattr.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/mm/pageattr.c 2017-04-18 17:38:08.038643196 +0200 -@@ -208,7 +208,15 @@ - int in_flags, struct page **pages) - { - unsigned int i, level; -+#ifdef CONFIG_PREEMPT -+ /* -+ * Avoid wbinvd() because it causes latencies on all CPUs, -+ * regardless of any CPU isolation that may be in effect. -+ */ -+ unsigned long do_wbinvd = 0; -+#else - unsigned long do_wbinvd = cache && numpages >= 1024; /* 4M threshold */ -+#endif + hrtimer_forward_now(hrtimer, pmu->timer_interval); - BUG_ON(irqs_disabled()); +@@ -263,9 +263,9 @@ static void rapl_pmu_event_start(struct perf_event *event, int mode) + struct rapl_pmu *pmu = __this_cpu_read(rapl_pmu); + unsigned long flags; -diff -Nur linux-4.4.62.orig/arch/x86/platform/uv/tlb_uv.c linux-4.4.62/arch/x86/platform/uv/tlb_uv.c ---- linux-4.4.62.orig/arch/x86/platform/uv/tlb_uv.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/platform/uv/tlb_uv.c 2017-04-18 17:38:08.038643196 +0200 -@@ -714,9 +714,9 @@ +- spin_lock_irqsave(&pmu->lock, flags); ++ raw_spin_lock_irqsave(&pmu->lock, flags); + __rapl_pmu_event_start(pmu, event); +- spin_unlock_irqrestore(&pmu->lock, flags); ++ raw_spin_unlock_irqrestore(&pmu->lock, flags); + } - quiesce_local_uvhub(hmaster); + static void rapl_pmu_event_stop(struct perf_event *event, int mode) +@@ -274,7 +274,7 @@ static void rapl_pmu_event_stop(struct perf_event *event, int mode) + struct hw_perf_event *hwc = &event->hw; + unsigned long flags; -- spin_lock(&hmaster->queue_lock); -+ raw_spin_lock(&hmaster->queue_lock); - reset_with_ipi(&bau_desc->distribution, bcp); -- spin_unlock(&hmaster->queue_lock); -+ raw_spin_unlock(&hmaster->queue_lock); +- spin_lock_irqsave(&pmu->lock, flags); ++ raw_spin_lock_irqsave(&pmu->lock, flags); - end_uvhub_quiesce(hmaster); + /* mark event as deactivated and stopped */ + if (!(hwc->state & PERF_HES_STOPPED)) { +@@ -299,7 +299,7 @@ static void rapl_pmu_event_stop(struct perf_event *event, int mode) + hwc->state |= PERF_HES_UPTODATE; + } -@@ -736,9 +736,9 @@ +- spin_unlock_irqrestore(&pmu->lock, flags); ++ raw_spin_unlock_irqrestore(&pmu->lock, flags); + } - quiesce_local_uvhub(hmaster); + static int rapl_pmu_event_add(struct perf_event *event, int mode) +@@ -308,14 +308,14 @@ static int rapl_pmu_event_add(struct perf_event *event, int mode) + struct hw_perf_event *hwc = &event->hw; + unsigned long flags; -- spin_lock(&hmaster->queue_lock); -+ raw_spin_lock(&hmaster->queue_lock); - reset_with_ipi(&bau_desc->distribution, bcp); -- spin_unlock(&hmaster->queue_lock); -+ raw_spin_unlock(&hmaster->queue_lock); +- spin_lock_irqsave(&pmu->lock, flags); ++ raw_spin_lock_irqsave(&pmu->lock, flags); - end_uvhub_quiesce(hmaster); + hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED; -@@ -759,7 +759,7 @@ - cycles_t tm1; + if (mode & PERF_EF_START) + __rapl_pmu_event_start(pmu, event); - hmaster = bcp->uvhub_master; -- spin_lock(&hmaster->disable_lock); -+ raw_spin_lock(&hmaster->disable_lock); - if (!bcp->baudisabled) { - stat->s_bau_disabled++; - tm1 = get_cycles(); -@@ -772,7 +772,7 @@ - } - } - } -- spin_unlock(&hmaster->disable_lock); -+ raw_spin_unlock(&hmaster->disable_lock); +- spin_unlock_irqrestore(&pmu->lock, flags); ++ raw_spin_unlock_irqrestore(&pmu->lock, flags); + + return 0; } +@@ -603,7 +603,7 @@ static int rapl_cpu_prepare(int cpu) + pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu)); + if (!pmu) + return -1; +- spin_lock_init(&pmu->lock); ++ raw_spin_lock_init(&pmu->lock); - static void count_max_concurr(int stat, struct bau_control *bcp, -@@ -835,7 +835,7 @@ - */ - static void uv1_throttle(struct bau_control *hmaster, struct ptc_stats *stat) - { -- spinlock_t *lock = &hmaster->uvhub_lock; -+ raw_spinlock_t *lock = &hmaster->uvhub_lock; - atomic_t *v; + INIT_LIST_HEAD(&pmu->active_list); - v = &hmaster->active_descriptor_count; -@@ -968,7 +968,7 @@ - struct bau_control *hmaster; +diff --git a/arch/x86/kernel/dumpstack_32.c b/arch/x86/kernel/dumpstack_32.c +index 464ffd69b92e..00db1aad1548 100644 +--- a/arch/x86/kernel/dumpstack_32.c ++++ b/arch/x86/kernel/dumpstack_32.c +@@ -42,7 +42,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, + unsigned long *stack, unsigned long bp, + const struct stacktrace_ops *ops, void *data) + { +- const unsigned cpu = get_cpu(); ++ const unsigned cpu = get_cpu_light(); + int graph = 0; + u32 *prev_esp; - hmaster = bcp->uvhub_master; -- spin_lock(&hmaster->disable_lock); -+ raw_spin_lock(&hmaster->disable_lock); - if (bcp->baudisabled && (get_cycles() >= bcp->set_bau_on_time)) { - stat->s_bau_reenabled++; - for_each_present_cpu(tcpu) { -@@ -980,10 +980,10 @@ - tbcp->period_giveups = 0; - } - } -- spin_unlock(&hmaster->disable_lock); -+ raw_spin_unlock(&hmaster->disable_lock); - return 0; +@@ -86,7 +86,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, + break; + touch_nmi_watchdog(); } -- spin_unlock(&hmaster->disable_lock); -+ raw_spin_unlock(&hmaster->disable_lock); - return -1; +- put_cpu(); ++ put_cpu_light(); } + EXPORT_SYMBOL(dump_trace); -@@ -1901,9 +1901,9 @@ - bcp->cong_reps = congested_reps; - bcp->disabled_period = sec_2_cycles(disabled_period); - bcp->giveup_limit = giveup_limit; -- spin_lock_init(&bcp->queue_lock); -- spin_lock_init(&bcp->uvhub_lock); -- spin_lock_init(&bcp->disable_lock); -+ raw_spin_lock_init(&bcp->queue_lock); -+ raw_spin_lock_init(&bcp->uvhub_lock); -+ raw_spin_lock_init(&bcp->disable_lock); - } +diff --git a/arch/x86/kernel/dumpstack_64.c b/arch/x86/kernel/dumpstack_64.c +index 5f1c6266eb30..c331e3fef465 100644 +--- a/arch/x86/kernel/dumpstack_64.c ++++ b/arch/x86/kernel/dumpstack_64.c +@@ -152,7 +152,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, + unsigned long *stack, unsigned long bp, + const struct stacktrace_ops *ops, void *data) + { +- const unsigned cpu = get_cpu(); ++ const unsigned cpu = get_cpu_light(); + struct thread_info *tinfo; + unsigned long *irq_stack = (unsigned long *)per_cpu(irq_stack_ptr, cpu); + unsigned long dummy; +@@ -241,7 +241,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, + * This handles the process stack: + */ + bp = ops->walk_stack(tinfo, stack, bp, ops, data, NULL, &graph); +- put_cpu(); ++ put_cpu_light(); } + EXPORT_SYMBOL(dump_trace); -diff -Nur linux-4.4.62.orig/arch/x86/platform/uv/uv_time.c linux-4.4.62/arch/x86/platform/uv/uv_time.c ---- linux-4.4.62.orig/arch/x86/platform/uv/uv_time.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/arch/x86/platform/uv/uv_time.c 2017-04-18 17:38:08.038643196 +0200 -@@ -57,7 +57,7 @@ - - /* There is one of these allocated per node */ - struct uv_rtc_timer_head { -- spinlock_t lock; -+ raw_spinlock_t lock; - /* next cpu waiting for timer, local node relative: */ - int next_cpu; - /* number of cpus on this node: */ -@@ -177,7 +177,7 @@ - uv_rtc_deallocate_timers(); - return -ENOMEM; - } -- spin_lock_init(&head->lock); -+ raw_spin_lock_init(&head->lock); - head->ncpus = uv_blade_nr_possible_cpus(bid); - head->next_cpu = -1; - blade_info[bid] = head; -@@ -231,7 +231,7 @@ - unsigned long flags; - int next_cpu; +@@ -255,7 +255,7 @@ show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs, + int cpu; + int i; -- spin_lock_irqsave(&head->lock, flags); -+ raw_spin_lock_irqsave(&head->lock, flags); +- preempt_disable(); ++ migrate_disable(); + cpu = smp_processor_id(); - next_cpu = head->next_cpu; - *t = expires; -@@ -243,12 +243,12 @@ - if (uv_setup_intr(cpu, expires)) { - *t = ULLONG_MAX; - uv_rtc_find_next_timer(head, pnode); -- spin_unlock_irqrestore(&head->lock, flags); -+ raw_spin_unlock_irqrestore(&head->lock, flags); - return -ETIME; - } + irq_stack_end = (unsigned long *)(per_cpu(irq_stack_ptr, cpu)); +@@ -291,7 +291,7 @@ show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs, + pr_cont(" %016lx", *stack++); + touch_nmi_watchdog(); } +- preempt_enable(); ++ migrate_enable(); -- spin_unlock_irqrestore(&head->lock, flags); -+ raw_spin_unlock_irqrestore(&head->lock, flags); - return 0; + pr_cont("\n"); + show_trace_log_lvl(task, regs, sp, bp, log_lvl); +diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c +index 38da8f29a9c8..ce71f7098f15 100644 +--- a/arch/x86/kernel/irq_32.c ++++ b/arch/x86/kernel/irq_32.c +@@ -128,6 +128,7 @@ void irq_ctx_init(int cpu) + cpu, per_cpu(hardirq_stack, cpu), per_cpu(softirq_stack, cpu)); } -@@ -267,7 +267,7 @@ - unsigned long flags; - int rc = 0; - -- spin_lock_irqsave(&head->lock, flags); -+ raw_spin_lock_irqsave(&head->lock, flags); - - if ((head->next_cpu == bcpu && uv_read_rtc(NULL) >= *t) || force) - rc = 1; -@@ -279,7 +279,7 @@ - uv_rtc_find_next_timer(head, pnode); - } - -- spin_unlock_irqrestore(&head->lock, flags); -+ raw_spin_unlock_irqrestore(&head->lock, flags); - - return rc; - } -@@ -299,13 +299,18 @@ - static cycle_t uv_read_rtc(struct clocksource *cs) ++#ifndef CONFIG_PREEMPT_RT_FULL + void do_softirq_own_stack(void) { - unsigned long offset; -+ cycle_t cycles; - -+ preempt_disable(); - if (uv_get_min_hub_revision_id() == 1) - offset = 0; - else - offset = (uv_blade_processor_id() * L1_CACHE_BYTES) % PAGE_SIZE; + struct thread_info *curstk; +@@ -146,6 +147,7 @@ void do_softirq_own_stack(void) -- return (cycle_t)uv_read_local_mmr(UVH_RTC | offset); -+ cycles = (cycle_t)uv_read_local_mmr(UVH_RTC | offset); -+ preempt_enable(); -+ -+ return cycles; + call_on_stack(__do_softirq, isp); } - - /* -diff -Nur linux-4.4.62.orig/block/blk-core.c linux-4.4.62/block/blk-core.c ---- linux-4.4.62.orig/block/blk-core.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/block/blk-core.c 2017-04-18 17:38:08.038643196 +0200 -@@ -125,6 +125,9 @@ - - INIT_LIST_HEAD(&rq->queuelist); - INIT_LIST_HEAD(&rq->timeout_list); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ INIT_WORK(&rq->work, __blk_mq_complete_request_remote_work); +#endif - rq->cpu = -1; - rq->q = q; - rq->__sector = (sector_t) -1; -@@ -233,7 +236,7 @@ - **/ - void blk_start_queue(struct request_queue *q) + + bool handle_irq(struct irq_desc *desc, struct pt_regs *regs) { -- WARN_ON(!irqs_disabled()); -+ WARN_ON_NONRT(!irqs_disabled()); +diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c +index 47190bd399e7..807950860fb7 100644 +--- a/arch/x86/kernel/kvm.c ++++ b/arch/x86/kernel/kvm.c +@@ -36,6 +36,7 @@ + #include + #include + #include ++#include + #include + #include + #include +@@ -91,14 +92,14 @@ static void kvm_io_delay(void) - queue_flag_clear(QUEUE_FLAG_STOPPED, q); - __blk_run_queue(q); -@@ -659,7 +662,7 @@ - if (!gfpflags_allow_blocking(gfp)) - return -EBUSY; + struct kvm_task_sleep_node { + struct hlist_node link; +- wait_queue_head_t wq; ++ struct swait_queue_head wq; + u32 token; + int cpu; + bool halted; + }; -- ret = wait_event_interruptible(q->mq_freeze_wq, -+ ret = swait_event_interruptible(q->mq_freeze_wq, - !atomic_read(&q->mq_freeze_depth) || - blk_queue_dying(q)); - if (blk_queue_dying(q)) -@@ -679,7 +682,7 @@ - struct request_queue *q = - container_of(ref, struct request_queue, q_usage_counter); + static struct kvm_task_sleep_head { +- spinlock_t lock; ++ raw_spinlock_t lock; + struct hlist_head list; + } async_pf_sleepers[KVM_TASK_SLEEP_HASHSIZE]; -- wake_up_all(&q->mq_freeze_wq); -+ swake_up_all(&q->mq_freeze_wq); - } +@@ -122,17 +123,17 @@ void kvm_async_pf_task_wait(u32 token) + u32 key = hash_32(token, KVM_TASK_SLEEP_HASHBITS); + struct kvm_task_sleep_head *b = &async_pf_sleepers[key]; + struct kvm_task_sleep_node n, *e; +- DEFINE_WAIT(wait); ++ DECLARE_SWAITQUEUE(wait); - struct request_queue *blk_alloc_queue_node(gfp_t gfp_mask, int node_id) -@@ -741,7 +744,7 @@ - q->bypass_depth = 1; - __set_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); + rcu_irq_enter(); -- init_waitqueue_head(&q->mq_freeze_wq); -+ init_swait_queue_head(&q->mq_freeze_wq); +- spin_lock(&b->lock); ++ raw_spin_lock(&b->lock); + e = _find_apf_task(b, token); + if (e) { + /* dummy entry exist -> wake up was delivered ahead of PF */ + hlist_del(&e->link); + kfree(e); +- spin_unlock(&b->lock); ++ raw_spin_unlock(&b->lock); - /* - * Init percpu_ref in atomic mode so that it's faster to shutdown. -@@ -3222,7 +3225,7 @@ - blk_run_queue_async(q); - else - __blk_run_queue(q); -- spin_unlock(q->queue_lock); -+ spin_unlock_irq(q->queue_lock); - } + rcu_irq_exit(); + return; +@@ -141,13 +142,13 @@ void kvm_async_pf_task_wait(u32 token) + n.token = token; + n.cpu = smp_processor_id(); + n.halted = is_idle_task(current) || preempt_count() > 1; +- init_waitqueue_head(&n.wq); ++ init_swait_queue_head(&n.wq); + hlist_add_head(&n.link, &b->list); +- spin_unlock(&b->lock); ++ raw_spin_unlock(&b->lock); - static void flush_plug_callbacks(struct blk_plug *plug, bool from_schedule) -@@ -3270,7 +3273,6 @@ - void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) - { - struct request_queue *q; -- unsigned long flags; - struct request *rq; - LIST_HEAD(list); - unsigned int depth; -@@ -3290,11 +3292,6 @@ - q = NULL; - depth = 0; + for (;;) { + if (!n.halted) +- prepare_to_wait(&n.wq, &wait, TASK_UNINTERRUPTIBLE); ++ prepare_to_swait(&n.wq, &wait, TASK_UNINTERRUPTIBLE); + if (hlist_unhashed(&n.link)) + break; -- /* -- * Save and disable interrupts here, to avoid doing it for every -- * queue lock we have to take. -- */ -- local_irq_save(flags); - while (!list_empty(&list)) { - rq = list_entry_rq(list.next); - list_del_init(&rq->queuelist); -@@ -3307,7 +3304,7 @@ - queue_unplugged(q, depth, from_schedule); - q = rq->q; - depth = 0; -- spin_lock(q->queue_lock); -+ spin_lock_irq(q->queue_lock); +@@ -166,7 +167,7 @@ void kvm_async_pf_task_wait(u32 token) } + } + if (!n.halted) +- finish_wait(&n.wq, &wait); ++ finish_swait(&n.wq, &wait); - /* -@@ -3334,8 +3331,6 @@ - */ - if (q) - queue_unplugged(q, depth, from_schedule); -- -- local_irq_restore(flags); + rcu_irq_exit(); + return; +@@ -178,8 +179,8 @@ static void apf_task_wake_one(struct kvm_task_sleep_node *n) + hlist_del_init(&n->link); + if (n->halted) + smp_send_reschedule(n->cpu); +- else if (waitqueue_active(&n->wq)) +- wake_up(&n->wq); ++ else if (swait_active(&n->wq)) ++ swake_up(&n->wq); } - void blk_finish_plug(struct blk_plug *plug) -diff -Nur linux-4.4.62.orig/block/blk-ioc.c linux-4.4.62/block/blk-ioc.c ---- linux-4.4.62.orig/block/blk-ioc.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/block/blk-ioc.c 2017-04-18 17:38:08.038643196 +0200 -@@ -7,6 +7,7 @@ - #include - #include - #include -+#include - - #include "blk.h" - -@@ -109,7 +110,7 @@ - spin_unlock(q->queue_lock); - } else { - spin_unlock_irqrestore(&ioc->lock, flags); -- cpu_relax(); -+ cpu_chill(); - spin_lock_irqsave_nested(&ioc->lock, flags, 1); - } - } -@@ -187,7 +188,7 @@ - spin_unlock(icq->q->queue_lock); - } else { - spin_unlock_irqrestore(&ioc->lock, flags); -- cpu_relax(); -+ cpu_chill(); - goto retry; + static void apf_task_wake_all(void) +@@ -189,14 +190,14 @@ static void apf_task_wake_all(void) + for (i = 0; i < KVM_TASK_SLEEP_HASHSIZE; i++) { + struct hlist_node *p, *next; + struct kvm_task_sleep_head *b = &async_pf_sleepers[i]; +- spin_lock(&b->lock); ++ raw_spin_lock(&b->lock); + hlist_for_each_safe(p, next, &b->list) { + struct kvm_task_sleep_node *n = + hlist_entry(p, typeof(*n), link); + if (n->cpu == smp_processor_id()) + apf_task_wake_one(n); } +- spin_unlock(&b->lock); ++ raw_spin_unlock(&b->lock); } -diff -Nur linux-4.4.62.orig/block/blk-iopoll.c linux-4.4.62/block/blk-iopoll.c ---- linux-4.4.62.orig/block/blk-iopoll.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/block/blk-iopoll.c 2017-04-18 17:38:08.038643196 +0200 -@@ -35,6 +35,7 @@ - list_add_tail(&iop->list, this_cpu_ptr(&blk_cpu_iopoll)); - __raise_softirq_irqoff(BLOCK_IOPOLL_SOFTIRQ); - local_irq_restore(flags); -+ preempt_check_resched_rt(); } - EXPORT_SYMBOL(blk_iopoll_sched); -@@ -132,6 +133,7 @@ - __raise_softirq_irqoff(BLOCK_IOPOLL_SOFTIRQ); +@@ -212,7 +213,7 @@ void kvm_async_pf_task_wake(u32 token) + } - local_irq_enable(); -+ preempt_check_resched_rt(); + again: +- spin_lock(&b->lock); ++ raw_spin_lock(&b->lock); + n = _find_apf_task(b, token); + if (!n) { + /* +@@ -225,17 +226,17 @@ again: + * Allocation failed! Busy wait while other cpu + * handles async PF. + */ +- spin_unlock(&b->lock); ++ raw_spin_unlock(&b->lock); + cpu_relax(); + goto again; + } + n->token = token; + n->cpu = smp_processor_id(); +- init_waitqueue_head(&n->wq); ++ init_swait_queue_head(&n->wq); + hlist_add_head(&n->link, &b->list); + } else + apf_task_wake_one(n); +- spin_unlock(&b->lock); ++ raw_spin_unlock(&b->lock); + return; } + EXPORT_SYMBOL_GPL(kvm_async_pf_task_wake); +@@ -486,7 +487,7 @@ void __init kvm_guest_init(void) + paravirt_ops_setup(); + register_reboot_notifier(&kvm_pv_reboot_nb); + for (i = 0; i < KVM_TASK_SLEEP_HASHSIZE; i++) +- spin_lock_init(&async_pf_sleepers[i].lock); ++ raw_spin_lock_init(&async_pf_sleepers[i].lock); + if (kvm_para_has_feature(KVM_FEATURE_ASYNC_PF)) + x86_init.irqs.trap_init = kvm_apf_trap_init; - /** -@@ -201,6 +203,7 @@ - this_cpu_ptr(&blk_cpu_iopoll)); - __raise_softirq_irqoff(BLOCK_IOPOLL_SOFTIRQ); - local_irq_enable(); -+ preempt_check_resched_rt(); - } - - return NOTIFY_OK; -diff -Nur linux-4.4.62.orig/block/blk-mq.c linux-4.4.62/block/blk-mq.c ---- linux-4.4.62.orig/block/blk-mq.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/block/blk-mq.c 2017-04-18 17:38:08.038643196 +0200 -@@ -92,7 +92,7 @@ +diff --git a/arch/x86/kernel/nmi.c b/arch/x86/kernel/nmi.c +index 697f90db0e37..424aec4a4c71 100644 +--- a/arch/x86/kernel/nmi.c ++++ b/arch/x86/kernel/nmi.c +@@ -231,7 +231,7 @@ pci_serr_error(unsigned char reason, struct pt_regs *regs) + #endif - static void blk_mq_freeze_queue_wait(struct request_queue *q) - { -- wait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->q_usage_counter)); -+ swait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->q_usage_counter)); - } + if (panic_on_unrecovered_nmi) +- panic("NMI: Not continuing"); ++ nmi_panic(regs, "NMI: Not continuing"); - /* -@@ -130,7 +130,7 @@ - WARN_ON_ONCE(freeze_depth < 0); - if (!freeze_depth) { - percpu_ref_reinit(&q->q_usage_counter); -- wake_up_all(&q->mq_freeze_wq); -+ swake_up_all(&q->mq_freeze_wq); - } - } - EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue); -@@ -149,7 +149,7 @@ - * dying, we need to ensure that processes currently waiting on - * the queue are notified as well. - */ -- wake_up_all(&q->mq_freeze_wq); -+ swake_up_all(&q->mq_freeze_wq); - } + pr_emerg("Dazed and confused, but trying to continue\n"); - bool blk_mq_can_queue(struct blk_mq_hw_ctx *hctx) -@@ -196,6 +196,9 @@ - rq->resid_len = 0; - rq->sense = NULL; +@@ -255,8 +255,16 @@ io_check_error(unsigned char reason, struct pt_regs *regs) + reason, smp_processor_id()); + show_regs(regs); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ INIT_WORK(&rq->work, __blk_mq_complete_request_remote_work); -+#endif - INIT_LIST_HEAD(&rq->timeout_list); - rq->timeout = 0; +- if (panic_on_io_nmi) +- panic("NMI IOCK error: Not continuing"); ++ if (panic_on_io_nmi) { ++ nmi_panic(regs, "NMI IOCK error: Not continuing"); ++ ++ /* ++ * If we end up here, it means we have received an NMI while ++ * processing panic(). Simply return without delaying and ++ * re-enabling NMIs. ++ */ ++ return; ++ } + + /* Re-enable the IOCK line, wait for a few seconds */ + reason = (reason & NMI_REASON_CLEAR_MASK) | NMI_REASON_CLEAR_IOCHK; +@@ -297,7 +305,7 @@ unknown_nmi_error(unsigned char reason, struct pt_regs *regs) + + pr_emerg("Do you have a strange power saving mode enabled?\n"); + if (unknown_nmi_panic || panic_on_unrecovered_nmi) +- panic("NMI: Not continuing"); ++ nmi_panic(regs, "NMI: Not continuing"); -@@ -325,6 +328,17 @@ + pr_emerg("Dazed and confused, but trying to continue\n"); } - EXPORT_SYMBOL(blk_mq_end_request); +diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c +index 9f950917528b..4dd4beae917a 100644 +--- a/arch/x86/kernel/process_32.c ++++ b/arch/x86/kernel/process_32.c +@@ -35,6 +35,7 @@ + #include + #include + #include ++#include + + #include + #include +@@ -210,6 +211,35 @@ start_thread(struct pt_regs *regs, unsigned long new_ip, unsigned long new_sp) + } + EXPORT_SYMBOL_GPL(start_thread); +#ifdef CONFIG_PREEMPT_RT_FULL -+ -+void __blk_mq_complete_request_remote_work(struct work_struct *work) ++static void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) +{ -+ struct request *rq = container_of(work, struct request, work); -+ -+ rq->q->softirq_done_fn(rq); -+} ++ int i; + -+#else ++ /* ++ * Clear @prev's kmap_atomic mappings ++ */ ++ for (i = 0; i < prev_p->kmap_idx; i++) { ++ int idx = i + KM_TYPE_NR * smp_processor_id(); ++ pte_t *ptep = kmap_pte - idx; + - static void __blk_mq_complete_request_remote(void *data) - { - struct request *rq = data; -@@ -332,6 +346,8 @@ - rq->q->softirq_done_fn(rq); - } - -+#endif ++ kpte_clear_flush(ptep, __fix_to_virt(FIX_KMAP_BEGIN + idx)); ++ } ++ /* ++ * Restore @next_p's kmap_atomic mappings ++ */ ++ for (i = 0; i < next_p->kmap_idx; i++) { ++ int idx = i + KM_TYPE_NR * smp_processor_id(); + - static void blk_mq_ipi_complete_request(struct request *rq) - { - struct blk_mq_ctx *ctx = rq->mq_ctx; -@@ -343,19 +359,23 @@ - return; - } - -- cpu = get_cpu(); -+ cpu = get_cpu_light(); - if (!test_bit(QUEUE_FLAG_SAME_FORCE, &rq->q->queue_flags)) - shared = cpus_share_cache(cpu, ctx->cpu); - - if (cpu != ctx->cpu && !shared && cpu_online(ctx->cpu)) { -+#ifdef CONFIG_PREEMPT_RT_FULL -+ schedule_work_on(ctx->cpu, &rq->work); ++ if (!pte_none(next_p->kmap_pte[i])) ++ set_pte(kmap_pte - idx, next_p->kmap_pte[i]); ++ } ++} +#else - rq->csd.func = __blk_mq_complete_request_remote; - rq->csd.info = rq; - rq->csd.flags = 0; - smp_call_function_single_async(ctx->cpu, &rq->csd); ++static inline void ++switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) { } +#endif - } else { - rq->q->softirq_done_fn(rq); - } -- put_cpu(); -+ put_cpu_light(); - } ++ - static void __blk_mq_complete_request(struct request *rq) -@@ -862,14 +882,14 @@ - return; + /* + * switch_to(x,y) should switch tasks from x to y. +@@ -286,6 +316,8 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) + task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT)) + __switch_to_xtra(prev_p, next_p, tss); - if (!async) { -- int cpu = get_cpu(); -+ int cpu = get_cpu_light(); - if (cpumask_test_cpu(cpu, hctx->cpumask)) { - __blk_mq_run_hw_queue(hctx); -- put_cpu(); -+ put_cpu_light(); - return; - } ++ switch_kmaps(prev_p, next_p); ++ + /* + * Leave lazy mode, flushing any hypercalls made here. + * This must be done before restoring TLS segments so +diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c +index f660d63f40fe..8384207adde2 100644 +--- a/arch/x86/kernel/reboot.c ++++ b/arch/x86/kernel/reboot.c +@@ -726,6 +726,7 @@ static int crashing_cpu; + static nmi_shootdown_cb shootdown_callback; -- put_cpu(); -+ put_cpu_light(); - } + static atomic_t waiting_for_crash_ipi; ++static int crash_ipi_issued; - kblockd_schedule_delayed_work_on(blk_mq_hctx_next_cpu(hctx), -@@ -1616,7 +1636,7 @@ + static int crash_nmi_callback(unsigned int val, struct pt_regs *regs) { - struct blk_mq_hw_ctx *hctx = data; - -- if (action == CPU_DEAD || action == CPU_DEAD_FROZEN) -+ if (action == CPU_POST_DEAD) - return blk_mq_hctx_cpu_offline(hctx, cpu); - - /* -diff -Nur linux-4.4.62.orig/block/blk-mq-cpu.c linux-4.4.62/block/blk-mq-cpu.c ---- linux-4.4.62.orig/block/blk-mq-cpu.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/block/blk-mq-cpu.c 2017-04-18 17:38:08.038643196 +0200 -@@ -16,7 +16,7 @@ - #include "blk-mq.h" - - static LIST_HEAD(blk_mq_cpu_notify_list); --static DEFINE_RAW_SPINLOCK(blk_mq_cpu_notify_lock); -+static DEFINE_SPINLOCK(blk_mq_cpu_notify_lock); +@@ -788,6 +789,9 @@ void nmi_shootdown_cpus(nmi_shootdown_cb callback) - static int blk_mq_main_cpu_notify(struct notifier_block *self, - unsigned long action, void *hcpu) -@@ -25,7 +25,10 @@ - struct blk_mq_cpu_notifier *notify; - int ret = NOTIFY_OK; + smp_send_nmi_allbutself(); -- raw_spin_lock(&blk_mq_cpu_notify_lock); -+ if (action != CPU_POST_DEAD) -+ return NOTIFY_OK; ++ /* Kick CPUs looping in NMI context. */ ++ WRITE_ONCE(crash_ipi_issued, 1); + -+ spin_lock(&blk_mq_cpu_notify_lock); - - list_for_each_entry(notify, &blk_mq_cpu_notify_list, list) { - ret = notify->notify(notify->data, action, cpu); -@@ -33,7 +36,7 @@ - break; - } + msecs = 1000; /* Wait at most a second for the other cpus to stop */ + while ((atomic_read(&waiting_for_crash_ipi) > 0) && msecs) { + mdelay(1); +@@ -796,6 +800,22 @@ void nmi_shootdown_cpus(nmi_shootdown_cb callback) -- raw_spin_unlock(&blk_mq_cpu_notify_lock); -+ spin_unlock(&blk_mq_cpu_notify_lock); - return ret; + /* Leave the nmi callback set */ } - -@@ -41,16 +44,16 @@ ++ ++/* Override the weak function in kernel/panic.c */ ++void nmi_panic_self_stop(struct pt_regs *regs) ++{ ++ while (1) { ++ /* ++ * Wait for the crash dumping IPI to be issued, and then ++ * call its callback directly. ++ */ ++ if (READ_ONCE(crash_ipi_issued)) ++ crash_nmi_callback(0, regs); /* Don't return */ ++ ++ cpu_relax(); ++ } ++} ++ + #else /* !CONFIG_SMP */ + void nmi_shootdown_cpus(nmi_shootdown_cb callback) { - BUG_ON(!notifier->notify); - -- raw_spin_lock(&blk_mq_cpu_notify_lock); -+ spin_lock(&blk_mq_cpu_notify_lock); - list_add_tail(¬ifier->list, &blk_mq_cpu_notify_list); -- raw_spin_unlock(&blk_mq_cpu_notify_lock); -+ spin_unlock(&blk_mq_cpu_notify_lock); - } - - void blk_mq_unregister_cpu_notifier(struct blk_mq_cpu_notifier *notifier) +diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c +index 1c96f09367ae..ffb6d9859122 100644 +--- a/arch/x86/kvm/lapic.c ++++ b/arch/x86/kvm/lapic.c +@@ -1195,7 +1195,7 @@ static void apic_update_lvtt(struct kvm_lapic *apic) + static void apic_timer_expired(struct kvm_lapic *apic) { -- raw_spin_lock(&blk_mq_cpu_notify_lock); -+ spin_lock(&blk_mq_cpu_notify_lock); - list_del(¬ifier->list); -- raw_spin_unlock(&blk_mq_cpu_notify_lock); -+ spin_unlock(&blk_mq_cpu_notify_lock); - } + struct kvm_vcpu *vcpu = apic->vcpu; +- wait_queue_head_t *q = &vcpu->wq; ++ struct swait_queue_head *q = &vcpu->wq; + struct kvm_timer *ktimer = &apic->lapic_timer; - void blk_mq_init_cpu_notifier(struct blk_mq_cpu_notifier *notifier, -diff -Nur linux-4.4.62.orig/block/blk-mq.h linux-4.4.62/block/blk-mq.h ---- linux-4.4.62.orig/block/blk-mq.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/block/blk-mq.h 2017-04-18 17:38:08.038643196 +0200 -@@ -74,7 +74,10 @@ - static inline struct blk_mq_ctx *__blk_mq_get_ctx(struct request_queue *q, - unsigned int cpu) - { -- return per_cpu_ptr(q->queue_ctx, cpu); -+ struct blk_mq_ctx *ctx; -+ -+ ctx = per_cpu_ptr(q->queue_ctx, cpu); -+ return ctx; - } + if (atomic_read(&apic->lapic_timer.pending)) +@@ -1204,8 +1204,8 @@ static void apic_timer_expired(struct kvm_lapic *apic) + atomic_inc(&apic->lapic_timer.pending); + kvm_set_pending_timer(vcpu); - /* -@@ -85,12 +88,12 @@ - */ - static inline struct blk_mq_ctx *blk_mq_get_ctx(struct request_queue *q) - { -- return __blk_mq_get_ctx(q, get_cpu()); -+ return __blk_mq_get_ctx(q, get_cpu_light()); - } +- if (waitqueue_active(q)) +- wake_up_interruptible(q); ++ if (swait_active(q)) ++ swake_up(q); - static inline void blk_mq_put_ctx(struct blk_mq_ctx *ctx) - { -- put_cpu(); -+ put_cpu_light(); - } - - struct blk_mq_alloc_data { -diff -Nur linux-4.4.62.orig/block/blk-softirq.c linux-4.4.62/block/blk-softirq.c ---- linux-4.4.62.orig/block/blk-softirq.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/block/blk-softirq.c 2017-04-18 17:38:08.038643196 +0200 -@@ -51,6 +51,7 @@ - raise_softirq_irqoff(BLOCK_SOFTIRQ); - - local_irq_restore(flags); -+ preempt_check_resched_rt(); - } + if (apic_lvtt_tscdeadline(apic)) + ktimer->expired_tscdeadline = ktimer->tscdeadline; +@@ -1801,6 +1801,7 @@ int kvm_create_lapic(struct kvm_vcpu *vcpu) + hrtimer_init(&apic->lapic_timer.timer, CLOCK_MONOTONIC, + HRTIMER_MODE_ABS); + apic->lapic_timer.timer.function = apic_timer_fn; ++ apic->lapic_timer.timer.irqsafe = 1; - /* -@@ -93,6 +94,7 @@ - this_cpu_ptr(&blk_cpu_done)); - raise_softirq_irqoff(BLOCK_SOFTIRQ); - local_irq_enable(); -+ preempt_check_resched_rt(); + /* + * APIC is created enabled. This will prevent kvm_lapic_set_base from +diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c +index ae2b9cd358f2..50c9861d4219 100644 +--- a/arch/x86/kvm/x86.c ++++ b/arch/x86/kvm/x86.c +@@ -5827,6 +5827,13 @@ int kvm_arch_init(void *opaque) + goto out; } - return NOTIFY_OK; -@@ -150,6 +152,7 @@ - goto do_local; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ if (!boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) { ++ printk(KERN_ERR "RT requires X86_FEATURE_CONSTANT_TSC\n"); ++ return -EOPNOTSUPP; ++ } ++#endif ++ + r = kvm_mmu_module_init(); + if (r) + goto out_free_percpu; +diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c +index a6d739258137..bd24ba1c4a86 100644 +--- a/arch/x86/mm/highmem_32.c ++++ b/arch/x86/mm/highmem_32.c +@@ -32,10 +32,11 @@ EXPORT_SYMBOL(kunmap); + */ + void *kmap_atomic_prot(struct page *page, pgprot_t prot) + { ++ pte_t pte = mk_pte(page, prot); + unsigned long vaddr; + int idx, type; - local_irq_restore(flags); -+ preempt_check_resched_rt(); - } +- preempt_disable(); ++ preempt_disable_nort(); + pagefault_disable(); - /** -diff -Nur linux-4.4.62.orig/block/bounce.c linux-4.4.62/block/bounce.c ---- linux-4.4.62.orig/block/bounce.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/block/bounce.c 2017-04-18 17:38:08.038643196 +0200 -@@ -55,11 +55,11 @@ - unsigned long flags; - unsigned char *vto; + if (!PageHighMem(page)) +@@ -45,7 +46,10 @@ void *kmap_atomic_prot(struct page *page, pgprot_t prot) + idx = type + KM_TYPE_NR*smp_processor_id(); + vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); + BUG_ON(!pte_none(*(kmap_pte-idx))); +- set_pte(kmap_pte-idx, mk_pte(page, prot)); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = pte; ++#endif ++ set_pte(kmap_pte-idx, pte); + arch_flush_lazy_mmu_mode(); -- local_irq_save(flags); -+ local_irq_save_nort(flags); - vto = kmap_atomic(to->bv_page); - memcpy(vto + to->bv_offset, vfrom, to->bv_len); - kunmap_atomic(vto); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); + return (void *)vaddr; +@@ -88,6 +92,9 @@ void __kunmap_atomic(void *kvaddr) + * is a bad idea also, in case the page changes cacheability + * attributes or becomes a protected page in a hypervisor. + */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = __pte(0); ++#endif + kpte_clear_flush(kmap_pte-idx, vaddr); + kmap_atomic_idx_pop(); + arch_flush_lazy_mmu_mode(); +@@ -100,7 +107,7 @@ void __kunmap_atomic(void *kvaddr) + #endif + + pagefault_enable(); +- preempt_enable(); ++ preempt_enable_nort(); } + EXPORT_SYMBOL(__kunmap_atomic); - #else /* CONFIG_HIGHMEM */ -diff -Nur linux-4.4.62.orig/crypto/algapi.c linux-4.4.62/crypto/algapi.c ---- linux-4.4.62.orig/crypto/algapi.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/crypto/algapi.c 2017-04-18 17:38:08.042643350 +0200 -@@ -720,13 +720,13 @@ +diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c +index 9c0ff045fdd4..dd25dd1671b6 100644 +--- a/arch/x86/mm/iomap_32.c ++++ b/arch/x86/mm/iomap_32.c +@@ -56,6 +56,7 @@ EXPORT_SYMBOL_GPL(iomap_free); - int crypto_register_notifier(struct notifier_block *nb) + void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot) { -- return blocking_notifier_chain_register(&crypto_chain, nb); -+ return srcu_notifier_chain_register(&crypto_chain, nb); - } - EXPORT_SYMBOL_GPL(crypto_register_notifier); ++ pte_t pte = pfn_pte(pfn, prot); + unsigned long vaddr; + int idx, type; - int crypto_unregister_notifier(struct notifier_block *nb) +@@ -65,7 +66,12 @@ void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot) + type = kmap_atomic_idx_push(); + idx = type + KM_TYPE_NR * smp_processor_id(); + vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); +- set_pte(kmap_pte - idx, pfn_pte(pfn, prot)); ++ WARN_ON(!pte_none(*(kmap_pte - idx))); ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = pte; ++#endif ++ set_pte(kmap_pte - idx, pte); + arch_flush_lazy_mmu_mode(); + + return (void *)vaddr; +@@ -113,6 +119,9 @@ iounmap_atomic(void __iomem *kvaddr) + * is a bad idea also, in case the page changes cacheability + * attributes or becomes a protected page in a hypervisor. + */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = __pte(0); ++#endif + kpte_clear_flush(kmap_pte-idx, vaddr); + kmap_atomic_idx_pop(); + } +diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c +index b599a780a5a9..2e85c4117daf 100644 +--- a/arch/x86/mm/pageattr.c ++++ b/arch/x86/mm/pageattr.c +@@ -208,7 +208,15 @@ static void cpa_flush_array(unsigned long *start, int numpages, int cache, + int in_flags, struct page **pages) { -- return blocking_notifier_chain_unregister(&crypto_chain, nb); -+ return srcu_notifier_chain_unregister(&crypto_chain, nb); - } - EXPORT_SYMBOL_GPL(crypto_unregister_notifier); + unsigned int i, level; ++#ifdef CONFIG_PREEMPT ++ /* ++ * Avoid wbinvd() because it causes latencies on all CPUs, ++ * regardless of any CPU isolation that may be in effect. ++ */ ++ unsigned long do_wbinvd = 0; ++#else + unsigned long do_wbinvd = cache && numpages >= 1024; /* 4M threshold */ ++#endif -diff -Nur linux-4.4.62.orig/crypto/api.c linux-4.4.62/crypto/api.c ---- linux-4.4.62.orig/crypto/api.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/crypto/api.c 2017-04-18 17:38:08.042643350 +0200 -@@ -31,7 +31,7 @@ - DECLARE_RWSEM(crypto_alg_sem); - EXPORT_SYMBOL_GPL(crypto_alg_sem); + BUG_ON(irqs_disabled()); --BLOCKING_NOTIFIER_HEAD(crypto_chain); -+SRCU_NOTIFIER_HEAD(crypto_chain); - EXPORT_SYMBOL_GPL(crypto_chain); +diff --git a/arch/x86/platform/uv/tlb_uv.c b/arch/x86/platform/uv/tlb_uv.c +index 3b6ec42718e4..7871083de089 100644 +--- a/arch/x86/platform/uv/tlb_uv.c ++++ b/arch/x86/platform/uv/tlb_uv.c +@@ -714,9 +714,9 @@ static void destination_plugged(struct bau_desc *bau_desc, - static struct crypto_alg *crypto_larval_wait(struct crypto_alg *alg); -@@ -236,10 +236,10 @@ - { - int ok; + quiesce_local_uvhub(hmaster); -- ok = blocking_notifier_call_chain(&crypto_chain, val, v); -+ ok = srcu_notifier_call_chain(&crypto_chain, val, v); - if (ok == NOTIFY_DONE) { - request_module("cryptomgr"); -- ok = blocking_notifier_call_chain(&crypto_chain, val, v); -+ ok = srcu_notifier_call_chain(&crypto_chain, val, v); - } +- spin_lock(&hmaster->queue_lock); ++ raw_spin_lock(&hmaster->queue_lock); + reset_with_ipi(&bau_desc->distribution, bcp); +- spin_unlock(&hmaster->queue_lock); ++ raw_spin_unlock(&hmaster->queue_lock); - return ok; -diff -Nur linux-4.4.62.orig/crypto/internal.h linux-4.4.62/crypto/internal.h ---- linux-4.4.62.orig/crypto/internal.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/crypto/internal.h 2017-04-18 17:38:08.042643350 +0200 -@@ -47,7 +47,7 @@ + end_uvhub_quiesce(hmaster); - extern struct list_head crypto_alg_list; - extern struct rw_semaphore crypto_alg_sem; --extern struct blocking_notifier_head crypto_chain; -+extern struct srcu_notifier_head crypto_chain; +@@ -736,9 +736,9 @@ static void destination_timeout(struct bau_desc *bau_desc, - #ifdef CONFIG_PROC_FS - void __init crypto_init_proc(void); -@@ -143,7 +143,7 @@ + quiesce_local_uvhub(hmaster); - static inline void crypto_notify(unsigned long val, void *v) +- spin_lock(&hmaster->queue_lock); ++ raw_spin_lock(&hmaster->queue_lock); + reset_with_ipi(&bau_desc->distribution, bcp); +- spin_unlock(&hmaster->queue_lock); ++ raw_spin_unlock(&hmaster->queue_lock); + + end_uvhub_quiesce(hmaster); + +@@ -759,7 +759,7 @@ static void disable_for_period(struct bau_control *bcp, struct ptc_stats *stat) + cycles_t tm1; + + hmaster = bcp->uvhub_master; +- spin_lock(&hmaster->disable_lock); ++ raw_spin_lock(&hmaster->disable_lock); + if (!bcp->baudisabled) { + stat->s_bau_disabled++; + tm1 = get_cycles(); +@@ -772,7 +772,7 @@ static void disable_for_period(struct bau_control *bcp, struct ptc_stats *stat) + } + } + } +- spin_unlock(&hmaster->disable_lock); ++ raw_spin_unlock(&hmaster->disable_lock); + } + + static void count_max_concurr(int stat, struct bau_control *bcp, +@@ -835,7 +835,7 @@ static void record_send_stats(cycles_t time1, cycles_t time2, + */ + static void uv1_throttle(struct bau_control *hmaster, struct ptc_stats *stat) { -- blocking_notifier_call_chain(&crypto_chain, val, v); -+ srcu_notifier_call_chain(&crypto_chain, val, v); +- spinlock_t *lock = &hmaster->uvhub_lock; ++ raw_spinlock_t *lock = &hmaster->uvhub_lock; + atomic_t *v; + + v = &hmaster->active_descriptor_count; +@@ -968,7 +968,7 @@ static int check_enable(struct bau_control *bcp, struct ptc_stats *stat) + struct bau_control *hmaster; + + hmaster = bcp->uvhub_master; +- spin_lock(&hmaster->disable_lock); ++ raw_spin_lock(&hmaster->disable_lock); + if (bcp->baudisabled && (get_cycles() >= bcp->set_bau_on_time)) { + stat->s_bau_reenabled++; + for_each_present_cpu(tcpu) { +@@ -980,10 +980,10 @@ static int check_enable(struct bau_control *bcp, struct ptc_stats *stat) + tbcp->period_giveups = 0; + } + } +- spin_unlock(&hmaster->disable_lock); ++ raw_spin_unlock(&hmaster->disable_lock); + return 0; + } +- spin_unlock(&hmaster->disable_lock); ++ raw_spin_unlock(&hmaster->disable_lock); + return -1; } - #endif /* _CRYPTO_INTERNAL_H */ -diff -Nur linux-4.4.62.orig/Documentation/hwlat_detector.txt linux-4.4.62/Documentation/hwlat_detector.txt ---- linux-4.4.62.orig/Documentation/hwlat_detector.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/Documentation/hwlat_detector.txt 2017-04-18 17:38:07.902637922 +0200 -@@ -0,0 +1,64 @@ -+Introduction: -+------------- -+ -+The module hwlat_detector is a special purpose kernel module that is used to -+detect large system latencies induced by the behavior of certain underlying -+hardware or firmware, independent of Linux itself. The code was developed -+originally to detect SMIs (System Management Interrupts) on x86 systems, -+however there is nothing x86 specific about this patchset. It was -+originally written for use by the "RT" patch since the Real Time -+kernel is highly latency sensitive. -+ -+SMIs are usually not serviced by the Linux kernel, which typically does not -+even know that they are occuring. SMIs are instead are set up by BIOS code -+and are serviced by BIOS code, usually for "critical" events such as -+management of thermal sensors and fans. Sometimes though, SMIs are used for -+other tasks and those tasks can spend an inordinate amount of time in the -+handler (sometimes measured in milliseconds). Obviously this is a problem if -+you are trying to keep event service latencies down in the microsecond range. -+ -+The hardware latency detector works by hogging all of the cpus for configurable -+amounts of time (by calling stop_machine()), polling the CPU Time Stamp Counter -+for some period, then looking for gaps in the TSC data. Any gap indicates a -+time when the polling was interrupted and since the machine is stopped and -+interrupts turned off the only thing that could do that would be an SMI. -+ -+Note that the SMI detector should *NEVER* be used in a production environment. -+It is intended to be run manually to determine if the hardware platform has a -+problem with long system firmware service routines. -+ -+Usage: -+------ -+ -+Loading the module hwlat_detector passing the parameter "enabled=1" (or by -+setting the "enable" entry in "hwlat_detector" debugfs toggled on) is the only -+step required to start the hwlat_detector. It is possible to redefine the -+threshold in microseconds (us) above which latency spikes will be taken -+into account (parameter "threshold="). -+ -+Example: -+ -+ # modprobe hwlat_detector enabled=1 threshold=100 -+ -+After the module is loaded, it creates a directory named "hwlat_detector" under -+the debugfs mountpoint, "/debug/hwlat_detector" for this text. It is necessary -+to have debugfs mounted, which might be on /sys/debug on your system. -+ -+The /debug/hwlat_detector interface contains the following files: -+ -+count - number of latency spikes observed since last reset -+enable - a global enable/disable toggle (0/1), resets count -+max - maximum hardware latency actually observed (usecs) -+sample - a pipe from which to read current raw sample data -+ in the format -+ (can be opened O_NONBLOCK for a single sample) -+threshold - minimum latency value to be considered (usecs) -+width - time period to sample with CPUs held (usecs) -+ must be less than the total window size (enforced) -+window - total period of sampling, width being inside (usecs) -+ -+By default we will set width to 500,000 and window to 1,000,000, meaning that -+we will sample every 1,000,000 usecs (1s) for 500,000 usecs (0.5s). If we -+observe any latencies that exceed the threshold (initially 100 usecs), -+then we write to a global sample ring buffer of 8K samples, which is -+consumed by reading from the "sample" (pipe) debugfs file interface. -diff -Nur linux-4.4.62.orig/Documentation/kernel-parameters.txt linux-4.4.62/Documentation/kernel-parameters.txt ---- linux-4.4.62.orig/Documentation/kernel-parameters.txt 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/Documentation/kernel-parameters.txt 2017-04-18 17:38:07.902637922 +0200 -@@ -1640,6 +1640,15 @@ - ip= [IP_PNP] - See Documentation/filesystems/nfs/nfsroot.txt. +@@ -1901,9 +1901,9 @@ static void __init init_per_cpu_tunables(void) + bcp->cong_reps = congested_reps; + bcp->disabled_period = sec_2_cycles(disabled_period); + bcp->giveup_limit = giveup_limit; +- spin_lock_init(&bcp->queue_lock); +- spin_lock_init(&bcp->uvhub_lock); +- spin_lock_init(&bcp->disable_lock); ++ raw_spin_lock_init(&bcp->queue_lock); ++ raw_spin_lock_init(&bcp->uvhub_lock); ++ raw_spin_lock_init(&bcp->disable_lock); + } + } -+ irqaffinity= [SMP] Set the default irq affinity mask -+ Format: -+ ,..., -+ or -+ - -+ (must be a positive range in ascending order) -+ or a mixture -+ ,...,- +diff --git a/arch/x86/platform/uv/uv_time.c b/arch/x86/platform/uv/uv_time.c +index 2b158a9fa1d7..5e0b122620cb 100644 +--- a/arch/x86/platform/uv/uv_time.c ++++ b/arch/x86/platform/uv/uv_time.c +@@ -57,7 +57,7 @@ static DEFINE_PER_CPU(struct clock_event_device, cpu_ced); + + /* There is one of these allocated per node */ + struct uv_rtc_timer_head { +- spinlock_t lock; ++ raw_spinlock_t lock; + /* next cpu waiting for timer, local node relative: */ + int next_cpu; + /* number of cpus on this node: */ +@@ -177,7 +177,7 @@ static __init int uv_rtc_allocate_timers(void) + uv_rtc_deallocate_timers(); + return -ENOMEM; + } +- spin_lock_init(&head->lock); ++ raw_spin_lock_init(&head->lock); + head->ncpus = uv_blade_nr_possible_cpus(bid); + head->next_cpu = -1; + blade_info[bid] = head; +@@ -231,7 +231,7 @@ static int uv_rtc_set_timer(int cpu, u64 expires) + unsigned long flags; + int next_cpu; + +- spin_lock_irqsave(&head->lock, flags); ++ raw_spin_lock_irqsave(&head->lock, flags); + + next_cpu = head->next_cpu; + *t = expires; +@@ -243,12 +243,12 @@ static int uv_rtc_set_timer(int cpu, u64 expires) + if (uv_setup_intr(cpu, expires)) { + *t = ULLONG_MAX; + uv_rtc_find_next_timer(head, pnode); +- spin_unlock_irqrestore(&head->lock, flags); ++ raw_spin_unlock_irqrestore(&head->lock, flags); + return -ETIME; + } + } + +- spin_unlock_irqrestore(&head->lock, flags); ++ raw_spin_unlock_irqrestore(&head->lock, flags); + return 0; + } + +@@ -267,7 +267,7 @@ static int uv_rtc_unset_timer(int cpu, int force) + unsigned long flags; + int rc = 0; + +- spin_lock_irqsave(&head->lock, flags); ++ raw_spin_lock_irqsave(&head->lock, flags); + + if ((head->next_cpu == bcpu && uv_read_rtc(NULL) >= *t) || force) + rc = 1; +@@ -279,7 +279,7 @@ static int uv_rtc_unset_timer(int cpu, int force) + uv_rtc_find_next_timer(head, pnode); + } + +- spin_unlock_irqrestore(&head->lock, flags); ++ raw_spin_unlock_irqrestore(&head->lock, flags); + + return rc; + } +@@ -299,13 +299,18 @@ static int uv_rtc_unset_timer(int cpu, int force) + static cycle_t uv_read_rtc(struct clocksource *cs) + { + unsigned long offset; ++ cycle_t cycles; + ++ preempt_disable(); + if (uv_get_min_hub_revision_id() == 1) + offset = 0; + else + offset = (uv_blade_processor_id() * L1_CACHE_BYTES) % PAGE_SIZE; + +- return (cycle_t)uv_read_local_mmr(UVH_RTC | offset); ++ cycles = (cycle_t)uv_read_local_mmr(UVH_RTC | offset); ++ preempt_enable(); + - irqfixup [HW] - When an interrupt is not handled search all handlers - for it. Intended to get systems with badly broken -diff -Nur linux-4.4.62.orig/Documentation/sysrq.txt linux-4.4.62/Documentation/sysrq.txt ---- linux-4.4.62.orig/Documentation/sysrq.txt 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/Documentation/sysrq.txt 2017-04-18 17:38:07.902637922 +0200 -@@ -59,10 +59,17 @@ - On other - If you know of the key combos for other architectures, please - let me know so I can add them to this section. ++ return cycles; + } --On all - write a character to /proc/sysrq-trigger. e.g.: + /* +diff --git a/block/blk-core.c b/block/blk-core.c +index ef083e7a37c5..7a7926b1b4b5 100644 +--- a/block/blk-core.c ++++ b/block/blk-core.c +@@ -125,6 +125,9 @@ void blk_rq_init(struct request_queue *q, struct request *rq) + + INIT_LIST_HEAD(&rq->queuelist); + INIT_LIST_HEAD(&rq->timeout_list); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ INIT_WORK(&rq->work, __blk_mq_complete_request_remote_work); ++#endif + rq->cpu = -1; + rq->q = q; + rq->__sector = (sector_t) -1; +@@ -233,7 +236,7 @@ EXPORT_SYMBOL(blk_start_queue_async); + **/ + void blk_start_queue(struct request_queue *q) + { +- WARN_ON(!irqs_disabled()); ++ WARN_ON_NONRT(!irqs_disabled()); + + queue_flag_clear(QUEUE_FLAG_STOPPED, q); + __blk_run_queue(q); +@@ -659,7 +662,7 @@ int blk_queue_enter(struct request_queue *q, gfp_t gfp) + if (!gfpflags_allow_blocking(gfp)) + return -EBUSY; + +- ret = wait_event_interruptible(q->mq_freeze_wq, ++ ret = swait_event_interruptible(q->mq_freeze_wq, + !atomic_read(&q->mq_freeze_depth) || + blk_queue_dying(q)); + if (blk_queue_dying(q)) +@@ -679,7 +682,7 @@ static void blk_queue_usage_counter_release(struct percpu_ref *ref) + struct request_queue *q = + container_of(ref, struct request_queue, q_usage_counter); + +- wake_up_all(&q->mq_freeze_wq); ++ swake_up_all(&q->mq_freeze_wq); + } + + struct request_queue *blk_alloc_queue_node(gfp_t gfp_mask, int node_id) +@@ -741,7 +744,7 @@ struct request_queue *blk_alloc_queue_node(gfp_t gfp_mask, int node_id) + q->bypass_depth = 1; + __set_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); + +- init_waitqueue_head(&q->mq_freeze_wq); ++ init_swait_queue_head(&q->mq_freeze_wq); + + /* + * Init percpu_ref in atomic mode so that it's faster to shutdown. +@@ -3222,7 +3225,7 @@ static void queue_unplugged(struct request_queue *q, unsigned int depth, + blk_run_queue_async(q); + else + __blk_run_queue(q); +- spin_unlock(q->queue_lock); ++ spin_unlock_irq(q->queue_lock); + } + + static void flush_plug_callbacks(struct blk_plug *plug, bool from_schedule) +@@ -3270,7 +3273,6 @@ EXPORT_SYMBOL(blk_check_plugged); + void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) + { + struct request_queue *q; +- unsigned long flags; + struct request *rq; + LIST_HEAD(list); + unsigned int depth; +@@ -3290,11 +3292,6 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) + q = NULL; + depth = 0; + +- /* +- * Save and disable interrupts here, to avoid doing it for every +- * queue lock we have to take. +- */ +- local_irq_save(flags); + while (!list_empty(&list)) { + rq = list_entry_rq(list.next); + list_del_init(&rq->queuelist); +@@ -3307,7 +3304,7 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) + queue_unplugged(q, depth, from_schedule); + q = rq->q; + depth = 0; +- spin_lock(q->queue_lock); ++ spin_lock_irq(q->queue_lock); + } + + /* +@@ -3334,8 +3331,6 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) + */ + if (q) + queue_unplugged(q, depth, from_schedule); - -+On all - write a character to /proc/sysrq-trigger, e.g.: - echo t > /proc/sysrq-trigger +- local_irq_restore(flags); + } -+On all - Enable network SysRq by writing a cookie to icmp_echo_sysrq, e.g. -+ echo 0x01020304 >/proc/sys/net/ipv4/icmp_echo_sysrq -+ Send an ICMP echo request with this pattern plus the particular -+ SysRq command key. Example: -+ # ping -c1 -s57 -p0102030468 -+ will trigger the SysRq-H (help) command. -+ -+ - * What are the 'command' keys? - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - 'b' - Will immediately reboot the system without syncing or unmounting -diff -Nur linux-4.4.62.orig/Documentation/trace/histograms.txt linux-4.4.62/Documentation/trace/histograms.txt ---- linux-4.4.62.orig/Documentation/trace/histograms.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/Documentation/trace/histograms.txt 2017-04-18 17:38:07.902637922 +0200 -@@ -0,0 +1,186 @@ -+ Using the Linux Kernel Latency Histograms -+ -+ -+This document gives a short explanation how to enable, configure and use -+latency histograms. Latency histograms are primarily relevant in the -+context of real-time enabled kernels (CONFIG_PREEMPT/CONFIG_PREEMPT_RT) -+and are used in the quality management of the Linux real-time -+capabilities. -+ -+ -+* Purpose of latency histograms -+ -+A latency histogram continuously accumulates the frequencies of latency -+data. There are two types of histograms -+- potential sources of latencies -+- effective latencies -+ -+ -+* Potential sources of latencies -+ -+Potential sources of latencies are code segments where interrupts, -+preemption or both are disabled (aka critical sections). To create -+histograms of potential sources of latency, the kernel stores the time -+stamp at the start of a critical section, determines the time elapsed -+when the end of the section is reached, and increments the frequency -+counter of that latency value - irrespective of whether any concurrently -+running process is affected by latency or not. -+- Configuration items (in the Kernel hacking/Tracers submenu) -+ CONFIG_INTERRUPT_OFF_LATENCY -+ CONFIG_PREEMPT_OFF_LATENCY -+ -+ -+* Effective latencies -+ -+Effective latencies are actually occuring during wakeup of a process. To -+determine effective latencies, the kernel stores the time stamp when a -+process is scheduled to be woken up, and determines the duration of the -+wakeup time shortly before control is passed over to this process. Note -+that the apparent latency in user space may be somewhat longer, since the -+process may be interrupted after control is passed over to it but before -+the execution in user space takes place. Simply measuring the interval -+between enqueuing and wakeup may also not appropriate in cases when a -+process is scheduled as a result of a timer expiration. The timer may have -+missed its deadline, e.g. due to disabled interrupts, but this latency -+would not be registered. Therefore, the offsets of missed timers are -+recorded in a separate histogram. If both wakeup latency and missed timer -+offsets are configured and enabled, a third histogram may be enabled that -+records the overall latency as a sum of the timer latency, if any, and the -+wakeup latency. This histogram is called "timerandwakeup". -+- Configuration items (in the Kernel hacking/Tracers submenu) -+ CONFIG_WAKEUP_LATENCY -+ CONFIG_MISSED_TIMER_OFSETS -+ -+ -+* Usage -+ -+The interface to the administration of the latency histograms is located -+in the debugfs file system. To mount it, either enter -+ -+mount -t sysfs nodev /sys -+mount -t debugfs nodev /sys/kernel/debug -+ -+from shell command line level, or add -+ -+nodev /sys sysfs defaults 0 0 -+nodev /sys/kernel/debug debugfs defaults 0 0 -+ -+to the file /etc/fstab. All latency histogram related files are then -+available in the directory /sys/kernel/debug/tracing/latency_hist. A -+particular histogram type is enabled by writing non-zero to the related -+variable in the /sys/kernel/debug/tracing/latency_hist/enable directory. -+Select "preemptirqsoff" for the histograms of potential sources of -+latencies and "wakeup" for histograms of effective latencies etc. The -+histogram data - one per CPU - are available in the files -+ -+/sys/kernel/debug/tracing/latency_hist/preemptoff/CPUx -+/sys/kernel/debug/tracing/latency_hist/irqsoff/CPUx -+/sys/kernel/debug/tracing/latency_hist/preemptirqsoff/CPUx -+/sys/kernel/debug/tracing/latency_hist/wakeup/CPUx -+/sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio/CPUx -+/sys/kernel/debug/tracing/latency_hist/missed_timer_offsets/CPUx -+/sys/kernel/debug/tracing/latency_hist/timerandwakeup/CPUx -+ -+The histograms are reset by writing non-zero to the file "reset" in a -+particular latency directory. To reset all latency data, use -+ -+#!/bin/sh -+ -+TRACINGDIR=/sys/kernel/debug/tracing -+HISTDIR=$TRACINGDIR/latency_hist -+ -+if test -d $HISTDIR -+then -+ cd $HISTDIR -+ for i in `find . | grep /reset$` -+ do -+ echo 1 >$i -+ done -+fi -+ -+ -+* Data format -+ -+Latency data are stored with a resolution of one microsecond. The -+maximum latency is 10,240 microseconds. The data are only valid, if the -+overflow register is empty. Every output line contains the latency in -+microseconds in the first row and the number of samples in the second -+row. To display only lines with a positive latency count, use, for -+example, -+ -+grep -v " 0$" /sys/kernel/debug/tracing/latency_hist/preemptoff/CPU0 -+ -+#Minimum latency: 0 microseconds. -+#Average latency: 0 microseconds. -+#Maximum latency: 25 microseconds. -+#Total samples: 3104770694 -+#There are 0 samples greater or equal than 10240 microseconds -+#usecs samples -+ 0 2984486876 -+ 1 49843506 -+ 2 58219047 -+ 3 5348126 -+ 4 2187960 -+ 5 3388262 -+ 6 959289 -+ 7 208294 -+ 8 40420 -+ 9 4485 -+ 10 14918 -+ 11 18340 -+ 12 25052 -+ 13 19455 -+ 14 5602 -+ 15 969 -+ 16 47 -+ 17 18 -+ 18 14 -+ 19 1 -+ 20 3 -+ 21 2 -+ 22 5 -+ 23 2 -+ 25 1 -+ -+ -+* Wakeup latency of a selected process -+ -+To only collect wakeup latency data of a particular process, write the -+PID of the requested process to -+ -+/sys/kernel/debug/tracing/latency_hist/wakeup/pid -+ -+PIDs are not considered, if this variable is set to 0. -+ -+ -+* Details of the process with the highest wakeup latency so far -+ -+Selected data of the process that suffered from the highest wakeup -+latency that occurred in a particular CPU are available in the file + void blk_finish_plug(struct blk_plug *plug) +diff --git a/block/blk-ioc.c b/block/blk-ioc.c +index 381cb50a673c..dc8785233d94 100644 +--- a/block/blk-ioc.c ++++ b/block/blk-ioc.c +@@ -7,6 +7,7 @@ + #include + #include + #include ++#include + + #include "blk.h" + +@@ -109,7 +110,7 @@ static void ioc_release_fn(struct work_struct *work) + spin_unlock(q->queue_lock); + } else { + spin_unlock_irqrestore(&ioc->lock, flags); +- cpu_relax(); ++ cpu_chill(); + spin_lock_irqsave_nested(&ioc->lock, flags, 1); + } + } +@@ -187,7 +188,7 @@ retry: + spin_unlock(icq->q->queue_lock); + } else { + spin_unlock_irqrestore(&ioc->lock, flags); +- cpu_relax(); ++ cpu_chill(); + goto retry; + } + } +diff --git a/block/blk-iopoll.c b/block/blk-iopoll.c +index 0736729d6494..3e21e31d0d7e 100644 +--- a/block/blk-iopoll.c ++++ b/block/blk-iopoll.c +@@ -35,6 +35,7 @@ void blk_iopoll_sched(struct blk_iopoll *iop) + list_add_tail(&iop->list, this_cpu_ptr(&blk_cpu_iopoll)); + __raise_softirq_irqoff(BLOCK_IOPOLL_SOFTIRQ); + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + EXPORT_SYMBOL(blk_iopoll_sched); + +@@ -132,6 +133,7 @@ static void blk_iopoll_softirq(struct softirq_action *h) + __raise_softirq_irqoff(BLOCK_IOPOLL_SOFTIRQ); + + local_irq_enable(); ++ preempt_check_resched_rt(); + } + + /** +@@ -201,6 +203,7 @@ static int blk_iopoll_cpu_notify(struct notifier_block *self, + this_cpu_ptr(&blk_cpu_iopoll)); + __raise_softirq_irqoff(BLOCK_IOPOLL_SOFTIRQ); + local_irq_enable(); ++ preempt_check_resched_rt(); + } + + return NOTIFY_OK; +diff --git a/block/blk-mq-cpu.c b/block/blk-mq-cpu.c +index bb3ed488f7b5..628c6c13c482 100644 +--- a/block/blk-mq-cpu.c ++++ b/block/blk-mq-cpu.c +@@ -16,7 +16,7 @@ + #include "blk-mq.h" + + static LIST_HEAD(blk_mq_cpu_notify_list); +-static DEFINE_RAW_SPINLOCK(blk_mq_cpu_notify_lock); ++static DEFINE_SPINLOCK(blk_mq_cpu_notify_lock); + + static int blk_mq_main_cpu_notify(struct notifier_block *self, + unsigned long action, void *hcpu) +@@ -25,7 +25,10 @@ static int blk_mq_main_cpu_notify(struct notifier_block *self, + struct blk_mq_cpu_notifier *notify; + int ret = NOTIFY_OK; + +- raw_spin_lock(&blk_mq_cpu_notify_lock); ++ if (action != CPU_POST_DEAD) ++ return NOTIFY_OK; + -+/sys/kernel/debug/tracing/latency_hist/wakeup/max_latency-CPUx. ++ spin_lock(&blk_mq_cpu_notify_lock); + + list_for_each_entry(notify, &blk_mq_cpu_notify_list, list) { + ret = notify->notify(notify->data, action, cpu); +@@ -33,7 +36,7 @@ static int blk_mq_main_cpu_notify(struct notifier_block *self, + break; + } + +- raw_spin_unlock(&blk_mq_cpu_notify_lock); ++ spin_unlock(&blk_mq_cpu_notify_lock); + return ret; + } + +@@ -41,16 +44,16 @@ void blk_mq_register_cpu_notifier(struct blk_mq_cpu_notifier *notifier) + { + BUG_ON(!notifier->notify); + +- raw_spin_lock(&blk_mq_cpu_notify_lock); ++ spin_lock(&blk_mq_cpu_notify_lock); + list_add_tail(¬ifier->list, &blk_mq_cpu_notify_list); +- raw_spin_unlock(&blk_mq_cpu_notify_lock); ++ spin_unlock(&blk_mq_cpu_notify_lock); + } + + void blk_mq_unregister_cpu_notifier(struct blk_mq_cpu_notifier *notifier) + { +- raw_spin_lock(&blk_mq_cpu_notify_lock); ++ spin_lock(&blk_mq_cpu_notify_lock); + list_del(¬ifier->list); +- raw_spin_unlock(&blk_mq_cpu_notify_lock); ++ spin_unlock(&blk_mq_cpu_notify_lock); + } + + void blk_mq_init_cpu_notifier(struct blk_mq_cpu_notifier *notifier, +diff --git a/block/blk-mq.c b/block/blk-mq.c +index 0d1af3e44efb..e4fc80184dd8 100644 +--- a/block/blk-mq.c ++++ b/block/blk-mq.c +@@ -92,7 +92,7 @@ EXPORT_SYMBOL_GPL(blk_mq_freeze_queue_start); + + static void blk_mq_freeze_queue_wait(struct request_queue *q) + { +- wait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->q_usage_counter)); ++ swait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->q_usage_counter)); + } + + /* +@@ -130,7 +130,7 @@ void blk_mq_unfreeze_queue(struct request_queue *q) + WARN_ON_ONCE(freeze_depth < 0); + if (!freeze_depth) { + percpu_ref_reinit(&q->q_usage_counter); +- wake_up_all(&q->mq_freeze_wq); ++ swake_up_all(&q->mq_freeze_wq); + } + } + EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue); +@@ -149,7 +149,7 @@ void blk_mq_wake_waiters(struct request_queue *q) + * dying, we need to ensure that processes currently waiting on + * the queue are notified as well. + */ +- wake_up_all(&q->mq_freeze_wq); ++ swake_up_all(&q->mq_freeze_wq); + } + + bool blk_mq_can_queue(struct blk_mq_hw_ctx *hctx) +@@ -196,6 +196,9 @@ static void blk_mq_rq_ctx_init(struct request_queue *q, struct blk_mq_ctx *ctx, + rq->resid_len = 0; + rq->sense = NULL; + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ INIT_WORK(&rq->work, __blk_mq_complete_request_remote_work); ++#endif + INIT_LIST_HEAD(&rq->timeout_list); + rq->timeout = 0; + +@@ -325,6 +328,17 @@ void blk_mq_end_request(struct request *rq, int error) + } + EXPORT_SYMBOL(blk_mq_end_request); + ++#ifdef CONFIG_PREEMPT_RT_FULL + -+In addition, other relevant system data at the time when the -+latency occurred are given. ++void __blk_mq_complete_request_remote_work(struct work_struct *work) ++{ ++ struct request *rq = container_of(work, struct request, work); + -+The format of the data is (all in one line): -+ () \ -+<- ++ rq->q->softirq_done_fn(rq); ++} + -+The value of is only relevant in the combined timer -+and wakeup latency recording. In the wakeup recording, it is -+always 0, in the missed_timer_offsets recording, it is the same -+as . ++#else + -+When retrospectively searching for the origin of a latency and -+tracing was not enabled, it may be helpful to know the name and -+some basic data of the task that (finally) was switching to the -+late real-tlme task. In addition to the victim's data, also the -+data of the possible culprit are therefore displayed after the -+"<-" symbol. + static void __blk_mq_complete_request_remote(void *data) + { + struct request *rq = data; +@@ -332,6 +346,8 @@ static void __blk_mq_complete_request_remote(void *data) + rq->q->softirq_done_fn(rq); + } + ++#endif + -+Finally, the timestamp of the time when the latency occurred -+in . after the most recent system boot -+is provided. + static void blk_mq_ipi_complete_request(struct request *rq) + { + struct blk_mq_ctx *ctx = rq->mq_ctx; +@@ -343,19 +359,23 @@ static void blk_mq_ipi_complete_request(struct request *rq) + return; + } + +- cpu = get_cpu(); ++ cpu = get_cpu_light(); + if (!test_bit(QUEUE_FLAG_SAME_FORCE, &rq->q->queue_flags)) + shared = cpus_share_cache(cpu, ctx->cpu); + + if (cpu != ctx->cpu && !shared && cpu_online(ctx->cpu)) { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ schedule_work_on(ctx->cpu, &rq->work); ++#else + rq->csd.func = __blk_mq_complete_request_remote; + rq->csd.info = rq; + rq->csd.flags = 0; + smp_call_function_single_async(ctx->cpu, &rq->csd); ++#endif + } else { + rq->q->softirq_done_fn(rq); + } +- put_cpu(); ++ put_cpu_light(); + } + + static void __blk_mq_complete_request(struct request *rq) +@@ -862,14 +882,14 @@ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async) + return; + + if (!async) { +- int cpu = get_cpu(); ++ int cpu = get_cpu_light(); + if (cpumask_test_cpu(cpu, hctx->cpumask)) { + __blk_mq_run_hw_queue(hctx); +- put_cpu(); ++ put_cpu_light(); + return; + } + +- put_cpu(); ++ put_cpu_light(); + } + + kblockd_schedule_delayed_work_on(blk_mq_hctx_next_cpu(hctx), +@@ -1616,7 +1636,7 @@ static int blk_mq_hctx_notify(void *data, unsigned long action, + { + struct blk_mq_hw_ctx *hctx = data; + +- if (action == CPU_DEAD || action == CPU_DEAD_FROZEN) ++ if (action == CPU_POST_DEAD) + return blk_mq_hctx_cpu_offline(hctx, cpu); + + /* +diff --git a/block/blk-mq.h b/block/blk-mq.h +index 713820b47b31..3cb6feb4fe23 100644 +--- a/block/blk-mq.h ++++ b/block/blk-mq.h +@@ -74,7 +74,10 @@ struct blk_align_bitmap { + static inline struct blk_mq_ctx *__blk_mq_get_ctx(struct request_queue *q, + unsigned int cpu) + { +- return per_cpu_ptr(q->queue_ctx, cpu); ++ struct blk_mq_ctx *ctx; + -+These data are also reset when the wakeup histogram is reset. -diff -Nur linux-4.4.62.orig/drivers/acpi/acpica/acglobal.h linux-4.4.62/drivers/acpi/acpica/acglobal.h ---- linux-4.4.62.orig/drivers/acpi/acpica/acglobal.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/acpi/acpica/acglobal.h 2017-04-18 17:38:08.042643350 +0200 -@@ -116,7 +116,7 @@ ++ ctx = per_cpu_ptr(q->queue_ctx, cpu); ++ return ctx; + } + + /* +@@ -85,12 +88,12 @@ static inline struct blk_mq_ctx *__blk_mq_get_ctx(struct request_queue *q, + */ + static inline struct blk_mq_ctx *blk_mq_get_ctx(struct request_queue *q) + { +- return __blk_mq_get_ctx(q, get_cpu()); ++ return __blk_mq_get_ctx(q, get_cpu_light()); + } + + static inline void blk_mq_put_ctx(struct blk_mq_ctx *ctx) + { +- put_cpu(); ++ put_cpu_light(); + } + + struct blk_mq_alloc_data { +diff --git a/block/blk-softirq.c b/block/blk-softirq.c +index 53b1737e978d..81c3c0a62edf 100644 +--- a/block/blk-softirq.c ++++ b/block/blk-softirq.c +@@ -51,6 +51,7 @@ static void trigger_softirq(void *data) + raise_softirq_irqoff(BLOCK_SOFTIRQ); + + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + + /* +@@ -93,6 +94,7 @@ static int blk_cpu_notify(struct notifier_block *self, unsigned long action, + this_cpu_ptr(&blk_cpu_done)); + raise_softirq_irqoff(BLOCK_SOFTIRQ); + local_irq_enable(); ++ preempt_check_resched_rt(); + } + + return NOTIFY_OK; +@@ -150,6 +152,7 @@ do_local: + goto do_local; + + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + + /** +diff --git a/block/bounce.c b/block/bounce.c +index 1cb5dd3a5da1..2f1ec8a67cbe 100644 +--- a/block/bounce.c ++++ b/block/bounce.c +@@ -55,11 +55,11 @@ static void bounce_copy_vec(struct bio_vec *to, unsigned char *vfrom) + unsigned long flags; + unsigned char *vto; + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + vto = kmap_atomic(to->bv_page); + memcpy(vto + to->bv_offset, vfrom, to->bv_len); + kunmap_atomic(vto); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + + #else /* CONFIG_HIGHMEM */ +diff --git a/crypto/algapi.c b/crypto/algapi.c +index 43f5bdb6b570..b54fd9187d12 100644 +--- a/crypto/algapi.c ++++ b/crypto/algapi.c +@@ -720,13 +720,13 @@ EXPORT_SYMBOL_GPL(crypto_spawn_tfm2); + + int crypto_register_notifier(struct notifier_block *nb) + { +- return blocking_notifier_chain_register(&crypto_chain, nb); ++ return srcu_notifier_chain_register(&crypto_chain, nb); + } + EXPORT_SYMBOL_GPL(crypto_register_notifier); + + int crypto_unregister_notifier(struct notifier_block *nb) + { +- return blocking_notifier_chain_unregister(&crypto_chain, nb); ++ return srcu_notifier_chain_unregister(&crypto_chain, nb); + } + EXPORT_SYMBOL_GPL(crypto_unregister_notifier); + +diff --git a/crypto/api.c b/crypto/api.c +index bbc147cb5dec..bc1a848f02ec 100644 +--- a/crypto/api.c ++++ b/crypto/api.c +@@ -31,7 +31,7 @@ EXPORT_SYMBOL_GPL(crypto_alg_list); + DECLARE_RWSEM(crypto_alg_sem); + EXPORT_SYMBOL_GPL(crypto_alg_sem); + +-BLOCKING_NOTIFIER_HEAD(crypto_chain); ++SRCU_NOTIFIER_HEAD(crypto_chain); + EXPORT_SYMBOL_GPL(crypto_chain); + + static struct crypto_alg *crypto_larval_wait(struct crypto_alg *alg); +@@ -236,10 +236,10 @@ int crypto_probing_notify(unsigned long val, void *v) + { + int ok; + +- ok = blocking_notifier_call_chain(&crypto_chain, val, v); ++ ok = srcu_notifier_call_chain(&crypto_chain, val, v); + if (ok == NOTIFY_DONE) { + request_module("cryptomgr"); +- ok = blocking_notifier_call_chain(&crypto_chain, val, v); ++ ok = srcu_notifier_call_chain(&crypto_chain, val, v); + } + + return ok; +diff --git a/crypto/internal.h b/crypto/internal.h +index 00e42a3ed814..2e85551e235f 100644 +--- a/crypto/internal.h ++++ b/crypto/internal.h +@@ -47,7 +47,7 @@ struct crypto_larval { + + extern struct list_head crypto_alg_list; + extern struct rw_semaphore crypto_alg_sem; +-extern struct blocking_notifier_head crypto_chain; ++extern struct srcu_notifier_head crypto_chain; + + #ifdef CONFIG_PROC_FS + void __init crypto_init_proc(void); +@@ -143,7 +143,7 @@ static inline int crypto_is_moribund(struct crypto_alg *alg) + + static inline void crypto_notify(unsigned long val, void *v) + { +- blocking_notifier_call_chain(&crypto_chain, val, v); ++ srcu_notifier_call_chain(&crypto_chain, val, v); + } + + #endif /* _CRYPTO_INTERNAL_H */ +diff --git a/drivers/acpi/acpica/acglobal.h b/drivers/acpi/acpica/acglobal.h +index faa97604d878..941497f31cf0 100644 +--- a/drivers/acpi/acpica/acglobal.h ++++ b/drivers/acpi/acpica/acglobal.h +@@ -116,7 +116,7 @@ ACPI_GLOBAL(u8, acpi_gbl_global_lock_pending); * interrupt level */ ACPI_GLOBAL(acpi_spinlock, acpi_gbl_gpe_lock); /* For GPE data structs and registers */ @@ -4248,10 +4367,11 @@ diff -Nur linux-4.4.62.orig/drivers/acpi/acpica/acglobal.h linux-4.4.62/drivers/ ACPI_GLOBAL(acpi_spinlock, acpi_gbl_reference_count_lock); /* Mutex for _OSI support */ -diff -Nur linux-4.4.62.orig/drivers/acpi/acpica/hwregs.c linux-4.4.62/drivers/acpi/acpica/hwregs.c ---- linux-4.4.62.orig/drivers/acpi/acpica/hwregs.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/acpi/acpica/hwregs.c 2017-04-18 17:38:08.042643350 +0200 -@@ -269,14 +269,14 @@ +diff --git a/drivers/acpi/acpica/hwregs.c b/drivers/acpi/acpica/hwregs.c +index 3cf77afd142c..dc32e72132f1 100644 +--- a/drivers/acpi/acpica/hwregs.c ++++ b/drivers/acpi/acpica/hwregs.c +@@ -269,14 +269,14 @@ acpi_status acpi_hw_clear_acpi_status(void) ACPI_BITMASK_ALL_FIXED_STATUS, ACPI_FORMAT_UINT64(acpi_gbl_xpm1a_status.address))); @@ -4268,10 +4388,11 @@ diff -Nur linux-4.4.62.orig/drivers/acpi/acpica/hwregs.c linux-4.4.62/drivers/ac if (ACPI_FAILURE(status)) { goto exit; -diff -Nur linux-4.4.62.orig/drivers/acpi/acpica/hwxface.c linux-4.4.62/drivers/acpi/acpica/hwxface.c ---- linux-4.4.62.orig/drivers/acpi/acpica/hwxface.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/acpi/acpica/hwxface.c 2017-04-18 17:38:08.042643350 +0200 -@@ -374,7 +374,7 @@ +diff --git a/drivers/acpi/acpica/hwxface.c b/drivers/acpi/acpica/hwxface.c +index b2e50d8007fe..ff007084dc48 100644 +--- a/drivers/acpi/acpica/hwxface.c ++++ b/drivers/acpi/acpica/hwxface.c +@@ -374,7 +374,7 @@ acpi_status acpi_write_bit_register(u32 register_id, u32 value) return_ACPI_STATUS(AE_BAD_PARAMETER); } @@ -4280,7 +4401,7 @@ diff -Nur linux-4.4.62.orig/drivers/acpi/acpica/hwxface.c linux-4.4.62/drivers/a /* * At this point, we know that the parent register is one of the -@@ -435,7 +435,7 @@ +@@ -435,7 +435,7 @@ acpi_status acpi_write_bit_register(u32 register_id, u32 value) unlock_and_exit: @@ -4289,10 +4410,11 @@ diff -Nur linux-4.4.62.orig/drivers/acpi/acpica/hwxface.c linux-4.4.62/drivers/a return_ACPI_STATUS(status); } -diff -Nur linux-4.4.62.orig/drivers/acpi/acpica/utmutex.c linux-4.4.62/drivers/acpi/acpica/utmutex.c ---- linux-4.4.62.orig/drivers/acpi/acpica/utmutex.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/acpi/acpica/utmutex.c 2017-04-18 17:38:08.042643350 +0200 -@@ -88,7 +88,7 @@ +diff --git a/drivers/acpi/acpica/utmutex.c b/drivers/acpi/acpica/utmutex.c +index ce406e39b669..41a75eb3ae9d 100644 +--- a/drivers/acpi/acpica/utmutex.c ++++ b/drivers/acpi/acpica/utmutex.c +@@ -88,7 +88,7 @@ acpi_status acpi_ut_mutex_initialize(void) return_ACPI_STATUS (status); } @@ -4301,7 +4423,7 @@ diff -Nur linux-4.4.62.orig/drivers/acpi/acpica/utmutex.c linux-4.4.62/drivers/a if (ACPI_FAILURE (status)) { return_ACPI_STATUS (status); } -@@ -156,7 +156,7 @@ +@@ -156,7 +156,7 @@ void acpi_ut_mutex_terminate(void) /* Delete the spinlocks */ acpi_os_delete_lock(acpi_gbl_gpe_lock); @@ -4310,10 +4432,11 @@ diff -Nur linux-4.4.62.orig/drivers/acpi/acpica/utmutex.c linux-4.4.62/drivers/a acpi_os_delete_lock(acpi_gbl_reference_count_lock); /* Delete the reader/writer lock */ -diff -Nur linux-4.4.62.orig/drivers/ata/libata-sff.c linux-4.4.62/drivers/ata/libata-sff.c ---- linux-4.4.62.orig/drivers/ata/libata-sff.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/ata/libata-sff.c 2017-04-18 17:38:08.042643350 +0200 -@@ -678,9 +678,9 @@ +diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c +index 7dbba387d12a..65beb7abb4e7 100644 +--- a/drivers/ata/libata-sff.c ++++ b/drivers/ata/libata-sff.c +@@ -678,9 +678,9 @@ unsigned int ata_sff_data_xfer_noirq(struct ata_device *dev, unsigned char *buf, unsigned long flags; unsigned int consumed; @@ -4325,7 +4448,7 @@ diff -Nur linux-4.4.62.orig/drivers/ata/libata-sff.c linux-4.4.62/drivers/ata/li return consumed; } -@@ -719,7 +719,7 @@ +@@ -719,7 +719,7 @@ static void ata_pio_sector(struct ata_queued_cmd *qc) unsigned long flags; /* FIXME: use a bounce buffer */ @@ -4334,7 +4457,7 @@ diff -Nur linux-4.4.62.orig/drivers/ata/libata-sff.c linux-4.4.62/drivers/ata/li buf = kmap_atomic(page); /* do the actual data transfer */ -@@ -727,7 +727,7 @@ +@@ -727,7 +727,7 @@ static void ata_pio_sector(struct ata_queued_cmd *qc) do_write); kunmap_atomic(buf); @@ -4343,7 +4466,7 @@ diff -Nur linux-4.4.62.orig/drivers/ata/libata-sff.c linux-4.4.62/drivers/ata/li } else { buf = page_address(page); ap->ops->sff_data_xfer(qc->dev, buf + offset, qc->sect_size, -@@ -864,7 +864,7 @@ +@@ -864,7 +864,7 @@ next_sg: unsigned long flags; /* FIXME: use bounce buffer */ @@ -4352,7 +4475,7 @@ diff -Nur linux-4.4.62.orig/drivers/ata/libata-sff.c linux-4.4.62/drivers/ata/li buf = kmap_atomic(page); /* do the actual data transfer */ -@@ -872,7 +872,7 @@ +@@ -872,7 +872,7 @@ next_sg: count, rw); kunmap_atomic(buf); @@ -4361,10 +4484,11 @@ diff -Nur linux-4.4.62.orig/drivers/ata/libata-sff.c linux-4.4.62/drivers/ata/li } else { buf = page_address(page); consumed = ap->ops->sff_data_xfer(dev, buf + offset, -diff -Nur linux-4.4.62.orig/drivers/block/zram/zram_drv.c linux-4.4.62/drivers/block/zram/zram_drv.c ---- linux-4.4.62.orig/drivers/block/zram/zram_drv.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/block/zram/zram_drv.c 2017-04-18 17:38:08.042643350 +0200 -@@ -520,6 +520,8 @@ +diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c +index 62a93b685c54..5ad3c4e652a8 100644 +--- a/drivers/block/zram/zram_drv.c ++++ b/drivers/block/zram/zram_drv.c +@@ -520,6 +520,8 @@ static struct zram_meta *zram_meta_alloc(char *pool_name, u64 disksize) goto out_error; } @@ -4373,7 +4497,7 @@ diff -Nur linux-4.4.62.orig/drivers/block/zram/zram_drv.c linux-4.4.62/drivers/b return meta; out_error: -@@ -568,12 +570,12 @@ +@@ -568,12 +570,12 @@ static int zram_decompress_page(struct zram *zram, char *mem, u32 index) unsigned long handle; size_t size; @@ -4385,10 +4509,10 @@ diff -Nur linux-4.4.62.orig/drivers/block/zram/zram_drv.c linux-4.4.62/drivers/b if (!handle || zram_test_flag(meta, index, ZRAM_ZERO)) { - bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); + zram_unlock_table(&meta->table[index]); - clear_page(mem); + memset(mem, 0, PAGE_SIZE); return 0; } -@@ -584,7 +586,7 @@ +@@ -584,7 +586,7 @@ static int zram_decompress_page(struct zram *zram, char *mem, u32 index) else ret = zcomp_decompress(zram->comp, cmem, size, mem); zs_unmap_object(meta->mem_pool, handle); @@ -4397,7 +4521,7 @@ diff -Nur linux-4.4.62.orig/drivers/block/zram/zram_drv.c linux-4.4.62/drivers/b /* Should NEVER happen. Return bio error if it does. */ if (unlikely(ret)) { -@@ -604,14 +606,14 @@ +@@ -604,14 +606,14 @@ static int zram_bvec_read(struct zram *zram, struct bio_vec *bvec, struct zram_meta *meta = zram->meta; page = bvec->bv_page; @@ -4415,7 +4539,7 @@ diff -Nur linux-4.4.62.orig/drivers/block/zram/zram_drv.c linux-4.4.62/drivers/b if (is_partial_io(bvec)) /* Use a temporary buffer to decompress the page */ -@@ -689,10 +691,10 @@ +@@ -689,10 +691,10 @@ static int zram_bvec_write(struct zram *zram, struct bio_vec *bvec, u32 index, if (user_mem) kunmap_atomic(user_mem); /* Free memory associated with this sector now. */ @@ -4428,7 +4552,7 @@ diff -Nur linux-4.4.62.orig/drivers/block/zram/zram_drv.c linux-4.4.62/drivers/b atomic64_inc(&zram->stats.zero_pages); ret = 0; -@@ -752,12 +754,12 @@ +@@ -752,12 +754,12 @@ static int zram_bvec_write(struct zram *zram, struct bio_vec *bvec, u32 index, * Free memory associated with this sector * before overwriting unused sectors. */ @@ -4443,7 +4567,7 @@ diff -Nur linux-4.4.62.orig/drivers/block/zram/zram_drv.c linux-4.4.62/drivers/b /* Update stats */ atomic64_add(clen, &zram->stats.compr_data_size); -@@ -800,9 +802,9 @@ +@@ -800,9 +802,9 @@ static void zram_bio_discard(struct zram *zram, u32 index, } while (n >= PAGE_SIZE) { @@ -4455,7 +4579,7 @@ diff -Nur linux-4.4.62.orig/drivers/block/zram/zram_drv.c linux-4.4.62/drivers/b atomic64_inc(&zram->stats.notify_free); index++; n -= PAGE_SIZE; -@@ -928,9 +930,9 @@ +@@ -928,9 +930,9 @@ static void zram_slot_free_notify(struct block_device *bdev, zram = bdev->bd_disk->private_data; meta = zram->meta; @@ -4467,10 +4591,11 @@ diff -Nur linux-4.4.62.orig/drivers/block/zram/zram_drv.c linux-4.4.62/drivers/b atomic64_inc(&zram->stats.notify_free); } -diff -Nur linux-4.4.62.orig/drivers/block/zram/zram_drv.h linux-4.4.62/drivers/block/zram/zram_drv.h ---- linux-4.4.62.orig/drivers/block/zram/zram_drv.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/block/zram/zram_drv.h 2017-04-18 17:38:08.042643350 +0200 -@@ -72,6 +72,9 @@ +diff --git a/drivers/block/zram/zram_drv.h b/drivers/block/zram/zram_drv.h +index 8e92339686d7..9e3e953d680e 100644 +--- a/drivers/block/zram/zram_drv.h ++++ b/drivers/block/zram/zram_drv.h +@@ -72,6 +72,9 @@ enum zram_pageflags { struct zram_table_entry { unsigned long handle; unsigned long value; @@ -4480,7 +4605,7 @@ diff -Nur linux-4.4.62.orig/drivers/block/zram/zram_drv.h linux-4.4.62/drivers/b }; struct zram_stats { -@@ -119,4 +122,42 @@ +@@ -119,4 +122,42 @@ struct zram { */ bool claim; /* Protected by bdev->bd_mutex */ }; @@ -4523,10 +4648,11 @@ diff -Nur linux-4.4.62.orig/drivers/block/zram/zram_drv.h linux-4.4.62/drivers/b +#endif /* CONFIG_PREEMPT_RT_BASE */ + #endif -diff -Nur linux-4.4.62.orig/drivers/char/random.c linux-4.4.62/drivers/char/random.c ---- linux-4.4.62.orig/drivers/char/random.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/char/random.c 2017-04-18 17:38:08.042643350 +0200 -@@ -799,8 +799,6 @@ +diff --git a/drivers/char/random.c b/drivers/char/random.c +index 491a4dce13fe..cf69b6b42208 100644 +--- a/drivers/char/random.c ++++ b/drivers/char/random.c +@@ -799,8 +799,6 @@ static void add_timer_randomness(struct timer_rand_state *state, unsigned num) } sample; long delta, delta2, delta3; @@ -4535,7 +4661,7 @@ diff -Nur linux-4.4.62.orig/drivers/char/random.c linux-4.4.62/drivers/char/rand sample.jiffies = jiffies; sample.cycles = random_get_entropy(); sample.num = num; -@@ -841,7 +839,6 @@ +@@ -841,7 +839,6 @@ static void add_timer_randomness(struct timer_rand_state *state, unsigned num) */ credit_entropy_bits(r, min_t(int, fls(delta>>1), 11)); } @@ -4543,7 +4669,7 @@ diff -Nur linux-4.4.62.orig/drivers/char/random.c linux-4.4.62/drivers/char/rand } void add_input_randomness(unsigned int type, unsigned int code, -@@ -894,28 +891,27 @@ +@@ -894,28 +891,27 @@ static __u32 get_reg(struct fast_pool *f, struct pt_regs *regs) return *(ptr + f->reg_idx++); } @@ -4577,9 +4703,10 @@ diff -Nur linux-4.4.62.orig/drivers/char/random.c linux-4.4.62/drivers/char/rand fast_mix(fast_pool); add_interrupt_bench(cycles); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-generated.c linux-4.4.62/drivers/clk/at91/clk-generated.c ---- linux-4.4.62.orig/drivers/clk/at91/clk-generated.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/clk-generated.c 2017-04-18 17:38:08.042643350 +0200 +diff --git a/drivers/clk/at91/clk-generated.c b/drivers/clk/at91/clk-generated.c +index abc80949e1dd..4ad3298eb372 100644 +--- a/drivers/clk/at91/clk-generated.c ++++ b/drivers/clk/at91/clk-generated.c @@ -15,8 +15,8 @@ #include #include @@ -4602,7 +4729,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-generated.c linux-4.4.62/driver u32 id; u32 gckdiv; u8 parent_id; -@@ -41,49 +42,52 @@ +@@ -41,49 +42,52 @@ struct clk_generated { static int clk_generated_enable(struct clk_hw *hw) { struct clk_generated *gck = to_clk_generated(hw); @@ -4641,13 +4768,14 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-generated.c linux-4.4.62/driver struct clk_generated *gck = to_clk_generated(hw); - struct at91_pmc *pmc = gck->pmc; - u32 tmp; -+ unsigned long flags; - +- - pmc_lock(pmc); - pmc_write(pmc, AT91_PMC_PCR, (gck->id & AT91_PMC_PCR_PID_MASK)); - tmp = pmc_read(pmc, AT91_PMC_PCR) & ~AT91_PMC_PCR_GCKEN; - pmc_write(pmc, AT91_PMC_PCR, tmp | AT91_PMC_PCR_CMD); - pmc_unlock(pmc); ++ unsigned long flags; ++ + spin_lock_irqsave(gck->lock, flags); + regmap_write(gck->regmap, AT91_PMC_PCR, + (gck->id & AT91_PMC_PCR_PID_MASK)); @@ -4680,7 +4808,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-generated.c linux-4.4.62/driver } static unsigned long -@@ -214,13 +218,14 @@ +@@ -214,13 +218,14 @@ static const struct clk_ops generated_ops = { */ static void clk_generated_startup(struct clk_generated *gck) { @@ -4700,7 +4828,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-generated.c linux-4.4.62/driver gck->parent_id = (tmp & AT91_PMC_PCR_GCKCSS_MASK) >> AT91_PMC_PCR_GCKCSS_OFFSET; -@@ -229,8 +234,8 @@ +@@ -229,8 +234,8 @@ static void clk_generated_startup(struct clk_generated *gck) } static struct clk * __init @@ -4711,7 +4839,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-generated.c linux-4.4.62/driver u8 id, const struct clk_range *range) { struct clk_generated *gck; -@@ -249,7 +254,8 @@ +@@ -249,7 +254,8 @@ at91_clk_register_generated(struct at91_pmc *pmc, const char *name, gck->id = id; gck->hw.init = &init; @@ -4721,7 +4849,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-generated.c linux-4.4.62/driver gck->range = *range; clk = clk_register(NULL, &gck->hw); -@@ -261,8 +267,7 @@ +@@ -261,8 +267,7 @@ at91_clk_register_generated(struct at91_pmc *pmc, const char *name, return clk; } @@ -4731,7 +4859,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-generated.c linux-4.4.62/driver { int num; u32 id; -@@ -272,6 +277,7 @@ +@@ -272,6 +277,7 @@ void __init of_sama5d2_clk_generated_setup(struct device_node *np, const char *parent_names[GENERATED_SOURCE_MAX]; struct device_node *gcknp; struct clk_range range = CLK_RANGE(0, 0); @@ -4739,7 +4867,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-generated.c linux-4.4.62/driver num_parents = of_clk_get_parent_count(np); if (num_parents <= 0 || num_parents > GENERATED_SOURCE_MAX) -@@ -283,6 +289,10 @@ +@@ -283,6 +289,10 @@ void __init of_sama5d2_clk_generated_setup(struct device_node *np, if (!num || num > PERIPHERAL_MAX) return; @@ -4750,7 +4878,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-generated.c linux-4.4.62/driver for_each_child_of_node(np, gcknp) { if (of_property_read_u32(gcknp, "reg", &id)) continue; -@@ -296,11 +306,14 @@ +@@ -296,11 +306,14 @@ void __init of_sama5d2_clk_generated_setup(struct device_node *np, of_at91_get_clk_range(gcknp, "atmel,clk-output-range", &range); @@ -4767,9 +4895,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-generated.c linux-4.4.62/driver } +CLK_OF_DECLARE(of_sama5d2_clk_generated_setup, "atmel,sama5d2-clk-generated", + of_sama5d2_clk_generated_setup); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-h32mx.c linux-4.4.62/drivers/clk/at91/clk-h32mx.c ---- linux-4.4.62.orig/drivers/clk/at91/clk-h32mx.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/clk-h32mx.c 2017-04-18 17:38:08.042643350 +0200 +diff --git a/drivers/clk/at91/clk-h32mx.c b/drivers/clk/at91/clk-h32mx.c +index a165230e7eda..8e20c8a76db7 100644 +--- a/drivers/clk/at91/clk-h32mx.c ++++ b/drivers/clk/at91/clk-h32mx.c @@ -15,15 +15,9 @@ #include #include @@ -4797,7 +4926,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-h32mx.c linux-4.4.62/drivers/cl }; #define to_clk_sama5d4_h32mx(hw) container_of(hw, struct clk_sama5d4_h32mx, hw) -@@ -40,8 +34,10 @@ +@@ -40,8 +34,10 @@ static unsigned long clk_sama5d4_h32mx_recalc_rate(struct clk_hw *hw, unsigned long parent_rate) { struct clk_sama5d4_h32mx *h32mxclk = to_clk_sama5d4_h32mx(hw); @@ -4809,7 +4938,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-h32mx.c linux-4.4.62/drivers/cl return parent_rate / 2; if (parent_rate > H32MX_MAX_FREQ) -@@ -70,18 +66,16 @@ +@@ -70,18 +66,16 @@ static int clk_sama5d4_h32mx_set_rate(struct clk_hw *hw, unsigned long rate, unsigned long parent_rate) { struct clk_sama5d4_h32mx *h32mxclk = to_clk_sama5d4_h32mx(hw); @@ -4833,7 +4962,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-h32mx.c linux-4.4.62/drivers/cl return 0; } -@@ -92,14 +86,18 @@ +@@ -92,14 +86,18 @@ static const struct clk_ops h32mx_ops = { .set_rate = clk_sama5d4_h32mx_set_rate, }; @@ -4854,7 +4983,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-h32mx.c linux-4.4.62/drivers/cl h32mxclk = kzalloc(sizeof(*h32mxclk), GFP_KERNEL); if (!h32mxclk) return; -@@ -113,7 +111,7 @@ +@@ -113,7 +111,7 @@ void __init of_sama5d4_clk_h32mx_setup(struct device_node *np, init.flags = CLK_SET_RATE_GATE; h32mxclk->hw.init = &init; @@ -4863,15 +4992,16 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-h32mx.c linux-4.4.62/drivers/cl clk = clk_register(NULL, &h32mxclk->hw); if (IS_ERR(clk)) { -@@ -123,3 +121,5 @@ +@@ -123,3 +121,5 @@ void __init of_sama5d4_clk_h32mx_setup(struct device_node *np, of_clk_add_provider(np, of_clk_src_simple_get, clk); } +CLK_OF_DECLARE(of_sama5d4_clk_h32mx_setup, "atmel,sama5d4-clk-h32mx", + of_sama5d4_clk_h32mx_setup); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk/at91/clk-main.c ---- linux-4.4.62.orig/drivers/clk/at91/clk-main.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/clk-main.c 2017-04-18 17:38:08.042643350 +0200 +diff --git a/drivers/clk/at91/clk-main.c b/drivers/clk/at91/clk-main.c +index fd7247deabdc..4bfc94d6c26e 100644 +--- a/drivers/clk/at91/clk-main.c ++++ b/drivers/clk/at91/clk-main.c @@ -13,13 +13,8 @@ #include #include @@ -4909,7 +5039,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk unsigned long frequency; unsigned long accuracy; }; -@@ -54,51 +45,47 @@ +@@ -54,51 +45,47 @@ struct clk_main_rc_osc { struct clk_rm9200_main { struct clk_hw hw; @@ -4974,7 +5104,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk return 0; } -@@ -106,9 +93,10 @@ +@@ -106,9 +93,10 @@ static int clk_main_osc_prepare(struct clk_hw *hw) static void clk_main_osc_unprepare(struct clk_hw *hw) { struct clk_main_osc *osc = to_clk_main_osc(hw); @@ -4987,7 +5117,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk if (tmp & AT91_PMC_OSCBYPASS) return; -@@ -116,20 +104,22 @@ +@@ -116,20 +104,22 @@ static void clk_main_osc_unprepare(struct clk_hw *hw) return; tmp &= ~(AT91_PMC_KEY | AT91_PMC_MOSCEN); @@ -5015,7 +5145,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk } static const struct clk_ops main_osc_ops = { -@@ -139,18 +129,16 @@ +@@ -139,18 +129,16 @@ static const struct clk_ops main_osc_ops = { }; static struct clk * __init @@ -5036,7 +5166,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk return ERR_PTR(-EINVAL); osc = kzalloc(sizeof(*osc), GFP_KERNEL); -@@ -164,85 +152,70 @@ +@@ -164,85 +152,70 @@ at91_clk_register_main_osc(struct at91_pmc *pmc, init.flags = CLK_IGNORE_UNUSED; osc->hw.init = &init; @@ -5148,7 +5278,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk return 0; } -@@ -250,23 +223,28 @@ +@@ -250,23 +223,28 @@ static int clk_main_rc_osc_prepare(struct clk_hw *hw) static void clk_main_rc_osc_unprepare(struct clk_hw *hw) { struct clk_main_rc_osc *osc = to_clk_main_rc_osc(hw); @@ -5156,10 +5286,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk - u32 tmp = pmc_read(pmc, AT91_CKGR_MOR); + struct regmap *regmap = osc->regmap; + unsigned int mor; ++ ++ regmap_read(regmap, AT91_CKGR_MOR, &mor); - if (!(tmp & AT91_PMC_MOSCRCEN)) -+ regmap_read(regmap, AT91_CKGR_MOR, &mor); -+ + if (!(mor & AT91_PMC_MOSCRCEN)) return; @@ -5175,17 +5305,17 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk - struct at91_pmc *pmc = osc->pmc; + struct regmap *regmap = osc->regmap; + unsigned int mor, status; -+ -+ regmap_read(regmap, AT91_CKGR_MOR, &mor); -+ regmap_read(regmap, AT91_PMC_SR, &status); - return !!((pmc_read(pmc, AT91_PMC_SR) & AT91_PMC_MOSCRCS) && - (pmc_read(pmc, AT91_CKGR_MOR) & AT91_PMC_MOSCRCEN)); ++ regmap_read(regmap, AT91_CKGR_MOR, &mor); ++ regmap_read(regmap, AT91_PMC_SR, &status); ++ + return (mor & AT91_PMC_MOSCRCEN) && (status & AT91_PMC_MOSCRCS); } static unsigned long clk_main_rc_osc_recalc_rate(struct clk_hw *hw, -@@ -294,17 +272,15 @@ +@@ -294,17 +272,15 @@ static const struct clk_ops main_rc_osc_ops = { }; static struct clk * __init @@ -5205,7 +5335,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk return ERR_PTR(-EINVAL); osc = kzalloc(sizeof(*osc), GFP_KERNEL); -@@ -318,63 +294,53 @@ +@@ -318,63 +294,53 @@ at91_clk_register_main_rc_osc(struct at91_pmc *pmc, init.flags = CLK_IS_ROOT | CLK_IGNORE_UNUSED; osc->hw.init = &init; @@ -5282,7 +5412,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk return 0; usleep_range(MAINF_LOOP_MIN_WAIT, MAINF_LOOP_MAX_WAIT); } while (time_before(prep_time, timeout)); -@@ -382,34 +348,37 @@ +@@ -382,34 +348,37 @@ static int clk_main_probe_frequency(struct at91_pmc *pmc) return -ETIMEDOUT; } @@ -5327,7 +5457,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk } static unsigned long clk_rm9200_main_recalc_rate(struct clk_hw *hw, -@@ -417,7 +386,7 @@ +@@ -417,7 +386,7 @@ static unsigned long clk_rm9200_main_recalc_rate(struct clk_hw *hw, { struct clk_rm9200_main *clkmain = to_clk_rm9200_main(hw); @@ -5336,7 +5466,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk } static const struct clk_ops rm9200_main_ops = { -@@ -427,7 +396,7 @@ +@@ -427,7 +396,7 @@ static const struct clk_ops rm9200_main_ops = { }; static struct clk * __init @@ -5345,7 +5475,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk const char *name, const char *parent_name) { -@@ -435,7 +404,7 @@ +@@ -435,7 +404,7 @@ at91_clk_register_rm9200_main(struct at91_pmc *pmc, struct clk *clk = NULL; struct clk_init_data init; @@ -5354,7 +5484,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk return ERR_PTR(-EINVAL); if (!parent_name) -@@ -452,7 +421,7 @@ +@@ -452,7 +421,7 @@ at91_clk_register_rm9200_main(struct at91_pmc *pmc, init.flags = 0; clkmain->hw.init = &init; @@ -5363,7 +5493,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk clk = clk_register(NULL, &clkmain->hw); if (IS_ERR(clk)) -@@ -461,52 +430,54 @@ +@@ -461,52 +430,54 @@ at91_clk_register_rm9200_main(struct at91_pmc *pmc, return clk; } @@ -5434,7 +5564,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk } static unsigned long clk_sam9x5_main_recalc_rate(struct clk_hw *hw, -@@ -514,30 +485,28 @@ +@@ -514,30 +485,28 @@ static unsigned long clk_sam9x5_main_recalc_rate(struct clk_hw *hw, { struct clk_sam9x5_main *clkmain = to_clk_sam9x5_main(hw); @@ -5474,7 +5604,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk return 0; } -@@ -545,8 +514,11 @@ +@@ -545,8 +514,11 @@ static int clk_sam9x5_main_set_parent(struct clk_hw *hw, u8 index) static u8 clk_sam9x5_main_get_parent(struct clk_hw *hw) { struct clk_sam9x5_main *clkmain = to_clk_sam9x5_main(hw); @@ -5487,7 +5617,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk } static const struct clk_ops sam9x5_main_ops = { -@@ -558,18 +530,17 @@ +@@ -558,18 +530,17 @@ static const struct clk_ops sam9x5_main_ops = { }; static struct clk * __init @@ -5509,7 +5639,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk return ERR_PTR(-EINVAL); if (!parent_names || !num_parents) -@@ -586,51 +557,42 @@ +@@ -586,51 +557,42 @@ at91_clk_register_sam9x5_main(struct at91_pmc *pmc, init.flags = CLK_SET_PARENT_GATE; clkmain->hw.init = &init; @@ -5573,9 +5703,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-main.c linux-4.4.62/drivers/clk } +CLK_OF_DECLARE(at91sam9x5_clk_main, "atmel,at91sam9x5-clk-main", + of_at91sam9x5_clk_main_setup); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-master.c linux-4.4.62/drivers/clk/at91/clk-master.c ---- linux-4.4.62.orig/drivers/clk/at91/clk-master.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/clk-master.c 2017-04-18 17:38:08.042643350 +0200 +diff --git a/drivers/clk/at91/clk-master.c b/drivers/clk/at91/clk-master.c +index 620ea323356b..7d4a1864ea7c 100644 +--- a/drivers/clk/at91/clk-master.c ++++ b/drivers/clk/at91/clk-master.c @@ -12,13 +12,8 @@ #include #include @@ -5592,7 +5723,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-master.c linux-4.4.62/drivers/c #include "pmc.h" -@@ -44,32 +39,26 @@ +@@ -44,32 +39,26 @@ struct clk_master_layout { struct clk_master { struct clk_hw hw; @@ -5633,7 +5764,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-master.c linux-4.4.62/drivers/c return 0; } -@@ -78,7 +67,7 @@ +@@ -78,7 +67,7 @@ static int clk_master_is_prepared(struct clk_hw *hw) { struct clk_master *master = to_clk_master(hw); @@ -5642,7 +5773,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-master.c linux-4.4.62/drivers/c } static unsigned long clk_master_recalc_rate(struct clk_hw *hw, -@@ -88,18 +77,16 @@ +@@ -88,18 +77,16 @@ static unsigned long clk_master_recalc_rate(struct clk_hw *hw, u8 div; unsigned long rate = parent_rate; struct clk_master *master = to_clk_master(hw); @@ -5666,7 +5797,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-master.c linux-4.4.62/drivers/c if (characteristics->have_div3_pres && pres == MASTER_PRES_MAX) rate /= 3; -@@ -119,9 +106,11 @@ +@@ -119,9 +106,11 @@ static unsigned long clk_master_recalc_rate(struct clk_hw *hw, static u8 clk_master_get_parent(struct clk_hw *hw) { struct clk_master *master = to_clk_master(hw); @@ -5680,7 +5811,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-master.c linux-4.4.62/drivers/c } static const struct clk_ops master_ops = { -@@ -132,18 +121,17 @@ +@@ -132,18 +121,17 @@ static const struct clk_ops master_ops = { }; static struct clk * __init @@ -5701,7 +5832,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-master.c linux-4.4.62/drivers/c return ERR_PTR(-EINVAL); master = kzalloc(sizeof(*master), GFP_KERNEL); -@@ -159,20 +147,10 @@ +@@ -159,20 +147,10 @@ at91_clk_register_master(struct at91_pmc *pmc, unsigned int irq, master->hw.init = &init; master->layout = layout; master->characteristics = characteristics; @@ -5723,7 +5854,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-master.c linux-4.4.62/drivers/c kfree(master); } -@@ -217,15 +195,15 @@ +@@ -217,15 +195,15 @@ out_free_characteristics: } static void __init @@ -5741,7 +5872,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-master.c linux-4.4.62/drivers/c num_parents = of_clk_get_parent_count(np); if (num_parents <= 0 || num_parents > MASTER_SOURCE_MAX) -@@ -239,11 +217,11 @@ +@@ -239,11 +217,11 @@ of_at91_clk_master_setup(struct device_node *np, struct at91_pmc *pmc, if (!characteristics) return; @@ -5757,7 +5888,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-master.c linux-4.4.62/drivers/c parent_names, layout, characteristics); if (IS_ERR(clk)) -@@ -256,14 +234,16 @@ +@@ -256,14 +234,16 @@ out_free_characteristics: kfree(characteristics); } @@ -5780,9 +5911,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-master.c linux-4.4.62/drivers/c } +CLK_OF_DECLARE(at91sam9x5_clk_master, "atmel,at91sam9x5-clk-master", + of_at91sam9x5_clk_master_setup); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drivers/clk/at91/clk-peripheral.c ---- linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/clk-peripheral.c 2017-04-18 17:38:08.042643350 +0200 +diff --git a/drivers/clk/at91/clk-peripheral.c b/drivers/clk/at91/clk-peripheral.c +index 58f3b568e9cb..d69cd2a121b1 100644 +--- a/drivers/clk/at91/clk-peripheral.c ++++ b/drivers/clk/at91/clk-peripheral.c @@ -12,11 +12,13 @@ #include #include @@ -5808,7 +5940,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive u32 id; }; -@@ -41,8 +43,9 @@ +@@ -41,8 +43,9 @@ struct clk_peripheral { struct clk_sam9x5_peripheral { struct clk_hw hw; @@ -5819,7 +5951,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive u32 id; u32 div; bool auto_div; -@@ -54,7 +57,6 @@ +@@ -54,7 +57,6 @@ struct clk_sam9x5_peripheral { static int clk_peripheral_enable(struct clk_hw *hw) { struct clk_peripheral *periph = to_clk_peripheral(hw); @@ -5827,7 +5959,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive int offset = AT91_PMC_PCER; u32 id = periph->id; -@@ -62,14 +64,14 @@ +@@ -62,14 +64,14 @@ static int clk_peripheral_enable(struct clk_hw *hw) return 0; if (id > PERIPHERAL_ID_MAX) offset = AT91_PMC_PCER1; @@ -5844,7 +5976,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive int offset = AT91_PMC_PCDR; u32 id = periph->id; -@@ -77,21 +79,23 @@ +@@ -77,21 +79,23 @@ static void clk_peripheral_disable(struct clk_hw *hw) return; if (id > PERIPHERAL_ID_MAX) offset = AT91_PMC_PCDR1; @@ -5871,7 +6003,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive } static const struct clk_ops peripheral_ops = { -@@ -101,14 +105,14 @@ +@@ -101,14 +105,14 @@ static const struct clk_ops peripheral_ops = { }; static struct clk * __init @@ -5888,7 +6020,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive return ERR_PTR(-EINVAL); periph = kzalloc(sizeof(*periph), GFP_KERNEL); -@@ -123,7 +127,7 @@ +@@ -123,7 +127,7 @@ at91_clk_register_peripheral(struct at91_pmc *pmc, const char *name, periph->id = id; periph->hw.init = &init; @@ -5897,7 +6029,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive clk = clk_register(NULL, &periph->hw); if (IS_ERR(clk)) -@@ -160,53 +164,58 @@ +@@ -160,53 +164,58 @@ static void clk_sam9x5_peripheral_autodiv(struct clk_sam9x5_peripheral *periph) static int clk_sam9x5_peripheral_enable(struct clk_hw *hw) { struct clk_sam9x5_peripheral *periph = to_clk_sam9x5_peripheral(hw); @@ -5979,7 +6111,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive } static unsigned long -@@ -214,19 +223,20 @@ +@@ -214,19 +223,20 @@ clk_sam9x5_peripheral_recalc_rate(struct clk_hw *hw, unsigned long parent_rate) { struct clk_sam9x5_peripheral *periph = to_clk_sam9x5_peripheral(hw); @@ -6008,7 +6140,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive periph->auto_div = false; } else { clk_sam9x5_peripheral_autodiv(periph); -@@ -318,15 +328,15 @@ +@@ -318,15 +328,15 @@ static const struct clk_ops sam9x5_peripheral_ops = { }; static struct clk * __init @@ -6028,7 +6160,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive return ERR_PTR(-EINVAL); periph = kzalloc(sizeof(*periph), GFP_KERNEL); -@@ -342,7 +352,8 @@ +@@ -342,7 +352,8 @@ at91_clk_register_sam9x5_peripheral(struct at91_pmc *pmc, const char *name, periph->id = id; periph->hw.init = &init; periph->div = 0; @@ -6038,7 +6170,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive periph->auto_div = true; periph->range = *range; -@@ -356,7 +367,7 @@ +@@ -356,7 +367,7 @@ at91_clk_register_sam9x5_peripheral(struct at91_pmc *pmc, const char *name, } static void __init @@ -6047,7 +6179,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive { int num; u32 id; -@@ -364,6 +375,7 @@ +@@ -364,6 +375,7 @@ of_at91_clk_periph_setup(struct device_node *np, struct at91_pmc *pmc, u8 type) const char *parent_name; const char *name; struct device_node *periphclknp; @@ -6055,7 +6187,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive parent_name = of_clk_get_parent_name(np, 0); if (!parent_name) -@@ -373,6 +385,10 @@ +@@ -373,6 +385,10 @@ of_at91_clk_periph_setup(struct device_node *np, struct at91_pmc *pmc, u8 type) if (!num || num > PERIPHERAL_MAX) return; @@ -6066,7 +6198,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive for_each_child_of_node(np, periphclknp) { if (of_property_read_u32(periphclknp, "reg", &id)) continue; -@@ -384,7 +400,7 @@ +@@ -384,7 +400,7 @@ of_at91_clk_periph_setup(struct device_node *np, struct at91_pmc *pmc, u8 type) name = periphclknp->name; if (type == PERIPHERAL_AT91RM9200) { @@ -6075,7 +6207,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive parent_name, id); } else { struct clk_range range = CLK_RANGE(0, 0); -@@ -393,7 +409,9 @@ +@@ -393,7 +409,9 @@ of_at91_clk_periph_setup(struct device_node *np, struct at91_pmc *pmc, u8 type) "atmel,clk-output-range", &range); @@ -6086,7 +6218,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive parent_name, id, &range); } -@@ -405,14 +423,16 @@ +@@ -405,14 +423,16 @@ of_at91_clk_periph_setup(struct device_node *np, struct at91_pmc *pmc, u8 type) } } @@ -6109,9 +6241,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-peripheral.c linux-4.4.62/drive } +CLK_OF_DECLARE(at91sam9x5_clk_periph, "atmel,at91sam9x5-clk-peripheral", + of_at91sam9x5_clk_periph_setup); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-pll.c linux-4.4.62/drivers/clk/at91/clk-pll.c ---- linux-4.4.62.orig/drivers/clk/at91/clk-pll.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/clk-pll.c 2017-04-18 17:38:08.042643350 +0200 +diff --git a/drivers/clk/at91/clk-pll.c b/drivers/clk/at91/clk-pll.c +index 18b60f4895a6..fb2e0b56d4b7 100644 +--- a/drivers/clk/at91/clk-pll.c ++++ b/drivers/clk/at91/clk-pll.c @@ -12,14 +12,8 @@ #include #include @@ -6129,7 +6262,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-pll.c linux-4.4.62/drivers/clk/ #include "pmc.h" -@@ -58,9 +52,7 @@ +@@ -58,9 +52,7 @@ struct clk_pll_layout { struct clk_pll { struct clk_hw hw; @@ -6140,7 +6273,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-pll.c linux-4.4.62/drivers/clk/ u8 id; u8 div; u8 range; -@@ -69,20 +61,19 @@ +@@ -69,20 +61,19 @@ struct clk_pll { const struct clk_pll_characteristics *characteristics; }; @@ -6166,7 +6299,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-pll.c linux-4.4.62/drivers/clk/ const struct clk_pll_layout *layout = pll->layout; const struct clk_pll_characteristics *characteristics = pll->characteristics; -@@ -90,39 +81,34 @@ +@@ -90,39 +81,34 @@ static int clk_pll_prepare(struct clk_hw *hw) u32 mask = PLL_STATUS_MASK(id); int offset = PLL_REG(id); u8 out = 0; @@ -6222,7 +6355,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-pll.c linux-4.4.62/drivers/clk/ return 0; } -@@ -130,32 +116,35 @@ +@@ -130,32 +116,35 @@ static int clk_pll_prepare(struct clk_hw *hw) static int clk_pll_is_prepared(struct clk_hw *hw) { struct clk_pll *pll = to_clk_pll(hw); @@ -6253,13 +6386,13 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-pll.c linux-4.4.62/drivers/clk/ + unsigned int pllr; + u16 mul; + u8 div; -+ + +- if (!pll->div || !pll->mul) + regmap_read(pll->regmap, PLL_REG(pll->id), &pllr); + + div = PLL_DIV(pllr); + mul = PLL_MUL(pllr, pll->layout); - -- if (!pll->div || !pll->mul) ++ + if (!div || !mul) return 0; @@ -6268,7 +6401,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-pll.c linux-4.4.62/drivers/clk/ } static long clk_pll_get_best_div_mul(struct clk_pll *pll, unsigned long rate, -@@ -308,7 +297,7 @@ +@@ -308,7 +297,7 @@ static const struct clk_ops pll_ops = { }; static struct clk * __init @@ -6277,7 +6410,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-pll.c linux-4.4.62/drivers/clk/ const char *parent_name, u8 id, const struct clk_pll_layout *layout, const struct clk_pll_characteristics *characteristics) -@@ -316,9 +305,8 @@ +@@ -316,9 +305,8 @@ at91_clk_register_pll(struct at91_pmc *pmc, unsigned int irq, const char *name, struct clk_pll *pll; struct clk *clk = NULL; struct clk_init_data init; @@ -6288,7 +6421,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-pll.c linux-4.4.62/drivers/clk/ if (id > PLL_MAX_ID) return ERR_PTR(-EINVAL); -@@ -337,23 +325,13 @@ +@@ -337,23 +325,13 @@ at91_clk_register_pll(struct at91_pmc *pmc, unsigned int irq, const char *name, pll->hw.init = &init; pll->layout = layout; pll->characteristics = characteristics; @@ -6316,7 +6449,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-pll.c linux-4.4.62/drivers/clk/ kfree(pll); } -@@ -483,12 +461,12 @@ +@@ -483,12 +461,12 @@ out_free_characteristics: } static void __init @@ -6331,7 +6464,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-pll.c linux-4.4.62/drivers/clk/ const char *parent_name; const char *name = np->name; struct clk_pll_characteristics *characteristics; -@@ -500,15 +478,15 @@ +@@ -500,15 +478,15 @@ of_at91_clk_pll_setup(struct device_node *np, struct at91_pmc *pmc, of_property_read_string(np, "clock-output-names", &name); @@ -6352,7 +6485,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-pll.c linux-4.4.62/drivers/clk/ characteristics); if (IS_ERR(clk)) goto out_free_characteristics; -@@ -520,26 +498,30 @@ +@@ -520,26 +498,30 @@ out_free_characteristics: kfree(characteristics); } @@ -6395,9 +6528,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-pll.c linux-4.4.62/drivers/clk/ } +CLK_OF_DECLARE(sama5d3_clk_pll, "atmel,sama5d3-clk-pll", + of_sama5d3_clk_pll_setup); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-plldiv.c linux-4.4.62/drivers/clk/at91/clk-plldiv.c ---- linux-4.4.62.orig/drivers/clk/at91/clk-plldiv.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/clk-plldiv.c 2017-04-18 17:38:08.042643350 +0200 +diff --git a/drivers/clk/at91/clk-plldiv.c b/drivers/clk/at91/clk-plldiv.c +index ea226562bb40..2bed26481027 100644 +--- a/drivers/clk/at91/clk-plldiv.c ++++ b/drivers/clk/at91/clk-plldiv.c @@ -12,8 +12,8 @@ #include #include @@ -6431,7 +6565,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-plldiv.c linux-4.4.62/drivers/c return parent_rate / 2; return parent_rate; -@@ -57,18 +59,12 @@ +@@ -57,18 +59,12 @@ static int clk_plldiv_set_rate(struct clk_hw *hw, unsigned long rate, unsigned long parent_rate) { struct clk_plldiv *plldiv = to_clk_plldiv(hw); @@ -6453,7 +6587,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-plldiv.c linux-4.4.62/drivers/c return 0; } -@@ -80,7 +76,7 @@ +@@ -80,7 +76,7 @@ static const struct clk_ops plldiv_ops = { }; static struct clk * __init @@ -6462,7 +6596,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-plldiv.c linux-4.4.62/drivers/c const char *parent_name) { struct clk_plldiv *plldiv; -@@ -98,7 +94,7 @@ +@@ -98,7 +94,7 @@ at91_clk_register_plldiv(struct at91_pmc *pmc, const char *name, init.flags = CLK_SET_RATE_GATE; plldiv->hw.init = &init; @@ -6471,7 +6605,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-plldiv.c linux-4.4.62/drivers/c clk = clk_register(NULL, &plldiv->hw); -@@ -109,27 +105,27 @@ +@@ -109,27 +105,27 @@ at91_clk_register_plldiv(struct at91_pmc *pmc, const char *name, } static void __init @@ -6507,9 +6641,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-plldiv.c linux-4.4.62/drivers/c -} +CLK_OF_DECLARE(at91sam9x5_clk_plldiv, "atmel,at91sam9x5-clk-plldiv", + of_at91sam9x5_clk_plldiv_setup); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c linux-4.4.62/drivers/clk/at91/clk-programmable.c ---- linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/clk-programmable.c 2017-04-18 17:38:08.046643506 +0200 +diff --git a/drivers/clk/at91/clk-programmable.c b/drivers/clk/at91/clk-programmable.c +index 14b270b85fec..bc0be629671b 100644 +--- a/drivers/clk/at91/clk-programmable.c ++++ b/drivers/clk/at91/clk-programmable.c @@ -12,10 +12,8 @@ #include #include @@ -6531,7 +6666,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c linux-4.4.62/dri #define PROG_MAX_RM9200_CSS 3 struct clk_programmable_layout { -@@ -34,7 +33,7 @@ +@@ -34,7 +33,7 @@ struct clk_programmable_layout { struct clk_programmable { struct clk_hw hw; @@ -6540,7 +6675,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c linux-4.4.62/dri u8 id; const struct clk_programmable_layout *layout; }; -@@ -44,14 +43,12 @@ +@@ -44,14 +43,12 @@ struct clk_programmable { static unsigned long clk_programmable_recalc_rate(struct clk_hw *hw, unsigned long parent_rate) { @@ -6559,7 +6694,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c linux-4.4.62/dri } static int clk_programmable_determine_rate(struct clk_hw *hw, -@@ -101,36 +98,36 @@ +@@ -101,36 +98,36 @@ static int clk_programmable_set_parent(struct clk_hw *hw, u8 index) { struct clk_programmable *prog = to_clk_programmable(hw); const struct clk_programmable_layout *layout = prog->layout; @@ -6611,7 +6746,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c linux-4.4.62/dri ret = PROG_MAX_RM9200_CSS + 1; return ret; -@@ -140,26 +137,27 @@ +@@ -140,26 +137,27 @@ static int clk_programmable_set_rate(struct clk_hw *hw, unsigned long rate, unsigned long parent_rate) { struct clk_programmable *prog = to_clk_programmable(hw); @@ -6645,7 +6780,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c linux-4.4.62/dri return 0; } -@@ -173,7 +171,7 @@ +@@ -173,7 +171,7 @@ static const struct clk_ops programmable_ops = { }; static struct clk * __init @@ -6654,7 +6789,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c linux-4.4.62/dri const char *name, const char **parent_names, u8 num_parents, u8 id, const struct clk_programmable_layout *layout) -@@ -198,7 +196,7 @@ +@@ -198,7 +196,7 @@ at91_clk_register_programmable(struct at91_pmc *pmc, prog->id = id; prog->layout = layout; prog->hw.init = &init; @@ -6663,7 +6798,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c linux-4.4.62/dri clk = clk_register(NULL, &prog->hw); if (IS_ERR(clk)) -@@ -226,7 +224,7 @@ +@@ -226,7 +224,7 @@ static const struct clk_programmable_layout at91sam9x5_programmable_layout = { }; static void __init @@ -6672,7 +6807,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c linux-4.4.62/dri const struct clk_programmable_layout *layout) { int num; -@@ -236,6 +234,7 @@ +@@ -236,6 +234,7 @@ of_at91_clk_prog_setup(struct device_node *np, struct at91_pmc *pmc, const char *parent_names[PROG_SOURCE_MAX]; const char *name; struct device_node *progclknp; @@ -6680,7 +6815,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c linux-4.4.62/dri num_parents = of_clk_get_parent_count(np); if (num_parents <= 0 || num_parents > PROG_SOURCE_MAX) -@@ -247,6 +246,10 @@ +@@ -247,6 +246,10 @@ of_at91_clk_prog_setup(struct device_node *np, struct at91_pmc *pmc, if (!num || num > (PROG_ID_MAX + 1)) return; @@ -6691,7 +6826,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c linux-4.4.62/dri for_each_child_of_node(np, progclknp) { if (of_property_read_u32(progclknp, "reg", &id)) continue; -@@ -254,7 +257,7 @@ +@@ -254,7 +257,7 @@ of_at91_clk_prog_setup(struct device_node *np, struct at91_pmc *pmc, if (of_property_read_string(np, "clock-output-names", &name)) name = progclknp->name; @@ -6700,7 +6835,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c linux-4.4.62/dri parent_names, num_parents, id, layout); if (IS_ERR(clk)) -@@ -265,20 +268,23 @@ +@@ -265,20 +268,23 @@ of_at91_clk_prog_setup(struct device_node *np, struct at91_pmc *pmc, } @@ -6733,9 +6868,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-programmable.c linux-4.4.62/dri } +CLK_OF_DECLARE(at91sam9x5_clk_prog, "atmel,at91sam9x5-clk-programmable", + of_at91sam9x5_clk_prog_setup); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-slow.c linux-4.4.62/drivers/clk/at91/clk-slow.c ---- linux-4.4.62.orig/drivers/clk/at91/clk-slow.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/clk-slow.c 2017-04-18 17:38:08.046643506 +0200 +diff --git a/drivers/clk/at91/clk-slow.c b/drivers/clk/at91/clk-slow.c +index d0d5076a9b94..221c09684ba3 100644 +--- a/drivers/clk/at91/clk-slow.c ++++ b/drivers/clk/at91/clk-slow.c @@ -13,17 +13,11 @@ #include #include @@ -6756,7 +6892,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-slow.c linux-4.4.62/drivers/clk #include "pmc.h" #include "sckc.h" -@@ -59,7 +53,7 @@ +@@ -59,7 +53,7 @@ struct clk_slow_rc_osc { struct clk_sam9260_slow { struct clk_hw hw; @@ -6765,7 +6901,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-slow.c linux-4.4.62/drivers/clk }; #define to_clk_sam9260_slow(hw) container_of(hw, struct clk_sam9260_slow, hw) -@@ -393,8 +387,11 @@ +@@ -393,8 +387,11 @@ void __init of_at91sam9x5_clk_slow_setup(struct device_node *np, static u8 clk_sam9260_slow_get_parent(struct clk_hw *hw) { struct clk_sam9260_slow *slowck = to_clk_sam9260_slow(hw); @@ -6778,7 +6914,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-slow.c linux-4.4.62/drivers/clk } static const struct clk_ops sam9260_slow_ops = { -@@ -402,7 +399,7 @@ +@@ -402,7 +399,7 @@ static const struct clk_ops sam9260_slow_ops = { }; static struct clk * __init @@ -6787,7 +6923,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-slow.c linux-4.4.62/drivers/clk const char *name, const char **parent_names, int num_parents) -@@ -411,7 +408,7 @@ +@@ -411,7 +408,7 @@ at91_clk_register_sam9260_slow(struct at91_pmc *pmc, struct clk *clk = NULL; struct clk_init_data init; @@ -6796,7 +6932,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-slow.c linux-4.4.62/drivers/clk return ERR_PTR(-EINVAL); if (!parent_names || !num_parents) -@@ -428,7 +425,7 @@ +@@ -428,7 +425,7 @@ at91_clk_register_sam9260_slow(struct at91_pmc *pmc, init.flags = 0; slowck->hw.init = &init; @@ -6805,7 +6941,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-slow.c linux-4.4.62/drivers/clk clk = clk_register(NULL, &slowck->hw); if (IS_ERR(clk)) -@@ -439,29 +436,34 @@ +@@ -439,29 +436,34 @@ at91_clk_register_sam9260_slow(struct at91_pmc *pmc, return clk; } @@ -6843,9 +6979,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-slow.c linux-4.4.62/drivers/clk /* * FIXME: All slow clk users are not properly claiming it (get + prepare + -diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-smd.c linux-4.4.62/drivers/clk/at91/clk-smd.c ---- linux-4.4.62.orig/drivers/clk/at91/clk-smd.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/clk-smd.c 2017-04-18 17:38:08.046643506 +0200 +diff --git a/drivers/clk/at91/clk-smd.c b/drivers/clk/at91/clk-smd.c +index a7f8501cfa05..e6948a52005a 100644 +--- a/drivers/clk/at91/clk-smd.c ++++ b/drivers/clk/at91/clk-smd.c @@ -12,8 +12,8 @@ #include #include @@ -6866,7 +7003,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-smd.c linux-4.4.62/drivers/clk/ }; #define to_at91sam9x5_clk_smd(hw) \ -@@ -33,13 +33,13 @@ +@@ -33,13 +33,13 @@ struct at91sam9x5_clk_smd { static unsigned long at91sam9x5_clk_smd_recalc_rate(struct clk_hw *hw, unsigned long parent_rate) { @@ -6885,7 +7022,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-smd.c linux-4.4.62/drivers/clk/ return parent_rate / (smddiv + 1); } -@@ -67,40 +67,38 @@ +@@ -67,40 +67,38 @@ static long at91sam9x5_clk_smd_round_rate(struct clk_hw *hw, unsigned long rate, static int at91sam9x5_clk_smd_set_parent(struct clk_hw *hw, u8 index) { @@ -6937,7 +7074,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-smd.c linux-4.4.62/drivers/clk/ return 0; } -@@ -114,7 +112,7 @@ +@@ -114,7 +112,7 @@ static const struct clk_ops at91sam9x5_smd_ops = { }; static struct clk * __init @@ -6946,7 +7083,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-smd.c linux-4.4.62/drivers/clk/ const char **parent_names, u8 num_parents) { struct at91sam9x5_clk_smd *smd; -@@ -132,7 +130,7 @@ +@@ -132,7 +130,7 @@ at91sam9x5_clk_register_smd(struct at91_pmc *pmc, const char *name, init.flags = CLK_SET_RATE_GATE | CLK_SET_PARENT_GATE; smd->hw.init = &init; @@ -6955,7 +7092,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-smd.c linux-4.4.62/drivers/clk/ clk = clk_register(NULL, &smd->hw); if (IS_ERR(clk)) -@@ -141,13 +139,13 @@ +@@ -141,13 +139,13 @@ at91sam9x5_clk_register_smd(struct at91_pmc *pmc, const char *name, return clk; } @@ -6971,7 +7108,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-smd.c linux-4.4.62/drivers/clk/ num_parents = of_clk_get_parent_count(np); if (num_parents <= 0 || num_parents > SMD_SOURCE_MAX) -@@ -157,10 +155,16 @@ +@@ -157,10 +155,16 @@ void __init of_at91sam9x5_clk_smd_setup(struct device_node *np, of_property_read_string(np, "clock-output-names", &name); @@ -6989,9 +7126,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-smd.c linux-4.4.62/drivers/clk/ } +CLK_OF_DECLARE(at91sam9x5_clk_smd, "atmel,at91sam9x5-clk-smd", + of_at91sam9x5_clk_smd_setup); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-system.c linux-4.4.62/drivers/clk/at91/clk-system.c ---- linux-4.4.62.orig/drivers/clk/at91/clk-system.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/clk-system.c 2017-04-18 17:38:08.046643506 +0200 +diff --git a/drivers/clk/at91/clk-system.c b/drivers/clk/at91/clk-system.c +index 3f5314344286..8f35d8172909 100644 +--- a/drivers/clk/at91/clk-system.c ++++ b/drivers/clk/at91/clk-system.c @@ -12,13 +12,8 @@ #include #include @@ -7019,7 +7157,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-system.c linux-4.4.62/drivers/c u8 id; }; -@@ -39,58 +32,54 @@ +@@ -39,58 +32,54 @@ static inline int is_pck(int id) { return (id >= 8) && (id <= 15); } @@ -7095,7 +7233,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-system.c linux-4.4.62/drivers/c } static const struct clk_ops system_ops = { -@@ -100,13 +89,12 @@ +@@ -100,13 +89,12 @@ static const struct clk_ops system_ops = { }; static struct clk * __init @@ -7111,7 +7249,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-system.c linux-4.4.62/drivers/c if (!parent_name || id > SYSTEM_MAX_ID) return ERR_PTR(-EINVAL); -@@ -123,44 +111,33 @@ +@@ -123,44 +111,33 @@ at91_clk_register_system(struct at91_pmc *pmc, const char *name, sys->id = id; sys->hw.init = &init; @@ -7164,7 +7302,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-system.c linux-4.4.62/drivers/c for_each_child_of_node(np, sysclknp) { if (of_property_read_u32(sysclknp, "reg", &id)) continue; -@@ -168,21 +145,14 @@ +@@ -168,21 +145,14 @@ of_at91_clk_sys_setup(struct device_node *np, struct at91_pmc *pmc) if (of_property_read_string(np, "clock-output-names", &name)) name = sysclknp->name; @@ -7189,9 +7327,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-system.c linux-4.4.62/drivers/c -} +CLK_OF_DECLARE(at91rm9200_clk_sys, "atmel,at91rm9200-clk-system", + of_at91rm9200_clk_sys_setup); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/at91/clk-usb.c ---- linux-4.4.62.orig/drivers/clk/at91/clk-usb.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/clk-usb.c 2017-04-18 17:38:08.046643506 +0200 +diff --git a/drivers/clk/at91/clk-usb.c b/drivers/clk/at91/clk-usb.c +index 8ab8502778a2..650ca45892c0 100644 +--- a/drivers/clk/at91/clk-usb.c ++++ b/drivers/clk/at91/clk-usb.c @@ -12,8 +12,8 @@ #include #include @@ -7212,7 +7351,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ }; #define to_at91sam9x5_clk_usb(hw) \ -@@ -35,7 +35,7 @@ +@@ -35,7 +35,7 @@ struct at91sam9x5_clk_usb { struct at91rm9200_clk_usb { struct clk_hw hw; @@ -7221,7 +7360,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ u32 divisors[4]; }; -@@ -45,13 +45,12 @@ +@@ -45,13 +45,12 @@ struct at91rm9200_clk_usb { static unsigned long at91sam9x5_clk_usb_recalc_rate(struct clk_hw *hw, unsigned long parent_rate) { @@ -7239,7 +7378,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ return DIV_ROUND_CLOSEST(parent_rate, (usbdiv + 1)); } -@@ -109,33 +108,31 @@ +@@ -109,33 +108,31 @@ static int at91sam9x5_clk_usb_determine_rate(struct clk_hw *hw, static int at91sam9x5_clk_usb_set_parent(struct clk_hw *hw, u8 index) { @@ -7281,7 +7420,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ unsigned long div; if (!rate) -@@ -145,9 +142,8 @@ +@@ -145,9 +142,8 @@ static int at91sam9x5_clk_usb_set_rate(struct clk_hw *hw, unsigned long rate, if (div > SAM9X5_USB_MAX_DIV + 1 || !div) return -EINVAL; @@ -7293,7 +7432,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ return 0; } -@@ -163,28 +159,28 @@ +@@ -163,28 +159,28 @@ static const struct clk_ops at91sam9x5_usb_ops = { static int at91sam9n12_clk_usb_enable(struct clk_hw *hw) { struct at91sam9x5_clk_usb *usb = to_at91sam9x5_clk_usb(hw); @@ -7330,7 +7469,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ } static const struct clk_ops at91sam9n12_usb_ops = { -@@ -197,7 +193,7 @@ +@@ -197,7 +193,7 @@ static const struct clk_ops at91sam9n12_usb_ops = { }; static struct clk * __init @@ -7339,7 +7478,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ const char **parent_names, u8 num_parents) { struct at91sam9x5_clk_usb *usb; -@@ -216,7 +212,7 @@ +@@ -216,7 +212,7 @@ at91sam9x5_clk_register_usb(struct at91_pmc *pmc, const char *name, CLK_SET_RATE_PARENT; usb->hw.init = &init; @@ -7348,7 +7487,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ clk = clk_register(NULL, &usb->hw); if (IS_ERR(clk)) -@@ -226,7 +222,7 @@ +@@ -226,7 +222,7 @@ at91sam9x5_clk_register_usb(struct at91_pmc *pmc, const char *name, } static struct clk * __init @@ -7357,7 +7496,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ const char *parent_name) { struct at91sam9x5_clk_usb *usb; -@@ -244,7 +240,7 @@ +@@ -244,7 +240,7 @@ at91sam9n12_clk_register_usb(struct at91_pmc *pmc, const char *name, init.flags = CLK_SET_RATE_GATE | CLK_SET_RATE_PARENT; usb->hw.init = &init; @@ -7366,7 +7505,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ clk = clk_register(NULL, &usb->hw); if (IS_ERR(clk)) -@@ -257,12 +253,12 @@ +@@ -257,12 +253,12 @@ static unsigned long at91rm9200_clk_usb_recalc_rate(struct clk_hw *hw, unsigned long parent_rate) { struct at91rm9200_clk_usb *usb = to_at91rm9200_clk_usb(hw); @@ -7383,7 +7522,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ if (usb->divisors[usbdiv]) return parent_rate / usb->divisors[usbdiv]; -@@ -310,10 +306,8 @@ +@@ -310,10 +306,8 @@ static long at91rm9200_clk_usb_round_rate(struct clk_hw *hw, unsigned long rate, static int at91rm9200_clk_usb_set_rate(struct clk_hw *hw, unsigned long rate, unsigned long parent_rate) { @@ -7394,7 +7533,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ unsigned long div; if (!rate) -@@ -323,10 +317,10 @@ +@@ -323,10 +317,10 @@ static int at91rm9200_clk_usb_set_rate(struct clk_hw *hw, unsigned long rate, for (i = 0; i < RM9200_USB_DIV_TAB_SIZE; i++) { if (usb->divisors[i] == div) { @@ -7409,7 +7548,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ return 0; } } -@@ -341,7 +335,7 @@ +@@ -341,7 +335,7 @@ static const struct clk_ops at91rm9200_usb_ops = { }; static struct clk * __init @@ -7418,7 +7557,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ const char *parent_name, const u32 *divisors) { struct at91rm9200_clk_usb *usb; -@@ -359,7 +353,7 @@ +@@ -359,7 +353,7 @@ at91rm9200_clk_register_usb(struct at91_pmc *pmc, const char *name, init.flags = CLK_SET_RATE_PARENT; usb->hw.init = &init; @@ -7427,7 +7566,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ memcpy(usb->divisors, divisors, sizeof(usb->divisors)); clk = clk_register(NULL, &usb->hw); -@@ -369,13 +363,13 @@ +@@ -369,13 +363,13 @@ at91rm9200_clk_register_usb(struct at91_pmc *pmc, const char *name, return clk; } @@ -7443,7 +7582,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ num_parents = of_clk_get_parent_count(np); if (num_parents <= 0 || num_parents > USB_SOURCE_MAX) -@@ -385,19 +379,26 @@ +@@ -385,19 +379,26 @@ void __init of_at91sam9x5_clk_usb_setup(struct device_node *np, of_property_read_string(np, "clock-output-names", &name); @@ -7473,7 +7612,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ parent_name = of_clk_get_parent_name(np, 0); if (!parent_name) -@@ -405,20 +406,26 @@ +@@ -405,20 +406,26 @@ void __init of_at91sam9n12_clk_usb_setup(struct device_node *np, of_property_read_string(np, "clock-output-names", &name); @@ -7503,7 +7642,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ parent_name = of_clk_get_parent_name(np, 0); if (!parent_name) -@@ -430,9 +437,15 @@ +@@ -430,9 +437,15 @@ void __init of_at91rm9200_clk_usb_setup(struct device_node *np, of_property_read_string(np, "clock-output-names", &name); @@ -7520,9 +7659,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-usb.c linux-4.4.62/drivers/clk/ } +CLK_OF_DECLARE(at91rm9200_clk_usb, "atmel,at91rm9200-clk-usb", + of_at91rm9200_clk_usb_setup); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-utmi.c linux-4.4.62/drivers/clk/at91/clk-utmi.c ---- linux-4.4.62.orig/drivers/clk/at91/clk-utmi.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/clk-utmi.c 2017-04-18 17:38:08.046643506 +0200 +diff --git a/drivers/clk/at91/clk-utmi.c b/drivers/clk/at91/clk-utmi.c +index ca561e90a60f..61fcf399e58c 100644 +--- a/drivers/clk/at91/clk-utmi.c ++++ b/drivers/clk/at91/clk-utmi.c @@ -11,14 +11,9 @@ #include #include @@ -7572,25 +7712,23 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-utmi.c linux-4.4.62/drivers/clk - struct at91_pmc *pmc = utmi->pmc; - u32 tmp = pmc_read(pmc, AT91_CKGR_UCKR) | AT91_PMC_UPLLEN | - AT91_PMC_UPLLCOUNT | AT91_PMC_BIASEN; -- ++ unsigned int uckr = AT91_PMC_UPLLEN | AT91_PMC_UPLLCOUNT | ++ AT91_PMC_BIASEN; + - pmc_write(pmc, AT91_CKGR_UCKR, tmp); -- ++ regmap_update_bits(utmi->regmap, AT91_CKGR_UCKR, uckr, uckr); + - while (!(pmc_read(pmc, AT91_PMC_SR) & AT91_PMC_LOCKU)) { - enable_irq(utmi->irq); - wait_event(utmi->wait, - pmc_read(pmc, AT91_PMC_SR) & AT91_PMC_LOCKU); - } -+ unsigned int uckr = AT91_PMC_UPLLEN | AT91_PMC_UPLLCOUNT | -+ AT91_PMC_BIASEN; -+ -+ regmap_update_bits(utmi->regmap, AT91_CKGR_UCKR, uckr, uckr); -+ + while (!clk_utmi_ready(utmi->regmap)) + cpu_relax(); return 0; } -@@ -64,18 +52,15 @@ +@@ -64,18 +52,15 @@ static int clk_utmi_prepare(struct clk_hw *hw) static int clk_utmi_is_prepared(struct clk_hw *hw) { struct clk_utmi *utmi = to_clk_utmi(hw); @@ -7611,7 +7749,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-utmi.c linux-4.4.62/drivers/clk } static unsigned long clk_utmi_recalc_rate(struct clk_hw *hw, -@@ -93,10 +78,9 @@ +@@ -93,10 +78,9 @@ static const struct clk_ops utmi_ops = { }; static struct clk * __init @@ -7623,7 +7761,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-utmi.c linux-4.4.62/drivers/clk struct clk_utmi *utmi; struct clk *clk = NULL; struct clk_init_data init; -@@ -112,52 +96,36 @@ +@@ -112,52 +96,36 @@ at91_clk_register_utmi(struct at91_pmc *pmc, unsigned int irq, init.flags = CLK_SET_RATE_GATE; utmi->hw.init = &init; @@ -7685,9 +7823,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/clk-utmi.c linux-4.4.62/drivers/clk -} +CLK_OF_DECLARE(at91sam9x5_clk_utmi, "atmel,at91sam9x5-clk-utmi", + of_at91sam9x5_clk_utmi_setup); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/pmc.c linux-4.4.62/drivers/clk/at91/pmc.c ---- linux-4.4.62.orig/drivers/clk/at91/pmc.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/pmc.c 2017-04-18 17:38:08.046643506 +0200 +diff --git a/drivers/clk/at91/pmc.c b/drivers/clk/at91/pmc.c +index 8476b570779b..526df5ba042d 100644 +--- a/drivers/clk/at91/pmc.c ++++ b/drivers/clk/at91/pmc.c @@ -12,36 +12,13 @@ #include #include @@ -7727,7 +7866,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/pmc.c linux-4.4.62/drivers/clk/at91 int of_at91_get_clk_range(struct device_node *np, const char *propname, struct clk_range *range) { -@@ -64,402 +41,3 @@ +@@ -64,402 +41,3 @@ int of_at91_get_clk_range(struct device_node *np, const char *propname, return 0; } EXPORT_SYMBOL_GPL(of_at91_get_clk_range); @@ -8130,9 +8269,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/pmc.c linux-4.4.62/drivers/clk/at91 -} -CLK_OF_DECLARE(sama5d3_clk_pmc, "atmel,sama5d3-pmc", - of_sama5d3_pmc_setup); -diff -Nur linux-4.4.62.orig/drivers/clk/at91/pmc.h linux-4.4.62/drivers/clk/at91/pmc.h ---- linux-4.4.62.orig/drivers/clk/at91/pmc.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clk/at91/pmc.h 2017-04-18 17:38:08.046643506 +0200 +diff --git a/drivers/clk/at91/pmc.h b/drivers/clk/at91/pmc.h +index f65739272779..5771fff0ee3f 100644 +--- a/drivers/clk/at91/pmc.h ++++ b/drivers/clk/at91/pmc.h @@ -14,8 +14,11 @@ #include @@ -8145,7 +8285,7 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/pmc.h linux-4.4.62/drivers/clk/at91 struct clk_range { unsigned long min; unsigned long max; -@@ -23,102 +26,7 @@ +@@ -23,102 +26,7 @@ struct clk_range { #define CLK_RANGE(MIN, MAX) {.min = MIN, .max = MAX,} @@ -8248,9 +8388,10 @@ diff -Nur linux-4.4.62.orig/drivers/clk/at91/pmc.h linux-4.4.62/drivers/clk/at91 - struct at91_pmc *pmc); - #endif /* __PMC_H_ */ -diff -Nur linux-4.4.62.orig/drivers/clocksource/tcb_clksrc.c linux-4.4.62/drivers/clocksource/tcb_clksrc.c ---- linux-4.4.62.orig/drivers/clocksource/tcb_clksrc.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clocksource/tcb_clksrc.c 2017-04-18 17:38:08.046643506 +0200 +diff --git a/drivers/clocksource/tcb_clksrc.c b/drivers/clocksource/tcb_clksrc.c +index 4da2af9694a2..5b6f57f500b8 100644 +--- a/drivers/clocksource/tcb_clksrc.c ++++ b/drivers/clocksource/tcb_clksrc.c @@ -23,8 +23,7 @@ * this 32 bit free-running counter. the second channel is not used. * @@ -8261,7 +8402,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/tcb_clksrc.c linux-4.4.62/driver * * A boot clocksource and clockevent source are also currently needed, * unless the relevant platforms (ARM/AT91, AVR32/AT32) are changed so -@@ -74,6 +73,8 @@ +@@ -74,6 +73,8 @@ static struct clocksource clksrc = { struct tc_clkevt_device { struct clock_event_device clkevt; struct clk *clk; @@ -8270,7 +8411,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/tcb_clksrc.c linux-4.4.62/driver void __iomem *regs; }; -@@ -82,15 +83,26 @@ +@@ -82,15 +83,26 @@ static struct tc_clkevt_device *to_tc_clkevt(struct clock_event_device *clkevt) return container_of(clkevt, struct tc_clkevt_device, clkevt); } @@ -8304,7 +8445,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/tcb_clksrc.c linux-4.4.62/driver static int tc_shutdown(struct clock_event_device *d) { struct tc_clkevt_device *tcd = to_tc_clkevt(d); -@@ -98,8 +110,14 @@ +@@ -98,8 +110,14 @@ static int tc_shutdown(struct clock_event_device *d) __raw_writel(0xff, regs + ATMEL_TC_REG(2, IDR)); __raw_writel(ATMEL_TC_CLKDIS, regs + ATMEL_TC_REG(2, CCR)); @@ -8320,7 +8461,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/tcb_clksrc.c linux-4.4.62/driver return 0; } -@@ -112,9 +130,9 @@ +@@ -112,9 +130,9 @@ static int tc_set_oneshot(struct clock_event_device *d) if (clockevent_state_oneshot(d) || clockevent_state_periodic(d)) tc_shutdown(d); @@ -8332,7 +8473,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/tcb_clksrc.c linux-4.4.62/driver __raw_writel(timer_clock | ATMEL_TC_CPCSTOP | ATMEL_TC_WAVE | ATMEL_TC_WAVESEL_UP_AUTO, regs + ATMEL_TC_REG(2, CMR)); __raw_writel(ATMEL_TC_CPCS, regs + ATMEL_TC_REG(2, IER)); -@@ -134,12 +152,12 @@ +@@ -134,12 +152,12 @@ static int tc_set_periodic(struct clock_event_device *d) /* By not making the gentime core emulate periodic mode on top * of oneshot, we get lower overhead and improved accuracy. */ @@ -8348,7 +8489,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/tcb_clksrc.c linux-4.4.62/driver /* Enable clock and interrupts on RC compare */ __raw_writel(ATMEL_TC_CPCS, regs + ATMEL_TC_REG(2, IER)); -@@ -166,9 +184,13 @@ +@@ -166,9 +184,13 @@ static struct tc_clkevt_device clkevt = { .features = CLOCK_EVT_FEAT_PERIODIC | CLOCK_EVT_FEAT_ONESHOT, /* Should be lower than at91rm9200's system timer */ @@ -8363,7 +8504,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/tcb_clksrc.c linux-4.4.62/driver .set_state_periodic = tc_set_periodic, .set_state_oneshot = tc_set_oneshot, }, -@@ -188,8 +210,9 @@ +@@ -188,8 +210,9 @@ static irqreturn_t ch2_irq(int irq, void *handle) return IRQ_NONE; } @@ -8374,7 +8515,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/tcb_clksrc.c linux-4.4.62/driver int ret; struct clk *t2_clk = tc->clk[2]; int irq = tc->irq[2]; -@@ -210,7 +233,11 @@ +@@ -210,7 +233,11 @@ static int __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx) clkevt.regs = tc->regs; clkevt.clk = t2_clk; @@ -8387,7 +8528,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/tcb_clksrc.c linux-4.4.62/driver clkevt.clkevt.cpumask = cpumask_of(0); -@@ -221,7 +248,7 @@ +@@ -221,7 +248,7 @@ static int __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx) return ret; } @@ -8396,7 +8537,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/tcb_clksrc.c linux-4.4.62/driver return ret; } -@@ -358,7 +385,11 @@ +@@ -358,7 +385,11 @@ static int __init tcb_clksrc_init(void) goto err_disable_t1; /* channel 2: periodic and oneshot timer support */ @@ -8408,10 +8549,11 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/tcb_clksrc.c linux-4.4.62/driver if (ret) goto err_unregister_clksrc; -diff -Nur linux-4.4.62.orig/drivers/clocksource/timer-atmel-pit.c linux-4.4.62/drivers/clocksource/timer-atmel-pit.c ---- linux-4.4.62.orig/drivers/clocksource/timer-atmel-pit.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clocksource/timer-atmel-pit.c 2017-04-18 17:38:08.046643506 +0200 -@@ -46,6 +46,7 @@ +diff --git a/drivers/clocksource/timer-atmel-pit.c b/drivers/clocksource/timer-atmel-pit.c +index d911c5dca8f1..7a40f7e88468 100644 +--- a/drivers/clocksource/timer-atmel-pit.c ++++ b/drivers/clocksource/timer-atmel-pit.c +@@ -46,6 +46,7 @@ struct pit_data { u32 cycle; u32 cnt; unsigned int irq; @@ -8419,7 +8561,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/timer-atmel-pit.c linux-4.4.62/d struct clk *mck; }; -@@ -96,15 +97,29 @@ +@@ -96,15 +97,29 @@ static int pit_clkevt_shutdown(struct clock_event_device *dev) /* disable irq, leaving the clocksource active */ pit_write(data->base, AT91_PIT_MR, (data->cycle - 1) | AT91_PIT_PITEN); @@ -8449,7 +8591,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/timer-atmel-pit.c linux-4.4.62/d /* update clocksource counter */ data->cnt += data->cycle * PIT_PICNT(pit_read(data->base, AT91_PIT_PIVR)); -@@ -181,7 +196,6 @@ +@@ -181,7 +196,6 @@ static void __init at91sam926x_pit_common_init(struct pit_data *data) { unsigned long pit_rate; unsigned bits; @@ -8457,7 +8599,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/timer-atmel-pit.c linux-4.4.62/d /* * Use our actual MCK to figure out how many MCK/16 ticks per -@@ -206,13 +220,6 @@ +@@ -206,13 +220,6 @@ static void __init at91sam926x_pit_common_init(struct pit_data *data) data->clksrc.flags = CLOCK_SOURCE_IS_CONTINUOUS; clocksource_register_hz(&data->clksrc, pit_rate); @@ -8471,10 +8613,11 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/timer-atmel-pit.c linux-4.4.62/d /* Set up and register clockevents */ data->clkevt.name = "pit"; data->clkevt.features = CLOCK_EVT_FEAT_PERIODIC; -diff -Nur linux-4.4.62.orig/drivers/clocksource/timer-atmel-st.c linux-4.4.62/drivers/clocksource/timer-atmel-st.c ---- linux-4.4.62.orig/drivers/clocksource/timer-atmel-st.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/clocksource/timer-atmel-st.c 2017-04-18 17:38:08.046643506 +0200 -@@ -115,18 +115,29 @@ +diff --git a/drivers/clocksource/timer-atmel-st.c b/drivers/clocksource/timer-atmel-st.c +index 29d21d68df5a..103d0fd70cc4 100644 +--- a/drivers/clocksource/timer-atmel-st.c ++++ b/drivers/clocksource/timer-atmel-st.c +@@ -115,18 +115,29 @@ static void clkdev32k_disable_and_flush_irq(void) last_crtr = read_CRTR(); } @@ -8504,7 +8647,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/timer-atmel-st.c linux-4.4.62/dr /* * ALM for oneshot irqs, set by next_event() * before 32 seconds have passed. -@@ -139,8 +150,16 @@ +@@ -139,8 +150,16 @@ static int clkevt32k_set_oneshot(struct clock_event_device *dev) static int clkevt32k_set_periodic(struct clock_event_device *dev) { @@ -8521,7 +8664,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/timer-atmel-st.c linux-4.4.62/dr /* PIT for periodic irqs; fixed rate of 1/HZ */ irqmask = AT91_ST_PITS; regmap_write(regmap_st, AT91_ST_PIMR, timer_latch); -@@ -198,7 +217,7 @@ +@@ -198,7 +217,7 @@ static void __init atmel_st_timer_init(struct device_node *node) { struct clk *sclk; unsigned int sclk_rate, val; @@ -8530,7 +8673,7 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/timer-atmel-st.c linux-4.4.62/dr regmap_st = syscon_node_to_regmap(node); if (IS_ERR(regmap_st)) -@@ -210,17 +229,10 @@ +@@ -210,17 +229,10 @@ static void __init atmel_st_timer_init(struct device_node *node) regmap_read(regmap_st, AT91_ST_SR, &val); /* Get the interrupts property */ @@ -8550,10 +8693,11 @@ diff -Nur linux-4.4.62.orig/drivers/clocksource/timer-atmel-st.c linux-4.4.62/dr sclk = of_clk_get(node, 0); if (IS_ERR(sclk)) panic(pr_fmt("Unable to get slow clock\n")); -diff -Nur linux-4.4.62.orig/drivers/cpufreq/Kconfig.x86 linux-4.4.62/drivers/cpufreq/Kconfig.x86 ---- linux-4.4.62.orig/drivers/cpufreq/Kconfig.x86 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/cpufreq/Kconfig.x86 2017-04-18 17:38:08.046643506 +0200 -@@ -123,7 +123,7 @@ +diff --git a/drivers/cpufreq/Kconfig.x86 b/drivers/cpufreq/Kconfig.x86 +index c59bdcb83217..8f23161d80be 100644 +--- a/drivers/cpufreq/Kconfig.x86 ++++ b/drivers/cpufreq/Kconfig.x86 +@@ -123,7 +123,7 @@ config X86_POWERNOW_K7_ACPI config X86_POWERNOW_K8 tristate "AMD Opteron/Athlon64 PowerNow!" @@ -8562,10 +8706,11 @@ diff -Nur linux-4.4.62.orig/drivers/cpufreq/Kconfig.x86 linux-4.4.62/drivers/cpu help This adds the CPUFreq driver for K8/early Opteron/Athlon64 processors. Support for K10 and newer processors is now in acpi-cpufreq. -diff -Nur linux-4.4.62.orig/drivers/cpuidle/coupled.c linux-4.4.62/drivers/cpuidle/coupled.c ---- linux-4.4.62.orig/drivers/cpuidle/coupled.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/cpuidle/coupled.c 2017-04-18 17:38:08.046643506 +0200 -@@ -119,7 +119,6 @@ +diff --git a/drivers/cpuidle/coupled.c b/drivers/cpuidle/coupled.c +index 344058f8501a..d5657d50ac40 100644 +--- a/drivers/cpuidle/coupled.c ++++ b/drivers/cpuidle/coupled.c +@@ -119,7 +119,6 @@ struct cpuidle_coupled { #define CPUIDLE_COUPLED_NOT_IDLE (-1) @@ -8573,10 +8718,11 @@ diff -Nur linux-4.4.62.orig/drivers/cpuidle/coupled.c linux-4.4.62/drivers/cpuid static DEFINE_PER_CPU(struct call_single_data, cpuidle_coupled_poke_cb); /* -diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/i915_gem_execbuffer.c linux-4.4.62/drivers/gpu/drm/i915/i915_gem_execbuffer.c ---- linux-4.4.62.orig/drivers/gpu/drm/i915/i915_gem_execbuffer.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/gpu/drm/i915/i915_gem_execbuffer.c 2017-04-18 17:38:08.046643506 +0200 -@@ -1264,7 +1264,9 @@ +diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c +index 6ed7d63a0688..9da7482ad256 100644 +--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c ++++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c +@@ -1264,7 +1264,9 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params, if (ret) return ret; @@ -8586,10 +8732,11 @@ diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/i915_gem_execbuffer.c linux-4.4 i915_gem_execbuffer_move_to_active(vmas, params->request); i915_gem_execbuffer_retire_commands(params); -diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/i915_gem_shrinker.c linux-4.4.62/drivers/gpu/drm/i915/i915_gem_shrinker.c ---- linux-4.4.62.orig/drivers/gpu/drm/i915/i915_gem_shrinker.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/gpu/drm/i915/i915_gem_shrinker.c 2017-04-18 17:38:08.046643506 +0200 -@@ -39,7 +39,7 @@ +diff --git a/drivers/gpu/drm/i915/i915_gem_shrinker.c b/drivers/gpu/drm/i915/i915_gem_shrinker.c +index c0a96f1ee18e..deb1e207fa3c 100644 +--- a/drivers/gpu/drm/i915/i915_gem_shrinker.c ++++ b/drivers/gpu/drm/i915/i915_gem_shrinker.c +@@ -39,7 +39,7 @@ static bool mutex_is_locked_by(struct mutex *mutex, struct task_struct *task) if (!mutex_is_locked(mutex)) return false; @@ -8598,10 +8745,11 @@ diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/i915_gem_shrinker.c linux-4.4.6 return mutex->owner == task; #else /* Since UP may be pre-empted, we cannot assume that we own the lock */ -diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/i915_irq.c linux-4.4.62/drivers/gpu/drm/i915/i915_irq.c ---- linux-4.4.62.orig/drivers/gpu/drm/i915/i915_irq.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/gpu/drm/i915/i915_irq.c 2017-04-18 17:38:08.046643506 +0200 -@@ -812,6 +812,7 @@ +diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c +index b7b0a38acd67..148dcb1349d5 100644 +--- a/drivers/gpu/drm/i915/i915_irq.c ++++ b/drivers/gpu/drm/i915/i915_irq.c +@@ -812,6 +812,7 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); /* preempt_disable_rt() should go right here in PREEMPT_RT patchset. */ @@ -8609,7 +8757,7 @@ diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/i915_irq.c linux-4.4.62/drivers /* Get optional system timestamp before query. */ if (stime) -@@ -863,6 +864,7 @@ +@@ -863,6 +864,7 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, *etime = ktime_get(); /* preempt_enable_rt() should go right here in PREEMPT_RT patchset. */ @@ -8617,10 +8765,11 @@ diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/i915_irq.c linux-4.4.62/drivers spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); -diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/intel_display.c linux-4.4.62/drivers/gpu/drm/i915/intel_display.c ---- linux-4.4.62.orig/drivers/gpu/drm/i915/intel_display.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/gpu/drm/i915/intel_display.c 2017-04-18 17:38:08.050643661 +0200 -@@ -11400,7 +11400,7 @@ +diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c +index 4f5d07bb3511..8ecd5c016dba 100644 +--- a/drivers/gpu/drm/i915/intel_display.c ++++ b/drivers/gpu/drm/i915/intel_display.c +@@ -11400,7 +11400,7 @@ void intel_check_page_flip(struct drm_device *dev, int pipe) struct intel_crtc *intel_crtc = to_intel_crtc(crtc); struct intel_unpin_work *work; @@ -8629,9 +8778,10 @@ diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/intel_display.c linux-4.4.62/dr if (crtc == NULL) return; -diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/intel_sprite.c linux-4.4.62/drivers/gpu/drm/i915/intel_sprite.c ---- linux-4.4.62.orig/drivers/gpu/drm/i915/intel_sprite.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/gpu/drm/i915/intel_sprite.c 2017-04-18 17:38:08.050643661 +0200 +diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c +index 2cc6aa072f4c..b79d33f14868 100644 +--- a/drivers/gpu/drm/i915/intel_sprite.c ++++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -38,6 +38,7 @@ #include "intel_drv.h" #include @@ -8640,7 +8790,7 @@ diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/intel_sprite.c linux-4.4.62/dri static bool format_is_yuv(uint32_t format) -@@ -64,6 +65,8 @@ +@@ -64,6 +65,8 @@ static int usecs_to_scanlines(const struct drm_display_mode *adjusted_mode, 1000 * adjusted_mode->crtc_htotal); } @@ -8649,7 +8799,7 @@ diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/intel_sprite.c linux-4.4.62/dri /** * intel_pipe_update_start() - start update of a set of display registers * @crtc: the crtc of which the registers are going to be updated -@@ -96,7 +99,7 @@ +@@ -96,7 +99,7 @@ void intel_pipe_update_start(struct intel_crtc *crtc) min = vblank_start - usecs_to_scanlines(adjusted_mode, 100); max = vblank_start - 1; @@ -8658,7 +8808,7 @@ diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/intel_sprite.c linux-4.4.62/dri if (min <= 0 || max <= 0) return; -@@ -126,11 +129,11 @@ +@@ -126,11 +129,11 @@ void intel_pipe_update_start(struct intel_crtc *crtc) break; } @@ -8672,7 +8822,7 @@ diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/intel_sprite.c linux-4.4.62/dri } finish_wait(wq, &wait); -@@ -164,7 +167,7 @@ +@@ -164,7 +167,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc) trace_i915_pipe_update_end(crtc, end_vbl_count, scanline_end); @@ -8681,10 +8831,11 @@ diff -Nur linux-4.4.62.orig/drivers/gpu/drm/i915/intel_sprite.c linux-4.4.62/dri if (crtc->debug.start_vbl_count && crtc->debug.start_vbl_count != end_vbl_count) { -diff -Nur linux-4.4.62.orig/drivers/gpu/drm/radeon/radeon_display.c linux-4.4.62/drivers/gpu/drm/radeon/radeon_display.c ---- linux-4.4.62.orig/drivers/gpu/drm/radeon/radeon_display.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/gpu/drm/radeon/radeon_display.c 2017-04-18 17:38:08.050643661 +0200 -@@ -1862,6 +1862,7 @@ +diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c +index 3645b223aa37..642854b2ed2c 100644 +--- a/drivers/gpu/drm/radeon/radeon_display.c ++++ b/drivers/gpu/drm/radeon/radeon_display.c +@@ -1862,6 +1862,7 @@ int radeon_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, struct radeon_device *rdev = dev->dev_private; /* preempt_disable_rt() should go right here in PREEMPT_RT patchset. */ @@ -8692,7 +8843,7 @@ diff -Nur linux-4.4.62.orig/drivers/gpu/drm/radeon/radeon_display.c linux-4.4.62 /* Get optional system timestamp before query. */ if (stime) -@@ -1954,6 +1955,7 @@ +@@ -1954,6 +1955,7 @@ int radeon_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, *etime = ktime_get(); /* preempt_enable_rt() should go right here in PREEMPT_RT patchset. */ @@ -8700,10 +8851,11 @@ diff -Nur linux-4.4.62.orig/drivers/gpu/drm/radeon/radeon_display.c linux-4.4.62 /* Decode into vertical and horizontal scanout position. */ *vpos = position & 0x1fff; -diff -Nur linux-4.4.62.orig/drivers/hv/vmbus_drv.c linux-4.4.62/drivers/hv/vmbus_drv.c ---- linux-4.4.62.orig/drivers/hv/vmbus_drv.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/hv/vmbus_drv.c 2017-04-18 17:38:08.050643661 +0200 -@@ -820,7 +820,7 @@ +diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c +index 802dcb409030..d6d9427860d8 100644 +--- a/drivers/hv/vmbus_drv.c ++++ b/drivers/hv/vmbus_drv.c +@@ -820,7 +820,7 @@ static void vmbus_isr(void) tasklet_schedule(&msg_dpc); } @@ -8712,10 +8864,11 @@ diff -Nur linux-4.4.62.orig/drivers/hv/vmbus_drv.c linux-4.4.62/drivers/hv/vmbus } -diff -Nur linux-4.4.62.orig/drivers/i2c/busses/i2c-omap.c linux-4.4.62/drivers/i2c/busses/i2c-omap.c ---- linux-4.4.62.orig/drivers/i2c/busses/i2c-omap.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/i2c/busses/i2c-omap.c 2017-04-18 17:38:08.050643661 +0200 -@@ -995,15 +995,12 @@ +diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c +index 08d26ba61ed3..46b89dd42b10 100644 +--- a/drivers/i2c/busses/i2c-omap.c ++++ b/drivers/i2c/busses/i2c-omap.c +@@ -995,15 +995,12 @@ omap_i2c_isr(int irq, void *dev_id) u16 mask; u16 stat; @@ -8732,10 +8885,11 @@ diff -Nur linux-4.4.62.orig/drivers/i2c/busses/i2c-omap.c linux-4.4.62/drivers/i return ret; } -diff -Nur linux-4.4.62.orig/drivers/ide/alim15x3.c linux-4.4.62/drivers/ide/alim15x3.c ---- linux-4.4.62.orig/drivers/ide/alim15x3.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/ide/alim15x3.c 2017-04-18 17:38:08.050643661 +0200 -@@ -234,7 +234,7 @@ +diff --git a/drivers/ide/alim15x3.c b/drivers/ide/alim15x3.c +index 36f76e28a0bf..394f142f90c7 100644 +--- a/drivers/ide/alim15x3.c ++++ b/drivers/ide/alim15x3.c +@@ -234,7 +234,7 @@ static int init_chipset_ali15x3(struct pci_dev *dev) isa_dev = pci_get_device(PCI_VENDOR_ID_AL, PCI_DEVICE_ID_AL_M1533, NULL); @@ -8744,7 +8898,7 @@ diff -Nur linux-4.4.62.orig/drivers/ide/alim15x3.c linux-4.4.62/drivers/ide/alim if (m5229_revision < 0xC2) { /* -@@ -325,7 +325,7 @@ +@@ -325,7 +325,7 @@ out: } pci_dev_put(north); pci_dev_put(isa_dev); @@ -8753,10 +8907,11 @@ diff -Nur linux-4.4.62.orig/drivers/ide/alim15x3.c linux-4.4.62/drivers/ide/alim return 0; } -diff -Nur linux-4.4.62.orig/drivers/ide/hpt366.c linux-4.4.62/drivers/ide/hpt366.c ---- linux-4.4.62.orig/drivers/ide/hpt366.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/ide/hpt366.c 2017-04-18 17:38:08.050643661 +0200 -@@ -1241,7 +1241,7 @@ +diff --git a/drivers/ide/hpt366.c b/drivers/ide/hpt366.c +index 696b6c1ec940..0d0a96629b73 100644 +--- a/drivers/ide/hpt366.c ++++ b/drivers/ide/hpt366.c +@@ -1241,7 +1241,7 @@ static int init_dma_hpt366(ide_hwif_t *hwif, dma_old = inb(base + 2); @@ -8765,7 +8920,7 @@ diff -Nur linux-4.4.62.orig/drivers/ide/hpt366.c linux-4.4.62/drivers/ide/hpt366 dma_new = dma_old; pci_read_config_byte(dev, hwif->channel ? 0x4b : 0x43, &masterdma); -@@ -1252,7 +1252,7 @@ +@@ -1252,7 +1252,7 @@ static int init_dma_hpt366(ide_hwif_t *hwif, if (dma_new != dma_old) outb(dma_new, base + 2); @@ -8774,40 +8929,11 @@ diff -Nur linux-4.4.62.orig/drivers/ide/hpt366.c linux-4.4.62/drivers/ide/hpt366 printk(KERN_INFO " %s: BM-DMA at 0x%04lx-0x%04lx\n", hwif->name, base, base + 7); -diff -Nur linux-4.4.62.orig/drivers/ide/ide-io.c linux-4.4.62/drivers/ide/ide-io.c ---- linux-4.4.62.orig/drivers/ide/ide-io.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/ide/ide-io.c 2017-04-18 17:38:08.050643661 +0200 -@@ -659,7 +659,7 @@ - /* disable_irq_nosync ?? */ - disable_irq(hwif->irq); - /* local CPU only, as if we were handling an interrupt */ -- local_irq_disable(); -+ local_irq_disable_nort(); - if (hwif->polling) { - startstop = handler(drive); - } else if (drive_is_ready(drive)) { -diff -Nur linux-4.4.62.orig/drivers/ide/ide-iops.c linux-4.4.62/drivers/ide/ide-iops.c ---- linux-4.4.62.orig/drivers/ide/ide-iops.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/ide/ide-iops.c 2017-04-18 17:38:08.050643661 +0200 -@@ -129,12 +129,12 @@ - if ((stat & ATA_BUSY) == 0) - break; - -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - *rstat = stat; - return -EBUSY; - } - } -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - } - /* - * Allow status to settle, then read it again. -diff -Nur linux-4.4.62.orig/drivers/ide/ide-io-std.c linux-4.4.62/drivers/ide/ide-io-std.c ---- linux-4.4.62.orig/drivers/ide/ide-io-std.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/ide/ide-io-std.c 2017-04-18 17:38:08.050643661 +0200 -@@ -175,7 +175,7 @@ +diff --git a/drivers/ide/ide-io-std.c b/drivers/ide/ide-io-std.c +index 19763977568c..4169433faab5 100644 +--- a/drivers/ide/ide-io-std.c ++++ b/drivers/ide/ide-io-std.c +@@ -175,7 +175,7 @@ void ide_input_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf, unsigned long uninitialized_var(flags); if ((io_32bit & 2) && !mmio) { @@ -8816,7 +8942,7 @@ diff -Nur linux-4.4.62.orig/drivers/ide/ide-io-std.c linux-4.4.62/drivers/ide/id ata_vlb_sync(io_ports->nsect_addr); } -@@ -186,7 +186,7 @@ +@@ -186,7 +186,7 @@ void ide_input_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf, insl(data_addr, buf, words); if ((io_32bit & 2) && !mmio) @@ -8825,7 +8951,7 @@ diff -Nur linux-4.4.62.orig/drivers/ide/ide-io-std.c linux-4.4.62/drivers/ide/id if (((len + 1) & 3) < 2) return; -@@ -219,7 +219,7 @@ +@@ -219,7 +219,7 @@ void ide_output_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf, unsigned long uninitialized_var(flags); if ((io_32bit & 2) && !mmio) { @@ -8834,7 +8960,7 @@ diff -Nur linux-4.4.62.orig/drivers/ide/ide-io-std.c linux-4.4.62/drivers/ide/id ata_vlb_sync(io_ports->nsect_addr); } -@@ -230,7 +230,7 @@ +@@ -230,7 +230,7 @@ void ide_output_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf, outsl(data_addr, buf, words); if ((io_32bit & 2) && !mmio) @@ -8843,10 +8969,43 @@ diff -Nur linux-4.4.62.orig/drivers/ide/ide-io-std.c linux-4.4.62/drivers/ide/id if (((len + 1) & 3) < 2) return; -diff -Nur linux-4.4.62.orig/drivers/ide/ide-probe.c linux-4.4.62/drivers/ide/ide-probe.c ---- linux-4.4.62.orig/drivers/ide/ide-probe.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/ide/ide-probe.c 2017-04-18 17:38:08.050643661 +0200 -@@ -196,10 +196,10 @@ +diff --git a/drivers/ide/ide-io.c b/drivers/ide/ide-io.c +index 669ea1e45795..e12e43e62245 100644 +--- a/drivers/ide/ide-io.c ++++ b/drivers/ide/ide-io.c +@@ -659,7 +659,7 @@ void ide_timer_expiry (unsigned long data) + /* disable_irq_nosync ?? */ + disable_irq(hwif->irq); + /* local CPU only, as if we were handling an interrupt */ +- local_irq_disable(); ++ local_irq_disable_nort(); + if (hwif->polling) { + startstop = handler(drive); + } else if (drive_is_ready(drive)) { +diff --git a/drivers/ide/ide-iops.c b/drivers/ide/ide-iops.c +index 376f2dc410c5..f014dd1b73dc 100644 +--- a/drivers/ide/ide-iops.c ++++ b/drivers/ide/ide-iops.c +@@ -129,12 +129,12 @@ int __ide_wait_stat(ide_drive_t *drive, u8 good, u8 bad, + if ((stat & ATA_BUSY) == 0) + break; + +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + *rstat = stat; + return -EBUSY; + } + } +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + /* + * Allow status to settle, then read it again. +diff --git a/drivers/ide/ide-probe.c b/drivers/ide/ide-probe.c +index 0b63facd1d87..4ceba37afc0c 100644 +--- a/drivers/ide/ide-probe.c ++++ b/drivers/ide/ide-probe.c +@@ -196,10 +196,10 @@ static void do_identify(ide_drive_t *drive, u8 cmd, u16 *id) int bswap = 1; /* local CPU only; some systems need this */ @@ -8859,10 +9018,11 @@ diff -Nur linux-4.4.62.orig/drivers/ide/ide-probe.c linux-4.4.62/drivers/ide/ide drive->dev_flags |= IDE_DFLAG_ID_READ; #ifdef DEBUG -diff -Nur linux-4.4.62.orig/drivers/ide/ide-taskfile.c linux-4.4.62/drivers/ide/ide-taskfile.c ---- linux-4.4.62.orig/drivers/ide/ide-taskfile.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/ide/ide-taskfile.c 2017-04-18 17:38:08.050643661 +0200 -@@ -250,7 +250,7 @@ +diff --git a/drivers/ide/ide-taskfile.c b/drivers/ide/ide-taskfile.c +index a716693417a3..be0568c722d6 100644 +--- a/drivers/ide/ide-taskfile.c ++++ b/drivers/ide/ide-taskfile.c +@@ -250,7 +250,7 @@ void ide_pio_bytes(ide_drive_t *drive, struct ide_cmd *cmd, page_is_high = PageHighMem(page); if (page_is_high) @@ -8871,7 +9031,7 @@ diff -Nur linux-4.4.62.orig/drivers/ide/ide-taskfile.c linux-4.4.62/drivers/ide/ buf = kmap_atomic(page) + offset; -@@ -271,7 +271,7 @@ +@@ -271,7 +271,7 @@ void ide_pio_bytes(ide_drive_t *drive, struct ide_cmd *cmd, kunmap_atomic(buf); if (page_is_high) @@ -8880,7 +9040,7 @@ diff -Nur linux-4.4.62.orig/drivers/ide/ide-taskfile.c linux-4.4.62/drivers/ide/ len -= nr_bytes; } -@@ -414,7 +414,7 @@ +@@ -414,7 +414,7 @@ static ide_startstop_t pre_task_out_intr(ide_drive_t *drive, } if ((drive->dev_flags & IDE_DFLAG_UNMASK) == 0) @@ -8889,10 +9049,11 @@ diff -Nur linux-4.4.62.orig/drivers/ide/ide-taskfile.c linux-4.4.62/drivers/ide/ ide_set_handler(drive, &task_pio_intr, WAIT_WORSTCASE); -diff -Nur linux-4.4.62.orig/drivers/infiniband/ulp/ipoib/ipoib_multicast.c linux-4.4.62/drivers/infiniband/ulp/ipoib/ipoib_multicast.c ---- linux-4.4.62.orig/drivers/infiniband/ulp/ipoib/ipoib_multicast.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/infiniband/ulp/ipoib/ipoib_multicast.c 2017-04-18 17:38:08.050643661 +0200 -@@ -862,7 +862,7 @@ +diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +index 5580ab0b5781..a123d0439c4c 100644 +--- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c ++++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +@@ -862,7 +862,7 @@ void ipoib_mcast_restart_task(struct work_struct *work) ipoib_dbg_mcast(priv, "restarting multicast task\n"); @@ -8901,7 +9062,7 @@ diff -Nur linux-4.4.62.orig/drivers/infiniband/ulp/ipoib/ipoib_multicast.c linux netif_addr_lock(dev); spin_lock(&priv->lock); -@@ -944,7 +944,7 @@ +@@ -944,7 +944,7 @@ void ipoib_mcast_restart_task(struct work_struct *work) spin_unlock(&priv->lock); netif_addr_unlock(dev); @@ -8910,10 +9071,11 @@ diff -Nur linux-4.4.62.orig/drivers/infiniband/ulp/ipoib/ipoib_multicast.c linux /* * make sure the in-flight joins have finished before we attempt -diff -Nur linux-4.4.62.orig/drivers/input/gameport/gameport.c linux-4.4.62/drivers/input/gameport/gameport.c ---- linux-4.4.62.orig/drivers/input/gameport/gameport.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/input/gameport/gameport.c 2017-04-18 17:38:08.054643816 +0200 -@@ -91,13 +91,13 @@ +diff --git a/drivers/input/gameport/gameport.c b/drivers/input/gameport/gameport.c +index 4a2a9e370be7..e970d9afd179 100644 +--- a/drivers/input/gameport/gameport.c ++++ b/drivers/input/gameport/gameport.c +@@ -91,13 +91,13 @@ static int gameport_measure_speed(struct gameport *gameport) tx = ~0; for (i = 0; i < 50; i++) { @@ -8929,7 +9091,7 @@ diff -Nur linux-4.4.62.orig/drivers/input/gameport/gameport.c linux-4.4.62/drive udelay(i * 10); t = (t2 - t1) - (t3 - t2); if (t < tx) -@@ -124,12 +124,12 @@ +@@ -124,12 +124,12 @@ static int old_gameport_measure_speed(struct gameport *gameport) tx = 1 << 30; for(i = 0; i < 50; i++) { @@ -8944,7 +9106,7 @@ diff -Nur linux-4.4.62.orig/drivers/input/gameport/gameport.c linux-4.4.62/drive udelay(i * 10); if ((t = DELTA(t2,t1) - DELTA(t3,t2)) < tx) tx = t; } -@@ -148,11 +148,11 @@ +@@ -148,11 +148,11 @@ static int old_gameport_measure_speed(struct gameport *gameport) tx = 1 << 30; for(i = 0; i < 50; i++) { @@ -8958,10 +9120,11 @@ diff -Nur linux-4.4.62.orig/drivers/input/gameport/gameport.c linux-4.4.62/drive udelay(i * 10); if (t2 - t1 < tx) tx = t2 - t1; } -diff -Nur linux-4.4.62.orig/drivers/iommu/amd_iommu.c linux-4.4.62/drivers/iommu/amd_iommu.c ---- linux-4.4.62.orig/drivers/iommu/amd_iommu.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/iommu/amd_iommu.c 2017-04-18 17:38:08.054643816 +0200 -@@ -2022,10 +2022,10 @@ +diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c +index a0ef57483ebb..b08732bae11a 100644 +--- a/drivers/iommu/amd_iommu.c ++++ b/drivers/iommu/amd_iommu.c +@@ -2022,10 +2022,10 @@ static int __attach_device(struct iommu_dev_data *dev_data, int ret; /* @@ -8975,7 +9138,7 @@ diff -Nur linux-4.4.62.orig/drivers/iommu/amd_iommu.c linux-4.4.62/drivers/iommu /* lock domain */ spin_lock(&domain->lock); -@@ -2188,10 +2188,10 @@ +@@ -2188,10 +2188,10 @@ static void __detach_device(struct iommu_dev_data *dev_data) struct protection_domain *domain; /* @@ -8989,10 +9152,11 @@ diff -Nur linux-4.4.62.orig/drivers/iommu/amd_iommu.c linux-4.4.62/drivers/iommu if (WARN_ON(!dev_data->domain)) return; -diff -Nur linux-4.4.62.orig/drivers/leds/trigger/Kconfig linux-4.4.62/drivers/leds/trigger/Kconfig ---- linux-4.4.62.orig/drivers/leds/trigger/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/leds/trigger/Kconfig 2017-04-18 17:38:08.054643816 +0200 -@@ -61,7 +61,7 @@ +diff --git a/drivers/leds/trigger/Kconfig b/drivers/leds/trigger/Kconfig +index 5bda6a9b56bb..d6286584c807 100644 +--- a/drivers/leds/trigger/Kconfig ++++ b/drivers/leds/trigger/Kconfig +@@ -61,7 +61,7 @@ config LEDS_TRIGGER_BACKLIGHT config LEDS_TRIGGER_CPU bool "LED CPU Trigger" @@ -9001,9 +9165,10 @@ diff -Nur linux-4.4.62.orig/drivers/leds/trigger/Kconfig linux-4.4.62/drivers/le help This allows LEDs to be controlled by active CPUs. This shows the active CPUs across an array of LEDs so you can see which -diff -Nur linux-4.4.62.orig/drivers/md/bcache/Kconfig linux-4.4.62/drivers/md/bcache/Kconfig ---- linux-4.4.62.orig/drivers/md/bcache/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/md/bcache/Kconfig 2017-04-18 17:38:08.054643816 +0200 +diff --git a/drivers/md/bcache/Kconfig b/drivers/md/bcache/Kconfig +index 4d200883c505..98b64ed5cb81 100644 +--- a/drivers/md/bcache/Kconfig ++++ b/drivers/md/bcache/Kconfig @@ -1,6 +1,7 @@ config BCACHE @@ -9012,10 +9177,11 @@ diff -Nur linux-4.4.62.orig/drivers/md/bcache/Kconfig linux-4.4.62/drivers/md/bc ---help--- Allows a block device to be used as cache for other devices; uses a btree for indexing and the layout is optimized for SSDs. -diff -Nur linux-4.4.62.orig/drivers/md/dm.c linux-4.4.62/drivers/md/dm.c ---- linux-4.4.62.orig/drivers/md/dm.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/md/dm.c 2017-04-18 17:38:08.054643816 +0200 -@@ -2185,7 +2185,7 @@ +diff --git a/drivers/md/dm.c b/drivers/md/dm.c +index 320eb3c4bb6b..aba2d71149e3 100644 +--- a/drivers/md/dm.c ++++ b/drivers/md/dm.c +@@ -2185,7 +2185,7 @@ static void dm_request_fn(struct request_queue *q) /* Establish tio->ti before queuing work (map_tio_request) */ tio->ti = ti; queue_kthread_work(&md->kworker, &tio->work); @@ -9024,10 +9190,11 @@ diff -Nur linux-4.4.62.orig/drivers/md/dm.c linux-4.4.62/drivers/md/dm.c } goto out; -diff -Nur linux-4.4.62.orig/drivers/md/raid5.c linux-4.4.62/drivers/md/raid5.c ---- linux-4.4.62.orig/drivers/md/raid5.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/md/raid5.c 2017-04-18 17:38:08.054643816 +0200 -@@ -1920,8 +1920,9 @@ +diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c +index 4384b46cee1a..6415f094a4eb 100644 +--- a/drivers/md/raid5.c ++++ b/drivers/md/raid5.c +@@ -1920,8 +1920,9 @@ static void raid_run_ops(struct stripe_head *sh, unsigned long ops_request) struct raid5_percpu *percpu; unsigned long cpu; @@ -9038,7 +9205,7 @@ diff -Nur linux-4.4.62.orig/drivers/md/raid5.c linux-4.4.62/drivers/md/raid5.c if (test_bit(STRIPE_OP_BIOFILL, &ops_request)) { ops_run_biofill(sh); overlap_clear++; -@@ -1977,7 +1978,8 @@ +@@ -1977,7 +1978,8 @@ static void raid_run_ops(struct stripe_head *sh, unsigned long ops_request) if (test_and_clear_bit(R5_Overlap, &dev->flags)) wake_up(&sh->raid_conf->wait_for_overlap); } @@ -9048,7 +9215,7 @@ diff -Nur linux-4.4.62.orig/drivers/md/raid5.c linux-4.4.62/drivers/md/raid5.c } static struct stripe_head *alloc_stripe(struct kmem_cache *sc, gfp_t gfp) -@@ -6414,6 +6416,7 @@ +@@ -6416,6 +6418,7 @@ static int raid5_alloc_percpu(struct r5conf *conf) __func__, cpu); break; } @@ -9056,10 +9223,11 @@ diff -Nur linux-4.4.62.orig/drivers/md/raid5.c linux-4.4.62/drivers/md/raid5.c } put_online_cpus(); -diff -Nur linux-4.4.62.orig/drivers/md/raid5.h linux-4.4.62/drivers/md/raid5.h ---- linux-4.4.62.orig/drivers/md/raid5.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/md/raid5.h 2017-04-18 17:38:08.054643816 +0200 -@@ -504,6 +504,7 @@ +diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h +index 517d4b68a1be..efe91887ecd7 100644 +--- a/drivers/md/raid5.h ++++ b/drivers/md/raid5.h +@@ -504,6 +504,7 @@ struct r5conf { int recovery_disabled; /* per cpu variables */ struct raid5_percpu { @@ -9067,10 +9235,11 @@ diff -Nur linux-4.4.62.orig/drivers/md/raid5.h linux-4.4.62/drivers/md/raid5.h struct page *spare_page; /* Used when checking P/Q in raid6 */ struct flex_array *scribble; /* space for constructing buffer * lists and performing address -diff -Nur linux-4.4.62.orig/drivers/media/platform/vsp1/vsp1_video.c linux-4.4.62/drivers/media/platform/vsp1/vsp1_video.c ---- linux-4.4.62.orig/drivers/media/platform/vsp1/vsp1_video.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/media/platform/vsp1/vsp1_video.c 2017-04-18 17:38:08.054643816 +0200 -@@ -520,7 +520,7 @@ +diff --git a/drivers/media/platform/vsp1/vsp1_video.c b/drivers/media/platform/vsp1/vsp1_video.c +index 5ce88e1f5d71..b4f8cd74ecb8 100644 +--- a/drivers/media/platform/vsp1/vsp1_video.c ++++ b/drivers/media/platform/vsp1/vsp1_video.c +@@ -520,7 +520,7 @@ static bool vsp1_pipeline_stopped(struct vsp1_pipeline *pipe) bool stopped; spin_lock_irqsave(&pipe->irqlock, flags); @@ -9079,9 +9248,97 @@ diff -Nur linux-4.4.62.orig/drivers/media/platform/vsp1/vsp1_video.c linux-4.4.6 spin_unlock_irqrestore(&pipe->irqlock, flags); return stopped; -diff -Nur linux-4.4.62.orig/drivers/misc/hwlat_detector.c linux-4.4.62/drivers/misc/hwlat_detector.c ---- linux-4.4.62.orig/drivers/misc/hwlat_detector.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/drivers/misc/hwlat_detector.c 2017-04-18 17:38:08.054643816 +0200 +diff --git a/drivers/misc/Kconfig b/drivers/misc/Kconfig +index 4bf7d50b1bc7..6f7e99ad6e29 100644 +--- a/drivers/misc/Kconfig ++++ b/drivers/misc/Kconfig +@@ -54,6 +54,7 @@ config AD525X_DPOT_SPI + config ATMEL_TCLIB + bool "Atmel AT32/AT91 Timer/Counter Library" + depends on (AVR32 || ARCH_AT91) ++ default y if PREEMPT_RT_FULL + help + Select this if you want a library to allocate the Timer/Counter + blocks found on many Atmel processors. This facilitates using +@@ -69,8 +70,7 @@ config ATMEL_TCB_CLKSRC + are combined to make a single 32-bit timer. + + When GENERIC_CLOCKEVENTS is defined, the third timer channel +- may be used as a clock event device supporting oneshot mode +- (delays of up to two seconds) based on the 32 KiHz clock. ++ may be used as a clock event device supporting oneshot mode. + + config ATMEL_TCB_CLKSRC_BLOCK + int +@@ -84,6 +84,15 @@ config ATMEL_TCB_CLKSRC_BLOCK + TC can be used for other purposes, such as PWM generation and + interval timing. + ++config ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK ++ bool "TC Block use 32 KiHz clock" ++ depends on ATMEL_TCB_CLKSRC ++ default y if !PREEMPT_RT_FULL ++ help ++ Select this to use 32 KiHz base clock rate as TC block clock ++ source for clock events. ++ ++ + config DUMMY_IRQ + tristate "Dummy IRQ handler" + default n +@@ -113,6 +122,35 @@ config IBM_ASM + for information on the specific driver level and support statement + for your IBM server. + ++config HWLAT_DETECTOR ++ tristate "Testing module to detect hardware-induced latencies" ++ depends on DEBUG_FS ++ depends on RING_BUFFER ++ default m ++ ---help--- ++ A simple hardware latency detector. Use this module to detect ++ large latencies introduced by the behavior of the underlying ++ system firmware external to Linux. We do this using periodic ++ use of stop_machine to grab all available CPUs and measure ++ for unexplainable gaps in the CPU timestamp counter(s). By ++ default, the module is not enabled until the "enable" file ++ within the "hwlat_detector" debugfs directory is toggled. ++ ++ This module is often used to detect SMI (System Management ++ Interrupts) on x86 systems, though is not x86 specific. To ++ this end, we default to using a sample window of 1 second, ++ during which we will sample for 0.5 seconds. If an SMI or ++ similar event occurs during that time, it is recorded ++ into an 8K samples global ring buffer until retreived. ++ ++ WARNING: This software should never be enabled (it can be built ++ but should not be turned on after it is loaded) in a production ++ environment where high latencies are a concern since the ++ sampling mechanism actually introduces latencies for ++ regular tasks while the CPU(s) are being held. ++ ++ If unsure, say N ++ + config PHANTOM + tristate "Sensable PHANToM (PCI)" + depends on PCI +diff --git a/drivers/misc/Makefile b/drivers/misc/Makefile +index 537d7f3b78da..ec4aecba0656 100644 +--- a/drivers/misc/Makefile ++++ b/drivers/misc/Makefile +@@ -39,6 +39,7 @@ obj-$(CONFIG_C2PORT) += c2port/ + obj-$(CONFIG_HMC6352) += hmc6352.o + obj-y += eeprom/ + obj-y += cb710/ ++obj-$(CONFIG_HWLAT_DETECTOR) += hwlat_detector.o + obj-$(CONFIG_SPEAR13XX_PCIE_GADGET) += spear13xx_pcie_gadget.o + obj-$(CONFIG_VMWARE_BALLOON) += vmw_balloon.o + obj-$(CONFIG_ARM_CHARLCD) += arm-charlcd.o +diff --git a/drivers/misc/hwlat_detector.c b/drivers/misc/hwlat_detector.c +new file mode 100644 +index 000000000000..52f5ad5fd9c0 +--- /dev/null ++++ b/drivers/misc/hwlat_detector.c @@ -0,0 +1,1240 @@ +/* + * hwlat_detector.c - A simple Hardware Latency detector. @@ -10318,99 +10575,16 @@ diff -Nur linux-4.4.62.orig/drivers/misc/hwlat_detector.c linux-4.4.62/drivers/m + + free_debugfs(); + ring_buffer_free(ring_buffer); /* free up the ring buffer */ -+ -+} -+ -+module_init(detector_init); -+module_exit(detector_exit); -diff -Nur linux-4.4.62.orig/drivers/misc/Kconfig linux-4.4.62/drivers/misc/Kconfig ---- linux-4.4.62.orig/drivers/misc/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/misc/Kconfig 2017-04-18 17:38:08.054643816 +0200 -@@ -54,6 +54,7 @@ - config ATMEL_TCLIB - bool "Atmel AT32/AT91 Timer/Counter Library" - depends on (AVR32 || ARCH_AT91) -+ default y if PREEMPT_RT_FULL - help - Select this if you want a library to allocate the Timer/Counter - blocks found on many Atmel processors. This facilitates using -@@ -69,8 +70,7 @@ - are combined to make a single 32-bit timer. - - When GENERIC_CLOCKEVENTS is defined, the third timer channel -- may be used as a clock event device supporting oneshot mode -- (delays of up to two seconds) based on the 32 KiHz clock. -+ may be used as a clock event device supporting oneshot mode. - - config ATMEL_TCB_CLKSRC_BLOCK - int -@@ -84,6 +84,15 @@ - TC can be used for other purposes, such as PWM generation and - interval timing. - -+config ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK -+ bool "TC Block use 32 KiHz clock" -+ depends on ATMEL_TCB_CLKSRC -+ default y if !PREEMPT_RT_FULL -+ help -+ Select this to use 32 KiHz base clock rate as TC block clock -+ source for clock events. -+ -+ - config DUMMY_IRQ - tristate "Dummy IRQ handler" - default n -@@ -113,6 +122,35 @@ - for information on the specific driver level and support statement - for your IBM server. - -+config HWLAT_DETECTOR -+ tristate "Testing module to detect hardware-induced latencies" -+ depends on DEBUG_FS -+ depends on RING_BUFFER -+ default m -+ ---help--- -+ A simple hardware latency detector. Use this module to detect -+ large latencies introduced by the behavior of the underlying -+ system firmware external to Linux. We do this using periodic -+ use of stop_machine to grab all available CPUs and measure -+ for unexplainable gaps in the CPU timestamp counter(s). By -+ default, the module is not enabled until the "enable" file -+ within the "hwlat_detector" debugfs directory is toggled. -+ -+ This module is often used to detect SMI (System Management -+ Interrupts) on x86 systems, though is not x86 specific. To -+ this end, we default to using a sample window of 1 second, -+ during which we will sample for 0.5 seconds. If an SMI or -+ similar event occurs during that time, it is recorded -+ into an 8K samples global ring buffer until retreived. -+ -+ WARNING: This software should never be enabled (it can be built -+ but should not be turned on after it is loaded) in a production -+ environment where high latencies are a concern since the -+ sampling mechanism actually introduces latencies for -+ regular tasks while the CPU(s) are being held. -+ -+ If unsure, say N -+ - config PHANTOM - tristate "Sensable PHANToM (PCI)" - depends on PCI -diff -Nur linux-4.4.62.orig/drivers/misc/Makefile linux-4.4.62/drivers/misc/Makefile ---- linux-4.4.62.orig/drivers/misc/Makefile 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/misc/Makefile 2017-04-18 17:38:08.054643816 +0200 -@@ -39,6 +39,7 @@ - obj-$(CONFIG_HMC6352) += hmc6352.o - obj-y += eeprom/ - obj-y += cb710/ -+obj-$(CONFIG_HWLAT_DETECTOR) += hwlat_detector.o - obj-$(CONFIG_SPEAR13XX_PCIE_GADGET) += spear13xx_pcie_gadget.o - obj-$(CONFIG_VMWARE_BALLOON) += vmw_balloon.o - obj-$(CONFIG_ARM_CHARLCD) += arm-charlcd.o -diff -Nur linux-4.4.62.orig/drivers/mmc/host/mmci.c linux-4.4.62/drivers/mmc/host/mmci.c ---- linux-4.4.62.orig/drivers/mmc/host/mmci.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/mmc/host/mmci.c 2017-04-18 17:38:08.054643816 +0200 -@@ -1155,15 +1155,12 @@ ++ ++} ++ ++module_init(detector_init); ++module_exit(detector_exit); +diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c +index acece3299756..58ea04a03fa9 100644 +--- a/drivers/mmc/host/mmci.c ++++ b/drivers/mmc/host/mmci.c +@@ -1155,15 +1155,12 @@ static irqreturn_t mmci_pio_irq(int irq, void *dev_id) struct sg_mapping_iter *sg_miter = &host->sg_miter; struct variant_data *variant = host->variant; void __iomem *base = host->base; @@ -10426,7 +10600,7 @@ diff -Nur linux-4.4.62.orig/drivers/mmc/host/mmci.c linux-4.4.62/drivers/mmc/hos do { unsigned int remain, len; char *buffer; -@@ -1203,8 +1200,6 @@ +@@ -1203,8 +1200,6 @@ static irqreturn_t mmci_pio_irq(int irq, void *dev_id) sg_miter_stop(sg_miter); @@ -10435,10 +10609,11 @@ diff -Nur linux-4.4.62.orig/drivers/mmc/host/mmci.c linux-4.4.62/drivers/mmc/hos /* * If we have less than the fifo 'half-full' threshold to transfer, * trigger a PIO interrupt as soon as any data is available. -diff -Nur linux-4.4.62.orig/drivers/net/ethernet/3com/3c59x.c linux-4.4.62/drivers/net/ethernet/3com/3c59x.c ---- linux-4.4.62.orig/drivers/net/ethernet/3com/3c59x.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/net/ethernet/3com/3c59x.c 2017-04-18 17:38:08.058643971 +0200 -@@ -842,9 +842,9 @@ +diff --git a/drivers/net/ethernet/3com/3c59x.c b/drivers/net/ethernet/3com/3c59x.c +index 2839af00f20c..4348b9c850d3 100644 +--- a/drivers/net/ethernet/3com/3c59x.c ++++ b/drivers/net/ethernet/3com/3c59x.c +@@ -842,9 +842,9 @@ static void poll_vortex(struct net_device *dev) { struct vortex_private *vp = netdev_priv(dev); unsigned long flags; @@ -10450,7 +10625,7 @@ diff -Nur linux-4.4.62.orig/drivers/net/ethernet/3com/3c59x.c linux-4.4.62/drive } #endif -@@ -1916,12 +1916,12 @@ +@@ -1916,12 +1916,12 @@ static void vortex_tx_timeout(struct net_device *dev) * Block interrupts because vortex_interrupt does a bare spin_lock() */ unsigned long flags; @@ -10465,10 +10640,11 @@ diff -Nur linux-4.4.62.orig/drivers/net/ethernet/3com/3c59x.c linux-4.4.62/drive } } -diff -Nur linux-4.4.62.orig/drivers/net/ethernet/atheros/atl1c/atl1c_main.c linux-4.4.62/drivers/net/ethernet/atheros/atl1c/atl1c_main.c ---- linux-4.4.62.orig/drivers/net/ethernet/atheros/atl1c/atl1c_main.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/net/ethernet/atheros/atl1c/atl1c_main.c 2017-04-18 17:38:08.058643971 +0200 -@@ -2221,11 +2221,7 @@ +diff --git a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c +index 8b5988e210d5..cf9928ccdd7e 100644 +--- a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c ++++ b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c +@@ -2221,11 +2221,7 @@ static netdev_tx_t atl1c_xmit_frame(struct sk_buff *skb, } tpd_req = atl1c_cal_tpd_req(skb); @@ -10481,10 +10657,11 @@ diff -Nur linux-4.4.62.orig/drivers/net/ethernet/atheros/atl1c/atl1c_main.c linu if (atl1c_tpd_avail(adapter, type) < tpd_req) { /* no enough descriptor, just stop queue */ -diff -Nur linux-4.4.62.orig/drivers/net/ethernet/atheros/atl1e/atl1e_main.c linux-4.4.62/drivers/net/ethernet/atheros/atl1e/atl1e_main.c ---- linux-4.4.62.orig/drivers/net/ethernet/atheros/atl1e/atl1e_main.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/net/ethernet/atheros/atl1e/atl1e_main.c 2017-04-18 17:38:08.058643971 +0200 -@@ -1880,8 +1880,7 @@ +diff --git a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c +index 59a03a193e83..734f7a7ad2c3 100644 +--- a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c ++++ b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c +@@ -1880,8 +1880,7 @@ static netdev_tx_t atl1e_xmit_frame(struct sk_buff *skb, return NETDEV_TX_OK; } tpd_req = atl1e_cal_tdp_req(skb); @@ -10494,10 +10671,11 @@ diff -Nur linux-4.4.62.orig/drivers/net/ethernet/atheros/atl1e/atl1e_main.c linu if (atl1e_tpd_avail(adapter) < tpd_req) { /* no enough descriptor, just stop queue */ -diff -Nur linux-4.4.62.orig/drivers/net/ethernet/chelsio/cxgb/sge.c linux-4.4.62/drivers/net/ethernet/chelsio/cxgb/sge.c ---- linux-4.4.62.orig/drivers/net/ethernet/chelsio/cxgb/sge.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/net/ethernet/chelsio/cxgb/sge.c 2017-04-18 17:38:08.058643971 +0200 -@@ -1664,8 +1664,7 @@ +diff --git a/drivers/net/ethernet/chelsio/cxgb/sge.c b/drivers/net/ethernet/chelsio/cxgb/sge.c +index 526ea74e82d9..86f467a2c485 100644 +--- a/drivers/net/ethernet/chelsio/cxgb/sge.c ++++ b/drivers/net/ethernet/chelsio/cxgb/sge.c +@@ -1664,8 +1664,7 @@ static int t1_sge_tx(struct sk_buff *skb, struct adapter *adapter, struct cmdQ *q = &sge->cmdQ[qid]; unsigned int credits, pidx, genbit, count, use_sched_skb = 0; @@ -10507,10 +10685,11 @@ diff -Nur linux-4.4.62.orig/drivers/net/ethernet/chelsio/cxgb/sge.c linux-4.4.62 reclaim_completed_tx(sge, q); -diff -Nur linux-4.4.62.orig/drivers/net/ethernet/neterion/s2io.c linux-4.4.62/drivers/net/ethernet/neterion/s2io.c ---- linux-4.4.62.orig/drivers/net/ethernet/neterion/s2io.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/net/ethernet/neterion/s2io.c 2017-04-18 17:38:08.058643971 +0200 -@@ -4084,12 +4084,7 @@ +diff --git a/drivers/net/ethernet/neterion/s2io.c b/drivers/net/ethernet/neterion/s2io.c +index 9ba975853ec6..813cfa698160 100644 +--- a/drivers/net/ethernet/neterion/s2io.c ++++ b/drivers/net/ethernet/neterion/s2io.c +@@ -4084,12 +4084,7 @@ static netdev_tx_t s2io_xmit(struct sk_buff *skb, struct net_device *dev) [skb->priority & (MAX_TX_FIFOS - 1)]; fifo = &mac_control->fifos[queue]; @@ -10524,10 +10703,11 @@ diff -Nur linux-4.4.62.orig/drivers/net/ethernet/neterion/s2io.c linux-4.4.62/dr if (sp->config.multiq) { if (__netif_subqueue_stopped(dev, fifo->fifo_no)) { -diff -Nur linux-4.4.62.orig/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c linux-4.4.62/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c ---- linux-4.4.62.orig/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c 2017-04-18 17:38:08.058643971 +0200 -@@ -2137,10 +2137,8 @@ +diff --git a/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c b/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c +index 3b98b263bad0..ca4add749410 100644 +--- a/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c ++++ b/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c +@@ -2137,10 +2137,8 @@ static int pch_gbe_xmit_frame(struct sk_buff *skb, struct net_device *netdev) struct pch_gbe_tx_ring *tx_ring = adapter->tx_ring; unsigned long flags; @@ -10540,10 +10720,11 @@ diff -Nur linux-4.4.62.orig/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c if (unlikely(!PCH_GBE_DESC_UNUSED(tx_ring))) { netif_stop_queue(netdev); spin_unlock_irqrestore(&tx_ring->tx_lock, flags); -diff -Nur linux-4.4.62.orig/drivers/net/ethernet/realtek/8139too.c linux-4.4.62/drivers/net/ethernet/realtek/8139too.c ---- linux-4.4.62.orig/drivers/net/ethernet/realtek/8139too.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/net/ethernet/realtek/8139too.c 2017-04-18 17:38:08.058643971 +0200 -@@ -2229,7 +2229,7 @@ +diff --git a/drivers/net/ethernet/realtek/8139too.c b/drivers/net/ethernet/realtek/8139too.c +index ef668d300800..d987d571fdd6 100644 +--- a/drivers/net/ethernet/realtek/8139too.c ++++ b/drivers/net/ethernet/realtek/8139too.c +@@ -2229,7 +2229,7 @@ static void rtl8139_poll_controller(struct net_device *dev) struct rtl8139_private *tp = netdev_priv(dev); const int irq = tp->pci_dev->irq; @@ -10552,10 +10733,11 @@ diff -Nur linux-4.4.62.orig/drivers/net/ethernet/realtek/8139too.c linux-4.4.62/ rtl8139_interrupt(irq, dev); enable_irq(irq); } -diff -Nur linux-4.4.62.orig/drivers/net/ethernet/tehuti/tehuti.c linux-4.4.62/drivers/net/ethernet/tehuti/tehuti.c ---- linux-4.4.62.orig/drivers/net/ethernet/tehuti/tehuti.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/net/ethernet/tehuti/tehuti.c 2017-04-18 17:38:08.058643971 +0200 -@@ -1629,13 +1629,8 @@ +diff --git a/drivers/net/ethernet/tehuti/tehuti.c b/drivers/net/ethernet/tehuti/tehuti.c +index 14c9d1baa85c..e1a5305418a8 100644 +--- a/drivers/net/ethernet/tehuti/tehuti.c ++++ b/drivers/net/ethernet/tehuti/tehuti.c +@@ -1629,13 +1629,8 @@ static netdev_tx_t bdx_tx_transmit(struct sk_buff *skb, unsigned long flags; ENTER; @@ -10571,10 +10753,11 @@ diff -Nur linux-4.4.62.orig/drivers/net/ethernet/tehuti/tehuti.c linux-4.4.62/dr /* build tx descriptor */ BDX_ASSERT(f->m.wptr >= f->m.memsz); /* started with valid wptr */ -diff -Nur linux-4.4.62.orig/drivers/net/rionet.c linux-4.4.62/drivers/net/rionet.c ---- linux-4.4.62.orig/drivers/net/rionet.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/net/rionet.c 2017-04-18 17:38:08.062644126 +0200 -@@ -174,11 +174,7 @@ +diff --git a/drivers/net/rionet.c b/drivers/net/rionet.c +index e7034c55e796..2e4ee0f912bf 100644 +--- a/drivers/net/rionet.c ++++ b/drivers/net/rionet.c +@@ -174,11 +174,7 @@ static int rionet_start_xmit(struct sk_buff *skb, struct net_device *ndev) unsigned long flags; int add_num = 1; @@ -10587,10 +10770,11 @@ diff -Nur linux-4.4.62.orig/drivers/net/rionet.c linux-4.4.62/drivers/net/rionet if (is_multicast_ether_addr(eth->h_dest)) add_num = nets[rnet->mport->id].nact; -diff -Nur linux-4.4.62.orig/drivers/net/wireless/orinoco/orinoco_usb.c linux-4.4.62/drivers/net/wireless/orinoco/orinoco_usb.c ---- linux-4.4.62.orig/drivers/net/wireless/orinoco/orinoco_usb.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/net/wireless/orinoco/orinoco_usb.c 2017-04-18 17:38:08.062644126 +0200 -@@ -697,7 +697,7 @@ +diff --git a/drivers/net/wireless/orinoco/orinoco_usb.c b/drivers/net/wireless/orinoco/orinoco_usb.c +index f2cd513d54b2..6c0f4c9638a2 100644 +--- a/drivers/net/wireless/orinoco/orinoco_usb.c ++++ b/drivers/net/wireless/orinoco/orinoco_usb.c +@@ -697,7 +697,7 @@ static void ezusb_req_ctx_wait(struct ezusb_priv *upriv, while (!ctx->done.done && msecs--) udelay(1000); } else { @@ -10599,10 +10783,11 @@ diff -Nur linux-4.4.62.orig/drivers/net/wireless/orinoco/orinoco_usb.c linux-4.4 ctx->done.done); } break; -diff -Nur linux-4.4.62.orig/drivers/pci/access.c linux-4.4.62/drivers/pci/access.c ---- linux-4.4.62.orig/drivers/pci/access.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/pci/access.c 2017-04-18 17:38:08.062644126 +0200 -@@ -561,7 +561,7 @@ +diff --git a/drivers/pci/access.c b/drivers/pci/access.c +index 59ac36fe7c42..7a45a20af78a 100644 +--- a/drivers/pci/access.c ++++ b/drivers/pci/access.c +@@ -561,7 +561,7 @@ void pci_cfg_access_unlock(struct pci_dev *dev) WARN_ON(!dev->block_cfg_access); dev->block_cfg_access = 0; @@ -10611,10 +10796,11 @@ diff -Nur linux-4.4.62.orig/drivers/pci/access.c linux-4.4.62/drivers/pci/access raw_spin_unlock_irqrestore(&pci_lock, flags); } EXPORT_SYMBOL_GPL(pci_cfg_access_unlock); -diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/drivers/pinctrl/qcom/pinctrl-msm.c ---- linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/pinctrl/qcom/pinctrl-msm.c 2017-04-18 17:38:08.062644126 +0200 -@@ -60,7 +60,7 @@ +diff --git a/drivers/pinctrl/qcom/pinctrl-msm.c b/drivers/pinctrl/qcom/pinctrl-msm.c +index 9736f9be5447..5fe9b173dcb3 100644 +--- a/drivers/pinctrl/qcom/pinctrl-msm.c ++++ b/drivers/pinctrl/qcom/pinctrl-msm.c +@@ -60,7 +60,7 @@ struct msm_pinctrl { struct notifier_block restart_nb; int irq; @@ -10623,7 +10809,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv DECLARE_BITMAP(dual_edge_irqs, MAX_NR_GPIO); DECLARE_BITMAP(enabled_irqs, MAX_NR_GPIO); -@@ -156,14 +156,14 @@ +@@ -156,14 +156,14 @@ static int msm_pinmux_set_mux(struct pinctrl_dev *pctldev, if (WARN_ON(i == g->nfuncs)) return -EINVAL; @@ -10640,7 +10826,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv return 0; } -@@ -326,14 +326,14 @@ +@@ -326,14 +326,14 @@ static int msm_config_group_set(struct pinctrl_dev *pctldev, break; case PIN_CONFIG_OUTPUT: /* set output value */ @@ -10657,7 +10843,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv /* enable output */ arg = 1; -@@ -354,12 +354,12 @@ +@@ -354,12 +354,12 @@ static int msm_config_group_set(struct pinctrl_dev *pctldev, return -EINVAL; } @@ -10672,7 +10858,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv } return 0; -@@ -387,13 +387,13 @@ +@@ -387,13 +387,13 @@ static int msm_gpio_direction_input(struct gpio_chip *chip, unsigned offset) g = &pctrl->soc->groups[offset]; @@ -10688,7 +10874,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv return 0; } -@@ -407,7 +407,7 @@ +@@ -407,7 +407,7 @@ static int msm_gpio_direction_output(struct gpio_chip *chip, unsigned offset, in g = &pctrl->soc->groups[offset]; @@ -10697,7 +10883,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv val = readl(pctrl->regs + g->io_reg); if (value) -@@ -420,7 +420,7 @@ +@@ -420,7 +420,7 @@ static int msm_gpio_direction_output(struct gpio_chip *chip, unsigned offset, in val |= BIT(g->oe_bit); writel(val, pctrl->regs + g->ctl_reg); @@ -10706,7 +10892,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv return 0; } -@@ -446,7 +446,7 @@ +@@ -446,7 +446,7 @@ static void msm_gpio_set(struct gpio_chip *chip, unsigned offset, int value) g = &pctrl->soc->groups[offset]; @@ -10715,7 +10901,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv val = readl(pctrl->regs + g->io_reg); if (value) -@@ -455,7 +455,7 @@ +@@ -455,7 +455,7 @@ static void msm_gpio_set(struct gpio_chip *chip, unsigned offset, int value) val &= ~BIT(g->out_bit); writel(val, pctrl->regs + g->io_reg); @@ -10724,7 +10910,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv } #ifdef CONFIG_DEBUG_FS -@@ -574,7 +574,7 @@ +@@ -574,7 +574,7 @@ static void msm_gpio_irq_mask(struct irq_data *d) g = &pctrl->soc->groups[d->hwirq]; @@ -10733,7 +10919,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv val = readl(pctrl->regs + g->intr_cfg_reg); val &= ~BIT(g->intr_enable_bit); -@@ -582,7 +582,7 @@ +@@ -582,7 +582,7 @@ static void msm_gpio_irq_mask(struct irq_data *d) clear_bit(d->hwirq, pctrl->enabled_irqs); @@ -10742,7 +10928,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv } static void msm_gpio_irq_unmask(struct irq_data *d) -@@ -595,7 +595,7 @@ +@@ -595,7 +595,7 @@ static void msm_gpio_irq_unmask(struct irq_data *d) g = &pctrl->soc->groups[d->hwirq]; @@ -10751,7 +10937,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv val = readl(pctrl->regs + g->intr_cfg_reg); val |= BIT(g->intr_enable_bit); -@@ -603,7 +603,7 @@ +@@ -603,7 +603,7 @@ static void msm_gpio_irq_unmask(struct irq_data *d) set_bit(d->hwirq, pctrl->enabled_irqs); @@ -10760,7 +10946,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv } static void msm_gpio_irq_ack(struct irq_data *d) -@@ -616,7 +616,7 @@ +@@ -616,7 +616,7 @@ static void msm_gpio_irq_ack(struct irq_data *d) g = &pctrl->soc->groups[d->hwirq]; @@ -10769,7 +10955,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv val = readl(pctrl->regs + g->intr_status_reg); if (g->intr_ack_high) -@@ -628,7 +628,7 @@ +@@ -628,7 +628,7 @@ static void msm_gpio_irq_ack(struct irq_data *d) if (test_bit(d->hwirq, pctrl->dual_edge_irqs)) msm_gpio_update_dual_edge_pos(pctrl, g, d); @@ -10778,7 +10964,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv } static int msm_gpio_irq_set_type(struct irq_data *d, unsigned int type) -@@ -641,7 +641,7 @@ +@@ -641,7 +641,7 @@ static int msm_gpio_irq_set_type(struct irq_data *d, unsigned int type) g = &pctrl->soc->groups[d->hwirq]; @@ -10787,7 +10973,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv /* * For hw without possibility of detecting both edges -@@ -715,7 +715,7 @@ +@@ -715,7 +715,7 @@ static int msm_gpio_irq_set_type(struct irq_data *d, unsigned int type) if (test_bit(d->hwirq, pctrl->dual_edge_irqs)) msm_gpio_update_dual_edge_pos(pctrl, g, d); @@ -10796,7 +10982,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv if (type & (IRQ_TYPE_LEVEL_LOW | IRQ_TYPE_LEVEL_HIGH)) irq_set_handler_locked(d, handle_level_irq); -@@ -731,11 +731,11 @@ +@@ -731,11 +731,11 @@ static int msm_gpio_irq_set_wake(struct irq_data *d, unsigned int on) struct msm_pinctrl *pctrl = to_msm_pinctrl(gc); unsigned long flags; @@ -10810,7 +10996,7 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv return 0; } -@@ -881,7 +881,7 @@ +@@ -881,7 +881,7 @@ int msm_pinctrl_probe(struct platform_device *pdev, pctrl->soc = soc_data; pctrl->chip = msm_gpio_template; @@ -10819,10 +11005,11 @@ diff -Nur linux-4.4.62.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.4.62/driv res = platform_get_resource(pdev, IORESOURCE_MEM, 0); pctrl->regs = devm_ioremap_resource(&pdev->dev, res); -diff -Nur linux-4.4.62.orig/drivers/scsi/fcoe/fcoe.c linux-4.4.62/drivers/scsi/fcoe/fcoe.c ---- linux-4.4.62.orig/drivers/scsi/fcoe/fcoe.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/scsi/fcoe/fcoe.c 2017-04-18 17:38:08.062644126 +0200 -@@ -1286,7 +1286,7 @@ +diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c +index f4424063b860..cbbbebd86c6e 100644 +--- a/drivers/scsi/fcoe/fcoe.c ++++ b/drivers/scsi/fcoe/fcoe.c +@@ -1286,7 +1286,7 @@ static void fcoe_percpu_thread_destroy(unsigned int cpu) struct sk_buff *skb; #ifdef CONFIG_SMP struct fcoe_percpu_s *p0; @@ -10831,7 +11018,7 @@ diff -Nur linux-4.4.62.orig/drivers/scsi/fcoe/fcoe.c linux-4.4.62/drivers/scsi/f #endif /* CONFIG_SMP */ FCOE_DBG("Destroying receive thread for CPU %d\n", cpu); -@@ -1342,7 +1342,7 @@ +@@ -1342,7 +1342,7 @@ static void fcoe_percpu_thread_destroy(unsigned int cpu) kfree_skb(skb); spin_unlock_bh(&p->fcoe_rx_list.lock); } @@ -10840,7 +11027,7 @@ diff -Nur linux-4.4.62.orig/drivers/scsi/fcoe/fcoe.c linux-4.4.62/drivers/scsi/f #else /* * This a non-SMP scenario where the singular Rx thread is -@@ -1566,11 +1566,11 @@ +@@ -1566,11 +1566,11 @@ err2: static int fcoe_alloc_paged_crc_eof(struct sk_buff *skb, int tlen) { struct fcoe_percpu_s *fps; @@ -10855,7 +11042,7 @@ diff -Nur linux-4.4.62.orig/drivers/scsi/fcoe/fcoe.c linux-4.4.62/drivers/scsi/f return rc; } -@@ -1766,11 +1766,11 @@ +@@ -1766,11 +1766,11 @@ static inline int fcoe_filter_frames(struct fc_lport *lport, return 0; } @@ -10869,7 +11056,7 @@ diff -Nur linux-4.4.62.orig/drivers/scsi/fcoe/fcoe.c linux-4.4.62/drivers/scsi/f return -EINVAL; } -@@ -1814,7 +1814,7 @@ +@@ -1814,7 +1814,7 @@ static void fcoe_recv_frame(struct sk_buff *skb) */ hp = (struct fcoe_hdr *) skb_network_header(skb); @@ -10878,7 +11065,7 @@ diff -Nur linux-4.4.62.orig/drivers/scsi/fcoe/fcoe.c linux-4.4.62/drivers/scsi/f if (unlikely(FC_FCOE_DECAPS_VER(hp) != FC_FCOE_VER)) { if (stats->ErrorFrames < 5) printk(KERN_WARNING "fcoe: FCoE version " -@@ -1846,13 +1846,13 @@ +@@ -1846,13 +1846,13 @@ static void fcoe_recv_frame(struct sk_buff *skb) goto drop; if (!fcoe_filter_frames(lport, fp)) { @@ -10894,10 +11081,11 @@ diff -Nur linux-4.4.62.orig/drivers/scsi/fcoe/fcoe.c linux-4.4.62/drivers/scsi/f kfree_skb(skb); } -diff -Nur linux-4.4.62.orig/drivers/scsi/fcoe/fcoe_ctlr.c linux-4.4.62/drivers/scsi/fcoe/fcoe_ctlr.c ---- linux-4.4.62.orig/drivers/scsi/fcoe/fcoe_ctlr.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/scsi/fcoe/fcoe_ctlr.c 2017-04-18 17:38:08.062644126 +0200 -@@ -831,7 +831,7 @@ +diff --git a/drivers/scsi/fcoe/fcoe_ctlr.c b/drivers/scsi/fcoe/fcoe_ctlr.c +index 34a1b1f333b4..d91131210695 100644 +--- a/drivers/scsi/fcoe/fcoe_ctlr.c ++++ b/drivers/scsi/fcoe/fcoe_ctlr.c +@@ -831,7 +831,7 @@ static unsigned long fcoe_ctlr_age_fcfs(struct fcoe_ctlr *fip) INIT_LIST_HEAD(&del_list); @@ -10906,7 +11094,7 @@ diff -Nur linux-4.4.62.orig/drivers/scsi/fcoe/fcoe_ctlr.c linux-4.4.62/drivers/s list_for_each_entry_safe(fcf, next, &fip->fcfs, list) { deadline = fcf->time + fcf->fka_period + fcf->fka_period / 2; -@@ -867,7 +867,7 @@ +@@ -867,7 +867,7 @@ static unsigned long fcoe_ctlr_age_fcfs(struct fcoe_ctlr *fip) sel_time = fcf->time; } } @@ -10915,10 +11103,11 @@ diff -Nur linux-4.4.62.orig/drivers/scsi/fcoe/fcoe_ctlr.c linux-4.4.62/drivers/s list_for_each_entry_safe(fcf, next, &del_list, list) { /* Removes fcf from current list */ -diff -Nur linux-4.4.62.orig/drivers/scsi/libfc/fc_exch.c linux-4.4.62/drivers/scsi/libfc/fc_exch.c ---- linux-4.4.62.orig/drivers/scsi/libfc/fc_exch.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/scsi/libfc/fc_exch.c 2017-04-18 17:38:08.062644126 +0200 -@@ -814,10 +814,10 @@ +diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c +index 30f9ef0c0d4f..6c686bc01a82 100644 +--- a/drivers/scsi/libfc/fc_exch.c ++++ b/drivers/scsi/libfc/fc_exch.c +@@ -814,10 +814,10 @@ static struct fc_exch *fc_exch_em_alloc(struct fc_lport *lport, } memset(ep, 0, sizeof(*ep)); @@ -10931,10 +11120,11 @@ diff -Nur linux-4.4.62.orig/drivers/scsi/libfc/fc_exch.c linux-4.4.62/drivers/sc /* peek cache of free slot */ if (pool->left != FC_XID_UNKNOWN) { -diff -Nur linux-4.4.62.orig/drivers/scsi/libsas/sas_ata.c linux-4.4.62/drivers/scsi/libsas/sas_ata.c ---- linux-4.4.62.orig/drivers/scsi/libsas/sas_ata.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/scsi/libsas/sas_ata.c 2017-04-18 17:38:08.062644126 +0200 -@@ -190,7 +190,7 @@ +diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c +index 6f5e2720ffad..ee8a8ed49b89 100644 +--- a/drivers/scsi/libsas/sas_ata.c ++++ b/drivers/scsi/libsas/sas_ata.c +@@ -190,7 +190,7 @@ static unsigned int sas_ata_qc_issue(struct ata_queued_cmd *qc) /* TODO: audit callers to ensure they are ready for qc_issue to * unconditionally re-enable interrupts */ @@ -10943,7 +11133,7 @@ diff -Nur linux-4.4.62.orig/drivers/scsi/libsas/sas_ata.c linux-4.4.62/drivers/s spin_unlock(ap->lock); /* If the device fell off, no sense in issuing commands */ -@@ -255,7 +255,7 @@ +@@ -255,7 +255,7 @@ static unsigned int sas_ata_qc_issue(struct ata_queued_cmd *qc) out: spin_lock(ap->lock); @@ -10952,10 +11142,11 @@ diff -Nur linux-4.4.62.orig/drivers/scsi/libsas/sas_ata.c linux-4.4.62/drivers/s return ret; } -diff -Nur linux-4.4.62.orig/drivers/scsi/qla2xxx/qla_inline.h linux-4.4.62/drivers/scsi/qla2xxx/qla_inline.h ---- linux-4.4.62.orig/drivers/scsi/qla2xxx/qla_inline.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/scsi/qla2xxx/qla_inline.h 2017-04-18 17:38:08.062644126 +0200 -@@ -59,12 +59,12 @@ +diff --git a/drivers/scsi/qla2xxx/qla_inline.h b/drivers/scsi/qla2xxx/qla_inline.h +index fee9eb7c8a60..b42d4adc42dc 100644 +--- a/drivers/scsi/qla2xxx/qla_inline.h ++++ b/drivers/scsi/qla2xxx/qla_inline.h +@@ -59,12 +59,12 @@ qla2x00_poll(struct rsp_que *rsp) { unsigned long flags; struct qla_hw_data *ha = rsp->hw; @@ -10970,9 +11161,10 @@ diff -Nur linux-4.4.62.orig/drivers/scsi/qla2xxx/qla_inline.h linux-4.4.62/drive } static inline uint8_t * -diff -Nur linux-4.4.62.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.4.62/drivers/thermal/x86_pkg_temp_thermal.c ---- linux-4.4.62.orig/drivers/thermal/x86_pkg_temp_thermal.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/thermal/x86_pkg_temp_thermal.c 2017-04-18 17:38:08.062644126 +0200 +diff --git a/drivers/thermal/x86_pkg_temp_thermal.c b/drivers/thermal/x86_pkg_temp_thermal.c +index 7fc919f7da4d..e03fa17b8670 100644 +--- a/drivers/thermal/x86_pkg_temp_thermal.c ++++ b/drivers/thermal/x86_pkg_temp_thermal.c @@ -29,6 +29,7 @@ #include #include @@ -10981,7 +11173,7 @@ diff -Nur linux-4.4.62.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.4.62/ #include #include -@@ -352,7 +353,7 @@ +@@ -352,7 +353,7 @@ static void pkg_temp_thermal_threshold_work_fn(struct work_struct *work) } } @@ -10990,7 +11182,7 @@ diff -Nur linux-4.4.62.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.4.62/ { unsigned long flags; int cpu = smp_processor_id(); -@@ -369,7 +370,7 @@ +@@ -369,7 +370,7 @@ static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) pkg_work_scheduled[phy_id]) { disable_pkg_thres_interrupt(); spin_unlock_irqrestore(&pkg_work_lock, flags); @@ -10999,7 +11191,7 @@ diff -Nur linux-4.4.62.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.4.62/ } pkg_work_scheduled[phy_id] = 1; spin_unlock_irqrestore(&pkg_work_lock, flags); -@@ -378,9 +379,48 @@ +@@ -378,9 +379,48 @@ static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) schedule_delayed_work_on(cpu, &per_cpu(pkg_temp_thermal_threshold_work, cpu), msecs_to_jiffies(notify_delay_ms)); @@ -11048,7 +11240,7 @@ diff -Nur linux-4.4.62.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.4.62/ static int find_siblings_cpu(int cpu) { int i; -@@ -584,6 +624,9 @@ +@@ -584,6 +624,9 @@ static int __init pkg_temp_thermal_init(void) if (!x86_match_cpu(pkg_temp_thermal_ids)) return -ENODEV; @@ -11058,7 +11250,7 @@ diff -Nur linux-4.4.62.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.4.62/ spin_lock_init(&pkg_work_lock); platform_thermal_package_notify = pkg_temp_thermal_platform_thermal_notify; -@@ -608,7 +651,7 @@ +@@ -608,7 +651,7 @@ err_ret: kfree(pkg_work_scheduled); platform_thermal_package_notify = NULL; platform_thermal_package_rate_control = NULL; @@ -11067,7 +11259,7 @@ diff -Nur linux-4.4.62.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.4.62/ return -ENODEV; } -@@ -633,6 +676,7 @@ +@@ -633,6 +676,7 @@ static void __exit pkg_temp_thermal_exit(void) mutex_unlock(&phy_dev_list_mutex); platform_thermal_package_notify = NULL; platform_thermal_package_rate_control = NULL; @@ -11075,10 +11267,11 @@ diff -Nur linux-4.4.62.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.4.62/ for_each_online_cpu(i) cancel_delayed_work_sync( &per_cpu(pkg_temp_thermal_threshold_work, i)); -diff -Nur linux-4.4.62.orig/drivers/tty/serial/8250/8250_core.c linux-4.4.62/drivers/tty/serial/8250/8250_core.c ---- linux-4.4.62.orig/drivers/tty/serial/8250/8250_core.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/tty/serial/8250/8250_core.c 2017-04-18 17:38:08.062644126 +0200 -@@ -58,7 +58,16 @@ +diff --git a/drivers/tty/serial/8250/8250_core.c b/drivers/tty/serial/8250/8250_core.c +index 39126460c1f5..af7701ca4d48 100644 +--- a/drivers/tty/serial/8250/8250_core.c ++++ b/drivers/tty/serial/8250/8250_core.c +@@ -58,7 +58,16 @@ static struct uart_driver serial8250_reg; static unsigned int skip_txen_test; /* force skip of txen test at init time */ @@ -11096,9 +11289,10 @@ diff -Nur linux-4.4.62.orig/drivers/tty/serial/8250/8250_core.c linux-4.4.62/dri #include /* -diff -Nur linux-4.4.62.orig/drivers/tty/serial/8250/8250_port.c linux-4.4.62/drivers/tty/serial/8250/8250_port.c ---- linux-4.4.62.orig/drivers/tty/serial/8250/8250_port.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/tty/serial/8250/8250_port.c 2017-04-18 17:38:08.062644126 +0200 +diff --git a/drivers/tty/serial/8250/8250_port.c b/drivers/tty/serial/8250/8250_port.c +index 56ccbcefdd85..a0b9e854672c 100644 +--- a/drivers/tty/serial/8250/8250_port.c ++++ b/drivers/tty/serial/8250/8250_port.c @@ -35,6 +35,7 @@ #include #include @@ -11107,7 +11301,7 @@ diff -Nur linux-4.4.62.orig/drivers/tty/serial/8250/8250_port.c linux-4.4.62/dri #include #include -@@ -2843,9 +2844,9 @@ +@@ -2843,9 +2844,9 @@ void serial8250_console_write(struct uart_8250_port *up, const char *s, serial8250_rpm_get(up); @@ -11119,10 +11313,11 @@ diff -Nur linux-4.4.62.orig/drivers/tty/serial/8250/8250_port.c linux-4.4.62/dri locked = spin_trylock_irqsave(&port->lock, flags); else spin_lock_irqsave(&port->lock, flags); -diff -Nur linux-4.4.62.orig/drivers/tty/serial/amba-pl011.c linux-4.4.62/drivers/tty/serial/amba-pl011.c ---- linux-4.4.62.orig/drivers/tty/serial/amba-pl011.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/tty/serial/amba-pl011.c 2017-04-18 17:38:08.062644126 +0200 -@@ -2067,13 +2067,19 @@ +diff --git a/drivers/tty/serial/amba-pl011.c b/drivers/tty/serial/amba-pl011.c +index 899a77187bde..3ff6363b3751 100644 +--- a/drivers/tty/serial/amba-pl011.c ++++ b/drivers/tty/serial/amba-pl011.c +@@ -2067,13 +2067,19 @@ pl011_console_write(struct console *co, const char *s, unsigned int count) clk_enable(uap->clk); @@ -11145,7 +11340,7 @@ diff -Nur linux-4.4.62.orig/drivers/tty/serial/amba-pl011.c linux-4.4.62/drivers /* * First save the CR then disable the interrupts -@@ -2098,8 +2104,7 @@ +@@ -2098,8 +2104,7 @@ pl011_console_write(struct console *co, const char *s, unsigned int count) writew(old_cr, uap->port.membase + UART011_CR); if (locked) @@ -11155,10 +11350,11 @@ diff -Nur linux-4.4.62.orig/drivers/tty/serial/amba-pl011.c linux-4.4.62/drivers clk_disable(uap->clk); } -diff -Nur linux-4.4.62.orig/drivers/tty/serial/omap-serial.c linux-4.4.62/drivers/tty/serial/omap-serial.c ---- linux-4.4.62.orig/drivers/tty/serial/omap-serial.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/tty/serial/omap-serial.c 2017-04-18 17:38:08.062644126 +0200 -@@ -1257,13 +1257,10 @@ +diff --git a/drivers/tty/serial/omap-serial.c b/drivers/tty/serial/omap-serial.c +index de1c143b475f..3fef536a1844 100644 +--- a/drivers/tty/serial/omap-serial.c ++++ b/drivers/tty/serial/omap-serial.c +@@ -1257,13 +1257,10 @@ serial_omap_console_write(struct console *co, const char *s, pm_runtime_get_sync(up->dev); @@ -11175,7 +11371,7 @@ diff -Nur linux-4.4.62.orig/drivers/tty/serial/omap-serial.c linux-4.4.62/driver /* * First save the IER then disable the interrupts -@@ -1292,8 +1289,7 @@ +@@ -1292,8 +1289,7 @@ serial_omap_console_write(struct console *co, const char *s, pm_runtime_mark_last_busy(up->dev); pm_runtime_put_autosuspend(up->dev); if (locked) @@ -11185,10 +11381,11 @@ diff -Nur linux-4.4.62.orig/drivers/tty/serial/omap-serial.c linux-4.4.62/driver } static int __init -diff -Nur linux-4.4.62.orig/drivers/usb/core/hcd.c linux-4.4.62/drivers/usb/core/hcd.c ---- linux-4.4.62.orig/drivers/usb/core/hcd.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/usb/core/hcd.c 2017-04-18 17:38:08.066644281 +0200 -@@ -1738,9 +1738,9 @@ +diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c +index c3f4f2ab7b33..ae1ffb5f7610 100644 +--- a/drivers/usb/core/hcd.c ++++ b/drivers/usb/core/hcd.c +@@ -1738,9 +1738,9 @@ static void __usb_hcd_giveback_urb(struct urb *urb) * and no one may trigger the above deadlock situation when * running complete() in tasklet. */ @@ -11200,10 +11397,11 @@ diff -Nur linux-4.4.62.orig/drivers/usb/core/hcd.c linux-4.4.62/drivers/usb/core usb_anchor_resume_wakeups(anchor); atomic_dec(&urb->use_count); -diff -Nur linux-4.4.62.orig/drivers/usb/gadget/function/f_fs.c linux-4.4.62/drivers/usb/gadget/function/f_fs.c ---- linux-4.4.62.orig/drivers/usb/gadget/function/f_fs.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/usb/gadget/function/f_fs.c 2017-04-18 17:38:08.066644281 +0200 -@@ -1404,7 +1404,7 @@ +diff --git a/drivers/usb/gadget/function/f_fs.c b/drivers/usb/gadget/function/f_fs.c +index 9ad5145d3103..bf883695cdfb 100644 +--- a/drivers/usb/gadget/function/f_fs.c ++++ b/drivers/usb/gadget/function/f_fs.c +@@ -1404,7 +1404,7 @@ static void ffs_data_put(struct ffs_data *ffs) pr_info("%s(): freeing\n", __func__); ffs_data_clear(ffs); BUG_ON(waitqueue_active(&ffs->ev.waitq) || @@ -11212,10 +11410,11 @@ diff -Nur linux-4.4.62.orig/drivers/usb/gadget/function/f_fs.c linux-4.4.62/driv kfree(ffs->dev_name); kfree(ffs); } -diff -Nur linux-4.4.62.orig/drivers/usb/gadget/legacy/inode.c linux-4.4.62/drivers/usb/gadget/legacy/inode.c ---- linux-4.4.62.orig/drivers/usb/gadget/legacy/inode.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/usb/gadget/legacy/inode.c 2017-04-18 17:38:08.066644281 +0200 -@@ -345,7 +345,7 @@ +diff --git a/drivers/usb/gadget/legacy/inode.c b/drivers/usb/gadget/legacy/inode.c +index de014436fb22..e52700c7555e 100644 +--- a/drivers/usb/gadget/legacy/inode.c ++++ b/drivers/usb/gadget/legacy/inode.c +@@ -345,7 +345,7 @@ ep_io (struct ep_data *epdata, void *buf, unsigned len) spin_unlock_irq (&epdata->dev->lock); if (likely (value == 0)) { @@ -11224,7 +11423,7 @@ diff -Nur linux-4.4.62.orig/drivers/usb/gadget/legacy/inode.c linux-4.4.62/drive if (value != 0) { spin_lock_irq (&epdata->dev->lock); if (likely (epdata->ep != NULL)) { -@@ -354,7 +354,7 @@ +@@ -354,7 +354,7 @@ ep_io (struct ep_data *epdata, void *buf, unsigned len) usb_ep_dequeue (epdata->ep, epdata->req); spin_unlock_irq (&epdata->dev->lock); @@ -11233,9 +11432,10 @@ diff -Nur linux-4.4.62.orig/drivers/usb/gadget/legacy/inode.c linux-4.4.62/drive if (epdata->status == -ECONNRESET) epdata->status = -EINTR; } else { -diff -Nur linux-4.4.62.orig/drivers/usb/gadget/udc/atmel_usba_udc.c linux-4.4.62/drivers/usb/gadget/udc/atmel_usba_udc.c ---- linux-4.4.62.orig/drivers/usb/gadget/udc/atmel_usba_udc.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/usb/gadget/udc/atmel_usba_udc.c 2017-04-18 17:38:08.066644281 +0200 +diff --git a/drivers/usb/gadget/udc/atmel_usba_udc.c b/drivers/usb/gadget/udc/atmel_usba_udc.c +index f92f5aff0dd5..f9bba26e3655 100644 +--- a/drivers/usb/gadget/udc/atmel_usba_udc.c ++++ b/drivers/usb/gadget/udc/atmel_usba_udc.c @@ -17,7 +17,9 @@ #include #include @@ -11246,7 +11446,7 @@ diff -Nur linux-4.4.62.orig/drivers/usb/gadget/udc/atmel_usba_udc.c linux-4.4.62 #include #include #include -@@ -1888,20 +1890,15 @@ +@@ -1888,20 +1890,15 @@ static int atmel_usba_stop(struct usb_gadget *gadget) #ifdef CONFIG_OF static void at91sam9rl_toggle_bias(struct usba_udc *udc, int is_on) { @@ -11272,7 +11472,7 @@ diff -Nur linux-4.4.62.orig/drivers/usb/gadget/udc/atmel_usba_udc.c linux-4.4.62 } static const struct usba_udc_errata at91sam9rl_errata = { -@@ -1938,6 +1935,9 @@ +@@ -1938,6 +1935,9 @@ static struct usba_ep * atmel_udc_of_init(struct platform_device *pdev, return ERR_PTR(-EINVAL); udc->errata = match->data; @@ -11282,10 +11482,11 @@ diff -Nur linux-4.4.62.orig/drivers/usb/gadget/udc/atmel_usba_udc.c linux-4.4.62 udc->num_ep = 0; -diff -Nur linux-4.4.62.orig/drivers/usb/gadget/udc/atmel_usba_udc.h linux-4.4.62/drivers/usb/gadget/udc/atmel_usba_udc.h ---- linux-4.4.62.orig/drivers/usb/gadget/udc/atmel_usba_udc.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/drivers/usb/gadget/udc/atmel_usba_udc.h 2017-04-18 17:38:08.066644281 +0200 -@@ -354,6 +354,8 @@ +diff --git a/drivers/usb/gadget/udc/atmel_usba_udc.h b/drivers/usb/gadget/udc/atmel_usba_udc.h +index ea448a344767..3e1c9d589dfa 100644 +--- a/drivers/usb/gadget/udc/atmel_usba_udc.h ++++ b/drivers/usb/gadget/udc/atmel_usba_udc.h +@@ -354,6 +354,8 @@ struct usba_udc { struct dentry *debugfs_root; struct dentry *debugfs_regs; #endif @@ -11294,9 +11495,10 @@ diff -Nur linux-4.4.62.orig/drivers/usb/gadget/udc/atmel_usba_udc.h linux-4.4.62 }; static inline struct usba_ep *to_usba_ep(struct usb_ep *ep) -diff -Nur linux-4.4.62.orig/fs/aio.c linux-4.4.62/fs/aio.c ---- linux-4.4.62.orig/fs/aio.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/fs/aio.c 2017-04-18 17:38:08.066644281 +0200 +diff --git a/fs/aio.c b/fs/aio.c +index fe4f49212b99..c3194afdc3df 100644 +--- a/fs/aio.c ++++ b/fs/aio.c @@ -40,6 +40,7 @@ #include #include @@ -11305,7 +11507,7 @@ diff -Nur linux-4.4.62.orig/fs/aio.c linux-4.4.62/fs/aio.c #include #include -@@ -115,7 +116,7 @@ +@@ -115,7 +116,7 @@ struct kioctx { struct page **ring_pages; long nr_pages; @@ -11314,7 +11516,7 @@ diff -Nur linux-4.4.62.orig/fs/aio.c linux-4.4.62/fs/aio.c /* * signals when all in-flight requests are done -@@ -258,6 +259,7 @@ +@@ -258,6 +259,7 @@ static int __init aio_setup(void) .mount = aio_mount, .kill_sb = kill_anon_super, }; @@ -11322,7 +11524,7 @@ diff -Nur linux-4.4.62.orig/fs/aio.c linux-4.4.62/fs/aio.c aio_mnt = kern_mount(&aio_fs); if (IS_ERR(aio_mnt)) panic("Failed to create aio fs mount."); -@@ -573,9 +575,9 @@ +@@ -573,9 +575,9 @@ static int kiocb_cancel(struct aio_kiocb *kiocb) return cancel(&kiocb->common); } @@ -11334,7 +11536,7 @@ diff -Nur linux-4.4.62.orig/fs/aio.c linux-4.4.62/fs/aio.c pr_debug("freeing %p\n", ctx); -@@ -594,8 +596,8 @@ +@@ -594,8 +596,8 @@ static void free_ioctx_reqs(struct percpu_ref *ref) if (ctx->rq_wait && atomic_dec_and_test(&ctx->rq_wait->count)) complete(&ctx->rq_wait->comp); @@ -11345,7 +11547,7 @@ diff -Nur linux-4.4.62.orig/fs/aio.c linux-4.4.62/fs/aio.c } /* -@@ -603,9 +605,9 @@ +@@ -603,9 +605,9 @@ static void free_ioctx_reqs(struct percpu_ref *ref) * and ctx->users has dropped to 0, so we know no more kiocbs can be submitted - * now it's safe to cancel any that need to be. */ @@ -11357,7 +11559,7 @@ diff -Nur linux-4.4.62.orig/fs/aio.c linux-4.4.62/fs/aio.c struct aio_kiocb *req; spin_lock_irq(&ctx->ctx_lock); -@@ -624,6 +626,14 @@ +@@ -624,6 +626,14 @@ static void free_ioctx_users(struct percpu_ref *ref) percpu_ref_put(&ctx->reqs); } @@ -11372,9 +11574,10 @@ diff -Nur linux-4.4.62.orig/fs/aio.c linux-4.4.62/fs/aio.c static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm) { unsigned i, new_nr; -diff -Nur linux-4.4.62.orig/fs/autofs4/autofs_i.h linux-4.4.62/fs/autofs4/autofs_i.h ---- linux-4.4.62.orig/fs/autofs4/autofs_i.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/fs/autofs4/autofs_i.h 2017-04-18 17:38:08.066644281 +0200 +diff --git a/fs/autofs4/autofs_i.h b/fs/autofs4/autofs_i.h +index 502d3892d8a4..05af8d3e6e88 100644 +--- a/fs/autofs4/autofs_i.h ++++ b/fs/autofs4/autofs_i.h @@ -34,6 +34,7 @@ #include #include @@ -11383,10 +11586,11 @@ diff -Nur linux-4.4.62.orig/fs/autofs4/autofs_i.h linux-4.4.62/fs/autofs4/autofs #include #include -diff -Nur linux-4.4.62.orig/fs/autofs4/expire.c linux-4.4.62/fs/autofs4/expire.c ---- linux-4.4.62.orig/fs/autofs4/expire.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/fs/autofs4/expire.c 2017-04-18 17:38:08.066644281 +0200 -@@ -150,7 +150,7 @@ +diff --git a/fs/autofs4/expire.c b/fs/autofs4/expire.c +index 7a5a598a2d94..d08bcdc30566 100644 +--- a/fs/autofs4/expire.c ++++ b/fs/autofs4/expire.c +@@ -150,7 +150,7 @@ again: parent = p->d_parent; if (!spin_trylock(&parent->d_lock)) { spin_unlock(&p->d_lock); @@ -11395,10 +11599,11 @@ diff -Nur linux-4.4.62.orig/fs/autofs4/expire.c linux-4.4.62/fs/autofs4/expire.c goto relock; } spin_unlock(&p->d_lock); -diff -Nur linux-4.4.62.orig/fs/buffer.c linux-4.4.62/fs/buffer.c ---- linux-4.4.62.orig/fs/buffer.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/fs/buffer.c 2017-04-18 17:38:08.066644281 +0200 -@@ -305,8 +305,7 @@ +diff --git a/fs/buffer.c b/fs/buffer.c +index 4f4cd959da7c..72b27e17b907 100644 +--- a/fs/buffer.c ++++ b/fs/buffer.c +@@ -305,8 +305,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate) * decide that the page is now completely done. */ first = page_buffers(page); @@ -11408,7 +11613,7 @@ diff -Nur linux-4.4.62.orig/fs/buffer.c linux-4.4.62/fs/buffer.c clear_buffer_async_read(bh); unlock_buffer(bh); tmp = bh; -@@ -319,8 +318,7 @@ +@@ -319,8 +318,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate) } tmp = tmp->b_this_page; } while (tmp != bh); @@ -11418,7 +11623,7 @@ diff -Nur linux-4.4.62.orig/fs/buffer.c linux-4.4.62/fs/buffer.c /* * If none of the buffers had errors and they are all -@@ -332,9 +330,7 @@ +@@ -332,9 +330,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate) return; still_busy: @@ -11429,7 +11634,7 @@ diff -Nur linux-4.4.62.orig/fs/buffer.c linux-4.4.62/fs/buffer.c } /* -@@ -362,8 +358,7 @@ +@@ -362,8 +358,7 @@ void end_buffer_async_write(struct buffer_head *bh, int uptodate) } first = page_buffers(page); @@ -11439,7 +11644,7 @@ diff -Nur linux-4.4.62.orig/fs/buffer.c linux-4.4.62/fs/buffer.c clear_buffer_async_write(bh); unlock_buffer(bh); -@@ -375,15 +370,12 @@ +@@ -375,15 +370,12 @@ void end_buffer_async_write(struct buffer_head *bh, int uptodate) } tmp = tmp->b_this_page; } @@ -11457,7 +11662,7 @@ diff -Nur linux-4.4.62.orig/fs/buffer.c linux-4.4.62/fs/buffer.c } EXPORT_SYMBOL(end_buffer_async_write); -@@ -3325,6 +3317,7 @@ +@@ -3325,6 +3317,7 @@ struct buffer_head *alloc_buffer_head(gfp_t gfp_flags) struct buffer_head *ret = kmem_cache_zalloc(bh_cachep, gfp_flags); if (ret) { INIT_LIST_HEAD(&ret->b_assoc_buffers); @@ -11465,9 +11670,10 @@ diff -Nur linux-4.4.62.orig/fs/buffer.c linux-4.4.62/fs/buffer.c preempt_disable(); __this_cpu_inc(bh_accounting.nr); recalc_bh_state(); -diff -Nur linux-4.4.62.orig/fs/dcache.c linux-4.4.62/fs/dcache.c ---- linux-4.4.62.orig/fs/dcache.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/fs/dcache.c 2017-04-18 17:38:08.066644281 +0200 +diff --git a/fs/dcache.c b/fs/dcache.c +index 849c1c1e787b..995995130191 100644 +--- a/fs/dcache.c ++++ b/fs/dcache.c @@ -19,6 +19,7 @@ #include #include @@ -11476,7 +11682,7 @@ diff -Nur linux-4.4.62.orig/fs/dcache.c linux-4.4.62/fs/dcache.c #include #include #include -@@ -747,6 +748,8 @@ +@@ -747,6 +748,8 @@ static inline bool fast_dput(struct dentry *dentry) */ void dput(struct dentry *dentry) { @@ -11485,7 +11691,7 @@ diff -Nur linux-4.4.62.orig/fs/dcache.c linux-4.4.62/fs/dcache.c if (unlikely(!dentry)) return; -@@ -783,9 +786,18 @@ +@@ -783,9 +786,18 @@ repeat: return; kill_it: @@ -11507,7 +11713,7 @@ diff -Nur linux-4.4.62.orig/fs/dcache.c linux-4.4.62/fs/dcache.c goto repeat; } } -@@ -2397,7 +2409,7 @@ +@@ -2397,7 +2409,7 @@ again: if (dentry->d_lockref.count == 1) { if (!spin_trylock(&inode->i_lock)) { spin_unlock(&dentry->d_lock); @@ -11516,10 +11722,11 @@ diff -Nur linux-4.4.62.orig/fs/dcache.c linux-4.4.62/fs/dcache.c goto again; } dentry->d_flags &= ~DCACHE_CANT_MOUNT; -diff -Nur linux-4.4.62.orig/fs/eventpoll.c linux-4.4.62/fs/eventpoll.c ---- linux-4.4.62.orig/fs/eventpoll.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/fs/eventpoll.c 2017-04-18 17:38:08.066644281 +0200 -@@ -505,12 +505,12 @@ +diff --git a/fs/eventpoll.c b/fs/eventpoll.c +index 1e009cad8d5c..d0c12504d3b4 100644 +--- a/fs/eventpoll.c ++++ b/fs/eventpoll.c +@@ -505,12 +505,12 @@ static int ep_poll_wakeup_proc(void *priv, void *cookie, int call_nests) */ static void ep_poll_safewake(wait_queue_head_t *wq) { @@ -11534,10 +11741,11 @@ diff -Nur linux-4.4.62.orig/fs/eventpoll.c linux-4.4.62/fs/eventpoll.c } static void ep_remove_wait_queue(struct eppoll_entry *pwq) -diff -Nur linux-4.4.62.orig/fs/exec.c linux-4.4.62/fs/exec.c ---- linux-4.4.62.orig/fs/exec.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/fs/exec.c 2017-04-18 17:38:08.066644281 +0200 -@@ -866,12 +866,14 @@ +diff --git a/fs/exec.c b/fs/exec.c +index 3a6de10d3891..74962ac0a738 100644 +--- a/fs/exec.c ++++ b/fs/exec.c +@@ -866,12 +866,14 @@ static int exec_mmap(struct mm_struct *mm) } } task_lock(tsk); @@ -11552,9 +11760,10 @@ diff -Nur linux-4.4.62.orig/fs/exec.c linux-4.4.62/fs/exec.c task_unlock(tsk); if (old_mm) { up_read(&old_mm->mmap_sem); -diff -Nur linux-4.4.62.orig/fs/f2fs/f2fs.h linux-4.4.62/fs/f2fs/f2fs.h ---- linux-4.4.62.orig/fs/f2fs/f2fs.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/fs/f2fs/f2fs.h 2017-04-18 17:38:08.086645056 +0200 +diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h +index 2871576fbca4..d1137790ea58 100644 +--- a/fs/f2fs/f2fs.h ++++ b/fs/f2fs/f2fs.h @@ -24,7 +24,6 @@ #ifdef CONFIG_F2FS_CHECK_FS @@ -11571,7 +11780,7 @@ diff -Nur linux-4.4.62.orig/fs/f2fs/f2fs.h linux-4.4.62/fs/f2fs/f2fs.h #endif /* -@@ -959,7 +957,7 @@ +@@ -959,7 +957,7 @@ static inline void f2fs_unlock_op(struct f2fs_sb_info *sbi) static inline void f2fs_lock_all(struct f2fs_sb_info *sbi) { @@ -11580,10 +11789,11 @@ diff -Nur linux-4.4.62.orig/fs/f2fs/f2fs.h linux-4.4.62/fs/f2fs/f2fs.h } static inline void f2fs_unlock_all(struct f2fs_sb_info *sbi) -diff -Nur linux-4.4.62.orig/fs/jbd2/checkpoint.c linux-4.4.62/fs/jbd2/checkpoint.c ---- linux-4.4.62.orig/fs/jbd2/checkpoint.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/fs/jbd2/checkpoint.c 2017-04-18 17:38:08.106645833 +0200 -@@ -116,6 +116,8 @@ +diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c +index 684996c8a3a4..6e18a06aaabe 100644 +--- a/fs/jbd2/checkpoint.c ++++ b/fs/jbd2/checkpoint.c +@@ -116,6 +116,8 @@ void __jbd2_log_wait_for_space(journal_t *journal) nblocks = jbd2_space_needed(journal); while (jbd2_log_space_left(journal) < nblocks) { write_unlock(&journal->j_state_lock); @@ -11592,9 +11802,10 @@ diff -Nur linux-4.4.62.orig/fs/jbd2/checkpoint.c linux-4.4.62/fs/jbd2/checkpoint mutex_lock(&journal->j_checkpoint_mutex); /* -diff -Nur linux-4.4.62.orig/fs/namespace.c linux-4.4.62/fs/namespace.c ---- linux-4.4.62.orig/fs/namespace.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/fs/namespace.c 2017-04-18 17:38:08.122646453 +0200 +diff --git a/fs/namespace.c b/fs/namespace.c +index f26d18d69712..6e0e024715e8 100644 +--- a/fs/namespace.c ++++ b/fs/namespace.c @@ -14,6 +14,7 @@ #include #include @@ -11603,7 +11814,7 @@ diff -Nur linux-4.4.62.orig/fs/namespace.c linux-4.4.62/fs/namespace.c #include #include #include /* init_rootfs */ -@@ -353,8 +354,11 @@ +@@ -356,8 +357,11 @@ int __mnt_want_write(struct vfsmount *m) * incremented count after it has set MNT_WRITE_HOLD. */ smp_mb(); @@ -11617,10 +11828,11 @@ diff -Nur linux-4.4.62.orig/fs/namespace.c linux-4.4.62/fs/namespace.c /* * After the slowpath clears MNT_WRITE_HOLD, mnt_is_readonly will * be set to match its requirements. So we must not load that until -diff -Nur linux-4.4.62.orig/fs/ntfs/aops.c linux-4.4.62/fs/ntfs/aops.c ---- linux-4.4.62.orig/fs/ntfs/aops.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/fs/ntfs/aops.c 2017-04-18 17:38:08.122646453 +0200 -@@ -107,8 +107,7 @@ +diff --git a/fs/ntfs/aops.c b/fs/ntfs/aops.c +index 7521e11db728..f0de4b6b8bf3 100644 +--- a/fs/ntfs/aops.c ++++ b/fs/ntfs/aops.c +@@ -107,8 +107,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) "0x%llx.", (unsigned long long)bh->b_blocknr); } first = page_buffers(page); @@ -11630,7 +11842,7 @@ diff -Nur linux-4.4.62.orig/fs/ntfs/aops.c linux-4.4.62/fs/ntfs/aops.c clear_buffer_async_read(bh); unlock_buffer(bh); tmp = bh; -@@ -123,8 +122,7 @@ +@@ -123,8 +122,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) } tmp = tmp->b_this_page; } while (tmp != bh); @@ -11640,7 +11852,7 @@ diff -Nur linux-4.4.62.orig/fs/ntfs/aops.c linux-4.4.62/fs/ntfs/aops.c /* * If none of the buffers had errors then we can set the page uptodate, * but we first have to perform the post read mst fixups, if the -@@ -145,13 +143,13 @@ +@@ -145,13 +143,13 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) recs = PAGE_CACHE_SIZE / rec_size; /* Should have been verified before we got here... */ BUG_ON(!recs); @@ -11656,7 +11868,7 @@ diff -Nur linux-4.4.62.orig/fs/ntfs/aops.c linux-4.4.62/fs/ntfs/aops.c flush_dcache_page(page); if (likely(page_uptodate && !PageError(page))) SetPageUptodate(page); -@@ -159,9 +157,7 @@ +@@ -159,9 +157,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) unlock_page(page); return; still_busy: @@ -11667,10 +11879,11 @@ diff -Nur linux-4.4.62.orig/fs/ntfs/aops.c linux-4.4.62/fs/ntfs/aops.c } /** -diff -Nur linux-4.4.62.orig/fs/timerfd.c linux-4.4.62/fs/timerfd.c ---- linux-4.4.62.orig/fs/timerfd.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/fs/timerfd.c 2017-04-18 17:38:08.122646453 +0200 -@@ -450,7 +450,10 @@ +diff --git a/fs/timerfd.c b/fs/timerfd.c +index 1327a02ec778..4260febcb029 100644 +--- a/fs/timerfd.c ++++ b/fs/timerfd.c +@@ -461,7 +461,10 @@ static int do_timerfd_settime(int ufd, int flags, break; } spin_unlock_irq(&ctx->wqh.lock); @@ -11682,9 +11895,10 @@ diff -Nur linux-4.4.62.orig/fs/timerfd.c linux-4.4.62/fs/timerfd.c } /* -diff -Nur linux-4.4.62.orig/include/acpi/platform/aclinux.h linux-4.4.62/include/acpi/platform/aclinux.h ---- linux-4.4.62.orig/include/acpi/platform/aclinux.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/acpi/platform/aclinux.h 2017-04-18 17:38:08.122646453 +0200 +diff --git a/include/acpi/platform/aclinux.h b/include/acpi/platform/aclinux.h +index 323e5daece54..cc5fbd534fd4 100644 +--- a/include/acpi/platform/aclinux.h ++++ b/include/acpi/platform/aclinux.h @@ -127,6 +127,7 @@ #define acpi_cache_t struct kmem_cache @@ -11714,10 +11928,11 @@ diff -Nur linux-4.4.62.orig/include/acpi/platform/aclinux.h linux-4.4.62/include /* * OSL interfaces used by debugger/disassembler */ -diff -Nur linux-4.4.62.orig/include/asm-generic/bug.h linux-4.4.62/include/asm-generic/bug.h ---- linux-4.4.62.orig/include/asm-generic/bug.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/asm-generic/bug.h 2017-04-18 17:38:08.122646453 +0200 -@@ -206,6 +206,20 @@ +diff --git a/include/asm-generic/bug.h b/include/asm-generic/bug.h +index 630dd2372238..850e4d993a88 100644 +--- a/include/asm-generic/bug.h ++++ b/include/asm-generic/bug.h +@@ -206,6 +206,20 @@ extern void warn_slowpath_null(const char *file, const int line); # define WARN_ON_SMP(x) ({0;}) #endif @@ -11738,9 +11953,10 @@ diff -Nur linux-4.4.62.orig/include/asm-generic/bug.h linux-4.4.62/include/asm-g #endif /* __ASSEMBLY__ */ #endif -diff -Nur linux-4.4.62.orig/include/asm-generic/preempt.h linux-4.4.62/include/asm-generic/preempt.h ---- linux-4.4.62.orig/include/asm-generic/preempt.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/asm-generic/preempt.h 2017-04-18 17:38:08.122646453 +0200 +diff --git a/include/asm-generic/preempt.h b/include/asm-generic/preempt.h +index 5d8ffa3e6f8c..c1cde3577551 100644 +--- a/include/asm-generic/preempt.h ++++ b/include/asm-generic/preempt.h @@ -7,10 +7,10 @@ static __always_inline int preempt_count(void) @@ -11754,10 +11970,23 @@ diff -Nur linux-4.4.62.orig/include/asm-generic/preempt.h linux-4.4.62/include/a { return ¤t_thread_info()->preempt_count; } -diff -Nur linux-4.4.62.orig/include/linux/blkdev.h linux-4.4.62/include/linux/blkdev.h ---- linux-4.4.62.orig/include/linux/blkdev.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/blkdev.h 2017-04-18 17:38:08.122646453 +0200 -@@ -89,6 +89,7 @@ +diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h +index daf17d70aeca..463df8954255 100644 +--- a/include/linux/blk-mq.h ++++ b/include/linux/blk-mq.h +@@ -212,6 +212,7 @@ static inline u16 blk_mq_unique_tag_to_tag(u32 unique_tag) + + struct blk_mq_hw_ctx *blk_mq_map_queue(struct request_queue *, const int ctx_index); + struct blk_mq_hw_ctx *blk_mq_alloc_single_hw_queue(struct blk_mq_tag_set *, unsigned int, int); ++void __blk_mq_complete_request_remote_work(struct work_struct *work); + + int blk_mq_request_started(struct request *rq); + void blk_mq_start_request(struct request *rq); +diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h +index fe14382f9664..a82143ad6702 100644 +--- a/include/linux/blkdev.h ++++ b/include/linux/blkdev.h +@@ -89,6 +89,7 @@ struct request { struct list_head queuelist; union { struct call_single_data csd; @@ -11765,7 +11994,7 @@ diff -Nur linux-4.4.62.orig/include/linux/blkdev.h linux-4.4.62/include/linux/bl unsigned long fifo_time; }; -@@ -455,7 +456,7 @@ +@@ -455,7 +456,7 @@ struct request_queue { struct throtl_data *td; #endif struct rcu_head rcu_head; @@ -11774,20 +12003,10 @@ diff -Nur linux-4.4.62.orig/include/linux/blkdev.h linux-4.4.62/include/linux/bl struct percpu_ref q_usage_counter; struct list_head all_q_node; -diff -Nur linux-4.4.62.orig/include/linux/blk-mq.h linux-4.4.62/include/linux/blk-mq.h ---- linux-4.4.62.orig/include/linux/blk-mq.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/blk-mq.h 2017-04-18 17:38:08.122646453 +0200 -@@ -212,6 +212,7 @@ - - struct blk_mq_hw_ctx *blk_mq_map_queue(struct request_queue *, const int ctx_index); - struct blk_mq_hw_ctx *blk_mq_alloc_single_hw_queue(struct blk_mq_tag_set *, unsigned int, int); -+void __blk_mq_complete_request_remote_work(struct work_struct *work); - - int blk_mq_request_started(struct request *rq); - void blk_mq_start_request(struct request *rq); -diff -Nur linux-4.4.62.orig/include/linux/bottom_half.h linux-4.4.62/include/linux/bottom_half.h ---- linux-4.4.62.orig/include/linux/bottom_half.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/bottom_half.h 2017-04-18 17:38:08.122646453 +0200 +diff --git a/include/linux/bottom_half.h b/include/linux/bottom_half.h +index 8fdcb783197d..d07dbeec7bc1 100644 +--- a/include/linux/bottom_half.h ++++ b/include/linux/bottom_half.h @@ -3,6 +3,39 @@ #include @@ -11828,17 +12047,18 @@ diff -Nur linux-4.4.62.orig/include/linux/bottom_half.h linux-4.4.62/include/lin #ifdef CONFIG_TRACE_IRQFLAGS extern void __local_bh_disable_ip(unsigned long ip, unsigned int cnt); #else -@@ -30,5 +63,6 @@ +@@ -30,5 +63,6 @@ static inline void local_bh_enable(void) { __local_bh_enable_ip(_THIS_IP_, SOFTIRQ_DISABLE_OFFSET); } +#endif #endif /* _LINUX_BH_H */ -diff -Nur linux-4.4.62.orig/include/linux/buffer_head.h linux-4.4.62/include/linux/buffer_head.h ---- linux-4.4.62.orig/include/linux/buffer_head.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/buffer_head.h 2017-04-18 17:38:08.122646453 +0200 -@@ -75,8 +75,50 @@ +diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h +index 89d9aa9e79bf..4a201008b02d 100644 +--- a/include/linux/buffer_head.h ++++ b/include/linux/buffer_head.h +@@ -75,8 +75,50 @@ struct buffer_head { struct address_space *b_assoc_map; /* mapping this buffer is associated with */ atomic_t b_count; /* users using this buffer_head */ @@ -11889,9 +12109,10 @@ diff -Nur linux-4.4.62.orig/include/linux/buffer_head.h linux-4.4.62/include/lin /* * macro tricks to expand the set_buffer_foo(), clear_buffer_foo() * and buffer_foo() functions. -diff -Nur linux-4.4.62.orig/include/linux/cgroup-defs.h linux-4.4.62/include/linux/cgroup-defs.h ---- linux-4.4.62.orig/include/linux/cgroup-defs.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/cgroup-defs.h 2017-04-18 17:38:08.122646453 +0200 +diff --git a/include/linux/cgroup-defs.h b/include/linux/cgroup-defs.h +index 8da263299754..0cc474291e08 100644 +--- a/include/linux/cgroup-defs.h ++++ b/include/linux/cgroup-defs.h @@ -16,6 +16,7 @@ #include #include @@ -11900,7 +12121,7 @@ diff -Nur linux-4.4.62.orig/include/linux/cgroup-defs.h linux-4.4.62/include/lin #ifdef CONFIG_CGROUPS -@@ -142,6 +143,7 @@ +@@ -142,6 +143,7 @@ struct cgroup_subsys_state { /* percpu_ref killing and RCU release */ struct rcu_head rcu_head; struct work_struct destroy_work; @@ -11908,9 +12129,10 @@ diff -Nur linux-4.4.62.orig/include/linux/cgroup-defs.h linux-4.4.62/include/lin }; /* -diff -Nur linux-4.4.62.orig/include/linux/clk/at91_pmc.h linux-4.4.62/include/linux/clk/at91_pmc.h ---- linux-4.4.62.orig/include/linux/clk/at91_pmc.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/clk/at91_pmc.h 2017-04-18 17:38:08.122646453 +0200 +diff --git a/include/linux/clk/at91_pmc.h b/include/linux/clk/at91_pmc.h +index 1e6932222e11..17f413bbbedf 100644 +--- a/include/linux/clk/at91_pmc.h ++++ b/include/linux/clk/at91_pmc.h @@ -16,18 +16,6 @@ #ifndef AT91_PMC_H #define AT91_PMC_H @@ -11930,9 +12152,10 @@ diff -Nur linux-4.4.62.orig/include/linux/clk/at91_pmc.h linux-4.4.62/include/li #define AT91_PMC_SCER 0x00 /* System Clock Enable Register */ #define AT91_PMC_SCDR 0x04 /* System Clock Disable Register */ -diff -Nur linux-4.4.62.orig/include/linux/completion.h linux-4.4.62/include/linux/completion.h ---- linux-4.4.62.orig/include/linux/completion.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/completion.h 2017-04-18 17:38:08.122646453 +0200 +diff --git a/include/linux/completion.h b/include/linux/completion.h +index 5d5aaae3af43..3bca1590e29f 100644 +--- a/include/linux/completion.h ++++ b/include/linux/completion.h @@ -7,8 +7,7 @@ * Atomic wait-for-completion handler data structures. * See kernel/sched/completion.c for details. @@ -11957,7 +12180,7 @@ diff -Nur linux-4.4.62.orig/include/linux/completion.h linux-4.4.62/include/linu #define COMPLETION_INITIALIZER_ONSTACK(work) \ ({ init_completion(&work); work; }) -@@ -73,7 +72,7 @@ +@@ -73,7 +72,7 @@ struct completion { static inline void init_completion(struct completion *x) { x->done = 0; @@ -11966,10 +12189,11 @@ diff -Nur linux-4.4.62.orig/include/linux/completion.h linux-4.4.62/include/linu } /** -diff -Nur linux-4.4.62.orig/include/linux/cpu.h linux-4.4.62/include/linux/cpu.h ---- linux-4.4.62.orig/include/linux/cpu.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/cpu.h 2017-04-18 17:38:08.122646453 +0200 -@@ -224,6 +224,8 @@ +diff --git a/include/linux/cpu.h b/include/linux/cpu.h +index 3ea9aae2387d..5de5f28a5ef3 100644 +--- a/include/linux/cpu.h ++++ b/include/linux/cpu.h +@@ -224,6 +224,8 @@ extern void get_online_cpus(void); extern void put_online_cpus(void); extern void cpu_hotplug_disable(void); extern void cpu_hotplug_enable(void); @@ -11978,7 +12202,7 @@ diff -Nur linux-4.4.62.orig/include/linux/cpu.h linux-4.4.62/include/linux/cpu.h #define hotcpu_notifier(fn, pri) cpu_notifier(fn, pri) #define __hotcpu_notifier(fn, pri) __cpu_notifier(fn, pri) #define register_hotcpu_notifier(nb) register_cpu_notifier(nb) -@@ -241,6 +243,8 @@ +@@ -241,6 +243,8 @@ static inline void cpu_hotplug_done(void) {} #define put_online_cpus() do { } while (0) #define cpu_hotplug_disable() do { } while (0) #define cpu_hotplug_enable() do { } while (0) @@ -11987,10 +12211,11 @@ diff -Nur linux-4.4.62.orig/include/linux/cpu.h linux-4.4.62/include/linux/cpu.h #define hotcpu_notifier(fn, pri) do { (void)(fn); } while (0) #define __hotcpu_notifier(fn, pri) do { (void)(fn); } while (0) /* These aren't inline functions due to a GCC bug. */ -diff -Nur linux-4.4.62.orig/include/linux/delay.h linux-4.4.62/include/linux/delay.h ---- linux-4.4.62.orig/include/linux/delay.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/delay.h 2017-04-18 17:38:08.122646453 +0200 -@@ -52,4 +52,10 @@ +diff --git a/include/linux/delay.h b/include/linux/delay.h +index a6ecb34cf547..37caab306336 100644 +--- a/include/linux/delay.h ++++ b/include/linux/delay.h +@@ -52,4 +52,10 @@ static inline void ssleep(unsigned int seconds) msleep(seconds * 1000); } @@ -12001,10 +12226,11 @@ diff -Nur linux-4.4.62.orig/include/linux/delay.h linux-4.4.62/include/linux/del +#endif + #endif /* defined(_LINUX_DELAY_H) */ -diff -Nur linux-4.4.62.orig/include/linux/ftrace.h linux-4.4.62/include/linux/ftrace.h ---- linux-4.4.62.orig/include/linux/ftrace.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/ftrace.h 2017-04-18 17:38:08.122646453 +0200 -@@ -694,6 +694,18 @@ +diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h +index 60048c50404e..f2cd67624f18 100644 +--- a/include/linux/ftrace.h ++++ b/include/linux/ftrace.h +@@ -694,6 +694,18 @@ static inline void __ftrace_enabled_restore(int enabled) #define CALLER_ADDR5 ((unsigned long)ftrace_return_address(5)) #define CALLER_ADDR6 ((unsigned long)ftrace_return_address(6)) @@ -12023,9 +12249,10 @@ diff -Nur linux-4.4.62.orig/include/linux/ftrace.h linux-4.4.62/include/linux/ft #ifdef CONFIG_IRQSOFF_TRACER extern void time_hardirqs_on(unsigned long a0, unsigned long a1); extern void time_hardirqs_off(unsigned long a0, unsigned long a1); -diff -Nur linux-4.4.62.orig/include/linux/highmem.h linux-4.4.62/include/linux/highmem.h ---- linux-4.4.62.orig/include/linux/highmem.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/highmem.h 2017-04-18 17:38:08.122646453 +0200 +diff --git a/include/linux/highmem.h b/include/linux/highmem.h +index bb3f3297062a..a117a33ef72c 100644 +--- a/include/linux/highmem.h ++++ b/include/linux/highmem.h @@ -7,6 +7,7 @@ #include #include @@ -12034,7 +12261,7 @@ diff -Nur linux-4.4.62.orig/include/linux/highmem.h linux-4.4.62/include/linux/h #include -@@ -65,7 +66,7 @@ +@@ -65,7 +66,7 @@ static inline void kunmap(struct page *page) static inline void *kmap_atomic(struct page *page) { @@ -12043,7 +12270,7 @@ diff -Nur linux-4.4.62.orig/include/linux/highmem.h linux-4.4.62/include/linux/h pagefault_disable(); return page_address(page); } -@@ -74,7 +75,7 @@ +@@ -74,7 +75,7 @@ static inline void *kmap_atomic(struct page *page) static inline void __kunmap_atomic(void *addr) { pagefault_enable(); @@ -12052,7 +12279,7 @@ diff -Nur linux-4.4.62.orig/include/linux/highmem.h linux-4.4.62/include/linux/h } #define kmap_atomic_pfn(pfn) kmap_atomic(pfn_to_page(pfn)) -@@ -86,32 +87,51 @@ +@@ -86,32 +87,51 @@ static inline void __kunmap_atomic(void *addr) #if defined(CONFIG_HIGHMEM) || defined(CONFIG_X86_32) @@ -12108,10 +12335,11 @@ diff -Nur linux-4.4.62.orig/include/linux/highmem.h linux-4.4.62/include/linux/h #endif } -diff -Nur linux-4.4.62.orig/include/linux/hrtimer.h linux-4.4.62/include/linux/hrtimer.h ---- linux-4.4.62.orig/include/linux/hrtimer.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/hrtimer.h 2017-04-18 17:38:08.122646453 +0200 -@@ -87,6 +87,9 @@ +diff --git a/include/linux/hrtimer.h b/include/linux/hrtimer.h +index 2ead22dd74a0..8fbcdfa5dc77 100644 +--- a/include/linux/hrtimer.h ++++ b/include/linux/hrtimer.h +@@ -87,6 +87,9 @@ enum hrtimer_restart { * @function: timer expiry callback function * @base: pointer to the timer base (per cpu and per clock) * @state: state information (See bit values above) @@ -12121,7 +12349,7 @@ diff -Nur linux-4.4.62.orig/include/linux/hrtimer.h linux-4.4.62/include/linux/h * @is_rel: Set if the timer was armed relative * @start_pid: timer statistics field to store the pid of the task which * started the timer -@@ -103,6 +106,11 @@ +@@ -103,6 +106,11 @@ struct hrtimer { enum hrtimer_restart (*function)(struct hrtimer *); struct hrtimer_clock_base *base; u8 state; @@ -12133,7 +12361,7 @@ diff -Nur linux-4.4.62.orig/include/linux/hrtimer.h linux-4.4.62/include/linux/h u8 is_rel; #ifdef CONFIG_TIMER_STATS int start_pid; -@@ -123,11 +131,7 @@ +@@ -123,11 +131,7 @@ struct hrtimer_sleeper { struct task_struct *task; }; @@ -12145,7 +12373,7 @@ diff -Nur linux-4.4.62.orig/include/linux/hrtimer.h linux-4.4.62/include/linux/h /** * struct hrtimer_clock_base - the timer base for a specific clock -@@ -136,6 +140,7 @@ +@@ -136,6 +140,7 @@ struct hrtimer_sleeper { * timer to a base on another cpu. * @clockid: clock id for per_cpu support * @active: red black tree root node for the active timers @@ -12153,7 +12381,7 @@ diff -Nur linux-4.4.62.orig/include/linux/hrtimer.h linux-4.4.62/include/linux/h * @get_time: function to retrieve the current time of the clock * @offset: offset of this clock to the monotonic base */ -@@ -144,6 +149,7 @@ +@@ -144,6 +149,7 @@ struct hrtimer_clock_base { int index; clockid_t clockid; struct timerqueue_head active; @@ -12161,7 +12389,7 @@ diff -Nur linux-4.4.62.orig/include/linux/hrtimer.h linux-4.4.62/include/linux/h ktime_t (*get_time)(void); ktime_t offset; } __attribute__((__aligned__(HRTIMER_CLOCK_BASE_ALIGN))); -@@ -187,6 +193,7 @@ +@@ -187,6 +193,7 @@ struct hrtimer_cpu_base { raw_spinlock_t lock; seqcount_t seq; struct hrtimer *running; @@ -12169,7 +12397,7 @@ diff -Nur linux-4.4.62.orig/include/linux/hrtimer.h linux-4.4.62/include/linux/h unsigned int cpu; unsigned int active_bases; unsigned int clock_was_set_seq; -@@ -203,6 +210,9 @@ +@@ -203,6 +210,9 @@ struct hrtimer_cpu_base { unsigned int nr_hangs; unsigned int max_hang_time; #endif @@ -12179,7 +12407,7 @@ diff -Nur linux-4.4.62.orig/include/linux/hrtimer.h linux-4.4.62/include/linux/h struct hrtimer_clock_base clock_base[HRTIMER_MAX_CLOCK_BASES]; } ____cacheline_aligned; -@@ -412,6 +422,13 @@ +@@ -412,6 +422,13 @@ static inline void hrtimer_restart(struct hrtimer *timer) hrtimer_start_expires(timer, HRTIMER_MODE_ABS); } @@ -12193,7 +12421,7 @@ diff -Nur linux-4.4.62.orig/include/linux/hrtimer.h linux-4.4.62/include/linux/h /* Query timers: */ extern ktime_t __hrtimer_get_remaining(const struct hrtimer *timer, bool adjust); -@@ -436,7 +453,7 @@ +@@ -436,7 +453,7 @@ static inline int hrtimer_is_queued(struct hrtimer *timer) * Helper function to check, whether the timer is running the callback * function */ @@ -12202,10 +12430,11 @@ diff -Nur linux-4.4.62.orig/include/linux/hrtimer.h linux-4.4.62/include/linux/h { return timer->base->cpu_base->running == timer; } -diff -Nur linux-4.4.62.orig/include/linux/idr.h linux-4.4.62/include/linux/idr.h ---- linux-4.4.62.orig/include/linux/idr.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/idr.h 2017-04-18 17:38:08.122646453 +0200 -@@ -95,10 +95,14 @@ +diff --git a/include/linux/idr.h b/include/linux/idr.h +index 013fd9bc4cb6..f62be0aec911 100644 +--- a/include/linux/idr.h ++++ b/include/linux/idr.h +@@ -95,10 +95,14 @@ bool idr_is_empty(struct idr *idp); * Each idr_preload() should be matched with an invocation of this * function. See idr_preload() for details. */ @@ -12220,10 +12449,11 @@ diff -Nur linux-4.4.62.orig/include/linux/idr.h linux-4.4.62/include/linux/idr.h /** * idr_find - return pointer for given id -diff -Nur linux-4.4.62.orig/include/linux/init_task.h linux-4.4.62/include/linux/init_task.h ---- linux-4.4.62.orig/include/linux/init_task.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/init_task.h 2017-04-18 17:38:08.122646453 +0200 -@@ -148,9 +148,15 @@ +diff --git a/include/linux/init_task.h b/include/linux/init_task.h +index 1c1ff7e4faa4..60fadde71a44 100644 +--- a/include/linux/init_task.h ++++ b/include/linux/init_task.h +@@ -148,9 +148,15 @@ extern struct task_group root_task_group; # define INIT_PERF_EVENTS(tsk) #endif @@ -12240,7 +12470,7 @@ diff -Nur linux-4.4.62.orig/include/linux/init_task.h linux-4.4.62/include/linux .vtime_snap = 0, \ .vtime_snap_whence = VTIME_SYS, #else -@@ -239,6 +245,7 @@ +@@ -239,6 +245,7 @@ extern struct task_group root_task_group; .cpu_timers = INIT_CPU_TIMERS(tsk.cpu_timers), \ .pi_lock = __RAW_SPIN_LOCK_UNLOCKED(tsk.pi_lock), \ .timer_slack_ns = 50000, /* 50 usec default slack */ \ @@ -12248,9 +12478,10 @@ diff -Nur linux-4.4.62.orig/include/linux/init_task.h linux-4.4.62/include/linux .pids = { \ [PIDTYPE_PID] = INIT_PID_LINK(PIDTYPE_PID), \ [PIDTYPE_PGID] = INIT_PID_LINK(PIDTYPE_PGID), \ -diff -Nur linux-4.4.62.orig/include/linux/interrupt.h linux-4.4.62/include/linux/interrupt.h ---- linux-4.4.62.orig/include/linux/interrupt.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/interrupt.h 2017-04-18 17:38:08.126646608 +0200 +diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h +index ad16809c8596..655cee096aed 100644 +--- a/include/linux/interrupt.h ++++ b/include/linux/interrupt.h @@ -61,6 +61,7 @@ * interrupt handler after suspending interrupts. For system * wakeup devices users need to implement wakeup detection in @@ -12267,7 +12498,7 @@ diff -Nur linux-4.4.62.orig/include/linux/interrupt.h linux-4.4.62/include/linux #define IRQF_TIMER (__IRQF_TIMER | IRQF_NO_SUSPEND | IRQF_NO_THREAD) -@@ -186,7 +188,7 @@ +@@ -186,7 +188,7 @@ extern void devm_free_irq(struct device *dev, unsigned int irq, void *dev_id); #ifdef CONFIG_LOCKDEP # define local_irq_enable_in_hardirq() do { } while (0) #else @@ -12276,7 +12507,7 @@ diff -Nur linux-4.4.62.orig/include/linux/interrupt.h linux-4.4.62/include/linux #endif extern void disable_irq_nosync(unsigned int irq); -@@ -206,6 +208,7 @@ +@@ -206,6 +208,7 @@ extern void resume_device_irqs(void); * @irq: Interrupt to which notification applies * @kref: Reference count, for internal use * @work: Work item, for internal use @@ -12284,7 +12515,7 @@ diff -Nur linux-4.4.62.orig/include/linux/interrupt.h linux-4.4.62/include/linux * @notify: Function to be called on change. This will be * called in process context. * @release: Function to be called on release. This will be -@@ -217,6 +220,7 @@ +@@ -217,6 +220,7 @@ struct irq_affinity_notify { unsigned int irq; struct kref kref; struct work_struct work; @@ -12292,7 +12523,7 @@ diff -Nur linux-4.4.62.orig/include/linux/interrupt.h linux-4.4.62/include/linux void (*notify)(struct irq_affinity_notify *, const cpumask_t *mask); void (*release)(struct kref *ref); }; -@@ -379,9 +383,13 @@ +@@ -379,9 +383,13 @@ extern int irq_set_irqchip_state(unsigned int irq, enum irqchip_irq_state which, bool state); #ifdef CONFIG_IRQ_FORCED_THREADING @@ -12307,7 +12538,7 @@ diff -Nur linux-4.4.62.orig/include/linux/interrupt.h linux-4.4.62/include/linux #endif #ifndef __ARCH_SET_SOFTIRQ_PENDING -@@ -438,9 +446,10 @@ +@@ -438,9 +446,10 @@ struct softirq_action void (*action)(struct softirq_action *); }; @@ -12319,7 +12550,7 @@ diff -Nur linux-4.4.62.orig/include/linux/interrupt.h linux-4.4.62/include/linux #ifdef __ARCH_HAS_DO_SOFTIRQ void do_softirq_own_stack(void); #else -@@ -449,13 +458,25 @@ +@@ -449,13 +458,25 @@ static inline void do_softirq_own_stack(void) __do_softirq(); } #endif @@ -12345,7 +12576,7 @@ diff -Nur linux-4.4.62.orig/include/linux/interrupt.h linux-4.4.62/include/linux DECLARE_PER_CPU(struct task_struct *, ksoftirqd); -@@ -477,8 +498,9 @@ +@@ -477,8 +498,9 @@ static inline struct task_struct *this_cpu_ksoftirqd(void) to be executed on some cpu at least once after this. * If the tasklet is already scheduled, but its execution is still not started, it will be executed only once. @@ -12357,7 +12588,7 @@ diff -Nur linux-4.4.62.orig/include/linux/interrupt.h linux-4.4.62/include/linux * Tasklet is strictly serialized wrt itself, but not wrt another tasklets. If client needs some intertask synchronization, he makes it with spinlocks. -@@ -503,27 +525,36 @@ +@@ -503,27 +525,36 @@ struct tasklet_struct name = { NULL, 0, ATOMIC_INIT(1), func, data } enum { TASKLET_STATE_SCHED, /* Tasklet is scheduled for execution */ @@ -12400,7 +12631,7 @@ diff -Nur linux-4.4.62.orig/include/linux/interrupt.h linux-4.4.62/include/linux #define tasklet_unlock_wait(t) do { } while (0) #define tasklet_unlock(t) do { } while (0) #endif -@@ -572,12 +603,7 @@ +@@ -572,12 +603,7 @@ static inline void tasklet_disable(struct tasklet_struct *t) smp_mb(); } @@ -12414,7 +12645,7 @@ diff -Nur linux-4.4.62.orig/include/linux/interrupt.h linux-4.4.62/include/linux extern void tasklet_kill(struct tasklet_struct *t); extern void tasklet_kill_immediate(struct tasklet_struct *t, unsigned int cpu); extern void tasklet_init(struct tasklet_struct *t, -@@ -608,6 +634,12 @@ +@@ -608,6 +634,12 @@ void tasklet_hrtimer_cancel(struct tasklet_hrtimer *ttimer) tasklet_kill(&ttimer->tasklet); } @@ -12427,10 +12658,62 @@ diff -Nur linux-4.4.62.orig/include/linux/interrupt.h linux-4.4.62/include/linux /* * Autoprobing for irqs: * -diff -Nur linux-4.4.62.orig/include/linux/irqdesc.h linux-4.4.62/include/linux/irqdesc.h ---- linux-4.4.62.orig/include/linux/irqdesc.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/irqdesc.h 2017-04-18 17:38:08.126646608 +0200 -@@ -61,6 +61,7 @@ +diff --git a/include/linux/irq.h b/include/linux/irq.h +index f7cade00c525..dac9e11ba037 100644 +--- a/include/linux/irq.h ++++ b/include/linux/irq.h +@@ -72,6 +72,7 @@ enum irqchip_irq_state; + * IRQ_IS_POLLED - Always polled by another interrupt. Exclude + * it from the spurious interrupt detection + * mechanism and from core side polling. ++ * IRQ_NO_SOFTIRQ_CALL - No softirq processing in the irq thread context (RT) + * IRQ_DISABLE_UNLAZY - Disable lazy irq disable + */ + enum { +@@ -99,13 +100,14 @@ enum { + IRQ_PER_CPU_DEVID = (1 << 17), + IRQ_IS_POLLED = (1 << 18), + IRQ_DISABLE_UNLAZY = (1 << 19), ++ IRQ_NO_SOFTIRQ_CALL = (1 << 20), + }; + + #define IRQF_MODIFY_MASK \ + (IRQ_TYPE_SENSE_MASK | IRQ_NOPROBE | IRQ_NOREQUEST | \ + IRQ_NOAUTOEN | IRQ_MOVE_PCNTXT | IRQ_LEVEL | IRQ_NO_BALANCING | \ + IRQ_PER_CPU | IRQ_NESTED_THREAD | IRQ_NOTHREAD | IRQ_PER_CPU_DEVID | \ +- IRQ_IS_POLLED | IRQ_DISABLE_UNLAZY) ++ IRQ_IS_POLLED | IRQ_DISABLE_UNLAZY | IRQ_NO_SOFTIRQ_CALL) + + #define IRQ_NO_BALANCING_MASK (IRQ_PER_CPU | IRQ_NO_BALANCING) + +diff --git a/include/linux/irq_work.h b/include/linux/irq_work.h +index 47b9ebd4a74f..2543aab05daa 100644 +--- a/include/linux/irq_work.h ++++ b/include/linux/irq_work.h +@@ -16,6 +16,7 @@ + #define IRQ_WORK_BUSY 2UL + #define IRQ_WORK_FLAGS 3UL + #define IRQ_WORK_LAZY 4UL /* Doesn't want IPI, wait for tick */ ++#define IRQ_WORK_HARD_IRQ 8UL /* Run hard IRQ context, even on RT */ + + struct irq_work { + unsigned long flags; +@@ -51,4 +52,10 @@ static inline bool irq_work_needs_cpu(void) { return false; } + static inline void irq_work_run(void) { } + #endif + ++#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) ++void irq_work_tick_soft(void); ++#else ++static inline void irq_work_tick_soft(void) { } ++#endif ++ + #endif /* _LINUX_IRQ_WORK_H */ +diff --git a/include/linux/irqdesc.h b/include/linux/irqdesc.h +index a587a33363c7..ad57402a242d 100644 +--- a/include/linux/irqdesc.h ++++ b/include/linux/irqdesc.h +@@ -61,6 +61,7 @@ struct irq_desc { unsigned int irqs_unhandled; atomic_t threads_handled; int threads_handled_last; @@ -12438,9 +12721,10 @@ diff -Nur linux-4.4.62.orig/include/linux/irqdesc.h linux-4.4.62/include/linux/i raw_spinlock_t lock; struct cpumask *percpu_enabled; #ifdef CONFIG_SMP -diff -Nur linux-4.4.62.orig/include/linux/irqflags.h linux-4.4.62/include/linux/irqflags.h ---- linux-4.4.62.orig/include/linux/irqflags.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/irqflags.h 2017-04-18 17:38:08.126646608 +0200 +diff --git a/include/linux/irqflags.h b/include/linux/irqflags.h +index 5dd1272d1ab2..9b77034f7c5e 100644 +--- a/include/linux/irqflags.h ++++ b/include/linux/irqflags.h @@ -25,8 +25,6 @@ # define trace_softirqs_enabled(p) ((p)->softirqs_enabled) # define trace_hardirq_enter() do { current->hardirq_context++; } while (0) @@ -12491,59 +12775,11 @@ diff -Nur linux-4.4.62.orig/include/linux/irqflags.h linux-4.4.62/include/linux/ +#endif + #endif -diff -Nur linux-4.4.62.orig/include/linux/irq.h linux-4.4.62/include/linux/irq.h ---- linux-4.4.62.orig/include/linux/irq.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/irq.h 2017-04-18 17:38:08.126646608 +0200 -@@ -72,6 +72,7 @@ - * IRQ_IS_POLLED - Always polled by another interrupt. Exclude - * it from the spurious interrupt detection - * mechanism and from core side polling. -+ * IRQ_NO_SOFTIRQ_CALL - No softirq processing in the irq thread context (RT) - * IRQ_DISABLE_UNLAZY - Disable lazy irq disable - */ - enum { -@@ -99,13 +100,14 @@ - IRQ_PER_CPU_DEVID = (1 << 17), - IRQ_IS_POLLED = (1 << 18), - IRQ_DISABLE_UNLAZY = (1 << 19), -+ IRQ_NO_SOFTIRQ_CALL = (1 << 20), - }; - - #define IRQF_MODIFY_MASK \ - (IRQ_TYPE_SENSE_MASK | IRQ_NOPROBE | IRQ_NOREQUEST | \ - IRQ_NOAUTOEN | IRQ_MOVE_PCNTXT | IRQ_LEVEL | IRQ_NO_BALANCING | \ - IRQ_PER_CPU | IRQ_NESTED_THREAD | IRQ_NOTHREAD | IRQ_PER_CPU_DEVID | \ -- IRQ_IS_POLLED | IRQ_DISABLE_UNLAZY) -+ IRQ_IS_POLLED | IRQ_DISABLE_UNLAZY | IRQ_NO_SOFTIRQ_CALL) - - #define IRQ_NO_BALANCING_MASK (IRQ_PER_CPU | IRQ_NO_BALANCING) - -diff -Nur linux-4.4.62.orig/include/linux/irq_work.h linux-4.4.62/include/linux/irq_work.h ---- linux-4.4.62.orig/include/linux/irq_work.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/irq_work.h 2017-04-18 17:38:08.126646608 +0200 -@@ -16,6 +16,7 @@ - #define IRQ_WORK_BUSY 2UL - #define IRQ_WORK_FLAGS 3UL - #define IRQ_WORK_LAZY 4UL /* Doesn't want IPI, wait for tick */ -+#define IRQ_WORK_HARD_IRQ 8UL /* Run hard IRQ context, even on RT */ - - struct irq_work { - unsigned long flags; -@@ -51,4 +52,10 @@ - static inline void irq_work_run(void) { } - #endif - -+#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) -+void irq_work_tick_soft(void); -+#else -+static inline void irq_work_tick_soft(void) { } -+#endif -+ - #endif /* _LINUX_IRQ_WORK_H */ -diff -Nur linux-4.4.62.orig/include/linux/jbd2.h linux-4.4.62/include/linux/jbd2.h ---- linux-4.4.62.orig/include/linux/jbd2.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/jbd2.h 2017-04-18 17:38:08.126646608 +0200 -@@ -352,32 +352,56 @@ +diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h +index 65407f6c9120..eb5aabe4e18c 100644 +--- a/include/linux/jbd2.h ++++ b/include/linux/jbd2.h +@@ -352,32 +352,56 @@ static inline struct journal_head *bh2jh(struct buffer_head *bh) static inline void jbd_lock_bh_state(struct buffer_head *bh) { @@ -12600,10 +12836,11 @@ diff -Nur linux-4.4.62.orig/include/linux/jbd2.h linux-4.4.62/include/linux/jbd2 } #define J_ASSERT(assert) BUG_ON(!(assert)) -diff -Nur linux-4.4.62.orig/include/linux/kdb.h linux-4.4.62/include/linux/kdb.h ---- linux-4.4.62.orig/include/linux/kdb.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/kdb.h 2017-04-18 17:38:08.126646608 +0200 -@@ -167,6 +167,7 @@ +diff --git a/include/linux/kdb.h b/include/linux/kdb.h +index a19bcf9e762e..897495386446 100644 +--- a/include/linux/kdb.h ++++ b/include/linux/kdb.h +@@ -167,6 +167,7 @@ extern __printf(2, 0) int vkdb_printf(enum kdb_msgsrc src, const char *fmt, extern __printf(1, 2) int kdb_printf(const char *, ...); typedef __printf(1, 2) int (*kdb_printf_t)(const char *, ...); @@ -12611,7 +12848,7 @@ diff -Nur linux-4.4.62.orig/include/linux/kdb.h linux-4.4.62/include/linux/kdb.h extern void kdb_init(int level); /* Access to kdb specific polling devices */ -@@ -201,6 +202,7 @@ +@@ -201,6 +202,7 @@ extern int kdb_register_flags(char *, kdb_func_t, char *, char *, extern int kdb_unregister(char *); #else /* ! CONFIG_KGDB_KDB */ static inline __printf(1, 2) int kdb_printf(const char *fmt, ...) { return 0; } @@ -12619,10 +12856,11 @@ diff -Nur linux-4.4.62.orig/include/linux/kdb.h linux-4.4.62/include/linux/kdb.h static inline void kdb_init(int level) {} static inline int kdb_register(char *cmd, kdb_func_t func, char *usage, char *help, short minlen) { return 0; } -diff -Nur linux-4.4.62.orig/include/linux/kernel.h linux-4.4.62/include/linux/kernel.h ---- linux-4.4.62.orig/include/linux/kernel.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/kernel.h 2017-04-18 17:38:08.126646608 +0200 -@@ -188,6 +188,9 @@ +diff --git a/include/linux/kernel.h b/include/linux/kernel.h +index 50220cab738c..d68f639f7330 100644 +--- a/include/linux/kernel.h ++++ b/include/linux/kernel.h +@@ -188,6 +188,9 @@ extern int _cond_resched(void); */ # define might_sleep() \ do { __might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0) @@ -12632,7 +12870,7 @@ diff -Nur linux-4.4.62.orig/include/linux/kernel.h linux-4.4.62/include/linux/ke # define sched_annotate_sleep() (current->task_state_change = 0) #else static inline void ___might_sleep(const char *file, int line, -@@ -195,6 +198,7 @@ +@@ -195,6 +198,7 @@ extern int _cond_resched(void); static inline void __might_sleep(const char *file, int line, int preempt_offset) { } # define might_sleep() do { might_resched(); } while (0) @@ -12640,7 +12878,7 @@ diff -Nur linux-4.4.62.orig/include/linux/kernel.h linux-4.4.62/include/linux/ke # define sched_annotate_sleep() do { } while (0) #endif -@@ -255,6 +259,7 @@ +@@ -255,6 +259,7 @@ extern long (*panic_blink)(int state); __printf(1, 2) void panic(const char *fmt, ...) __noreturn __cold; @@ -12648,7 +12886,7 @@ diff -Nur linux-4.4.62.orig/include/linux/kernel.h linux-4.4.62/include/linux/ke extern void oops_enter(void); extern void oops_exit(void); void print_oops_end_marker(void); -@@ -448,6 +453,14 @@ +@@ -448,6 +453,14 @@ extern int sysctl_panic_on_stackoverflow; extern bool crash_kexec_post_notifiers; /* @@ -12663,7 +12901,7 @@ diff -Nur linux-4.4.62.orig/include/linux/kernel.h linux-4.4.62/include/linux/ke * Only to be used by arch init code. If the user over-wrote the default * CONFIG_PANIC_TIMEOUT, honor it. */ -@@ -475,6 +488,7 @@ +@@ -475,6 +488,7 @@ extern enum system_states { SYSTEM_HALT, SYSTEM_POWER_OFF, SYSTEM_RESTART, @@ -12671,9 +12909,10 @@ diff -Nur linux-4.4.62.orig/include/linux/kernel.h linux-4.4.62/include/linux/ke } system_state; #define TAINT_PROPRIETARY_MODULE 0 -diff -Nur linux-4.4.62.orig/include/linux/kvm_host.h linux-4.4.62/include/linux/kvm_host.h ---- linux-4.4.62.orig/include/linux/kvm_host.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/kvm_host.h 2017-04-18 17:38:08.126646608 +0200 +diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h +index d7ce4e3280db..0b6d392b38e7 100644 +--- a/include/linux/kvm_host.h ++++ b/include/linux/kvm_host.h @@ -25,6 +25,7 @@ #include #include @@ -12682,7 +12921,7 @@ diff -Nur linux-4.4.62.orig/include/linux/kvm_host.h linux-4.4.62/include/linux/ #include #include -@@ -243,7 +244,7 @@ +@@ -243,7 +244,7 @@ struct kvm_vcpu { int fpu_active; int guest_fpu_loaded, guest_xcr0_loaded; unsigned char fpu_counter; @@ -12691,7 +12930,7 @@ diff -Nur linux-4.4.62.orig/include/linux/kvm_host.h linux-4.4.62/include/linux/ struct pid *pid; int sigset_active; sigset_t sigset; -@@ -794,7 +795,7 @@ +@@ -794,7 +795,7 @@ static inline bool kvm_arch_has_assigned_device(struct kvm *kvm) } #endif @@ -12700,9 +12939,10 @@ diff -Nur linux-4.4.62.orig/include/linux/kvm_host.h linux-4.4.62/include/linux/ { #ifdef __KVM_HAVE_ARCH_WQP return vcpu->arch.wqp; -diff -Nur linux-4.4.62.orig/include/linux/lglock.h linux-4.4.62/include/linux/lglock.h ---- linux-4.4.62.orig/include/linux/lglock.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/lglock.h 2017-04-18 17:38:08.126646608 +0200 +diff --git a/include/linux/lglock.h b/include/linux/lglock.h +index c92ebd100d9b..6f035f635d0e 100644 +--- a/include/linux/lglock.h ++++ b/include/linux/lglock.h @@ -34,13 +34,30 @@ #endif @@ -12734,7 +12974,7 @@ diff -Nur linux-4.4.62.orig/include/linux/lglock.h linux-4.4.62/include/linux/lg #define DEFINE_LGLOCK(name) \ static DEFINE_PER_CPU(arch_spinlock_t, name ## _lock) \ = __ARCH_SPIN_LOCK_UNLOCKED; \ -@@ -50,6 +67,7 @@ +@@ -50,6 +67,7 @@ struct lglock { static DEFINE_PER_CPU(arch_spinlock_t, name ## _lock) \ = __ARCH_SPIN_LOCK_UNLOCKED; \ static struct lglock name = { .lock = &name ## _lock } @@ -12742,7 +12982,7 @@ diff -Nur linux-4.4.62.orig/include/linux/lglock.h linux-4.4.62/include/linux/lg void lg_lock_init(struct lglock *lg, char *name); -@@ -64,6 +82,12 @@ +@@ -64,6 +82,12 @@ void lg_double_unlock(struct lglock *lg, int cpu1, int cpu2); void lg_global_lock(struct lglock *lg); void lg_global_unlock(struct lglock *lg); @@ -12755,9 +12995,10 @@ diff -Nur linux-4.4.62.orig/include/linux/lglock.h linux-4.4.62/include/linux/lg #else /* When !CONFIG_SMP, map lglock to spinlock */ #define lglock spinlock -diff -Nur linux-4.4.62.orig/include/linux/list_bl.h linux-4.4.62/include/linux/list_bl.h ---- linux-4.4.62.orig/include/linux/list_bl.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/list_bl.h 2017-04-18 17:38:08.126646608 +0200 +diff --git a/include/linux/list_bl.h b/include/linux/list_bl.h +index 8132214e8efd..89ffaa7bd342 100644 +--- a/include/linux/list_bl.h ++++ b/include/linux/list_bl.h @@ -2,6 +2,7 @@ #define _LINUX_LIST_BL_H @@ -12793,7 +13034,7 @@ diff -Nur linux-4.4.62.orig/include/linux/list_bl.h linux-4.4.62/include/linux/l static inline void INIT_HLIST_BL_NODE(struct hlist_bl_node *h) { -@@ -118,12 +130,26 @@ +@@ -118,12 +130,26 @@ static inline void hlist_bl_del_init(struct hlist_bl_node *n) static inline void hlist_bl_lock(struct hlist_bl_head *b) { @@ -12820,9 +13061,11 @@ diff -Nur linux-4.4.62.orig/include/linux/list_bl.h linux-4.4.62/include/linux/l } static inline bool hlist_bl_is_locked(struct hlist_bl_head *b) -diff -Nur linux-4.4.62.orig/include/linux/locallock.h linux-4.4.62/include/linux/locallock.h ---- linux-4.4.62.orig/include/linux/locallock.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/include/linux/locallock.h 2017-04-18 17:38:08.126646608 +0200 +diff --git a/include/linux/locallock.h b/include/linux/locallock.h +new file mode 100644 +index 000000000000..e572a3971631 +--- /dev/null ++++ b/include/linux/locallock.h @@ -0,0 +1,276 @@ +#ifndef _LINUX_LOCALLOCK_H +#define _LINUX_LOCALLOCK_H @@ -13100,9 +13343,10 @@ diff -Nur linux-4.4.62.orig/include/linux/locallock.h linux-4.4.62/include/linux +#endif + +#endif -diff -Nur linux-4.4.62.orig/include/linux/mm_types.h linux-4.4.62/include/linux/mm_types.h ---- linux-4.4.62.orig/include/linux/mm_types.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/mm_types.h 2017-04-18 17:38:08.126646608 +0200 +diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h +index 2ccccbfcd532..4b016eb6265b 100644 +--- a/include/linux/mm_types.h ++++ b/include/linux/mm_types.h @@ -11,6 +11,7 @@ #include #include @@ -13111,7 +13355,7 @@ diff -Nur linux-4.4.62.orig/include/linux/mm_types.h linux-4.4.62/include/linux/ #include #include #include -@@ -505,6 +506,9 @@ +@@ -505,6 +506,9 @@ struct mm_struct { bool tlb_flush_pending; #endif struct uprobes_state uprobes_state; @@ -13121,10 +13365,11 @@ diff -Nur linux-4.4.62.orig/include/linux/mm_types.h linux-4.4.62/include/linux/ #ifdef CONFIG_X86_INTEL_MPX /* address of the bounds directory */ void __user *bd_addr; -diff -Nur linux-4.4.62.orig/include/linux/module.h linux-4.4.62/include/linux/module.h ---- linux-4.4.62.orig/include/linux/module.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/module.h 2017-04-18 17:38:08.126646608 +0200 -@@ -500,6 +500,7 @@ +diff --git a/include/linux/module.h b/include/linux/module.h +index b229a9961d02..5fea847cf95c 100644 +--- a/include/linux/module.h ++++ b/include/linux/module.h +@@ -500,6 +500,7 @@ static inline int module_is_live(struct module *mod) struct module *__module_text_address(unsigned long addr); struct module *__module_address(unsigned long addr); bool is_module_address(unsigned long addr); @@ -13132,21 +13377,22 @@ diff -Nur linux-4.4.62.orig/include/linux/module.h linux-4.4.62/include/linux/mo bool is_module_percpu_address(unsigned long addr); bool is_module_text_address(unsigned long addr); -@@ -664,6 +665,11 @@ - { +@@ -665,6 +666,11 @@ static inline bool is_module_percpu_address(unsigned long addr) return false; } -+ + +static inline bool __is_module_percpu_address(unsigned long addr, unsigned long *can_addr) +{ + return false; +} - ++ static inline bool is_module_text_address(unsigned long addr) { -diff -Nur linux-4.4.62.orig/include/linux/mutex.h linux-4.4.62/include/linux/mutex.h ---- linux-4.4.62.orig/include/linux/mutex.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/mutex.h 2017-04-18 17:38:08.126646608 +0200 + return false; +diff --git a/include/linux/mutex.h b/include/linux/mutex.h +index 2cb7531e7d7a..b3fdfc820216 100644 +--- a/include/linux/mutex.h ++++ b/include/linux/mutex.h @@ -19,6 +19,17 @@ #include #include @@ -13165,7 +13411,7 @@ diff -Nur linux-4.4.62.orig/include/linux/mutex.h linux-4.4.62/include/linux/mut /* * Simple, straightforward mutexes with strict semantics: * -@@ -99,13 +110,6 @@ +@@ -99,13 +110,6 @@ do { \ static inline void mutex_destroy(struct mutex *lock) {} #endif @@ -13179,7 +13425,7 @@ diff -Nur linux-4.4.62.orig/include/linux/mutex.h linux-4.4.62/include/linux/mut #define __MUTEX_INITIALIZER(lockname) \ { .count = ATOMIC_INIT(1) \ , .wait_lock = __SPIN_LOCK_UNLOCKED(lockname.wait_lock) \ -@@ -173,6 +177,8 @@ +@@ -173,6 +177,8 @@ extern int __must_check mutex_lock_killable(struct mutex *lock); extern int mutex_trylock(struct mutex *lock); extern void mutex_unlock(struct mutex *lock); @@ -13188,9 +13434,11 @@ diff -Nur linux-4.4.62.orig/include/linux/mutex.h linux-4.4.62/include/linux/mut extern int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock); #endif /* __LINUX_MUTEX_H */ -diff -Nur linux-4.4.62.orig/include/linux/mutex_rt.h linux-4.4.62/include/linux/mutex_rt.h ---- linux-4.4.62.orig/include/linux/mutex_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/include/linux/mutex_rt.h 2017-04-18 17:38:08.194649247 +0200 +diff --git a/include/linux/mutex_rt.h b/include/linux/mutex_rt.h +new file mode 100644 +index 000000000000..e0284edec655 +--- /dev/null ++++ b/include/linux/mutex_rt.h @@ -0,0 +1,89 @@ +#ifndef __LINUX_MUTEX_RT_H +#define __LINUX_MUTEX_RT_H @@ -13281,10 +13529,11 @@ diff -Nur linux-4.4.62.orig/include/linux/mutex_rt.h linux-4.4.62/include/linux/ +} while (0) + +#endif -diff -Nur linux-4.4.62.orig/include/linux/netdevice.h linux-4.4.62/include/linux/netdevice.h ---- linux-4.4.62.orig/include/linux/netdevice.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/netdevice.h 2017-04-18 17:38:08.194649247 +0200 -@@ -390,7 +390,19 @@ +diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h +index 4035bbe40971..81ebe70d107d 100644 +--- a/include/linux/netdevice.h ++++ b/include/linux/netdevice.h +@@ -390,7 +390,19 @@ typedef enum rx_handler_result rx_handler_result_t; typedef rx_handler_result_t rx_handler_func_t(struct sk_buff **pskb); void __napi_schedule(struct napi_struct *n); @@ -13304,7 +13553,7 @@ diff -Nur linux-4.4.62.orig/include/linux/netdevice.h linux-4.4.62/include/linux static inline bool napi_disable_pending(struct napi_struct *n) { -@@ -2288,11 +2300,20 @@ +@@ -2288,11 +2300,20 @@ void netdev_freemem(struct net_device *dev); void synchronize_net(void); int init_dummy_netdev(struct net_device *dev); @@ -13325,7 +13574,7 @@ diff -Nur linux-4.4.62.orig/include/linux/netdevice.h linux-4.4.62/include/linux struct net_device *dev_get_by_index(struct net *net, int ifindex); struct net_device *__dev_get_by_index(struct net *net, int ifindex); -@@ -2610,6 +2631,7 @@ +@@ -2610,6 +2631,7 @@ struct softnet_data { unsigned int dropped; struct sk_buff_head input_pkt_queue; struct napi_struct backlog; @@ -13333,9 +13582,10 @@ diff -Nur linux-4.4.62.orig/include/linux/netdevice.h linux-4.4.62/include/linux }; -diff -Nur linux-4.4.62.orig/include/linux/netfilter/x_tables.h linux-4.4.62/include/linux/netfilter/x_tables.h ---- linux-4.4.62.orig/include/linux/netfilter/x_tables.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/netfilter/x_tables.h 2017-04-18 17:38:08.194649247 +0200 +diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h +index 04078e8a4803..a61c9609e32f 100644 +--- a/include/linux/netfilter/x_tables.h ++++ b/include/linux/netfilter/x_tables.h @@ -4,6 +4,7 @@ #include @@ -13344,7 +13594,7 @@ diff -Nur linux-4.4.62.orig/include/linux/netfilter/x_tables.h linux-4.4.62/incl #include /** -@@ -289,6 +290,8 @@ +@@ -289,6 +290,8 @@ void xt_free_table_info(struct xt_table_info *info); */ DECLARE_PER_CPU(seqcount_t, xt_recseq); @@ -13353,7 +13603,7 @@ diff -Nur linux-4.4.62.orig/include/linux/netfilter/x_tables.h linux-4.4.62/incl /* xt_tee_enabled - true if x_tables needs to handle reentrancy * * Enabled if current ip(6)tables ruleset has at least one -j TEE rule. -@@ -309,6 +312,9 @@ +@@ -309,6 +312,9 @@ static inline unsigned int xt_write_recseq_begin(void) { unsigned int addend; @@ -13363,7 +13613,7 @@ diff -Nur linux-4.4.62.orig/include/linux/netfilter/x_tables.h linux-4.4.62/incl /* * Low order bit of sequence is set if we already * called xt_write_recseq_begin(). -@@ -339,6 +345,7 @@ +@@ -339,6 +345,7 @@ static inline void xt_write_recseq_end(unsigned int addend) /* this is kind of a write_seqcount_end(), but addend is 0 or 1 */ smp_wmb(); __this_cpu_add(xt_recseq.sequence, addend); @@ -13371,9 +13621,10 @@ diff -Nur linux-4.4.62.orig/include/linux/netfilter/x_tables.h linux-4.4.62/incl } /* -diff -Nur linux-4.4.62.orig/include/linux/notifier.h linux-4.4.62/include/linux/notifier.h ---- linux-4.4.62.orig/include/linux/notifier.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/notifier.h 2017-04-18 17:38:08.194649247 +0200 +diff --git a/include/linux/notifier.h b/include/linux/notifier.h +index d14a4c362465..2e4414a0c1c4 100644 +--- a/include/linux/notifier.h ++++ b/include/linux/notifier.h @@ -6,7 +6,7 @@ * * Alan Cox @@ -13394,7 +13645,7 @@ diff -Nur linux-4.4.62.orig/include/linux/notifier.h linux-4.4.62/include/linux/ */ typedef int (*notifier_fn_t)(struct notifier_block *nb, -@@ -88,7 +86,7 @@ +@@ -88,7 +86,7 @@ struct srcu_notifier_head { (name)->head = NULL; \ } while (0) @@ -13403,7 +13654,7 @@ diff -Nur linux-4.4.62.orig/include/linux/notifier.h linux-4.4.62/include/linux/ extern void srcu_init_notifier_head(struct srcu_notifier_head *nh); #define srcu_cleanup_notifier_head(name) \ cleanup_srcu_struct(&(name)->srcu); -@@ -101,7 +99,13 @@ +@@ -101,7 +99,13 @@ extern void srcu_init_notifier_head(struct srcu_notifier_head *nh); .head = NULL } #define RAW_NOTIFIER_INIT(name) { \ .head = NULL } @@ -13418,7 +13669,7 @@ diff -Nur linux-4.4.62.orig/include/linux/notifier.h linux-4.4.62/include/linux/ #define ATOMIC_NOTIFIER_HEAD(name) \ struct atomic_notifier_head name = \ -@@ -113,6 +117,18 @@ +@@ -113,6 +117,18 @@ extern void srcu_init_notifier_head(struct srcu_notifier_head *nh); struct raw_notifier_head name = \ RAW_NOTIFIER_INIT(name) @@ -13437,7 +13688,7 @@ diff -Nur linux-4.4.62.orig/include/linux/notifier.h linux-4.4.62/include/linux/ #ifdef __KERNEL__ extern int atomic_notifier_chain_register(struct atomic_notifier_head *nh, -@@ -182,12 +198,12 @@ +@@ -182,12 +198,12 @@ static inline int notifier_to_errno(int ret) /* * Declared notifiers so far. I can imagine quite a few more chains @@ -13453,9 +13704,10 @@ diff -Nur linux-4.4.62.orig/include/linux/notifier.h linux-4.4.62/include/linux/ /* CPU notfiers are defined in include/linux/cpu.h. */ /* netdevice notifiers are defined in include/linux/netdevice.h */ -diff -Nur linux-4.4.62.orig/include/linux/percpu.h linux-4.4.62/include/linux/percpu.h ---- linux-4.4.62.orig/include/linux/percpu.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/percpu.h 2017-04-18 17:38:08.194649247 +0200 +diff --git a/include/linux/percpu.h b/include/linux/percpu.h +index caebf2a758dc..4ecc057b6e27 100644 +--- a/include/linux/percpu.h ++++ b/include/linux/percpu.h @@ -24,6 +24,35 @@ PERCPU_MODULE_RESERVE) #endif @@ -13492,7 +13744,7 @@ diff -Nur linux-4.4.62.orig/include/linux/percpu.h linux-4.4.62/include/linux/pe /* minimum unit size, also is the maximum supported allocation size */ #define PCPU_MIN_UNIT_SIZE PFN_ALIGN(32 << 10) -@@ -116,6 +145,7 @@ +@@ -116,6 +145,7 @@ extern int __init pcpu_page_first_chunk(size_t reserved_size, #endif extern void __percpu *__alloc_reserved_percpu(size_t size, size_t align); @@ -13500,9 +13752,10 @@ diff -Nur linux-4.4.62.orig/include/linux/percpu.h linux-4.4.62/include/linux/pe extern bool is_kernel_percpu_address(unsigned long addr); #if !defined(CONFIG_SMP) || !defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) -diff -Nur linux-4.4.62.orig/include/linux/pid.h linux-4.4.62/include/linux/pid.h ---- linux-4.4.62.orig/include/linux/pid.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/pid.h 2017-04-18 17:38:08.194649247 +0200 +diff --git a/include/linux/pid.h b/include/linux/pid.h +index 23705a53abba..2cc64b779f03 100644 +--- a/include/linux/pid.h ++++ b/include/linux/pid.h @@ -2,6 +2,7 @@ #define _LINUX_PID_H @@ -13511,9 +13764,10 @@ diff -Nur linux-4.4.62.orig/include/linux/pid.h linux-4.4.62/include/linux/pid.h enum pid_type { -diff -Nur linux-4.4.62.orig/include/linux/preempt.h linux-4.4.62/include/linux/preempt.h ---- linux-4.4.62.orig/include/linux/preempt.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/preempt.h 2017-04-18 17:38:08.194649247 +0200 +diff --git a/include/linux/preempt.h b/include/linux/preempt.h +index 75e4e30677f1..1cfb1cb72354 100644 +--- a/include/linux/preempt.h ++++ b/include/linux/preempt.h @@ -50,7 +50,11 @@ #define HARDIRQ_OFFSET (1UL << HARDIRQ_SHIFT) #define NMI_OFFSET (1UL << NMI_SHIFT) @@ -13564,7 +13818,7 @@ diff -Nur linux-4.4.62.orig/include/linux/preempt.h linux-4.4.62/include/linux/p /* * The preempt_count offset needed for things like: -@@ -140,6 +153,20 @@ +@@ -140,6 +153,20 @@ extern void preempt_count_sub(int val); #define preempt_count_inc() preempt_count_add(1) #define preempt_count_dec() preempt_count_sub(1) @@ -13585,7 +13839,7 @@ diff -Nur linux-4.4.62.orig/include/linux/preempt.h linux-4.4.62/include/linux/p #ifdef CONFIG_PREEMPT_COUNT #define preempt_disable() \ -@@ -148,13 +175,25 @@ +@@ -148,13 +175,25 @@ do { \ barrier(); \ } while (0) @@ -13612,7 +13866,7 @@ diff -Nur linux-4.4.62.orig/include/linux/preempt.h linux-4.4.62/include/linux/p #define preemptible() (preempt_count() == 0 && !irqs_disabled()) -@@ -179,6 +218,13 @@ +@@ -179,6 +218,13 @@ do { \ __preempt_schedule(); \ } while (0) @@ -13626,7 +13880,7 @@ diff -Nur linux-4.4.62.orig/include/linux/preempt.h linux-4.4.62/include/linux/p #else /* !CONFIG_PREEMPT */ #define preempt_enable() \ do { \ -@@ -224,6 +270,7 @@ +@@ -224,6 +270,7 @@ do { \ #define preempt_disable_notrace() barrier() #define preempt_enable_no_resched_notrace() barrier() #define preempt_enable_notrace() barrier() @@ -13634,7 +13888,7 @@ diff -Nur linux-4.4.62.orig/include/linux/preempt.h linux-4.4.62/include/linux/p #define preemptible() 0 #endif /* CONFIG_PREEMPT_COUNT */ -@@ -244,10 +291,31 @@ +@@ -244,10 +291,31 @@ do { \ } while (0) #define preempt_fold_need_resched() \ do { \ @@ -13667,10 +13921,11 @@ diff -Nur linux-4.4.62.orig/include/linux/preempt.h linux-4.4.62/include/linux/p #ifdef CONFIG_PREEMPT_NOTIFIERS struct preempt_notifier; -diff -Nur linux-4.4.62.orig/include/linux/printk.h linux-4.4.62/include/linux/printk.h ---- linux-4.4.62.orig/include/linux/printk.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/printk.h 2017-04-18 17:38:08.194649247 +0200 -@@ -117,9 +117,11 @@ +diff --git a/include/linux/printk.h b/include/linux/printk.h +index 9729565c25ff..9cdca696b718 100644 +--- a/include/linux/printk.h ++++ b/include/linux/printk.h +@@ -117,9 +117,11 @@ int no_printk(const char *fmt, ...) #ifdef CONFIG_EARLY_PRINTK extern asmlinkage __printf(1, 2) void early_printk(const char *fmt, ...); @@ -13682,10 +13937,11 @@ diff -Nur linux-4.4.62.orig/include/linux/printk.h linux-4.4.62/include/linux/pr #endif typedef __printf(1, 0) int (*printk_func_t)(const char *fmt, va_list args); -diff -Nur linux-4.4.62.orig/include/linux/radix-tree.h linux-4.4.62/include/linux/radix-tree.h ---- linux-4.4.62.orig/include/linux/radix-tree.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/radix-tree.h 2017-04-18 17:38:08.194649247 +0200 -@@ -279,6 +279,8 @@ +diff --git a/include/linux/radix-tree.h b/include/linux/radix-tree.h +index 5d5174b59802..327dddaf4c8f 100644 +--- a/include/linux/radix-tree.h ++++ b/include/linux/radix-tree.h +@@ -279,6 +279,8 @@ unsigned int radix_tree_gang_lookup_slot(struct radix_tree_root *root, unsigned long first_index, unsigned int max_items); int radix_tree_preload(gfp_t gfp_mask); int radix_tree_maybe_preload(gfp_t gfp_mask); @@ -13694,7 +13950,7 @@ diff -Nur linux-4.4.62.orig/include/linux/radix-tree.h linux-4.4.62/include/linu void radix_tree_init(void); void *radix_tree_tag_set(struct radix_tree_root *root, unsigned long index, unsigned int tag); -@@ -301,11 +303,6 @@ +@@ -301,11 +303,6 @@ unsigned long radix_tree_range_tag_if_tagged(struct radix_tree_root *root, int radix_tree_tagged(struct radix_tree_root *root, unsigned int tag); unsigned long radix_tree_locate_item(struct radix_tree_root *root, void *item); @@ -13706,10 +13962,11 @@ diff -Nur linux-4.4.62.orig/include/linux/radix-tree.h linux-4.4.62/include/linu /** * struct radix_tree_iter - radix tree iterator state * -diff -Nur linux-4.4.62.orig/include/linux/random.h linux-4.4.62/include/linux/random.h ---- linux-4.4.62.orig/include/linux/random.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/random.h 2017-04-18 17:38:08.194649247 +0200 -@@ -20,7 +20,7 @@ +diff --git a/include/linux/random.h b/include/linux/random.h +index a75840c1aa71..1a804361670c 100644 +--- a/include/linux/random.h ++++ b/include/linux/random.h +@@ -20,7 +20,7 @@ struct random_ready_callback { extern void add_device_randomness(const void *, unsigned int); extern void add_input_randomness(unsigned int type, unsigned int code, unsigned int value); @@ -13718,9 +13975,10 @@ diff -Nur linux-4.4.62.orig/include/linux/random.h linux-4.4.62/include/linux/ra extern void get_random_bytes(void *buf, int nbytes); extern int add_random_ready_callback(struct random_ready_callback *rdy); -diff -Nur linux-4.4.62.orig/include/linux/rbtree.h linux-4.4.62/include/linux/rbtree.h ---- linux-4.4.62.orig/include/linux/rbtree.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/rbtree.h 2017-04-18 17:38:08.194649247 +0200 +diff --git a/include/linux/rbtree.h b/include/linux/rbtree.h +index a5aa7ae671f4..24ddffd25492 100644 +--- a/include/linux/rbtree.h ++++ b/include/linux/rbtree.h @@ -31,7 +31,6 @@ #include @@ -13729,7 +13987,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rbtree.h linux-4.4.62/include/linux/rb struct rb_node { unsigned long __rb_parent_color; -@@ -86,14 +85,8 @@ +@@ -86,14 +85,8 @@ static inline void rb_link_node(struct rb_node *node, struct rb_node *parent, *rb_link = node; } @@ -13746,10 +14004,11 @@ diff -Nur linux-4.4.62.orig/include/linux/rbtree.h linux-4.4.62/include/linux/rb #define rb_entry_safe(ptr, type, member) \ ({ typeof(ptr) ____ptr = (ptr); \ -diff -Nur linux-4.4.62.orig/include/linux/rcupdate.h linux-4.4.62/include/linux/rcupdate.h ---- linux-4.4.62.orig/include/linux/rcupdate.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/rcupdate.h 2017-04-18 17:38:08.194649247 +0200 -@@ -169,6 +169,9 @@ +diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h +index a0189ba67fde..c2f5f955163d 100644 +--- a/include/linux/rcupdate.h ++++ b/include/linux/rcupdate.h +@@ -169,6 +169,9 @@ void call_rcu(struct rcu_head *head, #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ @@ -13759,7 +14018,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rcupdate.h linux-4.4.62/include/linux/ /** * call_rcu_bh() - Queue an RCU for invocation after a quicker grace period. * @head: structure to be used for queueing the RCU updates. -@@ -192,6 +195,7 @@ +@@ -192,6 +195,7 @@ void call_rcu(struct rcu_head *head, */ void call_rcu_bh(struct rcu_head *head, rcu_callback_t func); @@ -13767,7 +14026,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rcupdate.h linux-4.4.62/include/linux/ /** * call_rcu_sched() - Queue an RCU for invocation after sched grace period. -@@ -292,6 +296,11 @@ +@@ -292,6 +296,11 @@ void synchronize_rcu(void); * types of kernel builds, the rcu_read_lock() nesting depth is unknowable. */ #define rcu_preempt_depth() (current->rcu_read_lock_nesting) @@ -13779,7 +14038,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rcupdate.h linux-4.4.62/include/linux/ #else /* #ifdef CONFIG_PREEMPT_RCU */ -@@ -317,6 +326,8 @@ +@@ -317,6 +326,8 @@ static inline int rcu_preempt_depth(void) return 0; } @@ -13788,7 +14047,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rcupdate.h linux-4.4.62/include/linux/ #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ /* Internal to kernel */ -@@ -489,7 +500,14 @@ +@@ -489,7 +500,14 @@ extern struct lockdep_map rcu_callback_map; int debug_lockdep_rcu_enabled(void); int rcu_read_lock_held(void); @@ -13803,7 +14062,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rcupdate.h linux-4.4.62/include/linux/ /** * rcu_read_lock_sched_held() - might we be in RCU-sched read-side critical section? -@@ -937,10 +955,14 @@ +@@ -937,10 +955,14 @@ static inline void rcu_read_unlock(void) static inline void rcu_read_lock_bh(void) { local_bh_disable(); @@ -13818,7 +14077,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rcupdate.h linux-4.4.62/include/linux/ } /* -@@ -950,10 +972,14 @@ +@@ -950,10 +972,14 @@ static inline void rcu_read_lock_bh(void) */ static inline void rcu_read_unlock_bh(void) { @@ -13833,10 +14092,11 @@ diff -Nur linux-4.4.62.orig/include/linux/rcupdate.h linux-4.4.62/include/linux/ local_bh_enable(); } -diff -Nur linux-4.4.62.orig/include/linux/rcutree.h linux-4.4.62/include/linux/rcutree.h ---- linux-4.4.62.orig/include/linux/rcutree.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/rcutree.h 2017-04-18 17:38:08.194649247 +0200 -@@ -44,7 +44,11 @@ +diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h +index 60d15a080d7c..436c9e62bfc6 100644 +--- a/include/linux/rcutree.h ++++ b/include/linux/rcutree.h +@@ -44,7 +44,11 @@ static inline void rcu_virt_note_context_switch(int cpu) rcu_note_context_switch(); } @@ -13848,7 +14108,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rcutree.h linux-4.4.62/include/linux/r void synchronize_sched_expedited(void); void synchronize_rcu_expedited(void); -@@ -72,7 +76,11 @@ +@@ -72,7 +76,11 @@ static inline void synchronize_rcu_bh_expedited(void) } void rcu_barrier(void); @@ -13860,7 +14120,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rcutree.h linux-4.4.62/include/linux/r void rcu_barrier_sched(void); unsigned long get_state_synchronize_rcu(void); void cond_synchronize_rcu(unsigned long oldstate); -@@ -85,12 +93,10 @@ +@@ -85,12 +93,10 @@ unsigned long rcu_batches_started(void); unsigned long rcu_batches_started_bh(void); unsigned long rcu_batches_started_sched(void); unsigned long rcu_batches_completed(void); @@ -13873,7 +14133,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rcutree.h linux-4.4.62/include/linux/r void rcu_sched_force_quiescent_state(void); void rcu_idle_enter(void); -@@ -105,6 +111,14 @@ +@@ -105,6 +111,14 @@ extern int rcu_scheduler_active __read_mostly; bool rcu_is_watching(void); @@ -13888,9 +14148,10 @@ diff -Nur linux-4.4.62.orig/include/linux/rcutree.h linux-4.4.62/include/linux/r void rcu_all_qs(void); #endif /* __LINUX_RCUTREE_H */ -diff -Nur linux-4.4.62.orig/include/linux/rtmutex.h linux-4.4.62/include/linux/rtmutex.h ---- linux-4.4.62.orig/include/linux/rtmutex.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/rtmutex.h 2017-04-18 17:38:08.194649247 +0200 +diff --git a/include/linux/rtmutex.h b/include/linux/rtmutex.h +index 1abba5ce2a2f..30211c627511 100644 +--- a/include/linux/rtmutex.h ++++ b/include/linux/rtmutex.h @@ -13,11 +13,15 @@ #define __LINUX_RT_MUTEX_H @@ -13908,7 +14169,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rtmutex.h linux-4.4.62/include/linux/r /** * The rt_mutex structure * -@@ -31,8 +35,8 @@ +@@ -31,8 +35,8 @@ struct rt_mutex { struct rb_root waiters; struct rb_node *waiters_leftmost; struct task_struct *owner; @@ -13918,7 +14179,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rtmutex.h linux-4.4.62/include/linux/r const char *name, *file; int line; void *magic; -@@ -55,22 +59,33 @@ +@@ -55,22 +59,33 @@ struct hrtimer_sleeper; # define rt_mutex_debug_check_no_locks_held(task) do { } while (0) #endif @@ -13957,7 +14218,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rtmutex.h linux-4.4.62/include/linux/r #define DEFINE_RT_MUTEX(mutexname) \ struct rt_mutex mutexname = __RT_MUTEX_INITIALIZER(mutexname) -@@ -91,6 +106,7 @@ +@@ -91,6 +106,7 @@ extern void rt_mutex_destroy(struct rt_mutex *lock); extern void rt_mutex_lock(struct rt_mutex *lock); extern int rt_mutex_lock_interruptible(struct rt_mutex *lock); @@ -13965,9 +14226,11 @@ diff -Nur linux-4.4.62.orig/include/linux/rtmutex.h linux-4.4.62/include/linux/r extern int rt_mutex_timed_lock(struct rt_mutex *lock, struct hrtimer_sleeper *timeout); -diff -Nur linux-4.4.62.orig/include/linux/rwlock_rt.h linux-4.4.62/include/linux/rwlock_rt.h ---- linux-4.4.62.orig/include/linux/rwlock_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/include/linux/rwlock_rt.h 2017-04-18 17:38:08.194649247 +0200 +diff --git a/include/linux/rwlock_rt.h b/include/linux/rwlock_rt.h +new file mode 100644 +index 000000000000..49ed2d45d3be +--- /dev/null ++++ b/include/linux/rwlock_rt.h @@ -0,0 +1,99 @@ +#ifndef __LINUX_RWLOCK_RT_H +#define __LINUX_RWLOCK_RT_H @@ -14068,9 +14331,10 @@ diff -Nur linux-4.4.62.orig/include/linux/rwlock_rt.h linux-4.4.62/include/linux + } while (0) + +#endif -diff -Nur linux-4.4.62.orig/include/linux/rwlock_types.h linux-4.4.62/include/linux/rwlock_types.h ---- linux-4.4.62.orig/include/linux/rwlock_types.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/rwlock_types.h 2017-04-18 17:38:08.194649247 +0200 +diff --git a/include/linux/rwlock_types.h b/include/linux/rwlock_types.h +index cc0072e93e36..d0da966ad7a0 100644 +--- a/include/linux/rwlock_types.h ++++ b/include/linux/rwlock_types.h @@ -1,6 +1,10 @@ #ifndef __LINUX_RWLOCK_TYPES_H #define __LINUX_RWLOCK_TYPES_H @@ -14082,7 +14346,7 @@ diff -Nur linux-4.4.62.orig/include/linux/rwlock_types.h linux-4.4.62/include/li /* * include/linux/rwlock_types.h - generic rwlock type definitions * and initializers -@@ -43,6 +47,7 @@ +@@ -43,6 +47,7 @@ typedef struct { RW_DEP_MAP_INIT(lockname) } #endif @@ -14091,9 +14355,11 @@ diff -Nur linux-4.4.62.orig/include/linux/rwlock_types.h linux-4.4.62/include/li + rwlock_t name __cacheline_aligned_in_smp = __RW_LOCK_UNLOCKED(name) #endif /* __LINUX_RWLOCK_TYPES_H */ -diff -Nur linux-4.4.62.orig/include/linux/rwlock_types_rt.h linux-4.4.62/include/linux/rwlock_types_rt.h ---- linux-4.4.62.orig/include/linux/rwlock_types_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/include/linux/rwlock_types_rt.h 2017-04-18 17:38:08.194649247 +0200 +diff --git a/include/linux/rwlock_types_rt.h b/include/linux/rwlock_types_rt.h +new file mode 100644 +index 000000000000..b13832119591 +--- /dev/null ++++ b/include/linux/rwlock_types_rt.h @@ -0,0 +1,33 @@ +#ifndef __LINUX_RWLOCK_TYPES_RT_H +#define __LINUX_RWLOCK_TYPES_RT_H @@ -14128,9 +14394,10 @@ diff -Nur linux-4.4.62.orig/include/linux/rwlock_types_rt.h linux-4.4.62/include + rwlock_t name __cacheline_aligned_in_smp = __RW_LOCK_UNLOCKED(name) + +#endif -diff -Nur linux-4.4.62.orig/include/linux/rwsem.h linux-4.4.62/include/linux/rwsem.h ---- linux-4.4.62.orig/include/linux/rwsem.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/rwsem.h 2017-04-18 17:38:08.198649401 +0200 +diff --git a/include/linux/rwsem.h b/include/linux/rwsem.h +index 8f498cdde280..2b2148431f14 100644 +--- a/include/linux/rwsem.h ++++ b/include/linux/rwsem.h @@ -18,6 +18,10 @@ #include #endif @@ -14142,16 +14409,18 @@ diff -Nur linux-4.4.62.orig/include/linux/rwsem.h linux-4.4.62/include/linux/rws struct rw_semaphore; #ifdef CONFIG_RWSEM_GENERIC_SPINLOCK -@@ -177,4 +181,6 @@ +@@ -177,4 +181,6 @@ extern void up_read_non_owner(struct rw_semaphore *sem); # define up_read_non_owner(sem) up_read(sem) #endif +#endif /* !PREEMPT_RT_FULL */ + #endif /* _LINUX_RWSEM_H */ -diff -Nur linux-4.4.62.orig/include/linux/rwsem_rt.h linux-4.4.62/include/linux/rwsem_rt.h ---- linux-4.4.62.orig/include/linux/rwsem_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/include/linux/rwsem_rt.h 2017-04-18 17:38:08.198649401 +0200 +diff --git a/include/linux/rwsem_rt.h b/include/linux/rwsem_rt.h +new file mode 100644 +index 000000000000..f97860b2e2a4 +--- /dev/null ++++ b/include/linux/rwsem_rt.h @@ -0,0 +1,152 @@ +#ifndef _LINUX_RWSEM_RT_H +#define _LINUX_RWSEM_RT_H @@ -14305,10 +14574,11 @@ diff -Nur linux-4.4.62.orig/include/linux/rwsem_rt.h linux-4.4.62/include/linux/ +} +#endif +#endif -diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sched.h ---- linux-4.4.62.orig/include/linux/sched.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/sched.h 2017-04-18 17:38:08.198649401 +0200 -@@ -26,6 +26,7 @@ +diff --git a/include/linux/sched.h b/include/linux/sched.h +index 352213b360d7..1c7aa9446684 100644 +--- a/include/linux/sched.h ++++ b/include/linux/sched.h +@@ -26,6 +26,7 @@ struct sched_param { #include #include #include @@ -14316,7 +14586,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch #include #include -@@ -182,8 +183,6 @@ +@@ -182,8 +183,6 @@ extern void update_cpu_load_nohz(void); static inline void update_cpu_load_nohz(void) { } #endif @@ -14325,7 +14595,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch extern void dump_cpu_task(int cpu); struct seq_file; -@@ -242,10 +241,7 @@ +@@ -242,10 +241,7 @@ extern char ___assert_task_state[1 - 2*!!( TASK_UNINTERRUPTIBLE | __TASK_STOPPED | \ __TASK_TRACED | EXIT_ZOMBIE | EXIT_DEAD) @@ -14336,7 +14606,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch #define task_contributes_to_load(task) \ ((task->state & TASK_UNINTERRUPTIBLE) != 0 && \ (task->flags & PF_FROZEN) == 0 && \ -@@ -311,6 +307,11 @@ +@@ -311,6 +307,11 @@ extern char ___assert_task_state[1 - 2*!!( #endif @@ -14348,7 +14618,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch /* Task command name length */ #define TASK_COMM_LEN 16 -@@ -970,8 +971,18 @@ +@@ -970,8 +971,18 @@ struct wake_q_head { struct wake_q_head name = { WAKE_Q_TAIL, &name.first } extern void wake_q_add(struct wake_q_head *head, @@ -14369,7 +14639,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch /* * sched-domains (multiprocessor balancing) declarations: -@@ -1379,6 +1390,7 @@ +@@ -1379,6 +1390,7 @@ struct tlbflush_unmap_batch { struct task_struct { volatile long state; /* -1 unrunnable, 0 runnable, >0 stopped */ @@ -14377,7 +14647,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch void *stack; atomic_t usage; unsigned int flags; /* per process flags, defined below */ -@@ -1415,6 +1427,12 @@ +@@ -1415,6 +1427,12 @@ struct task_struct { #endif unsigned int policy; @@ -14390,7 +14660,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch int nr_cpus_allowed; cpumask_t cpus_allowed; -@@ -1522,11 +1540,14 @@ +@@ -1526,11 +1544,14 @@ struct task_struct { cputime_t gtime; struct prev_cputime prev_cputime; #ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN @@ -14407,7 +14677,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch VTIME_SYS, } vtime_snap_whence; #endif -@@ -1538,6 +1559,9 @@ +@@ -1542,6 +1563,9 @@ struct task_struct { struct task_cputime cputime_expires; struct list_head cpu_timers[3]; @@ -14417,7 +14687,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch /* process credentials */ const struct cred __rcu *ptracer_cred; /* Tracer's credentials at attach */ -@@ -1569,10 +1593,15 @@ +@@ -1573,10 +1597,15 @@ struct task_struct { /* signal handlers */ struct signal_struct *signal; struct sighand_struct *sighand; @@ -14433,7 +14703,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch unsigned long sas_ss_sp; size_t sas_ss_size; -@@ -1796,6 +1825,12 @@ +@@ -1800,6 +1829,12 @@ struct task_struct { unsigned long trace; /* bitmask and counter of trace recursion */ unsigned long trace_recursion; @@ -14446,7 +14716,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch #endif /* CONFIG_TRACING */ #ifdef CONFIG_MEMCG struct mem_cgroup *memcg_in_oom; -@@ -1812,9 +1847,23 @@ +@@ -1816,9 +1851,23 @@ struct task_struct { unsigned int sequential_io; unsigned int sequential_io_avg; #endif @@ -14470,7 +14740,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch int pagefault_disabled; /* CPU-specific state of this task */ struct thread_struct thread; -@@ -1832,9 +1881,6 @@ +@@ -1836,9 +1885,6 @@ extern int arch_task_struct_size __read_mostly; # define arch_task_struct_size (sizeof(struct task_struct)) #endif @@ -14480,7 +14750,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch #define TNF_MIGRATED 0x01 #define TNF_NO_GROUP 0x02 #define TNF_SHARED 0x04 -@@ -2024,6 +2070,15 @@ +@@ -2028,6 +2074,15 @@ extern struct pid *cad_pid; extern void free_task(struct task_struct *tsk); #define get_task_struct(tsk) do { atomic_inc(&(tsk)->usage); } while(0) @@ -14496,7 +14766,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch extern void __put_task_struct(struct task_struct *t); static inline void put_task_struct(struct task_struct *t) -@@ -2031,6 +2086,7 @@ +@@ -2035,6 +2090,7 @@ static inline void put_task_struct(struct task_struct *t) if (atomic_dec_and_test(&t->usage)) __put_task_struct(t); } @@ -14504,7 +14774,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch #ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN extern void task_cputime(struct task_struct *t, -@@ -2069,6 +2125,7 @@ +@@ -2073,6 +2129,7 @@ extern void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, /* * Per process flags */ @@ -14512,7 +14782,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch #define PF_EXITING 0x00000004 /* getting shut down */ #define PF_EXITPIDONE 0x00000008 /* pi exit done on shut down */ #define PF_VCPU 0x00000010 /* I'm a virtual CPU */ -@@ -2233,6 +2290,10 @@ +@@ -2237,6 +2294,10 @@ extern void do_set_cpus_allowed(struct task_struct *p, extern int set_cpus_allowed_ptr(struct task_struct *p, const struct cpumask *new_mask); @@ -14523,7 +14793,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch #else static inline void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask) -@@ -2245,6 +2306,9 @@ +@@ -2249,6 +2310,9 @@ static inline int set_cpus_allowed_ptr(struct task_struct *p, return -EINVAL; return 0; } @@ -14533,7 +14803,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch #endif #ifdef CONFIG_NO_HZ_COMMON -@@ -2454,6 +2518,7 @@ +@@ -2458,6 +2522,7 @@ extern void xtime_update(unsigned long ticks); extern int wake_up_state(struct task_struct *tsk, unsigned int state); extern int wake_up_process(struct task_struct *tsk); @@ -14541,7 +14811,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch extern void wake_up_new_task(struct task_struct *tsk); #ifdef CONFIG_SMP extern void kick_process(struct task_struct *tsk); -@@ -2577,12 +2642,24 @@ +@@ -2581,12 +2646,24 @@ extern struct mm_struct * mm_alloc(void); /* mmdrop drops the mm and the page tables */ extern void __mmdrop(struct mm_struct *); @@ -14566,7 +14836,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch /* mmput gets rid of the mappings and all user-space */ extern void mmput(struct mm_struct *); /* Grab a reference to a task's mm, if it is not already going away */ -@@ -2892,6 +2969,43 @@ +@@ -2896,6 +2973,43 @@ static inline int test_tsk_need_resched(struct task_struct *tsk) return unlikely(test_tsk_thread_flag(tsk,TIF_NEED_RESCHED)); } @@ -14610,7 +14880,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch static inline int restart_syscall(void) { set_tsk_thread_flag(current, TIF_SIGPENDING); -@@ -2923,6 +3037,51 @@ +@@ -2927,6 +3041,51 @@ static inline int signal_pending_state(long state, struct task_struct *p) return (state & TASK_INTERRUPTIBLE) || __fatal_signal_pending(p); } @@ -14662,7 +14932,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch /* * cond_resched() and cond_resched_lock(): latency reduction via * explicit rescheduling in places that are safe. The return -@@ -2944,12 +3103,16 @@ +@@ -2948,12 +3107,16 @@ extern int __cond_resched_lock(spinlock_t *lock); __cond_resched_lock(lock); \ }) @@ -14679,7 +14949,7 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch static inline void cond_resched_rcu(void) { -@@ -3111,6 +3274,31 @@ +@@ -3115,6 +3278,31 @@ static inline void set_task_cpu(struct task_struct *p, unsigned int cpu) #endif /* CONFIG_SMP */ @@ -14711,10 +14981,11 @@ diff -Nur linux-4.4.62.orig/include/linux/sched.h linux-4.4.62/include/linux/sch extern long sched_setaffinity(pid_t pid, const struct cpumask *new_mask); extern long sched_getaffinity(pid_t pid, struct cpumask *mask); -diff -Nur linux-4.4.62.orig/include/linux/seqlock.h linux-4.4.62/include/linux/seqlock.h ---- linux-4.4.62.orig/include/linux/seqlock.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/seqlock.h 2017-04-18 17:38:08.198649401 +0200 -@@ -220,20 +220,30 @@ +diff --git a/include/linux/seqlock.h b/include/linux/seqlock.h +index e0582106ef4f..b14f4d2368aa 100644 +--- a/include/linux/seqlock.h ++++ b/include/linux/seqlock.h +@@ -220,20 +220,30 @@ static inline int read_seqcount_retry(const seqcount_t *s, unsigned start) return __read_seqcount_retry(s, start); } @@ -14749,7 +15020,7 @@ diff -Nur linux-4.4.62.orig/include/linux/seqlock.h linux-4.4.62/include/linux/s /** * raw_write_seqcount_barrier - do a seq write barrier * @s: pointer to seqcount_t -@@ -425,10 +435,32 @@ +@@ -425,10 +435,32 @@ typedef struct { /* * Read side functions for starting and finalizing a read side section. */ @@ -14782,7 +15053,7 @@ diff -Nur linux-4.4.62.orig/include/linux/seqlock.h linux-4.4.62/include/linux/s static inline unsigned read_seqretry(const seqlock_t *sl, unsigned start) { -@@ -443,36 +475,36 @@ +@@ -443,36 +475,36 @@ static inline unsigned read_seqretry(const seqlock_t *sl, unsigned start) static inline void write_seqlock(seqlock_t *sl) { spin_lock(&sl->lock); @@ -14825,7 +15096,7 @@ diff -Nur linux-4.4.62.orig/include/linux/seqlock.h linux-4.4.62/include/linux/s spin_unlock_irq(&sl->lock); } -@@ -481,7 +513,7 @@ +@@ -481,7 +513,7 @@ static inline unsigned long __write_seqlock_irqsave(seqlock_t *sl) unsigned long flags; spin_lock_irqsave(&sl->lock, flags); @@ -14834,7 +15105,7 @@ diff -Nur linux-4.4.62.orig/include/linux/seqlock.h linux-4.4.62/include/linux/s return flags; } -@@ -491,7 +523,7 @@ +@@ -491,7 +523,7 @@ static inline unsigned long __write_seqlock_irqsave(seqlock_t *sl) static inline void write_sequnlock_irqrestore(seqlock_t *sl, unsigned long flags) { @@ -14843,10 +15114,11 @@ diff -Nur linux-4.4.62.orig/include/linux/seqlock.h linux-4.4.62/include/linux/s spin_unlock_irqrestore(&sl->lock, flags); } -diff -Nur linux-4.4.62.orig/include/linux/signal.h linux-4.4.62/include/linux/signal.h ---- linux-4.4.62.orig/include/linux/signal.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/signal.h 2017-04-18 17:38:08.198649401 +0200 -@@ -233,6 +233,7 @@ +diff --git a/include/linux/signal.h b/include/linux/signal.h +index d80259afb9e5..ddd1e6866a54 100644 +--- a/include/linux/signal.h ++++ b/include/linux/signal.h +@@ -233,6 +233,7 @@ static inline void init_sigpending(struct sigpending *sig) } extern void flush_sigqueue(struct sigpending *queue); @@ -14854,10 +15126,11 @@ diff -Nur linux-4.4.62.orig/include/linux/signal.h linux-4.4.62/include/linux/si /* Test if 'sig' is valid signal. Use this instead of testing _NSIG directly */ static inline int valid_signal(unsigned long sig) -diff -Nur linux-4.4.62.orig/include/linux/skbuff.h linux-4.4.62/include/linux/skbuff.h ---- linux-4.4.62.orig/include/linux/skbuff.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/skbuff.h 2017-04-18 17:38:08.198649401 +0200 -@@ -203,6 +203,7 @@ +diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h +index d443d9ab0236..2d1c7f9b7fd0 100644 +--- a/include/linux/skbuff.h ++++ b/include/linux/skbuff.h +@@ -203,6 +203,7 @@ struct sk_buff_head { __u32 qlen; spinlock_t lock; @@ -14865,7 +15138,7 @@ diff -Nur linux-4.4.62.orig/include/linux/skbuff.h linux-4.4.62/include/linux/sk }; struct sk_buff; -@@ -1465,6 +1466,12 @@ +@@ -1465,6 +1466,12 @@ static inline void skb_queue_head_init(struct sk_buff_head *list) __skb_queue_head_init(list); } @@ -14878,10 +15151,11 @@ diff -Nur linux-4.4.62.orig/include/linux/skbuff.h linux-4.4.62/include/linux/sk static inline void skb_queue_head_init_class(struct sk_buff_head *list, struct lock_class_key *class) { -diff -Nur linux-4.4.62.orig/include/linux/smp.h linux-4.4.62/include/linux/smp.h ---- linux-4.4.62.orig/include/linux/smp.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/smp.h 2017-04-18 17:38:08.198649401 +0200 -@@ -185,6 +185,9 @@ +diff --git a/include/linux/smp.h b/include/linux/smp.h +index c4414074bd88..e6ab36aeaaab 100644 +--- a/include/linux/smp.h ++++ b/include/linux/smp.h +@@ -185,6 +185,9 @@ static inline void smp_init(void) { } #define get_cpu() ({ preempt_disable(); smp_processor_id(); }) #define put_cpu() preempt_enable() @@ -14891,23 +15165,11 @@ diff -Nur linux-4.4.62.orig/include/linux/smp.h linux-4.4.62/include/linux/smp.h /* * Callback to arch code if there's nosmp or maxcpus=0 on the * boot command line: -diff -Nur linux-4.4.62.orig/include/linux/spinlock_api_smp.h linux-4.4.62/include/linux/spinlock_api_smp.h ---- linux-4.4.62.orig/include/linux/spinlock_api_smp.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/spinlock_api_smp.h 2017-04-18 17:38:08.198649401 +0200 -@@ -189,6 +189,8 @@ - return 0; - } - --#include -+#ifndef CONFIG_PREEMPT_RT_FULL -+# include -+#endif - - #endif /* __LINUX_SPINLOCK_API_SMP_H */ -diff -Nur linux-4.4.62.orig/include/linux/spinlock.h linux-4.4.62/include/linux/spinlock.h ---- linux-4.4.62.orig/include/linux/spinlock.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/spinlock.h 2017-04-18 17:38:08.198649401 +0200 -@@ -271,7 +271,11 @@ +diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h +index 47dd0cebd204..02928fa5499d 100644 +--- a/include/linux/spinlock.h ++++ b/include/linux/spinlock.h +@@ -271,7 +271,11 @@ static inline void do_raw_spin_unlock(raw_spinlock_t *lock) __releases(lock) #define raw_spin_can_lock(lock) (!raw_spin_is_locked(lock)) /* Include rwlock functions */ @@ -14920,7 +15182,7 @@ diff -Nur linux-4.4.62.orig/include/linux/spinlock.h linux-4.4.62/include/linux/ /* * Pull the _spin_*()/_read_*()/_write_*() functions/declarations: -@@ -282,6 +286,10 @@ +@@ -282,6 +286,10 @@ static inline void do_raw_spin_unlock(raw_spinlock_t *lock) __releases(lock) # include #endif @@ -14931,7 +15193,7 @@ diff -Nur linux-4.4.62.orig/include/linux/spinlock.h linux-4.4.62/include/linux/ /* * Map the spin_lock functions to the raw variants for PREEMPT_RT=n */ -@@ -347,6 +355,12 @@ +@@ -347,6 +355,12 @@ static __always_inline void spin_unlock(spinlock_t *lock) raw_spin_unlock(&lock->rlock); } @@ -14944,16 +15206,32 @@ diff -Nur linux-4.4.62.orig/include/linux/spinlock.h linux-4.4.62/include/linux/ static __always_inline void spin_unlock_bh(spinlock_t *lock) { raw_spin_unlock_bh(&lock->rlock); -@@ -416,4 +430,6 @@ +@@ -416,4 +430,6 @@ extern int _atomic_dec_and_lock(atomic_t *atomic, spinlock_t *lock); #define atomic_dec_and_lock(atomic, lock) \ __cond_lock(lock, _atomic_dec_and_lock(atomic, lock)) +#endif /* !PREEMPT_RT_FULL */ + #endif /* __LINUX_SPINLOCK_H */ -diff -Nur linux-4.4.62.orig/include/linux/spinlock_rt.h linux-4.4.62/include/linux/spinlock_rt.h ---- linux-4.4.62.orig/include/linux/spinlock_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/include/linux/spinlock_rt.h 2017-04-18 17:38:08.198649401 +0200 +diff --git a/include/linux/spinlock_api_smp.h b/include/linux/spinlock_api_smp.h +index 5344268e6e62..043263f30e81 100644 +--- a/include/linux/spinlock_api_smp.h ++++ b/include/linux/spinlock_api_smp.h +@@ -189,6 +189,8 @@ static inline int __raw_spin_trylock_bh(raw_spinlock_t *lock) + return 0; + } + +-#include ++#ifndef CONFIG_PREEMPT_RT_FULL ++# include ++#endif + + #endif /* __LINUX_SPINLOCK_API_SMP_H */ +diff --git a/include/linux/spinlock_rt.h b/include/linux/spinlock_rt.h +new file mode 100644 +index 000000000000..7eb87584e843 +--- /dev/null ++++ b/include/linux/spinlock_rt.h @@ -0,0 +1,165 @@ +#ifndef __LINUX_SPINLOCK_RT_H +#define __LINUX_SPINLOCK_RT_H @@ -15120,9 +15398,10 @@ diff -Nur linux-4.4.62.orig/include/linux/spinlock_rt.h linux-4.4.62/include/lin + atomic_dec_and_spin_lock(atomic, lock) + +#endif -diff -Nur linux-4.4.62.orig/include/linux/spinlock_types.h linux-4.4.62/include/linux/spinlock_types.h ---- linux-4.4.62.orig/include/linux/spinlock_types.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/spinlock_types.h 2017-04-18 17:38:08.198649401 +0200 +diff --git a/include/linux/spinlock_types.h b/include/linux/spinlock_types.h +index 73548eb13a5d..10bac715ea96 100644 +--- a/include/linux/spinlock_types.h ++++ b/include/linux/spinlock_types.h @@ -9,80 +9,15 @@ * Released under the General Public License (GPL). */ @@ -15211,9 +15490,11 @@ diff -Nur linux-4.4.62.orig/include/linux/spinlock_types.h linux-4.4.62/include/ -#include - #endif /* __LINUX_SPINLOCK_TYPES_H */ -diff -Nur linux-4.4.62.orig/include/linux/spinlock_types_nort.h linux-4.4.62/include/linux/spinlock_types_nort.h ---- linux-4.4.62.orig/include/linux/spinlock_types_nort.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/include/linux/spinlock_types_nort.h 2017-04-18 17:38:08.198649401 +0200 +diff --git a/include/linux/spinlock_types_nort.h b/include/linux/spinlock_types_nort.h +new file mode 100644 +index 000000000000..f1dac1fb1d6a +--- /dev/null ++++ b/include/linux/spinlock_types_nort.h @@ -0,0 +1,33 @@ +#ifndef __LINUX_SPINLOCK_TYPES_NORT_H +#define __LINUX_SPINLOCK_TYPES_NORT_H @@ -15248,9 +15529,11 @@ diff -Nur linux-4.4.62.orig/include/linux/spinlock_types_nort.h linux-4.4.62/inc +#define DEFINE_SPINLOCK(x) spinlock_t x = __SPIN_LOCK_UNLOCKED(x) + +#endif -diff -Nur linux-4.4.62.orig/include/linux/spinlock_types_raw.h linux-4.4.62/include/linux/spinlock_types_raw.h ---- linux-4.4.62.orig/include/linux/spinlock_types_raw.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/include/linux/spinlock_types_raw.h 2017-04-18 17:38:08.198649401 +0200 +diff --git a/include/linux/spinlock_types_raw.h b/include/linux/spinlock_types_raw.h +new file mode 100644 +index 000000000000..edffc4d53fc9 +--- /dev/null ++++ b/include/linux/spinlock_types_raw.h @@ -0,0 +1,56 @@ +#ifndef __LINUX_SPINLOCK_TYPES_RAW_H +#define __LINUX_SPINLOCK_TYPES_RAW_H @@ -15308,9 +15591,11 @@ diff -Nur linux-4.4.62.orig/include/linux/spinlock_types_raw.h linux-4.4.62/incl +#define DEFINE_RAW_SPINLOCK(x) raw_spinlock_t x = __RAW_SPIN_LOCK_UNLOCKED(x) + +#endif -diff -Nur linux-4.4.62.orig/include/linux/spinlock_types_rt.h linux-4.4.62/include/linux/spinlock_types_rt.h ---- linux-4.4.62.orig/include/linux/spinlock_types_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/include/linux/spinlock_types_rt.h 2017-04-18 17:38:08.198649401 +0200 +diff --git a/include/linux/spinlock_types_rt.h b/include/linux/spinlock_types_rt.h +new file mode 100644 +index 000000000000..9fd431967abc +--- /dev/null ++++ b/include/linux/spinlock_types_rt.h @@ -0,0 +1,51 @@ +#ifndef __LINUX_SPINLOCK_TYPES_RT_H +#define __LINUX_SPINLOCK_TYPES_RT_H @@ -15363,10 +15648,11 @@ diff -Nur linux-4.4.62.orig/include/linux/spinlock_types_rt.h linux-4.4.62/inclu + spinlock_t name __cacheline_aligned_in_smp = __SPIN_LOCK_UNLOCKED(name) + +#endif -diff -Nur linux-4.4.62.orig/include/linux/srcu.h linux-4.4.62/include/linux/srcu.h ---- linux-4.4.62.orig/include/linux/srcu.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/srcu.h 2017-04-18 17:38:08.198649401 +0200 -@@ -84,10 +84,10 @@ +diff --git a/include/linux/srcu.h b/include/linux/srcu.h +index f5f80c5643ac..ec1a8f01563c 100644 +--- a/include/linux/srcu.h ++++ b/include/linux/srcu.h +@@ -84,10 +84,10 @@ int init_srcu_struct(struct srcu_struct *sp); void process_srcu(struct work_struct *work); @@ -15379,7 +15665,7 @@ diff -Nur linux-4.4.62.orig/include/linux/srcu.h linux-4.4.62/include/linux/srcu .queue_lock = __SPIN_LOCK_UNLOCKED(name.queue_lock), \ .running = false, \ .batch_queue = RCU_BATCH_INIT(name.batch_queue), \ -@@ -104,7 +104,7 @@ +@@ -104,7 +104,7 @@ void process_srcu(struct work_struct *work); */ #define __DEFINE_SRCU(name, is_static) \ static DEFINE_PER_CPU(struct srcu_struct_array, name##_srcu_array);\ @@ -15388,10 +15674,11 @@ diff -Nur linux-4.4.62.orig/include/linux/srcu.h linux-4.4.62/include/linux/srcu #define DEFINE_SRCU(name) __DEFINE_SRCU(name, /* not static */) #define DEFINE_STATIC_SRCU(name) __DEFINE_SRCU(name, static) -diff -Nur linux-4.4.62.orig/include/linux/suspend.h linux-4.4.62/include/linux/suspend.h ---- linux-4.4.62.orig/include/linux/suspend.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/suspend.h 2017-04-18 17:38:08.198649401 +0200 -@@ -194,6 +194,12 @@ +diff --git a/include/linux/suspend.h b/include/linux/suspend.h +index 8b6ec7ef0854..9b77d4cc929f 100644 +--- a/include/linux/suspend.h ++++ b/include/linux/suspend.h +@@ -194,6 +194,12 @@ struct platform_freeze_ops { void (*end)(void); }; @@ -15404,9 +15691,11 @@ diff -Nur linux-4.4.62.orig/include/linux/suspend.h linux-4.4.62/include/linux/s #ifdef CONFIG_SUSPEND /** * suspend_set_ops - set platform dependent suspend operations -diff -Nur linux-4.4.62.orig/include/linux/swait.h linux-4.4.62/include/linux/swait.h ---- linux-4.4.62.orig/include/linux/swait.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/include/linux/swait.h 2017-04-18 17:38:08.198649401 +0200 +diff --git a/include/linux/swait.h b/include/linux/swait.h +new file mode 100644 +index 000000000000..83f004a72320 +--- /dev/null ++++ b/include/linux/swait.h @@ -0,0 +1,173 @@ +#ifndef _LINUX_SWAIT_H +#define _LINUX_SWAIT_H @@ -15581,9 +15870,10 @@ diff -Nur linux-4.4.62.orig/include/linux/swait.h linux-4.4.62/include/linux/swa +}) + +#endif /* _LINUX_SWAIT_H */ -diff -Nur linux-4.4.62.orig/include/linux/swap.h linux-4.4.62/include/linux/swap.h ---- linux-4.4.62.orig/include/linux/swap.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/swap.h 2017-04-18 17:38:08.198649401 +0200 +diff --git a/include/linux/swap.h b/include/linux/swap.h +index d8ca2eaa3a8b..19e038054914 100644 +--- a/include/linux/swap.h ++++ b/include/linux/swap.h @@ -11,6 +11,7 @@ #include #include @@ -15592,7 +15882,7 @@ diff -Nur linux-4.4.62.orig/include/linux/swap.h linux-4.4.62/include/linux/swap #include struct notifier_block; -@@ -252,7 +253,8 @@ +@@ -252,7 +253,8 @@ struct swap_info_struct { void *workingset_eviction(struct address_space *mapping, struct page *page); bool workingset_refault(void *shadow); void workingset_activation(struct page *page); @@ -15602,7 +15892,7 @@ diff -Nur linux-4.4.62.orig/include/linux/swap.h linux-4.4.62/include/linux/swap static inline unsigned int workingset_node_pages(struct radix_tree_node *node) { -@@ -298,6 +300,7 @@ +@@ -298,6 +300,7 @@ extern unsigned long nr_free_pagecache_pages(void); /* linux/mm/swap.c */ @@ -15610,9 +15900,11 @@ diff -Nur linux-4.4.62.orig/include/linux/swap.h linux-4.4.62/include/linux/swap extern void lru_cache_add(struct page *); extern void lru_cache_add_anon(struct page *page); extern void lru_cache_add_file(struct page *page); -diff -Nur linux-4.4.62.orig/include/linux/swork.h linux-4.4.62/include/linux/swork.h ---- linux-4.4.62.orig/include/linux/swork.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/include/linux/swork.h 2017-04-18 17:38:08.198649401 +0200 +diff --git a/include/linux/swork.h b/include/linux/swork.h +new file mode 100644 +index 000000000000..f175fa9a6016 +--- /dev/null ++++ b/include/linux/swork.h @@ -0,0 +1,24 @@ +#ifndef _LINUX_SWORK_H +#define _LINUX_SWORK_H @@ -15638,10 +15930,11 @@ diff -Nur linux-4.4.62.orig/include/linux/swork.h linux-4.4.62/include/linux/swo +void swork_put(void); + +#endif /* _LINUX_SWORK_H */ -diff -Nur linux-4.4.62.orig/include/linux/thread_info.h linux-4.4.62/include/linux/thread_info.h ---- linux-4.4.62.orig/include/linux/thread_info.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/thread_info.h 2017-04-18 17:38:08.198649401 +0200 -@@ -102,7 +102,17 @@ +diff --git a/include/linux/thread_info.h b/include/linux/thread_info.h +index ff307b548ed3..be9f9dc6a4e1 100644 +--- a/include/linux/thread_info.h ++++ b/include/linux/thread_info.h +@@ -102,7 +102,17 @@ static inline int test_ti_thread_flag(struct thread_info *ti, int flag) #define test_thread_flag(flag) \ test_ti_thread_flag(current_thread_info(), flag) @@ -15660,10 +15953,11 @@ diff -Nur linux-4.4.62.orig/include/linux/thread_info.h linux-4.4.62/include/lin #if defined TIF_RESTORE_SIGMASK && !defined HAVE_SET_RESTORE_SIGMASK /* -diff -Nur linux-4.4.62.orig/include/linux/timer.h linux-4.4.62/include/linux/timer.h ---- linux-4.4.62.orig/include/linux/timer.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/timer.h 2017-04-18 17:38:08.198649401 +0200 -@@ -225,7 +225,7 @@ +diff --git a/include/linux/timer.h b/include/linux/timer.h +index 61aa61dc410c..299d2b78591f 100644 +--- a/include/linux/timer.h ++++ b/include/linux/timer.h +@@ -225,7 +225,7 @@ extern void add_timer(struct timer_list *timer); extern int try_to_del_timer_sync(struct timer_list *timer); @@ -15672,10 +15966,11 @@ diff -Nur linux-4.4.62.orig/include/linux/timer.h linux-4.4.62/include/linux/tim extern int del_timer_sync(struct timer_list *timer); #else # define del_timer_sync(t) del_timer(t) -diff -Nur linux-4.4.62.orig/include/linux/trace_events.h linux-4.4.62/include/linux/trace_events.h ---- linux-4.4.62.orig/include/linux/trace_events.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/trace_events.h 2017-04-18 17:38:08.198649401 +0200 -@@ -66,6 +66,9 @@ +diff --git a/include/linux/trace_events.h b/include/linux/trace_events.h +index 925730bc9fc1..a591f414da6f 100644 +--- a/include/linux/trace_events.h ++++ b/include/linux/trace_events.h +@@ -66,6 +66,9 @@ struct trace_entry { unsigned char flags; unsigned char preempt_count; int pid; @@ -15685,10 +15980,11 @@ diff -Nur linux-4.4.62.orig/include/linux/trace_events.h linux-4.4.62/include/li }; #define TRACE_EVENT_TYPE_MAX \ -diff -Nur linux-4.4.62.orig/include/linux/uaccess.h linux-4.4.62/include/linux/uaccess.h ---- linux-4.4.62.orig/include/linux/uaccess.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/uaccess.h 2017-04-18 17:38:08.198649401 +0200 -@@ -24,6 +24,7 @@ +diff --git a/include/linux/uaccess.h b/include/linux/uaccess.h +index 558129af828a..cf5c472bbc79 100644 +--- a/include/linux/uaccess.h ++++ b/include/linux/uaccess.h +@@ -24,6 +24,7 @@ static __always_inline void pagefault_disabled_dec(void) */ static inline void pagefault_disable(void) { @@ -15696,7 +15992,7 @@ diff -Nur linux-4.4.62.orig/include/linux/uaccess.h linux-4.4.62/include/linux/u pagefault_disabled_inc(); /* * make sure to have issued the store before a pagefault -@@ -40,6 +41,7 @@ +@@ -40,6 +41,7 @@ static inline void pagefault_enable(void) */ barrier(); pagefault_disabled_dec(); @@ -15704,9 +16000,10 @@ diff -Nur linux-4.4.62.orig/include/linux/uaccess.h linux-4.4.62/include/linux/u } /* -diff -Nur linux-4.4.62.orig/include/linux/uprobes.h linux-4.4.62/include/linux/uprobes.h ---- linux-4.4.62.orig/include/linux/uprobes.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/uprobes.h 2017-04-18 17:38:08.198649401 +0200 +diff --git a/include/linux/uprobes.h b/include/linux/uprobes.h +index 4a29c75b146e..0a294e950df8 100644 +--- a/include/linux/uprobes.h ++++ b/include/linux/uprobes.h @@ -27,6 +27,7 @@ #include #include @@ -15715,10 +16012,11 @@ diff -Nur linux-4.4.62.orig/include/linux/uprobes.h linux-4.4.62/include/linux/u struct vm_area_struct; struct mm_struct; -diff -Nur linux-4.4.62.orig/include/linux/vmstat.h linux-4.4.62/include/linux/vmstat.h ---- linux-4.4.62.orig/include/linux/vmstat.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/vmstat.h 2017-04-18 17:38:08.202649555 +0200 -@@ -33,7 +33,9 @@ +diff --git a/include/linux/vmstat.h b/include/linux/vmstat.h +index 3e5d9075960f..7eaa847cd5a5 100644 +--- a/include/linux/vmstat.h ++++ b/include/linux/vmstat.h +@@ -33,7 +33,9 @@ DECLARE_PER_CPU(struct vm_event_state, vm_event_states); */ static inline void __count_vm_event(enum vm_event_item item) { @@ -15728,7 +16026,7 @@ diff -Nur linux-4.4.62.orig/include/linux/vmstat.h linux-4.4.62/include/linux/vm } static inline void count_vm_event(enum vm_event_item item) -@@ -43,7 +45,9 @@ +@@ -43,7 +45,9 @@ static inline void count_vm_event(enum vm_event_item item) static inline void __count_vm_events(enum vm_event_item item, long delta) { @@ -15738,9 +16036,10 @@ diff -Nur linux-4.4.62.orig/include/linux/vmstat.h linux-4.4.62/include/linux/vm } static inline void count_vm_events(enum vm_event_item item, long delta) -diff -Nur linux-4.4.62.orig/include/linux/wait.h linux-4.4.62/include/linux/wait.h ---- linux-4.4.62.orig/include/linux/wait.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/linux/wait.h 2017-04-18 17:38:08.202649555 +0200 +diff --git a/include/linux/wait.h b/include/linux/wait.h +index 513b36f04dfd..981c8a840f96 100644 +--- a/include/linux/wait.h ++++ b/include/linux/wait.h @@ -8,6 +8,7 @@ #include #include @@ -15749,10 +16048,11 @@ diff -Nur linux-4.4.62.orig/include/linux/wait.h linux-4.4.62/include/linux/wait typedef struct __wait_queue wait_queue_t; typedef int (*wait_queue_func_t)(wait_queue_t *wait, unsigned mode, int flags, void *key); -diff -Nur linux-4.4.62.orig/include/net/dst.h linux-4.4.62/include/net/dst.h ---- linux-4.4.62.orig/include/net/dst.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/net/dst.h 2017-04-18 17:38:08.202649555 +0200 -@@ -437,7 +437,7 @@ +diff --git a/include/net/dst.h b/include/net/dst.h +index c7329dcd90cc..35c3dba16728 100644 +--- a/include/net/dst.h ++++ b/include/net/dst.h +@@ -437,7 +437,7 @@ static inline void dst_confirm(struct dst_entry *dst) static inline int dst_neigh_output(struct dst_entry *dst, struct neighbour *n, struct sk_buff *skb) { @@ -15761,10 +16061,11 @@ diff -Nur linux-4.4.62.orig/include/net/dst.h linux-4.4.62/include/net/dst.h if (dst->pending_confirm) { unsigned long now = jiffies; -diff -Nur linux-4.4.62.orig/include/net/neighbour.h linux-4.4.62/include/net/neighbour.h ---- linux-4.4.62.orig/include/net/neighbour.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/net/neighbour.h 2017-04-18 17:38:08.202649555 +0200 -@@ -446,7 +446,7 @@ +diff --git a/include/net/neighbour.h b/include/net/neighbour.h +index 8b683841e574..bf656008f6e7 100644 +--- a/include/net/neighbour.h ++++ b/include/net/neighbour.h +@@ -446,7 +446,7 @@ static inline int neigh_hh_bridge(struct hh_cache *hh, struct sk_buff *skb) } #endif @@ -15773,7 +16074,7 @@ diff -Nur linux-4.4.62.orig/include/net/neighbour.h linux-4.4.62/include/net/nei { unsigned int seq; int hh_len; -@@ -501,7 +501,7 @@ +@@ -501,7 +501,7 @@ struct neighbour_cb { #define NEIGH_CB(skb) ((struct neighbour_cb *)(skb)->cb) @@ -15782,10 +16083,11 @@ diff -Nur linux-4.4.62.orig/include/net/neighbour.h linux-4.4.62/include/net/nei const struct net_device *dev) { unsigned int seq; -diff -Nur linux-4.4.62.orig/include/net/netns/ipv4.h linux-4.4.62/include/net/netns/ipv4.h ---- linux-4.4.62.orig/include/net/netns/ipv4.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/net/netns/ipv4.h 2017-04-18 17:38:08.202649555 +0200 -@@ -70,6 +70,7 @@ +diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h +index c68926b4899c..dd0751e76065 100644 +--- a/include/net/netns/ipv4.h ++++ b/include/net/netns/ipv4.h +@@ -70,6 +70,7 @@ struct netns_ipv4 { int sysctl_icmp_echo_ignore_all; int sysctl_icmp_echo_ignore_broadcasts; @@ -15793,9 +16095,11 @@ diff -Nur linux-4.4.62.orig/include/net/netns/ipv4.h linux-4.4.62/include/net/ne int sysctl_icmp_ignore_bogus_error_responses; int sysctl_icmp_ratelimit; int sysctl_icmp_ratemask; -diff -Nur linux-4.4.62.orig/include/trace/events/hist.h linux-4.4.62/include/trace/events/hist.h ---- linux-4.4.62.orig/include/trace/events/hist.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/include/trace/events/hist.h 2017-04-18 17:38:08.202649555 +0200 +diff --git a/include/trace/events/hist.h b/include/trace/events/hist.h +new file mode 100644 +index 000000000000..f7710de1b1f3 +--- /dev/null ++++ b/include/trace/events/hist.h @@ -0,0 +1,73 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM hist @@ -15870,9 +16174,11 @@ diff -Nur linux-4.4.62.orig/include/trace/events/hist.h linux-4.4.62/include/tra + +/* This part must be outside protection */ +#include -diff -Nur linux-4.4.62.orig/include/trace/events/latency_hist.h linux-4.4.62/include/trace/events/latency_hist.h ---- linux-4.4.62.orig/include/trace/events/latency_hist.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/include/trace/events/latency_hist.h 2017-04-18 17:38:08.202649555 +0200 +diff --git a/include/trace/events/latency_hist.h b/include/trace/events/latency_hist.h +new file mode 100644 +index 000000000000..d3f2fbd560b1 +--- /dev/null ++++ b/include/trace/events/latency_hist.h @@ -0,0 +1,29 @@ +#ifndef _LATENCY_HIST_H +#define _LATENCY_HIST_H @@ -15903,10 +16209,11 @@ diff -Nur linux-4.4.62.orig/include/trace/events/latency_hist.h linux-4.4.62/inc +} + +#endif /* _LATENCY_HIST_H */ -diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/include/trace/events/writeback.h ---- linux-4.4.62.orig/include/trace/events/writeback.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/include/trace/events/writeback.h 2017-04-18 17:38:08.202649555 +0200 -@@ -134,58 +134,28 @@ +diff --git a/include/trace/events/writeback.h b/include/trace/events/writeback.h +index fff846b512e6..73614ce1d204 100644 +--- a/include/trace/events/writeback.h ++++ b/include/trace/events/writeback.h +@@ -134,58 +134,28 @@ DEFINE_EVENT(writeback_dirty_inode_template, writeback_dirty_inode, #ifdef CREATE_TRACE_POINTS #ifdef CONFIG_CGROUP_WRITEBACK @@ -15974,7 +16281,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ } #endif /* CONFIG_CGROUP_WRITEBACK */ -@@ -201,7 +171,7 @@ +@@ -201,7 +171,7 @@ DECLARE_EVENT_CLASS(writeback_write_inode_template, __array(char, name, 32) __field(unsigned long, ino) __field(int, sync_mode) @@ -15983,7 +16290,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ), TP_fast_assign( -@@ -209,14 +179,14 @@ +@@ -209,14 +179,14 @@ DECLARE_EVENT_CLASS(writeback_write_inode_template, dev_name(inode_to_bdi(inode)->dev), 32); __entry->ino = inode->i_ino; __entry->sync_mode = wbc->sync_mode; @@ -16001,7 +16308,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ) ); -@@ -246,7 +216,7 @@ +@@ -246,7 +216,7 @@ DECLARE_EVENT_CLASS(writeback_work_class, __field(int, range_cyclic) __field(int, for_background) __field(int, reason) @@ -16010,7 +16317,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ), TP_fast_assign( strncpy(__entry->name, -@@ -258,10 +228,10 @@ +@@ -258,10 +228,10 @@ DECLARE_EVENT_CLASS(writeback_work_class, __entry->range_cyclic = work->range_cyclic; __entry->for_background = work->for_background; __entry->reason = work->reason; @@ -16023,7 +16330,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ __entry->name, MAJOR(__entry->sb_dev), MINOR(__entry->sb_dev), __entry->nr_pages, -@@ -270,7 +240,7 @@ +@@ -270,7 +240,7 @@ DECLARE_EVENT_CLASS(writeback_work_class, __entry->range_cyclic, __entry->for_background, __print_symbolic(__entry->reason, WB_WORK_REASON), @@ -16032,7 +16339,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ) ); #define DEFINE_WRITEBACK_WORK_EVENT(name) \ -@@ -300,15 +270,15 @@ +@@ -300,15 +270,15 @@ DECLARE_EVENT_CLASS(writeback_class, TP_ARGS(wb), TP_STRUCT__entry( __array(char, name, 32) @@ -16052,7 +16359,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ) ); #define DEFINE_WRITEBACK_EVENT(name) \ -@@ -347,7 +317,7 @@ +@@ -347,7 +317,7 @@ DECLARE_EVENT_CLASS(wbc_class, __field(int, range_cyclic) __field(long, range_start) __field(long, range_end) @@ -16061,7 +16368,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ), TP_fast_assign( -@@ -361,12 +331,12 @@ +@@ -361,12 +331,12 @@ DECLARE_EVENT_CLASS(wbc_class, __entry->range_cyclic = wbc->range_cyclic; __entry->range_start = (long)wbc->range_start; __entry->range_end = (long)wbc->range_end; @@ -16076,7 +16383,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ __entry->name, __entry->nr_to_write, __entry->pages_skipped, -@@ -377,7 +347,7 @@ +@@ -377,7 +347,7 @@ DECLARE_EVENT_CLASS(wbc_class, __entry->range_cyclic, __entry->range_start, __entry->range_end, @@ -16085,7 +16392,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ) ) -@@ -398,7 +368,7 @@ +@@ -398,7 +368,7 @@ TRACE_EVENT(writeback_queue_io, __field(long, age) __field(int, moved) __field(int, reason) @@ -16094,7 +16401,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ), TP_fast_assign( unsigned long *older_than_this = work->older_than_this; -@@ -408,15 +378,15 @@ +@@ -408,15 +378,15 @@ TRACE_EVENT(writeback_queue_io, (jiffies - *older_than_this) * 1000 / HZ : -1; __entry->moved = moved; __entry->reason = work->reason; @@ -16113,7 +16420,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ) ); -@@ -484,7 +454,7 @@ +@@ -484,7 +454,7 @@ TRACE_EVENT(bdi_dirty_ratelimit, __field(unsigned long, dirty_ratelimit) __field(unsigned long, task_ratelimit) __field(unsigned long, balanced_dirty_ratelimit) @@ -16122,7 +16429,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ), TP_fast_assign( -@@ -496,13 +466,13 @@ +@@ -496,13 +466,13 @@ TRACE_EVENT(bdi_dirty_ratelimit, __entry->task_ratelimit = KBps(task_ratelimit); __entry->balanced_dirty_ratelimit = KBps(wb->balanced_dirty_ratelimit); @@ -16138,7 +16445,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ __entry->bdi, __entry->write_bw, /* write bandwidth */ __entry->avg_write_bw, /* avg write bandwidth */ -@@ -510,7 +480,7 @@ +@@ -510,7 +480,7 @@ TRACE_EVENT(bdi_dirty_ratelimit, __entry->dirty_ratelimit, /* base ratelimit */ __entry->task_ratelimit, /* ratelimit with position control */ __entry->balanced_dirty_ratelimit, /* the balanced ratelimit */ @@ -16147,7 +16454,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ) ); -@@ -548,7 +518,7 @@ +@@ -548,7 +518,7 @@ TRACE_EVENT(balance_dirty_pages, __field( long, pause) __field(unsigned long, period) __field( long, think) @@ -16156,7 +16463,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ), TP_fast_assign( -@@ -571,7 +541,7 @@ +@@ -571,7 +541,7 @@ TRACE_EVENT(balance_dirty_pages, __entry->period = period * 1000 / HZ; __entry->pause = pause * 1000 / HZ; __entry->paused = (jiffies - start_time) * 1000 / HZ; @@ -16165,7 +16472,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ), -@@ -580,7 +550,7 @@ +@@ -580,7 +550,7 @@ TRACE_EVENT(balance_dirty_pages, "bdi_setpoint=%lu bdi_dirty=%lu " "dirty_ratelimit=%lu task_ratelimit=%lu " "dirtied=%u dirtied_pause=%u " @@ -16174,7 +16481,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ __entry->bdi, __entry->limit, __entry->setpoint, -@@ -595,7 +565,7 @@ +@@ -595,7 +565,7 @@ TRACE_EVENT(balance_dirty_pages, __entry->pause, /* ms */ __entry->period, /* ms */ __entry->think, /* ms */ @@ -16183,7 +16490,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ) ); -@@ -609,8 +579,7 @@ +@@ -609,8 +579,7 @@ TRACE_EVENT(writeback_sb_inodes_requeue, __field(unsigned long, ino) __field(unsigned long, state) __field(unsigned long, dirtied_when) @@ -16193,7 +16500,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ), TP_fast_assign( -@@ -619,16 +588,16 @@ +@@ -619,16 +588,16 @@ TRACE_EVENT(writeback_sb_inodes_requeue, __entry->ino = inode->i_ino; __entry->state = inode->i_state; __entry->dirtied_when = inode->dirtied_when; @@ -16213,7 +16520,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ) ); -@@ -684,7 +653,7 @@ +@@ -684,7 +653,7 @@ DECLARE_EVENT_CLASS(writeback_single_inode_template, __field(unsigned long, writeback_index) __field(long, nr_to_write) __field(unsigned long, wrote) @@ -16222,7 +16529,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ), TP_fast_assign( -@@ -696,11 +665,11 @@ +@@ -696,11 +665,11 @@ DECLARE_EVENT_CLASS(writeback_single_inode_template, __entry->writeback_index = inode->i_mapping->writeback_index; __entry->nr_to_write = nr_to_write; __entry->wrote = nr_to_write - wbc->nr_to_write; @@ -16236,7 +16543,7 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ __entry->name, __entry->ino, show_inode_state(__entry->state), -@@ -709,7 +678,7 @@ +@@ -709,7 +678,7 @@ DECLARE_EVENT_CLASS(writeback_single_inode_template, __entry->writeback_index, __entry->nr_to_write, __entry->wrote, @@ -16245,10 +16552,11 @@ diff -Nur linux-4.4.62.orig/include/trace/events/writeback.h linux-4.4.62/includ ) ); -diff -Nur linux-4.4.62.orig/init/Kconfig linux-4.4.62/init/Kconfig ---- linux-4.4.62.orig/init/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/init/Kconfig 2017-04-18 17:38:08.202649555 +0200 -@@ -498,7 +498,7 @@ +diff --git a/init/Kconfig b/init/Kconfig +index 235c7a2c0d20..a7c81c0911da 100644 +--- a/init/Kconfig ++++ b/init/Kconfig +@@ -498,7 +498,7 @@ config TINY_RCU config RCU_EXPERT bool "Make expert-level adjustments to RCU configuration" @@ -16257,7 +16565,7 @@ diff -Nur linux-4.4.62.orig/init/Kconfig linux-4.4.62/init/Kconfig help This option needs to be enabled if you wish to make expert-level adjustments to RCU configuration. By default, -@@ -614,7 +614,7 @@ +@@ -614,7 +614,7 @@ config RCU_FANOUT_LEAF config RCU_FAST_NO_HZ bool "Accelerate last non-dyntick-idle CPU's grace periods" @@ -16266,7 +16574,7 @@ diff -Nur linux-4.4.62.orig/init/Kconfig linux-4.4.62/init/Kconfig default n help This option permits CPUs to enter dynticks-idle state even if -@@ -641,7 +641,7 @@ +@@ -641,7 +641,7 @@ config TREE_RCU_TRACE config RCU_BOOST bool "Enable RCU priority boosting" depends on RT_MUTEXES && PREEMPT_RCU && RCU_EXPERT @@ -16275,7 +16583,7 @@ diff -Nur linux-4.4.62.orig/init/Kconfig linux-4.4.62/init/Kconfig help This option boosts the priority of preempted RCU readers that block the current preemptible RCU grace period for too long. -@@ -1106,6 +1106,7 @@ +@@ -1106,6 +1106,7 @@ config CFS_BANDWIDTH config RT_GROUP_SCHED bool "Group scheduling for SCHED_RR/FIFO" depends on CGROUP_SCHED @@ -16283,7 +16591,7 @@ diff -Nur linux-4.4.62.orig/init/Kconfig linux-4.4.62/init/Kconfig default n help This feature lets you explicitly allocate real CPU bandwidth -@@ -1719,6 +1720,7 @@ +@@ -1719,6 +1720,7 @@ choice config SLAB bool "SLAB" @@ -16291,7 +16599,7 @@ diff -Nur linux-4.4.62.orig/init/Kconfig linux-4.4.62/init/Kconfig help The regular slab allocator that is established and known to work well in all environments. It organizes cache hot objects in -@@ -1737,6 +1739,7 @@ +@@ -1737,6 +1739,7 @@ config SLUB config SLOB depends on EXPERT bool "SLOB (Simple Allocator)" @@ -16299,7 +16607,7 @@ diff -Nur linux-4.4.62.orig/init/Kconfig linux-4.4.62/init/Kconfig help SLOB replaces the stock allocator with a drastically simpler allocator. SLOB is generally more space efficient but -@@ -1746,7 +1749,7 @@ +@@ -1746,7 +1749,7 @@ endchoice config SLUB_CPU_PARTIAL default y @@ -16308,10 +16616,21 @@ diff -Nur linux-4.4.62.orig/init/Kconfig linux-4.4.62/init/Kconfig bool "SLUB per cpu partial cache" help Per cpu partial caches accellerate objects allocation and freeing -diff -Nur linux-4.4.62.orig/init/main.c linux-4.4.62/init/main.c ---- linux-4.4.62.orig/init/main.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/init/main.c 2017-04-18 17:38:08.202649555 +0200 -@@ -530,6 +530,7 @@ +diff --git a/init/Makefile b/init/Makefile +index 7bc47ee31c36..88cf473554e0 100644 +--- a/init/Makefile ++++ b/init/Makefile +@@ -33,4 +33,4 @@ silent_chk_compile.h = : + include/generated/compile.h: FORCE + @$($(quiet)chk_compile.h) + $(Q)$(CONFIG_SHELL) $(srctree)/scripts/mkcompile_h $@ \ +- "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CC) $(KBUILD_CFLAGS)" ++ "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CONFIG_PREEMPT_RT_FULL)" "$(CC) $(KBUILD_CFLAGS)" +diff --git a/init/main.c b/init/main.c +index 9e64d7097f1a..4a76e629c137 100644 +--- a/init/main.c ++++ b/init/main.c +@@ -530,6 +530,7 @@ asmlinkage __visible void __init start_kernel(void) setup_command_line(command_line); setup_nr_cpu_ids(); setup_per_cpu_areas(); @@ -16319,19 +16638,11 @@ diff -Nur linux-4.4.62.orig/init/main.c linux-4.4.62/init/main.c smp_prepare_boot_cpu(); /* arch-specific boot-cpu hooks */ build_all_zonelists(NULL, NULL); -diff -Nur linux-4.4.62.orig/init/Makefile linux-4.4.62/init/Makefile ---- linux-4.4.62.orig/init/Makefile 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/init/Makefile 2017-04-18 17:38:08.202649555 +0200 -@@ -33,4 +33,4 @@ - include/generated/compile.h: FORCE - @$($(quiet)chk_compile.h) - $(Q)$(CONFIG_SHELL) $(srctree)/scripts/mkcompile_h $@ \ -- "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CC) $(KBUILD_CFLAGS)" -+ "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CONFIG_PREEMPT_RT_FULL)" "$(CC) $(KBUILD_CFLAGS)" -diff -Nur linux-4.4.62.orig/ipc/msg.c linux-4.4.62/ipc/msg.c ---- linux-4.4.62.orig/ipc/msg.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/ipc/msg.c 2017-04-18 17:38:08.202649555 +0200 -@@ -183,20 +183,14 @@ +diff --git a/ipc/msg.c b/ipc/msg.c +index c6521c205cb4..996d89023552 100644 +--- a/ipc/msg.c ++++ b/ipc/msg.c +@@ -183,20 +183,14 @@ static void ss_wakeup(struct list_head *h, int kill) } } @@ -16356,7 +16667,7 @@ diff -Nur linux-4.4.62.orig/ipc/msg.c linux-4.4.62/ipc/msg.c msr->r_msg = ERR_PTR(res); } } -@@ -213,11 +207,13 @@ +@@ -213,11 +207,13 @@ static void freeque(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp) { struct msg_msg *msg, *t; struct msg_queue *msq = container_of(ipcp, struct msg_queue, q_perm); @@ -16371,7 +16682,7 @@ diff -Nur linux-4.4.62.orig/ipc/msg.c linux-4.4.62/ipc/msg.c rcu_read_unlock(); list_for_each_entry_safe(msg, t, &msq->q_messages, m_list) { -@@ -342,6 +338,7 @@ +@@ -342,6 +338,7 @@ static int msgctl_down(struct ipc_namespace *ns, int msqid, int cmd, struct kern_ipc_perm *ipcp; struct msqid64_ds uninitialized_var(msqid64); struct msg_queue *msq; @@ -16379,7 +16690,7 @@ diff -Nur linux-4.4.62.orig/ipc/msg.c linux-4.4.62/ipc/msg.c int err; if (cmd == IPC_SET) { -@@ -389,7 +386,7 @@ +@@ -389,7 +386,7 @@ static int msgctl_down(struct ipc_namespace *ns, int msqid, int cmd, /* sleeping receivers might be excluded by * stricter permissions. */ @@ -16388,7 +16699,7 @@ diff -Nur linux-4.4.62.orig/ipc/msg.c linux-4.4.62/ipc/msg.c /* sleeping senders might be able to send * due to a larger queue size. */ -@@ -402,6 +399,7 @@ +@@ -402,6 +399,7 @@ static int msgctl_down(struct ipc_namespace *ns, int msqid, int cmd, out_unlock0: ipc_unlock_object(&msq->q_perm); @@ -16396,7 +16707,7 @@ diff -Nur linux-4.4.62.orig/ipc/msg.c linux-4.4.62/ipc/msg.c out_unlock1: rcu_read_unlock(); out_up: -@@ -566,7 +564,8 @@ +@@ -566,7 +564,8 @@ static int testmsg(struct msg_msg *msg, long type, int mode) return 0; } @@ -16406,7 +16717,7 @@ diff -Nur linux-4.4.62.orig/ipc/msg.c linux-4.4.62/ipc/msg.c { struct msg_receiver *msr, *t; -@@ -577,27 +576,13 @@ +@@ -577,27 +576,13 @@ static inline int pipelined_send(struct msg_queue *msq, struct msg_msg *msg) list_del(&msr->r_list); if (msr->r_maxsize < msg->m_ts) { @@ -16436,7 +16747,7 @@ diff -Nur linux-4.4.62.orig/ipc/msg.c linux-4.4.62/ipc/msg.c return 1; } } -@@ -613,6 +598,7 @@ +@@ -613,6 +598,7 @@ long do_msgsnd(int msqid, long mtype, void __user *mtext, struct msg_msg *msg; int err; struct ipc_namespace *ns; @@ -16444,7 +16755,7 @@ diff -Nur linux-4.4.62.orig/ipc/msg.c linux-4.4.62/ipc/msg.c ns = current->nsproxy->ipc_ns; -@@ -698,7 +684,7 @@ +@@ -698,7 +684,7 @@ long do_msgsnd(int msqid, long mtype, void __user *mtext, msq->q_lspid = task_tgid_vnr(current); msq->q_stime = get_seconds(); @@ -16453,7 +16764,7 @@ diff -Nur linux-4.4.62.orig/ipc/msg.c linux-4.4.62/ipc/msg.c /* no one is waiting for this message, enqueue it */ list_add_tail(&msg->m_list, &msq->q_messages); msq->q_cbytes += msgsz; -@@ -712,6 +698,7 @@ +@@ -712,6 +698,7 @@ long do_msgsnd(int msqid, long mtype, void __user *mtext, out_unlock0: ipc_unlock_object(&msq->q_perm); @@ -16461,7 +16772,7 @@ diff -Nur linux-4.4.62.orig/ipc/msg.c linux-4.4.62/ipc/msg.c out_unlock1: rcu_read_unlock(); if (msg != NULL) -@@ -932,57 +919,25 @@ +@@ -932,57 +919,25 @@ long do_msgrcv(int msqid, void __user *buf, size_t bufsz, long msgtyp, int msgfl rcu_read_lock(); /* Lockless receive, part 2: @@ -16530,10 +16841,11 @@ diff -Nur linux-4.4.62.orig/ipc/msg.c linux-4.4.62/ipc/msg.c msg = (struct msg_msg *)msr_d.r_msg; if (msg != ERR_PTR(-EAGAIN)) goto out_unlock0; -diff -Nur linux-4.4.62.orig/ipc/sem.c linux-4.4.62/ipc/sem.c ---- linux-4.4.62.orig/ipc/sem.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/ipc/sem.c 2017-04-18 17:38:08.202649555 +0200 -@@ -708,6 +708,13 @@ +diff --git a/ipc/sem.c b/ipc/sem.c +index 9862c3d1c26d..ef34d7376697 100644 +--- a/ipc/sem.c ++++ b/ipc/sem.c +@@ -708,6 +708,13 @@ undo: static void wake_up_sem_queue_prepare(struct list_head *pt, struct sem_queue *q, int error) { @@ -16547,7 +16859,7 @@ diff -Nur linux-4.4.62.orig/ipc/sem.c linux-4.4.62/ipc/sem.c if (list_empty(pt)) { /* * Hold preempt off so that we don't get preempted and have the -@@ -719,6 +726,7 @@ +@@ -719,6 +726,7 @@ static void wake_up_sem_queue_prepare(struct list_head *pt, q->pid = error; list_add_tail(&q->list, pt); @@ -16555,7 +16867,7 @@ diff -Nur linux-4.4.62.orig/ipc/sem.c linux-4.4.62/ipc/sem.c } /** -@@ -732,6 +740,7 @@ +@@ -732,6 +740,7 @@ static void wake_up_sem_queue_prepare(struct list_head *pt, */ static void wake_up_sem_queue_do(struct list_head *pt) { @@ -16563,7 +16875,7 @@ diff -Nur linux-4.4.62.orig/ipc/sem.c linux-4.4.62/ipc/sem.c struct sem_queue *q, *t; int did_something; -@@ -744,6 +753,7 @@ +@@ -744,6 +753,7 @@ static void wake_up_sem_queue_do(struct list_head *pt) } if (did_something) preempt_enable(); @@ -16571,10 +16883,85 @@ diff -Nur linux-4.4.62.orig/ipc/sem.c linux-4.4.62/ipc/sem.c } static void unlink_queue(struct sem_array *sma, struct sem_queue *q) -diff -Nur linux-4.4.62.orig/kernel/cgroup.c linux-4.4.62/kernel/cgroup.c ---- linux-4.4.62.orig/kernel/cgroup.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/cgroup.c 2017-04-18 17:38:08.202649555 +0200 -@@ -4740,10 +4740,10 @@ +diff --git a/kernel/Kconfig.locks b/kernel/Kconfig.locks +index ebdb0043203a..b9e6aa7e5aa6 100644 +--- a/kernel/Kconfig.locks ++++ b/kernel/Kconfig.locks +@@ -225,11 +225,11 @@ config ARCH_SUPPORTS_ATOMIC_RMW + + config MUTEX_SPIN_ON_OWNER + def_bool y +- depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW ++ depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL + + config RWSEM_SPIN_ON_OWNER + def_bool y +- depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW ++ depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL + + config LOCK_SPIN_ON_OWNER + def_bool y +diff --git a/kernel/Kconfig.preempt b/kernel/Kconfig.preempt +index 3f9c97419f02..11dbe26a8279 100644 +--- a/kernel/Kconfig.preempt ++++ b/kernel/Kconfig.preempt +@@ -1,3 +1,16 @@ ++config PREEMPT ++ bool ++ select PREEMPT_COUNT ++ ++config PREEMPT_RT_BASE ++ bool ++ select PREEMPT ++ ++config HAVE_PREEMPT_LAZY ++ bool ++ ++config PREEMPT_LAZY ++ def_bool y if HAVE_PREEMPT_LAZY && PREEMPT_RT_FULL + + choice + prompt "Preemption Model" +@@ -33,9 +46,9 @@ config PREEMPT_VOLUNTARY + + Select this if you are building a kernel for a desktop system. + +-config PREEMPT ++config PREEMPT__LL + bool "Preemptible Kernel (Low-Latency Desktop)" +- select PREEMPT_COUNT ++ select PREEMPT + select UNINLINE_SPIN_UNLOCK if !ARCH_INLINE_SPIN_UNLOCK + help + This option reduces the latency of the kernel by making +@@ -52,6 +65,22 @@ config PREEMPT + embedded system with latency requirements in the milliseconds + range. + ++config PREEMPT_RTB ++ bool "Preemptible Kernel (Basic RT)" ++ select PREEMPT_RT_BASE ++ help ++ This option is basically the same as (Low-Latency Desktop) but ++ enables changes which are preliminary for the full preemptible ++ RT kernel. ++ ++config PREEMPT_RT_FULL ++ bool "Fully Preemptible Kernel (RT)" ++ depends on IRQ_FORCED_THREADING ++ select PREEMPT_RT_BASE ++ select PREEMPT_RCU ++ help ++ All and everything ++ + endchoice + + config PREEMPT_COUNT +diff --git a/kernel/cgroup.c b/kernel/cgroup.c +index 4cb94b678e9f..8c41ee8a6fee 100644 +--- a/kernel/cgroup.c ++++ b/kernel/cgroup.c +@@ -4741,10 +4741,10 @@ static void css_free_rcu_fn(struct rcu_head *rcu_head) queue_work(cgroup_destroy_wq, &css->destroy_work); } @@ -16587,7 +16974,7 @@ diff -Nur linux-4.4.62.orig/kernel/cgroup.c linux-4.4.62/kernel/cgroup.c struct cgroup_subsys *ss = css->ss; struct cgroup *cgrp = css->cgroup; -@@ -4782,8 +4782,8 @@ +@@ -4783,8 +4783,8 @@ static void css_release(struct percpu_ref *ref) struct cgroup_subsys_state *css = container_of(ref, struct cgroup_subsys_state, refcnt); @@ -16598,7 +16985,7 @@ diff -Nur linux-4.4.62.orig/kernel/cgroup.c linux-4.4.62/kernel/cgroup.c } static void init_and_link_css(struct cgroup_subsys_state *css, -@@ -5400,6 +5400,7 @@ +@@ -5401,6 +5401,7 @@ static int __init cgroup_wq_init(void) */ cgroup_destroy_wq = alloc_workqueue("cgroup_destroy", 0, 1); BUG_ON(!cgroup_destroy_wq); @@ -16606,10 +16993,11 @@ diff -Nur linux-4.4.62.orig/kernel/cgroup.c linux-4.4.62/kernel/cgroup.c /* * Used to destroy pidlists and separate to serve as flush domain. -diff -Nur linux-4.4.62.orig/kernel/cpu.c linux-4.4.62/kernel/cpu.c ---- linux-4.4.62.orig/kernel/cpu.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/cpu.c 2017-04-18 17:38:08.202649555 +0200 -@@ -75,8 +75,8 @@ +diff --git a/kernel/cpu.c b/kernel/cpu.c +index 40d20bf5de28..0be18c1684d8 100644 +--- a/kernel/cpu.c ++++ b/kernel/cpu.c +@@ -75,8 +75,8 @@ static struct { #endif } cpu_hotplug = { .active_writer = NULL, @@ -16619,7 +17007,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpu.c linux-4.4.62/kernel/cpu.c #ifdef CONFIG_DEBUG_LOCK_ALLOC .dep_map = {.name = "cpu_hotplug.lock" }, #endif -@@ -89,6 +89,289 @@ +@@ -89,6 +89,289 @@ static struct { #define cpuhp_lock_acquire() lock_map_acquire(&cpu_hotplug.dep_map) #define cpuhp_lock_release() lock_map_release(&cpu_hotplug.dep_map) @@ -16909,7 +17297,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpu.c linux-4.4.62/kernel/cpu.c void get_online_cpus(void) { -@@ -338,13 +621,15 @@ +@@ -338,13 +621,15 @@ static int take_cpu_down(void *_param) /* Requires cpu_add_remove_lock to be held */ static int _cpu_down(unsigned int cpu, int tasks_frozen) { @@ -16926,7 +17314,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpu.c linux-4.4.62/kernel/cpu.c if (num_online_cpus() == 1) return -EBUSY; -@@ -352,7 +637,34 @@ +@@ -352,7 +637,34 @@ static int _cpu_down(unsigned int cpu, int tasks_frozen) if (!cpu_online(cpu)) return -EINVAL; @@ -16961,7 +17349,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpu.c linux-4.4.62/kernel/cpu.c err = __cpu_notify(CPU_DOWN_PREPARE | mod, hcpu, -1, &nr_calls); if (err) { -@@ -378,8 +690,12 @@ +@@ -378,8 +690,12 @@ static int _cpu_down(unsigned int cpu, int tasks_frozen) else synchronize_rcu(); @@ -16974,7 +17362,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpu.c linux-4.4.62/kernel/cpu.c /* * Prevent irq alloc/free while the dying cpu reorganizes the * interrupt affinities. -@@ -424,9 +740,14 @@ +@@ -424,9 +740,14 @@ static int _cpu_down(unsigned int cpu, int tasks_frozen) check_for_tasks(cpu); out_release: @@ -16989,10 +17377,11 @@ diff -Nur linux-4.4.62.orig/kernel/cpu.c linux-4.4.62/kernel/cpu.c return err; } -diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c ---- linux-4.4.62.orig/kernel/cpuset.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/cpuset.c 2017-04-18 17:38:08.206649710 +0200 -@@ -283,7 +283,7 @@ +diff --git a/kernel/cpuset.c b/kernel/cpuset.c +index b271353d5202..dd7b87b7f618 100644 +--- a/kernel/cpuset.c ++++ b/kernel/cpuset.c +@@ -283,7 +283,7 @@ static struct cpuset top_cpuset = { */ static DEFINE_MUTEX(cpuset_mutex); @@ -17001,7 +17390,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c static struct workqueue_struct *cpuset_migrate_mm_wq; -@@ -906,9 +906,9 @@ +@@ -906,9 +906,9 @@ static void update_cpumasks_hier(struct cpuset *cs, struct cpumask *new_cpus) continue; rcu_read_unlock(); @@ -17013,7 +17402,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c WARN_ON(!cgroup_subsys_on_dfl(cpuset_cgrp_subsys) && !cpumask_equal(cp->cpus_allowed, cp->effective_cpus)); -@@ -973,9 +973,9 @@ +@@ -973,9 +973,9 @@ static int update_cpumask(struct cpuset *cs, struct cpuset *trialcs, if (retval < 0) return retval; @@ -17025,7 +17414,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c /* use trialcs->cpus_allowed as a temp variable */ update_cpumasks_hier(cs, trialcs->cpus_allowed); -@@ -1184,9 +1184,9 @@ +@@ -1184,9 +1184,9 @@ static void update_nodemasks_hier(struct cpuset *cs, nodemask_t *new_mems) continue; rcu_read_unlock(); @@ -17037,7 +17426,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c WARN_ON(!cgroup_subsys_on_dfl(cpuset_cgrp_subsys) && !nodes_equal(cp->mems_allowed, cp->effective_mems)); -@@ -1254,9 +1254,9 @@ +@@ -1254,9 +1254,9 @@ static int update_nodemask(struct cpuset *cs, struct cpuset *trialcs, if (retval < 0) goto done; @@ -17049,7 +17438,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c /* use trialcs->mems_allowed as a temp variable */ update_nodemasks_hier(cs, &trialcs->mems_allowed); -@@ -1347,9 +1347,9 @@ +@@ -1347,9 +1347,9 @@ static int update_flag(cpuset_flagbits_t bit, struct cpuset *cs, spread_flag_changed = ((is_spread_slab(cs) != is_spread_slab(trialcs)) || (is_spread_page(cs) != is_spread_page(trialcs))); @@ -17061,7 +17450,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c if (!cpumask_empty(trialcs->cpus_allowed) && balance_flag_changed) rebuild_sched_domains_locked(); -@@ -1761,7 +1761,7 @@ +@@ -1761,7 +1761,7 @@ static int cpuset_common_seq_show(struct seq_file *sf, void *v) cpuset_filetype_t type = seq_cft(sf)->private; int ret = 0; @@ -17070,7 +17459,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c switch (type) { case FILE_CPULIST: -@@ -1780,7 +1780,7 @@ +@@ -1780,7 +1780,7 @@ static int cpuset_common_seq_show(struct seq_file *sf, void *v) ret = -EINVAL; } @@ -17079,7 +17468,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c return ret; } -@@ -1994,12 +1994,12 @@ +@@ -1994,12 +1994,12 @@ static int cpuset_css_online(struct cgroup_subsys_state *css) cpuset_inc(); @@ -17094,7 +17483,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c if (!test_bit(CGRP_CPUSET_CLONE_CHILDREN, &css->cgroup->flags)) goto out_unlock; -@@ -2026,12 +2026,12 @@ +@@ -2026,12 +2026,12 @@ static int cpuset_css_online(struct cgroup_subsys_state *css) } rcu_read_unlock(); @@ -17109,7 +17498,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c out_unlock: mutex_unlock(&cpuset_mutex); return 0; -@@ -2070,7 +2070,7 @@ +@@ -2070,7 +2070,7 @@ static void cpuset_css_free(struct cgroup_subsys_state *css) static void cpuset_bind(struct cgroup_subsys_state *root_css) { mutex_lock(&cpuset_mutex); @@ -17118,7 +17507,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c if (cgroup_subsys_on_dfl(cpuset_cgrp_subsys)) { cpumask_copy(top_cpuset.cpus_allowed, cpu_possible_mask); -@@ -2081,7 +2081,7 @@ +@@ -2081,7 +2081,7 @@ static void cpuset_bind(struct cgroup_subsys_state *root_css) top_cpuset.mems_allowed = top_cpuset.effective_mems; } @@ -17127,7 +17516,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c mutex_unlock(&cpuset_mutex); } -@@ -2182,12 +2182,12 @@ +@@ -2182,12 +2182,12 @@ hotplug_update_tasks_legacy(struct cpuset *cs, { bool is_empty; @@ -17142,7 +17531,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c /* * Don't call update_tasks_cpumask() if the cpuset becomes empty, -@@ -2224,10 +2224,10 @@ +@@ -2224,10 +2224,10 @@ hotplug_update_tasks(struct cpuset *cs, if (nodes_empty(*new_mems)) *new_mems = parent_cs(cs)->effective_mems; @@ -17155,7 +17544,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c if (cpus_updated) update_tasks_cpumask(cs); -@@ -2313,21 +2313,21 @@ +@@ -2313,21 +2313,21 @@ static void cpuset_hotplug_workfn(struct work_struct *work) /* synchronize cpus_allowed to cpu_active_mask */ if (cpus_updated) { @@ -17181,7 +17570,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c update_tasks_nodemask(&top_cpuset); } -@@ -2425,11 +2425,11 @@ +@@ -2425,11 +2425,11 @@ void cpuset_cpus_allowed(struct task_struct *tsk, struct cpumask *pmask) { unsigned long flags; @@ -17195,7 +17584,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c } void cpuset_cpus_allowed_fallback(struct task_struct *tsk) -@@ -2477,11 +2477,11 @@ +@@ -2477,11 +2477,11 @@ nodemask_t cpuset_mems_allowed(struct task_struct *tsk) nodemask_t mask; unsigned long flags; @@ -17209,7 +17598,7 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c return mask; } -@@ -2573,14 +2573,14 @@ +@@ -2573,14 +2573,14 @@ int __cpuset_node_allowed(int node, gfp_t gfp_mask) return 1; /* Not hardwall and node outside mems_allowed: scan up cpusets */ @@ -17226,10 +17615,11 @@ diff -Nur linux-4.4.62.orig/kernel/cpuset.c linux-4.4.62/kernel/cpuset.c return allowed; } -diff -Nur linux-4.4.62.orig/kernel/debug/kdb/kdb_io.c linux-4.4.62/kernel/debug/kdb/kdb_io.c ---- linux-4.4.62.orig/kernel/debug/kdb/kdb_io.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/debug/kdb/kdb_io.c 2017-04-18 17:38:08.206649710 +0200 -@@ -554,7 +554,6 @@ +diff --git a/kernel/debug/kdb/kdb_io.c b/kernel/debug/kdb/kdb_io.c +index fc1ef736253c..83c666537a7a 100644 +--- a/kernel/debug/kdb/kdb_io.c ++++ b/kernel/debug/kdb/kdb_io.c +@@ -554,7 +554,6 @@ int vkdb_printf(enum kdb_msgsrc src, const char *fmt, va_list ap) int linecount; int colcount; int logging, saved_loglevel = 0; @@ -17237,7 +17627,7 @@ diff -Nur linux-4.4.62.orig/kernel/debug/kdb/kdb_io.c linux-4.4.62/kernel/debug/ int got_printf_lock = 0; int retlen = 0; int fnd, len; -@@ -565,8 +564,6 @@ +@@ -565,8 +564,6 @@ int vkdb_printf(enum kdb_msgsrc src, const char *fmt, va_list ap) unsigned long uninitialized_var(flags); preempt_disable(); @@ -17246,7 +17636,7 @@ diff -Nur linux-4.4.62.orig/kernel/debug/kdb/kdb_io.c linux-4.4.62/kernel/debug/ /* Serialize kdb_printf if multiple cpus try to write at once. * But if any cpu goes recursive in kdb, just print the output, -@@ -855,7 +852,6 @@ +@@ -855,7 +852,6 @@ kdb_print_out: } else { __release(kdb_printf_lock); } @@ -17254,7 +17644,7 @@ diff -Nur linux-4.4.62.orig/kernel/debug/kdb/kdb_io.c linux-4.4.62/kernel/debug/ preempt_enable(); return retlen; } -@@ -865,9 +861,11 @@ +@@ -865,9 +861,11 @@ int kdb_printf(const char *fmt, ...) va_list ap; int r; @@ -17266,10 +17656,11 @@ diff -Nur linux-4.4.62.orig/kernel/debug/kdb/kdb_io.c linux-4.4.62/kernel/debug/ return r; } -diff -Nur linux-4.4.62.orig/kernel/events/core.c linux-4.4.62/kernel/events/core.c ---- linux-4.4.62.orig/kernel/events/core.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/events/core.c 2017-04-18 17:38:08.206649710 +0200 -@@ -802,6 +802,7 @@ +diff --git a/kernel/events/core.c b/kernel/events/core.c +index 784ab8fe8714..03166fb2364e 100644 +--- a/kernel/events/core.c ++++ b/kernel/events/core.c +@@ -802,6 +802,7 @@ static void __perf_mux_hrtimer_init(struct perf_cpu_context *cpuctx, int cpu) raw_spin_lock_init(&cpuctx->hrtimer_lock); hrtimer_init(timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS_PINNED); timer->function = perf_mux_hrtimer_handler; @@ -17277,7 +17668,7 @@ diff -Nur linux-4.4.62.orig/kernel/events/core.c linux-4.4.62/kernel/events/core } static int perf_mux_hrtimer_restart(struct perf_cpu_context *cpuctx) -@@ -7240,6 +7241,7 @@ +@@ -7240,6 +7241,7 @@ static void perf_swevent_init_hrtimer(struct perf_event *event) hrtimer_init(&hwc->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); hwc->hrtimer.function = perf_swevent_hrtimer; @@ -17285,10 +17676,11 @@ diff -Nur linux-4.4.62.orig/kernel/events/core.c linux-4.4.62/kernel/events/core /* * Since hrtimers have a fixed rate, we can do a static freq->period -diff -Nur linux-4.4.62.orig/kernel/exit.c linux-4.4.62/kernel/exit.c ---- linux-4.4.62.orig/kernel/exit.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/exit.c 2017-04-18 17:38:08.206649710 +0200 -@@ -144,7 +144,7 @@ +diff --git a/kernel/exit.c b/kernel/exit.c +index ffba5df4abd5..e199407f8831 100644 +--- a/kernel/exit.c ++++ b/kernel/exit.c +@@ -144,7 +144,7 @@ static void __exit_signal(struct task_struct *tsk) * Do this under ->siglock, we can race with another thread * doing sigqueue_free() if we have SIGQUEUE_PREALLOC signals. */ @@ -17297,10 +17689,11 @@ diff -Nur linux-4.4.62.orig/kernel/exit.c linux-4.4.62/kernel/exit.c tsk->sighand = NULL; spin_unlock(&sighand->siglock); -diff -Nur linux-4.4.62.orig/kernel/fork.c linux-4.4.62/kernel/fork.c ---- linux-4.4.62.orig/kernel/fork.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/fork.c 2017-04-18 17:38:08.206649710 +0200 -@@ -108,7 +108,7 @@ +diff --git a/kernel/fork.c b/kernel/fork.c +index 0ee630f3ad4b..0fe2c188e94d 100644 +--- a/kernel/fork.c ++++ b/kernel/fork.c +@@ -108,7 +108,7 @@ int max_threads; /* tunable limit on nr_threads */ DEFINE_PER_CPU(unsigned long, process_counts) = 0; @@ -17309,7 +17702,7 @@ diff -Nur linux-4.4.62.orig/kernel/fork.c linux-4.4.62/kernel/fork.c #ifdef CONFIG_PROVE_RCU int lockdep_tasklist_lock_is_held(void) -@@ -244,7 +244,9 @@ +@@ -244,7 +244,9 @@ static inline void put_signal_struct(struct signal_struct *sig) if (atomic_dec_and_test(&sig->sigcnt)) free_signal_struct(sig); } @@ -17320,7 +17713,7 @@ diff -Nur linux-4.4.62.orig/kernel/fork.c linux-4.4.62/kernel/fork.c void __put_task_struct(struct task_struct *tsk) { WARN_ON(!tsk->exit_state); -@@ -261,7 +263,18 @@ +@@ -261,7 +263,18 @@ void __put_task_struct(struct task_struct *tsk) if (!profile_handoff_task(tsk)) free_task(tsk); } @@ -17339,7 +17732,7 @@ diff -Nur linux-4.4.62.orig/kernel/fork.c linux-4.4.62/kernel/fork.c void __init __weak arch_task_cache_init(void) { } -@@ -693,6 +706,19 @@ +@@ -693,6 +706,19 @@ void __mmdrop(struct mm_struct *mm) } EXPORT_SYMBOL_GPL(__mmdrop); @@ -17359,7 +17752,7 @@ diff -Nur linux-4.4.62.orig/kernel/fork.c linux-4.4.62/kernel/fork.c /* * Decrement the use count and release all resources for an mm. */ -@@ -1243,6 +1269,9 @@ +@@ -1243,6 +1269,9 @@ static void rt_mutex_init_task(struct task_struct *p) */ static void posix_cpu_timers_init(struct task_struct *tsk) { @@ -17369,7 +17762,7 @@ diff -Nur linux-4.4.62.orig/kernel/fork.c linux-4.4.62/kernel/fork.c tsk->cputime_expires.prof_exp = 0; tsk->cputime_expires.virt_exp = 0; tsk->cputime_expires.sched_exp = 0; -@@ -1369,15 +1398,16 @@ +@@ -1369,15 +1398,16 @@ static struct task_struct *copy_process(unsigned long clone_flags, spin_lock_init(&p->alloc_lock); init_sigpending(&p->pending); @@ -17388,10 +17781,11 @@ diff -Nur linux-4.4.62.orig/kernel/fork.c linux-4.4.62/kernel/fork.c #endif #if defined(SPLIT_RSS_COUNTING) -diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c ---- linux-4.4.62.orig/kernel/futex.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/futex.c 2017-04-18 17:38:08.206649710 +0200 -@@ -815,7 +815,9 @@ +diff --git a/kernel/futex.c b/kernel/futex.c +index 3057dabf726f..9604b73ada47 100644 +--- a/kernel/futex.c ++++ b/kernel/futex.c +@@ -815,7 +815,9 @@ void exit_pi_state_list(struct task_struct *curr) * task still owns the PI-state: */ if (head->next != next) { @@ -17401,7 +17795,7 @@ diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c continue; } -@@ -1210,6 +1212,7 @@ +@@ -1210,6 +1212,7 @@ static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, struct futex_pi_state *pi_state = this->pi_state; u32 uninitialized_var(curval), newval; WAKE_Q(wake_q); @@ -17409,7 +17803,7 @@ diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c bool deboost; int ret = 0; -@@ -1223,7 +1226,7 @@ +@@ -1223,7 +1226,7 @@ static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, if (pi_state->owner != current) return -EINVAL; @@ -17418,7 +17812,7 @@ diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c new_owner = rt_mutex_next_owner(&pi_state->pi_mutex); /* -@@ -1259,24 +1262,25 @@ +@@ -1259,24 +1262,25 @@ static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, ret = -EINVAL; } if (ret) { @@ -17451,7 +17845,7 @@ diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c /* * First unlock HB so the waiter does not spin on it once he got woken -@@ -1284,8 +1288,9 @@ +@@ -1284,8 +1288,9 @@ static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, * deboost first (and lose our higher priority), then the task might get * scheduled away before the wake up can take place. */ @@ -17462,7 +17856,7 @@ diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c if (deboost) rt_mutex_adjust_prio(current); -@@ -1822,6 +1827,16 @@ +@@ -1822,6 +1827,16 @@ retry_private: requeue_pi_wake_futex(this, &key2, hb2); drop_count++; continue; @@ -17479,7 +17873,7 @@ diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c } else if (ret) { /* -EDEADLK */ this->pi_state = NULL; -@@ -2139,11 +2154,11 @@ +@@ -2139,11 +2154,11 @@ static int fixup_owner(u32 __user *uaddr, struct futex_q *q, int locked) * we returned due to timeout or signal without taking the * rt_mutex. Too late. */ @@ -17493,7 +17887,7 @@ diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c ret = fixup_pi_state_owner(uaddr, q, owner); goto out; } -@@ -2690,7 +2705,7 @@ +@@ -2690,7 +2705,7 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, { struct hrtimer_sleeper timeout, *to = NULL; struct rt_mutex_waiter rt_waiter; @@ -17502,7 +17896,7 @@ diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c union futex_key key2 = FUTEX_KEY_INIT; struct futex_q q = futex_q_init; int res, ret; -@@ -2715,10 +2730,7 @@ +@@ -2715,10 +2730,7 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, * The waiter is allocated on our stack, manipulated by the requeue * code while we sleep on uaddr. */ @@ -17514,7 +17908,7 @@ diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2, VERIFY_WRITE); if (unlikely(ret != 0)) -@@ -2749,20 +2761,55 @@ +@@ -2749,20 +2761,55 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, /* Queue the futex_q, drop the hb lock, wait for wakeup. */ futex_wait_queue_me(hb, &q, to); @@ -17581,7 +17975,7 @@ diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c /* Check if the requeue code acquired the second futex for us. */ if (!q.rt_waiter) { -@@ -2771,7 +2818,8 @@ +@@ -2771,7 +2818,8 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, * did a lock-steal - fix up the PI-state in that case. */ if (q.pi_state && (q.pi_state->owner != current)) { @@ -17591,7 +17985,7 @@ diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c ret = fixup_pi_state_owner(uaddr2, &q, current); if (ret && rt_mutex_owner(&q.pi_state->pi_mutex) == current) rt_mutex_unlock(&q.pi_state->pi_mutex); -@@ -2780,7 +2828,7 @@ +@@ -2780,7 +2828,7 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, * the requeue_pi() code acquired for us. */ free_pi_state(q.pi_state); @@ -17600,7 +17994,7 @@ diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c } } else { struct rt_mutex *pi_mutex; -@@ -2795,7 +2843,8 @@ +@@ -2795,7 +2843,8 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, ret = rt_mutex_finish_proxy_lock(pi_mutex, to, &rt_waiter); debug_rt_mutex_free_waiter(&rt_waiter); @@ -17610,10 +18004,11 @@ diff -Nur linux-4.4.62.orig/kernel/futex.c linux-4.4.62/kernel/futex.c /* * Fixup the pi_state owner and possibly acquire the lock if we * haven't already. -diff -Nur linux-4.4.62.orig/kernel/irq/handle.c linux-4.4.62/kernel/irq/handle.c ---- linux-4.4.62.orig/kernel/irq/handle.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/irq/handle.c 2017-04-18 17:38:08.206649710 +0200 -@@ -134,6 +134,8 @@ +diff --git a/kernel/irq/handle.c b/kernel/irq/handle.c +index 57bff7857e87..6c65c9252991 100644 +--- a/kernel/irq/handle.c ++++ b/kernel/irq/handle.c +@@ -134,6 +134,8 @@ void __irq_wake_thread(struct irq_desc *desc, struct irqaction *action) irqreturn_t handle_irq_event_percpu(struct irq_desc *desc) { @@ -17622,7 +18017,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/handle.c linux-4.4.62/kernel/irq/handle.c irqreturn_t retval = IRQ_NONE; unsigned int flags = 0, irq = desc->irq_data.irq; struct irqaction *action = desc->action; -@@ -176,7 +178,11 @@ +@@ -176,7 +178,11 @@ irqreturn_t handle_irq_event_percpu(struct irq_desc *desc) action = action->next; } @@ -17635,9 +18030,10 @@ diff -Nur linux-4.4.62.orig/kernel/irq/handle.c linux-4.4.62/kernel/irq/handle.c if (!noirqdebug) note_interrupt(desc, retval); -diff -Nur linux-4.4.62.orig/kernel/irq/irqdesc.c linux-4.4.62/kernel/irq/irqdesc.c ---- linux-4.4.62.orig/kernel/irq/irqdesc.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/irq/irqdesc.c 2017-04-18 17:38:08.206649710 +0200 +diff --git a/kernel/irq/irqdesc.c b/kernel/irq/irqdesc.c +index 239e2ae2c947..0b73349a42d5 100644 +--- a/kernel/irq/irqdesc.c ++++ b/kernel/irq/irqdesc.c @@ -24,10 +24,27 @@ static struct lock_class_key irq_desc_lock_class; @@ -17668,9 +18064,10 @@ diff -Nur linux-4.4.62.orig/kernel/irq/irqdesc.c linux-4.4.62/kernel/irq/irqdesc } #else static void __init init_irq_default_affinity(void) -diff -Nur linux-4.4.62.orig/kernel/irq/manage.c linux-4.4.62/kernel/irq/manage.c ---- linux-4.4.62.orig/kernel/irq/manage.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/irq/manage.c 2017-04-18 17:38:08.206649710 +0200 +diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c +index 6ead200370da..8e89554aa345 100644 +--- a/kernel/irq/manage.c ++++ b/kernel/irq/manage.c @@ -22,6 +22,7 @@ #include "internals.h" @@ -17679,7 +18076,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/manage.c linux-4.4.62/kernel/irq/manage.c __read_mostly bool force_irqthreads; static int __init setup_forced_irqthreads(char *arg) -@@ -30,6 +31,7 @@ +@@ -30,6 +31,7 @@ static int __init setup_forced_irqthreads(char *arg) return 0; } early_param("threadirqs", setup_forced_irqthreads); @@ -17687,7 +18084,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/manage.c linux-4.4.62/kernel/irq/manage.c #endif static void __synchronize_hardirq(struct irq_desc *desc) -@@ -181,6 +183,62 @@ +@@ -181,6 +183,62 @@ static inline void irq_get_pending(struct cpumask *mask, struct irq_desc *desc) { } #endif @@ -17750,7 +18147,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/manage.c linux-4.4.62/kernel/irq/manage.c int irq_do_set_affinity(struct irq_data *data, const struct cpumask *mask, bool force) { -@@ -220,7 +278,17 @@ +@@ -220,7 +278,17 @@ int irq_set_affinity_locked(struct irq_data *data, const struct cpumask *mask, if (desc->affinity_notify) { kref_get(&desc->affinity_notify->kref); @@ -17768,7 +18165,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/manage.c linux-4.4.62/kernel/irq/manage.c } irqd_set(data, IRQD_AFFINITY_SET); -@@ -258,10 +326,8 @@ +@@ -258,10 +326,8 @@ int irq_set_affinity_hint(unsigned int irq, const struct cpumask *m) } EXPORT_SYMBOL_GPL(irq_set_affinity_hint); @@ -17780,7 +18177,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/manage.c linux-4.4.62/kernel/irq/manage.c struct irq_desc *desc = irq_to_desc(notify->irq); cpumask_var_t cpumask; unsigned long flags; -@@ -283,6 +349,13 @@ +@@ -283,6 +349,13 @@ out: kref_put(¬ify->kref, notify->release); } @@ -17794,7 +18191,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/manage.c linux-4.4.62/kernel/irq/manage.c /** * irq_set_affinity_notifier - control notification of IRQ affinity changes * @irq: Interrupt for which to enable/disable notification -@@ -312,6 +385,8 @@ +@@ -312,6 +385,8 @@ irq_set_affinity_notifier(unsigned int irq, struct irq_affinity_notify *notify) notify->irq = irq; kref_init(¬ify->kref); INIT_WORK(¬ify->work, irq_affinity_notify); @@ -17803,7 +18200,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/manage.c linux-4.4.62/kernel/irq/manage.c } raw_spin_lock_irqsave(&desc->lock, flags); -@@ -865,7 +940,15 @@ +@@ -865,7 +940,15 @@ irq_forced_thread_fn(struct irq_desc *desc, struct irqaction *action) local_bh_disable(); ret = action->thread_fn(action->irq, action->dev_id); irq_finalize_oneshot(desc, action); @@ -17820,7 +18217,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/manage.c linux-4.4.62/kernel/irq/manage.c return ret; } -@@ -962,6 +1045,12 @@ +@@ -962,6 +1045,12 @@ static int irq_thread(void *data) if (action_ret == IRQ_WAKE_THREAD) irq_wake_secondary(desc, action); @@ -17833,7 +18230,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/manage.c linux-4.4.62/kernel/irq/manage.c wake_threads_waitq(desc); } -@@ -1315,6 +1404,9 @@ +@@ -1315,6 +1404,9 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new) irqd_set(&desc->irq_data, IRQD_NO_BALANCING); } @@ -17843,7 +18240,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/manage.c linux-4.4.62/kernel/irq/manage.c /* Set default affinity mask once everything is setup */ setup_affinity(desc, mask); -@@ -1968,7 +2060,7 @@ +@@ -1968,7 +2060,7 @@ EXPORT_SYMBOL_GPL(irq_get_irqchip_state); * This call sets the internal irqchip state of an interrupt, * depending on the value of @which. * @@ -17852,10 +18249,11 @@ diff -Nur linux-4.4.62.orig/kernel/irq/manage.c linux-4.4.62/kernel/irq/manage.c * interrupt controller has per-cpu registers. */ int irq_set_irqchip_state(unsigned int irq, enum irqchip_irq_state which, -diff -Nur linux-4.4.62.orig/kernel/irq/settings.h linux-4.4.62/kernel/irq/settings.h ---- linux-4.4.62.orig/kernel/irq/settings.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/irq/settings.h 2017-04-18 17:38:08.206649710 +0200 -@@ -16,6 +16,7 @@ +diff --git a/kernel/irq/settings.h b/kernel/irq/settings.h +index 320579d89091..2df2d4445b1e 100644 +--- a/kernel/irq/settings.h ++++ b/kernel/irq/settings.h +@@ -16,6 +16,7 @@ enum { _IRQ_PER_CPU_DEVID = IRQ_PER_CPU_DEVID, _IRQ_IS_POLLED = IRQ_IS_POLLED, _IRQ_DISABLE_UNLAZY = IRQ_DISABLE_UNLAZY, @@ -17863,7 +18261,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/settings.h linux-4.4.62/kernel/irq/settin _IRQF_MODIFY_MASK = IRQF_MODIFY_MASK, }; -@@ -30,6 +31,7 @@ +@@ -30,6 +31,7 @@ enum { #define IRQ_PER_CPU_DEVID GOT_YOU_MORON #define IRQ_IS_POLLED GOT_YOU_MORON #define IRQ_DISABLE_UNLAZY GOT_YOU_MORON @@ -17871,7 +18269,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/settings.h linux-4.4.62/kernel/irq/settin #undef IRQF_MODIFY_MASK #define IRQF_MODIFY_MASK GOT_YOU_MORON -@@ -40,6 +42,16 @@ +@@ -40,6 +42,16 @@ irq_settings_clr_and_set(struct irq_desc *desc, u32 clr, u32 set) desc->status_use_accessors |= (set & _IRQF_MODIFY_MASK); } @@ -17888,10 +18286,11 @@ diff -Nur linux-4.4.62.orig/kernel/irq/settings.h linux-4.4.62/kernel/irq/settin static inline bool irq_settings_is_per_cpu(struct irq_desc *desc) { return desc->status_use_accessors & _IRQ_PER_CPU; -diff -Nur linux-4.4.62.orig/kernel/irq/spurious.c linux-4.4.62/kernel/irq/spurious.c ---- linux-4.4.62.orig/kernel/irq/spurious.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/irq/spurious.c 2017-04-18 17:38:08.206649710 +0200 -@@ -444,6 +444,10 @@ +diff --git a/kernel/irq/spurious.c b/kernel/irq/spurious.c +index 32144175458d..ed26f2554972 100644 +--- a/kernel/irq/spurious.c ++++ b/kernel/irq/spurious.c +@@ -444,6 +444,10 @@ MODULE_PARM_DESC(noirqdebug, "Disable irq lockup detection when true"); static int __init irqfixup_setup(char *str) { @@ -17902,7 +18301,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq/spurious.c linux-4.4.62/kernel/irq/spurio irqfixup = 1; printk(KERN_WARNING "Misrouted IRQ fixup support enabled.\n"); printk(KERN_WARNING "This may impact system performance.\n"); -@@ -456,6 +460,10 @@ +@@ -456,6 +460,10 @@ module_param(irqfixup, int, 0644); static int __init irqpoll_setup(char *str) { @@ -17913,9 +18312,10 @@ diff -Nur linux-4.4.62.orig/kernel/irq/spurious.c linux-4.4.62/kernel/irq/spurio irqfixup = 2; printk(KERN_WARNING "Misrouted IRQ fixup and polling support " "enabled\n"); -diff -Nur linux-4.4.62.orig/kernel/irq_work.c linux-4.4.62/kernel/irq_work.c ---- linux-4.4.62.orig/kernel/irq_work.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/irq_work.c 2017-04-18 17:38:08.210649865 +0200 +diff --git a/kernel/irq_work.c b/kernel/irq_work.c +index bcf107ce0854..2899ba0d23d1 100644 +--- a/kernel/irq_work.c ++++ b/kernel/irq_work.c @@ -17,6 +17,7 @@ #include #include @@ -17924,7 +18324,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq_work.c linux-4.4.62/kernel/irq_work.c #include -@@ -65,6 +66,8 @@ +@@ -65,6 +66,8 @@ void __weak arch_irq_work_raise(void) */ bool irq_work_queue_on(struct irq_work *work, int cpu) { @@ -17933,7 +18333,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq_work.c linux-4.4.62/kernel/irq_work.c /* All work should have been flushed before going offline */ WARN_ON_ONCE(cpu_is_offline(cpu)); -@@ -75,7 +78,12 @@ +@@ -75,7 +78,12 @@ bool irq_work_queue_on(struct irq_work *work, int cpu) if (!irq_work_claim(work)) return false; @@ -17947,7 +18347,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq_work.c linux-4.4.62/kernel/irq_work.c arch_send_call_function_single_ipi(cpu); return true; -@@ -86,6 +94,9 @@ +@@ -86,6 +94,9 @@ EXPORT_SYMBOL_GPL(irq_work_queue_on); /* Enqueue the irq work @work on the current CPU */ bool irq_work_queue(struct irq_work *work) { @@ -17957,7 +18357,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq_work.c linux-4.4.62/kernel/irq_work.c /* Only queue if not already pending */ if (!irq_work_claim(work)) return false; -@@ -93,13 +104,15 @@ +@@ -93,13 +104,15 @@ bool irq_work_queue(struct irq_work *work) /* Queue the entry and raise the IPI if needed. */ preempt_disable(); @@ -17980,7 +18380,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq_work.c linux-4.4.62/kernel/irq_work.c arch_irq_work_raise(); } -@@ -116,9 +129,8 @@ +@@ -116,9 +129,8 @@ bool irq_work_needs_cpu(void) raised = this_cpu_ptr(&raised_list); lazy = this_cpu_ptr(&lazy_list); @@ -17992,7 +18392,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq_work.c linux-4.4.62/kernel/irq_work.c /* All work should have been flushed before going offline */ WARN_ON_ONCE(cpu_is_offline(smp_processor_id())); -@@ -132,7 +144,7 @@ +@@ -132,7 +144,7 @@ static void irq_work_run_list(struct llist_head *list) struct irq_work *work; struct llist_node *llnode; @@ -18001,7 +18401,7 @@ diff -Nur linux-4.4.62.orig/kernel/irq_work.c linux-4.4.62/kernel/irq_work.c if (llist_empty(list)) return; -@@ -169,7 +181,16 @@ +@@ -169,7 +181,16 @@ static void irq_work_run_list(struct llist_head *list) void irq_work_run(void) { irq_work_run_list(this_cpu_ptr(&raised_list)); @@ -18010,109 +18410,38 @@ diff -Nur linux-4.4.62.orig/kernel/irq_work.c linux-4.4.62/kernel/irq_work.c + /* + * NOTE: we raise softirq via IPI for safety, + * and execute in irq_work_tick() to move the -+ * overhead from hard to soft irq context. -+ */ -+ if (!llist_empty(this_cpu_ptr(&lazy_list))) -+ raise_softirq(TIMER_SOFTIRQ); -+ } else -+ irq_work_run_list(this_cpu_ptr(&lazy_list)); - } - EXPORT_SYMBOL_GPL(irq_work_run); - -@@ -179,8 +200,17 @@ - - if (!llist_empty(raised) && !arch_irq_work_has_interrupt()) - irq_work_run_list(raised); -+ -+ if (!IS_ENABLED(CONFIG_PREEMPT_RT_FULL)) -+ irq_work_run_list(this_cpu_ptr(&lazy_list)); -+} -+ -+#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) -+void irq_work_tick_soft(void) -+{ - irq_work_run_list(this_cpu_ptr(&lazy_list)); - } -+#endif - - /* - * Synchronize against the irq_work @entry, ensures the entry is not -diff -Nur linux-4.4.62.orig/kernel/Kconfig.locks linux-4.4.62/kernel/Kconfig.locks ---- linux-4.4.62.orig/kernel/Kconfig.locks 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/Kconfig.locks 2017-04-18 17:38:08.202649555 +0200 -@@ -225,11 +225,11 @@ - - config MUTEX_SPIN_ON_OWNER - def_bool y -- depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW -+ depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL - - config RWSEM_SPIN_ON_OWNER - def_bool y -- depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW -+ depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL - - config LOCK_SPIN_ON_OWNER - def_bool y -diff -Nur linux-4.4.62.orig/kernel/Kconfig.preempt linux-4.4.62/kernel/Kconfig.preempt ---- linux-4.4.62.orig/kernel/Kconfig.preempt 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/Kconfig.preempt 2017-04-18 17:38:08.202649555 +0200 -@@ -1,3 +1,16 @@ -+config PREEMPT -+ bool -+ select PREEMPT_COUNT -+ -+config PREEMPT_RT_BASE -+ bool -+ select PREEMPT -+ -+config HAVE_PREEMPT_LAZY -+ bool -+ -+config PREEMPT_LAZY -+ def_bool y if HAVE_PREEMPT_LAZY && PREEMPT_RT_FULL - - choice - prompt "Preemption Model" -@@ -33,9 +46,9 @@ - - Select this if you are building a kernel for a desktop system. ++ * overhead from hard to soft irq context. ++ */ ++ if (!llist_empty(this_cpu_ptr(&lazy_list))) ++ raise_softirq(TIMER_SOFTIRQ); ++ } else ++ irq_work_run_list(this_cpu_ptr(&lazy_list)); + } + EXPORT_SYMBOL_GPL(irq_work_run); --config PREEMPT -+config PREEMPT__LL - bool "Preemptible Kernel (Low-Latency Desktop)" -- select PREEMPT_COUNT -+ select PREEMPT - select UNINLINE_SPIN_UNLOCK if !ARCH_INLINE_SPIN_UNLOCK - help - This option reduces the latency of the kernel by making -@@ -52,6 +65,22 @@ - embedded system with latency requirements in the milliseconds - range. +@@ -179,8 +200,17 @@ void irq_work_tick(void) -+config PREEMPT_RTB -+ bool "Preemptible Kernel (Basic RT)" -+ select PREEMPT_RT_BASE -+ help -+ This option is basically the same as (Low-Latency Desktop) but -+ enables changes which are preliminary for the full preemptible -+ RT kernel. + if (!llist_empty(raised) && !arch_irq_work_has_interrupt()) + irq_work_run_list(raised); + -+config PREEMPT_RT_FULL -+ bool "Fully Preemptible Kernel (RT)" -+ depends on IRQ_FORCED_THREADING -+ select PREEMPT_RT_BASE -+ select PREEMPT_RCU -+ help -+ All and everything ++ if (!IS_ENABLED(CONFIG_PREEMPT_RT_FULL)) ++ irq_work_run_list(this_cpu_ptr(&lazy_list)); ++} + - endchoice ++#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) ++void irq_work_tick_soft(void) ++{ + irq_work_run_list(this_cpu_ptr(&lazy_list)); + } ++#endif - config PREEMPT_COUNT -diff -Nur linux-4.4.62.orig/kernel/ksysfs.c linux-4.4.62/kernel/ksysfs.c ---- linux-4.4.62.orig/kernel/ksysfs.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/ksysfs.c 2017-04-18 17:38:08.210649865 +0200 -@@ -136,6 +136,15 @@ + /* + * Synchronize against the irq_work @entry, ensures the entry is not +diff --git a/kernel/ksysfs.c b/kernel/ksysfs.c +index e83b26464061..c0e08d1cf33e 100644 +--- a/kernel/ksysfs.c ++++ b/kernel/ksysfs.c +@@ -136,6 +136,15 @@ KERNEL_ATTR_RO(vmcoreinfo); #endif /* CONFIG_KEXEC_CORE */ @@ -18128,7 +18457,7 @@ diff -Nur linux-4.4.62.orig/kernel/ksysfs.c linux-4.4.62/kernel/ksysfs.c /* whether file capabilities are enabled */ static ssize_t fscaps_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf) -@@ -203,6 +212,9 @@ +@@ -203,6 +212,9 @@ static struct attribute * kernel_attrs[] = { &vmcoreinfo_attr.attr, #endif &rcu_expedited_attr.attr, @@ -18138,9 +18467,44 @@ diff -Nur linux-4.4.62.orig/kernel/ksysfs.c linux-4.4.62/kernel/ksysfs.c NULL }; -diff -Nur linux-4.4.62.orig/kernel/locking/lglock.c linux-4.4.62/kernel/locking/lglock.c ---- linux-4.4.62.orig/kernel/locking/lglock.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/locking/lglock.c 2017-04-18 17:38:08.210649865 +0200 +diff --git a/kernel/locking/Makefile b/kernel/locking/Makefile +index 8e96f6cc2a4a..447b03082d88 100644 +--- a/kernel/locking/Makefile ++++ b/kernel/locking/Makefile +@@ -1,5 +1,5 @@ + +-obj-y += mutex.o semaphore.o rwsem.o percpu-rwsem.o ++obj-y += semaphore.o percpu-rwsem.o + + ifdef CONFIG_FUNCTION_TRACER + CFLAGS_REMOVE_lockdep.o = $(CC_FLAGS_FTRACE) +@@ -8,7 +8,11 @@ CFLAGS_REMOVE_mutex-debug.o = $(CC_FLAGS_FTRACE) + CFLAGS_REMOVE_rtmutex-debug.o = $(CC_FLAGS_FTRACE) + endif + ++ifneq ($(CONFIG_PREEMPT_RT_FULL),y) ++obj-y += mutex.o + obj-$(CONFIG_DEBUG_MUTEXES) += mutex-debug.o ++obj-y += rwsem.o ++endif + obj-$(CONFIG_LOCKDEP) += lockdep.o + ifeq ($(CONFIG_PROC_FS),y) + obj-$(CONFIG_LOCKDEP) += lockdep_proc.o +@@ -22,7 +26,10 @@ obj-$(CONFIG_RT_MUTEXES) += rtmutex.o + obj-$(CONFIG_DEBUG_RT_MUTEXES) += rtmutex-debug.o + obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock.o + obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock_debug.o ++ifneq ($(CONFIG_PREEMPT_RT_FULL),y) + obj-$(CONFIG_RWSEM_GENERIC_SPINLOCK) += rwsem-spinlock.o + obj-$(CONFIG_RWSEM_XCHGADD_ALGORITHM) += rwsem-xadd.o ++endif ++obj-$(CONFIG_PREEMPT_RT_FULL) += rt.o + obj-$(CONFIG_QUEUED_RWLOCKS) += qrwlock.o + obj-$(CONFIG_LOCK_TORTURE_TEST) += locktorture.o +diff --git a/kernel/locking/lglock.c b/kernel/locking/lglock.c +index 951cfcd10b4a..57e0ea72c28a 100644 +--- a/kernel/locking/lglock.c ++++ b/kernel/locking/lglock.c @@ -4,6 +4,15 @@ #include #include @@ -18230,7 +18594,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/lglock.c linux-4.4.62/kernel/locking/ } EXPORT_SYMBOL(lg_local_unlock_cpu); -@@ -68,30 +86,30 @@ +@@ -68,30 +86,30 @@ void lg_double_lock(struct lglock *lg, int cpu1, int cpu2) if (cpu2 < cpu1) swap(cpu1, cpu2); @@ -18270,7 +18634,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/lglock.c linux-4.4.62/kernel/locking/ } } EXPORT_SYMBOL(lg_global_lock); -@@ -102,10 +120,35 @@ +@@ -102,10 +120,35 @@ void lg_global_unlock(struct lglock *lg) lock_release(&lg->lock_dep_map, 1, _RET_IP_); for_each_possible_cpu(i) { @@ -18309,10 +18673,11 @@ diff -Nur linux-4.4.62.orig/kernel/locking/lglock.c linux-4.4.62/kernel/locking/ + } +} +#endif -diff -Nur linux-4.4.62.orig/kernel/locking/lockdep.c linux-4.4.62/kernel/locking/lockdep.c ---- linux-4.4.62.orig/kernel/locking/lockdep.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/locking/lockdep.c 2017-04-18 17:38:08.210649865 +0200 -@@ -668,6 +668,7 @@ +diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c +index 60ace56618f6..fd54dbf686cc 100644 +--- a/kernel/locking/lockdep.c ++++ b/kernel/locking/lockdep.c +@@ -668,6 +668,7 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) struct lockdep_subclass_key *key; struct list_head *hash_head; struct lock_class *class; @@ -18320,18 +18685,17 @@ diff -Nur linux-4.4.62.orig/kernel/locking/lockdep.c linux-4.4.62/kernel/locking #ifdef CONFIG_DEBUG_LOCKDEP /* -@@ -695,10 +696,23 @@ +@@ -695,10 +696,23 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) /* * Static locks do not have their class-keys yet - for them the key - * is the lock object itself: -- */ -- if (unlikely(!lock->key)) -- lock->key = (void *)lock; + * is the lock object itself. If the lock is in the per cpu area, + * the canonical address of the lock (per cpu offset removed) is + * used. -+ */ + */ +- if (unlikely(!lock->key)) +- lock->key = (void *)lock; + if (unlikely(!lock->key)) { + unsigned long can_addr, addr = (unsigned long)lock; + @@ -18348,7 +18712,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/lockdep.c linux-4.4.62/kernel/locking /* * NOTE: the class-key must be unique. For dynamic locks, a static -@@ -730,7 +744,7 @@ +@@ -730,7 +744,7 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) } } @@ -18357,7 +18721,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/lockdep.c linux-4.4.62/kernel/locking } /* -@@ -748,19 +762,18 @@ +@@ -748,19 +762,18 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) DEBUG_LOCKS_WARN_ON(!irqs_disabled()); class = look_up_lock_class(lock, subclass); @@ -18380,7 +18744,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/lockdep.c linux-4.4.62/kernel/locking return NULL; } -@@ -3278,7 +3291,7 @@ +@@ -3278,7 +3291,7 @@ static int match_held_lock(struct held_lock *hlock, struct lockdep_map *lock) * Clearly if the lock hasn't been acquired _ever_, we're not * holding it either, so report failure. */ @@ -18389,7 +18753,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/lockdep.c linux-4.4.62/kernel/locking return 0; /* -@@ -3525,6 +3538,7 @@ +@@ -3525,6 +3538,7 @@ static void check_flags(unsigned long flags) } } @@ -18397,7 +18761,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/lockdep.c linux-4.4.62/kernel/locking /* * We dont accurately track softirq state in e.g. * hardirq contexts (such as on 4KSTACKS), so only -@@ -3539,6 +3553,7 @@ +@@ -3539,6 +3553,7 @@ static void check_flags(unsigned long flags) DEBUG_LOCKS_WARN_ON(!current->softirqs_enabled); } } @@ -18405,7 +18769,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/lockdep.c linux-4.4.62/kernel/locking if (!debug_locks) print_irqtrace_events(current); -@@ -3977,7 +3992,7 @@ +@@ -3977,7 +3992,7 @@ void lockdep_reset_lock(struct lockdep_map *lock) * If the class exists we look it up and zap it: */ class = look_up_lock_class(lock, j); @@ -18414,9 +18778,10 @@ diff -Nur linux-4.4.62.orig/kernel/locking/lockdep.c linux-4.4.62/kernel/locking zap_class(class); } /* -diff -Nur linux-4.4.62.orig/kernel/locking/locktorture.c linux-4.4.62/kernel/locking/locktorture.c ---- linux-4.4.62.orig/kernel/locking/locktorture.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/locking/locktorture.c 2017-04-18 17:38:08.210649865 +0200 +diff --git a/kernel/locking/locktorture.c b/kernel/locking/locktorture.c +index 8ef1919d63b2..291fc19e28e0 100644 +--- a/kernel/locking/locktorture.c ++++ b/kernel/locking/locktorture.c @@ -26,7 +26,6 @@ #include #include @@ -18425,42 +18790,11 @@ diff -Nur linux-4.4.62.orig/kernel/locking/locktorture.c linux-4.4.62/kernel/loc #include #include #include -diff -Nur linux-4.4.62.orig/kernel/locking/Makefile linux-4.4.62/kernel/locking/Makefile ---- linux-4.4.62.orig/kernel/locking/Makefile 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/locking/Makefile 2017-04-18 17:38:08.210649865 +0200 -@@ -1,5 +1,5 @@ - --obj-y += mutex.o semaphore.o rwsem.o percpu-rwsem.o -+obj-y += semaphore.o percpu-rwsem.o - - ifdef CONFIG_FUNCTION_TRACER - CFLAGS_REMOVE_lockdep.o = $(CC_FLAGS_FTRACE) -@@ -8,7 +8,11 @@ - CFLAGS_REMOVE_rtmutex-debug.o = $(CC_FLAGS_FTRACE) - endif - -+ifneq ($(CONFIG_PREEMPT_RT_FULL),y) -+obj-y += mutex.o - obj-$(CONFIG_DEBUG_MUTEXES) += mutex-debug.o -+obj-y += rwsem.o -+endif - obj-$(CONFIG_LOCKDEP) += lockdep.o - ifeq ($(CONFIG_PROC_FS),y) - obj-$(CONFIG_LOCKDEP) += lockdep_proc.o -@@ -22,7 +26,10 @@ - obj-$(CONFIG_DEBUG_RT_MUTEXES) += rtmutex-debug.o - obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock.o - obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock_debug.o -+ifneq ($(CONFIG_PREEMPT_RT_FULL),y) - obj-$(CONFIG_RWSEM_GENERIC_SPINLOCK) += rwsem-spinlock.o - obj-$(CONFIG_RWSEM_XCHGADD_ALGORITHM) += rwsem-xadd.o -+endif -+obj-$(CONFIG_PREEMPT_RT_FULL) += rt.o - obj-$(CONFIG_QUEUED_RWLOCKS) += qrwlock.o - obj-$(CONFIG_LOCK_TORTURE_TEST) += locktorture.o -diff -Nur linux-4.4.62.orig/kernel/locking/rt.c linux-4.4.62/kernel/locking/rt.c ---- linux-4.4.62.orig/kernel/locking/rt.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/kernel/locking/rt.c 2017-04-18 17:38:08.210649865 +0200 +diff --git a/kernel/locking/rt.c b/kernel/locking/rt.c +new file mode 100644 +index 000000000000..d4ab61c1848b +--- /dev/null ++++ b/kernel/locking/rt.c @@ -0,0 +1,474 @@ +/* + * kernel/rt.c @@ -18936,9 +19270,10 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rt.c linux-4.4.62/kernel/locking/rt.c + return 1; +} +EXPORT_SYMBOL(atomic_dec_and_mutex_lock); -diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking/rtmutex.c ---- linux-4.4.62.orig/kernel/locking/rtmutex.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/locking/rtmutex.c 2017-04-18 17:38:08.210649865 +0200 +diff --git a/kernel/locking/rtmutex.c b/kernel/locking/rtmutex.c +index b066724d7a5b..0e9a6260441d 100644 +--- a/kernel/locking/rtmutex.c ++++ b/kernel/locking/rtmutex.c @@ -7,6 +7,11 @@ * Copyright (C) 2005-2006 Timesys Corp., Thomas Gleixner * Copyright (C) 2005 Kihon Technologies Inc., Steven Rostedt @@ -18959,7 +19294,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking #include "rtmutex_common.h" -@@ -133,6 +139,12 @@ +@@ -133,6 +139,12 @@ static void fixup_rt_mutex_waiters(struct rt_mutex *lock) WRITE_ONCE(*p, owner & ~RT_MUTEX_HAS_WAITERS); } @@ -18972,7 +19307,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* * We can speed up the acquire/release, if there's no debugging state to be * set up. -@@ -163,13 +175,14 @@ +@@ -163,13 +175,14 @@ static inline void mark_rt_mutex_waiters(struct rt_mutex *lock) * 2) Drop lock->wait_lock * 3) Try to unlock the lock with cmpxchg */ @@ -18989,7 +19324,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* * If a new waiter comes in between the unlock and the cmpxchg * we have two situations: -@@ -211,11 +224,12 @@ +@@ -211,11 +224,12 @@ static inline void mark_rt_mutex_waiters(struct rt_mutex *lock) /* * Simple slow path only version: lock->owner is protected by lock->wait_lock. */ @@ -19004,7 +19339,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking return true; } #endif -@@ -412,6 +426,14 @@ +@@ -412,6 +426,14 @@ static bool rt_mutex_cond_detect_deadlock(struct rt_mutex_waiter *waiter, return debug_rt_mutex_detect_deadlock(waiter, chwalk); } @@ -19019,7 +19354,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* * Max number of times we'll walk the boosting chain: */ -@@ -419,7 +441,8 @@ +@@ -419,7 +441,8 @@ int max_lock_depth = 1024; static inline struct rt_mutex *task_blocked_on_lock(struct task_struct *p) { @@ -19029,7 +19364,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking } /* -@@ -497,7 +520,6 @@ +@@ -497,7 +520,6 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, int ret = 0, depth = 0; struct rt_mutex *lock; bool detect_deadlock; @@ -19037,7 +19372,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking bool requeue = true; detect_deadlock = rt_mutex_cond_detect_deadlock(orig_waiter, chwalk); -@@ -540,7 +562,7 @@ +@@ -540,7 +562,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, /* * [1] Task cannot go away as we did a get_task() before ! */ @@ -19046,7 +19381,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* * [2] Get the waiter on which @task is blocked on. -@@ -556,7 +578,7 @@ +@@ -556,7 +578,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, * reached or the state of the chain has changed while we * dropped the locks. */ @@ -19055,7 +19390,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking goto out_unlock_pi; /* -@@ -624,7 +646,7 @@ +@@ -624,7 +646,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, * operations. */ if (!raw_spin_trylock(&lock->wait_lock)) { @@ -19064,7 +19399,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking cpu_relax(); goto retry; } -@@ -655,7 +677,7 @@ +@@ -655,7 +677,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, /* * No requeue[7] here. Just release @task [8] */ @@ -19073,7 +19408,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking put_task_struct(task); /* -@@ -663,14 +685,14 @@ +@@ -663,14 +685,14 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, * If there is no owner of the lock, end of chain. */ if (!rt_mutex_owner(lock)) { @@ -19090,7 +19425,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* * No requeue [11] here. We just do deadlock detection. -@@ -685,8 +707,8 @@ +@@ -685,8 +707,8 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, top_waiter = rt_mutex_top_waiter(lock); /* [13] Drop locks */ @@ -19101,7 +19436,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* If owner is not blocked, end of chain. */ if (!next_lock) -@@ -707,7 +729,7 @@ +@@ -707,7 +729,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, rt_mutex_enqueue(lock, waiter); /* [8] Release the task */ @@ -19110,7 +19445,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking put_task_struct(task); /* -@@ -718,21 +740,24 @@ +@@ -718,21 +740,24 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, * follow here. This is the end of the chain we are walking. */ if (!rt_mutex_owner(lock)) { @@ -19139,7 +19474,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* [11] requeue the pi waiters if necessary */ if (waiter == rt_mutex_top_waiter(lock)) { -@@ -786,8 +811,8 @@ +@@ -786,8 +811,8 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, top_waiter = rt_mutex_top_waiter(lock); /* [13] Drop the locks */ @@ -19150,7 +19485,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* * Make the actual exit decisions [12], based on the stored -@@ -810,28 +835,46 @@ +@@ -810,28 +835,46 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, goto again; out_unlock_pi: @@ -19203,7 +19538,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* * Before testing whether we can acquire @lock, we set the * RT_MUTEX_HAS_WAITERS bit in @lock->owner. This forces all -@@ -867,8 +910,10 @@ +@@ -867,8 +910,10 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, * If waiter is not the highest priority waiter of * @lock, give up. */ @@ -19215,7 +19550,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* * We can acquire the lock. Remove the waiter from the -@@ -886,14 +931,10 @@ +@@ -886,14 +931,10 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, * not need to be dequeued. */ if (rt_mutex_has_waiters(lock)) { @@ -19233,7 +19568,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* * The current top waiter stays enqueued. We * don't have to change anything in the lock -@@ -916,7 +957,7 @@ +@@ -916,7 +957,7 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, * case, but conditionals are more expensive than a redundant * store. */ @@ -19242,7 +19577,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking task->pi_blocked_on = NULL; /* * Finish the lock acquisition. @task is the new owner. If -@@ -925,7 +966,7 @@ +@@ -925,7 +966,7 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, */ if (rt_mutex_has_waiters(lock)) rt_mutex_enqueue_pi(task, rt_mutex_top_waiter(lock)); @@ -19251,7 +19586,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking takeit: /* We got the lock. */ -@@ -942,12 +983,444 @@ +@@ -942,12 +983,444 @@ takeit: return 1; } @@ -19697,7 +20032,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking */ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, struct rt_mutex_waiter *waiter, -@@ -958,7 +1431,6 @@ +@@ -958,7 +1431,6 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, struct rt_mutex_waiter *top_waiter = waiter; struct rt_mutex *next_lock; int chain_walk = 0, res; @@ -19705,7 +20040,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* * Early deadlock detection. We really don't want the task to -@@ -972,7 +1444,24 @@ +@@ -972,7 +1444,24 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, if (owner == task) return -EDEADLK; @@ -19731,7 +20066,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking __rt_mutex_adjust_prio(task); waiter->task = task; waiter->lock = lock; -@@ -985,18 +1474,18 @@ +@@ -985,18 +1474,18 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, task->pi_blocked_on = waiter; @@ -19753,7 +20088,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking chain_walk = 1; } else if (rt_mutex_cond_detect_deadlock(waiter, chwalk)) { chain_walk = 1; -@@ -1005,7 +1494,7 @@ +@@ -1005,7 +1494,7 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, /* Store the lock on which owner is blocked or NULL */ next_lock = task_blocked_on_lock(owner); @@ -19762,7 +20097,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* * Even if full deadlock detection is on, if the owner is not * blocked itself, we can avoid finding this out in the chain -@@ -1021,12 +1510,12 @@ +@@ -1021,12 +1510,12 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, */ get_task_struct(owner); @@ -19777,7 +20112,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking return res; } -@@ -1035,15 +1524,15 @@ +@@ -1035,15 +1524,15 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, * Remove the top waiter from the current tasks pi waiter tree and * queue it up. * @@ -19796,7 +20131,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking waiter = rt_mutex_top_waiter(lock); -@@ -1065,15 +1554,18 @@ +@@ -1065,15 +1554,18 @@ static void mark_wakeup_next_waiter(struct wake_q_head *wake_q, */ lock->owner = (void *) RT_MUTEX_HAS_WAITERS; @@ -19818,7 +20153,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking * have just failed to try_to_take_rt_mutex(). */ static void remove_waiter(struct rt_mutex *lock, -@@ -1081,13 +1573,12 @@ +@@ -1081,13 +1573,12 @@ static void remove_waiter(struct rt_mutex *lock, { bool is_top_waiter = (waiter == rt_mutex_top_waiter(lock)); struct task_struct *owner = rt_mutex_owner(lock); @@ -19835,7 +20170,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* * Only update priority if the waiter was the highest priority -@@ -1096,7 +1587,7 @@ +@@ -1096,7 +1587,7 @@ static void remove_waiter(struct rt_mutex *lock, if (!owner || !is_top_waiter) return; @@ -19844,7 +20179,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking rt_mutex_dequeue_pi(owner, waiter); -@@ -1106,9 +1597,10 @@ +@@ -1106,9 +1597,10 @@ static void remove_waiter(struct rt_mutex *lock, __rt_mutex_adjust_prio(owner); /* Store the lock on which owner is blocked or NULL */ @@ -19857,7 +20192,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* * Don't walk the chain, if the owner task is not blocked -@@ -1120,12 +1612,12 @@ +@@ -1120,12 +1612,12 @@ static void remove_waiter(struct rt_mutex *lock, /* gets dropped in rt_mutex_adjust_prio_chain()! */ get_task_struct(owner); @@ -19872,7 +20207,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking } /* -@@ -1142,17 +1634,17 @@ +@@ -1142,17 +1634,17 @@ void rt_mutex_adjust_pi(struct task_struct *task) raw_spin_lock_irqsave(&task->pi_lock, flags); waiter = task->pi_blocked_on; @@ -19892,7 +20227,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking rt_mutex_adjust_prio_chain(task, RT_MUTEX_MIN_CHAINWALK, NULL, next_lock, NULL, task); } -@@ -1161,16 +1653,17 @@ +@@ -1161,16 +1653,17 @@ void rt_mutex_adjust_pi(struct task_struct *task) * __rt_mutex_slowlock() - Perform the wait-wake-try-to-take loop * @lock: the rt_mutex to take * @state: the state the task should block in (TASK_INTERRUPTIBLE @@ -19913,7 +20248,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking { int ret = 0; -@@ -1193,13 +1686,19 @@ +@@ -1193,13 +1686,19 @@ __rt_mutex_slowlock(struct rt_mutex *lock, int state, break; } @@ -19935,7 +20270,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking set_current_state(state); } -@@ -1227,26 +1726,112 @@ +@@ -1227,26 +1726,112 @@ static void rt_mutex_handle_deadlock(int res, int detect_deadlock, } } @@ -20054,7 +20389,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking return 0; } -@@ -1260,13 +1845,23 @@ +@@ -1260,13 +1845,23 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state, if (likely(!ret)) /* sleep on the mutex */ @@ -20080,7 +20415,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking } /* -@@ -1275,7 +1870,7 @@ +@@ -1275,7 +1870,7 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state, */ fixup_rt_mutex_waiters(lock); @@ -20089,7 +20424,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* Remove pending timer: */ if (unlikely(timeout)) -@@ -1291,6 +1886,7 @@ +@@ -1291,6 +1886,7 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state, */ static inline int rt_mutex_slowtrylock(struct rt_mutex *lock) { @@ -20097,7 +20432,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking int ret; /* -@@ -1302,10 +1898,10 @@ +@@ -1302,10 +1898,10 @@ static inline int rt_mutex_slowtrylock(struct rt_mutex *lock) return 0; /* @@ -20111,7 +20446,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking ret = try_to_take_rt_mutex(lock, current, NULL); -@@ -1315,7 +1911,7 @@ +@@ -1315,7 +1911,7 @@ static inline int rt_mutex_slowtrylock(struct rt_mutex *lock) */ fixup_rt_mutex_waiters(lock); @@ -20120,7 +20455,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking return ret; } -@@ -1325,9 +1921,13 @@ +@@ -1325,9 +1921,13 @@ static inline int rt_mutex_slowtrylock(struct rt_mutex *lock) * Return whether the current task needs to undo a potential priority boosting. */ static bool __sched rt_mutex_slowunlock(struct rt_mutex *lock, @@ -20136,7 +20471,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking debug_rt_mutex_unlock(lock); -@@ -1366,10 +1966,10 @@ +@@ -1366,10 +1966,10 @@ static bool __sched rt_mutex_slowunlock(struct rt_mutex *lock, */ while (!rt_mutex_has_waiters(lock)) { /* Drops lock->wait_lock ! */ @@ -20149,7 +20484,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking } /* -@@ -1378,9 +1978,9 @@ +@@ -1378,9 +1978,9 @@ static bool __sched rt_mutex_slowunlock(struct rt_mutex *lock, * * Queue the next waiter for wakeup once we release the wait_lock. */ @@ -20161,7 +20496,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* check PI boosting */ return true; -@@ -1394,31 +1994,36 @@ +@@ -1394,31 +1994,36 @@ static bool __sched rt_mutex_slowunlock(struct rt_mutex *lock, */ static inline int rt_mutex_fastlock(struct rt_mutex *lock, int state, @@ -20202,7 +20537,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking } static inline int -@@ -1435,17 +2040,20 @@ +@@ -1435,17 +2040,20 @@ rt_mutex_fasttrylock(struct rt_mutex *lock, static inline void rt_mutex_fastunlock(struct rt_mutex *lock, bool (*slowfn)(struct rt_mutex *lock, @@ -20225,7 +20560,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* Undo pi boosting if necessary: */ if (deboost) -@@ -1462,7 +2070,7 @@ +@@ -1462,7 +2070,7 @@ void __sched rt_mutex_lock(struct rt_mutex *lock) { might_sleep(); @@ -20234,7 +20569,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking } EXPORT_SYMBOL_GPL(rt_mutex_lock); -@@ -1479,7 +2087,7 @@ +@@ -1479,7 +2087,7 @@ int __sched rt_mutex_lock_interruptible(struct rt_mutex *lock) { might_sleep(); @@ -20243,7 +20578,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking } EXPORT_SYMBOL_GPL(rt_mutex_lock_interruptible); -@@ -1492,11 +2100,30 @@ +@@ -1492,11 +2100,30 @@ int rt_mutex_timed_futex_lock(struct rt_mutex *lock, might_sleep(); return rt_mutex_timed_fastlock(lock, TASK_INTERRUPTIBLE, timeout, @@ -20275,7 +20610,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking * rt_mutex_timed_lock - lock a rt_mutex interruptible * the timeout structure is provided * by the caller -@@ -1516,6 +2143,7 @@ +@@ -1516,6 +2143,7 @@ rt_mutex_timed_lock(struct rt_mutex *lock, struct hrtimer_sleeper *timeout) return rt_mutex_timed_fastlock(lock, TASK_INTERRUPTIBLE, timeout, RT_MUTEX_MIN_CHAINWALK, @@ -20283,7 +20618,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking rt_mutex_slowlock); } EXPORT_SYMBOL_GPL(rt_mutex_timed_lock); -@@ -1533,7 +2161,11 @@ +@@ -1533,7 +2161,11 @@ EXPORT_SYMBOL_GPL(rt_mutex_timed_lock); */ int __sched rt_mutex_trylock(struct rt_mutex *lock) { @@ -20295,7 +20630,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking return 0; return rt_mutex_fasttrylock(lock, rt_mutex_slowtrylock); -@@ -1559,13 +2191,14 @@ +@@ -1559,13 +2191,14 @@ EXPORT_SYMBOL_GPL(rt_mutex_unlock); * required or not. */ bool __sched rt_mutex_futex_unlock(struct rt_mutex *lock, @@ -20312,7 +20647,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking } /** -@@ -1598,13 +2231,12 @@ +@@ -1598,13 +2231,12 @@ EXPORT_SYMBOL_GPL(rt_mutex_destroy); void __rt_mutex_init(struct rt_mutex *lock, const char *name) { lock->owner = NULL; @@ -20327,7 +20662,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /** * rt_mutex_init_proxy_locked - initialize and lock a rt_mutex on behalf of a -@@ -1619,7 +2251,7 @@ +@@ -1619,7 +2251,7 @@ EXPORT_SYMBOL_GPL(__rt_mutex_init); void rt_mutex_init_proxy_locked(struct rt_mutex *lock, struct task_struct *proxy_owner) { @@ -20336,7 +20671,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking debug_rt_mutex_proxy_lock(lock, proxy_owner); rt_mutex_set_owner(lock, proxy_owner); rt_mutex_deadlock_account_lock(lock, proxy_owner); -@@ -1660,13 +2292,42 @@ +@@ -1660,13 +2292,42 @@ int rt_mutex_start_proxy_lock(struct rt_mutex *lock, { int ret; @@ -20381,7 +20716,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking /* We enforce deadlock detection for futexes */ ret = task_blocks_on_rt_mutex(lock, waiter, task, RT_MUTEX_FULL_CHAINWALK); -@@ -1681,10 +2342,10 @@ +@@ -1681,10 +2342,10 @@ int rt_mutex_start_proxy_lock(struct rt_mutex *lock, ret = 0; } @@ -20394,7 +20729,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking debug_rt_mutex_print_deadlock(waiter); -@@ -1732,12 +2393,12 @@ +@@ -1732,12 +2393,12 @@ int rt_mutex_finish_proxy_lock(struct rt_mutex *lock, { int ret; @@ -20409,15 +20744,15 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking if (unlikely(ret)) remove_waiter(lock, waiter); -@@ -1748,7 +2409,93 @@ +@@ -1748,7 +2409,93 @@ int rt_mutex_finish_proxy_lock(struct rt_mutex *lock, */ fixup_rt_mutex_waiters(lock); - raw_spin_unlock(&lock->wait_lock); + raw_spin_unlock_irq(&lock->wait_lock); -+ -+ return ret; -+} + + return ret; + } + +static inline int +ww_mutex_deadlock_injection(struct ww_mutex *lock, struct ww_acquire_ctx *ctx) @@ -20459,9 +20794,9 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking + mutex_release(&lock->base.dep_map, 1, _RET_IP_); + else if (!ret && ww_ctx->acquired > 1) + return ww_mutex_deadlock_injection(lock, ww_ctx); - - return ret; - } ++ ++ return ret; ++} +EXPORT_SYMBOL_GPL(__ww_mutex_lock_interruptible); + +int __sched @@ -20504,10 +20839,11 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex.c linux-4.4.62/kernel/locking +} +EXPORT_SYMBOL(ww_mutex_unlock); +#endif -diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex_common.h linux-4.4.62/kernel/locking/rtmutex_common.h ---- linux-4.4.62.orig/kernel/locking/rtmutex_common.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/locking/rtmutex_common.h 2017-04-18 17:38:08.210649865 +0200 -@@ -27,6 +27,7 @@ +diff --git a/kernel/locking/rtmutex_common.h b/kernel/locking/rtmutex_common.h +index e317e1cbb3eb..f457c7574920 100644 +--- a/kernel/locking/rtmutex_common.h ++++ b/kernel/locking/rtmutex_common.h +@@ -27,6 +27,7 @@ struct rt_mutex_waiter { struct rb_node pi_tree_entry; struct task_struct *task; struct rt_mutex *lock; @@ -20515,7 +20851,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex_common.h linux-4.4.62/kernel/ #ifdef CONFIG_DEBUG_RT_MUTEXES unsigned long ip; struct pid *deadlock_task_pid; -@@ -98,6 +99,9 @@ +@@ -98,6 +99,9 @@ enum rtmutex_chainwalk { /* * PI-futex support (proxy locking functions, etc.): */ @@ -20525,7 +20861,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex_common.h linux-4.4.62/kernel/ extern struct task_struct *rt_mutex_next_owner(struct rt_mutex *lock); extern void rt_mutex_init_proxy_locked(struct rt_mutex *lock, struct task_struct *proxy_owner); -@@ -111,7 +115,8 @@ +@@ -111,7 +115,8 @@ extern int rt_mutex_finish_proxy_lock(struct rt_mutex *lock, struct rt_mutex_waiter *waiter); extern int rt_mutex_timed_futex_lock(struct rt_mutex *l, struct hrtimer_sleeper *to); extern bool rt_mutex_futex_unlock(struct rt_mutex *lock, @@ -20535,7 +20871,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex_common.h linux-4.4.62/kernel/ extern void rt_mutex_adjust_prio(struct task_struct *task); #ifdef CONFIG_DEBUG_RT_MUTEXES -@@ -120,4 +125,14 @@ +@@ -120,4 +125,14 @@ extern void rt_mutex_adjust_prio(struct task_struct *task); # include "rtmutex.h" #endif @@ -20550,10 +20886,11 @@ diff -Nur linux-4.4.62.orig/kernel/locking/rtmutex_common.h linux-4.4.62/kernel/ +} + #endif -diff -Nur linux-4.4.62.orig/kernel/locking/spinlock.c linux-4.4.62/kernel/locking/spinlock.c ---- linux-4.4.62.orig/kernel/locking/spinlock.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/locking/spinlock.c 2017-04-18 17:38:08.210649865 +0200 -@@ -124,8 +124,11 @@ +diff --git a/kernel/locking/spinlock.c b/kernel/locking/spinlock.c +index db3ccb1dd614..909779647bd1 100644 +--- a/kernel/locking/spinlock.c ++++ b/kernel/locking/spinlock.c +@@ -124,8 +124,11 @@ void __lockfunc __raw_##op##_lock_bh(locktype##_t *lock) \ * __[spin|read|write]_lock_bh() */ BUILD_LOCK_OPS(spin, raw_spinlock); @@ -20565,7 +20902,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/spinlock.c linux-4.4.62/kernel/lockin #endif -@@ -209,6 +212,8 @@ +@@ -209,6 +212,8 @@ void __lockfunc _raw_spin_unlock_bh(raw_spinlock_t *lock) EXPORT_SYMBOL(_raw_spin_unlock_bh); #endif @@ -20574,7 +20911,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/spinlock.c linux-4.4.62/kernel/lockin #ifndef CONFIG_INLINE_READ_TRYLOCK int __lockfunc _raw_read_trylock(rwlock_t *lock) { -@@ -353,6 +358,8 @@ +@@ -353,6 +358,8 @@ void __lockfunc _raw_write_unlock_bh(rwlock_t *lock) EXPORT_SYMBOL(_raw_write_unlock_bh); #endif @@ -20583,10 +20920,11 @@ diff -Nur linux-4.4.62.orig/kernel/locking/spinlock.c linux-4.4.62/kernel/lockin #ifdef CONFIG_DEBUG_LOCK_ALLOC void __lockfunc _raw_spin_lock_nested(raw_spinlock_t *lock, int subclass) -diff -Nur linux-4.4.62.orig/kernel/locking/spinlock_debug.c linux-4.4.62/kernel/locking/spinlock_debug.c ---- linux-4.4.62.orig/kernel/locking/spinlock_debug.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/locking/spinlock_debug.c 2017-04-18 17:38:08.210649865 +0200 -@@ -31,6 +31,7 @@ +diff --git a/kernel/locking/spinlock_debug.c b/kernel/locking/spinlock_debug.c +index 0374a596cffa..94970338d518 100644 +--- a/kernel/locking/spinlock_debug.c ++++ b/kernel/locking/spinlock_debug.c +@@ -31,6 +31,7 @@ void __raw_spin_lock_init(raw_spinlock_t *lock, const char *name, EXPORT_SYMBOL(__raw_spin_lock_init); @@ -20594,7 +20932,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/spinlock_debug.c linux-4.4.62/kernel/ void __rwlock_init(rwlock_t *lock, const char *name, struct lock_class_key *key) { -@@ -48,6 +49,7 @@ +@@ -48,6 +49,7 @@ void __rwlock_init(rwlock_t *lock, const char *name, } EXPORT_SYMBOL(__rwlock_init); @@ -20602,7 +20940,7 @@ diff -Nur linux-4.4.62.orig/kernel/locking/spinlock_debug.c linux-4.4.62/kernel/ static void spin_dump(raw_spinlock_t *lock, const char *msg) { -@@ -159,6 +161,7 @@ +@@ -159,6 +161,7 @@ void do_raw_spin_unlock(raw_spinlock_t *lock) arch_spin_unlock(&lock->raw_lock); } @@ -20610,16 +20948,17 @@ diff -Nur linux-4.4.62.orig/kernel/locking/spinlock_debug.c linux-4.4.62/kernel/ static void rwlock_bug(rwlock_t *lock, const char *msg) { if (!debug_locks_off()) -@@ -300,3 +303,5 @@ +@@ -300,3 +303,5 @@ void do_raw_write_unlock(rwlock_t *lock) debug_write_unlock(lock); arch_write_unlock(&lock->raw_lock); } + +#endif -diff -Nur linux-4.4.62.orig/kernel/module.c linux-4.4.62/kernel/module.c ---- linux-4.4.62.orig/kernel/module.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/module.c 2017-04-18 17:38:08.210649865 +0200 -@@ -682,16 +682,7 @@ +diff --git a/kernel/module.c b/kernel/module.c +index b14a4f31221f..be3f497a089d 100644 +--- a/kernel/module.c ++++ b/kernel/module.c +@@ -682,16 +682,7 @@ static void percpu_modcopy(struct module *mod, memcpy(per_cpu_ptr(mod->percpu, cpu), from, size); } @@ -20637,7 +20976,7 @@ diff -Nur linux-4.4.62.orig/kernel/module.c linux-4.4.62/kernel/module.c { struct module *mod; unsigned int cpu; -@@ -705,9 +696,11 @@ +@@ -705,9 +696,11 @@ bool is_module_percpu_address(unsigned long addr) continue; for_each_possible_cpu(cpu) { void *start = per_cpu_ptr(mod->percpu, cpu); @@ -20651,7 +20990,7 @@ diff -Nur linux-4.4.62.orig/kernel/module.c linux-4.4.62/kernel/module.c preempt_enable(); return true; } -@@ -718,6 +711,20 @@ +@@ -718,6 +711,20 @@ bool is_module_percpu_address(unsigned long addr) return false; } @@ -20672,22 +21011,23 @@ diff -Nur linux-4.4.62.orig/kernel/module.c linux-4.4.62/kernel/module.c #else /* ... !CONFIG_SMP */ static inline void __percpu *mod_percpu(struct module *mod) -@@ -748,6 +755,11 @@ - { +@@ -749,6 +756,11 @@ bool is_module_percpu_address(unsigned long addr) return false; } -+ + +bool __is_module_percpu_address(unsigned long addr, unsigned long *can_addr) +{ + return false; +} - ++ #endif /* CONFIG_SMP */ -diff -Nur linux-4.4.62.orig/kernel/panic.c linux-4.4.62/kernel/panic.c ---- linux-4.4.62.orig/kernel/panic.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/panic.c 2017-04-18 17:38:08.210649865 +0200 -@@ -61,6 +61,37 @@ + #define MODINFO_ATTR(field) \ +diff --git a/kernel/panic.c b/kernel/panic.c +index 41e2b54f36b5..3535f802953a 100644 +--- a/kernel/panic.c ++++ b/kernel/panic.c +@@ -61,6 +61,37 @@ void __weak panic_smp_self_stop(void) cpu_relax(); } @@ -20725,7 +21065,7 @@ diff -Nur linux-4.4.62.orig/kernel/panic.c linux-4.4.62/kernel/panic.c /** * panic - halt the system * @fmt: The text string to print -@@ -71,17 +102,17 @@ +@@ -71,17 +102,17 @@ void __weak panic_smp_self_stop(void) */ void panic(const char *fmt, ...) { @@ -20745,7 +21085,7 @@ diff -Nur linux-4.4.62.orig/kernel/panic.c linux-4.4.62/kernel/panic.c */ local_irq_disable(); -@@ -94,8 +125,16 @@ +@@ -94,8 +125,16 @@ void panic(const char *fmt, ...) * multiple parallel invocations of panic, all other CPUs either * stop themself or will wait until they are stopped by the 1st CPU * with smp_send_stop(). @@ -20763,7 +21103,7 @@ diff -Nur linux-4.4.62.orig/kernel/panic.c linux-4.4.62/kernel/panic.c panic_smp_self_stop(); console_verbose(); -@@ -400,9 +439,11 @@ +@@ -400,9 +439,11 @@ static u64 oops_id; static int init_oops_id(void) { @@ -20775,10 +21115,11 @@ diff -Nur linux-4.4.62.orig/kernel/panic.c linux-4.4.62/kernel/panic.c oops_id++; return 0; -diff -Nur linux-4.4.62.orig/kernel/power/hibernate.c linux-4.4.62/kernel/power/hibernate.c ---- linux-4.4.62.orig/kernel/power/hibernate.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/power/hibernate.c 2017-04-18 17:38:08.210649865 +0200 -@@ -285,6 +285,8 @@ +diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c +index 3124cebaec31..c1b981521dd0 100644 +--- a/kernel/power/hibernate.c ++++ b/kernel/power/hibernate.c +@@ -285,6 +285,8 @@ static int create_image(int platform_mode) local_irq_disable(); @@ -20787,7 +21128,7 @@ diff -Nur linux-4.4.62.orig/kernel/power/hibernate.c linux-4.4.62/kernel/power/h error = syscore_suspend(); if (error) { printk(KERN_ERR "PM: Some system devices failed to power down, " -@@ -314,6 +316,7 @@ +@@ -314,6 +316,7 @@ static int create_image(int platform_mode) syscore_resume(); Enable_irqs: @@ -20795,7 +21136,7 @@ diff -Nur linux-4.4.62.orig/kernel/power/hibernate.c linux-4.4.62/kernel/power/h local_irq_enable(); Enable_cpus: -@@ -438,6 +441,7 @@ +@@ -438,6 +441,7 @@ static int resume_target_kernel(bool platform_mode) goto Enable_cpus; local_irq_disable(); @@ -20803,7 +21144,7 @@ diff -Nur linux-4.4.62.orig/kernel/power/hibernate.c linux-4.4.62/kernel/power/h error = syscore_suspend(); if (error) -@@ -471,6 +475,7 @@ +@@ -471,6 +475,7 @@ static int resume_target_kernel(bool platform_mode) syscore_resume(); Enable_irqs: @@ -20811,7 +21152,7 @@ diff -Nur linux-4.4.62.orig/kernel/power/hibernate.c linux-4.4.62/kernel/power/h local_irq_enable(); Enable_cpus: -@@ -556,6 +561,7 @@ +@@ -556,6 +561,7 @@ int hibernation_platform_enter(void) goto Enable_cpus; local_irq_disable(); @@ -20819,7 +21160,7 @@ diff -Nur linux-4.4.62.orig/kernel/power/hibernate.c linux-4.4.62/kernel/power/h syscore_suspend(); if (pm_wakeup_pending()) { error = -EAGAIN; -@@ -568,6 +574,7 @@ +@@ -568,6 +574,7 @@ int hibernation_platform_enter(void) Power_up: syscore_resume(); @@ -20827,7 +21168,7 @@ diff -Nur linux-4.4.62.orig/kernel/power/hibernate.c linux-4.4.62/kernel/power/h local_irq_enable(); Enable_cpus: -@@ -642,6 +649,10 @@ +@@ -642,6 +649,10 @@ static void power_down(void) cpu_relax(); } @@ -20838,7 +21179,7 @@ diff -Nur linux-4.4.62.orig/kernel/power/hibernate.c linux-4.4.62/kernel/power/h /** * hibernate - Carry out system hibernation, including saving the image. */ -@@ -654,6 +665,8 @@ +@@ -654,6 +665,8 @@ int hibernate(void) return -EPERM; } @@ -20847,7 +21188,7 @@ diff -Nur linux-4.4.62.orig/kernel/power/hibernate.c linux-4.4.62/kernel/power/h lock_system_sleep(); /* The snapshot device should not be opened while we're running */ if (!atomic_add_unless(&snapshot_device_available, -1, 0)) { -@@ -719,6 +732,7 @@ +@@ -719,6 +732,7 @@ int hibernate(void) atomic_inc(&snapshot_device_available); Unlock: unlock_system_sleep(); @@ -20855,10 +21196,11 @@ diff -Nur linux-4.4.62.orig/kernel/power/hibernate.c linux-4.4.62/kernel/power/h return error; } -diff -Nur linux-4.4.62.orig/kernel/power/suspend.c linux-4.4.62/kernel/power/suspend.c ---- linux-4.4.62.orig/kernel/power/suspend.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/power/suspend.c 2017-04-18 17:38:08.210649865 +0200 -@@ -359,6 +359,8 @@ +diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c +index f9fe133c13e2..393bc342c586 100644 +--- a/kernel/power/suspend.c ++++ b/kernel/power/suspend.c +@@ -359,6 +359,8 @@ static int suspend_enter(suspend_state_t state, bool *wakeup) arch_suspend_disable_irqs(); BUG_ON(!irqs_disabled()); @@ -20867,7 +21209,7 @@ diff -Nur linux-4.4.62.orig/kernel/power/suspend.c linux-4.4.62/kernel/power/sus error = syscore_suspend(); if (!error) { *wakeup = pm_wakeup_pending(); -@@ -375,6 +377,8 @@ +@@ -375,6 +377,8 @@ static int suspend_enter(suspend_state_t state, bool *wakeup) syscore_resume(); } @@ -20876,7 +21218,7 @@ diff -Nur linux-4.4.62.orig/kernel/power/suspend.c linux-4.4.62/kernel/power/sus arch_suspend_enable_irqs(); BUG_ON(irqs_disabled()); -@@ -518,6 +522,8 @@ +@@ -518,6 +522,8 @@ static int enter_state(suspend_state_t state) return error; } @@ -20885,7 +21227,7 @@ diff -Nur linux-4.4.62.orig/kernel/power/suspend.c linux-4.4.62/kernel/power/sus /** * pm_suspend - Externally visible function for suspending the system. * @state: System sleep state to enter. -@@ -532,6 +538,8 @@ +@@ -532,6 +538,8 @@ int pm_suspend(suspend_state_t state) if (state <= PM_SUSPEND_ON || state >= PM_SUSPEND_MAX) return -EINVAL; @@ -20894,7 +21236,7 @@ diff -Nur linux-4.4.62.orig/kernel/power/suspend.c linux-4.4.62/kernel/power/sus error = enter_state(state); if (error) { suspend_stats.fail++; -@@ -539,6 +547,7 @@ +@@ -539,6 +547,7 @@ int pm_suspend(suspend_state_t state) } else { suspend_stats.success++; } @@ -20902,10 +21244,11 @@ diff -Nur linux-4.4.62.orig/kernel/power/suspend.c linux-4.4.62/kernel/power/sus return error; } EXPORT_SYMBOL(pm_suspend); -diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/printk.c ---- linux-4.4.62.orig/kernel/printk/printk.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/printk/printk.c 2017-04-18 17:38:08.214650020 +0200 -@@ -241,6 +241,65 @@ +diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c +index 0b5613554769..99deb2617308 100644 +--- a/kernel/printk/printk.c ++++ b/kernel/printk/printk.c +@@ -241,6 +241,65 @@ struct printk_log { */ static DEFINE_RAW_SPINLOCK(logbuf_lock); @@ -20971,7 +21314,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr #ifdef CONFIG_PRINTK DECLARE_WAIT_QUEUE_HEAD(log_wait); /* the next printk record to read by syslog(READ) or /proc/kmsg */ -@@ -1203,6 +1262,7 @@ +@@ -1203,6 +1262,7 @@ static int syslog_print_all(char __user *buf, int size, bool clear) { char *text; int len = 0; @@ -20979,7 +21322,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr text = kmalloc(LOG_LINE_MAX + PREFIX_MAX, GFP_KERNEL); if (!text) -@@ -1214,7 +1274,14 @@ +@@ -1214,7 +1274,14 @@ static int syslog_print_all(char __user *buf, int size, bool clear) u64 seq; u32 idx; enum log_flags prev; @@ -20995,7 +21338,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr if (clear_seq < log_first_seq) { /* messages are gone, move to first available one */ clear_seq = log_first_seq; -@@ -1235,6 +1302,14 @@ +@@ -1235,6 +1302,14 @@ static int syslog_print_all(char __user *buf, int size, bool clear) prev = msg->flags; idx = log_next(idx); seq++; @@ -21010,7 +21353,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr } /* move first record forward until length fits into the buffer */ -@@ -1248,6 +1323,14 @@ +@@ -1248,6 +1323,14 @@ static int syslog_print_all(char __user *buf, int size, bool clear) prev = msg->flags; idx = log_next(idx); seq++; @@ -21025,7 +21368,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr } /* last message fitting into this dump */ -@@ -1288,6 +1371,7 @@ +@@ -1288,6 +1371,7 @@ static int syslog_print_all(char __user *buf, int size, bool clear) clear_seq = log_next_seq; clear_idx = log_next_idx; } @@ -21033,7 +21376,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr raw_spin_unlock_irq(&logbuf_lock); kfree(text); -@@ -1443,6 +1527,12 @@ +@@ -1443,6 +1527,12 @@ static void call_console_drivers(int level, if (!console_drivers) return; @@ -21046,7 +21389,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr for_each_console(con) { if (exclusive_console && con != exclusive_console) continue; -@@ -1458,6 +1548,7 @@ +@@ -1458,6 +1548,7 @@ static void call_console_drivers(int level, else con->write(con, text, len); } @@ -21054,7 +21397,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr } /* -@@ -1518,6 +1609,15 @@ +@@ -1518,6 +1609,15 @@ static inline int can_use_console(unsigned int cpu) static int console_trylock_for_printk(void) { unsigned int cpu = smp_processor_id(); @@ -21070,7 +21413,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr if (!console_trylock()) return 0; -@@ -1672,6 +1772,13 @@ +@@ -1672,6 +1772,13 @@ asmlinkage int vprintk_emit(int facility, int level, /* cpu currently holding logbuf_lock in this function */ static unsigned int logbuf_cpu = UINT_MAX; @@ -21084,7 +21427,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr if (level == LOGLEVEL_SCHED) { level = LOGLEVEL_DEFAULT; in_sched = true; -@@ -1813,8 +1920,7 @@ +@@ -1813,8 +1920,7 @@ asmlinkage int vprintk_emit(int facility, int level, * console_sem which would prevent anyone from printing to * console */ @@ -21094,7 +21437,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr /* * Try to acquire and then immediately release the console * semaphore. The release will print out buffers and wake up -@@ -1822,7 +1928,7 @@ +@@ -1822,7 +1928,7 @@ asmlinkage int vprintk_emit(int facility, int level, */ if (console_trylock_for_printk()) console_unlock(); @@ -21103,7 +21446,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr lockdep_on(); } -@@ -1961,26 +2067,6 @@ +@@ -1961,26 +2067,6 @@ DEFINE_PER_CPU(printk_func_t, printk_func); #endif /* CONFIG_PRINTK */ @@ -21130,7 +21473,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr static int __add_preferred_console(char *name, int idx, char *options, char *brl_options) { -@@ -2202,11 +2288,16 @@ +@@ -2202,11 +2288,16 @@ static void console_cont_flush(char *text, size_t size) goto out; len = cont_print_text(text, size); @@ -21147,7 +21490,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr return; out: raw_spin_unlock_irqrestore(&logbuf_lock, flags); -@@ -2316,13 +2407,17 @@ +@@ -2316,13 +2407,17 @@ skip: console_idx = log_next(console_idx); console_seq++; console_prev = msg->flags; @@ -21166,7 +21509,7 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr if (do_cond_resched) cond_resched(); } -@@ -2374,6 +2469,11 @@ +@@ -2374,6 +2469,11 @@ void console_unblank(void) { struct console *c; @@ -21178,10 +21521,11 @@ diff -Nur linux-4.4.62.orig/kernel/printk/printk.c linux-4.4.62/kernel/printk/pr /* * console_unblank can no longer be called in interrupt context unless * oops_in_progress is set to 1.. -diff -Nur linux-4.4.62.orig/kernel/ptrace.c linux-4.4.62/kernel/ptrace.c ---- linux-4.4.62.orig/kernel/ptrace.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/ptrace.c 2017-04-18 17:38:08.214650020 +0200 -@@ -136,7 +136,14 @@ +diff --git a/kernel/ptrace.c b/kernel/ptrace.c +index c7e8ed99c953..2a19cc5001b4 100644 +--- a/kernel/ptrace.c ++++ b/kernel/ptrace.c +@@ -136,7 +136,14 @@ static bool ptrace_freeze_traced(struct task_struct *task) spin_lock_irq(&task->sighand->siglock); if (task_is_traced(task) && !__fatal_signal_pending(task)) { @@ -21197,10 +21541,11 @@ diff -Nur linux-4.4.62.orig/kernel/ptrace.c linux-4.4.62/kernel/ptrace.c ret = true; } spin_unlock_irq(&task->sighand->siglock); -diff -Nur linux-4.4.62.orig/kernel/rcu/rcutorture.c linux-4.4.62/kernel/rcu/rcutorture.c ---- linux-4.4.62.orig/kernel/rcu/rcutorture.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/rcu/rcutorture.c 2017-04-18 17:38:08.214650020 +0200 -@@ -390,6 +390,7 @@ +diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c +index d89328e260df..5bb3364a6284 100644 +--- a/kernel/rcu/rcutorture.c ++++ b/kernel/rcu/rcutorture.c +@@ -390,6 +390,7 @@ static struct rcu_torture_ops rcu_ops = { .name = "rcu" }; @@ -21208,7 +21553,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/rcutorture.c linux-4.4.62/kernel/rcu/rcut /* * Definitions for rcu_bh torture testing. */ -@@ -429,6 +430,12 @@ +@@ -429,6 +430,12 @@ static struct rcu_torture_ops rcu_bh_ops = { .name = "rcu_bh" }; @@ -21221,9 +21566,10 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/rcutorture.c linux-4.4.62/kernel/rcu/rcut /* * Don't even think about trying any of these in real life!!! * The names includes "busted", and they really means it! -diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c ---- linux-4.4.62.orig/kernel/rcu/tree.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/rcu/tree.c 2017-04-18 17:38:08.214650020 +0200 +diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c +index f07343b54fe5..d862a203fce0 100644 +--- a/kernel/rcu/tree.c ++++ b/kernel/rcu/tree.c @@ -56,6 +56,11 @@ #include #include @@ -21236,7 +21582,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c #include "tree.h" #include "rcu.h" -@@ -266,6 +271,19 @@ +@@ -266,6 +271,19 @@ void rcu_sched_qs(void) } } @@ -21256,7 +21602,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c void rcu_bh_qs(void) { if (__this_cpu_read(rcu_bh_data.cpu_no_qs.s)) { -@@ -275,6 +293,7 @@ +@@ -275,6 +293,7 @@ void rcu_bh_qs(void) __this_cpu_write(rcu_bh_data.cpu_no_qs.b.norm, false); } } @@ -21264,7 +21610,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c static DEFINE_PER_CPU(int, rcu_sched_qs_mask); -@@ -435,11 +454,13 @@ +@@ -435,11 +454,13 @@ EXPORT_SYMBOL_GPL(rcu_batches_started_sched); /* * Return the number of RCU BH batches started thus far for debug & stats. */ @@ -21278,7 +21624,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c /* * Return the number of RCU batches completed thus far for debug & stats. -@@ -459,6 +480,7 @@ +@@ -459,6 +480,7 @@ unsigned long rcu_batches_completed_sched(void) } EXPORT_SYMBOL_GPL(rcu_batches_completed_sched); @@ -21286,7 +21632,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c /* * Return the number of RCU BH batches completed thus far for debug & stats. */ -@@ -486,6 +508,13 @@ +@@ -486,6 +508,13 @@ void rcu_bh_force_quiescent_state(void) } EXPORT_SYMBOL_GPL(rcu_bh_force_quiescent_state); @@ -21300,7 +21646,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c /* * Force a quiescent state for RCU-sched. */ -@@ -536,9 +565,11 @@ +@@ -536,9 +565,11 @@ void rcutorture_get_gp_data(enum rcutorture_type test_type, int *flags, case RCU_FLAVOR: rsp = rcu_state_p; break; @@ -21312,7 +21658,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c case RCU_SCHED_FLAVOR: rsp = &rcu_sched_state; break; -@@ -1590,7 +1621,6 @@ +@@ -1590,7 +1621,6 @@ static int rcu_future_gp_cleanup(struct rcu_state *rsp, struct rcu_node *rnp) int needmore; struct rcu_data *rdp = this_cpu_ptr(rsp->rda); @@ -21320,7 +21666,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c rnp->need_future_gp[c & 0x1] = 0; needmore = rnp->need_future_gp[(c + 1) & 0x1]; trace_rcu_future_gp(rnp, rdp, c, -@@ -1611,7 +1641,7 @@ +@@ -1611,7 +1641,7 @@ static void rcu_gp_kthread_wake(struct rcu_state *rsp) !READ_ONCE(rsp->gp_flags) || !rsp->gp_kthread) return; @@ -21329,7 +21675,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c } /* -@@ -1991,6 +2021,7 @@ +@@ -1991,6 +2021,7 @@ static void rcu_gp_cleanup(struct rcu_state *rsp) int nocb = 0; struct rcu_data *rdp; struct rcu_node *rnp = rcu_get_root(rsp); @@ -21337,7 +21683,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c WRITE_ONCE(rsp->gp_activity, jiffies); raw_spin_lock_irq(&rnp->lock); -@@ -2029,7 +2060,9 @@ +@@ -2029,7 +2060,9 @@ static void rcu_gp_cleanup(struct rcu_state *rsp) needgp = __note_gp_changes(rsp, rnp, rdp) || needgp; /* smp_mb() provided by prior unlock-lock pair. */ nocb += rcu_future_gp_cleanup(rsp, rnp); @@ -21347,7 +21693,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c cond_resched_rcu_qs(); WRITE_ONCE(rsp->gp_activity, jiffies); rcu_gp_slow(rsp, gp_cleanup_delay); -@@ -2076,7 +2109,7 @@ +@@ -2076,7 +2109,7 @@ static int __noreturn rcu_gp_kthread(void *arg) READ_ONCE(rsp->gpnum), TPS("reqwait")); rsp->gp_state = RCU_GP_WAIT_GPS; @@ -21356,7 +21702,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c READ_ONCE(rsp->gp_flags) & RCU_GP_FLAG_INIT); rsp->gp_state = RCU_GP_DONE_GPS; -@@ -2106,7 +2139,7 @@ +@@ -2106,7 +2139,7 @@ static int __noreturn rcu_gp_kthread(void *arg) READ_ONCE(rsp->gpnum), TPS("fqswait")); rsp->gp_state = RCU_GP_WAIT_FQS; @@ -21365,7 +21711,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c rcu_gp_fqs_check_wake(rsp, &gf), j); rsp->gp_state = RCU_GP_DOING_FQS; /* Locking provides needed memory barriers. */ -@@ -2230,7 +2263,7 @@ +@@ -2230,7 +2263,7 @@ static void rcu_report_qs_rsp(struct rcu_state *rsp, unsigned long flags) WARN_ON_ONCE(!rcu_gp_in_progress(rsp)); WRITE_ONCE(rsp->gp_flags, READ_ONCE(rsp->gp_flags) | RCU_GP_FLAG_FQS); raw_spin_unlock_irqrestore(&rcu_get_root(rsp)->lock, flags); @@ -21374,7 +21720,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c } /* -@@ -2891,7 +2924,7 @@ +@@ -2891,7 +2924,7 @@ static void force_quiescent_state(struct rcu_state *rsp) } WRITE_ONCE(rsp->gp_flags, READ_ONCE(rsp->gp_flags) | RCU_GP_FLAG_FQS); raw_spin_unlock_irqrestore(&rnp_old->lock, flags); @@ -21383,7 +21729,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c } /* -@@ -2934,18 +2967,17 @@ +@@ -2934,18 +2967,17 @@ __rcu_process_callbacks(struct rcu_state *rsp) /* * Do RCU core processing for the current CPU. */ @@ -21404,15 +21750,18 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c /* * Schedule RCU callback invocation. If the specified type of RCU * does not support RCU priority boosting, just do a direct call, -@@ -2957,18 +2989,105 @@ +@@ -2957,18 +2989,105 @@ static void invoke_rcu_callbacks(struct rcu_state *rsp, struct rcu_data *rdp) { if (unlikely(!READ_ONCE(rcu_scheduler_fully_active))) return; - if (likely(!rsp->boost)) { - rcu_do_batch(rsp, rdp); +- return; +- } +- invoke_rcu_callbacks_kthread(); + rcu_do_batch(rsp, rdp); -+} -+ + } + +static void rcu_wake_cond(struct task_struct *t, int status) +{ + /* @@ -21426,13 +21775,15 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c +/* + * Wake up this CPU's rcuc kthread to do RCU core processing. + */ -+static void invoke_rcu_core(void) -+{ + static void invoke_rcu_core(void) + { +- if (cpu_online(smp_processor_id())) +- raise_softirq(RCU_SOFTIRQ); + unsigned long flags; + struct task_struct *t; + + if (!cpu_online(smp_processor_id())) - return; ++ return; + local_irq_save(flags); + __this_cpu_write(rcu_cpu_has_work, 1); + t = __this_cpu_read(rcu_cpu_kthread_task); @@ -21479,16 +21830,14 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c + *statusp = RCU_KTHREAD_WAITING; + return; + } - } -- invoke_rcu_callbacks_kthread(); ++ } + *statusp = RCU_KTHREAD_YIELDING; + trace_rcu_utilization(TPS("Start CPU kthread@rcu_yield")); + schedule_timeout_interruptible(2); + trace_rcu_utilization(TPS("End CPU kthread@rcu_yield")); + *statusp = RCU_KTHREAD_WAITING; - } - --static void invoke_rcu_core(void) ++} ++ +static struct smp_hotplug_thread rcu_cpu_thread_spec = { + .store = &rcu_cpu_kthread_task, + .thread_should_run = rcu_cpu_kthread_should_run, @@ -21502,9 +21851,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c + * Spawn per-CPU RCU core processing kthreads. + */ +static int __init rcu_spawn_core_kthreads(void) - { -- if (cpu_online(smp_processor_id())) -- raise_softirq(RCU_SOFTIRQ); ++{ + int cpu; + + for_each_possible_cpu(cpu) @@ -21516,7 +21863,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c /* * Handle any core-RCU processing required by a call_rcu() invocation. -@@ -3114,6 +3233,7 @@ +@@ -3114,6 +3233,7 @@ void call_rcu_sched(struct rcu_head *head, rcu_callback_t func) } EXPORT_SYMBOL_GPL(call_rcu_sched); @@ -21524,7 +21871,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c /* * Queue an RCU callback for invocation after a quicker grace period. */ -@@ -3122,6 +3242,7 @@ +@@ -3122,6 +3242,7 @@ void call_rcu_bh(struct rcu_head *head, rcu_callback_t func) __call_rcu(head, func, &rcu_bh_state, -1, 0); } EXPORT_SYMBOL_GPL(call_rcu_bh); @@ -21532,7 +21879,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c /* * Queue an RCU callback for lazy invocation after a grace period. -@@ -3213,6 +3334,7 @@ +@@ -3213,6 +3334,7 @@ void synchronize_sched(void) } EXPORT_SYMBOL_GPL(synchronize_sched); @@ -21540,7 +21887,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c /** * synchronize_rcu_bh - wait until an rcu_bh grace period has elapsed. * -@@ -3239,6 +3361,7 @@ +@@ -3239,6 +3361,7 @@ void synchronize_rcu_bh(void) wait_rcu_gp(call_rcu_bh); } EXPORT_SYMBOL_GPL(synchronize_rcu_bh); @@ -21548,7 +21895,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c /** * get_state_synchronize_rcu - Snapshot current RCU state -@@ -3524,7 +3647,7 @@ +@@ -3524,7 +3647,7 @@ static void __rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp, raw_spin_unlock_irqrestore(&rnp->lock, flags); if (wake) { smp_mb(); /* EGP done before wake_up(). */ @@ -21557,7 +21904,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c } break; } -@@ -3781,7 +3904,7 @@ +@@ -3781,7 +3904,7 @@ static void synchronize_sched_expedited_wait(struct rcu_state *rsp) jiffies_start = jiffies; for (;;) { @@ -21566,7 +21913,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c rsp->expedited_wq, sync_rcu_preempt_exp_done(rnp_root), jiffies_stall); -@@ -3789,7 +3912,7 @@ +@@ -3789,7 +3912,7 @@ static void synchronize_sched_expedited_wait(struct rcu_state *rsp) return; if (ret < 0) { /* Hit a signal, disable CPU stall warnings. */ @@ -21575,7 +21922,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c sync_rcu_preempt_exp_done(rnp_root)); return; } -@@ -4101,6 +4224,7 @@ +@@ -4101,6 +4224,7 @@ static void _rcu_barrier(struct rcu_state *rsp) mutex_unlock(&rsp->barrier_mutex); } @@ -21583,7 +21930,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c /** * rcu_barrier_bh - Wait until all in-flight call_rcu_bh() callbacks complete. */ -@@ -4109,6 +4233,7 @@ +@@ -4109,6 +4233,7 @@ void rcu_barrier_bh(void) _rcu_barrier(&rcu_bh_state); } EXPORT_SYMBOL_GPL(rcu_barrier_bh); @@ -21591,7 +21938,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c /** * rcu_barrier_sched - Wait for in-flight call_rcu_sched() callbacks. -@@ -4455,8 +4580,8 @@ +@@ -4455,8 +4580,8 @@ static void __init rcu_init_one(struct rcu_state *rsp, } } @@ -21602,7 +21949,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c rnp = rsp->level[rcu_num_lvls - 1]; for_each_possible_cpu(i) { while (i > rnp->grphi) -@@ -4576,12 +4701,13 @@ +@@ -4576,12 +4701,13 @@ void __init rcu_init(void) rcu_bootup_announce(); rcu_init_geometry(); @@ -21617,9 +21964,10 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.c linux-4.4.62/kernel/rcu/tree.c /* * We don't need protection against CPU-hotplug here because -diff -Nur linux-4.4.62.orig/kernel/rcu/tree.h linux-4.4.62/kernel/rcu/tree.h ---- linux-4.4.62.orig/kernel/rcu/tree.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/rcu/tree.h 2017-04-18 17:38:08.214650020 +0200 +diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h +index 9fb4e238d4dc..c75834d8de24 100644 +--- a/kernel/rcu/tree.h ++++ b/kernel/rcu/tree.h @@ -27,6 +27,7 @@ #include #include @@ -21628,7 +21976,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.h linux-4.4.62/kernel/rcu/tree.h #include /* -@@ -241,7 +242,7 @@ +@@ -241,7 +242,7 @@ struct rcu_node { /* Refused to boost: not sure why, though. */ /* This can happen due to race conditions. */ #ifdef CONFIG_RCU_NOCB_CPU @@ -21637,7 +21985,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.h linux-4.4.62/kernel/rcu/tree.h /* Place for rcu_nocb_kthread() to wait GP. */ #endif /* #ifdef CONFIG_RCU_NOCB_CPU */ int need_future_gp[2]; -@@ -393,7 +394,7 @@ +@@ -393,7 +394,7 @@ struct rcu_data { atomic_long_t nocb_q_count_lazy; /* invocation (all stages). */ struct rcu_head *nocb_follower_head; /* CBs ready to invoke. */ struct rcu_head **nocb_follower_tail; @@ -21646,7 +21994,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.h linux-4.4.62/kernel/rcu/tree.h struct task_struct *nocb_kthread; int nocb_defer_wakeup; /* Defer wakeup of nocb_kthread. */ -@@ -472,7 +473,7 @@ +@@ -472,7 +473,7 @@ struct rcu_state { unsigned long gpnum; /* Current gp number. */ unsigned long completed; /* # of last completed gp. */ struct task_struct *gp_kthread; /* Task for grace periods. */ @@ -21655,7 +22003,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.h linux-4.4.62/kernel/rcu/tree.h short gp_flags; /* Commands for GP task. */ short gp_state; /* GP kthread sleep state. */ -@@ -504,7 +505,7 @@ +@@ -504,7 +505,7 @@ struct rcu_state { atomic_long_t expedited_workdone3; /* # done by others #3. */ atomic_long_t expedited_normal; /* # fallbacks to normal. */ atomic_t expedited_need_qs; /* # CPUs left to check in. */ @@ -21664,7 +22012,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.h linux-4.4.62/kernel/rcu/tree.h int ncpus_snap; /* # CPUs seen last time. */ unsigned long jiffies_force_qs; /* Time at which to invoke */ -@@ -556,18 +557,18 @@ +@@ -556,18 +557,18 @@ extern struct list_head rcu_struct_flavors; */ extern struct rcu_state rcu_sched_state; @@ -21685,7 +22033,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.h linux-4.4.62/kernel/rcu/tree.h #ifndef RCU_TREE_NONCORE -@@ -587,10 +588,9 @@ +@@ -587,10 +588,9 @@ void call_rcu(struct rcu_head *head, rcu_callback_t func); static void __init __rcu_init_preempt(void); static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags); static void rcu_preempt_boost_start_gp(struct rcu_node *rnp); @@ -21697,7 +22045,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.h linux-4.4.62/kernel/rcu/tree.h static int rcu_spawn_one_boost_kthread(struct rcu_state *rsp, struct rcu_node *rnp); #endif /* #ifdef CONFIG_RCU_BOOST */ -@@ -607,7 +607,8 @@ +@@ -607,7 +607,8 @@ static void zero_cpu_stall_ticks(struct rcu_data *rdp); static void increment_cpu_stall_ticks(void); static bool rcu_nocb_cpu_needs_barrier(struct rcu_state *rsp, int cpu); static void rcu_nocb_gp_set(struct rcu_node *rnp, int nrq); @@ -21707,9 +22055,10 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree.h linux-4.4.62/kernel/rcu/tree.h static void rcu_init_one_nocb(struct rcu_node *rnp); static bool __call_rcu_nocb(struct rcu_data *rdp, struct rcu_head *rhp, bool lazy, unsigned long flags); -diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tree_plugin.h ---- linux-4.4.62.orig/kernel/rcu/tree_plugin.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/rcu/tree_plugin.h 2017-04-18 17:38:08.214650020 +0200 +diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h +index 32cbe72bf545..45e3e3e02a5c 100644 +--- a/kernel/rcu/tree_plugin.h ++++ b/kernel/rcu/tree_plugin.h @@ -24,25 +24,10 @@ * Paul E. McKenney */ @@ -21736,7 +22085,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre #else /* #ifdef CONFIG_RCU_BOOST */ /* -@@ -55,6 +40,14 @@ +@@ -55,6 +40,14 @@ DEFINE_PER_CPU(char, rcu_cpu_has_work); #endif /* #else #ifdef CONFIG_RCU_BOOST */ @@ -21751,7 +22100,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre #ifdef CONFIG_RCU_NOCB_CPU static cpumask_var_t rcu_nocb_mask; /* CPUs to have callbacks offloaded. */ static bool have_rcu_nocb_mask; /* Was rcu_nocb_mask allocated? */ -@@ -432,7 +425,7 @@ +@@ -432,7 +425,7 @@ void rcu_read_unlock_special(struct task_struct *t) } /* Hardware IRQ handlers cannot block, complain if they get here. */ @@ -21760,7 +22109,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre lockdep_rcu_suspicious(__FILE__, __LINE__, "rcu_read_unlock() from irq or softirq with blocking in critical section!!!\n"); pr_alert("->rcu_read_unlock_special: %#x (b: %d, enq: %d nq: %d)\n", -@@ -645,15 +638,6 @@ +@@ -645,15 +638,6 @@ static void rcu_preempt_check_callbacks(void) t->rcu_read_unlock_special.b.need_qs = true; } @@ -21776,7 +22125,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre /* * Queue a preemptible-RCU callback for invocation after a grace period. */ -@@ -930,6 +914,19 @@ +@@ -930,6 +914,19 @@ void exit_rcu(void) #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ @@ -21796,7 +22145,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre #ifdef CONFIG_RCU_BOOST #include "../locking/rtmutex_common.h" -@@ -961,16 +958,6 @@ +@@ -961,16 +958,6 @@ static void rcu_initiate_boost_trace(struct rcu_node *rnp) #endif /* #else #ifdef CONFIG_RCU_TRACE */ @@ -21813,7 +22162,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre /* * Carry out RCU priority boosting on the task indicated by ->exp_tasks * or ->boost_tasks, advancing the pointer to the next task in the -@@ -1115,23 +1102,6 @@ +@@ -1115,23 +1102,6 @@ static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags) } /* @@ -21837,7 +22186,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre * Is the current CPU running the RCU-callbacks kthread? * Caller must have preemption disabled. */ -@@ -1186,67 +1156,6 @@ +@@ -1186,67 +1156,6 @@ static int rcu_spawn_one_boost_kthread(struct rcu_state *rsp, return 0; } @@ -21905,7 +22254,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre /* * Set the per-rcu_node kthread's affinity to cover all CPUs that are * served by the rcu_node in question. The CPU hotplug lock is still -@@ -1276,26 +1185,12 @@ +@@ -1276,26 +1185,12 @@ static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu) free_cpumask_var(cm); } @@ -21932,7 +22281,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre rcu_for_each_leaf_node(rcu_state_p, rnp) (void)rcu_spawn_one_boost_kthread(rcu_state_p, rnp); } -@@ -1318,11 +1213,6 @@ +@@ -1318,11 +1213,6 @@ static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags) raw_spin_unlock_irqrestore(&rnp->lock, flags); } @@ -21944,7 +22293,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre static bool rcu_is_callbacks_kthread(void) { return false; -@@ -1346,7 +1236,7 @@ +@@ -1346,7 +1236,7 @@ static void rcu_prepare_kthreads(int cpu) #endif /* #else #ifdef CONFIG_RCU_BOOST */ @@ -21953,7 +22302,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre /* * Check to see if any future RCU-related work will need to be done -@@ -1363,7 +1253,9 @@ +@@ -1363,7 +1253,9 @@ int rcu_needs_cpu(u64 basemono, u64 *nextevt) return IS_ENABLED(CONFIG_RCU_NOCB_CPU_ALL) ? 0 : rcu_cpu_has_callbacks(NULL); } @@ -21963,7 +22312,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre /* * Because we do not have RCU_FAST_NO_HZ, don't bother cleaning up * after it. -@@ -1459,6 +1351,8 @@ +@@ -1459,6 +1351,8 @@ static bool __maybe_unused rcu_try_advance_all_cbs(void) return cbs_ready; } @@ -21972,7 +22321,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre /* * Allow the CPU to enter dyntick-idle mode unless it has callbacks ready * to invoke. If the CPU has callbacks, try to advance them. Tell the -@@ -1504,6 +1398,7 @@ +@@ -1504,6 +1398,7 @@ int rcu_needs_cpu(u64 basemono, u64 *nextevt) *nextevt = basemono + dj * TICK_NSEC; return 0; } @@ -21980,7 +22329,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre /* * Prepare a CPU for idle from an RCU perspective. The first major task -@@ -1822,9 +1717,9 @@ +@@ -1822,9 +1717,9 @@ early_param("rcu_nocb_poll", parse_rcu_nocb_poll); * Wake up any no-CBs CPUs' kthreads that were waiting on the just-ended * grace period. */ @@ -21992,7 +22341,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre } /* -@@ -1840,10 +1735,15 @@ +@@ -1840,10 +1735,15 @@ static void rcu_nocb_gp_set(struct rcu_node *rnp, int nrq) rnp->need_future_gp[(rnp->completed + 1) & 0x1] += nrq; } @@ -22010,7 +22359,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre } #ifndef CONFIG_RCU_NOCB_CPU_ALL -@@ -1868,7 +1768,7 @@ +@@ -1868,7 +1768,7 @@ static void wake_nocb_leader(struct rcu_data *rdp, bool force) if (READ_ONCE(rdp_leader->nocb_leader_sleep) || force) { /* Prior smp_mb__after_atomic() orders against prior enqueue. */ WRITE_ONCE(rdp_leader->nocb_leader_sleep, false); @@ -22019,7 +22368,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre } } -@@ -2081,7 +1981,7 @@ +@@ -2081,7 +1981,7 @@ static void rcu_nocb_wait_gp(struct rcu_data *rdp) */ trace_rcu_future_gp(rnp, rdp, c, TPS("StartWait")); for (;;) { @@ -22028,7 +22377,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre rnp->nocb_gp_wq[c & 0x1], (d = ULONG_CMP_GE(READ_ONCE(rnp->completed), c))); if (likely(d)) -@@ -2109,7 +2009,7 @@ +@@ -2109,7 +2009,7 @@ wait_again: /* Wait for callbacks to appear. */ if (!rcu_nocb_poll) { trace_rcu_nocb_wake(my_rdp->rsp->name, my_rdp->cpu, "Sleep"); @@ -22037,7 +22386,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre !READ_ONCE(my_rdp->nocb_leader_sleep)); /* Memory barrier handled by smp_mb() calls below and repoll. */ } else if (firsttime) { -@@ -2184,7 +2084,7 @@ +@@ -2184,7 +2084,7 @@ wait_again: * List was empty, wake up the follower. * Memory barriers supplied by atomic_long_add(). */ @@ -22046,7 +22395,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre } } -@@ -2205,7 +2105,7 @@ +@@ -2205,7 +2105,7 @@ static void nocb_follower_wait(struct rcu_data *rdp) if (!rcu_nocb_poll) { trace_rcu_nocb_wake(rdp->rsp->name, rdp->cpu, "FollowerSleep"); @@ -22055,7 +22404,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre READ_ONCE(rdp->nocb_follower_head)); } else if (firsttime) { /* Don't drown trace log with "Poll"! */ -@@ -2365,7 +2265,7 @@ +@@ -2365,7 +2265,7 @@ void __init rcu_init_nohz(void) static void __init rcu_boot_init_nocb_percpu_data(struct rcu_data *rdp) { rdp->nocb_tail = &rdp->nocb_head; @@ -22064,7 +22413,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre rdp->nocb_follower_tail = &rdp->nocb_follower_head; } -@@ -2515,7 +2415,7 @@ +@@ -2515,7 +2415,7 @@ static bool rcu_nocb_cpu_needs_barrier(struct rcu_state *rsp, int cpu) return false; } @@ -22073,7 +22422,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre { } -@@ -2523,6 +2423,11 @@ +@@ -2523,6 +2423,11 @@ static void rcu_nocb_gp_set(struct rcu_node *rnp, int nrq) { } @@ -22085,10 +22434,11 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/tree_plugin.h linux-4.4.62/kernel/rcu/tre static void rcu_init_one_nocb(struct rcu_node *rnp) { } -diff -Nur linux-4.4.62.orig/kernel/rcu/update.c linux-4.4.62/kernel/rcu/update.c ---- linux-4.4.62.orig/kernel/rcu/update.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/rcu/update.c 2017-04-18 17:38:08.214650020 +0200 -@@ -276,6 +276,7 @@ +diff --git a/kernel/rcu/update.c b/kernel/rcu/update.c +index 5f748c5a40f0..9a3904603ff6 100644 +--- a/kernel/rcu/update.c ++++ b/kernel/rcu/update.c +@@ -276,6 +276,7 @@ int rcu_read_lock_held(void) } EXPORT_SYMBOL_GPL(rcu_read_lock_held); @@ -22096,7 +22446,7 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/update.c linux-4.4.62/kernel/rcu/update.c /** * rcu_read_lock_bh_held() - might we be in RCU-bh read-side critical section? * -@@ -302,6 +303,7 @@ +@@ -302,6 +303,7 @@ int rcu_read_lock_bh_held(void) return in_softirq() || irqs_disabled(); } EXPORT_SYMBOL_GPL(rcu_read_lock_bh_held); @@ -22104,10 +22454,11 @@ diff -Nur linux-4.4.62.orig/kernel/rcu/update.c linux-4.4.62/kernel/rcu/update.c #endif /* #ifdef CONFIG_DEBUG_LOCK_ALLOC */ -diff -Nur linux-4.4.62.orig/kernel/relay.c linux-4.4.62/kernel/relay.c ---- linux-4.4.62.orig/kernel/relay.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/relay.c 2017-04-18 17:38:08.214650020 +0200 -@@ -336,6 +336,10 @@ +diff --git a/kernel/relay.c b/kernel/relay.c +index 0b4570cfacae..60684be39f22 100644 +--- a/kernel/relay.c ++++ b/kernel/relay.c +@@ -336,6 +336,10 @@ static void wakeup_readers(unsigned long data) { struct rchan_buf *buf = (struct rchan_buf *)data; wake_up_interruptible(&buf->read_wait); @@ -22118,7 +22469,7 @@ diff -Nur linux-4.4.62.orig/kernel/relay.c linux-4.4.62/kernel/relay.c } /** -@@ -353,6 +357,7 @@ +@@ -353,6 +357,7 @@ static void __relay_reset(struct rchan_buf *buf, unsigned int init) init_waitqueue_head(&buf->read_wait); kref_init(&buf->kref); setup_timer(&buf->timer, wakeup_readers, (unsigned long)buf); @@ -22126,7 +22477,7 @@ diff -Nur linux-4.4.62.orig/kernel/relay.c linux-4.4.62/kernel/relay.c } else del_timer_sync(&buf->timer); -@@ -736,15 +741,6 @@ +@@ -736,15 +741,6 @@ size_t relay_switch_subbuf(struct rchan_buf *buf, size_t length) else buf->early_bytes += buf->chan->subbuf_size - buf->padding[old_subbuf]; @@ -22142,10 +22493,24 @@ diff -Nur linux-4.4.62.orig/kernel/relay.c linux-4.4.62/kernel/relay.c } old = buf->data; -diff -Nur linux-4.4.62.orig/kernel/sched/completion.c linux-4.4.62/kernel/sched/completion.c ---- linux-4.4.62.orig/kernel/sched/completion.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/sched/completion.c 2017-04-18 17:38:08.214650020 +0200 -@@ -30,10 +30,10 @@ +diff --git a/kernel/sched/Makefile b/kernel/sched/Makefile +index 67687973ce80..01b9994b367a 100644 +--- a/kernel/sched/Makefile ++++ b/kernel/sched/Makefile +@@ -13,7 +13,7 @@ endif + + obj-y += core.o loadavg.o clock.o cputime.o + obj-y += idle_task.o fair.o rt.o deadline.o stop_task.o +-obj-y += wait.o completion.o idle.o ++obj-y += wait.o swait.o swork.o completion.o idle.o + obj-$(CONFIG_SMP) += cpupri.o cpudeadline.o + obj-$(CONFIG_SCHED_AUTOGROUP) += auto_group.o + obj-$(CONFIG_SCHEDSTATS) += stats.o +diff --git a/kernel/sched/completion.c b/kernel/sched/completion.c +index 8d0f35debf35..b62cf6400fe0 100644 +--- a/kernel/sched/completion.c ++++ b/kernel/sched/completion.c +@@ -30,10 +30,10 @@ void complete(struct completion *x) { unsigned long flags; @@ -22159,7 +22524,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/completion.c linux-4.4.62/kernel/sched/ } EXPORT_SYMBOL(complete); -@@ -50,10 +50,10 @@ +@@ -50,10 +50,10 @@ void complete_all(struct completion *x) { unsigned long flags; @@ -22173,7 +22538,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/completion.c linux-4.4.62/kernel/sched/ } EXPORT_SYMBOL(complete_all); -@@ -62,20 +62,20 @@ +@@ -62,20 +62,20 @@ do_wait_for_common(struct completion *x, long (*action)(long), long timeout, int state) { if (!x->done) { @@ -22199,7 +22564,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/completion.c linux-4.4.62/kernel/sched/ if (!x->done) return timeout; } -@@ -89,9 +89,9 @@ +@@ -89,9 +89,9 @@ __wait_for_common(struct completion *x, { might_sleep(); @@ -22211,7 +22576,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/completion.c linux-4.4.62/kernel/sched/ return timeout; } -@@ -277,12 +277,12 @@ +@@ -277,12 +277,12 @@ bool try_wait_for_completion(struct completion *x) if (!READ_ONCE(x->done)) return 0; @@ -22226,7 +22591,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/completion.c linux-4.4.62/kernel/sched/ return ret; } EXPORT_SYMBOL(try_wait_for_completion); -@@ -311,7 +311,7 @@ +@@ -311,7 +311,7 @@ bool completion_done(struct completion *x) * after it's acquired the lock. */ smp_rmb(); @@ -22235,10 +22600,11 @@ diff -Nur linux-4.4.62.orig/kernel/sched/completion.c linux-4.4.62/kernel/sched/ return true; } EXPORT_SYMBOL(completion_done); -diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c ---- linux-4.4.62.orig/kernel/sched/core.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/sched/core.c 2017-04-18 17:38:08.218650175 +0200 -@@ -260,7 +260,11 @@ +diff --git a/kernel/sched/core.c b/kernel/sched/core.c +index 20253dbc8610..e9b8d518202e 100644 +--- a/kernel/sched/core.c ++++ b/kernel/sched/core.c +@@ -260,7 +260,11 @@ late_initcall(sched_init_debug); * Number of tasks to iterate in a single balance run. * Limited because this is done with IRQs disabled. */ @@ -22250,7 +22616,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c /* * period over which we average the RT time consumption, measured -@@ -438,6 +442,7 @@ +@@ -438,6 +442,7 @@ static void init_rq_hrtick(struct rq *rq) hrtimer_init(&rq->hrtick_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); rq->hrtick_timer.function = hrtick; @@ -22258,7 +22624,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c } #else /* CONFIG_SCHED_HRTICK */ static inline void hrtick_clear(struct rq *rq) -@@ -542,7 +547,7 @@ +@@ -542,7 +547,7 @@ void wake_q_add(struct wake_q_head *head, struct task_struct *task) head->lastp = &node->next; } @@ -22267,7 +22633,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c { struct wake_q_node *node = head->first; -@@ -559,7 +564,10 @@ +@@ -559,7 +564,10 @@ void wake_up_q(struct wake_q_head *head) * wake_up_process() implies a wmb() to pair with the queueing * in wake_q_add() so as not to miss wakeups. */ @@ -22279,7 +22645,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c put_task_struct(task); } } -@@ -595,6 +603,38 @@ +@@ -595,6 +603,38 @@ void resched_curr(struct rq *rq) trace_sched_wake_idle_without_ipi(cpu); } @@ -22318,7 +22684,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c void resched_cpu(int cpu) { struct rq *rq = cpu_rq(cpu); -@@ -618,11 +658,14 @@ +@@ -618,11 +658,14 @@ void resched_cpu(int cpu) */ int get_nohz_timer_target(void) { @@ -22335,7 +22701,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c rcu_read_lock(); for_each_domain(cpu, sd) { -@@ -641,6 +684,8 @@ +@@ -641,6 +684,8 @@ int get_nohz_timer_target(void) cpu = housekeeping_any_cpu(); unlock: rcu_read_unlock(); @@ -22344,7 +22710,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c return cpu; } /* -@@ -1174,6 +1219,11 @@ +@@ -1174,6 +1219,11 @@ void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask) lockdep_assert_held(&p->pi_lock); @@ -22356,7 +22722,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c queued = task_on_rq_queued(p); running = task_current(rq, p); -@@ -1196,6 +1246,84 @@ +@@ -1196,6 +1246,84 @@ void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask) enqueue_task(rq, p, ENQUEUE_RESTORE); } @@ -22441,7 +22807,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c /* * Change a given task's CPU affinity. Migrate the thread to a * proper CPU and schedule it away if the CPU it's executing on -@@ -1235,7 +1363,7 @@ +@@ -1235,7 +1363,7 @@ static int __set_cpus_allowed_ptr(struct task_struct *p, do_set_cpus_allowed(p, new_mask); /* Can the task run on the task's current CPU? If so, we're done */ @@ -22450,7 +22816,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c goto out; dest_cpu = cpumask_any_and(cpu_active_mask, new_mask); -@@ -1411,6 +1539,18 @@ +@@ -1411,6 +1539,18 @@ out: return ret; } @@ -22469,7 +22835,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c /* * wait_task_inactive - wait for a thread to unschedule. * -@@ -1455,7 +1595,7 @@ +@@ -1455,7 +1595,7 @@ unsigned long wait_task_inactive(struct task_struct *p, long match_state) * is actually now running somewhere else! */ while (task_running(rq, p)) { @@ -22478,7 +22844,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c return 0; cpu_relax(); } -@@ -1470,7 +1610,8 @@ +@@ -1470,7 +1610,8 @@ unsigned long wait_task_inactive(struct task_struct *p, long match_state) running = task_running(rq, p); queued = task_on_rq_queued(p); ncsw = 0; @@ -22488,7 +22854,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c ncsw = p->nvcsw | LONG_MIN; /* sets MSB */ task_rq_unlock(rq, p, &flags); -@@ -1627,7 +1768,7 @@ +@@ -1627,7 +1768,7 @@ int select_task_rq(struct task_struct *p, int cpu, int sd_flags, int wake_flags) { lockdep_assert_held(&p->pi_lock); @@ -22497,7 +22863,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c cpu = p->sched_class->select_task_rq(p, cpu, sd_flags, wake_flags); /* -@@ -1707,10 +1848,6 @@ +@@ -1707,10 +1848,6 @@ static inline void ttwu_activate(struct rq *rq, struct task_struct *p, int en_fl { activate_task(rq, p, en_flags); p->on_rq = TASK_ON_RQ_QUEUED; @@ -22508,7 +22874,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c } /* -@@ -1937,8 +2074,27 @@ +@@ -1937,8 +2074,27 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags) */ smp_mb__before_spinlock(); raw_spin_lock_irqsave(&p->pi_lock, flags); @@ -22537,7 +22903,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c trace_sched_waking(p); -@@ -2030,52 +2186,6 @@ +@@ -2030,52 +2186,6 @@ out: } /** @@ -22590,7 +22956,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c * wake_up_process - Wake up a specific process * @p: The process to be woken up. * -@@ -2093,6 +2203,18 @@ +@@ -2093,6 +2203,18 @@ int wake_up_process(struct task_struct *p) } EXPORT_SYMBOL(wake_up_process); @@ -22609,7 +22975,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c int wake_up_state(struct task_struct *p, unsigned int state) { return try_to_wake_up(p, state, 0); -@@ -2279,6 +2401,9 @@ +@@ -2279,6 +2401,9 @@ int sched_fork(unsigned long clone_flags, struct task_struct *p) p->on_cpu = 0; #endif init_task_preempt_count(p); @@ -22619,7 +22985,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c #ifdef CONFIG_SMP plist_node_init(&p->pushable_tasks, MAX_PRIO); RB_CLEAR_NODE(&p->pushable_dl_tasks); -@@ -2603,8 +2728,12 @@ +@@ -2603,8 +2728,12 @@ static struct rq *finish_task_switch(struct task_struct *prev) finish_arch_post_lock_switch(); fire_sched_in_preempt_notifiers(current); @@ -22633,7 +22999,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c if (unlikely(prev_state == TASK_DEAD)) { if (prev->sched_class->task_dead) prev->sched_class->task_dead(prev); -@@ -2935,16 +3064,6 @@ +@@ -2935,16 +3064,6 @@ u64 scheduler_tick_max_deferment(void) } #endif @@ -22650,7 +23016,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c #if defined(CONFIG_PREEMPT) && (defined(CONFIG_DEBUG_PREEMPT) || \ defined(CONFIG_PREEMPT_TRACER)) -@@ -2966,7 +3085,7 @@ +@@ -2966,7 +3085,7 @@ void preempt_count_add(int val) PREEMPT_MASK - 10); #endif if (preempt_count() == val) { @@ -22659,7 +23025,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c #ifdef CONFIG_DEBUG_PREEMPT current->preempt_disable_ip = ip; #endif -@@ -2993,7 +3112,7 @@ +@@ -2993,7 +3112,7 @@ void preempt_count_sub(int val) #endif if (preempt_count() == val) @@ -22668,7 +23034,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c __preempt_count_sub(val); } EXPORT_SYMBOL(preempt_count_sub); -@@ -3048,6 +3167,77 @@ +@@ -3048,6 +3167,77 @@ static inline void schedule_debug(struct task_struct *prev) schedstat_inc(this_rq(), sched_count); } @@ -22746,7 +23112,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c /* * Pick up the highest-prio task: */ -@@ -3172,19 +3362,6 @@ +@@ -3172,19 +3362,6 @@ static void __sched notrace __schedule(bool preempt) } else { deactivate_task(rq, prev, DEQUEUE_SLEEP); prev->on_rq = 0; @@ -22766,7 +23132,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c } switch_count = &prev->nvcsw; } -@@ -3194,6 +3371,7 @@ +@@ -3194,6 +3371,7 @@ static void __sched notrace __schedule(bool preempt) next = pick_next_task(rq, prev); clear_tsk_need_resched(prev); @@ -22774,7 +23140,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c clear_preempt_need_resched(); rq->clock_skip_update = 0; -@@ -3215,9 +3393,20 @@ +@@ -3215,9 +3393,20 @@ static void __sched notrace __schedule(bool preempt) static inline void sched_submit_work(struct task_struct *tsk) { @@ -22796,7 +23162,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c * If we are going to sleep and we have plugged IO queued, * make sure to submit it to avoid deadlocks. */ -@@ -3225,6 +3414,12 @@ +@@ -3225,6 +3414,12 @@ static inline void sched_submit_work(struct task_struct *tsk) blk_schedule_flush_plug(tsk); } @@ -22809,7 +23175,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c asmlinkage __visible void __sched schedule(void) { struct task_struct *tsk = current; -@@ -3235,6 +3430,7 @@ +@@ -3235,6 +3430,7 @@ asmlinkage __visible void __sched schedule(void) __schedule(false); sched_preempt_enable_no_resched(); } while (need_resched()); @@ -22817,7 +23183,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c } EXPORT_SYMBOL(schedule); -@@ -3283,6 +3479,30 @@ +@@ -3283,6 +3479,30 @@ static void __sched notrace preempt_schedule_common(void) } while (need_resched()); } @@ -22848,7 +23214,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c #ifdef CONFIG_PREEMPT /* * this is the entry point to schedule() from in-kernel preemption -@@ -3297,6 +3517,8 @@ +@@ -3297,6 +3517,8 @@ asmlinkage __visible void __sched notrace preempt_schedule(void) */ if (likely(!preemptible())) return; @@ -22857,7 +23223,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c preempt_schedule_common(); } -@@ -3323,6 +3545,8 @@ +@@ -3323,6 +3545,8 @@ asmlinkage __visible void __sched notrace preempt_schedule_notrace(void) if (likely(!preemptible())) return; @@ -22866,7 +23232,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c do { preempt_disable_notrace(); -@@ -3332,7 +3556,16 @@ +@@ -3332,7 +3556,16 @@ asmlinkage __visible void __sched notrace preempt_schedule_notrace(void) * an infinite recursion. */ prev_ctx = exception_enter(); @@ -22883,7 +23249,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c exception_exit(prev_ctx); preempt_enable_no_resched_notrace(); -@@ -4676,6 +4909,7 @@ +@@ -4676,6 +4909,7 @@ int __cond_resched_lock(spinlock_t *lock) } EXPORT_SYMBOL(__cond_resched_lock); @@ -22891,7 +23257,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c int __sched __cond_resched_softirq(void) { BUG_ON(!in_softirq()); -@@ -4689,6 +4923,7 @@ +@@ -4689,6 +4923,7 @@ int __sched __cond_resched_softirq(void) return 0; } EXPORT_SYMBOL(__cond_resched_softirq); @@ -22899,7 +23265,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c /** * yield - yield the current processor to other threads. -@@ -5055,7 +5290,9 @@ +@@ -5055,7 +5290,9 @@ void init_idle(struct task_struct *idle, int cpu) /* Set the preempt count _outside_ the spinlocks! */ init_idle_preempt_count(idle, cpu); @@ -22910,7 +23276,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c /* * The idle tasks have their own, simple scheduling class: */ -@@ -5196,6 +5433,8 @@ +@@ -5196,6 +5433,8 @@ void sched_setnuma(struct task_struct *p, int nid) #endif /* CONFIG_NUMA_BALANCING */ #ifdef CONFIG_HOTPLUG_CPU @@ -22919,7 +23285,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c /* * Ensures that the idle task is using init_mm right before its cpu goes * offline. -@@ -5210,7 +5449,11 @@ +@@ -5210,7 +5449,11 @@ void idle_task_exit(void) switch_mm(mm, &init_mm, current); finish_arch_post_lock_switch(); } @@ -22932,7 +23298,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c } /* -@@ -5583,6 +5826,10 @@ +@@ -5583,6 +5826,10 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) case CPU_DEAD: calc_load_migrate(rq); @@ -22943,7 +23309,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c break; #endif } -@@ -7566,7 +7813,7 @@ +@@ -7566,7 +7813,7 @@ void __init sched_init(void) #ifdef CONFIG_DEBUG_ATOMIC_SLEEP static inline int preempt_count_equals(int preempt_offset) { @@ -22952,10 +23318,11 @@ diff -Nur linux-4.4.62.orig/kernel/sched/core.c linux-4.4.62/kernel/sched/core.c return (nested == preempt_offset); } -diff -Nur linux-4.4.62.orig/kernel/sched/cpudeadline.c linux-4.4.62/kernel/sched/cpudeadline.c ---- linux-4.4.62.orig/kernel/sched/cpudeadline.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/sched/cpudeadline.c 2017-04-18 17:38:08.218650175 +0200 -@@ -103,10 +103,10 @@ +diff --git a/kernel/sched/cpudeadline.c b/kernel/sched/cpudeadline.c +index 5a75b08cfd85..5be58820465c 100644 +--- a/kernel/sched/cpudeadline.c ++++ b/kernel/sched/cpudeadline.c +@@ -103,10 +103,10 @@ int cpudl_find(struct cpudl *cp, struct task_struct *p, const struct sched_dl_entity *dl_se = &p->dl; if (later_mask && @@ -22968,10 +23335,11 @@ diff -Nur linux-4.4.62.orig/kernel/sched/cpudeadline.c linux-4.4.62/kernel/sched dl_time_before(dl_se->deadline, cp->elements[0].dl)) { best_cpu = cpudl_maximum(cp); if (later_mask) -diff -Nur linux-4.4.62.orig/kernel/sched/cpupri.c linux-4.4.62/kernel/sched/cpupri.c ---- linux-4.4.62.orig/kernel/sched/cpupri.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/sched/cpupri.c 2017-04-18 17:38:08.218650175 +0200 -@@ -103,11 +103,11 @@ +diff --git a/kernel/sched/cpupri.c b/kernel/sched/cpupri.c +index 981fcd7dc394..11e9705bf937 100644 +--- a/kernel/sched/cpupri.c ++++ b/kernel/sched/cpupri.c +@@ -103,11 +103,11 @@ int cpupri_find(struct cpupri *cp, struct task_struct *p, if (skip) continue; @@ -22985,10 +23353,11 @@ diff -Nur linux-4.4.62.orig/kernel/sched/cpupri.c linux-4.4.62/kernel/sched/cpup /* * We have to ensure that we have at least one bit -diff -Nur linux-4.4.62.orig/kernel/sched/cputime.c linux-4.4.62/kernel/sched/cputime.c ---- linux-4.4.62.orig/kernel/sched/cputime.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/sched/cputime.c 2017-04-18 17:38:08.218650175 +0200 -@@ -685,7 +685,7 @@ +diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c +index a1aecbedf5b1..558b98af241d 100644 +--- a/kernel/sched/cputime.c ++++ b/kernel/sched/cputime.c +@@ -685,7 +685,7 @@ static cputime_t get_vtime_delta(struct task_struct *tsk) { unsigned long long delta = vtime_delta(tsk); @@ -22997,7 +23366,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/cputime.c linux-4.4.62/kernel/sched/cpu tsk->vtime_snap += delta; /* CHECKME: always safe to convert nsecs to cputime? */ -@@ -701,37 +701,37 @@ +@@ -701,37 +701,37 @@ static void __vtime_account_system(struct task_struct *tsk) void vtime_account_system(struct task_struct *tsk) { @@ -23043,7 +23412,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/cputime.c linux-4.4.62/kernel/sched/cpu } void vtime_guest_enter(struct task_struct *tsk) -@@ -743,19 +743,19 @@ +@@ -743,19 +743,19 @@ void vtime_guest_enter(struct task_struct *tsk) * synchronization against the reader (task_gtime()) * that can thus safely catch up with a tickless delta. */ @@ -23067,7 +23436,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/cputime.c linux-4.4.62/kernel/sched/cpu } EXPORT_SYMBOL_GPL(vtime_guest_exit); -@@ -768,24 +768,26 @@ +@@ -768,24 +768,26 @@ void vtime_account_idle(struct task_struct *tsk) void arch_vtime_task_switch(struct task_struct *prev) { @@ -23101,7 +23470,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/cputime.c linux-4.4.62/kernel/sched/cpu } cputime_t task_gtime(struct task_struct *t) -@@ -797,13 +799,13 @@ +@@ -797,13 +799,13 @@ cputime_t task_gtime(struct task_struct *t) return t->gtime; do { @@ -23117,7 +23486,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/cputime.c linux-4.4.62/kernel/sched/cpu return gtime; } -@@ -826,7 +828,7 @@ +@@ -826,7 +828,7 @@ fetch_task_cputime(struct task_struct *t, *udelta = 0; *sdelta = 0; @@ -23126,7 +23495,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/cputime.c linux-4.4.62/kernel/sched/cpu if (u_dst) *u_dst = *u_src; -@@ -834,7 +836,7 @@ +@@ -834,7 +836,7 @@ fetch_task_cputime(struct task_struct *t, *s_dst = *s_src; /* Task is sleeping, nothing to add */ @@ -23135,7 +23504,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/cputime.c linux-4.4.62/kernel/sched/cpu is_idle_task(t)) continue; -@@ -850,7 +852,7 @@ +@@ -850,7 +852,7 @@ fetch_task_cputime(struct task_struct *t, if (t->vtime_snap_whence == VTIME_SYS) *sdelta = delta; } @@ -23144,10 +23513,11 @@ diff -Nur linux-4.4.62.orig/kernel/sched/cputime.c linux-4.4.62/kernel/sched/cpu } -diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/deadline.c ---- linux-4.4.62.orig/kernel/sched/deadline.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/sched/deadline.c 2017-04-18 17:38:08.218650175 +0200 -@@ -134,7 +134,7 @@ +diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c +index e984f059e5fc..20e0c9b9ace5 100644 +--- a/kernel/sched/deadline.c ++++ b/kernel/sched/deadline.c +@@ -134,7 +134,7 @@ static void inc_dl_migration(struct sched_dl_entity *dl_se, struct dl_rq *dl_rq) { struct task_struct *p = dl_task_of(dl_se); @@ -23156,7 +23526,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de dl_rq->dl_nr_migratory++; update_dl_migration(dl_rq); -@@ -144,7 +144,7 @@ +@@ -144,7 +144,7 @@ static void dec_dl_migration(struct sched_dl_entity *dl_se, struct dl_rq *dl_rq) { struct task_struct *p = dl_task_of(dl_se); @@ -23165,7 +23535,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de dl_rq->dl_nr_migratory--; update_dl_migration(dl_rq); -@@ -697,6 +697,7 @@ +@@ -697,6 +697,7 @@ void init_dl_task_timer(struct sched_dl_entity *dl_se) hrtimer_init(timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); timer->function = dl_task_timer; @@ -23173,7 +23543,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de } static -@@ -989,7 +990,7 @@ +@@ -989,7 +990,7 @@ static void enqueue_task_dl(struct rq *rq, struct task_struct *p, int flags) enqueue_dl_entity(&p->dl, pi_se, flags); @@ -23182,7 +23552,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de enqueue_pushable_dl_task(rq, p); } -@@ -1067,9 +1068,9 @@ +@@ -1067,9 +1068,9 @@ select_task_rq_dl(struct task_struct *p, int cpu, int sd_flag, int flags) * try to make it stay here, it might be important. */ if (unlikely(dl_task(curr)) && @@ -23194,7 +23564,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de int target = find_later_rq(p); if (target != -1 && -@@ -1090,7 +1091,7 @@ +@@ -1090,7 +1091,7 @@ static void check_preempt_equal_dl(struct rq *rq, struct task_struct *p) * Current can't be migrated, useless to reschedule, * let's hope p can move out. */ @@ -23203,7 +23573,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de cpudl_find(&rq->rd->cpudl, rq->curr, NULL) == -1) return; -@@ -1098,7 +1099,7 @@ +@@ -1098,7 +1099,7 @@ static void check_preempt_equal_dl(struct rq *rq, struct task_struct *p) * p is migratable, so let's not schedule it and * see if it is pushed or pulled somewhere else. */ @@ -23212,7 +23582,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de cpudl_find(&rq->rd->cpudl, p, NULL) != -1) return; -@@ -1212,7 +1213,7 @@ +@@ -1212,7 +1213,7 @@ static void put_prev_task_dl(struct rq *rq, struct task_struct *p) { update_curr_dl(rq); @@ -23221,7 +23591,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de enqueue_pushable_dl_task(rq, p); } -@@ -1335,7 +1336,7 @@ +@@ -1335,7 +1336,7 @@ static int find_later_rq(struct task_struct *task) if (unlikely(!later_mask)) return -1; @@ -23230,7 +23600,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de return -1; /* -@@ -1441,7 +1442,7 @@ +@@ -1441,7 +1442,7 @@ static struct rq *find_lock_later_rq(struct task_struct *task, struct rq *rq) if (double_lock_balance(rq, later_rq)) { if (unlikely(task_rq(task) != rq || !cpumask_test_cpu(later_rq->cpu, @@ -23239,7 +23609,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de task_running(rq, task) || !task_on_rq_queued(task))) { double_unlock_balance(rq, later_rq); -@@ -1480,7 +1481,7 @@ +@@ -1480,7 +1481,7 @@ static struct task_struct *pick_next_pushable_dl_task(struct rq *rq) BUG_ON(rq->cpu != task_cpu(p)); BUG_ON(task_current(rq, p)); @@ -23248,7 +23618,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de BUG_ON(!task_on_rq_queued(p)); BUG_ON(!dl_task(p)); -@@ -1519,7 +1520,7 @@ +@@ -1519,7 +1520,7 @@ retry: */ if (dl_task(rq->curr) && dl_time_before(next_task->dl.deadline, rq->curr->dl.deadline) && @@ -23257,7 +23627,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de resched_curr(rq); return 0; } -@@ -1666,9 +1667,9 @@ +@@ -1666,9 +1667,9 @@ static void task_woken_dl(struct rq *rq, struct task_struct *p) { if (!task_running(rq, p) && !test_tsk_need_resched(rq->curr) && @@ -23269,7 +23639,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de !dl_entity_preempt(&p->dl, &rq->curr->dl))) { push_dl_tasks(rq); } -@@ -1769,7 +1770,7 @@ +@@ -1769,7 +1770,7 @@ static void switched_to_dl(struct rq *rq, struct task_struct *p) { if (task_on_rq_queued(p) && rq->curr != p) { #ifdef CONFIG_SMP @@ -23278,10 +23648,11 @@ diff -Nur linux-4.4.62.orig/kernel/sched/deadline.c linux-4.4.62/kernel/sched/de queue_push_tasks(rq); #endif if (dl_task(rq->curr)) -diff -Nur linux-4.4.62.orig/kernel/sched/debug.c linux-4.4.62/kernel/sched/debug.c ---- linux-4.4.62.orig/kernel/sched/debug.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/sched/debug.c 2017-04-18 17:38:08.218650175 +0200 -@@ -251,6 +251,9 @@ +diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c +index 641511771ae6..a2d69b883623 100644 +--- a/kernel/sched/debug.c ++++ b/kernel/sched/debug.c +@@ -251,6 +251,9 @@ void print_rt_rq(struct seq_file *m, int cpu, struct rt_rq *rt_rq) P(rt_throttled); PN(rt_time); PN(rt_runtime); @@ -23291,7 +23662,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/debug.c linux-4.4.62/kernel/sched/debug #undef PN #undef P -@@ -635,6 +638,10 @@ +@@ -635,6 +638,10 @@ void proc_sched_show_task(struct task_struct *p, struct seq_file *m) #endif P(policy); P(prio); @@ -23302,10 +23673,11 @@ diff -Nur linux-4.4.62.orig/kernel/sched/debug.c linux-4.4.62/kernel/sched/debug #undef PN #undef __PN #undef P -diff -Nur linux-4.4.62.orig/kernel/sched/fair.c linux-4.4.62/kernel/sched/fair.c ---- linux-4.4.62.orig/kernel/sched/fair.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/sched/fair.c 2017-04-18 17:38:08.218650175 +0200 -@@ -3166,7 +3166,7 @@ +diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c +index 812069b66f47..ddf1424bcc78 100644 +--- a/kernel/sched/fair.c ++++ b/kernel/sched/fair.c +@@ -3166,7 +3166,7 @@ check_preempt_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr) ideal_runtime = sched_slice(cfs_rq, curr); delta_exec = curr->sum_exec_runtime - curr->prev_sum_exec_runtime; if (delta_exec > ideal_runtime) { @@ -23314,7 +23686,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/fair.c linux-4.4.62/kernel/sched/fair.c /* * The current task ran long enough, ensure it doesn't get * re-elected due to buddy favours. -@@ -3190,7 +3190,7 @@ +@@ -3190,7 +3190,7 @@ check_preempt_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr) return; if (delta > ideal_runtime) @@ -23323,7 +23695,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/fair.c linux-4.4.62/kernel/sched/fair.c } static void -@@ -3330,7 +3330,7 @@ +@@ -3330,7 +3330,7 @@ entity_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr, int queued) * validating it and just reschedule. */ if (queued) { @@ -23332,7 +23704,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/fair.c linux-4.4.62/kernel/sched/fair.c return; } /* -@@ -3512,7 +3512,7 @@ +@@ -3512,7 +3512,7 @@ static void __account_cfs_rq_runtime(struct cfs_rq *cfs_rq, u64 delta_exec) * hierarchy can be throttled */ if (!assign_cfs_rq_runtime(cfs_rq) && likely(cfs_rq->curr)) @@ -23341,7 +23713,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/fair.c linux-4.4.62/kernel/sched/fair.c } static __always_inline -@@ -4124,7 +4124,7 @@ +@@ -4144,7 +4144,7 @@ static void hrtick_start_fair(struct rq *rq, struct task_struct *p) if (delta < 0) { if (rq->curr == p) @@ -23350,7 +23722,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/fair.c linux-4.4.62/kernel/sched/fair.c return; } hrtick_start(rq, delta); -@@ -5213,7 +5213,7 @@ +@@ -5232,7 +5232,7 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_ return; preempt: @@ -23359,7 +23731,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/fair.c linux-4.4.62/kernel/sched/fair.c /* * Only set the backward buddy when the current task is still * on the rq. This can happen when a wakeup gets interleaved -@@ -7964,7 +7964,7 @@ +@@ -7983,7 +7983,7 @@ static void task_fork_fair(struct task_struct *p) * 'current' within the tree based on its new key value. */ swap(curr->vruntime, se->vruntime); @@ -23368,7 +23740,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/fair.c linux-4.4.62/kernel/sched/fair.c } se->vruntime -= cfs_rq->min_vruntime; -@@ -7989,7 +7989,7 @@ +@@ -8008,7 +8008,7 @@ prio_changed_fair(struct rq *rq, struct task_struct *p, int oldprio) */ if (rq->curr == p) { if (p->prio > oldprio) @@ -23377,10 +23749,11 @@ diff -Nur linux-4.4.62.orig/kernel/sched/fair.c linux-4.4.62/kernel/sched/fair.c } else check_preempt_curr(rq, p, 0); } -diff -Nur linux-4.4.62.orig/kernel/sched/features.h linux-4.4.62/kernel/sched/features.h ---- linux-4.4.62.orig/kernel/sched/features.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/sched/features.h 2017-04-18 17:38:08.218650175 +0200 -@@ -45,11 +45,19 @@ +diff --git a/kernel/sched/features.h b/kernel/sched/features.h +index 69631fa46c2f..6d28fcd08872 100644 +--- a/kernel/sched/features.h ++++ b/kernel/sched/features.h +@@ -45,11 +45,19 @@ SCHED_FEAT(LB_BIAS, true) */ SCHED_FEAT(NONTASK_CAPACITY, true) @@ -23400,22 +23773,11 @@ diff -Nur linux-4.4.62.orig/kernel/sched/features.h linux-4.4.62/kernel/sched/fe #ifdef HAVE_RT_PUSH_IPI /* -diff -Nur linux-4.4.62.orig/kernel/sched/Makefile linux-4.4.62/kernel/sched/Makefile ---- linux-4.4.62.orig/kernel/sched/Makefile 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/sched/Makefile 2017-04-18 17:38:08.214650020 +0200 -@@ -13,7 +13,7 @@ - - obj-y += core.o loadavg.o clock.o cputime.o - obj-y += idle_task.o fair.o rt.o deadline.o stop_task.o --obj-y += wait.o completion.o idle.o -+obj-y += wait.o swait.o swork.o completion.o idle.o - obj-$(CONFIG_SMP) += cpupri.o cpudeadline.o - obj-$(CONFIG_SCHED_AUTOGROUP) += auto_group.o - obj-$(CONFIG_SCHEDSTATS) += stats.o -diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c ---- linux-4.4.62.orig/kernel/sched/rt.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/sched/rt.c 2017-04-18 17:38:08.218650175 +0200 -@@ -47,6 +47,7 @@ +diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c +index 78ae5c1d9412..4ac6937f4a65 100644 +--- a/kernel/sched/rt.c ++++ b/kernel/sched/rt.c +@@ -47,6 +47,7 @@ void init_rt_bandwidth(struct rt_bandwidth *rt_b, u64 period, u64 runtime) hrtimer_init(&rt_b->rt_period_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); @@ -23423,7 +23785,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c rt_b->rt_period_timer.function = sched_rt_period_timer; } -@@ -93,6 +94,7 @@ +@@ -93,6 +94,7 @@ void init_rt_rq(struct rt_rq *rt_rq) rt_rq->push_cpu = nr_cpu_ids; raw_spin_lock_init(&rt_rq->push_lock); init_irq_work(&rt_rq->push_work, push_irq_work_func); @@ -23431,7 +23793,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c #endif #endif /* CONFIG_SMP */ /* We start is dequeued state, because no RT tasks are queued */ -@@ -326,7 +328,7 @@ +@@ -326,7 +328,7 @@ static void inc_rt_migration(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq) rt_rq = &rq_of_rt_rq(rt_rq)->rt; rt_rq->rt_nr_total++; @@ -23440,7 +23802,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c rt_rq->rt_nr_migratory++; update_rt_migration(rt_rq); -@@ -343,7 +345,7 @@ +@@ -343,7 +345,7 @@ static void dec_rt_migration(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq) rt_rq = &rq_of_rt_rq(rt_rq)->rt; rt_rq->rt_nr_total--; @@ -23449,7 +23811,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c rt_rq->rt_nr_migratory--; update_rt_migration(rt_rq); -@@ -1262,7 +1264,7 @@ +@@ -1262,7 +1264,7 @@ enqueue_task_rt(struct rq *rq, struct task_struct *p, int flags) enqueue_rt_entity(rt_se, flags & ENQUEUE_HEAD); @@ -23458,7 +23820,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c enqueue_pushable_task(rq, p); } -@@ -1351,7 +1353,7 @@ +@@ -1351,7 +1353,7 @@ select_task_rq_rt(struct task_struct *p, int cpu, int sd_flag, int flags) * will have to sort it out. */ if (curr && unlikely(rt_task(curr)) && @@ -23467,7 +23829,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c curr->prio <= p->prio)) { int target = find_lowest_rq(p); -@@ -1375,7 +1377,7 @@ +@@ -1375,7 +1377,7 @@ static void check_preempt_equal_prio(struct rq *rq, struct task_struct *p) * Current can't be migrated, useless to reschedule, * let's hope p can move out. */ @@ -23476,7 +23838,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c !cpupri_find(&rq->rd->cpupri, rq->curr, NULL)) return; -@@ -1383,7 +1385,7 @@ +@@ -1383,7 +1385,7 @@ static void check_preempt_equal_prio(struct rq *rq, struct task_struct *p) * p is migratable, so let's not schedule it and * see if it is pushed or pulled somewhere else. */ @@ -23485,7 +23847,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c && cpupri_find(&rq->rd->cpupri, p, NULL)) return; -@@ -1517,7 +1519,7 @@ +@@ -1517,7 +1519,7 @@ static void put_prev_task_rt(struct rq *rq, struct task_struct *p) * The previous task needs to be made eligible for pushing * if it is still active */ @@ -23494,7 +23856,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c enqueue_pushable_task(rq, p); } -@@ -1567,7 +1569,7 @@ +@@ -1567,7 +1569,7 @@ static int find_lowest_rq(struct task_struct *task) if (unlikely(!lowest_mask)) return -1; @@ -23503,7 +23865,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c return -1; /* No other targets possible */ if (!cpupri_find(&task_rq(task)->rd->cpupri, task, lowest_mask)) -@@ -1699,7 +1701,7 @@ +@@ -1699,7 +1701,7 @@ static struct task_struct *pick_next_pushable_task(struct rq *rq) BUG_ON(rq->cpu != task_cpu(p)); BUG_ON(task_current(rq, p)); @@ -23512,7 +23874,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c BUG_ON(!task_on_rq_queued(p)); BUG_ON(!rt_task(p)); -@@ -2059,9 +2061,9 @@ +@@ -2059,9 +2061,9 @@ static void task_woken_rt(struct rq *rq, struct task_struct *p) { if (!task_running(rq, p) && !test_tsk_need_resched(rq->curr) && @@ -23524,7 +23886,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c rq->curr->prio <= p->prio)) push_rt_tasks(rq); } -@@ -2134,7 +2136,7 @@ +@@ -2134,7 +2136,7 @@ static void switched_to_rt(struct rq *rq, struct task_struct *p) */ if (task_on_rq_queued(p) && rq->curr != p) { #ifdef CONFIG_SMP @@ -23533,10 +23895,11 @@ diff -Nur linux-4.4.62.orig/kernel/sched/rt.c linux-4.4.62/kernel/sched/rt.c queue_push_tasks(rq); #endif /* CONFIG_SMP */ if (p->prio < rq->curr->prio) -diff -Nur linux-4.4.62.orig/kernel/sched/sched.h linux-4.4.62/kernel/sched/sched.h ---- linux-4.4.62.orig/kernel/sched/sched.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/sched/sched.h 2017-04-18 17:38:08.218650175 +0200 -@@ -1100,6 +1100,7 @@ +diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h +index 4e5db65d1aab..ae26df3bd63f 100644 +--- a/kernel/sched/sched.h ++++ b/kernel/sched/sched.h +@@ -1100,6 +1100,7 @@ static inline void finish_lock_switch(struct rq *rq, struct task_struct *prev) #define WF_SYNC 0x01 /* waker goes to sleep after wakeup */ #define WF_FORK 0x02 /* child wakeup after fork */ #define WF_MIGRATED 0x4 /* internal use, task got migrated */ @@ -23544,7 +23907,7 @@ diff -Nur linux-4.4.62.orig/kernel/sched/sched.h linux-4.4.62/kernel/sched/sched /* * To aid in avoiding the subversion of "niceness" due to uneven distribution -@@ -1299,6 +1300,15 @@ +@@ -1299,6 +1300,15 @@ extern void init_sched_fair_class(void); extern void resched_curr(struct rq *rq); extern void resched_cpu(int cpu); @@ -23560,9 +23923,11 @@ diff -Nur linux-4.4.62.orig/kernel/sched/sched.h linux-4.4.62/kernel/sched/sched extern struct rt_bandwidth def_rt_bandwidth; extern void init_rt_bandwidth(struct rt_bandwidth *rt_b, u64 period, u64 runtime); -diff -Nur linux-4.4.62.orig/kernel/sched/swait.c linux-4.4.62/kernel/sched/swait.c ---- linux-4.4.62.orig/kernel/sched/swait.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/kernel/sched/swait.c 2017-04-18 17:38:08.218650175 +0200 +diff --git a/kernel/sched/swait.c b/kernel/sched/swait.c +new file mode 100644 +index 000000000000..205fe36868f9 +--- /dev/null ++++ b/kernel/sched/swait.c @@ -0,0 +1,143 @@ +#include +#include @@ -23707,9 +24072,11 @@ diff -Nur linux-4.4.62.orig/kernel/sched/swait.c linux-4.4.62/kernel/sched/swait + } +} +EXPORT_SYMBOL(finish_swait); -diff -Nur linux-4.4.62.orig/kernel/sched/swork.c linux-4.4.62/kernel/sched/swork.c ---- linux-4.4.62.orig/kernel/sched/swork.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/kernel/sched/swork.c 2017-04-18 17:38:08.218650175 +0200 +diff --git a/kernel/sched/swork.c b/kernel/sched/swork.c +new file mode 100644 +index 000000000000..1950f40ca725 +--- /dev/null ++++ b/kernel/sched/swork.c @@ -0,0 +1,173 @@ +/* + * Copyright (C) 2014 BMW Car IT GmbH, Daniel Wagner daniel.wagner@bmw-carit.de @@ -23884,9 +24251,10 @@ diff -Nur linux-4.4.62.orig/kernel/sched/swork.c linux-4.4.62/kernel/sched/swork + mutex_unlock(&worker_mutex); +} +EXPORT_SYMBOL_GPL(swork_put); -diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c ---- linux-4.4.62.orig/kernel/signal.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/signal.c 2017-04-18 17:38:08.222650330 +0200 +diff --git a/kernel/signal.c b/kernel/signal.c +index f3f1f7a972fd..bc2c990f3f63 100644 +--- a/kernel/signal.c ++++ b/kernel/signal.c @@ -14,6 +14,7 @@ #include #include @@ -23895,7 +24263,7 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c #include #include #include -@@ -352,13 +353,30 @@ +@@ -352,13 +353,30 @@ static bool task_participate_group_stop(struct task_struct *task) return false; } @@ -23927,7 +24295,7 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c { struct sigqueue *q = NULL; struct user_struct *user; -@@ -375,7 +393,10 @@ +@@ -375,7 +393,10 @@ __sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimi if (override_rlimit || atomic_read(&user->sigpending) <= task_rlimit(t, RLIMIT_SIGPENDING)) { @@ -23939,7 +24307,7 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c } else { print_dropped_signal(sig); } -@@ -392,6 +413,13 @@ +@@ -392,6 +413,13 @@ __sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimi return q; } @@ -23953,7 +24321,7 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c static void __sigqueue_free(struct sigqueue *q) { if (q->flags & SIGQUEUE_PREALLOC) -@@ -401,6 +429,21 @@ +@@ -401,6 +429,21 @@ static void __sigqueue_free(struct sigqueue *q) kmem_cache_free(sigqueue_cachep, q); } @@ -23975,7 +24343,7 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c void flush_sigqueue(struct sigpending *queue) { struct sigqueue *q; -@@ -414,6 +457,21 @@ +@@ -414,6 +457,21 @@ void flush_sigqueue(struct sigpending *queue) } /* @@ -23997,7 +24365,7 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c * Flush all pending signals for this kthread. */ void flush_signals(struct task_struct *t) -@@ -525,7 +583,7 @@ +@@ -525,7 +583,7 @@ static void collect_signal(int sig, struct sigpending *list, siginfo_t *info) still_pending: list_del_init(&first->list); copy_siginfo(info, &first->info); @@ -24006,7 +24374,7 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c } else { /* * Ok, it wasn't in the queue. This must be -@@ -560,6 +618,8 @@ +@@ -560,6 +618,8 @@ int dequeue_signal(struct task_struct *tsk, sigset_t *mask, siginfo_t *info) { int signr; @@ -24015,7 +24383,7 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c /* We only dequeue private signals from ourselves, we don't let * signalfd steal them */ -@@ -1156,8 +1216,8 @@ +@@ -1156,8 +1216,8 @@ int do_send_sig_info(int sig, struct siginfo *info, struct task_struct *p, * We don't want to have recursive SIGSEGV's etc, for example, * that is why we also clear SIGNAL_UNKILLABLE. */ @@ -24026,7 +24394,7 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c { unsigned long int flags; int ret, blocked, ignored; -@@ -1182,6 +1242,39 @@ +@@ -1182,6 +1242,39 @@ force_sig_info(int sig, struct siginfo *info, struct task_struct *t) return ret; } @@ -24066,7 +24434,7 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c /* * Nuke all other threads in the group. */ -@@ -1216,12 +1309,12 @@ +@@ -1216,12 +1309,12 @@ struct sighand_struct *__lock_task_sighand(struct task_struct *tsk, * Disable interrupts early to avoid deadlocks. * See rcu_read_unlock() comment header for details. */ @@ -24081,7 +24449,7 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c break; } /* -@@ -1242,7 +1335,7 @@ +@@ -1242,7 +1335,7 @@ struct sighand_struct *__lock_task_sighand(struct task_struct *tsk, } spin_unlock(&sighand->siglock); rcu_read_unlock(); @@ -24090,7 +24458,7 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c } return sighand; -@@ -1485,7 +1578,8 @@ +@@ -1485,7 +1578,8 @@ EXPORT_SYMBOL(kill_pid); */ struct sigqueue *sigqueue_alloc(void) { @@ -24100,7 +24468,7 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c if (q) q->flags |= SIGQUEUE_PREALLOC; -@@ -1846,15 +1940,7 @@ +@@ -1846,15 +1940,7 @@ static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info) if (gstop_done && ptrace_reparented(current)) do_notify_parent_cldstop(current, false, why); @@ -24116,9 +24484,10 @@ diff -Nur linux-4.4.62.orig/kernel/signal.c linux-4.4.62/kernel/signal.c freezable_schedule(); } else { /* -diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c ---- linux-4.4.62.orig/kernel/softirq.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/softirq.c 2017-04-18 17:38:08.222650330 +0200 +diff --git a/kernel/softirq.c b/kernel/softirq.c +index 479e4436f787..cb9c1d5dee10 100644 +--- a/kernel/softirq.c ++++ b/kernel/softirq.c @@ -21,10 +21,12 @@ #include #include @@ -24132,7 +24501,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c #include #define CREATE_TRACE_POINTS -@@ -56,12 +58,108 @@ +@@ -56,12 +58,108 @@ EXPORT_SYMBOL(irq_stat); static struct softirq_action softirq_vec[NR_SOFTIRQS] __cacheline_aligned_in_smp; DEFINE_PER_CPU(struct task_struct *, ksoftirqd); @@ -24241,7 +24610,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c /* * we cannot loop indefinitely here to avoid userspace starvation, * but we also don't want to introduce a worst case 1/HZ latency -@@ -77,6 +175,79 @@ +@@ -77,6 +175,79 @@ static void wakeup_softirqd(void) wake_up_process(tsk); } @@ -24321,7 +24690,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c /* * preempt_count and SOFTIRQ_OFFSET usage: * - preempt_count is changed by SOFTIRQ_OFFSET on entering or leaving -@@ -116,9 +287,9 @@ +@@ -116,9 +287,9 @@ void __local_bh_disable_ip(unsigned long ip, unsigned int cnt) if (preempt_count() == cnt) { #ifdef CONFIG_DEBUG_PREEMPT @@ -24333,7 +24702,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c } } EXPORT_SYMBOL(__local_bh_disable_ip); -@@ -232,10 +403,8 @@ +@@ -232,10 +403,8 @@ asmlinkage __visible void __do_softirq(void) unsigned long end = jiffies + MAX_SOFTIRQ_TIME; unsigned long old_flags = current->flags; int max_restart = MAX_SOFTIRQ_RESTART; @@ -24344,7 +24713,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c /* * Mask out PF_MEMALLOC s current task context is borrowed for the -@@ -254,36 +423,7 @@ +@@ -254,36 +423,7 @@ restart: /* Reset the pending bitmask before enabling irqs */ set_softirq_pending(0); @@ -24382,7 +24751,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c pending = local_softirq_pending(); if (pending) { -@@ -320,6 +460,310 @@ +@@ -320,6 +460,310 @@ asmlinkage __visible void do_softirq(void) } /* @@ -24693,7 +25062,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c * Enter an interrupt context. */ void irq_enter(void) -@@ -330,9 +774,9 @@ +@@ -330,9 +774,9 @@ void irq_enter(void) * Prevent raise_softirq from needlessly waking up ksoftirqd * here, as softirq will be serviced on return from interrupt. */ @@ -24705,7 +25074,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c } __irq_enter(); -@@ -340,6 +784,7 @@ +@@ -340,6 +784,7 @@ void irq_enter(void) static inline void invoke_softirq(void) { @@ -24713,7 +25082,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c if (!force_irqthreads) { #ifdef CONFIG_HAVE_IRQ_EXIT_ON_IRQ_STACK /* -@@ -359,6 +804,18 @@ +@@ -359,6 +804,18 @@ static inline void invoke_softirq(void) } else { wakeup_softirqd(); } @@ -24732,7 +25101,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c } static inline void tick_irq_exit(void) -@@ -395,26 +852,6 @@ +@@ -395,26 +852,6 @@ void irq_exit(void) trace_hardirq_exit(); /* must be last! */ } @@ -24759,7 +25128,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c void raise_softirq(unsigned int nr) { unsigned long flags; -@@ -424,12 +861,6 @@ +@@ -424,12 +861,6 @@ void raise_softirq(unsigned int nr) local_irq_restore(flags); } @@ -24772,7 +25141,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c void open_softirq(int nr, void (*action)(struct softirq_action *)) { softirq_vec[nr].action = action; -@@ -446,15 +877,45 @@ +@@ -446,15 +877,45 @@ struct tasklet_head { static DEFINE_PER_CPU(struct tasklet_head, tasklet_vec); static DEFINE_PER_CPU(struct tasklet_head, tasklet_hi_vec); @@ -24822,7 +25191,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c local_irq_restore(flags); } EXPORT_SYMBOL(__tasklet_schedule); -@@ -464,10 +925,7 @@ +@@ -464,10 +925,7 @@ void __tasklet_hi_schedule(struct tasklet_struct *t) unsigned long flags; local_irq_save(flags); @@ -24834,7 +25203,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c local_irq_restore(flags); } EXPORT_SYMBOL(__tasklet_hi_schedule); -@@ -476,82 +934,122 @@ +@@ -476,82 +934,122 @@ void __tasklet_hi_schedule_first(struct tasklet_struct *t) { BUG_ON(!irqs_disabled()); @@ -25006,7 +25375,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c } void tasklet_init(struct tasklet_struct *t, -@@ -572,7 +1070,7 @@ +@@ -572,7 +1070,7 @@ void tasklet_kill(struct tasklet_struct *t) while (test_and_set_bit(TASKLET_STATE_SCHED, &t->state)) { do { @@ -25015,7 +25384,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c } while (test_bit(TASKLET_STATE_SCHED, &t->state)); } tasklet_unlock_wait(t); -@@ -646,25 +1144,26 @@ +@@ -646,25 +1144,26 @@ void __init softirq_init(void) open_softirq(HI_SOFTIRQ, tasklet_hi_action); } @@ -25057,7 +25426,7 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c } #ifdef CONFIG_HOTPLUG_CPU -@@ -746,16 +1245,31 @@ +@@ -746,16 +1245,31 @@ static struct notifier_block cpu_nfb = { static struct smp_hotplug_thread softirq_threads = { .store = &ksoftirqd, @@ -25089,10 +25458,11 @@ diff -Nur linux-4.4.62.orig/kernel/softirq.c linux-4.4.62/kernel/softirq.c return 0; } -diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machine.c ---- linux-4.4.62.orig/kernel/stop_machine.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/stop_machine.c 2017-04-18 17:38:08.222650330 +0200 -@@ -37,7 +37,7 @@ +diff --git a/kernel/stop_machine.c b/kernel/stop_machine.c +index a3bbaee77c58..f84d3b45cda7 100644 +--- a/kernel/stop_machine.c ++++ b/kernel/stop_machine.c +@@ -37,7 +37,7 @@ struct cpu_stop_done { struct cpu_stopper { struct task_struct *thread; @@ -25101,7 +25471,7 @@ diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machi bool enabled; /* is this stopper enabled? */ struct list_head works; /* list of pending works */ -@@ -86,12 +86,12 @@ +@@ -86,12 +86,12 @@ static void cpu_stop_queue_work(unsigned int cpu, struct cpu_stop_work *work) struct cpu_stopper *stopper = &per_cpu(cpu_stopper, cpu); unsigned long flags; @@ -25116,7 +25486,7 @@ diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machi } /** -@@ -224,8 +224,8 @@ +@@ -224,8 +224,8 @@ static int cpu_stop_queue_two_works(int cpu1, struct cpu_stop_work *work1, int err; lg_double_lock(&stop_cpus_lock, cpu1, cpu2); @@ -25127,7 +25497,7 @@ diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machi err = -ENOENT; if (!stopper1->enabled || !stopper2->enabled) -@@ -235,8 +235,8 @@ +@@ -235,8 +235,8 @@ static int cpu_stop_queue_two_works(int cpu1, struct cpu_stop_work *work1, __cpu_stop_queue_work(stopper1, work1); __cpu_stop_queue_work(stopper2, work2); unlock: @@ -25138,7 +25508,7 @@ diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machi lg_double_unlock(&stop_cpus_lock, cpu1, cpu2); return err; -@@ -258,7 +258,7 @@ +@@ -258,7 +258,7 @@ int stop_two_cpus(unsigned int cpu1, unsigned int cpu2, cpu_stop_fn_t fn, void * struct cpu_stop_work work1, work2; struct multi_stop_data msdata; @@ -25147,7 +25517,7 @@ diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machi msdata = (struct multi_stop_data){ .fn = fn, .data = arg, -@@ -278,11 +278,11 @@ +@@ -278,11 +278,11 @@ int stop_two_cpus(unsigned int cpu1, unsigned int cpu2, cpu_stop_fn_t fn, void * if (cpu1 > cpu2) swap(cpu1, cpu2); if (cpu_stop_queue_two_works(cpu1, &work1, cpu2, &work2)) { @@ -25161,7 +25531,7 @@ diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machi wait_for_completion(&done.completion); -@@ -315,17 +315,20 @@ +@@ -315,17 +315,20 @@ static DEFINE_MUTEX(stop_cpus_mutex); static void queue_stop_cpus_work(const struct cpumask *cpumask, cpu_stop_fn_t fn, void *arg, @@ -25187,7 +25557,7 @@ diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machi for_each_cpu(cpu, cpumask) { work = &per_cpu(cpu_stopper.stop_work, cpu); work->fn = fn; -@@ -342,7 +345,7 @@ +@@ -342,7 +345,7 @@ static int __stop_cpus(const struct cpumask *cpumask, struct cpu_stop_done done; cpu_stop_init_done(&done, cpumask_weight(cpumask)); @@ -25196,7 +25566,7 @@ diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machi wait_for_completion(&done.completion); return done.executed ? done.ret : -ENOENT; } -@@ -422,9 +425,9 @@ +@@ -422,9 +425,9 @@ static int cpu_stop_should_run(unsigned int cpu) unsigned long flags; int run; @@ -25208,7 +25578,7 @@ diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machi return run; } -@@ -436,13 +439,13 @@ +@@ -436,13 +439,13 @@ static void cpu_stopper_thread(unsigned int cpu) repeat: work = NULL; @@ -25224,7 +25594,7 @@ diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machi if (work) { cpu_stop_fn_t fn = work->fn; -@@ -450,6 +453,16 @@ +@@ -450,6 +453,16 @@ repeat: struct cpu_stop_done *done = work->done; char ksym_buf[KSYM_NAME_LEN] __maybe_unused; @@ -25241,7 +25611,7 @@ diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machi /* cpu stop callbacks are not allowed to sleep */ preempt_disable(); -@@ -520,10 +533,12 @@ +@@ -520,10 +533,12 @@ static int __init cpu_stop_init(void) for_each_possible_cpu(cpu) { struct cpu_stopper *stopper = &per_cpu(cpu_stopper, cpu); @@ -25255,7 +25625,7 @@ diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machi BUG_ON(smpboot_register_percpu_thread(&cpu_stop_threads)); stop_machine_unpark(raw_smp_processor_id()); stop_machine_initialized = true; -@@ -620,7 +635,7 @@ +@@ -620,7 +635,7 @@ int stop_machine_from_inactive_cpu(cpu_stop_fn_t fn, void *data, set_state(&msdata, MULTI_STOP_PREPARE); cpu_stop_init_done(&done, num_active_cpus()); queue_stop_cpus_work(cpu_active_mask, multi_cpu_stop, &msdata, @@ -25264,9 +25634,10 @@ diff -Nur linux-4.4.62.orig/kernel/stop_machine.c linux-4.4.62/kernel/stop_machi ret = multi_cpu_stop(&msdata); /* Busy wait for completion. */ -diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtimer.c ---- linux-4.4.62.orig/kernel/time/hrtimer.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/time/hrtimer.c 2017-04-18 17:38:08.222650330 +0200 +diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c +index 17f7bcff1e02..ba3d60144838 100644 +--- a/kernel/time/hrtimer.c ++++ b/kernel/time/hrtimer.c @@ -48,11 +48,13 @@ #include #include @@ -25281,7 +25652,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim #include "tick-internal.h" -@@ -717,6 +719,44 @@ +@@ -717,6 +719,44 @@ static void clock_was_set_work(struct work_struct *work) static DECLARE_WORK(hrtimer_work, clock_was_set_work); @@ -25326,7 +25697,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim /* * Called from timekeeping and resume code to reprogramm the hrtimer * interrupt device on all cpus. -@@ -725,6 +765,7 @@ +@@ -725,6 +765,7 @@ void clock_was_set_delayed(void) { schedule_work(&hrtimer_work); } @@ -25334,7 +25705,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim #else -@@ -734,11 +775,8 @@ +@@ -734,11 +775,8 @@ static inline int hrtimer_is_hres_enabled(void) { return 0; } static inline void hrtimer_switch_to_hres(void) { } static inline void hrtimer_force_reprogram(struct hrtimer_cpu_base *base, int skip_equal) { } @@ -25348,7 +25719,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim static inline void hrtimer_init_hres(struct hrtimer_cpu_base *base) { } static inline void retrigger_next_event(void *arg) { } -@@ -870,6 +908,32 @@ +@@ -870,6 +908,32 @@ u64 hrtimer_forward(struct hrtimer *timer, ktime_t now, ktime_t interval) } EXPORT_SYMBOL_GPL(hrtimer_forward); @@ -25381,7 +25752,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim /* * enqueue_hrtimer - internal function to (re)start a timer * -@@ -911,6 +975,11 @@ +@@ -911,6 +975,11 @@ static void __remove_hrtimer(struct hrtimer *timer, if (!(state & HRTIMER_STATE_ENQUEUED)) return; @@ -25393,7 +25764,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim if (!timerqueue_del(&base->active, &timer->node)) cpu_base->active_bases &= ~(1 << base->index); -@@ -1006,7 +1075,16 @@ +@@ -1006,7 +1075,16 @@ void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, new_base = switch_hrtimer_base(timer, base, mode & HRTIMER_MODE_PINNED); timer_stats_hrtimer_set_start_info(timer); @@ -25410,7 +25781,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim leftmost = enqueue_hrtimer(timer, new_base); if (!leftmost) goto unlock; -@@ -1078,7 +1156,7 @@ +@@ -1078,7 +1156,7 @@ int hrtimer_cancel(struct hrtimer *timer) if (ret >= 0) return ret; @@ -25419,7 +25790,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim } } EXPORT_SYMBOL_GPL(hrtimer_cancel); -@@ -1142,6 +1220,7 @@ +@@ -1142,6 +1220,7 @@ static void __hrtimer_init(struct hrtimer *timer, clockid_t clock_id, base = hrtimer_clockid_to_base(clock_id); timer->base = &cpu_base->clock_base[base]; @@ -25427,7 +25798,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim timerqueue_init(&timer->node); #ifdef CONFIG_TIMER_STATS -@@ -1182,6 +1261,7 @@ +@@ -1182,6 +1261,7 @@ bool hrtimer_active(const struct hrtimer *timer) seq = raw_read_seqcount_begin(&cpu_base->seq); if (timer->state != HRTIMER_STATE_INACTIVE || @@ -25435,7 +25806,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim cpu_base->running == timer) return true; -@@ -1280,10 +1360,112 @@ +@@ -1280,10 +1360,112 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, cpu_base->running = NULL; } @@ -25548,7 +25919,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim for (; active; base++, active >>= 1) { struct timerqueue_node *node; -@@ -1299,6 +1481,15 @@ +@@ -1299,6 +1481,15 @@ static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now) timer = container_of(node, struct hrtimer, node); @@ -25564,7 +25935,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim /* * The immediate goal for using the softexpires is * minimizing wakeups, not running timers at the -@@ -1314,9 +1505,14 @@ +@@ -1314,9 +1505,14 @@ static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now) if (basenow.tv64 < hrtimer_get_softexpires_tv64(timer)) break; @@ -25580,7 +25951,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim } #ifdef CONFIG_HIGH_RES_TIMERS -@@ -1479,16 +1675,18 @@ +@@ -1479,16 +1675,18 @@ static enum hrtimer_restart hrtimer_wakeup(struct hrtimer *timer) void hrtimer_init_sleeper(struct hrtimer_sleeper *sl, struct task_struct *task) { sl->timer.function = hrtimer_wakeup; @@ -25601,7 +25972,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim hrtimer_start_expires(&t->timer, mode); if (likely(t->task)) -@@ -1530,7 +1728,8 @@ +@@ -1530,7 +1728,8 @@ long __sched hrtimer_nanosleep_restart(struct restart_block *restart) HRTIMER_MODE_ABS); hrtimer_set_expires_tv64(&t.timer, restart->nanosleep.expires); @@ -25611,7 +25982,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim goto out; rmtp = restart->nanosleep.rmtp; -@@ -1547,8 +1746,10 @@ +@@ -1547,8 +1746,10 @@ out: return ret; } @@ -25624,7 +25995,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim { struct restart_block *restart; struct hrtimer_sleeper t; -@@ -1561,7 +1762,7 @@ +@@ -1561,7 +1762,7 @@ long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, hrtimer_init_on_stack(&t.timer, clockid, mode); hrtimer_set_expires_range_ns(&t.timer, timespec_to_ktime(*rqtp), slack); @@ -25633,7 +26004,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim goto out; /* Absolute timers do not update the rmtp value and restart: */ -@@ -1588,6 +1789,12 @@ +@@ -1588,6 +1789,12 @@ out: return ret; } @@ -25646,7 +26017,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim SYSCALL_DEFINE2(nanosleep, struct timespec __user *, rqtp, struct timespec __user *, rmtp) { -@@ -1602,6 +1809,26 @@ +@@ -1602,6 +1809,26 @@ SYSCALL_DEFINE2(nanosleep, struct timespec __user *, rqtp, return hrtimer_nanosleep(&tu, rmtp, HRTIMER_MODE_REL, CLOCK_MONOTONIC); } @@ -25673,7 +26044,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim /* * Functions related to boot-time initialization: */ -@@ -1613,10 +1840,14 @@ +@@ -1613,10 +1840,14 @@ static void init_hrtimers_cpu(int cpu) for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) { cpu_base->clock_base[i].cpu_base = cpu_base; timerqueue_init_head(&cpu_base->clock_base[i].active); @@ -25688,7 +26059,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim } #ifdef CONFIG_HOTPLUG_CPU -@@ -1714,11 +1945,21 @@ +@@ -1714,11 +1945,21 @@ static struct notifier_block hrtimers_nb = { .notifier_call = hrtimer_cpu_notify, }; @@ -25710,10 +26081,11 @@ diff -Nur linux-4.4.62.orig/kernel/time/hrtimer.c linux-4.4.62/kernel/time/hrtim } /** -diff -Nur linux-4.4.62.orig/kernel/time/itimer.c linux-4.4.62/kernel/time/itimer.c ---- linux-4.4.62.orig/kernel/time/itimer.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/time/itimer.c 2017-04-18 17:38:08.222650330 +0200 -@@ -213,6 +213,7 @@ +diff --git a/kernel/time/itimer.c b/kernel/time/itimer.c +index 1d5c7204ddc9..184de6751180 100644 +--- a/kernel/time/itimer.c ++++ b/kernel/time/itimer.c +@@ -213,6 +213,7 @@ again: /* We are sharing ->siglock with it_real_fn() */ if (hrtimer_try_to_cancel(timer) < 0) { spin_unlock_irq(&tsk->sighand->siglock); @@ -25721,10 +26093,11 @@ diff -Nur linux-4.4.62.orig/kernel/time/itimer.c linux-4.4.62/kernel/time/itimer goto again; } expires = timeval_to_ktime(value->it_value); -diff -Nur linux-4.4.62.orig/kernel/time/jiffies.c linux-4.4.62/kernel/time/jiffies.c ---- linux-4.4.62.orig/kernel/time/jiffies.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/time/jiffies.c 2017-04-18 17:38:08.222650330 +0200 -@@ -74,7 +74,8 @@ +diff --git a/kernel/time/jiffies.c b/kernel/time/jiffies.c +index 347fecf86a3f..2ede47408a3e 100644 +--- a/kernel/time/jiffies.c ++++ b/kernel/time/jiffies.c +@@ -74,7 +74,8 @@ static struct clocksource clocksource_jiffies = { .max_cycles = 10, }; @@ -25734,7 +26107,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/jiffies.c linux-4.4.62/kernel/time/jiffi #if (BITS_PER_LONG < 64) u64 get_jiffies_64(void) -@@ -83,9 +84,9 @@ +@@ -83,9 +84,9 @@ u64 get_jiffies_64(void) u64 ret; do { @@ -25746,9 +26119,10 @@ diff -Nur linux-4.4.62.orig/kernel/time/jiffies.c linux-4.4.62/kernel/time/jiffi return ret; } EXPORT_SYMBOL(get_jiffies_64); -diff -Nur linux-4.4.62.orig/kernel/time/ntp.c linux-4.4.62/kernel/time/ntp.c ---- linux-4.4.62.orig/kernel/time/ntp.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/time/ntp.c 2017-04-18 17:38:08.222650330 +0200 +diff --git a/kernel/time/ntp.c b/kernel/time/ntp.c +index ab861771e37f..0f6868fd2de6 100644 +--- a/kernel/time/ntp.c ++++ b/kernel/time/ntp.c @@ -10,6 +10,7 @@ #include #include @@ -25757,7 +26131,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/ntp.c linux-4.4.62/kernel/time/ntp.c #include #include #include -@@ -562,10 +563,52 @@ +@@ -562,10 +563,52 @@ static void sync_cmos_clock(struct work_struct *work) &sync_cmos_work, timespec64_to_jiffies(&next)); } @@ -25810,9 +26184,10 @@ diff -Nur linux-4.4.62.orig/kernel/time/ntp.c linux-4.4.62/kernel/time/ntp.c #else void ntp_notify_cmos_timer(void) { } -diff -Nur linux-4.4.62.orig/kernel/time/posix-cpu-timers.c linux-4.4.62/kernel/time/posix-cpu-timers.c ---- linux-4.4.62.orig/kernel/time/posix-cpu-timers.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/time/posix-cpu-timers.c 2017-04-18 17:38:08.222650330 +0200 +diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c +index 80016b329d94..b7342b6e6a5a 100644 +--- a/kernel/time/posix-cpu-timers.c ++++ b/kernel/time/posix-cpu-timers.c @@ -3,6 +3,7 @@ */ @@ -25821,7 +26196,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/posix-cpu-timers.c linux-4.4.62/kernel/t #include #include #include -@@ -650,7 +651,7 @@ +@@ -650,7 +651,7 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int timer_flags, /* * Disarm any old timer after extracting its expiry time. */ @@ -25830,7 +26205,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/posix-cpu-timers.c linux-4.4.62/kernel/t ret = 0; old_incr = timer->it.cpu.incr; -@@ -1092,7 +1093,7 @@ +@@ -1092,7 +1093,7 @@ void posix_cpu_timer_schedule(struct k_itimer *timer) /* * Now re-arm for the new expiry time. */ @@ -25839,7 +26214,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/posix-cpu-timers.c linux-4.4.62/kernel/t arm_timer(timer); unlock_task_sighand(p, &flags); -@@ -1183,13 +1184,13 @@ +@@ -1183,13 +1184,13 @@ static inline int fastpath_timer_check(struct task_struct *tsk) * already updated our counts. We need to check if any timers fire now. * Interrupts are disabled. */ @@ -25855,7 +26230,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/posix-cpu-timers.c linux-4.4.62/kernel/t /* * The fast path checks that there are no expired thread or thread -@@ -1243,6 +1244,190 @@ +@@ -1243,6 +1244,190 @@ void run_posix_cpu_timers(struct task_struct *tsk) } } @@ -26046,10 +26421,11 @@ diff -Nur linux-4.4.62.orig/kernel/time/posix-cpu-timers.c linux-4.4.62/kernel/t /* * Set one of the process-wide special case CPU timers or RLIMIT_CPU. * The tsk->sighand->siglock must be held by the caller. -diff -Nur linux-4.4.62.orig/kernel/time/posix-timers.c linux-4.4.62/kernel/time/posix-timers.c ---- linux-4.4.62.orig/kernel/time/posix-timers.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/time/posix-timers.c 2017-04-18 17:38:08.222650330 +0200 -@@ -506,6 +506,7 @@ +diff --git a/kernel/time/posix-timers.c b/kernel/time/posix-timers.c +index f2826c35e918..464a98155a0e 100644 +--- a/kernel/time/posix-timers.c ++++ b/kernel/time/posix-timers.c +@@ -506,6 +506,7 @@ static enum hrtimer_restart posix_timer_fn(struct hrtimer *timer) static struct pid *good_sigevent(sigevent_t * event) { struct task_struct *rtn = current->group_leader; @@ -26057,7 +26433,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/posix-timers.c linux-4.4.62/kernel/time/ if ((event->sigev_notify & SIGEV_THREAD_ID ) && (!(rtn = find_task_by_vpid(event->sigev_notify_thread_id)) || -@@ -514,7 +515,8 @@ +@@ -514,7 +515,8 @@ static struct pid *good_sigevent(sigevent_t * event) return NULL; if (((event->sigev_notify & ~SIGEV_THREAD_ID) != SIGEV_NONE) && @@ -26067,7 +26443,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/posix-timers.c linux-4.4.62/kernel/time/ return NULL; return task_pid(rtn); -@@ -826,6 +828,20 @@ +@@ -826,6 +828,20 @@ SYSCALL_DEFINE1(timer_getoverrun, timer_t, timer_id) return overrun; } @@ -26088,7 +26464,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/posix-timers.c linux-4.4.62/kernel/time/ /* Set a POSIX.1b interval timer. */ /* timr->it_lock is taken. */ static int -@@ -903,6 +919,7 @@ +@@ -903,6 +919,7 @@ retry: if (!timr) return -EINVAL; @@ -26096,7 +26472,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/posix-timers.c linux-4.4.62/kernel/time/ kc = clockid_to_kclock(timr->it_clock); if (WARN_ON_ONCE(!kc || !kc->timer_set)) error = -EINVAL; -@@ -911,9 +928,12 @@ +@@ -911,9 +928,12 @@ retry: unlock_timer(timr, flag); if (error == TIMER_RETRY) { @@ -26109,7 +26485,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/posix-timers.c linux-4.4.62/kernel/time/ if (old_setting && !error && copy_to_user(old_setting, &old_spec, sizeof (old_spec))) -@@ -951,10 +971,15 @@ +@@ -951,10 +971,15 @@ retry_delete: if (!timer) return -EINVAL; @@ -26125,7 +26501,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/posix-timers.c linux-4.4.62/kernel/time/ spin_lock(¤t->sighand->siglock); list_del(&timer->list); -@@ -980,8 +1005,18 @@ +@@ -980,8 +1005,18 @@ static void itimer_delete(struct k_itimer *timer) retry_delete: spin_lock_irqsave(&timer->it_lock, flags); @@ -26144,20 +26520,22 @@ diff -Nur linux-4.4.62.orig/kernel/time/posix-timers.c linux-4.4.62/kernel/time/ goto retry_delete; } list_del(&timer->list); -diff -Nur linux-4.4.62.orig/kernel/time/tick-broadcast-hrtimer.c linux-4.4.62/kernel/time/tick-broadcast-hrtimer.c ---- linux-4.4.62.orig/kernel/time/tick-broadcast-hrtimer.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/time/tick-broadcast-hrtimer.c 2017-04-18 17:38:08.222650330 +0200 -@@ -106,5 +106,6 @@ +diff --git a/kernel/time/tick-broadcast-hrtimer.c b/kernel/time/tick-broadcast-hrtimer.c +index 53d7184da0be..1b4ac3361c3f 100644 +--- a/kernel/time/tick-broadcast-hrtimer.c ++++ b/kernel/time/tick-broadcast-hrtimer.c +@@ -106,5 +106,6 @@ void tick_setup_hrtimer_broadcast(void) { hrtimer_init(&bctimer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); bctimer.function = bc_handler; + bctimer.irqsafe = true; clockevents_register_device(&ce_broadcast_hrtimer); } -diff -Nur linux-4.4.62.orig/kernel/time/tick-common.c linux-4.4.62/kernel/time/tick-common.c ---- linux-4.4.62.orig/kernel/time/tick-common.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/time/tick-common.c 2017-04-18 17:38:08.222650330 +0200 -@@ -79,13 +79,15 @@ +diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c +index 4fcd99e12aa0..5a47f2e98faf 100644 +--- a/kernel/time/tick-common.c ++++ b/kernel/time/tick-common.c +@@ -79,13 +79,15 @@ int tick_is_oneshot_available(void) static void tick_periodic(int cpu) { if (tick_do_timer_cpu == cpu) { @@ -26175,7 +26553,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/tick-common.c linux-4.4.62/kernel/time/t update_wall_time(); } -@@ -157,9 +159,9 @@ +@@ -157,9 +159,9 @@ void tick_setup_periodic(struct clock_event_device *dev, int broadcast) ktime_t next; do { @@ -26187,10 +26565,11 @@ diff -Nur linux-4.4.62.orig/kernel/time/tick-common.c linux-4.4.62/kernel/time/t clockevents_switch_state(dev, CLOCK_EVT_STATE_ONESHOT); -diff -Nur linux-4.4.62.orig/kernel/time/tick-sched.c linux-4.4.62/kernel/time/tick-sched.c ---- linux-4.4.62.orig/kernel/time/tick-sched.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/time/tick-sched.c 2017-04-18 17:38:08.222650330 +0200 -@@ -62,7 +62,8 @@ +diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c +index 22c57e191a23..d536824cbd36 100644 +--- a/kernel/time/tick-sched.c ++++ b/kernel/time/tick-sched.c +@@ -62,7 +62,8 @@ static void tick_do_update_jiffies64(ktime_t now) return; /* Reevalute with jiffies_lock held */ @@ -26200,7 +26579,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/tick-sched.c linux-4.4.62/kernel/time/ti delta = ktime_sub(now, last_jiffies_update); if (delta.tv64 >= tick_period.tv64) { -@@ -85,10 +86,12 @@ +@@ -85,10 +86,12 @@ static void tick_do_update_jiffies64(ktime_t now) /* Keep the tick_next_period variable up to date */ tick_next_period = ktime_add(last_jiffies_update, tick_period); } else { @@ -26215,7 +26594,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/tick-sched.c linux-4.4.62/kernel/time/ti update_wall_time(); } -@@ -99,12 +102,14 @@ +@@ -99,12 +102,14 @@ static ktime_t tick_init_jiffy_update(void) { ktime_t period; @@ -26232,7 +26611,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/tick-sched.c linux-4.4.62/kernel/time/ti return period; } -@@ -176,6 +181,11 @@ +@@ -176,6 +181,11 @@ static bool can_stop_full_tick(void) return false; } @@ -26244,7 +26623,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/tick-sched.c linux-4.4.62/kernel/time/ti /* sched_clock_tick() needs us? */ #ifdef CONFIG_HAVE_UNSTABLE_SCHED_CLOCK /* -@@ -204,6 +214,7 @@ +@@ -204,6 +214,7 @@ static void nohz_full_kick_work_func(struct irq_work *work) static DEFINE_PER_CPU(struct irq_work, nohz_full_kick_work) = { .func = nohz_full_kick_work_func, @@ -26252,7 +26631,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/tick-sched.c linux-4.4.62/kernel/time/ti }; /* -@@ -578,10 +589,10 @@ +@@ -578,10 +589,10 @@ static ktime_t tick_nohz_stop_sched_tick(struct tick_sched *ts, /* Read jiffies and the time when jiffies were updated last */ do { @@ -26265,7 +26644,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/tick-sched.c linux-4.4.62/kernel/time/ti ts->last_jiffies = basejiff; if (rcu_needs_cpu(basemono, &next_rcu) || -@@ -753,14 +764,7 @@ +@@ -753,14 +764,7 @@ static bool can_stop_idle_tick(int cpu, struct tick_sched *ts) return false; if (unlikely(local_softirq_pending() && cpu_online(cpu))) { @@ -26281,7 +26660,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/tick-sched.c linux-4.4.62/kernel/time/ti return false; } -@@ -1100,6 +1104,7 @@ +@@ -1100,6 +1104,7 @@ void tick_setup_sched_timer(void) * Emulate tick processing via per-CPU hrtimers: */ hrtimer_init(&ts->sched_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); @@ -26289,10 +26668,11 @@ diff -Nur linux-4.4.62.orig/kernel/time/tick-sched.c linux-4.4.62/kernel/time/ti ts->sched_timer.function = tick_sched_timer; /* Get the next period (per cpu) */ -diff -Nur linux-4.4.62.orig/kernel/time/timekeeping.c linux-4.4.62/kernel/time/timekeeping.c ---- linux-4.4.62.orig/kernel/time/timekeeping.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/time/timekeeping.c 2017-04-18 17:38:08.222650330 +0200 -@@ -2070,8 +2070,10 @@ +diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c +index 738012d68117..e060b34d5603 100644 +--- a/kernel/time/timekeeping.c ++++ b/kernel/time/timekeeping.c +@@ -2070,8 +2070,10 @@ EXPORT_SYMBOL(hardpps); */ void xtime_update(unsigned long ticks) { @@ -26305,10 +26685,11 @@ diff -Nur linux-4.4.62.orig/kernel/time/timekeeping.c linux-4.4.62/kernel/time/t + raw_spin_unlock(&jiffies_lock); update_wall_time(); } -diff -Nur linux-4.4.62.orig/kernel/time/timekeeping.h linux-4.4.62/kernel/time/timekeeping.h ---- linux-4.4.62.orig/kernel/time/timekeeping.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/time/timekeeping.h 2017-04-18 17:38:08.222650330 +0200 -@@ -19,7 +19,8 @@ +diff --git a/kernel/time/timekeeping.h b/kernel/time/timekeeping.h +index 704f595ce83f..763a3e5121ff 100644 +--- a/kernel/time/timekeeping.h ++++ b/kernel/time/timekeeping.h +@@ -19,7 +19,8 @@ extern void timekeeping_resume(void); extern void do_timer(unsigned long ticks); extern void update_wall_time(void); @@ -26318,10 +26699,11 @@ diff -Nur linux-4.4.62.orig/kernel/time/timekeeping.h linux-4.4.62/kernel/time/t #define CS_NAME_LEN 32 -diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c ---- linux-4.4.62.orig/kernel/time/timer.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/time/timer.c 2017-04-18 17:38:08.226650485 +0200 -@@ -80,6 +80,9 @@ +diff --git a/kernel/time/timer.c b/kernel/time/timer.c +index bbc5d1114583..603699ff9411 100644 +--- a/kernel/time/timer.c ++++ b/kernel/time/timer.c +@@ -80,6 +80,9 @@ struct tvec_root { struct tvec_base { spinlock_t lock; struct timer_list *running_timer; @@ -26331,7 +26713,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c unsigned long timer_jiffies; unsigned long next_timer; unsigned long active_timers; -@@ -777,6 +780,39 @@ +@@ -777,6 +780,39 @@ static struct tvec_base *lock_timer_base(struct timer_list *timer, cpu_relax(); } } @@ -26371,7 +26753,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c static inline int __mod_timer(struct timer_list *timer, unsigned long expires, -@@ -807,16 +843,8 @@ +@@ -807,16 +843,8 @@ __mod_timer(struct timer_list *timer, unsigned long expires, * handler yet has not finished. This also guarantees that * the timer is serialized wrt itself. */ @@ -26390,7 +26772,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c } timer->expires = expires; -@@ -1006,6 +1034,33 @@ +@@ -1006,6 +1034,33 @@ void add_timer_on(struct timer_list *timer, int cpu) } EXPORT_SYMBOL_GPL(add_timer_on); @@ -26424,7 +26806,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c /** * del_timer - deactive a timer. * @timer: the timer to be deactivated -@@ -1063,7 +1118,7 @@ +@@ -1063,7 +1118,7 @@ int try_to_del_timer_sync(struct timer_list *timer) } EXPORT_SYMBOL(try_to_del_timer_sync); @@ -26433,7 +26815,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c /** * del_timer_sync - deactivate a timer and wait for the handler to finish. * @timer: the timer to be deactivated -@@ -1123,7 +1178,7 @@ +@@ -1123,7 +1178,7 @@ int del_timer_sync(struct timer_list *timer) int ret = try_to_del_timer_sync(timer); if (ret >= 0) return ret; @@ -26442,7 +26824,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c } } EXPORT_SYMBOL(del_timer_sync); -@@ -1248,16 +1303,18 @@ +@@ -1248,16 +1303,18 @@ static inline void __run_timers(struct tvec_base *base) if (irqsafe) { spin_unlock(&base->lock); call_timer_fn(timer, fn, data); @@ -26462,7 +26844,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c } #ifdef CONFIG_NO_HZ_COMMON -@@ -1390,6 +1447,14 @@ +@@ -1390,6 +1447,14 @@ u64 get_next_timer_interrupt(unsigned long basej, u64 basem) if (cpu_is_offline(smp_processor_id())) return expires; @@ -26477,7 +26859,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c spin_lock(&base->lock); if (base->active_timers) { if (time_before_eq(base->next_timer, base->timer_jiffies)) -@@ -1416,13 +1481,13 @@ +@@ -1416,13 +1481,13 @@ void update_process_times(int user_tick) /* Note: this timer irq context must be accounted for as well. */ account_process_tick(p, user_tick); @@ -26493,7 +26875,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c run_posix_cpu_timers(p); } -@@ -1433,6 +1498,8 @@ +@@ -1433,6 +1498,8 @@ static void run_timer_softirq(struct softirq_action *h) { struct tvec_base *base = this_cpu_ptr(&tvec_bases); @@ -26502,7 +26884,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c if (time_after_eq(jiffies, base->timer_jiffies)) __run_timers(base); } -@@ -1589,7 +1656,7 @@ +@@ -1589,7 +1656,7 @@ static void migrate_timers(int cpu) BUG_ON(cpu_online(cpu)); old_base = per_cpu_ptr(&tvec_bases, cpu); @@ -26511,7 +26893,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c /* * The caller is globally serialized and nobody else * takes two locks at once, deadlock is not possible. -@@ -1613,7 +1680,7 @@ +@@ -1613,7 +1680,7 @@ static void migrate_timers(int cpu) spin_unlock(&old_base->lock); spin_unlock_irq(&new_base->lock); @@ -26520,7 +26902,7 @@ diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c } static int timer_cpu_notify(struct notifier_block *self, -@@ -1645,6 +1712,9 @@ +@@ -1645,6 +1712,9 @@ static void __init init_timer_cpu(int cpu) base->cpu = cpu; spin_lock_init(&base->lock); @@ -26530,10 +26912,11 @@ diff -Nur linux-4.4.62.orig/kernel/time/timer.c linux-4.4.62/kernel/time/timer.c base->timer_jiffies = jiffies; base->next_timer = base->timer_jiffies; -diff -Nur linux-4.4.62.orig/kernel/trace/Kconfig linux-4.4.62/kernel/trace/Kconfig ---- linux-4.4.62.orig/kernel/trace/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/trace/Kconfig 2017-04-18 17:38:08.226650485 +0200 -@@ -187,6 +187,24 @@ +diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig +index e45db6b0d878..364ccd0eb57b 100644 +--- a/kernel/trace/Kconfig ++++ b/kernel/trace/Kconfig +@@ -187,6 +187,24 @@ config IRQSOFF_TRACER enabled. This option and the preempt-off timing option can be used together or separately.) @@ -26558,7 +26941,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/Kconfig linux-4.4.62/kernel/trace/Kconf config PREEMPT_TRACER bool "Preemption-off Latency Tracer" default n -@@ -211,6 +229,24 @@ +@@ -211,6 +229,24 @@ config PREEMPT_TRACER enabled. This option and the irqs-off timing option can be used together or separately.) @@ -26583,7 +26966,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/Kconfig linux-4.4.62/kernel/trace/Kconf config SCHED_TRACER bool "Scheduling Latency Tracer" select GENERIC_TRACER -@@ -221,6 +257,74 @@ +@@ -221,6 +257,74 @@ config SCHED_TRACER This tracer tracks the latency of the highest priority task to be scheduled in, starting from the point it has woken up. @@ -26658,9 +27041,26 @@ diff -Nur linux-4.4.62.orig/kernel/trace/Kconfig linux-4.4.62/kernel/trace/Kconf config ENABLE_DEFAULT_TRACERS bool "Trace process context switches and events" depends on !GENERIC_TRACER -diff -Nur linux-4.4.62.orig/kernel/trace/latency_hist.c linux-4.4.62/kernel/trace/latency_hist.c ---- linux-4.4.62.orig/kernel/trace/latency_hist.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.4.62/kernel/trace/latency_hist.c 2017-04-18 17:38:08.226650485 +0200 +diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile +index 05ea5167e6bb..bc08c67301ae 100644 +--- a/kernel/trace/Makefile ++++ b/kernel/trace/Makefile +@@ -40,6 +40,10 @@ obj-$(CONFIG_FUNCTION_TRACER) += trace_functions.o + obj-$(CONFIG_IRQSOFF_TRACER) += trace_irqsoff.o + obj-$(CONFIG_PREEMPT_TRACER) += trace_irqsoff.o + obj-$(CONFIG_SCHED_TRACER) += trace_sched_wakeup.o ++obj-$(CONFIG_INTERRUPT_OFF_HIST) += latency_hist.o ++obj-$(CONFIG_PREEMPT_OFF_HIST) += latency_hist.o ++obj-$(CONFIG_WAKEUP_LATENCY_HIST) += latency_hist.o ++obj-$(CONFIG_MISSED_TIMER_OFFSETS_HIST) += latency_hist.o + obj-$(CONFIG_NOP_TRACER) += trace_nop.o + obj-$(CONFIG_STACK_TRACER) += trace_stack.o + obj-$(CONFIG_MMIOTRACE) += trace_mmiotrace.o +diff --git a/kernel/trace/latency_hist.c b/kernel/trace/latency_hist.c +new file mode 100644 +index 000000000000..7f6ee70dea41 +--- /dev/null ++++ b/kernel/trace/latency_hist.c @@ -0,0 +1,1178 @@ +/* + * kernel/trace/latency_hist.c @@ -27840,24 +28240,11 @@ diff -Nur linux-4.4.62.orig/kernel/trace/latency_hist.c linux-4.4.62/kernel/trac +} + +device_initcall(latency_hist_init); -diff -Nur linux-4.4.62.orig/kernel/trace/Makefile linux-4.4.62/kernel/trace/Makefile ---- linux-4.4.62.orig/kernel/trace/Makefile 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/trace/Makefile 2017-04-18 17:38:08.226650485 +0200 -@@ -40,6 +40,10 @@ - obj-$(CONFIG_IRQSOFF_TRACER) += trace_irqsoff.o - obj-$(CONFIG_PREEMPT_TRACER) += trace_irqsoff.o - obj-$(CONFIG_SCHED_TRACER) += trace_sched_wakeup.o -+obj-$(CONFIG_INTERRUPT_OFF_HIST) += latency_hist.o -+obj-$(CONFIG_PREEMPT_OFF_HIST) += latency_hist.o -+obj-$(CONFIG_WAKEUP_LATENCY_HIST) += latency_hist.o -+obj-$(CONFIG_MISSED_TIMER_OFFSETS_HIST) += latency_hist.o - obj-$(CONFIG_NOP_TRACER) += trace_nop.o - obj-$(CONFIG_STACK_TRACER) += trace_stack.o - obj-$(CONFIG_MMIOTRACE) += trace_mmiotrace.o -diff -Nur linux-4.4.62.orig/kernel/trace/trace.c linux-4.4.62/kernel/trace/trace.c ---- linux-4.4.62.orig/kernel/trace/trace.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/trace/trace.c 2017-04-18 17:38:08.226650485 +0200 -@@ -1652,6 +1652,7 @@ +diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c +index 4c21c0b7dc91..d0e52efed27a 100644 +--- a/kernel/trace/trace.c ++++ b/kernel/trace/trace.c +@@ -1652,6 +1652,7 @@ tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags, struct task_struct *tsk = current; entry->preempt_count = pc & 0xff; @@ -27865,7 +28252,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace.c linux-4.4.62/kernel/trace/trace entry->pid = (tsk) ? tsk->pid : 0; entry->flags = #ifdef CONFIG_TRACE_IRQFLAGS_SUPPORT -@@ -1661,8 +1662,11 @@ +@@ -1661,8 +1662,11 @@ tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags, #endif ((pc & HARDIRQ_MASK) ? TRACE_FLAG_HARDIRQ : 0) | ((pc & SOFTIRQ_MASK) ? TRACE_FLAG_SOFTIRQ : 0) | @@ -27878,7 +28265,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace.c linux-4.4.62/kernel/trace/trace } EXPORT_SYMBOL_GPL(tracing_generic_entry_update); -@@ -2555,14 +2559,17 @@ +@@ -2555,14 +2559,17 @@ get_total_entries(struct trace_buffer *buf, static void print_lat_help_header(struct seq_file *m) { @@ -27904,7 +28291,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace.c linux-4.4.62/kernel/trace/trace } static void print_event_info(struct trace_buffer *buf, struct seq_file *m) -@@ -2588,11 +2595,14 @@ +@@ -2588,11 +2595,14 @@ static void print_func_help_header_irq(struct trace_buffer *buf, struct seq_file print_event_info(buf, m); seq_puts(m, "# _-----=> irqs-off\n" "# / _----=> need-resched\n" @@ -27924,10 +28311,31 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace.c linux-4.4.62/kernel/trace/trace } void -diff -Nur linux-4.4.62.orig/kernel/trace/trace_events.c linux-4.4.62/kernel/trace/trace_events.c ---- linux-4.4.62.orig/kernel/trace/trace_events.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/trace/trace_events.c 2017-04-18 17:38:08.226650485 +0200 -@@ -188,6 +188,8 @@ +diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h +index 919d9d07686f..3bf86ece683c 100644 +--- a/kernel/trace/trace.h ++++ b/kernel/trace/trace.h +@@ -117,6 +117,7 @@ struct kretprobe_trace_entry_head { + * NEED_RESCHED - reschedule is requested + * HARDIRQ - inside an interrupt handler + * SOFTIRQ - inside a softirq handler ++ * NEED_RESCHED_LAZY - lazy reschedule is requested + */ + enum trace_flag_type { + TRACE_FLAG_IRQS_OFF = 0x01, +@@ -125,6 +126,7 @@ enum trace_flag_type { + TRACE_FLAG_HARDIRQ = 0x08, + TRACE_FLAG_SOFTIRQ = 0x10, + TRACE_FLAG_PREEMPT_RESCHED = 0x20, ++ TRACE_FLAG_NEED_RESCHED_LAZY = 0x40, + }; + + #define TRACE_BUF_SIZE 1024 +diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c +index 996f0fd34312..5bd79b347398 100644 +--- a/kernel/trace/trace_events.c ++++ b/kernel/trace/trace_events.c +@@ -188,6 +188,8 @@ static int trace_define_common_fields(void) __common_field(unsigned char, flags); __common_field(unsigned char, preempt_count); __common_field(int, pid); @@ -27936,7 +28344,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace_events.c linux-4.4.62/kernel/trac return ret; } -@@ -244,6 +246,14 @@ +@@ -244,6 +246,14 @@ void *trace_event_buffer_reserve(struct trace_event_buffer *fbuffer, local_save_flags(fbuffer->flags); fbuffer->pc = preempt_count(); @@ -27951,28 +28359,10 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace_events.c linux-4.4.62/kernel/trac fbuffer->trace_file = trace_file; fbuffer->event = -diff -Nur linux-4.4.62.orig/kernel/trace/trace.h linux-4.4.62/kernel/trace/trace.h ---- linux-4.4.62.orig/kernel/trace/trace.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/trace/trace.h 2017-04-18 17:38:08.226650485 +0200 -@@ -117,6 +117,7 @@ - * NEED_RESCHED - reschedule is requested - * HARDIRQ - inside an interrupt handler - * SOFTIRQ - inside a softirq handler -+ * NEED_RESCHED_LAZY - lazy reschedule is requested - */ - enum trace_flag_type { - TRACE_FLAG_IRQS_OFF = 0x01, -@@ -125,6 +126,7 @@ - TRACE_FLAG_HARDIRQ = 0x08, - TRACE_FLAG_SOFTIRQ = 0x10, - TRACE_FLAG_PREEMPT_RESCHED = 0x20, -+ TRACE_FLAG_NEED_RESCHED_LAZY = 0x40, - }; - - #define TRACE_BUF_SIZE 1024 -diff -Nur linux-4.4.62.orig/kernel/trace/trace_irqsoff.c linux-4.4.62/kernel/trace/trace_irqsoff.c ---- linux-4.4.62.orig/kernel/trace/trace_irqsoff.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/trace/trace_irqsoff.c 2017-04-18 17:38:08.226650485 +0200 +diff --git a/kernel/trace/trace_irqsoff.c b/kernel/trace/trace_irqsoff.c +index be3222b7d72e..553e71254ad6 100644 +--- a/kernel/trace/trace_irqsoff.c ++++ b/kernel/trace/trace_irqsoff.c @@ -13,6 +13,7 @@ #include #include @@ -27981,7 +28371,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace_irqsoff.c linux-4.4.62/kernel/tra #include "trace.h" -@@ -424,11 +425,13 @@ +@@ -424,11 +425,13 @@ void start_critical_timings(void) { if (preempt_trace() || irq_trace()) start_critical_timing(CALLER_ADDR0, CALLER_ADDR1); @@ -27995,7 +28385,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace_irqsoff.c linux-4.4.62/kernel/tra if (preempt_trace() || irq_trace()) stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1); } -@@ -438,6 +441,7 @@ +@@ -438,6 +441,7 @@ EXPORT_SYMBOL_GPL(stop_critical_timings); #ifdef CONFIG_PROVE_LOCKING void time_hardirqs_on(unsigned long a0, unsigned long a1) { @@ -28003,7 +28393,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace_irqsoff.c linux-4.4.62/kernel/tra if (!preempt_trace() && irq_trace()) stop_critical_timing(a0, a1); } -@@ -446,6 +450,7 @@ +@@ -446,6 +450,7 @@ void time_hardirqs_off(unsigned long a0, unsigned long a1) { if (!preempt_trace() && irq_trace()) start_critical_timing(a0, a1); @@ -28011,7 +28401,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace_irqsoff.c linux-4.4.62/kernel/tra } #else /* !CONFIG_PROVE_LOCKING */ -@@ -471,6 +476,7 @@ +@@ -471,6 +476,7 @@ inline void print_irqtrace_events(struct task_struct *curr) */ void trace_hardirqs_on(void) { @@ -28019,7 +28409,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace_irqsoff.c linux-4.4.62/kernel/tra if (!preempt_trace() && irq_trace()) stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1); } -@@ -480,11 +486,13 @@ +@@ -480,11 +486,13 @@ void trace_hardirqs_off(void) { if (!preempt_trace() && irq_trace()) start_critical_timing(CALLER_ADDR0, CALLER_ADDR1); @@ -28033,7 +28423,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace_irqsoff.c linux-4.4.62/kernel/tra if (!preempt_trace() && irq_trace()) stop_critical_timing(CALLER_ADDR0, caller_addr); } -@@ -494,6 +502,7 @@ +@@ -494,6 +502,7 @@ __visible void trace_hardirqs_off_caller(unsigned long caller_addr) { if (!preempt_trace() && irq_trace()) start_critical_timing(CALLER_ADDR0, caller_addr); @@ -28041,7 +28431,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace_irqsoff.c linux-4.4.62/kernel/tra } EXPORT_SYMBOL(trace_hardirqs_off_caller); -@@ -503,12 +512,14 @@ +@@ -503,12 +512,14 @@ EXPORT_SYMBOL(trace_hardirqs_off_caller); #ifdef CONFIG_PREEMPT_TRACER void trace_preempt_on(unsigned long a0, unsigned long a1) { @@ -28056,10 +28446,11 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace_irqsoff.c linux-4.4.62/kernel/tra if (preempt_trace() && !irq_trace()) start_critical_timing(a0, a1); } -diff -Nur linux-4.4.62.orig/kernel/trace/trace_output.c linux-4.4.62/kernel/trace/trace_output.c ---- linux-4.4.62.orig/kernel/trace/trace_output.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/trace/trace_output.c 2017-04-18 17:38:08.226650485 +0200 -@@ -386,6 +386,7 @@ +diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c +index 282982195e09..9f19d839a756 100644 +--- a/kernel/trace/trace_output.c ++++ b/kernel/trace/trace_output.c +@@ -386,6 +386,7 @@ int trace_print_lat_fmt(struct trace_seq *s, struct trace_entry *entry) { char hardsoft_irq; char need_resched; @@ -28067,7 +28458,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace_output.c linux-4.4.62/kernel/trac char irqs_off; int hardirq; int softirq; -@@ -413,6 +414,8 @@ +@@ -413,6 +414,8 @@ int trace_print_lat_fmt(struct trace_seq *s, struct trace_entry *entry) need_resched = '.'; break; } @@ -28076,7 +28467,7 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace_output.c linux-4.4.62/kernel/trac hardsoft_irq = (hardirq && softirq) ? 'H' : -@@ -420,14 +423,25 @@ +@@ -420,14 +423,25 @@ int trace_print_lat_fmt(struct trace_seq *s, struct trace_entry *entry) softirq ? 's' : '.'; @@ -28104,10 +28495,11 @@ diff -Nur linux-4.4.62.orig/kernel/trace/trace_output.c linux-4.4.62/kernel/trac return !trace_seq_has_overflowed(s); } -diff -Nur linux-4.4.62.orig/kernel/user.c linux-4.4.62/kernel/user.c ---- linux-4.4.62.orig/kernel/user.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/user.c 2017-04-18 17:38:08.226650485 +0200 -@@ -161,11 +161,11 @@ +diff --git a/kernel/user.c b/kernel/user.c +index b069ccbfb0b0..1a2e88e98b5e 100644 +--- a/kernel/user.c ++++ b/kernel/user.c +@@ -161,11 +161,11 @@ void free_uid(struct user_struct *up) if (!up) return; @@ -28121,10 +28513,11 @@ diff -Nur linux-4.4.62.orig/kernel/user.c linux-4.4.62/kernel/user.c } struct user_struct *alloc_uid(kuid_t uid) -diff -Nur linux-4.4.62.orig/kernel/watchdog.c linux-4.4.62/kernel/watchdog.c ---- linux-4.4.62.orig/kernel/watchdog.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/watchdog.c 2017-04-18 17:38:08.226650485 +0200 -@@ -299,6 +299,8 @@ +diff --git a/kernel/watchdog.c b/kernel/watchdog.c +index c1e0b5f429b6..fa2e079cc314 100644 +--- a/kernel/watchdog.c ++++ b/kernel/watchdog.c +@@ -299,6 +299,8 @@ static int is_softlockup(unsigned long touch_ts) #ifdef CONFIG_HARDLOCKUP_DETECTOR @@ -28133,7 +28526,7 @@ diff -Nur linux-4.4.62.orig/kernel/watchdog.c linux-4.4.62/kernel/watchdog.c static struct perf_event_attr wd_hw_attr = { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CPU_CYCLES, -@@ -332,6 +334,13 @@ +@@ -332,6 +334,13 @@ static void watchdog_overflow_callback(struct perf_event *event, /* only print hardlockups once */ if (__this_cpu_read(hard_watchdog_warn) == true) return; @@ -28147,7 +28540,7 @@ diff -Nur linux-4.4.62.orig/kernel/watchdog.c linux-4.4.62/kernel/watchdog.c pr_emerg("Watchdog detected hard LOCKUP on cpu %d", this_cpu); print_modules(); -@@ -349,8 +358,9 @@ +@@ -349,8 +358,9 @@ static void watchdog_overflow_callback(struct perf_event *event, !test_and_set_bit(0, &hardlockup_allcpu_dumped)) trigger_allbutself_cpu_backtrace(); @@ -28158,7 +28551,7 @@ diff -Nur linux-4.4.62.orig/kernel/watchdog.c linux-4.4.62/kernel/watchdog.c __this_cpu_write(hard_watchdog_warn, true); return; -@@ -496,6 +506,7 @@ +@@ -496,6 +506,7 @@ static void watchdog_enable(unsigned int cpu) /* kick off the timer for the hardlockup detector */ hrtimer_init(hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); hrtimer->function = watchdog_timer_fn; @@ -28166,9 +28559,10 @@ diff -Nur linux-4.4.62.orig/kernel/watchdog.c linux-4.4.62/kernel/watchdog.c /* Enable the perf event */ watchdog_nmi_enable(cpu); -diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c ---- linux-4.4.62.orig/kernel/workqueue.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/workqueue.c 2017-04-18 17:38:08.230650641 +0200 +diff --git a/kernel/workqueue.c b/kernel/workqueue.c +index 2c2f971f3e75..d5b0f4fc0eb0 100644 +--- a/kernel/workqueue.c ++++ b/kernel/workqueue.c @@ -48,6 +48,8 @@ #include #include @@ -28178,7 +28572,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c #include "workqueue_internal.h" -@@ -121,11 +123,16 @@ +@@ -121,11 +123,16 @@ enum { * cpu or grabbing pool->lock is enough for read access. If * POOL_DISASSOCIATED is set, it's identical to L. * @@ -28196,7 +28590,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c * * PW: wq_pool_mutex and wq->mutex protected for writes. Either for reads. * -@@ -134,7 +141,7 @@ +@@ -134,7 +141,7 @@ enum { * * WQ: wq->mutex protected. * @@ -28205,7 +28599,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c * * MD: wq_mayday_lock protected. */ -@@ -183,7 +190,7 @@ +@@ -183,7 +190,7 @@ struct worker_pool { atomic_t nr_running ____cacheline_aligned_in_smp; /* @@ -28214,7 +28608,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c * from get_work_pool(). */ struct rcu_head rcu; -@@ -212,7 +219,7 @@ +@@ -212,7 +219,7 @@ struct pool_workqueue { /* * Release of unbound pwq is punted to system_wq. See put_pwq() * and pwq_unbound_release_workfn() for details. pool_workqueue @@ -28223,7 +28617,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c * determined without grabbing wq->mutex. */ struct work_struct unbound_release_work; -@@ -331,6 +338,8 @@ +@@ -331,6 +338,8 @@ EXPORT_SYMBOL_GPL(system_power_efficient_wq); struct workqueue_struct *system_freezable_power_efficient_wq __read_mostly; EXPORT_SYMBOL_GPL(system_freezable_power_efficient_wq); @@ -28232,7 +28626,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c static int worker_thread(void *__worker); static void workqueue_sysfs_unregister(struct workqueue_struct *wq); -@@ -338,20 +347,20 @@ +@@ -338,20 +347,20 @@ static void workqueue_sysfs_unregister(struct workqueue_struct *wq); #include #define assert_rcu_or_pool_mutex() \ @@ -28259,7 +28653,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c #define for_each_cpu_worker_pool(pool, cpu) \ for ((pool) = &per_cpu(cpu_worker_pools, cpu)[0]; \ -@@ -363,7 +372,7 @@ +@@ -363,7 +372,7 @@ static void workqueue_sysfs_unregister(struct workqueue_struct *wq); * @pool: iteration cursor * @pi: integer used for iteration * @@ -28268,7 +28662,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c * locked. If the pool needs to be used beyond the locking in effect, the * caller is responsible for guaranteeing that the pool stays online. * -@@ -395,7 +404,7 @@ +@@ -395,7 +404,7 @@ static void workqueue_sysfs_unregister(struct workqueue_struct *wq); * @pwq: iteration cursor * @wq: the target workqueue * @@ -28277,7 +28671,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c * If the pwq needs to be used beyond the locking in effect, the caller is * responsible for guaranteeing that the pwq stays online. * -@@ -407,6 +416,31 @@ +@@ -407,6 +416,31 @@ static void workqueue_sysfs_unregister(struct workqueue_struct *wq); if (({ assert_rcu_or_wq_mutex(wq); false; })) { } \ else @@ -28309,7 +28703,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c #ifdef CONFIG_DEBUG_OBJECTS_WORK static struct debug_obj_descr work_debug_descr; -@@ -557,7 +591,7 @@ +@@ -557,7 +591,7 @@ static int worker_pool_assign_id(struct worker_pool *pool) * @wq: the target workqueue * @node: the node ID * @@ -28318,7 +28712,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c * read locked. * If the pwq needs to be used beyond the locking in effect, the caller is * responsible for guaranteeing that the pwq stays online. -@@ -701,8 +735,8 @@ +@@ -701,8 +735,8 @@ static struct pool_workqueue *get_work_pwq(struct work_struct *work) * @work: the work item of interest * * Pools are created and destroyed under wq_pool_mutex, and allows read @@ -28329,7 +28723,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c * * All fields of the returned pool are accessible as long as the above * mentioned locking is in effect. If the returned pool needs to be used -@@ -839,51 +873,44 @@ +@@ -839,51 +873,44 @@ static struct worker *first_idle_worker(struct worker_pool *pool) */ static void wake_up_worker(struct worker_pool *pool) { @@ -28400,7 +28794,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c struct worker_pool *pool; /* -@@ -892,29 +919,26 @@ +@@ -892,29 +919,26 @@ struct task_struct *wq_worker_sleeping(struct task_struct *task, int cpu) * checking NOT_RUNNING. */ if (worker->flags & WORKER_NOT_RUNNING) @@ -28440,7 +28834,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c } /** -@@ -1108,12 +1132,14 @@ +@@ -1108,12 +1132,14 @@ static void put_pwq_unlocked(struct pool_workqueue *pwq) { if (pwq) { /* @@ -28458,7 +28852,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c } } -@@ -1215,7 +1241,7 @@ +@@ -1215,7 +1241,7 @@ static int try_to_grab_pending(struct work_struct *work, bool is_dwork, struct worker_pool *pool; struct pool_workqueue *pwq; @@ -28467,7 +28861,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c /* try to steal the timer if it exists */ if (is_dwork) { -@@ -1234,6 +1260,7 @@ +@@ -1234,6 +1260,7 @@ static int try_to_grab_pending(struct work_struct *work, bool is_dwork, if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) return 0; @@ -28475,7 +28869,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c /* * The queueing is in progress, or it is already queued. Try to * steal it from ->worklist without clearing WORK_STRUCT_PENDING. -@@ -1272,14 +1299,16 @@ +@@ -1272,14 +1299,16 @@ static int try_to_grab_pending(struct work_struct *work, bool is_dwork, set_work_pool_and_keep_pending(work, pool->id); spin_unlock(&pool->lock); @@ -28494,7 +28888,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c return -EAGAIN; } -@@ -1348,7 +1377,7 @@ +@@ -1348,7 +1377,7 @@ static void __queue_work(int cpu, struct workqueue_struct *wq, * queued or lose PENDING. Grabbing PENDING and queueing should * happen with IRQ disabled. */ @@ -28503,7 +28897,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c debug_work_activate(work); -@@ -1356,6 +1385,8 @@ +@@ -1356,6 +1385,8 @@ static void __queue_work(int cpu, struct workqueue_struct *wq, if (unlikely(wq->flags & __WQ_DRAINING) && WARN_ON_ONCE(!is_chained_work(wq))) return; @@ -28512,7 +28906,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c retry: if (req_cpu == WORK_CPU_UNBOUND) cpu = raw_smp_processor_id(); -@@ -1412,10 +1443,8 @@ +@@ -1412,10 +1443,8 @@ retry: /* pwq determined, queue */ trace_workqueue_queue_work(req_cpu, pwq, work); @@ -28525,7 +28919,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c pwq->nr_in_flight[pwq->work_color]++; work_flags = work_color_to_flags(pwq->work_color); -@@ -1431,7 +1460,9 @@ +@@ -1431,7 +1460,9 @@ retry: insert_work(pwq, work, worklist, work_flags); @@ -28535,7 +28929,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c } /** -@@ -1451,14 +1482,14 @@ +@@ -1451,14 +1482,14 @@ bool queue_work_on(int cpu, struct workqueue_struct *wq, bool ret = false; unsigned long flags; @@ -28552,7 +28946,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c return ret; } EXPORT_SYMBOL(queue_work_on); -@@ -1525,14 +1556,14 @@ +@@ -1525,14 +1556,14 @@ bool queue_delayed_work_on(int cpu, struct workqueue_struct *wq, unsigned long flags; /* read the comment in __queue_work() */ @@ -28569,7 +28963,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c return ret; } EXPORT_SYMBOL(queue_delayed_work_on); -@@ -1567,7 +1598,7 @@ +@@ -1567,7 +1598,7 @@ bool mod_delayed_work_on(int cpu, struct workqueue_struct *wq, if (likely(ret >= 0)) { __queue_delayed_work(cpu, wq, dwork, delay); @@ -28578,7 +28972,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c } /* -ENOENT from try_to_grab_pending() becomes %true */ -@@ -1600,7 +1631,9 @@ +@@ -1600,7 +1631,9 @@ static void worker_enter_idle(struct worker *worker) worker->last_active = jiffies; /* idle_list is LIFO */ @@ -28588,7 +28982,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c if (too_many_workers(pool) && !timer_pending(&pool->idle_timer)) mod_timer(&pool->idle_timer, jiffies + IDLE_WORKER_TIMEOUT); -@@ -1633,7 +1666,9 @@ +@@ -1633,7 +1666,9 @@ static void worker_leave_idle(struct worker *worker) return; worker_clr_flags(worker, WORKER_IDLE); pool->nr_idle--; @@ -28598,7 +28992,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c } static struct worker *alloc_worker(int node) -@@ -1799,7 +1834,9 @@ +@@ -1799,7 +1834,9 @@ static void destroy_worker(struct worker *worker) pool->nr_workers--; pool->nr_idle--; @@ -28608,7 +29002,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c worker->flags |= WORKER_DIE; wake_up_process(worker->task); } -@@ -2716,14 +2753,14 @@ +@@ -2716,14 +2753,14 @@ static bool start_flush_work(struct work_struct *work, struct wq_barrier *barr) might_sleep(); @@ -28626,7 +29020,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c /* see the comment in try_to_grab_pending() with the same code */ pwq = get_work_pwq(work); if (pwq) { -@@ -2750,10 +2787,11 @@ +@@ -2750,10 +2787,11 @@ static bool start_flush_work(struct work_struct *work, struct wq_barrier *barr) else lock_map_acquire_read(&pwq->wq->lockdep_map); lock_map_release(&pwq->wq->lockdep_map); @@ -28639,7 +29033,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c return false; } -@@ -2840,7 +2878,7 @@ +@@ -2840,7 +2878,7 @@ static bool __cancel_work_timer(struct work_struct *work, bool is_dwork) /* tell other tasks trying to grab @work to back off */ mark_work_canceling(work); @@ -28648,7 +29042,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c flush_work(work); clear_work_data(work); -@@ -2895,10 +2933,10 @@ +@@ -2895,10 +2933,10 @@ EXPORT_SYMBOL_GPL(cancel_work_sync); */ bool flush_delayed_work(struct delayed_work *dwork) { @@ -28661,7 +29055,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c return flush_work(&dwork->work); } EXPORT_SYMBOL(flush_delayed_work); -@@ -2933,7 +2971,7 @@ +@@ -2933,7 +2971,7 @@ bool cancel_delayed_work(struct delayed_work *dwork) set_work_pool_and_clear_pending(&dwork->work, get_work_pool_id(&dwork->work)); @@ -28670,7 +29064,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c return ret; } EXPORT_SYMBOL(cancel_delayed_work); -@@ -3161,7 +3199,7 @@ +@@ -3161,7 +3199,7 @@ static void rcu_free_pool(struct rcu_head *rcu) * put_unbound_pool - put a worker_pool * @pool: worker_pool to put * @@ -28679,7 +29073,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c * safe manner. get_unbound_pool() calls this function on its failure path * and this function should be able to release pools which went through, * successfully or not, init_worker_pool(). -@@ -3215,8 +3253,8 @@ +@@ -3215,8 +3253,8 @@ static void put_unbound_pool(struct worker_pool *pool) del_timer_sync(&pool->idle_timer); del_timer_sync(&pool->mayday_timer); @@ -28690,7 +29084,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c } /** -@@ -3323,14 +3361,14 @@ +@@ -3323,14 +3361,14 @@ static void pwq_unbound_release_workfn(struct work_struct *work) put_unbound_pool(pool); mutex_unlock(&wq_pool_mutex); @@ -28707,7 +29101,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c } /** -@@ -3983,7 +4021,7 @@ +@@ -3983,7 +4021,7 @@ void destroy_workqueue(struct workqueue_struct *wq) * The base ref is never dropped on per-cpu pwqs. Directly * schedule RCU free. */ @@ -28716,7 +29110,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c } else { /* * We're the sole accessor of @wq at this point. Directly -@@ -4076,7 +4114,8 @@ +@@ -4076,7 +4114,8 @@ bool workqueue_congested(int cpu, struct workqueue_struct *wq) struct pool_workqueue *pwq; bool ret; @@ -28726,7 +29120,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c if (cpu == WORK_CPU_UNBOUND) cpu = smp_processor_id(); -@@ -4087,7 +4126,8 @@ +@@ -4087,7 +4126,8 @@ bool workqueue_congested(int cpu, struct workqueue_struct *wq) pwq = unbound_pwq_by_node(wq, cpu_to_node(cpu)); ret = !list_empty(&pwq->delayed_works); @@ -28736,7 +29130,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c return ret; } -@@ -4113,15 +4153,15 @@ +@@ -4113,15 +4153,15 @@ unsigned int work_busy(struct work_struct *work) if (work_pending(work)) ret |= WORK_BUSY_PENDING; @@ -28756,7 +29150,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c return ret; } -@@ -4310,7 +4350,7 @@ +@@ -4310,7 +4350,7 @@ void show_workqueue_state(void) unsigned long flags; int pi; @@ -28765,7 +29159,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c pr_info("Showing busy workqueues and worker pools:\n"); -@@ -4361,7 +4401,7 @@ +@@ -4361,7 +4401,7 @@ void show_workqueue_state(void) spin_unlock_irqrestore(&pool->lock, flags); } @@ -28774,7 +29168,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c } /* -@@ -4722,16 +4762,16 @@ +@@ -4722,16 +4762,16 @@ bool freeze_workqueues_busy(void) * nr_active is monotonically decreasing. It's safe * to peek without lock. */ @@ -28794,7 +29188,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c } out_unlock: mutex_unlock(&wq_pool_mutex); -@@ -4921,7 +4961,8 @@ +@@ -4921,7 +4961,8 @@ static ssize_t wq_pool_ids_show(struct device *dev, const char *delim = ""; int node, written = 0; @@ -28804,7 +29198,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c for_each_node(node) { written += scnprintf(buf + written, PAGE_SIZE - written, "%s%d:%d", delim, node, -@@ -4929,7 +4970,8 @@ +@@ -4929,7 +4970,8 @@ static ssize_t wq_pool_ids_show(struct device *dev, delim = " "; } written += scnprintf(buf + written, PAGE_SIZE - written, "\n"); @@ -28814,10 +29208,11 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue.c linux-4.4.62/kernel/workqueue.c return written; } -diff -Nur linux-4.4.62.orig/kernel/workqueue_internal.h linux-4.4.62/kernel/workqueue_internal.h ---- linux-4.4.62.orig/kernel/workqueue_internal.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/kernel/workqueue_internal.h 2017-04-18 17:38:08.230650641 +0200 -@@ -43,6 +43,7 @@ +diff --git a/kernel/workqueue_internal.h b/kernel/workqueue_internal.h +index 45215870ac6c..f000c4d6917e 100644 +--- a/kernel/workqueue_internal.h ++++ b/kernel/workqueue_internal.h +@@ -43,6 +43,7 @@ struct worker { unsigned long last_active; /* L: last active timestamp */ unsigned int flags; /* X: flags */ int id; /* I: worker id */ @@ -28825,7 +29220,7 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue_internal.h linux-4.4.62/kernel/work /* * Opaque string set with work_set_desc(). Printed out with task -@@ -68,7 +69,7 @@ +@@ -68,7 +69,7 @@ static inline struct worker *current_wq_worker(void) * Scheduler hooks for concurrency managed workqueue. Only to be used from * sched/core.c and workqueue.c. */ @@ -28835,10 +29230,23 @@ diff -Nur linux-4.4.62.orig/kernel/workqueue_internal.h linux-4.4.62/kernel/work +void wq_worker_sleeping(struct task_struct *task); #endif /* _KERNEL_WORKQUEUE_INTERNAL_H */ -diff -Nur linux-4.4.62.orig/lib/debugobjects.c linux-4.4.62/lib/debugobjects.c ---- linux-4.4.62.orig/lib/debugobjects.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/lib/debugobjects.c 2017-04-18 17:38:08.230650641 +0200 -@@ -309,7 +309,10 @@ +diff --git a/lib/Kconfig b/lib/Kconfig +index 1a48744253d7..f75de578cca8 100644 +--- a/lib/Kconfig ++++ b/lib/Kconfig +@@ -397,6 +397,7 @@ config CHECK_SIGNATURE + + config CPUMASK_OFFSTACK + bool "Force CPU masks off stack" if DEBUG_PER_CPU_MAPS ++ depends on !PREEMPT_RT_FULL + help + Use dynamic allocation for cpumask_var_t, instead of putting + them on the stack. This is a bit more expensive, but avoids +diff --git a/lib/debugobjects.c b/lib/debugobjects.c +index 547f7f923dbc..8fcdbc2fc6d0 100644 +--- a/lib/debugobjects.c ++++ b/lib/debugobjects.c +@@ -309,7 +309,10 @@ __debug_object_init(void *addr, struct debug_obj_descr *descr, int onstack) struct debug_obj *obj; unsigned long flags; @@ -28850,9 +29258,10 @@ diff -Nur linux-4.4.62.orig/lib/debugobjects.c linux-4.4.62/lib/debugobjects.c db = get_bucket((unsigned long) addr); -diff -Nur linux-4.4.62.orig/lib/idr.c linux-4.4.62/lib/idr.c ---- linux-4.4.62.orig/lib/idr.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/lib/idr.c 2017-04-18 17:38:08.230650641 +0200 +diff --git a/lib/idr.c b/lib/idr.c +index 6098336df267..9decbe914595 100644 +--- a/lib/idr.c ++++ b/lib/idr.c @@ -30,6 +30,7 @@ #include #include @@ -28861,7 +29270,7 @@ diff -Nur linux-4.4.62.orig/lib/idr.c linux-4.4.62/lib/idr.c #define MAX_IDR_SHIFT (sizeof(int) * 8 - 1) #define MAX_IDR_BIT (1U << MAX_IDR_SHIFT) -@@ -45,6 +46,37 @@ +@@ -45,6 +46,37 @@ static DEFINE_PER_CPU(struct idr_layer *, idr_preload_head); static DEFINE_PER_CPU(int, idr_preload_cnt); static DEFINE_SPINLOCK(simple_ida_lock); @@ -28899,7 +29308,7 @@ diff -Nur linux-4.4.62.orig/lib/idr.c linux-4.4.62/lib/idr.c /* the maximum ID which can be allocated given idr->layers */ static int idr_max(int layers) { -@@ -115,14 +147,14 @@ +@@ -115,14 +147,14 @@ static struct idr_layer *idr_layer_alloc(gfp_t gfp_mask, struct idr *layer_idr) * context. See idr_preload() for details. */ if (!in_interrupt()) { @@ -28916,7 +29325,7 @@ diff -Nur linux-4.4.62.orig/lib/idr.c linux-4.4.62/lib/idr.c if (new) return new; } -@@ -366,7 +398,6 @@ +@@ -366,7 +398,6 @@ static void idr_fill_slot(struct idr *idr, void *ptr, int id, idr_mark_full(pa, id); } @@ -28924,7 +29333,7 @@ diff -Nur linux-4.4.62.orig/lib/idr.c linux-4.4.62/lib/idr.c /** * idr_preload - preload for idr_alloc() * @gfp_mask: allocation mask to use for preloading -@@ -401,7 +432,7 @@ +@@ -401,7 +432,7 @@ void idr_preload(gfp_t gfp_mask) WARN_ON_ONCE(in_interrupt()); might_sleep_if(gfpflags_allow_blocking(gfp_mask)); @@ -28933,7 +29342,7 @@ diff -Nur linux-4.4.62.orig/lib/idr.c linux-4.4.62/lib/idr.c /* * idr_alloc() is likely to succeed w/o full idr_layer buffer and -@@ -413,9 +444,9 @@ +@@ -413,9 +444,9 @@ void idr_preload(gfp_t gfp_mask) while (__this_cpu_read(idr_preload_cnt) < MAX_IDR_FREE) { struct idr_layer *new; @@ -28945,21 +29354,11 @@ diff -Nur linux-4.4.62.orig/lib/idr.c linux-4.4.62/lib/idr.c if (!new) break; -diff -Nur linux-4.4.62.orig/lib/Kconfig linux-4.4.62/lib/Kconfig ---- linux-4.4.62.orig/lib/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/lib/Kconfig 2017-04-18 17:38:08.230650641 +0200 -@@ -397,6 +397,7 @@ - - config CPUMASK_OFFSTACK - bool "Force CPU masks off stack" if DEBUG_PER_CPU_MAPS -+ depends on !PREEMPT_RT_FULL - help - Use dynamic allocation for cpumask_var_t, instead of putting - them on the stack. This is a bit more expensive, but avoids -diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-selftest.c ---- linux-4.4.62.orig/lib/locking-selftest.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/lib/locking-selftest.c 2017-04-18 17:38:08.230650641 +0200 -@@ -590,6 +590,8 @@ +diff --git a/lib/locking-selftest.c b/lib/locking-selftest.c +index 872a15a2a637..b93a6103fa4d 100644 +--- a/lib/locking-selftest.c ++++ b/lib/locking-selftest.c +@@ -590,6 +590,8 @@ GENERATE_TESTCASE(init_held_rsem) #include "locking-selftest-spin-hardirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_hard_spin) @@ -28968,7 +29367,7 @@ diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-self #include "locking-selftest-rlock-hardirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_hard_rlock) -@@ -605,9 +607,12 @@ +@@ -605,9 +607,12 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_soft_rlock) #include "locking-selftest-wlock-softirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_soft_wlock) @@ -28981,7 +29380,7 @@ diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-self /* * Enabling hardirqs with a softirq-safe lock held: */ -@@ -640,6 +645,8 @@ +@@ -640,6 +645,8 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2A_rlock) #undef E1 #undef E2 @@ -28990,7 +29389,7 @@ diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-self /* * Enabling irqs with an irq-safe lock held: */ -@@ -663,6 +670,8 @@ +@@ -663,6 +670,8 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2A_rlock) #include "locking-selftest-spin-hardirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_hard_spin) @@ -28999,7 +29398,7 @@ diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-self #include "locking-selftest-rlock-hardirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_hard_rlock) -@@ -678,6 +687,8 @@ +@@ -678,6 +687,8 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_rlock) #include "locking-selftest-wlock-softirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_wlock) @@ -29008,7 +29407,7 @@ diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-self #undef E1 #undef E2 -@@ -709,6 +720,8 @@ +@@ -709,6 +720,8 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_wlock) #include "locking-selftest-spin-hardirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_hard_spin) @@ -29017,7 +29416,7 @@ diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-self #include "locking-selftest-rlock-hardirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_hard_rlock) -@@ -724,6 +737,8 @@ +@@ -724,6 +737,8 @@ GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_rlock) #include "locking-selftest-wlock-softirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_wlock) @@ -29026,7 +29425,7 @@ diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-self #undef E1 #undef E2 #undef E3 -@@ -757,6 +772,8 @@ +@@ -757,6 +772,8 @@ GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_wlock) #include "locking-selftest-spin-hardirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_hard_spin) @@ -29035,7 +29434,7 @@ diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-self #include "locking-selftest-rlock-hardirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_hard_rlock) -@@ -772,10 +789,14 @@ +@@ -772,10 +789,14 @@ GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_soft_rlock) #include "locking-selftest-wlock-softirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_soft_wlock) @@ -29050,7 +29449,7 @@ diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-self /* * read-lock / write-lock irq inversion. * -@@ -838,6 +859,10 @@ +@@ -838,6 +859,10 @@ GENERATE_PERMUTATIONS_3_EVENTS(irq_inversion_soft_wlock) #undef E2 #undef E3 @@ -29061,7 +29460,7 @@ diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-self /* * read-lock / write-lock recursion that is actually safe. */ -@@ -876,6 +901,8 @@ +@@ -876,6 +901,8 @@ GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion_soft) #undef E2 #undef E3 @@ -29070,7 +29469,7 @@ diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-self /* * read-lock / write-lock recursion that is unsafe. */ -@@ -1858,6 +1885,7 @@ +@@ -1858,6 +1885,7 @@ void locking_selftest(void) printk(" --------------------------------------------------------------------------\n"); @@ -29078,7 +29477,7 @@ diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-self /* * irq-context testcases: */ -@@ -1870,6 +1898,28 @@ +@@ -1870,6 +1898,28 @@ void locking_selftest(void) DO_TESTCASE_6x2("irq read-recursion", irq_read_recursion); // DO_TESTCASE_6x2B("irq read-recursion #2", irq_read_recursion2); @@ -29107,9 +29506,10 @@ diff -Nur linux-4.4.62.orig/lib/locking-selftest.c linux-4.4.62/lib/locking-self ww_tests(); -diff -Nur linux-4.4.62.orig/lib/percpu_ida.c linux-4.4.62/lib/percpu_ida.c ---- linux-4.4.62.orig/lib/percpu_ida.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/lib/percpu_ida.c 2017-04-18 17:38:08.230650641 +0200 +diff --git a/lib/percpu_ida.c b/lib/percpu_ida.c +index 6d40944960de..822a2c027e72 100644 +--- a/lib/percpu_ida.c ++++ b/lib/percpu_ida.c @@ -26,6 +26,9 @@ #include #include @@ -29120,7 +29520,7 @@ diff -Nur linux-4.4.62.orig/lib/percpu_ida.c linux-4.4.62/lib/percpu_ida.c struct percpu_ida_cpu { /* -@@ -148,13 +151,13 @@ +@@ -148,13 +151,13 @@ int percpu_ida_alloc(struct percpu_ida *pool, int state) unsigned long flags; int tag; @@ -29136,7 +29536,7 @@ diff -Nur linux-4.4.62.orig/lib/percpu_ida.c linux-4.4.62/lib/percpu_ida.c return tag; } -@@ -173,6 +176,7 @@ +@@ -173,6 +176,7 @@ int percpu_ida_alloc(struct percpu_ida *pool, int state) if (!tags->nr_free) alloc_global_tags(pool, tags); @@ -29144,7 +29544,7 @@ diff -Nur linux-4.4.62.orig/lib/percpu_ida.c linux-4.4.62/lib/percpu_ida.c if (!tags->nr_free) steal_tags(pool, tags); -@@ -184,7 +188,7 @@ +@@ -184,7 +188,7 @@ int percpu_ida_alloc(struct percpu_ida *pool, int state) } spin_unlock(&pool->lock); @@ -29153,7 +29553,7 @@ diff -Nur linux-4.4.62.orig/lib/percpu_ida.c linux-4.4.62/lib/percpu_ida.c if (tag >= 0 || state == TASK_RUNNING) break; -@@ -196,7 +200,7 @@ +@@ -196,7 +200,7 @@ int percpu_ida_alloc(struct percpu_ida *pool, int state) schedule(); @@ -29162,7 +29562,7 @@ diff -Nur linux-4.4.62.orig/lib/percpu_ida.c linux-4.4.62/lib/percpu_ida.c tags = this_cpu_ptr(pool->tag_cpu); } if (state != TASK_RUNNING) -@@ -221,7 +225,7 @@ +@@ -221,7 +225,7 @@ void percpu_ida_free(struct percpu_ida *pool, unsigned tag) BUG_ON(tag >= pool->nr_tags); @@ -29171,7 +29571,7 @@ diff -Nur linux-4.4.62.orig/lib/percpu_ida.c linux-4.4.62/lib/percpu_ida.c tags = this_cpu_ptr(pool->tag_cpu); spin_lock(&tags->lock); -@@ -253,7 +257,7 @@ +@@ -253,7 +257,7 @@ void percpu_ida_free(struct percpu_ida *pool, unsigned tag) spin_unlock(&pool->lock); } @@ -29180,7 +29580,7 @@ diff -Nur linux-4.4.62.orig/lib/percpu_ida.c linux-4.4.62/lib/percpu_ida.c } EXPORT_SYMBOL_GPL(percpu_ida_free); -@@ -345,7 +349,7 @@ +@@ -345,7 +349,7 @@ int percpu_ida_for_each_free(struct percpu_ida *pool, percpu_ida_cb fn, struct percpu_ida_cpu *remote; unsigned cpu, i, err = 0; @@ -29189,7 +29589,7 @@ diff -Nur linux-4.4.62.orig/lib/percpu_ida.c linux-4.4.62/lib/percpu_ida.c for_each_possible_cpu(cpu) { remote = per_cpu_ptr(pool->tag_cpu, cpu); spin_lock(&remote->lock); -@@ -367,7 +371,7 @@ +@@ -367,7 +371,7 @@ int percpu_ida_for_each_free(struct percpu_ida *pool, percpu_ida_cb fn, } spin_unlock(&pool->lock); out: @@ -29198,9 +29598,10 @@ diff -Nur linux-4.4.62.orig/lib/percpu_ida.c linux-4.4.62/lib/percpu_ida.c return err; } EXPORT_SYMBOL_GPL(percpu_ida_for_each_free); -diff -Nur linux-4.4.62.orig/lib/radix-tree.c linux-4.4.62/lib/radix-tree.c ---- linux-4.4.62.orig/lib/radix-tree.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/lib/radix-tree.c 2017-04-18 17:38:08.230650641 +0200 +diff --git a/lib/radix-tree.c b/lib/radix-tree.c +index 6b79e9026e24..44bf36a396a9 100644 +--- a/lib/radix-tree.c ++++ b/lib/radix-tree.c @@ -34,7 +34,7 @@ #include #include @@ -29210,7 +29611,7 @@ diff -Nur linux-4.4.62.orig/lib/radix-tree.c linux-4.4.62/lib/radix-tree.c /* * The height_to_maxindex array needs to be one deeper than the maximum -@@ -69,6 +69,7 @@ +@@ -69,6 +69,7 @@ struct radix_tree_preload { struct radix_tree_node *nodes; }; static DEFINE_PER_CPU(struct radix_tree_preload, radix_tree_preloads) = { 0, }; @@ -29218,7 +29619,7 @@ diff -Nur linux-4.4.62.orig/lib/radix-tree.c linux-4.4.62/lib/radix-tree.c static inline void *ptr_to_indirect(void *ptr) { -@@ -196,13 +197,14 @@ +@@ -196,13 +197,14 @@ radix_tree_node_alloc(struct radix_tree_root *root) * succeed in getting a node here (and never reach * kmem_cache_alloc) */ @@ -29234,7 +29635,7 @@ diff -Nur linux-4.4.62.orig/lib/radix-tree.c linux-4.4.62/lib/radix-tree.c /* * Update the allocation stack trace as this is more useful * for debugging. -@@ -257,14 +259,14 @@ +@@ -257,14 +259,14 @@ static int __radix_tree_preload(gfp_t gfp_mask) struct radix_tree_node *node; int ret = -ENOMEM; @@ -29252,7 +29653,7 @@ diff -Nur linux-4.4.62.orig/lib/radix-tree.c linux-4.4.62/lib/radix-tree.c rtp = this_cpu_ptr(&radix_tree_preloads); if (rtp->nr < RADIX_TREE_PRELOAD_SIZE) { node->private_data = rtp->nodes; -@@ -306,11 +308,17 @@ +@@ -306,11 +308,17 @@ int radix_tree_maybe_preload(gfp_t gfp_mask) if (gfpflags_allow_blocking(gfp_mask)) return __radix_tree_preload(gfp_mask); /* Preloading doesn't help anything with this gfp mask, skip it */ @@ -29271,9 +29672,10 @@ diff -Nur linux-4.4.62.orig/lib/radix-tree.c linux-4.4.62/lib/radix-tree.c /* * Return the maximum key which can be store into a * radix tree with height HEIGHT. -diff -Nur linux-4.4.62.orig/lib/rbtree.c linux-4.4.62/lib/rbtree.c ---- linux-4.4.62.orig/lib/rbtree.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/lib/rbtree.c 2017-04-18 17:38:08.230650641 +0200 +diff --git a/lib/rbtree.c b/lib/rbtree.c +index 1356454e36de..d15d6c4327f1 100644 +--- a/lib/rbtree.c ++++ b/lib/rbtree.c @@ -23,6 +23,7 @@ #include @@ -29282,7 +29684,7 @@ diff -Nur linux-4.4.62.orig/lib/rbtree.c linux-4.4.62/lib/rbtree.c /* * red-black trees properties: http://en.wikipedia.org/wiki/Rbtree -@@ -590,3 +591,13 @@ +@@ -590,3 +591,13 @@ struct rb_node *rb_first_postorder(const struct rb_root *root) return rb_left_deepest_node(root->rb_node); } EXPORT_SYMBOL(rb_first_postorder); @@ -29296,10 +29698,11 @@ diff -Nur linux-4.4.62.orig/lib/rbtree.c linux-4.4.62/lib/rbtree.c + rcu_assign_pointer(*rb_link, node); +} +EXPORT_SYMBOL(rb_link_node_rcu); -diff -Nur linux-4.4.62.orig/lib/scatterlist.c linux-4.4.62/lib/scatterlist.c ---- linux-4.4.62.orig/lib/scatterlist.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/lib/scatterlist.c 2017-04-18 17:38:08.230650641 +0200 -@@ -620,7 +620,7 @@ +diff --git a/lib/scatterlist.c b/lib/scatterlist.c +index bafa9933fa76..ebe3b7edd086 100644 +--- a/lib/scatterlist.c ++++ b/lib/scatterlist.c +@@ -620,7 +620,7 @@ void sg_miter_stop(struct sg_mapping_iter *miter) flush_kernel_dcache_page(miter->page); if (miter->__flags & SG_MITER_ATOMIC) { @@ -29308,7 +29711,7 @@ diff -Nur linux-4.4.62.orig/lib/scatterlist.c linux-4.4.62/lib/scatterlist.c kunmap_atomic(miter->addr); } else kunmap(miter->page); -@@ -664,7 +664,7 @@ +@@ -664,7 +664,7 @@ size_t sg_copy_buffer(struct scatterlist *sgl, unsigned int nents, void *buf, if (!sg_miter_skip(&miter, skip)) return false; @@ -29317,7 +29720,7 @@ diff -Nur linux-4.4.62.orig/lib/scatterlist.c linux-4.4.62/lib/scatterlist.c while (sg_miter_next(&miter) && offset < buflen) { unsigned int len; -@@ -681,7 +681,7 @@ +@@ -681,7 +681,7 @@ size_t sg_copy_buffer(struct scatterlist *sgl, unsigned int nents, void *buf, sg_miter_stop(&miter); @@ -29326,10 +29729,11 @@ diff -Nur linux-4.4.62.orig/lib/scatterlist.c linux-4.4.62/lib/scatterlist.c return offset; } EXPORT_SYMBOL(sg_copy_buffer); -diff -Nur linux-4.4.62.orig/lib/smp_processor_id.c linux-4.4.62/lib/smp_processor_id.c ---- linux-4.4.62.orig/lib/smp_processor_id.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/lib/smp_processor_id.c 2017-04-18 17:38:08.230650641 +0200 -@@ -39,8 +39,9 @@ +diff --git a/lib/smp_processor_id.c b/lib/smp_processor_id.c +index 1afec32de6f2..11fa431046a8 100644 +--- a/lib/smp_processor_id.c ++++ b/lib/smp_processor_id.c +@@ -39,8 +39,9 @@ notrace static unsigned int check_preemption_disabled(const char *what1, if (!printk_ratelimit()) goto out_enable; @@ -29341,23 +29745,24 @@ diff -Nur linux-4.4.62.orig/lib/smp_processor_id.c linux-4.4.62/lib/smp_processo print_symbol("caller is %s\n", (long)__builtin_return_address(0)); dump_stack(); -diff -Nur linux-4.4.62.orig/Makefile linux-4.4.62/Makefile ---- linux-4.4.62.orig/Makefile 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/Makefile 2017-04-18 17:38:07.902637922 +0200 -@@ -785,6 +785,9 @@ - # Prohibit date/time macros, which would make the build non-deterministic - KBUILD_CFLAGS += $(call cc-option,-Werror=date-time) - -+# enforce correct pointer usage -+KBUILD_CFLAGS += $(call cc-option,-Werror=incompatible-pointer-types) -+ - # use the deterministic mode of AR if available - KBUILD_ARFLAGS := $(call ar-option,D) +diff --git a/mm/Kconfig b/mm/Kconfig +index 97a4e06b15c0..9614351e68b8 100644 +--- a/mm/Kconfig ++++ b/mm/Kconfig +@@ -392,7 +392,7 @@ config NOMMU_INITIAL_TRIM_EXCESS -diff -Nur linux-4.4.62.orig/mm/backing-dev.c linux-4.4.62/mm/backing-dev.c ---- linux-4.4.62.orig/mm/backing-dev.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/backing-dev.c 2017-04-18 17:38:08.230650641 +0200 -@@ -457,9 +457,9 @@ + config TRANSPARENT_HUGEPAGE + bool "Transparent Hugepage Support" +- depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE ++ depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE && !PREEMPT_RT_FULL + select COMPACTION + help + Transparent Hugepages allows the kernel to use huge pages and +diff --git a/mm/backing-dev.c b/mm/backing-dev.c +index a988d4ef39da..f2c2ee1d5191 100644 +--- a/mm/backing-dev.c ++++ b/mm/backing-dev.c +@@ -457,9 +457,9 @@ void wb_congested_put(struct bdi_writeback_congested *congested) { unsigned long flags; @@ -29369,10 +29774,11 @@ diff -Nur linux-4.4.62.orig/mm/backing-dev.c linux-4.4.62/mm/backing-dev.c return; } -diff -Nur linux-4.4.62.orig/mm/compaction.c linux-4.4.62/mm/compaction.c ---- linux-4.4.62.orig/mm/compaction.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/compaction.c 2017-04-18 17:38:08.230650641 +0200 -@@ -1430,10 +1430,12 @@ +diff --git a/mm/compaction.c b/mm/compaction.c +index dba02dec7195..51963f58a29b 100644 +--- a/mm/compaction.c ++++ b/mm/compaction.c +@@ -1430,10 +1430,12 @@ check_drain: cc->migrate_pfn & ~((1UL << cc->order) - 1); if (cc->last_migrated_pfn < current_block_start) { @@ -29387,10 +29793,11 @@ diff -Nur linux-4.4.62.orig/mm/compaction.c linux-4.4.62/mm/compaction.c /* No more flushing until we migrate again */ cc->last_migrated_pfn = 0; } -diff -Nur linux-4.4.62.orig/mm/filemap.c linux-4.4.62/mm/filemap.c ---- linux-4.4.62.orig/mm/filemap.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/filemap.c 2017-04-18 17:38:08.230650641 +0200 -@@ -144,9 +144,12 @@ +diff --git a/mm/filemap.c b/mm/filemap.c +index 69f75c77c098..b203169ca0b4 100644 +--- a/mm/filemap.c ++++ b/mm/filemap.c +@@ -144,9 +144,12 @@ static int page_cache_tree_insert(struct address_space *mapping, * node->private_list is protected by * mapping->tree_lock. */ @@ -29405,7 +29812,7 @@ diff -Nur linux-4.4.62.orig/mm/filemap.c linux-4.4.62/mm/filemap.c } return 0; } -@@ -218,7 +221,9 @@ +@@ -218,7 +221,9 @@ static void page_cache_tree_delete(struct address_space *mapping, if (!workingset_node_pages(node) && list_empty(&node->private_list)) { node->private_data = mapping; @@ -29416,9 +29823,10 @@ diff -Nur linux-4.4.62.orig/mm/filemap.c linux-4.4.62/mm/filemap.c } } -diff -Nur linux-4.4.62.orig/mm/highmem.c linux-4.4.62/mm/highmem.c ---- linux-4.4.62.orig/mm/highmem.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/highmem.c 2017-04-18 17:38:08.230650641 +0200 +diff --git a/mm/highmem.c b/mm/highmem.c +index 123bcd3ed4f2..16e8cf26d38a 100644 +--- a/mm/highmem.c ++++ b/mm/highmem.c @@ -29,10 +29,11 @@ #include #include @@ -29432,7 +29840,7 @@ diff -Nur linux-4.4.62.orig/mm/highmem.c linux-4.4.62/mm/highmem.c /* * Virtual_count is not a pure "count". -@@ -107,8 +108,9 @@ +@@ -107,8 +108,9 @@ static inline wait_queue_head_t *get_pkmap_wait_queue_head(unsigned int color) unsigned long totalhigh_pages __read_mostly; EXPORT_SYMBOL(totalhigh_pages); @@ -29443,21 +29851,10 @@ diff -Nur linux-4.4.62.orig/mm/highmem.c linux-4.4.62/mm/highmem.c unsigned int nr_free_highpages (void) { -diff -Nur linux-4.4.62.orig/mm/Kconfig linux-4.4.62/mm/Kconfig ---- linux-4.4.62.orig/mm/Kconfig 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/Kconfig 2017-04-18 17:38:08.230650641 +0200 -@@ -392,7 +392,7 @@ - - config TRANSPARENT_HUGEPAGE - bool "Transparent Hugepage Support" -- depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE -+ depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE && !PREEMPT_RT_FULL - select COMPACTION - help - Transparent Hugepages allows the kernel to use huge pages and -diff -Nur linux-4.4.62.orig/mm/memcontrol.c linux-4.4.62/mm/memcontrol.c ---- linux-4.4.62.orig/mm/memcontrol.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/memcontrol.c 2017-04-18 17:38:08.234650796 +0200 +diff --git a/mm/memcontrol.c b/mm/memcontrol.c +index e25b93a4267d..1c619267d9da 100644 +--- a/mm/memcontrol.c ++++ b/mm/memcontrol.c @@ -67,6 +67,8 @@ #include #include @@ -29467,7 +29864,7 @@ diff -Nur linux-4.4.62.orig/mm/memcontrol.c linux-4.4.62/mm/memcontrol.c #include "slab.h" #include -@@ -87,6 +89,7 @@ +@@ -87,6 +89,7 @@ int do_swap_account __read_mostly; #define do_swap_account 0 #endif @@ -29475,7 +29872,7 @@ diff -Nur linux-4.4.62.orig/mm/memcontrol.c linux-4.4.62/mm/memcontrol.c static const char * const mem_cgroup_stat_names[] = { "cache", "rss", -@@ -1922,14 +1925,17 @@ +@@ -1922,14 +1925,17 @@ static void drain_local_stock(struct work_struct *dummy) */ static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages) { @@ -29495,7 +29892,7 @@ diff -Nur linux-4.4.62.orig/mm/memcontrol.c linux-4.4.62/mm/memcontrol.c } /* -@@ -1945,7 +1951,7 @@ +@@ -1945,7 +1951,7 @@ static void drain_all_stock(struct mem_cgroup *root_memcg) return; /* Notify other cpus that system-wide "drain" is running */ get_online_cpus(); @@ -29504,7 +29901,7 @@ diff -Nur linux-4.4.62.orig/mm/memcontrol.c linux-4.4.62/mm/memcontrol.c for_each_online_cpu(cpu) { struct memcg_stock_pcp *stock = &per_cpu(memcg_stock, cpu); struct mem_cgroup *memcg; -@@ -1962,7 +1968,7 @@ +@@ -1962,7 +1968,7 @@ static void drain_all_stock(struct mem_cgroup *root_memcg) schedule_work_on(cpu, &stock->work); } } @@ -29513,7 +29910,7 @@ diff -Nur linux-4.4.62.orig/mm/memcontrol.c linux-4.4.62/mm/memcontrol.c put_online_cpus(); mutex_unlock(&percpu_charge_mutex); } -@@ -4691,12 +4697,12 @@ +@@ -4691,12 +4697,12 @@ static int mem_cgroup_move_account(struct page *page, ret = 0; @@ -29528,7 +29925,7 @@ diff -Nur linux-4.4.62.orig/mm/memcontrol.c linux-4.4.62/mm/memcontrol.c out_unlock: unlock_page(page); out: -@@ -5486,10 +5492,10 @@ +@@ -5486,10 +5492,10 @@ void mem_cgroup_commit_charge(struct page *page, struct mem_cgroup *memcg, VM_BUG_ON_PAGE(!PageTransHuge(page), page); } @@ -29541,7 +29938,7 @@ diff -Nur linux-4.4.62.orig/mm/memcontrol.c linux-4.4.62/mm/memcontrol.c if (do_swap_account && PageSwapCache(page)) { swp_entry_t entry = { .val = page_private(page) }; -@@ -5545,14 +5551,14 @@ +@@ -5545,14 +5551,14 @@ static void uncharge_batch(struct mem_cgroup *memcg, unsigned long pgpgout, memcg_oom_recover(memcg); } @@ -29558,7 +29955,7 @@ diff -Nur linux-4.4.62.orig/mm/memcontrol.c linux-4.4.62/mm/memcontrol.c if (!mem_cgroup_is_root(memcg)) css_put_many(&memcg->css, nr_pages); -@@ -5762,6 +5768,7 @@ +@@ -5762,6 +5768,7 @@ void mem_cgroup_swapout(struct page *page, swp_entry_t entry) { struct mem_cgroup *memcg, *swap_memcg; unsigned short oldid; @@ -29566,7 +29963,7 @@ diff -Nur linux-4.4.62.orig/mm/memcontrol.c linux-4.4.62/mm/memcontrol.c VM_BUG_ON_PAGE(PageLRU(page), page); VM_BUG_ON_PAGE(page_count(page), page); -@@ -5802,12 +5809,16 @@ +@@ -5802,12 +5809,16 @@ void mem_cgroup_swapout(struct page *page, swp_entry_t entry) * important here to have the interrupts disabled because it is the * only synchronisation we have for udpating the per-CPU variables. */ @@ -29583,10 +29980,11 @@ diff -Nur linux-4.4.62.orig/mm/memcontrol.c linux-4.4.62/mm/memcontrol.c } /** -diff -Nur linux-4.4.62.orig/mm/mmu_context.c linux-4.4.62/mm/mmu_context.c ---- linux-4.4.62.orig/mm/mmu_context.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/mmu_context.c 2017-04-18 17:38:08.234650796 +0200 -@@ -23,6 +23,7 @@ +diff --git a/mm/mmu_context.c b/mm/mmu_context.c +index f802c2d216a7..b1b6f238e42d 100644 +--- a/mm/mmu_context.c ++++ b/mm/mmu_context.c +@@ -23,6 +23,7 @@ void use_mm(struct mm_struct *mm) struct task_struct *tsk = current; task_lock(tsk); @@ -29594,7 +29992,7 @@ diff -Nur linux-4.4.62.orig/mm/mmu_context.c linux-4.4.62/mm/mmu_context.c active_mm = tsk->active_mm; if (active_mm != mm) { atomic_inc(&mm->mm_count); -@@ -30,6 +31,7 @@ +@@ -30,6 +31,7 @@ void use_mm(struct mm_struct *mm) } tsk->mm = mm; switch_mm(active_mm, mm, tsk); @@ -29602,9 +30000,10 @@ diff -Nur linux-4.4.62.orig/mm/mmu_context.c linux-4.4.62/mm/mmu_context.c task_unlock(tsk); #ifdef finish_arch_post_lock_switch finish_arch_post_lock_switch(); -diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c ---- linux-4.4.62.orig/mm/page_alloc.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/page_alloc.c 2017-04-18 17:38:08.234650796 +0200 +diff --git a/mm/page_alloc.c b/mm/page_alloc.c +index 6f9005dcca2e..e8ec72251705 100644 +--- a/mm/page_alloc.c ++++ b/mm/page_alloc.c @@ -60,6 +60,7 @@ #include #include @@ -29613,7 +30012,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c #include #include -@@ -264,6 +265,18 @@ +@@ -264,6 +265,18 @@ EXPORT_SYMBOL(nr_node_ids); EXPORT_SYMBOL(nr_online_nodes); #endif @@ -29632,7 +30031,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c int page_group_by_mobility_disabled __read_mostly; #ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT -@@ -786,7 +799,7 @@ +@@ -786,7 +799,7 @@ static inline int free_pages_check(struct page *page) } /* @@ -29641,7 +30040,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c * Assumes all pages on list are in same zone, and of same order. * count is the number of pages to free. * -@@ -797,18 +810,53 @@ +@@ -797,18 +810,53 @@ static inline int free_pages_check(struct page *page) * pinned" detection logic. */ static void free_pcppages_bulk(struct zone *zone, int count, @@ -29699,7 +30098,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c while (to_free) { struct page *page; struct list_head *list; -@@ -824,7 +872,7 @@ +@@ -824,7 +872,7 @@ static void free_pcppages_bulk(struct zone *zone, int count, batch_free++; if (++migratetype == MIGRATE_PCPTYPES) migratetype = 0; @@ -29708,7 +30107,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c } while (list_empty(list)); /* This is the only non-empty list. Free them all. */ -@@ -832,24 +880,12 @@ +@@ -832,24 +880,12 @@ static void free_pcppages_bulk(struct zone *zone, int count, batch_free = to_free; do { @@ -29735,7 +30134,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c } static void free_one_page(struct zone *zone, -@@ -858,7 +894,9 @@ +@@ -858,7 +894,9 @@ static void free_one_page(struct zone *zone, int migratetype) { unsigned long nr_scanned; @@ -29746,7 +30145,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c nr_scanned = zone_page_state(zone, NR_PAGES_SCANNED); if (nr_scanned) __mod_zone_page_state(zone, NR_PAGES_SCANNED, -nr_scanned); -@@ -868,7 +906,7 @@ +@@ -868,7 +906,7 @@ static void free_one_page(struct zone *zone, migratetype = get_pfnblock_migratetype(page, pfn); } __free_one_page(page, pfn, zone, order, migratetype); @@ -29755,7 +30154,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c } static int free_tail_pages_check(struct page *head_page, struct page *page) -@@ -1019,10 +1057,10 @@ +@@ -1019,10 +1057,10 @@ static void __free_pages_ok(struct page *page, unsigned int order) return; migratetype = get_pfnblock_migratetype(page, pfn); @@ -29768,7 +30167,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c } static void __init __free_pages_boot_core(struct page *page, -@@ -1879,16 +1917,18 @@ +@@ -1879,16 +1917,18 @@ static int rmqueue_bulk(struct zone *zone, unsigned int order, void drain_zone_pages(struct zone *zone, struct per_cpu_pages *pcp) { unsigned long flags; @@ -29790,7 +30189,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c } #endif -@@ -1904,16 +1944,21 @@ +@@ -1904,16 +1944,21 @@ static void drain_pages_zone(unsigned int cpu, struct zone *zone) unsigned long flags; struct per_cpu_pageset *pset; struct per_cpu_pages *pcp; @@ -29816,7 +30215,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c } /* -@@ -1999,8 +2044,17 @@ +@@ -1999,8 +2044,17 @@ void drain_all_pages(struct zone *zone) else cpumask_clear_cpu(cpu, &cpus_with_pcps); } @@ -29834,7 +30233,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c } #ifdef CONFIG_HIBERNATION -@@ -2056,7 +2110,7 @@ +@@ -2056,7 +2110,7 @@ void free_hot_cold_page(struct page *page, bool cold) migratetype = get_pfnblock_migratetype(page, pfn); set_pcppage_migratetype(page, migratetype); @@ -29843,7 +30242,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c __count_vm_event(PGFREE); /* -@@ -2082,12 +2136,17 @@ +@@ -2082,12 +2136,17 @@ void free_hot_cold_page(struct page *page, bool cold) pcp->count++; if (pcp->count >= pcp->high) { unsigned long batch = READ_ONCE(pcp->batch); @@ -29863,7 +30262,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c } /* -@@ -2222,7 +2281,7 @@ +@@ -2222,7 +2281,7 @@ struct page *buffered_rmqueue(struct zone *preferred_zone, struct per_cpu_pages *pcp; struct list_head *list; @@ -29872,7 +30271,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c pcp = &this_cpu_ptr(zone->pageset)->pcp; list = &pcp->lists[migratetype]; if (list_empty(list)) { -@@ -2254,7 +2313,7 @@ +@@ -2254,7 +2313,7 @@ struct page *buffered_rmqueue(struct zone *preferred_zone, */ WARN_ON_ONCE(order > 1); } @@ -29881,7 +30280,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c page = NULL; if (alloc_flags & ALLOC_HARDER) { -@@ -2264,11 +2323,13 @@ +@@ -2264,11 +2323,13 @@ struct page *buffered_rmqueue(struct zone *preferred_zone, } if (!page) page = __rmqueue(zone, order, migratetype, gfp_flags); @@ -29897,7 +30296,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c } __mod_zone_page_state(zone, NR_ALLOC_BATCH, -(1 << order)); -@@ -2278,13 +2339,13 @@ +@@ -2278,13 +2339,13 @@ struct page *buffered_rmqueue(struct zone *preferred_zone, __count_zone_vm_events(PGALLOC, zone, 1 << order); zone_statistics(preferred_zone, zone, gfp_flags); @@ -29913,7 +30312,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c return NULL; } -@@ -5953,6 +6014,7 @@ +@@ -5953,6 +6014,7 @@ static int page_alloc_cpu_notify(struct notifier_block *self, void __init page_alloc_init(void) { hotcpu_notifier(page_alloc_cpu_notify, 0); @@ -29921,7 +30320,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c } /* -@@ -6847,7 +6909,7 @@ +@@ -6847,7 +6909,7 @@ void zone_pcp_reset(struct zone *zone) struct per_cpu_pageset *pset; /* avoid races with drain_pages() */ @@ -29930,7 +30329,7 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c if (zone->pageset != &boot_pageset) { for_each_online_cpu(cpu) { pset = per_cpu_ptr(zone->pageset, cpu); -@@ -6856,7 +6918,7 @@ +@@ -6856,7 +6918,7 @@ void zone_pcp_reset(struct zone *zone) free_percpu(zone->pageset); zone->pageset = &boot_pageset; } @@ -29939,10 +30338,11 @@ diff -Nur linux-4.4.62.orig/mm/page_alloc.c linux-4.4.62/mm/page_alloc.c } #ifdef CONFIG_MEMORY_HOTREMOVE -diff -Nur linux-4.4.62.orig/mm/percpu.c linux-4.4.62/mm/percpu.c ---- linux-4.4.62.orig/mm/percpu.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/percpu.c 2017-04-18 17:38:08.234650796 +0200 -@@ -1285,18 +1285,7 @@ +diff --git a/mm/percpu.c b/mm/percpu.c +index ef6353f0adbd..33ccbac7cdb8 100644 +--- a/mm/percpu.c ++++ b/mm/percpu.c +@@ -1285,18 +1285,7 @@ void free_percpu(void __percpu *ptr) } EXPORT_SYMBOL_GPL(free_percpu); @@ -29962,7 +30362,7 @@ diff -Nur linux-4.4.62.orig/mm/percpu.c linux-4.4.62/mm/percpu.c { #ifdef CONFIG_SMP const size_t static_size = __per_cpu_end - __per_cpu_start; -@@ -1305,16 +1294,36 @@ +@@ -1305,16 +1294,36 @@ bool is_kernel_percpu_address(unsigned long addr) for_each_possible_cpu(cpu) { void *start = per_cpu_ptr(base, cpu); @@ -30001,10 +30401,11 @@ diff -Nur linux-4.4.62.orig/mm/percpu.c linux-4.4.62/mm/percpu.c * per_cpu_ptr_to_phys - convert translated percpu address to physical address * @addr: the address to be converted to physical address * -diff -Nur linux-4.4.62.orig/mm/slab.h linux-4.4.62/mm/slab.h ---- linux-4.4.62.orig/mm/slab.h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/slab.h 2017-04-18 17:38:08.234650796 +0200 -@@ -324,7 +324,11 @@ +diff --git a/mm/slab.h b/mm/slab.h +index 7b6087197997..afdc57941179 100644 +--- a/mm/slab.h ++++ b/mm/slab.h +@@ -324,7 +324,11 @@ static inline struct kmem_cache *cache_from_obj(struct kmem_cache *s, void *x) * The slab lists for all objects. */ struct kmem_cache_node { @@ -30016,10 +30417,11 @@ diff -Nur linux-4.4.62.orig/mm/slab.h linux-4.4.62/mm/slab.h #ifdef CONFIG_SLAB struct list_head slabs_partial; /* partial list first, better asm code */ -diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c ---- linux-4.4.62.orig/mm/slub.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/slub.c 2017-04-18 17:38:08.234650796 +0200 -@@ -1075,7 +1075,7 @@ +diff --git a/mm/slub.c b/mm/slub.c +index 65d5f92d51d2..feb4a445a546 100644 +--- a/mm/slub.c ++++ b/mm/slub.c +@@ -1075,7 +1075,7 @@ static noinline struct kmem_cache_node *free_debug_processing( void *object = head; int cnt = 0; @@ -30028,7 +30430,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c slab_lock(page); if (!check_slab(s, page)) -@@ -1136,7 +1136,7 @@ +@@ -1136,7 +1136,7 @@ out: fail: slab_unlock(page); @@ -30037,7 +30439,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c slab_fix(s, "Object at 0x%p not freed", object); return NULL; } -@@ -1263,6 +1263,12 @@ +@@ -1263,6 +1263,12 @@ static inline void dec_slabs_node(struct kmem_cache *s, int node, #endif /* CONFIG_SLUB_DEBUG */ @@ -30050,7 +30452,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c /* * Hooks for other subsystems that check memory allocations. In a typical * production configuration these hooks all should produce no code at all. -@@ -1399,10 +1405,17 @@ +@@ -1399,10 +1405,17 @@ static struct page *allocate_slab(struct kmem_cache *s, gfp_t flags, int node) gfp_t alloc_gfp; void *start, *p; int idx, order; @@ -30068,7 +30470,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c local_irq_enable(); flags |= s->allocflags; -@@ -1473,7 +1486,7 @@ +@@ -1473,7 +1486,7 @@ static struct page *allocate_slab(struct kmem_cache *s, gfp_t flags, int node) page->frozen = 1; out: @@ -30077,7 +30479,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c local_irq_disable(); if (!page) return NULL; -@@ -1529,6 +1542,16 @@ +@@ -1529,6 +1542,16 @@ static void __free_slab(struct kmem_cache *s, struct page *page) __free_kmem_pages(page, order); } @@ -30094,7 +30496,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c #define need_reserve_slab_rcu \ (sizeof(((struct page *)NULL)->lru) < sizeof(struct rcu_head)) -@@ -1560,6 +1583,12 @@ +@@ -1560,6 +1583,12 @@ static void free_slab(struct kmem_cache *s, struct page *page) } call_rcu(head, rcu_free_slab); @@ -30107,7 +30509,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c } else __free_slab(s, page); } -@@ -1673,7 +1702,7 @@ +@@ -1673,7 +1702,7 @@ static void *get_partial_node(struct kmem_cache *s, struct kmem_cache_node *n, if (!n || !n->nr_partial) return NULL; @@ -30116,7 +30518,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c list_for_each_entry_safe(page, page2, &n->partial, lru) { void *t; -@@ -1698,7 +1727,7 @@ +@@ -1698,7 +1727,7 @@ static void *get_partial_node(struct kmem_cache *s, struct kmem_cache_node *n, break; } @@ -30125,7 +30527,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c return object; } -@@ -1944,7 +1973,7 @@ +@@ -1944,7 +1973,7 @@ redo: * that acquire_slab() will see a slab page that * is frozen */ @@ -30134,7 +30536,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c } } else { m = M_FULL; -@@ -1955,7 +1984,7 @@ +@@ -1955,7 +1984,7 @@ redo: * slabs from diagnostic functions will not see * any frozen slabs. */ @@ -30143,7 +30545,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c } } -@@ -1990,7 +2019,7 @@ +@@ -1990,7 +2019,7 @@ redo: goto redo; if (lock) @@ -30152,7 +30554,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c if (m == M_FREE) { stat(s, DEACTIVATE_EMPTY); -@@ -2022,10 +2051,10 @@ +@@ -2022,10 +2051,10 @@ static void unfreeze_partials(struct kmem_cache *s, n2 = get_node(s, page_to_nid(page)); if (n != n2) { if (n) @@ -30165,7 +30567,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c } do { -@@ -2054,7 +2083,7 @@ +@@ -2054,7 +2083,7 @@ static void unfreeze_partials(struct kmem_cache *s, } if (n) @@ -30174,7 +30576,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c while (discard_page) { page = discard_page; -@@ -2093,14 +2122,21 @@ +@@ -2093,14 +2122,21 @@ static void put_cpu_partial(struct kmem_cache *s, struct page *page, int drain) pobjects = oldpage->pobjects; pages = oldpage->pages; if (drain && pobjects > s->cpu_partial) { @@ -30196,7 +30598,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c oldpage = NULL; pobjects = 0; pages = 0; -@@ -2172,7 +2208,22 @@ +@@ -2172,7 +2208,22 @@ static bool has_cpu_slab(int cpu, void *info) static void flush_all(struct kmem_cache *s) { @@ -30219,7 +30621,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c } /* -@@ -2208,10 +2259,10 @@ +@@ -2208,10 +2259,10 @@ static unsigned long count_partial(struct kmem_cache_node *n, unsigned long x = 0; struct page *page; @@ -30232,7 +30634,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c return x; } #endif /* CONFIG_SLUB_DEBUG || CONFIG_SYSFS */ -@@ -2349,8 +2400,10 @@ +@@ -2349,8 +2400,10 @@ static inline void *get_freelist(struct kmem_cache *s, struct page *page) * already disabled (which is the case for bulk allocation). */ static void *___slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, @@ -30244,7 +30646,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c void *freelist; struct page *page; -@@ -2410,6 +2463,13 @@ +@@ -2410,6 +2463,13 @@ load_freelist: VM_BUG_ON(!c->page->frozen); c->freelist = get_freepointer(s, freelist); c->tid = next_tid(c->tid); @@ -30258,7 +30660,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c return freelist; new_slab: -@@ -2441,7 +2501,7 @@ +@@ -2441,7 +2501,7 @@ new_slab: deactivate_slab(s, page, get_freepointer(s, freelist)); c->page = NULL; c->freelist = NULL; @@ -30267,7 +30669,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c } /* -@@ -2453,6 +2513,7 @@ +@@ -2453,6 +2513,7 @@ static void *__slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, { void *p; unsigned long flags; @@ -30275,7 +30677,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c local_irq_save(flags); #ifdef CONFIG_PREEMPT -@@ -2464,8 +2525,9 @@ +@@ -2464,8 +2525,9 @@ static void *__slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, c = this_cpu_ptr(s->cpu_slab); #endif @@ -30286,7 +30688,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c return p; } -@@ -2652,7 +2714,7 @@ +@@ -2652,7 +2714,7 @@ static void __slab_free(struct kmem_cache *s, struct page *page, do { if (unlikely(n)) { @@ -30295,7 +30697,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c n = NULL; } prior = page->freelist; -@@ -2684,7 +2746,7 @@ +@@ -2684,7 +2746,7 @@ static void __slab_free(struct kmem_cache *s, struct page *page, * Otherwise the list_lock will synchronize with * other processors updating the list of slabs. */ @@ -30304,7 +30706,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c } } -@@ -2726,7 +2788,7 @@ +@@ -2726,7 +2788,7 @@ static void __slab_free(struct kmem_cache *s, struct page *page, add_partial(n, page, DEACTIVATE_TO_TAIL); stat(s, FREE_ADD_PARTIAL); } @@ -30313,7 +30715,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c return; slab_empty: -@@ -2741,7 +2803,7 @@ +@@ -2741,7 +2803,7 @@ slab_empty: remove_full(s, n, page); } @@ -30322,7 +30724,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c stat(s, FREE_SLAB); discard_slab(s, page); } -@@ -2913,6 +2975,7 @@ +@@ -2913,6 +2975,7 @@ int kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, void **p) { struct kmem_cache_cpu *c; @@ -30330,7 +30732,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c int i; /* memcg and kmem_cache debug support */ -@@ -2936,7 +2999,7 @@ +@@ -2936,7 +2999,7 @@ int kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, * of re-populating per CPU c->freelist */ p[i] = ___slab_alloc(s, flags, NUMA_NO_NODE, @@ -30339,7 +30741,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c if (unlikely(!p[i])) goto error; -@@ -2948,6 +3011,7 @@ +@@ -2948,6 +3011,7 @@ int kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, } c->tid = next_tid(c->tid); local_irq_enable(); @@ -30347,7 +30749,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c /* Clear memory outside IRQ disabled fastpath loop */ if (unlikely(flags & __GFP_ZERO)) { -@@ -3095,7 +3159,7 @@ +@@ -3095,7 +3159,7 @@ static void init_kmem_cache_node(struct kmem_cache_node *n) { n->nr_partial = 0; @@ -30356,7 +30758,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c INIT_LIST_HEAD(&n->partial); #ifdef CONFIG_SLUB_DEBUG atomic_long_set(&n->nr_slabs, 0); -@@ -3677,7 +3741,7 @@ +@@ -3677,7 +3741,7 @@ int __kmem_cache_shrink(struct kmem_cache *s, bool deactivate) for (i = 0; i < SHRINK_PROMOTE_MAX; i++) INIT_LIST_HEAD(promote + i); @@ -30365,7 +30767,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c /* * Build lists of slabs to discard or promote. -@@ -3708,7 +3772,7 @@ +@@ -3708,7 +3772,7 @@ int __kmem_cache_shrink(struct kmem_cache *s, bool deactivate) for (i = SHRINK_PROMOTE_MAX - 1; i >= 0; i--) list_splice(promote + i, &n->partial); @@ -30374,7 +30776,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c /* Release empty slabs */ list_for_each_entry_safe(page, t, &discard, lru) -@@ -3884,6 +3948,12 @@ +@@ -3884,6 +3948,12 @@ void __init kmem_cache_init(void) { static __initdata struct kmem_cache boot_kmem_cache, boot_kmem_cache_node; @@ -30387,7 +30789,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c if (debug_guardpage_minorder()) slub_max_order = 0; -@@ -4127,7 +4197,7 @@ +@@ -4127,7 +4197,7 @@ static int validate_slab_node(struct kmem_cache *s, struct page *page; unsigned long flags; @@ -30396,7 +30798,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c list_for_each_entry(page, &n->partial, lru) { validate_slab_slab(s, page, map); -@@ -4149,7 +4219,7 @@ +@@ -4149,7 +4219,7 @@ static int validate_slab_node(struct kmem_cache *s, s->name, count, atomic_long_read(&n->nr_slabs)); out: @@ -30405,7 +30807,7 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c return count; } -@@ -4337,12 +4407,12 @@ +@@ -4337,12 +4407,12 @@ static int list_locations(struct kmem_cache *s, char *buf, if (!atomic_long_read(&n->nr_slabs)) continue; @@ -30420,9 +30822,10 @@ diff -Nur linux-4.4.62.orig/mm/slub.c linux-4.4.62/mm/slub.c } for (i = 0; i < t.count; i++) { -diff -Nur linux-4.4.62.orig/mm/swap.c linux-4.4.62/mm/swap.c ---- linux-4.4.62.orig/mm/swap.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/swap.c 2017-04-18 17:38:08.234650796 +0200 +diff --git a/mm/swap.c b/mm/swap.c +index 39395fb549c0..ad16649221d7 100644 +--- a/mm/swap.c ++++ b/mm/swap.c @@ -31,6 +31,7 @@ #include #include @@ -30431,7 +30834,7 @@ diff -Nur linux-4.4.62.orig/mm/swap.c linux-4.4.62/mm/swap.c #include #include -@@ -46,6 +47,9 @@ +@@ -46,6 +47,9 @@ static DEFINE_PER_CPU(struct pagevec, lru_add_pvec); static DEFINE_PER_CPU(struct pagevec, lru_rotate_pvecs); static DEFINE_PER_CPU(struct pagevec, lru_deactivate_file_pvecs); @@ -30441,7 +30844,7 @@ diff -Nur linux-4.4.62.orig/mm/swap.c linux-4.4.62/mm/swap.c /* * This path almost never happens for VM activity - pages are normally * freed via pagevecs. But it gets used by networking. -@@ -481,11 +485,11 @@ +@@ -481,11 +485,11 @@ void rotate_reclaimable_page(struct page *page) unsigned long flags; page_cache_get(page); @@ -30455,7 +30858,7 @@ diff -Nur linux-4.4.62.orig/mm/swap.c linux-4.4.62/mm/swap.c } } -@@ -536,12 +540,13 @@ +@@ -536,12 +540,13 @@ static bool need_activate_page_drain(int cpu) void activate_page(struct page *page) { if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) { @@ -30471,7 +30874,7 @@ diff -Nur linux-4.4.62.orig/mm/swap.c linux-4.4.62/mm/swap.c } } -@@ -567,7 +572,7 @@ +@@ -567,7 +572,7 @@ void activate_page(struct page *page) static void __lru_cache_activate_page(struct page *page) { @@ -30480,7 +30883,7 @@ diff -Nur linux-4.4.62.orig/mm/swap.c linux-4.4.62/mm/swap.c int i; /* -@@ -589,7 +594,7 @@ +@@ -589,7 +594,7 @@ static void __lru_cache_activate_page(struct page *page) } } @@ -30489,7 +30892,7 @@ diff -Nur linux-4.4.62.orig/mm/swap.c linux-4.4.62/mm/swap.c } /* -@@ -630,13 +635,13 @@ +@@ -630,13 +635,13 @@ EXPORT_SYMBOL(mark_page_accessed); static void __lru_cache_add(struct page *page) { @@ -30505,7 +30908,7 @@ diff -Nur linux-4.4.62.orig/mm/swap.c linux-4.4.62/mm/swap.c } /** -@@ -816,9 +821,15 @@ +@@ -816,9 +821,15 @@ void lru_add_drain_cpu(int cpu) unsigned long flags; /* No harm done if a racing interrupt already did this */ @@ -30523,7 +30926,7 @@ diff -Nur linux-4.4.62.orig/mm/swap.c linux-4.4.62/mm/swap.c } pvec = &per_cpu(lru_deactivate_file_pvecs, cpu); -@@ -846,26 +857,47 @@ +@@ -846,26 +857,47 @@ void deactivate_file_page(struct page *page) return; if (likely(get_page_unless_zero(page))) { @@ -30575,7 +30978,7 @@ diff -Nur linux-4.4.62.orig/mm/swap.c linux-4.4.62/mm/swap.c void lru_add_drain_all(void) { -@@ -878,20 +910,17 @@ +@@ -878,20 +910,17 @@ void lru_add_drain_all(void) cpumask_clear(&has_work); for_each_online_cpu(cpu) { @@ -30600,10 +31003,11 @@ diff -Nur linux-4.4.62.orig/mm/swap.c linux-4.4.62/mm/swap.c put_online_cpus(); mutex_unlock(&lock); -diff -Nur linux-4.4.62.orig/mm/truncate.c linux-4.4.62/mm/truncate.c ---- linux-4.4.62.orig/mm/truncate.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/truncate.c 2017-04-18 17:38:08.234650796 +0200 -@@ -56,8 +56,11 @@ +diff --git a/mm/truncate.c b/mm/truncate.c +index 76e35ad97102..5f196420020c 100644 +--- a/mm/truncate.c ++++ b/mm/truncate.c +@@ -56,8 +56,11 @@ static void clear_exceptional_entry(struct address_space *mapping, * protected by mapping->tree_lock. */ if (!workingset_node_shadows(node) && @@ -30617,10 +31021,11 @@ diff -Nur linux-4.4.62.orig/mm/truncate.c linux-4.4.62/mm/truncate.c __radix_tree_delete_node(&mapping->page_tree, node); unlock: spin_unlock_irq(&mapping->tree_lock); -diff -Nur linux-4.4.62.orig/mm/vmalloc.c linux-4.4.62/mm/vmalloc.c ---- linux-4.4.62.orig/mm/vmalloc.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/vmalloc.c 2017-04-18 17:38:08.234650796 +0200 -@@ -821,7 +821,7 @@ +diff --git a/mm/vmalloc.c b/mm/vmalloc.c +index 8e3c9c5a3042..68740314ad54 100644 +--- a/mm/vmalloc.c ++++ b/mm/vmalloc.c +@@ -821,7 +821,7 @@ static void *new_vmap_block(unsigned int order, gfp_t gfp_mask) struct vmap_block *vb; struct vmap_area *va; unsigned long vb_idx; @@ -30629,7 +31034,7 @@ diff -Nur linux-4.4.62.orig/mm/vmalloc.c linux-4.4.62/mm/vmalloc.c void *vaddr; node = numa_node_id(); -@@ -864,11 +864,12 @@ +@@ -864,11 +864,12 @@ static void *new_vmap_block(unsigned int order, gfp_t gfp_mask) BUG_ON(err); radix_tree_preload_end(); @@ -30644,7 +31049,7 @@ diff -Nur linux-4.4.62.orig/mm/vmalloc.c linux-4.4.62/mm/vmalloc.c return vaddr; } -@@ -937,6 +938,7 @@ +@@ -937,6 +938,7 @@ static void *vb_alloc(unsigned long size, gfp_t gfp_mask) struct vmap_block *vb; void *vaddr = NULL; unsigned int order; @@ -30652,7 +31057,7 @@ diff -Nur linux-4.4.62.orig/mm/vmalloc.c linux-4.4.62/mm/vmalloc.c BUG_ON(offset_in_page(size)); BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC); -@@ -951,7 +953,8 @@ +@@ -951,7 +953,8 @@ static void *vb_alloc(unsigned long size, gfp_t gfp_mask) order = get_order(size); rcu_read_lock(); @@ -30662,7 +31067,7 @@ diff -Nur linux-4.4.62.orig/mm/vmalloc.c linux-4.4.62/mm/vmalloc.c list_for_each_entry_rcu(vb, &vbq->free, free_list) { unsigned long pages_off; -@@ -974,7 +977,7 @@ +@@ -974,7 +977,7 @@ static void *vb_alloc(unsigned long size, gfp_t gfp_mask) break; } @@ -30671,10 +31076,11 @@ diff -Nur linux-4.4.62.orig/mm/vmalloc.c linux-4.4.62/mm/vmalloc.c rcu_read_unlock(); /* Allocate new block if nothing was found */ -diff -Nur linux-4.4.62.orig/mm/vmstat.c linux-4.4.62/mm/vmstat.c ---- linux-4.4.62.orig/mm/vmstat.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/vmstat.c 2017-04-18 17:38:08.238650951 +0200 -@@ -226,6 +226,7 @@ +diff --git a/mm/vmstat.c b/mm/vmstat.c +index c54fd2924f25..64416fd7c209 100644 +--- a/mm/vmstat.c ++++ b/mm/vmstat.c +@@ -226,6 +226,7 @@ void __mod_zone_page_state(struct zone *zone, enum zone_stat_item item, long x; long t; @@ -30682,7 +31088,7 @@ diff -Nur linux-4.4.62.orig/mm/vmstat.c linux-4.4.62/mm/vmstat.c x = delta + __this_cpu_read(*p); t = __this_cpu_read(pcp->stat_threshold); -@@ -235,6 +236,7 @@ +@@ -235,6 +236,7 @@ void __mod_zone_page_state(struct zone *zone, enum zone_stat_item item, x = 0; } __this_cpu_write(*p, x); @@ -30690,7 +31096,7 @@ diff -Nur linux-4.4.62.orig/mm/vmstat.c linux-4.4.62/mm/vmstat.c } EXPORT_SYMBOL(__mod_zone_page_state); -@@ -267,6 +269,7 @@ +@@ -267,6 +269,7 @@ void __inc_zone_state(struct zone *zone, enum zone_stat_item item) s8 __percpu *p = pcp->vm_stat_diff + item; s8 v, t; @@ -30698,7 +31104,7 @@ diff -Nur linux-4.4.62.orig/mm/vmstat.c linux-4.4.62/mm/vmstat.c v = __this_cpu_inc_return(*p); t = __this_cpu_read(pcp->stat_threshold); if (unlikely(v > t)) { -@@ -275,6 +278,7 @@ +@@ -275,6 +278,7 @@ void __inc_zone_state(struct zone *zone, enum zone_stat_item item) zone_page_state_add(v + overstep, zone, item); __this_cpu_write(*p, -overstep); } @@ -30706,7 +31112,7 @@ diff -Nur linux-4.4.62.orig/mm/vmstat.c linux-4.4.62/mm/vmstat.c } void __inc_zone_page_state(struct page *page, enum zone_stat_item item) -@@ -289,6 +293,7 @@ +@@ -289,6 +293,7 @@ void __dec_zone_state(struct zone *zone, enum zone_stat_item item) s8 __percpu *p = pcp->vm_stat_diff + item; s8 v, t; @@ -30714,7 +31120,7 @@ diff -Nur linux-4.4.62.orig/mm/vmstat.c linux-4.4.62/mm/vmstat.c v = __this_cpu_dec_return(*p); t = __this_cpu_read(pcp->stat_threshold); if (unlikely(v < - t)) { -@@ -297,6 +302,7 @@ +@@ -297,6 +302,7 @@ void __dec_zone_state(struct zone *zone, enum zone_stat_item item) zone_page_state_add(v - overstep, zone, item); __this_cpu_write(*p, overstep); } @@ -30722,10 +31128,11 @@ diff -Nur linux-4.4.62.orig/mm/vmstat.c linux-4.4.62/mm/vmstat.c } void __dec_zone_page_state(struct page *page, enum zone_stat_item item) -diff -Nur linux-4.4.62.orig/mm/workingset.c linux-4.4.62/mm/workingset.c ---- linux-4.4.62.orig/mm/workingset.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/workingset.c 2017-04-18 17:38:08.238650951 +0200 -@@ -264,7 +264,8 @@ +diff --git a/mm/workingset.c b/mm/workingset.c +index df66f426fdcf..6db7b243fa0d 100644 +--- a/mm/workingset.c ++++ b/mm/workingset.c +@@ -264,7 +264,8 @@ void workingset_activation(struct page *page) * point where they would still be useful. */ @@ -30735,7 +31142,7 @@ diff -Nur linux-4.4.62.orig/mm/workingset.c linux-4.4.62/mm/workingset.c static unsigned long count_shadow_nodes(struct shrinker *shrinker, struct shrink_control *sc) -@@ -274,9 +275,9 @@ +@@ -274,9 +275,9 @@ static unsigned long count_shadow_nodes(struct shrinker *shrinker, unsigned long pages; /* list_lru lock nests inside IRQ-safe mapping->tree_lock */ @@ -30748,7 +31155,7 @@ diff -Nur linux-4.4.62.orig/mm/workingset.c linux-4.4.62/mm/workingset.c pages = node_present_pages(sc->nid); /* -@@ -361,9 +362,9 @@ +@@ -361,9 +362,9 @@ static enum lru_status shadow_lru_isolate(struct list_head *item, spin_unlock(&mapping->tree_lock); ret = LRU_REMOVED_RETRY; out: @@ -30760,7 +31167,7 @@ diff -Nur linux-4.4.62.orig/mm/workingset.c linux-4.4.62/mm/workingset.c spin_lock(lru_lock); return ret; } -@@ -374,10 +375,10 @@ +@@ -374,10 +375,10 @@ static unsigned long scan_shadow_nodes(struct shrinker *shrinker, unsigned long ret; /* list_lru lock nests inside IRQ-safe mapping->tree_lock */ @@ -30774,7 +31181,7 @@ diff -Nur linux-4.4.62.orig/mm/workingset.c linux-4.4.62/mm/workingset.c return ret; } -@@ -398,7 +399,7 @@ +@@ -398,7 +399,7 @@ static int __init workingset_init(void) { int ret; @@ -30783,7 +31190,7 @@ diff -Nur linux-4.4.62.orig/mm/workingset.c linux-4.4.62/mm/workingset.c if (ret) goto err; ret = register_shrinker(&workingset_shadow_shrinker); -@@ -406,7 +407,7 @@ +@@ -406,7 +407,7 @@ static int __init workingset_init(void) goto err_list_lru; return 0; err_list_lru: @@ -30792,9 +31199,10 @@ diff -Nur linux-4.4.62.orig/mm/workingset.c linux-4.4.62/mm/workingset.c err: return ret; } -diff -Nur linux-4.4.62.orig/mm/zsmalloc.c linux-4.4.62/mm/zsmalloc.c ---- linux-4.4.62.orig/mm/zsmalloc.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/mm/zsmalloc.c 2017-04-18 17:38:08.238650951 +0200 +diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c +index c1ea19478119..529552c3716d 100644 +--- a/mm/zsmalloc.c ++++ b/mm/zsmalloc.c @@ -64,6 +64,7 @@ #include #include @@ -30803,7 +31211,7 @@ diff -Nur linux-4.4.62.orig/mm/zsmalloc.c linux-4.4.62/mm/zsmalloc.c /* * This must be power of 2 and greater than of equal to sizeof(link_free). -@@ -403,6 +404,7 @@ +@@ -403,6 +404,7 @@ static unsigned int get_maxobj_per_zspage(int size, int pages_per_zspage) /* per-cpu VM mapping areas for zspage accesses that cross page boundaries */ static DEFINE_PER_CPU(struct mapping_area, zs_map_area); @@ -30811,7 +31219,7 @@ diff -Nur linux-4.4.62.orig/mm/zsmalloc.c linux-4.4.62/mm/zsmalloc.c static int is_first_page(struct page *page) { -@@ -1289,7 +1291,7 @@ +@@ -1289,7 +1291,7 @@ void *zs_map_object(struct zs_pool *pool, unsigned long handle, class = pool->size_class[class_idx]; off = obj_idx_to_offset(page, obj_idx, class->size); @@ -30820,7 +31228,7 @@ diff -Nur linux-4.4.62.orig/mm/zsmalloc.c linux-4.4.62/mm/zsmalloc.c area->vm_mm = mm; if (off + class->size <= PAGE_SIZE) { /* this object is contained entirely within a page */ -@@ -1342,7 +1344,7 @@ +@@ -1342,7 +1344,7 @@ void zs_unmap_object(struct zs_pool *pool, unsigned long handle) __zs_unmap_object(area, pages, off, class->size); } @@ -30829,10 +31237,11 @@ diff -Nur linux-4.4.62.orig/mm/zsmalloc.c linux-4.4.62/mm/zsmalloc.c unpin_tag(handle); } EXPORT_SYMBOL_GPL(zs_unmap_object); -diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c ---- linux-4.4.62.orig/net/core/dev.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/net/core/dev.c 2017-04-18 17:38:08.238650951 +0200 -@@ -186,6 +186,7 @@ +diff --git a/net/core/dev.c b/net/core/dev.c +index 48399d8ce614..2ccc02633e97 100644 +--- a/net/core/dev.c ++++ b/net/core/dev.c +@@ -186,6 +186,7 @@ static unsigned int napi_gen_id; static DEFINE_HASHTABLE(napi_hash, 8); static seqcount_t devnet_rename_seq; @@ -30840,7 +31249,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c static inline void dev_base_seq_inc(struct net *net) { -@@ -207,14 +208,14 @@ +@@ -207,14 +208,14 @@ static inline struct hlist_head *dev_index_hash(struct net *net, int ifindex) static inline void rps_lock(struct softnet_data *sd) { #ifdef CONFIG_RPS @@ -30857,7 +31266,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c #endif } -@@ -884,7 +885,8 @@ +@@ -884,7 +885,8 @@ retry: strcpy(name, dev->name); rcu_read_unlock(); if (read_seqcount_retry(&devnet_rename_seq, seq)) { @@ -30867,7 +31276,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c goto retry; } -@@ -1153,20 +1155,17 @@ +@@ -1153,20 +1155,17 @@ int dev_change_name(struct net_device *dev, const char *newname) if (dev->flags & IFF_UP) return -EBUSY; @@ -30894,7 +31303,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c if (oldname[0] && !strchr(oldname, '%')) netdev_info(dev, "renamed from %s\n", oldname); -@@ -1179,11 +1178,12 @@ +@@ -1179,11 +1178,12 @@ rollback: if (ret) { memcpy(dev->name, oldname, IFNAMSIZ); dev->name_assign_type = old_assign_type; @@ -30910,7 +31319,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c netdev_adjacent_rename_links(dev, oldname); -@@ -1204,7 +1204,8 @@ +@@ -1204,7 +1204,8 @@ rollback: /* err >= 0 after dev_alloc_name() or stores the first errno */ if (err >= 0) { err = ret; @@ -30920,7 +31329,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c memcpy(dev->name, oldname, IFNAMSIZ); memcpy(oldname, newname, IFNAMSIZ); dev->name_assign_type = old_assign_type; -@@ -1217,6 +1218,11 @@ +@@ -1217,6 +1218,11 @@ rollback: } return err; @@ -30932,7 +31341,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c } /** -@@ -2268,6 +2274,7 @@ +@@ -2268,6 +2274,7 @@ static inline void __netif_reschedule(struct Qdisc *q) sd->output_queue_tailp = &q->next_sched; raise_softirq_irqoff(NET_TX_SOFTIRQ); local_irq_restore(flags); @@ -30940,7 +31349,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c } void __netif_schedule(struct Qdisc *q) -@@ -2349,6 +2356,7 @@ +@@ -2349,6 +2356,7 @@ void __dev_kfree_skb_irq(struct sk_buff *skb, enum skb_free_reason reason) __this_cpu_write(softnet_data.completion_queue, skb); raise_softirq_irqoff(NET_TX_SOFTIRQ); local_irq_restore(flags); @@ -30948,7 +31357,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c } EXPORT_SYMBOL(__dev_kfree_skb_irq); -@@ -2906,7 +2914,11 @@ +@@ -2906,7 +2914,11 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, * This permits __QDISC___STATE_RUNNING owner to get the lock more * often and dequeue packets faster. */ @@ -30960,7 +31369,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c if (unlikely(contended)) spin_lock(&q->busylock); -@@ -2966,9 +2978,44 @@ +@@ -2966,9 +2978,44 @@ static void skb_update_prio(struct sk_buff *skb) #define skb_update_prio(skb) #endif @@ -31005,7 +31414,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c #define RECURSION_LIMIT 10 /** -@@ -3161,7 +3208,7 @@ +@@ -3161,7 +3208,7 @@ static int __dev_queue_xmit(struct sk_buff *skb, void *accel_priv) if (txq->xmit_lock_owner != cpu) { @@ -31014,7 +31423,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c goto recursion_alert; skb = validate_xmit_skb(skb, dev); -@@ -3171,9 +3218,9 @@ +@@ -3171,9 +3218,9 @@ static int __dev_queue_xmit(struct sk_buff *skb, void *accel_priv) HARD_TX_LOCK(dev, txq, cpu); if (!netif_xmit_stopped(txq)) { @@ -31026,7 +31435,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c if (dev_xmit_complete(rc)) { HARD_TX_UNLOCK(dev, txq); goto out; -@@ -3547,6 +3594,7 @@ +@@ -3547,6 +3594,7 @@ drop: rps_unlock(sd); local_irq_restore(flags); @@ -31034,7 +31443,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c atomic_long_inc(&skb->dev->rx_dropped); kfree_skb(skb); -@@ -3565,7 +3613,7 @@ +@@ -3565,7 +3613,7 @@ static int netif_rx_internal(struct sk_buff *skb) struct rps_dev_flow voidflow, *rflow = &voidflow; int cpu; @@ -31043,7 +31452,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c rcu_read_lock(); cpu = get_rps_cpu(skb->dev, skb, &rflow); -@@ -3575,13 +3623,13 @@ +@@ -3575,13 +3623,13 @@ static int netif_rx_internal(struct sk_buff *skb) ret = enqueue_to_backlog(skb, cpu, &rflow->last_qtail); rcu_read_unlock(); @@ -31060,7 +31469,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c } return ret; } -@@ -3615,16 +3663,44 @@ +@@ -3615,16 +3663,44 @@ int netif_rx_ni(struct sk_buff *skb) trace_netif_rx_ni_entry(skb); @@ -31109,7 +31518,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c static void net_tx_action(struct softirq_action *h) { struct softnet_data *sd = this_cpu_ptr(&softnet_data); -@@ -3666,7 +3742,7 @@ +@@ -3666,7 +3742,7 @@ static void net_tx_action(struct softirq_action *h) head = head->next_sched; root_lock = qdisc_lock(q); @@ -31118,7 +31527,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c smp_mb__before_atomic(); clear_bit(__QDISC_STATE_SCHED, &q->state); -@@ -4088,7 +4164,7 @@ +@@ -4088,7 +4164,7 @@ static void flush_backlog(void *arg) skb_queue_walk_safe(&sd->input_pkt_queue, skb, tmp) { if (skb->dev == dev) { __skb_unlink(skb, &sd->input_pkt_queue); @@ -31127,7 +31536,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c input_queue_head_incr(sd); } } -@@ -4097,10 +4173,13 @@ +@@ -4097,10 +4173,13 @@ static void flush_backlog(void *arg) skb_queue_walk_safe(&sd->process_queue, skb, tmp) { if (skb->dev == dev) { __skb_unlink(skb, &sd->process_queue); @@ -31142,7 +31551,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c } static int napi_gro_complete(struct sk_buff *skb) -@@ -4557,6 +4636,7 @@ +@@ -4557,6 +4636,7 @@ static void net_rps_action_and_irq_enable(struct softnet_data *sd) sd->rps_ipi_list = NULL; local_irq_enable(); @@ -31150,7 +31559,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c /* Send pending IPI's to kick RPS processing on remote cpus. */ while (remsd) { -@@ -4570,6 +4650,7 @@ +@@ -4570,6 +4650,7 @@ static void net_rps_action_and_irq_enable(struct softnet_data *sd) } else #endif local_irq_enable(); @@ -31158,7 +31567,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c } static bool sd_has_rps_ipi_waiting(struct softnet_data *sd) -@@ -4651,9 +4732,11 @@ +@@ -4651,9 +4732,11 @@ void __napi_schedule(struct napi_struct *n) local_irq_save(flags); ____napi_schedule(this_cpu_ptr(&softnet_data), n); local_irq_restore(flags); @@ -31170,7 +31579,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c /** * __napi_schedule_irqoff - schedule for receive * @n: entry to schedule -@@ -4665,6 +4748,7 @@ +@@ -4665,6 +4748,7 @@ void __napi_schedule_irqoff(struct napi_struct *n) ____napi_schedule(this_cpu_ptr(&softnet_data), n); } EXPORT_SYMBOL(__napi_schedule_irqoff); @@ -31178,7 +31587,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c void __napi_complete(struct napi_struct *n) { -@@ -4891,13 +4975,21 @@ +@@ -4891,13 +4975,21 @@ static void net_rx_action(struct softirq_action *h) struct softnet_data *sd = this_cpu_ptr(&softnet_data); unsigned long time_limit = jiffies + 2; int budget = netdev_budget; @@ -31200,7 +31609,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c for (;;) { struct napi_struct *n; -@@ -4927,7 +5019,7 @@ +@@ -4927,7 +5019,7 @@ static void net_rx_action(struct softirq_action *h) list_splice_tail(&repoll, &list); list_splice(&list, &sd->poll_list); if (!list_empty(&sd->poll_list)) @@ -31209,7 +31618,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c net_rps_action_and_irq_enable(sd); } -@@ -7266,7 +7358,7 @@ +@@ -7266,7 +7358,7 @@ EXPORT_SYMBOL(free_netdev); void synchronize_net(void) { might_sleep(); @@ -31218,7 +31627,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c synchronize_rcu_expedited(); else synchronize_rcu(); -@@ -7507,16 +7599,20 @@ +@@ -7507,16 +7599,20 @@ static int dev_cpu_callback(struct notifier_block *nfb, raise_softirq_irqoff(NET_TX_SOFTIRQ); local_irq_enable(); @@ -31240,7 +31649,7 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c return NOTIFY_OK; } -@@ -7818,8 +7914,9 @@ +@@ -7818,8 +7914,9 @@ static int __init net_dev_init(void) for_each_possible_cpu(i) { struct softnet_data *sd = &per_cpu(softnet_data, i); @@ -31252,9 +31661,10 @@ diff -Nur linux-4.4.62.orig/net/core/dev.c linux-4.4.62/net/core/dev.c INIT_LIST_HEAD(&sd->poll_list); sd->output_queue_tailp = &sd->output_queue; #ifdef CONFIG_RPS -diff -Nur linux-4.4.62.orig/net/core/skbuff.c linux-4.4.62/net/core/skbuff.c ---- linux-4.4.62.orig/net/core/skbuff.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/net/core/skbuff.c 2017-04-18 17:38:08.238650951 +0200 +diff --git a/net/core/skbuff.c b/net/core/skbuff.c +index 73dfd7729bc9..62f62810cddd 100644 +--- a/net/core/skbuff.c ++++ b/net/core/skbuff.c @@ -63,6 +63,7 @@ #include #include @@ -31263,7 +31673,7 @@ diff -Nur linux-4.4.62.orig/net/core/skbuff.c linux-4.4.62/net/core/skbuff.c #include #include -@@ -351,6 +352,8 @@ +@@ -351,6 +352,8 @@ EXPORT_SYMBOL(build_skb); static DEFINE_PER_CPU(struct page_frag_cache, netdev_alloc_cache); static DEFINE_PER_CPU(struct page_frag_cache, napi_alloc_cache); @@ -31272,7 +31682,7 @@ diff -Nur linux-4.4.62.orig/net/core/skbuff.c linux-4.4.62/net/core/skbuff.c static void *__netdev_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) { -@@ -358,10 +361,10 @@ +@@ -358,10 +361,10 @@ static void *__netdev_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) unsigned long flags; void *data; @@ -31285,7 +31695,7 @@ diff -Nur linux-4.4.62.orig/net/core/skbuff.c linux-4.4.62/net/core/skbuff.c return data; } -@@ -380,9 +383,13 @@ +@@ -380,9 +383,13 @@ EXPORT_SYMBOL(netdev_alloc_frag); static void *__napi_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) { @@ -31301,7 +31711,7 @@ diff -Nur linux-4.4.62.orig/net/core/skbuff.c linux-4.4.62/net/core/skbuff.c } void *napi_alloc_frag(unsigned int fragsz) -@@ -429,13 +436,13 @@ +@@ -429,13 +436,13 @@ struct sk_buff *__netdev_alloc_skb(struct net_device *dev, unsigned int len, if (sk_memalloc_socks()) gfp_mask |= __GFP_MEMALLOC; @@ -31317,7 +31727,7 @@ diff -Nur linux-4.4.62.orig/net/core/skbuff.c linux-4.4.62/net/core/skbuff.c if (unlikely(!data)) return NULL; -@@ -476,9 +483,10 @@ +@@ -476,9 +483,10 @@ EXPORT_SYMBOL(__netdev_alloc_skb); struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len, gfp_t gfp_mask) { @@ -31329,7 +31739,7 @@ diff -Nur linux-4.4.62.orig/net/core/skbuff.c linux-4.4.62/net/core/skbuff.c len += NET_SKB_PAD + NET_IP_ALIGN; -@@ -496,7 +504,11 @@ +@@ -496,7 +504,11 @@ struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len, if (sk_memalloc_socks()) gfp_mask |= __GFP_MEMALLOC; @@ -31341,7 +31751,7 @@ diff -Nur linux-4.4.62.orig/net/core/skbuff.c linux-4.4.62/net/core/skbuff.c if (unlikely(!data)) return NULL; -@@ -507,7 +519,7 @@ +@@ -507,7 +519,7 @@ struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len, } /* use OR instead of assignment to avoid clearing of bits in mask */ @@ -31350,10 +31760,11 @@ diff -Nur linux-4.4.62.orig/net/core/skbuff.c linux-4.4.62/net/core/skbuff.c skb->pfmemalloc = 1; skb->head_frag = 1; -diff -Nur linux-4.4.62.orig/net/core/sock.c linux-4.4.62/net/core/sock.c ---- linux-4.4.62.orig/net/core/sock.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/net/core/sock.c 2017-04-18 17:38:08.238650951 +0200 -@@ -2447,12 +2447,11 @@ +diff --git a/net/core/sock.c b/net/core/sock.c +index 9c708a5fb751..823377a7e63a 100644 +--- a/net/core/sock.c ++++ b/net/core/sock.c +@@ -2447,12 +2447,11 @@ void lock_sock_nested(struct sock *sk, int subclass) if (sk->sk_lock.owned) __lock_sock(sk); sk->sk_lock.owned = 1; @@ -31367,9 +31778,10 @@ diff -Nur linux-4.4.62.orig/net/core/sock.c linux-4.4.62/net/core/sock.c } EXPORT_SYMBOL(lock_sock_nested); -diff -Nur linux-4.4.62.orig/net/ipv4/icmp.c linux-4.4.62/net/ipv4/icmp.c ---- linux-4.4.62.orig/net/ipv4/icmp.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/net/ipv4/icmp.c 2017-04-18 17:38:08.238650951 +0200 +diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c +index 36e26977c908..ff2593269089 100644 +--- a/net/ipv4/icmp.c ++++ b/net/ipv4/icmp.c @@ -69,6 +69,7 @@ #include #include @@ -31386,7 +31798,7 @@ diff -Nur linux-4.4.62.orig/net/ipv4/icmp.c linux-4.4.62/net/ipv4/icmp.c #include #include #include -@@ -204,6 +206,8 @@ +@@ -204,6 +206,8 @@ static const struct icmp_control icmp_pointers[NR_ICMP_TYPES+1]; * * On SMP we have one ICMP socket per-cpu. */ @@ -31395,7 +31807,7 @@ diff -Nur linux-4.4.62.orig/net/ipv4/icmp.c linux-4.4.62/net/ipv4/icmp.c static struct sock *icmp_sk(struct net *net) { return *this_cpu_ptr(net->ipv4.icmp_sk); -@@ -215,12 +219,14 @@ +@@ -215,12 +219,14 @@ static inline struct sock *icmp_xmit_lock(struct net *net) local_bh_disable(); @@ -31410,7 +31822,7 @@ diff -Nur linux-4.4.62.orig/net/ipv4/icmp.c linux-4.4.62/net/ipv4/icmp.c local_bh_enable(); return NULL; } -@@ -230,6 +236,7 @@ +@@ -230,6 +236,7 @@ static inline struct sock *icmp_xmit_lock(struct net *net) static inline void icmp_xmit_unlock(struct sock *sk) { spin_unlock_bh(&sk->sk_lock.slock); @@ -31418,7 +31830,7 @@ diff -Nur linux-4.4.62.orig/net/ipv4/icmp.c linux-4.4.62/net/ipv4/icmp.c } int sysctl_icmp_msgs_per_sec __read_mostly = 1000; -@@ -358,6 +365,7 @@ +@@ -358,6 +365,7 @@ static void icmp_push_reply(struct icmp_bxm *icmp_param, struct sock *sk; struct sk_buff *skb; @@ -31426,7 +31838,7 @@ diff -Nur linux-4.4.62.orig/net/ipv4/icmp.c linux-4.4.62/net/ipv4/icmp.c sk = icmp_sk(dev_net((*rt)->dst.dev)); if (ip_append_data(sk, fl4, icmp_glue_bits, icmp_param, icmp_param->data_len+icmp_param->head_len, -@@ -380,6 +388,7 @@ +@@ -380,6 +388,7 @@ static void icmp_push_reply(struct icmp_bxm *icmp_param, skb->ip_summed = CHECKSUM_NONE; ip_push_pending_frames(sk, fl4); } @@ -31434,7 +31846,7 @@ diff -Nur linux-4.4.62.orig/net/ipv4/icmp.c linux-4.4.62/net/ipv4/icmp.c } /* -@@ -891,6 +900,30 @@ +@@ -891,6 +900,30 @@ static bool icmp_redirect(struct sk_buff *skb) } /* @@ -31465,7 +31877,7 @@ diff -Nur linux-4.4.62.orig/net/ipv4/icmp.c linux-4.4.62/net/ipv4/icmp.c * Handle ICMP_ECHO ("ping") requests. * * RFC 1122: 3.2.2.6 MUST have an echo server that answers ICMP echo -@@ -917,6 +950,11 @@ +@@ -917,6 +950,11 @@ static bool icmp_echo(struct sk_buff *skb) icmp_param.data_len = skb->len; icmp_param.head_len = sizeof(struct icmphdr); icmp_reply(&icmp_param, skb); @@ -31477,10 +31889,11 @@ diff -Nur linux-4.4.62.orig/net/ipv4/icmp.c linux-4.4.62/net/ipv4/icmp.c } /* should there be an ICMP stat for ignored echos? */ return true; -diff -Nur linux-4.4.62.orig/net/ipv4/sysctl_net_ipv4.c linux-4.4.62/net/ipv4/sysctl_net_ipv4.c ---- linux-4.4.62.orig/net/ipv4/sysctl_net_ipv4.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/net/ipv4/sysctl_net_ipv4.c 2017-04-18 17:38:08.238650951 +0200 -@@ -818,6 +818,13 @@ +diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c +index 70fb352e317f..1bcd436709a4 100644 +--- a/net/ipv4/sysctl_net_ipv4.c ++++ b/net/ipv4/sysctl_net_ipv4.c +@@ -818,6 +818,13 @@ static struct ctl_table ipv4_net_table[] = { .proc_handler = proc_dointvec }, { @@ -31494,9 +31907,10 @@ diff -Nur linux-4.4.62.orig/net/ipv4/sysctl_net_ipv4.c linux-4.4.62/net/ipv4/sys .procname = "icmp_ignore_bogus_error_responses", .data = &init_net.ipv4.sysctl_icmp_ignore_bogus_error_responses, .maxlen = sizeof(int), -diff -Nur linux-4.4.62.orig/net/ipv4/tcp_ipv4.c linux-4.4.62/net/ipv4/tcp_ipv4.c ---- linux-4.4.62.orig/net/ipv4/tcp_ipv4.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/net/ipv4/tcp_ipv4.c 2017-04-18 17:38:08.238650951 +0200 +diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c +index 198fc2314c82..6d84e7c8c8a9 100644 +--- a/net/ipv4/tcp_ipv4.c ++++ b/net/ipv4/tcp_ipv4.c @@ -62,6 +62,7 @@ #include #include @@ -31505,7 +31919,7 @@ diff -Nur linux-4.4.62.orig/net/ipv4/tcp_ipv4.c linux-4.4.62/net/ipv4/tcp_ipv4.c #include #include -@@ -570,6 +571,7 @@ +@@ -570,6 +571,7 @@ void tcp_v4_send_check(struct sock *sk, struct sk_buff *skb) } EXPORT_SYMBOL(tcp_v4_send_check); @@ -31513,7 +31927,7 @@ diff -Nur linux-4.4.62.orig/net/ipv4/tcp_ipv4.c linux-4.4.62/net/ipv4/tcp_ipv4.c /* * This routine will send an RST to the other tcp. * -@@ -691,10 +693,13 @@ +@@ -691,10 +693,13 @@ static void tcp_v4_send_reset(const struct sock *sk, struct sk_buff *skb) arg.bound_dev_if = sk->sk_bound_dev_if; arg.tos = ip_hdr(skb)->tos; @@ -31527,7 +31941,7 @@ diff -Nur linux-4.4.62.orig/net/ipv4/tcp_ipv4.c linux-4.4.62/net/ipv4/tcp_ipv4.c TCP_INC_STATS_BH(net, TCP_MIB_OUTSEGS); TCP_INC_STATS_BH(net, TCP_MIB_OUTRSTS); -@@ -776,10 +781,12 @@ +@@ -776,10 +781,12 @@ static void tcp_v4_send_ack(struct net *net, if (oif) arg.bound_dev_if = oif; arg.tos = tos; @@ -31540,10 +31954,11 @@ diff -Nur linux-4.4.62.orig/net/ipv4/tcp_ipv4.c linux-4.4.62/net/ipv4/tcp_ipv4.c TCP_INC_STATS_BH(net, TCP_MIB_OUTSEGS); } -diff -Nur linux-4.4.62.orig/net/mac80211/rx.c linux-4.4.62/net/mac80211/rx.c ---- linux-4.4.62.orig/net/mac80211/rx.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/net/mac80211/rx.c 2017-04-18 17:38:08.242651106 +0200 -@@ -3580,7 +3580,7 @@ +diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c +index 9f0915f72702..28d234c03cb5 100644 +--- a/net/mac80211/rx.c ++++ b/net/mac80211/rx.c +@@ -3601,7 +3601,7 @@ void ieee80211_rx_napi(struct ieee80211_hw *hw, struct sk_buff *skb, struct ieee80211_supported_band *sband; struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(skb); @@ -31552,9 +31967,10 @@ diff -Nur linux-4.4.62.orig/net/mac80211/rx.c linux-4.4.62/net/mac80211/rx.c if (WARN_ON(status->band >= IEEE80211_NUM_BANDS)) goto drop; -diff -Nur linux-4.4.62.orig/net/netfilter/core.c linux-4.4.62/net/netfilter/core.c ---- linux-4.4.62.orig/net/netfilter/core.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/net/netfilter/core.c 2017-04-18 17:38:08.242651106 +0200 +diff --git a/net/netfilter/core.c b/net/netfilter/core.c +index f39276d1c2d7..10880c89d62f 100644 +--- a/net/netfilter/core.c ++++ b/net/netfilter/core.c @@ -22,11 +22,17 @@ #include #include @@ -31573,9 +31989,10 @@ diff -Nur linux-4.4.62.orig/net/netfilter/core.c linux-4.4.62/net/netfilter/core static DEFINE_MUTEX(afinfo_mutex); const struct nf_afinfo __rcu *nf_afinfo[NFPROTO_NUMPROTO] __read_mostly; -diff -Nur linux-4.4.62.orig/net/packet/af_packet.c linux-4.4.62/net/packet/af_packet.c ---- linux-4.4.62.orig/net/packet/af_packet.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/net/packet/af_packet.c 2017-04-18 17:38:08.242651106 +0200 +diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c +index f8d6a0ca9c03..f6be45510cf6 100644 +--- a/net/packet/af_packet.c ++++ b/net/packet/af_packet.c @@ -63,6 +63,7 @@ #include #include @@ -31584,7 +32001,7 @@ diff -Nur linux-4.4.62.orig/net/packet/af_packet.c linux-4.4.62/net/packet/af_pa #include #include #include -@@ -694,7 +695,7 @@ +@@ -694,7 +695,7 @@ static void prb_retire_rx_blk_timer_expired(unsigned long data) if (BLOCK_NUM_PKTS(pbd)) { while (atomic_read(&pkc->blk_fill_in_prog)) { /* Waiting for skb_copy_bits to finish... */ @@ -31593,7 +32010,7 @@ diff -Nur linux-4.4.62.orig/net/packet/af_packet.c linux-4.4.62/net/packet/af_pa } } -@@ -956,7 +957,7 @@ +@@ -956,7 +957,7 @@ static void prb_retire_current_block(struct tpacket_kbdq_core *pkc, if (!(status & TP_STATUS_BLK_TMO)) { while (atomic_read(&pkc->blk_fill_in_prog)) { /* Waiting for skb_copy_bits to finish... */ @@ -31602,9 +32019,10 @@ diff -Nur linux-4.4.62.orig/net/packet/af_packet.c linux-4.4.62/net/packet/af_pa } } prb_close_block(pkc, pbd, po, status); -diff -Nur linux-4.4.62.orig/net/rds/ib_rdma.c linux-4.4.62/net/rds/ib_rdma.c ---- linux-4.4.62.orig/net/rds/ib_rdma.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/net/rds/ib_rdma.c 2017-04-18 17:38:08.242651106 +0200 +diff --git a/net/rds/ib_rdma.c b/net/rds/ib_rdma.c +index a2340748ec86..19123a97b354 100644 +--- a/net/rds/ib_rdma.c ++++ b/net/rds/ib_rdma.c @@ -34,6 +34,7 @@ #include #include @@ -31613,7 +32031,7 @@ diff -Nur linux-4.4.62.orig/net/rds/ib_rdma.c linux-4.4.62/net/rds/ib_rdma.c #include "rds.h" #include "ib.h" -@@ -313,7 +314,7 @@ +@@ -313,7 +314,7 @@ static inline void wait_clean_list_grace(void) for_each_online_cpu(cpu) { flag = &per_cpu(clean_list_grace, cpu); while (test_bit(CLEAN_LIST_BUSY_BIT, flag)) @@ -31622,10 +32040,11 @@ diff -Nur linux-4.4.62.orig/net/rds/ib_rdma.c linux-4.4.62/net/rds/ib_rdma.c } } -diff -Nur linux-4.4.62.orig/net/sched/sch_generic.c linux-4.4.62/net/sched/sch_generic.c ---- linux-4.4.62.orig/net/sched/sch_generic.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/net/sched/sch_generic.c 2017-04-18 17:38:08.242651106 +0200 -@@ -893,7 +893,7 @@ +diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c +index aa4725038f94..00b81cab28f3 100644 +--- a/net/sched/sch_generic.c ++++ b/net/sched/sch_generic.c +@@ -893,7 +893,7 @@ void dev_deactivate_many(struct list_head *head) /* Wait for outstanding qdisc_run calls. */ list_for_each_entry(dev, head, close_list) while (some_qdisc_is_busy(dev)) @@ -31634,10 +32053,11 @@ diff -Nur linux-4.4.62.orig/net/sched/sch_generic.c linux-4.4.62/net/sched/sch_g } void dev_deactivate(struct net_device *dev) -diff -Nur linux-4.4.62.orig/net/sunrpc/svc_xprt.c linux-4.4.62/net/sunrpc/svc_xprt.c ---- linux-4.4.62.orig/net/sunrpc/svc_xprt.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/net/sunrpc/svc_xprt.c 2017-04-18 17:38:08.242651106 +0200 -@@ -340,7 +340,7 @@ +diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c +index a6cbb2104667..5b69bb580617 100644 +--- a/net/sunrpc/svc_xprt.c ++++ b/net/sunrpc/svc_xprt.c +@@ -340,7 +340,7 @@ void svc_xprt_do_enqueue(struct svc_xprt *xprt) goto out; } @@ -31646,7 +32066,7 @@ diff -Nur linux-4.4.62.orig/net/sunrpc/svc_xprt.c linux-4.4.62/net/sunrpc/svc_xp pool = svc_pool_for_cpu(xprt->xpt_server, cpu); atomic_long_inc(&pool->sp_stats.packets); -@@ -376,7 +376,7 @@ +@@ -376,7 +376,7 @@ redo_search: atomic_long_inc(&pool->sp_stats.threads_woken); wake_up_process(rqstp->rq_task); @@ -31655,7 +32075,7 @@ diff -Nur linux-4.4.62.orig/net/sunrpc/svc_xprt.c linux-4.4.62/net/sunrpc/svc_xp goto out; } rcu_read_unlock(); -@@ -397,7 +397,7 @@ +@@ -397,7 +397,7 @@ redo_search: goto redo_search; } rqstp = NULL; @@ -31664,10 +32084,11 @@ diff -Nur linux-4.4.62.orig/net/sunrpc/svc_xprt.c linux-4.4.62/net/sunrpc/svc_xp out: trace_svc_xprt_do_enqueue(xprt, rqstp); } -diff -Nur linux-4.4.62.orig/scripts/mkcompile_h linux-4.4.62/scripts/mkcompile_h ---- linux-4.4.62.orig/scripts/mkcompile_h 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/scripts/mkcompile_h 2017-04-18 17:38:08.242651106 +0200 -@@ -4,7 +4,8 @@ +diff --git a/scripts/mkcompile_h b/scripts/mkcompile_h +index 6fdc97ef6023..523e0420d7f0 100755 +--- a/scripts/mkcompile_h ++++ b/scripts/mkcompile_h +@@ -4,7 +4,8 @@ TARGET=$1 ARCH=$2 SMP=$3 PREEMPT=$4 @@ -31677,7 +32098,7 @@ diff -Nur linux-4.4.62.orig/scripts/mkcompile_h linux-4.4.62/scripts/mkcompile_h vecho() { [ "${quiet}" = "silent_" ] || echo "$@" ; } -@@ -57,6 +58,7 @@ +@@ -57,6 +58,7 @@ UTS_VERSION="#$VERSION" CONFIG_FLAGS="" if [ -n "$SMP" ] ; then CONFIG_FLAGS="SMP"; fi if [ -n "$PREEMPT" ] ; then CONFIG_FLAGS="$CONFIG_FLAGS PREEMPT"; fi @@ -31685,10 +32106,11 @@ diff -Nur linux-4.4.62.orig/scripts/mkcompile_h linux-4.4.62/scripts/mkcompile_h UTS_VERSION="$UTS_VERSION $CONFIG_FLAGS $TIMESTAMP" # Truncate to maximum length -diff -Nur linux-4.4.62.orig/sound/core/pcm_native.c linux-4.4.62/sound/core/pcm_native.c ---- linux-4.4.62.orig/sound/core/pcm_native.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/sound/core/pcm_native.c 2017-04-18 17:38:08.242651106 +0200 -@@ -135,7 +135,7 @@ +diff --git a/sound/core/pcm_native.c b/sound/core/pcm_native.c +index 4ba64fd49759..34e50186885d 100644 +--- a/sound/core/pcm_native.c ++++ b/sound/core/pcm_native.c +@@ -135,7 +135,7 @@ EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock); void snd_pcm_stream_lock_irq(struct snd_pcm_substream *substream) { if (!substream->pcm->nonatomic) @@ -31697,7 +32119,7 @@ diff -Nur linux-4.4.62.orig/sound/core/pcm_native.c linux-4.4.62/sound/core/pcm_ snd_pcm_stream_lock(substream); } EXPORT_SYMBOL_GPL(snd_pcm_stream_lock_irq); -@@ -150,7 +150,7 @@ +@@ -150,7 +150,7 @@ void snd_pcm_stream_unlock_irq(struct snd_pcm_substream *substream) { snd_pcm_stream_unlock(substream); if (!substream->pcm->nonatomic) @@ -31706,7 +32128,7 @@ diff -Nur linux-4.4.62.orig/sound/core/pcm_native.c linux-4.4.62/sound/core/pcm_ } EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock_irq); -@@ -158,7 +158,7 @@ +@@ -158,7 +158,7 @@ unsigned long _snd_pcm_stream_lock_irqsave(struct snd_pcm_substream *substream) { unsigned long flags = 0; if (!substream->pcm->nonatomic) @@ -31715,7 +32137,7 @@ diff -Nur linux-4.4.62.orig/sound/core/pcm_native.c linux-4.4.62/sound/core/pcm_ snd_pcm_stream_lock(substream); return flags; } -@@ -176,7 +176,7 @@ +@@ -176,7 +176,7 @@ void snd_pcm_stream_unlock_irqrestore(struct snd_pcm_substream *substream, { snd_pcm_stream_unlock(substream); if (!substream->pcm->nonatomic) @@ -31724,10 +32146,11 @@ diff -Nur linux-4.4.62.orig/sound/core/pcm_native.c linux-4.4.62/sound/core/pcm_ } EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock_irqrestore); -diff -Nur linux-4.4.62.orig/virt/kvm/async_pf.c linux-4.4.62/virt/kvm/async_pf.c ---- linux-4.4.62.orig/virt/kvm/async_pf.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/virt/kvm/async_pf.c 2017-04-18 17:38:08.242651106 +0200 -@@ -98,8 +98,8 @@ +diff --git a/virt/kvm/async_pf.c b/virt/kvm/async_pf.c +index 4f70d12e392d..9378d0919ed8 100644 +--- a/virt/kvm/async_pf.c ++++ b/virt/kvm/async_pf.c +@@ -98,8 +98,8 @@ static void async_pf_execute(struct work_struct *work) * This memory barrier pairs with prepare_to_wait's set_current_state() */ smp_mb(); @@ -31738,10 +32161,11 @@ diff -Nur linux-4.4.62.orig/virt/kvm/async_pf.c linux-4.4.62/virt/kvm/async_pf.c mmput(mm); kvm_put_kvm(vcpu->kvm); -diff -Nur linux-4.4.62.orig/virt/kvm/kvm_main.c linux-4.4.62/virt/kvm/kvm_main.c ---- linux-4.4.62.orig/virt/kvm/kvm_main.c 2017-04-18 07:15:37.000000000 +0200 -+++ linux-4.4.62/virt/kvm/kvm_main.c 2017-04-18 17:38:08.242651106 +0200 -@@ -228,8 +228,7 @@ +diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c +index cb092bd9965b..133e7fa2ec8b 100644 +--- a/virt/kvm/kvm_main.c ++++ b/virt/kvm/kvm_main.c +@@ -228,8 +228,7 @@ int kvm_vcpu_init(struct kvm_vcpu *vcpu, struct kvm *kvm, unsigned id) vcpu->kvm = kvm; vcpu->vcpu_id = id; vcpu->pid = NULL; @@ -31751,7 +32175,7 @@ diff -Nur linux-4.4.62.orig/virt/kvm/kvm_main.c linux-4.4.62/virt/kvm/kvm_main.c kvm_async_pf_vcpu_init(vcpu); vcpu->pre_pcpu = -1; -@@ -2008,7 +2007,7 @@ +@@ -2008,7 +2007,7 @@ static int kvm_vcpu_check_block(struct kvm_vcpu *vcpu) void kvm_vcpu_block(struct kvm_vcpu *vcpu) { ktime_t start, cur; @@ -31760,7 +32184,7 @@ diff -Nur linux-4.4.62.orig/virt/kvm/kvm_main.c linux-4.4.62/virt/kvm/kvm_main.c bool waited = false; u64 block_ns; -@@ -2033,7 +2032,7 @@ +@@ -2033,7 +2032,7 @@ void kvm_vcpu_block(struct kvm_vcpu *vcpu) kvm_arch_vcpu_blocking(vcpu); for (;;) { @@ -31769,7 +32193,7 @@ diff -Nur linux-4.4.62.orig/virt/kvm/kvm_main.c linux-4.4.62/virt/kvm/kvm_main.c if (kvm_vcpu_check_block(vcpu) < 0) break; -@@ -2042,7 +2041,7 @@ +@@ -2042,7 +2041,7 @@ void kvm_vcpu_block(struct kvm_vcpu *vcpu) schedule(); } @@ -31778,7 +32202,7 @@ diff -Nur linux-4.4.62.orig/virt/kvm/kvm_main.c linux-4.4.62/virt/kvm/kvm_main.c cur = ktime_get(); kvm_arch_vcpu_unblocking(vcpu); -@@ -2074,11 +2073,11 @@ +@@ -2074,11 +2073,11 @@ void kvm_vcpu_kick(struct kvm_vcpu *vcpu) { int me; int cpu = vcpu->cpu; @@ -31793,7 +32217,7 @@ diff -Nur linux-4.4.62.orig/virt/kvm/kvm_main.c linux-4.4.62/virt/kvm/kvm_main.c ++vcpu->stat.halt_wakeup; } -@@ -2179,7 +2178,7 @@ +@@ -2179,7 +2178,7 @@ void kvm_vcpu_on_spin(struct kvm_vcpu *me) continue; if (vcpu == me) continue; diff --git a/target/linux/patches/4.9.32/patch-realtime b/target/linux/patches/4.9.32/patch-realtime index ab9028ddd..7abba16fe 100644 --- a/target/linux/patches/4.9.32/patch-realtime +++ b/target/linux/patches/4.9.32/patch-realtime @@ -1,299 +1,575 @@ -diff -Nur linux-4.9.28.orig/arch/arm/include/asm/irq.h linux-4.9.28/arch/arm/include/asm/irq.h ---- linux-4.9.28.orig/arch/arm/include/asm/irq.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/include/asm/irq.h 2017-05-19 03:37:25.122174217 +0200 -@@ -22,6 +22,8 @@ - #endif - - #ifndef __ASSEMBLY__ -+#include -+ - struct irqaction; - struct pt_regs; - extern void migrate_irqs(void); -diff -Nur linux-4.9.28.orig/arch/arm/include/asm/switch_to.h linux-4.9.28/arch/arm/include/asm/switch_to.h ---- linux-4.9.28.orig/arch/arm/include/asm/switch_to.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/include/asm/switch_to.h 2017-05-19 03:37:25.122174217 +0200 -@@ -3,6 +3,13 @@ +diff --git a/Documentation/sysrq.txt b/Documentation/sysrq.txt +index 3a3b30ac2a75..9e0745cafbd8 100644 +--- a/Documentation/sysrq.txt ++++ b/Documentation/sysrq.txt +@@ -59,10 +59,17 @@ On PowerPC - Press 'ALT - Print Screen (or F13) - , + On other - If you know of the key combos for other architectures, please + let me know so I can add them to this section. - #include +-On all - write a character to /proc/sysrq-trigger. e.g.: +- ++On all - write a character to /proc/sysrq-trigger, e.g.: + echo t > /proc/sysrq-trigger -+#if defined CONFIG_PREEMPT_RT_FULL && defined CONFIG_HIGHMEM -+void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p); -+#else -+static inline void -+switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) { } -+#endif ++On all - Enable network SysRq by writing a cookie to icmp_echo_sysrq, e.g. ++ echo 0x01020304 >/proc/sys/net/ipv4/icmp_echo_sysrq ++ Send an ICMP echo request with this pattern plus the particular ++ SysRq command key. Example: ++ # ping -c1 -s57 -p0102030468 ++ will trigger the SysRq-H (help) command. + - /* - * For v7 SMP cores running a preemptible kernel we may be pre-empted - * during a TLB maintenance operation, so execute an inner-shareable dsb -@@ -25,6 +32,7 @@ - #define switch_to(prev,next,last) \ - do { \ - __complete_pending_tlbi(); \ -+ switch_kmaps(prev, next); \ - last = __switch_to(prev,task_thread_info(prev), task_thread_info(next)); \ - } while (0) - -diff -Nur linux-4.9.28.orig/arch/arm/include/asm/thread_info.h linux-4.9.28/arch/arm/include/asm/thread_info.h ---- linux-4.9.28.orig/arch/arm/include/asm/thread_info.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/include/asm/thread_info.h 2017-05-19 03:37:25.122174217 +0200 -@@ -49,6 +49,7 @@ - struct thread_info { - unsigned long flags; /* low level flags */ - int preempt_count; /* 0 => preemptable, <0 => bug */ -+ int preempt_lazy_count; /* 0 => preemptable, <0 => bug */ - mm_segment_t addr_limit; /* address limit */ - struct task_struct *task; /* main task structure */ - __u32 cpu; /* cpu */ -@@ -142,7 +143,8 @@ - #define TIF_SYSCALL_TRACE 4 /* syscall trace active */ - #define TIF_SYSCALL_AUDIT 5 /* syscall auditing active */ - #define TIF_SYSCALL_TRACEPOINT 6 /* syscall tracepoint instrumentation */ --#define TIF_SECCOMP 7 /* seccomp syscall filtering active */ -+#define TIF_SECCOMP 8 /* seccomp syscall filtering active */ -+#define TIF_NEED_RESCHED_LAZY 7 - - #define TIF_NOHZ 12 /* in adaptive nohz mode */ - #define TIF_USING_IWMMXT 17 -@@ -152,6 +154,7 @@ - #define _TIF_SIGPENDING (1 << TIF_SIGPENDING) - #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) - #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) -+#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) - #define _TIF_UPROBE (1 << TIF_UPROBE) - #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) - #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) -@@ -167,7 +170,8 @@ - * Change these and you break ASM code in entry-common.S - */ - #define _TIF_WORK_MASK (_TIF_NEED_RESCHED | _TIF_SIGPENDING | \ -- _TIF_NOTIFY_RESUME | _TIF_UPROBE) -+ _TIF_NOTIFY_RESUME | _TIF_UPROBE | \ -+ _TIF_NEED_RESCHED_LAZY) - - #endif /* __KERNEL__ */ - #endif /* __ASM_ARM_THREAD_INFO_H */ -diff -Nur linux-4.9.28.orig/arch/arm/Kconfig linux-4.9.28/arch/arm/Kconfig ---- linux-4.9.28.orig/arch/arm/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/Kconfig 2017-05-19 03:37:25.122174217 +0200 -@@ -36,7 +36,7 @@ - select HAVE_ARCH_AUDITSYSCALL if (AEABI && !OABI_COMPAT) - select HAVE_ARCH_BITREVERSE if (CPU_32v7M || CPU_32v7) && !CPU_32v6 - select HAVE_ARCH_HARDENED_USERCOPY -- select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL && !CPU_ENDIAN_BE32 && MMU -+ select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL && !CPU_ENDIAN_BE32 && MMU && !PREEMPT_RT_BASE - select HAVE_ARCH_KGDB if !CPU_ENDIAN_BE32 && MMU - select HAVE_ARCH_MMAP_RND_BITS if MMU - select HAVE_ARCH_SECCOMP_FILTER if (AEABI && !OABI_COMPAT) -@@ -75,6 +75,7 @@ - select HAVE_PERF_EVENTS - select HAVE_PERF_REGS - select HAVE_PERF_USER_STACK_DUMP -+ select HAVE_PREEMPT_LAZY - select HAVE_RCU_TABLE_FREE if (SMP && ARM_LPAE) - select HAVE_REGS_AND_STACK_ACCESS_API - select HAVE_SYSCALL_TRACEPOINTS -diff -Nur linux-4.9.28.orig/arch/arm/kernel/asm-offsets.c linux-4.9.28/arch/arm/kernel/asm-offsets.c ---- linux-4.9.28.orig/arch/arm/kernel/asm-offsets.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/kernel/asm-offsets.c 2017-05-19 03:37:25.122174217 +0200 -@@ -65,6 +65,7 @@ - BLANK(); - DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); - DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); -+ DEFINE(TI_PREEMPT_LAZY, offsetof(struct thread_info, preempt_lazy_count)); - DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit)); - DEFINE(TI_TASK, offsetof(struct thread_info, task)); - DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); -diff -Nur linux-4.9.28.orig/arch/arm/kernel/entry-armv.S linux-4.9.28/arch/arm/kernel/entry-armv.S ---- linux-4.9.28.orig/arch/arm/kernel/entry-armv.S 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/kernel/entry-armv.S 2017-05-19 03:37:25.122174217 +0200 -@@ -220,11 +220,18 @@ - - #ifdef CONFIG_PREEMPT - ldr r8, [tsk, #TI_PREEMPT] @ get preempt count -- ldr r0, [tsk, #TI_FLAGS] @ get flags - teq r8, #0 @ if preempt count != 0 -+ bne 1f @ return from exeption -+ ldr r0, [tsk, #TI_FLAGS] @ get flags -+ tst r0, #_TIF_NEED_RESCHED @ if NEED_RESCHED is set -+ blne svc_preempt @ preempt! + -+ ldr r8, [tsk, #TI_PREEMPT_LAZY] @ get preempt lazy count -+ teq r8, #0 @ if preempt lazy count != 0 - movne r0, #0 @ force flags to 0 -- tst r0, #_TIF_NEED_RESCHED -+ tst r0, #_TIF_NEED_RESCHED_LAZY - blne svc_preempt -+1: - #endif - - svc_exit r5, irq = 1 @ return from exception -@@ -239,8 +246,14 @@ - 1: bl preempt_schedule_irq @ irq en/disable is done inside - ldr r0, [tsk, #TI_FLAGS] @ get new tasks TI_FLAGS - tst r0, #_TIF_NEED_RESCHED -+ bne 1b -+ tst r0, #_TIF_NEED_RESCHED_LAZY - reteq r8 @ go again -- b 1b -+ ldr r0, [tsk, #TI_PREEMPT_LAZY] @ get preempt lazy count -+ teq r0, #0 @ if preempt lazy count != 0 -+ beq 1b -+ ret r8 @ go again + * What are the 'command' keys? + ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + 'b' - Will immediately reboot the system without syncing or unmounting +diff --git a/Documentation/trace/histograms.txt b/Documentation/trace/histograms.txt +new file mode 100644 +index 000000000000..6f2aeabf7faa +--- /dev/null ++++ b/Documentation/trace/histograms.txt +@@ -0,0 +1,186 @@ ++ Using the Linux Kernel Latency Histograms + - #endif - - __und_fault: -diff -Nur linux-4.9.28.orig/arch/arm/kernel/entry-common.S linux-4.9.28/arch/arm/kernel/entry-common.S ---- linux-4.9.28.orig/arch/arm/kernel/entry-common.S 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/kernel/entry-common.S 2017-05-19 03:37:25.122174217 +0200 -@@ -36,7 +36,9 @@ - UNWIND(.cantunwind ) - disable_irq_notrace @ disable interrupts - ldr r1, [tsk, #TI_FLAGS] @ re-check for syscall tracing -- tst r1, #_TIF_SYSCALL_WORK | _TIF_WORK_MASK -+ tst r1, #((_TIF_SYSCALL_WORK | _TIF_WORK_MASK) & ~_TIF_SECCOMP) -+ bne fast_work_pending -+ tst r1, #_TIF_SECCOMP - bne fast_work_pending - - /* perform architecture specific actions before user return */ -@@ -62,8 +64,11 @@ - str r0, [sp, #S_R0 + S_OFF]! @ save returned r0 - disable_irq_notrace @ disable interrupts - ldr r1, [tsk, #TI_FLAGS] @ re-check for syscall tracing -- tst r1, #_TIF_SYSCALL_WORK | _TIF_WORK_MASK -+ tst r1, #((_TIF_SYSCALL_WORK | _TIF_WORK_MASK) & ~_TIF_SECCOMP) -+ bne do_slower_path -+ tst r1, #_TIF_SECCOMP - beq no_work_pending -+do_slower_path: - UNWIND(.fnend ) - ENDPROC(ret_fast_syscall) - -diff -Nur linux-4.9.28.orig/arch/arm/kernel/patch.c linux-4.9.28/arch/arm/kernel/patch.c ---- linux-4.9.28.orig/arch/arm/kernel/patch.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/kernel/patch.c 2017-05-19 03:37:25.122174217 +0200 -@@ -15,7 +15,7 @@ - unsigned int insn; - }; - --static DEFINE_SPINLOCK(patch_lock); -+static DEFINE_RAW_SPINLOCK(patch_lock); - - static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags) - __acquires(&patch_lock) -@@ -32,7 +32,7 @@ - return addr; - - if (flags) -- spin_lock_irqsave(&patch_lock, *flags); -+ raw_spin_lock_irqsave(&patch_lock, *flags); - else - __acquire(&patch_lock); - -@@ -47,7 +47,7 @@ - clear_fixmap(fixmap); - - if (flags) -- spin_unlock_irqrestore(&patch_lock, *flags); -+ raw_spin_unlock_irqrestore(&patch_lock, *flags); - else - __release(&patch_lock); - } -diff -Nur linux-4.9.28.orig/arch/arm/kernel/process.c linux-4.9.28/arch/arm/kernel/process.c ---- linux-4.9.28.orig/arch/arm/kernel/process.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/kernel/process.c 2017-05-19 03:37:25.122174217 +0200 -@@ -322,6 +322,30 @@ - } - - #ifdef CONFIG_MMU -+/* -+ * CONFIG_SPLIT_PTLOCK_CPUS results in a page->ptl lock. If the lock is not -+ * initialized by pgtable_page_ctor() then a coredump of the vector page will -+ * fail. -+ */ -+static int __init vectors_user_mapping_init_page(void) -+{ -+ struct page *page; -+ unsigned long addr = 0xffff0000; -+ pgd_t *pgd; -+ pud_t *pud; -+ pmd_t *pmd; + -+ pgd = pgd_offset_k(addr); -+ pud = pud_offset(pgd, addr); -+ pmd = pmd_offset(pud, addr); -+ page = pmd_page(*(pmd)); ++This document gives a short explanation how to enable, configure and use ++latency histograms. Latency histograms are primarily relevant in the ++context of real-time enabled kernels (CONFIG_PREEMPT/CONFIG_PREEMPT_RT) ++and are used in the quality management of the Linux real-time ++capabilities. + -+ pgtable_page_ctor(page); + -+ return 0; -+} -+late_initcall(vectors_user_mapping_init_page); ++* Purpose of latency histograms + - #ifdef CONFIG_KUSER_HELPERS - /* - * The vectors page is always readable from user space for the -diff -Nur linux-4.9.28.orig/arch/arm/kernel/signal.c linux-4.9.28/arch/arm/kernel/signal.c ---- linux-4.9.28.orig/arch/arm/kernel/signal.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/kernel/signal.c 2017-05-19 03:37:25.122174217 +0200 -@@ -572,7 +572,8 @@ - */ - trace_hardirqs_off(); - do { -- if (likely(thread_flags & _TIF_NEED_RESCHED)) { -+ if (likely(thread_flags & (_TIF_NEED_RESCHED | -+ _TIF_NEED_RESCHED_LAZY))) { - schedule(); - } else { - if (unlikely(!user_mode(regs))) -diff -Nur linux-4.9.28.orig/arch/arm/kernel/smp.c linux-4.9.28/arch/arm/kernel/smp.c ---- linux-4.9.28.orig/arch/arm/kernel/smp.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/kernel/smp.c 2017-05-19 03:37:25.122174217 +0200 -@@ -234,8 +234,6 @@ - flush_cache_louis(); - local_flush_tlb_all(); - -- clear_tasks_mm_cpumask(cpu); -- - return 0; - } - -@@ -251,6 +249,9 @@ - pr_err("CPU%u: cpu didn't die\n", cpu); - return; - } ++A latency histogram continuously accumulates the frequencies of latency ++data. There are two types of histograms ++- potential sources of latencies ++- effective latencies + -+ clear_tasks_mm_cpumask(cpu); + - pr_notice("CPU%u: shutdown\n", cpu); - - /* -diff -Nur linux-4.9.28.orig/arch/arm/kernel/unwind.c linux-4.9.28/arch/arm/kernel/unwind.c ---- linux-4.9.28.orig/arch/arm/kernel/unwind.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/kernel/unwind.c 2017-05-19 03:37:25.122174217 +0200 -@@ -93,7 +93,7 @@ - static const struct unwind_idx *__origin_unwind_idx; - extern const struct unwind_idx __stop_unwind_idx[]; - --static DEFINE_SPINLOCK(unwind_lock); -+static DEFINE_RAW_SPINLOCK(unwind_lock); - static LIST_HEAD(unwind_tables); - - /* Convert a prel31 symbol to an absolute address */ -@@ -201,7 +201,7 @@ - /* module unwind tables */ - struct unwind_table *table; - -- spin_lock_irqsave(&unwind_lock, flags); -+ raw_spin_lock_irqsave(&unwind_lock, flags); - list_for_each_entry(table, &unwind_tables, list) { - if (addr >= table->begin_addr && - addr < table->end_addr) { -@@ -213,7 +213,7 @@ - break; ++* Potential sources of latencies ++ ++Potential sources of latencies are code segments where interrupts, ++preemption or both are disabled (aka critical sections). To create ++histograms of potential sources of latency, the kernel stores the time ++stamp at the start of a critical section, determines the time elapsed ++when the end of the section is reached, and increments the frequency ++counter of that latency value - irrespective of whether any concurrently ++running process is affected by latency or not. ++- Configuration items (in the Kernel hacking/Tracers submenu) ++ CONFIG_INTERRUPT_OFF_LATENCY ++ CONFIG_PREEMPT_OFF_LATENCY ++ ++ ++* Effective latencies ++ ++Effective latencies are actually occuring during wakeup of a process. To ++determine effective latencies, the kernel stores the time stamp when a ++process is scheduled to be woken up, and determines the duration of the ++wakeup time shortly before control is passed over to this process. Note ++that the apparent latency in user space may be somewhat longer, since the ++process may be interrupted after control is passed over to it but before ++the execution in user space takes place. Simply measuring the interval ++between enqueuing and wakeup may also not appropriate in cases when a ++process is scheduled as a result of a timer expiration. The timer may have ++missed its deadline, e.g. due to disabled interrupts, but this latency ++would not be registered. Therefore, the offsets of missed timers are ++recorded in a separate histogram. If both wakeup latency and missed timer ++offsets are configured and enabled, a third histogram may be enabled that ++records the overall latency as a sum of the timer latency, if any, and the ++wakeup latency. This histogram is called "timerandwakeup". ++- Configuration items (in the Kernel hacking/Tracers submenu) ++ CONFIG_WAKEUP_LATENCY ++ CONFIG_MISSED_TIMER_OFSETS ++ ++ ++* Usage ++ ++The interface to the administration of the latency histograms is located ++in the debugfs file system. To mount it, either enter ++ ++mount -t sysfs nodev /sys ++mount -t debugfs nodev /sys/kernel/debug ++ ++from shell command line level, or add ++ ++nodev /sys sysfs defaults 0 0 ++nodev /sys/kernel/debug debugfs defaults 0 0 ++ ++to the file /etc/fstab. All latency histogram related files are then ++available in the directory /sys/kernel/debug/tracing/latency_hist. A ++particular histogram type is enabled by writing non-zero to the related ++variable in the /sys/kernel/debug/tracing/latency_hist/enable directory. ++Select "preemptirqsoff" for the histograms of potential sources of ++latencies and "wakeup" for histograms of effective latencies etc. The ++histogram data - one per CPU - are available in the files ++ ++/sys/kernel/debug/tracing/latency_hist/preemptoff/CPUx ++/sys/kernel/debug/tracing/latency_hist/irqsoff/CPUx ++/sys/kernel/debug/tracing/latency_hist/preemptirqsoff/CPUx ++/sys/kernel/debug/tracing/latency_hist/wakeup/CPUx ++/sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio/CPUx ++/sys/kernel/debug/tracing/latency_hist/missed_timer_offsets/CPUx ++/sys/kernel/debug/tracing/latency_hist/timerandwakeup/CPUx ++ ++The histograms are reset by writing non-zero to the file "reset" in a ++particular latency directory. To reset all latency data, use ++ ++#!/bin/sh ++ ++TRACINGDIR=/sys/kernel/debug/tracing ++HISTDIR=$TRACINGDIR/latency_hist ++ ++if test -d $HISTDIR ++then ++ cd $HISTDIR ++ for i in `find . | grep /reset$` ++ do ++ echo 1 >$i ++ done ++fi ++ ++ ++* Data format ++ ++Latency data are stored with a resolution of one microsecond. The ++maximum latency is 10,240 microseconds. The data are only valid, if the ++overflow register is empty. Every output line contains the latency in ++microseconds in the first row and the number of samples in the second ++row. To display only lines with a positive latency count, use, for ++example, ++ ++grep -v " 0$" /sys/kernel/debug/tracing/latency_hist/preemptoff/CPU0 ++ ++#Minimum latency: 0 microseconds. ++#Average latency: 0 microseconds. ++#Maximum latency: 25 microseconds. ++#Total samples: 3104770694 ++#There are 0 samples greater or equal than 10240 microseconds ++#usecs samples ++ 0 2984486876 ++ 1 49843506 ++ 2 58219047 ++ 3 5348126 ++ 4 2187960 ++ 5 3388262 ++ 6 959289 ++ 7 208294 ++ 8 40420 ++ 9 4485 ++ 10 14918 ++ 11 18340 ++ 12 25052 ++ 13 19455 ++ 14 5602 ++ 15 969 ++ 16 47 ++ 17 18 ++ 18 14 ++ 19 1 ++ 20 3 ++ 21 2 ++ 22 5 ++ 23 2 ++ 25 1 ++ ++ ++* Wakeup latency of a selected process ++ ++To only collect wakeup latency data of a particular process, write the ++PID of the requested process to ++ ++/sys/kernel/debug/tracing/latency_hist/wakeup/pid ++ ++PIDs are not considered, if this variable is set to 0. ++ ++ ++* Details of the process with the highest wakeup latency so far ++ ++Selected data of the process that suffered from the highest wakeup ++latency that occurred in a particular CPU are available in the file ++ ++/sys/kernel/debug/tracing/latency_hist/wakeup/max_latency-CPUx. ++ ++In addition, other relevant system data at the time when the ++latency occurred are given. ++ ++The format of the data is (all in one line): ++ () \ ++<- ++ ++The value of is only relevant in the combined timer ++and wakeup latency recording. In the wakeup recording, it is ++always 0, in the missed_timer_offsets recording, it is the same ++as . ++ ++When retrospectively searching for the origin of a latency and ++tracing was not enabled, it may be helpful to know the name and ++some basic data of the task that (finally) was switching to the ++late real-tlme task. In addition to the victim's data, also the ++data of the possible culprit are therefore displayed after the ++"<-" symbol. ++ ++Finally, the timestamp of the time when the latency occurred ++in . after the most recent system boot ++is provided. ++ ++These data are also reset when the wakeup histogram is reset. +diff --git a/MAINTAINERS b/MAINTAINERS +index 63cefa62324c..be0ea1e5c4cc 100644 +--- a/MAINTAINERS ++++ b/MAINTAINERS +@@ -5196,6 +5196,23 @@ F: fs/fuse/ + F: include/uapi/linux/fuse.h + F: Documentation/filesystems/fuse.txt + ++FUTEX SUBSYSTEM ++M: Thomas Gleixner ++M: Ingo Molnar ++R: Peter Zijlstra ++R: Darren Hart ++L: linux-kernel@vger.kernel.org ++T: git git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git locking/core ++S: Maintained ++F: kernel/futex.c ++F: kernel/futex_compat.c ++F: include/asm-generic/futex.h ++F: include/linux/futex.h ++F: include/uapi/linux/futex.h ++F: tools/testing/selftests/futex/ ++F: tools/perf/bench/futex* ++F: Documentation/*futex* ++ + FUTURE DOMAIN TMC-16x0 SCSI DRIVER (16-bit) + M: Rik Faith + L: linux-scsi@vger.kernel.org +diff --git a/arch/Kconfig b/arch/Kconfig +index 659bdd079277..099fc0f5155e 100644 +--- a/arch/Kconfig ++++ b/arch/Kconfig +@@ -9,6 +9,7 @@ config OPROFILE + tristate "OProfile system profiling" + depends on PROFILING + depends on HAVE_OPROFILE ++ depends on !PREEMPT_RT_FULL + select RING_BUFFER + select RING_BUFFER_ALLOW_SWAP + help +@@ -52,6 +53,7 @@ config KPROBES + config JUMP_LABEL + bool "Optimize very unlikely/likely branches" + depends on HAVE_ARCH_JUMP_LABEL ++ depends on (!INTERRUPT_OFF_HIST && !PREEMPT_OFF_HIST && !WAKEUP_LATENCY_HIST && !MISSED_TIMER_OFFSETS_HIST) + help + This option enables a transparent branch optimization that + makes certain almost-always-true or almost-always-false branch +diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig +index b5d529fdffab..5715844e83e3 100644 +--- a/arch/arm/Kconfig ++++ b/arch/arm/Kconfig +@@ -36,7 +36,7 @@ config ARM + select HAVE_ARCH_AUDITSYSCALL if (AEABI && !OABI_COMPAT) + select HAVE_ARCH_BITREVERSE if (CPU_32v7M || CPU_32v7) && !CPU_32v6 + select HAVE_ARCH_HARDENED_USERCOPY +- select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL && !CPU_ENDIAN_BE32 && MMU ++ select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL && !CPU_ENDIAN_BE32 && MMU && !PREEMPT_RT_BASE + select HAVE_ARCH_KGDB if !CPU_ENDIAN_BE32 && MMU + select HAVE_ARCH_MMAP_RND_BITS if MMU + select HAVE_ARCH_SECCOMP_FILTER if (AEABI && !OABI_COMPAT) +@@ -75,6 +75,7 @@ config ARM + select HAVE_PERF_EVENTS + select HAVE_PERF_REGS + select HAVE_PERF_USER_STACK_DUMP ++ select HAVE_PREEMPT_LAZY + select HAVE_RCU_TABLE_FREE if (SMP && ARM_LPAE) + select HAVE_REGS_AND_STACK_ACCESS_API + select HAVE_SYSCALL_TRACEPOINTS +diff --git a/arch/arm/include/asm/irq.h b/arch/arm/include/asm/irq.h +index e53638c8ed8a..6095a1649865 100644 +--- a/arch/arm/include/asm/irq.h ++++ b/arch/arm/include/asm/irq.h +@@ -22,6 +22,8 @@ + #endif + + #ifndef __ASSEMBLY__ ++#include ++ + struct irqaction; + struct pt_regs; + extern void migrate_irqs(void); +diff --git a/arch/arm/include/asm/switch_to.h b/arch/arm/include/asm/switch_to.h +index 12ebfcc1d539..c962084605bc 100644 +--- a/arch/arm/include/asm/switch_to.h ++++ b/arch/arm/include/asm/switch_to.h +@@ -3,6 +3,13 @@ + + #include + ++#if defined CONFIG_PREEMPT_RT_FULL && defined CONFIG_HIGHMEM ++void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p); ++#else ++static inline void ++switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) { } ++#endif ++ + /* + * For v7 SMP cores running a preemptible kernel we may be pre-empted + * during a TLB maintenance operation, so execute an inner-shareable dsb +@@ -25,6 +32,7 @@ extern struct task_struct *__switch_to(struct task_struct *, struct thread_info + #define switch_to(prev,next,last) \ + do { \ + __complete_pending_tlbi(); \ ++ switch_kmaps(prev, next); \ + last = __switch_to(prev,task_thread_info(prev), task_thread_info(next)); \ + } while (0) + +diff --git a/arch/arm/include/asm/thread_info.h b/arch/arm/include/asm/thread_info.h +index 776757d1604a..1f36a4eccc72 100644 +--- a/arch/arm/include/asm/thread_info.h ++++ b/arch/arm/include/asm/thread_info.h +@@ -49,6 +49,7 @@ struct cpu_context_save { + struct thread_info { + unsigned long flags; /* low level flags */ + int preempt_count; /* 0 => preemptable, <0 => bug */ ++ int preempt_lazy_count; /* 0 => preemptable, <0 => bug */ + mm_segment_t addr_limit; /* address limit */ + struct task_struct *task; /* main task structure */ + __u32 cpu; /* cpu */ +@@ -142,7 +143,8 @@ extern int vfp_restore_user_hwstate(struct user_vfp __user *, + #define TIF_SYSCALL_TRACE 4 /* syscall trace active */ + #define TIF_SYSCALL_AUDIT 5 /* syscall auditing active */ + #define TIF_SYSCALL_TRACEPOINT 6 /* syscall tracepoint instrumentation */ +-#define TIF_SECCOMP 7 /* seccomp syscall filtering active */ ++#define TIF_SECCOMP 8 /* seccomp syscall filtering active */ ++#define TIF_NEED_RESCHED_LAZY 7 + + #define TIF_NOHZ 12 /* in adaptive nohz mode */ + #define TIF_USING_IWMMXT 17 +@@ -152,6 +154,7 @@ extern int vfp_restore_user_hwstate(struct user_vfp __user *, + #define _TIF_SIGPENDING (1 << TIF_SIGPENDING) + #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) + #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) ++#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) + #define _TIF_UPROBE (1 << TIF_UPROBE) + #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) + #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) +@@ -167,7 +170,8 @@ extern int vfp_restore_user_hwstate(struct user_vfp __user *, + * Change these and you break ASM code in entry-common.S + */ + #define _TIF_WORK_MASK (_TIF_NEED_RESCHED | _TIF_SIGPENDING | \ +- _TIF_NOTIFY_RESUME | _TIF_UPROBE) ++ _TIF_NOTIFY_RESUME | _TIF_UPROBE | \ ++ _TIF_NEED_RESCHED_LAZY) + + #endif /* __KERNEL__ */ + #endif /* __ASM_ARM_THREAD_INFO_H */ +diff --git a/arch/arm/kernel/asm-offsets.c b/arch/arm/kernel/asm-offsets.c +index 608008229c7d..3866da3f7bb7 100644 +--- a/arch/arm/kernel/asm-offsets.c ++++ b/arch/arm/kernel/asm-offsets.c +@@ -65,6 +65,7 @@ int main(void) + BLANK(); + DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); + DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); ++ DEFINE(TI_PREEMPT_LAZY, offsetof(struct thread_info, preempt_lazy_count)); + DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit)); + DEFINE(TI_TASK, offsetof(struct thread_info, task)); + DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); +diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S +index 9f157e7c51e7..468e224d76aa 100644 +--- a/arch/arm/kernel/entry-armv.S ++++ b/arch/arm/kernel/entry-armv.S +@@ -220,11 +220,18 @@ ENDPROC(__dabt_svc) + + #ifdef CONFIG_PREEMPT + ldr r8, [tsk, #TI_PREEMPT] @ get preempt count +- ldr r0, [tsk, #TI_FLAGS] @ get flags + teq r8, #0 @ if preempt count != 0 ++ bne 1f @ return from exeption ++ ldr r0, [tsk, #TI_FLAGS] @ get flags ++ tst r0, #_TIF_NEED_RESCHED @ if NEED_RESCHED is set ++ blne svc_preempt @ preempt! ++ ++ ldr r8, [tsk, #TI_PREEMPT_LAZY] @ get preempt lazy count ++ teq r8, #0 @ if preempt lazy count != 0 + movne r0, #0 @ force flags to 0 +- tst r0, #_TIF_NEED_RESCHED ++ tst r0, #_TIF_NEED_RESCHED_LAZY + blne svc_preempt ++1: + #endif + + svc_exit r5, irq = 1 @ return from exception +@@ -239,8 +246,14 @@ ENDPROC(__irq_svc) + 1: bl preempt_schedule_irq @ irq en/disable is done inside + ldr r0, [tsk, #TI_FLAGS] @ get new tasks TI_FLAGS + tst r0, #_TIF_NEED_RESCHED ++ bne 1b ++ tst r0, #_TIF_NEED_RESCHED_LAZY + reteq r8 @ go again +- b 1b ++ ldr r0, [tsk, #TI_PREEMPT_LAZY] @ get preempt lazy count ++ teq r0, #0 @ if preempt lazy count != 0 ++ beq 1b ++ ret r8 @ go again ++ + #endif + + __und_fault: +diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S +index 10c3283d6c19..8872937862cc 100644 +--- a/arch/arm/kernel/entry-common.S ++++ b/arch/arm/kernel/entry-common.S +@@ -36,7 +36,9 @@ + UNWIND(.cantunwind ) + disable_irq_notrace @ disable interrupts + ldr r1, [tsk, #TI_FLAGS] @ re-check for syscall tracing +- tst r1, #_TIF_SYSCALL_WORK | _TIF_WORK_MASK ++ tst r1, #((_TIF_SYSCALL_WORK | _TIF_WORK_MASK) & ~_TIF_SECCOMP) ++ bne fast_work_pending ++ tst r1, #_TIF_SECCOMP + bne fast_work_pending + + /* perform architecture specific actions before user return */ +@@ -62,8 +64,11 @@ ENDPROC(ret_fast_syscall) + str r0, [sp, #S_R0 + S_OFF]! @ save returned r0 + disable_irq_notrace @ disable interrupts + ldr r1, [tsk, #TI_FLAGS] @ re-check for syscall tracing +- tst r1, #_TIF_SYSCALL_WORK | _TIF_WORK_MASK ++ tst r1, #((_TIF_SYSCALL_WORK | _TIF_WORK_MASK) & ~_TIF_SECCOMP) ++ bne do_slower_path ++ tst r1, #_TIF_SECCOMP + beq no_work_pending ++do_slower_path: + UNWIND(.fnend ) + ENDPROC(ret_fast_syscall) + +diff --git a/arch/arm/kernel/patch.c b/arch/arm/kernel/patch.c +index 69bda1a5707e..1f665acaa6a9 100644 +--- a/arch/arm/kernel/patch.c ++++ b/arch/arm/kernel/patch.c +@@ -15,7 +15,7 @@ struct patch { + unsigned int insn; + }; + +-static DEFINE_SPINLOCK(patch_lock); ++static DEFINE_RAW_SPINLOCK(patch_lock); + + static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags) + __acquires(&patch_lock) +@@ -32,7 +32,7 @@ static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags) + return addr; + + if (flags) +- spin_lock_irqsave(&patch_lock, *flags); ++ raw_spin_lock_irqsave(&patch_lock, *flags); + else + __acquire(&patch_lock); + +@@ -47,7 +47,7 @@ static void __kprobes patch_unmap(int fixmap, unsigned long *flags) + clear_fixmap(fixmap); + + if (flags) +- spin_unlock_irqrestore(&patch_lock, *flags); ++ raw_spin_unlock_irqrestore(&patch_lock, *flags); + else + __release(&patch_lock); + } +diff --git a/arch/arm/kernel/process.c b/arch/arm/kernel/process.c +index 91d2d5b01414..750550098b59 100644 +--- a/arch/arm/kernel/process.c ++++ b/arch/arm/kernel/process.c +@@ -322,6 +322,30 @@ unsigned long arch_randomize_brk(struct mm_struct *mm) + } + + #ifdef CONFIG_MMU ++/* ++ * CONFIG_SPLIT_PTLOCK_CPUS results in a page->ptl lock. If the lock is not ++ * initialized by pgtable_page_ctor() then a coredump of the vector page will ++ * fail. ++ */ ++static int __init vectors_user_mapping_init_page(void) ++{ ++ struct page *page; ++ unsigned long addr = 0xffff0000; ++ pgd_t *pgd; ++ pud_t *pud; ++ pmd_t *pmd; ++ ++ pgd = pgd_offset_k(addr); ++ pud = pud_offset(pgd, addr); ++ pmd = pmd_offset(pud, addr); ++ page = pmd_page(*(pmd)); ++ ++ pgtable_page_ctor(page); ++ ++ return 0; ++} ++late_initcall(vectors_user_mapping_init_page); ++ + #ifdef CONFIG_KUSER_HELPERS + /* + * The vectors page is always readable from user space for the +diff --git a/arch/arm/kernel/signal.c b/arch/arm/kernel/signal.c +index 7b8f2141427b..96541e00b74a 100644 +--- a/arch/arm/kernel/signal.c ++++ b/arch/arm/kernel/signal.c +@@ -572,7 +572,8 @@ do_work_pending(struct pt_regs *regs, unsigned int thread_flags, int syscall) + */ + trace_hardirqs_off(); + do { +- if (likely(thread_flags & _TIF_NEED_RESCHED)) { ++ if (likely(thread_flags & (_TIF_NEED_RESCHED | ++ _TIF_NEED_RESCHED_LAZY))) { + schedule(); + } else { + if (unlikely(!user_mode(regs))) +diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c +index 7dd14e8395e6..4cd7e3d98035 100644 +--- a/arch/arm/kernel/smp.c ++++ b/arch/arm/kernel/smp.c +@@ -234,8 +234,6 @@ int __cpu_disable(void) + flush_cache_louis(); + local_flush_tlb_all(); + +- clear_tasks_mm_cpumask(cpu); +- + return 0; + } + +@@ -251,6 +249,9 @@ void __cpu_die(unsigned int cpu) + pr_err("CPU%u: cpu didn't die\n", cpu); + return; + } ++ ++ clear_tasks_mm_cpumask(cpu); ++ + pr_notice("CPU%u: shutdown\n", cpu); + + /* +diff --git a/arch/arm/kernel/unwind.c b/arch/arm/kernel/unwind.c +index 0bee233fef9a..314cfb232a63 100644 +--- a/arch/arm/kernel/unwind.c ++++ b/arch/arm/kernel/unwind.c +@@ -93,7 +93,7 @@ extern const struct unwind_idx __start_unwind_idx[]; + static const struct unwind_idx *__origin_unwind_idx; + extern const struct unwind_idx __stop_unwind_idx[]; + +-static DEFINE_SPINLOCK(unwind_lock); ++static DEFINE_RAW_SPINLOCK(unwind_lock); + static LIST_HEAD(unwind_tables); + + /* Convert a prel31 symbol to an absolute address */ +@@ -201,7 +201,7 @@ static const struct unwind_idx *unwind_find_idx(unsigned long addr) + /* module unwind tables */ + struct unwind_table *table; + +- spin_lock_irqsave(&unwind_lock, flags); ++ raw_spin_lock_irqsave(&unwind_lock, flags); + list_for_each_entry(table, &unwind_tables, list) { + if (addr >= table->begin_addr && + addr < table->end_addr) { +@@ -213,7 +213,7 @@ static const struct unwind_idx *unwind_find_idx(unsigned long addr) + break; } } - spin_unlock_irqrestore(&unwind_lock, flags); @@ -301,7 +577,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/kernel/unwind.c linux-4.9.28/arch/arm/kerne } pr_debug("%s: idx = %p\n", __func__, idx); -@@ -529,9 +529,9 @@ +@@ -529,9 +529,9 @@ struct unwind_table *unwind_table_add(unsigned long start, unsigned long size, tab->begin_addr = text_addr; tab->end_addr = text_addr + text_size; @@ -313,7 +589,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/kernel/unwind.c linux-4.9.28/arch/arm/kerne return tab; } -@@ -543,9 +543,9 @@ +@@ -543,9 +543,9 @@ void unwind_table_del(struct unwind_table *tab) if (!tab) return; @@ -325,10 +601,11 @@ diff -Nur linux-4.9.28.orig/arch/arm/kernel/unwind.c linux-4.9.28/arch/arm/kerne kfree(tab); } -diff -Nur linux-4.9.28.orig/arch/arm/kvm/arm.c linux-4.9.28/arch/arm/kvm/arm.c ---- linux-4.9.28.orig/arch/arm/kvm/arm.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/kvm/arm.c 2017-05-19 03:37:25.122174217 +0200 -@@ -619,7 +619,7 @@ +diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c +index 19b5f5c1c0ff..82aa639e6737 100644 +--- a/arch/arm/kvm/arm.c ++++ b/arch/arm/kvm/arm.c +@@ -619,7 +619,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) * involves poking the GIC, which must be done in a * non-preemptible context. */ @@ -337,7 +614,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/kvm/arm.c linux-4.9.28/arch/arm/kvm/arm.c kvm_pmu_flush_hwstate(vcpu); kvm_timer_flush_hwstate(vcpu); kvm_vgic_flush_hwstate(vcpu); -@@ -640,7 +640,7 @@ +@@ -640,7 +640,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) kvm_pmu_sync_hwstate(vcpu); kvm_timer_sync_hwstate(vcpu); kvm_vgic_sync_hwstate(vcpu); @@ -346,7 +623,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/kvm/arm.c linux-4.9.28/arch/arm/kvm/arm.c continue; } -@@ -696,7 +696,7 @@ +@@ -696,7 +696,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) kvm_vgic_sync_hwstate(vcpu); @@ -355,10 +632,11 @@ diff -Nur linux-4.9.28.orig/arch/arm/kvm/arm.c linux-4.9.28/arch/arm/kvm/arm.c ret = handle_exit(vcpu, run, ret); } -diff -Nur linux-4.9.28.orig/arch/arm/mach-exynos/platsmp.c linux-4.9.28/arch/arm/mach-exynos/platsmp.c ---- linux-4.9.28.orig/arch/arm/mach-exynos/platsmp.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/mach-exynos/platsmp.c 2017-05-19 03:37:25.122174217 +0200 -@@ -229,7 +229,7 @@ +diff --git a/arch/arm/mach-exynos/platsmp.c b/arch/arm/mach-exynos/platsmp.c +index 98ffe1e62ad5..df9769ddece5 100644 +--- a/arch/arm/mach-exynos/platsmp.c ++++ b/arch/arm/mach-exynos/platsmp.c +@@ -229,7 +229,7 @@ static void __iomem *scu_base_addr(void) return (void __iomem *)(S5P_VA_SCU); } @@ -367,7 +645,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-exynos/platsmp.c linux-4.9.28/arch/arm static void exynos_secondary_init(unsigned int cpu) { -@@ -242,8 +242,8 @@ +@@ -242,8 +242,8 @@ static void exynos_secondary_init(unsigned int cpu) /* * Synchronise with the boot thread. */ @@ -378,7 +656,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-exynos/platsmp.c linux-4.9.28/arch/arm } int exynos_set_boot_addr(u32 core_id, unsigned long boot_addr) -@@ -307,7 +307,7 @@ +@@ -307,7 +307,7 @@ static int exynos_boot_secondary(unsigned int cpu, struct task_struct *idle) * Set synchronisation state between this boot processor * and the secondary one */ @@ -387,7 +665,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-exynos/platsmp.c linux-4.9.28/arch/arm /* * The secondary processor is waiting to be released from -@@ -334,7 +334,7 @@ +@@ -334,7 +334,7 @@ static int exynos_boot_secondary(unsigned int cpu, struct task_struct *idle) if (timeout == 0) { printk(KERN_ERR "cpu1 power enable failed"); @@ -396,7 +674,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-exynos/platsmp.c linux-4.9.28/arch/arm return -ETIMEDOUT; } } -@@ -380,7 +380,7 @@ +@@ -380,7 +380,7 @@ static int exynos_boot_secondary(unsigned int cpu, struct task_struct *idle) * calibrations, then wait for it to finish */ fail: @@ -405,9 +683,10 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-exynos/platsmp.c linux-4.9.28/arch/arm return pen_release != -1 ? ret : 0; } -diff -Nur linux-4.9.28.orig/arch/arm/mach-hisi/platmcpm.c linux-4.9.28/arch/arm/mach-hisi/platmcpm.c ---- linux-4.9.28.orig/arch/arm/mach-hisi/platmcpm.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/mach-hisi/platmcpm.c 2017-05-19 03:37:25.122174217 +0200 +diff --git a/arch/arm/mach-hisi/platmcpm.c b/arch/arm/mach-hisi/platmcpm.c +index 4b653a8cb75c..b03d5a922cb1 100644 +--- a/arch/arm/mach-hisi/platmcpm.c ++++ b/arch/arm/mach-hisi/platmcpm.c @@ -61,7 +61,7 @@ static void __iomem *sysctrl, *fabric; @@ -417,7 +696,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-hisi/platmcpm.c linux-4.9.28/arch/arm/ static u32 fabric_phys_addr; /* * [0]: bootwrapper physical address -@@ -113,7 +113,7 @@ +@@ -113,7 +113,7 @@ static int hip04_boot_secondary(unsigned int l_cpu, struct task_struct *idle) if (cluster >= HIP04_MAX_CLUSTERS || cpu >= HIP04_MAX_CPUS_PER_CLUSTER) return -EINVAL; @@ -426,7 +705,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-hisi/platmcpm.c linux-4.9.28/arch/arm/ if (hip04_cpu_table[cluster][cpu]) goto out; -@@ -147,7 +147,7 @@ +@@ -147,7 +147,7 @@ static int hip04_boot_secondary(unsigned int l_cpu, struct task_struct *idle) out: hip04_cpu_table[cluster][cpu]++; @@ -435,7 +714,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-hisi/platmcpm.c linux-4.9.28/arch/arm/ return 0; } -@@ -162,11 +162,11 @@ +@@ -162,11 +162,11 @@ static void hip04_cpu_die(unsigned int l_cpu) cpu = MPIDR_AFFINITY_LEVEL(mpidr, 0); cluster = MPIDR_AFFINITY_LEVEL(mpidr, 1); @@ -449,7 +728,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-hisi/platmcpm.c linux-4.9.28/arch/arm/ return; } else if (hip04_cpu_table[cluster][cpu] > 1) { pr_err("Cluster %d CPU%d boots multiple times\n", cluster, cpu); -@@ -174,7 +174,7 @@ +@@ -174,7 +174,7 @@ static void hip04_cpu_die(unsigned int l_cpu) } last_man = hip04_cluster_is_down(cluster); @@ -458,7 +737,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-hisi/platmcpm.c linux-4.9.28/arch/arm/ if (last_man) { /* Since it's Cortex A15, disable L2 prefetching. */ asm volatile( -@@ -203,7 +203,7 @@ +@@ -203,7 +203,7 @@ static int hip04_cpu_kill(unsigned int l_cpu) cpu >= HIP04_MAX_CPUS_PER_CLUSTER); count = TIMEOUT_MSEC / POLL_MSEC; @@ -467,7 +746,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-hisi/platmcpm.c linux-4.9.28/arch/arm/ for (tries = 0; tries < count; tries++) { if (hip04_cpu_table[cluster][cpu]) goto err; -@@ -211,10 +211,10 @@ +@@ -211,10 +211,10 @@ static int hip04_cpu_kill(unsigned int l_cpu) data = readl_relaxed(sysctrl + SC_CPU_RESET_STATUS(cluster)); if (data & CORE_WFI_STATUS(cpu)) break; @@ -480,7 +759,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-hisi/platmcpm.c linux-4.9.28/arch/arm/ } if (tries >= count) goto err; -@@ -231,10 +231,10 @@ +@@ -231,10 +231,10 @@ static int hip04_cpu_kill(unsigned int l_cpu) goto err; if (hip04_cluster_is_down(cluster)) hip04_set_snoop_filter(cluster, 0); @@ -493,10 +772,11 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-hisi/platmcpm.c linux-4.9.28/arch/arm/ return 0; } #endif -diff -Nur linux-4.9.28.orig/arch/arm/mach-omap2/omap-smp.c linux-4.9.28/arch/arm/mach-omap2/omap-smp.c ---- linux-4.9.28.orig/arch/arm/mach-omap2/omap-smp.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/mach-omap2/omap-smp.c 2017-05-19 03:37:25.122174217 +0200 -@@ -64,7 +64,7 @@ +diff --git a/arch/arm/mach-omap2/omap-smp.c b/arch/arm/mach-omap2/omap-smp.c +index b4de3da6dffa..b52893319d75 100644 +--- a/arch/arm/mach-omap2/omap-smp.c ++++ b/arch/arm/mach-omap2/omap-smp.c +@@ -64,7 +64,7 @@ static const struct omap_smp_config omap5_cfg __initconst = { .startup_addr = omap5_secondary_startup, }; @@ -505,7 +785,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-omap2/omap-smp.c linux-4.9.28/arch/arm void __iomem *omap4_get_scu_base(void) { -@@ -131,8 +131,8 @@ +@@ -131,8 +131,8 @@ static void omap4_secondary_init(unsigned int cpu) /* * Synchronise with the boot thread. */ @@ -516,7 +796,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-omap2/omap-smp.c linux-4.9.28/arch/arm } static int omap4_boot_secondary(unsigned int cpu, struct task_struct *idle) -@@ -146,7 +146,7 @@ +@@ -146,7 +146,7 @@ static int omap4_boot_secondary(unsigned int cpu, struct task_struct *idle) * Set synchronisation state between this boot processor * and the secondary one */ @@ -525,7 +805,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-omap2/omap-smp.c linux-4.9.28/arch/arm /* * Update the AuxCoreBoot0 with boot state for secondary core. -@@ -223,7 +223,7 @@ +@@ -223,7 +223,7 @@ static int omap4_boot_secondary(unsigned int cpu, struct task_struct *idle) * Now the secondary core is starting up let it run its * calibrations, then wait for it to finish */ @@ -534,9 +814,10 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-omap2/omap-smp.c linux-4.9.28/arch/arm return 0; } -diff -Nur linux-4.9.28.orig/arch/arm/mach-prima2/platsmp.c linux-4.9.28/arch/arm/mach-prima2/platsmp.c ---- linux-4.9.28.orig/arch/arm/mach-prima2/platsmp.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/mach-prima2/platsmp.c 2017-05-19 03:37:25.122174217 +0200 +diff --git a/arch/arm/mach-prima2/platsmp.c b/arch/arm/mach-prima2/platsmp.c +index 0875b99add18..18b6d98d2581 100644 +--- a/arch/arm/mach-prima2/platsmp.c ++++ b/arch/arm/mach-prima2/platsmp.c @@ -22,7 +22,7 @@ static void __iomem *clk_base; @@ -546,7 +827,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-prima2/platsmp.c linux-4.9.28/arch/arm static void sirfsoc_secondary_init(unsigned int cpu) { -@@ -36,8 +36,8 @@ +@@ -36,8 +36,8 @@ static void sirfsoc_secondary_init(unsigned int cpu) /* * Synchronise with the boot thread. */ @@ -557,7 +838,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-prima2/platsmp.c linux-4.9.28/arch/arm } static const struct of_device_id clk_ids[] = { -@@ -75,7 +75,7 @@ +@@ -75,7 +75,7 @@ static int sirfsoc_boot_secondary(unsigned int cpu, struct task_struct *idle) /* make sure write buffer is drained */ mb(); @@ -566,7 +847,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-prima2/platsmp.c linux-4.9.28/arch/arm /* * The secondary processor is waiting to be released from -@@ -107,7 +107,7 @@ +@@ -107,7 +107,7 @@ static int sirfsoc_boot_secondary(unsigned int cpu, struct task_struct *idle) * now the secondary core is starting up let it run its * calibrations, then wait for it to finish */ @@ -575,9 +856,10 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-prima2/platsmp.c linux-4.9.28/arch/arm return pen_release != -1 ? -ENOSYS : 0; } -diff -Nur linux-4.9.28.orig/arch/arm/mach-qcom/platsmp.c linux-4.9.28/arch/arm/mach-qcom/platsmp.c ---- linux-4.9.28.orig/arch/arm/mach-qcom/platsmp.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/mach-qcom/platsmp.c 2017-05-19 03:37:25.122174217 +0200 +diff --git a/arch/arm/mach-qcom/platsmp.c b/arch/arm/mach-qcom/platsmp.c +index 5494c9e0c909..e8ce157d3548 100644 +--- a/arch/arm/mach-qcom/platsmp.c ++++ b/arch/arm/mach-qcom/platsmp.c @@ -46,7 +46,7 @@ extern void secondary_startup_arm(void); @@ -587,7 +869,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-qcom/platsmp.c linux-4.9.28/arch/arm/m #ifdef CONFIG_HOTPLUG_CPU static void qcom_cpu_die(unsigned int cpu) -@@ -60,8 +60,8 @@ +@@ -60,8 +60,8 @@ static void qcom_secondary_init(unsigned int cpu) /* * Synchronise with the boot thread. */ @@ -598,7 +880,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-qcom/platsmp.c linux-4.9.28/arch/arm/m } static int scss_release_secondary(unsigned int cpu) -@@ -284,7 +284,7 @@ +@@ -284,7 +284,7 @@ static int qcom_boot_secondary(unsigned int cpu, int (*func)(unsigned int)) * set synchronisation state between this boot processor * and the secondary one */ @@ -607,7 +889,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-qcom/platsmp.c linux-4.9.28/arch/arm/m /* * Send the secondary CPU a soft interrupt, thereby causing -@@ -297,7 +297,7 @@ +@@ -297,7 +297,7 @@ static int qcom_boot_secondary(unsigned int cpu, int (*func)(unsigned int)) * now the secondary core is starting up let it run its * calibrations, then wait for it to finish */ @@ -616,10 +898,11 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-qcom/platsmp.c linux-4.9.28/arch/arm/m return ret; } -diff -Nur linux-4.9.28.orig/arch/arm/mach-spear/platsmp.c linux-4.9.28/arch/arm/mach-spear/platsmp.c ---- linux-4.9.28.orig/arch/arm/mach-spear/platsmp.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/mach-spear/platsmp.c 2017-05-19 03:37:25.122174217 +0200 -@@ -32,7 +32,7 @@ +diff --git a/arch/arm/mach-spear/platsmp.c b/arch/arm/mach-spear/platsmp.c +index 8d1e2d551786..7fa56cc78118 100644 +--- a/arch/arm/mach-spear/platsmp.c ++++ b/arch/arm/mach-spear/platsmp.c +@@ -32,7 +32,7 @@ static void write_pen_release(int val) sync_cache_w(&pen_release); } @@ -628,7 +911,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-spear/platsmp.c linux-4.9.28/arch/arm/ static void __iomem *scu_base = IOMEM(VA_SCU_BASE); -@@ -47,8 +47,8 @@ +@@ -47,8 +47,8 @@ static void spear13xx_secondary_init(unsigned int cpu) /* * Synchronise with the boot thread. */ @@ -639,7 +922,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-spear/platsmp.c linux-4.9.28/arch/arm/ } static int spear13xx_boot_secondary(unsigned int cpu, struct task_struct *idle) -@@ -59,7 +59,7 @@ +@@ -59,7 +59,7 @@ static int spear13xx_boot_secondary(unsigned int cpu, struct task_struct *idle) * set synchronisation state between this boot processor * and the secondary one */ @@ -648,7 +931,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-spear/platsmp.c linux-4.9.28/arch/arm/ /* * The secondary processor is waiting to be released from -@@ -84,7 +84,7 @@ +@@ -84,7 +84,7 @@ static int spear13xx_boot_secondary(unsigned int cpu, struct task_struct *idle) * now the secondary core is starting up let it run its * calibrations, then wait for it to finish */ @@ -657,10 +940,11 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-spear/platsmp.c linux-4.9.28/arch/arm/ return pen_release != -1 ? -ENOSYS : 0; } -diff -Nur linux-4.9.28.orig/arch/arm/mach-sti/platsmp.c linux-4.9.28/arch/arm/mach-sti/platsmp.c ---- linux-4.9.28.orig/arch/arm/mach-sti/platsmp.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/mach-sti/platsmp.c 2017-05-19 03:37:25.122174217 +0200 -@@ -35,7 +35,7 @@ +diff --git a/arch/arm/mach-sti/platsmp.c b/arch/arm/mach-sti/platsmp.c +index ea5a2277ee46..b988e081ac79 100644 +--- a/arch/arm/mach-sti/platsmp.c ++++ b/arch/arm/mach-sti/platsmp.c +@@ -35,7 +35,7 @@ static void write_pen_release(int val) sync_cache_w(&pen_release); } @@ -669,7 +953,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-sti/platsmp.c linux-4.9.28/arch/arm/ma static void sti_secondary_init(unsigned int cpu) { -@@ -48,8 +48,8 @@ +@@ -48,8 +48,8 @@ static void sti_secondary_init(unsigned int cpu) /* * Synchronise with the boot thread. */ @@ -680,7 +964,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-sti/platsmp.c linux-4.9.28/arch/arm/ma } static int sti_boot_secondary(unsigned int cpu, struct task_struct *idle) -@@ -60,7 +60,7 @@ +@@ -60,7 +60,7 @@ static int sti_boot_secondary(unsigned int cpu, struct task_struct *idle) * set synchronisation state between this boot processor * and the secondary one */ @@ -689,7 +973,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-sti/platsmp.c linux-4.9.28/arch/arm/ma /* * The secondary processor is waiting to be released from -@@ -91,7 +91,7 @@ +@@ -91,7 +91,7 @@ static int sti_boot_secondary(unsigned int cpu, struct task_struct *idle) * now the secondary core is starting up let it run its * calibrations, then wait for it to finish */ @@ -698,10 +982,11 @@ diff -Nur linux-4.9.28.orig/arch/arm/mach-sti/platsmp.c linux-4.9.28/arch/arm/ma return pen_release != -1 ? -ENOSYS : 0; } -diff -Nur linux-4.9.28.orig/arch/arm/mm/fault.c linux-4.9.28/arch/arm/mm/fault.c ---- linux-4.9.28.orig/arch/arm/mm/fault.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/mm/fault.c 2017-05-19 03:37:25.122174217 +0200 -@@ -430,6 +430,9 @@ +diff --git a/arch/arm/mm/fault.c b/arch/arm/mm/fault.c +index 0122ad1a6027..926b1be48043 100644 +--- a/arch/arm/mm/fault.c ++++ b/arch/arm/mm/fault.c +@@ -430,6 +430,9 @@ do_translation_fault(unsigned long addr, unsigned int fsr, if (addr < TASK_SIZE) return do_page_fault(addr, fsr, regs); @@ -711,7 +996,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mm/fault.c linux-4.9.28/arch/arm/mm/fault.c if (user_mode(regs)) goto bad_area; -@@ -497,6 +500,9 @@ +@@ -497,6 +500,9 @@ do_translation_fault(unsigned long addr, unsigned int fsr, static int do_sect_fault(unsigned long addr, unsigned int fsr, struct pt_regs *regs) { @@ -721,10 +1006,11 @@ diff -Nur linux-4.9.28.orig/arch/arm/mm/fault.c linux-4.9.28/arch/arm/mm/fault.c do_bad_area(addr, fsr, regs); return 0; } -diff -Nur linux-4.9.28.orig/arch/arm/mm/highmem.c linux-4.9.28/arch/arm/mm/highmem.c ---- linux-4.9.28.orig/arch/arm/mm/highmem.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/mm/highmem.c 2017-05-19 03:37:25.126174470 +0200 -@@ -34,6 +34,11 @@ +diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c +index d02f8187b1cc..542692dbd40a 100644 +--- a/arch/arm/mm/highmem.c ++++ b/arch/arm/mm/highmem.c +@@ -34,6 +34,11 @@ static inline pte_t get_fixmap_pte(unsigned long vaddr) return *ptep; } @@ -736,7 +1022,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mm/highmem.c linux-4.9.28/arch/arm/mm/highm void *kmap(struct page *page) { might_sleep(); -@@ -54,12 +59,13 @@ +@@ -54,12 +59,13 @@ EXPORT_SYMBOL(kunmap); void *kmap_atomic(struct page *page) { @@ -751,7 +1037,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mm/highmem.c linux-4.9.28/arch/arm/mm/highm pagefault_disable(); if (!PageHighMem(page)) return page_address(page); -@@ -79,7 +85,7 @@ +@@ -79,7 +85,7 @@ void *kmap_atomic(struct page *page) type = kmap_atomic_idx_push(); @@ -760,7 +1046,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mm/highmem.c linux-4.9.28/arch/arm/mm/highm vaddr = __fix_to_virt(idx); #ifdef CONFIG_DEBUG_HIGHMEM /* -@@ -93,7 +99,10 @@ +@@ -93,7 +99,10 @@ void *kmap_atomic(struct page *page) * in place, so the contained TLB flush ensures the TLB is updated * with the new mapping. */ @@ -772,7 +1058,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/mm/highmem.c linux-4.9.28/arch/arm/mm/highm return (void *)vaddr; } -@@ -106,44 +115,75 @@ +@@ -106,44 +115,75 @@ void __kunmap_atomic(void *kvaddr) if (kvaddr >= (void *)FIXADDR_START) { type = kmap_atomic_idx(); @@ -854,10 +1140,11 @@ diff -Nur linux-4.9.28.orig/arch/arm/mm/highmem.c linux-4.9.28/arch/arm/mm/highm + } +} +#endif -diff -Nur linux-4.9.28.orig/arch/arm/plat-versatile/platsmp.c linux-4.9.28/arch/arm/plat-versatile/platsmp.c ---- linux-4.9.28.orig/arch/arm/plat-versatile/platsmp.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm/plat-versatile/platsmp.c 2017-05-19 03:37:25.126174470 +0200 -@@ -32,7 +32,7 @@ +diff --git a/arch/arm/plat-versatile/platsmp.c b/arch/arm/plat-versatile/platsmp.c +index c2366510187a..6b60f582b738 100644 +--- a/arch/arm/plat-versatile/platsmp.c ++++ b/arch/arm/plat-versatile/platsmp.c +@@ -32,7 +32,7 @@ static void write_pen_release(int val) sync_cache_w(&pen_release); } @@ -866,7 +1153,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/plat-versatile/platsmp.c linux-4.9.28/arch/ void versatile_secondary_init(unsigned int cpu) { -@@ -45,8 +45,8 @@ +@@ -45,8 +45,8 @@ void versatile_secondary_init(unsigned int cpu) /* * Synchronise with the boot thread. */ @@ -877,7 +1164,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/plat-versatile/platsmp.c linux-4.9.28/arch/ } int versatile_boot_secondary(unsigned int cpu, struct task_struct *idle) -@@ -57,7 +57,7 @@ +@@ -57,7 +57,7 @@ int versatile_boot_secondary(unsigned int cpu, struct task_struct *idle) * Set synchronisation state between this boot processor * and the secondary one */ @@ -886,7 +1173,7 @@ diff -Nur linux-4.9.28.orig/arch/arm/plat-versatile/platsmp.c linux-4.9.28/arch/ /* * This is really belt and braces; we hold unintended secondary -@@ -87,7 +87,7 @@ +@@ -87,7 +87,7 @@ int versatile_boot_secondary(unsigned int cpu, struct task_struct *idle) * now the secondary core is starting up let it run its * calibrations, then wait for it to finish */ @@ -895,10 +1182,32 @@ diff -Nur linux-4.9.28.orig/arch/arm/plat-versatile/platsmp.c linux-4.9.28/arch/ return pen_release != -1 ? -ENOSYS : 0; } -diff -Nur linux-4.9.28.orig/arch/arm64/include/asm/thread_info.h linux-4.9.28/arch/arm64/include/asm/thread_info.h ---- linux-4.9.28.orig/arch/arm64/include/asm/thread_info.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm64/include/asm/thread_info.h 2017-05-19 03:37:25.126174470 +0200 -@@ -49,6 +49,7 @@ +diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig +index cf57a7799a0f..78d1b49fbed5 100644 +--- a/arch/arm64/Kconfig ++++ b/arch/arm64/Kconfig +@@ -91,6 +91,7 @@ config ARM64 + select HAVE_PERF_EVENTS + select HAVE_PERF_REGS + select HAVE_PERF_USER_STACK_DUMP ++ select HAVE_PREEMPT_LAZY + select HAVE_REGS_AND_STACK_ACCESS_API + select HAVE_RCU_TABLE_FREE + select HAVE_SYSCALL_TRACEPOINTS +@@ -704,7 +705,7 @@ config XEN_DOM0 + + config XEN + bool "Xen guest support on ARM64" +- depends on ARM64 && OF ++ depends on ARM64 && OF && !PREEMPT_RT_FULL + select SWIOTLB_XEN + select PARAVIRT + help +diff --git a/arch/arm64/include/asm/thread_info.h b/arch/arm64/include/asm/thread_info.h +index e9ea5a6bd449..6c500ad63c6a 100644 +--- a/arch/arm64/include/asm/thread_info.h ++++ b/arch/arm64/include/asm/thread_info.h +@@ -49,6 +49,7 @@ struct thread_info { mm_segment_t addr_limit; /* address limit */ struct task_struct *task; /* main task structure */ int preempt_count; /* 0 => preemptable, <0 => bug */ @@ -906,7 +1215,7 @@ diff -Nur linux-4.9.28.orig/arch/arm64/include/asm/thread_info.h linux-4.9.28/ar int cpu; /* cpu */ }; -@@ -112,6 +113,7 @@ +@@ -112,6 +113,7 @@ static inline struct thread_info *current_thread_info(void) #define TIF_NEED_RESCHED 1 #define TIF_NOTIFY_RESUME 2 /* callback before returning to user */ #define TIF_FOREIGN_FPSTATE 3 /* CPU's FP state is not current's */ @@ -914,7 +1223,7 @@ diff -Nur linux-4.9.28.orig/arch/arm64/include/asm/thread_info.h linux-4.9.28/ar #define TIF_NOHZ 7 #define TIF_SYSCALL_TRACE 8 #define TIF_SYSCALL_AUDIT 9 -@@ -127,6 +129,7 @@ +@@ -127,6 +129,7 @@ static inline struct thread_info *current_thread_info(void) #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) #define _TIF_FOREIGN_FPSTATE (1 << TIF_FOREIGN_FPSTATE) @@ -922,7 +1231,7 @@ diff -Nur linux-4.9.28.orig/arch/arm64/include/asm/thread_info.h linux-4.9.28/ar #define _TIF_NOHZ (1 << TIF_NOHZ) #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) -@@ -135,7 +138,9 @@ +@@ -135,7 +138,9 @@ static inline struct thread_info *current_thread_info(void) #define _TIF_32BIT (1 << TIF_32BIT) #define _TIF_WORK_MASK (_TIF_NEED_RESCHED | _TIF_SIGPENDING | \ @@ -933,30 +1242,11 @@ diff -Nur linux-4.9.28.orig/arch/arm64/include/asm/thread_info.h linux-4.9.28/ar #define _TIF_SYSCALL_WORK (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | \ _TIF_SYSCALL_TRACEPOINT | _TIF_SECCOMP | \ -diff -Nur linux-4.9.28.orig/arch/arm64/Kconfig linux-4.9.28/arch/arm64/Kconfig ---- linux-4.9.28.orig/arch/arm64/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm64/Kconfig 2017-05-19 03:37:25.126174470 +0200 -@@ -91,6 +91,7 @@ - select HAVE_PERF_EVENTS - select HAVE_PERF_REGS - select HAVE_PERF_USER_STACK_DUMP -+ select HAVE_PREEMPT_LAZY - select HAVE_REGS_AND_STACK_ACCESS_API - select HAVE_RCU_TABLE_FREE - select HAVE_SYSCALL_TRACEPOINTS -@@ -704,7 +705,7 @@ - - config XEN - bool "Xen guest support on ARM64" -- depends on ARM64 && OF -+ depends on ARM64 && OF && !PREEMPT_RT_FULL - select SWIOTLB_XEN - select PARAVIRT - help -diff -Nur linux-4.9.28.orig/arch/arm64/kernel/asm-offsets.c linux-4.9.28/arch/arm64/kernel/asm-offsets.c ---- linux-4.9.28.orig/arch/arm64/kernel/asm-offsets.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm64/kernel/asm-offsets.c 2017-05-19 03:37:25.126174470 +0200 -@@ -38,6 +38,7 @@ +diff --git a/arch/arm64/kernel/asm-offsets.c b/arch/arm64/kernel/asm-offsets.c +index c58ddf8c4062..a8f2f7c1fe12 100644 +--- a/arch/arm64/kernel/asm-offsets.c ++++ b/arch/arm64/kernel/asm-offsets.c +@@ -38,6 +38,7 @@ int main(void) BLANK(); DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); @@ -964,10 +1254,11 @@ diff -Nur linux-4.9.28.orig/arch/arm64/kernel/asm-offsets.c linux-4.9.28/arch/ar DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit)); DEFINE(TI_TASK, offsetof(struct thread_info, task)); DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); -diff -Nur linux-4.9.28.orig/arch/arm64/kernel/entry.S linux-4.9.28/arch/arm64/kernel/entry.S ---- linux-4.9.28.orig/arch/arm64/kernel/entry.S 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm64/kernel/entry.S 2017-05-19 03:37:25.126174470 +0200 -@@ -428,11 +428,16 @@ +diff --git a/arch/arm64/kernel/entry.S b/arch/arm64/kernel/entry.S +index 79b0fe24d5b7..f3c959ade308 100644 +--- a/arch/arm64/kernel/entry.S ++++ b/arch/arm64/kernel/entry.S +@@ -428,11 +428,16 @@ ENDPROC(el1_sync) #ifdef CONFIG_PREEMPT ldr w24, [tsk, #TI_PREEMPT] // get preempt count @@ -987,7 +1278,7 @@ diff -Nur linux-4.9.28.orig/arch/arm64/kernel/entry.S linux-4.9.28/arch/arm64/ke #endif #ifdef CONFIG_TRACE_IRQFLAGS bl trace_hardirqs_on -@@ -446,6 +451,7 @@ +@@ -446,6 +451,7 @@ ENDPROC(el1_irq) 1: bl preempt_schedule_irq // irq en/disable is done inside ldr x0, [tsk, #TI_FLAGS] // get new tasks TI_FLAGS tbnz x0, #TIF_NEED_RESCHED, 1b // needs rescheduling? @@ -995,10 +1286,11 @@ diff -Nur linux-4.9.28.orig/arch/arm64/kernel/entry.S linux-4.9.28/arch/arm64/ke ret x24 #endif -diff -Nur linux-4.9.28.orig/arch/arm64/kernel/signal.c linux-4.9.28/arch/arm64/kernel/signal.c ---- linux-4.9.28.orig/arch/arm64/kernel/signal.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/arm64/kernel/signal.c 2017-05-19 03:37:25.126174470 +0200 -@@ -409,7 +409,7 @@ +diff --git a/arch/arm64/kernel/signal.c b/arch/arm64/kernel/signal.c +index 404dd67080b9..639dc6d12e72 100644 +--- a/arch/arm64/kernel/signal.c ++++ b/arch/arm64/kernel/signal.c +@@ -409,7 +409,7 @@ asmlinkage void do_notify_resume(struct pt_regs *regs, */ trace_hardirqs_off(); do { @@ -1007,29 +1299,11 @@ diff -Nur linux-4.9.28.orig/arch/arm64/kernel/signal.c linux-4.9.28/arch/arm64/k schedule(); } else { local_irq_enable(); -diff -Nur linux-4.9.28.orig/arch/Kconfig linux-4.9.28/arch/Kconfig ---- linux-4.9.28.orig/arch/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/Kconfig 2017-05-19 03:37:25.122174217 +0200 -@@ -9,6 +9,7 @@ - tristate "OProfile system profiling" - depends on PROFILING - depends on HAVE_OPROFILE -+ depends on !PREEMPT_RT_FULL - select RING_BUFFER - select RING_BUFFER_ALLOW_SWAP - help -@@ -52,6 +53,7 @@ - config JUMP_LABEL - bool "Optimize very unlikely/likely branches" - depends on HAVE_ARCH_JUMP_LABEL -+ depends on (!INTERRUPT_OFF_HIST && !PREEMPT_OFF_HIST && !WAKEUP_LATENCY_HIST && !MISSED_TIMER_OFFSETS_HIST) - help - This option enables a transparent branch optimization that - makes certain almost-always-true or almost-always-false branch -diff -Nur linux-4.9.28.orig/arch/mips/Kconfig linux-4.9.28/arch/mips/Kconfig ---- linux-4.9.28.orig/arch/mips/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/mips/Kconfig 2017-05-19 03:37:25.126174470 +0200 -@@ -2515,7 +2515,7 @@ +diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig +index 5e844f68e847..dc613cc10f54 100644 +--- a/arch/mips/Kconfig ++++ b/arch/mips/Kconfig +@@ -2516,7 +2516,7 @@ config MIPS_ASID_BITS_VARIABLE # config HIGHMEM bool "High Memory Support" @@ -1038,10 +1312,45 @@ diff -Nur linux-4.9.28.orig/arch/mips/Kconfig linux-4.9.28/arch/mips/Kconfig config CPU_SUPPORTS_HIGHMEM bool -diff -Nur linux-4.9.28.orig/arch/powerpc/include/asm/thread_info.h linux-4.9.28/arch/powerpc/include/asm/thread_info.h ---- linux-4.9.28.orig/arch/powerpc/include/asm/thread_info.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/powerpc/include/asm/thread_info.h 2017-05-19 03:37:25.126174470 +0200 -@@ -43,6 +43,8 @@ +diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig +index 8f01f21e78f1..619485b777d2 100644 +--- a/arch/powerpc/Kconfig ++++ b/arch/powerpc/Kconfig +@@ -52,10 +52,11 @@ config LOCKDEP_SUPPORT + + config RWSEM_GENERIC_SPINLOCK + bool ++ default y if PREEMPT_RT_FULL + + config RWSEM_XCHGADD_ALGORITHM + bool +- default y ++ default y if !PREEMPT_RT_FULL + + config GENERIC_LOCKBREAK + bool +@@ -134,6 +135,7 @@ config PPC + select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST + select GENERIC_STRNCPY_FROM_USER + select GENERIC_STRNLEN_USER ++ select HAVE_PREEMPT_LAZY + select HAVE_MOD_ARCH_SPECIFIC + select MODULES_USE_ELF_RELA + select CLONE_BACKWARDS +@@ -321,7 +323,7 @@ menu "Kernel options" + + config HIGHMEM + bool "High memory support" +- depends on PPC32 ++ depends on PPC32 && !PREEMPT_RT_FULL + + source kernel/Kconfig.hz + source kernel/Kconfig.preempt +diff --git a/arch/powerpc/include/asm/thread_info.h b/arch/powerpc/include/asm/thread_info.h +index 87e4b2d8dcd4..981e501a4359 100644 +--- a/arch/powerpc/include/asm/thread_info.h ++++ b/arch/powerpc/include/asm/thread_info.h +@@ -43,6 +43,8 @@ struct thread_info { int cpu; /* cpu we're on */ int preempt_count; /* 0 => preemptable, <0 => BUG */ @@ -1050,7 +1359,7 @@ diff -Nur linux-4.9.28.orig/arch/powerpc/include/asm/thread_info.h linux-4.9.28/ unsigned long local_flags; /* private flags for thread */ #ifdef CONFIG_LIVEPATCH unsigned long *livepatch_sp; -@@ -88,8 +90,7 @@ +@@ -88,8 +90,7 @@ static inline struct thread_info *current_thread_info(void) #define TIF_SYSCALL_TRACE 0 /* syscall trace active */ #define TIF_SIGPENDING 1 /* signal pending */ #define TIF_NEED_RESCHED 2 /* rescheduling necessary */ @@ -1060,7 +1369,7 @@ diff -Nur linux-4.9.28.orig/arch/powerpc/include/asm/thread_info.h linux-4.9.28/ #define TIF_32BIT 4 /* 32 bit binary */ #define TIF_RESTORE_TM 5 /* need to restore TM FP/VEC/VSX */ #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ -@@ -107,6 +108,8 @@ +@@ -107,6 +108,8 @@ static inline struct thread_info *current_thread_info(void) #if defined(CONFIG_PPC64) #define TIF_ELF2ABI 18 /* function descriptors must die! */ #endif @@ -1069,7 +1378,7 @@ diff -Nur linux-4.9.28.orig/arch/powerpc/include/asm/thread_info.h linux-4.9.28/ /* as above, but as bit values */ #define _TIF_SYSCALL_TRACE (1<flags) set_bits(irqtp->flags, &curtp->flags); } @@ -1242,9 +1524,10 @@ diff -Nur linux-4.9.28.orig/arch/powerpc/kernel/irq.c linux-4.9.28/arch/powerpc/ irq_hw_number_t virq_to_hw(unsigned int virq) { -diff -Nur linux-4.9.28.orig/arch/powerpc/kernel/misc_32.S linux-4.9.28/arch/powerpc/kernel/misc_32.S ---- linux-4.9.28.orig/arch/powerpc/kernel/misc_32.S 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/powerpc/kernel/misc_32.S 2017-05-19 03:37:25.126174470 +0200 +diff --git a/arch/powerpc/kernel/misc_32.S b/arch/powerpc/kernel/misc_32.S +index 030d72df5dd5..b471a709e100 100644 +--- a/arch/powerpc/kernel/misc_32.S ++++ b/arch/powerpc/kernel/misc_32.S @@ -41,6 +41,7 @@ * We store the saved ksp_limit in the unused part * of the STACK_FRAME_OVERHEAD @@ -1253,7 +1536,7 @@ diff -Nur linux-4.9.28.orig/arch/powerpc/kernel/misc_32.S linux-4.9.28/arch/powe _GLOBAL(call_do_softirq) mflr r0 stw r0,4(r1) -@@ -57,6 +58,7 @@ +@@ -57,6 +58,7 @@ _GLOBAL(call_do_softirq) stw r10,THREAD+KSP_LIMIT(r2) mtlr r0 blr @@ -1261,9 +1544,10 @@ diff -Nur linux-4.9.28.orig/arch/powerpc/kernel/misc_32.S linux-4.9.28/arch/powe /* * void call_do_irq(struct pt_regs *regs, struct thread_info *irqtp); -diff -Nur linux-4.9.28.orig/arch/powerpc/kernel/misc_64.S linux-4.9.28/arch/powerpc/kernel/misc_64.S ---- linux-4.9.28.orig/arch/powerpc/kernel/misc_64.S 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/powerpc/kernel/misc_64.S 2017-05-19 03:37:25.126174470 +0200 +diff --git a/arch/powerpc/kernel/misc_64.S b/arch/powerpc/kernel/misc_64.S +index 4cefe6888b18..cb2ee4be999a 100644 +--- a/arch/powerpc/kernel/misc_64.S ++++ b/arch/powerpc/kernel/misc_64.S @@ -31,6 +31,7 @@ .text @@ -1272,7 +1556,7 @@ diff -Nur linux-4.9.28.orig/arch/powerpc/kernel/misc_64.S linux-4.9.28/arch/powe _GLOBAL(call_do_softirq) mflr r0 std r0,16(r1) -@@ -41,6 +42,7 @@ +@@ -41,6 +42,7 @@ _GLOBAL(call_do_softirq) ld r0,16(r1) mtlr r0 blr @@ -1280,10 +1564,11 @@ diff -Nur linux-4.9.28.orig/arch/powerpc/kernel/misc_64.S linux-4.9.28/arch/powe _GLOBAL(call_do_irq) mflr r0 -diff -Nur linux-4.9.28.orig/arch/powerpc/kvm/Kconfig linux-4.9.28/arch/powerpc/kvm/Kconfig ---- linux-4.9.28.orig/arch/powerpc/kvm/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/powerpc/kvm/Kconfig 2017-05-19 03:37:25.126174470 +0200 -@@ -175,6 +175,7 @@ +diff --git a/arch/powerpc/kvm/Kconfig b/arch/powerpc/kvm/Kconfig +index 029be26b5a17..9528089ea142 100644 +--- a/arch/powerpc/kvm/Kconfig ++++ b/arch/powerpc/kvm/Kconfig +@@ -175,6 +175,7 @@ config KVM_E500MC config KVM_MPIC bool "KVM in-kernel MPIC emulation" depends on KVM && E500 @@ -1291,10 +1576,11 @@ diff -Nur linux-4.9.28.orig/arch/powerpc/kvm/Kconfig linux-4.9.28/arch/powerpc/k select HAVE_KVM_IRQCHIP select HAVE_KVM_IRQFD select HAVE_KVM_IRQ_ROUTING -diff -Nur linux-4.9.28.orig/arch/powerpc/platforms/ps3/device-init.c linux-4.9.28/arch/powerpc/platforms/ps3/device-init.c ---- linux-4.9.28.orig/arch/powerpc/platforms/ps3/device-init.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/powerpc/platforms/ps3/device-init.c 2017-05-19 03:37:25.126174470 +0200 -@@ -752,7 +752,7 @@ +diff --git a/arch/powerpc/platforms/ps3/device-init.c b/arch/powerpc/platforms/ps3/device-init.c +index e48462447ff0..2670cee66064 100644 +--- a/arch/powerpc/platforms/ps3/device-init.c ++++ b/arch/powerpc/platforms/ps3/device-init.c +@@ -752,7 +752,7 @@ static int ps3_notification_read_write(struct ps3_notification_device *dev, } pr_debug("%s:%u: notification %s issued\n", __func__, __LINE__, op); @@ -1303,10 +1589,11 @@ diff -Nur linux-4.9.28.orig/arch/powerpc/platforms/ps3/device-init.c linux-4.9.2 dev->done.done || kthread_should_stop()); if (kthread_should_stop()) res = -EINTR; -diff -Nur linux-4.9.28.orig/arch/sh/kernel/irq.c linux-4.9.28/arch/sh/kernel/irq.c ---- linux-4.9.28.orig/arch/sh/kernel/irq.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/sh/kernel/irq.c 2017-05-19 03:37:25.126174470 +0200 -@@ -147,6 +147,7 @@ +diff --git a/arch/sh/kernel/irq.c b/arch/sh/kernel/irq.c +index 6c0378c0b8b5..abd58b4dff97 100644 +--- a/arch/sh/kernel/irq.c ++++ b/arch/sh/kernel/irq.c +@@ -147,6 +147,7 @@ void irq_ctx_exit(int cpu) hardirq_ctx[cpu] = NULL; } @@ -1314,7 +1601,7 @@ diff -Nur linux-4.9.28.orig/arch/sh/kernel/irq.c linux-4.9.28/arch/sh/kernel/irq void do_softirq_own_stack(void) { struct thread_info *curctx; -@@ -174,6 +175,7 @@ +@@ -174,6 +175,7 @@ void do_softirq_own_stack(void) "r5", "r6", "r7", "r8", "r9", "r15", "t", "pr" ); } @@ -1322,10 +1609,11 @@ diff -Nur linux-4.9.28.orig/arch/sh/kernel/irq.c linux-4.9.28/arch/sh/kernel/irq #else static inline void handle_one_irq(unsigned int irq) { -diff -Nur linux-4.9.28.orig/arch/sparc/Kconfig linux-4.9.28/arch/sparc/Kconfig ---- linux-4.9.28.orig/arch/sparc/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/sparc/Kconfig 2017-05-19 03:37:25.126174470 +0200 -@@ -194,12 +194,10 @@ +diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig +index 165ecdd24d22..b68a464a22be 100644 +--- a/arch/sparc/Kconfig ++++ b/arch/sparc/Kconfig +@@ -194,12 +194,10 @@ config NR_CPUS source kernel/Kconfig.hz config RWSEM_GENERIC_SPINLOCK @@ -1340,10 +1628,11 @@ diff -Nur linux-4.9.28.orig/arch/sparc/Kconfig linux-4.9.28/arch/sparc/Kconfig config GENERIC_HWEIGHT bool -diff -Nur linux-4.9.28.orig/arch/sparc/kernel/irq_64.c linux-4.9.28/arch/sparc/kernel/irq_64.c ---- linux-4.9.28.orig/arch/sparc/kernel/irq_64.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/sparc/kernel/irq_64.c 2017-05-19 03:37:25.126174470 +0200 -@@ -854,6 +854,7 @@ +diff --git a/arch/sparc/kernel/irq_64.c b/arch/sparc/kernel/irq_64.c +index 34a7930b76ef..773740521008 100644 +--- a/arch/sparc/kernel/irq_64.c ++++ b/arch/sparc/kernel/irq_64.c +@@ -854,6 +854,7 @@ void __irq_entry handler_irq(int pil, struct pt_regs *regs) set_irq_regs(old_regs); } @@ -1351,7 +1640,7 @@ diff -Nur linux-4.9.28.orig/arch/sparc/kernel/irq_64.c linux-4.9.28/arch/sparc/k void do_softirq_own_stack(void) { void *orig_sp, *sp = softirq_stack[smp_processor_id()]; -@@ -868,6 +869,7 @@ +@@ -868,6 +869,7 @@ void do_softirq_own_stack(void) __asm__ __volatile__("mov %0, %%sp" : : "r" (orig_sp)); } @@ -1359,10 +1648,45 @@ diff -Nur linux-4.9.28.orig/arch/sparc/kernel/irq_64.c linux-4.9.28/arch/sparc/k #ifdef CONFIG_HOTPLUG_CPU void fixup_irqs(void) -diff -Nur linux-4.9.28.orig/arch/x86/crypto/aesni-intel_glue.c linux-4.9.28/arch/x86/crypto/aesni-intel_glue.c ---- linux-4.9.28.orig/arch/x86/crypto/aesni-intel_glue.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/crypto/aesni-intel_glue.c 2017-05-19 03:37:25.130174719 +0200 -@@ -372,14 +372,14 @@ +diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig +index bada636d1065..f8a995c90c01 100644 +--- a/arch/x86/Kconfig ++++ b/arch/x86/Kconfig +@@ -17,6 +17,7 @@ config X86_64 + ### Arch settings + config X86 + def_bool y ++ select HAVE_PREEMPT_LAZY + select ACPI_LEGACY_TABLES_LOOKUP if ACPI + select ACPI_SYSTEM_POWER_STATES_SUPPORT if ACPI + select ANON_INODES +@@ -232,8 +233,11 @@ config ARCH_MAY_HAVE_PC_FDC + def_bool y + depends on ISA_DMA_API + ++config RWSEM_GENERIC_SPINLOCK ++ def_bool PREEMPT_RT_FULL ++ + config RWSEM_XCHGADD_ALGORITHM +- def_bool y ++ def_bool !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL + + config GENERIC_CALIBRATE_DELAY + def_bool y +@@ -897,7 +901,7 @@ config IOMMU_HELPER + config MAXSMP + bool "Enable Maximum number of SMP Processors and NUMA Nodes" + depends on X86_64 && SMP && DEBUG_KERNEL +- select CPUMASK_OFFSTACK ++ select CPUMASK_OFFSTACK if !PREEMPT_RT_FULL + ---help--- + Enable maximum number of CPUS and NUMA Nodes for this architecture. + If unsure, say N. +diff --git a/arch/x86/crypto/aesni-intel_glue.c b/arch/x86/crypto/aesni-intel_glue.c +index aa8b0672f87a..2429414bfc71 100644 +--- a/arch/x86/crypto/aesni-intel_glue.c ++++ b/arch/x86/crypto/aesni-intel_glue.c +@@ -372,14 +372,14 @@ static int ecb_encrypt(struct blkcipher_desc *desc, err = blkcipher_walk_virt(desc, &walk); desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; @@ -1380,7 +1704,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/aesni-intel_glue.c linux-4.9.28/arch return err; } -@@ -396,14 +396,14 @@ +@@ -396,14 +396,14 @@ static int ecb_decrypt(struct blkcipher_desc *desc, err = blkcipher_walk_virt(desc, &walk); desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; @@ -1397,7 +1721,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/aesni-intel_glue.c linux-4.9.28/arch return err; } -@@ -420,14 +420,14 @@ +@@ -420,14 +420,14 @@ static int cbc_encrypt(struct blkcipher_desc *desc, err = blkcipher_walk_virt(desc, &walk); desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; @@ -1414,7 +1738,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/aesni-intel_glue.c linux-4.9.28/arch return err; } -@@ -444,14 +444,14 @@ +@@ -444,14 +444,14 @@ static int cbc_decrypt(struct blkcipher_desc *desc, err = blkcipher_walk_virt(desc, &walk); desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; @@ -1431,7 +1755,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/aesni-intel_glue.c linux-4.9.28/arch return err; } -@@ -503,18 +503,20 @@ +@@ -503,18 +503,20 @@ static int ctr_crypt(struct blkcipher_desc *desc, err = blkcipher_walk_virt_block(desc, &walk, AES_BLOCK_SIZE); desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; @@ -1454,10 +1778,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/aesni-intel_glue.c linux-4.9.28/arch return err; } -diff -Nur linux-4.9.28.orig/arch/x86/crypto/cast5_avx_glue.c linux-4.9.28/arch/x86/crypto/cast5_avx_glue.c ---- linux-4.9.28.orig/arch/x86/crypto/cast5_avx_glue.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/crypto/cast5_avx_glue.c 2017-05-19 03:37:25.130174719 +0200 -@@ -59,7 +59,7 @@ +diff --git a/arch/x86/crypto/cast5_avx_glue.c b/arch/x86/crypto/cast5_avx_glue.c +index 8648158f3916..d7699130ee36 100644 +--- a/arch/x86/crypto/cast5_avx_glue.c ++++ b/arch/x86/crypto/cast5_avx_glue.c +@@ -59,7 +59,7 @@ static inline void cast5_fpu_end(bool fpu_enabled) static int ecb_crypt(struct blkcipher_desc *desc, struct blkcipher_walk *walk, bool enc) { @@ -1466,7 +1791,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/cast5_avx_glue.c linux-4.9.28/arch/x struct cast5_ctx *ctx = crypto_blkcipher_ctx(desc->tfm); const unsigned int bsize = CAST5_BLOCK_SIZE; unsigned int nbytes; -@@ -75,7 +75,7 @@ +@@ -75,7 +75,7 @@ static int ecb_crypt(struct blkcipher_desc *desc, struct blkcipher_walk *walk, u8 *wsrc = walk->src.virt.addr; u8 *wdst = walk->dst.virt.addr; @@ -1475,7 +1800,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/cast5_avx_glue.c linux-4.9.28/arch/x /* Process multi-block batch */ if (nbytes >= bsize * CAST5_PARALLEL_BLOCKS) { -@@ -103,10 +103,9 @@ +@@ -103,10 +103,9 @@ static int ecb_crypt(struct blkcipher_desc *desc, struct blkcipher_walk *walk, } while (nbytes >= bsize); done: @@ -1487,7 +1812,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/cast5_avx_glue.c linux-4.9.28/arch/x return err; } -@@ -227,7 +226,7 @@ +@@ -227,7 +226,7 @@ static unsigned int __cbc_decrypt(struct blkcipher_desc *desc, static int cbc_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst, struct scatterlist *src, unsigned int nbytes) { @@ -1496,7 +1821,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/cast5_avx_glue.c linux-4.9.28/arch/x struct blkcipher_walk walk; int err; -@@ -236,12 +235,11 @@ +@@ -236,12 +235,11 @@ static int cbc_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst, desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; while ((nbytes = walk.nbytes)) { @@ -1511,7 +1836,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/cast5_avx_glue.c linux-4.9.28/arch/x return err; } -@@ -311,7 +309,7 @@ +@@ -311,7 +309,7 @@ static unsigned int __ctr_crypt(struct blkcipher_desc *desc, static int ctr_crypt(struct blkcipher_desc *desc, struct scatterlist *dst, struct scatterlist *src, unsigned int nbytes) { @@ -1520,7 +1845,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/cast5_avx_glue.c linux-4.9.28/arch/x struct blkcipher_walk walk; int err; -@@ -320,13 +318,12 @@ +@@ -320,13 +318,12 @@ static int ctr_crypt(struct blkcipher_desc *desc, struct scatterlist *dst, desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; while ((nbytes = walk.nbytes) >= CAST5_BLOCK_SIZE) { @@ -1536,10 +1861,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/cast5_avx_glue.c linux-4.9.28/arch/x if (walk.nbytes) { ctr_crypt_final(desc, &walk); err = blkcipher_walk_done(desc, &walk, 0); -diff -Nur linux-4.9.28.orig/arch/x86/crypto/glue_helper.c linux-4.9.28/arch/x86/crypto/glue_helper.c ---- linux-4.9.28.orig/arch/x86/crypto/glue_helper.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/crypto/glue_helper.c 2017-05-19 03:37:25.130174719 +0200 -@@ -39,7 +39,7 @@ +diff --git a/arch/x86/crypto/glue_helper.c b/arch/x86/crypto/glue_helper.c +index 6a85598931b5..3a506ce7ed93 100644 +--- a/arch/x86/crypto/glue_helper.c ++++ b/arch/x86/crypto/glue_helper.c +@@ -39,7 +39,7 @@ static int __glue_ecb_crypt_128bit(const struct common_glue_ctx *gctx, void *ctx = crypto_blkcipher_ctx(desc->tfm); const unsigned int bsize = 128 / 8; unsigned int nbytes, i, func_bytes; @@ -1548,7 +1874,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/glue_helper.c linux-4.9.28/arch/x86/ int err; err = blkcipher_walk_virt(desc, walk); -@@ -49,7 +49,7 @@ +@@ -49,7 +49,7 @@ static int __glue_ecb_crypt_128bit(const struct common_glue_ctx *gctx, u8 *wdst = walk->dst.virt.addr; fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, @@ -1557,7 +1883,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/glue_helper.c linux-4.9.28/arch/x86/ for (i = 0; i < gctx->num_funcs; i++) { func_bytes = bsize * gctx->funcs[i].num_blocks; -@@ -71,10 +71,10 @@ +@@ -71,10 +71,10 @@ static int __glue_ecb_crypt_128bit(const struct common_glue_ctx *gctx, } done: @@ -1569,7 +1895,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/glue_helper.c linux-4.9.28/arch/x86/ return err; } -@@ -194,7 +194,7 @@ +@@ -194,7 +194,7 @@ int glue_cbc_decrypt_128bit(const struct common_glue_ctx *gctx, struct scatterlist *src, unsigned int nbytes) { const unsigned int bsize = 128 / 8; @@ -1578,7 +1904,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/glue_helper.c linux-4.9.28/arch/x86/ struct blkcipher_walk walk; int err; -@@ -203,12 +203,12 @@ +@@ -203,12 +203,12 @@ int glue_cbc_decrypt_128bit(const struct common_glue_ctx *gctx, while ((nbytes = walk.nbytes)) { fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, @@ -1593,7 +1919,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/glue_helper.c linux-4.9.28/arch/x86/ return err; } EXPORT_SYMBOL_GPL(glue_cbc_decrypt_128bit); -@@ -277,7 +277,7 @@ +@@ -277,7 +277,7 @@ int glue_ctr_crypt_128bit(const struct common_glue_ctx *gctx, struct scatterlist *src, unsigned int nbytes) { const unsigned int bsize = 128 / 8; @@ -1602,7 +1928,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/glue_helper.c linux-4.9.28/arch/x86/ struct blkcipher_walk walk; int err; -@@ -286,13 +286,12 @@ +@@ -286,13 +286,12 @@ int glue_ctr_crypt_128bit(const struct common_glue_ctx *gctx, while ((nbytes = walk.nbytes) >= bsize) { fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, @@ -1618,7 +1944,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/glue_helper.c linux-4.9.28/arch/x86/ if (walk.nbytes) { glue_ctr_crypt_final_128bit( gctx->funcs[gctx->num_funcs - 1].fn_u.ctr, desc, &walk); -@@ -347,7 +346,7 @@ +@@ -347,7 +346,7 @@ int glue_xts_crypt_128bit(const struct common_glue_ctx *gctx, void *tweak_ctx, void *crypt_ctx) { const unsigned int bsize = 128 / 8; @@ -1627,7 +1953,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/glue_helper.c linux-4.9.28/arch/x86/ struct blkcipher_walk walk; int err; -@@ -360,21 +359,21 @@ +@@ -360,21 +359,21 @@ int glue_xts_crypt_128bit(const struct common_glue_ctx *gctx, /* set minimum length to bsize, for tweak_fn */ fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, @@ -1654,10 +1980,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/crypto/glue_helper.c linux-4.9.28/arch/x86/ return err; } EXPORT_SYMBOL_GPL(glue_xts_crypt_128bit); -diff -Nur linux-4.9.28.orig/arch/x86/entry/common.c linux-4.9.28/arch/x86/entry/common.c ---- linux-4.9.28.orig/arch/x86/entry/common.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/entry/common.c 2017-05-19 03:37:25.130174719 +0200 -@@ -129,7 +129,7 @@ +diff --git a/arch/x86/entry/common.c b/arch/x86/entry/common.c +index bdd9cc59d20f..56d01a339ba4 100644 +--- a/arch/x86/entry/common.c ++++ b/arch/x86/entry/common.c +@@ -129,7 +129,7 @@ static long syscall_trace_enter(struct pt_regs *regs) #define EXIT_TO_USERMODE_LOOP_FLAGS \ (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME | _TIF_UPROBE | \ @@ -1666,7 +1993,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/entry/common.c linux-4.9.28/arch/x86/entry/ static void exit_to_usermode_loop(struct pt_regs *regs, u32 cached_flags) { -@@ -145,9 +145,16 @@ +@@ -145,9 +145,16 @@ static void exit_to_usermode_loop(struct pt_regs *regs, u32 cached_flags) /* We have work to do. */ local_irq_enable(); @@ -1684,10 +2011,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/entry/common.c linux-4.9.28/arch/x86/entry/ if (cached_flags & _TIF_UPROBE) uprobe_notify_resume(regs); -diff -Nur linux-4.9.28.orig/arch/x86/entry/entry_32.S linux-4.9.28/arch/x86/entry/entry_32.S ---- linux-4.9.28.orig/arch/x86/entry/entry_32.S 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/entry/entry_32.S 2017-05-19 03:37:25.130174719 +0200 -@@ -308,8 +308,25 @@ +diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S +index edba8606b99a..4a3389535fc6 100644 +--- a/arch/x86/entry/entry_32.S ++++ b/arch/x86/entry/entry_32.S +@@ -308,8 +308,25 @@ END(ret_from_exception) ENTRY(resume_kernel) DISABLE_INTERRUPTS(CLBR_ANY) need_resched: @@ -1713,10 +2041,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/entry/entry_32.S linux-4.9.28/arch/x86/entr testl $X86_EFLAGS_IF, PT_EFLAGS(%esp) # interrupts off (exception path) ? jz restore_all call preempt_schedule_irq -diff -Nur linux-4.9.28.orig/arch/x86/entry/entry_64.S linux-4.9.28/arch/x86/entry/entry_64.S ---- linux-4.9.28.orig/arch/x86/entry/entry_64.S 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/entry/entry_64.S 2017-05-19 03:37:25.130174719 +0200 -@@ -546,7 +546,23 @@ +diff --git a/arch/x86/entry/entry_64.S b/arch/x86/entry/entry_64.S +index ef766a358b37..28401f826ab1 100644 +--- a/arch/x86/entry/entry_64.S ++++ b/arch/x86/entry/entry_64.S +@@ -546,7 +546,23 @@ GLOBAL(retint_user) bt $9, EFLAGS(%rsp) /* were interrupts off? */ jnc 1f 0: cmpl $0, PER_CPU_VAR(__preempt_count) @@ -1740,7 +2069,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/entry/entry_64.S linux-4.9.28/arch/x86/entr call preempt_schedule_irq jmp 0b 1: -@@ -894,6 +910,7 @@ +@@ -894,6 +910,7 @@ EXPORT_SYMBOL(native_load_gs_index) jmp 2b .previous @@ -1748,7 +2077,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/entry/entry_64.S linux-4.9.28/arch/x86/entr /* Call softirq on interrupt stack. Interrupts are off. */ ENTRY(do_softirq_own_stack) pushq %rbp -@@ -906,6 +923,7 @@ +@@ -906,6 +923,7 @@ ENTRY(do_softirq_own_stack) decl PER_CPU_VAR(irq_count) ret END(do_softirq_own_stack) @@ -1756,10 +2085,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/entry/entry_64.S linux-4.9.28/arch/x86/entr #ifdef CONFIG_XEN idtentry xen_hypervisor_callback xen_do_hypervisor_callback has_error_code=0 -diff -Nur linux-4.9.28.orig/arch/x86/include/asm/preempt.h linux-4.9.28/arch/x86/include/asm/preempt.h ---- linux-4.9.28.orig/arch/x86/include/asm/preempt.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/include/asm/preempt.h 2017-05-19 03:37:25.130174719 +0200 -@@ -79,17 +79,46 @@ +diff --git a/arch/x86/include/asm/preempt.h b/arch/x86/include/asm/preempt.h +index 17f218645701..11bd1b7ee6eb 100644 +--- a/arch/x86/include/asm/preempt.h ++++ b/arch/x86/include/asm/preempt.h +@@ -79,17 +79,46 @@ static __always_inline void __preempt_count_sub(int val) * a decrement which hits zero means we have no preempt_count and should * reschedule. */ @@ -1807,10 +2137,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/include/asm/preempt.h linux-4.9.28/arch/x86 } #ifdef CONFIG_PREEMPT -diff -Nur linux-4.9.28.orig/arch/x86/include/asm/signal.h linux-4.9.28/arch/x86/include/asm/signal.h ---- linux-4.9.28.orig/arch/x86/include/asm/signal.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/include/asm/signal.h 2017-05-19 03:37:25.130174719 +0200 -@@ -27,6 +27,19 @@ +diff --git a/arch/x86/include/asm/signal.h b/arch/x86/include/asm/signal.h +index 8af22be0fe61..d1328789b759 100644 +--- a/arch/x86/include/asm/signal.h ++++ b/arch/x86/include/asm/signal.h +@@ -27,6 +27,19 @@ typedef struct { #define SA_IA32_ABI 0x02000000u #define SA_X32_ABI 0x01000000u @@ -1830,9 +2161,10 @@ diff -Nur linux-4.9.28.orig/arch/x86/include/asm/signal.h linux-4.9.28/arch/x86/ #ifndef CONFIG_COMPAT typedef sigset_t compat_sigset_t; #endif -diff -Nur linux-4.9.28.orig/arch/x86/include/asm/stackprotector.h linux-4.9.28/arch/x86/include/asm/stackprotector.h ---- linux-4.9.28.orig/arch/x86/include/asm/stackprotector.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/include/asm/stackprotector.h 2017-05-19 03:37:25.130174719 +0200 +diff --git a/arch/x86/include/asm/stackprotector.h b/arch/x86/include/asm/stackprotector.h +index 58505f01962f..02fa39652cd6 100644 +--- a/arch/x86/include/asm/stackprotector.h ++++ b/arch/x86/include/asm/stackprotector.h @@ -59,7 +59,7 @@ */ static __always_inline void boot_init_stack_canary(void) @@ -1842,7 +2174,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/include/asm/stackprotector.h linux-4.9.28/a u64 tsc; #ifdef CONFIG_X86_64 -@@ -70,8 +70,15 @@ +@@ -70,8 +70,15 @@ static __always_inline void boot_init_stack_canary(void) * of randomness. The TSC only matters for very early init, * there it already has some randomness on most systems. Later * on during the bootup the random pool has true entropy too. @@ -1858,10 +2190,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/include/asm/stackprotector.h linux-4.9.28/a tsc = rdtsc(); canary += tsc + (tsc << 32UL); -diff -Nur linux-4.9.28.orig/arch/x86/include/asm/thread_info.h linux-4.9.28/arch/x86/include/asm/thread_info.h ---- linux-4.9.28.orig/arch/x86/include/asm/thread_info.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/include/asm/thread_info.h 2017-05-19 03:37:25.130174719 +0200 -@@ -54,11 +54,14 @@ +diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h +index ad6f5eb07a95..5ceb3a1c2b1a 100644 +--- a/arch/x86/include/asm/thread_info.h ++++ b/arch/x86/include/asm/thread_info.h +@@ -54,11 +54,14 @@ struct task_struct; struct thread_info { unsigned long flags; /* low level flags */ @@ -1876,7 +2209,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/include/asm/thread_info.h linux-4.9.28/arch } #define init_stack (init_thread_union.stack) -@@ -67,6 +70,10 @@ +@@ -67,6 +70,10 @@ struct thread_info { #include @@ -1887,7 +2220,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/include/asm/thread_info.h linux-4.9.28/arch #endif /* -@@ -85,6 +92,7 @@ +@@ -85,6 +92,7 @@ struct thread_info { #define TIF_SYSCALL_EMU 6 /* syscall emulation active */ #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ #define TIF_SECCOMP 8 /* secure computing */ @@ -1895,7 +2228,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/include/asm/thread_info.h linux-4.9.28/arch #define TIF_USER_RETURN_NOTIFY 11 /* notify kernel of userspace return */ #define TIF_UPROBE 12 /* breakpointed or singlestepping */ #define TIF_NOTSC 16 /* TSC is not accessible in userland */ -@@ -108,6 +116,7 @@ +@@ -108,6 +116,7 @@ struct thread_info { #define _TIF_SYSCALL_EMU (1 << TIF_SYSCALL_EMU) #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) #define _TIF_SECCOMP (1 << TIF_SECCOMP) @@ -1903,7 +2236,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/include/asm/thread_info.h linux-4.9.28/arch #define _TIF_USER_RETURN_NOTIFY (1 << TIF_USER_RETURN_NOTIFY) #define _TIF_UPROBE (1 << TIF_UPROBE) #define _TIF_NOTSC (1 << TIF_NOTSC) -@@ -143,6 +152,8 @@ +@@ -143,6 +152,8 @@ struct thread_info { #define _TIF_WORK_CTXSW_PREV (_TIF_WORK_CTXSW|_TIF_USER_RETURN_NOTIFY) #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW) @@ -1912,10 +2245,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/include/asm/thread_info.h linux-4.9.28/arch #define STACK_WARN (THREAD_SIZE/8) /* -diff -Nur linux-4.9.28.orig/arch/x86/include/asm/uv/uv_bau.h linux-4.9.28/arch/x86/include/asm/uv/uv_bau.h ---- linux-4.9.28.orig/arch/x86/include/asm/uv/uv_bau.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/include/asm/uv/uv_bau.h 2017-05-19 03:37:25.130174719 +0200 -@@ -624,9 +624,9 @@ +diff --git a/arch/x86/include/asm/uv/uv_bau.h b/arch/x86/include/asm/uv/uv_bau.h +index 57ab86d94d64..35d25e27180f 100644 +--- a/arch/x86/include/asm/uv/uv_bau.h ++++ b/arch/x86/include/asm/uv/uv_bau.h +@@ -624,9 +624,9 @@ struct bau_control { cycles_t send_message; cycles_t period_end; cycles_t period_time; @@ -1928,7 +2262,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/include/asm/uv/uv_bau.h linux-4.9.28/arch/x /* tunables */ int max_concurr; int max_concurr_const; -@@ -815,15 +815,15 @@ +@@ -815,15 +815,15 @@ static inline int atom_asr(short i, struct atomic_short *v) * to be lowered below the current 'v'. atomic_add_unless can only stop * on equal. */ @@ -1948,43 +2282,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/include/asm/uv/uv_bau.h linux-4.9.28/arch/x return 1; } -diff -Nur linux-4.9.28.orig/arch/x86/Kconfig linux-4.9.28/arch/x86/Kconfig ---- linux-4.9.28.orig/arch/x86/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/Kconfig 2017-05-19 03:37:25.130174719 +0200 -@@ -17,6 +17,7 @@ - ### Arch settings - config X86 - def_bool y -+ select HAVE_PREEMPT_LAZY - select ACPI_LEGACY_TABLES_LOOKUP if ACPI - select ACPI_SYSTEM_POWER_STATES_SUPPORT if ACPI - select ANON_INODES -@@ -232,8 +233,11 @@ - def_bool y - depends on ISA_DMA_API - -+config RWSEM_GENERIC_SPINLOCK -+ def_bool PREEMPT_RT_FULL -+ - config RWSEM_XCHGADD_ALGORITHM -- def_bool y -+ def_bool !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL - - config GENERIC_CALIBRATE_DELAY - def_bool y -@@ -897,7 +901,7 @@ - config MAXSMP - bool "Enable Maximum number of SMP Processors and NUMA Nodes" - depends on X86_64 && SMP && DEBUG_KERNEL -- select CPUMASK_OFFSTACK -+ select CPUMASK_OFFSTACK if !PREEMPT_RT_FULL - ---help--- - Enable maximum number of CPUS and NUMA Nodes for this architecture. - If unsure, say N. -diff -Nur linux-4.9.28.orig/arch/x86/kernel/acpi/boot.c linux-4.9.28/arch/x86/kernel/acpi/boot.c ---- linux-4.9.28.orig/arch/x86/kernel/acpi/boot.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/kernel/acpi/boot.c 2017-05-19 03:37:25.130174719 +0200 -@@ -87,7 +87,9 @@ +diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c +index 931ced8ca345..167975ac8af7 100644 +--- a/arch/x86/kernel/acpi/boot.c ++++ b/arch/x86/kernel/acpi/boot.c +@@ -87,7 +87,9 @@ static u64 acpi_lapic_addr __initdata = APIC_DEFAULT_PHYS_BASE; * ->ioapic_mutex * ->ioapic_lock */ @@ -1994,10 +2296,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/acpi/boot.c linux-4.9.28/arch/x86/ke /* -------------------------------------------------------------------------- Boot-time Configuration -diff -Nur linux-4.9.28.orig/arch/x86/kernel/apic/io_apic.c linux-4.9.28/arch/x86/kernel/apic/io_apic.c ---- linux-4.9.28.orig/arch/x86/kernel/apic/io_apic.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/kernel/apic/io_apic.c 2017-05-19 03:37:25.130174719 +0200 -@@ -1712,7 +1712,8 @@ +diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c +index 7249f1500bcb..a79d5c224004 100644 +--- a/arch/x86/kernel/apic/io_apic.c ++++ b/arch/x86/kernel/apic/io_apic.c +@@ -1712,7 +1712,8 @@ static bool io_apic_level_ack_pending(struct mp_chip_data *data) static inline bool ioapic_irqd_mask(struct irq_data *data) { /* If we are moving the irq we need to mask it */ @@ -2007,10 +2310,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/apic/io_apic.c linux-4.9.28/arch/x86 mask_ioapic_irq(data); return true; } -diff -Nur linux-4.9.28.orig/arch/x86/kernel/asm-offsets.c linux-4.9.28/arch/x86/kernel/asm-offsets.c ---- linux-4.9.28.orig/arch/x86/kernel/asm-offsets.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/kernel/asm-offsets.c 2017-05-19 03:37:25.130174719 +0200 -@@ -36,6 +36,7 @@ +diff --git a/arch/x86/kernel/asm-offsets.c b/arch/x86/kernel/asm-offsets.c +index c62e015b126c..0cc71257fca6 100644 +--- a/arch/x86/kernel/asm-offsets.c ++++ b/arch/x86/kernel/asm-offsets.c +@@ -36,6 +36,7 @@ void common(void) { BLANK(); OFFSET(TASK_TI_flags, task_struct, thread_info.flags); @@ -2018,15 +2322,16 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/asm-offsets.c linux-4.9.28/arch/x86/ OFFSET(TASK_addr_limit, task_struct, thread.addr_limit); BLANK(); -@@ -91,4 +92,5 @@ +@@ -91,4 +92,5 @@ void common(void) { BLANK(); DEFINE(PTREGS_SIZE, sizeof(struct pt_regs)); + DEFINE(_PREEMPT_ENABLED, PREEMPT_ENABLED); } -diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x86/kernel/cpu/mcheck/mce.c ---- linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/kernel/cpu/mcheck/mce.c 2017-05-19 03:37:25.130174719 +0200 +diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c +index 22cda29d654e..57c85e3af092 100644 +--- a/arch/x86/kernel/cpu/mcheck/mce.c ++++ b/arch/x86/kernel/cpu/mcheck/mce.c @@ -41,6 +41,8 @@ #include #include @@ -2036,7 +2341,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x #include #include -@@ -1307,7 +1309,7 @@ +@@ -1307,7 +1309,7 @@ void mce_log_therm_throt_event(__u64 status) static unsigned long check_interval = INITIAL_CHECK_INTERVAL; static DEFINE_PER_CPU(unsigned long, mce_next_interval); /* in jiffies */ @@ -2045,7 +2350,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x static unsigned long mce_adjust_timer_default(unsigned long interval) { -@@ -1316,32 +1318,18 @@ +@@ -1316,32 +1318,18 @@ static unsigned long mce_adjust_timer_default(unsigned long interval) static unsigned long (*mce_adjust_timer)(unsigned long interval) = mce_adjust_timer_default; @@ -2084,7 +2389,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x iv = __this_cpu_read(mce_next_interval); if (mce_available(this_cpu_ptr(&cpu_info))) { -@@ -1364,7 +1352,7 @@ +@@ -1364,7 +1352,7 @@ static void mce_timer_fn(unsigned long data) done: __this_cpu_write(mce_next_interval, iv); @@ -2093,7 +2398,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x } /* -@@ -1372,7 +1360,7 @@ +@@ -1372,7 +1360,7 @@ static void mce_timer_fn(unsigned long data) */ void mce_timer_kick(unsigned long interval) { @@ -2102,7 +2407,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x unsigned long iv = __this_cpu_read(mce_next_interval); __restart_timer(t, interval); -@@ -1387,7 +1375,7 @@ +@@ -1387,7 +1375,7 @@ static void mce_timer_delete_all(void) int cpu; for_each_online_cpu(cpu) @@ -2111,7 +2416,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x } static void mce_do_trigger(struct work_struct *work) -@@ -1397,6 +1385,56 @@ +@@ -1397,6 +1385,56 @@ static void mce_do_trigger(struct work_struct *work) static DECLARE_WORK(mce_trigger_work, mce_do_trigger); @@ -2168,7 +2473,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x /* * Notify the user(s) about new machine check events. * Can be called from interrupt context, but not from machine check/NMI -@@ -1404,19 +1442,8 @@ +@@ -1404,19 +1442,8 @@ static DECLARE_WORK(mce_trigger_work, mce_do_trigger); */ int mce_notify_irq(void) { @@ -2189,7 +2494,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x return 1; } return 0; -@@ -1722,7 +1749,7 @@ +@@ -1722,7 +1749,7 @@ static void __mcheck_cpu_clear_vendor(struct cpuinfo_x86 *c) } } @@ -2198,7 +2503,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x { unsigned long iv = check_interval * HZ; -@@ -1731,16 +1758,17 @@ +@@ -1731,16 +1758,17 @@ static void mce_start_timer(unsigned int cpu, struct timer_list *t) per_cpu(mce_next_interval, cpu) = iv; @@ -2220,7 +2525,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x mce_start_timer(cpu, t); } -@@ -2465,6 +2493,8 @@ +@@ -2465,6 +2493,8 @@ static void mce_disable_cpu(void *h) if (!mce_available(raw_cpu_ptr(&cpu_info))) return; @@ -2229,7 +2534,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x if (!(action & CPU_TASKS_FROZEN)) cmci_clear(); -@@ -2487,6 +2517,7 @@ +@@ -2487,6 +2517,7 @@ static void mce_reenable_cpu(void *h) if (b->init) wrmsrl(msr_ops.ctl(i), b->ctl); } @@ -2237,7 +2542,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x } /* Get notified when a cpu comes on/off. Be hotplug friendly. */ -@@ -2494,7 +2525,6 @@ +@@ -2494,7 +2525,6 @@ static int mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) { unsigned int cpu = (unsigned long)hcpu; @@ -2245,7 +2550,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x switch (action & ~CPU_TASKS_FROZEN) { case CPU_ONLINE: -@@ -2514,11 +2544,9 @@ +@@ -2514,11 +2544,9 @@ mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) break; case CPU_DOWN_PREPARE: smp_call_function_single(cpu, mce_disable_cpu, &action, 1); @@ -2257,7 +2562,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x break; } -@@ -2557,6 +2585,10 @@ +@@ -2557,6 +2585,10 @@ static __init int mcheck_init_device(void) goto err_out; } @@ -2268,10 +2573,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.9.28/arch/x if (!zalloc_cpumask_var(&mce_device_initialized, GFP_KERNEL)) { err = -ENOMEM; goto err_out; -diff -Nur linux-4.9.28.orig/arch/x86/kernel/irq_32.c linux-4.9.28/arch/x86/kernel/irq_32.c ---- linux-4.9.28.orig/arch/x86/kernel/irq_32.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/kernel/irq_32.c 2017-05-19 03:37:25.130174719 +0200 -@@ -127,6 +127,7 @@ +diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c +index 1f38d9a4d9de..053bf3b2ef39 100644 +--- a/arch/x86/kernel/irq_32.c ++++ b/arch/x86/kernel/irq_32.c +@@ -127,6 +127,7 @@ void irq_ctx_init(int cpu) cpu, per_cpu(hardirq_stack, cpu), per_cpu(softirq_stack, cpu)); } @@ -2279,7 +2585,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/irq_32.c linux-4.9.28/arch/x86/kerne void do_softirq_own_stack(void) { struct irq_stack *irqstk; -@@ -143,6 +144,7 @@ +@@ -143,6 +144,7 @@ void do_softirq_own_stack(void) call_on_stack(__do_softirq, isp); } @@ -2287,9 +2593,10 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/irq_32.c linux-4.9.28/arch/x86/kerne bool handle_irq(struct irq_desc *desc, struct pt_regs *regs) { -diff -Nur linux-4.9.28.orig/arch/x86/kernel/process_32.c linux-4.9.28/arch/x86/kernel/process_32.c ---- linux-4.9.28.orig/arch/x86/kernel/process_32.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/kernel/process_32.c 2017-05-19 03:37:25.130174719 +0200 +diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c +index bd7be8efdc4c..b3b0a7f7b1ca 100644 +--- a/arch/x86/kernel/process_32.c ++++ b/arch/x86/kernel/process_32.c @@ -35,6 +35,7 @@ #include #include @@ -2298,7 +2605,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/process_32.c linux-4.9.28/arch/x86/k #include #include -@@ -195,6 +196,35 @@ +@@ -195,6 +196,35 @@ start_thread(struct pt_regs *regs, unsigned long new_ip, unsigned long new_sp) } EXPORT_SYMBOL_GPL(start_thread); @@ -2334,7 +2641,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/process_32.c linux-4.9.28/arch/x86/k /* * switch_to(x,y) should switch tasks from x to y. -@@ -271,6 +301,8 @@ +@@ -271,6 +301,8 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT)) __switch_to_xtra(prev_p, next_p, tss); @@ -2343,10 +2650,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/kernel/process_32.c linux-4.9.28/arch/x86/k /* * Leave lazy mode, flushing any hypercalls made here. * This must be done before restoring TLS segments so -diff -Nur linux-4.9.28.orig/arch/x86/kvm/lapic.c linux-4.9.28/arch/x86/kvm/lapic.c ---- linux-4.9.28.orig/arch/x86/kvm/lapic.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/kvm/lapic.c 2017-05-19 03:37:25.134174924 +0200 -@@ -1939,6 +1939,7 @@ +diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c +index 3f05c044720b..fe68afd37162 100644 +--- a/arch/x86/kvm/lapic.c ++++ b/arch/x86/kvm/lapic.c +@@ -1939,6 +1939,7 @@ int kvm_create_lapic(struct kvm_vcpu *vcpu) hrtimer_init(&apic->lapic_timer.timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS_PINNED); apic->lapic_timer.timer.function = apic_timer_fn; @@ -2354,10 +2662,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/kvm/lapic.c linux-4.9.28/arch/x86/kvm/lapic /* * APIC is created enabled. This will prevent kvm_lapic_set_base from -diff -Nur linux-4.9.28.orig/arch/x86/kvm/x86.c linux-4.9.28/arch/x86/kvm/x86.c ---- linux-4.9.28.orig/arch/x86/kvm/x86.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/kvm/x86.c 2017-05-19 03:37:25.134174924 +0200 -@@ -5933,6 +5933,13 @@ +diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c +index 81bba3c2137d..fcb84512e85d 100644 +--- a/arch/x86/kvm/x86.c ++++ b/arch/x86/kvm/x86.c +@@ -5958,6 +5958,13 @@ int kvm_arch_init(void *opaque) goto out; } @@ -2371,10 +2680,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/kvm/x86.c linux-4.9.28/arch/x86/kvm/x86.c r = kvm_mmu_module_init(); if (r) goto out_free_percpu; -diff -Nur linux-4.9.28.orig/arch/x86/mm/highmem_32.c linux-4.9.28/arch/x86/mm/highmem_32.c ---- linux-4.9.28.orig/arch/x86/mm/highmem_32.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/mm/highmem_32.c 2017-05-19 03:37:25.134174924 +0200 -@@ -32,10 +32,11 @@ +diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c +index 6d18b70ed5a9..f752724c22e8 100644 +--- a/arch/x86/mm/highmem_32.c ++++ b/arch/x86/mm/highmem_32.c +@@ -32,10 +32,11 @@ EXPORT_SYMBOL(kunmap); */ void *kmap_atomic_prot(struct page *page, pgprot_t prot) { @@ -2387,7 +2697,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/mm/highmem_32.c linux-4.9.28/arch/x86/mm/hi pagefault_disable(); if (!PageHighMem(page)) -@@ -45,7 +46,10 @@ +@@ -45,7 +46,10 @@ void *kmap_atomic_prot(struct page *page, pgprot_t prot) idx = type + KM_TYPE_NR*smp_processor_id(); vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); BUG_ON(!pte_none(*(kmap_pte-idx))); @@ -2399,7 +2709,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/mm/highmem_32.c linux-4.9.28/arch/x86/mm/hi arch_flush_lazy_mmu_mode(); return (void *)vaddr; -@@ -88,6 +92,9 @@ +@@ -88,6 +92,9 @@ void __kunmap_atomic(void *kvaddr) * is a bad idea also, in case the page changes cacheability * attributes or becomes a protected page in a hypervisor. */ @@ -2409,7 +2719,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/mm/highmem_32.c linux-4.9.28/arch/x86/mm/hi kpte_clear_flush(kmap_pte-idx, vaddr); kmap_atomic_idx_pop(); arch_flush_lazy_mmu_mode(); -@@ -100,7 +107,7 @@ +@@ -100,7 +107,7 @@ void __kunmap_atomic(void *kvaddr) #endif pagefault_enable(); @@ -2418,10 +2728,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/mm/highmem_32.c linux-4.9.28/arch/x86/mm/hi } EXPORT_SYMBOL(__kunmap_atomic); -diff -Nur linux-4.9.28.orig/arch/x86/mm/iomap_32.c linux-4.9.28/arch/x86/mm/iomap_32.c ---- linux-4.9.28.orig/arch/x86/mm/iomap_32.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/mm/iomap_32.c 2017-05-19 03:37:25.134174924 +0200 -@@ -56,6 +56,7 @@ +diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c +index ada98b39b8ad..585f6829653b 100644 +--- a/arch/x86/mm/iomap_32.c ++++ b/arch/x86/mm/iomap_32.c +@@ -56,6 +56,7 @@ EXPORT_SYMBOL_GPL(iomap_free); void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot) { @@ -2429,7 +2740,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/mm/iomap_32.c linux-4.9.28/arch/x86/mm/ioma unsigned long vaddr; int idx, type; -@@ -65,7 +66,12 @@ +@@ -65,7 +66,12 @@ void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot) type = kmap_atomic_idx_push(); idx = type + KM_TYPE_NR * smp_processor_id(); vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); @@ -2443,7 +2754,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/mm/iomap_32.c linux-4.9.28/arch/x86/mm/ioma arch_flush_lazy_mmu_mode(); return (void *)vaddr; -@@ -113,6 +119,9 @@ +@@ -113,6 +119,9 @@ iounmap_atomic(void __iomem *kvaddr) * is a bad idea also, in case the page changes cacheability * attributes or becomes a protected page in a hypervisor. */ @@ -2453,10 +2764,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/mm/iomap_32.c linux-4.9.28/arch/x86/mm/ioma kpte_clear_flush(kmap_pte-idx, vaddr); kmap_atomic_idx_pop(); } -diff -Nur linux-4.9.28.orig/arch/x86/mm/pageattr.c linux-4.9.28/arch/x86/mm/pageattr.c ---- linux-4.9.28.orig/arch/x86/mm/pageattr.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/mm/pageattr.c 2017-05-19 03:37:25.134174924 +0200 -@@ -214,7 +214,15 @@ +diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c +index e3353c97d086..01664968555c 100644 +--- a/arch/x86/mm/pageattr.c ++++ b/arch/x86/mm/pageattr.c +@@ -214,7 +214,15 @@ static void cpa_flush_array(unsigned long *start, int numpages, int cache, int in_flags, struct page **pages) { unsigned int i, level; @@ -2472,10 +2784,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/mm/pageattr.c linux-4.9.28/arch/x86/mm/page BUG_ON(irqs_disabled()); -diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/tlb_uv.c linux-4.9.28/arch/x86/platform/uv/tlb_uv.c ---- linux-4.9.28.orig/arch/x86/platform/uv/tlb_uv.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/platform/uv/tlb_uv.c 2017-05-19 03:37:25.134174924 +0200 -@@ -748,9 +748,9 @@ +diff --git a/arch/x86/platform/uv/tlb_uv.c b/arch/x86/platform/uv/tlb_uv.c +index 9e42842e924a..5398f97172f9 100644 +--- a/arch/x86/platform/uv/tlb_uv.c ++++ b/arch/x86/platform/uv/tlb_uv.c +@@ -748,9 +748,9 @@ static void destination_plugged(struct bau_desc *bau_desc, quiesce_local_uvhub(hmaster); @@ -2487,7 +2800,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/tlb_uv.c linux-4.9.28/arch/x86/ end_uvhub_quiesce(hmaster); -@@ -770,9 +770,9 @@ +@@ -770,9 +770,9 @@ static void destination_timeout(struct bau_desc *bau_desc, quiesce_local_uvhub(hmaster); @@ -2499,7 +2812,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/tlb_uv.c linux-4.9.28/arch/x86/ end_uvhub_quiesce(hmaster); -@@ -793,7 +793,7 @@ +@@ -793,7 +793,7 @@ static void disable_for_period(struct bau_control *bcp, struct ptc_stats *stat) cycles_t tm1; hmaster = bcp->uvhub_master; @@ -2508,7 +2821,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/tlb_uv.c linux-4.9.28/arch/x86/ if (!bcp->baudisabled) { stat->s_bau_disabled++; tm1 = get_cycles(); -@@ -806,7 +806,7 @@ +@@ -806,7 +806,7 @@ static void disable_for_period(struct bau_control *bcp, struct ptc_stats *stat) } } } @@ -2517,7 +2830,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/tlb_uv.c linux-4.9.28/arch/x86/ } static void count_max_concurr(int stat, struct bau_control *bcp, -@@ -869,7 +869,7 @@ +@@ -869,7 +869,7 @@ static void record_send_stats(cycles_t time1, cycles_t time2, */ static void uv1_throttle(struct bau_control *hmaster, struct ptc_stats *stat) { @@ -2526,7 +2839,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/tlb_uv.c linux-4.9.28/arch/x86/ atomic_t *v; v = &hmaster->active_descriptor_count; -@@ -1002,7 +1002,7 @@ +@@ -1002,7 +1002,7 @@ static int check_enable(struct bau_control *bcp, struct ptc_stats *stat) struct bau_control *hmaster; hmaster = bcp->uvhub_master; @@ -2535,7 +2848,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/tlb_uv.c linux-4.9.28/arch/x86/ if (bcp->baudisabled && (get_cycles() >= bcp->set_bau_on_time)) { stat->s_bau_reenabled++; for_each_present_cpu(tcpu) { -@@ -1014,10 +1014,10 @@ +@@ -1014,10 +1014,10 @@ static int check_enable(struct bau_control *bcp, struct ptc_stats *stat) tbcp->period_giveups = 0; } } @@ -2548,7 +2861,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/tlb_uv.c linux-4.9.28/arch/x86/ return -1; } -@@ -1940,9 +1940,9 @@ +@@ -1940,9 +1940,9 @@ static void __init init_per_cpu_tunables(void) bcp->cong_reps = congested_reps; bcp->disabled_period = sec_2_cycles(disabled_period); bcp->giveup_limit = giveup_limit; @@ -2561,10 +2874,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/tlb_uv.c linux-4.9.28/arch/x86/ } } -diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/uv_time.c linux-4.9.28/arch/x86/platform/uv/uv_time.c ---- linux-4.9.28.orig/arch/x86/platform/uv/uv_time.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/arch/x86/platform/uv/uv_time.c 2017-05-19 03:37:25.134174924 +0200 -@@ -57,7 +57,7 @@ +diff --git a/arch/x86/platform/uv/uv_time.c b/arch/x86/platform/uv/uv_time.c +index b333fc45f9ec..8b85916e6986 100644 +--- a/arch/x86/platform/uv/uv_time.c ++++ b/arch/x86/platform/uv/uv_time.c +@@ -57,7 +57,7 @@ static DEFINE_PER_CPU(struct clock_event_device, cpu_ced); /* There is one of these allocated per node */ struct uv_rtc_timer_head { @@ -2573,7 +2887,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/uv_time.c linux-4.9.28/arch/x86 /* next cpu waiting for timer, local node relative: */ int next_cpu; /* number of cpus on this node: */ -@@ -177,7 +177,7 @@ +@@ -177,7 +177,7 @@ static __init int uv_rtc_allocate_timers(void) uv_rtc_deallocate_timers(); return -ENOMEM; } @@ -2582,7 +2896,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/uv_time.c linux-4.9.28/arch/x86 head->ncpus = uv_blade_nr_possible_cpus(bid); head->next_cpu = -1; blade_info[bid] = head; -@@ -231,7 +231,7 @@ +@@ -231,7 +231,7 @@ static int uv_rtc_set_timer(int cpu, u64 expires) unsigned long flags; int next_cpu; @@ -2591,7 +2905,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/uv_time.c linux-4.9.28/arch/x86 next_cpu = head->next_cpu; *t = expires; -@@ -243,12 +243,12 @@ +@@ -243,12 +243,12 @@ static int uv_rtc_set_timer(int cpu, u64 expires) if (uv_setup_intr(cpu, expires)) { *t = ULLONG_MAX; uv_rtc_find_next_timer(head, pnode); @@ -2606,7 +2920,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/uv_time.c linux-4.9.28/arch/x86 return 0; } -@@ -267,7 +267,7 @@ +@@ -267,7 +267,7 @@ static int uv_rtc_unset_timer(int cpu, int force) unsigned long flags; int rc = 0; @@ -2615,7 +2929,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/uv_time.c linux-4.9.28/arch/x86 if ((head->next_cpu == bcpu && uv_read_rtc(NULL) >= *t) || force) rc = 1; -@@ -279,7 +279,7 @@ +@@ -279,7 +279,7 @@ static int uv_rtc_unset_timer(int cpu, int force) uv_rtc_find_next_timer(head, pnode); } @@ -2624,7 +2938,7 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/uv_time.c linux-4.9.28/arch/x86 return rc; } -@@ -299,13 +299,18 @@ +@@ -299,13 +299,18 @@ static int uv_rtc_unset_timer(int cpu, int force) static cycle_t uv_read_rtc(struct clocksource *cs) { unsigned long offset; @@ -2644,10 +2958,11 @@ diff -Nur linux-4.9.28.orig/arch/x86/platform/uv/uv_time.c linux-4.9.28/arch/x86 } /* -diff -Nur linux-4.9.28.orig/block/blk-core.c linux-4.9.28/block/blk-core.c ---- linux-4.9.28.orig/block/blk-core.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/block/blk-core.c 2017-05-19 03:37:25.134174924 +0200 -@@ -125,6 +125,9 @@ +diff --git a/block/blk-core.c b/block/blk-core.c +index d1f2801ce836..6f945bb0fa1a 100644 +--- a/block/blk-core.c ++++ b/block/blk-core.c +@@ -125,6 +125,9 @@ void blk_rq_init(struct request_queue *q, struct request *rq) INIT_LIST_HEAD(&rq->queuelist); INIT_LIST_HEAD(&rq->timeout_list); @@ -2657,7 +2972,7 @@ diff -Nur linux-4.9.28.orig/block/blk-core.c linux-4.9.28/block/blk-core.c rq->cpu = -1; rq->q = q; rq->__sector = (sector_t) -1; -@@ -233,7 +236,7 @@ +@@ -233,7 +236,7 @@ EXPORT_SYMBOL(blk_start_queue_async); **/ void blk_start_queue(struct request_queue *q) { @@ -2666,7 +2981,7 @@ diff -Nur linux-4.9.28.orig/block/blk-core.c linux-4.9.28/block/blk-core.c queue_flag_clear(QUEUE_FLAG_STOPPED, q); __blk_run_queue(q); -@@ -659,7 +662,7 @@ +@@ -659,7 +662,7 @@ int blk_queue_enter(struct request_queue *q, bool nowait) if (nowait) return -EBUSY; @@ -2675,7 +2990,7 @@ diff -Nur linux-4.9.28.orig/block/blk-core.c linux-4.9.28/block/blk-core.c !atomic_read(&q->mq_freeze_depth) || blk_queue_dying(q)); if (blk_queue_dying(q)) -@@ -679,7 +682,7 @@ +@@ -679,7 +682,7 @@ static void blk_queue_usage_counter_release(struct percpu_ref *ref) struct request_queue *q = container_of(ref, struct request_queue, q_usage_counter); @@ -2684,7 +2999,7 @@ diff -Nur linux-4.9.28.orig/block/blk-core.c linux-4.9.28/block/blk-core.c } static void blk_rq_timed_out_timer(unsigned long data) -@@ -748,7 +751,7 @@ +@@ -748,7 +751,7 @@ struct request_queue *blk_alloc_queue_node(gfp_t gfp_mask, int node_id) q->bypass_depth = 1; __set_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); @@ -2693,7 +3008,7 @@ diff -Nur linux-4.9.28.orig/block/blk-core.c linux-4.9.28/block/blk-core.c /* * Init percpu_ref in atomic mode so that it's faster to shutdown. -@@ -3200,7 +3203,7 @@ +@@ -3200,7 +3203,7 @@ static void queue_unplugged(struct request_queue *q, unsigned int depth, blk_run_queue_async(q); else __blk_run_queue(q); @@ -2702,7 +3017,7 @@ diff -Nur linux-4.9.28.orig/block/blk-core.c linux-4.9.28/block/blk-core.c } static void flush_plug_callbacks(struct blk_plug *plug, bool from_schedule) -@@ -3248,7 +3251,6 @@ +@@ -3248,7 +3251,6 @@ EXPORT_SYMBOL(blk_check_plugged); void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) { struct request_queue *q; @@ -2710,7 +3025,7 @@ diff -Nur linux-4.9.28.orig/block/blk-core.c linux-4.9.28/block/blk-core.c struct request *rq; LIST_HEAD(list); unsigned int depth; -@@ -3268,11 +3270,6 @@ +@@ -3268,11 +3270,6 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) q = NULL; depth = 0; @@ -2722,508 +3037,304 @@ diff -Nur linux-4.9.28.orig/block/blk-core.c linux-4.9.28/block/blk-core.c while (!list_empty(&list)) { rq = list_entry_rq(list.next); list_del_init(&rq->queuelist); -@@ -3285,7 +3282,7 @@ +@@ -3285,7 +3282,7 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) queue_unplugged(q, depth, from_schedule); q = rq->q; depth = 0; -- spin_lock(q->queue_lock); -+ spin_lock_irq(q->queue_lock); - } - - /* -@@ -3312,8 +3309,6 @@ - */ - if (q) - queue_unplugged(q, depth, from_schedule); -- -- local_irq_restore(flags); - } - - void blk_finish_plug(struct blk_plug *plug) -diff -Nur linux-4.9.28.orig/block/blk-ioc.c linux-4.9.28/block/blk-ioc.c ---- linux-4.9.28.orig/block/blk-ioc.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/block/blk-ioc.c 2017-05-19 03:37:25.134174924 +0200 -@@ -7,6 +7,7 @@ - #include - #include - #include -+#include - - #include "blk.h" - -@@ -109,7 +110,7 @@ - spin_unlock(q->queue_lock); - } else { - spin_unlock_irqrestore(&ioc->lock, flags); -- cpu_relax(); -+ cpu_chill(); - spin_lock_irqsave_nested(&ioc->lock, flags, 1); - } - } -@@ -187,7 +188,7 @@ - spin_unlock(icq->q->queue_lock); - } else { - spin_unlock_irqrestore(&ioc->lock, flags); -- cpu_relax(); -+ cpu_chill(); - goto retry; - } - } -diff -Nur linux-4.9.28.orig/block/blk-mq.c linux-4.9.28/block/blk-mq.c ---- linux-4.9.28.orig/block/blk-mq.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/block/blk-mq.c 2017-05-19 03:37:25.134174924 +0200 -@@ -72,7 +72,7 @@ - - static void blk_mq_freeze_queue_wait(struct request_queue *q) - { -- wait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->q_usage_counter)); -+ swait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->q_usage_counter)); - } - - /* -@@ -110,7 +110,7 @@ - WARN_ON_ONCE(freeze_depth < 0); - if (!freeze_depth) { - percpu_ref_reinit(&q->q_usage_counter); -- wake_up_all(&q->mq_freeze_wq); -+ swake_up_all(&q->mq_freeze_wq); - } - } - EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue); -@@ -129,7 +129,7 @@ - * dying, we need to ensure that processes currently waiting on - * the queue are notified as well. - */ -- wake_up_all(&q->mq_freeze_wq); -+ swake_up_all(&q->mq_freeze_wq); - } - - bool blk_mq_can_queue(struct blk_mq_hw_ctx *hctx) -@@ -177,6 +177,9 @@ - rq->resid_len = 0; - rq->sense = NULL; - -+#ifdef CONFIG_PREEMPT_RT_FULL -+ INIT_WORK(&rq->work, __blk_mq_complete_request_remote_work); -+#endif - INIT_LIST_HEAD(&rq->timeout_list); - rq->timeout = 0; - -@@ -345,6 +348,17 @@ - } - EXPORT_SYMBOL(blk_mq_end_request); - -+#ifdef CONFIG_PREEMPT_RT_FULL -+ -+void __blk_mq_complete_request_remote_work(struct work_struct *work) -+{ -+ struct request *rq = container_of(work, struct request, work); -+ -+ rq->q->softirq_done_fn(rq); -+} -+ -+#else -+ - static void __blk_mq_complete_request_remote(void *data) - { - struct request *rq = data; -@@ -352,6 +366,8 @@ - rq->q->softirq_done_fn(rq); - } - -+#endif -+ - static void blk_mq_ipi_complete_request(struct request *rq) - { - struct blk_mq_ctx *ctx = rq->mq_ctx; -@@ -363,19 +379,23 @@ - return; - } - -- cpu = get_cpu(); -+ cpu = get_cpu_light(); - if (!test_bit(QUEUE_FLAG_SAME_FORCE, &rq->q->queue_flags)) - shared = cpus_share_cache(cpu, ctx->cpu); - - if (cpu != ctx->cpu && !shared && cpu_online(ctx->cpu)) { -+#ifdef CONFIG_PREEMPT_RT_FULL -+ schedule_work_on(ctx->cpu, &rq->work); -+#else - rq->csd.func = __blk_mq_complete_request_remote; - rq->csd.info = rq; - rq->csd.flags = 0; - smp_call_function_single_async(ctx->cpu, &rq->csd); -+#endif - } else { - rq->q->softirq_done_fn(rq); - } -- put_cpu(); -+ put_cpu_light(); - } - - static void __blk_mq_complete_request(struct request *rq) -@@ -906,14 +926,14 @@ - return; - - if (!async && !(hctx->flags & BLK_MQ_F_BLOCKING)) { -- int cpu = get_cpu(); -+ int cpu = get_cpu_light(); - if (cpumask_test_cpu(cpu, hctx->cpumask)) { - __blk_mq_run_hw_queue(hctx); -- put_cpu(); -+ put_cpu_light(); - return; - } - -- put_cpu(); -+ put_cpu_light(); - } - - kblockd_schedule_work_on(blk_mq_hctx_next_cpu(hctx), &hctx->run_work); -diff -Nur linux-4.9.28.orig/block/blk-mq.h linux-4.9.28/block/blk-mq.h ---- linux-4.9.28.orig/block/blk-mq.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/block/blk-mq.h 2017-05-19 03:37:25.134174924 +0200 -@@ -72,12 +72,12 @@ - */ - static inline struct blk_mq_ctx *blk_mq_get_ctx(struct request_queue *q) - { -- return __blk_mq_get_ctx(q, get_cpu()); -+ return __blk_mq_get_ctx(q, get_cpu_light()); - } - - static inline void blk_mq_put_ctx(struct blk_mq_ctx *ctx) - { -- put_cpu(); -+ put_cpu_light(); - } - - struct blk_mq_alloc_data { -diff -Nur linux-4.9.28.orig/block/blk-softirq.c linux-4.9.28/block/blk-softirq.c ---- linux-4.9.28.orig/block/blk-softirq.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/block/blk-softirq.c 2017-05-19 03:37:25.134174924 +0200 -@@ -51,6 +51,7 @@ - raise_softirq_irqoff(BLOCK_SOFTIRQ); - - local_irq_restore(flags); -+ preempt_check_resched_rt(); - } - - /* -@@ -89,6 +90,7 @@ - this_cpu_ptr(&blk_cpu_done)); - raise_softirq_irqoff(BLOCK_SOFTIRQ); - local_irq_enable(); -+ preempt_check_resched_rt(); - - return 0; - } -@@ -141,6 +143,7 @@ - goto do_local; - - local_irq_restore(flags); -+ preempt_check_resched_rt(); - } - - /** -diff -Nur linux-4.9.28.orig/block/bounce.c linux-4.9.28/block/bounce.c ---- linux-4.9.28.orig/block/bounce.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/block/bounce.c 2017-05-19 03:37:25.134174924 +0200 -@@ -55,11 +55,11 @@ - unsigned long flags; - unsigned char *vto; - -- local_irq_save(flags); -+ local_irq_save_nort(flags); - vto = kmap_atomic(to->bv_page); - memcpy(vto + to->bv_offset, vfrom, to->bv_len); - kunmap_atomic(vto); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - } - - #else /* CONFIG_HIGHMEM */ -diff -Nur linux-4.9.28.orig/crypto/algapi.c linux-4.9.28/crypto/algapi.c ---- linux-4.9.28.orig/crypto/algapi.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/crypto/algapi.c 2017-05-19 03:37:25.134174924 +0200 -@@ -719,13 +719,13 @@ - - int crypto_register_notifier(struct notifier_block *nb) - { -- return blocking_notifier_chain_register(&crypto_chain, nb); -+ return srcu_notifier_chain_register(&crypto_chain, nb); - } - EXPORT_SYMBOL_GPL(crypto_register_notifier); - - int crypto_unregister_notifier(struct notifier_block *nb) - { -- return blocking_notifier_chain_unregister(&crypto_chain, nb); -+ return srcu_notifier_chain_unregister(&crypto_chain, nb); - } - EXPORT_SYMBOL_GPL(crypto_unregister_notifier); +- spin_lock(q->queue_lock); ++ spin_lock_irq(q->queue_lock); + } -diff -Nur linux-4.9.28.orig/crypto/api.c linux-4.9.28/crypto/api.c ---- linux-4.9.28.orig/crypto/api.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/crypto/api.c 2017-05-19 03:37:25.134174924 +0200 -@@ -31,7 +31,7 @@ - DECLARE_RWSEM(crypto_alg_sem); - EXPORT_SYMBOL_GPL(crypto_alg_sem); + /* +@@ -3312,8 +3309,6 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) + */ + if (q) + queue_unplugged(q, depth, from_schedule); +- +- local_irq_restore(flags); + } --BLOCKING_NOTIFIER_HEAD(crypto_chain); -+SRCU_NOTIFIER_HEAD(crypto_chain); - EXPORT_SYMBOL_GPL(crypto_chain); + void blk_finish_plug(struct blk_plug *plug) +diff --git a/block/blk-ioc.c b/block/blk-ioc.c +index 381cb50a673c..dc8785233d94 100644 +--- a/block/blk-ioc.c ++++ b/block/blk-ioc.c +@@ -7,6 +7,7 @@ + #include + #include + #include ++#include - static struct crypto_alg *crypto_larval_wait(struct crypto_alg *alg); -@@ -236,10 +236,10 @@ - { - int ok; + #include "blk.h" -- ok = blocking_notifier_call_chain(&crypto_chain, val, v); -+ ok = srcu_notifier_call_chain(&crypto_chain, val, v); - if (ok == NOTIFY_DONE) { - request_module("cryptomgr"); -- ok = blocking_notifier_call_chain(&crypto_chain, val, v); -+ ok = srcu_notifier_call_chain(&crypto_chain, val, v); +@@ -109,7 +110,7 @@ static void ioc_release_fn(struct work_struct *work) + spin_unlock(q->queue_lock); + } else { + spin_unlock_irqrestore(&ioc->lock, flags); +- cpu_relax(); ++ cpu_chill(); + spin_lock_irqsave_nested(&ioc->lock, flags, 1); + } } +@@ -187,7 +188,7 @@ void put_io_context_active(struct io_context *ioc) + spin_unlock(icq->q->queue_lock); + } else { + spin_unlock_irqrestore(&ioc->lock, flags); +- cpu_relax(); ++ cpu_chill(); + goto retry; + } + } +diff --git a/block/blk-mq.c b/block/blk-mq.c +index 7b597ec4e9c5..48c9652a701c 100644 +--- a/block/blk-mq.c ++++ b/block/blk-mq.c +@@ -72,7 +72,7 @@ EXPORT_SYMBOL_GPL(blk_mq_freeze_queue_start); - return ok; -diff -Nur linux-4.9.28.orig/crypto/internal.h linux-4.9.28/crypto/internal.h ---- linux-4.9.28.orig/crypto/internal.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/crypto/internal.h 2017-05-19 03:37:25.138175108 +0200 -@@ -47,7 +47,7 @@ - - extern struct list_head crypto_alg_list; - extern struct rw_semaphore crypto_alg_sem; --extern struct blocking_notifier_head crypto_chain; -+extern struct srcu_notifier_head crypto_chain; - - #ifdef CONFIG_PROC_FS - void __init crypto_init_proc(void); -@@ -146,7 +146,7 @@ - - static inline void crypto_notify(unsigned long val, void *v) + static void blk_mq_freeze_queue_wait(struct request_queue *q) { -- blocking_notifier_call_chain(&crypto_chain, val, v); -+ srcu_notifier_call_chain(&crypto_chain, val, v); +- wait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->q_usage_counter)); ++ swait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->q_usage_counter)); } - #endif /* _CRYPTO_INTERNAL_H */ -diff -Nur linux-4.9.28.orig/Documentation/sysrq.txt linux-4.9.28/Documentation/sysrq.txt ---- linux-4.9.28.orig/Documentation/sysrq.txt 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/Documentation/sysrq.txt 2017-05-19 03:37:25.118174016 +0200 -@@ -59,10 +59,17 @@ - On other - If you know of the key combos for other architectures, please - let me know so I can add them to this section. - --On all - write a character to /proc/sysrq-trigger. e.g.: -- -+On all - write a character to /proc/sysrq-trigger, e.g.: - echo t > /proc/sysrq-trigger - -+On all - Enable network SysRq by writing a cookie to icmp_echo_sysrq, e.g. -+ echo 0x01020304 >/proc/sys/net/ipv4/icmp_echo_sysrq -+ Send an ICMP echo request with this pattern plus the particular -+ SysRq command key. Example: -+ # ping -c1 -s57 -p0102030468 -+ will trigger the SysRq-H (help) command. -+ -+ - * What are the 'command' keys? - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - 'b' - Will immediately reboot the system without syncing or unmounting -diff -Nur linux-4.9.28.orig/Documentation/trace/histograms.txt linux-4.9.28/Documentation/trace/histograms.txt ---- linux-4.9.28.orig/Documentation/trace/histograms.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/Documentation/trace/histograms.txt 2017-05-19 03:37:25.118174016 +0200 -@@ -0,0 +1,186 @@ -+ Using the Linux Kernel Latency Histograms -+ -+ -+This document gives a short explanation how to enable, configure and use -+latency histograms. Latency histograms are primarily relevant in the -+context of real-time enabled kernels (CONFIG_PREEMPT/CONFIG_PREEMPT_RT) -+and are used in the quality management of the Linux real-time -+capabilities. -+ -+ -+* Purpose of latency histograms -+ -+A latency histogram continuously accumulates the frequencies of latency -+data. There are two types of histograms -+- potential sources of latencies -+- effective latencies -+ -+ -+* Potential sources of latencies -+ -+Potential sources of latencies are code segments where interrupts, -+preemption or both are disabled (aka critical sections). To create -+histograms of potential sources of latency, the kernel stores the time -+stamp at the start of a critical section, determines the time elapsed -+when the end of the section is reached, and increments the frequency -+counter of that latency value - irrespective of whether any concurrently -+running process is affected by latency or not. -+- Configuration items (in the Kernel hacking/Tracers submenu) -+ CONFIG_INTERRUPT_OFF_LATENCY -+ CONFIG_PREEMPT_OFF_LATENCY -+ -+ -+* Effective latencies -+ -+Effective latencies are actually occuring during wakeup of a process. To -+determine effective latencies, the kernel stores the time stamp when a -+process is scheduled to be woken up, and determines the duration of the -+wakeup time shortly before control is passed over to this process. Note -+that the apparent latency in user space may be somewhat longer, since the -+process may be interrupted after control is passed over to it but before -+the execution in user space takes place. Simply measuring the interval -+between enqueuing and wakeup may also not appropriate in cases when a -+process is scheduled as a result of a timer expiration. The timer may have -+missed its deadline, e.g. due to disabled interrupts, but this latency -+would not be registered. Therefore, the offsets of missed timers are -+recorded in a separate histogram. If both wakeup latency and missed timer -+offsets are configured and enabled, a third histogram may be enabled that -+records the overall latency as a sum of the timer latency, if any, and the -+wakeup latency. This histogram is called "timerandwakeup". -+- Configuration items (in the Kernel hacking/Tracers submenu) -+ CONFIG_WAKEUP_LATENCY -+ CONFIG_MISSED_TIMER_OFSETS -+ -+ -+* Usage -+ -+The interface to the administration of the latency histograms is located -+in the debugfs file system. To mount it, either enter -+ -+mount -t sysfs nodev /sys -+mount -t debugfs nodev /sys/kernel/debug -+ -+from shell command line level, or add -+ -+nodev /sys sysfs defaults 0 0 -+nodev /sys/kernel/debug debugfs defaults 0 0 -+ -+to the file /etc/fstab. All latency histogram related files are then -+available in the directory /sys/kernel/debug/tracing/latency_hist. A -+particular histogram type is enabled by writing non-zero to the related -+variable in the /sys/kernel/debug/tracing/latency_hist/enable directory. -+Select "preemptirqsoff" for the histograms of potential sources of -+latencies and "wakeup" for histograms of effective latencies etc. The -+histogram data - one per CPU - are available in the files -+ -+/sys/kernel/debug/tracing/latency_hist/preemptoff/CPUx -+/sys/kernel/debug/tracing/latency_hist/irqsoff/CPUx -+/sys/kernel/debug/tracing/latency_hist/preemptirqsoff/CPUx -+/sys/kernel/debug/tracing/latency_hist/wakeup/CPUx -+/sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio/CPUx -+/sys/kernel/debug/tracing/latency_hist/missed_timer_offsets/CPUx -+/sys/kernel/debug/tracing/latency_hist/timerandwakeup/CPUx -+ -+The histograms are reset by writing non-zero to the file "reset" in a -+particular latency directory. To reset all latency data, use -+ -+#!/bin/sh -+ -+TRACINGDIR=/sys/kernel/debug/tracing -+HISTDIR=$TRACINGDIR/latency_hist -+ -+if test -d $HISTDIR -+then -+ cd $HISTDIR -+ for i in `find . | grep /reset$` -+ do -+ echo 1 >$i -+ done -+fi -+ -+ -+* Data format -+ -+Latency data are stored with a resolution of one microsecond. The -+maximum latency is 10,240 microseconds. The data are only valid, if the -+overflow register is empty. Every output line contains the latency in -+microseconds in the first row and the number of samples in the second -+row. To display only lines with a positive latency count, use, for -+example, -+ -+grep -v " 0$" /sys/kernel/debug/tracing/latency_hist/preemptoff/CPU0 -+ -+#Minimum latency: 0 microseconds. -+#Average latency: 0 microseconds. -+#Maximum latency: 25 microseconds. -+#Total samples: 3104770694 -+#There are 0 samples greater or equal than 10240 microseconds -+#usecs samples -+ 0 2984486876 -+ 1 49843506 -+ 2 58219047 -+ 3 5348126 -+ 4 2187960 -+ 5 3388262 -+ 6 959289 -+ 7 208294 -+ 8 40420 -+ 9 4485 -+ 10 14918 -+ 11 18340 -+ 12 25052 -+ 13 19455 -+ 14 5602 -+ 15 969 -+ 16 47 -+ 17 18 -+ 18 14 -+ 19 1 -+ 20 3 -+ 21 2 -+ 22 5 -+ 23 2 -+ 25 1 -+ -+ -+* Wakeup latency of a selected process -+ -+To only collect wakeup latency data of a particular process, write the -+PID of the requested process to -+ -+/sys/kernel/debug/tracing/latency_hist/wakeup/pid -+ -+PIDs are not considered, if this variable is set to 0. -+ -+ -+* Details of the process with the highest wakeup latency so far -+ -+Selected data of the process that suffered from the highest wakeup -+latency that occurred in a particular CPU are available in the file -+ -+/sys/kernel/debug/tracing/latency_hist/wakeup/max_latency-CPUx. -+ -+In addition, other relevant system data at the time when the -+latency occurred are given. + /* +@@ -110,7 +110,7 @@ void blk_mq_unfreeze_queue(struct request_queue *q) + WARN_ON_ONCE(freeze_depth < 0); + if (!freeze_depth) { + percpu_ref_reinit(&q->q_usage_counter); +- wake_up_all(&q->mq_freeze_wq); ++ swake_up_all(&q->mq_freeze_wq); + } + } + EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue); +@@ -129,7 +129,7 @@ void blk_mq_wake_waiters(struct request_queue *q) + * dying, we need to ensure that processes currently waiting on + * the queue are notified as well. + */ +- wake_up_all(&q->mq_freeze_wq); ++ swake_up_all(&q->mq_freeze_wq); + } + + bool blk_mq_can_queue(struct blk_mq_hw_ctx *hctx) +@@ -177,6 +177,9 @@ static void blk_mq_rq_ctx_init(struct request_queue *q, struct blk_mq_ctx *ctx, + rq->resid_len = 0; + rq->sense = NULL; + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ INIT_WORK(&rq->work, __blk_mq_complete_request_remote_work); ++#endif + INIT_LIST_HEAD(&rq->timeout_list); + rq->timeout = 0; + +@@ -345,6 +348,17 @@ void blk_mq_end_request(struct request *rq, int error) + } + EXPORT_SYMBOL(blk_mq_end_request); + ++#ifdef CONFIG_PREEMPT_RT_FULL + -+The format of the data is (all in one line): -+ () \ -+<- ++void __blk_mq_complete_request_remote_work(struct work_struct *work) ++{ ++ struct request *rq = container_of(work, struct request, work); + -+The value of is only relevant in the combined timer -+and wakeup latency recording. In the wakeup recording, it is -+always 0, in the missed_timer_offsets recording, it is the same -+as . ++ rq->q->softirq_done_fn(rq); ++} + -+When retrospectively searching for the origin of a latency and -+tracing was not enabled, it may be helpful to know the name and -+some basic data of the task that (finally) was switching to the -+late real-tlme task. In addition to the victim's data, also the -+data of the possible culprit are therefore displayed after the -+"<-" symbol. ++#else + -+Finally, the timestamp of the time when the latency occurred -+in . after the most recent system boot -+is provided. + static void __blk_mq_complete_request_remote(void *data) + { + struct request *rq = data; +@@ -352,6 +366,8 @@ static void __blk_mq_complete_request_remote(void *data) + rq->q->softirq_done_fn(rq); + } + ++#endif + -+These data are also reset when the wakeup histogram is reset. -diff -Nur linux-4.9.28.orig/drivers/acpi/acpica/acglobal.h linux-4.9.28/drivers/acpi/acpica/acglobal.h ---- linux-4.9.28.orig/drivers/acpi/acpica/acglobal.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/acpi/acpica/acglobal.h 2017-05-19 03:37:25.138175108 +0200 -@@ -116,7 +116,7 @@ + static void blk_mq_ipi_complete_request(struct request *rq) + { + struct blk_mq_ctx *ctx = rq->mq_ctx; +@@ -363,19 +379,23 @@ static void blk_mq_ipi_complete_request(struct request *rq) + return; + } + +- cpu = get_cpu(); ++ cpu = get_cpu_light(); + if (!test_bit(QUEUE_FLAG_SAME_FORCE, &rq->q->queue_flags)) + shared = cpus_share_cache(cpu, ctx->cpu); + + if (cpu != ctx->cpu && !shared && cpu_online(ctx->cpu)) { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ schedule_work_on(ctx->cpu, &rq->work); ++#else + rq->csd.func = __blk_mq_complete_request_remote; + rq->csd.info = rq; + rq->csd.flags = 0; + smp_call_function_single_async(ctx->cpu, &rq->csd); ++#endif + } else { + rq->q->softirq_done_fn(rq); + } +- put_cpu(); ++ put_cpu_light(); + } + + static void __blk_mq_complete_request(struct request *rq) +@@ -906,14 +926,14 @@ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async) + return; + + if (!async && !(hctx->flags & BLK_MQ_F_BLOCKING)) { +- int cpu = get_cpu(); ++ int cpu = get_cpu_light(); + if (cpumask_test_cpu(cpu, hctx->cpumask)) { + __blk_mq_run_hw_queue(hctx); +- put_cpu(); ++ put_cpu_light(); + return; + } + +- put_cpu(); ++ put_cpu_light(); + } + + kblockd_schedule_work_on(blk_mq_hctx_next_cpu(hctx), &hctx->run_work); +diff --git a/block/blk-mq.h b/block/blk-mq.h +index e5d25249028c..1e846b842eab 100644 +--- a/block/blk-mq.h ++++ b/block/blk-mq.h +@@ -72,12 +72,12 @@ static inline struct blk_mq_ctx *__blk_mq_get_ctx(struct request_queue *q, + */ + static inline struct blk_mq_ctx *blk_mq_get_ctx(struct request_queue *q) + { +- return __blk_mq_get_ctx(q, get_cpu()); ++ return __blk_mq_get_ctx(q, get_cpu_light()); + } + + static inline void blk_mq_put_ctx(struct blk_mq_ctx *ctx) + { +- put_cpu(); ++ put_cpu_light(); + } + + struct blk_mq_alloc_data { +diff --git a/block/blk-softirq.c b/block/blk-softirq.c +index 06cf9807f49a..c40342643ca0 100644 +--- a/block/blk-softirq.c ++++ b/block/blk-softirq.c +@@ -51,6 +51,7 @@ static void trigger_softirq(void *data) + raise_softirq_irqoff(BLOCK_SOFTIRQ); + + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + + /* +@@ -89,6 +90,7 @@ static int blk_softirq_cpu_dead(unsigned int cpu) + this_cpu_ptr(&blk_cpu_done)); + raise_softirq_irqoff(BLOCK_SOFTIRQ); + local_irq_enable(); ++ preempt_check_resched_rt(); + + return 0; + } +@@ -141,6 +143,7 @@ void __blk_complete_request(struct request *req) + goto do_local; + + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + + /** +diff --git a/block/bounce.c b/block/bounce.c +index 1cb5dd3a5da1..2f1ec8a67cbe 100644 +--- a/block/bounce.c ++++ b/block/bounce.c +@@ -55,11 +55,11 @@ static void bounce_copy_vec(struct bio_vec *to, unsigned char *vfrom) + unsigned long flags; + unsigned char *vto; + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + vto = kmap_atomic(to->bv_page); + memcpy(vto + to->bv_offset, vfrom, to->bv_len); + kunmap_atomic(vto); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + + #else /* CONFIG_HIGHMEM */ +diff --git a/crypto/algapi.c b/crypto/algapi.c +index 1fad2a6b3bbb..ecb7315426a9 100644 +--- a/crypto/algapi.c ++++ b/crypto/algapi.c +@@ -719,13 +719,13 @@ EXPORT_SYMBOL_GPL(crypto_spawn_tfm2); + + int crypto_register_notifier(struct notifier_block *nb) + { +- return blocking_notifier_chain_register(&crypto_chain, nb); ++ return srcu_notifier_chain_register(&crypto_chain, nb); + } + EXPORT_SYMBOL_GPL(crypto_register_notifier); + + int crypto_unregister_notifier(struct notifier_block *nb) + { +- return blocking_notifier_chain_unregister(&crypto_chain, nb); ++ return srcu_notifier_chain_unregister(&crypto_chain, nb); + } + EXPORT_SYMBOL_GPL(crypto_unregister_notifier); + +diff --git a/crypto/api.c b/crypto/api.c +index bbc147cb5dec..bc1a848f02ec 100644 +--- a/crypto/api.c ++++ b/crypto/api.c +@@ -31,7 +31,7 @@ EXPORT_SYMBOL_GPL(crypto_alg_list); + DECLARE_RWSEM(crypto_alg_sem); + EXPORT_SYMBOL_GPL(crypto_alg_sem); + +-BLOCKING_NOTIFIER_HEAD(crypto_chain); ++SRCU_NOTIFIER_HEAD(crypto_chain); + EXPORT_SYMBOL_GPL(crypto_chain); + + static struct crypto_alg *crypto_larval_wait(struct crypto_alg *alg); +@@ -236,10 +236,10 @@ int crypto_probing_notify(unsigned long val, void *v) + { + int ok; + +- ok = blocking_notifier_call_chain(&crypto_chain, val, v); ++ ok = srcu_notifier_call_chain(&crypto_chain, val, v); + if (ok == NOTIFY_DONE) { + request_module("cryptomgr"); +- ok = blocking_notifier_call_chain(&crypto_chain, val, v); ++ ok = srcu_notifier_call_chain(&crypto_chain, val, v); + } + + return ok; +diff --git a/crypto/internal.h b/crypto/internal.h +index 7eefcdb00227..0ecc7f5a2f40 100644 +--- a/crypto/internal.h ++++ b/crypto/internal.h +@@ -47,7 +47,7 @@ struct crypto_larval { + + extern struct list_head crypto_alg_list; + extern struct rw_semaphore crypto_alg_sem; +-extern struct blocking_notifier_head crypto_chain; ++extern struct srcu_notifier_head crypto_chain; + + #ifdef CONFIG_PROC_FS + void __init crypto_init_proc(void); +@@ -146,7 +146,7 @@ static inline int crypto_is_moribund(struct crypto_alg *alg) + + static inline void crypto_notify(unsigned long val, void *v) + { +- blocking_notifier_call_chain(&crypto_chain, val, v); ++ srcu_notifier_call_chain(&crypto_chain, val, v); + } + + #endif /* _CRYPTO_INTERNAL_H */ +diff --git a/drivers/acpi/acpica/acglobal.h b/drivers/acpi/acpica/acglobal.h +index 750fa824d42c..441edf51484a 100644 +--- a/drivers/acpi/acpica/acglobal.h ++++ b/drivers/acpi/acpica/acglobal.h +@@ -116,7 +116,7 @@ ACPI_GLOBAL(u8, acpi_gbl_global_lock_pending); * interrupt level */ ACPI_GLOBAL(acpi_spinlock, acpi_gbl_gpe_lock); /* For GPE data structs and registers */ @@ -3232,10 +3343,11 @@ diff -Nur linux-4.9.28.orig/drivers/acpi/acpica/acglobal.h linux-4.9.28/drivers/ ACPI_GLOBAL(acpi_spinlock, acpi_gbl_reference_count_lock); /* Mutex for _OSI support */ -diff -Nur linux-4.9.28.orig/drivers/acpi/acpica/hwregs.c linux-4.9.28/drivers/acpi/acpica/hwregs.c ---- linux-4.9.28.orig/drivers/acpi/acpica/hwregs.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/acpi/acpica/hwregs.c 2017-05-19 03:37:25.138175108 +0200 -@@ -363,14 +363,14 @@ +diff --git a/drivers/acpi/acpica/hwregs.c b/drivers/acpi/acpica/hwregs.c +index 3b7fb99362b6..696bf8e62afb 100644 +--- a/drivers/acpi/acpica/hwregs.c ++++ b/drivers/acpi/acpica/hwregs.c +@@ -363,14 +363,14 @@ acpi_status acpi_hw_clear_acpi_status(void) ACPI_BITMASK_ALL_FIXED_STATUS, ACPI_FORMAT_UINT64(acpi_gbl_xpm1a_status.address))); @@ -3252,10 +3364,11 @@ diff -Nur linux-4.9.28.orig/drivers/acpi/acpica/hwregs.c linux-4.9.28/drivers/ac if (ACPI_FAILURE(status)) { goto exit; -diff -Nur linux-4.9.28.orig/drivers/acpi/acpica/hwxface.c linux-4.9.28/drivers/acpi/acpica/hwxface.c ---- linux-4.9.28.orig/drivers/acpi/acpica/hwxface.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/acpi/acpica/hwxface.c 2017-05-19 03:37:25.138175108 +0200 -@@ -373,7 +373,7 @@ +diff --git a/drivers/acpi/acpica/hwxface.c b/drivers/acpi/acpica/hwxface.c +index 98c26ff39409..6e236f2ea791 100644 +--- a/drivers/acpi/acpica/hwxface.c ++++ b/drivers/acpi/acpica/hwxface.c +@@ -373,7 +373,7 @@ acpi_status acpi_write_bit_register(u32 register_id, u32 value) return_ACPI_STATUS(AE_BAD_PARAMETER); } @@ -3264,7 +3377,7 @@ diff -Nur linux-4.9.28.orig/drivers/acpi/acpica/hwxface.c linux-4.9.28/drivers/a /* * At this point, we know that the parent register is one of the -@@ -434,7 +434,7 @@ +@@ -434,7 +434,7 @@ acpi_status acpi_write_bit_register(u32 register_id, u32 value) unlock_and_exit: @@ -3273,10 +3386,11 @@ diff -Nur linux-4.9.28.orig/drivers/acpi/acpica/hwxface.c linux-4.9.28/drivers/a return_ACPI_STATUS(status); } -diff -Nur linux-4.9.28.orig/drivers/acpi/acpica/utmutex.c linux-4.9.28/drivers/acpi/acpica/utmutex.c ---- linux-4.9.28.orig/drivers/acpi/acpica/utmutex.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/acpi/acpica/utmutex.c 2017-05-19 03:37:25.138175108 +0200 -@@ -88,7 +88,7 @@ +diff --git a/drivers/acpi/acpica/utmutex.c b/drivers/acpi/acpica/utmutex.c +index 15073375bd00..357e7ca5a587 100644 +--- a/drivers/acpi/acpica/utmutex.c ++++ b/drivers/acpi/acpica/utmutex.c +@@ -88,7 +88,7 @@ acpi_status acpi_ut_mutex_initialize(void) return_ACPI_STATUS (status); } @@ -3285,7 +3399,7 @@ diff -Nur linux-4.9.28.orig/drivers/acpi/acpica/utmutex.c linux-4.9.28/drivers/a if (ACPI_FAILURE (status)) { return_ACPI_STATUS (status); } -@@ -145,7 +145,7 @@ +@@ -145,7 +145,7 @@ void acpi_ut_mutex_terminate(void) /* Delete the spinlocks */ acpi_os_delete_lock(acpi_gbl_gpe_lock); @@ -3294,10 +3408,11 @@ diff -Nur linux-4.9.28.orig/drivers/acpi/acpica/utmutex.c linux-4.9.28/drivers/a acpi_os_delete_lock(acpi_gbl_reference_count_lock); /* Delete the reader/writer lock */ -diff -Nur linux-4.9.28.orig/drivers/ata/libata-sff.c linux-4.9.28/drivers/ata/libata-sff.c ---- linux-4.9.28.orig/drivers/ata/libata-sff.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/ata/libata-sff.c 2017-05-19 03:37:25.138175108 +0200 -@@ -678,9 +678,9 @@ +diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c +index 051b6158d1b7..7ad293bef6ed 100644 +--- a/drivers/ata/libata-sff.c ++++ b/drivers/ata/libata-sff.c +@@ -678,9 +678,9 @@ unsigned int ata_sff_data_xfer_noirq(struct ata_device *dev, unsigned char *buf, unsigned long flags; unsigned int consumed; @@ -3309,7 +3424,7 @@ diff -Nur linux-4.9.28.orig/drivers/ata/libata-sff.c linux-4.9.28/drivers/ata/li return consumed; } -@@ -719,7 +719,7 @@ +@@ -719,7 +719,7 @@ static void ata_pio_sector(struct ata_queued_cmd *qc) unsigned long flags; /* FIXME: use a bounce buffer */ @@ -3318,7 +3433,7 @@ diff -Nur linux-4.9.28.orig/drivers/ata/libata-sff.c linux-4.9.28/drivers/ata/li buf = kmap_atomic(page); /* do the actual data transfer */ -@@ -727,7 +727,7 @@ +@@ -727,7 +727,7 @@ static void ata_pio_sector(struct ata_queued_cmd *qc) do_write); kunmap_atomic(buf); @@ -3327,7 +3442,7 @@ diff -Nur linux-4.9.28.orig/drivers/ata/libata-sff.c linux-4.9.28/drivers/ata/li } else { buf = page_address(page); ap->ops->sff_data_xfer(qc->dev, buf + offset, qc->sect_size, -@@ -864,7 +864,7 @@ +@@ -864,7 +864,7 @@ static int __atapi_pio_bytes(struct ata_queued_cmd *qc, unsigned int bytes) unsigned long flags; /* FIXME: use bounce buffer */ @@ -3336,7 +3451,7 @@ diff -Nur linux-4.9.28.orig/drivers/ata/libata-sff.c linux-4.9.28/drivers/ata/li buf = kmap_atomic(page); /* do the actual data transfer */ -@@ -872,7 +872,7 @@ +@@ -872,7 +872,7 @@ static int __atapi_pio_bytes(struct ata_queued_cmd *qc, unsigned int bytes) count, rw); kunmap_atomic(buf); @@ -3345,10 +3460,11 @@ diff -Nur linux-4.9.28.orig/drivers/ata/libata-sff.c linux-4.9.28/drivers/ata/li } else { buf = page_address(page); consumed = ap->ops->sff_data_xfer(dev, buf + offset, -diff -Nur linux-4.9.28.orig/drivers/block/zram/zcomp.c linux-4.9.28/drivers/block/zram/zcomp.c ---- linux-4.9.28.orig/drivers/block/zram/zcomp.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/block/zram/zcomp.c 2017-05-19 03:37:25.138175108 +0200 -@@ -118,12 +118,19 @@ +diff --git a/drivers/block/zram/zcomp.c b/drivers/block/zram/zcomp.c +index 4b5cd3a7b2b6..fa8329ad79fd 100644 +--- a/drivers/block/zram/zcomp.c ++++ b/drivers/block/zram/zcomp.c +@@ -118,12 +118,19 @@ ssize_t zcomp_available_show(const char *comp, char *buf) struct zcomp_strm *zcomp_stream_get(struct zcomp *comp) { @@ -3370,7 +3486,7 @@ diff -Nur linux-4.9.28.orig/drivers/block/zram/zcomp.c linux-4.9.28/drivers/bloc } int zcomp_compress(struct zcomp_strm *zstrm, -@@ -174,6 +181,7 @@ +@@ -174,6 +181,7 @@ static int __zcomp_cpu_notifier(struct zcomp *comp, pr_err("Can't allocate a compression stream\n"); return NOTIFY_BAD; } @@ -3378,10 +3494,11 @@ diff -Nur linux-4.9.28.orig/drivers/block/zram/zcomp.c linux-4.9.28/drivers/bloc *per_cpu_ptr(comp->stream, cpu) = zstrm; break; case CPU_DEAD: -diff -Nur linux-4.9.28.orig/drivers/block/zram/zcomp.h linux-4.9.28/drivers/block/zram/zcomp.h ---- linux-4.9.28.orig/drivers/block/zram/zcomp.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/block/zram/zcomp.h 2017-05-19 03:37:25.138175108 +0200 -@@ -14,6 +14,7 @@ +diff --git a/drivers/block/zram/zcomp.h b/drivers/block/zram/zcomp.h +index 478cac2ed465..f7a6efdc3285 100644 +--- a/drivers/block/zram/zcomp.h ++++ b/drivers/block/zram/zcomp.h +@@ -14,6 +14,7 @@ struct zcomp_strm { /* compression/decompression buffer */ void *buffer; struct crypto_comp *tfm; @@ -3389,10 +3506,11 @@ diff -Nur linux-4.9.28.orig/drivers/block/zram/zcomp.h linux-4.9.28/drivers/bloc }; /* dynamic per-device compression frontend */ -diff -Nur linux-4.9.28.orig/drivers/block/zram/zram_drv.c linux-4.9.28/drivers/block/zram/zram_drv.c ---- linux-4.9.28.orig/drivers/block/zram/zram_drv.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/block/zram/zram_drv.c 2017-05-19 03:37:25.138175108 +0200 -@@ -528,6 +528,8 @@ +diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c +index c9914d653968..2038d138f286 100644 +--- a/drivers/block/zram/zram_drv.c ++++ b/drivers/block/zram/zram_drv.c +@@ -528,6 +528,8 @@ static struct zram_meta *zram_meta_alloc(char *pool_name, u64 disksize) goto out_error; } @@ -3401,7 +3519,7 @@ diff -Nur linux-4.9.28.orig/drivers/block/zram/zram_drv.c linux-4.9.28/drivers/b return meta; out_error: -@@ -575,28 +577,28 @@ +@@ -575,28 +577,28 @@ static int zram_decompress_page(struct zram *zram, char *mem, u32 index) struct zram_meta *meta = zram->meta; unsigned long handle; unsigned int size; @@ -3436,7 +3554,7 @@ diff -Nur linux-4.9.28.orig/drivers/block/zram/zram_drv.c linux-4.9.28/drivers/b /* Should NEVER happen. Return bio error if it does. */ if (unlikely(ret)) { -@@ -616,14 +618,14 @@ +@@ -616,14 +618,14 @@ static int zram_bvec_read(struct zram *zram, struct bio_vec *bvec, struct zram_meta *meta = zram->meta; page = bvec->bv_page; @@ -3454,7 +3572,7 @@ diff -Nur linux-4.9.28.orig/drivers/block/zram/zram_drv.c linux-4.9.28/drivers/b if (is_partial_io(bvec)) /* Use a temporary buffer to decompress the page */ -@@ -700,10 +702,10 @@ +@@ -700,10 +702,10 @@ static int zram_bvec_write(struct zram *zram, struct bio_vec *bvec, u32 index, if (user_mem) kunmap_atomic(user_mem); /* Free memory associated with this sector now. */ @@ -3467,7 +3585,7 @@ diff -Nur linux-4.9.28.orig/drivers/block/zram/zram_drv.c linux-4.9.28/drivers/b atomic64_inc(&zram->stats.zero_pages); ret = 0; -@@ -794,12 +796,12 @@ +@@ -794,12 +796,12 @@ static int zram_bvec_write(struct zram *zram, struct bio_vec *bvec, u32 index, * Free memory associated with this sector * before overwriting unused sectors. */ @@ -3482,7 +3600,7 @@ diff -Nur linux-4.9.28.orig/drivers/block/zram/zram_drv.c linux-4.9.28/drivers/b /* Update stats */ atomic64_add(clen, &zram->stats.compr_data_size); -@@ -842,9 +844,9 @@ +@@ -842,9 +844,9 @@ static void zram_bio_discard(struct zram *zram, u32 index, } while (n >= PAGE_SIZE) { @@ -3494,7 +3612,7 @@ diff -Nur linux-4.9.28.orig/drivers/block/zram/zram_drv.c linux-4.9.28/drivers/b atomic64_inc(&zram->stats.notify_free); index++; n -= PAGE_SIZE; -@@ -973,9 +975,9 @@ +@@ -973,9 +975,9 @@ static void zram_slot_free_notify(struct block_device *bdev, zram = bdev->bd_disk->private_data; meta = zram->meta; @@ -3506,10 +3624,11 @@ diff -Nur linux-4.9.28.orig/drivers/block/zram/zram_drv.c linux-4.9.28/drivers/b atomic64_inc(&zram->stats.notify_free); } -diff -Nur linux-4.9.28.orig/drivers/block/zram/zram_drv.h linux-4.9.28/drivers/block/zram/zram_drv.h ---- linux-4.9.28.orig/drivers/block/zram/zram_drv.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/block/zram/zram_drv.h 2017-05-19 03:37:25.138175108 +0200 -@@ -73,6 +73,9 @@ +diff --git a/drivers/block/zram/zram_drv.h b/drivers/block/zram/zram_drv.h +index 74fcf10da374..fd4020c99b9e 100644 +--- a/drivers/block/zram/zram_drv.h ++++ b/drivers/block/zram/zram_drv.h +@@ -73,6 +73,9 @@ enum zram_pageflags { struct zram_table_entry { unsigned long handle; unsigned long value; @@ -3519,7 +3638,7 @@ diff -Nur linux-4.9.28.orig/drivers/block/zram/zram_drv.h linux-4.9.28/drivers/b }; struct zram_stats { -@@ -120,4 +123,42 @@ +@@ -120,4 +123,42 @@ struct zram { */ bool claim; /* Protected by bdev->bd_mutex */ }; @@ -3562,9 +3681,10 @@ diff -Nur linux-4.9.28.orig/drivers/block/zram/zram_drv.h linux-4.9.28/drivers/b +#endif /* CONFIG_PREEMPT_RT_BASE */ + #endif -diff -Nur linux-4.9.28.orig/drivers/char/random.c linux-4.9.28/drivers/char/random.c ---- linux-4.9.28.orig/drivers/char/random.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/char/random.c 2017-05-19 03:37:25.138175108 +0200 +diff --git a/drivers/char/random.c b/drivers/char/random.c +index 08d1dd58c0d2..25ee319dc8e3 100644 +--- a/drivers/char/random.c ++++ b/drivers/char/random.c @@ -262,6 +262,7 @@ #include #include @@ -3573,7 +3693,7 @@ diff -Nur linux-4.9.28.orig/drivers/char/random.c linux-4.9.28/drivers/char/rand #include #include -@@ -1028,8 +1029,6 @@ +@@ -1028,8 +1029,6 @@ static void add_timer_randomness(struct timer_rand_state *state, unsigned num) } sample; long delta, delta2, delta3; @@ -3582,7 +3702,7 @@ diff -Nur linux-4.9.28.orig/drivers/char/random.c linux-4.9.28/drivers/char/rand sample.jiffies = jiffies; sample.cycles = random_get_entropy(); sample.num = num; -@@ -1070,7 +1069,6 @@ +@@ -1070,7 +1069,6 @@ static void add_timer_randomness(struct timer_rand_state *state, unsigned num) */ credit_entropy_bits(r, min_t(int, fls(delta>>1), 11)); } @@ -3590,7 +3710,7 @@ diff -Nur linux-4.9.28.orig/drivers/char/random.c linux-4.9.28/drivers/char/rand } void add_input_randomness(unsigned int type, unsigned int code, -@@ -1123,28 +1121,27 @@ +@@ -1123,28 +1121,27 @@ static __u32 get_reg(struct fast_pool *f, struct pt_regs *regs) return *(ptr + f->reg_idx++); } @@ -3624,7 +3744,7 @@ diff -Nur linux-4.9.28.orig/drivers/char/random.c linux-4.9.28/drivers/char/rand fast_mix(fast_pool); add_interrupt_bench(cycles); -@@ -2056,6 +2053,7 @@ +@@ -2056,6 +2053,7 @@ struct batched_entropy { * goal of being quite fast and not depleting entropy. */ static DEFINE_PER_CPU(struct batched_entropy, batched_entropy_long); @@ -3632,7 +3752,7 @@ diff -Nur linux-4.9.28.orig/drivers/char/random.c linux-4.9.28/drivers/char/rand unsigned long get_random_long(void) { unsigned long ret; -@@ -2064,13 +2062,13 @@ +@@ -2064,13 +2062,13 @@ unsigned long get_random_long(void) if (arch_get_random_long(&ret)) return ret; @@ -3648,7 +3768,7 @@ diff -Nur linux-4.9.28.orig/drivers/char/random.c linux-4.9.28/drivers/char/rand return ret; } EXPORT_SYMBOL(get_random_long); -@@ -2082,6 +2080,8 @@ +@@ -2082,6 +2080,8 @@ unsigned int get_random_int(void) } #else static DEFINE_PER_CPU(struct batched_entropy, batched_entropy_int); @@ -3657,7 +3777,7 @@ diff -Nur linux-4.9.28.orig/drivers/char/random.c linux-4.9.28/drivers/char/rand unsigned int get_random_int(void) { unsigned int ret; -@@ -2090,13 +2090,13 @@ +@@ -2090,13 +2090,13 @@ unsigned int get_random_int(void) if (arch_get_random_int(&ret)) return ret; @@ -3673,9 +3793,10 @@ diff -Nur linux-4.9.28.orig/drivers/char/random.c linux-4.9.28/drivers/char/rand return ret; } #endif -diff -Nur linux-4.9.28.orig/drivers/clocksource/tcb_clksrc.c linux-4.9.28/drivers/clocksource/tcb_clksrc.c ---- linux-4.9.28.orig/drivers/clocksource/tcb_clksrc.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/clocksource/tcb_clksrc.c 2017-05-19 03:37:25.138175108 +0200 +diff --git a/drivers/clocksource/tcb_clksrc.c b/drivers/clocksource/tcb_clksrc.c +index 4da2af9694a2..5b6f57f500b8 100644 +--- a/drivers/clocksource/tcb_clksrc.c ++++ b/drivers/clocksource/tcb_clksrc.c @@ -23,8 +23,7 @@ * this 32 bit free-running counter. the second channel is not used. * @@ -3686,7 +3807,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/tcb_clksrc.c linux-4.9.28/driver * * A boot clocksource and clockevent source are also currently needed, * unless the relevant platforms (ARM/AT91, AVR32/AT32) are changed so -@@ -74,6 +73,8 @@ +@@ -74,6 +73,8 @@ static struct clocksource clksrc = { struct tc_clkevt_device { struct clock_event_device clkevt; struct clk *clk; @@ -3695,7 +3816,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/tcb_clksrc.c linux-4.9.28/driver void __iomem *regs; }; -@@ -82,15 +83,26 @@ +@@ -82,15 +83,26 @@ static struct tc_clkevt_device *to_tc_clkevt(struct clock_event_device *clkevt) return container_of(clkevt, struct tc_clkevt_device, clkevt); } @@ -3729,7 +3850,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/tcb_clksrc.c linux-4.9.28/driver static int tc_shutdown(struct clock_event_device *d) { struct tc_clkevt_device *tcd = to_tc_clkevt(d); -@@ -98,8 +110,14 @@ +@@ -98,8 +110,14 @@ static int tc_shutdown(struct clock_event_device *d) __raw_writel(0xff, regs + ATMEL_TC_REG(2, IDR)); __raw_writel(ATMEL_TC_CLKDIS, regs + ATMEL_TC_REG(2, CCR)); @@ -3745,7 +3866,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/tcb_clksrc.c linux-4.9.28/driver return 0; } -@@ -112,9 +130,9 @@ +@@ -112,9 +130,9 @@ static int tc_set_oneshot(struct clock_event_device *d) if (clockevent_state_oneshot(d) || clockevent_state_periodic(d)) tc_shutdown(d); @@ -3757,7 +3878,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/tcb_clksrc.c linux-4.9.28/driver __raw_writel(timer_clock | ATMEL_TC_CPCSTOP | ATMEL_TC_WAVE | ATMEL_TC_WAVESEL_UP_AUTO, regs + ATMEL_TC_REG(2, CMR)); __raw_writel(ATMEL_TC_CPCS, regs + ATMEL_TC_REG(2, IER)); -@@ -134,12 +152,12 @@ +@@ -134,12 +152,12 @@ static int tc_set_periodic(struct clock_event_device *d) /* By not making the gentime core emulate periodic mode on top * of oneshot, we get lower overhead and improved accuracy. */ @@ -3773,7 +3894,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/tcb_clksrc.c linux-4.9.28/driver /* Enable clock and interrupts on RC compare */ __raw_writel(ATMEL_TC_CPCS, regs + ATMEL_TC_REG(2, IER)); -@@ -166,9 +184,13 @@ +@@ -166,9 +184,13 @@ static struct tc_clkevt_device clkevt = { .features = CLOCK_EVT_FEAT_PERIODIC | CLOCK_EVT_FEAT_ONESHOT, /* Should be lower than at91rm9200's system timer */ @@ -3788,7 +3909,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/tcb_clksrc.c linux-4.9.28/driver .set_state_periodic = tc_set_periodic, .set_state_oneshot = tc_set_oneshot, }, -@@ -188,8 +210,9 @@ +@@ -188,8 +210,9 @@ static irqreturn_t ch2_irq(int irq, void *handle) return IRQ_NONE; } @@ -3799,7 +3920,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/tcb_clksrc.c linux-4.9.28/driver int ret; struct clk *t2_clk = tc->clk[2]; int irq = tc->irq[2]; -@@ -210,7 +233,11 @@ +@@ -210,7 +233,11 @@ static int __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx) clkevt.regs = tc->regs; clkevt.clk = t2_clk; @@ -3812,7 +3933,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/tcb_clksrc.c linux-4.9.28/driver clkevt.clkevt.cpumask = cpumask_of(0); -@@ -221,7 +248,7 @@ +@@ -221,7 +248,7 @@ static int __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx) return ret; } @@ -3821,7 +3942,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/tcb_clksrc.c linux-4.9.28/driver return ret; } -@@ -358,7 +385,11 @@ +@@ -358,7 +385,11 @@ static int __init tcb_clksrc_init(void) goto err_disable_t1; /* channel 2: periodic and oneshot timer support */ @@ -3833,10 +3954,11 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/tcb_clksrc.c linux-4.9.28/driver if (ret) goto err_unregister_clksrc; -diff -Nur linux-4.9.28.orig/drivers/clocksource/timer-atmel-pit.c linux-4.9.28/drivers/clocksource/timer-atmel-pit.c ---- linux-4.9.28.orig/drivers/clocksource/timer-atmel-pit.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/clocksource/timer-atmel-pit.c 2017-05-19 03:37:25.138175108 +0200 -@@ -46,6 +46,7 @@ +diff --git a/drivers/clocksource/timer-atmel-pit.c b/drivers/clocksource/timer-atmel-pit.c +index 6555821bbdae..93288849b2bd 100644 +--- a/drivers/clocksource/timer-atmel-pit.c ++++ b/drivers/clocksource/timer-atmel-pit.c +@@ -46,6 +46,7 @@ struct pit_data { u32 cycle; u32 cnt; unsigned int irq; @@ -3844,7 +3966,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/timer-atmel-pit.c linux-4.9.28/d struct clk *mck; }; -@@ -96,15 +97,29 @@ +@@ -96,15 +97,29 @@ static int pit_clkevt_shutdown(struct clock_event_device *dev) /* disable irq, leaving the clocksource active */ pit_write(data->base, AT91_PIT_MR, (data->cycle - 1) | AT91_PIT_PITEN); @@ -3874,7 +3996,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/timer-atmel-pit.c linux-4.9.28/d /* update clocksource counter */ data->cnt += data->cycle * PIT_PICNT(pit_read(data->base, AT91_PIT_PIVR)); -@@ -230,15 +245,6 @@ +@@ -230,15 +245,6 @@ static int __init at91sam926x_pit_dt_init(struct device_node *node) return ret; } @@ -3890,10 +4012,11 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/timer-atmel-pit.c linux-4.9.28/d /* Set up and register clockevents */ data->clkevt.name = "pit"; data->clkevt.features = CLOCK_EVT_FEAT_PERIODIC; -diff -Nur linux-4.9.28.orig/drivers/clocksource/timer-atmel-st.c linux-4.9.28/drivers/clocksource/timer-atmel-st.c ---- linux-4.9.28.orig/drivers/clocksource/timer-atmel-st.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/clocksource/timer-atmel-st.c 2017-05-19 03:37:25.138175108 +0200 -@@ -115,18 +115,29 @@ +diff --git a/drivers/clocksource/timer-atmel-st.c b/drivers/clocksource/timer-atmel-st.c +index e90ab5b63a90..9e124087c55f 100644 +--- a/drivers/clocksource/timer-atmel-st.c ++++ b/drivers/clocksource/timer-atmel-st.c +@@ -115,18 +115,29 @@ static void clkdev32k_disable_and_flush_irq(void) last_crtr = read_CRTR(); } @@ -3923,7 +4046,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/timer-atmel-st.c linux-4.9.28/dr /* * ALM for oneshot irqs, set by next_event() * before 32 seconds have passed. -@@ -139,8 +150,16 @@ +@@ -139,8 +150,16 @@ static int clkevt32k_set_oneshot(struct clock_event_device *dev) static int clkevt32k_set_periodic(struct clock_event_device *dev) { @@ -3940,7 +4063,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/timer-atmel-st.c linux-4.9.28/dr /* PIT for periodic irqs; fixed rate of 1/HZ */ irqmask = AT91_ST_PITS; regmap_write(regmap_st, AT91_ST_PIMR, timer_latch); -@@ -198,7 +217,7 @@ +@@ -198,7 +217,7 @@ static int __init atmel_st_timer_init(struct device_node *node) { struct clk *sclk; unsigned int sclk_rate, val; @@ -3949,7 +4072,7 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/timer-atmel-st.c linux-4.9.28/dr regmap_st = syscon_node_to_regmap(node); if (IS_ERR(regmap_st)) { -@@ -212,21 +231,12 @@ +@@ -212,21 +231,12 @@ static int __init atmel_st_timer_init(struct device_node *node) regmap_read(regmap_st, AT91_ST_SR, &val); /* Get the interrupts property */ @@ -3973,9 +4096,10 @@ diff -Nur linux-4.9.28.orig/drivers/clocksource/timer-atmel-st.c linux-4.9.28/dr sclk = of_clk_get(node, 0); if (IS_ERR(sclk)) { pr_err("Unable to get slow clock\n"); -diff -Nur linux-4.9.28.orig/drivers/connector/cn_proc.c linux-4.9.28/drivers/connector/cn_proc.c ---- linux-4.9.28.orig/drivers/connector/cn_proc.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/connector/cn_proc.c 2017-05-19 03:37:25.138175108 +0200 +diff --git a/drivers/connector/cn_proc.c b/drivers/connector/cn_proc.c +index a782ce87715c..19d265948526 100644 +--- a/drivers/connector/cn_proc.c ++++ b/drivers/connector/cn_proc.c @@ -32,6 +32,7 @@ #include @@ -3984,7 +4108,7 @@ diff -Nur linux-4.9.28.orig/drivers/connector/cn_proc.c linux-4.9.28/drivers/con /* * Size of a cn_msg followed by a proc_event structure. Since the -@@ -54,10 +55,11 @@ +@@ -54,10 +55,11 @@ static struct cb_id cn_proc_event_id = { CN_IDX_PROC, CN_VAL_PROC }; /* proc_event_counts is used as the sequence number of the netlink message */ static DEFINE_PER_CPU(__u32, proc_event_counts) = { 0 }; @@ -3997,7 +4121,7 @@ diff -Nur linux-4.9.28.orig/drivers/connector/cn_proc.c linux-4.9.28/drivers/con msg->seq = __this_cpu_inc_return(proc_event_counts) - 1; ((struct proc_event *)msg->data)->cpu = smp_processor_id(); -@@ -70,7 +72,7 @@ +@@ -70,7 +72,7 @@ static inline void send_msg(struct cn_msg *msg) */ cn_netlink_send(msg, 0, CN_IDX_PROC, GFP_NOWAIT); @@ -4006,10 +4130,11 @@ diff -Nur linux-4.9.28.orig/drivers/connector/cn_proc.c linux-4.9.28/drivers/con } void proc_fork_connector(struct task_struct *task) -diff -Nur linux-4.9.28.orig/drivers/cpufreq/Kconfig.x86 linux-4.9.28/drivers/cpufreq/Kconfig.x86 ---- linux-4.9.28.orig/drivers/cpufreq/Kconfig.x86 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/cpufreq/Kconfig.x86 2017-05-19 03:37:25.138175108 +0200 -@@ -124,7 +124,7 @@ +diff --git a/drivers/cpufreq/Kconfig.x86 b/drivers/cpufreq/Kconfig.x86 +index adbd1de1cea5..1fac5074f2cf 100644 +--- a/drivers/cpufreq/Kconfig.x86 ++++ b/drivers/cpufreq/Kconfig.x86 +@@ -124,7 +124,7 @@ config X86_POWERNOW_K7_ACPI config X86_POWERNOW_K8 tristate "AMD Opteron/Athlon64 PowerNow!" @@ -4018,10 +4143,11 @@ diff -Nur linux-4.9.28.orig/drivers/cpufreq/Kconfig.x86 linux-4.9.28/drivers/cpu help This adds the CPUFreq driver for K8/early Opteron/Athlon64 processors. Support for K10 and newer processors is now in acpi-cpufreq. -diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/i915_gem_execbuffer.c linux-4.9.28/drivers/gpu/drm/i915/i915_gem_execbuffer.c ---- linux-4.9.28.orig/drivers/gpu/drm/i915/i915_gem_execbuffer.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/gpu/drm/i915/i915_gem_execbuffer.c 2017-05-19 03:37:25.138175108 +0200 -@@ -1489,7 +1489,9 @@ +diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c +index 2117f172d7a2..96c15501b0c8 100644 +--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c ++++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c +@@ -1489,7 +1489,9 @@ execbuf_submit(struct i915_execbuffer_params *params, if (ret) return ret; @@ -4031,10 +4157,11 @@ diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/i915_gem_execbuffer.c linux-4.9 i915_gem_execbuffer_move_to_active(vmas, params->request); -diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/i915_gem_shrinker.c linux-4.9.28/drivers/gpu/drm/i915/i915_gem_shrinker.c ---- linux-4.9.28.orig/drivers/gpu/drm/i915/i915_gem_shrinker.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/gpu/drm/i915/i915_gem_shrinker.c 2017-05-19 03:37:25.138175108 +0200 -@@ -40,7 +40,7 @@ +diff --git a/drivers/gpu/drm/i915/i915_gem_shrinker.c b/drivers/gpu/drm/i915/i915_gem_shrinker.c +index 755d78832a66..97fb03dc4971 100644 +--- a/drivers/gpu/drm/i915/i915_gem_shrinker.c ++++ b/drivers/gpu/drm/i915/i915_gem_shrinker.c +@@ -40,7 +40,7 @@ static bool mutex_is_locked_by(struct mutex *mutex, struct task_struct *task) if (!mutex_is_locked(mutex)) return false; @@ -4043,10 +4170,11 @@ diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/i915_gem_shrinker.c linux-4.9.2 return mutex->owner == task; #else /* Since UP may be pre-empted, we cannot assume that we own the lock */ -diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/i915_irq.c linux-4.9.28/drivers/gpu/drm/i915/i915_irq.c ---- linux-4.9.28.orig/drivers/gpu/drm/i915/i915_irq.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/gpu/drm/i915/i915_irq.c 2017-05-19 03:37:25.142175268 +0200 -@@ -812,6 +812,7 @@ +diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c +index 02908e37c228..05c0480576e1 100644 +--- a/drivers/gpu/drm/i915/i915_irq.c ++++ b/drivers/gpu/drm/i915/i915_irq.c +@@ -812,6 +812,7 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); /* preempt_disable_rt() should go right here in PREEMPT_RT patchset. */ @@ -4054,7 +4182,7 @@ diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/i915_irq.c linux-4.9.28/drivers /* Get optional system timestamp before query. */ if (stime) -@@ -863,6 +864,7 @@ +@@ -863,6 +864,7 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, *etime = ktime_get(); /* preempt_enable_rt() should go right here in PREEMPT_RT patchset. */ @@ -4062,10 +4190,11 @@ diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/i915_irq.c linux-4.9.28/drivers spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); -diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/intel_display.c linux-4.9.28/drivers/gpu/drm/i915/intel_display.c ---- linux-4.9.28.orig/drivers/gpu/drm/i915/intel_display.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/gpu/drm/i915/intel_display.c 2017-05-19 03:37:25.142175268 +0200 -@@ -12131,7 +12131,7 @@ +diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c +index 5dc6082639db..c32458fb3be2 100644 +--- a/drivers/gpu/drm/i915/intel_display.c ++++ b/drivers/gpu/drm/i915/intel_display.c +@@ -12131,7 +12131,7 @@ void intel_check_page_flip(struct drm_i915_private *dev_priv, int pipe) struct intel_crtc *intel_crtc = to_intel_crtc(crtc); struct intel_flip_work *work; @@ -4074,9 +4203,10 @@ diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/intel_display.c linux-4.9.28/dr if (crtc == NULL) return; -diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/intel_sprite.c linux-4.9.28/drivers/gpu/drm/i915/intel_sprite.c ---- linux-4.9.28.orig/drivers/gpu/drm/i915/intel_sprite.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/gpu/drm/i915/intel_sprite.c 2017-05-19 03:37:25.142175268 +0200 +diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c +index dbed12c484c9..5c540b78e8b5 100644 +--- a/drivers/gpu/drm/i915/intel_sprite.c ++++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -35,6 +35,7 @@ #include #include @@ -4085,7 +4215,7 @@ diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/intel_sprite.c linux-4.9.28/dri #include "intel_drv.h" #include "intel_frontbuffer.h" #include -@@ -65,6 +66,8 @@ +@@ -65,6 +66,8 @@ int intel_usecs_to_scanlines(const struct drm_display_mode *adjusted_mode, 1000 * adjusted_mode->crtc_htotal); } @@ -4094,7 +4224,7 @@ diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/intel_sprite.c linux-4.9.28/dri /** * intel_pipe_update_start() - start update of a set of display registers * @crtc: the crtc of which the registers are going to be updated -@@ -95,7 +98,7 @@ +@@ -95,7 +98,7 @@ void intel_pipe_update_start(struct intel_crtc *crtc) min = vblank_start - intel_usecs_to_scanlines(adjusted_mode, 100); max = vblank_start - 1; @@ -4103,7 +4233,7 @@ diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/intel_sprite.c linux-4.9.28/dri if (min <= 0 || max <= 0) return; -@@ -125,11 +128,11 @@ +@@ -125,11 +128,11 @@ void intel_pipe_update_start(struct intel_crtc *crtc) break; } @@ -4117,7 +4247,7 @@ diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/intel_sprite.c linux-4.9.28/dri } finish_wait(wq, &wait); -@@ -181,7 +184,7 @@ +@@ -181,7 +184,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work crtc->base.state->event = NULL; } @@ -4126,10 +4256,11 @@ diff -Nur linux-4.9.28.orig/drivers/gpu/drm/i915/intel_sprite.c linux-4.9.28/dri if (crtc->debug.start_vbl_count && crtc->debug.start_vbl_count != end_vbl_count) { -diff -Nur linux-4.9.28.orig/drivers/gpu/drm/msm/msm_gem_shrinker.c linux-4.9.28/drivers/gpu/drm/msm/msm_gem_shrinker.c ---- linux-4.9.28.orig/drivers/gpu/drm/msm/msm_gem_shrinker.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/gpu/drm/msm/msm_gem_shrinker.c 2017-05-19 03:37:25.142175268 +0200 -@@ -23,7 +23,7 @@ +diff --git a/drivers/gpu/drm/msm/msm_gem_shrinker.c b/drivers/gpu/drm/msm/msm_gem_shrinker.c +index 192b2d3a79cb..d5372a207326 100644 +--- a/drivers/gpu/drm/msm/msm_gem_shrinker.c ++++ b/drivers/gpu/drm/msm/msm_gem_shrinker.c +@@ -23,7 +23,7 @@ static bool mutex_is_locked_by(struct mutex *mutex, struct task_struct *task) if (!mutex_is_locked(mutex)) return false; @@ -4138,10 +4269,11 @@ diff -Nur linux-4.9.28.orig/drivers/gpu/drm/msm/msm_gem_shrinker.c linux-4.9.28/ return mutex->owner == task; #else /* Since UP may be pre-empted, we cannot assume that we own the lock */ -diff -Nur linux-4.9.28.orig/drivers/gpu/drm/radeon/radeon_display.c linux-4.9.28/drivers/gpu/drm/radeon/radeon_display.c ---- linux-4.9.28.orig/drivers/gpu/drm/radeon/radeon_display.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/gpu/drm/radeon/radeon_display.c 2017-05-19 03:37:25.142175268 +0200 -@@ -1845,6 +1845,7 @@ +diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c +index cdb8cb568c15..b6d7fd964cbc 100644 +--- a/drivers/gpu/drm/radeon/radeon_display.c ++++ b/drivers/gpu/drm/radeon/radeon_display.c +@@ -1845,6 +1845,7 @@ int radeon_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, struct radeon_device *rdev = dev->dev_private; /* preempt_disable_rt() should go right here in PREEMPT_RT patchset. */ @@ -4149,7 +4281,7 @@ diff -Nur linux-4.9.28.orig/drivers/gpu/drm/radeon/radeon_display.c linux-4.9.28 /* Get optional system timestamp before query. */ if (stime) -@@ -1937,6 +1938,7 @@ +@@ -1937,6 +1938,7 @@ int radeon_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, *etime = ktime_get(); /* preempt_enable_rt() should go right here in PREEMPT_RT patchset. */ @@ -4157,10 +4289,11 @@ diff -Nur linux-4.9.28.orig/drivers/gpu/drm/radeon/radeon_display.c linux-4.9.28 /* Decode into vertical and horizontal scanout position. */ *vpos = position & 0x1fff; -diff -Nur linux-4.9.28.orig/drivers/hv/vmbus_drv.c linux-4.9.28/drivers/hv/vmbus_drv.c ---- linux-4.9.28.orig/drivers/hv/vmbus_drv.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/hv/vmbus_drv.c 2017-05-19 03:37:25.142175268 +0200 -@@ -761,6 +761,8 @@ +diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c +index 0276d2ef06ee..8868045eabde 100644 +--- a/drivers/hv/vmbus_drv.c ++++ b/drivers/hv/vmbus_drv.c +@@ -761,6 +761,8 @@ static void vmbus_isr(void) void *page_addr; struct hv_message *msg; union hv_synic_event_flags *event; @@ -4169,7 +4302,7 @@ diff -Nur linux-4.9.28.orig/drivers/hv/vmbus_drv.c linux-4.9.28/drivers/hv/vmbus bool handled = false; page_addr = hv_context.synic_event_page[cpu]; -@@ -808,7 +810,7 @@ +@@ -808,7 +810,7 @@ static void vmbus_isr(void) tasklet_schedule(hv_context.msg_dpc[cpu]); } @@ -4178,10 +4311,11 @@ diff -Nur linux-4.9.28.orig/drivers/hv/vmbus_drv.c linux-4.9.28/drivers/hv/vmbus } -diff -Nur linux-4.9.28.orig/drivers/ide/alim15x3.c linux-4.9.28/drivers/ide/alim15x3.c ---- linux-4.9.28.orig/drivers/ide/alim15x3.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/ide/alim15x3.c 2017-05-19 03:37:25.142175268 +0200 -@@ -234,7 +234,7 @@ +diff --git a/drivers/ide/alim15x3.c b/drivers/ide/alim15x3.c +index 36f76e28a0bf..394f142f90c7 100644 +--- a/drivers/ide/alim15x3.c ++++ b/drivers/ide/alim15x3.c +@@ -234,7 +234,7 @@ static int init_chipset_ali15x3(struct pci_dev *dev) isa_dev = pci_get_device(PCI_VENDOR_ID_AL, PCI_DEVICE_ID_AL_M1533, NULL); @@ -4190,7 +4324,7 @@ diff -Nur linux-4.9.28.orig/drivers/ide/alim15x3.c linux-4.9.28/drivers/ide/alim if (m5229_revision < 0xC2) { /* -@@ -325,7 +325,7 @@ +@@ -325,7 +325,7 @@ static int init_chipset_ali15x3(struct pci_dev *dev) } pci_dev_put(north); pci_dev_put(isa_dev); @@ -4199,10 +4333,11 @@ diff -Nur linux-4.9.28.orig/drivers/ide/alim15x3.c linux-4.9.28/drivers/ide/alim return 0; } -diff -Nur linux-4.9.28.orig/drivers/ide/hpt366.c linux-4.9.28/drivers/ide/hpt366.c ---- linux-4.9.28.orig/drivers/ide/hpt366.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/ide/hpt366.c 2017-05-19 03:37:25.142175268 +0200 -@@ -1236,7 +1236,7 @@ +diff --git a/drivers/ide/hpt366.c b/drivers/ide/hpt366.c +index 0ceae5cbd89a..c212e85d7f3e 100644 +--- a/drivers/ide/hpt366.c ++++ b/drivers/ide/hpt366.c +@@ -1236,7 +1236,7 @@ static int init_dma_hpt366(ide_hwif_t *hwif, dma_old = inb(base + 2); @@ -4211,7 +4346,7 @@ diff -Nur linux-4.9.28.orig/drivers/ide/hpt366.c linux-4.9.28/drivers/ide/hpt366 dma_new = dma_old; pci_read_config_byte(dev, hwif->channel ? 0x4b : 0x43, &masterdma); -@@ -1247,7 +1247,7 @@ +@@ -1247,7 +1247,7 @@ static int init_dma_hpt366(ide_hwif_t *hwif, if (dma_new != dma_old) outb(dma_new, base + 2); @@ -4220,40 +4355,11 @@ diff -Nur linux-4.9.28.orig/drivers/ide/hpt366.c linux-4.9.28/drivers/ide/hpt366 printk(KERN_INFO " %s: BM-DMA at 0x%04lx-0x%04lx\n", hwif->name, base, base + 7); -diff -Nur linux-4.9.28.orig/drivers/ide/ide-io.c linux-4.9.28/drivers/ide/ide-io.c ---- linux-4.9.28.orig/drivers/ide/ide-io.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/ide/ide-io.c 2017-05-19 03:37:25.146175423 +0200 -@@ -659,7 +659,7 @@ - /* disable_irq_nosync ?? */ - disable_irq(hwif->irq); - /* local CPU only, as if we were handling an interrupt */ -- local_irq_disable(); -+ local_irq_disable_nort(); - if (hwif->polling) { - startstop = handler(drive); - } else if (drive_is_ready(drive)) { -diff -Nur linux-4.9.28.orig/drivers/ide/ide-iops.c linux-4.9.28/drivers/ide/ide-iops.c ---- linux-4.9.28.orig/drivers/ide/ide-iops.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/ide/ide-iops.c 2017-05-19 03:37:25.146175423 +0200 -@@ -129,12 +129,12 @@ - if ((stat & ATA_BUSY) == 0) - break; - -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - *rstat = stat; - return -EBUSY; - } - } -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - } - /* - * Allow status to settle, then read it again. -diff -Nur linux-4.9.28.orig/drivers/ide/ide-io-std.c linux-4.9.28/drivers/ide/ide-io-std.c ---- linux-4.9.28.orig/drivers/ide/ide-io-std.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/ide/ide-io-std.c 2017-05-19 03:37:25.142175268 +0200 -@@ -175,7 +175,7 @@ +diff --git a/drivers/ide/ide-io-std.c b/drivers/ide/ide-io-std.c +index 19763977568c..4169433faab5 100644 +--- a/drivers/ide/ide-io-std.c ++++ b/drivers/ide/ide-io-std.c +@@ -175,7 +175,7 @@ void ide_input_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf, unsigned long uninitialized_var(flags); if ((io_32bit & 2) && !mmio) { @@ -4262,7 +4368,7 @@ diff -Nur linux-4.9.28.orig/drivers/ide/ide-io-std.c linux-4.9.28/drivers/ide/id ata_vlb_sync(io_ports->nsect_addr); } -@@ -186,7 +186,7 @@ +@@ -186,7 +186,7 @@ void ide_input_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf, insl(data_addr, buf, words); if ((io_32bit & 2) && !mmio) @@ -4271,7 +4377,7 @@ diff -Nur linux-4.9.28.orig/drivers/ide/ide-io-std.c linux-4.9.28/drivers/ide/id if (((len + 1) & 3) < 2) return; -@@ -219,7 +219,7 @@ +@@ -219,7 +219,7 @@ void ide_output_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf, unsigned long uninitialized_var(flags); if ((io_32bit & 2) && !mmio) { @@ -4280,7 +4386,7 @@ diff -Nur linux-4.9.28.orig/drivers/ide/ide-io-std.c linux-4.9.28/drivers/ide/id ata_vlb_sync(io_ports->nsect_addr); } -@@ -230,7 +230,7 @@ +@@ -230,7 +230,7 @@ void ide_output_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf, outsl(data_addr, buf, words); if ((io_32bit & 2) && !mmio) @@ -4289,10 +4395,43 @@ diff -Nur linux-4.9.28.orig/drivers/ide/ide-io-std.c linux-4.9.28/drivers/ide/id if (((len + 1) & 3) < 2) return; -diff -Nur linux-4.9.28.orig/drivers/ide/ide-probe.c linux-4.9.28/drivers/ide/ide-probe.c ---- linux-4.9.28.orig/drivers/ide/ide-probe.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/ide/ide-probe.c 2017-05-19 03:37:25.146175423 +0200 -@@ -196,10 +196,10 @@ +diff --git a/drivers/ide/ide-io.c b/drivers/ide/ide-io.c +index 669ea1e45795..e12e43e62245 100644 +--- a/drivers/ide/ide-io.c ++++ b/drivers/ide/ide-io.c +@@ -659,7 +659,7 @@ void ide_timer_expiry (unsigned long data) + /* disable_irq_nosync ?? */ + disable_irq(hwif->irq); + /* local CPU only, as if we were handling an interrupt */ +- local_irq_disable(); ++ local_irq_disable_nort(); + if (hwif->polling) { + startstop = handler(drive); + } else if (drive_is_ready(drive)) { +diff --git a/drivers/ide/ide-iops.c b/drivers/ide/ide-iops.c +index 376f2dc410c5..f014dd1b73dc 100644 +--- a/drivers/ide/ide-iops.c ++++ b/drivers/ide/ide-iops.c +@@ -129,12 +129,12 @@ int __ide_wait_stat(ide_drive_t *drive, u8 good, u8 bad, + if ((stat & ATA_BUSY) == 0) + break; + +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + *rstat = stat; + return -EBUSY; + } + } +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + /* + * Allow status to settle, then read it again. +diff --git a/drivers/ide/ide-probe.c b/drivers/ide/ide-probe.c +index 0b63facd1d87..4ceba37afc0c 100644 +--- a/drivers/ide/ide-probe.c ++++ b/drivers/ide/ide-probe.c +@@ -196,10 +196,10 @@ static void do_identify(ide_drive_t *drive, u8 cmd, u16 *id) int bswap = 1; /* local CPU only; some systems need this */ @@ -4305,10 +4444,11 @@ diff -Nur linux-4.9.28.orig/drivers/ide/ide-probe.c linux-4.9.28/drivers/ide/ide drive->dev_flags |= IDE_DFLAG_ID_READ; #ifdef DEBUG -diff -Nur linux-4.9.28.orig/drivers/ide/ide-taskfile.c linux-4.9.28/drivers/ide/ide-taskfile.c ---- linux-4.9.28.orig/drivers/ide/ide-taskfile.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/ide/ide-taskfile.c 2017-05-19 03:37:25.146175423 +0200 -@@ -250,7 +250,7 @@ +diff --git a/drivers/ide/ide-taskfile.c b/drivers/ide/ide-taskfile.c +index a716693417a3..be0568c722d6 100644 +--- a/drivers/ide/ide-taskfile.c ++++ b/drivers/ide/ide-taskfile.c +@@ -250,7 +250,7 @@ void ide_pio_bytes(ide_drive_t *drive, struct ide_cmd *cmd, page_is_high = PageHighMem(page); if (page_is_high) @@ -4317,7 +4457,7 @@ diff -Nur linux-4.9.28.orig/drivers/ide/ide-taskfile.c linux-4.9.28/drivers/ide/ buf = kmap_atomic(page) + offset; -@@ -271,7 +271,7 @@ +@@ -271,7 +271,7 @@ void ide_pio_bytes(ide_drive_t *drive, struct ide_cmd *cmd, kunmap_atomic(buf); if (page_is_high) @@ -4326,7 +4466,7 @@ diff -Nur linux-4.9.28.orig/drivers/ide/ide-taskfile.c linux-4.9.28/drivers/ide/ len -= nr_bytes; } -@@ -414,7 +414,7 @@ +@@ -414,7 +414,7 @@ static ide_startstop_t pre_task_out_intr(ide_drive_t *drive, } if ((drive->dev_flags & IDE_DFLAG_UNMASK) == 0) @@ -4335,10 +4475,11 @@ diff -Nur linux-4.9.28.orig/drivers/ide/ide-taskfile.c linux-4.9.28/drivers/ide/ ide_set_handler(drive, &task_pio_intr, WAIT_WORSTCASE); -diff -Nur linux-4.9.28.orig/drivers/infiniband/ulp/ipoib/ipoib_multicast.c linux-4.9.28/drivers/infiniband/ulp/ipoib/ipoib_multicast.c ---- linux-4.9.28.orig/drivers/infiniband/ulp/ipoib/ipoib_multicast.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/infiniband/ulp/ipoib/ipoib_multicast.c 2017-05-19 03:37:25.146175423 +0200 -@@ -902,7 +902,7 @@ +diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +index fddff403d5d2..cca1bb4fbfe3 100644 +--- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c ++++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +@@ -902,7 +902,7 @@ void ipoib_mcast_restart_task(struct work_struct *work) ipoib_dbg_mcast(priv, "restarting multicast task\n"); @@ -4347,7 +4488,7 @@ diff -Nur linux-4.9.28.orig/drivers/infiniband/ulp/ipoib/ipoib_multicast.c linux netif_addr_lock(dev); spin_lock(&priv->lock); -@@ -984,7 +984,7 @@ +@@ -984,7 +984,7 @@ void ipoib_mcast_restart_task(struct work_struct *work) spin_unlock(&priv->lock); netif_addr_unlock(dev); @@ -4356,10 +4497,11 @@ diff -Nur linux-4.9.28.orig/drivers/infiniband/ulp/ipoib/ipoib_multicast.c linux /* * make sure the in-flight joins have finished before we attempt -diff -Nur linux-4.9.28.orig/drivers/input/gameport/gameport.c linux-4.9.28/drivers/input/gameport/gameport.c ---- linux-4.9.28.orig/drivers/input/gameport/gameport.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/input/gameport/gameport.c 2017-05-19 03:37:25.146175423 +0200 -@@ -91,13 +91,13 @@ +diff --git a/drivers/input/gameport/gameport.c b/drivers/input/gameport/gameport.c +index 4a2a9e370be7..e970d9afd179 100644 +--- a/drivers/input/gameport/gameport.c ++++ b/drivers/input/gameport/gameport.c +@@ -91,13 +91,13 @@ static int gameport_measure_speed(struct gameport *gameport) tx = ~0; for (i = 0; i < 50; i++) { @@ -4375,7 +4517,7 @@ diff -Nur linux-4.9.28.orig/drivers/input/gameport/gameport.c linux-4.9.28/drive udelay(i * 10); t = (t2 - t1) - (t3 - t2); if (t < tx) -@@ -124,12 +124,12 @@ +@@ -124,12 +124,12 @@ static int old_gameport_measure_speed(struct gameport *gameport) tx = 1 << 30; for(i = 0; i < 50; i++) { @@ -4390,7 +4532,7 @@ diff -Nur linux-4.9.28.orig/drivers/input/gameport/gameport.c linux-4.9.28/drive udelay(i * 10); if ((t = DELTA(t2,t1) - DELTA(t3,t2)) < tx) tx = t; } -@@ -148,11 +148,11 @@ +@@ -148,11 +148,11 @@ static int old_gameport_measure_speed(struct gameport *gameport) tx = 1 << 30; for(i = 0; i < 50; i++) { @@ -4404,10 +4546,11 @@ diff -Nur linux-4.9.28.orig/drivers/input/gameport/gameport.c linux-4.9.28/drive udelay(i * 10); if (t2 - t1 < tx) tx = t2 - t1; } -diff -Nur linux-4.9.28.orig/drivers/iommu/amd_iommu.c linux-4.9.28/drivers/iommu/amd_iommu.c ---- linux-4.9.28.orig/drivers/iommu/amd_iommu.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/iommu/amd_iommu.c 2017-05-19 03:37:25.146175423 +0200 -@@ -1923,10 +1923,10 @@ +diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c +index 11a13b5be73a..baaed0ac274b 100644 +--- a/drivers/iommu/amd_iommu.c ++++ b/drivers/iommu/amd_iommu.c +@@ -1923,10 +1923,10 @@ static int __attach_device(struct iommu_dev_data *dev_data, int ret; /* @@ -4421,7 +4564,7 @@ diff -Nur linux-4.9.28.orig/drivers/iommu/amd_iommu.c linux-4.9.28/drivers/iommu /* lock domain */ spin_lock(&domain->lock); -@@ -2094,10 +2094,10 @@ +@@ -2094,10 +2094,10 @@ static void __detach_device(struct iommu_dev_data *dev_data) struct protection_domain *domain; /* @@ -4435,10 +4578,11 @@ diff -Nur linux-4.9.28.orig/drivers/iommu/amd_iommu.c linux-4.9.28/drivers/iommu if (WARN_ON(!dev_data->domain)) return; -diff -Nur linux-4.9.28.orig/drivers/iommu/intel-iommu.c linux-4.9.28/drivers/iommu/intel-iommu.c ---- linux-4.9.28.orig/drivers/iommu/intel-iommu.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/iommu/intel-iommu.c 2017-05-19 03:37:25.146175423 +0200 -@@ -479,7 +479,7 @@ +diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c +index 87fcbf71b85a..674c82b61f36 100644 +--- a/drivers/iommu/intel-iommu.c ++++ b/drivers/iommu/intel-iommu.c +@@ -479,7 +479,7 @@ struct deferred_flush_data { struct deferred_flush_table *tables; }; @@ -4447,7 +4591,7 @@ diff -Nur linux-4.9.28.orig/drivers/iommu/intel-iommu.c linux-4.9.28/drivers/iom /* bitmap for indexing intel_iommus */ static int g_num_of_iommus; -@@ -3716,10 +3716,8 @@ +@@ -3719,10 +3719,8 @@ static void add_unmap(struct dmar_domain *dom, unsigned long iova_pfn, struct intel_iommu *iommu; struct deferred_flush_entry *entry; struct deferred_flush_data *flush_data; @@ -4459,7 +4603,7 @@ diff -Nur linux-4.9.28.orig/drivers/iommu/intel-iommu.c linux-4.9.28/drivers/iom /* Flush all CPUs' entries to avoid deferring too much. If * this becomes a bottleneck, can just flush us, and rely on -@@ -3752,8 +3750,6 @@ +@@ -3755,8 +3753,6 @@ static void add_unmap(struct dmar_domain *dom, unsigned long iova_pfn, } flush_data->size++; spin_unlock_irqrestore(&flush_data->lock, flags); @@ -4468,9 +4612,10 @@ diff -Nur linux-4.9.28.orig/drivers/iommu/intel-iommu.c linux-4.9.28/drivers/iom } static void intel_unmap(struct device *dev, dma_addr_t dev_addr, size_t size) -diff -Nur linux-4.9.28.orig/drivers/iommu/iova.c linux-4.9.28/drivers/iommu/iova.c ---- linux-4.9.28.orig/drivers/iommu/iova.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/iommu/iova.c 2017-05-19 03:37:25.146175423 +0200 +diff --git a/drivers/iommu/iova.c b/drivers/iommu/iova.c +index e23001bfcfee..359d5d169ec0 100644 +--- a/drivers/iommu/iova.c ++++ b/drivers/iommu/iova.c @@ -22,6 +22,7 @@ #include #include @@ -4479,7 +4624,7 @@ diff -Nur linux-4.9.28.orig/drivers/iommu/iova.c linux-4.9.28/drivers/iommu/iova static bool iova_rcache_insert(struct iova_domain *iovad, unsigned long pfn, -@@ -420,10 +421,8 @@ +@@ -420,10 +421,8 @@ alloc_iova_fast(struct iova_domain *iovad, unsigned long size, /* Try replenishing IOVAs by flushing rcache. */ flushed_rcache = true; @@ -4490,7 +4635,7 @@ diff -Nur linux-4.9.28.orig/drivers/iommu/iova.c linux-4.9.28/drivers/iommu/iova goto retry; } -@@ -751,7 +750,7 @@ +@@ -751,7 +750,7 @@ static bool __iova_rcache_insert(struct iova_domain *iovad, bool can_insert = false; unsigned long flags; @@ -4499,7 +4644,7 @@ diff -Nur linux-4.9.28.orig/drivers/iommu/iova.c linux-4.9.28/drivers/iommu/iova spin_lock_irqsave(&cpu_rcache->lock, flags); if (!iova_magazine_full(cpu_rcache->loaded)) { -@@ -781,7 +780,6 @@ +@@ -781,7 +780,6 @@ static bool __iova_rcache_insert(struct iova_domain *iovad, iova_magazine_push(cpu_rcache->loaded, iova_pfn); spin_unlock_irqrestore(&cpu_rcache->lock, flags); @@ -4507,7 +4652,7 @@ diff -Nur linux-4.9.28.orig/drivers/iommu/iova.c linux-4.9.28/drivers/iommu/iova if (mag_to_free) { iova_magazine_free_pfns(mag_to_free, iovad); -@@ -815,7 +813,7 @@ +@@ -815,7 +813,7 @@ static unsigned long __iova_rcache_get(struct iova_rcache *rcache, bool has_pfn = false; unsigned long flags; @@ -4516,7 +4661,7 @@ diff -Nur linux-4.9.28.orig/drivers/iommu/iova.c linux-4.9.28/drivers/iommu/iova spin_lock_irqsave(&cpu_rcache->lock, flags); if (!iova_magazine_empty(cpu_rcache->loaded)) { -@@ -837,7 +835,6 @@ +@@ -837,7 +835,6 @@ static unsigned long __iova_rcache_get(struct iova_rcache *rcache, iova_pfn = iova_magazine_pop(cpu_rcache->loaded, limit_pfn); spin_unlock_irqrestore(&cpu_rcache->lock, flags); @@ -4524,10 +4669,11 @@ diff -Nur linux-4.9.28.orig/drivers/iommu/iova.c linux-4.9.28/drivers/iommu/iova return iova_pfn; } -diff -Nur linux-4.9.28.orig/drivers/leds/trigger/Kconfig linux-4.9.28/drivers/leds/trigger/Kconfig ---- linux-4.9.28.orig/drivers/leds/trigger/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/leds/trigger/Kconfig 2017-05-19 03:37:25.146175423 +0200 -@@ -69,7 +69,7 @@ +diff --git a/drivers/leds/trigger/Kconfig b/drivers/leds/trigger/Kconfig +index 3f9ddb9fafa7..09da5b6b44a1 100644 +--- a/drivers/leds/trigger/Kconfig ++++ b/drivers/leds/trigger/Kconfig +@@ -69,7 +69,7 @@ config LEDS_TRIGGER_BACKLIGHT config LEDS_TRIGGER_CPU bool "LED CPU Trigger" @@ -4536,9 +4682,10 @@ diff -Nur linux-4.9.28.orig/drivers/leds/trigger/Kconfig linux-4.9.28/drivers/le help This allows LEDs to be controlled by active CPUs. This shows the active CPUs across an array of LEDs so you can see which -diff -Nur linux-4.9.28.orig/drivers/md/bcache/Kconfig linux-4.9.28/drivers/md/bcache/Kconfig ---- linux-4.9.28.orig/drivers/md/bcache/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/md/bcache/Kconfig 2017-05-19 03:37:25.146175423 +0200 +diff --git a/drivers/md/bcache/Kconfig b/drivers/md/bcache/Kconfig +index 4d200883c505..98b64ed5cb81 100644 +--- a/drivers/md/bcache/Kconfig ++++ b/drivers/md/bcache/Kconfig @@ -1,6 +1,7 @@ config BCACHE @@ -4547,10 +4694,11 @@ diff -Nur linux-4.9.28.orig/drivers/md/bcache/Kconfig linux-4.9.28/drivers/md/bc ---help--- Allows a block device to be used as cache for other devices; uses a btree for indexing and the layout is optimized for SSDs. -diff -Nur linux-4.9.28.orig/drivers/md/dm-rq.c linux-4.9.28/drivers/md/dm-rq.c ---- linux-4.9.28.orig/drivers/md/dm-rq.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/md/dm-rq.c 2017-05-19 03:37:25.146175423 +0200 -@@ -842,7 +842,7 @@ +diff --git a/drivers/md/dm-rq.c b/drivers/md/dm-rq.c +index ba7c4c685db3..834ec328f217 100644 +--- a/drivers/md/dm-rq.c ++++ b/drivers/md/dm-rq.c +@@ -842,7 +842,7 @@ static void dm_old_request_fn(struct request_queue *q) /* Establish tio->ti before queuing work (map_tio_request) */ tio->ti = ti; kthread_queue_work(&md->kworker, &tio->work); @@ -4559,10 +4707,11 @@ diff -Nur linux-4.9.28.orig/drivers/md/dm-rq.c linux-4.9.28/drivers/md/dm-rq.c } } -diff -Nur linux-4.9.28.orig/drivers/md/raid5.c linux-4.9.28/drivers/md/raid5.c ---- linux-4.9.28.orig/drivers/md/raid5.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/md/raid5.c 2017-05-19 03:37:25.146175423 +0200 -@@ -1928,8 +1928,9 @@ +diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c +index f34ad2be66a1..123469f7b560 100644 +--- a/drivers/md/raid5.c ++++ b/drivers/md/raid5.c +@@ -1928,8 +1928,9 @@ static void raid_run_ops(struct stripe_head *sh, unsigned long ops_request) struct raid5_percpu *percpu; unsigned long cpu; @@ -4573,7 +4722,7 @@ diff -Nur linux-4.9.28.orig/drivers/md/raid5.c linux-4.9.28/drivers/md/raid5.c if (test_bit(STRIPE_OP_BIOFILL, &ops_request)) { ops_run_biofill(sh); overlap_clear++; -@@ -1985,7 +1986,8 @@ +@@ -1985,7 +1986,8 @@ static void raid_run_ops(struct stripe_head *sh, unsigned long ops_request) if (test_and_clear_bit(R5_Overlap, &dev->flags)) wake_up(&sh->raid_conf->wait_for_overlap); } @@ -4583,7 +4732,7 @@ diff -Nur linux-4.9.28.orig/drivers/md/raid5.c linux-4.9.28/drivers/md/raid5.c } static struct stripe_head *alloc_stripe(struct kmem_cache *sc, gfp_t gfp, -@@ -6391,6 +6393,7 @@ +@@ -6393,6 +6395,7 @@ static int raid456_cpu_up_prepare(unsigned int cpu, struct hlist_node *node) __func__, cpu); return -ENOMEM; } @@ -4591,7 +4740,7 @@ diff -Nur linux-4.9.28.orig/drivers/md/raid5.c linux-4.9.28/drivers/md/raid5.c return 0; } -@@ -6401,7 +6404,6 @@ +@@ -6403,7 +6406,6 @@ static int raid5_alloc_percpu(struct r5conf *conf) conf->percpu = alloc_percpu(struct raid5_percpu); if (!conf->percpu) return -ENOMEM; @@ -4599,10 +4748,11 @@ diff -Nur linux-4.9.28.orig/drivers/md/raid5.c linux-4.9.28/drivers/md/raid5.c err = cpuhp_state_add_instance(CPUHP_MD_RAID5_PREPARE, &conf->node); if (!err) { conf->scribble_disks = max(conf->raid_disks, -diff -Nur linux-4.9.28.orig/drivers/md/raid5.h linux-4.9.28/drivers/md/raid5.h ---- linux-4.9.28.orig/drivers/md/raid5.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/md/raid5.h 2017-05-19 03:37:25.150175574 +0200 -@@ -504,6 +504,7 @@ +diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h +index 57ec49f0839e..0739604990b7 100644 +--- a/drivers/md/raid5.h ++++ b/drivers/md/raid5.h +@@ -504,6 +504,7 @@ struct r5conf { int recovery_disabled; /* per cpu variables */ struct raid5_percpu { @@ -4610,10 +4760,11 @@ diff -Nur linux-4.9.28.orig/drivers/md/raid5.h linux-4.9.28/drivers/md/raid5.h struct page *spare_page; /* Used when checking P/Q in raid6 */ struct flex_array *scribble; /* space for constructing buffer * lists and performing address -diff -Nur linux-4.9.28.orig/drivers/misc/Kconfig linux-4.9.28/drivers/misc/Kconfig ---- linux-4.9.28.orig/drivers/misc/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/misc/Kconfig 2017-05-19 03:37:25.150175574 +0200 -@@ -54,6 +54,7 @@ +diff --git a/drivers/misc/Kconfig b/drivers/misc/Kconfig +index 64971baf11fa..215e91e36198 100644 +--- a/drivers/misc/Kconfig ++++ b/drivers/misc/Kconfig +@@ -54,6 +54,7 @@ config AD525X_DPOT_SPI config ATMEL_TCLIB bool "Atmel AT32/AT91 Timer/Counter Library" depends on (AVR32 || ARCH_AT91) @@ -4621,7 +4772,7 @@ diff -Nur linux-4.9.28.orig/drivers/misc/Kconfig linux-4.9.28/drivers/misc/Kconf help Select this if you want a library to allocate the Timer/Counter blocks found on many Atmel processors. This facilitates using -@@ -69,8 +70,7 @@ +@@ -69,8 +70,7 @@ config ATMEL_TCB_CLKSRC are combined to make a single 32-bit timer. When GENERIC_CLOCKEVENTS is defined, the third timer channel @@ -4631,7 +4782,7 @@ diff -Nur linux-4.9.28.orig/drivers/misc/Kconfig linux-4.9.28/drivers/misc/Kconf config ATMEL_TCB_CLKSRC_BLOCK int -@@ -84,6 +84,15 @@ +@@ -84,6 +84,15 @@ config ATMEL_TCB_CLKSRC_BLOCK TC can be used for other purposes, such as PWM generation and interval timing. @@ -4647,10 +4798,11 @@ diff -Nur linux-4.9.28.orig/drivers/misc/Kconfig linux-4.9.28/drivers/misc/Kconf config DUMMY_IRQ tristate "Dummy IRQ handler" default n -diff -Nur linux-4.9.28.orig/drivers/mmc/host/mmci.c linux-4.9.28/drivers/mmc/host/mmci.c ---- linux-4.9.28.orig/drivers/mmc/host/mmci.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/mmc/host/mmci.c 2017-05-19 03:37:25.150175574 +0200 -@@ -1147,15 +1147,12 @@ +diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c +index df990bb8c873..1a162709a85e 100644 +--- a/drivers/mmc/host/mmci.c ++++ b/drivers/mmc/host/mmci.c +@@ -1147,15 +1147,12 @@ static irqreturn_t mmci_pio_irq(int irq, void *dev_id) struct sg_mapping_iter *sg_miter = &host->sg_miter; struct variant_data *variant = host->variant; void __iomem *base = host->base; @@ -4666,7 +4818,7 @@ diff -Nur linux-4.9.28.orig/drivers/mmc/host/mmci.c linux-4.9.28/drivers/mmc/hos do { unsigned int remain, len; char *buffer; -@@ -1195,8 +1192,6 @@ +@@ -1195,8 +1192,6 @@ static irqreturn_t mmci_pio_irq(int irq, void *dev_id) sg_miter_stop(sg_miter); @@ -4675,10 +4827,11 @@ diff -Nur linux-4.9.28.orig/drivers/mmc/host/mmci.c linux-4.9.28/drivers/mmc/hos /* * If we have less than the fifo 'half-full' threshold to transfer, * trigger a PIO interrupt as soon as any data is available. -diff -Nur linux-4.9.28.orig/drivers/net/ethernet/3com/3c59x.c linux-4.9.28/drivers/net/ethernet/3com/3c59x.c ---- linux-4.9.28.orig/drivers/net/ethernet/3com/3c59x.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/net/ethernet/3com/3c59x.c 2017-05-19 03:37:25.150175574 +0200 -@@ -842,9 +842,9 @@ +diff --git a/drivers/net/ethernet/3com/3c59x.c b/drivers/net/ethernet/3com/3c59x.c +index 9133e7926da5..63afb921ed40 100644 +--- a/drivers/net/ethernet/3com/3c59x.c ++++ b/drivers/net/ethernet/3com/3c59x.c +@@ -842,9 +842,9 @@ static void poll_vortex(struct net_device *dev) { struct vortex_private *vp = netdev_priv(dev); unsigned long flags; @@ -4690,7 +4843,7 @@ diff -Nur linux-4.9.28.orig/drivers/net/ethernet/3com/3c59x.c linux-4.9.28/drive } #endif -@@ -1910,12 +1910,12 @@ +@@ -1910,12 +1910,12 @@ static void vortex_tx_timeout(struct net_device *dev) * Block interrupts because vortex_interrupt does a bare spin_lock() */ unsigned long flags; @@ -4705,10 +4858,11 @@ diff -Nur linux-4.9.28.orig/drivers/net/ethernet/3com/3c59x.c linux-4.9.28/drive } } -diff -Nur linux-4.9.28.orig/drivers/net/ethernet/realtek/8139too.c linux-4.9.28/drivers/net/ethernet/realtek/8139too.c ---- linux-4.9.28.orig/drivers/net/ethernet/realtek/8139too.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/net/ethernet/realtek/8139too.c 2017-05-19 03:37:25.150175574 +0200 -@@ -2233,7 +2233,7 @@ +diff --git a/drivers/net/ethernet/realtek/8139too.c b/drivers/net/ethernet/realtek/8139too.c +index da4c2d8a4173..1420dfb56bac 100644 +--- a/drivers/net/ethernet/realtek/8139too.c ++++ b/drivers/net/ethernet/realtek/8139too.c +@@ -2233,7 +2233,7 @@ static void rtl8139_poll_controller(struct net_device *dev) struct rtl8139_private *tp = netdev_priv(dev); const int irq = tp->pci_dev->irq; @@ -4717,10 +4871,11 @@ diff -Nur linux-4.9.28.orig/drivers/net/ethernet/realtek/8139too.c linux-4.9.28/ rtl8139_interrupt(irq, dev); enable_irq(irq); } -diff -Nur linux-4.9.28.orig/drivers/net/wireless/intersil/orinoco/orinoco_usb.c linux-4.9.28/drivers/net/wireless/intersil/orinoco/orinoco_usb.c ---- linux-4.9.28.orig/drivers/net/wireless/intersil/orinoco/orinoco_usb.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/net/wireless/intersil/orinoco/orinoco_usb.c 2017-05-19 03:37:25.150175574 +0200 -@@ -697,7 +697,7 @@ +diff --git a/drivers/net/wireless/intersil/orinoco/orinoco_usb.c b/drivers/net/wireless/intersil/orinoco/orinoco_usb.c +index bca6935a94db..d7a35ee34d03 100644 +--- a/drivers/net/wireless/intersil/orinoco/orinoco_usb.c ++++ b/drivers/net/wireless/intersil/orinoco/orinoco_usb.c +@@ -697,7 +697,7 @@ static void ezusb_req_ctx_wait(struct ezusb_priv *upriv, while (!ctx->done.done && msecs--) udelay(1000); } else { @@ -4729,10 +4884,11 @@ diff -Nur linux-4.9.28.orig/drivers/net/wireless/intersil/orinoco/orinoco_usb.c ctx->done.done); } break; -diff -Nur linux-4.9.28.orig/drivers/pci/access.c linux-4.9.28/drivers/pci/access.c ---- linux-4.9.28.orig/drivers/pci/access.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/pci/access.c 2017-05-19 03:37:25.150175574 +0200 -@@ -672,7 +672,7 @@ +diff --git a/drivers/pci/access.c b/drivers/pci/access.c +index d11cdbb8fba3..223bbb9acb03 100644 +--- a/drivers/pci/access.c ++++ b/drivers/pci/access.c +@@ -672,7 +672,7 @@ void pci_cfg_access_unlock(struct pci_dev *dev) WARN_ON(!dev->block_cfg_access); dev->block_cfg_access = 0; @@ -4741,10 +4897,11 @@ diff -Nur linux-4.9.28.orig/drivers/pci/access.c linux-4.9.28/drivers/pci/access raw_spin_unlock_irqrestore(&pci_lock, flags); } EXPORT_SYMBOL_GPL(pci_cfg_access_unlock); -diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/drivers/pinctrl/qcom/pinctrl-msm.c ---- linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/pinctrl/qcom/pinctrl-msm.c 2017-05-19 03:37:25.150175574 +0200 -@@ -61,7 +61,7 @@ +diff --git a/drivers/pinctrl/qcom/pinctrl-msm.c b/drivers/pinctrl/qcom/pinctrl-msm.c +index bedce3453dd3..faf038978650 100644 +--- a/drivers/pinctrl/qcom/pinctrl-msm.c ++++ b/drivers/pinctrl/qcom/pinctrl-msm.c +@@ -61,7 +61,7 @@ struct msm_pinctrl { struct notifier_block restart_nb; int irq; @@ -4753,7 +4910,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv DECLARE_BITMAP(dual_edge_irqs, MAX_NR_GPIO); DECLARE_BITMAP(enabled_irqs, MAX_NR_GPIO); -@@ -153,14 +153,14 @@ +@@ -153,14 +153,14 @@ static int msm_pinmux_set_mux(struct pinctrl_dev *pctldev, if (WARN_ON(i == g->nfuncs)) return -EINVAL; @@ -4770,7 +4927,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv return 0; } -@@ -323,14 +323,14 @@ +@@ -323,14 +323,14 @@ static int msm_config_group_set(struct pinctrl_dev *pctldev, break; case PIN_CONFIG_OUTPUT: /* set output value */ @@ -4787,7 +4944,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv /* enable output */ arg = 1; -@@ -351,12 +351,12 @@ +@@ -351,12 +351,12 @@ static int msm_config_group_set(struct pinctrl_dev *pctldev, return -EINVAL; } @@ -4802,7 +4959,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv } return 0; -@@ -384,13 +384,13 @@ +@@ -384,13 +384,13 @@ static int msm_gpio_direction_input(struct gpio_chip *chip, unsigned offset) g = &pctrl->soc->groups[offset]; @@ -4818,7 +4975,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv return 0; } -@@ -404,7 +404,7 @@ +@@ -404,7 +404,7 @@ static int msm_gpio_direction_output(struct gpio_chip *chip, unsigned offset, in g = &pctrl->soc->groups[offset]; @@ -4827,7 +4984,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv val = readl(pctrl->regs + g->io_reg); if (value) -@@ -417,7 +417,7 @@ +@@ -417,7 +417,7 @@ static int msm_gpio_direction_output(struct gpio_chip *chip, unsigned offset, in val |= BIT(g->oe_bit); writel(val, pctrl->regs + g->ctl_reg); @@ -4836,7 +4993,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv return 0; } -@@ -443,7 +443,7 @@ +@@ -443,7 +443,7 @@ static void msm_gpio_set(struct gpio_chip *chip, unsigned offset, int value) g = &pctrl->soc->groups[offset]; @@ -4845,7 +5002,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv val = readl(pctrl->regs + g->io_reg); if (value) -@@ -452,7 +452,7 @@ +@@ -452,7 +452,7 @@ static void msm_gpio_set(struct gpio_chip *chip, unsigned offset, int value) val &= ~BIT(g->out_bit); writel(val, pctrl->regs + g->io_reg); @@ -4854,7 +5011,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv } #ifdef CONFIG_DEBUG_FS -@@ -571,7 +571,7 @@ +@@ -571,7 +571,7 @@ static void msm_gpio_irq_mask(struct irq_data *d) g = &pctrl->soc->groups[d->hwirq]; @@ -4863,7 +5020,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv val = readl(pctrl->regs + g->intr_cfg_reg); val &= ~BIT(g->intr_enable_bit); -@@ -579,7 +579,7 @@ +@@ -579,7 +579,7 @@ static void msm_gpio_irq_mask(struct irq_data *d) clear_bit(d->hwirq, pctrl->enabled_irqs); @@ -4872,7 +5029,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv } static void msm_gpio_irq_unmask(struct irq_data *d) -@@ -592,7 +592,7 @@ +@@ -592,7 +592,7 @@ static void msm_gpio_irq_unmask(struct irq_data *d) g = &pctrl->soc->groups[d->hwirq]; @@ -4881,7 +5038,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv val = readl(pctrl->regs + g->intr_cfg_reg); val |= BIT(g->intr_enable_bit); -@@ -600,7 +600,7 @@ +@@ -600,7 +600,7 @@ static void msm_gpio_irq_unmask(struct irq_data *d) set_bit(d->hwirq, pctrl->enabled_irqs); @@ -4890,7 +5047,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv } static void msm_gpio_irq_ack(struct irq_data *d) -@@ -613,7 +613,7 @@ +@@ -613,7 +613,7 @@ static void msm_gpio_irq_ack(struct irq_data *d) g = &pctrl->soc->groups[d->hwirq]; @@ -4899,7 +5056,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv val = readl(pctrl->regs + g->intr_status_reg); if (g->intr_ack_high) -@@ -625,7 +625,7 @@ +@@ -625,7 +625,7 @@ static void msm_gpio_irq_ack(struct irq_data *d) if (test_bit(d->hwirq, pctrl->dual_edge_irqs)) msm_gpio_update_dual_edge_pos(pctrl, g, d); @@ -4908,7 +5065,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv } static int msm_gpio_irq_set_type(struct irq_data *d, unsigned int type) -@@ -638,7 +638,7 @@ +@@ -638,7 +638,7 @@ static int msm_gpio_irq_set_type(struct irq_data *d, unsigned int type) g = &pctrl->soc->groups[d->hwirq]; @@ -4917,7 +5074,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv /* * For hw without possibility of detecting both edges -@@ -712,7 +712,7 @@ +@@ -712,7 +712,7 @@ static int msm_gpio_irq_set_type(struct irq_data *d, unsigned int type) if (test_bit(d->hwirq, pctrl->dual_edge_irqs)) msm_gpio_update_dual_edge_pos(pctrl, g, d); @@ -4926,7 +5083,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv if (type & (IRQ_TYPE_LEVEL_LOW | IRQ_TYPE_LEVEL_HIGH)) irq_set_handler_locked(d, handle_level_irq); -@@ -728,11 +728,11 @@ +@@ -728,11 +728,11 @@ static int msm_gpio_irq_set_wake(struct irq_data *d, unsigned int on) struct msm_pinctrl *pctrl = gpiochip_get_data(gc); unsigned long flags; @@ -4940,7 +5097,7 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv return 0; } -@@ -878,7 +878,7 @@ +@@ -878,7 +878,7 @@ int msm_pinctrl_probe(struct platform_device *pdev, pctrl->soc = soc_data; pctrl->chip = msm_gpio_template; @@ -4949,10 +5106,11 @@ diff -Nur linux-4.9.28.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-4.9.28/driv res = platform_get_resource(pdev, IORESOURCE_MEM, 0); pctrl->regs = devm_ioremap_resource(&pdev->dev, res); -diff -Nur linux-4.9.28.orig/drivers/scsi/fcoe/fcoe.c linux-4.9.28/drivers/scsi/fcoe/fcoe.c ---- linux-4.9.28.orig/drivers/scsi/fcoe/fcoe.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/scsi/fcoe/fcoe.c 2017-05-19 03:37:25.150175574 +0200 -@@ -1455,11 +1455,11 @@ +diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c +index 9bd41a35a78a..8e2d436c2e3f 100644 +--- a/drivers/scsi/fcoe/fcoe.c ++++ b/drivers/scsi/fcoe/fcoe.c +@@ -1455,11 +1455,11 @@ static int fcoe_rcv(struct sk_buff *skb, struct net_device *netdev, static int fcoe_alloc_paged_crc_eof(struct sk_buff *skb, int tlen) { struct fcoe_percpu_s *fps; @@ -4967,7 +5125,7 @@ diff -Nur linux-4.9.28.orig/drivers/scsi/fcoe/fcoe.c linux-4.9.28/drivers/scsi/f return rc; } -@@ -1646,11 +1646,11 @@ +@@ -1646,11 +1646,11 @@ static inline int fcoe_filter_frames(struct fc_lport *lport, return 0; } @@ -4981,7 +5139,7 @@ diff -Nur linux-4.9.28.orig/drivers/scsi/fcoe/fcoe.c linux-4.9.28/drivers/scsi/f return -EINVAL; } -@@ -1693,7 +1693,7 @@ +@@ -1693,7 +1693,7 @@ static void fcoe_recv_frame(struct sk_buff *skb) */ hp = (struct fcoe_hdr *) skb_network_header(skb); @@ -4990,7 +5148,7 @@ diff -Nur linux-4.9.28.orig/drivers/scsi/fcoe/fcoe.c linux-4.9.28/drivers/scsi/f if (unlikely(FC_FCOE_DECAPS_VER(hp) != FC_FCOE_VER)) { if (stats->ErrorFrames < 5) printk(KERN_WARNING "fcoe: FCoE version " -@@ -1725,13 +1725,13 @@ +@@ -1725,13 +1725,13 @@ static void fcoe_recv_frame(struct sk_buff *skb) goto drop; if (!fcoe_filter_frames(lport, fp)) { @@ -5006,10 +5164,11 @@ diff -Nur linux-4.9.28.orig/drivers/scsi/fcoe/fcoe.c linux-4.9.28/drivers/scsi/f kfree_skb(skb); } -diff -Nur linux-4.9.28.orig/drivers/scsi/fcoe/fcoe_ctlr.c linux-4.9.28/drivers/scsi/fcoe/fcoe_ctlr.c ---- linux-4.9.28.orig/drivers/scsi/fcoe/fcoe_ctlr.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/scsi/fcoe/fcoe_ctlr.c 2017-05-19 03:37:25.150175574 +0200 -@@ -834,7 +834,7 @@ +diff --git a/drivers/scsi/fcoe/fcoe_ctlr.c b/drivers/scsi/fcoe/fcoe_ctlr.c +index dcf36537a767..1a1f2e46452c 100644 +--- a/drivers/scsi/fcoe/fcoe_ctlr.c ++++ b/drivers/scsi/fcoe/fcoe_ctlr.c +@@ -834,7 +834,7 @@ static unsigned long fcoe_ctlr_age_fcfs(struct fcoe_ctlr *fip) INIT_LIST_HEAD(&del_list); @@ -5018,7 +5177,7 @@ diff -Nur linux-4.9.28.orig/drivers/scsi/fcoe/fcoe_ctlr.c linux-4.9.28/drivers/s list_for_each_entry_safe(fcf, next, &fip->fcfs, list) { deadline = fcf->time + fcf->fka_period + fcf->fka_period / 2; -@@ -870,7 +870,7 @@ +@@ -870,7 +870,7 @@ static unsigned long fcoe_ctlr_age_fcfs(struct fcoe_ctlr *fip) sel_time = fcf->time; } } @@ -5027,10 +5186,11 @@ diff -Nur linux-4.9.28.orig/drivers/scsi/fcoe/fcoe_ctlr.c linux-4.9.28/drivers/s list_for_each_entry_safe(fcf, next, &del_list, list) { /* Removes fcf from current list */ -diff -Nur linux-4.9.28.orig/drivers/scsi/libfc/fc_exch.c linux-4.9.28/drivers/scsi/libfc/fc_exch.c ---- linux-4.9.28.orig/drivers/scsi/libfc/fc_exch.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/scsi/libfc/fc_exch.c 2017-05-19 03:37:25.150175574 +0200 -@@ -814,10 +814,10 @@ +diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c +index 16ca31ad5ec0..c3987347e762 100644 +--- a/drivers/scsi/libfc/fc_exch.c ++++ b/drivers/scsi/libfc/fc_exch.c +@@ -814,10 +814,10 @@ static struct fc_exch *fc_exch_em_alloc(struct fc_lport *lport, } memset(ep, 0, sizeof(*ep)); @@ -5043,10 +5203,11 @@ diff -Nur linux-4.9.28.orig/drivers/scsi/libfc/fc_exch.c linux-4.9.28/drivers/sc /* peek cache of free slot */ if (pool->left != FC_XID_UNKNOWN) { -diff -Nur linux-4.9.28.orig/drivers/scsi/libsas/sas_ata.c linux-4.9.28/drivers/scsi/libsas/sas_ata.c ---- linux-4.9.28.orig/drivers/scsi/libsas/sas_ata.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/scsi/libsas/sas_ata.c 2017-05-19 03:37:25.150175574 +0200 -@@ -190,7 +190,7 @@ +diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c +index 87f5e694dbed..23c0a50fb6aa 100644 +--- a/drivers/scsi/libsas/sas_ata.c ++++ b/drivers/scsi/libsas/sas_ata.c +@@ -190,7 +190,7 @@ static unsigned int sas_ata_qc_issue(struct ata_queued_cmd *qc) /* TODO: audit callers to ensure they are ready for qc_issue to * unconditionally re-enable interrupts */ @@ -5055,7 +5216,7 @@ diff -Nur linux-4.9.28.orig/drivers/scsi/libsas/sas_ata.c linux-4.9.28/drivers/s spin_unlock(ap->lock); /* If the device fell off, no sense in issuing commands */ -@@ -252,7 +252,7 @@ +@@ -252,7 +252,7 @@ static unsigned int sas_ata_qc_issue(struct ata_queued_cmd *qc) out: spin_lock(ap->lock); @@ -5064,10 +5225,11 @@ diff -Nur linux-4.9.28.orig/drivers/scsi/libsas/sas_ata.c linux-4.9.28/drivers/s return ret; } -diff -Nur linux-4.9.28.orig/drivers/scsi/qla2xxx/qla_inline.h linux-4.9.28/drivers/scsi/qla2xxx/qla_inline.h ---- linux-4.9.28.orig/drivers/scsi/qla2xxx/qla_inline.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/scsi/qla2xxx/qla_inline.h 2017-05-19 03:37:25.150175574 +0200 -@@ -59,12 +59,12 @@ +diff --git a/drivers/scsi/qla2xxx/qla_inline.h b/drivers/scsi/qla2xxx/qla_inline.h +index edc48f3b8230..ee5c6f9dfb6f 100644 +--- a/drivers/scsi/qla2xxx/qla_inline.h ++++ b/drivers/scsi/qla2xxx/qla_inline.h +@@ -59,12 +59,12 @@ qla2x00_poll(struct rsp_que *rsp) { unsigned long flags; struct qla_hw_data *ha = rsp->hw; @@ -5082,10 +5244,11 @@ diff -Nur linux-4.9.28.orig/drivers/scsi/qla2xxx/qla_inline.h linux-4.9.28/drive } static inline uint8_t * -diff -Nur linux-4.9.28.orig/drivers/scsi/qla2xxx/qla_isr.c linux-4.9.28/drivers/scsi/qla2xxx/qla_isr.c ---- linux-4.9.28.orig/drivers/scsi/qla2xxx/qla_isr.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/scsi/qla2xxx/qla_isr.c 2017-05-19 03:37:25.154175725 +0200 -@@ -3125,7 +3125,11 @@ +diff --git a/drivers/scsi/qla2xxx/qla_isr.c b/drivers/scsi/qla2xxx/qla_isr.c +index 068c4e47fac9..a2090f640397 100644 +--- a/drivers/scsi/qla2xxx/qla_isr.c ++++ b/drivers/scsi/qla2xxx/qla_isr.c +@@ -3125,7 +3125,11 @@ qla24xx_enable_msix(struct qla_hw_data *ha, struct rsp_que *rsp) * kref_put(). */ kref_get(&qentry->irq_notify.kref); @@ -5097,9 +5260,10 @@ diff -Nur linux-4.9.28.orig/drivers/scsi/qla2xxx/qla_isr.c linux-4.9.28/drivers/ } /* -diff -Nur linux-4.9.28.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.9.28/drivers/thermal/x86_pkg_temp_thermal.c ---- linux-4.9.28.orig/drivers/thermal/x86_pkg_temp_thermal.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/thermal/x86_pkg_temp_thermal.c 2017-05-19 03:37:25.154175725 +0200 +diff --git a/drivers/thermal/x86_pkg_temp_thermal.c b/drivers/thermal/x86_pkg_temp_thermal.c +index 95f4c1bcdb4c..0be934799bff 100644 +--- a/drivers/thermal/x86_pkg_temp_thermal.c ++++ b/drivers/thermal/x86_pkg_temp_thermal.c @@ -29,6 +29,7 @@ #include #include @@ -5108,7 +5272,7 @@ diff -Nur linux-4.9.28.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.9.28/ #include #include -@@ -353,7 +354,7 @@ +@@ -353,7 +354,7 @@ static void pkg_temp_thermal_threshold_work_fn(struct work_struct *work) } } @@ -5117,7 +5281,7 @@ diff -Nur linux-4.9.28.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.9.28/ { unsigned long flags; int cpu = smp_processor_id(); -@@ -370,7 +371,7 @@ +@@ -370,7 +371,7 @@ static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) pkg_work_scheduled[phy_id]) { disable_pkg_thres_interrupt(); spin_unlock_irqrestore(&pkg_work_lock, flags); @@ -5126,7 +5290,7 @@ diff -Nur linux-4.9.28.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.9.28/ } pkg_work_scheduled[phy_id] = 1; spin_unlock_irqrestore(&pkg_work_lock, flags); -@@ -379,9 +380,48 @@ +@@ -379,9 +380,48 @@ static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) schedule_delayed_work_on(cpu, &per_cpu(pkg_temp_thermal_threshold_work, cpu), msecs_to_jiffies(notify_delay_ms)); @@ -5175,7 +5339,7 @@ diff -Nur linux-4.9.28.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.9.28/ static int find_siblings_cpu(int cpu) { int i; -@@ -585,6 +625,9 @@ +@@ -585,6 +625,9 @@ static int __init pkg_temp_thermal_init(void) if (!x86_match_cpu(pkg_temp_thermal_ids)) return -ENODEV; @@ -5185,7 +5349,7 @@ diff -Nur linux-4.9.28.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.9.28/ spin_lock_init(&pkg_work_lock); platform_thermal_package_notify = pkg_temp_thermal_platform_thermal_notify; -@@ -609,7 +652,7 @@ +@@ -609,7 +652,7 @@ static int __init pkg_temp_thermal_init(void) kfree(pkg_work_scheduled); platform_thermal_package_notify = NULL; platform_thermal_package_rate_control = NULL; @@ -5194,7 +5358,7 @@ diff -Nur linux-4.9.28.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.9.28/ return -ENODEV; } -@@ -634,6 +677,7 @@ +@@ -634,6 +677,7 @@ static void __exit pkg_temp_thermal_exit(void) mutex_unlock(&phy_dev_list_mutex); platform_thermal_package_notify = NULL; platform_thermal_package_rate_control = NULL; @@ -5202,10 +5366,11 @@ diff -Nur linux-4.9.28.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.9.28/ for_each_online_cpu(i) cancel_delayed_work_sync( &per_cpu(pkg_temp_thermal_threshold_work, i)); -diff -Nur linux-4.9.28.orig/drivers/tty/serial/8250/8250_core.c linux-4.9.28/drivers/tty/serial/8250/8250_core.c ---- linux-4.9.28.orig/drivers/tty/serial/8250/8250_core.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/tty/serial/8250/8250_core.c 2017-05-19 03:37:25.154175725 +0200 -@@ -58,7 +58,16 @@ +diff --git a/drivers/tty/serial/8250/8250_core.c b/drivers/tty/serial/8250/8250_core.c +index e8819aa20415..dd7f9bf45d6c 100644 +--- a/drivers/tty/serial/8250/8250_core.c ++++ b/drivers/tty/serial/8250/8250_core.c +@@ -58,7 +58,16 @@ static struct uart_driver serial8250_reg; static unsigned int skip_txen_test; /* force skip of txen test at init time */ @@ -5223,9 +5388,10 @@ diff -Nur linux-4.9.28.orig/drivers/tty/serial/8250/8250_core.c linux-4.9.28/dri #include /* -diff -Nur linux-4.9.28.orig/drivers/tty/serial/8250/8250_port.c linux-4.9.28/drivers/tty/serial/8250/8250_port.c ---- linux-4.9.28.orig/drivers/tty/serial/8250/8250_port.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/tty/serial/8250/8250_port.c 2017-05-19 03:37:25.154175725 +0200 +diff --git a/drivers/tty/serial/8250/8250_port.c b/drivers/tty/serial/8250/8250_port.c +index 080d5a59d0a7..eecc4f111473 100644 +--- a/drivers/tty/serial/8250/8250_port.c ++++ b/drivers/tty/serial/8250/8250_port.c @@ -35,6 +35,7 @@ #include #include @@ -5234,7 +5400,7 @@ diff -Nur linux-4.9.28.orig/drivers/tty/serial/8250/8250_port.c linux-4.9.28/dri #include #include #include -@@ -3144,9 +3145,9 @@ +@@ -3144,9 +3145,9 @@ void serial8250_console_write(struct uart_8250_port *up, const char *s, serial8250_rpm_get(up); @@ -5246,10 +5412,11 @@ diff -Nur linux-4.9.28.orig/drivers/tty/serial/8250/8250_port.c linux-4.9.28/dri locked = spin_trylock_irqsave(&port->lock, flags); else spin_lock_irqsave(&port->lock, flags); -diff -Nur linux-4.9.28.orig/drivers/tty/serial/amba-pl011.c linux-4.9.28/drivers/tty/serial/amba-pl011.c ---- linux-4.9.28.orig/drivers/tty/serial/amba-pl011.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/tty/serial/amba-pl011.c 2017-05-19 03:37:25.154175725 +0200 -@@ -2194,13 +2194,19 @@ +diff --git a/drivers/tty/serial/amba-pl011.c b/drivers/tty/serial/amba-pl011.c +index e2c33b9528d8..53af53c43e8c 100644 +--- a/drivers/tty/serial/amba-pl011.c ++++ b/drivers/tty/serial/amba-pl011.c +@@ -2194,13 +2194,19 @@ pl011_console_write(struct console *co, const char *s, unsigned int count) clk_enable(uap->clk); @@ -5272,7 +5439,7 @@ diff -Nur linux-4.9.28.orig/drivers/tty/serial/amba-pl011.c linux-4.9.28/drivers /* * First save the CR then disable the interrupts -@@ -2224,8 +2230,7 @@ +@@ -2224,8 +2230,7 @@ pl011_console_write(struct console *co, const char *s, unsigned int count) pl011_write(old_cr, uap, REG_CR); if (locked) @@ -5282,10 +5449,11 @@ diff -Nur linux-4.9.28.orig/drivers/tty/serial/amba-pl011.c linux-4.9.28/drivers clk_disable(uap->clk); } -diff -Nur linux-4.9.28.orig/drivers/tty/serial/omap-serial.c linux-4.9.28/drivers/tty/serial/omap-serial.c ---- linux-4.9.28.orig/drivers/tty/serial/omap-serial.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/tty/serial/omap-serial.c 2017-05-19 03:37:25.154175725 +0200 -@@ -1257,13 +1257,10 @@ +diff --git a/drivers/tty/serial/omap-serial.c b/drivers/tty/serial/omap-serial.c +index 44e5b5bf713b..400140d1dfff 100644 +--- a/drivers/tty/serial/omap-serial.c ++++ b/drivers/tty/serial/omap-serial.c +@@ -1257,13 +1257,10 @@ serial_omap_console_write(struct console *co, const char *s, pm_runtime_get_sync(up->dev); @@ -5302,7 +5470,7 @@ diff -Nur linux-4.9.28.orig/drivers/tty/serial/omap-serial.c linux-4.9.28/driver /* * First save the IER then disable the interrupts -@@ -1292,8 +1289,7 @@ +@@ -1292,8 +1289,7 @@ serial_omap_console_write(struct console *co, const char *s, pm_runtime_mark_last_busy(up->dev); pm_runtime_put_autosuspend(up->dev); if (locked) @@ -5312,10 +5480,11 @@ diff -Nur linux-4.9.28.orig/drivers/tty/serial/omap-serial.c linux-4.9.28/driver } static int __init -diff -Nur linux-4.9.28.orig/drivers/usb/core/hcd.c linux-4.9.28/drivers/usb/core/hcd.c ---- linux-4.9.28.orig/drivers/usb/core/hcd.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/usb/core/hcd.c 2017-05-19 03:37:25.154175725 +0200 -@@ -1764,9 +1764,9 @@ +diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c +index 3b9735abf2e0..73ba3239869e 100644 +--- a/drivers/usb/core/hcd.c ++++ b/drivers/usb/core/hcd.c +@@ -1764,9 +1764,9 @@ static void __usb_hcd_giveback_urb(struct urb *urb) * and no one may trigger the above deadlock situation when * running complete() in tasklet. */ @@ -5327,10 +5496,11 @@ diff -Nur linux-4.9.28.orig/drivers/usb/core/hcd.c linux-4.9.28/drivers/usb/core usb_anchor_resume_wakeups(anchor); atomic_dec(&urb->use_count); -diff -Nur linux-4.9.28.orig/drivers/usb/gadget/function/f_fs.c linux-4.9.28/drivers/usb/gadget/function/f_fs.c ---- linux-4.9.28.orig/drivers/usb/gadget/function/f_fs.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/usb/gadget/function/f_fs.c 2017-05-19 03:37:25.154175725 +0200 -@@ -1593,7 +1593,7 @@ +diff --git a/drivers/usb/gadget/function/f_fs.c b/drivers/usb/gadget/function/f_fs.c +index 89081b834615..90b231b7ad0a 100644 +--- a/drivers/usb/gadget/function/f_fs.c ++++ b/drivers/usb/gadget/function/f_fs.c +@@ -1593,7 +1593,7 @@ static void ffs_data_put(struct ffs_data *ffs) pr_info("%s(): freeing\n", __func__); ffs_data_clear(ffs); BUG_ON(waitqueue_active(&ffs->ev.waitq) || @@ -5339,10 +5509,11 @@ diff -Nur linux-4.9.28.orig/drivers/usb/gadget/function/f_fs.c linux-4.9.28/driv kfree(ffs->dev_name); kfree(ffs); } -diff -Nur linux-4.9.28.orig/drivers/usb/gadget/legacy/inode.c linux-4.9.28/drivers/usb/gadget/legacy/inode.c ---- linux-4.9.28.orig/drivers/usb/gadget/legacy/inode.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/drivers/usb/gadget/legacy/inode.c 2017-05-19 03:37:25.154175725 +0200 -@@ -346,7 +346,7 @@ +diff --git a/drivers/usb/gadget/legacy/inode.c b/drivers/usb/gadget/legacy/inode.c +index 1468d8f085a3..6aae3ae25c18 100644 +--- a/drivers/usb/gadget/legacy/inode.c ++++ b/drivers/usb/gadget/legacy/inode.c +@@ -346,7 +346,7 @@ ep_io (struct ep_data *epdata, void *buf, unsigned len) spin_unlock_irq (&epdata->dev->lock); if (likely (value == 0)) { @@ -5351,7 +5522,7 @@ diff -Nur linux-4.9.28.orig/drivers/usb/gadget/legacy/inode.c linux-4.9.28/drive if (value != 0) { spin_lock_irq (&epdata->dev->lock); if (likely (epdata->ep != NULL)) { -@@ -355,7 +355,7 @@ +@@ -355,7 +355,7 @@ ep_io (struct ep_data *epdata, void *buf, unsigned len) usb_ep_dequeue (epdata->ep, epdata->req); spin_unlock_irq (&epdata->dev->lock); @@ -5360,9 +5531,10 @@ diff -Nur linux-4.9.28.orig/drivers/usb/gadget/legacy/inode.c linux-4.9.28/drive if (epdata->status == -ECONNRESET) epdata->status = -EINTR; } else { -diff -Nur linux-4.9.28.orig/fs/aio.c linux-4.9.28/fs/aio.c ---- linux-4.9.28.orig/fs/aio.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/aio.c 2017-05-19 03:37:25.154175725 +0200 +diff --git a/fs/aio.c b/fs/aio.c +index 428484f2f841..2b02e2eb2158 100644 +--- a/fs/aio.c ++++ b/fs/aio.c @@ -40,6 +40,7 @@ #include #include @@ -5371,7 +5543,7 @@ diff -Nur linux-4.9.28.orig/fs/aio.c linux-4.9.28/fs/aio.c #include #include -@@ -115,7 +116,7 @@ +@@ -115,7 +116,7 @@ struct kioctx { struct page **ring_pages; long nr_pages; @@ -5380,7 +5552,7 @@ diff -Nur linux-4.9.28.orig/fs/aio.c linux-4.9.28/fs/aio.c /* * signals when all in-flight requests are done -@@ -258,6 +259,7 @@ +@@ -258,6 +259,7 @@ static int __init aio_setup(void) .mount = aio_mount, .kill_sb = kill_anon_super, }; @@ -5388,7 +5560,7 @@ diff -Nur linux-4.9.28.orig/fs/aio.c linux-4.9.28/fs/aio.c aio_mnt = kern_mount(&aio_fs); if (IS_ERR(aio_mnt)) panic("Failed to create aio fs mount."); -@@ -581,9 +583,9 @@ +@@ -581,9 +583,9 @@ static int kiocb_cancel(struct aio_kiocb *kiocb) return cancel(&kiocb->common); } @@ -5400,7 +5572,7 @@ diff -Nur linux-4.9.28.orig/fs/aio.c linux-4.9.28/fs/aio.c pr_debug("freeing %p\n", ctx); -@@ -602,8 +604,8 @@ +@@ -602,8 +604,8 @@ static void free_ioctx_reqs(struct percpu_ref *ref) if (ctx->rq_wait && atomic_dec_and_test(&ctx->rq_wait->count)) complete(&ctx->rq_wait->comp); @@ -5411,7 +5583,7 @@ diff -Nur linux-4.9.28.orig/fs/aio.c linux-4.9.28/fs/aio.c } /* -@@ -611,9 +613,9 @@ +@@ -611,9 +613,9 @@ static void free_ioctx_reqs(struct percpu_ref *ref) * and ctx->users has dropped to 0, so we know no more kiocbs can be submitted - * now it's safe to cancel any that need to be. */ @@ -5423,7 +5595,7 @@ diff -Nur linux-4.9.28.orig/fs/aio.c linux-4.9.28/fs/aio.c struct aio_kiocb *req; spin_lock_irq(&ctx->ctx_lock); -@@ -632,6 +634,14 @@ +@@ -632,6 +634,14 @@ static void free_ioctx_users(struct percpu_ref *ref) percpu_ref_put(&ctx->reqs); } @@ -5438,9 +5610,10 @@ diff -Nur linux-4.9.28.orig/fs/aio.c linux-4.9.28/fs/aio.c static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm) { unsigned i, new_nr; -diff -Nur linux-4.9.28.orig/fs/autofs4/autofs_i.h linux-4.9.28/fs/autofs4/autofs_i.h ---- linux-4.9.28.orig/fs/autofs4/autofs_i.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/autofs4/autofs_i.h 2017-05-19 03:37:25.154175725 +0200 +diff --git a/fs/autofs4/autofs_i.h b/fs/autofs4/autofs_i.h +index a1fba4285277..3796769b4cd1 100644 +--- a/fs/autofs4/autofs_i.h ++++ b/fs/autofs4/autofs_i.h @@ -31,6 +31,7 @@ #include #include @@ -5449,10 +5622,11 @@ diff -Nur linux-4.9.28.orig/fs/autofs4/autofs_i.h linux-4.9.28/fs/autofs4/autofs #include #include -diff -Nur linux-4.9.28.orig/fs/autofs4/expire.c linux-4.9.28/fs/autofs4/expire.c ---- linux-4.9.28.orig/fs/autofs4/expire.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/autofs4/expire.c 2017-05-19 03:37:25.154175725 +0200 -@@ -148,7 +148,7 @@ +diff --git a/fs/autofs4/expire.c b/fs/autofs4/expire.c +index d8e6d421c27f..2e689ab1306b 100644 +--- a/fs/autofs4/expire.c ++++ b/fs/autofs4/expire.c +@@ -148,7 +148,7 @@ static struct dentry *get_next_positive_dentry(struct dentry *prev, parent = p->d_parent; if (!spin_trylock(&parent->d_lock)) { spin_unlock(&p->d_lock); @@ -5461,10 +5635,11 @@ diff -Nur linux-4.9.28.orig/fs/autofs4/expire.c linux-4.9.28/fs/autofs4/expire.c goto relock; } spin_unlock(&p->d_lock); -diff -Nur linux-4.9.28.orig/fs/buffer.c linux-4.9.28/fs/buffer.c ---- linux-4.9.28.orig/fs/buffer.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/buffer.c 2017-05-19 03:37:25.158175874 +0200 -@@ -301,8 +301,7 @@ +diff --git a/fs/buffer.c b/fs/buffer.c +index b205a629001d..5646afc022ba 100644 +--- a/fs/buffer.c ++++ b/fs/buffer.c +@@ -301,8 +301,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate) * decide that the page is now completely done. */ first = page_buffers(page); @@ -5474,7 +5649,7 @@ diff -Nur linux-4.9.28.orig/fs/buffer.c linux-4.9.28/fs/buffer.c clear_buffer_async_read(bh); unlock_buffer(bh); tmp = bh; -@@ -315,8 +314,7 @@ +@@ -315,8 +314,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate) } tmp = tmp->b_this_page; } while (tmp != bh); @@ -5484,7 +5659,7 @@ diff -Nur linux-4.9.28.orig/fs/buffer.c linux-4.9.28/fs/buffer.c /* * If none of the buffers had errors and they are all -@@ -328,9 +326,7 @@ +@@ -328,9 +326,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate) return; still_busy: @@ -5495,7 +5670,7 @@ diff -Nur linux-4.9.28.orig/fs/buffer.c linux-4.9.28/fs/buffer.c } /* -@@ -358,8 +354,7 @@ +@@ -358,8 +354,7 @@ void end_buffer_async_write(struct buffer_head *bh, int uptodate) } first = page_buffers(page); @@ -5505,7 +5680,7 @@ diff -Nur linux-4.9.28.orig/fs/buffer.c linux-4.9.28/fs/buffer.c clear_buffer_async_write(bh); unlock_buffer(bh); -@@ -371,15 +366,12 @@ +@@ -371,15 +366,12 @@ void end_buffer_async_write(struct buffer_head *bh, int uptodate) } tmp = tmp->b_this_page; } @@ -5523,7 +5698,7 @@ diff -Nur linux-4.9.28.orig/fs/buffer.c linux-4.9.28/fs/buffer.c } EXPORT_SYMBOL(end_buffer_async_write); -@@ -3383,6 +3375,7 @@ +@@ -3383,6 +3375,7 @@ struct buffer_head *alloc_buffer_head(gfp_t gfp_flags) struct buffer_head *ret = kmem_cache_zalloc(bh_cachep, gfp_flags); if (ret) { INIT_LIST_HEAD(&ret->b_assoc_buffers); @@ -5531,10 +5706,11 @@ diff -Nur linux-4.9.28.orig/fs/buffer.c linux-4.9.28/fs/buffer.c preempt_disable(); __this_cpu_inc(bh_accounting.nr); recalc_bh_state(); -diff -Nur linux-4.9.28.orig/fs/cifs/readdir.c linux-4.9.28/fs/cifs/readdir.c ---- linux-4.9.28.orig/fs/cifs/readdir.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/cifs/readdir.c 2017-05-19 03:37:25.158175874 +0200 -@@ -80,7 +80,7 @@ +diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c +index a27fc8791551..791aecb7c1ac 100644 +--- a/fs/cifs/readdir.c ++++ b/fs/cifs/readdir.c +@@ -80,7 +80,7 @@ cifs_prime_dcache(struct dentry *parent, struct qstr *name, struct inode *inode; struct super_block *sb = parent->d_sb; struct cifs_sb_info *cifs_sb = CIFS_SB(sb); @@ -5543,9 +5719,10 @@ diff -Nur linux-4.9.28.orig/fs/cifs/readdir.c linux-4.9.28/fs/cifs/readdir.c cifs_dbg(FYI, "%s: for %s\n", __func__, name->name); -diff -Nur linux-4.9.28.orig/fs/dcache.c linux-4.9.28/fs/dcache.c ---- linux-4.9.28.orig/fs/dcache.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/dcache.c 2017-05-19 03:37:25.158175874 +0200 +diff --git a/fs/dcache.c b/fs/dcache.c +index 4485a48f4091..691039a6a872 100644 +--- a/fs/dcache.c ++++ b/fs/dcache.c @@ -19,6 +19,7 @@ #include #include @@ -5554,7 +5731,7 @@ diff -Nur linux-4.9.28.orig/fs/dcache.c linux-4.9.28/fs/dcache.c #include #include #include -@@ -750,6 +751,8 @@ +@@ -750,6 +751,8 @@ static inline bool fast_dput(struct dentry *dentry) */ void dput(struct dentry *dentry) { @@ -5563,7 +5740,7 @@ diff -Nur linux-4.9.28.orig/fs/dcache.c linux-4.9.28/fs/dcache.c if (unlikely(!dentry)) return; -@@ -788,9 +791,18 @@ +@@ -788,9 +791,18 @@ void dput(struct dentry *dentry) return; kill_it: @@ -5585,7 +5762,7 @@ diff -Nur linux-4.9.28.orig/fs/dcache.c linux-4.9.28/fs/dcache.c goto repeat; } } -@@ -2324,7 +2336,7 @@ +@@ -2324,7 +2336,7 @@ void d_delete(struct dentry * dentry) if (dentry->d_lockref.count == 1) { if (!spin_trylock(&inode->i_lock)) { spin_unlock(&dentry->d_lock); @@ -5594,7 +5771,7 @@ diff -Nur linux-4.9.28.orig/fs/dcache.c linux-4.9.28/fs/dcache.c goto again; } dentry->d_flags &= ~DCACHE_CANT_MOUNT; -@@ -2384,21 +2396,24 @@ +@@ -2384,21 +2396,24 @@ static inline void end_dir_add(struct inode *dir, unsigned n) static void d_wait_lookup(struct dentry *dentry) { @@ -5630,7 +5807,7 @@ diff -Nur linux-4.9.28.orig/fs/dcache.c linux-4.9.28/fs/dcache.c { unsigned int hash = name->hash; struct hlist_bl_head *b = in_lookup_hash(parent, hash); -@@ -2507,7 +2522,7 @@ +@@ -2507,7 +2522,7 @@ void __d_lookup_done(struct dentry *dentry) hlist_bl_lock(b); dentry->d_flags &= ~DCACHE_PAR_LOOKUP; __hlist_bl_del(&dentry->d_u.d_in_lookup_hash); @@ -5639,7 +5816,7 @@ diff -Nur linux-4.9.28.orig/fs/dcache.c linux-4.9.28/fs/dcache.c dentry->d_wait = NULL; hlist_bl_unlock(b); INIT_HLIST_NODE(&dentry->d_u.d_alias); -@@ -3604,6 +3619,11 @@ +@@ -3604,6 +3619,11 @@ EXPORT_SYMBOL(d_genocide); void __init vfs_caches_init_early(void) { @@ -5651,10 +5828,11 @@ diff -Nur linux-4.9.28.orig/fs/dcache.c linux-4.9.28/fs/dcache.c dcache_init_early(); inode_init_early(); } -diff -Nur linux-4.9.28.orig/fs/eventpoll.c linux-4.9.28/fs/eventpoll.c ---- linux-4.9.28.orig/fs/eventpoll.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/eventpoll.c 2017-05-19 03:37:25.158175874 +0200 -@@ -510,12 +510,12 @@ +diff --git a/fs/eventpoll.c b/fs/eventpoll.c +index 10db91218933..42af0a06f657 100644 +--- a/fs/eventpoll.c ++++ b/fs/eventpoll.c +@@ -510,12 +510,12 @@ static int ep_poll_wakeup_proc(void *priv, void *cookie, int call_nests) */ static void ep_poll_safewake(wait_queue_head_t *wq) { @@ -5669,10 +5847,11 @@ diff -Nur linux-4.9.28.orig/fs/eventpoll.c linux-4.9.28/fs/eventpoll.c } static void ep_remove_wait_queue(struct eppoll_entry *pwq) -diff -Nur linux-4.9.28.orig/fs/exec.c linux-4.9.28/fs/exec.c ---- linux-4.9.28.orig/fs/exec.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/exec.c 2017-05-19 03:37:25.158175874 +0200 -@@ -1017,12 +1017,14 @@ +diff --git a/fs/exec.c b/fs/exec.c +index 67e86571685a..fe14cdd84016 100644 +--- a/fs/exec.c ++++ b/fs/exec.c +@@ -1017,12 +1017,14 @@ static int exec_mmap(struct mm_struct *mm) } } task_lock(tsk); @@ -5687,10 +5866,11 @@ diff -Nur linux-4.9.28.orig/fs/exec.c linux-4.9.28/fs/exec.c task_unlock(tsk); if (old_mm) { up_read(&old_mm->mmap_sem); -diff -Nur linux-4.9.28.orig/fs/fuse/dir.c linux-4.9.28/fs/fuse/dir.c ---- linux-4.9.28.orig/fs/fuse/dir.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/fuse/dir.c 2017-05-19 03:37:25.158175874 +0200 -@@ -1191,7 +1191,7 @@ +diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c +index 642c57b8de7b..8494b9308333 100644 +--- a/fs/fuse/dir.c ++++ b/fs/fuse/dir.c +@@ -1191,7 +1191,7 @@ static int fuse_direntplus_link(struct file *file, struct inode *dir = d_inode(parent); struct fuse_conn *fc; struct inode *inode; @@ -5699,10 +5879,11 @@ diff -Nur linux-4.9.28.orig/fs/fuse/dir.c linux-4.9.28/fs/fuse/dir.c if (!o->nodeid) { /* -diff -Nur linux-4.9.28.orig/fs/jbd2/checkpoint.c linux-4.9.28/fs/jbd2/checkpoint.c ---- linux-4.9.28.orig/fs/jbd2/checkpoint.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/jbd2/checkpoint.c 2017-05-19 03:37:25.158175874 +0200 -@@ -116,6 +116,8 @@ +diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c +index 684996c8a3a4..6e18a06aaabe 100644 +--- a/fs/jbd2/checkpoint.c ++++ b/fs/jbd2/checkpoint.c +@@ -116,6 +116,8 @@ void __jbd2_log_wait_for_space(journal_t *journal) nblocks = jbd2_space_needed(journal); while (jbd2_log_space_left(journal) < nblocks) { write_unlock(&journal->j_state_lock); @@ -5711,10 +5892,11 @@ diff -Nur linux-4.9.28.orig/fs/jbd2/checkpoint.c linux-4.9.28/fs/jbd2/checkpoint mutex_lock(&journal->j_checkpoint_mutex); /* -diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c ---- linux-4.9.28.orig/fs/locks.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/locks.c 2017-05-19 03:37:25.158175874 +0200 -@@ -935,7 +935,7 @@ +diff --git a/fs/locks.c b/fs/locks.c +index 22c5b4aa4961..269c6a44449a 100644 +--- a/fs/locks.c ++++ b/fs/locks.c +@@ -935,7 +935,7 @@ static int flock_lock_inode(struct inode *inode, struct file_lock *request) return -ENOMEM; } @@ -5723,7 +5905,7 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c spin_lock(&ctx->flc_lock); if (request->fl_flags & FL_ACCESS) goto find_conflict; -@@ -976,7 +976,7 @@ +@@ -976,7 +976,7 @@ static int flock_lock_inode(struct inode *inode, struct file_lock *request) out: spin_unlock(&ctx->flc_lock); @@ -5732,7 +5914,7 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c if (new_fl) locks_free_lock(new_fl); locks_dispose_list(&dispose); -@@ -1013,7 +1013,7 @@ +@@ -1013,7 +1013,7 @@ static int posix_lock_inode(struct inode *inode, struct file_lock *request, new_fl2 = locks_alloc_lock(); } @@ -5741,7 +5923,7 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c spin_lock(&ctx->flc_lock); /* * New lock request. Walk all POSIX locks and look for conflicts. If -@@ -1185,7 +1185,7 @@ +@@ -1185,7 +1185,7 @@ static int posix_lock_inode(struct inode *inode, struct file_lock *request, } out: spin_unlock(&ctx->flc_lock); @@ -5750,7 +5932,7 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c /* * Free any unused locks. */ -@@ -1460,7 +1460,7 @@ +@@ -1460,7 +1460,7 @@ int __break_lease(struct inode *inode, unsigned int mode, unsigned int type) return error; } @@ -5759,7 +5941,7 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c spin_lock(&ctx->flc_lock); time_out_leases(inode, &dispose); -@@ -1512,13 +1512,13 @@ +@@ -1512,13 +1512,13 @@ int __break_lease(struct inode *inode, unsigned int mode, unsigned int type) locks_insert_block(fl, new_fl); trace_break_lease_block(inode, new_fl); spin_unlock(&ctx->flc_lock); @@ -5775,7 +5957,7 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c spin_lock(&ctx->flc_lock); trace_break_lease_unblock(inode, new_fl); locks_delete_block(new_fl); -@@ -1535,7 +1535,7 @@ +@@ -1535,7 +1535,7 @@ int __break_lease(struct inode *inode, unsigned int mode, unsigned int type) } out: spin_unlock(&ctx->flc_lock); @@ -5784,7 +5966,7 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c locks_dispose_list(&dispose); locks_free_lock(new_fl); return error; -@@ -1609,7 +1609,7 @@ +@@ -1609,7 +1609,7 @@ int fcntl_getlease(struct file *filp) ctx = smp_load_acquire(&inode->i_flctx); if (ctx && !list_empty_careful(&ctx->flc_lease)) { @@ -5793,7 +5975,7 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c spin_lock(&ctx->flc_lock); time_out_leases(inode, &dispose); list_for_each_entry(fl, &ctx->flc_lease, fl_list) { -@@ -1619,7 +1619,7 @@ +@@ -1619,7 +1619,7 @@ int fcntl_getlease(struct file *filp) break; } spin_unlock(&ctx->flc_lock); @@ -5802,7 +5984,7 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c locks_dispose_list(&dispose); } -@@ -1694,7 +1694,7 @@ +@@ -1694,7 +1694,7 @@ generic_add_lease(struct file *filp, long arg, struct file_lock **flp, void **pr return -EINVAL; } @@ -5811,7 +5993,7 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c spin_lock(&ctx->flc_lock); time_out_leases(inode, &dispose); error = check_conflicting_open(dentry, arg, lease->fl_flags); -@@ -1765,7 +1765,7 @@ +@@ -1765,7 +1765,7 @@ generic_add_lease(struct file *filp, long arg, struct file_lock **flp, void **pr lease->fl_lmops->lm_setup(lease, priv); out: spin_unlock(&ctx->flc_lock); @@ -5820,7 +6002,7 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c locks_dispose_list(&dispose); if (is_deleg) inode_unlock(inode); -@@ -1788,7 +1788,7 @@ +@@ -1788,7 +1788,7 @@ static int generic_delete_lease(struct file *filp, void *owner) return error; } @@ -5829,7 +6011,7 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c spin_lock(&ctx->flc_lock); list_for_each_entry(fl, &ctx->flc_lease, fl_list) { if (fl->fl_file == filp && -@@ -1801,7 +1801,7 @@ +@@ -1801,7 +1801,7 @@ static int generic_delete_lease(struct file *filp, void *owner) if (victim) error = fl->fl_lmops->lm_change(victim, F_UNLCK, &dispose); spin_unlock(&ctx->flc_lock); @@ -5838,7 +6020,7 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c locks_dispose_list(&dispose); return error; } -@@ -2532,13 +2532,13 @@ +@@ -2532,13 +2532,13 @@ locks_remove_lease(struct file *filp, struct file_lock_context *ctx) if (list_empty(&ctx->flc_lease)) return; @@ -5854,10 +6036,11 @@ diff -Nur linux-4.9.28.orig/fs/locks.c linux-4.9.28/fs/locks.c locks_dispose_list(&dispose); } -diff -Nur linux-4.9.28.orig/fs/namei.c linux-4.9.28/fs/namei.c ---- linux-4.9.28.orig/fs/namei.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/namei.c 2017-05-19 03:37:25.158175874 +0200 -@@ -1626,7 +1626,7 @@ +diff --git a/fs/namei.c b/fs/namei.c +index d5e5140c1045..150fbdd8e04c 100644 +--- a/fs/namei.c ++++ b/fs/namei.c +@@ -1626,7 +1626,7 @@ static struct dentry *lookup_slow(const struct qstr *name, { struct dentry *dentry = ERR_PTR(-ENOENT), *old; struct inode *inode = dir->d_inode; @@ -5866,7 +6049,7 @@ diff -Nur linux-4.9.28.orig/fs/namei.c linux-4.9.28/fs/namei.c inode_lock_shared(inode); /* Don't go there if it's already dead */ -@@ -3083,7 +3083,7 @@ +@@ -3083,7 +3083,7 @@ static int lookup_open(struct nameidata *nd, struct path *path, struct dentry *dentry; int error, create_error = 0; umode_t mode = op->mode; @@ -5875,9 +6058,10 @@ diff -Nur linux-4.9.28.orig/fs/namei.c linux-4.9.28/fs/namei.c if (unlikely(IS_DEADDIR(dir_inode))) return -ENOENT; -diff -Nur linux-4.9.28.orig/fs/namespace.c linux-4.9.28/fs/namespace.c ---- linux-4.9.28.orig/fs/namespace.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/namespace.c 2017-05-19 03:37:25.158175874 +0200 +diff --git a/fs/namespace.c b/fs/namespace.c +index 5e35057f07ac..843d274ba167 100644 +--- a/fs/namespace.c ++++ b/fs/namespace.c @@ -14,6 +14,7 @@ #include #include @@ -5886,7 +6070,7 @@ diff -Nur linux-4.9.28.orig/fs/namespace.c linux-4.9.28/fs/namespace.c #include #include #include /* init_rootfs */ -@@ -356,8 +357,11 @@ +@@ -356,8 +357,11 @@ int __mnt_want_write(struct vfsmount *m) * incremented count after it has set MNT_WRITE_HOLD. */ smp_mb(); @@ -5900,10 +6084,11 @@ diff -Nur linux-4.9.28.orig/fs/namespace.c linux-4.9.28/fs/namespace.c /* * After the slowpath clears MNT_WRITE_HOLD, mnt_is_readonly will * be set to match its requirements. So we must not load that until -diff -Nur linux-4.9.28.orig/fs/nfs/delegation.c linux-4.9.28/fs/nfs/delegation.c ---- linux-4.9.28.orig/fs/nfs/delegation.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/nfs/delegation.c 2017-05-19 03:37:25.158175874 +0200 -@@ -150,11 +150,11 @@ +diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c +index dff600ae0d74..d726d2e09353 100644 +--- a/fs/nfs/delegation.c ++++ b/fs/nfs/delegation.c +@@ -150,11 +150,11 @@ static int nfs_delegation_claim_opens(struct inode *inode, sp = state->owner; /* Block nfs4_proc_unlck */ mutex_lock(&sp->so_delegreturn_mutex); @@ -5917,10 +6102,11 @@ diff -Nur linux-4.9.28.orig/fs/nfs/delegation.c linux-4.9.28/fs/nfs/delegation.c err = -EAGAIN; mutex_unlock(&sp->so_delegreturn_mutex); put_nfs_open_context(ctx); -diff -Nur linux-4.9.28.orig/fs/nfs/dir.c linux-4.9.28/fs/nfs/dir.c ---- linux-4.9.28.orig/fs/nfs/dir.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/nfs/dir.c 2017-05-19 03:37:25.162176025 +0200 -@@ -485,7 +485,7 @@ +diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c +index 53e02b8bd9bd..a66e7d77cfbb 100644 +--- a/fs/nfs/dir.c ++++ b/fs/nfs/dir.c +@@ -485,7 +485,7 @@ static void nfs_prime_dcache(struct dentry *parent, struct nfs_entry *entry) { struct qstr filename = QSTR_INIT(entry->name, entry->len); @@ -5929,7 +6115,7 @@ diff -Nur linux-4.9.28.orig/fs/nfs/dir.c linux-4.9.28/fs/nfs/dir.c struct dentry *dentry; struct dentry *alias; struct inode *dir = d_inode(parent); -@@ -1487,7 +1487,7 @@ +@@ -1487,7 +1487,7 @@ int nfs_atomic_open(struct inode *dir, struct dentry *dentry, struct file *file, unsigned open_flags, umode_t mode, int *opened) { @@ -5938,7 +6124,7 @@ diff -Nur linux-4.9.28.orig/fs/nfs/dir.c linux-4.9.28/fs/nfs/dir.c struct nfs_open_context *ctx; struct dentry *res; struct iattr attr = { .ia_valid = ATTR_OPEN }; -@@ -1802,7 +1802,11 @@ +@@ -1802,7 +1802,11 @@ int nfs_rmdir(struct inode *dir, struct dentry *dentry) trace_nfs_rmdir_enter(dir, dentry); if (d_really_is_positive(dentry)) { @@ -5950,7 +6136,7 @@ diff -Nur linux-4.9.28.orig/fs/nfs/dir.c linux-4.9.28/fs/nfs/dir.c error = NFS_PROTO(dir)->rmdir(dir, &dentry->d_name); /* Ensure the VFS deletes this inode */ switch (error) { -@@ -1812,7 +1816,11 @@ +@@ -1812,7 +1816,11 @@ int nfs_rmdir(struct inode *dir, struct dentry *dentry) case -ENOENT: nfs_dentry_handle_enoent(dentry); } @@ -5962,10 +6148,11 @@ diff -Nur linux-4.9.28.orig/fs/nfs/dir.c linux-4.9.28/fs/nfs/dir.c } else error = NFS_PROTO(dir)->rmdir(dir, &dentry->d_name); trace_nfs_rmdir_exit(dir, dentry, error); -diff -Nur linux-4.9.28.orig/fs/nfs/inode.c linux-4.9.28/fs/nfs/inode.c ---- linux-4.9.28.orig/fs/nfs/inode.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/nfs/inode.c 2017-05-19 03:37:25.162176025 +0200 -@@ -1957,7 +1957,11 @@ +diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c +index bf4ec5ecc97e..36cd5fc9192c 100644 +--- a/fs/nfs/inode.c ++++ b/fs/nfs/inode.c +@@ -1957,7 +1957,11 @@ static void init_once(void *foo) nfsi->nrequests = 0; nfsi->commit_info.ncommit = 0; atomic_set(&nfsi->commit_info.rpcs_out, 0); @@ -5977,10 +6164,11 @@ diff -Nur linux-4.9.28.orig/fs/nfs/inode.c linux-4.9.28/fs/nfs/inode.c nfs4_init_once(nfsi); } -diff -Nur linux-4.9.28.orig/fs/nfs/nfs4_fs.h linux-4.9.28/fs/nfs/nfs4_fs.h ---- linux-4.9.28.orig/fs/nfs/nfs4_fs.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/nfs/nfs4_fs.h 2017-05-19 03:37:25.162176025 +0200 -@@ -111,7 +111,7 @@ +diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h +index 1452177c822d..f43b01d54c59 100644 +--- a/fs/nfs/nfs4_fs.h ++++ b/fs/nfs/nfs4_fs.h +@@ -111,7 +111,7 @@ struct nfs4_state_owner { unsigned long so_flags; struct list_head so_states; struct nfs_seqid_counter so_seqid; @@ -5989,10 +6177,11 @@ diff -Nur linux-4.9.28.orig/fs/nfs/nfs4_fs.h linux-4.9.28/fs/nfs/nfs4_fs.h struct mutex so_delegreturn_mutex; }; -diff -Nur linux-4.9.28.orig/fs/nfs/nfs4proc.c linux-4.9.28/fs/nfs/nfs4proc.c ---- linux-4.9.28.orig/fs/nfs/nfs4proc.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/nfs/nfs4proc.c 2017-05-19 03:37:25.162176025 +0200 -@@ -2695,7 +2695,7 @@ +diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c +index fc9b04941739..7c9bc1c7efe7 100644 +--- a/fs/nfs/nfs4proc.c ++++ b/fs/nfs/nfs4proc.c +@@ -2697,7 +2697,7 @@ static int _nfs4_open_and_get_state(struct nfs4_opendata *opendata, unsigned int seq; int ret; @@ -6001,7 +6190,7 @@ diff -Nur linux-4.9.28.orig/fs/nfs/nfs4proc.c linux-4.9.28/fs/nfs/nfs4proc.c ret = _nfs4_proc_open(opendata); if (ret != 0) -@@ -2733,7 +2733,7 @@ +@@ -2735,7 +2735,7 @@ static int _nfs4_open_and_get_state(struct nfs4_opendata *opendata, if (d_inode(dentry) == state->inode) { nfs_inode_attach_open_context(ctx); @@ -6010,10 +6199,11 @@ diff -Nur linux-4.9.28.orig/fs/nfs/nfs4proc.c linux-4.9.28/fs/nfs/nfs4proc.c nfs4_schedule_stateid_recovery(server, state); } out: -diff -Nur linux-4.9.28.orig/fs/nfs/nfs4state.c linux-4.9.28/fs/nfs/nfs4state.c ---- linux-4.9.28.orig/fs/nfs/nfs4state.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/nfs/nfs4state.c 2017-05-19 03:37:25.162176025 +0200 -@@ -488,7 +488,7 @@ +diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c +index 0959c9661662..dabd834d7686 100644 +--- a/fs/nfs/nfs4state.c ++++ b/fs/nfs/nfs4state.c +@@ -488,7 +488,7 @@ nfs4_alloc_state_owner(struct nfs_server *server, nfs4_init_seqid_counter(&sp->so_seqid); atomic_set(&sp->so_count, 1); INIT_LIST_HEAD(&sp->so_lru); @@ -6022,7 +6212,7 @@ diff -Nur linux-4.9.28.orig/fs/nfs/nfs4state.c linux-4.9.28/fs/nfs/nfs4state.c mutex_init(&sp->so_delegreturn_mutex); return sp; } -@@ -1497,8 +1497,12 @@ +@@ -1497,8 +1497,12 @@ static int nfs4_reclaim_open_state(struct nfs4_state_owner *sp, const struct nfs * recovering after a network partition or a reboot from a * server that doesn't support a grace period. */ @@ -6036,7 +6226,7 @@ diff -Nur linux-4.9.28.orig/fs/nfs/nfs4state.c linux-4.9.28/fs/nfs/nfs4state.c restart: list_for_each_entry(state, &sp->so_states, open_states) { if (!test_and_clear_bit(ops->state_flag_bit, &state->flags)) -@@ -1567,14 +1571,20 @@ +@@ -1567,14 +1571,20 @@ static int nfs4_reclaim_open_state(struct nfs4_state_owner *sp, const struct nfs spin_lock(&sp->so_lock); goto restart; } @@ -6061,9 +6251,10 @@ diff -Nur linux-4.9.28.orig/fs/nfs/nfs4state.c linux-4.9.28/fs/nfs/nfs4state.c return status; } -diff -Nur linux-4.9.28.orig/fs/nfs/unlink.c linux-4.9.28/fs/nfs/unlink.c ---- linux-4.9.28.orig/fs/nfs/unlink.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/nfs/unlink.c 2017-05-19 03:37:25.162176025 +0200 +diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c +index 191aa577dd1f..58990c8f52e0 100644 +--- a/fs/nfs/unlink.c ++++ b/fs/nfs/unlink.c @@ -12,7 +12,7 @@ #include #include @@ -6073,7 +6264,7 @@ diff -Nur linux-4.9.28.orig/fs/nfs/unlink.c linux-4.9.28/fs/nfs/unlink.c #include #include -@@ -51,6 +51,29 @@ +@@ -51,6 +51,29 @@ static void nfs_async_unlink_done(struct rpc_task *task, void *calldata) rpc_restart_call_prepare(task); } @@ -6103,7 +6294,7 @@ diff -Nur linux-4.9.28.orig/fs/nfs/unlink.c linux-4.9.28/fs/nfs/unlink.c /** * nfs_async_unlink_release - Release the sillydelete data. * @task: rpc_task of the sillydelete -@@ -64,7 +87,7 @@ +@@ -64,7 +87,7 @@ static void nfs_async_unlink_release(void *calldata) struct dentry *dentry = data->dentry; struct super_block *sb = dentry->d_sb; @@ -6112,7 +6303,7 @@ diff -Nur linux-4.9.28.orig/fs/nfs/unlink.c linux-4.9.28/fs/nfs/unlink.c d_lookup_done(dentry); nfs_free_unlinkdata(data); dput(dentry); -@@ -117,10 +140,10 @@ +@@ -117,10 +140,10 @@ static int nfs_call_unlink(struct dentry *dentry, struct nfs_unlinkdata *data) struct inode *dir = d_inode(dentry->d_parent); struct dentry *alias; @@ -6125,7 +6316,7 @@ diff -Nur linux-4.9.28.orig/fs/nfs/unlink.c linux-4.9.28/fs/nfs/unlink.c return 0; } if (!d_in_lookup(alias)) { -@@ -142,7 +165,7 @@ +@@ -142,7 +165,7 @@ static int nfs_call_unlink(struct dentry *dentry, struct nfs_unlinkdata *data) ret = 0; spin_unlock(&alias->d_lock); dput(alias); @@ -6134,7 +6325,7 @@ diff -Nur linux-4.9.28.orig/fs/nfs/unlink.c linux-4.9.28/fs/nfs/unlink.c /* * If we'd displaced old cached devname, free it. At that * point dentry is definitely not a root, so we won't need -@@ -182,7 +205,7 @@ +@@ -182,7 +205,7 @@ nfs_async_unlink(struct dentry *dentry, const struct qstr *name) goto out_free_name; } data->res.dir_attr = &data->dir_attr; @@ -6143,10 +6334,11 @@ diff -Nur linux-4.9.28.orig/fs/nfs/unlink.c linux-4.9.28/fs/nfs/unlink.c status = -EBUSY; spin_lock(&dentry->d_lock); -diff -Nur linux-4.9.28.orig/fs/ntfs/aops.c linux-4.9.28/fs/ntfs/aops.c ---- linux-4.9.28.orig/fs/ntfs/aops.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/ntfs/aops.c 2017-05-19 03:37:25.162176025 +0200 -@@ -92,13 +92,13 @@ +diff --git a/fs/ntfs/aops.c b/fs/ntfs/aops.c +index fe251f187ff8..e89da4fb14c2 100644 +--- a/fs/ntfs/aops.c ++++ b/fs/ntfs/aops.c +@@ -92,13 +92,13 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) ofs = 0; if (file_ofs < init_size) ofs = init_size - file_ofs; @@ -6162,7 +6354,7 @@ diff -Nur linux-4.9.28.orig/fs/ntfs/aops.c linux-4.9.28/fs/ntfs/aops.c } } else { clear_buffer_uptodate(bh); -@@ -107,8 +107,7 @@ +@@ -107,8 +107,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) "0x%llx.", (unsigned long long)bh->b_blocknr); } first = page_buffers(page); @@ -6172,7 +6364,7 @@ diff -Nur linux-4.9.28.orig/fs/ntfs/aops.c linux-4.9.28/fs/ntfs/aops.c clear_buffer_async_read(bh); unlock_buffer(bh); tmp = bh; -@@ -123,8 +122,7 @@ +@@ -123,8 +122,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) } tmp = tmp->b_this_page; } while (tmp != bh); @@ -6182,7 +6374,7 @@ diff -Nur linux-4.9.28.orig/fs/ntfs/aops.c linux-4.9.28/fs/ntfs/aops.c /* * If none of the buffers had errors then we can set the page uptodate, * but we first have to perform the post read mst fixups, if the -@@ -145,13 +143,13 @@ +@@ -145,13 +143,13 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) recs = PAGE_SIZE / rec_size; /* Should have been verified before we got here... */ BUG_ON(!recs); @@ -6198,7 +6390,7 @@ diff -Nur linux-4.9.28.orig/fs/ntfs/aops.c linux-4.9.28/fs/ntfs/aops.c flush_dcache_page(page); if (likely(page_uptodate && !PageError(page))) SetPageUptodate(page); -@@ -159,9 +157,7 @@ +@@ -159,9 +157,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) unlock_page(page); return; still_busy: @@ -6209,10 +6401,11 @@ diff -Nur linux-4.9.28.orig/fs/ntfs/aops.c linux-4.9.28/fs/ntfs/aops.c } /** -diff -Nur linux-4.9.28.orig/fs/proc/base.c linux-4.9.28/fs/proc/base.c ---- linux-4.9.28.orig/fs/proc/base.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/proc/base.c 2017-05-19 03:37:25.162176025 +0200 -@@ -1834,7 +1834,7 @@ +diff --git a/fs/proc/base.c b/fs/proc/base.c +index ca651ac00660..41d9dc789285 100644 +--- a/fs/proc/base.c ++++ b/fs/proc/base.c +@@ -1834,7 +1834,7 @@ bool proc_fill_cache(struct file *file, struct dir_context *ctx, child = d_hash_and_lookup(dir, &qname); if (!child) { @@ -6221,10 +6414,11 @@ diff -Nur linux-4.9.28.orig/fs/proc/base.c linux-4.9.28/fs/proc/base.c child = d_alloc_parallel(dir, &qname, &wq); if (IS_ERR(child)) goto end_instantiate; -diff -Nur linux-4.9.28.orig/fs/proc/proc_sysctl.c linux-4.9.28/fs/proc/proc_sysctl.c ---- linux-4.9.28.orig/fs/proc/proc_sysctl.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/proc/proc_sysctl.c 2017-05-19 03:37:25.162176025 +0200 -@@ -632,7 +632,7 @@ +diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c +index d4e37acd4821..000cea46434a 100644 +--- a/fs/proc/proc_sysctl.c ++++ b/fs/proc/proc_sysctl.c +@@ -632,7 +632,7 @@ static bool proc_sys_fill_cache(struct file *file, child = d_lookup(dir, &qname); if (!child) { @@ -6233,10 +6427,11 @@ diff -Nur linux-4.9.28.orig/fs/proc/proc_sysctl.c linux-4.9.28/fs/proc/proc_sysc child = d_alloc_parallel(dir, &qname, &wq); if (IS_ERR(child)) return false; -diff -Nur linux-4.9.28.orig/fs/timerfd.c linux-4.9.28/fs/timerfd.c ---- linux-4.9.28.orig/fs/timerfd.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/fs/timerfd.c 2017-05-19 03:37:25.162176025 +0200 -@@ -471,7 +471,10 @@ +diff --git a/fs/timerfd.c b/fs/timerfd.c +index ab8dd1538381..5580853f57dd 100644 +--- a/fs/timerfd.c ++++ b/fs/timerfd.c +@@ -471,7 +471,10 @@ static int do_timerfd_settime(int ufd, int flags, break; } spin_unlock_irq(&ctx->wqh.lock); @@ -6248,9 +6443,10 @@ diff -Nur linux-4.9.28.orig/fs/timerfd.c linux-4.9.28/fs/timerfd.c } /* -diff -Nur linux-4.9.28.orig/include/acpi/platform/aclinux.h linux-4.9.28/include/acpi/platform/aclinux.h ---- linux-4.9.28.orig/include/acpi/platform/aclinux.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/acpi/platform/aclinux.h 2017-05-19 03:37:25.166176182 +0200 +diff --git a/include/acpi/platform/aclinux.h b/include/acpi/platform/aclinux.h +index e861a24f06f2..b5c97d3059c7 100644 +--- a/include/acpi/platform/aclinux.h ++++ b/include/acpi/platform/aclinux.h @@ -133,6 +133,7 @@ #define acpi_cache_t struct kmem_cache @@ -6280,10 +6476,11 @@ diff -Nur linux-4.9.28.orig/include/acpi/platform/aclinux.h linux-4.9.28/include /* * OSL interfaces used by debugger/disassembler */ -diff -Nur linux-4.9.28.orig/include/asm-generic/bug.h linux-4.9.28/include/asm-generic/bug.h ---- linux-4.9.28.orig/include/asm-generic/bug.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/asm-generic/bug.h 2017-05-19 03:37:25.166176182 +0200 -@@ -215,6 +215,20 @@ +diff --git a/include/asm-generic/bug.h b/include/asm-generic/bug.h +index 6f96247226a4..fa53a21263c2 100644 +--- a/include/asm-generic/bug.h ++++ b/include/asm-generic/bug.h +@@ -215,6 +215,20 @@ void __warn(const char *file, int line, void *caller, unsigned taint, # define WARN_ON_SMP(x) ({0;}) #endif @@ -6304,10 +6501,24 @@ diff -Nur linux-4.9.28.orig/include/asm-generic/bug.h linux-4.9.28/include/asm-g #endif /* __ASSEMBLY__ */ #endif -diff -Nur linux-4.9.28.orig/include/linux/blkdev.h linux-4.9.28/include/linux/blkdev.h ---- linux-4.9.28.orig/include/linux/blkdev.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/blkdev.h 2017-05-19 03:37:25.166176182 +0200 -@@ -89,6 +89,7 @@ +diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h +index 535ab2e13d2e..cfc246899473 100644 +--- a/include/linux/blk-mq.h ++++ b/include/linux/blk-mq.h +@@ -209,7 +209,7 @@ static inline u16 blk_mq_unique_tag_to_tag(u32 unique_tag) + return unique_tag & BLK_MQ_UNIQUE_TAG_MASK; + } + +- ++void __blk_mq_complete_request_remote_work(struct work_struct *work); + int blk_mq_request_started(struct request *rq); + void blk_mq_start_request(struct request *rq); + void blk_mq_end_request(struct request *rq, int error); +diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h +index f6a816129856..ec7a4676f8a8 100644 +--- a/include/linux/blkdev.h ++++ b/include/linux/blkdev.h +@@ -89,6 +89,7 @@ struct request { struct list_head queuelist; union { struct call_single_data csd; @@ -6315,30 +6526,19 @@ diff -Nur linux-4.9.28.orig/include/linux/blkdev.h linux-4.9.28/include/linux/bl u64 fifo_time; }; -@@ -467,7 +468,7 @@ +@@ -467,7 +468,7 @@ struct request_queue { struct throtl_data *td; #endif struct rcu_head rcu_head; - wait_queue_head_t mq_freeze_wq; + struct swait_queue_head mq_freeze_wq; - struct percpu_ref q_usage_counter; - struct list_head all_q_node; - -diff -Nur linux-4.9.28.orig/include/linux/blk-mq.h linux-4.9.28/include/linux/blk-mq.h ---- linux-4.9.28.orig/include/linux/blk-mq.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/blk-mq.h 2017-05-19 03:37:25.166176182 +0200 -@@ -209,7 +209,7 @@ - return unique_tag & BLK_MQ_UNIQUE_TAG_MASK; - } + struct percpu_ref q_usage_counter; + struct list_head all_q_node; -- -+void __blk_mq_complete_request_remote_work(struct work_struct *work); - int blk_mq_request_started(struct request *rq); - void blk_mq_start_request(struct request *rq); - void blk_mq_end_request(struct request *rq, int error); -diff -Nur linux-4.9.28.orig/include/linux/bottom_half.h linux-4.9.28/include/linux/bottom_half.h ---- linux-4.9.28.orig/include/linux/bottom_half.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/bottom_half.h 2017-05-19 03:37:25.166176182 +0200 +diff --git a/include/linux/bottom_half.h b/include/linux/bottom_half.h +index 8fdcb783197d..d07dbeec7bc1 100644 +--- a/include/linux/bottom_half.h ++++ b/include/linux/bottom_half.h @@ -3,6 +3,39 @@ #include @@ -6379,17 +6579,18 @@ diff -Nur linux-4.9.28.orig/include/linux/bottom_half.h linux-4.9.28/include/lin #ifdef CONFIG_TRACE_IRQFLAGS extern void __local_bh_disable_ip(unsigned long ip, unsigned int cnt); #else -@@ -30,5 +63,6 @@ +@@ -30,5 +63,6 @@ static inline void local_bh_enable(void) { __local_bh_enable_ip(_THIS_IP_, SOFTIRQ_DISABLE_OFFSET); } +#endif #endif /* _LINUX_BH_H */ -diff -Nur linux-4.9.28.orig/include/linux/buffer_head.h linux-4.9.28/include/linux/buffer_head.h ---- linux-4.9.28.orig/include/linux/buffer_head.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/buffer_head.h 2017-05-19 03:37:25.166176182 +0200 -@@ -75,8 +75,50 @@ +diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h +index ebbacd14d450..be5e87f6360a 100644 +--- a/include/linux/buffer_head.h ++++ b/include/linux/buffer_head.h +@@ -75,8 +75,50 @@ struct buffer_head { struct address_space *b_assoc_map; /* mapping this buffer is associated with */ atomic_t b_count; /* users using this buffer_head */ @@ -6440,9 +6641,10 @@ diff -Nur linux-4.9.28.orig/include/linux/buffer_head.h linux-4.9.28/include/lin /* * macro tricks to expand the set_buffer_foo(), clear_buffer_foo() * and buffer_foo() functions. -diff -Nur linux-4.9.28.orig/include/linux/cgroup-defs.h linux-4.9.28/include/linux/cgroup-defs.h ---- linux-4.9.28.orig/include/linux/cgroup-defs.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/cgroup-defs.h 2017-05-19 03:37:25.166176182 +0200 +diff --git a/include/linux/cgroup-defs.h b/include/linux/cgroup-defs.h +index 5b17de62c962..56027cc01a56 100644 +--- a/include/linux/cgroup-defs.h ++++ b/include/linux/cgroup-defs.h @@ -16,6 +16,7 @@ #include #include @@ -6451,7 +6653,7 @@ diff -Nur linux-4.9.28.orig/include/linux/cgroup-defs.h linux-4.9.28/include/lin #ifdef CONFIG_CGROUPS -@@ -137,6 +138,7 @@ +@@ -137,6 +138,7 @@ struct cgroup_subsys_state { /* percpu_ref killing and RCU release */ struct rcu_head rcu_head; struct work_struct destroy_work; @@ -6459,9 +6661,10 @@ diff -Nur linux-4.9.28.orig/include/linux/cgroup-defs.h linux-4.9.28/include/lin }; /* -diff -Nur linux-4.9.28.orig/include/linux/completion.h linux-4.9.28/include/linux/completion.h ---- linux-4.9.28.orig/include/linux/completion.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/completion.h 2017-05-19 03:37:25.166176182 +0200 +diff --git a/include/linux/completion.h b/include/linux/completion.h +index 5d5aaae3af43..3bca1590e29f 100644 +--- a/include/linux/completion.h ++++ b/include/linux/completion.h @@ -7,8 +7,7 @@ * Atomic wait-for-completion handler data structures. * See kernel/sched/completion.c for details. @@ -6486,7 +6689,7 @@ diff -Nur linux-4.9.28.orig/include/linux/completion.h linux-4.9.28/include/linu #define COMPLETION_INITIALIZER_ONSTACK(work) \ ({ init_completion(&work); work; }) -@@ -73,7 +72,7 @@ +@@ -73,7 +72,7 @@ struct completion { static inline void init_completion(struct completion *x) { x->done = 0; @@ -6495,10 +6698,11 @@ diff -Nur linux-4.9.28.orig/include/linux/completion.h linux-4.9.28/include/linu } /** -diff -Nur linux-4.9.28.orig/include/linux/cpu.h linux-4.9.28/include/linux/cpu.h ---- linux-4.9.28.orig/include/linux/cpu.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/cpu.h 2017-05-19 03:37:25.166176182 +0200 -@@ -182,6 +182,8 @@ +diff --git a/include/linux/cpu.h b/include/linux/cpu.h +index e571128ad99a..5e52d28c20c1 100644 +--- a/include/linux/cpu.h ++++ b/include/linux/cpu.h +@@ -182,6 +182,8 @@ extern void get_online_cpus(void); extern void put_online_cpus(void); extern void cpu_hotplug_disable(void); extern void cpu_hotplug_enable(void); @@ -6507,7 +6711,7 @@ diff -Nur linux-4.9.28.orig/include/linux/cpu.h linux-4.9.28/include/linux/cpu.h #define hotcpu_notifier(fn, pri) cpu_notifier(fn, pri) #define __hotcpu_notifier(fn, pri) __cpu_notifier(fn, pri) #define register_hotcpu_notifier(nb) register_cpu_notifier(nb) -@@ -199,6 +201,8 @@ +@@ -199,6 +201,8 @@ static inline void cpu_hotplug_done(void) {} #define put_online_cpus() do { } while (0) #define cpu_hotplug_disable() do { } while (0) #define cpu_hotplug_enable() do { } while (0) @@ -6516,9 +6720,10 @@ diff -Nur linux-4.9.28.orig/include/linux/cpu.h linux-4.9.28/include/linux/cpu.h #define hotcpu_notifier(fn, pri) do { (void)(fn); } while (0) #define __hotcpu_notifier(fn, pri) do { (void)(fn); } while (0) /* These aren't inline functions due to a GCC bug. */ -diff -Nur linux-4.9.28.orig/include/linux/dcache.h linux-4.9.28/include/linux/dcache.h ---- linux-4.9.28.orig/include/linux/dcache.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/dcache.h 2017-05-19 03:37:25.166176182 +0200 +diff --git a/include/linux/dcache.h b/include/linux/dcache.h +index 5beed7b30561..61cab7ef458e 100644 +--- a/include/linux/dcache.h ++++ b/include/linux/dcache.h @@ -11,6 +11,7 @@ #include #include @@ -6527,7 +6732,7 @@ diff -Nur linux-4.9.28.orig/include/linux/dcache.h linux-4.9.28/include/linux/dc struct path; struct vfsmount; -@@ -100,7 +101,7 @@ +@@ -100,7 +101,7 @@ struct dentry { union { struct list_head d_lru; /* LRU list */ @@ -6536,7 +6741,7 @@ diff -Nur linux-4.9.28.orig/include/linux/dcache.h linux-4.9.28/include/linux/dc }; struct list_head d_child; /* child of parent list */ struct list_head d_subdirs; /* our children */ -@@ -230,7 +231,7 @@ +@@ -230,7 +231,7 @@ extern void d_set_d_op(struct dentry *dentry, const struct dentry_operations *op extern struct dentry * d_alloc(struct dentry *, const struct qstr *); extern struct dentry * d_alloc_pseudo(struct super_block *, const struct qstr *); extern struct dentry * d_alloc_parallel(struct dentry *, const struct qstr *, @@ -6545,10 +6750,11 @@ diff -Nur linux-4.9.28.orig/include/linux/dcache.h linux-4.9.28/include/linux/dc extern struct dentry * d_splice_alias(struct inode *, struct dentry *); extern struct dentry * d_add_ci(struct dentry *, struct inode *, struct qstr *); extern struct dentry * d_exact_alias(struct dentry *, struct inode *); -diff -Nur linux-4.9.28.orig/include/linux/delay.h linux-4.9.28/include/linux/delay.h ---- linux-4.9.28.orig/include/linux/delay.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/delay.h 2017-05-19 03:37:25.166176182 +0200 -@@ -52,4 +52,10 @@ +diff --git a/include/linux/delay.h b/include/linux/delay.h +index a6ecb34cf547..37caab306336 100644 +--- a/include/linux/delay.h ++++ b/include/linux/delay.h +@@ -52,4 +52,10 @@ static inline void ssleep(unsigned int seconds) msleep(seconds * 1000); } @@ -6559,9 +6765,10 @@ diff -Nur linux-4.9.28.orig/include/linux/delay.h linux-4.9.28/include/linux/del +#endif + #endif /* defined(_LINUX_DELAY_H) */ -diff -Nur linux-4.9.28.orig/include/linux/highmem.h linux-4.9.28/include/linux/highmem.h ---- linux-4.9.28.orig/include/linux/highmem.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/highmem.h 2017-05-19 03:37:25.166176182 +0200 +diff --git a/include/linux/highmem.h b/include/linux/highmem.h +index bb3f3297062a..a117a33ef72c 100644 +--- a/include/linux/highmem.h ++++ b/include/linux/highmem.h @@ -7,6 +7,7 @@ #include #include @@ -6570,7 +6777,7 @@ diff -Nur linux-4.9.28.orig/include/linux/highmem.h linux-4.9.28/include/linux/h #include -@@ -65,7 +66,7 @@ +@@ -65,7 +66,7 @@ static inline void kunmap(struct page *page) static inline void *kmap_atomic(struct page *page) { @@ -6579,7 +6786,7 @@ diff -Nur linux-4.9.28.orig/include/linux/highmem.h linux-4.9.28/include/linux/h pagefault_disable(); return page_address(page); } -@@ -74,7 +75,7 @@ +@@ -74,7 +75,7 @@ static inline void *kmap_atomic(struct page *page) static inline void __kunmap_atomic(void *addr) { pagefault_enable(); @@ -6588,7 +6795,7 @@ diff -Nur linux-4.9.28.orig/include/linux/highmem.h linux-4.9.28/include/linux/h } #define kmap_atomic_pfn(pfn) kmap_atomic(pfn_to_page(pfn)) -@@ -86,32 +87,51 @@ +@@ -86,32 +87,51 @@ static inline void __kunmap_atomic(void *addr) #if defined(CONFIG_HIGHMEM) || defined(CONFIG_X86_32) @@ -6644,10 +6851,11 @@ diff -Nur linux-4.9.28.orig/include/linux/highmem.h linux-4.9.28/include/linux/h #endif } -diff -Nur linux-4.9.28.orig/include/linux/hrtimer.h linux-4.9.28/include/linux/hrtimer.h ---- linux-4.9.28.orig/include/linux/hrtimer.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/hrtimer.h 2017-05-19 03:37:25.166176182 +0200 -@@ -87,6 +87,9 @@ +diff --git a/include/linux/hrtimer.h b/include/linux/hrtimer.h +index 5e00f80b1535..a34e10b55cde 100644 +--- a/include/linux/hrtimer.h ++++ b/include/linux/hrtimer.h +@@ -87,6 +87,9 @@ enum hrtimer_restart { * @function: timer expiry callback function * @base: pointer to the timer base (per cpu and per clock) * @state: state information (See bit values above) @@ -6657,7 +6865,7 @@ diff -Nur linux-4.9.28.orig/include/linux/hrtimer.h linux-4.9.28/include/linux/h * @is_rel: Set if the timer was armed relative * @start_pid: timer statistics field to store the pid of the task which * started the timer -@@ -103,6 +106,11 @@ +@@ -103,6 +106,11 @@ struct hrtimer { enum hrtimer_restart (*function)(struct hrtimer *); struct hrtimer_clock_base *base; u8 state; @@ -6669,7 +6877,7 @@ diff -Nur linux-4.9.28.orig/include/linux/hrtimer.h linux-4.9.28/include/linux/h u8 is_rel; #ifdef CONFIG_TIMER_STATS int start_pid; -@@ -123,11 +131,7 @@ +@@ -123,11 +131,7 @@ struct hrtimer_sleeper { struct task_struct *task; }; @@ -6681,7 +6889,7 @@ diff -Nur linux-4.9.28.orig/include/linux/hrtimer.h linux-4.9.28/include/linux/h /** * struct hrtimer_clock_base - the timer base for a specific clock -@@ -136,6 +140,7 @@ +@@ -136,6 +140,7 @@ struct hrtimer_sleeper { * timer to a base on another cpu. * @clockid: clock id for per_cpu support * @active: red black tree root node for the active timers @@ -6689,7 +6897,7 @@ diff -Nur linux-4.9.28.orig/include/linux/hrtimer.h linux-4.9.28/include/linux/h * @get_time: function to retrieve the current time of the clock * @offset: offset of this clock to the monotonic base */ -@@ -144,6 +149,7 @@ +@@ -144,6 +149,7 @@ struct hrtimer_clock_base { int index; clockid_t clockid; struct timerqueue_head active; @@ -6697,7 +6905,7 @@ diff -Nur linux-4.9.28.orig/include/linux/hrtimer.h linux-4.9.28/include/linux/h ktime_t (*get_time)(void); ktime_t offset; } __attribute__((__aligned__(HRTIMER_CLOCK_BASE_ALIGN))); -@@ -187,6 +193,7 @@ +@@ -187,6 +193,7 @@ struct hrtimer_cpu_base { raw_spinlock_t lock; seqcount_t seq; struct hrtimer *running; @@ -6705,7 +6913,7 @@ diff -Nur linux-4.9.28.orig/include/linux/hrtimer.h linux-4.9.28/include/linux/h unsigned int cpu; unsigned int active_bases; unsigned int clock_was_set_seq; -@@ -203,6 +210,9 @@ +@@ -203,6 +210,9 @@ struct hrtimer_cpu_base { unsigned int nr_hangs; unsigned int max_hang_time; #endif @@ -6715,7 +6923,7 @@ diff -Nur linux-4.9.28.orig/include/linux/hrtimer.h linux-4.9.28/include/linux/h struct hrtimer_clock_base clock_base[HRTIMER_MAX_CLOCK_BASES]; } ____cacheline_aligned; -@@ -412,6 +422,13 @@ +@@ -412,6 +422,13 @@ static inline void hrtimer_restart(struct hrtimer *timer) hrtimer_start_expires(timer, HRTIMER_MODE_ABS); } @@ -6729,7 +6937,7 @@ diff -Nur linux-4.9.28.orig/include/linux/hrtimer.h linux-4.9.28/include/linux/h /* Query timers: */ extern ktime_t __hrtimer_get_remaining(const struct hrtimer *timer, bool adjust); -@@ -436,9 +453,15 @@ +@@ -436,9 +453,15 @@ static inline int hrtimer_is_queued(struct hrtimer *timer) * Helper function to check, whether the timer is running the callback * function */ @@ -6747,10 +6955,11 @@ diff -Nur linux-4.9.28.orig/include/linux/hrtimer.h linux-4.9.28/include/linux/h } /* Forward a hrtimer so it expires after now: */ -diff -Nur linux-4.9.28.orig/include/linux/idr.h linux-4.9.28/include/linux/idr.h ---- linux-4.9.28.orig/include/linux/idr.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/idr.h 2017-05-19 03:37:25.166176182 +0200 -@@ -95,10 +95,14 @@ +diff --git a/include/linux/idr.h b/include/linux/idr.h +index 083d61e92706..5899796f50cb 100644 +--- a/include/linux/idr.h ++++ b/include/linux/idr.h +@@ -95,10 +95,14 @@ bool idr_is_empty(struct idr *idp); * Each idr_preload() should be matched with an invocation of this * function. See idr_preload() for details. */ @@ -6765,10 +6974,11 @@ diff -Nur linux-4.9.28.orig/include/linux/idr.h linux-4.9.28/include/linux/idr.h /** * idr_find - return pointer for given id -diff -Nur linux-4.9.28.orig/include/linux/init_task.h linux-4.9.28/include/linux/init_task.h ---- linux-4.9.28.orig/include/linux/init_task.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/init_task.h 2017-05-19 03:37:25.166176182 +0200 -@@ -150,6 +150,12 @@ +diff --git a/include/linux/init_task.h b/include/linux/init_task.h +index 325f649d77ff..a56e263f5005 100644 +--- a/include/linux/init_task.h ++++ b/include/linux/init_task.h +@@ -150,6 +150,12 @@ extern struct task_group root_task_group; # define INIT_PERF_EVENTS(tsk) #endif @@ -6781,7 +6991,7 @@ diff -Nur linux-4.9.28.orig/include/linux/init_task.h linux-4.9.28/include/linux #ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN # define INIT_VTIME(tsk) \ .vtime_seqcount = SEQCNT_ZERO(tsk.vtime_seqcount), \ -@@ -164,6 +170,7 @@ +@@ -164,6 +170,7 @@ extern struct task_group root_task_group; #ifdef CONFIG_RT_MUTEXES # define INIT_RT_MUTEXES(tsk) \ .pi_waiters = RB_ROOT, \ @@ -6789,7 +6999,7 @@ diff -Nur linux-4.9.28.orig/include/linux/init_task.h linux-4.9.28/include/linux .pi_waiters_leftmost = NULL, #else # define INIT_RT_MUTEXES(tsk) -@@ -250,6 +257,7 @@ +@@ -250,6 +257,7 @@ extern struct task_group root_task_group; .cpu_timers = INIT_CPU_TIMERS(tsk.cpu_timers), \ .pi_lock = __RAW_SPIN_LOCK_UNLOCKED(tsk.pi_lock), \ .timer_slack_ns = 50000, /* 50 usec default slack */ \ @@ -6797,9 +7007,10 @@ diff -Nur linux-4.9.28.orig/include/linux/init_task.h linux-4.9.28/include/linux .pids = { \ [PIDTYPE_PID] = INIT_PID_LINK(PIDTYPE_PID), \ [PIDTYPE_PGID] = INIT_PID_LINK(PIDTYPE_PGID), \ -diff -Nur linux-4.9.28.orig/include/linux/interrupt.h linux-4.9.28/include/linux/interrupt.h ---- linux-4.9.28.orig/include/linux/interrupt.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/interrupt.h 2017-05-19 03:37:25.166176182 +0200 +diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h +index 72f0721f75e7..480972ae47d3 100644 +--- a/include/linux/interrupt.h ++++ b/include/linux/interrupt.h @@ -14,6 +14,7 @@ #include #include @@ -6824,7 +7035,7 @@ diff -Nur linux-4.9.28.orig/include/linux/interrupt.h linux-4.9.28/include/linux #define IRQF_TIMER (__IRQF_TIMER | IRQF_NO_SUSPEND | IRQF_NO_THREAD) -@@ -196,7 +199,7 @@ +@@ -196,7 +199,7 @@ extern void devm_free_irq(struct device *dev, unsigned int irq, void *dev_id); #ifdef CONFIG_LOCKDEP # define local_irq_enable_in_hardirq() do { } while (0) #else @@ -6833,7 +7044,7 @@ diff -Nur linux-4.9.28.orig/include/linux/interrupt.h linux-4.9.28/include/linux #endif extern void disable_irq_nosync(unsigned int irq); -@@ -216,6 +219,7 @@ +@@ -216,6 +219,7 @@ extern void resume_device_irqs(void); * struct irq_affinity_notify - context for notification of IRQ affinity changes * @irq: Interrupt to which notification applies * @kref: Reference count, for internal use @@ -6841,7 +7052,7 @@ diff -Nur linux-4.9.28.orig/include/linux/interrupt.h linux-4.9.28/include/linux * @work: Work item, for internal use * @notify: Function to be called on change. This will be * called in process context. -@@ -227,7 +231,11 @@ +@@ -227,7 +231,11 @@ extern void resume_device_irqs(void); struct irq_affinity_notify { unsigned int irq; struct kref kref; @@ -6853,7 +7064,7 @@ diff -Nur linux-4.9.28.orig/include/linux/interrupt.h linux-4.9.28/include/linux void (*notify)(struct irq_affinity_notify *, const cpumask_t *mask); void (*release)(struct kref *ref); }; -@@ -406,9 +414,13 @@ +@@ -406,9 +414,13 @@ extern int irq_set_irqchip_state(unsigned int irq, enum irqchip_irq_state which, bool state); #ifdef CONFIG_IRQ_FORCED_THREADING @@ -6868,7 +7079,7 @@ diff -Nur linux-4.9.28.orig/include/linux/interrupt.h linux-4.9.28/include/linux #endif #ifndef __ARCH_SET_SOFTIRQ_PENDING -@@ -465,9 +477,10 @@ +@@ -465,9 +477,10 @@ struct softirq_action void (*action)(struct softirq_action *); }; @@ -6880,7 +7091,7 @@ diff -Nur linux-4.9.28.orig/include/linux/interrupt.h linux-4.9.28/include/linux #ifdef __ARCH_HAS_DO_SOFTIRQ void do_softirq_own_stack(void); #else -@@ -476,13 +489,25 @@ +@@ -476,13 +489,25 @@ static inline void do_softirq_own_stack(void) __do_softirq(); } #endif @@ -6906,7 +7117,7 @@ diff -Nur linux-4.9.28.orig/include/linux/interrupt.h linux-4.9.28/include/linux DECLARE_PER_CPU(struct task_struct *, ksoftirqd); -@@ -504,8 +529,9 @@ +@@ -504,8 +529,9 @@ static inline struct task_struct *this_cpu_ksoftirqd(void) to be executed on some cpu at least once after this. * If the tasklet is already scheduled, but its execution is still not started, it will be executed only once. @@ -6918,7 +7129,7 @@ diff -Nur linux-4.9.28.orig/include/linux/interrupt.h linux-4.9.28/include/linux * Tasklet is strictly serialized wrt itself, but not wrt another tasklets. If client needs some intertask synchronization, he makes it with spinlocks. -@@ -530,27 +556,36 @@ +@@ -530,27 +556,36 @@ struct tasklet_struct name = { NULL, 0, ATOMIC_INIT(1), func, data } enum { TASKLET_STATE_SCHED, /* Tasklet is scheduled for execution */ @@ -6961,7 +7172,7 @@ diff -Nur linux-4.9.28.orig/include/linux/interrupt.h linux-4.9.28/include/linux #define tasklet_unlock_wait(t) do { } while (0) #define tasklet_unlock(t) do { } while (0) #endif -@@ -599,12 +634,7 @@ +@@ -599,12 +634,7 @@ static inline void tasklet_disable(struct tasklet_struct *t) smp_mb(); } @@ -6975,7 +7186,7 @@ diff -Nur linux-4.9.28.orig/include/linux/interrupt.h linux-4.9.28/include/linux extern void tasklet_kill(struct tasklet_struct *t); extern void tasklet_kill_immediate(struct tasklet_struct *t, unsigned int cpu); extern void tasklet_init(struct tasklet_struct *t, -@@ -635,6 +665,12 @@ +@@ -635,6 +665,12 @@ void tasklet_hrtimer_cancel(struct tasklet_hrtimer *ttimer) tasklet_kill(&ttimer->tasklet); } @@ -6988,10 +7199,62 @@ diff -Nur linux-4.9.28.orig/include/linux/interrupt.h linux-4.9.28/include/linux /* * Autoprobing for irqs: * -diff -Nur linux-4.9.28.orig/include/linux/irqdesc.h linux-4.9.28/include/linux/irqdesc.h ---- linux-4.9.28.orig/include/linux/irqdesc.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/irqdesc.h 2017-05-19 03:37:25.166176182 +0200 -@@ -66,6 +66,7 @@ +diff --git a/include/linux/irq.h b/include/linux/irq.h +index 39e3254e5769..8ebac94fbb9f 100644 +--- a/include/linux/irq.h ++++ b/include/linux/irq.h +@@ -72,6 +72,7 @@ enum irqchip_irq_state; + * IRQ_IS_POLLED - Always polled by another interrupt. Exclude + * it from the spurious interrupt detection + * mechanism and from core side polling. ++ * IRQ_NO_SOFTIRQ_CALL - No softirq processing in the irq thread context (RT) + * IRQ_DISABLE_UNLAZY - Disable lazy irq disable + */ + enum { +@@ -99,13 +100,14 @@ enum { + IRQ_PER_CPU_DEVID = (1 << 17), + IRQ_IS_POLLED = (1 << 18), + IRQ_DISABLE_UNLAZY = (1 << 19), ++ IRQ_NO_SOFTIRQ_CALL = (1 << 20), + }; + + #define IRQF_MODIFY_MASK \ + (IRQ_TYPE_SENSE_MASK | IRQ_NOPROBE | IRQ_NOREQUEST | \ + IRQ_NOAUTOEN | IRQ_MOVE_PCNTXT | IRQ_LEVEL | IRQ_NO_BALANCING | \ + IRQ_PER_CPU | IRQ_NESTED_THREAD | IRQ_NOTHREAD | IRQ_PER_CPU_DEVID | \ +- IRQ_IS_POLLED | IRQ_DISABLE_UNLAZY) ++ IRQ_IS_POLLED | IRQ_DISABLE_UNLAZY | IRQ_NO_SOFTIRQ_CALL) + + #define IRQ_NO_BALANCING_MASK (IRQ_PER_CPU | IRQ_NO_BALANCING) + +diff --git a/include/linux/irq_work.h b/include/linux/irq_work.h +index 47b9ebd4a74f..2543aab05daa 100644 +--- a/include/linux/irq_work.h ++++ b/include/linux/irq_work.h +@@ -16,6 +16,7 @@ + #define IRQ_WORK_BUSY 2UL + #define IRQ_WORK_FLAGS 3UL + #define IRQ_WORK_LAZY 4UL /* Doesn't want IPI, wait for tick */ ++#define IRQ_WORK_HARD_IRQ 8UL /* Run hard IRQ context, even on RT */ + + struct irq_work { + unsigned long flags; +@@ -51,4 +52,10 @@ static inline bool irq_work_needs_cpu(void) { return false; } + static inline void irq_work_run(void) { } + #endif + ++#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) ++void irq_work_tick_soft(void); ++#else ++static inline void irq_work_tick_soft(void) { } ++#endif ++ + #endif /* _LINUX_IRQ_WORK_H */ +diff --git a/include/linux/irqdesc.h b/include/linux/irqdesc.h +index c9be57931b58..eeeb540971ae 100644 +--- a/include/linux/irqdesc.h ++++ b/include/linux/irqdesc.h +@@ -66,6 +66,7 @@ struct irq_desc { unsigned int irqs_unhandled; atomic_t threads_handled; int threads_handled_last; @@ -6999,9 +7262,10 @@ diff -Nur linux-4.9.28.orig/include/linux/irqdesc.h linux-4.9.28/include/linux/i raw_spinlock_t lock; struct cpumask *percpu_enabled; const struct cpumask *percpu_affinity; -diff -Nur linux-4.9.28.orig/include/linux/irqflags.h linux-4.9.28/include/linux/irqflags.h ---- linux-4.9.28.orig/include/linux/irqflags.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/irqflags.h 2017-05-19 03:37:25.166176182 +0200 +diff --git a/include/linux/irqflags.h b/include/linux/irqflags.h +index 5dd1272d1ab2..9b77034f7c5e 100644 +--- a/include/linux/irqflags.h ++++ b/include/linux/irqflags.h @@ -25,8 +25,6 @@ # define trace_softirqs_enabled(p) ((p)->softirqs_enabled) # define trace_hardirq_enter() do { current->hardirq_context++; } while (0) @@ -7052,59 +7316,11 @@ diff -Nur linux-4.9.28.orig/include/linux/irqflags.h linux-4.9.28/include/linux/ +#endif + #endif -diff -Nur linux-4.9.28.orig/include/linux/irq.h linux-4.9.28/include/linux/irq.h ---- linux-4.9.28.orig/include/linux/irq.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/irq.h 2017-05-19 03:37:25.166176182 +0200 -@@ -72,6 +72,7 @@ - * IRQ_IS_POLLED - Always polled by another interrupt. Exclude - * it from the spurious interrupt detection - * mechanism and from core side polling. -+ * IRQ_NO_SOFTIRQ_CALL - No softirq processing in the irq thread context (RT) - * IRQ_DISABLE_UNLAZY - Disable lazy irq disable - */ - enum { -@@ -99,13 +100,14 @@ - IRQ_PER_CPU_DEVID = (1 << 17), - IRQ_IS_POLLED = (1 << 18), - IRQ_DISABLE_UNLAZY = (1 << 19), -+ IRQ_NO_SOFTIRQ_CALL = (1 << 20), - }; - - #define IRQF_MODIFY_MASK \ - (IRQ_TYPE_SENSE_MASK | IRQ_NOPROBE | IRQ_NOREQUEST | \ - IRQ_NOAUTOEN | IRQ_MOVE_PCNTXT | IRQ_LEVEL | IRQ_NO_BALANCING | \ - IRQ_PER_CPU | IRQ_NESTED_THREAD | IRQ_NOTHREAD | IRQ_PER_CPU_DEVID | \ -- IRQ_IS_POLLED | IRQ_DISABLE_UNLAZY) -+ IRQ_IS_POLLED | IRQ_DISABLE_UNLAZY | IRQ_NO_SOFTIRQ_CALL) - - #define IRQ_NO_BALANCING_MASK (IRQ_PER_CPU | IRQ_NO_BALANCING) - -diff -Nur linux-4.9.28.orig/include/linux/irq_work.h linux-4.9.28/include/linux/irq_work.h ---- linux-4.9.28.orig/include/linux/irq_work.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/irq_work.h 2017-05-19 03:37:25.166176182 +0200 -@@ -16,6 +16,7 @@ - #define IRQ_WORK_BUSY 2UL - #define IRQ_WORK_FLAGS 3UL - #define IRQ_WORK_LAZY 4UL /* Doesn't want IPI, wait for tick */ -+#define IRQ_WORK_HARD_IRQ 8UL /* Run hard IRQ context, even on RT */ - - struct irq_work { - unsigned long flags; -@@ -51,4 +52,10 @@ - static inline void irq_work_run(void) { } - #endif - -+#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) -+void irq_work_tick_soft(void); -+#else -+static inline void irq_work_tick_soft(void) { } -+#endif -+ - #endif /* _LINUX_IRQ_WORK_H */ -diff -Nur linux-4.9.28.orig/include/linux/jbd2.h linux-4.9.28/include/linux/jbd2.h ---- linux-4.9.28.orig/include/linux/jbd2.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/jbd2.h 2017-05-19 03:37:25.166176182 +0200 -@@ -347,32 +347,56 @@ +diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h +index dfaa1f4dcb0c..d57dd06544a1 100644 +--- a/include/linux/jbd2.h ++++ b/include/linux/jbd2.h +@@ -347,32 +347,56 @@ static inline struct journal_head *bh2jh(struct buffer_head *bh) static inline void jbd_lock_bh_state(struct buffer_head *bh) { @@ -7161,10 +7377,11 @@ diff -Nur linux-4.9.28.orig/include/linux/jbd2.h linux-4.9.28/include/linux/jbd2 } #define J_ASSERT(assert) BUG_ON(!(assert)) -diff -Nur linux-4.9.28.orig/include/linux/kdb.h linux-4.9.28/include/linux/kdb.h ---- linux-4.9.28.orig/include/linux/kdb.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/kdb.h 2017-05-19 03:37:25.166176182 +0200 -@@ -167,6 +167,7 @@ +diff --git a/include/linux/kdb.h b/include/linux/kdb.h +index 410decacff8f..0861bebfc188 100644 +--- a/include/linux/kdb.h ++++ b/include/linux/kdb.h +@@ -167,6 +167,7 @@ extern __printf(2, 0) int vkdb_printf(enum kdb_msgsrc src, const char *fmt, extern __printf(1, 2) int kdb_printf(const char *, ...); typedef __printf(1, 2) int (*kdb_printf_t)(const char *, ...); @@ -7172,7 +7389,7 @@ diff -Nur linux-4.9.28.orig/include/linux/kdb.h linux-4.9.28/include/linux/kdb.h extern void kdb_init(int level); /* Access to kdb specific polling devices */ -@@ -201,6 +202,7 @@ +@@ -201,6 +202,7 @@ extern int kdb_register_flags(char *, kdb_func_t, char *, char *, extern int kdb_unregister(char *); #else /* ! CONFIG_KGDB_KDB */ static inline __printf(1, 2) int kdb_printf(const char *fmt, ...) { return 0; } @@ -7180,10 +7397,11 @@ diff -Nur linux-4.9.28.orig/include/linux/kdb.h linux-4.9.28/include/linux/kdb.h static inline void kdb_init(int level) {} static inline int kdb_register(char *cmd, kdb_func_t func, char *usage, char *help, short minlen) { return 0; } -diff -Nur linux-4.9.28.orig/include/linux/kernel.h linux-4.9.28/include/linux/kernel.h ---- linux-4.9.28.orig/include/linux/kernel.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/kernel.h 2017-05-19 03:37:25.166176182 +0200 -@@ -194,6 +194,9 @@ +diff --git a/include/linux/kernel.h b/include/linux/kernel.h +index bc6ed52a39b9..7894d55e4998 100644 +--- a/include/linux/kernel.h ++++ b/include/linux/kernel.h +@@ -194,6 +194,9 @@ extern int _cond_resched(void); */ # define might_sleep() \ do { __might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0) @@ -7193,7 +7411,7 @@ diff -Nur linux-4.9.28.orig/include/linux/kernel.h linux-4.9.28/include/linux/ke # define sched_annotate_sleep() (current->task_state_change = 0) #else static inline void ___might_sleep(const char *file, int line, -@@ -201,6 +204,7 @@ +@@ -201,6 +204,7 @@ extern int _cond_resched(void); static inline void __might_sleep(const char *file, int line, int preempt_offset) { } # define might_sleep() do { might_resched(); } while (0) @@ -7201,7 +7419,7 @@ diff -Nur linux-4.9.28.orig/include/linux/kernel.h linux-4.9.28/include/linux/ke # define sched_annotate_sleep() do { } while (0) #endif -@@ -488,6 +492,7 @@ +@@ -488,6 +492,7 @@ extern enum system_states { SYSTEM_HALT, SYSTEM_POWER_OFF, SYSTEM_RESTART, @@ -7209,9 +7427,10 @@ diff -Nur linux-4.9.28.orig/include/linux/kernel.h linux-4.9.28/include/linux/ke } system_state; #define TAINT_PROPRIETARY_MODULE 0 -diff -Nur linux-4.9.28.orig/include/linux/list_bl.h linux-4.9.28/include/linux/list_bl.h ---- linux-4.9.28.orig/include/linux/list_bl.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/list_bl.h 2017-05-19 03:37:25.166176182 +0200 +diff --git a/include/linux/list_bl.h b/include/linux/list_bl.h +index cb483305e1f5..4e5062316bb6 100644 +--- a/include/linux/list_bl.h ++++ b/include/linux/list_bl.h @@ -2,6 +2,7 @@ #define _LINUX_LIST_BL_H @@ -7247,7 +7466,7 @@ diff -Nur linux-4.9.28.orig/include/linux/list_bl.h linux-4.9.28/include/linux/l static inline void INIT_HLIST_BL_NODE(struct hlist_bl_node *h) { -@@ -118,12 +130,26 @@ +@@ -118,12 +130,26 @@ static inline void hlist_bl_del_init(struct hlist_bl_node *n) static inline void hlist_bl_lock(struct hlist_bl_head *b) { @@ -7274,9 +7493,11 @@ diff -Nur linux-4.9.28.orig/include/linux/list_bl.h linux-4.9.28/include/linux/l } static inline bool hlist_bl_is_locked(struct hlist_bl_head *b) -diff -Nur linux-4.9.28.orig/include/linux/locallock.h linux-4.9.28/include/linux/locallock.h ---- linux-4.9.28.orig/include/linux/locallock.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/linux/locallock.h 2017-05-19 03:37:25.166176182 +0200 +diff --git a/include/linux/locallock.h b/include/linux/locallock.h +new file mode 100644 +index 000000000000..845c77f1a5ca +--- /dev/null ++++ b/include/linux/locallock.h @@ -0,0 +1,278 @@ +#ifndef _LINUX_LOCALLOCK_H +#define _LINUX_LOCALLOCK_H @@ -7556,9 +7777,10 @@ diff -Nur linux-4.9.28.orig/include/linux/locallock.h linux-4.9.28/include/linux +#endif + +#endif -diff -Nur linux-4.9.28.orig/include/linux/mm_types.h linux-4.9.28/include/linux/mm_types.h ---- linux-4.9.28.orig/include/linux/mm_types.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/mm_types.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h +index 08d947fc4c59..705fb564a605 100644 +--- a/include/linux/mm_types.h ++++ b/include/linux/mm_types.h @@ -11,6 +11,7 @@ #include #include @@ -7567,7 +7789,7 @@ diff -Nur linux-4.9.28.orig/include/linux/mm_types.h linux-4.9.28/include/linux/ #include #include #include -@@ -509,6 +510,9 @@ +@@ -509,6 +510,9 @@ struct mm_struct { bool tlb_flush_pending; #endif struct uprobes_state uprobes_state; @@ -7577,10 +7799,11 @@ diff -Nur linux-4.9.28.orig/include/linux/mm_types.h linux-4.9.28/include/linux/ #ifdef CONFIG_X86_INTEL_MPX /* address of the bounds directory */ void __user *bd_addr; -diff -Nur linux-4.9.28.orig/include/linux/module.h linux-4.9.28/include/linux/module.h ---- linux-4.9.28.orig/include/linux/module.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/module.h 2017-05-19 03:37:25.170176339 +0200 -@@ -496,6 +496,7 @@ +diff --git a/include/linux/module.h b/include/linux/module.h +index 0c3207d26ac0..5944baaa3f28 100644 +--- a/include/linux/module.h ++++ b/include/linux/module.h +@@ -496,6 +496,7 @@ static inline int module_is_live(struct module *mod) struct module *__module_text_address(unsigned long addr); struct module *__module_address(unsigned long addr); bool is_module_address(unsigned long addr); @@ -7588,21 +7811,22 @@ diff -Nur linux-4.9.28.orig/include/linux/module.h linux-4.9.28/include/linux/mo bool is_module_percpu_address(unsigned long addr); bool is_module_text_address(unsigned long addr); -@@ -662,6 +663,11 @@ - { +@@ -663,6 +664,11 @@ static inline bool is_module_percpu_address(unsigned long addr) return false; } -+ + +static inline bool __is_module_percpu_address(unsigned long addr, unsigned long *can_addr) +{ + return false; +} - ++ static inline bool is_module_text_address(unsigned long addr) { -diff -Nur linux-4.9.28.orig/include/linux/mutex.h linux-4.9.28/include/linux/mutex.h ---- linux-4.9.28.orig/include/linux/mutex.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/mutex.h 2017-05-19 03:37:25.170176339 +0200 + return false; +diff --git a/include/linux/mutex.h b/include/linux/mutex.h +index 2cb7531e7d7a..b3fdfc820216 100644 +--- a/include/linux/mutex.h ++++ b/include/linux/mutex.h @@ -19,6 +19,17 @@ #include #include @@ -7621,7 +7845,7 @@ diff -Nur linux-4.9.28.orig/include/linux/mutex.h linux-4.9.28/include/linux/mut /* * Simple, straightforward mutexes with strict semantics: * -@@ -99,13 +110,6 @@ +@@ -99,13 +110,6 @@ do { \ static inline void mutex_destroy(struct mutex *lock) {} #endif @@ -7635,7 +7859,7 @@ diff -Nur linux-4.9.28.orig/include/linux/mutex.h linux-4.9.28/include/linux/mut #define __MUTEX_INITIALIZER(lockname) \ { .count = ATOMIC_INIT(1) \ , .wait_lock = __SPIN_LOCK_UNLOCKED(lockname.wait_lock) \ -@@ -173,6 +177,8 @@ +@@ -173,6 +177,8 @@ extern int __must_check mutex_lock_killable(struct mutex *lock); extern int mutex_trylock(struct mutex *lock); extern void mutex_unlock(struct mutex *lock); @@ -7644,9 +7868,11 @@ diff -Nur linux-4.9.28.orig/include/linux/mutex.h linux-4.9.28/include/linux/mut extern int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock); #endif /* __LINUX_MUTEX_H */ -diff -Nur linux-4.9.28.orig/include/linux/mutex_rt.h linux-4.9.28/include/linux/mutex_rt.h ---- linux-4.9.28.orig/include/linux/mutex_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/linux/mutex_rt.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/mutex_rt.h b/include/linux/mutex_rt.h +new file mode 100644 +index 000000000000..e0284edec655 +--- /dev/null ++++ b/include/linux/mutex_rt.h @@ -0,0 +1,89 @@ +#ifndef __LINUX_MUTEX_RT_H +#define __LINUX_MUTEX_RT_H @@ -7737,10 +7963,11 @@ diff -Nur linux-4.9.28.orig/include/linux/mutex_rt.h linux-4.9.28/include/linux/ +} while (0) + +#endif -diff -Nur linux-4.9.28.orig/include/linux/netdevice.h linux-4.9.28/include/linux/netdevice.h ---- linux-4.9.28.orig/include/linux/netdevice.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/netdevice.h 2017-05-19 03:37:25.170176339 +0200 -@@ -396,7 +396,19 @@ +diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h +index bb9b102c15cd..a5b12b8ad196 100644 +--- a/include/linux/netdevice.h ++++ b/include/linux/netdevice.h +@@ -396,7 +396,19 @@ typedef enum rx_handler_result rx_handler_result_t; typedef rx_handler_result_t rx_handler_func_t(struct sk_buff **pskb); void __napi_schedule(struct napi_struct *n); @@ -7760,7 +7987,7 @@ diff -Nur linux-4.9.28.orig/include/linux/netdevice.h linux-4.9.28/include/linux static inline bool napi_disable_pending(struct napi_struct *n) { -@@ -2463,14 +2475,53 @@ +@@ -2463,14 +2475,53 @@ void netdev_freemem(struct net_device *dev); void synchronize_net(void); int init_dummy_netdev(struct net_device *dev); @@ -7815,7 +8042,7 @@ diff -Nur linux-4.9.28.orig/include/linux/netdevice.h linux-4.9.28/include/linux struct net_device *dev_get_by_index(struct net *net, int ifindex); struct net_device *__dev_get_by_index(struct net *net, int ifindex); struct net_device *dev_get_by_index_rcu(struct net *net, int ifindex); -@@ -2855,6 +2906,7 @@ +@@ -2855,6 +2906,7 @@ struct softnet_data { unsigned int dropped; struct sk_buff_head input_pkt_queue; struct napi_struct backlog; @@ -7823,9 +8050,10 @@ diff -Nur linux-4.9.28.orig/include/linux/netdevice.h linux-4.9.28/include/linux }; -diff -Nur linux-4.9.28.orig/include/linux/netfilter/x_tables.h linux-4.9.28/include/linux/netfilter/x_tables.h ---- linux-4.9.28.orig/include/linux/netfilter/x_tables.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/netfilter/x_tables.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h +index 2ad1a2b289b5..b4d10155af54 100644 +--- a/include/linux/netfilter/x_tables.h ++++ b/include/linux/netfilter/x_tables.h @@ -4,6 +4,7 @@ #include @@ -7834,7 +8062,7 @@ diff -Nur linux-4.9.28.orig/include/linux/netfilter/x_tables.h linux-4.9.28/incl #include /* Test a struct->invflags and a boolean for inequality */ -@@ -300,6 +301,8 @@ +@@ -300,6 +301,8 @@ void xt_free_table_info(struct xt_table_info *info); */ DECLARE_PER_CPU(seqcount_t, xt_recseq); @@ -7843,7 +8071,7 @@ diff -Nur linux-4.9.28.orig/include/linux/netfilter/x_tables.h linux-4.9.28/incl /* xt_tee_enabled - true if x_tables needs to handle reentrancy * * Enabled if current ip(6)tables ruleset has at least one -j TEE rule. -@@ -320,6 +323,9 @@ +@@ -320,6 +323,9 @@ static inline unsigned int xt_write_recseq_begin(void) { unsigned int addend; @@ -7853,7 +8081,7 @@ diff -Nur linux-4.9.28.orig/include/linux/netfilter/x_tables.h linux-4.9.28/incl /* * Low order bit of sequence is set if we already * called xt_write_recseq_begin(). -@@ -350,6 +356,7 @@ +@@ -350,6 +356,7 @@ static inline void xt_write_recseq_end(unsigned int addend) /* this is kind of a write_seqcount_end(), but addend is 0 or 1 */ smp_wmb(); __this_cpu_add(xt_recseq.sequence, addend); @@ -7861,10 +8089,11 @@ diff -Nur linux-4.9.28.orig/include/linux/netfilter/x_tables.h linux-4.9.28/incl } /* -diff -Nur linux-4.9.28.orig/include/linux/nfs_fs.h linux-4.9.28/include/linux/nfs_fs.h ---- linux-4.9.28.orig/include/linux/nfs_fs.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/nfs_fs.h 2017-05-19 03:37:25.170176339 +0200 -@@ -165,7 +165,11 @@ +diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h +index 810124b33327..d54ca43d571f 100644 +--- a/include/linux/nfs_fs.h ++++ b/include/linux/nfs_fs.h +@@ -165,7 +165,11 @@ struct nfs_inode { /* Readers: in-flight sillydelete RPC calls */ /* Writers: rmdir */ @@ -7876,10 +8105,11 @@ diff -Nur linux-4.9.28.orig/include/linux/nfs_fs.h linux-4.9.28/include/linux/nf #if IS_ENABLED(CONFIG_NFS_V4) struct nfs4_cached_acl *nfs4_acl; -diff -Nur linux-4.9.28.orig/include/linux/nfs_xdr.h linux-4.9.28/include/linux/nfs_xdr.h ---- linux-4.9.28.orig/include/linux/nfs_xdr.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/nfs_xdr.h 2017-05-19 03:37:25.170176339 +0200 -@@ -1490,7 +1490,7 @@ +diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h +index beb1e10f446e..ebaf2e7bfe29 100644 +--- a/include/linux/nfs_xdr.h ++++ b/include/linux/nfs_xdr.h +@@ -1490,7 +1490,7 @@ struct nfs_unlinkdata { struct nfs_removeargs args; struct nfs_removeres res; struct dentry *dentry; @@ -7888,9 +8118,10 @@ diff -Nur linux-4.9.28.orig/include/linux/nfs_xdr.h linux-4.9.28/include/linux/n struct rpc_cred *cred; struct nfs_fattr dir_attr; long timeout; -diff -Nur linux-4.9.28.orig/include/linux/notifier.h linux-4.9.28/include/linux/notifier.h ---- linux-4.9.28.orig/include/linux/notifier.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/notifier.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/notifier.h b/include/linux/notifier.h +index 4149868de4e6..babe5b9bcb91 100644 +--- a/include/linux/notifier.h ++++ b/include/linux/notifier.h @@ -6,7 +6,7 @@ * * Alan Cox @@ -7911,7 +8142,7 @@ diff -Nur linux-4.9.28.orig/include/linux/notifier.h linux-4.9.28/include/linux/ */ struct notifier_block; -@@ -90,7 +88,7 @@ +@@ -90,7 +88,7 @@ struct srcu_notifier_head { (name)->head = NULL; \ } while (0) @@ -7920,7 +8151,7 @@ diff -Nur linux-4.9.28.orig/include/linux/notifier.h linux-4.9.28/include/linux/ extern void srcu_init_notifier_head(struct srcu_notifier_head *nh); #define srcu_cleanup_notifier_head(name) \ cleanup_srcu_struct(&(name)->srcu); -@@ -103,7 +101,13 @@ +@@ -103,7 +101,13 @@ extern void srcu_init_notifier_head(struct srcu_notifier_head *nh); .head = NULL } #define RAW_NOTIFIER_INIT(name) { \ .head = NULL } @@ -7935,7 +8166,7 @@ diff -Nur linux-4.9.28.orig/include/linux/notifier.h linux-4.9.28/include/linux/ #define ATOMIC_NOTIFIER_HEAD(name) \ struct atomic_notifier_head name = \ -@@ -115,6 +119,18 @@ +@@ -115,6 +119,18 @@ extern void srcu_init_notifier_head(struct srcu_notifier_head *nh); struct raw_notifier_head name = \ RAW_NOTIFIER_INIT(name) @@ -7954,7 +8185,7 @@ diff -Nur linux-4.9.28.orig/include/linux/notifier.h linux-4.9.28/include/linux/ #ifdef __KERNEL__ extern int atomic_notifier_chain_register(struct atomic_notifier_head *nh, -@@ -184,12 +200,12 @@ +@@ -184,12 +200,12 @@ static inline int notifier_to_errno(int ret) /* * Declared notifiers so far. I can imagine quite a few more chains @@ -7970,56 +8201,10 @@ diff -Nur linux-4.9.28.orig/include/linux/notifier.h linux-4.9.28/include/linux/ /* CPU notfiers are defined in include/linux/cpu.h. */ /* netdevice notifiers are defined in include/linux/netdevice.h */ -diff -Nur linux-4.9.28.orig/include/linux/percpu.h linux-4.9.28/include/linux/percpu.h ---- linux-4.9.28.orig/include/linux/percpu.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/percpu.h 2017-05-19 03:37:25.170176339 +0200 -@@ -18,6 +18,35 @@ - #define PERCPU_MODULE_RESERVE 0 - #endif - -+#ifdef CONFIG_PREEMPT_RT_FULL -+ -+#define get_local_var(var) (*({ \ -+ migrate_disable(); \ -+ this_cpu_ptr(&var); })) -+ -+#define put_local_var(var) do { \ -+ (void)&(var); \ -+ migrate_enable(); \ -+} while (0) -+ -+# define get_local_ptr(var) ({ \ -+ migrate_disable(); \ -+ this_cpu_ptr(var); }) -+ -+# define put_local_ptr(var) do { \ -+ (void)(var); \ -+ migrate_enable(); \ -+} while (0) -+ -+#else -+ -+#define get_local_var(var) get_cpu_var(var) -+#define put_local_var(var) put_cpu_var(var) -+#define get_local_ptr(var) get_cpu_ptr(var) -+#define put_local_ptr(var) put_cpu_ptr(var) -+ -+#endif -+ - /* minimum unit size, also is the maximum supported allocation size */ - #define PCPU_MIN_UNIT_SIZE PFN_ALIGN(32 << 10) - -@@ -110,6 +139,7 @@ - #endif - - extern void __percpu *__alloc_reserved_percpu(size_t size, size_t align); -+extern bool __is_kernel_percpu_address(unsigned long addr, unsigned long *can_addr); - extern bool is_kernel_percpu_address(unsigned long addr); - - #if !defined(CONFIG_SMP) || !defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) -diff -Nur linux-4.9.28.orig/include/linux/percpu-rwsem.h linux-4.9.28/include/linux/percpu-rwsem.h ---- linux-4.9.28.orig/include/linux/percpu-rwsem.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/percpu-rwsem.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/percpu-rwsem.h b/include/linux/percpu-rwsem.h +index 5b2e6159b744..ea940f451606 100644 +--- a/include/linux/percpu-rwsem.h ++++ b/include/linux/percpu-rwsem.h @@ -4,7 +4,7 @@ #include #include @@ -8029,7 +8214,7 @@ diff -Nur linux-4.9.28.orig/include/linux/percpu-rwsem.h linux-4.9.28/include/li #include #include -@@ -12,7 +12,7 @@ +@@ -12,7 +12,7 @@ struct percpu_rw_semaphore { struct rcu_sync rss; unsigned int __percpu *read_count; struct rw_semaphore rw_sem; @@ -8038,7 +8223,7 @@ diff -Nur linux-4.9.28.orig/include/linux/percpu-rwsem.h linux-4.9.28/include/li int readers_block; }; -@@ -22,13 +22,13 @@ +@@ -22,13 +22,13 @@ static struct percpu_rw_semaphore name = { \ .rss = __RCU_SYNC_INITIALIZER(name.rss, RCU_SCHED_SYNC), \ .read_count = &__percpu_rwsem_rc_##name, \ .rw_sem = __RWSEM_INITIALIZER(name.rw_sem), \ @@ -8054,7 +8239,7 @@ diff -Nur linux-4.9.28.orig/include/linux/percpu-rwsem.h linux-4.9.28/include/li { might_sleep(); -@@ -46,16 +46,10 @@ +@@ -46,16 +46,10 @@ static inline void percpu_down_read_preempt_disable(struct percpu_rw_semaphore * __this_cpu_inc(*sem->read_count); if (unlikely(!rcu_sync_is_idle(&sem->rss))) __percpu_down_read(sem, false); /* Unconditional memory barrier */ @@ -8072,7 +8257,7 @@ diff -Nur linux-4.9.28.orig/include/linux/percpu-rwsem.h linux-4.9.28/include/li preempt_enable(); } -@@ -82,13 +76,9 @@ +@@ -82,13 +76,9 @@ static inline int percpu_down_read_trylock(struct percpu_rw_semaphore *sem) return ret; } @@ -8088,7 +8273,7 @@ diff -Nur linux-4.9.28.orig/include/linux/percpu-rwsem.h linux-4.9.28/include/li /* * Same as in percpu_down_read(). */ -@@ -101,12 +91,6 @@ +@@ -101,12 +91,6 @@ static inline void percpu_up_read_preempt_enable(struct percpu_rw_semaphore *sem rwsem_release(&sem->rw_sem.dep_map, 1, _RET_IP_); } @@ -8101,9 +8286,58 @@ diff -Nur linux-4.9.28.orig/include/linux/percpu-rwsem.h linux-4.9.28/include/li extern void percpu_down_write(struct percpu_rw_semaphore *); extern void percpu_up_write(struct percpu_rw_semaphore *); -diff -Nur linux-4.9.28.orig/include/linux/pid.h linux-4.9.28/include/linux/pid.h ---- linux-4.9.28.orig/include/linux/pid.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/pid.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/percpu.h b/include/linux/percpu.h +index 56939d3f6e53..b988bf40ad3e 100644 +--- a/include/linux/percpu.h ++++ b/include/linux/percpu.h +@@ -18,6 +18,35 @@ + #define PERCPU_MODULE_RESERVE 0 + #endif + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ ++#define get_local_var(var) (*({ \ ++ migrate_disable(); \ ++ this_cpu_ptr(&var); })) ++ ++#define put_local_var(var) do { \ ++ (void)&(var); \ ++ migrate_enable(); \ ++} while (0) ++ ++# define get_local_ptr(var) ({ \ ++ migrate_disable(); \ ++ this_cpu_ptr(var); }) ++ ++# define put_local_ptr(var) do { \ ++ (void)(var); \ ++ migrate_enable(); \ ++} while (0) ++ ++#else ++ ++#define get_local_var(var) get_cpu_var(var) ++#define put_local_var(var) put_cpu_var(var) ++#define get_local_ptr(var) get_cpu_ptr(var) ++#define put_local_ptr(var) put_cpu_ptr(var) ++ ++#endif ++ + /* minimum unit size, also is the maximum supported allocation size */ + #define PCPU_MIN_UNIT_SIZE PFN_ALIGN(32 << 10) + +@@ -110,6 +139,7 @@ extern int __init pcpu_page_first_chunk(size_t reserved_size, + #endif + + extern void __percpu *__alloc_reserved_percpu(size_t size, size_t align); ++extern bool __is_kernel_percpu_address(unsigned long addr, unsigned long *can_addr); + extern bool is_kernel_percpu_address(unsigned long addr); + + #if !defined(CONFIG_SMP) || !defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) +diff --git a/include/linux/pid.h b/include/linux/pid.h +index 23705a53abba..2cc64b779f03 100644 +--- a/include/linux/pid.h ++++ b/include/linux/pid.h @@ -2,6 +2,7 @@ #define _LINUX_PID_H @@ -8112,9 +8346,10 @@ diff -Nur linux-4.9.28.orig/include/linux/pid.h linux-4.9.28/include/linux/pid.h enum pid_type { -diff -Nur linux-4.9.28.orig/include/linux/preempt.h linux-4.9.28/include/linux/preempt.h ---- linux-4.9.28.orig/include/linux/preempt.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/preempt.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/preempt.h b/include/linux/preempt.h +index 75e4e30677f1..1cfb1cb72354 100644 +--- a/include/linux/preempt.h ++++ b/include/linux/preempt.h @@ -50,7 +50,11 @@ #define HARDIRQ_OFFSET (1UL << HARDIRQ_SHIFT) #define NMI_OFFSET (1UL << NMI_SHIFT) @@ -8165,7 +8400,7 @@ diff -Nur linux-4.9.28.orig/include/linux/preempt.h linux-4.9.28/include/linux/p /* * The preempt_count offset needed for things like: -@@ -140,6 +153,20 @@ +@@ -140,6 +153,20 @@ extern void preempt_count_sub(int val); #define preempt_count_inc() preempt_count_add(1) #define preempt_count_dec() preempt_count_sub(1) @@ -8186,7 +8421,7 @@ diff -Nur linux-4.9.28.orig/include/linux/preempt.h linux-4.9.28/include/linux/p #ifdef CONFIG_PREEMPT_COUNT #define preempt_disable() \ -@@ -148,13 +175,25 @@ +@@ -148,13 +175,25 @@ do { \ barrier(); \ } while (0) @@ -8213,7 +8448,7 @@ diff -Nur linux-4.9.28.orig/include/linux/preempt.h linux-4.9.28/include/linux/p #define preemptible() (preempt_count() == 0 && !irqs_disabled()) -@@ -179,6 +218,13 @@ +@@ -179,6 +218,13 @@ do { \ __preempt_schedule(); \ } while (0) @@ -8227,7 +8462,7 @@ diff -Nur linux-4.9.28.orig/include/linux/preempt.h linux-4.9.28/include/linux/p #else /* !CONFIG_PREEMPT */ #define preempt_enable() \ do { \ -@@ -224,6 +270,7 @@ +@@ -224,6 +270,7 @@ do { \ #define preempt_disable_notrace() barrier() #define preempt_enable_no_resched_notrace() barrier() #define preempt_enable_notrace() barrier() @@ -8235,7 +8470,7 @@ diff -Nur linux-4.9.28.orig/include/linux/preempt.h linux-4.9.28/include/linux/p #define preemptible() 0 #endif /* CONFIG_PREEMPT_COUNT */ -@@ -244,10 +291,31 @@ +@@ -244,10 +291,31 @@ do { \ } while (0) #define preempt_fold_need_resched() \ do { \ @@ -8268,10 +8503,11 @@ diff -Nur linux-4.9.28.orig/include/linux/preempt.h linux-4.9.28/include/linux/p #ifdef CONFIG_PREEMPT_NOTIFIERS struct preempt_notifier; -diff -Nur linux-4.9.28.orig/include/linux/printk.h linux-4.9.28/include/linux/printk.h ---- linux-4.9.28.orig/include/linux/printk.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/printk.h 2017-05-19 03:37:25.170176339 +0200 -@@ -126,9 +126,11 @@ +diff --git a/include/linux/printk.h b/include/linux/printk.h +index eac1af8502bb..37e647af0b0b 100644 +--- a/include/linux/printk.h ++++ b/include/linux/printk.h +@@ -126,9 +126,11 @@ struct va_format { #ifdef CONFIG_EARLY_PRINTK extern asmlinkage __printf(1, 2) void early_printk(const char *fmt, ...); @@ -8283,10 +8519,11 @@ diff -Nur linux-4.9.28.orig/include/linux/printk.h linux-4.9.28/include/linux/pr #endif #ifdef CONFIG_PRINTK_NMI -diff -Nur linux-4.9.28.orig/include/linux/radix-tree.h linux-4.9.28/include/linux/radix-tree.h ---- linux-4.9.28.orig/include/linux/radix-tree.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/radix-tree.h 2017-05-19 03:37:25.170176339 +0200 -@@ -292,6 +292,8 @@ +diff --git a/include/linux/radix-tree.h b/include/linux/radix-tree.h +index af3581b8a451..277295039c8f 100644 +--- a/include/linux/radix-tree.h ++++ b/include/linux/radix-tree.h +@@ -292,6 +292,8 @@ unsigned int radix_tree_gang_lookup_slot(struct radix_tree_root *root, int radix_tree_preload(gfp_t gfp_mask); int radix_tree_maybe_preload(gfp_t gfp_mask); int radix_tree_maybe_preload_order(gfp_t gfp_mask, int order); @@ -8295,7 +8532,7 @@ diff -Nur linux-4.9.28.orig/include/linux/radix-tree.h linux-4.9.28/include/linu void radix_tree_init(void); void *radix_tree_tag_set(struct radix_tree_root *root, unsigned long index, unsigned int tag); -@@ -314,11 +316,6 @@ +@@ -314,11 +316,6 @@ unsigned long radix_tree_range_tag_if_tagged(struct radix_tree_root *root, int radix_tree_tagged(struct radix_tree_root *root, unsigned int tag); unsigned long radix_tree_locate_item(struct radix_tree_root *root, void *item); @@ -8307,10 +8544,11 @@ diff -Nur linux-4.9.28.orig/include/linux/radix-tree.h linux-4.9.28/include/linu /** * struct radix_tree_iter - radix tree iterator state * -diff -Nur linux-4.9.28.orig/include/linux/random.h linux-4.9.28/include/linux/random.h ---- linux-4.9.28.orig/include/linux/random.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/random.h 2017-05-19 03:37:25.170176339 +0200 -@@ -31,7 +31,7 @@ +diff --git a/include/linux/random.h b/include/linux/random.h +index 16ab429735a7..9d0fecb5b6c2 100644 +--- a/include/linux/random.h ++++ b/include/linux/random.h +@@ -31,7 +31,7 @@ static inline void add_latent_entropy(void) {} extern void add_input_randomness(unsigned int type, unsigned int code, unsigned int value) __latent_entropy; @@ -8319,20 +8557,10 @@ diff -Nur linux-4.9.28.orig/include/linux/random.h linux-4.9.28/include/linux/ra extern void get_random_bytes(void *buf, int nbytes); extern int add_random_ready_callback(struct random_ready_callback *rdy); -diff -Nur linux-4.9.28.orig/include/linux/rbtree_augmented.h linux-4.9.28/include/linux/rbtree_augmented.h ---- linux-4.9.28.orig/include/linux/rbtree_augmented.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/rbtree_augmented.h 2017-05-19 03:37:25.170176339 +0200 -@@ -26,6 +26,7 @@ - - #include - #include -+#include - - /* - * Please note - only struct rb_augment_callbacks and the prototypes for -diff -Nur linux-4.9.28.orig/include/linux/rbtree.h linux-4.9.28/include/linux/rbtree.h ---- linux-4.9.28.orig/include/linux/rbtree.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/rbtree.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/rbtree.h b/include/linux/rbtree.h +index e585018498d5..25c64474fc27 100644 +--- a/include/linux/rbtree.h ++++ b/include/linux/rbtree.h @@ -31,7 +31,7 @@ #include @@ -8342,9 +8570,23 @@ diff -Nur linux-4.9.28.orig/include/linux/rbtree.h linux-4.9.28/include/linux/rb struct rb_node { unsigned long __rb_parent_color; -diff -Nur linux-4.9.28.orig/include/linux/rcu_assign_pointer.h linux-4.9.28/include/linux/rcu_assign_pointer.h ---- linux-4.9.28.orig/include/linux/rcu_assign_pointer.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/linux/rcu_assign_pointer.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/rbtree_augmented.h b/include/linux/rbtree_augmented.h +index d076183e49be..36bfb4dd57ae 100644 +--- a/include/linux/rbtree_augmented.h ++++ b/include/linux/rbtree_augmented.h +@@ -26,6 +26,7 @@ + + #include + #include ++#include + + /* + * Please note - only struct rb_augment_callbacks and the prototypes for +diff --git a/include/linux/rcu_assign_pointer.h b/include/linux/rcu_assign_pointer.h +new file mode 100644 +index 000000000000..7066962a4379 +--- /dev/null ++++ b/include/linux/rcu_assign_pointer.h @@ -0,0 +1,54 @@ +#ifndef __LINUX_RCU_ASSIGN_POINTER_H__ +#define __LINUX_RCU_ASSIGN_POINTER_H__ @@ -8400,9 +8642,10 @@ diff -Nur linux-4.9.28.orig/include/linux/rcu_assign_pointer.h linux-4.9.28/incl +}) + +#endif -diff -Nur linux-4.9.28.orig/include/linux/rcupdate.h linux-4.9.28/include/linux/rcupdate.h ---- linux-4.9.28.orig/include/linux/rcupdate.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/rcupdate.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h +index 01f71e1d2e94..30cc001d0d5a 100644 +--- a/include/linux/rcupdate.h ++++ b/include/linux/rcupdate.h @@ -46,6 +46,7 @@ #include #include @@ -8411,7 +8654,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rcupdate.h linux-4.9.28/include/linux/ #include -@@ -178,6 +179,9 @@ +@@ -178,6 +179,9 @@ void call_rcu(struct rcu_head *head, #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ @@ -8421,7 +8664,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rcupdate.h linux-4.9.28/include/linux/ /** * call_rcu_bh() - Queue an RCU for invocation after a quicker grace period. * @head: structure to be used for queueing the RCU updates. -@@ -201,6 +205,7 @@ +@@ -201,6 +205,7 @@ void call_rcu(struct rcu_head *head, */ void call_rcu_bh(struct rcu_head *head, rcu_callback_t func); @@ -8429,7 +8672,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rcupdate.h linux-4.9.28/include/linux/ /** * call_rcu_sched() - Queue an RCU for invocation after sched grace period. -@@ -301,6 +306,11 @@ +@@ -301,6 +306,11 @@ void synchronize_rcu(void); * types of kernel builds, the rcu_read_lock() nesting depth is unknowable. */ #define rcu_preempt_depth() (current->rcu_read_lock_nesting) @@ -8441,7 +8684,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rcupdate.h linux-4.9.28/include/linux/ #else /* #ifdef CONFIG_PREEMPT_RCU */ -@@ -326,6 +336,8 @@ +@@ -326,6 +336,8 @@ static inline int rcu_preempt_depth(void) return 0; } @@ -8450,7 +8693,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rcupdate.h linux-4.9.28/include/linux/ #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ /* Internal to kernel */ -@@ -505,7 +517,14 @@ +@@ -505,7 +517,14 @@ extern struct lockdep_map rcu_callback_map; int debug_lockdep_rcu_enabled(void); int rcu_read_lock_held(void); @@ -8465,7 +8708,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rcupdate.h linux-4.9.28/include/linux/ /** * rcu_read_lock_sched_held() - might we be in RCU-sched read-side critical section? -@@ -626,54 +645,6 @@ +@@ -626,54 +645,6 @@ static inline void rcu_preempt_sleep_check(void) }) /** @@ -8520,7 +8763,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rcupdate.h linux-4.9.28/include/linux/ * rcu_access_pointer() - fetch RCU pointer with no dereferencing * @p: The pointer to read * -@@ -951,10 +922,14 @@ +@@ -951,10 +922,14 @@ static inline void rcu_read_unlock(void) static inline void rcu_read_lock_bh(void) { local_bh_disable(); @@ -8535,7 +8778,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rcupdate.h linux-4.9.28/include/linux/ } /* -@@ -964,10 +939,14 @@ +@@ -964,10 +939,14 @@ static inline void rcu_read_lock_bh(void) */ static inline void rcu_read_unlock_bh(void) { @@ -8550,10 +8793,11 @@ diff -Nur linux-4.9.28.orig/include/linux/rcupdate.h linux-4.9.28/include/linux/ local_bh_enable(); } -diff -Nur linux-4.9.28.orig/include/linux/rcutree.h linux-4.9.28/include/linux/rcutree.h ---- linux-4.9.28.orig/include/linux/rcutree.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/rcutree.h 2017-05-19 03:37:25.170176339 +0200 -@@ -44,7 +44,11 @@ +diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h +index 63a4e4cf40a5..08ab12df2863 100644 +--- a/include/linux/rcutree.h ++++ b/include/linux/rcutree.h +@@ -44,7 +44,11 @@ static inline void rcu_virt_note_context_switch(int cpu) rcu_note_context_switch(); } @@ -8565,7 +8809,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rcutree.h linux-4.9.28/include/linux/r void synchronize_sched_expedited(void); void synchronize_rcu_expedited(void); -@@ -72,7 +76,11 @@ +@@ -72,7 +76,11 @@ static inline void synchronize_rcu_bh_expedited(void) } void rcu_barrier(void); @@ -8577,7 +8821,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rcutree.h linux-4.9.28/include/linux/r void rcu_barrier_sched(void); unsigned long get_state_synchronize_rcu(void); void cond_synchronize_rcu(unsigned long oldstate); -@@ -82,17 +90,14 @@ +@@ -82,17 +90,14 @@ void cond_synchronize_sched(unsigned long oldstate); extern unsigned long rcutorture_testseq; extern unsigned long rcutorture_vernum; unsigned long rcu_batches_started(void); @@ -8595,7 +8839,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rcutree.h linux-4.9.28/include/linux/r void rcu_sched_force_quiescent_state(void); void rcu_idle_enter(void); -@@ -109,6 +114,16 @@ +@@ -109,6 +114,16 @@ extern int rcu_scheduler_active __read_mostly; bool rcu_is_watching(void); @@ -8612,9 +8856,10 @@ diff -Nur linux-4.9.28.orig/include/linux/rcutree.h linux-4.9.28/include/linux/r void rcu_all_qs(void); /* RCUtree hotplug events */ -diff -Nur linux-4.9.28.orig/include/linux/rtmutex.h linux-4.9.28/include/linux/rtmutex.h ---- linux-4.9.28.orig/include/linux/rtmutex.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/rtmutex.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/rtmutex.h b/include/linux/rtmutex.h +index 1abba5ce2a2f..294a8b4875f1 100644 +--- a/include/linux/rtmutex.h ++++ b/include/linux/rtmutex.h @@ -13,11 +13,15 @@ #define __LINUX_RT_MUTEX_H @@ -8632,7 +8877,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rtmutex.h linux-4.9.28/include/linux/r /** * The rt_mutex structure * -@@ -31,8 +35,8 @@ +@@ -31,8 +35,8 @@ struct rt_mutex { struct rb_root waiters; struct rb_node *waiters_leftmost; struct task_struct *owner; @@ -8642,7 +8887,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rtmutex.h linux-4.9.28/include/linux/r const char *name, *file; int line; void *magic; -@@ -55,22 +59,33 @@ +@@ -55,22 +59,33 @@ struct hrtimer_sleeper; # define rt_mutex_debug_check_no_locks_held(task) do { } while (0) #endif @@ -8681,7 +8926,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rtmutex.h linux-4.9.28/include/linux/r #define DEFINE_RT_MUTEX(mutexname) \ struct rt_mutex mutexname = __RT_MUTEX_INITIALIZER(mutexname) -@@ -90,7 +105,9 @@ +@@ -90,7 +105,9 @@ extern void __rt_mutex_init(struct rt_mutex *lock, const char *name); extern void rt_mutex_destroy(struct rt_mutex *lock); extern void rt_mutex_lock(struct rt_mutex *lock); @@ -8691,9 +8936,11 @@ diff -Nur linux-4.9.28.orig/include/linux/rtmutex.h linux-4.9.28/include/linux/r extern int rt_mutex_timed_lock(struct rt_mutex *lock, struct hrtimer_sleeper *timeout); -diff -Nur linux-4.9.28.orig/include/linux/rwlock_rt.h linux-4.9.28/include/linux/rwlock_rt.h ---- linux-4.9.28.orig/include/linux/rwlock_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/linux/rwlock_rt.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/rwlock_rt.h b/include/linux/rwlock_rt.h +new file mode 100644 +index 000000000000..49ed2d45d3be +--- /dev/null ++++ b/include/linux/rwlock_rt.h @@ -0,0 +1,99 @@ +#ifndef __LINUX_RWLOCK_RT_H +#define __LINUX_RWLOCK_RT_H @@ -8794,9 +9041,10 @@ diff -Nur linux-4.9.28.orig/include/linux/rwlock_rt.h linux-4.9.28/include/linux + } while (0) + +#endif -diff -Nur linux-4.9.28.orig/include/linux/rwlock_types.h linux-4.9.28/include/linux/rwlock_types.h ---- linux-4.9.28.orig/include/linux/rwlock_types.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/rwlock_types.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/rwlock_types.h b/include/linux/rwlock_types.h +index cc0072e93e36..5317cd957292 100644 +--- a/include/linux/rwlock_types.h ++++ b/include/linux/rwlock_types.h @@ -1,6 +1,10 @@ #ifndef __LINUX_RWLOCK_TYPES_H #define __LINUX_RWLOCK_TYPES_H @@ -8808,9 +9056,11 @@ diff -Nur linux-4.9.28.orig/include/linux/rwlock_types.h linux-4.9.28/include/li /* * include/linux/rwlock_types.h - generic rwlock type definitions * and initializers -diff -Nur linux-4.9.28.orig/include/linux/rwlock_types_rt.h linux-4.9.28/include/linux/rwlock_types_rt.h ---- linux-4.9.28.orig/include/linux/rwlock_types_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/linux/rwlock_types_rt.h 2017-05-19 03:37:25.170176339 +0200 +diff --git a/include/linux/rwlock_types_rt.h b/include/linux/rwlock_types_rt.h +new file mode 100644 +index 000000000000..51b28d775fe1 +--- /dev/null ++++ b/include/linux/rwlock_types_rt.h @@ -0,0 +1,33 @@ +#ifndef __LINUX_RWLOCK_TYPES_RT_H +#define __LINUX_RWLOCK_TYPES_RT_H @@ -8845,9 +9095,10 @@ diff -Nur linux-4.9.28.orig/include/linux/rwlock_types_rt.h linux-4.9.28/include + rwlock_t name = __RW_LOCK_UNLOCKED(name) + +#endif -diff -Nur linux-4.9.28.orig/include/linux/rwsem.h linux-4.9.28/include/linux/rwsem.h ---- linux-4.9.28.orig/include/linux/rwsem.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/rwsem.h 2017-05-19 03:37:25.174176493 +0200 +diff --git a/include/linux/rwsem.h b/include/linux/rwsem.h +index dd1d14250340..aa2ac1f65c2d 100644 +--- a/include/linux/rwsem.h ++++ b/include/linux/rwsem.h @@ -19,6 +19,10 @@ #include #endif @@ -8859,7 +9110,7 @@ diff -Nur linux-4.9.28.orig/include/linux/rwsem.h linux-4.9.28/include/linux/rws struct rw_semaphore; #ifdef CONFIG_RWSEM_GENERIC_SPINLOCK -@@ -106,6 +110,13 @@ +@@ -106,6 +110,13 @@ static inline int rwsem_is_contended(struct rw_semaphore *sem) return !list_empty(&sem->wait_list); } @@ -8873,9 +9124,11 @@ diff -Nur linux-4.9.28.orig/include/linux/rwsem.h linux-4.9.28/include/linux/rws /* * lock for reading */ -diff -Nur linux-4.9.28.orig/include/linux/rwsem_rt.h linux-4.9.28/include/linux/rwsem_rt.h ---- linux-4.9.28.orig/include/linux/rwsem_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/linux/rwsem_rt.h 2017-05-19 03:37:25.174176493 +0200 +diff --git a/include/linux/rwsem_rt.h b/include/linux/rwsem_rt.h +new file mode 100644 +index 000000000000..2ffbf093ae92 +--- /dev/null ++++ b/include/linux/rwsem_rt.h @@ -0,0 +1,67 @@ +#ifndef _LINUX_RWSEM_RT_H +#define _LINUX_RWSEM_RT_H @@ -8921,72 +9174,34 @@ diff -Nur linux-4.9.28.orig/include/linux/rwsem_rt.h linux-4.9.28/include/linux/ +do { \ + static struct lock_class_key __key; \ + \ -+ __init_rwsem((sem), #sem, &__key); \ -+} while (0) -+ -+static inline int rwsem_is_locked(struct rw_semaphore *sem) -+{ -+ return atomic_read(&sem->readers) != READER_BIAS; -+} -+ -+static inline int rwsem_is_contended(struct rw_semaphore *sem) -+{ -+ return atomic_read(&sem->readers) > 0; -+} -+ -+extern void __down_read(struct rw_semaphore *sem); -+extern int __down_read_trylock(struct rw_semaphore *sem); -+extern void __down_write(struct rw_semaphore *sem); -+extern int __must_check __down_write_killable(struct rw_semaphore *sem); -+extern int __down_write_trylock(struct rw_semaphore *sem); -+extern void __up_read(struct rw_semaphore *sem); -+extern void __up_write(struct rw_semaphore *sem); -+extern void __downgrade_write(struct rw_semaphore *sem); -+ -+#endif -diff -Nur linux-4.9.28.orig/include/linux/sched/rt.h linux-4.9.28/include/linux/sched/rt.h ---- linux-4.9.28.orig/include/linux/sched/rt.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/sched/rt.h 2017-05-19 03:37:25.174176493 +0200 -@@ -16,27 +16,20 @@ - } - - #ifdef CONFIG_RT_MUTEXES --extern int rt_mutex_getprio(struct task_struct *p); --extern void rt_mutex_setprio(struct task_struct *p, int prio); --extern int rt_mutex_get_effective_prio(struct task_struct *task, int newprio); --extern struct task_struct *rt_mutex_get_top_task(struct task_struct *task); -+/* -+ * Must hold either p->pi_lock or task_rq(p)->lock. -+ */ -+static inline struct task_struct *rt_mutex_get_top_task(struct task_struct *p) -+{ -+ return p->pi_top_task; -+} -+extern void rt_mutex_setprio(struct task_struct *p, struct task_struct *pi_task); - extern void rt_mutex_adjust_pi(struct task_struct *p); - static inline bool tsk_is_pi_blocked(struct task_struct *tsk) - { - return tsk->pi_blocked_on != NULL; - } - #else --static inline int rt_mutex_getprio(struct task_struct *p) --{ -- return p->normal_prio; --} -- --static inline int rt_mutex_get_effective_prio(struct task_struct *task, -- int newprio) --{ -- return newprio; --} -- - static inline struct task_struct *rt_mutex_get_top_task(struct task_struct *task) - { - return NULL; -diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sched.h ---- linux-4.9.28.orig/include/linux/sched.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/sched.h 2017-05-19 03:37:25.174176493 +0200 -@@ -26,6 +26,7 @@ ++ __init_rwsem((sem), #sem, &__key); \ ++} while (0) ++ ++static inline int rwsem_is_locked(struct rw_semaphore *sem) ++{ ++ return atomic_read(&sem->readers) != READER_BIAS; ++} ++ ++static inline int rwsem_is_contended(struct rw_semaphore *sem) ++{ ++ return atomic_read(&sem->readers) > 0; ++} ++ ++extern void __down_read(struct rw_semaphore *sem); ++extern int __down_read_trylock(struct rw_semaphore *sem); ++extern void __down_write(struct rw_semaphore *sem); ++extern int __must_check __down_write_killable(struct rw_semaphore *sem); ++extern int __down_write_trylock(struct rw_semaphore *sem); ++extern void __up_read(struct rw_semaphore *sem); ++extern void __up_write(struct rw_semaphore *sem); ++extern void __downgrade_write(struct rw_semaphore *sem); ++ ++#endif +diff --git a/include/linux/sched.h b/include/linux/sched.h +index f425eb3318ab..4d779486ad6b 100644 +--- a/include/linux/sched.h ++++ b/include/linux/sched.h +@@ -26,6 +26,7 @@ struct sched_param { #include #include #include @@ -8994,7 +9209,14 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch #include #include -@@ -243,10 +244,7 @@ +@@ -236,17 +237,13 @@ extern char ___assert_task_state[1 - 2*!!( + + /* Convenience macros for the sake of wake_up */ + #define TASK_NORMAL (TASK_INTERRUPTIBLE | TASK_UNINTERRUPTIBLE) +-#define TASK_ALL (TASK_NORMAL | __TASK_STOPPED | __TASK_TRACED) + + /* get_task_state() */ + #define TASK_REPORT (TASK_RUNNING | TASK_INTERRUPTIBLE | \ TASK_UNINTERRUPTIBLE | __TASK_STOPPED | \ __TASK_TRACED | EXIT_ZOMBIE | EXIT_DEAD) @@ -9005,7 +9227,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch #define task_contributes_to_load(task) \ ((task->state & TASK_UNINTERRUPTIBLE) != 0 && \ (task->flags & PF_FROZEN) == 0 && \ -@@ -312,6 +310,11 @@ +@@ -312,6 +309,11 @@ extern char ___assert_task_state[1 - 2*!!( #endif @@ -9017,7 +9239,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch /* Task command name length */ #define TASK_COMM_LEN 16 -@@ -1013,8 +1016,18 @@ +@@ -1013,8 +1015,18 @@ struct wake_q_head { struct wake_q_head name = { WAKE_Q_TAIL, &name.first } extern void wake_q_add(struct wake_q_head *head, @@ -9038,7 +9260,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch /* * sched-domains (multiprocessor balancing) declarations: -@@ -1481,6 +1494,7 @@ +@@ -1481,6 +1493,7 @@ struct task_struct { struct thread_info thread_info; #endif volatile long state; /* -1 unrunnable, 0 runnable, >0 stopped */ @@ -9046,7 +9268,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch void *stack; atomic_t usage; unsigned int flags; /* per process flags, defined below */ -@@ -1520,6 +1534,12 @@ +@@ -1520,6 +1533,12 @@ struct task_struct { #endif unsigned int policy; @@ -9059,7 +9281,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch int nr_cpus_allowed; cpumask_t cpus_allowed; -@@ -1658,6 +1678,9 @@ +@@ -1658,6 +1677,9 @@ struct task_struct { struct task_cputime cputime_expires; struct list_head cpu_timers[3]; @@ -9069,7 +9291,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch /* process credentials */ const struct cred __rcu *ptracer_cred; /* Tracer's credentials at attach */ -@@ -1689,10 +1712,15 @@ +@@ -1689,10 +1711,15 @@ struct task_struct { /* signal handlers */ struct signal_struct *signal; struct sighand_struct *sighand; @@ -9085,7 +9307,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch unsigned long sas_ss_sp; size_t sas_ss_size; -@@ -1723,6 +1751,8 @@ +@@ -1723,6 +1750,8 @@ struct task_struct { /* PI waiters blocked on a rt_mutex held by this task */ struct rb_root pi_waiters; struct rb_node *pi_waiters_leftmost; @@ -9094,7 +9316,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch /* Deadlock detection and priority inheritance handling */ struct rt_mutex_waiter *pi_blocked_on; #endif -@@ -1921,6 +1951,12 @@ +@@ -1921,6 +1950,12 @@ struct task_struct { /* bitmask and counter of trace recursion */ unsigned long trace_recursion; #endif /* CONFIG_TRACING */ @@ -9107,7 +9329,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch #ifdef CONFIG_KCOV /* Coverage collection mode enabled for this task (0 if disabled). */ enum kcov_mode kcov_mode; -@@ -1946,9 +1982,23 @@ +@@ -1946,9 +1981,23 @@ struct task_struct { unsigned int sequential_io; unsigned int sequential_io_avg; #endif @@ -9131,7 +9353,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch int pagefault_disabled; #ifdef CONFIG_MMU struct task_struct *oom_reaper_list; -@@ -1988,14 +2038,6 @@ +@@ -1988,14 +2037,6 @@ static inline struct vm_struct *task_stack_vm_area(const struct task_struct *t) } #endif @@ -9146,7 +9368,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch #define TNF_MIGRATED 0x01 #define TNF_NO_GROUP 0x02 #define TNF_SHARED 0x04 -@@ -2211,6 +2253,15 @@ +@@ -2211,6 +2252,15 @@ extern struct pid *cad_pid; extern void free_task(struct task_struct *tsk); #define get_task_struct(tsk) do { atomic_inc(&(tsk)->usage); } while(0) @@ -9162,7 +9384,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch extern void __put_task_struct(struct task_struct *t); static inline void put_task_struct(struct task_struct *t) -@@ -2218,6 +2269,7 @@ +@@ -2218,6 +2268,7 @@ static inline void put_task_struct(struct task_struct *t) if (atomic_dec_and_test(&t->usage)) __put_task_struct(t); } @@ -9170,7 +9392,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch struct task_struct *task_rcu_dereference(struct task_struct **ptask); struct task_struct *try_get_task_struct(struct task_struct **ptask); -@@ -2259,6 +2311,7 @@ +@@ -2259,6 +2310,7 @@ extern void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, /* * Per process flags */ @@ -9178,7 +9400,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch #define PF_EXITING 0x00000004 /* getting shut down */ #define PF_EXITPIDONE 0x00000008 /* pi exit done on shut down */ #define PF_VCPU 0x00000010 /* I'm a virtual CPU */ -@@ -2427,6 +2480,10 @@ +@@ -2427,6 +2479,10 @@ extern void do_set_cpus_allowed(struct task_struct *p, extern int set_cpus_allowed_ptr(struct task_struct *p, const struct cpumask *new_mask); @@ -9189,7 +9411,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch #else static inline void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask) -@@ -2439,6 +2496,9 @@ +@@ -2439,6 +2495,9 @@ static inline int set_cpus_allowed_ptr(struct task_struct *p, return -EINVAL; return 0; } @@ -9199,7 +9421,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch #endif #ifdef CONFIG_NO_HZ_COMMON -@@ -2677,6 +2737,7 @@ +@@ -2677,6 +2736,7 @@ extern void xtime_update(unsigned long ticks); extern int wake_up_state(struct task_struct *tsk, unsigned int state); extern int wake_up_process(struct task_struct *tsk); @@ -9207,7 +9429,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch extern void wake_up_new_task(struct task_struct *tsk); #ifdef CONFIG_SMP extern void kick_process(struct task_struct *tsk); -@@ -2885,6 +2946,17 @@ +@@ -2885,6 +2945,17 @@ static inline void mmdrop(struct mm_struct *mm) __mmdrop(mm); } @@ -9225,7 +9447,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch static inline void mmdrop_async_fn(struct work_struct *work) { struct mm_struct *mm = container_of(work, struct mm_struct, async_put_work); -@@ -3277,6 +3349,43 @@ +@@ -3277,6 +3348,43 @@ static inline int test_tsk_need_resched(struct task_struct *tsk) return unlikely(test_tsk_thread_flag(tsk,TIF_NEED_RESCHED)); } @@ -9269,7 +9491,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch static inline int restart_syscall(void) { set_tsk_thread_flag(current, TIF_SIGPENDING); -@@ -3308,6 +3417,51 @@ +@@ -3308,6 +3416,51 @@ static inline int signal_pending_state(long state, struct task_struct *p) return (state & TASK_INTERRUPTIBLE) || __fatal_signal_pending(p); } @@ -9321,7 +9543,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch /* * cond_resched() and cond_resched_lock(): latency reduction via * explicit rescheduling in places that are safe. The return -@@ -3333,12 +3487,16 @@ +@@ -3333,12 +3486,16 @@ extern int __cond_resched_lock(spinlock_t *lock); __cond_resched_lock(lock); \ }) @@ -9338,7 +9560,7 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch static inline void cond_resched_rcu(void) { -@@ -3513,6 +3671,31 @@ +@@ -3513,6 +3670,31 @@ static inline void set_task_cpu(struct task_struct *p, unsigned int cpu) #endif /* CONFIG_SMP */ @@ -9370,10 +9592,51 @@ diff -Nur linux-4.9.28.orig/include/linux/sched.h linux-4.9.28/include/linux/sch extern long sched_setaffinity(pid_t pid, const struct cpumask *new_mask); extern long sched_getaffinity(pid_t pid, struct cpumask *mask); -diff -Nur linux-4.9.28.orig/include/linux/seqlock.h linux-4.9.28/include/linux/seqlock.h ---- linux-4.9.28.orig/include/linux/seqlock.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/seqlock.h 2017-05-19 03:37:25.174176493 +0200 -@@ -220,20 +220,30 @@ +diff --git a/include/linux/sched/rt.h b/include/linux/sched/rt.h +index a30b172df6e1..db3e91f2bc03 100644 +--- a/include/linux/sched/rt.h ++++ b/include/linux/sched/rt.h +@@ -16,27 +16,20 @@ static inline int rt_task(struct task_struct *p) + } + + #ifdef CONFIG_RT_MUTEXES +-extern int rt_mutex_getprio(struct task_struct *p); +-extern void rt_mutex_setprio(struct task_struct *p, int prio); +-extern int rt_mutex_get_effective_prio(struct task_struct *task, int newprio); +-extern struct task_struct *rt_mutex_get_top_task(struct task_struct *task); ++/* ++ * Must hold either p->pi_lock or task_rq(p)->lock. ++ */ ++static inline struct task_struct *rt_mutex_get_top_task(struct task_struct *p) ++{ ++ return p->pi_top_task; ++} ++extern void rt_mutex_setprio(struct task_struct *p, struct task_struct *pi_task); + extern void rt_mutex_adjust_pi(struct task_struct *p); + static inline bool tsk_is_pi_blocked(struct task_struct *tsk) + { + return tsk->pi_blocked_on != NULL; + } + #else +-static inline int rt_mutex_getprio(struct task_struct *p) +-{ +- return p->normal_prio; +-} +- +-static inline int rt_mutex_get_effective_prio(struct task_struct *task, +- int newprio) +-{ +- return newprio; +-} +- + static inline struct task_struct *rt_mutex_get_top_task(struct task_struct *task) + { + return NULL; +diff --git a/include/linux/seqlock.h b/include/linux/seqlock.h +index ead97654c4e9..3d7223ffdd3b 100644 +--- a/include/linux/seqlock.h ++++ b/include/linux/seqlock.h +@@ -220,20 +220,30 @@ static inline int read_seqcount_retry(const seqcount_t *s, unsigned start) return __read_seqcount_retry(s, start); } @@ -9408,7 +9671,7 @@ diff -Nur linux-4.9.28.orig/include/linux/seqlock.h linux-4.9.28/include/linux/s /** * raw_write_seqcount_barrier - do a seq write barrier * @s: pointer to seqcount_t -@@ -428,10 +438,32 @@ +@@ -428,10 +438,32 @@ typedef struct { /* * Read side functions for starting and finalizing a read side section. */ @@ -9441,7 +9704,7 @@ diff -Nur linux-4.9.28.orig/include/linux/seqlock.h linux-4.9.28/include/linux/s static inline unsigned read_seqretry(const seqlock_t *sl, unsigned start) { -@@ -446,36 +478,45 @@ +@@ -446,36 +478,45 @@ static inline unsigned read_seqretry(const seqlock_t *sl, unsigned start) static inline void write_seqlock(seqlock_t *sl) { spin_lock(&sl->lock); @@ -9493,7 +9756,7 @@ diff -Nur linux-4.9.28.orig/include/linux/seqlock.h linux-4.9.28/include/linux/s spin_unlock_irq(&sl->lock); } -@@ -484,7 +525,7 @@ +@@ -484,7 +525,7 @@ static inline unsigned long __write_seqlock_irqsave(seqlock_t *sl) unsigned long flags; spin_lock_irqsave(&sl->lock, flags); @@ -9502,7 +9765,7 @@ diff -Nur linux-4.9.28.orig/include/linux/seqlock.h linux-4.9.28/include/linux/s return flags; } -@@ -494,7 +535,7 @@ +@@ -494,7 +535,7 @@ static inline unsigned long __write_seqlock_irqsave(seqlock_t *sl) static inline void write_sequnlock_irqrestore(seqlock_t *sl, unsigned long flags) { @@ -9511,10 +9774,11 @@ diff -Nur linux-4.9.28.orig/include/linux/seqlock.h linux-4.9.28/include/linux/s spin_unlock_irqrestore(&sl->lock, flags); } -diff -Nur linux-4.9.28.orig/include/linux/signal.h linux-4.9.28/include/linux/signal.h ---- linux-4.9.28.orig/include/linux/signal.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/signal.h 2017-05-19 03:37:25.174176493 +0200 -@@ -233,6 +233,7 @@ +diff --git a/include/linux/signal.h b/include/linux/signal.h +index b63f63eaa39c..295540fdfc72 100644 +--- a/include/linux/signal.h ++++ b/include/linux/signal.h +@@ -233,6 +233,7 @@ static inline void init_sigpending(struct sigpending *sig) } extern void flush_sigqueue(struct sigpending *queue); @@ -9522,10 +9786,11 @@ diff -Nur linux-4.9.28.orig/include/linux/signal.h linux-4.9.28/include/linux/si /* Test if 'sig' is valid signal. Use this instead of testing _NSIG directly */ static inline int valid_signal(unsigned long sig) -diff -Nur linux-4.9.28.orig/include/linux/skbuff.h linux-4.9.28/include/linux/skbuff.h ---- linux-4.9.28.orig/include/linux/skbuff.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/skbuff.h 2017-05-19 03:37:25.174176493 +0200 -@@ -284,6 +284,7 @@ +diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h +index 32810f279f8e..0db6e31161f6 100644 +--- a/include/linux/skbuff.h ++++ b/include/linux/skbuff.h +@@ -284,6 +284,7 @@ struct sk_buff_head { __u32 qlen; spinlock_t lock; @@ -9533,7 +9798,7 @@ diff -Nur linux-4.9.28.orig/include/linux/skbuff.h linux-4.9.28/include/linux/sk }; struct sk_buff; -@@ -1573,6 +1574,12 @@ +@@ -1573,6 +1574,12 @@ static inline void skb_queue_head_init(struct sk_buff_head *list) __skb_queue_head_init(list); } @@ -9546,10 +9811,11 @@ diff -Nur linux-4.9.28.orig/include/linux/skbuff.h linux-4.9.28/include/linux/sk static inline void skb_queue_head_init_class(struct sk_buff_head *list, struct lock_class_key *class) { -diff -Nur linux-4.9.28.orig/include/linux/smp.h linux-4.9.28/include/linux/smp.h ---- linux-4.9.28.orig/include/linux/smp.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/smp.h 2017-05-19 03:37:25.174176493 +0200 -@@ -120,6 +120,13 @@ +diff --git a/include/linux/smp.h b/include/linux/smp.h +index 8e0cb7a0f836..891c533724f5 100644 +--- a/include/linux/smp.h ++++ b/include/linux/smp.h +@@ -120,6 +120,13 @@ extern unsigned int setup_max_cpus; extern void __init setup_nr_cpu_ids(void); extern void __init smp_init(void); @@ -9563,7 +9829,7 @@ diff -Nur linux-4.9.28.orig/include/linux/smp.h linux-4.9.28/include/linux/smp.h #else /* !SMP */ static inline void smp_send_stop(void) { } -@@ -158,6 +165,11 @@ +@@ -158,6 +165,11 @@ static inline void smp_init(void) { up_late_init(); } static inline void smp_init(void) { } #endif @@ -9575,7 +9841,7 @@ diff -Nur linux-4.9.28.orig/include/linux/smp.h linux-4.9.28/include/linux/smp.h #endif /* !SMP */ /* -@@ -185,6 +197,9 @@ +@@ -185,6 +197,9 @@ static inline void smp_init(void) { } #define get_cpu() ({ preempt_disable(); smp_processor_id(); }) #define put_cpu() preempt_enable() @@ -9585,23 +9851,11 @@ diff -Nur linux-4.9.28.orig/include/linux/smp.h linux-4.9.28/include/linux/smp.h /* * Callback to arch code if there's nosmp or maxcpus=0 on the * boot command line: -diff -Nur linux-4.9.28.orig/include/linux/spinlock_api_smp.h linux-4.9.28/include/linux/spinlock_api_smp.h ---- linux-4.9.28.orig/include/linux/spinlock_api_smp.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/spinlock_api_smp.h 2017-05-19 03:37:25.174176493 +0200 -@@ -189,6 +189,8 @@ - return 0; - } - --#include -+#ifndef CONFIG_PREEMPT_RT_FULL -+# include -+#endif - - #endif /* __LINUX_SPINLOCK_API_SMP_H */ -diff -Nur linux-4.9.28.orig/include/linux/spinlock.h linux-4.9.28/include/linux/spinlock.h ---- linux-4.9.28.orig/include/linux/spinlock.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/spinlock.h 2017-05-19 03:37:25.174176493 +0200 -@@ -271,7 +271,11 @@ +diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h +index 47dd0cebd204..b241cc044bd3 100644 +--- a/include/linux/spinlock.h ++++ b/include/linux/spinlock.h +@@ -271,7 +271,11 @@ static inline void do_raw_spin_unlock(raw_spinlock_t *lock) __releases(lock) #define raw_spin_can_lock(lock) (!raw_spin_is_locked(lock)) /* Include rwlock functions */ @@ -9614,7 +9868,7 @@ diff -Nur linux-4.9.28.orig/include/linux/spinlock.h linux-4.9.28/include/linux/ /* * Pull the _spin_*()/_read_*()/_write_*() functions/declarations: -@@ -282,6 +286,10 @@ +@@ -282,6 +286,10 @@ static inline void do_raw_spin_unlock(raw_spinlock_t *lock) __releases(lock) # include #endif @@ -9625,16 +9879,32 @@ diff -Nur linux-4.9.28.orig/include/linux/spinlock.h linux-4.9.28/include/linux/ /* * Map the spin_lock functions to the raw variants for PREEMPT_RT=n */ -@@ -416,4 +424,6 @@ +@@ -416,4 +424,6 @@ extern int _atomic_dec_and_lock(atomic_t *atomic, spinlock_t *lock); #define atomic_dec_and_lock(atomic, lock) \ __cond_lock(lock, _atomic_dec_and_lock(atomic, lock)) +#endif /* !PREEMPT_RT_FULL */ + #endif /* __LINUX_SPINLOCK_H */ -diff -Nur linux-4.9.28.orig/include/linux/spinlock_rt.h linux-4.9.28/include/linux/spinlock_rt.h ---- linux-4.9.28.orig/include/linux/spinlock_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/linux/spinlock_rt.h 2017-05-19 03:37:25.174176493 +0200 +diff --git a/include/linux/spinlock_api_smp.h b/include/linux/spinlock_api_smp.h +index 5344268e6e62..043263f30e81 100644 +--- a/include/linux/spinlock_api_smp.h ++++ b/include/linux/spinlock_api_smp.h +@@ -189,6 +189,8 @@ static inline int __raw_spin_trylock_bh(raw_spinlock_t *lock) + return 0; + } + +-#include ++#ifndef CONFIG_PREEMPT_RT_FULL ++# include ++#endif + + #endif /* __LINUX_SPINLOCK_API_SMP_H */ +diff --git a/include/linux/spinlock_rt.h b/include/linux/spinlock_rt.h +new file mode 100644 +index 000000000000..43ca841b913a +--- /dev/null ++++ b/include/linux/spinlock_rt.h @@ -0,0 +1,162 @@ +#ifndef __LINUX_SPINLOCK_RT_H +#define __LINUX_SPINLOCK_RT_H @@ -9798,19 +10068,28 @@ diff -Nur linux-4.9.28.orig/include/linux/spinlock_rt.h linux-4.9.28/include/lin + atomic_dec_and_spin_lock(atomic, lock) + +#endif -diff -Nur linux-4.9.28.orig/include/linux/spinlock_types.h linux-4.9.28/include/linux/spinlock_types.h ---- linux-4.9.28.orig/include/linux/spinlock_types.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/spinlock_types.h 2017-05-19 03:37:25.174176493 +0200 +diff --git a/include/linux/spinlock_types.h b/include/linux/spinlock_types.h +index 73548eb13a5d..10bac715ea96 100644 +--- a/include/linux/spinlock_types.h ++++ b/include/linux/spinlock_types.h @@ -9,80 +9,15 @@ * Released under the General Public License (GPL). */ -#if defined(CONFIG_SMP) -# include --#else ++#include ++ ++#ifndef CONFIG_PREEMPT_RT_FULL ++# include ++# include + #else -# include --#endif -- ++# include ++# include ++# include + #endif + -#include - -typedef struct raw_spinlock { @@ -9836,23 +10115,16 @@ diff -Nur linux-4.9.28.orig/include/linux/spinlock_types.h linux-4.9.28/include/ -#else -# define SPIN_DEP_MAP_INIT(lockname) -#endif -+#include - +- -#ifdef CONFIG_DEBUG_SPINLOCK -# define SPIN_DEBUG_INIT(lockname) \ - .magic = SPINLOCK_MAGIC, \ - .owner_cpu = -1, \ - .owner = SPINLOCK_OWNER_INIT, -+#ifndef CONFIG_PREEMPT_RT_FULL -+# include -+# include - #else +-#else -# define SPIN_DEBUG_INIT(lockname) -+# include -+# include -+# include - #endif - +-#endif +- -#define __RAW_SPIN_LOCK_INITIALIZER(lockname) \ - { \ - .raw_lock = __ARCH_SPIN_LOCK_UNLOCKED, \ @@ -9889,9 +10161,11 @@ diff -Nur linux-4.9.28.orig/include/linux/spinlock_types.h linux-4.9.28/include/ -#include - #endif /* __LINUX_SPINLOCK_TYPES_H */ -diff -Nur linux-4.9.28.orig/include/linux/spinlock_types_nort.h linux-4.9.28/include/linux/spinlock_types_nort.h ---- linux-4.9.28.orig/include/linux/spinlock_types_nort.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/linux/spinlock_types_nort.h 2017-05-19 03:37:25.174176493 +0200 +diff --git a/include/linux/spinlock_types_nort.h b/include/linux/spinlock_types_nort.h +new file mode 100644 +index 000000000000..f1dac1fb1d6a +--- /dev/null ++++ b/include/linux/spinlock_types_nort.h @@ -0,0 +1,33 @@ +#ifndef __LINUX_SPINLOCK_TYPES_NORT_H +#define __LINUX_SPINLOCK_TYPES_NORT_H @@ -9926,9 +10200,11 @@ diff -Nur linux-4.9.28.orig/include/linux/spinlock_types_nort.h linux-4.9.28/inc +#define DEFINE_SPINLOCK(x) spinlock_t x = __SPIN_LOCK_UNLOCKED(x) + +#endif -diff -Nur linux-4.9.28.orig/include/linux/spinlock_types_raw.h linux-4.9.28/include/linux/spinlock_types_raw.h ---- linux-4.9.28.orig/include/linux/spinlock_types_raw.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/linux/spinlock_types_raw.h 2017-05-19 03:37:25.174176493 +0200 +diff --git a/include/linux/spinlock_types_raw.h b/include/linux/spinlock_types_raw.h +new file mode 100644 +index 000000000000..edffc4d53fc9 +--- /dev/null ++++ b/include/linux/spinlock_types_raw.h @@ -0,0 +1,56 @@ +#ifndef __LINUX_SPINLOCK_TYPES_RAW_H +#define __LINUX_SPINLOCK_TYPES_RAW_H @@ -9986,9 +10262,11 @@ diff -Nur linux-4.9.28.orig/include/linux/spinlock_types_raw.h linux-4.9.28/incl +#define DEFINE_RAW_SPINLOCK(x) raw_spinlock_t x = __RAW_SPIN_LOCK_UNLOCKED(x) + +#endif -diff -Nur linux-4.9.28.orig/include/linux/spinlock_types_rt.h linux-4.9.28/include/linux/spinlock_types_rt.h ---- linux-4.9.28.orig/include/linux/spinlock_types_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/linux/spinlock_types_rt.h 2017-05-19 03:37:25.174176493 +0200 +diff --git a/include/linux/spinlock_types_rt.h b/include/linux/spinlock_types_rt.h +new file mode 100644 +index 000000000000..3e3d8c5f7a9a +--- /dev/null ++++ b/include/linux/spinlock_types_rt.h @@ -0,0 +1,48 @@ +#ifndef __LINUX_SPINLOCK_TYPES_RT_H +#define __LINUX_SPINLOCK_TYPES_RT_H @@ -10038,10 +10316,11 @@ diff -Nur linux-4.9.28.orig/include/linux/spinlock_types_rt.h linux-4.9.28/inclu + spinlock_t name = __SPIN_LOCK_UNLOCKED(name) + +#endif -diff -Nur linux-4.9.28.orig/include/linux/srcu.h linux-4.9.28/include/linux/srcu.h ---- linux-4.9.28.orig/include/linux/srcu.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/srcu.h 2017-05-19 03:37:25.174176493 +0200 -@@ -84,10 +84,10 @@ +diff --git a/include/linux/srcu.h b/include/linux/srcu.h +index dc8eb63c6568..e793d3a257da 100644 +--- a/include/linux/srcu.h ++++ b/include/linux/srcu.h +@@ -84,10 +84,10 @@ int init_srcu_struct(struct srcu_struct *sp); void process_srcu(struct work_struct *work); @@ -10054,7 +10333,7 @@ diff -Nur linux-4.9.28.orig/include/linux/srcu.h linux-4.9.28/include/linux/srcu .queue_lock = __SPIN_LOCK_UNLOCKED(name.queue_lock), \ .running = false, \ .batch_queue = RCU_BATCH_INIT(name.batch_queue), \ -@@ -119,7 +119,7 @@ +@@ -119,7 +119,7 @@ void process_srcu(struct work_struct *work); */ #define __DEFINE_SRCU(name, is_static) \ static DEFINE_PER_CPU(struct srcu_struct_array, name##_srcu_array);\ @@ -10063,10 +10342,11 @@ diff -Nur linux-4.9.28.orig/include/linux/srcu.h linux-4.9.28/include/linux/srcu #define DEFINE_SRCU(name) __DEFINE_SRCU(name, /* not static */) #define DEFINE_STATIC_SRCU(name) __DEFINE_SRCU(name, static) -diff -Nur linux-4.9.28.orig/include/linux/suspend.h linux-4.9.28/include/linux/suspend.h ---- linux-4.9.28.orig/include/linux/suspend.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/suspend.h 2017-05-19 03:37:25.174176493 +0200 -@@ -193,6 +193,12 @@ +diff --git a/include/linux/suspend.h b/include/linux/suspend.h +index d9718378a8be..e81e6dc7dcb1 100644 +--- a/include/linux/suspend.h ++++ b/include/linux/suspend.h +@@ -193,6 +193,12 @@ struct platform_freeze_ops { void (*end)(void); }; @@ -10079,10 +10359,11 @@ diff -Nur linux-4.9.28.orig/include/linux/suspend.h linux-4.9.28/include/linux/s #ifdef CONFIG_SUSPEND /** * suspend_set_ops - set platform dependent suspend operations -diff -Nur linux-4.9.28.orig/include/linux/swait.h linux-4.9.28/include/linux/swait.h ---- linux-4.9.28.orig/include/linux/swait.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/swait.h 2017-05-19 03:37:25.174176493 +0200 -@@ -87,6 +87,7 @@ +diff --git a/include/linux/swait.h b/include/linux/swait.h +index c1f9c62a8a50..83f004a72320 100644 +--- a/include/linux/swait.h ++++ b/include/linux/swait.h +@@ -87,6 +87,7 @@ static inline int swait_active(struct swait_queue_head *q) extern void swake_up(struct swait_queue_head *q); extern void swake_up_all(struct swait_queue_head *q); extern void swake_up_locked(struct swait_queue_head *q); @@ -10090,9 +10371,10 @@ diff -Nur linux-4.9.28.orig/include/linux/swait.h linux-4.9.28/include/linux/swa extern void __prepare_to_swait(struct swait_queue_head *q, struct swait_queue *wait); extern void prepare_to_swait(struct swait_queue_head *q, struct swait_queue *wait, int state); -diff -Nur linux-4.9.28.orig/include/linux/swap.h linux-4.9.28/include/linux/swap.h ---- linux-4.9.28.orig/include/linux/swap.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/swap.h 2017-05-19 03:37:25.174176493 +0200 +diff --git a/include/linux/swap.h b/include/linux/swap.h +index 55ff5593c193..52bf5477dc92 100644 +--- a/include/linux/swap.h ++++ b/include/linux/swap.h @@ -11,6 +11,7 @@ #include #include @@ -10101,7 +10383,7 @@ diff -Nur linux-4.9.28.orig/include/linux/swap.h linux-4.9.28/include/linux/swap #include struct notifier_block; -@@ -247,7 +248,8 @@ +@@ -247,7 +248,8 @@ struct swap_info_struct { void *workingset_eviction(struct address_space *mapping, struct page *page); bool workingset_refault(void *shadow); void workingset_activation(struct page *page); @@ -10111,7 +10393,7 @@ diff -Nur linux-4.9.28.orig/include/linux/swap.h linux-4.9.28/include/linux/swap static inline unsigned int workingset_node_pages(struct radix_tree_node *node) { -@@ -292,6 +294,7 @@ +@@ -292,6 +294,7 @@ extern unsigned long nr_free_pagecache_pages(void); /* linux/mm/swap.c */ @@ -10119,9 +10401,11 @@ diff -Nur linux-4.9.28.orig/include/linux/swap.h linux-4.9.28/include/linux/swap extern void lru_cache_add(struct page *); extern void lru_cache_add_anon(struct page *page); extern void lru_cache_add_file(struct page *page); -diff -Nur linux-4.9.28.orig/include/linux/swork.h linux-4.9.28/include/linux/swork.h ---- linux-4.9.28.orig/include/linux/swork.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/linux/swork.h 2017-05-19 03:37:25.174176493 +0200 +diff --git a/include/linux/swork.h b/include/linux/swork.h +new file mode 100644 +index 000000000000..f175fa9a6016 +--- /dev/null ++++ b/include/linux/swork.h @@ -0,0 +1,24 @@ +#ifndef _LINUX_SWORK_H +#define _LINUX_SWORK_H @@ -10147,10 +10431,11 @@ diff -Nur linux-4.9.28.orig/include/linux/swork.h linux-4.9.28/include/linux/swo +void swork_put(void); + +#endif /* _LINUX_SWORK_H */ -diff -Nur linux-4.9.28.orig/include/linux/thread_info.h linux-4.9.28/include/linux/thread_info.h ---- linux-4.9.28.orig/include/linux/thread_info.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/thread_info.h 2017-05-19 03:37:25.174176493 +0200 -@@ -107,7 +107,17 @@ +diff --git a/include/linux/thread_info.h b/include/linux/thread_info.h +index 2873baf5372a..eb1a108f17ca 100644 +--- a/include/linux/thread_info.h ++++ b/include/linux/thread_info.h +@@ -107,7 +107,17 @@ static inline int test_ti_thread_flag(struct thread_info *ti, int flag) #define test_thread_flag(flag) \ test_ti_thread_flag(current_thread_info(), flag) @@ -10169,10 +10454,11 @@ diff -Nur linux-4.9.28.orig/include/linux/thread_info.h linux-4.9.28/include/lin #ifndef CONFIG_HAVE_ARCH_WITHIN_STACK_FRAMES static inline int arch_within_stack_frames(const void * const stack, -diff -Nur linux-4.9.28.orig/include/linux/timer.h linux-4.9.28/include/linux/timer.h ---- linux-4.9.28.orig/include/linux/timer.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/timer.h 2017-05-19 03:37:25.174176493 +0200 -@@ -241,7 +241,7 @@ +diff --git a/include/linux/timer.h b/include/linux/timer.h +index 51d601f192d4..83cea629efe1 100644 +--- a/include/linux/timer.h ++++ b/include/linux/timer.h +@@ -241,7 +241,7 @@ extern void add_timer(struct timer_list *timer); extern int try_to_del_timer_sync(struct timer_list *timer); @@ -10181,10 +10467,11 @@ diff -Nur linux-4.9.28.orig/include/linux/timer.h linux-4.9.28/include/linux/tim extern int del_timer_sync(struct timer_list *timer); #else # define del_timer_sync(t) del_timer(t) -diff -Nur linux-4.9.28.orig/include/linux/trace_events.h linux-4.9.28/include/linux/trace_events.h ---- linux-4.9.28.orig/include/linux/trace_events.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/trace_events.h 2017-05-19 03:37:25.174176493 +0200 -@@ -56,6 +56,9 @@ +diff --git a/include/linux/trace_events.h b/include/linux/trace_events.h +index be007610ceb0..15154b13a53b 100644 +--- a/include/linux/trace_events.h ++++ b/include/linux/trace_events.h +@@ -56,6 +56,9 @@ struct trace_entry { unsigned char flags; unsigned char preempt_count; int pid; @@ -10194,10 +10481,11 @@ diff -Nur linux-4.9.28.orig/include/linux/trace_events.h linux-4.9.28/include/li }; #define TRACE_EVENT_TYPE_MAX \ -diff -Nur linux-4.9.28.orig/include/linux/uaccess.h linux-4.9.28/include/linux/uaccess.h ---- linux-4.9.28.orig/include/linux/uaccess.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/uaccess.h 2017-05-19 03:37:25.178176648 +0200 -@@ -24,6 +24,7 @@ +diff --git a/include/linux/uaccess.h b/include/linux/uaccess.h +index f30c187ed785..83bf0f798426 100644 +--- a/include/linux/uaccess.h ++++ b/include/linux/uaccess.h +@@ -24,6 +24,7 @@ static __always_inline void pagefault_disabled_dec(void) */ static inline void pagefault_disable(void) { @@ -10205,7 +10493,7 @@ diff -Nur linux-4.9.28.orig/include/linux/uaccess.h linux-4.9.28/include/linux/u pagefault_disabled_inc(); /* * make sure to have issued the store before a pagefault -@@ -40,6 +41,7 @@ +@@ -40,6 +41,7 @@ static inline void pagefault_enable(void) */ barrier(); pagefault_disabled_dec(); @@ -10213,9 +10501,10 @@ diff -Nur linux-4.9.28.orig/include/linux/uaccess.h linux-4.9.28/include/linux/u } /* -diff -Nur linux-4.9.28.orig/include/linux/uprobes.h linux-4.9.28/include/linux/uprobes.h ---- linux-4.9.28.orig/include/linux/uprobes.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/uprobes.h 2017-05-19 03:37:25.178176648 +0200 +diff --git a/include/linux/uprobes.h b/include/linux/uprobes.h +index 4a29c75b146e..0a294e950df8 100644 +--- a/include/linux/uprobes.h ++++ b/include/linux/uprobes.h @@ -27,6 +27,7 @@ #include #include @@ -10224,10 +10513,11 @@ diff -Nur linux-4.9.28.orig/include/linux/uprobes.h linux-4.9.28/include/linux/u struct vm_area_struct; struct mm_struct; -diff -Nur linux-4.9.28.orig/include/linux/vmstat.h linux-4.9.28/include/linux/vmstat.h ---- linux-4.9.28.orig/include/linux/vmstat.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/vmstat.h 2017-05-19 03:37:25.178176648 +0200 -@@ -33,7 +33,9 @@ +diff --git a/include/linux/vmstat.h b/include/linux/vmstat.h +index 613771909b6e..e28c5a43229d 100644 +--- a/include/linux/vmstat.h ++++ b/include/linux/vmstat.h +@@ -33,7 +33,9 @@ DECLARE_PER_CPU(struct vm_event_state, vm_event_states); */ static inline void __count_vm_event(enum vm_event_item item) { @@ -10237,7 +10527,7 @@ diff -Nur linux-4.9.28.orig/include/linux/vmstat.h linux-4.9.28/include/linux/vm } static inline void count_vm_event(enum vm_event_item item) -@@ -43,7 +45,9 @@ +@@ -43,7 +45,9 @@ static inline void count_vm_event(enum vm_event_item item) static inline void __count_vm_events(enum vm_event_item item, long delta) { @@ -10247,9 +10537,10 @@ diff -Nur linux-4.9.28.orig/include/linux/vmstat.h linux-4.9.28/include/linux/vm } static inline void count_vm_events(enum vm_event_item item, long delta) -diff -Nur linux-4.9.28.orig/include/linux/wait.h linux-4.9.28/include/linux/wait.h ---- linux-4.9.28.orig/include/linux/wait.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/linux/wait.h 2017-05-19 03:37:25.178176648 +0200 +diff --git a/include/linux/wait.h b/include/linux/wait.h +index 2408e8d5c05c..db50d6609195 100644 +--- a/include/linux/wait.h ++++ b/include/linux/wait.h @@ -8,6 +8,7 @@ #include #include @@ -10258,10 +10549,11 @@ diff -Nur linux-4.9.28.orig/include/linux/wait.h linux-4.9.28/include/linux/wait typedef struct __wait_queue wait_queue_t; typedef int (*wait_queue_func_t)(wait_queue_t *wait, unsigned mode, int flags, void *key); -diff -Nur linux-4.9.28.orig/include/net/dst.h linux-4.9.28/include/net/dst.h ---- linux-4.9.28.orig/include/net/dst.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/net/dst.h 2017-05-19 03:37:25.178176648 +0200 -@@ -446,7 +446,7 @@ +diff --git a/include/net/dst.h b/include/net/dst.h +index 6835d224d47b..55a5a9698f14 100644 +--- a/include/net/dst.h ++++ b/include/net/dst.h +@@ -446,7 +446,7 @@ static inline void dst_confirm(struct dst_entry *dst) static inline int dst_neigh_output(struct dst_entry *dst, struct neighbour *n, struct sk_buff *skb) { @@ -10270,9 +10562,10 @@ diff -Nur linux-4.9.28.orig/include/net/dst.h linux-4.9.28/include/net/dst.h if (dst->pending_confirm) { unsigned long now = jiffies; -diff -Nur linux-4.9.28.orig/include/net/gen_stats.h linux-4.9.28/include/net/gen_stats.h ---- linux-4.9.28.orig/include/net/gen_stats.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/net/gen_stats.h 2017-05-19 03:37:25.178176648 +0200 +diff --git a/include/net/gen_stats.h b/include/net/gen_stats.h +index 231e121cc7d9..d125222b979d 100644 +--- a/include/net/gen_stats.h ++++ b/include/net/gen_stats.h @@ -5,6 +5,7 @@ #include #include @@ -10281,7 +10574,7 @@ diff -Nur linux-4.9.28.orig/include/net/gen_stats.h linux-4.9.28/include/net/gen struct gnet_stats_basic_cpu { struct gnet_stats_basic_packed bstats; -@@ -33,11 +34,11 @@ +@@ -33,11 +34,11 @@ int gnet_stats_start_copy_compat(struct sk_buff *skb, int type, spinlock_t *lock, struct gnet_dump *d, int padattr); @@ -10295,7 +10588,7 @@ diff -Nur linux-4.9.28.orig/include/net/gen_stats.h linux-4.9.28/include/net/gen struct gnet_stats_basic_packed *bstats, struct gnet_stats_basic_cpu __percpu *cpu, struct gnet_stats_basic_packed *b); -@@ -55,14 +56,14 @@ +@@ -55,14 +56,14 @@ int gen_new_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_basic_cpu __percpu *cpu_bstats, struct gnet_stats_rate_est64 *rate_est, spinlock_t *stats_lock, @@ -10312,10 +10605,11 @@ diff -Nur linux-4.9.28.orig/include/net/gen_stats.h linux-4.9.28/include/net/gen bool gen_estimator_active(const struct gnet_stats_basic_packed *bstats, const struct gnet_stats_rate_est64 *rate_est); #endif -diff -Nur linux-4.9.28.orig/include/net/neighbour.h linux-4.9.28/include/net/neighbour.h ---- linux-4.9.28.orig/include/net/neighbour.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/net/neighbour.h 2017-05-19 03:37:25.178176648 +0200 -@@ -446,7 +446,7 @@ +diff --git a/include/net/neighbour.h b/include/net/neighbour.h +index 8b683841e574..bf656008f6e7 100644 +--- a/include/net/neighbour.h ++++ b/include/net/neighbour.h +@@ -446,7 +446,7 @@ static inline int neigh_hh_bridge(struct hh_cache *hh, struct sk_buff *skb) } #endif @@ -10324,7 +10618,7 @@ diff -Nur linux-4.9.28.orig/include/net/neighbour.h linux-4.9.28/include/net/nei { unsigned int seq; int hh_len; -@@ -501,7 +501,7 @@ +@@ -501,7 +501,7 @@ struct neighbour_cb { #define NEIGH_CB(skb) ((struct neighbour_cb *)(skb)->cb) @@ -10333,20 +10627,11 @@ diff -Nur linux-4.9.28.orig/include/net/neighbour.h linux-4.9.28/include/net/nei const struct net_device *dev) { unsigned int seq; -diff -Nur linux-4.9.28.orig/include/net/netns/ipv4.h linux-4.9.28/include/net/netns/ipv4.h ---- linux-4.9.28.orig/include/net/netns/ipv4.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/net/netns/ipv4.h 2017-05-19 03:37:25.178176648 +0200 -@@ -69,6 +69,7 @@ - - int sysctl_icmp_echo_ignore_all; - int sysctl_icmp_echo_ignore_broadcasts; -+ int sysctl_icmp_echo_sysrq; - int sysctl_icmp_ignore_bogus_error_responses; - int sysctl_icmp_ratelimit; - int sysctl_icmp_ratemask; -diff -Nur linux-4.9.28.orig/include/net/net_seq_lock.h linux-4.9.28/include/net/net_seq_lock.h ---- linux-4.9.28.orig/include/net/net_seq_lock.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/net/net_seq_lock.h 2017-05-19 03:37:25.178176648 +0200 +diff --git a/include/net/net_seq_lock.h b/include/net/net_seq_lock.h +new file mode 100644 +index 000000000000..a7034298a82a +--- /dev/null ++++ b/include/net/net_seq_lock.h @@ -0,0 +1,15 @@ +#ifndef __NET_NET_SEQ_LOCK_H__ +#define __NET_NET_SEQ_LOCK_H__ @@ -10363,9 +10648,22 @@ diff -Nur linux-4.9.28.orig/include/net/net_seq_lock.h linux-4.9.28/include/net/ +#endif + +#endif -diff -Nur linux-4.9.28.orig/include/net/sch_generic.h linux-4.9.28/include/net/sch_generic.h ---- linux-4.9.28.orig/include/net/sch_generic.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/net/sch_generic.h 2017-05-19 03:37:25.178176648 +0200 +diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h +index 7adf4386ac8f..d3fd5c357268 100644 +--- a/include/net/netns/ipv4.h ++++ b/include/net/netns/ipv4.h +@@ -69,6 +69,7 @@ struct netns_ipv4 { + + int sysctl_icmp_echo_ignore_all; + int sysctl_icmp_echo_ignore_broadcasts; ++ int sysctl_icmp_echo_sysrq; + int sysctl_icmp_ignore_bogus_error_responses; + int sysctl_icmp_ratelimit; + int sysctl_icmp_ratemask; +diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h +index e6aa0a249672..b57736f2a8a3 100644 +--- a/include/net/sch_generic.h ++++ b/include/net/sch_generic.h @@ -10,6 +10,7 @@ #include #include @@ -10374,7 +10672,7 @@ diff -Nur linux-4.9.28.orig/include/net/sch_generic.h linux-4.9.28/include/net/s struct Qdisc_ops; struct qdisc_walker; -@@ -86,7 +87,7 @@ +@@ -86,7 +87,7 @@ struct Qdisc { struct sk_buff *gso_skb ____cacheline_aligned_in_smp; struct qdisc_skb_head q; struct gnet_stats_basic_packed bstats; @@ -10383,7 +10681,7 @@ diff -Nur linux-4.9.28.orig/include/net/sch_generic.h linux-4.9.28/include/net/s struct gnet_stats_queue qstats; unsigned long state; struct Qdisc *next_sched; -@@ -98,13 +99,22 @@ +@@ -98,13 +99,22 @@ struct Qdisc { spinlock_t busylock ____cacheline_aligned_in_smp; }; @@ -10407,7 +10705,7 @@ diff -Nur linux-4.9.28.orig/include/net/sch_generic.h linux-4.9.28/include/net/s if (qdisc_is_running(qdisc)) return false; /* Variant of write_seqcount_begin() telling lockdep a trylock -@@ -113,11 +123,16 @@ +@@ -113,11 +123,16 @@ static inline bool qdisc_run_begin(struct Qdisc *qdisc) raw_write_seqcount_begin(&qdisc->running); seqcount_acquire(&qdisc->running.dep_map, 0, 1, _RET_IP_); return true; @@ -10424,7 +10722,7 @@ diff -Nur linux-4.9.28.orig/include/net/sch_generic.h linux-4.9.28/include/net/s } static inline bool qdisc_may_bulk(const struct Qdisc *qdisc) -@@ -308,7 +323,7 @@ +@@ -308,7 +323,7 @@ static inline spinlock_t *qdisc_root_sleeping_lock(const struct Qdisc *qdisc) return qdisc_lock(root); } @@ -10433,9 +10731,11 @@ diff -Nur linux-4.9.28.orig/include/net/sch_generic.h linux-4.9.28/include/net/s { struct Qdisc *root = qdisc_root_sleeping(qdisc); -diff -Nur linux-4.9.28.orig/include/trace/events/hist.h linux-4.9.28/include/trace/events/hist.h ---- linux-4.9.28.orig/include/trace/events/hist.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/trace/events/hist.h 2017-05-19 03:37:25.178176648 +0200 +diff --git a/include/trace/events/hist.h b/include/trace/events/hist.h +new file mode 100644 +index 000000000000..f7710de1b1f3 +--- /dev/null ++++ b/include/trace/events/hist.h @@ -0,0 +1,73 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM hist @@ -10510,9 +10810,11 @@ diff -Nur linux-4.9.28.orig/include/trace/events/hist.h linux-4.9.28/include/tra + +/* This part must be outside protection */ +#include -diff -Nur linux-4.9.28.orig/include/trace/events/latency_hist.h linux-4.9.28/include/trace/events/latency_hist.h ---- linux-4.9.28.orig/include/trace/events/latency_hist.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/include/trace/events/latency_hist.h 2017-05-19 03:37:25.178176648 +0200 +diff --git a/include/trace/events/latency_hist.h b/include/trace/events/latency_hist.h +new file mode 100644 +index 000000000000..d3f2fbd560b1 +--- /dev/null ++++ b/include/trace/events/latency_hist.h @@ -0,0 +1,29 @@ +#ifndef _LATENCY_HIST_H +#define _LATENCY_HIST_H @@ -10543,10 +10845,11 @@ diff -Nur linux-4.9.28.orig/include/trace/events/latency_hist.h linux-4.9.28/inc +} + +#endif /* _LATENCY_HIST_H */ -diff -Nur linux-4.9.28.orig/include/trace/events/sched.h linux-4.9.28/include/trace/events/sched.h ---- linux-4.9.28.orig/include/trace/events/sched.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/include/trace/events/sched.h 2017-05-19 03:37:25.178176648 +0200 -@@ -70,7 +70,7 @@ +diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h +index 9b90c57517a9..516ae88cddf4 100644 +--- a/include/trace/events/sched.h ++++ b/include/trace/events/sched.h +@@ -70,7 +70,7 @@ DECLARE_EVENT_CLASS(sched_wakeup_template, TP_fast_assign( memcpy(__entry->comm, p->comm, TASK_COMM_LEN); __entry->pid = p->pid; @@ -10555,7 +10858,7 @@ diff -Nur linux-4.9.28.orig/include/trace/events/sched.h linux-4.9.28/include/tr __entry->success = 1; /* rudiment, kill when possible */ __entry->target_cpu = task_cpu(p); ), -@@ -147,6 +147,7 @@ +@@ -147,6 +147,7 @@ TRACE_EVENT(sched_switch, memcpy(__entry->prev_comm, prev->comm, TASK_COMM_LEN); __entry->next_pid = next->pid; __entry->next_prio = next->prio; @@ -10563,7 +10866,7 @@ diff -Nur linux-4.9.28.orig/include/trace/events/sched.h linux-4.9.28/include/tr ), TP_printk("prev_comm=%s prev_pid=%d prev_prio=%d prev_state=%s%s ==> next_comm=%s next_pid=%d next_prio=%d", -@@ -181,7 +182,7 @@ +@@ -181,7 +182,7 @@ TRACE_EVENT(sched_migrate_task, TP_fast_assign( memcpy(__entry->comm, p->comm, TASK_COMM_LEN); __entry->pid = p->pid; @@ -10572,7 +10875,7 @@ diff -Nur linux-4.9.28.orig/include/trace/events/sched.h linux-4.9.28/include/tr __entry->orig_cpu = task_cpu(p); __entry->dest_cpu = dest_cpu; ), -@@ -206,7 +207,7 @@ +@@ -206,7 +207,7 @@ DECLARE_EVENT_CLASS(sched_process_template, TP_fast_assign( memcpy(__entry->comm, p->comm, TASK_COMM_LEN); __entry->pid = p->pid; @@ -10581,7 +10884,7 @@ diff -Nur linux-4.9.28.orig/include/trace/events/sched.h linux-4.9.28/include/tr ), TP_printk("comm=%s pid=%d prio=%d", -@@ -253,7 +254,7 @@ +@@ -253,7 +254,7 @@ TRACE_EVENT(sched_process_wait, TP_fast_assign( memcpy(__entry->comm, current->comm, TASK_COMM_LEN); __entry->pid = pid_nr(pid); @@ -10590,7 +10893,7 @@ diff -Nur linux-4.9.28.orig/include/trace/events/sched.h linux-4.9.28/include/tr ), TP_printk("comm=%s pid=%d prio=%d", -@@ -413,9 +414,9 @@ +@@ -413,9 +414,9 @@ DEFINE_EVENT(sched_stat_runtime, sched_stat_runtime, */ TRACE_EVENT(sched_pi_setprio, @@ -10602,7 +10905,7 @@ diff -Nur linux-4.9.28.orig/include/trace/events/sched.h linux-4.9.28/include/tr TP_STRUCT__entry( __array( char, comm, TASK_COMM_LEN ) -@@ -428,7 +429,8 @@ +@@ -428,7 +429,8 @@ TRACE_EVENT(sched_pi_setprio, memcpy(__entry->comm, tsk->comm, TASK_COMM_LEN); __entry->pid = tsk->pid; __entry->oldprio = tsk->prio; @@ -10612,10 +10915,11 @@ diff -Nur linux-4.9.28.orig/include/trace/events/sched.h linux-4.9.28/include/tr ), TP_printk("comm=%s pid=%d oldprio=%d newprio=%d", -diff -Nur linux-4.9.28.orig/init/Kconfig linux-4.9.28/init/Kconfig ---- linux-4.9.28.orig/init/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/init/Kconfig 2017-05-19 03:37:25.178176648 +0200 -@@ -506,7 +506,7 @@ +diff --git a/init/Kconfig b/init/Kconfig +index 34407f15e6d3..2ce33a32e65d 100644 +--- a/init/Kconfig ++++ b/init/Kconfig +@@ -506,7 +506,7 @@ config TINY_RCU config RCU_EXPERT bool "Make expert-level adjustments to RCU configuration" @@ -10624,7 +10928,7 @@ diff -Nur linux-4.9.28.orig/init/Kconfig linux-4.9.28/init/Kconfig help This option needs to be enabled if you wish to make expert-level adjustments to RCU configuration. By default, -@@ -623,7 +623,7 @@ +@@ -623,7 +623,7 @@ config RCU_FANOUT_LEAF config RCU_FAST_NO_HZ bool "Accelerate last non-dyntick-idle CPU's grace periods" @@ -10633,7 +10937,7 @@ diff -Nur linux-4.9.28.orig/init/Kconfig linux-4.9.28/init/Kconfig default n help This option permits CPUs to enter dynticks-idle state even if -@@ -650,7 +650,7 @@ +@@ -650,7 +650,7 @@ config TREE_RCU_TRACE config RCU_BOOST bool "Enable RCU priority boosting" depends on RT_MUTEXES && PREEMPT_RCU && RCU_EXPERT @@ -10642,7 +10946,7 @@ diff -Nur linux-4.9.28.orig/init/Kconfig linux-4.9.28/init/Kconfig help This option boosts the priority of preempted RCU readers that block the current preemptible RCU grace period for too long. -@@ -781,19 +781,6 @@ +@@ -781,19 +781,6 @@ config RCU_NOCB_CPU_ALL endchoice @@ -10662,7 +10966,7 @@ diff -Nur linux-4.9.28.orig/init/Kconfig linux-4.9.28/init/Kconfig endmenu # "RCU Subsystem" config BUILD_BIN2C -@@ -1064,6 +1051,7 @@ +@@ -1064,6 +1051,7 @@ config CFS_BANDWIDTH config RT_GROUP_SCHED bool "Group scheduling for SCHED_RR/FIFO" depends on CGROUP_SCHED @@ -10670,7 +10974,7 @@ diff -Nur linux-4.9.28.orig/init/Kconfig linux-4.9.28/init/Kconfig default n help This feature lets you explicitly allocate real CPU bandwidth -@@ -1772,6 +1760,7 @@ +@@ -1772,6 +1760,7 @@ choice config SLAB bool "SLAB" @@ -10678,7 +10982,7 @@ diff -Nur linux-4.9.28.orig/init/Kconfig linux-4.9.28/init/Kconfig select HAVE_HARDENED_USERCOPY_ALLOCATOR help The regular slab allocator that is established and known to work -@@ -1792,6 +1781,7 @@ +@@ -1792,6 +1781,7 @@ config SLUB config SLOB depends on EXPERT bool "SLOB (Simple Allocator)" @@ -10686,7 +10990,7 @@ diff -Nur linux-4.9.28.orig/init/Kconfig linux-4.9.28/init/Kconfig help SLOB replaces the stock allocator with a drastically simpler allocator. SLOB is generally more space efficient but -@@ -1810,7 +1800,7 @@ +@@ -1810,7 +1800,7 @@ config SLAB_FREELIST_RANDOM config SLUB_CPU_PARTIAL default y @@ -10695,10 +10999,21 @@ diff -Nur linux-4.9.28.orig/init/Kconfig linux-4.9.28/init/Kconfig bool "SLUB per cpu partial cache" help Per cpu partial caches accellerate objects allocation and freeing -diff -Nur linux-4.9.28.orig/init/main.c linux-4.9.28/init/main.c ---- linux-4.9.28.orig/init/main.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/init/main.c 2017-05-19 03:37:25.178176648 +0200 -@@ -507,6 +507,7 @@ +diff --git a/init/Makefile b/init/Makefile +index c4fb45525d08..821190dfaa75 100644 +--- a/init/Makefile ++++ b/init/Makefile +@@ -35,4 +35,4 @@ $(obj)/version.o: include/generated/compile.h + include/generated/compile.h: FORCE + @$($(quiet)chk_compile.h) + $(Q)$(CONFIG_SHELL) $(srctree)/scripts/mkcompile_h $@ \ +- "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CC) $(KBUILD_CFLAGS)" ++ "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CONFIG_PREEMPT_RT_FULL)" "$(CC) $(KBUILD_CFLAGS)" +diff --git a/init/main.c b/init/main.c +index ae3996ae9bac..6470deef01c9 100644 +--- a/init/main.c ++++ b/init/main.c +@@ -507,6 +507,7 @@ asmlinkage __visible void __init start_kernel(void) setup_command_line(command_line); setup_nr_cpu_ids(); setup_per_cpu_areas(); @@ -10706,19 +11021,11 @@ diff -Nur linux-4.9.28.orig/init/main.c linux-4.9.28/init/main.c boot_cpu_state_init(); smp_prepare_boot_cpu(); /* arch-specific boot-cpu hooks */ -diff -Nur linux-4.9.28.orig/init/Makefile linux-4.9.28/init/Makefile ---- linux-4.9.28.orig/init/Makefile 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/init/Makefile 2017-05-19 03:37:25.178176648 +0200 -@@ -35,4 +35,4 @@ - include/generated/compile.h: FORCE - @$($(quiet)chk_compile.h) - $(Q)$(CONFIG_SHELL) $(srctree)/scripts/mkcompile_h $@ \ -- "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CC) $(KBUILD_CFLAGS)" -+ "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CONFIG_PREEMPT_RT_FULL)" "$(CC) $(KBUILD_CFLAGS)" -diff -Nur linux-4.9.28.orig/ipc/sem.c linux-4.9.28/ipc/sem.c ---- linux-4.9.28.orig/ipc/sem.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/ipc/sem.c 2017-05-19 03:37:25.178176648 +0200 -@@ -712,6 +712,13 @@ +diff --git a/ipc/sem.c b/ipc/sem.c +index 10b94bc59d4a..b8360eaacc7a 100644 +--- a/ipc/sem.c ++++ b/ipc/sem.c +@@ -712,6 +712,13 @@ static int perform_atomic_semop(struct sem_array *sma, struct sem_queue *q) static void wake_up_sem_queue_prepare(struct list_head *pt, struct sem_queue *q, int error) { @@ -10732,7 +11039,7 @@ diff -Nur linux-4.9.28.orig/ipc/sem.c linux-4.9.28/ipc/sem.c if (list_empty(pt)) { /* * Hold preempt off so that we don't get preempted and have the -@@ -723,6 +730,7 @@ +@@ -723,6 +730,7 @@ static void wake_up_sem_queue_prepare(struct list_head *pt, q->pid = error; list_add_tail(&q->list, pt); @@ -10740,7 +11047,7 @@ diff -Nur linux-4.9.28.orig/ipc/sem.c linux-4.9.28/ipc/sem.c } /** -@@ -736,6 +744,7 @@ +@@ -736,6 +744,7 @@ static void wake_up_sem_queue_prepare(struct list_head *pt, */ static void wake_up_sem_queue_do(struct list_head *pt) { @@ -10748,7 +11055,7 @@ diff -Nur linux-4.9.28.orig/ipc/sem.c linux-4.9.28/ipc/sem.c struct sem_queue *q, *t; int did_something; -@@ -748,6 +757,7 @@ +@@ -748,6 +757,7 @@ static void wake_up_sem_queue_do(struct list_head *pt) } if (did_something) preempt_enable(); @@ -10756,10 +11063,85 @@ diff -Nur linux-4.9.28.orig/ipc/sem.c linux-4.9.28/ipc/sem.c } static void unlink_queue(struct sem_array *sma, struct sem_queue *q) -diff -Nur linux-4.9.28.orig/kernel/cgroup.c linux-4.9.28/kernel/cgroup.c ---- linux-4.9.28.orig/kernel/cgroup.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/cgroup.c 2017-05-19 03:37:25.178176648 +0200 -@@ -5041,10 +5041,10 @@ +diff --git a/kernel/Kconfig.locks b/kernel/Kconfig.locks +index ebdb0043203a..b9e6aa7e5aa6 100644 +--- a/kernel/Kconfig.locks ++++ b/kernel/Kconfig.locks +@@ -225,11 +225,11 @@ config ARCH_SUPPORTS_ATOMIC_RMW + + config MUTEX_SPIN_ON_OWNER + def_bool y +- depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW ++ depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL + + config RWSEM_SPIN_ON_OWNER + def_bool y +- depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW ++ depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL + + config LOCK_SPIN_ON_OWNER + def_bool y +diff --git a/kernel/Kconfig.preempt b/kernel/Kconfig.preempt +index 3f9c97419f02..11dbe26a8279 100644 +--- a/kernel/Kconfig.preempt ++++ b/kernel/Kconfig.preempt +@@ -1,3 +1,16 @@ ++config PREEMPT ++ bool ++ select PREEMPT_COUNT ++ ++config PREEMPT_RT_BASE ++ bool ++ select PREEMPT ++ ++config HAVE_PREEMPT_LAZY ++ bool ++ ++config PREEMPT_LAZY ++ def_bool y if HAVE_PREEMPT_LAZY && PREEMPT_RT_FULL + + choice + prompt "Preemption Model" +@@ -33,9 +46,9 @@ config PREEMPT_VOLUNTARY + + Select this if you are building a kernel for a desktop system. + +-config PREEMPT ++config PREEMPT__LL + bool "Preemptible Kernel (Low-Latency Desktop)" +- select PREEMPT_COUNT ++ select PREEMPT + select UNINLINE_SPIN_UNLOCK if !ARCH_INLINE_SPIN_UNLOCK + help + This option reduces the latency of the kernel by making +@@ -52,6 +65,22 @@ config PREEMPT + embedded system with latency requirements in the milliseconds + range. + ++config PREEMPT_RTB ++ bool "Preemptible Kernel (Basic RT)" ++ select PREEMPT_RT_BASE ++ help ++ This option is basically the same as (Low-Latency Desktop) but ++ enables changes which are preliminary for the full preemptible ++ RT kernel. ++ ++config PREEMPT_RT_FULL ++ bool "Fully Preemptible Kernel (RT)" ++ depends on IRQ_FORCED_THREADING ++ select PREEMPT_RT_BASE ++ select PREEMPT_RCU ++ help ++ All and everything ++ + endchoice + + config PREEMPT_COUNT +diff --git a/kernel/cgroup.c b/kernel/cgroup.c +index a3d2aad2443f..bb6b252648ff 100644 +--- a/kernel/cgroup.c ++++ b/kernel/cgroup.c +@@ -5041,10 +5041,10 @@ static void css_free_rcu_fn(struct rcu_head *rcu_head) queue_work(cgroup_destroy_wq, &css->destroy_work); } @@ -10772,7 +11154,7 @@ diff -Nur linux-4.9.28.orig/kernel/cgroup.c linux-4.9.28/kernel/cgroup.c struct cgroup_subsys *ss = css->ss; struct cgroup *cgrp = css->cgroup; -@@ -5087,8 +5087,8 @@ +@@ -5087,8 +5087,8 @@ static void css_release(struct percpu_ref *ref) struct cgroup_subsys_state *css = container_of(ref, struct cgroup_subsys_state, refcnt); @@ -10783,7 +11165,7 @@ diff -Nur linux-4.9.28.orig/kernel/cgroup.c linux-4.9.28/kernel/cgroup.c } static void init_and_link_css(struct cgroup_subsys_state *css, -@@ -5740,6 +5740,7 @@ +@@ -5740,6 +5740,7 @@ static int __init cgroup_wq_init(void) */ cgroup_destroy_wq = alloc_workqueue("cgroup_destroy", 0, 1); BUG_ON(!cgroup_destroy_wq); @@ -10791,10 +11173,11 @@ diff -Nur linux-4.9.28.orig/kernel/cgroup.c linux-4.9.28/kernel/cgroup.c /* * Used to destroy pidlists and separate to serve as flush domain. -diff -Nur linux-4.9.28.orig/kernel/cpu.c linux-4.9.28/kernel/cpu.c ---- linux-4.9.28.orig/kernel/cpu.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/cpu.c 2017-05-19 03:37:25.182176801 +0200 -@@ -239,6 +239,289 @@ +diff --git a/kernel/cpu.c b/kernel/cpu.c +index 99c6c568bc55..f1c64e563970 100644 +--- a/kernel/cpu.c ++++ b/kernel/cpu.c +@@ -239,6 +239,289 @@ static struct { #define cpuhp_lock_acquire() lock_map_acquire(&cpu_hotplug.dep_map) #define cpuhp_lock_release() lock_map_release(&cpu_hotplug.dep_map) @@ -11084,7 +11467,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpu.c linux-4.9.28/kernel/cpu.c void get_online_cpus(void) { -@@ -789,10 +1072,14 @@ +@@ -789,10 +1072,14 @@ static int takedown_cpu(unsigned int cpu) struct cpuhp_cpu_state *st = per_cpu_ptr(&cpuhp_state, cpu); int err; @@ -11099,7 +11482,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpu.c linux-4.9.28/kernel/cpu.c /* * Prevent irq alloc/free while the dying cpu reorganizes the * interrupt affinities. -@@ -877,6 +1164,9 @@ +@@ -877,6 +1164,9 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen, struct cpuhp_cpu_state *st = per_cpu_ptr(&cpuhp_state, cpu); int prev_state, ret = 0; bool hasdied = false; @@ -11109,7 +11492,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpu.c linux-4.9.28/kernel/cpu.c if (num_online_cpus() == 1) return -EBUSY; -@@ -884,7 +1174,34 @@ +@@ -884,7 +1174,34 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen, if (!cpu_present(cpu)) return -EINVAL; @@ -11144,7 +11527,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpu.c linux-4.9.28/kernel/cpu.c cpuhp_tasks_frozen = tasks_frozen; -@@ -923,10 +1240,15 @@ +@@ -923,10 +1240,15 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen, hasdied = prev_state != st->state && st->state == CPUHP_OFFLINE; out: @@ -11160,7 +11543,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpu.c linux-4.9.28/kernel/cpu.c return ret; } -@@ -1240,6 +1562,8 @@ +@@ -1240,6 +1562,8 @@ core_initcall(cpu_hotplug_pm_sync_init); #endif /* CONFIG_PM_SLEEP_SMP */ @@ -11169,7 +11552,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpu.c linux-4.9.28/kernel/cpu.c #endif /* CONFIG_SMP */ /* Boot processor state steps */ -@@ -1924,6 +2248,10 @@ +@@ -1924,6 +2248,10 @@ void __init boot_cpu_init(void) set_cpu_active(cpu, true); set_cpu_present(cpu, true); set_cpu_possible(cpu, true); @@ -11180,10 +11563,11 @@ diff -Nur linux-4.9.28.orig/kernel/cpu.c linux-4.9.28/kernel/cpu.c } /* -diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c ---- linux-4.9.28.orig/kernel/cpuset.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/cpuset.c 2017-05-19 03:37:25.182176801 +0200 -@@ -284,7 +284,7 @@ +diff --git a/kernel/cpuset.c b/kernel/cpuset.c +index 29f815d2ef7e..341b17f24f95 100644 +--- a/kernel/cpuset.c ++++ b/kernel/cpuset.c +@@ -284,7 +284,7 @@ static struct cpuset top_cpuset = { */ static DEFINE_MUTEX(cpuset_mutex); @@ -11192,7 +11576,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c static struct workqueue_struct *cpuset_migrate_mm_wq; -@@ -907,9 +907,9 @@ +@@ -907,9 +907,9 @@ static void update_cpumasks_hier(struct cpuset *cs, struct cpumask *new_cpus) continue; rcu_read_unlock(); @@ -11204,7 +11588,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c WARN_ON(!cgroup_subsys_on_dfl(cpuset_cgrp_subsys) && !cpumask_equal(cp->cpus_allowed, cp->effective_cpus)); -@@ -974,9 +974,9 @@ +@@ -974,9 +974,9 @@ static int update_cpumask(struct cpuset *cs, struct cpuset *trialcs, if (retval < 0) return retval; @@ -11216,7 +11600,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c /* use trialcs->cpus_allowed as a temp variable */ update_cpumasks_hier(cs, trialcs->cpus_allowed); -@@ -1176,9 +1176,9 @@ +@@ -1176,9 +1176,9 @@ static void update_nodemasks_hier(struct cpuset *cs, nodemask_t *new_mems) continue; rcu_read_unlock(); @@ -11228,7 +11612,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c WARN_ON(!cgroup_subsys_on_dfl(cpuset_cgrp_subsys) && !nodes_equal(cp->mems_allowed, cp->effective_mems)); -@@ -1246,9 +1246,9 @@ +@@ -1246,9 +1246,9 @@ static int update_nodemask(struct cpuset *cs, struct cpuset *trialcs, if (retval < 0) goto done; @@ -11240,7 +11624,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c /* use trialcs->mems_allowed as a temp variable */ update_nodemasks_hier(cs, &trialcs->mems_allowed); -@@ -1339,9 +1339,9 @@ +@@ -1339,9 +1339,9 @@ static int update_flag(cpuset_flagbits_t bit, struct cpuset *cs, spread_flag_changed = ((is_spread_slab(cs) != is_spread_slab(trialcs)) || (is_spread_page(cs) != is_spread_page(trialcs))); @@ -11252,7 +11636,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c if (!cpumask_empty(trialcs->cpus_allowed) && balance_flag_changed) rebuild_sched_domains_locked(); -@@ -1756,7 +1756,7 @@ +@@ -1756,7 +1756,7 @@ static int cpuset_common_seq_show(struct seq_file *sf, void *v) cpuset_filetype_t type = seq_cft(sf)->private; int ret = 0; @@ -11261,7 +11645,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c switch (type) { case FILE_CPULIST: -@@ -1775,7 +1775,7 @@ +@@ -1775,7 +1775,7 @@ static int cpuset_common_seq_show(struct seq_file *sf, void *v) ret = -EINVAL; } @@ -11270,7 +11654,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c return ret; } -@@ -1989,12 +1989,12 @@ +@@ -1989,12 +1989,12 @@ static int cpuset_css_online(struct cgroup_subsys_state *css) cpuset_inc(); @@ -11285,7 +11669,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c if (!test_bit(CGRP_CPUSET_CLONE_CHILDREN, &css->cgroup->flags)) goto out_unlock; -@@ -2021,12 +2021,12 @@ +@@ -2021,12 +2021,12 @@ static int cpuset_css_online(struct cgroup_subsys_state *css) } rcu_read_unlock(); @@ -11300,7 +11684,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c out_unlock: mutex_unlock(&cpuset_mutex); return 0; -@@ -2065,7 +2065,7 @@ +@@ -2065,7 +2065,7 @@ static void cpuset_css_free(struct cgroup_subsys_state *css) static void cpuset_bind(struct cgroup_subsys_state *root_css) { mutex_lock(&cpuset_mutex); @@ -11309,7 +11693,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c if (cgroup_subsys_on_dfl(cpuset_cgrp_subsys)) { cpumask_copy(top_cpuset.cpus_allowed, cpu_possible_mask); -@@ -2076,7 +2076,7 @@ +@@ -2076,7 +2076,7 @@ static void cpuset_bind(struct cgroup_subsys_state *root_css) top_cpuset.mems_allowed = top_cpuset.effective_mems; } @@ -11318,7 +11702,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c mutex_unlock(&cpuset_mutex); } -@@ -2177,12 +2177,12 @@ +@@ -2177,12 +2177,12 @@ hotplug_update_tasks_legacy(struct cpuset *cs, { bool is_empty; @@ -11333,7 +11717,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c /* * Don't call update_tasks_cpumask() if the cpuset becomes empty, -@@ -2219,10 +2219,10 @@ +@@ -2219,10 +2219,10 @@ hotplug_update_tasks(struct cpuset *cs, if (nodes_empty(*new_mems)) *new_mems = parent_cs(cs)->effective_mems; @@ -11346,7 +11730,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c if (cpus_updated) update_tasks_cpumask(cs); -@@ -2308,21 +2308,21 @@ +@@ -2308,21 +2308,21 @@ static void cpuset_hotplug_workfn(struct work_struct *work) /* synchronize cpus_allowed to cpu_active_mask */ if (cpus_updated) { @@ -11372,7 +11756,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c update_tasks_nodemask(&top_cpuset); } -@@ -2420,11 +2420,11 @@ +@@ -2420,11 +2420,11 @@ void cpuset_cpus_allowed(struct task_struct *tsk, struct cpumask *pmask) { unsigned long flags; @@ -11386,7 +11770,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c } void cpuset_cpus_allowed_fallback(struct task_struct *tsk) -@@ -2472,11 +2472,11 @@ +@@ -2472,11 +2472,11 @@ nodemask_t cpuset_mems_allowed(struct task_struct *tsk) nodemask_t mask; unsigned long flags; @@ -11400,7 +11784,7 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c return mask; } -@@ -2568,14 +2568,14 @@ +@@ -2568,14 +2568,14 @@ bool __cpuset_node_allowed(int node, gfp_t gfp_mask) return true; /* Not hardwall and node outside mems_allowed: scan up cpusets */ @@ -11417,10 +11801,11 @@ diff -Nur linux-4.9.28.orig/kernel/cpuset.c linux-4.9.28/kernel/cpuset.c return allowed; } -diff -Nur linux-4.9.28.orig/kernel/debug/kdb/kdb_io.c linux-4.9.28/kernel/debug/kdb/kdb_io.c ---- linux-4.9.28.orig/kernel/debug/kdb/kdb_io.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/debug/kdb/kdb_io.c 2017-05-19 03:37:25.182176801 +0200 -@@ -554,7 +554,6 @@ +diff --git a/kernel/debug/kdb/kdb_io.c b/kernel/debug/kdb/kdb_io.c +index fc1ef736253c..83c666537a7a 100644 +--- a/kernel/debug/kdb/kdb_io.c ++++ b/kernel/debug/kdb/kdb_io.c +@@ -554,7 +554,6 @@ int vkdb_printf(enum kdb_msgsrc src, const char *fmt, va_list ap) int linecount; int colcount; int logging, saved_loglevel = 0; @@ -11428,7 +11813,7 @@ diff -Nur linux-4.9.28.orig/kernel/debug/kdb/kdb_io.c linux-4.9.28/kernel/debug/ int got_printf_lock = 0; int retlen = 0; int fnd, len; -@@ -565,8 +564,6 @@ +@@ -565,8 +564,6 @@ int vkdb_printf(enum kdb_msgsrc src, const char *fmt, va_list ap) unsigned long uninitialized_var(flags); preempt_disable(); @@ -11437,7 +11822,7 @@ diff -Nur linux-4.9.28.orig/kernel/debug/kdb/kdb_io.c linux-4.9.28/kernel/debug/ /* Serialize kdb_printf if multiple cpus try to write at once. * But if any cpu goes recursive in kdb, just print the output, -@@ -855,7 +852,6 @@ +@@ -855,7 +852,6 @@ int vkdb_printf(enum kdb_msgsrc src, const char *fmt, va_list ap) } else { __release(kdb_printf_lock); } @@ -11445,7 +11830,7 @@ diff -Nur linux-4.9.28.orig/kernel/debug/kdb/kdb_io.c linux-4.9.28/kernel/debug/ preempt_enable(); return retlen; } -@@ -865,9 +861,11 @@ +@@ -865,9 +861,11 @@ int kdb_printf(const char *fmt, ...) va_list ap; int r; @@ -11457,10 +11842,11 @@ diff -Nur linux-4.9.28.orig/kernel/debug/kdb/kdb_io.c linux-4.9.28/kernel/debug/ return r; } -diff -Nur linux-4.9.28.orig/kernel/events/core.c linux-4.9.28/kernel/events/core.c ---- linux-4.9.28.orig/kernel/events/core.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/events/core.c 2017-05-19 03:37:25.182176801 +0200 -@@ -1050,6 +1050,7 @@ +diff --git a/kernel/events/core.c b/kernel/events/core.c +index 07c0dc806dfc..baf1a2867d74 100644 +--- a/kernel/events/core.c ++++ b/kernel/events/core.c +@@ -1050,6 +1050,7 @@ static void __perf_mux_hrtimer_init(struct perf_cpu_context *cpuctx, int cpu) raw_spin_lock_init(&cpuctx->hrtimer_lock); hrtimer_init(timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS_PINNED); timer->function = perf_mux_hrtimer_handler; @@ -11468,7 +11854,7 @@ diff -Nur linux-4.9.28.orig/kernel/events/core.c linux-4.9.28/kernel/events/core } static int perf_mux_hrtimer_restart(struct perf_cpu_context *cpuctx) -@@ -8363,6 +8364,7 @@ +@@ -8363,6 +8364,7 @@ static void perf_swevent_init_hrtimer(struct perf_event *event) hrtimer_init(&hwc->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); hwc->hrtimer.function = perf_swevent_hrtimer; @@ -11476,10 +11862,11 @@ diff -Nur linux-4.9.28.orig/kernel/events/core.c linux-4.9.28/kernel/events/core /* * Since hrtimers have a fixed rate, we can do a static freq->period -diff -Nur linux-4.9.28.orig/kernel/exit.c linux-4.9.28/kernel/exit.c ---- linux-4.9.28.orig/kernel/exit.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/exit.c 2017-05-19 03:37:25.182176801 +0200 -@@ -143,7 +143,7 @@ +diff --git a/kernel/exit.c b/kernel/exit.c +index 3076f3089919..fb2ebcf3ca7c 100644 +--- a/kernel/exit.c ++++ b/kernel/exit.c +@@ -143,7 +143,7 @@ static void __exit_signal(struct task_struct *tsk) * Do this under ->siglock, we can race with another thread * doing sigqueue_free() if we have SIGQUEUE_PREALLOC signals. */ @@ -11488,9 +11875,10 @@ diff -Nur linux-4.9.28.orig/kernel/exit.c linux-4.9.28/kernel/exit.c tsk->sighand = NULL; spin_unlock(&sighand->siglock); -diff -Nur linux-4.9.28.orig/kernel/fork.c linux-4.9.28/kernel/fork.c ---- linux-4.9.28.orig/kernel/fork.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/fork.c 2017-05-19 03:37:25.182176801 +0200 +diff --git a/kernel/fork.c b/kernel/fork.c +index 59faac4de181..0edb0f3c1db8 100644 +--- a/kernel/fork.c ++++ b/kernel/fork.c @@ -76,6 +76,7 @@ #include #include @@ -11499,7 +11887,7 @@ diff -Nur linux-4.9.28.orig/kernel/fork.c linux-4.9.28/kernel/fork.c #include #include -@@ -376,13 +377,24 @@ +@@ -376,13 +377,24 @@ static inline void put_signal_struct(struct signal_struct *sig) if (atomic_dec_and_test(&sig->sigcnt)) free_signal_struct(sig); } @@ -11525,7 +11913,7 @@ diff -Nur linux-4.9.28.orig/kernel/fork.c linux-4.9.28/kernel/fork.c cgroup_free(tsk); task_numa_free(tsk); security_task_free(tsk); -@@ -393,7 +405,18 @@ +@@ -393,7 +405,18 @@ void __put_task_struct(struct task_struct *tsk) if (!profile_handoff_task(tsk)) free_task(tsk); } @@ -11544,7 +11932,7 @@ diff -Nur linux-4.9.28.orig/kernel/fork.c linux-4.9.28/kernel/fork.c void __init __weak arch_task_cache_init(void) { } -@@ -852,6 +875,19 @@ +@@ -852,6 +875,19 @@ void __mmdrop(struct mm_struct *mm) } EXPORT_SYMBOL_GPL(__mmdrop); @@ -11564,7 +11952,7 @@ diff -Nur linux-4.9.28.orig/kernel/fork.c linux-4.9.28/kernel/fork.c static inline void __mmput(struct mm_struct *mm) { VM_BUG_ON(atomic_read(&mm->mm_users)); -@@ -1417,6 +1453,7 @@ +@@ -1417,6 +1453,7 @@ static void rt_mutex_init_task(struct task_struct *p) #ifdef CONFIG_RT_MUTEXES p->pi_waiters = RB_ROOT; p->pi_waiters_leftmost = NULL; @@ -11572,7 +11960,7 @@ diff -Nur linux-4.9.28.orig/kernel/fork.c linux-4.9.28/kernel/fork.c p->pi_blocked_on = NULL; #endif } -@@ -1426,6 +1463,9 @@ +@@ -1426,6 +1463,9 @@ static void rt_mutex_init_task(struct task_struct *p) */ static void posix_cpu_timers_init(struct task_struct *tsk) { @@ -11582,7 +11970,7 @@ diff -Nur linux-4.9.28.orig/kernel/fork.c linux-4.9.28/kernel/fork.c tsk->cputime_expires.prof_exp = 0; tsk->cputime_expires.virt_exp = 0; tsk->cputime_expires.sched_exp = 0; -@@ -1552,6 +1592,7 @@ +@@ -1552,6 +1592,7 @@ static __latent_entropy struct task_struct *copy_process( spin_lock_init(&p->alloc_lock); init_sigpending(&p->pending); @@ -11590,10 +11978,11 @@ diff -Nur linux-4.9.28.orig/kernel/fork.c linux-4.9.28/kernel/fork.c p->utime = p->stime = p->gtime = 0; p->utimescaled = p->stimescaled = 0; -diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c ---- linux-4.9.28.orig/kernel/futex.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/futex.c 2017-05-19 03:37:25.182176801 +0200 -@@ -800,7 +800,7 @@ +diff --git a/kernel/futex.c b/kernel/futex.c +index 4c6b6e697b73..d9bab63efccb 100644 +--- a/kernel/futex.c ++++ b/kernel/futex.c +@@ -800,7 +800,7 @@ static int refill_pi_state_cache(void) return 0; } @@ -11602,7 +11991,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c { struct futex_pi_state *pi_state = current->pi_state_cache; -@@ -810,6 +810,11 @@ +@@ -810,6 +810,11 @@ static struct futex_pi_state * alloc_pi_state(void) return pi_state; } @@ -11614,7 +12003,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c /* * Drops a reference to the pi_state object and frees or caches it * when the last reference is gone. -@@ -854,7 +859,7 @@ +@@ -854,7 +859,7 @@ static void put_pi_state(struct futex_pi_state *pi_state) * Look up the task based on what TID userspace gave us. * We dont trust it. */ @@ -11623,7 +12012,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c { struct task_struct *p; -@@ -904,7 +909,9 @@ +@@ -904,7 +909,9 @@ void exit_pi_state_list(struct task_struct *curr) * task still owns the PI-state: */ if (head->next != next) { @@ -11633,7 +12022,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c continue; } -@@ -914,10 +921,12 @@ +@@ -914,10 +921,12 @@ void exit_pi_state_list(struct task_struct *curr) pi_state->owner = NULL; raw_spin_unlock_irq(&curr->pi_lock); @@ -11648,7 +12037,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c raw_spin_lock_irq(&curr->pi_lock); } raw_spin_unlock_irq(&curr->pi_lock); -@@ -971,6 +980,39 @@ +@@ -971,6 +980,39 @@ void exit_pi_state_list(struct task_struct *curr) * * [10] There is no transient state which leaves owner and user space * TID out of sync. @@ -11688,7 +12077,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c */ /* -@@ -978,10 +1020,13 @@ +@@ -978,10 +1020,13 @@ void exit_pi_state_list(struct task_struct *curr) * the pi_state against the user space value. If correct, attach to * it. */ @@ -11703,7 +12092,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c /* * Userspace might have messed up non-PI and PI futexes [3] -@@ -989,9 +1034,39 @@ +@@ -989,9 +1034,39 @@ static int attach_to_pi_state(u32 uval, struct futex_pi_state *pi_state, if (unlikely(!pi_state)) return -EINVAL; @@ -11743,7 +12132,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c * Handle the owner died case: */ if (uval & FUTEX_OWNER_DIED) { -@@ -1006,11 +1081,11 @@ +@@ -1006,11 +1081,11 @@ static int attach_to_pi_state(u32 uval, struct futex_pi_state *pi_state, * is not 0. Inconsistent state. [5] */ if (pid) @@ -11757,7 +12146,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c } /* -@@ -1022,14 +1097,14 @@ +@@ -1022,14 +1097,14 @@ static int attach_to_pi_state(u32 uval, struct futex_pi_state *pi_state, * Take a ref on the state and return success. [6] */ if (!pid) @@ -11774,7 +12163,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c } /* -@@ -1038,11 +1113,29 @@ +@@ -1038,11 +1113,29 @@ static int attach_to_pi_state(u32 uval, struct futex_pi_state *pi_state, * user space TID. [9/10] */ if (pid != task_pid_vnr(pi_state->owner)) @@ -11807,7 +12196,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c } /* -@@ -1093,6 +1186,9 @@ +@@ -1093,6 +1186,9 @@ static int attach_to_pi_owner(u32 uval, union futex_key *key, /* * No existing pi state. First waiter. [2] @@ -11817,7 +12206,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c */ pi_state = alloc_pi_state(); -@@ -1117,17 +1213,18 @@ +@@ -1117,17 +1213,18 @@ static int attach_to_pi_owner(u32 uval, union futex_key *key, return 0; } @@ -11840,7 +12229,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c /* * We are the first waiter - try to look up the owner based on -@@ -1146,7 +1243,7 @@ +@@ -1146,7 +1243,7 @@ static int lock_pi_update_atomic(u32 __user *uaddr, u32 uval, u32 newval) if (unlikely(cmpxchg_futex_value_locked(&curval, uaddr, uval, newval))) return -EFAULT; @@ -11849,7 +12238,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c return curval != uval ? -EAGAIN : 0; } -@@ -1174,7 +1271,7 @@ +@@ -1174,7 +1271,7 @@ static int futex_lock_pi_atomic(u32 __user *uaddr, struct futex_hash_bucket *hb, struct task_struct *task, int set_waiters) { u32 uval, newval, vpid = task_pid_vnr(task); @@ -11858,7 +12247,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c int ret; /* -@@ -1200,9 +1297,9 @@ +@@ -1200,9 +1297,9 @@ static int futex_lock_pi_atomic(u32 __user *uaddr, struct futex_hash_bucket *hb, * Lookup existing state first. If it exists, try to attach to * its pi_state. */ @@ -11871,7 +12260,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c /* * No waiter and user TID is 0. We are here because the -@@ -1283,50 +1380,45 @@ +@@ -1283,50 +1380,45 @@ static void mark_wake_futex(struct wake_q_head *wake_q, struct futex_q *q) wake_q_add(wake_q, p); __unqueue_futex(q); /* @@ -11950,7 +12339,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c */ newval = FUTEX_WAITERS | task_pid_vnr(new_owner); -@@ -1335,6 +1427,7 @@ +@@ -1335,6 +1427,7 @@ static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, if (cmpxchg_futex_value_locked(&curval, uaddr, uval, newval)) { ret = -EFAULT; @@ -11958,7 +12347,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c } else if (curval != uval) { /* * If a unconditional UNLOCK_PI operation (user space did not -@@ -1347,10 +1440,14 @@ +@@ -1347,10 +1440,14 @@ static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, else ret = -EINVAL; } @@ -11977,7 +12366,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c raw_spin_lock(&pi_state->owner->pi_lock); WARN_ON(list_empty(&pi_state->list)); -@@ -1363,22 +1460,15 @@ +@@ -1363,22 +1460,15 @@ static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, pi_state->owner = new_owner; raw_spin_unlock(&new_owner->pi_lock); @@ -12006,7 +12395,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c } /* -@@ -1824,7 +1914,7 @@ +@@ -1824,7 +1914,7 @@ static int futex_requeue(u32 __user *uaddr1, unsigned int flags, * If that call succeeds then we have pi_state and an * initial refcount on it. */ @@ -12015,7 +12404,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c } switch (ret) { -@@ -1907,7 +1997,7 @@ +@@ -1907,7 +1997,7 @@ static int futex_requeue(u32 __user *uaddr1, unsigned int flags, * refcount on the pi_state and store the pointer in * the futex_q object of the waiter. */ @@ -12024,7 +12413,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c this->pi_state = pi_state; ret = rt_mutex_start_proxy_lock(&pi_state->pi_mutex, this->rt_waiter, -@@ -1924,6 +2014,16 @@ +@@ -1924,6 +2014,16 @@ static int futex_requeue(u32 __user *uaddr1, unsigned int flags, requeue_pi_wake_futex(this, &key2, hb2); drop_count++; continue; @@ -12041,7 +12430,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c } else if (ret) { /* * rt_mutex_start_proxy_lock() detected a -@@ -2007,20 +2107,7 @@ +@@ -2007,20 +2107,7 @@ queue_unlock(struct futex_hash_bucket *hb) hb_waiters_dec(hb); } @@ -12063,7 +12452,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c { int prio; -@@ -2037,6 +2124,24 @@ +@@ -2037,6 +2124,24 @@ static inline void queue_me(struct futex_q *q, struct futex_hash_bucket *hb) plist_node_init(&q->list, prio); plist_add(&q->list, &hb->chain); q->task = current; @@ -12088,7 +12477,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c spin_unlock(&hb->lock); } -@@ -2123,10 +2228,13 @@ +@@ -2123,10 +2228,13 @@ static int fixup_pi_state_owner(u32 __user *uaddr, struct futex_q *q, { u32 newtid = task_pid_vnr(newowner) | FUTEX_WAITERS; struct futex_pi_state *pi_state = q->pi_state; @@ -12103,7 +12492,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c /* Owner died? */ if (!pi_state->owner) newtid |= FUTEX_OWNER_DIED; -@@ -2134,7 +2242,8 @@ +@@ -2134,7 +2242,8 @@ static int fixup_pi_state_owner(u32 __user *uaddr, struct futex_q *q, /* * We are here either because we stole the rtmutex from the * previous highest priority waiter or we are the highest priority @@ -12113,7 +12502,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c * We have to replace the newowner TID in the user space variable. * This must be atomic as we have to preserve the owner died bit here. * -@@ -2142,17 +2251,16 @@ +@@ -2142,17 +2251,16 @@ static int fixup_pi_state_owner(u32 __user *uaddr, struct futex_q *q, * because we can fault here. Imagine swapped out pages or a fork * that marked all the anonymous memory readonly for cow. * @@ -12136,7 +12525,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c newval = (uval & FUTEX_OWNER_DIED) | newtid; if (cmpxchg_futex_value_locked(&curval, uaddr, uval, newval)) -@@ -2167,47 +2275,60 @@ +@@ -2167,47 +2275,60 @@ static int fixup_pi_state_owner(u32 __user *uaddr, struct futex_q *q, * itself. */ if (pi_state->owner != NULL) { @@ -12212,7 +12601,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c } static long futex_wait_restart(struct restart_block *restart); -@@ -2229,13 +2350,16 @@ +@@ -2229,13 +2350,16 @@ static long futex_wait_restart(struct restart_block *restart); */ static int fixup_owner(u32 __user *uaddr, struct futex_q *q, int locked) { @@ -12230,7 +12619,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c */ if (q->pi_state->owner != current) ret = fixup_pi_state_owner(uaddr, q, current); -@@ -2243,43 +2367,15 @@ +@@ -2243,43 +2367,15 @@ static int fixup_owner(u32 __user *uaddr, struct futex_q *q, int locked) } /* @@ -12276,7 +12665,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c out: return ret ? ret : locked; -@@ -2503,6 +2599,8 @@ +@@ -2503,6 +2599,8 @@ static int futex_lock_pi(u32 __user *uaddr, unsigned int flags, ktime_t *time, int trylock) { struct hrtimer_sleeper timeout, *to = NULL; @@ -12285,7 +12674,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c struct futex_hash_bucket *hb; struct futex_q q = futex_q_init; int res, ret; -@@ -2555,25 +2653,77 @@ +@@ -2555,25 +2653,77 @@ static int futex_lock_pi(u32 __user *uaddr, unsigned int flags, } } @@ -12372,7 +12761,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c * Fixup the pi_state owner and possibly acquire the lock if we * haven't already. */ -@@ -2589,12 +2739,19 @@ +@@ -2589,12 +2739,19 @@ static int futex_lock_pi(u32 __user *uaddr, unsigned int flags, * If fixup_owner() faulted and was unable to handle the fault, unlock * it and return the fault to userspace. */ @@ -12394,7 +12783,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c goto out_put_key; out_unlock_put_key: -@@ -2603,8 +2760,10 @@ +@@ -2603,8 +2760,10 @@ static int futex_lock_pi(u32 __user *uaddr, unsigned int flags, out_put_key: put_futex_key(&q.key); out: @@ -12406,7 +12795,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c return ret != -EINTR ? ret : -ERESTARTNOINTR; uaddr_faulted: -@@ -2631,7 +2790,7 @@ +@@ -2631,7 +2790,7 @@ static int futex_unlock_pi(u32 __user *uaddr, unsigned int flags) u32 uninitialized_var(curval), uval, vpid = task_pid_vnr(current); union futex_key key = FUTEX_KEY_INIT; struct futex_hash_bucket *hb; @@ -12415,7 +12804,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c int ret; retry: -@@ -2655,12 +2814,48 @@ +@@ -2655,12 +2814,48 @@ static int futex_unlock_pi(u32 __user *uaddr, unsigned int flags) * all and we at least want to know if user space fiddled * with the futex value instead of blindly unlocking. */ @@ -12469,7 +12858,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c */ if (!ret) goto out_putkey; -@@ -2675,7 +2870,6 @@ +@@ -2675,7 +2870,6 @@ static int futex_unlock_pi(u32 __user *uaddr, unsigned int flags) * setting the FUTEX_WAITERS bit. Try again. */ if (ret == -EAGAIN) { @@ -12477,7 +12866,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c put_futex_key(&key); goto retry; } -@@ -2683,7 +2877,7 @@ +@@ -2683,7 +2877,7 @@ static int futex_unlock_pi(u32 __user *uaddr, unsigned int flags) * wake_futex_pi has detected invalid state. Tell user * space. */ @@ -12486,7 +12875,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c } /* -@@ -2693,8 +2887,10 @@ +@@ -2693,8 +2887,10 @@ static int futex_unlock_pi(u32 __user *uaddr, unsigned int flags) * preserve the WAITERS bit not the OWNER_DIED one. We are the * owner. */ @@ -12498,7 +12887,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c /* * If uval has changed, let user space handle it. -@@ -2708,7 +2904,6 @@ +@@ -2708,7 +2904,6 @@ static int futex_unlock_pi(u32 __user *uaddr, unsigned int flags) return ret; pi_faulted: @@ -12506,7 +12895,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c put_futex_key(&key); ret = fault_in_user_writeable(uaddr); -@@ -2812,8 +3007,9 @@ +@@ -2812,8 +3007,9 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, u32 __user *uaddr2) { struct hrtimer_sleeper timeout, *to = NULL; @@ -12517,7 +12906,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c union futex_key key2 = FUTEX_KEY_INIT; struct futex_q q = futex_q_init; int res, ret; -@@ -2838,10 +3034,7 @@ +@@ -2838,10 +3034,7 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, * The waiter is allocated on our stack, manipulated by the requeue * code while we sleep on uaddr. */ @@ -12529,7 +12918,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2, VERIFY_WRITE); if (unlikely(ret != 0)) -@@ -2872,20 +3065,55 @@ +@@ -2872,20 +3065,55 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, /* Queue the futex_q, drop the hb lock, wait for wakeup. */ futex_wait_queue_me(hb, &q, to); @@ -12596,7 +12985,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c /* Check if the requeue code acquired the second futex for us. */ if (!q.rt_waiter) { -@@ -2894,16 +3122,19 @@ +@@ -2894,16 +3122,19 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, * did a lock-steal - fix up the PI-state in that case. */ if (q.pi_state && (q.pi_state->owner != current)) { @@ -12620,7 +13009,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c } } else { struct rt_mutex *pi_mutex; -@@ -2915,10 +3146,14 @@ +@@ -2915,10 +3146,14 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, */ WARN_ON(!q.pi_state); pi_mutex = &q.pi_state->pi_mutex; @@ -12638,7 +13027,7 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c /* * Fixup the pi_state owner and possibly acquire the lock if we * haven't already. -@@ -2936,13 +3171,20 @@ +@@ -2936,13 +3171,20 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, * the fault, unlock the rt_mutex and return the fault to * userspace. */ @@ -12661,10 +13050,11 @@ diff -Nur linux-4.9.28.orig/kernel/futex.c linux-4.9.28/kernel/futex.c if (ret == -EINTR) { /* * We've already been requeued, but cannot restart by calling -diff -Nur linux-4.9.28.orig/kernel/irq/handle.c linux-4.9.28/kernel/irq/handle.c ---- linux-4.9.28.orig/kernel/irq/handle.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/irq/handle.c 2017-05-19 03:37:25.186176955 +0200 -@@ -181,10 +181,16 @@ +diff --git a/kernel/irq/handle.c b/kernel/irq/handle.c +index d3f24905852c..f87aa8fdcc51 100644 +--- a/kernel/irq/handle.c ++++ b/kernel/irq/handle.c +@@ -181,10 +181,16 @@ irqreturn_t handle_irq_event_percpu(struct irq_desc *desc) { irqreturn_t retval; unsigned int flags = 0; @@ -12682,9 +13072,10 @@ diff -Nur linux-4.9.28.orig/kernel/irq/handle.c linux-4.9.28/kernel/irq/handle.c if (!noirqdebug) note_interrupt(desc, retval); -diff -Nur linux-4.9.28.orig/kernel/irq/manage.c linux-4.9.28/kernel/irq/manage.c ---- linux-4.9.28.orig/kernel/irq/manage.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/irq/manage.c 2017-05-19 03:37:25.186176955 +0200 +diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c +index 6b669593e7eb..e357bf6c59d5 100644 +--- a/kernel/irq/manage.c ++++ b/kernel/irq/manage.c @@ -22,6 +22,7 @@ #include "internals.h" @@ -12693,7 +13084,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq/manage.c linux-4.9.28/kernel/irq/manage.c __read_mostly bool force_irqthreads; static int __init setup_forced_irqthreads(char *arg) -@@ -30,6 +31,7 @@ +@@ -30,6 +31,7 @@ static int __init setup_forced_irqthreads(char *arg) return 0; } early_param("threadirqs", setup_forced_irqthreads); @@ -12701,7 +13092,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq/manage.c linux-4.9.28/kernel/irq/manage.c #endif static void __synchronize_hardirq(struct irq_desc *desc) -@@ -233,7 +235,12 @@ +@@ -233,7 +235,12 @@ int irq_set_affinity_locked(struct irq_data *data, const struct cpumask *mask, if (desc->affinity_notify) { kref_get(&desc->affinity_notify->kref); @@ -12714,7 +13105,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq/manage.c linux-4.9.28/kernel/irq/manage.c } irqd_set(data, IRQD_AFFINITY_SET); -@@ -271,10 +278,8 @@ +@@ -271,10 +278,8 @@ int irq_set_affinity_hint(unsigned int irq, const struct cpumask *m) } EXPORT_SYMBOL_GPL(irq_set_affinity_hint); @@ -12726,7 +13117,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq/manage.c linux-4.9.28/kernel/irq/manage.c struct irq_desc *desc = irq_to_desc(notify->irq); cpumask_var_t cpumask; unsigned long flags; -@@ -296,6 +301,35 @@ +@@ -296,6 +301,35 @@ static void irq_affinity_notify(struct work_struct *work) kref_put(¬ify->kref, notify->release); } @@ -12762,7 +13153,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq/manage.c linux-4.9.28/kernel/irq/manage.c /** * irq_set_affinity_notifier - control notification of IRQ affinity changes * @irq: Interrupt for which to enable/disable notification -@@ -324,7 +358,12 @@ +@@ -324,7 +358,12 @@ irq_set_affinity_notifier(unsigned int irq, struct irq_affinity_notify *notify) if (notify) { notify->irq = irq; kref_init(¬ify->kref); @@ -12775,7 +13166,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq/manage.c linux-4.9.28/kernel/irq/manage.c } raw_spin_lock_irqsave(&desc->lock, flags); -@@ -879,7 +918,15 @@ +@@ -879,7 +918,15 @@ irq_forced_thread_fn(struct irq_desc *desc, struct irqaction *action) local_bh_disable(); ret = action->thread_fn(action->irq, action->dev_id); irq_finalize_oneshot(desc, action); @@ -12792,7 +13183,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq/manage.c linux-4.9.28/kernel/irq/manage.c return ret; } -@@ -976,6 +1023,12 @@ +@@ -976,6 +1023,12 @@ static int irq_thread(void *data) if (action_ret == IRQ_WAKE_THREAD) irq_wake_secondary(desc, action); @@ -12805,7 +13196,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq/manage.c linux-4.9.28/kernel/irq/manage.c wake_threads_waitq(desc); } -@@ -1336,6 +1389,9 @@ +@@ -1336,6 +1389,9 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new) irqd_set(&desc->irq_data, IRQD_NO_BALANCING); } @@ -12815,7 +13206,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq/manage.c linux-4.9.28/kernel/irq/manage.c /* Set default affinity mask once everything is setup */ setup_affinity(desc, mask); -@@ -2061,7 +2117,7 @@ +@@ -2061,7 +2117,7 @@ EXPORT_SYMBOL_GPL(irq_get_irqchip_state); * This call sets the internal irqchip state of an interrupt, * depending on the value of @which. * @@ -12824,10 +13215,11 @@ diff -Nur linux-4.9.28.orig/kernel/irq/manage.c linux-4.9.28/kernel/irq/manage.c * interrupt controller has per-cpu registers. */ int irq_set_irqchip_state(unsigned int irq, enum irqchip_irq_state which, -diff -Nur linux-4.9.28.orig/kernel/irq/settings.h linux-4.9.28/kernel/irq/settings.h ---- linux-4.9.28.orig/kernel/irq/settings.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/irq/settings.h 2017-05-19 03:37:25.186176955 +0200 -@@ -16,6 +16,7 @@ +diff --git a/kernel/irq/settings.h b/kernel/irq/settings.h +index 320579d89091..2df2d4445b1e 100644 +--- a/kernel/irq/settings.h ++++ b/kernel/irq/settings.h +@@ -16,6 +16,7 @@ enum { _IRQ_PER_CPU_DEVID = IRQ_PER_CPU_DEVID, _IRQ_IS_POLLED = IRQ_IS_POLLED, _IRQ_DISABLE_UNLAZY = IRQ_DISABLE_UNLAZY, @@ -12835,7 +13227,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq/settings.h linux-4.9.28/kernel/irq/settin _IRQF_MODIFY_MASK = IRQF_MODIFY_MASK, }; -@@ -30,6 +31,7 @@ +@@ -30,6 +31,7 @@ enum { #define IRQ_PER_CPU_DEVID GOT_YOU_MORON #define IRQ_IS_POLLED GOT_YOU_MORON #define IRQ_DISABLE_UNLAZY GOT_YOU_MORON @@ -12843,7 +13235,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq/settings.h linux-4.9.28/kernel/irq/settin #undef IRQF_MODIFY_MASK #define IRQF_MODIFY_MASK GOT_YOU_MORON -@@ -40,6 +42,16 @@ +@@ -40,6 +42,16 @@ irq_settings_clr_and_set(struct irq_desc *desc, u32 clr, u32 set) desc->status_use_accessors |= (set & _IRQF_MODIFY_MASK); } @@ -12860,10 +13252,11 @@ diff -Nur linux-4.9.28.orig/kernel/irq/settings.h linux-4.9.28/kernel/irq/settin static inline bool irq_settings_is_per_cpu(struct irq_desc *desc) { return desc->status_use_accessors & _IRQ_PER_CPU; -diff -Nur linux-4.9.28.orig/kernel/irq/spurious.c linux-4.9.28/kernel/irq/spurious.c ---- linux-4.9.28.orig/kernel/irq/spurious.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/irq/spurious.c 2017-05-19 03:37:25.186176955 +0200 -@@ -442,6 +442,10 @@ +diff --git a/kernel/irq/spurious.c b/kernel/irq/spurious.c +index 5707f97a3e6a..73f38dc7a7fb 100644 +--- a/kernel/irq/spurious.c ++++ b/kernel/irq/spurious.c +@@ -442,6 +442,10 @@ MODULE_PARM_DESC(noirqdebug, "Disable irq lockup detection when true"); static int __init irqfixup_setup(char *str) { @@ -12874,7 +13267,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq/spurious.c linux-4.9.28/kernel/irq/spurio irqfixup = 1; printk(KERN_WARNING "Misrouted IRQ fixup support enabled.\n"); printk(KERN_WARNING "This may impact system performance.\n"); -@@ -454,6 +458,10 @@ +@@ -454,6 +458,10 @@ module_param(irqfixup, int, 0644); static int __init irqpoll_setup(char *str) { @@ -12885,9 +13278,10 @@ diff -Nur linux-4.9.28.orig/kernel/irq/spurious.c linux-4.9.28/kernel/irq/spurio irqfixup = 2; printk(KERN_WARNING "Misrouted IRQ fixup and polling support " "enabled\n"); -diff -Nur linux-4.9.28.orig/kernel/irq_work.c linux-4.9.28/kernel/irq_work.c ---- linux-4.9.28.orig/kernel/irq_work.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/irq_work.c 2017-05-19 03:37:25.186176955 +0200 +diff --git a/kernel/irq_work.c b/kernel/irq_work.c +index bcf107ce0854..2899ba0d23d1 100644 +--- a/kernel/irq_work.c ++++ b/kernel/irq_work.c @@ -17,6 +17,7 @@ #include #include @@ -12896,7 +13290,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq_work.c linux-4.9.28/kernel/irq_work.c #include -@@ -65,6 +66,8 @@ +@@ -65,6 +66,8 @@ void __weak arch_irq_work_raise(void) */ bool irq_work_queue_on(struct irq_work *work, int cpu) { @@ -12905,7 +13299,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq_work.c linux-4.9.28/kernel/irq_work.c /* All work should have been flushed before going offline */ WARN_ON_ONCE(cpu_is_offline(cpu)); -@@ -75,7 +78,12 @@ +@@ -75,7 +78,12 @@ bool irq_work_queue_on(struct irq_work *work, int cpu) if (!irq_work_claim(work)) return false; @@ -12919,7 +13313,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq_work.c linux-4.9.28/kernel/irq_work.c arch_send_call_function_single_ipi(cpu); return true; -@@ -86,6 +94,9 @@ +@@ -86,6 +94,9 @@ EXPORT_SYMBOL_GPL(irq_work_queue_on); /* Enqueue the irq work @work on the current CPU */ bool irq_work_queue(struct irq_work *work) { @@ -12929,7 +13323,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq_work.c linux-4.9.28/kernel/irq_work.c /* Only queue if not already pending */ if (!irq_work_claim(work)) return false; -@@ -93,13 +104,15 @@ +@@ -93,13 +104,15 @@ bool irq_work_queue(struct irq_work *work) /* Queue the entry and raise the IPI if needed. */ preempt_disable(); @@ -12952,7 +13346,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq_work.c linux-4.9.28/kernel/irq_work.c arch_irq_work_raise(); } -@@ -116,9 +129,8 @@ +@@ -116,9 +129,8 @@ bool irq_work_needs_cpu(void) raised = this_cpu_ptr(&raised_list); lazy = this_cpu_ptr(&lazy_list); @@ -12964,7 +13358,7 @@ diff -Nur linux-4.9.28.orig/kernel/irq_work.c linux-4.9.28/kernel/irq_work.c /* All work should have been flushed before going offline */ WARN_ON_ONCE(cpu_is_offline(smp_processor_id())); -@@ -132,7 +144,7 @@ +@@ -132,7 +144,7 @@ static void irq_work_run_list(struct llist_head *list) struct irq_work *work; struct llist_node *llnode; @@ -12973,118 +13367,47 @@ diff -Nur linux-4.9.28.orig/kernel/irq_work.c linux-4.9.28/kernel/irq_work.c if (llist_empty(list)) return; -@@ -169,7 +181,16 @@ +@@ -169,7 +181,16 @@ static void irq_work_run_list(struct llist_head *list) void irq_work_run(void) { irq_work_run_list(this_cpu_ptr(&raised_list)); - irq_work_run_list(this_cpu_ptr(&lazy_list)); + if (IS_ENABLED(CONFIG_PREEMPT_RT_FULL)) { + /* -+ * NOTE: we raise softirq via IPI for safety, -+ * and execute in irq_work_tick() to move the -+ * overhead from hard to soft irq context. -+ */ -+ if (!llist_empty(this_cpu_ptr(&lazy_list))) -+ raise_softirq(TIMER_SOFTIRQ); -+ } else -+ irq_work_run_list(this_cpu_ptr(&lazy_list)); - } - EXPORT_SYMBOL_GPL(irq_work_run); - -@@ -179,8 +200,17 @@ - - if (!llist_empty(raised) && !arch_irq_work_has_interrupt()) - irq_work_run_list(raised); -+ -+ if (!IS_ENABLED(CONFIG_PREEMPT_RT_FULL)) -+ irq_work_run_list(this_cpu_ptr(&lazy_list)); -+} -+ -+#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) -+void irq_work_tick_soft(void) -+{ - irq_work_run_list(this_cpu_ptr(&lazy_list)); - } -+#endif - - /* - * Synchronize against the irq_work @entry, ensures the entry is not -diff -Nur linux-4.9.28.orig/kernel/Kconfig.locks linux-4.9.28/kernel/Kconfig.locks ---- linux-4.9.28.orig/kernel/Kconfig.locks 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/Kconfig.locks 2017-05-19 03:37:25.178176648 +0200 -@@ -225,11 +225,11 @@ - - config MUTEX_SPIN_ON_OWNER - def_bool y -- depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW -+ depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL - - config RWSEM_SPIN_ON_OWNER - def_bool y -- depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW -+ depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL - - config LOCK_SPIN_ON_OWNER - def_bool y -diff -Nur linux-4.9.28.orig/kernel/Kconfig.preempt linux-4.9.28/kernel/Kconfig.preempt ---- linux-4.9.28.orig/kernel/Kconfig.preempt 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/Kconfig.preempt 2017-05-19 03:37:25.178176648 +0200 -@@ -1,3 +1,16 @@ -+config PREEMPT -+ bool -+ select PREEMPT_COUNT -+ -+config PREEMPT_RT_BASE -+ bool -+ select PREEMPT -+ -+config HAVE_PREEMPT_LAZY -+ bool -+ -+config PREEMPT_LAZY -+ def_bool y if HAVE_PREEMPT_LAZY && PREEMPT_RT_FULL - - choice - prompt "Preemption Model" -@@ -33,9 +46,9 @@ - - Select this if you are building a kernel for a desktop system. ++ * NOTE: we raise softirq via IPI for safety, ++ * and execute in irq_work_tick() to move the ++ * overhead from hard to soft irq context. ++ */ ++ if (!llist_empty(this_cpu_ptr(&lazy_list))) ++ raise_softirq(TIMER_SOFTIRQ); ++ } else ++ irq_work_run_list(this_cpu_ptr(&lazy_list)); + } + EXPORT_SYMBOL_GPL(irq_work_run); --config PREEMPT -+config PREEMPT__LL - bool "Preemptible Kernel (Low-Latency Desktop)" -- select PREEMPT_COUNT -+ select PREEMPT - select UNINLINE_SPIN_UNLOCK if !ARCH_INLINE_SPIN_UNLOCK - help - This option reduces the latency of the kernel by making -@@ -52,6 +65,22 @@ - embedded system with latency requirements in the milliseconds - range. +@@ -179,8 +200,17 @@ void irq_work_tick(void) -+config PREEMPT_RTB -+ bool "Preemptible Kernel (Basic RT)" -+ select PREEMPT_RT_BASE -+ help -+ This option is basically the same as (Low-Latency Desktop) but -+ enables changes which are preliminary for the full preemptible -+ RT kernel. + if (!llist_empty(raised) && !arch_irq_work_has_interrupt()) + irq_work_run_list(raised); + -+config PREEMPT_RT_FULL -+ bool "Fully Preemptible Kernel (RT)" -+ depends on IRQ_FORCED_THREADING -+ select PREEMPT_RT_BASE -+ select PREEMPT_RCU -+ help -+ All and everything ++ if (!IS_ENABLED(CONFIG_PREEMPT_RT_FULL)) ++ irq_work_run_list(this_cpu_ptr(&lazy_list)); ++} + - endchoice ++#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) ++void irq_work_tick_soft(void) ++{ + irq_work_run_list(this_cpu_ptr(&lazy_list)); + } ++#endif - config PREEMPT_COUNT -diff -Nur linux-4.9.28.orig/kernel/ksysfs.c linux-4.9.28/kernel/ksysfs.c ---- linux-4.9.28.orig/kernel/ksysfs.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/ksysfs.c 2017-05-19 03:37:25.186176955 +0200 -@@ -136,6 +136,15 @@ + /* + * Synchronize against the irq_work @entry, ensures the entry is not +diff --git a/kernel/ksysfs.c b/kernel/ksysfs.c +index ee1bc1bb8feb..ddef07958840 100644 +--- a/kernel/ksysfs.c ++++ b/kernel/ksysfs.c +@@ -136,6 +136,15 @@ KERNEL_ATTR_RO(vmcoreinfo); #endif /* CONFIG_KEXEC_CORE */ @@ -13100,7 +13423,7 @@ diff -Nur linux-4.9.28.orig/kernel/ksysfs.c linux-4.9.28/kernel/ksysfs.c /* whether file capabilities are enabled */ static ssize_t fscaps_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf) -@@ -225,6 +234,9 @@ +@@ -225,6 +234,9 @@ static struct attribute * kernel_attrs[] = { &rcu_expedited_attr.attr, &rcu_normal_attr.attr, #endif @@ -13110,10 +13433,47 @@ diff -Nur linux-4.9.28.orig/kernel/ksysfs.c linux-4.9.28/kernel/ksysfs.c NULL }; -diff -Nur linux-4.9.28.orig/kernel/locking/lockdep.c linux-4.9.28/kernel/locking/lockdep.c ---- linux-4.9.28.orig/kernel/locking/lockdep.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/locking/lockdep.c 2017-05-19 03:37:25.186176955 +0200 -@@ -658,6 +658,7 @@ +diff --git a/kernel/locking/Makefile b/kernel/locking/Makefile +index 6f88e352cd4f..6ff9e8011dd0 100644 +--- a/kernel/locking/Makefile ++++ b/kernel/locking/Makefile +@@ -2,7 +2,7 @@ + # and is generally not a function of system call inputs. + KCOV_INSTRUMENT := n + +-obj-y += mutex.o semaphore.o rwsem.o percpu-rwsem.o ++obj-y += semaphore.o percpu-rwsem.o + + ifdef CONFIG_FUNCTION_TRACER + CFLAGS_REMOVE_lockdep.o = $(CC_FLAGS_FTRACE) +@@ -11,7 +11,11 @@ CFLAGS_REMOVE_mutex-debug.o = $(CC_FLAGS_FTRACE) + CFLAGS_REMOVE_rtmutex-debug.o = $(CC_FLAGS_FTRACE) + endif + ++ifneq ($(CONFIG_PREEMPT_RT_FULL),y) ++obj-y += mutex.o + obj-$(CONFIG_DEBUG_MUTEXES) += mutex-debug.o ++endif ++obj-y += rwsem.o + obj-$(CONFIG_LOCKDEP) += lockdep.o + ifeq ($(CONFIG_PROC_FS),y) + obj-$(CONFIG_LOCKDEP) += lockdep_proc.o +@@ -24,7 +28,10 @@ obj-$(CONFIG_RT_MUTEXES) += rtmutex.o + obj-$(CONFIG_DEBUG_RT_MUTEXES) += rtmutex-debug.o + obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock.o + obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock_debug.o ++ifneq ($(CONFIG_PREEMPT_RT_FULL),y) + obj-$(CONFIG_RWSEM_GENERIC_SPINLOCK) += rwsem-spinlock.o + obj-$(CONFIG_RWSEM_XCHGADD_ALGORITHM) += rwsem-xadd.o ++endif ++obj-$(CONFIG_PREEMPT_RT_FULL) += rt.o rwsem-rt.o + obj-$(CONFIG_QUEUED_RWLOCKS) += qrwlock.o + obj-$(CONFIG_LOCK_TORTURE_TEST) += locktorture.o +diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c +index 4d7ffc0a0d00..3d157b3128eb 100644 +--- a/kernel/locking/lockdep.c ++++ b/kernel/locking/lockdep.c +@@ -658,6 +658,7 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) struct lockdep_subclass_key *key; struct hlist_head *hash_head; struct lock_class *class; @@ -13121,18 +13481,17 @@ diff -Nur linux-4.9.28.orig/kernel/locking/lockdep.c linux-4.9.28/kernel/locking if (unlikely(subclass >= MAX_LOCKDEP_SUBCLASSES)) { debug_locks_off(); -@@ -671,10 +672,23 @@ +@@ -671,10 +672,23 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) /* * Static locks do not have their class-keys yet - for them the key - * is the lock object itself: -- */ -- if (unlikely(!lock->key)) -- lock->key = (void *)lock; + * is the lock object itself. If the lock is in the per cpu area, + * the canonical address of the lock (per cpu offset removed) is + * used. -+ */ + */ +- if (unlikely(!lock->key)) +- lock->key = (void *)lock; + if (unlikely(!lock->key)) { + unsigned long can_addr, addr = (unsigned long)lock; + @@ -13149,7 +13508,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/lockdep.c linux-4.9.28/kernel/locking /* * NOTE: the class-key must be unique. For dynamic locks, a static -@@ -706,7 +720,7 @@ +@@ -706,7 +720,7 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) } } @@ -13158,7 +13517,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/lockdep.c linux-4.9.28/kernel/locking } /* -@@ -724,19 +738,18 @@ +@@ -724,19 +738,18 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) DEBUG_LOCKS_WARN_ON(!irqs_disabled()); class = look_up_lock_class(lock, subclass); @@ -13181,7 +13540,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/lockdep.c linux-4.9.28/kernel/locking return NULL; } -@@ -3410,7 +3423,7 @@ +@@ -3410,7 +3423,7 @@ static int match_held_lock(struct held_lock *hlock, struct lockdep_map *lock) * Clearly if the lock hasn't been acquired _ever_, we're not * holding it either, so report failure. */ @@ -13190,7 +13549,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/lockdep.c linux-4.9.28/kernel/locking return 0; /* -@@ -3689,6 +3702,7 @@ +@@ -3689,6 +3702,7 @@ static void check_flags(unsigned long flags) } } @@ -13198,7 +13557,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/lockdep.c linux-4.9.28/kernel/locking /* * We dont accurately track softirq state in e.g. * hardirq contexts (such as on 4KSTACKS), so only -@@ -3703,6 +3717,7 @@ +@@ -3703,6 +3717,7 @@ static void check_flags(unsigned long flags) DEBUG_LOCKS_WARN_ON(!current->softirqs_enabled); } } @@ -13206,7 +13565,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/lockdep.c linux-4.9.28/kernel/locking if (!debug_locks) print_irqtrace_events(current); -@@ -4159,7 +4174,7 @@ +@@ -4159,7 +4174,7 @@ void lockdep_reset_lock(struct lockdep_map *lock) * If the class exists we look it up and zap it: */ class = look_up_lock_class(lock, j); @@ -13215,9 +13574,10 @@ diff -Nur linux-4.9.28.orig/kernel/locking/lockdep.c linux-4.9.28/kernel/locking zap_class(class); } /* -diff -Nur linux-4.9.28.orig/kernel/locking/locktorture.c linux-4.9.28/kernel/locking/locktorture.c ---- linux-4.9.28.orig/kernel/locking/locktorture.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/locking/locktorture.c 2017-05-19 03:37:25.186176955 +0200 +diff --git a/kernel/locking/locktorture.c b/kernel/locking/locktorture.c +index f8c5af52a131..788068773e61 100644 +--- a/kernel/locking/locktorture.c ++++ b/kernel/locking/locktorture.c @@ -26,7 +26,6 @@ #include #include @@ -13226,45 +13586,11 @@ diff -Nur linux-4.9.28.orig/kernel/locking/locktorture.c linux-4.9.28/kernel/loc #include #include #include -diff -Nur linux-4.9.28.orig/kernel/locking/Makefile linux-4.9.28/kernel/locking/Makefile ---- linux-4.9.28.orig/kernel/locking/Makefile 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/locking/Makefile 2017-05-19 03:37:25.186176955 +0200 -@@ -2,7 +2,7 @@ - # and is generally not a function of system call inputs. - KCOV_INSTRUMENT := n - --obj-y += mutex.o semaphore.o rwsem.o percpu-rwsem.o -+obj-y += semaphore.o percpu-rwsem.o - - ifdef CONFIG_FUNCTION_TRACER - CFLAGS_REMOVE_lockdep.o = $(CC_FLAGS_FTRACE) -@@ -11,7 +11,11 @@ - CFLAGS_REMOVE_rtmutex-debug.o = $(CC_FLAGS_FTRACE) - endif - -+ifneq ($(CONFIG_PREEMPT_RT_FULL),y) -+obj-y += mutex.o - obj-$(CONFIG_DEBUG_MUTEXES) += mutex-debug.o -+endif -+obj-y += rwsem.o - obj-$(CONFIG_LOCKDEP) += lockdep.o - ifeq ($(CONFIG_PROC_FS),y) - obj-$(CONFIG_LOCKDEP) += lockdep_proc.o -@@ -24,7 +28,10 @@ - obj-$(CONFIG_DEBUG_RT_MUTEXES) += rtmutex-debug.o - obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock.o - obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock_debug.o -+ifneq ($(CONFIG_PREEMPT_RT_FULL),y) - obj-$(CONFIG_RWSEM_GENERIC_SPINLOCK) += rwsem-spinlock.o - obj-$(CONFIG_RWSEM_XCHGADD_ALGORITHM) += rwsem-xadd.o -+endif -+obj-$(CONFIG_PREEMPT_RT_FULL) += rt.o rwsem-rt.o - obj-$(CONFIG_QUEUED_RWLOCKS) += qrwlock.o - obj-$(CONFIG_LOCK_TORTURE_TEST) += locktorture.o -diff -Nur linux-4.9.28.orig/kernel/locking/percpu-rwsem.c linux-4.9.28/kernel/locking/percpu-rwsem.c ---- linux-4.9.28.orig/kernel/locking/percpu-rwsem.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/locking/percpu-rwsem.c 2017-05-19 03:37:25.186176955 +0200 -@@ -18,7 +18,7 @@ +diff --git a/kernel/locking/percpu-rwsem.c b/kernel/locking/percpu-rwsem.c +index ce182599cf2e..2ad3a1e8344c 100644 +--- a/kernel/locking/percpu-rwsem.c ++++ b/kernel/locking/percpu-rwsem.c +@@ -18,7 +18,7 @@ int __percpu_init_rwsem(struct percpu_rw_semaphore *sem, /* ->rw_sem represents the whole percpu_rw_semaphore for lockdep */ rcu_sync_init(&sem->rss, RCU_SCHED_SYNC); __init_rwsem(&sem->rw_sem, name, rwsem_key); @@ -13273,7 +13599,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/percpu-rwsem.c linux-4.9.28/kernel/lo sem->readers_block = 0; return 0; } -@@ -103,7 +103,7 @@ +@@ -103,7 +103,7 @@ void __percpu_up_read(struct percpu_rw_semaphore *sem) __this_cpu_dec(*sem->read_count); /* Prod writer to recheck readers_active */ @@ -13282,7 +13608,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/percpu-rwsem.c linux-4.9.28/kernel/lo } EXPORT_SYMBOL_GPL(__percpu_up_read); -@@ -160,7 +160,7 @@ +@@ -160,7 +160,7 @@ void percpu_down_write(struct percpu_rw_semaphore *sem) */ /* Wait for all now active readers to complete. */ @@ -13291,9 +13617,11 @@ diff -Nur linux-4.9.28.orig/kernel/locking/percpu-rwsem.c linux-4.9.28/kernel/lo } EXPORT_SYMBOL_GPL(percpu_down_write); -diff -Nur linux-4.9.28.orig/kernel/locking/rt.c linux-4.9.28/kernel/locking/rt.c ---- linux-4.9.28.orig/kernel/locking/rt.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/kernel/locking/rt.c 2017-05-19 03:37:25.186176955 +0200 +diff --git a/kernel/locking/rt.c b/kernel/locking/rt.c +new file mode 100644 +index 000000000000..6284e3b15091 +--- /dev/null ++++ b/kernel/locking/rt.c @@ -0,0 +1,331 @@ +/* + * kernel/rt.c @@ -13626,9 +13954,41 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rt.c linux-4.9.28/kernel/locking/rt.c + return 1; +} +EXPORT_SYMBOL(atomic_dec_and_mutex_lock); -diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking/rtmutex.c ---- linux-4.9.28.orig/kernel/locking/rtmutex.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/locking/rtmutex.c 2017-05-19 03:37:25.186176955 +0200 +diff --git a/kernel/locking/rtmutex-debug.c b/kernel/locking/rtmutex-debug.c +index 62b6cee8ea7f..0613c4b1d059 100644 +--- a/kernel/locking/rtmutex-debug.c ++++ b/kernel/locking/rtmutex-debug.c +@@ -173,12 +173,3 @@ void debug_rt_mutex_init(struct rt_mutex *lock, const char *name) + lock->name = name; + } + +-void +-rt_mutex_deadlock_account_lock(struct rt_mutex *lock, struct task_struct *task) +-{ +-} +- +-void rt_mutex_deadlock_account_unlock(struct task_struct *task) +-{ +-} +- +diff --git a/kernel/locking/rtmutex-debug.h b/kernel/locking/rtmutex-debug.h +index d0519c3432b6..b585af9a1b50 100644 +--- a/kernel/locking/rtmutex-debug.h ++++ b/kernel/locking/rtmutex-debug.h +@@ -9,9 +9,6 @@ + * This file contains macros used solely by rtmutex.c. Debug version. + */ + +-extern void +-rt_mutex_deadlock_account_lock(struct rt_mutex *lock, struct task_struct *task); +-extern void rt_mutex_deadlock_account_unlock(struct task_struct *task); + extern void debug_rt_mutex_init_waiter(struct rt_mutex_waiter *waiter); + extern void debug_rt_mutex_free_waiter(struct rt_mutex_waiter *waiter); + extern void debug_rt_mutex_init(struct rt_mutex *lock, const char *name); +diff --git a/kernel/locking/rtmutex.c b/kernel/locking/rtmutex.c +index 2c49d76f96c3..218f1d26afe7 100644 +--- a/kernel/locking/rtmutex.c ++++ b/kernel/locking/rtmutex.c @@ -7,6 +7,11 @@ * Copyright (C) 2005-2006 Timesys Corp., Thomas Gleixner * Copyright (C) 2005 Kihon Technologies Inc., Steven Rostedt @@ -13649,7 +14009,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking #include "rtmutex_common.h" -@@ -133,6 +139,12 @@ +@@ -133,6 +139,12 @@ static void fixup_rt_mutex_waiters(struct rt_mutex *lock) WRITE_ONCE(*p, owner & ~RT_MUTEX_HAS_WAITERS); } @@ -13662,7 +14022,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking /* * We can speed up the acquire/release, if there's no debugging state to be * set up. -@@ -222,12 +234,25 @@ +@@ -222,12 +234,25 @@ static inline bool unlock_rt_mutex_safe(struct rt_mutex *lock, } #endif @@ -13691,7 +14051,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking /* * If both waiters have dl_prio(), we check the deadlines of the -@@ -236,12 +261,30 @@ +@@ -236,12 +261,30 @@ rt_mutex_waiter_less(struct rt_mutex_waiter *left, * then right waiter has a dl_prio() too. */ if (dl_prio(left->prio)) @@ -13724,7 +14084,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking static void rt_mutex_enqueue(struct rt_mutex *lock, struct rt_mutex_waiter *waiter) { -@@ -253,7 +296,7 @@ +@@ -253,7 +296,7 @@ rt_mutex_enqueue(struct rt_mutex *lock, struct rt_mutex_waiter *waiter) while (*link) { parent = *link; entry = rb_entry(parent, struct rt_mutex_waiter, tree_entry); @@ -13733,7 +14093,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking link = &parent->rb_left; } else { link = &parent->rb_right; -@@ -292,7 +335,7 @@ +@@ -292,7 +335,7 @@ rt_mutex_enqueue_pi(struct task_struct *task, struct rt_mutex_waiter *waiter) while (*link) { parent = *link; entry = rb_entry(parent, struct rt_mutex_waiter, pi_tree_entry); @@ -13742,7 +14102,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking link = &parent->rb_left; } else { link = &parent->rb_right; -@@ -320,72 +363,16 @@ +@@ -320,72 +363,16 @@ rt_mutex_dequeue_pi(struct task_struct *task, struct rt_mutex_waiter *waiter) RB_CLEAR_NODE(&waiter->pi_tree_entry); } @@ -13753,19 +14113,24 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking - * the waiter is not allowed to do priority boosting - */ -int rt_mutex_getprio(struct task_struct *task) --{ ++static void rt_mutex_adjust_prio(struct task_struct *p) + { - if (likely(!task_has_pi_waiters(task))) - return task->normal_prio; -- ++ struct task_struct *pi_task = NULL; + - return min(task_top_pi_waiter(task)->prio, - task->normal_prio); -} -- ++ lockdep_assert_held(&p->pi_lock); + -struct task_struct *rt_mutex_get_top_task(struct task_struct *task) -{ - if (likely(!task_has_pi_waiters(task))) - return NULL; -- ++ if (task_has_pi_waiters(p)) ++ pi_task = task_top_pi_waiter(p)->task; + - return task_top_pi_waiter(task)->task; -} - @@ -13789,16 +14154,13 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking - * This can be both boosting and unboosting. task->pi_lock must be held. - */ -static void __rt_mutex_adjust_prio(struct task_struct *task) -+static void rt_mutex_adjust_prio(struct task_struct *p) - { +-{ - int prio = rt_mutex_getprio(task); -+ struct task_struct *pi_task = NULL; - +- - if (task->prio != prio || dl_prio(prio)) - rt_mutex_setprio(task, prio); -} -+ lockdep_assert_held(&p->pi_lock); - +- -/* - * Adjust task priority (undo boosting). Called from the exit path of - * rt_mutex_slowunlock() and rt_mutex_slowlock(). @@ -13811,9 +14173,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking -void rt_mutex_adjust_prio(struct task_struct *task) -{ - unsigned long flags; -+ if (task_has_pi_waiters(p)) -+ pi_task = task_top_pi_waiter(p)->task; - +- - raw_spin_lock_irqsave(&task->pi_lock, flags); - __rt_mutex_adjust_prio(task); - raw_spin_unlock_irqrestore(&task->pi_lock, flags); @@ -13821,7 +14181,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking } /* -@@ -414,6 +401,14 @@ +@@ -414,6 +401,14 @@ static bool rt_mutex_cond_detect_deadlock(struct rt_mutex_waiter *waiter, return debug_rt_mutex_detect_deadlock(waiter, chwalk); } @@ -13836,7 +14196,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking /* * Max number of times we'll walk the boosting chain: */ -@@ -421,7 +416,8 @@ +@@ -421,7 +416,8 @@ int max_lock_depth = 1024; static inline struct rt_mutex *task_blocked_on_lock(struct task_struct *p) { @@ -13846,7 +14206,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking } /* -@@ -557,7 +553,7 @@ +@@ -557,7 +553,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, * reached or the state of the chain has changed while we * dropped the locks. */ @@ -13855,7 +14215,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking goto out_unlock_pi; /* -@@ -608,7 +604,7 @@ +@@ -608,7 +604,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, * enabled we continue, but stop the requeueing in the chain * walk. */ @@ -13864,7 +14224,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking if (!detect_deadlock) goto out_unlock_pi; else -@@ -704,7 +700,26 @@ +@@ -704,7 +700,26 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, /* [7] Requeue the waiter in the lock waiter tree. */ rt_mutex_dequeue(lock, waiter); @@ -13891,7 +14251,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking rt_mutex_enqueue(lock, waiter); /* [8] Release the task */ -@@ -719,13 +734,16 @@ +@@ -719,13 +734,16 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, * follow here. This is the end of the chain we are walking. */ if (!rt_mutex_owner(lock)) { @@ -13910,7 +14270,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking raw_spin_unlock_irq(&lock->wait_lock); return 0; } -@@ -745,7 +763,7 @@ +@@ -745,7 +763,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, */ rt_mutex_dequeue_pi(task, prerequeue_top_waiter); rt_mutex_enqueue_pi(task, waiter); @@ -13919,7 +14279,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking } else if (prerequeue_top_waiter == waiter) { /* -@@ -761,7 +779,7 @@ +@@ -761,7 +779,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, rt_mutex_dequeue_pi(task, waiter); waiter = rt_mutex_top_waiter(lock); rt_mutex_enqueue_pi(task, waiter); @@ -13928,7 +14288,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking } else { /* * Nothing changed. No need to do any priority -@@ -818,6 +836,7 @@ +@@ -818,6 +836,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, return ret; } @@ -13936,7 +14296,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking /* * Try to take an rt-mutex * -@@ -828,9 +847,12 @@ +@@ -828,9 +847,12 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, * @waiter: The waiter that is queued to the lock's wait tree if the * callsite called task_blocked_on_lock(), otherwise NULL */ @@ -13951,7 +14311,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking /* * Before testing whether we can acquire @lock, we set the * RT_MUTEX_HAS_WAITERS bit in @lock->owner. This forces all -@@ -866,8 +888,10 @@ +@@ -866,8 +888,10 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, * If waiter is not the highest priority waiter of * @lock, give up. */ @@ -13963,7 +14323,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking /* * We can acquire the lock. Remove the waiter from the -@@ -885,14 +909,26 @@ +@@ -885,14 +909,26 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, * not need to be dequeued. */ if (rt_mutex_has_waiters(lock)) { @@ -13992,7 +14352,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking /* * The current top waiter stays enqueued. We * don't have to change anything in the lock -@@ -936,180 +972,592 @@ +@@ -936,11 +972,384 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, */ rt_mutex_set_owner(lock, task); @@ -14002,58 +14362,25 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking } +#ifdef CONFIG_PREEMPT_RT_FULL - /* -- * Task blocks on lock. -- * -- * Prepare waiter and propagate pi chain -- * -- * This must be called with lock->wait_lock held and interrupts disabled ++/* + * preemptible spin_lock functions: - */ --static int task_blocks_on_rt_mutex(struct rt_mutex *lock, -- struct rt_mutex_waiter *waiter, -- struct task_struct *task, -- enum rtmutex_chainwalk chwalk) ++ */ +static inline void rt_spin_lock_fastlock(struct rt_mutex *lock, + void (*slowfn)(struct rt_mutex *lock, + bool mg_off), + bool do_mig_dis) - { -- struct task_struct *owner = rt_mutex_owner(lock); -- struct rt_mutex_waiter *top_waiter = waiter; -- struct rt_mutex *next_lock; -- int chain_walk = 0, res; ++{ + might_sleep_no_state_check(); - -- /* -- * Early deadlock detection. We really don't want the task to -- * enqueue on itself just to untangle the mess later. It's not -- * only an optimization. We drop the locks, so another waiter -- * can come in before the chain walk detects the deadlock. So -- * the other will detect the deadlock and return -EDEADLOCK, -- * which is wrong, as the other waiter is not in a deadlock -- * situation. -- */ -- if (owner == task) -- return -EDEADLK; ++ + if (do_mig_dis) + migrate_disable(); - -- raw_spin_lock(&task->pi_lock); -- __rt_mutex_adjust_prio(task); -- waiter->task = task; -- waiter->lock = lock; -- waiter->prio = task->prio; ++ + if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) + return; + else + slowfn(lock, do_mig_dis); +} - -- /* Get the top priority waiter on the lock */ -- if (rt_mutex_has_waiters(lock)) -- top_waiter = rt_mutex_top_waiter(lock); -- rt_mutex_enqueue(lock, waiter); ++ +static inline void rt_spin_lock_fastunlock(struct rt_mutex *lock, + void (*slowfn)(struct rt_mutex *lock)) +{ @@ -14071,8 +14398,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + struct task_struct *owner) +{ + int res = 0; - -- task->pi_blocked_on = waiter; ++ + rcu_read_lock(); + for (;;) { + if (owner != rt_mutex_owner(lock)) @@ -14098,8 +14424,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + return 1; +} +#endif - -- raw_spin_unlock(&task->pi_lock); ++ +static int task_blocks_on_rt_mutex(struct rt_mutex *lock, + struct rt_mutex_waiter *waiter, + struct task_struct *task, @@ -14118,85 +14443,44 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + struct rt_mutex_waiter waiter, *top_waiter; + unsigned long flags; + int ret; - -- if (!owner) -- return 0; ++ + rt_mutex_init_waiter(&waiter, true); - -- raw_spin_lock(&owner->pi_lock); -- if (waiter == rt_mutex_top_waiter(lock)) { -- rt_mutex_dequeue_pi(owner, top_waiter); -- rt_mutex_enqueue_pi(owner, waiter); ++ + raw_spin_lock_irqsave(&lock->wait_lock, flags); - -- __rt_mutex_adjust_prio(owner); -- if (owner->pi_blocked_on) -- chain_walk = 1; -- } else if (rt_mutex_cond_detect_deadlock(waiter, chwalk)) { -- chain_walk = 1; ++ + if (__try_to_take_rt_mutex(lock, self, NULL, STEAL_LATERAL)) { + raw_spin_unlock_irqrestore(&lock->wait_lock, flags); + return; - } - -- /* Store the lock on which owner is blocked or NULL */ -- next_lock = task_blocked_on_lock(owner); ++ } ++ + BUG_ON(rt_mutex_owner(lock) == self); - -- raw_spin_unlock(&owner->pi_lock); - /* -- * Even if full deadlock detection is on, if the owner is not -- * blocked itself, we can avoid finding this out in the chain -- * walk. ++ ++ /* + * We save whatever state the task is in and we'll restore it + * after acquiring the lock taking real wakeups into account + * as well. We are serialized via pi_lock against wakeups. See + * try_to_wake_up(). - */ -- if (!chain_walk || !next_lock) -- return 0; ++ */ + raw_spin_lock(&self->pi_lock); + self->saved_state = self->state; + __set_current_state_no_track(TASK_UNINTERRUPTIBLE); + raw_spin_unlock(&self->pi_lock); - -- /* -- * The owner can't disappear while holding a lock, -- * so the owner struct is protected by wait_lock. -- * Gets dropped in rt_mutex_adjust_prio_chain()! -- */ -- get_task_struct(owner); ++ + ret = task_blocks_on_rt_mutex(lock, &waiter, self, RT_MUTEX_MIN_CHAINWALK); + BUG_ON(ret); - -- raw_spin_unlock_irq(&lock->wait_lock); ++ + for (;;) { + /* Try to acquire the lock again. */ + if (__try_to_take_rt_mutex(lock, self, &waiter, STEAL_LATERAL)) + break; - -- res = rt_mutex_adjust_prio_chain(owner, chwalk, lock, -- next_lock, waiter, task); ++ + top_waiter = rt_mutex_top_waiter(lock); + lock_owner = rt_mutex_owner(lock); - -- raw_spin_lock_irq(&lock->wait_lock); ++ + raw_spin_unlock_irqrestore(&lock->wait_lock, flags); - -- return res; --} ++ + debug_rt_mutex_print_deadlock(&waiter); - --/* -- * Remove the top waiter from the current tasks pi waiter tree and -- * queue it up. -- * -- * Called with lock->wait_lock held and interrupts disabled. -- */ --static void mark_wakeup_next_waiter(struct wake_q_head *wake_q, -- struct rt_mutex *lock) --{ -- struct rt_mutex_waiter *waiter; ++ + if (top_waiter != &waiter || adaptive_wait(lock, lock_owner)) { + if (mg_off) + migrate_enable(); @@ -14204,23 +14488,15 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + if (mg_off) + migrate_disable(); + } - -- raw_spin_lock(¤t->pi_lock); ++ + raw_spin_lock_irqsave(&lock->wait_lock, flags); - -- waiter = rt_mutex_top_waiter(lock); ++ + raw_spin_lock(&self->pi_lock); + __set_current_state_no_track(TASK_UNINTERRUPTIBLE); + raw_spin_unlock(&self->pi_lock); + } - - /* -- * Remove it from current->pi_waiters. We do not adjust a -- * possible priority boost right now. We execute wakeup in the -- * boosted mode and go back to normal after releasing -- * lock->wait_lock. -- */ -- rt_mutex_dequeue_pi(current, waiter); ++ ++ /* + * Restore the task state to current->saved_state. We set it + * to the original state above and the try_to_wake_up() code + * has possibly updated it when a real (non-rtmutex) wakeup @@ -14231,103 +14507,68 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + __set_current_state_no_track(self->saved_state); + self->saved_state = TASK_RUNNING; + raw_spin_unlock(&self->pi_lock); - - /* -- * As we are waking up the top waiter, and the waiter stays -- * queued on the lock until it gets the lock, this lock -- * obviously has waiters. Just set the bit here and this has -- * the added benefit of forcing all new tasks into the -- * slow path making sure no task of lower priority than -- * the top waiter can steal this lock. ++ ++ /* + * try_to_take_rt_mutex() sets the waiter bit + * unconditionally. We might have to fix that up: - */ -- lock->owner = (void *) RT_MUTEX_HAS_WAITERS; ++ */ + fixup_rt_mutex_waiters(lock); - -- raw_spin_unlock(¤t->pi_lock); ++ + BUG_ON(rt_mutex_has_waiters(lock) && &waiter == rt_mutex_top_waiter(lock)); + BUG_ON(!RB_EMPTY_NODE(&waiter.tree_entry)); + + raw_spin_unlock_irqrestore(&lock->wait_lock, flags); - -- wake_q_add(wake_q, waiter->task); ++ + debug_rt_mutex_free_waiter(&waiter); - } - ++} ++ +static bool __sched __rt_mutex_unlock_common(struct rt_mutex *lock, + struct wake_q_head *wake_q, + struct wake_q_head *wq_sleeper); - /* -- * Remove a waiter from a lock and give up -- * -- * Must be called with lock->wait_lock held and interrupts disabled. I must -- * have just failed to try_to_take_rt_mutex(). ++/* + * Slow path to release a rt_mutex spin_lock style - */ --static void remove_waiter(struct rt_mutex *lock, -- struct rt_mutex_waiter *waiter) ++ */ +static void noinline __sched rt_spin_lock_slowunlock(struct rt_mutex *lock) - { -- bool is_top_waiter = (waiter == rt_mutex_top_waiter(lock)); -- struct task_struct *owner = rt_mutex_owner(lock); -- struct rt_mutex *next_lock; ++{ + unsigned long flags; + WAKE_Q(wake_q); + WAKE_Q(wake_sleeper_q); + bool postunlock; - -- raw_spin_lock(¤t->pi_lock); -- rt_mutex_dequeue(lock, waiter); -- current->pi_blocked_on = NULL; -- raw_spin_unlock(¤t->pi_lock); ++ + raw_spin_lock_irqsave(&lock->wait_lock, flags); + postunlock = __rt_mutex_unlock_common(lock, &wake_q, &wake_sleeper_q); + raw_spin_unlock_irqrestore(&lock->wait_lock, flags); - -- /* -- * Only update priority if the waiter was the highest priority -- * waiter of the lock and there is an owner to update. -- */ -- if (!owner || !is_top_waiter) -- return; ++ + if (postunlock) + rt_mutex_postunlock(&wake_q, &wake_sleeper_q); +} - -- raw_spin_lock(&owner->pi_lock); ++ +void __lockfunc rt_spin_lock__no_mg(spinlock_t *lock) +{ + rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock, false); + spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); +} +EXPORT_SYMBOL(rt_spin_lock__no_mg); - -- rt_mutex_dequeue_pi(owner, waiter); ++ +void __lockfunc rt_spin_lock(spinlock_t *lock) +{ + rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock, true); + spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); +} +EXPORT_SYMBOL(rt_spin_lock); - -- if (rt_mutex_has_waiters(lock)) -- rt_mutex_enqueue_pi(owner, rt_mutex_top_waiter(lock)); ++ +void __lockfunc __rt_spin_lock(struct rt_mutex *lock) +{ + rt_spin_lock_fastlock(lock, rt_spin_lock_slowlock, true); +} +EXPORT_SYMBOL(__rt_spin_lock); - -- __rt_mutex_adjust_prio(owner); ++ +void __lockfunc __rt_spin_lock__no_mg(struct rt_mutex *lock) +{ + rt_spin_lock_fastlock(lock, rt_spin_lock_slowlock, false); +} +EXPORT_SYMBOL(__rt_spin_lock__no_mg); - -- /* Store the lock on which owner is blocked or NULL */ -- next_lock = task_blocked_on_lock(owner); ++ +#ifdef CONFIG_DEBUG_LOCK_ALLOC +void __lockfunc rt_spin_lock_nested(spinlock_t *lock, int subclass) +{ @@ -14336,8 +14577,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking +} +EXPORT_SYMBOL(rt_spin_lock_nested); +#endif - -- raw_spin_unlock(&owner->pi_lock); ++ +void __lockfunc rt_spin_unlock__no_mg(spinlock_t *lock) +{ + /* NOTE: we always pass in '1' for nested, for simplicity */ @@ -14345,11 +14585,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + rt_spin_lock_fastunlock(&lock->lock, rt_spin_lock_slowunlock); +} +EXPORT_SYMBOL(rt_spin_unlock__no_mg); - -- /* -- * Don't walk the chain, if the owner task is not blocked -- * itself. -- */ ++ +void __lockfunc rt_spin_unlock(spinlock_t *lock) +{ + /* NOTE: we always pass in '1' for nested, for simplicity */ @@ -14500,38 +14736,23 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + return __try_to_take_rt_mutex(lock, task, waiter, STEAL_NORMAL); +} + -+/* -+ * Task blocks on lock. -+ * -+ * Prepare waiter and propagate pi chain -+ * -+ * This must be called with lock->wait_lock held and interrupts disabled -+ */ -+static int task_blocks_on_rt_mutex(struct rt_mutex *lock, -+ struct rt_mutex_waiter *waiter, -+ struct task_struct *task, -+ enum rtmutex_chainwalk chwalk) -+{ -+ struct task_struct *owner = rt_mutex_owner(lock); -+ struct rt_mutex_waiter *top_waiter = waiter; -+ struct rt_mutex *next_lock; -+ int chain_walk = 0, res; -+ + /* + * Task blocks on lock. + * +@@ -958,6 +1367,8 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, + struct rt_mutex *next_lock; + int chain_walk = 0, res; + + lockdep_assert_held(&lock->wait_lock); + -+ /* -+ * Early deadlock detection. We really don't want the task to -+ * enqueue on itself just to untangle the mess later. It's not -+ * only an optimization. We drop the locks, so another waiter -+ * can come in before the chain walk detects the deadlock. So -+ * the other will detect the deadlock and return -EDEADLOCK, -+ * which is wrong, as the other waiter is not in a deadlock -+ * situation. -+ */ -+ if (owner == task) -+ return -EDEADLK; -+ -+ raw_spin_lock(&task->pi_lock); + /* + * Early deadlock detection. We really don't want the task to + * enqueue on itself just to untangle the mess later. It's not +@@ -971,10 +1382,28 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, + return -EDEADLK; + + raw_spin_lock(&task->pi_lock); +- __rt_mutex_adjust_prio(task); + + /* + * In the case of futex requeue PI, this will be a proxy @@ -14550,100 +14771,55 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + BUG_ON(rt_mutex_real_waiter(task->pi_blocked_on)); + + rt_mutex_adjust_prio(task); -+ waiter->task = task; -+ waiter->lock = lock; -+ waiter->prio = task->prio; + waiter->task = task; + waiter->lock = lock; + waiter->prio = task->prio; + waiter->deadline = task->dl.deadline; -+ -+ /* Get the top priority waiter on the lock */ -+ if (rt_mutex_has_waiters(lock)) -+ top_waiter = rt_mutex_top_waiter(lock); -+ rt_mutex_enqueue(lock, waiter); -+ -+ task->pi_blocked_on = waiter; -+ -+ raw_spin_unlock(&task->pi_lock); -+ -+ if (!owner) -+ return 0; -+ -+ raw_spin_lock(&owner->pi_lock); -+ if (waiter == rt_mutex_top_waiter(lock)) { -+ rt_mutex_dequeue_pi(owner, top_waiter); -+ rt_mutex_enqueue_pi(owner, waiter); -+ + + /* Get the top priority waiter on the lock */ + if (rt_mutex_has_waiters(lock)) +@@ -993,8 +1422,8 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, + rt_mutex_dequeue_pi(owner, top_waiter); + rt_mutex_enqueue_pi(owner, waiter); + +- __rt_mutex_adjust_prio(owner); +- if (owner->pi_blocked_on) + rt_mutex_adjust_prio(owner); + if (rt_mutex_real_waiter(owner->pi_blocked_on)) -+ chain_walk = 1; -+ } else if (rt_mutex_cond_detect_deadlock(waiter, chwalk)) { -+ chain_walk = 1; -+ } -+ -+ /* Store the lock on which owner is blocked or NULL */ -+ next_lock = task_blocked_on_lock(owner); -+ -+ raw_spin_unlock(&owner->pi_lock); -+ /* -+ * Even if full deadlock detection is on, if the owner is not -+ * blocked itself, we can avoid finding this out in the chain -+ * walk. -+ */ -+ if (!chain_walk || !next_lock) -+ return 0; -+ -+ /* -+ * The owner can't disappear while holding a lock, -+ * so the owner struct is protected by wait_lock. -+ * Gets dropped in rt_mutex_adjust_prio_chain()! -+ */ -+ get_task_struct(owner); -+ -+ raw_spin_unlock_irq(&lock->wait_lock); -+ -+ res = rt_mutex_adjust_prio_chain(owner, chwalk, lock, -+ next_lock, waiter, task); -+ -+ raw_spin_lock_irq(&lock->wait_lock); -+ -+ return res; -+} -+ -+/* -+ * Remove the top waiter from the current tasks pi waiter tree and -+ * queue it up. -+ * -+ * Called with lock->wait_lock held and interrupts disabled. -+ */ -+static void mark_wakeup_next_waiter(struct wake_q_head *wake_q, + chain_walk = 1; + } else if (rt_mutex_cond_detect_deadlock(waiter, chwalk)) { + chain_walk = 1; +@@ -1036,6 +1465,7 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, + * Called with lock->wait_lock held and interrupts disabled. + */ + static void mark_wakeup_next_waiter(struct wake_q_head *wake_q, + struct wake_q_head *wake_sleeper_q, -+ struct rt_mutex *lock) -+{ -+ struct rt_mutex_waiter *waiter; -+ -+ raw_spin_lock(¤t->pi_lock); -+ -+ waiter = rt_mutex_top_waiter(lock); -+ -+ /* + struct rt_mutex *lock) + { + struct rt_mutex_waiter *waiter; +@@ -1045,12 +1475,14 @@ static void mark_wakeup_next_waiter(struct wake_q_head *wake_q, + waiter = rt_mutex_top_waiter(lock); + + /* +- * Remove it from current->pi_waiters. We do not adjust a +- * possible priority boost right now. We execute wakeup in the +- * boosted mode and go back to normal after releasing +- * lock->wait_lock. + * Remove it from current->pi_waiters and deboost. + * + * We must in fact deboost here in order to ensure we call + * rt_mutex_setprio() to update p->pi_top_task before the + * task unblocks. -+ */ -+ rt_mutex_dequeue_pi(current, waiter); + */ + rt_mutex_dequeue_pi(current, waiter); + rt_mutex_adjust_prio(current); -+ -+ /* -+ * As we are waking up the top waiter, and the waiter stays -+ * queued on the lock until it gets the lock, this lock -+ * obviously has waiters. Just set the bit here and this has -+ * the added benefit of forcing all new tasks into the -+ * slow path making sure no task of lower priority than -+ * the top waiter can steal this lock. -+ */ -+ lock->owner = (void *) RT_MUTEX_HAS_WAITERS; -+ + + /* + * As we are waking up the top waiter, and the waiter stays +@@ -1062,9 +1494,22 @@ static void mark_wakeup_next_waiter(struct wake_q_head *wake_q, + */ + lock->owner = (void *) RT_MUTEX_HAS_WAITERS; + + /* + * We deboosted before waking the top waiter task such that we don't + * run two tasks with the 'same' priority (and ensure the @@ -14659,59 +14835,38 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + wake_q_add(wake_sleeper_q, waiter->task); + else + wake_q_add(wake_q, waiter->task); -+ raw_spin_unlock(¤t->pi_lock); -+} -+ -+/* -+ * Remove a waiter from a lock and give up -+ * -+ * Must be called with lock->wait_lock held and interrupts disabled. I must -+ * have just failed to try_to_take_rt_mutex(). -+ */ -+static void remove_waiter(struct rt_mutex *lock, -+ struct rt_mutex_waiter *waiter) -+{ -+ bool is_top_waiter = (waiter == rt_mutex_top_waiter(lock)); -+ struct task_struct *owner = rt_mutex_owner(lock); + raw_spin_unlock(¤t->pi_lock); +- +- wake_q_add(wake_q, waiter->task); + } + + /* +@@ -1078,7 +1523,9 @@ static void remove_waiter(struct rt_mutex *lock, + { + bool is_top_waiter = (waiter == rt_mutex_top_waiter(lock)); + struct task_struct *owner = rt_mutex_owner(lock); +- struct rt_mutex *next_lock; + struct rt_mutex *next_lock = NULL; + + lockdep_assert_held(&lock->wait_lock); -+ -+ raw_spin_lock(¤t->pi_lock); -+ rt_mutex_dequeue(lock, waiter); -+ current->pi_blocked_on = NULL; -+ raw_spin_unlock(¤t->pi_lock); -+ -+ /* -+ * Only update priority if the waiter was the highest priority -+ * waiter of the lock and there is an owner to update. -+ */ -+ if (!owner || !is_top_waiter) -+ return; -+ -+ raw_spin_lock(&owner->pi_lock); -+ -+ rt_mutex_dequeue_pi(owner, waiter); -+ -+ if (rt_mutex_has_waiters(lock)) -+ rt_mutex_enqueue_pi(owner, rt_mutex_top_waiter(lock)); -+ + + raw_spin_lock(¤t->pi_lock); + rt_mutex_dequeue(lock, waiter); +@@ -1099,10 +1546,11 @@ static void remove_waiter(struct rt_mutex *lock, + if (rt_mutex_has_waiters(lock)) + rt_mutex_enqueue_pi(owner, rt_mutex_top_waiter(lock)); + +- __rt_mutex_adjust_prio(owner); + rt_mutex_adjust_prio(owner); -+ -+ /* Store the lock on which owner is blocked or NULL */ + + /* Store the lock on which owner is blocked or NULL */ +- next_lock = task_blocked_on_lock(owner); + if (rt_mutex_real_waiter(owner->pi_blocked_on)) + next_lock = task_blocked_on_lock(owner); -+ -+ raw_spin_unlock(&owner->pi_lock); -+ -+ /* -+ * Don't walk the chain, if the owner task is not blocked -+ * itself. -+ */ - if (!next_lock) - return; -@@ -1138,21 +1586,30 @@ + raw_spin_unlock(&owner->pi_lock); + +@@ -1138,21 +1586,30 @@ void rt_mutex_adjust_pi(struct task_struct *task) raw_spin_lock_irqsave(&task->pi_lock, flags); waiter = task->pi_blocked_on; @@ -14745,7 +14900,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking /** * __rt_mutex_slowlock() - Perform the wait-wake-try-to-take loop * @lock: the rt_mutex to take -@@ -1166,7 +1623,8 @@ +@@ -1166,7 +1623,8 @@ void rt_mutex_adjust_pi(struct task_struct *task) static int __sched __rt_mutex_slowlock(struct rt_mutex *lock, int state, struct hrtimer_sleeper *timeout, @@ -14755,7 +14910,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking { int ret = 0; -@@ -1175,16 +1633,17 @@ +@@ -1175,16 +1633,17 @@ __rt_mutex_slowlock(struct rt_mutex *lock, int state, if (try_to_take_rt_mutex(lock, current, waiter)) break; @@ -14783,36 +14938,10 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking if (ret) break; } -@@ -1199,59 +1658,118 @@ - set_current_state(state); +@@ -1223,21 +1682,148 @@ static void rt_mutex_handle_deadlock(int res, int detect_deadlock, } + } -- __set_current_state(TASK_RUNNING); -- return ret; -+ __set_current_state(TASK_RUNNING); -+ return ret; -+} -+ -+static void rt_mutex_handle_deadlock(int res, int detect_deadlock, -+ struct rt_mutex_waiter *w) -+{ -+ /* -+ * If the result is not -EDEADLOCK or the caller requested -+ * deadlock detection, nothing to do here. -+ */ -+ if (res != -EDEADLOCK || detect_deadlock) -+ return; -+ -+ /* -+ * Yell lowdly and stop the task right here. -+ */ -+ rt_mutex_print_deadlock(w); -+ while (1) { -+ set_current_state(TASK_INTERRUPTIBLE); -+ schedule(); -+ } -+} -+ +static __always_inline void ww_mutex_lock_acquired(struct ww_mutex *ww, + struct ww_acquire_ctx *ww_ctx) +{ @@ -14851,78 +14980,44 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + DEBUG_LOCKS_WARN_ON(ww_ctx->ww_class != ww->ww_class); +#endif + ww_ctx->acquired++; - } - --static void rt_mutex_handle_deadlock(int res, int detect_deadlock, -- struct rt_mutex_waiter *w) ++} ++ +#ifdef CONFIG_PREEMPT_RT_FULL +static void ww_mutex_account_lock(struct rt_mutex *lock, + struct ww_acquire_ctx *ww_ctx) - { ++{ + struct ww_mutex *ww = container_of(lock, struct ww_mutex, base.lock); + struct rt_mutex_waiter *waiter, *n; + - /* -- * If the result is not -EDEADLOCK or the caller requested -- * deadlock detection, nothing to do here. ++ /* + * This branch gets optimized out for the common case, + * and is only important for ww_mutex_lock. - */ -- if (res != -EDEADLOCK || detect_deadlock) -- return; ++ */ + ww_mutex_lock_acquired(ww, ww_ctx); + ww->ctx = ww_ctx; - - /* -- * Yell lowdly and stop the task right here. ++ ++ /* + * Give any possible sleeping processes the chance to wake up, + * so they can recheck if they have to back off. - */ -- rt_mutex_print_deadlock(w); -- while (1) { -- set_current_state(TASK_INTERRUPTIBLE); -- schedule(); ++ */ + rbtree_postorder_for_each_entry_safe(waiter, n, &lock->waiters, + tree_entry) { + /* XXX debug rt mutex waiter wakeup */ + + BUG_ON(waiter->lock != lock); + rt_mutex_wake_waiter(waiter); - } - } - --/* -- * Slow path lock function: -- */ --static int __sched --rt_mutex_slowlock(struct rt_mutex *lock, int state, -- struct hrtimer_sleeper *timeout, -- enum rtmutex_chainwalk chwalk) --{ -- struct rt_mutex_waiter waiter; -- unsigned long flags; -- int ret = 0; ++ } ++} ++ +#else - -- debug_rt_mutex_init_waiter(&waiter); -- RB_CLEAR_NODE(&waiter.pi_tree_entry); -- RB_CLEAR_NODE(&waiter.tree_entry); ++ +static void ww_mutex_account_lock(struct rt_mutex *lock, + struct ww_acquire_ctx *ww_ctx) +{ + BUG(); +} +#endif - -- /* -- * Technically we could use raw_spin_[un]lock_irq() here, but this can -- * be called in early boot if the cmpxchg() fast path is disabled -- * (debug, no architecture support). In this case we will acquire the -- * rtmutex with lock->wait_lock held. But we cannot unconditionally -- * enable interrupts in that early boot case. So we need to use the -- * irqsave/restore variants. -- */ -- raw_spin_lock_irqsave(&lock->wait_lock, flags); ++ +int __sched rt_mutex_slowlock_locked(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, + enum rtmutex_chainwalk chwalk, @@ -14930,26 +15025,24 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + struct rt_mutex_waiter *waiter) +{ + int ret; - - /* Try to acquire the lock again: */ - if (try_to_take_rt_mutex(lock, current, NULL)) { -- raw_spin_unlock_irqrestore(&lock->wait_lock, flags); ++ ++ /* Try to acquire the lock again: */ ++ if (try_to_take_rt_mutex(lock, current, NULL)) { + if (ww_ctx) + ww_mutex_account_lock(lock, ww_ctx); - return 0; - } - -@@ -1261,17 +1779,27 @@ - if (unlikely(timeout)) - hrtimer_start_expires(&timeout->timer, HRTIMER_MODE_ABS); - -- ret = task_blocks_on_rt_mutex(lock, &waiter, current, chwalk); ++ return 0; ++ } ++ ++ set_current_state(state); ++ ++ /* Setup the timer, when timeout != NULL */ ++ if (unlikely(timeout)) ++ hrtimer_start_expires(&timeout->timer, HRTIMER_MODE_ABS); ++ + ret = task_blocks_on_rt_mutex(lock, waiter, current, chwalk); - -- if (likely(!ret)) ++ + if (likely(!ret)) { - /* sleep on the mutex */ -- ret = __rt_mutex_slowlock(lock, state, timeout, &waiter); ++ /* sleep on the mutex */ + ret = __rt_mutex_slowlock(lock, state, timeout, waiter, + ww_ctx); + } else if (ww_ctx) { @@ -14957,59 +15050,87 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + ret = __mutex_lock_check_stamp(lock, ww_ctx); + BUG_ON(!ret); + } - - if (unlikely(ret)) { - __set_current_state(TASK_RUNNING); - if (rt_mutex_has_waiters(lock)) -- remove_waiter(lock, &waiter); -- rt_mutex_handle_deadlock(ret, chwalk, &waiter); ++ ++ if (unlikely(ret)) { ++ __set_current_state(TASK_RUNNING); ++ if (rt_mutex_has_waiters(lock)) + remove_waiter(lock, waiter); + /* ww_mutex want to report EDEADLK/EALREADY, let them */ + if (!ww_ctx) + rt_mutex_handle_deadlock(ret, chwalk, waiter); + } else if (ww_ctx) { + ww_mutex_account_lock(lock, ww_ctx); - } - - /* -@@ -1279,6 +1807,36 @@ - * unconditionally. We might have to fix that up. - */ - fixup_rt_mutex_waiters(lock); ++ } ++ ++ /* ++ * try_to_take_rt_mutex() sets the waiter bit ++ * unconditionally. We might have to fix that up. ++ */ ++ fixup_rt_mutex_waiters(lock); + return ret; +} + -+/* -+ * Slow path lock function: -+ */ -+static int __sched -+rt_mutex_slowlock(struct rt_mutex *lock, int state, -+ struct hrtimer_sleeper *timeout, + /* + * Slow path lock function: + */ + static int __sched + rt_mutex_slowlock(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, +- enum rtmutex_chainwalk chwalk) + enum rtmutex_chainwalk chwalk, + struct ww_acquire_ctx *ww_ctx) -+{ -+ struct rt_mutex_waiter waiter; -+ unsigned long flags; -+ int ret = 0; -+ -+ rt_mutex_init_waiter(&waiter, false); -+ -+ /* -+ * Technically we could use raw_spin_[un]lock_irq() here, but this can -+ * be called in early boot if the cmpxchg() fast path is disabled -+ * (debug, no architecture support). In this case we will acquire the -+ * rtmutex with lock->wait_lock held. But we cannot unconditionally -+ * enable interrupts in that early boot case. So we need to use the -+ * irqsave/restore variants. -+ */ -+ raw_spin_lock_irqsave(&lock->wait_lock, flags); -+ + { + struct rt_mutex_waiter waiter; + unsigned long flags; + int ret = 0; + +- debug_rt_mutex_init_waiter(&waiter); +- RB_CLEAR_NODE(&waiter.pi_tree_entry); +- RB_CLEAR_NODE(&waiter.tree_entry); ++ rt_mutex_init_waiter(&waiter, false); + + /* + * Technically we could use raw_spin_[un]lock_irq() here, but this can +@@ -1249,36 +1835,8 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state, + */ + raw_spin_lock_irqsave(&lock->wait_lock, flags); + +- /* Try to acquire the lock again: */ +- if (try_to_take_rt_mutex(lock, current, NULL)) { +- raw_spin_unlock_irqrestore(&lock->wait_lock, flags); +- return 0; +- } +- +- set_current_state(state); +- +- /* Setup the timer, when timeout != NULL */ +- if (unlikely(timeout)) +- hrtimer_start_expires(&timeout->timer, HRTIMER_MODE_ABS); +- +- ret = task_blocks_on_rt_mutex(lock, &waiter, current, chwalk); +- +- if (likely(!ret)) +- /* sleep on the mutex */ +- ret = __rt_mutex_slowlock(lock, state, timeout, &waiter); +- +- if (unlikely(ret)) { +- __set_current_state(TASK_RUNNING); +- if (rt_mutex_has_waiters(lock)) +- remove_waiter(lock, &waiter); +- rt_mutex_handle_deadlock(ret, chwalk, &waiter); +- } +- +- /* +- * try_to_take_rt_mutex() sets the waiter bit +- * unconditionally. We might have to fix that up. +- */ +- fixup_rt_mutex_waiters(lock); + ret = rt_mutex_slowlock_locked(lock, state, timeout, chwalk, ww_ctx, + &waiter); raw_spin_unlock_irqrestore(&lock->wait_lock, flags); -@@ -1328,10 +1886,12 @@ +@@ -1328,10 +1886,12 @@ static inline int rt_mutex_slowtrylock(struct rt_mutex *lock) /* * Slow path to release a rt-mutex. @@ -15024,7 +15145,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking { unsigned long flags; -@@ -1340,8 +1900,6 @@ +@@ -1340,8 +1900,6 @@ static bool __sched rt_mutex_slowunlock(struct rt_mutex *lock, debug_rt_mutex_unlock(lock); @@ -15033,7 +15154,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking /* * We must be careful here if the fast path is enabled. If we * have no waiters queued we cannot set owner to NULL here -@@ -1387,12 +1945,10 @@ +@@ -1387,12 +1945,10 @@ static bool __sched rt_mutex_slowunlock(struct rt_mutex *lock, * * Queue the next waiter for wakeup once we release the wait_lock. */ @@ -15048,7 +15169,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking } /* -@@ -1403,63 +1959,85 @@ +@@ -1403,63 +1959,85 @@ static bool __sched rt_mutex_slowunlock(struct rt_mutex *lock, */ static inline int rt_mutex_fastlock(struct rt_mutex *lock, int state, @@ -15157,7 +15278,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking } /** -@@ -1469,15 +2047,13 @@ +@@ -1469,15 +2047,13 @@ rt_mutex_fastunlock(struct rt_mutex *lock, */ void __sched rt_mutex_lock(struct rt_mutex *lock) { @@ -15175,7 +15296,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking * @lock: the rt_mutex to be locked * * Returns: -@@ -1486,23 +2062,32 @@ +@@ -1486,23 +2062,32 @@ EXPORT_SYMBOL_GPL(rt_mutex_lock); */ int __sched rt_mutex_lock_interruptible(struct rt_mutex *lock) { @@ -15219,7 +15340,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking } /** -@@ -1525,6 +2110,7 @@ +@@ -1525,6 +2110,7 @@ rt_mutex_timed_lock(struct rt_mutex *lock, struct hrtimer_sleeper *timeout) return rt_mutex_timed_fastlock(lock, TASK_INTERRUPTIBLE, timeout, RT_MUTEX_MIN_CHAINWALK, @@ -15227,7 +15348,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking rt_mutex_slowlock); } EXPORT_SYMBOL_GPL(rt_mutex_timed_lock); -@@ -1542,7 +2128,11 @@ +@@ -1542,7 +2128,11 @@ EXPORT_SYMBOL_GPL(rt_mutex_timed_lock); */ int __sched rt_mutex_trylock(struct rt_mutex *lock) { @@ -15239,14 +15360,26 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking return 0; return rt_mutex_fasttrylock(lock, rt_mutex_slowtrylock); -@@ -1560,21 +2150,53 @@ +@@ -1560,21 +2150,53 @@ void __sched rt_mutex_unlock(struct rt_mutex *lock) } EXPORT_SYMBOL_GPL(rt_mutex_unlock); +-/** +- * rt_mutex_futex_unlock - Futex variant of rt_mutex_unlock +- * @lock: the rt_mutex to be unlocked +- * +- * Returns: true/false indicating whether priority adjustment is +- * required or not. +- */ +-bool __sched rt_mutex_futex_unlock(struct rt_mutex *lock, +- struct wake_q_head *wqh) +static bool __sched __rt_mutex_unlock_common(struct rt_mutex *lock, + struct wake_q_head *wake_q, + struct wake_q_head *wq_sleeper) -+{ + { +- if (likely(rt_mutex_cmpxchg_release(lock, current, NULL))) { +- rt_mutex_deadlock_account_unlock(current); +- return false; + lockdep_assert_held(&lock->wait_lock); + + debug_rt_mutex_unlock(lock); @@ -15254,7 +15387,8 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + if (!rt_mutex_has_waiters(lock)) { + lock->owner = NULL; + return false; /* done */ -+ } + } +- return rt_mutex_slowunlock(lock, wqh); + + /* + * We've already deboosted, mark_wakeup_next_waiter() will @@ -15267,26 +15401,14 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + return true; /* call postunlock() */ +} + - /** -- * rt_mutex_futex_unlock - Futex variant of rt_mutex_unlock -- * @lock: the rt_mutex to be unlocked -- * -- * Returns: true/false indicating whether priority adjustment is -- * required or not. ++/** + * Futex variant, that since futex variants do not use the fast-path, can be + * simple and will not need to retry. - */ --bool __sched rt_mutex_futex_unlock(struct rt_mutex *lock, -- struct wake_q_head *wqh) ++ */ +bool __sched __rt_mutex_futex_unlock(struct rt_mutex *lock, + struct wake_q_head *wake_q, + struct wake_q_head *wq_sleeper) - { -- if (likely(rt_mutex_cmpxchg_release(lock, current, NULL))) { -- rt_mutex_deadlock_account_unlock(current); -- return false; -- } -- return rt_mutex_slowunlock(lock, wqh); ++{ + return __rt_mutex_unlock_common(lock, wake_q, wq_sleeper); +} + @@ -15305,7 +15427,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking } /** -@@ -1607,13 +2229,12 @@ +@@ -1607,13 +2229,12 @@ EXPORT_SYMBOL_GPL(rt_mutex_destroy); void __rt_mutex_init(struct rt_mutex *lock, const char *name) { lock->owner = NULL; @@ -15320,7 +15442,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking /** * rt_mutex_init_proxy_locked - initialize and lock a rt_mutex on behalf of a -@@ -1628,10 +2249,9 @@ +@@ -1628,10 +2249,9 @@ EXPORT_SYMBOL_GPL(__rt_mutex_init); void rt_mutex_init_proxy_locked(struct rt_mutex *lock, struct task_struct *proxy_owner) { @@ -15332,38 +15454,22 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking } /** -@@ -1647,34 +2267,45 @@ +@@ -1647,7 +2267,66 @@ void rt_mutex_proxy_unlock(struct rt_mutex *lock, { debug_rt_mutex_proxy_unlock(lock); rt_mutex_set_owner(lock, NULL); - rt_mutex_deadlock_account_unlock(proxy_owner); - } - --/** -- * rt_mutex_start_proxy_lock() - Start lock acquisition for another task -- * @lock: the rt_mutex to take -- * @waiter: the pre-initialized rt_mutex_waiter -- * @task: the task to prepare -- * -- * Returns: -- * 0 - task blocked on lock -- * 1 - acquired the lock for task, caller should wake it up -- * <0 - error -- * -- * Special API call for FUTEX_REQUEUE_PI support. -- */ --int rt_mutex_start_proxy_lock(struct rt_mutex *lock, ++} ++ +int __rt_mutex_start_proxy_lock(struct rt_mutex *lock, - struct rt_mutex_waiter *waiter, - struct task_struct *task) - { - int ret; - -- raw_spin_lock_irq(&lock->wait_lock); ++ struct rt_mutex_waiter *waiter, ++ struct task_struct *task) ++{ ++ int ret; ++ + if (try_to_take_rt_mutex(lock, task, NULL)) + return 1; - -- if (try_to_take_rt_mutex(lock, task, NULL)) { ++ +#ifdef CONFIG_PREEMPT_RT_FULL + /* + * In PREEMPT_RT there's an added race. @@ -15386,62 +15492,72 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + raw_spin_lock(&task->pi_lock); + if (task->pi_blocked_on) { + raw_spin_unlock(&task->pi_lock); - raw_spin_unlock_irq(&lock->wait_lock); -- return 1; ++ raw_spin_unlock_irq(&lock->wait_lock); + return -EAGAIN; - } ++ } + task->pi_blocked_on = PI_REQUEUE_INPROGRESS; + raw_spin_unlock(&task->pi_lock); +#endif ++ ++ /* We enforce deadlock detection for futexes */ ++ ret = task_blocks_on_rt_mutex(lock, waiter, task, ++ RT_MUTEX_FULL_CHAINWALK); ++ ++ if (ret && !rt_mutex_owner(lock)) { ++ /* ++ * Reset the return value. We might have ++ * returned with -EDEADLK and the owner ++ * released the lock while we were walking the ++ * pi chain. Let the waiter sort it out. ++ */ ++ ret = 0; ++ } ++ ++ if (ret && rt_mutex_has_waiters(lock)) ++ remove_waiter(lock, waiter); ++ ++ debug_rt_mutex_print_deadlock(waiter); ++ ++ return ret; + } - /* We enforce deadlock detection for futexes */ - ret = task_blocks_on_rt_mutex(lock, waiter, task, -@@ -1690,17 +2321,41 @@ - ret = 0; - } + /** +@@ -1670,33 +2349,9 @@ int rt_mutex_start_proxy_lock(struct rt_mutex *lock, + int ret; + raw_spin_lock_irq(&lock->wait_lock); +- +- if (try_to_take_rt_mutex(lock, task, NULL)) { +- raw_spin_unlock_irq(&lock->wait_lock); +- return 1; +- } +- +- /* We enforce deadlock detection for futexes */ +- ret = task_blocks_on_rt_mutex(lock, waiter, task, +- RT_MUTEX_FULL_CHAINWALK); +- +- if (ret && !rt_mutex_owner(lock)) { +- /* +- * Reset the return value. We might have +- * returned with -EDEADLK and the owner +- * released the lock while we were walking the +- * pi chain. Let the waiter sort it out. +- */ +- ret = 0; +- } +- - if (unlikely(ret)) -+ if (ret && rt_mutex_has_waiters(lock)) - remove_waiter(lock, waiter); - -- raw_spin_unlock_irq(&lock->wait_lock); +- remove_waiter(lock, waiter); - - debug_rt_mutex_print_deadlock(waiter); ++ ret = __rt_mutex_start_proxy_lock(lock, waiter, task); + raw_spin_unlock_irq(&lock->wait_lock); +- debug_rt_mutex_print_deadlock(waiter); +- return ret; } - /** -+ * rt_mutex_start_proxy_lock() - Start lock acquisition for another task -+ * @lock: the rt_mutex to take -+ * @waiter: the pre-initialized rt_mutex_waiter -+ * @task: the task to prepare -+ * -+ * Returns: -+ * 0 - task blocked on lock -+ * 1 - acquired the lock for task, caller should wake it up -+ * <0 - error -+ * -+ * Special API call for FUTEX_REQUEUE_PI support. -+ */ -+int rt_mutex_start_proxy_lock(struct rt_mutex *lock, -+ struct rt_mutex_waiter *waiter, -+ struct task_struct *task) -+{ -+ int ret; -+ -+ raw_spin_lock_irq(&lock->wait_lock); -+ ret = __rt_mutex_start_proxy_lock(lock, waiter, task); -+ raw_spin_unlock_irq(&lock->wait_lock); -+ -+ return ret; -+} -+ -+/** - * rt_mutex_next_owner - return the next owner of the lock - * - * @lock: the rt lock query -@@ -1721,24 +2376,27 @@ +@@ -1721,36 +2376,106 @@ struct task_struct *rt_mutex_next_owner(struct rt_mutex *lock) } /** @@ -15473,13 +15589,20 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking int ret; raw_spin_lock_irq(&lock->wait_lock); -@@ -1746,10 +2404,65 @@ - set_current_state(TASK_INTERRUPTIBLE); - +- +- set_current_state(TASK_INTERRUPTIBLE); +- /* sleep on the mutex */ - ret = __rt_mutex_slowlock(lock, TASK_INTERRUPTIBLE, to, waiter); ++ set_current_state(TASK_INTERRUPTIBLE); + ret = __rt_mutex_slowlock(lock, TASK_INTERRUPTIBLE, to, waiter, NULL); -+ ++ /* ++ * try_to_take_rt_mutex() sets the waiter bit unconditionally. We might ++ * have to fix that up. ++ */ ++ fixup_rt_mutex_waiters(lock); + +- if (unlikely(ret)) + /* + * RT has a problem here when the wait got interrupted by a timeout + * or a signal. task->pi_blocked_on is still set. The task must @@ -15497,7 +15620,6 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + tsk->pi_blocked_on = NULL; + raw_spin_unlock(&tsk->pi_lock); + } -+ + raw_spin_unlock_irq(&lock->wait_lock); + + return ret; @@ -15526,22 +15648,33 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + struct rt_mutex_waiter *waiter) +{ + bool cleanup = false; - -- if (unlikely(ret)) ++ + raw_spin_lock_irq(&lock->wait_lock); + /* ++ * Do an unconditional try-lock, this deals with the lock stealing ++ * state where __rt_mutex_futex_unlock() -> mark_wakeup_next_waiter() ++ * sets a NULL owner. ++ * ++ * We're not interested in the return value, because the subsequent ++ * test on rt_mutex_owner() will infer that. If the trylock succeeded, ++ * we will own the lock and it will have removed the waiter. If we ++ * failed the trylock, we're still not owner and we need to remove ++ * ourselves. ++ */ ++ try_to_take_rt_mutex(lock, current, waiter); ++ /* + * Unless we're the owner; we're still enqueued on the wait_list. + * So check if we became owner, if not, take us off the wait_list. + */ + if (rt_mutex_owner(lock) != current) { remove_waiter(lock, waiter); -+ fixup_rt_mutex_waiters(lock); +- + cleanup = true; + } - /* * try_to_take_rt_mutex() sets the waiter bit unconditionally. We might -@@ -1759,5 +2472,91 @@ + * have to fix that up. +@@ -1759,5 +2484,91 @@ int rt_mutex_finish_proxy_lock(struct rt_mutex *lock, raw_spin_unlock_irq(&lock->wait_lock); @@ -15589,8 +15722,8 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + else if (!ret && ww_ctx->acquired > 1) + return ww_mutex_deadlock_injection(lock, ww_ctx); + -+ return ret; -+} + return ret; + } +EXPORT_SYMBOL_GPL(__ww_mutex_lock_interruptible); + +int __sched @@ -15607,8 +15740,8 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking + else if (!ret && ww_ctx->acquired > 1) + return ww_mutex_deadlock_injection(lock, ww_ctx); + - return ret; - } ++ return ret; ++} +EXPORT_SYMBOL_GPL(__ww_mutex_lock); + +void __sched ww_mutex_unlock(struct ww_mutex *lock) @@ -15633,10 +15766,24 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.c linux-4.9.28/kernel/locking +} +EXPORT_SYMBOL(ww_mutex_unlock); +#endif -diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex_common.h linux-4.9.28/kernel/locking/rtmutex_common.h ---- linux-4.9.28.orig/kernel/locking/rtmutex_common.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/locking/rtmutex_common.h 2017-05-19 03:37:25.186176955 +0200 -@@ -27,12 +27,14 @@ +diff --git a/kernel/locking/rtmutex.h b/kernel/locking/rtmutex.h +index c4060584c407..6607802efa8b 100644 +--- a/kernel/locking/rtmutex.h ++++ b/kernel/locking/rtmutex.h +@@ -11,8 +11,6 @@ + */ + + #define rt_mutex_deadlock_check(l) (0) +-#define rt_mutex_deadlock_account_lock(m, t) do { } while (0) +-#define rt_mutex_deadlock_account_unlock(l) do { } while (0) + #define debug_rt_mutex_init_waiter(w) do { } while (0) + #define debug_rt_mutex_free_waiter(w) do { } while (0) + #define debug_rt_mutex_lock(l) do { } while (0) +diff --git a/kernel/locking/rtmutex_common.h b/kernel/locking/rtmutex_common.h +index e317e1cbb3eb..64d89d780059 100644 +--- a/kernel/locking/rtmutex_common.h ++++ b/kernel/locking/rtmutex_common.h +@@ -27,12 +27,14 @@ struct rt_mutex_waiter { struct rb_node pi_tree_entry; struct task_struct *task; struct rt_mutex *lock; @@ -15651,7 +15798,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex_common.h linux-4.9.28/kernel/ }; /* -@@ -98,21 +100,45 @@ +@@ -98,21 +100,45 @@ enum rtmutex_chainwalk { /* * PI-futex support (proxy locking functions, etc.): */ @@ -15704,50 +15851,11 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex_common.h linux-4.9.28/kernel/ #ifdef CONFIG_DEBUG_RT_MUTEXES # include "rtmutex-debug.h" -diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex-debug.c linux-4.9.28/kernel/locking/rtmutex-debug.c ---- linux-4.9.28.orig/kernel/locking/rtmutex-debug.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/locking/rtmutex-debug.c 2017-05-19 03:37:25.186176955 +0200 -@@ -173,12 +173,3 @@ - lock->name = name; - } - --void --rt_mutex_deadlock_account_lock(struct rt_mutex *lock, struct task_struct *task) --{ --} -- --void rt_mutex_deadlock_account_unlock(struct task_struct *task) --{ --} -- -diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex-debug.h linux-4.9.28/kernel/locking/rtmutex-debug.h ---- linux-4.9.28.orig/kernel/locking/rtmutex-debug.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/locking/rtmutex-debug.h 2017-05-19 03:37:25.186176955 +0200 -@@ -9,9 +9,6 @@ - * This file contains macros used solely by rtmutex.c. Debug version. - */ - --extern void --rt_mutex_deadlock_account_lock(struct rt_mutex *lock, struct task_struct *task); --extern void rt_mutex_deadlock_account_unlock(struct task_struct *task); - extern void debug_rt_mutex_init_waiter(struct rt_mutex_waiter *waiter); - extern void debug_rt_mutex_free_waiter(struct rt_mutex_waiter *waiter); - extern void debug_rt_mutex_init(struct rt_mutex *lock, const char *name); -diff -Nur linux-4.9.28.orig/kernel/locking/rtmutex.h linux-4.9.28/kernel/locking/rtmutex.h ---- linux-4.9.28.orig/kernel/locking/rtmutex.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/locking/rtmutex.h 2017-05-19 03:37:25.186176955 +0200 -@@ -11,8 +11,6 @@ - */ - - #define rt_mutex_deadlock_check(l) (0) --#define rt_mutex_deadlock_account_lock(m, t) do { } while (0) --#define rt_mutex_deadlock_account_unlock(l) do { } while (0) - #define debug_rt_mutex_init_waiter(w) do { } while (0) - #define debug_rt_mutex_free_waiter(w) do { } while (0) - #define debug_rt_mutex_lock(l) do { } while (0) -diff -Nur linux-4.9.28.orig/kernel/locking/rwsem-rt.c linux-4.9.28/kernel/locking/rwsem-rt.c ---- linux-4.9.28.orig/kernel/locking/rwsem-rt.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/kernel/locking/rwsem-rt.c 2017-05-19 03:37:25.186176955 +0200 +diff --git a/kernel/locking/rwsem-rt.c b/kernel/locking/rwsem-rt.c +new file mode 100644 +index 000000000000..4a708ffcded6 +--- /dev/null ++++ b/kernel/locking/rwsem-rt.c @@ -0,0 +1,268 @@ +/* + */ @@ -16017,10 +16125,11 @@ diff -Nur linux-4.9.28.orig/kernel/locking/rwsem-rt.c linux-4.9.28/kernel/lockin + /* Release it and account current as reader */ + __up_write_unlock(sem, WRITER_BIAS - 1, flags); +} -diff -Nur linux-4.9.28.orig/kernel/locking/spinlock.c linux-4.9.28/kernel/locking/spinlock.c ---- linux-4.9.28.orig/kernel/locking/spinlock.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/locking/spinlock.c 2017-05-19 03:37:25.186176955 +0200 -@@ -124,8 +124,11 @@ +diff --git a/kernel/locking/spinlock.c b/kernel/locking/spinlock.c +index db3ccb1dd614..909779647bd1 100644 +--- a/kernel/locking/spinlock.c ++++ b/kernel/locking/spinlock.c +@@ -124,8 +124,11 @@ void __lockfunc __raw_##op##_lock_bh(locktype##_t *lock) \ * __[spin|read|write]_lock_bh() */ BUILD_LOCK_OPS(spin, raw_spinlock); @@ -16032,7 +16141,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/spinlock.c linux-4.9.28/kernel/lockin #endif -@@ -209,6 +212,8 @@ +@@ -209,6 +212,8 @@ void __lockfunc _raw_spin_unlock_bh(raw_spinlock_t *lock) EXPORT_SYMBOL(_raw_spin_unlock_bh); #endif @@ -16041,7 +16150,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/spinlock.c linux-4.9.28/kernel/lockin #ifndef CONFIG_INLINE_READ_TRYLOCK int __lockfunc _raw_read_trylock(rwlock_t *lock) { -@@ -353,6 +358,8 @@ +@@ -353,6 +358,8 @@ void __lockfunc _raw_write_unlock_bh(rwlock_t *lock) EXPORT_SYMBOL(_raw_write_unlock_bh); #endif @@ -16050,10 +16159,11 @@ diff -Nur linux-4.9.28.orig/kernel/locking/spinlock.c linux-4.9.28/kernel/lockin #ifdef CONFIG_DEBUG_LOCK_ALLOC void __lockfunc _raw_spin_lock_nested(raw_spinlock_t *lock, int subclass) -diff -Nur linux-4.9.28.orig/kernel/locking/spinlock_debug.c linux-4.9.28/kernel/locking/spinlock_debug.c ---- linux-4.9.28.orig/kernel/locking/spinlock_debug.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/locking/spinlock_debug.c 2017-05-19 03:37:25.190177110 +0200 -@@ -31,6 +31,7 @@ +diff --git a/kernel/locking/spinlock_debug.c b/kernel/locking/spinlock_debug.c +index 0374a596cffa..94970338d518 100644 +--- a/kernel/locking/spinlock_debug.c ++++ b/kernel/locking/spinlock_debug.c +@@ -31,6 +31,7 @@ void __raw_spin_lock_init(raw_spinlock_t *lock, const char *name, EXPORT_SYMBOL(__raw_spin_lock_init); @@ -16061,7 +16171,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/spinlock_debug.c linux-4.9.28/kernel/ void __rwlock_init(rwlock_t *lock, const char *name, struct lock_class_key *key) { -@@ -48,6 +49,7 @@ +@@ -48,6 +49,7 @@ void __rwlock_init(rwlock_t *lock, const char *name, } EXPORT_SYMBOL(__rwlock_init); @@ -16069,7 +16179,7 @@ diff -Nur linux-4.9.28.orig/kernel/locking/spinlock_debug.c linux-4.9.28/kernel/ static void spin_dump(raw_spinlock_t *lock, const char *msg) { -@@ -159,6 +161,7 @@ +@@ -159,6 +161,7 @@ void do_raw_spin_unlock(raw_spinlock_t *lock) arch_spin_unlock(&lock->raw_lock); } @@ -16077,16 +16187,17 @@ diff -Nur linux-4.9.28.orig/kernel/locking/spinlock_debug.c linux-4.9.28/kernel/ static void rwlock_bug(rwlock_t *lock, const char *msg) { if (!debug_locks_off()) -@@ -300,3 +303,5 @@ +@@ -300,3 +303,5 @@ void do_raw_write_unlock(rwlock_t *lock) debug_write_unlock(lock); arch_write_unlock(&lock->raw_lock); } + +#endif -diff -Nur linux-4.9.28.orig/kernel/module.c linux-4.9.28/kernel/module.c ---- linux-4.9.28.orig/kernel/module.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/module.c 2017-05-19 03:37:25.190177110 +0200 -@@ -660,16 +660,7 @@ +diff --git a/kernel/module.c b/kernel/module.c +index 0e54d5bf0097..f27764fbfa24 100644 +--- a/kernel/module.c ++++ b/kernel/module.c +@@ -660,16 +660,7 @@ static void percpu_modcopy(struct module *mod, memcpy(per_cpu_ptr(mod->percpu, cpu), from, size); } @@ -16104,7 +16215,7 @@ diff -Nur linux-4.9.28.orig/kernel/module.c linux-4.9.28/kernel/module.c { struct module *mod; unsigned int cpu; -@@ -683,9 +674,15 @@ +@@ -683,9 +674,15 @@ bool is_module_percpu_address(unsigned long addr) continue; for_each_possible_cpu(cpu) { void *start = per_cpu_ptr(mod->percpu, cpu); @@ -16122,7 +16233,7 @@ diff -Nur linux-4.9.28.orig/kernel/module.c linux-4.9.28/kernel/module.c preempt_enable(); return true; } -@@ -696,6 +693,20 @@ +@@ -696,6 +693,20 @@ bool is_module_percpu_address(unsigned long addr) return false; } @@ -16143,22 +16254,23 @@ diff -Nur linux-4.9.28.orig/kernel/module.c linux-4.9.28/kernel/module.c #else /* ... !CONFIG_SMP */ static inline void __percpu *mod_percpu(struct module *mod) -@@ -726,6 +737,11 @@ - { +@@ -727,6 +738,11 @@ bool is_module_percpu_address(unsigned long addr) return false; } -+ + +bool __is_module_percpu_address(unsigned long addr, unsigned long *can_addr) +{ + return false; +} - ++ #endif /* CONFIG_SMP */ -diff -Nur linux-4.9.28.orig/kernel/panic.c linux-4.9.28/kernel/panic.c ---- linux-4.9.28.orig/kernel/panic.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/panic.c 2017-05-19 03:37:25.190177110 +0200 -@@ -482,9 +482,11 @@ + #define MODINFO_ATTR(field) \ +diff --git a/kernel/panic.c b/kernel/panic.c +index e6480e20379e..7e9c1918a94e 100644 +--- a/kernel/panic.c ++++ b/kernel/panic.c +@@ -482,9 +482,11 @@ static u64 oops_id; static int init_oops_id(void) { @@ -16170,10 +16282,11 @@ diff -Nur linux-4.9.28.orig/kernel/panic.c linux-4.9.28/kernel/panic.c oops_id++; return 0; -diff -Nur linux-4.9.28.orig/kernel/power/hibernate.c linux-4.9.28/kernel/power/hibernate.c ---- linux-4.9.28.orig/kernel/power/hibernate.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/power/hibernate.c 2017-05-19 03:37:25.190177110 +0200 -@@ -286,6 +286,8 @@ +diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c +index b26dbc48c75b..968255f27a33 100644 +--- a/kernel/power/hibernate.c ++++ b/kernel/power/hibernate.c +@@ -286,6 +286,8 @@ static int create_image(int platform_mode) local_irq_disable(); @@ -16182,7 +16295,7 @@ diff -Nur linux-4.9.28.orig/kernel/power/hibernate.c linux-4.9.28/kernel/power/h error = syscore_suspend(); if (error) { printk(KERN_ERR "PM: Some system devices failed to power down, " -@@ -317,6 +319,7 @@ +@@ -317,6 +319,7 @@ static int create_image(int platform_mode) syscore_resume(); Enable_irqs: @@ -16190,7 +16303,7 @@ diff -Nur linux-4.9.28.orig/kernel/power/hibernate.c linux-4.9.28/kernel/power/h local_irq_enable(); Enable_cpus: -@@ -446,6 +449,7 @@ +@@ -446,6 +449,7 @@ static int resume_target_kernel(bool platform_mode) goto Enable_cpus; local_irq_disable(); @@ -16198,7 +16311,7 @@ diff -Nur linux-4.9.28.orig/kernel/power/hibernate.c linux-4.9.28/kernel/power/h error = syscore_suspend(); if (error) -@@ -479,6 +483,7 @@ +@@ -479,6 +483,7 @@ static int resume_target_kernel(bool platform_mode) syscore_resume(); Enable_irqs: @@ -16206,7 +16319,7 @@ diff -Nur linux-4.9.28.orig/kernel/power/hibernate.c linux-4.9.28/kernel/power/h local_irq_enable(); Enable_cpus: -@@ -564,6 +569,7 @@ +@@ -564,6 +569,7 @@ int hibernation_platform_enter(void) goto Enable_cpus; local_irq_disable(); @@ -16214,7 +16327,7 @@ diff -Nur linux-4.9.28.orig/kernel/power/hibernate.c linux-4.9.28/kernel/power/h syscore_suspend(); if (pm_wakeup_pending()) { error = -EAGAIN; -@@ -576,6 +582,7 @@ +@@ -576,6 +582,7 @@ int hibernation_platform_enter(void) Power_up: syscore_resume(); @@ -16222,7 +16335,7 @@ diff -Nur linux-4.9.28.orig/kernel/power/hibernate.c linux-4.9.28/kernel/power/h local_irq_enable(); Enable_cpus: -@@ -676,6 +683,10 @@ +@@ -676,6 +683,10 @@ static int load_image_and_restore(void) return error; } @@ -16233,7 +16346,7 @@ diff -Nur linux-4.9.28.orig/kernel/power/hibernate.c linux-4.9.28/kernel/power/h /** * hibernate - Carry out system hibernation, including saving the image. */ -@@ -689,6 +700,8 @@ +@@ -689,6 +700,8 @@ int hibernate(void) return -EPERM; } @@ -16242,7 +16355,7 @@ diff -Nur linux-4.9.28.orig/kernel/power/hibernate.c linux-4.9.28/kernel/power/h lock_system_sleep(); /* The snapshot device should not be opened while we're running */ if (!atomic_add_unless(&snapshot_device_available, -1, 0)) { -@@ -766,6 +779,7 @@ +@@ -766,6 +779,7 @@ int hibernate(void) atomic_inc(&snapshot_device_available); Unlock: unlock_system_sleep(); @@ -16250,10 +16363,11 @@ diff -Nur linux-4.9.28.orig/kernel/power/hibernate.c linux-4.9.28/kernel/power/h return error; } -diff -Nur linux-4.9.28.orig/kernel/power/suspend.c linux-4.9.28/kernel/power/suspend.c ---- linux-4.9.28.orig/kernel/power/suspend.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/power/suspend.c 2017-05-19 03:37:25.190177110 +0200 -@@ -369,6 +369,8 @@ +diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c +index 6ccb08f57fcb..c8cbb5ed2fe3 100644 +--- a/kernel/power/suspend.c ++++ b/kernel/power/suspend.c +@@ -369,6 +369,8 @@ static int suspend_enter(suspend_state_t state, bool *wakeup) arch_suspend_disable_irqs(); BUG_ON(!irqs_disabled()); @@ -16262,7 +16376,7 @@ diff -Nur linux-4.9.28.orig/kernel/power/suspend.c linux-4.9.28/kernel/power/sus error = syscore_suspend(); if (!error) { *wakeup = pm_wakeup_pending(); -@@ -385,6 +387,8 @@ +@@ -385,6 +387,8 @@ static int suspend_enter(suspend_state_t state, bool *wakeup) syscore_resume(); } @@ -16271,7 +16385,7 @@ diff -Nur linux-4.9.28.orig/kernel/power/suspend.c linux-4.9.28/kernel/power/sus arch_suspend_enable_irqs(); BUG_ON(irqs_disabled()); -@@ -527,6 +531,8 @@ +@@ -527,6 +531,8 @@ static int enter_state(suspend_state_t state) return error; } @@ -16280,7 +16394,7 @@ diff -Nur linux-4.9.28.orig/kernel/power/suspend.c linux-4.9.28/kernel/power/sus /** * pm_suspend - Externally visible function for suspending the system. * @state: System sleep state to enter. -@@ -541,6 +547,8 @@ +@@ -541,6 +547,8 @@ int pm_suspend(suspend_state_t state) if (state <= PM_SUSPEND_ON || state >= PM_SUSPEND_MAX) return -EINVAL; @@ -16289,7 +16403,7 @@ diff -Nur linux-4.9.28.orig/kernel/power/suspend.c linux-4.9.28/kernel/power/sus error = enter_state(state); if (error) { suspend_stats.fail++; -@@ -548,6 +556,7 @@ +@@ -548,6 +556,7 @@ int pm_suspend(suspend_state_t state) } else { suspend_stats.success++; } @@ -16297,10 +16411,11 @@ diff -Nur linux-4.9.28.orig/kernel/power/suspend.c linux-4.9.28/kernel/power/sus return error; } EXPORT_SYMBOL(pm_suspend); -diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/printk.c ---- linux-4.9.28.orig/kernel/printk/printk.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/printk/printk.c 2017-05-19 03:37:25.190177110 +0200 -@@ -351,6 +351,65 @@ +diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c +index 9c5b231684d0..cf15bdb6855b 100644 +--- a/kernel/printk/printk.c ++++ b/kernel/printk/printk.c +@@ -351,6 +351,65 @@ __packed __aligned(4) */ DEFINE_RAW_SPINLOCK(logbuf_lock); @@ -16366,7 +16481,7 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr #ifdef CONFIG_PRINTK DECLARE_WAIT_QUEUE_HEAD(log_wait); /* the next printk record to read by syslog(READ) or /proc/kmsg */ -@@ -1337,6 +1396,7 @@ +@@ -1337,6 +1396,7 @@ static int syslog_print_all(char __user *buf, int size, bool clear) { char *text; int len = 0; @@ -16374,7 +16489,7 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr text = kmalloc(LOG_LINE_MAX + PREFIX_MAX, GFP_KERNEL); if (!text) -@@ -1348,6 +1408,14 @@ +@@ -1348,6 +1408,14 @@ static int syslog_print_all(char __user *buf, int size, bool clear) u64 seq; u32 idx; enum log_flags prev; @@ -16389,7 +16504,7 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr /* * Find first record that fits, including all following records, -@@ -1363,6 +1431,14 @@ +@@ -1363,6 +1431,14 @@ static int syslog_print_all(char __user *buf, int size, bool clear) prev = msg->flags; idx = log_next(idx); seq++; @@ -16404,7 +16519,7 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr } /* move first record forward until length fits into the buffer */ -@@ -1376,6 +1452,14 @@ +@@ -1376,6 +1452,14 @@ static int syslog_print_all(char __user *buf, int size, bool clear) prev = msg->flags; idx = log_next(idx); seq++; @@ -16419,7 +16534,7 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr } /* last message fitting into this dump */ -@@ -1416,6 +1500,7 @@ +@@ -1416,6 +1500,7 @@ static int syslog_print_all(char __user *buf, int size, bool clear) clear_seq = log_next_seq; clear_idx = log_next_idx; } @@ -16427,7 +16542,7 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr raw_spin_unlock_irq(&logbuf_lock); kfree(text); -@@ -1569,6 +1654,12 @@ +@@ -1569,6 +1654,12 @@ static void call_console_drivers(int level, if (!console_drivers) return; @@ -16440,7 +16555,7 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr for_each_console(con) { if (exclusive_console && con != exclusive_console) continue; -@@ -1584,6 +1675,7 @@ +@@ -1584,6 +1675,7 @@ static void call_console_drivers(int level, else con->write(con, text, len); } @@ -16448,7 +16563,7 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr } /* -@@ -1781,6 +1873,13 @@ +@@ -1781,6 +1873,13 @@ asmlinkage int vprintk_emit(int facility, int level, /* cpu currently holding logbuf_lock in this function */ static unsigned int logbuf_cpu = UINT_MAX; @@ -16462,7 +16577,7 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr if (level == LOGLEVEL_SCHED) { level = LOGLEVEL_DEFAULT; in_sched = true; -@@ -1885,13 +1984,23 @@ +@@ -1885,13 +1984,23 @@ asmlinkage int vprintk_emit(int facility, int level, /* If called from the scheduler, we can not call up(). */ if (!in_sched) { @@ -16487,7 +16602,7 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr console_unlock(); lockdep_on(); } -@@ -2014,26 +2123,6 @@ +@@ -2014,26 +2123,6 @@ DEFINE_PER_CPU(printk_func_t, printk_func); #endif /* CONFIG_PRINTK */ @@ -16514,7 +16629,7 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr static int __add_preferred_console(char *name, int idx, char *options, char *brl_options) { -@@ -2303,11 +2392,16 @@ +@@ -2303,11 +2392,16 @@ static void console_cont_flush(char *text, size_t size) goto out; len = cont_print_text(text, size); @@ -16531,7 +16646,7 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr return; out: raw_spin_unlock_irqrestore(&logbuf_lock, flags); -@@ -2431,13 +2525,17 @@ +@@ -2431,13 +2525,17 @@ void console_unlock(void) console_idx = log_next(console_idx); console_seq++; console_prev = msg->flags; @@ -16550,7 +16665,7 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr if (do_cond_resched) cond_resched(); } -@@ -2489,6 +2587,11 @@ +@@ -2489,6 +2587,11 @@ void console_unblank(void) { struct console *c; @@ -16562,10 +16677,11 @@ diff -Nur linux-4.9.28.orig/kernel/printk/printk.c linux-4.9.28/kernel/printk/pr /* * console_unblank can no longer be called in interrupt context unless * oops_in_progress is set to 1.. -diff -Nur linux-4.9.28.orig/kernel/ptrace.c linux-4.9.28/kernel/ptrace.c ---- linux-4.9.28.orig/kernel/ptrace.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/ptrace.c 2017-05-19 03:37:25.190177110 +0200 -@@ -166,7 +166,14 @@ +diff --git a/kernel/ptrace.c b/kernel/ptrace.c +index a5caecef88be..61e7c5e2183c 100644 +--- a/kernel/ptrace.c ++++ b/kernel/ptrace.c +@@ -166,7 +166,14 @@ static bool ptrace_freeze_traced(struct task_struct *task) spin_lock_irq(&task->sighand->siglock); if (task_is_traced(task) && !__fatal_signal_pending(task)) { @@ -16581,10 +16697,11 @@ diff -Nur linux-4.9.28.orig/kernel/ptrace.c linux-4.9.28/kernel/ptrace.c ret = true; } spin_unlock_irq(&task->sighand->siglock); -diff -Nur linux-4.9.28.orig/kernel/rcu/rcutorture.c linux-4.9.28/kernel/rcu/rcutorture.c ---- linux-4.9.28.orig/kernel/rcu/rcutorture.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/rcu/rcutorture.c 2017-05-19 03:37:25.190177110 +0200 -@@ -404,6 +404,7 @@ +diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c +index bf08fee53dc7..eeb8ce4ad7b6 100644 +--- a/kernel/rcu/rcutorture.c ++++ b/kernel/rcu/rcutorture.c +@@ -404,6 +404,7 @@ static struct rcu_torture_ops rcu_ops = { .name = "rcu" }; @@ -16592,7 +16709,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/rcutorture.c linux-4.9.28/kernel/rcu/rcut /* * Definitions for rcu_bh torture testing. */ -@@ -443,6 +444,12 @@ +@@ -443,6 +444,12 @@ static struct rcu_torture_ops rcu_bh_ops = { .name = "rcu_bh" }; @@ -16605,9 +16722,10 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/rcutorture.c linux-4.9.28/kernel/rcu/rcut /* * Don't even think about trying any of these in real life!!! * The names includes "busted", and they really means it! -diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c ---- linux-4.9.28.orig/kernel/rcu/tree.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/rcu/tree.c 2017-05-19 03:37:25.190177110 +0200 +diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c +index 10f62c6f48e7..dbee19478f09 100644 +--- a/kernel/rcu/tree.c ++++ b/kernel/rcu/tree.c @@ -55,6 +55,11 @@ #include #include @@ -16620,7 +16738,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c #include "tree.h" #include "rcu.h" -@@ -260,6 +265,19 @@ +@@ -260,6 +265,19 @@ void rcu_sched_qs(void) this_cpu_ptr(&rcu_sched_data), true); } @@ -16640,7 +16758,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c void rcu_bh_qs(void) { if (__this_cpu_read(rcu_bh_data.cpu_no_qs.s)) { -@@ -269,6 +287,7 @@ +@@ -269,6 +287,7 @@ void rcu_bh_qs(void) __this_cpu_write(rcu_bh_data.cpu_no_qs.b.norm, false); } } @@ -16648,7 +16766,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c static DEFINE_PER_CPU(int, rcu_sched_qs_mask); -@@ -449,11 +468,13 @@ +@@ -449,11 +468,13 @@ EXPORT_SYMBOL_GPL(rcu_batches_started_sched); /* * Return the number of RCU BH batches started thus far for debug & stats. */ @@ -16662,7 +16780,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c /* * Return the number of RCU batches completed thus far for debug & stats. -@@ -473,6 +494,7 @@ +@@ -473,6 +494,7 @@ unsigned long rcu_batches_completed_sched(void) } EXPORT_SYMBOL_GPL(rcu_batches_completed_sched); @@ -16670,7 +16788,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c /* * Return the number of RCU BH batches completed thus far for debug & stats. */ -@@ -481,6 +503,7 @@ +@@ -481,6 +503,7 @@ unsigned long rcu_batches_completed_bh(void) return rcu_bh_state.completed; } EXPORT_SYMBOL_GPL(rcu_batches_completed_bh); @@ -16678,7 +16796,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c /* * Return the number of RCU expedited batches completed thus far for -@@ -504,6 +527,7 @@ +@@ -504,6 +527,7 @@ unsigned long rcu_exp_batches_completed_sched(void) } EXPORT_SYMBOL_GPL(rcu_exp_batches_completed_sched); @@ -16686,7 +16804,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c /* * Force a quiescent state. */ -@@ -522,6 +546,13 @@ +@@ -522,6 +546,13 @@ void rcu_bh_force_quiescent_state(void) } EXPORT_SYMBOL_GPL(rcu_bh_force_quiescent_state); @@ -16700,7 +16818,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c /* * Force a quiescent state for RCU-sched. */ -@@ -572,9 +603,11 @@ +@@ -572,9 +603,11 @@ void rcutorture_get_gp_data(enum rcutorture_type test_type, int *flags, case RCU_FLAVOR: rsp = rcu_state_p; break; @@ -16712,7 +16830,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c case RCU_SCHED_FLAVOR: rsp = &rcu_sched_state; break; -@@ -3016,18 +3049,17 @@ +@@ -3016,18 +3049,17 @@ __rcu_process_callbacks(struct rcu_state *rsp) /* * Do RCU core processing for the current CPU. */ @@ -16733,15 +16851,18 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c /* * Schedule RCU callback invocation. If the specified type of RCU * does not support RCU priority boosting, just do a direct call, -@@ -3039,18 +3071,105 @@ +@@ -3039,19 +3071,106 @@ static void invoke_rcu_callbacks(struct rcu_state *rsp, struct rcu_data *rdp) { if (unlikely(!READ_ONCE(rcu_scheduler_fully_active))) return; - if (likely(!rsp->boost)) { - rcu_do_batch(rsp, rdp); +- return; +- } +- invoke_rcu_callbacks_kthread(); + rcu_do_batch(rsp, rdp); -+} -+ + } + +static void rcu_wake_cond(struct task_struct *t, int status) +{ + /* @@ -16755,21 +16876,23 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c +/* + * Wake up this CPU's rcuc kthread to do RCU core processing. + */ -+static void invoke_rcu_core(void) -+{ + static void invoke_rcu_core(void) + { +- if (cpu_online(smp_processor_id())) +- raise_softirq(RCU_SOFTIRQ); + unsigned long flags; + struct task_struct *t; + + if (!cpu_online(smp_processor_id())) - return; ++ return; + local_irq_save(flags); + __this_cpu_write(rcu_cpu_has_work, 1); + t = __this_cpu_read(rcu_cpu_kthread_task); + if (t != NULL && current != t) + rcu_wake_cond(t, __this_cpu_read(rcu_cpu_kthread_status)); + local_irq_restore(flags); -+} -+ + } + +static void rcu_cpu_kthread_park(unsigned int cpu) +{ + per_cpu(rcu_cpu_kthread_status, cpu) = RCU_KTHREAD_OFFCPU; @@ -16808,16 +16931,14 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c + *statusp = RCU_KTHREAD_WAITING; + return; + } - } -- invoke_rcu_callbacks_kthread(); ++ } + *statusp = RCU_KTHREAD_YIELDING; + trace_rcu_utilization(TPS("Start CPU kthread@rcu_yield")); + schedule_timeout_interruptible(2); + trace_rcu_utilization(TPS("End CPU kthread@rcu_yield")); + *statusp = RCU_KTHREAD_WAITING; - } - --static void invoke_rcu_core(void) ++} ++ +static struct smp_hotplug_thread rcu_cpu_thread_spec = { + .store = &rcu_cpu_kthread_task, + .thread_should_run = rcu_cpu_kthread_should_run, @@ -16831,21 +16952,20 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c + * Spawn per-CPU RCU core processing kthreads. + */ +static int __init rcu_spawn_core_kthreads(void) - { -- if (cpu_online(smp_processor_id())) -- raise_softirq(RCU_SOFTIRQ); ++{ + int cpu; + + for_each_possible_cpu(cpu) + per_cpu(rcu_cpu_has_work, cpu) = 0; + BUG_ON(smpboot_register_percpu_thread(&rcu_cpu_thread_spec)); + return 0; - } ++} +early_initcall(rcu_spawn_core_kthreads); - ++ /* * Handle any core-RCU processing required by a call_rcu() invocation. -@@ -3195,6 +3314,7 @@ + */ +@@ -3195,6 +3314,7 @@ void call_rcu_sched(struct rcu_head *head, rcu_callback_t func) } EXPORT_SYMBOL_GPL(call_rcu_sched); @@ -16853,7 +16973,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c /* * Queue an RCU callback for invocation after a quicker grace period. */ -@@ -3203,6 +3323,7 @@ +@@ -3203,6 +3323,7 @@ void call_rcu_bh(struct rcu_head *head, rcu_callback_t func) __call_rcu(head, func, &rcu_bh_state, -1, 0); } EXPORT_SYMBOL_GPL(call_rcu_bh); @@ -16861,7 +16981,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c /* * Queue an RCU callback for lazy invocation after a grace period. -@@ -3294,6 +3415,7 @@ +@@ -3294,6 +3415,7 @@ void synchronize_sched(void) } EXPORT_SYMBOL_GPL(synchronize_sched); @@ -16869,7 +16989,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c /** * synchronize_rcu_bh - wait until an rcu_bh grace period has elapsed. * -@@ -3320,6 +3442,7 @@ +@@ -3320,6 +3442,7 @@ void synchronize_rcu_bh(void) wait_rcu_gp(call_rcu_bh); } EXPORT_SYMBOL_GPL(synchronize_rcu_bh); @@ -16877,7 +16997,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c /** * get_state_synchronize_rcu - Snapshot current RCU state -@@ -3698,6 +3821,7 @@ +@@ -3698,6 +3821,7 @@ static void _rcu_barrier(struct rcu_state *rsp) mutex_unlock(&rsp->barrier_mutex); } @@ -16885,7 +17005,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c /** * rcu_barrier_bh - Wait until all in-flight call_rcu_bh() callbacks complete. */ -@@ -3706,6 +3830,7 @@ +@@ -3706,6 +3830,7 @@ void rcu_barrier_bh(void) _rcu_barrier(&rcu_bh_state); } EXPORT_SYMBOL_GPL(rcu_barrier_bh); @@ -16893,7 +17013,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c /** * rcu_barrier_sched - Wait for in-flight call_rcu_sched() callbacks. -@@ -4227,12 +4352,13 @@ +@@ -4227,12 +4352,13 @@ void __init rcu_init(void) rcu_bootup_announce(); rcu_init_geometry(); @@ -16908,10 +17028,11 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.c linux-4.9.28/kernel/rcu/tree.c /* * We don't need protection against CPU-hotplug here because -diff -Nur linux-4.9.28.orig/kernel/rcu/tree.h linux-4.9.28/kernel/rcu/tree.h ---- linux-4.9.28.orig/kernel/rcu/tree.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/rcu/tree.h 2017-05-19 03:37:25.190177110 +0200 -@@ -588,18 +588,18 @@ +diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h +index e99a5234d9ed..958ac107062c 100644 +--- a/kernel/rcu/tree.h ++++ b/kernel/rcu/tree.h +@@ -588,18 +588,18 @@ extern struct list_head rcu_struct_flavors; */ extern struct rcu_state rcu_sched_state; @@ -16932,7 +17053,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.h linux-4.9.28/kernel/rcu/tree.h #ifndef RCU_TREE_NONCORE -@@ -619,10 +619,9 @@ +@@ -619,10 +619,9 @@ void call_rcu(struct rcu_head *head, rcu_callback_t func); static void __init __rcu_init_preempt(void); static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags); static void rcu_preempt_boost_start_gp(struct rcu_node *rnp); @@ -16944,9 +17065,10 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree.h linux-4.9.28/kernel/rcu/tree.h static int rcu_spawn_one_boost_kthread(struct rcu_state *rsp, struct rcu_node *rnp); #endif /* #ifdef CONFIG_RCU_BOOST */ -diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tree_plugin.h ---- linux-4.9.28.orig/kernel/rcu/tree_plugin.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/rcu/tree_plugin.h 2017-05-19 03:37:25.190177110 +0200 +diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h +index 56583e764ebf..7c656f8e192f 100644 +--- a/kernel/rcu/tree_plugin.h ++++ b/kernel/rcu/tree_plugin.h @@ -24,25 +24,10 @@ * Paul E. McKenney */ @@ -16973,7 +17095,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre #else /* #ifdef CONFIG_RCU_BOOST */ /* -@@ -55,6 +40,14 @@ +@@ -55,6 +40,14 @@ DEFINE_PER_CPU(char, rcu_cpu_has_work); #endif /* #else #ifdef CONFIG_RCU_BOOST */ @@ -16988,7 +17110,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre #ifdef CONFIG_RCU_NOCB_CPU static cpumask_var_t rcu_nocb_mask; /* CPUs to have callbacks offloaded. */ static bool have_rcu_nocb_mask; /* Was rcu_nocb_mask allocated? */ -@@ -426,7 +419,7 @@ +@@ -426,7 +419,7 @@ void rcu_read_unlock_special(struct task_struct *t) } /* Hardware IRQ handlers cannot block, complain if they get here. */ @@ -16997,7 +17119,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre lockdep_rcu_suspicious(__FILE__, __LINE__, "rcu_read_unlock() from irq or softirq with blocking in critical section!!!\n"); pr_alert("->rcu_read_unlock_special: %#x (b: %d, enq: %d nq: %d)\n", -@@ -632,15 +625,6 @@ +@@ -632,15 +625,6 @@ static void rcu_preempt_check_callbacks(void) t->rcu_read_unlock_special.b.need_qs = true; } @@ -17013,7 +17135,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre /* * Queue a preemptible-RCU callback for invocation after a grace period. */ -@@ -829,6 +813,19 @@ +@@ -829,6 +813,19 @@ void exit_rcu(void) #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ @@ -17033,7 +17155,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre #ifdef CONFIG_RCU_BOOST #include "../locking/rtmutex_common.h" -@@ -860,16 +857,6 @@ +@@ -860,16 +857,6 @@ static void rcu_initiate_boost_trace(struct rcu_node *rnp) #endif /* #else #ifdef CONFIG_RCU_TRACE */ @@ -17050,7 +17172,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre /* * Carry out RCU priority boosting on the task indicated by ->exp_tasks * or ->boost_tasks, advancing the pointer to the next task in the -@@ -1013,23 +1000,6 @@ +@@ -1013,23 +1000,6 @@ static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags) } /* @@ -17074,7 +17196,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre * Is the current CPU running the RCU-callbacks kthread? * Caller must have preemption disabled. */ -@@ -1083,67 +1053,6 @@ +@@ -1083,67 +1053,6 @@ static int rcu_spawn_one_boost_kthread(struct rcu_state *rsp, return 0; } @@ -17142,7 +17264,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre /* * Set the per-rcu_node kthread's affinity to cover all CPUs that are * served by the rcu_node in question. The CPU hotplug lock is still -@@ -1174,26 +1083,12 @@ +@@ -1174,26 +1083,12 @@ static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu) free_cpumask_var(cm); } @@ -17169,7 +17291,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre rcu_for_each_leaf_node(rcu_state_p, rnp) (void)rcu_spawn_one_boost_kthread(rcu_state_p, rnp); } -@@ -1216,11 +1111,6 @@ +@@ -1216,11 +1111,6 @@ static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags) raw_spin_unlock_irqrestore_rcu_node(rnp, flags); } @@ -17181,7 +17303,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre static bool rcu_is_callbacks_kthread(void) { return false; -@@ -1244,7 +1134,7 @@ +@@ -1244,7 +1134,7 @@ static void rcu_prepare_kthreads(int cpu) #endif /* #else #ifdef CONFIG_RCU_BOOST */ @@ -17190,7 +17312,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre /* * Check to see if any future RCU-related work will need to be done -@@ -1261,7 +1151,9 @@ +@@ -1261,7 +1151,9 @@ int rcu_needs_cpu(u64 basemono, u64 *nextevt) return IS_ENABLED(CONFIG_RCU_NOCB_CPU_ALL) ? 0 : rcu_cpu_has_callbacks(NULL); } @@ -17200,7 +17322,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre /* * Because we do not have RCU_FAST_NO_HZ, don't bother cleaning up * after it. -@@ -1357,6 +1249,8 @@ +@@ -1357,6 +1249,8 @@ static bool __maybe_unused rcu_try_advance_all_cbs(void) return cbs_ready; } @@ -17209,7 +17331,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre /* * Allow the CPU to enter dyntick-idle mode unless it has callbacks ready * to invoke. If the CPU has callbacks, try to advance them. Tell the -@@ -1402,6 +1296,7 @@ +@@ -1402,6 +1296,7 @@ int rcu_needs_cpu(u64 basemono, u64 *nextevt) *nextevt = basemono + dj * TICK_NSEC; return 0; } @@ -17217,9 +17339,10 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/tree_plugin.h linux-4.9.28/kernel/rcu/tre /* * Prepare a CPU for idle from an RCU perspective. The first major task -diff -Nur linux-4.9.28.orig/kernel/rcu/update.c linux-4.9.28/kernel/rcu/update.c ---- linux-4.9.28.orig/kernel/rcu/update.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/rcu/update.c 2017-05-19 03:37:25.190177110 +0200 +diff --git a/kernel/rcu/update.c b/kernel/rcu/update.c +index 4f6db7e6a117..ee02e1e1b3e5 100644 +--- a/kernel/rcu/update.c ++++ b/kernel/rcu/update.c @@ -62,7 +62,7 @@ #ifndef CONFIG_TINY_RCU module_param(rcu_expedited, int, 0); @@ -17229,7 +17352,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/update.c linux-4.9.28/kernel/rcu/update.c module_param(rcu_normal_after_boot, int, 0); #endif /* #ifndef CONFIG_TINY_RCU */ -@@ -132,8 +132,7 @@ +@@ -132,8 +132,7 @@ bool rcu_gp_is_normal(void) } EXPORT_SYMBOL_GPL(rcu_gp_is_normal); @@ -17239,7 +17362,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/update.c linux-4.9.28/kernel/rcu/update.c /* * Should normal grace-period primitives be expedited? Intended for -@@ -182,8 +181,7 @@ +@@ -182,8 +181,7 @@ EXPORT_SYMBOL_GPL(rcu_unexpedite_gp); */ void rcu_end_inkernel_boot(void) { @@ -17249,7 +17372,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/update.c linux-4.9.28/kernel/rcu/update.c if (rcu_normal_after_boot) WRITE_ONCE(rcu_normal, 1); } -@@ -298,6 +296,7 @@ +@@ -298,6 +296,7 @@ int rcu_read_lock_held(void) } EXPORT_SYMBOL_GPL(rcu_read_lock_held); @@ -17257,7 +17380,7 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/update.c linux-4.9.28/kernel/rcu/update.c /** * rcu_read_lock_bh_held() - might we be in RCU-bh read-side critical section? * -@@ -324,6 +323,7 @@ +@@ -324,6 +323,7 @@ int rcu_read_lock_bh_held(void) return in_softirq() || irqs_disabled(); } EXPORT_SYMBOL_GPL(rcu_read_lock_bh_held); @@ -17265,10 +17388,24 @@ diff -Nur linux-4.9.28.orig/kernel/rcu/update.c linux-4.9.28/kernel/rcu/update.c #endif /* #ifdef CONFIG_DEBUG_LOCK_ALLOC */ -diff -Nur linux-4.9.28.orig/kernel/sched/completion.c linux-4.9.28/kernel/sched/completion.c ---- linux-4.9.28.orig/kernel/sched/completion.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/sched/completion.c 2017-05-19 03:37:25.194177264 +0200 -@@ -30,10 +30,10 @@ +diff --git a/kernel/sched/Makefile b/kernel/sched/Makefile +index 5e59b832ae2b..7337a7f60e3f 100644 +--- a/kernel/sched/Makefile ++++ b/kernel/sched/Makefile +@@ -17,7 +17,7 @@ endif + + obj-y += core.o loadavg.o clock.o cputime.o + obj-y += idle_task.o fair.o rt.o deadline.o stop_task.o +-obj-y += wait.o swait.o completion.o idle.o ++obj-y += wait.o swait.o swork.o completion.o idle.o + obj-$(CONFIG_SMP) += cpupri.o cpudeadline.o + obj-$(CONFIG_SCHED_AUTOGROUP) += auto_group.o + obj-$(CONFIG_SCHEDSTATS) += stats.o +diff --git a/kernel/sched/completion.c b/kernel/sched/completion.c +index 8d0f35debf35..b62cf6400fe0 100644 +--- a/kernel/sched/completion.c ++++ b/kernel/sched/completion.c +@@ -30,10 +30,10 @@ void complete(struct completion *x) { unsigned long flags; @@ -17282,7 +17419,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/completion.c linux-4.9.28/kernel/sched/ } EXPORT_SYMBOL(complete); -@@ -50,10 +50,10 @@ +@@ -50,10 +50,10 @@ void complete_all(struct completion *x) { unsigned long flags; @@ -17296,7 +17433,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/completion.c linux-4.9.28/kernel/sched/ } EXPORT_SYMBOL(complete_all); -@@ -62,20 +62,20 @@ +@@ -62,20 +62,20 @@ do_wait_for_common(struct completion *x, long (*action)(long), long timeout, int state) { if (!x->done) { @@ -17322,7 +17459,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/completion.c linux-4.9.28/kernel/sched/ if (!x->done) return timeout; } -@@ -89,9 +89,9 @@ +@@ -89,9 +89,9 @@ __wait_for_common(struct completion *x, { might_sleep(); @@ -17334,7 +17471,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/completion.c linux-4.9.28/kernel/sched/ return timeout; } -@@ -277,12 +277,12 @@ +@@ -277,12 +277,12 @@ bool try_wait_for_completion(struct completion *x) if (!READ_ONCE(x->done)) return 0; @@ -17349,7 +17486,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/completion.c linux-4.9.28/kernel/sched/ return ret; } EXPORT_SYMBOL(try_wait_for_completion); -@@ -311,7 +311,7 @@ +@@ -311,7 +311,7 @@ bool completion_done(struct completion *x) * after it's acquired the lock. */ smp_rmb(); @@ -17358,10 +17495,11 @@ diff -Nur linux-4.9.28.orig/kernel/sched/completion.c linux-4.9.28/kernel/sched/ return true; } EXPORT_SYMBOL(completion_done); -diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c ---- linux-4.9.28.orig/kernel/sched/core.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/sched/core.c 2017-05-19 03:37:25.194177264 +0200 -@@ -129,7 +129,11 @@ +diff --git a/kernel/sched/core.c b/kernel/sched/core.c +index 154fd689fe02..10e832da70b6 100644 +--- a/kernel/sched/core.c ++++ b/kernel/sched/core.c +@@ -129,7 +129,11 @@ const_debug unsigned int sysctl_sched_features = * Number of tasks to iterate in a single balance run. * Limited because this is done with IRQs disabled. */ @@ -17373,7 +17511,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c /* * period over which we average the RT time consumption, measured -@@ -345,6 +349,7 @@ +@@ -345,6 +349,7 @@ static void init_rq_hrtick(struct rq *rq) hrtimer_init(&rq->hrtick_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); rq->hrtick_timer.function = hrtick; @@ -17381,7 +17519,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c } #else /* CONFIG_SCHED_HRTICK */ static inline void hrtick_clear(struct rq *rq) -@@ -449,7 +454,7 @@ +@@ -449,7 +454,7 @@ void wake_q_add(struct wake_q_head *head, struct task_struct *task) head->lastp = &node->next; } @@ -17390,7 +17528,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c { struct wake_q_node *node = head->first; -@@ -466,7 +471,10 @@ +@@ -466,7 +471,10 @@ void wake_up_q(struct wake_q_head *head) * wake_up_process() implies a wmb() to pair with the queueing * in wake_q_add() so as not to miss wakeups. */ @@ -17402,7 +17540,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c put_task_struct(task); } } -@@ -502,6 +510,38 @@ +@@ -502,6 +510,38 @@ void resched_curr(struct rq *rq) trace_sched_wake_idle_without_ipi(cpu); } @@ -17441,7 +17579,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c void resched_cpu(int cpu) { struct rq *rq = cpu_rq(cpu); -@@ -525,11 +565,14 @@ +@@ -525,11 +565,14 @@ void resched_cpu(int cpu) */ int get_nohz_timer_target(void) { @@ -17458,7 +17596,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c rcu_read_lock(); for_each_domain(cpu, sd) { -@@ -548,6 +591,8 @@ +@@ -548,6 +591,8 @@ int get_nohz_timer_target(void) cpu = housekeeping_any_cpu(); unlock: rcu_read_unlock(); @@ -17467,7 +17605,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c return cpu; } /* -@@ -1100,6 +1145,11 @@ +@@ -1100,6 +1145,11 @@ void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask) lockdep_assert_held(&p->pi_lock); @@ -17479,7 +17617,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c queued = task_on_rq_queued(p); running = task_current(rq, p); -@@ -1122,6 +1172,84 @@ +@@ -1122,6 +1172,84 @@ void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask) set_curr_task(rq, p); } @@ -17564,7 +17702,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c /* * Change a given task's CPU affinity. Migrate the thread to a * proper CPU and schedule it away if the CPU it's executing on -@@ -1179,7 +1307,7 @@ +@@ -1179,7 +1307,7 @@ static int __set_cpus_allowed_ptr(struct task_struct *p, } /* Can the task run on the task's current CPU? If so, we're done */ @@ -17573,7 +17711,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c goto out; dest_cpu = cpumask_any_and(cpu_valid_mask, new_mask); -@@ -1366,6 +1494,18 @@ +@@ -1366,6 +1494,18 @@ int migrate_swap(struct task_struct *cur, struct task_struct *p) return ret; } @@ -17592,7 +17730,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c /* * wait_task_inactive - wait for a thread to unschedule. * -@@ -1410,7 +1550,7 @@ +@@ -1410,7 +1550,7 @@ unsigned long wait_task_inactive(struct task_struct *p, long match_state) * is actually now running somewhere else! */ while (task_running(rq, p)) { @@ -17601,7 +17739,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c return 0; cpu_relax(); } -@@ -1425,7 +1565,8 @@ +@@ -1425,7 +1565,8 @@ unsigned long wait_task_inactive(struct task_struct *p, long match_state) running = task_running(rq, p); queued = task_on_rq_queued(p); ncsw = 0; @@ -17611,7 +17749,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c ncsw = p->nvcsw | LONG_MIN; /* sets MSB */ task_rq_unlock(rq, p, &rf); -@@ -1680,10 +1821,6 @@ +@@ -1680,10 +1821,6 @@ static inline void ttwu_activate(struct rq *rq, struct task_struct *p, int en_fl { activate_task(rq, p, en_flags); p->on_rq = TASK_ON_RQ_QUEUED; @@ -17622,7 +17760,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c } /* -@@ -2018,8 +2155,27 @@ +@@ -2018,8 +2155,27 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags) */ smp_mb__before_spinlock(); raw_spin_lock_irqsave(&p->pi_lock, flags); @@ -17651,7 +17789,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c trace_sched_waking(p); -@@ -2102,53 +2258,6 @@ +@@ -2102,53 +2258,6 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags) } /** @@ -17705,7 +17843,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c * wake_up_process - Wake up a specific process * @p: The process to be woken up. * -@@ -2166,6 +2275,18 @@ +@@ -2166,6 +2275,18 @@ int wake_up_process(struct task_struct *p) } EXPORT_SYMBOL(wake_up_process); @@ -17718,13 +17856,13 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c + */ +int wake_up_lock_sleeper(struct task_struct *p) +{ -+ return try_to_wake_up(p, TASK_ALL, WF_LOCK_SLEEPER); ++ return try_to_wake_up(p, TASK_UNINTERRUPTIBLE, WF_LOCK_SLEEPER); +} + int wake_up_state(struct task_struct *p, unsigned int state) { return try_to_wake_up(p, state, 0); -@@ -2442,6 +2563,9 @@ +@@ -2442,6 +2563,9 @@ int sched_fork(unsigned long clone_flags, struct task_struct *p) p->on_cpu = 0; #endif init_task_preempt_count(p); @@ -17734,7 +17872,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c #ifdef CONFIG_SMP plist_node_init(&p->pushable_tasks, MAX_PRIO); RB_CLEAR_NODE(&p->pushable_dl_tasks); -@@ -2770,21 +2894,16 @@ +@@ -2770,21 +2894,16 @@ static struct rq *finish_task_switch(struct task_struct *prev) finish_arch_post_lock_switch(); fire_sched_in_preempt_notifiers(current); @@ -17761,7 +17899,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c put_task_struct(prev); } -@@ -3252,6 +3371,77 @@ +@@ -3252,6 +3371,77 @@ static inline void schedule_debug(struct task_struct *prev) schedstat_inc(this_rq()->sched_count); } @@ -17839,7 +17977,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c /* * Pick up the highest-prio task: */ -@@ -3368,19 +3558,6 @@ +@@ -3368,19 +3558,6 @@ static void __sched notrace __schedule(bool preempt) } else { deactivate_task(rq, prev, DEQUEUE_SLEEP); prev->on_rq = 0; @@ -17859,7 +17997,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c } switch_count = &prev->nvcsw; } -@@ -3390,6 +3567,7 @@ +@@ -3390,6 +3567,7 @@ static void __sched notrace __schedule(bool preempt) next = pick_next_task(rq, prev, cookie); clear_tsk_need_resched(prev); @@ -17867,7 +18005,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c clear_preempt_need_resched(); rq->clock_skip_update = 0; -@@ -3437,9 +3615,20 @@ +@@ -3437,9 +3615,20 @@ void __noreturn do_task_dead(void) static inline void sched_submit_work(struct task_struct *tsk) { @@ -17889,7 +18027,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c * If we are going to sleep and we have plugged IO queued, * make sure to submit it to avoid deadlocks. */ -@@ -3447,6 +3636,12 @@ +@@ -3447,6 +3636,12 @@ static inline void sched_submit_work(struct task_struct *tsk) blk_schedule_flush_plug(tsk); } @@ -17902,7 +18040,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c asmlinkage __visible void __sched schedule(void) { struct task_struct *tsk = current; -@@ -3457,6 +3652,7 @@ +@@ -3457,6 +3652,7 @@ asmlinkage __visible void __sched schedule(void) __schedule(false); sched_preempt_enable_no_resched(); } while (need_resched()); @@ -17910,7 +18048,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c } EXPORT_SYMBOL(schedule); -@@ -3520,6 +3716,30 @@ +@@ -3520,6 +3716,30 @@ static void __sched notrace preempt_schedule_common(void) } while (need_resched()); } @@ -17941,7 +18079,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c #ifdef CONFIG_PREEMPT /* * this is the entry point to schedule() from in-kernel preemption -@@ -3534,7 +3754,8 @@ +@@ -3534,7 +3754,8 @@ asmlinkage __visible void __sched notrace preempt_schedule(void) */ if (likely(!preemptible())) return; @@ -17951,7 +18089,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c preempt_schedule_common(); } NOKPROBE_SYMBOL(preempt_schedule); -@@ -3561,6 +3782,9 @@ +@@ -3561,6 +3782,9 @@ asmlinkage __visible void __sched notrace preempt_schedule_notrace(void) if (likely(!preemptible())) return; @@ -17961,7 +18099,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c do { /* * Because the function tracer can trace preempt_count_sub() -@@ -3583,7 +3807,16 @@ +@@ -3583,7 +3807,16 @@ asmlinkage __visible void __sched notrace preempt_schedule_notrace(void) * an infinite recursion. */ prev_ctx = exception_enter(); @@ -17978,7 +18116,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c exception_exit(prev_ctx); preempt_latency_stop(1); -@@ -3629,10 +3862,25 @@ +@@ -3629,10 +3862,25 @@ EXPORT_SYMBOL(default_wake_function); #ifdef CONFIG_RT_MUTEXES @@ -18006,7 +18144,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c * * This function changes the 'effective' priority of a task. It does * not touch ->normal_prio like __setscheduler(). -@@ -3640,16 +3888,40 @@ +@@ -3640,16 +3888,40 @@ EXPORT_SYMBOL(default_wake_function); * Used by the rt_mutex code to implement priority inheritance * logic. Call site only calls if the priority of the task changed. */ @@ -18050,7 +18188,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c /* * Idle task boosting is a nono in general. There is one -@@ -3669,7 +3941,7 @@ +@@ -3669,7 +3941,7 @@ void rt_mutex_setprio(struct task_struct *p, int prio) goto out_unlock; } @@ -18059,7 +18197,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c oldprio = p->prio; if (oldprio == prio) -@@ -3693,7 +3965,6 @@ +@@ -3693,7 +3965,6 @@ void rt_mutex_setprio(struct task_struct *p, int prio) * running task */ if (dl_prio(prio)) { @@ -18067,7 +18205,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c if (!dl_prio(p->normal_prio) || (pi_task && dl_entity_preempt(&pi_task->dl, &p->dl))) { p->dl.dl_boosted = 1; -@@ -3730,6 +4001,11 @@ +@@ -3730,6 +4001,11 @@ void rt_mutex_setprio(struct task_struct *p, int prio) balance_callback(rq); preempt_enable(); } @@ -18079,7 +18217,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c #endif void set_user_nice(struct task_struct *p, long nice) -@@ -3974,10 +4250,9 @@ +@@ -3974,10 +4250,9 @@ static void __setscheduler(struct rq *rq, struct task_struct *p, * Keep a potential priority boosting if called from * sched_setscheduler(). */ @@ -18092,7 +18230,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c if (dl_prio(p->prio)) p->sched_class = &dl_sched_class; -@@ -4264,7 +4539,7 @@ +@@ -4264,7 +4539,7 @@ static int __sched_setscheduler(struct task_struct *p, * the runqueue. This will be done when the task deboost * itself. */ @@ -18101,7 +18239,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c if (new_effective_prio == oldprio) queue_flags &= ~DEQUEUE_MOVE; } -@@ -4939,6 +5214,7 @@ +@@ -4939,6 +5214,7 @@ int __cond_resched_lock(spinlock_t *lock) } EXPORT_SYMBOL(__cond_resched_lock); @@ -18109,7 +18247,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c int __sched __cond_resched_softirq(void) { BUG_ON(!in_softirq()); -@@ -4952,6 +5228,7 @@ +@@ -4952,6 +5228,7 @@ int __sched __cond_resched_softirq(void) return 0; } EXPORT_SYMBOL(__cond_resched_softirq); @@ -18117,7 +18255,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c /** * yield - yield the current processor to other threads. -@@ -5315,7 +5592,9 @@ +@@ -5315,7 +5592,9 @@ void init_idle(struct task_struct *idle, int cpu) /* Set the preempt count _outside_ the spinlocks! */ init_idle_preempt_count(idle, cpu); @@ -18128,7 +18266,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c /* * The idle tasks have their own, simple scheduling class: */ -@@ -5458,6 +5737,8 @@ +@@ -5458,6 +5737,8 @@ void sched_setnuma(struct task_struct *p, int nid) #endif /* CONFIG_NUMA_BALANCING */ #ifdef CONFIG_HOTPLUG_CPU @@ -18137,7 +18275,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c /* * Ensures that the idle task is using init_mm right before its cpu goes * offline. -@@ -5472,7 +5753,12 @@ +@@ -5472,7 +5753,12 @@ void idle_task_exit(void) switch_mm_irqs_off(mm, &init_mm, current); finish_arch_post_lock_switch(); } @@ -18151,7 +18289,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c } /* -@@ -7418,6 +7704,10 @@ +@@ -7418,6 +7704,10 @@ int sched_cpu_dying(unsigned int cpu) update_max_interval(); nohz_balance_exit_idle(cpu); hrtick_clear(rq); @@ -18162,7 +18300,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c return 0; } #endif -@@ -7698,7 +7988,7 @@ +@@ -7698,7 +7988,7 @@ void __init sched_init(void) #ifdef CONFIG_DEBUG_ATOMIC_SLEEP static inline int preempt_count_equals(int preempt_offset) { @@ -18171,10 +18309,11 @@ diff -Nur linux-4.9.28.orig/kernel/sched/core.c linux-4.9.28/kernel/sched/core.c return (nested == preempt_offset); } -diff -Nur linux-4.9.28.orig/kernel/sched/deadline.c linux-4.9.28/kernel/sched/deadline.c ---- linux-4.9.28.orig/kernel/sched/deadline.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/sched/deadline.c 2017-05-19 03:37:25.194177264 +0200 -@@ -687,6 +687,7 @@ +diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c +index c95c5122b105..e00accf92a4b 100644 +--- a/kernel/sched/deadline.c ++++ b/kernel/sched/deadline.c +@@ -687,6 +687,7 @@ void init_dl_task_timer(struct sched_dl_entity *dl_se) hrtimer_init(timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); timer->function = dl_task_timer; @@ -18182,10 +18321,11 @@ diff -Nur linux-4.9.28.orig/kernel/sched/deadline.c linux-4.9.28/kernel/sched/de } static -diff -Nur linux-4.9.28.orig/kernel/sched/debug.c linux-4.9.28/kernel/sched/debug.c ---- linux-4.9.28.orig/kernel/sched/debug.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/sched/debug.c 2017-05-19 03:37:25.194177264 +0200 -@@ -558,6 +558,9 @@ +diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c +index fa178b62ea79..935224123441 100644 +--- a/kernel/sched/debug.c ++++ b/kernel/sched/debug.c +@@ -558,6 +558,9 @@ void print_rt_rq(struct seq_file *m, int cpu, struct rt_rq *rt_rq) P(rt_throttled); PN(rt_time); PN(rt_runtime); @@ -18195,7 +18335,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/debug.c linux-4.9.28/kernel/sched/debug #undef PN #undef P -@@ -953,6 +956,10 @@ +@@ -953,6 +956,10 @@ void proc_sched_show_task(struct task_struct *p, struct seq_file *m) #endif P(policy); P(prio); @@ -18206,10 +18346,11 @@ diff -Nur linux-4.9.28.orig/kernel/sched/debug.c linux-4.9.28/kernel/sched/debug #undef PN_SCHEDSTAT #undef PN #undef __PN -diff -Nur linux-4.9.28.orig/kernel/sched/fair.c linux-4.9.28/kernel/sched/fair.c ---- linux-4.9.28.orig/kernel/sched/fair.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/sched/fair.c 2017-05-19 03:37:25.194177264 +0200 -@@ -3518,7 +3518,7 @@ +diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c +index c242944f5cbd..4aeb2e2e41bc 100644 +--- a/kernel/sched/fair.c ++++ b/kernel/sched/fair.c +@@ -3518,7 +3518,7 @@ check_preempt_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr) ideal_runtime = sched_slice(cfs_rq, curr); delta_exec = curr->sum_exec_runtime - curr->prev_sum_exec_runtime; if (delta_exec > ideal_runtime) { @@ -18218,7 +18359,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/fair.c linux-4.9.28/kernel/sched/fair.c /* * The current task ran long enough, ensure it doesn't get * re-elected due to buddy favours. -@@ -3542,7 +3542,7 @@ +@@ -3542,7 +3542,7 @@ check_preempt_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr) return; if (delta > ideal_runtime) @@ -18227,7 +18368,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/fair.c linux-4.9.28/kernel/sched/fair.c } static void -@@ -3684,7 +3684,7 @@ +@@ -3684,7 +3684,7 @@ entity_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr, int queued) * validating it and just reschedule. */ if (queued) { @@ -18236,7 +18377,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/fair.c linux-4.9.28/kernel/sched/fair.c return; } /* -@@ -3866,7 +3866,7 @@ +@@ -3866,7 +3866,7 @@ static void __account_cfs_rq_runtime(struct cfs_rq *cfs_rq, u64 delta_exec) * hierarchy can be throttled */ if (!assign_cfs_rq_runtime(cfs_rq) && likely(cfs_rq->curr)) @@ -18245,7 +18386,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/fair.c linux-4.9.28/kernel/sched/fair.c } static __always_inline -@@ -4494,7 +4494,7 @@ +@@ -4494,7 +4494,7 @@ static void hrtick_start_fair(struct rq *rq, struct task_struct *p) if (delta < 0) { if (rq->curr == p) @@ -18254,7 +18395,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/fair.c linux-4.9.28/kernel/sched/fair.c return; } hrtick_start(rq, delta); -@@ -5905,7 +5905,7 @@ +@@ -5905,7 +5905,7 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_ return; preempt: @@ -18263,7 +18404,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/fair.c linux-4.9.28/kernel/sched/fair.c /* * Only set the backward buddy when the current task is still * on the rq. This can happen when a wakeup gets interleaved -@@ -8631,7 +8631,7 @@ +@@ -8631,7 +8631,7 @@ static void task_fork_fair(struct task_struct *p) * 'current' within the tree based on its new key value. */ swap(curr->vruntime, se->vruntime); @@ -18272,7 +18413,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/fair.c linux-4.9.28/kernel/sched/fair.c } se->vruntime -= cfs_rq->min_vruntime; -@@ -8655,7 +8655,7 @@ +@@ -8655,7 +8655,7 @@ prio_changed_fair(struct rq *rq, struct task_struct *p, int oldprio) */ if (rq->curr == p) { if (p->prio > oldprio) @@ -18281,10 +18422,11 @@ diff -Nur linux-4.9.28.orig/kernel/sched/fair.c linux-4.9.28/kernel/sched/fair.c } else check_preempt_curr(rq, p, 0); } -diff -Nur linux-4.9.28.orig/kernel/sched/features.h linux-4.9.28/kernel/sched/features.h ---- linux-4.9.28.orig/kernel/sched/features.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/sched/features.h 2017-05-19 03:37:25.194177264 +0200 -@@ -45,11 +45,19 @@ +diff --git a/kernel/sched/features.h b/kernel/sched/features.h +index 69631fa46c2f..6d28fcd08872 100644 +--- a/kernel/sched/features.h ++++ b/kernel/sched/features.h +@@ -45,11 +45,19 @@ SCHED_FEAT(LB_BIAS, true) */ SCHED_FEAT(NONTASK_CAPACITY, true) @@ -18303,23 +18445,12 @@ diff -Nur linux-4.9.28.orig/kernel/sched/features.h linux-4.9.28/kernel/sched/fe +#endif #ifdef HAVE_RT_PUSH_IPI - /* -diff -Nur linux-4.9.28.orig/kernel/sched/Makefile linux-4.9.28/kernel/sched/Makefile ---- linux-4.9.28.orig/kernel/sched/Makefile 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/sched/Makefile 2017-05-19 03:37:25.194177264 +0200 -@@ -17,7 +17,7 @@ - - obj-y += core.o loadavg.o clock.o cputime.o - obj-y += idle_task.o fair.o rt.o deadline.o stop_task.o --obj-y += wait.o swait.o completion.o idle.o -+obj-y += wait.o swait.o swork.o completion.o idle.o - obj-$(CONFIG_SMP) += cpupri.o cpudeadline.o - obj-$(CONFIG_SCHED_AUTOGROUP) += auto_group.o - obj-$(CONFIG_SCHEDSTATS) += stats.o -diff -Nur linux-4.9.28.orig/kernel/sched/rt.c linux-4.9.28/kernel/sched/rt.c ---- linux-4.9.28.orig/kernel/sched/rt.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/sched/rt.c 2017-05-19 03:37:25.194177264 +0200 -@@ -47,6 +47,7 @@ + /* +diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c +index f139f22ce30d..b0691f4e7d49 100644 +--- a/kernel/sched/rt.c ++++ b/kernel/sched/rt.c +@@ -47,6 +47,7 @@ void init_rt_bandwidth(struct rt_bandwidth *rt_b, u64 period, u64 runtime) hrtimer_init(&rt_b->rt_period_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); @@ -18327,7 +18458,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/rt.c linux-4.9.28/kernel/sched/rt.c rt_b->rt_period_timer.function = sched_rt_period_timer; } -@@ -101,6 +102,7 @@ +@@ -101,6 +102,7 @@ void init_rt_rq(struct rt_rq *rt_rq) rt_rq->push_cpu = nr_cpu_ids; raw_spin_lock_init(&rt_rq->push_lock); init_irq_work(&rt_rq->push_work, push_irq_work_func); @@ -18335,10 +18466,11 @@ diff -Nur linux-4.9.28.orig/kernel/sched/rt.c linux-4.9.28/kernel/sched/rt.c #endif #endif /* CONFIG_SMP */ /* We start is dequeued state, because no RT tasks are queued */ -diff -Nur linux-4.9.28.orig/kernel/sched/sched.h linux-4.9.28/kernel/sched/sched.h ---- linux-4.9.28.orig/kernel/sched/sched.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/sched/sched.h 2017-05-19 03:37:25.194177264 +0200 -@@ -1163,6 +1163,7 @@ +diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h +index 055f935d4421..19324ac27026 100644 +--- a/kernel/sched/sched.h ++++ b/kernel/sched/sched.h +@@ -1163,6 +1163,7 @@ static inline void finish_lock_switch(struct rq *rq, struct task_struct *prev) #define WF_SYNC 0x01 /* waker goes to sleep after wakeup */ #define WF_FORK 0x02 /* child wakeup after fork */ #define WF_MIGRATED 0x4 /* internal use, task got migrated */ @@ -18346,7 +18478,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/sched.h linux-4.9.28/kernel/sched/sched /* * To aid in avoiding the subversion of "niceness" due to uneven distribution -@@ -1346,6 +1347,15 @@ +@@ -1346,6 +1347,15 @@ extern void init_sched_fair_class(void); extern void resched_curr(struct rq *rq); extern void resched_cpu(int cpu); @@ -18362,9 +18494,10 @@ diff -Nur linux-4.9.28.orig/kernel/sched/sched.h linux-4.9.28/kernel/sched/sched extern struct rt_bandwidth def_rt_bandwidth; extern void init_rt_bandwidth(struct rt_bandwidth *rt_b, u64 period, u64 runtime); -diff -Nur linux-4.9.28.orig/kernel/sched/swait.c linux-4.9.28/kernel/sched/swait.c ---- linux-4.9.28.orig/kernel/sched/swait.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/sched/swait.c 2017-05-19 03:37:25.198177419 +0200 +diff --git a/kernel/sched/swait.c b/kernel/sched/swait.c +index 82f0dff90030..ef027ff3250a 100644 +--- a/kernel/sched/swait.c ++++ b/kernel/sched/swait.c @@ -1,5 +1,6 @@ #include #include @@ -18372,7 +18505,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/swait.c linux-4.9.28/kernel/sched/swait void __init_swait_queue_head(struct swait_queue_head *q, const char *name, struct lock_class_key *key) -@@ -29,6 +30,25 @@ +@@ -29,6 +30,25 @@ void swake_up_locked(struct swait_queue_head *q) } EXPORT_SYMBOL(swake_up_locked); @@ -18398,7 +18531,7 @@ diff -Nur linux-4.9.28.orig/kernel/sched/swait.c linux-4.9.28/kernel/sched/swait void swake_up(struct swait_queue_head *q) { unsigned long flags; -@@ -54,6 +74,7 @@ +@@ -54,6 +74,7 @@ void swake_up_all(struct swait_queue_head *q) if (!swait_active(q)) return; @@ -18406,9 +18539,11 @@ diff -Nur linux-4.9.28.orig/kernel/sched/swait.c linux-4.9.28/kernel/sched/swait raw_spin_lock_irq(&q->lock); list_splice_init(&q->task_list, &tmp); while (!list_empty(&tmp)) { -diff -Nur linux-4.9.28.orig/kernel/sched/swork.c linux-4.9.28/kernel/sched/swork.c ---- linux-4.9.28.orig/kernel/sched/swork.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/kernel/sched/swork.c 2017-05-19 03:37:25.198177419 +0200 +diff --git a/kernel/sched/swork.c b/kernel/sched/swork.c +new file mode 100644 +index 000000000000..1950f40ca725 +--- /dev/null ++++ b/kernel/sched/swork.c @@ -0,0 +1,173 @@ +/* + * Copyright (C) 2014 BMW Car IT GmbH, Daniel Wagner daniel.wagner@bmw-carit.de @@ -18583,9 +18718,10 @@ diff -Nur linux-4.9.28.orig/kernel/sched/swork.c linux-4.9.28/kernel/sched/swork + mutex_unlock(&worker_mutex); +} +EXPORT_SYMBOL_GPL(swork_put); -diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c ---- linux-4.9.28.orig/kernel/signal.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/signal.c 2017-05-19 03:37:25.198177419 +0200 +diff --git a/kernel/signal.c b/kernel/signal.c +index 0b1415720a15..c884647951f7 100644 +--- a/kernel/signal.c ++++ b/kernel/signal.c @@ -14,6 +14,7 @@ #include #include @@ -18594,7 +18730,7 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c #include #include #include -@@ -352,13 +353,30 @@ +@@ -352,13 +353,30 @@ static bool task_participate_group_stop(struct task_struct *task) return false; } @@ -18626,7 +18762,7 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c { struct sigqueue *q = NULL; struct user_struct *user; -@@ -375,7 +393,10 @@ +@@ -375,7 +393,10 @@ __sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimi if (override_rlimit || atomic_read(&user->sigpending) <= task_rlimit(t, RLIMIT_SIGPENDING)) { @@ -18638,7 +18774,7 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c } else { print_dropped_signal(sig); } -@@ -392,6 +413,13 @@ +@@ -392,6 +413,13 @@ __sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimi return q; } @@ -18652,7 +18788,7 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c static void __sigqueue_free(struct sigqueue *q) { if (q->flags & SIGQUEUE_PREALLOC) -@@ -401,6 +429,21 @@ +@@ -401,6 +429,21 @@ static void __sigqueue_free(struct sigqueue *q) kmem_cache_free(sigqueue_cachep, q); } @@ -18674,7 +18810,7 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c void flush_sigqueue(struct sigpending *queue) { struct sigqueue *q; -@@ -414,6 +457,21 @@ +@@ -414,6 +457,21 @@ void flush_sigqueue(struct sigpending *queue) } /* @@ -18696,7 +18832,7 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c * Flush all pending signals for this kthread. */ void flush_signals(struct task_struct *t) -@@ -525,7 +583,7 @@ +@@ -525,7 +583,7 @@ static void collect_signal(int sig, struct sigpending *list, siginfo_t *info) still_pending: list_del_init(&first->list); copy_siginfo(info, &first->info); @@ -18705,7 +18841,7 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c } else { /* * Ok, it wasn't in the queue. This must be -@@ -560,6 +618,8 @@ +@@ -560,6 +618,8 @@ int dequeue_signal(struct task_struct *tsk, sigset_t *mask, siginfo_t *info) { int signr; @@ -18714,7 +18850,7 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c /* We only dequeue private signals from ourselves, we don't let * signalfd steal them */ -@@ -1156,8 +1216,8 @@ +@@ -1156,8 +1216,8 @@ int do_send_sig_info(int sig, struct siginfo *info, struct task_struct *p, * We don't want to have recursive SIGSEGV's etc, for example, * that is why we also clear SIGNAL_UNKILLABLE. */ @@ -18725,7 +18861,7 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c { unsigned long int flags; int ret, blocked, ignored; -@@ -1182,6 +1242,39 @@ +@@ -1182,6 +1242,39 @@ force_sig_info(int sig, struct siginfo *info, struct task_struct *t) return ret; } @@ -18765,7 +18901,7 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c /* * Nuke all other threads in the group. */ -@@ -1216,12 +1309,12 @@ +@@ -1216,12 +1309,12 @@ struct sighand_struct *__lock_task_sighand(struct task_struct *tsk, * Disable interrupts early to avoid deadlocks. * See rcu_read_unlock() comment header for details. */ @@ -18780,7 +18916,7 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c break; } /* -@@ -1242,7 +1335,7 @@ +@@ -1242,7 +1335,7 @@ struct sighand_struct *__lock_task_sighand(struct task_struct *tsk, } spin_unlock(&sighand->siglock); rcu_read_unlock(); @@ -18789,7 +18925,7 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c } return sighand; -@@ -1485,7 +1578,8 @@ +@@ -1485,7 +1578,8 @@ EXPORT_SYMBOL(kill_pid); */ struct sigqueue *sigqueue_alloc(void) { @@ -18799,7 +18935,7 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c if (q) q->flags |= SIGQUEUE_PREALLOC; -@@ -1846,15 +1940,7 @@ +@@ -1846,15 +1940,7 @@ static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info) if (gstop_done && ptrace_reparented(current)) do_notify_parent_cldstop(current, false, why); @@ -18815,9 +18951,10 @@ diff -Nur linux-4.9.28.orig/kernel/signal.c linux-4.9.28/kernel/signal.c freezable_schedule(); } else { /* -diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c ---- linux-4.9.28.orig/kernel/softirq.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/softirq.c 2017-05-19 03:37:25.198177419 +0200 +diff --git a/kernel/softirq.c b/kernel/softirq.c +index 744fa611cae0..819bd7cf5ad0 100644 +--- a/kernel/softirq.c ++++ b/kernel/softirq.c @@ -21,10 +21,12 @@ #include #include @@ -18831,7 +18968,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c #include #define CREATE_TRACE_POINTS -@@ -56,12 +58,108 @@ +@@ -56,12 +58,108 @@ EXPORT_SYMBOL(irq_stat); static struct softirq_action softirq_vec[NR_SOFTIRQS] __cacheline_aligned_in_smp; DEFINE_PER_CPU(struct task_struct *, ksoftirqd); @@ -18940,7 +19077,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c /* * we cannot loop indefinitely here to avoid userspace starvation, * but we also don't want to introduce a worst case 1/HZ latency -@@ -77,6 +175,38 @@ +@@ -77,6 +175,38 @@ static void wakeup_softirqd(void) wake_up_process(tsk); } @@ -18979,7 +19116,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c /* * If ksoftirqd is scheduled, we do not want to process pending softirqs * right now. Let ksoftirqd handle this at its own rate, to get fairness. -@@ -88,6 +218,47 @@ +@@ -88,6 +218,47 @@ static bool ksoftirqd_running(void) return tsk && (tsk->state == TASK_RUNNING); } @@ -19027,7 +19164,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c /* * preempt_count and SOFTIRQ_OFFSET usage: * - preempt_count is changed by SOFTIRQ_OFFSET on entering or leaving -@@ -243,10 +414,8 @@ +@@ -243,10 +414,8 @@ asmlinkage __visible void __softirq_entry __do_softirq(void) unsigned long end = jiffies + MAX_SOFTIRQ_TIME; unsigned long old_flags = current->flags; int max_restart = MAX_SOFTIRQ_RESTART; @@ -19038,7 +19175,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c /* * Mask out PF_MEMALLOC s current task context is borrowed for the -@@ -265,36 +434,7 @@ +@@ -265,36 +434,7 @@ asmlinkage __visible void __softirq_entry __do_softirq(void) /* Reset the pending bitmask before enabling irqs */ set_softirq_pending(0); @@ -19076,7 +19213,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c pending = local_softirq_pending(); if (pending) { -@@ -331,6 +471,309 @@ +@@ -331,6 +471,309 @@ asmlinkage __visible void do_softirq(void) } /* @@ -19386,7 +19523,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c * Enter an interrupt context. */ void irq_enter(void) -@@ -341,9 +784,9 @@ +@@ -341,9 +784,9 @@ void irq_enter(void) * Prevent raise_softirq from needlessly waking up ksoftirqd * here, as softirq will be serviced on return from interrupt. */ @@ -19398,7 +19535,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c } __irq_enter(); -@@ -351,6 +794,7 @@ +@@ -351,6 +794,7 @@ void irq_enter(void) static inline void invoke_softirq(void) { @@ -19406,7 +19543,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c if (ksoftirqd_running()) return; -@@ -373,6 +817,18 @@ +@@ -373,6 +817,18 @@ static inline void invoke_softirq(void) } else { wakeup_softirqd(); } @@ -19425,7 +19562,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c } static inline void tick_irq_exit(void) -@@ -409,26 +865,6 @@ +@@ -409,26 +865,6 @@ void irq_exit(void) trace_hardirq_exit(); /* must be last! */ } @@ -19452,7 +19589,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c void raise_softirq(unsigned int nr) { unsigned long flags; -@@ -438,12 +874,6 @@ +@@ -438,12 +874,6 @@ void raise_softirq(unsigned int nr) local_irq_restore(flags); } @@ -19465,7 +19602,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c void open_softirq(int nr, void (*action)(struct softirq_action *)) { softirq_vec[nr].action = action; -@@ -460,15 +890,45 @@ +@@ -460,15 +890,45 @@ struct tasklet_head { static DEFINE_PER_CPU(struct tasklet_head, tasklet_vec); static DEFINE_PER_CPU(struct tasklet_head, tasklet_hi_vec); @@ -19515,7 +19652,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c local_irq_restore(flags); } EXPORT_SYMBOL(__tasklet_schedule); -@@ -478,10 +938,7 @@ +@@ -478,10 +938,7 @@ void __tasklet_hi_schedule(struct tasklet_struct *t) unsigned long flags; local_irq_save(flags); @@ -19527,7 +19664,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c local_irq_restore(flags); } EXPORT_SYMBOL(__tasklet_hi_schedule); -@@ -490,82 +947,122 @@ +@@ -490,82 +947,122 @@ void __tasklet_hi_schedule_first(struct tasklet_struct *t) { BUG_ON(!irqs_disabled()); @@ -19668,7 +19805,8 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c list = __this_cpu_read(tasklet_hi_vec.head); __this_cpu_write(tasklet_hi_vec.head, NULL); __this_cpu_write(tasklet_hi_vec.tail, this_cpu_ptr(&tasklet_hi_vec.head)); -- local_irq_enable(); ++ + local_irq_enable(); - while (list) { - struct tasklet_struct *t = list; @@ -19686,8 +19824,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c - } - tasklet_unlock(t); - } -+ local_irq_enable(); - +- - local_irq_disable(); - t->next = NULL; - *__this_cpu_read(tasklet_hi_vec.tail) = t; @@ -19699,7 +19836,7 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c } void tasklet_init(struct tasklet_struct *t, -@@ -586,7 +1083,7 @@ +@@ -586,7 +1083,7 @@ void tasklet_kill(struct tasklet_struct *t) while (test_and_set_bit(TASKLET_STATE_SCHED, &t->state)) { do { @@ -19708,13 +19845,28 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c } while (test_bit(TASKLET_STATE_SCHED, &t->state)); } tasklet_unlock_wait(t); -@@ -660,25 +1157,26 @@ +@@ -660,25 +1157,26 @@ void __init softirq_init(void) open_softirq(HI_SOFTIRQ, tasklet_hi_action); } --static int ksoftirqd_should_run(unsigned int cpu) +#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL) +void tasklet_unlock_wait(struct tasklet_struct *t) ++{ ++ while (test_bit(TASKLET_STATE_RUN, &(t)->state)) { ++ /* ++ * Hack for now to avoid this busy-loop: ++ */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ msleep(1); ++#else ++ barrier(); ++#endif ++ } ++} ++EXPORT_SYMBOL(tasklet_unlock_wait); ++#endif ++ + static int ksoftirqd_should_run(unsigned int cpu) { - return local_softirq_pending(); -} @@ -19723,34 +19875,21 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c -{ - local_irq_disable(); - if (local_softirq_pending()) { -+ while (test_bit(TASKLET_STATE_RUN, &(t)->state)) { - /* +- /* - * We can safely run softirq on inline stack, as we are not deep - * in the task stack here. -+ * Hack for now to avoid this busy-loop: - */ +- */ - __do_softirq(); - local_irq_enable(); - cond_resched_rcu_qs(); - return; -+#ifdef CONFIG_PREEMPT_RT_FULL -+ msleep(1); -+#else -+ barrier(); -+#endif - } +- } - local_irq_enable(); -+} -+EXPORT_SYMBOL(tasklet_unlock_wait); -+#endif -+ -+static int ksoftirqd_should_run(unsigned int cpu) -+{ + return ksoftirqd_softirq_pending(); } #ifdef CONFIG_HOTPLUG_CPU -@@ -745,17 +1243,31 @@ +@@ -745,17 +1243,31 @@ static int takeover_tasklets(unsigned int cpu) static struct smp_hotplug_thread softirq_threads = { .store = &ksoftirqd, @@ -19783,10 +19922,11 @@ diff -Nur linux-4.9.28.orig/kernel/softirq.c linux-4.9.28/kernel/softirq.c return 0; } early_initcall(spawn_ksoftirqd); -diff -Nur linux-4.9.28.orig/kernel/stop_machine.c linux-4.9.28/kernel/stop_machine.c ---- linux-4.9.28.orig/kernel/stop_machine.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/stop_machine.c 2017-05-19 03:37:25.198177419 +0200 -@@ -36,7 +36,7 @@ +diff --git a/kernel/stop_machine.c b/kernel/stop_machine.c +index ec9ab2f01489..8b89dbedeaff 100644 +--- a/kernel/stop_machine.c ++++ b/kernel/stop_machine.c +@@ -36,7 +36,7 @@ struct cpu_stop_done { struct cpu_stopper { struct task_struct *thread; @@ -19795,7 +19935,7 @@ diff -Nur linux-4.9.28.orig/kernel/stop_machine.c linux-4.9.28/kernel/stop_machi bool enabled; /* is this stopper enabled? */ struct list_head works; /* list of pending works */ -@@ -78,14 +78,14 @@ +@@ -78,14 +78,14 @@ static bool cpu_stop_queue_work(unsigned int cpu, struct cpu_stop_work *work) unsigned long flags; bool enabled; @@ -19812,7 +19952,7 @@ diff -Nur linux-4.9.28.orig/kernel/stop_machine.c linux-4.9.28/kernel/stop_machi return enabled; } -@@ -231,8 +231,8 @@ +@@ -231,8 +231,8 @@ static int cpu_stop_queue_two_works(int cpu1, struct cpu_stop_work *work1, struct cpu_stopper *stopper2 = per_cpu_ptr(&cpu_stopper, cpu2); int err; retry: @@ -19823,7 +19963,7 @@ diff -Nur linux-4.9.28.orig/kernel/stop_machine.c linux-4.9.28/kernel/stop_machi err = -ENOENT; if (!stopper1->enabled || !stopper2->enabled) -@@ -255,8 +255,8 @@ +@@ -255,8 +255,8 @@ static int cpu_stop_queue_two_works(int cpu1, struct cpu_stop_work *work1, __cpu_stop_queue_work(stopper1, work1); __cpu_stop_queue_work(stopper2, work2); unlock: @@ -19834,7 +19974,7 @@ diff -Nur linux-4.9.28.orig/kernel/stop_machine.c linux-4.9.28/kernel/stop_machi if (unlikely(err == -EDEADLK)) { while (stop_cpus_in_progress) -@@ -448,9 +448,9 @@ +@@ -448,9 +448,9 @@ static int cpu_stop_should_run(unsigned int cpu) unsigned long flags; int run; @@ -19846,7 +19986,7 @@ diff -Nur linux-4.9.28.orig/kernel/stop_machine.c linux-4.9.28/kernel/stop_machi return run; } -@@ -461,13 +461,13 @@ +@@ -461,13 +461,13 @@ static void cpu_stopper_thread(unsigned int cpu) repeat: work = NULL; @@ -19862,7 +20002,7 @@ diff -Nur linux-4.9.28.orig/kernel/stop_machine.c linux-4.9.28/kernel/stop_machi if (work) { cpu_stop_fn_t fn = work->fn; -@@ -475,6 +475,8 @@ +@@ -475,6 +475,8 @@ static void cpu_stopper_thread(unsigned int cpu) struct cpu_stop_done *done = work->done; int ret; @@ -19871,7 +20011,7 @@ diff -Nur linux-4.9.28.orig/kernel/stop_machine.c linux-4.9.28/kernel/stop_machi /* cpu stop callbacks must not sleep, make in_atomic() == T */ preempt_count_inc(); ret = fn(arg); -@@ -541,7 +543,7 @@ +@@ -541,7 +543,7 @@ static int __init cpu_stop_init(void) for_each_possible_cpu(cpu) { struct cpu_stopper *stopper = &per_cpu(cpu_stopper, cpu); @@ -19880,9 +20020,10 @@ diff -Nur linux-4.9.28.orig/kernel/stop_machine.c linux-4.9.28/kernel/stop_machi INIT_LIST_HEAD(&stopper->works); } -diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtimer.c ---- linux-4.9.28.orig/kernel/time/hrtimer.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/time/hrtimer.c 2017-05-19 03:37:25.198177419 +0200 +diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c +index bb5ec425dfe0..8338b14ed3a3 100644 +--- a/kernel/time/hrtimer.c ++++ b/kernel/time/hrtimer.c @@ -53,6 +53,7 @@ #include @@ -19891,7 +20032,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim #include "tick-internal.h" -@@ -695,6 +696,29 @@ +@@ -695,6 +696,29 @@ static void hrtimer_switch_to_hres(void) retrigger_next_event(NULL); } @@ -19921,7 +20062,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim static void clock_was_set_work(struct work_struct *work) { clock_was_set(); -@@ -710,6 +734,7 @@ +@@ -710,6 +734,7 @@ void clock_was_set_delayed(void) { schedule_work(&hrtimer_work); } @@ -19929,7 +20070,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim #else -@@ -719,11 +744,8 @@ +@@ -719,11 +744,8 @@ static inline int hrtimer_is_hres_enabled(void) { return 0; } static inline void hrtimer_switch_to_hres(void) { } static inline void hrtimer_force_reprogram(struct hrtimer_cpu_base *base, int skip_equal) { } @@ -19943,7 +20084,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim static inline void hrtimer_init_hres(struct hrtimer_cpu_base *base) { } static inline void retrigger_next_event(void *arg) { } -@@ -855,6 +877,32 @@ +@@ -855,6 +877,32 @@ u64 hrtimer_forward(struct hrtimer *timer, ktime_t now, ktime_t interval) } EXPORT_SYMBOL_GPL(hrtimer_forward); @@ -19976,7 +20117,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim /* * enqueue_hrtimer - internal function to (re)start a timer * -@@ -896,6 +944,11 @@ +@@ -896,6 +944,11 @@ static void __remove_hrtimer(struct hrtimer *timer, if (!(state & HRTIMER_STATE_ENQUEUED)) return; @@ -19988,7 +20129,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim if (!timerqueue_del(&base->active, &timer->node)) cpu_base->active_bases &= ~(1 << base->index); -@@ -991,7 +1044,16 @@ +@@ -991,7 +1044,16 @@ void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, new_base = switch_hrtimer_base(timer, base, mode & HRTIMER_MODE_PINNED); timer_stats_hrtimer_set_start_info(timer); @@ -20005,7 +20146,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim leftmost = enqueue_hrtimer(timer, new_base); if (!leftmost) goto unlock; -@@ -1063,7 +1125,7 @@ +@@ -1063,7 +1125,7 @@ int hrtimer_cancel(struct hrtimer *timer) if (ret >= 0) return ret; @@ -20014,7 +20155,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim } } EXPORT_SYMBOL_GPL(hrtimer_cancel); -@@ -1127,6 +1189,7 @@ +@@ -1127,6 +1189,7 @@ static void __hrtimer_init(struct hrtimer *timer, clockid_t clock_id, base = hrtimer_clockid_to_base(clock_id); timer->base = &cpu_base->clock_base[base]; @@ -20022,7 +20163,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim timerqueue_init(&timer->node); #ifdef CONFIG_TIMER_STATS -@@ -1167,6 +1230,7 @@ +@@ -1167,6 +1230,7 @@ bool hrtimer_active(const struct hrtimer *timer) seq = raw_read_seqcount_begin(&cpu_base->seq); if (timer->state != HRTIMER_STATE_INACTIVE || @@ -20030,7 +20171,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim cpu_base->running == timer) return true; -@@ -1265,10 +1329,112 @@ +@@ -1265,10 +1329,112 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, cpu_base->running = NULL; } @@ -20143,7 +20284,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim for (; active; base++, active >>= 1) { struct timerqueue_node *node; -@@ -1284,6 +1450,15 @@ +@@ -1284,6 +1450,15 @@ static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now) timer = container_of(node, struct hrtimer, node); @@ -20159,7 +20300,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim /* * The immediate goal for using the softexpires is * minimizing wakeups, not running timers at the -@@ -1299,9 +1474,14 @@ +@@ -1299,9 +1474,14 @@ static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now) if (basenow.tv64 < hrtimer_get_softexpires_tv64(timer)) break; @@ -20175,7 +20316,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim } #ifdef CONFIG_HIGH_RES_TIMERS -@@ -1464,16 +1644,18 @@ +@@ -1464,16 +1644,18 @@ static enum hrtimer_restart hrtimer_wakeup(struct hrtimer *timer) void hrtimer_init_sleeper(struct hrtimer_sleeper *sl, struct task_struct *task) { sl->timer.function = hrtimer_wakeup; @@ -20196,7 +20337,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim hrtimer_start_expires(&t->timer, mode); if (likely(t->task)) -@@ -1515,7 +1697,8 @@ +@@ -1515,7 +1697,8 @@ long __sched hrtimer_nanosleep_restart(struct restart_block *restart) HRTIMER_MODE_ABS); hrtimer_set_expires_tv64(&t.timer, restart->nanosleep.expires); @@ -20206,7 +20347,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim goto out; rmtp = restart->nanosleep.rmtp; -@@ -1532,8 +1715,10 @@ +@@ -1532,8 +1715,10 @@ long __sched hrtimer_nanosleep_restart(struct restart_block *restart) return ret; } @@ -20219,7 +20360,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim { struct restart_block *restart; struct hrtimer_sleeper t; -@@ -1546,7 +1731,7 @@ +@@ -1546,7 +1731,7 @@ long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, hrtimer_init_on_stack(&t.timer, clockid, mode); hrtimer_set_expires_range_ns(&t.timer, timespec_to_ktime(*rqtp), slack); @@ -20228,7 +20369,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim goto out; /* Absolute timers do not update the rmtp value and restart: */ -@@ -1573,6 +1758,12 @@ +@@ -1573,6 +1758,12 @@ long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, return ret; } @@ -20241,7 +20382,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim SYSCALL_DEFINE2(nanosleep, struct timespec __user *, rqtp, struct timespec __user *, rmtp) { -@@ -1587,6 +1778,26 @@ +@@ -1587,6 +1778,26 @@ SYSCALL_DEFINE2(nanosleep, struct timespec __user *, rqtp, return hrtimer_nanosleep(&tu, rmtp, HRTIMER_MODE_REL, CLOCK_MONOTONIC); } @@ -20268,7 +20409,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim /* * Functions related to boot-time initialization: */ -@@ -1598,10 +1809,14 @@ +@@ -1598,10 +1809,14 @@ int hrtimers_prepare_cpu(unsigned int cpu) for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) { cpu_base->clock_base[i].cpu_base = cpu_base; timerqueue_init_head(&cpu_base->clock_base[i].active); @@ -20283,7 +20424,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim return 0; } -@@ -1671,9 +1886,26 @@ +@@ -1671,9 +1886,26 @@ int hrtimers_dead_cpu(unsigned int scpu) #endif /* CONFIG_HOTPLUG_CPU */ @@ -20310,10 +20451,11 @@ diff -Nur linux-4.9.28.orig/kernel/time/hrtimer.c linux-4.9.28/kernel/time/hrtim } /** -diff -Nur linux-4.9.28.orig/kernel/time/itimer.c linux-4.9.28/kernel/time/itimer.c ---- linux-4.9.28.orig/kernel/time/itimer.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/time/itimer.c 2017-05-19 03:37:25.198177419 +0200 -@@ -213,6 +213,7 @@ +diff --git a/kernel/time/itimer.c b/kernel/time/itimer.c +index 1d5c7204ddc9..184de6751180 100644 +--- a/kernel/time/itimer.c ++++ b/kernel/time/itimer.c +@@ -213,6 +213,7 @@ int do_setitimer(int which, struct itimerval *value, struct itimerval *ovalue) /* We are sharing ->siglock with it_real_fn() */ if (hrtimer_try_to_cancel(timer) < 0) { spin_unlock_irq(&tsk->sighand->siglock); @@ -20321,10 +20463,11 @@ diff -Nur linux-4.9.28.orig/kernel/time/itimer.c linux-4.9.28/kernel/time/itimer goto again; } expires = timeval_to_ktime(value->it_value); -diff -Nur linux-4.9.28.orig/kernel/time/jiffies.c linux-4.9.28/kernel/time/jiffies.c ---- linux-4.9.28.orig/kernel/time/jiffies.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/time/jiffies.c 2017-05-19 03:37:25.198177419 +0200 -@@ -74,7 +74,8 @@ +diff --git a/kernel/time/jiffies.c b/kernel/time/jiffies.c +index 555e21f7b966..a5d6435fabbb 100644 +--- a/kernel/time/jiffies.c ++++ b/kernel/time/jiffies.c +@@ -74,7 +74,8 @@ static struct clocksource clocksource_jiffies = { .max_cycles = 10, }; @@ -20334,7 +20477,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/jiffies.c linux-4.9.28/kernel/time/jiffi #if (BITS_PER_LONG < 64) u64 get_jiffies_64(void) -@@ -83,9 +84,9 @@ +@@ -83,9 +84,9 @@ u64 get_jiffies_64(void) u64 ret; do { @@ -20346,9 +20489,10 @@ diff -Nur linux-4.9.28.orig/kernel/time/jiffies.c linux-4.9.28/kernel/time/jiffi return ret; } EXPORT_SYMBOL(get_jiffies_64); -diff -Nur linux-4.9.28.orig/kernel/time/ntp.c linux-4.9.28/kernel/time/ntp.c ---- linux-4.9.28.orig/kernel/time/ntp.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/time/ntp.c 2017-05-19 03:37:25.198177419 +0200 +diff --git a/kernel/time/ntp.c b/kernel/time/ntp.c +index 6df8927c58a5..05b7391bf9bd 100644 +--- a/kernel/time/ntp.c ++++ b/kernel/time/ntp.c @@ -17,6 +17,7 @@ #include #include @@ -20357,7 +20501,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/ntp.c linux-4.9.28/kernel/time/ntp.c #include "ntp_internal.h" #include "timekeeping_internal.h" -@@ -568,10 +569,35 @@ +@@ -568,10 +569,35 @@ static void sync_cmos_clock(struct work_struct *work) &sync_cmos_work, timespec64_to_jiffies(&next)); } @@ -20393,9 +20537,10 @@ diff -Nur linux-4.9.28.orig/kernel/time/ntp.c linux-4.9.28/kernel/time/ntp.c #else void ntp_notify_cmos_timer(void) { } -diff -Nur linux-4.9.28.orig/kernel/time/posix-cpu-timers.c linux-4.9.28/kernel/time/posix-cpu-timers.c ---- linux-4.9.28.orig/kernel/time/posix-cpu-timers.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/time/posix-cpu-timers.c 2017-05-19 03:37:25.198177419 +0200 +diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c +index 39008d78927a..633f4eaca9e7 100644 +--- a/kernel/time/posix-cpu-timers.c ++++ b/kernel/time/posix-cpu-timers.c @@ -3,6 +3,7 @@ */ @@ -20404,7 +20549,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/posix-cpu-timers.c linux-4.9.28/kernel/t #include #include #include -@@ -620,7 +621,7 @@ +@@ -620,7 +621,7 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int timer_flags, /* * Disarm any old timer after extracting its expiry time. */ @@ -20413,7 +20558,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/posix-cpu-timers.c linux-4.9.28/kernel/t ret = 0; old_incr = timer->it.cpu.incr; -@@ -1064,7 +1065,7 @@ +@@ -1064,7 +1065,7 @@ void posix_cpu_timer_schedule(struct k_itimer *timer) /* * Now re-arm for the new expiry time. */ @@ -20422,7 +20567,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/posix-cpu-timers.c linux-4.9.28/kernel/t arm_timer(timer); unlock_task_sighand(p, &flags); -@@ -1153,13 +1154,13 @@ +@@ -1153,13 +1154,13 @@ static inline int fastpath_timer_check(struct task_struct *tsk) * already updated our counts. We need to check if any timers fire now. * Interrupts are disabled. */ @@ -20438,7 +20583,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/posix-cpu-timers.c linux-4.9.28/kernel/t /* * The fast path checks that there are no expired thread or thread -@@ -1213,6 +1214,190 @@ +@@ -1213,6 +1214,190 @@ void run_posix_cpu_timers(struct task_struct *tsk) } } @@ -20629,10 +20774,11 @@ diff -Nur linux-4.9.28.orig/kernel/time/posix-cpu-timers.c linux-4.9.28/kernel/t /* * Set one of the process-wide special case CPU timers or RLIMIT_CPU. * The tsk->sighand->siglock must be held by the caller. -diff -Nur linux-4.9.28.orig/kernel/time/posix-timers.c linux-4.9.28/kernel/time/posix-timers.c ---- linux-4.9.28.orig/kernel/time/posix-timers.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/time/posix-timers.c 2017-05-19 03:37:25.198177419 +0200 -@@ -506,6 +506,7 @@ +diff --git a/kernel/time/posix-timers.c b/kernel/time/posix-timers.c +index f2826c35e918..464a98155a0e 100644 +--- a/kernel/time/posix-timers.c ++++ b/kernel/time/posix-timers.c +@@ -506,6 +506,7 @@ static enum hrtimer_restart posix_timer_fn(struct hrtimer *timer) static struct pid *good_sigevent(sigevent_t * event) { struct task_struct *rtn = current->group_leader; @@ -20640,7 +20786,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/posix-timers.c linux-4.9.28/kernel/time/ if ((event->sigev_notify & SIGEV_THREAD_ID ) && (!(rtn = find_task_by_vpid(event->sigev_notify_thread_id)) || -@@ -514,7 +515,8 @@ +@@ -514,7 +515,8 @@ static struct pid *good_sigevent(sigevent_t * event) return NULL; if (((event->sigev_notify & ~SIGEV_THREAD_ID) != SIGEV_NONE) && @@ -20650,7 +20796,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/posix-timers.c linux-4.9.28/kernel/time/ return NULL; return task_pid(rtn); -@@ -826,6 +828,20 @@ +@@ -826,6 +828,20 @@ SYSCALL_DEFINE1(timer_getoverrun, timer_t, timer_id) return overrun; } @@ -20671,7 +20817,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/posix-timers.c linux-4.9.28/kernel/time/ /* Set a POSIX.1b interval timer. */ /* timr->it_lock is taken. */ static int -@@ -903,6 +919,7 @@ +@@ -903,6 +919,7 @@ SYSCALL_DEFINE4(timer_settime, timer_t, timer_id, int, flags, if (!timr) return -EINVAL; @@ -20679,7 +20825,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/posix-timers.c linux-4.9.28/kernel/time/ kc = clockid_to_kclock(timr->it_clock); if (WARN_ON_ONCE(!kc || !kc->timer_set)) error = -EINVAL; -@@ -911,9 +928,12 @@ +@@ -911,9 +928,12 @@ SYSCALL_DEFINE4(timer_settime, timer_t, timer_id, int, flags, unlock_timer(timr, flag); if (error == TIMER_RETRY) { @@ -20692,7 +20838,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/posix-timers.c linux-4.9.28/kernel/time/ if (old_setting && !error && copy_to_user(old_setting, &old_spec, sizeof (old_spec))) -@@ -951,10 +971,15 @@ +@@ -951,10 +971,15 @@ SYSCALL_DEFINE1(timer_delete, timer_t, timer_id) if (!timer) return -EINVAL; @@ -20708,39 +20854,42 @@ diff -Nur linux-4.9.28.orig/kernel/time/posix-timers.c linux-4.9.28/kernel/time/ spin_lock(¤t->sighand->siglock); list_del(&timer->list); -@@ -980,8 +1005,18 @@ +@@ -980,8 +1005,18 @@ static void itimer_delete(struct k_itimer *timer) retry_delete: spin_lock_irqsave(&timer->it_lock, flags); +- if (timer_delete_hook(timer) == TIMER_RETRY) { + /* On RT we can race with a deletion */ + if (!timer->it_signal) { -+ unlock_timer(timer, flags); + unlock_timer(timer, flags); + return; + } + - if (timer_delete_hook(timer) == TIMER_RETRY) { ++ if (timer_delete_hook(timer) == TIMER_RETRY) { + rcu_read_lock(); - unlock_timer(timer, flags); ++ unlock_timer(timer, flags); + timer_wait_for_callback(clockid_to_kclock(timer->it_clock), + timer); + rcu_read_unlock(); goto retry_delete; } list_del(&timer->list); -diff -Nur linux-4.9.28.orig/kernel/time/tick-broadcast-hrtimer.c linux-4.9.28/kernel/time/tick-broadcast-hrtimer.c ---- linux-4.9.28.orig/kernel/time/tick-broadcast-hrtimer.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/time/tick-broadcast-hrtimer.c 2017-05-19 03:37:25.198177419 +0200 -@@ -107,5 +107,6 @@ +diff --git a/kernel/time/tick-broadcast-hrtimer.c b/kernel/time/tick-broadcast-hrtimer.c +index 690b797f522e..fe8ba1619879 100644 +--- a/kernel/time/tick-broadcast-hrtimer.c ++++ b/kernel/time/tick-broadcast-hrtimer.c +@@ -107,5 +107,6 @@ void tick_setup_hrtimer_broadcast(void) { hrtimer_init(&bctimer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); bctimer.function = bc_handler; + bctimer.irqsafe = true; clockevents_register_device(&ce_broadcast_hrtimer); } -diff -Nur linux-4.9.28.orig/kernel/time/tick-common.c linux-4.9.28/kernel/time/tick-common.c ---- linux-4.9.28.orig/kernel/time/tick-common.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/time/tick-common.c 2017-05-19 03:37:25.198177419 +0200 -@@ -79,13 +79,15 @@ +diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c +index 4fcd99e12aa0..5a47f2e98faf 100644 +--- a/kernel/time/tick-common.c ++++ b/kernel/time/tick-common.c +@@ -79,13 +79,15 @@ int tick_is_oneshot_available(void) static void tick_periodic(int cpu) { if (tick_do_timer_cpu == cpu) { @@ -20758,7 +20907,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/tick-common.c linux-4.9.28/kernel/time/t update_wall_time(); } -@@ -157,9 +159,9 @@ +@@ -157,9 +159,9 @@ void tick_setup_periodic(struct clock_event_device *dev, int broadcast) ktime_t next; do { @@ -20770,10 +20919,11 @@ diff -Nur linux-4.9.28.orig/kernel/time/tick-common.c linux-4.9.28/kernel/time/t clockevents_switch_state(dev, CLOCK_EVT_STATE_ONESHOT); -diff -Nur linux-4.9.28.orig/kernel/time/tick-sched.c linux-4.9.28/kernel/time/tick-sched.c ---- linux-4.9.28.orig/kernel/time/tick-sched.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/time/tick-sched.c 2017-05-19 03:37:25.198177419 +0200 -@@ -62,7 +62,8 @@ +diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c +index 3bcb61b52f6c..66d85482a96e 100644 +--- a/kernel/time/tick-sched.c ++++ b/kernel/time/tick-sched.c +@@ -62,7 +62,8 @@ static void tick_do_update_jiffies64(ktime_t now) return; /* Reevaluate with jiffies_lock held */ @@ -20783,7 +20933,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/tick-sched.c linux-4.9.28/kernel/time/ti delta = ktime_sub(now, last_jiffies_update); if (delta.tv64 >= tick_period.tv64) { -@@ -85,10 +86,12 @@ +@@ -85,10 +86,12 @@ static void tick_do_update_jiffies64(ktime_t now) /* Keep the tick_next_period variable up to date */ tick_next_period = ktime_add(last_jiffies_update, tick_period); } else { @@ -20798,7 +20948,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/tick-sched.c linux-4.9.28/kernel/time/ti update_wall_time(); } -@@ -99,12 +102,14 @@ +@@ -99,12 +102,14 @@ static ktime_t tick_init_jiffy_update(void) { ktime_t period; @@ -20815,7 +20965,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/tick-sched.c linux-4.9.28/kernel/time/ti return period; } -@@ -215,6 +220,7 @@ +@@ -215,6 +220,7 @@ static void nohz_full_kick_func(struct irq_work *work) static DEFINE_PER_CPU(struct irq_work, nohz_full_kick_work) = { .func = nohz_full_kick_func, @@ -20823,7 +20973,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/tick-sched.c linux-4.9.28/kernel/time/ti }; /* -@@ -673,10 +679,10 @@ +@@ -673,10 +679,10 @@ static ktime_t tick_nohz_stop_sched_tick(struct tick_sched *ts, /* Read jiffies and the time when jiffies were updated last */ do { @@ -20836,7 +20986,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/tick-sched.c linux-4.9.28/kernel/time/ti ts->last_jiffies = basejiff; if (rcu_needs_cpu(basemono, &next_rcu) || -@@ -877,14 +883,7 @@ +@@ -877,14 +883,7 @@ static bool can_stop_idle_tick(int cpu, struct tick_sched *ts) return false; if (unlikely(local_softirq_pending() && cpu_online(cpu))) { @@ -20852,7 +21002,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/tick-sched.c linux-4.9.28/kernel/time/ti return false; } -@@ -1193,6 +1192,7 @@ +@@ -1193,6 +1192,7 @@ void tick_setup_sched_timer(void) * Emulate tick processing via per-CPU hrtimers: */ hrtimer_init(&ts->sched_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); @@ -20860,10 +21010,11 @@ diff -Nur linux-4.9.28.orig/kernel/time/tick-sched.c linux-4.9.28/kernel/time/ti ts->sched_timer.function = tick_sched_timer; /* Get the next period (per-CPU) */ -diff -Nur linux-4.9.28.orig/kernel/time/timekeeping.c linux-4.9.28/kernel/time/timekeeping.c ---- linux-4.9.28.orig/kernel/time/timekeeping.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/time/timekeeping.c 2017-05-19 03:37:25.198177419 +0200 -@@ -2328,8 +2328,10 @@ +diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c +index 46e312e9be38..fa75cf5d9253 100644 +--- a/kernel/time/timekeeping.c ++++ b/kernel/time/timekeeping.c +@@ -2328,8 +2328,10 @@ EXPORT_SYMBOL(hardpps); */ void xtime_update(unsigned long ticks) { @@ -20876,10 +21027,11 @@ diff -Nur linux-4.9.28.orig/kernel/time/timekeeping.c linux-4.9.28/kernel/time/t + raw_spin_unlock(&jiffies_lock); update_wall_time(); } -diff -Nur linux-4.9.28.orig/kernel/time/timekeeping.h linux-4.9.28/kernel/time/timekeeping.h ---- linux-4.9.28.orig/kernel/time/timekeeping.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/time/timekeeping.h 2017-05-19 03:37:25.202177573 +0200 -@@ -19,7 +19,8 @@ +diff --git a/kernel/time/timekeeping.h b/kernel/time/timekeeping.h +index 704f595ce83f..763a3e5121ff 100644 +--- a/kernel/time/timekeeping.h ++++ b/kernel/time/timekeeping.h +@@ -19,7 +19,8 @@ extern void timekeeping_resume(void); extern void do_timer(unsigned long ticks); extern void update_wall_time(void); @@ -20889,10 +21041,11 @@ diff -Nur linux-4.9.28.orig/kernel/time/timekeeping.h linux-4.9.28/kernel/time/t #define CS_NAME_LEN 32 -diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c ---- linux-4.9.28.orig/kernel/time/timer.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/time/timer.c 2017-05-19 03:37:25.202177573 +0200 -@@ -193,8 +193,11 @@ +diff --git a/kernel/time/timer.c b/kernel/time/timer.c +index c611c47de884..08a5ab762495 100644 +--- a/kernel/time/timer.c ++++ b/kernel/time/timer.c +@@ -193,8 +193,11 @@ EXPORT_SYMBOL(jiffies_64); #endif struct timer_base { @@ -20905,16 +21058,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c unsigned long clk; unsigned long next_expiry; unsigned int cpu; -@@ -203,6 +206,8 @@ - bool is_idle; - DECLARE_BITMAP(pending_map, WHEEL_SIZE); - struct hlist_head vectors[WHEEL_SIZE]; -+ struct hlist_head expired_lists[LVL_DEPTH]; -+ int expired_count; - } ____cacheline_aligned; - - static DEFINE_PER_CPU(struct timer_base, timer_bases[NR_BASES]); -@@ -948,10 +953,10 @@ +@@ -948,10 +951,10 @@ static struct timer_base *lock_timer_base(struct timer_list *timer, if (!(tf & TIMER_MIGRATING)) { base = get_timer_base(tf); @@ -20927,7 +21071,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c } cpu_relax(); } -@@ -1023,9 +1028,9 @@ +@@ -1023,9 +1026,9 @@ __mod_timer(struct timer_list *timer, unsigned long expires, bool pending_only) /* See the comment in lock_timer_base() */ timer->flags |= TIMER_MIGRATING; @@ -20939,7 +21083,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c WRITE_ONCE(timer->flags, (timer->flags & ~TIMER_BASEMASK) | base->cpu); } -@@ -1050,7 +1055,7 @@ +@@ -1050,7 +1053,7 @@ __mod_timer(struct timer_list *timer, unsigned long expires, bool pending_only) } out_unlock: @@ -20948,7 +21092,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c return ret; } -@@ -1144,19 +1149,46 @@ +@@ -1144,19 +1147,46 @@ void add_timer_on(struct timer_list *timer, int cpu) if (base != new_base) { timer->flags |= TIMER_MIGRATING; @@ -20998,7 +21142,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c /** * del_timer - deactive a timer. * @timer: the timer to be deactivated -@@ -1180,7 +1212,7 @@ +@@ -1180,7 +1210,7 @@ int del_timer(struct timer_list *timer) if (timer_pending(timer)) { base = lock_timer_base(timer, &flags); ret = detach_if_pending(timer, base, true); @@ -21007,7 +21151,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c } return ret; -@@ -1208,13 +1240,13 @@ +@@ -1208,13 +1238,13 @@ int try_to_del_timer_sync(struct timer_list *timer) timer_stats_timer_clear_start_info(timer); ret = detach_if_pending(timer, base, true); } @@ -21023,7 +21167,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c /** * del_timer_sync - deactivate a timer and wait for the handler to finish. * @timer: the timer to be deactivated -@@ -1274,7 +1306,7 @@ +@@ -1274,7 +1304,7 @@ int del_timer_sync(struct timer_list *timer) int ret = try_to_del_timer_sync(timer); if (ret >= 0) return ret; @@ -21032,17 +21176,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c } } EXPORT_SYMBOL(del_timer_sync); -@@ -1323,7 +1355,8 @@ - } - } - --static void expire_timers(struct timer_base *base, struct hlist_head *head) -+static inline void __expire_timers(struct timer_base *base, -+ struct hlist_head *head) - { - while (!hlist_empty(head)) { - struct timer_list *timer; -@@ -1339,33 +1372,53 @@ +@@ -1339,14 +1369,17 @@ static void expire_timers(struct timer_base *base, struct hlist_head *head) fn = timer->function; data = timer->data; @@ -21065,56 +21199,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c } } } - --static int __collect_expired_timers(struct timer_base *base, -- struct hlist_head *heads) -+static void expire_timers(struct timer_base *base) -+{ -+ struct hlist_head *head; -+ -+ while (base->expired_count--) { -+ head = base->expired_lists + base->expired_count; -+ __expire_timers(base, head); -+ } -+ base->expired_count = 0; -+} -+ -+static void __collect_expired_timers(struct timer_base *base) - { - unsigned long clk = base->clk; - struct hlist_head *vec; -- int i, levels = 0; -+ int i; - unsigned int idx; - -+ /* -+ * expire_timers() must be called at least once before we can -+ * collect more timers -+ */ -+ if (WARN_ON(base->expired_count)) -+ return; -+ - for (i = 0; i < LVL_DEPTH; i++) { - idx = (clk & LVL_MASK) + i * LVL_SIZE; - - if (__test_and_clear_bit(idx, base->pending_map)) { - vec = base->vectors + idx; -- hlist_move_list(vec, heads++); -- levels++; -+ hlist_move_list(vec, -+ &base->expired_lists[base->expired_count++]); - } - /* Is it time to look at the next level? */ - if (clk & LVL_CLK_MASK) -@@ -1373,7 +1426,6 @@ - /* Shift clock for the next level granularity */ - clk >>= LVL_CLK_SHIFT; - } -- return levels; - } - - #ifdef CONFIG_NO_HZ_COMMON -@@ -1515,7 +1567,7 @@ +@@ -1515,7 +1548,7 @@ u64 get_next_timer_interrupt(unsigned long basej, u64 basem) if (cpu_is_offline(smp_processor_id())) return expires; @@ -21123,7 +21208,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c nextevt = __next_timer_interrupt(base); is_max_delta = (nextevt == base->clk + NEXT_TIMER_MAX_DELTA); base->next_expiry = nextevt; -@@ -1543,7 +1595,7 @@ +@@ -1543,7 +1576,7 @@ u64 get_next_timer_interrupt(unsigned long basej, u64 basem) if ((expires - basem) > TICK_NSEC) base->is_idle = true; } @@ -21132,72 +21217,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c return cmp_next_hrtimer_event(basem, expires); } -@@ -1566,8 +1618,7 @@ - base->is_idle = false; - } - --static int collect_expired_timers(struct timer_base *base, -- struct hlist_head *heads) -+static void collect_expired_timers(struct timer_base *base) - { - /* - * NOHZ optimization. After a long idle sleep we need to forward the -@@ -1584,20 +1635,49 @@ - if (time_after(next, jiffies)) { - /* The call site will increment clock! */ - base->clk = jiffies - 1; -- return 0; -+ return; - } - base->clk = next; - } -- return __collect_expired_timers(base, heads); -+ __collect_expired_timers(base); - } - #else --static inline int collect_expired_timers(struct timer_base *base, -- struct hlist_head *heads) -+static inline void collect_expired_timers(struct timer_base *base) - { -- return __collect_expired_timers(base, heads); -+ __collect_expired_timers(base); - } - #endif - -+static int find_expired_timers(struct timer_base *base) -+{ -+ const unsigned long int end_clk = jiffies; -+ -+ while (!base->expired_count && time_after_eq(end_clk, base->clk)) { -+ collect_expired_timers(base); -+ base->clk++; -+ } -+ -+ return base->expired_count; -+} -+ -+/* Called from CPU tick routine to quickly collect expired timers */ -+static int tick_find_expired(struct timer_base *base) -+{ -+ int count; -+ -+ raw_spin_lock(&base->lock); -+ -+ if (unlikely(time_after(jiffies, base->clk + HZ))) { -+ /* defer to ktimersoftd; don't spend too long in irq context */ -+ count = -1; -+ } else -+ count = find_expired_timers(base); -+ -+ raw_spin_unlock(&base->lock); -+ -+ return count; -+} -+ - /* - * Called from the timer interrupt handler to charge one tick to the current - * process. user_tick is 1 if the tick is user time, 0 for system. -@@ -1608,13 +1688,13 @@ +@@ -1608,13 +1641,13 @@ void update_process_times(int user_tick) /* Note: this timer irq context must be accounted for as well. */ account_process_tick(p, user_tick); @@ -21213,29 +21233,19 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c run_posix_cpu_timers(p); } -@@ -1624,24 +1704,13 @@ - */ - static inline void __run_timers(struct timer_base *base) - { -- struct hlist_head heads[LVL_DEPTH]; -- int levels; -- -- if (!time_after_eq(jiffies, base->clk)) -- return; -+ raw_spin_lock_irq(&base->lock); +@@ -1630,7 +1663,7 @@ static inline void __run_timers(struct timer_base *base) + if (!time_after_eq(jiffies, base->clk)) + return; - spin_lock_irq(&base->lock); -+ while (find_expired_timers(base)) -+ expire_timers(base); ++ raw_spin_lock_irq(&base->lock); -- while (time_after_eq(jiffies, base->clk)) { -- -- levels = collect_expired_timers(base, heads); -- base->clk++; -- -- while (levels--) -- expire_timers(base, heads + levels); -- } + while (time_after_eq(jiffies, base->clk)) { + +@@ -1640,8 +1673,8 @@ static inline void __run_timers(struct timer_base *base) + while (levels--) + expire_timers(base, heads + levels); + } - base->running_timer = NULL; - spin_unlock_irq(&base->lock); + raw_spin_unlock_irq(&base->lock); @@ -21243,7 +21253,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c } /* -@@ -1651,6 +1720,8 @@ +@@ -1651,6 +1684,8 @@ static __latent_entropy void run_timer_softirq(struct softirq_action *h) { struct timer_base *base = this_cpu_ptr(&timer_bases[BASE_STD]); @@ -21252,22 +21262,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c __run_timers(base); if (IS_ENABLED(CONFIG_NO_HZ_COMMON) && base->nohz_active) __run_timers(this_cpu_ptr(&timer_bases[BASE_DEF])); -@@ -1665,12 +1736,12 @@ - - hrtimer_run_queues(); - /* Raise the softirq only if required. */ -- if (time_before(jiffies, base->clk)) { -+ if (time_before(jiffies, base->clk) || !tick_find_expired(base)) { - if (!IS_ENABLED(CONFIG_NO_HZ_COMMON) || !base->nohz_active) - return; - /* CPU is awake, so check the deferrable base. */ - base++; -- if (time_before(jiffies, base->clk)) -+ if (time_before(jiffies, base->clk) || !tick_find_expired(base)) - return; - } - raise_softirq(TIMER_SOFTIRQ); -@@ -1836,16 +1907,17 @@ +@@ -1836,16 +1871,16 @@ int timers_dead_cpu(unsigned int cpu) * The caller is globally serialized and nobody else * takes two locks at once, deadlock is not possible. */ @@ -21277,7 +21272,6 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c + raw_spin_lock_nested(&old_base->lock, SINGLE_DEPTH_NESTING); BUG_ON(old_base->running_timer); -+ BUG_ON(old_base->expired_count); for (i = 0; i < WHEEL_SIZE; i++) migrate_timer_list(new_base, old_base->vectors + i); @@ -21289,7 +21283,7 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c put_cpu_ptr(&timer_bases); } return 0; -@@ -1861,8 +1933,12 @@ +@@ -1861,8 +1896,11 @@ static void __init init_timer_cpu(int cpu) for (i = 0; i < NR_BASES; i++) { base = per_cpu_ptr(&timer_bases[i], cpu); base->cpu = cpu; @@ -21299,14 +21293,14 @@ diff -Nur linux-4.9.28.orig/kernel/time/timer.c linux-4.9.28/kernel/time/timer.c +#ifdef CONFIG_PREEMPT_RT_FULL + init_swait_queue_head(&base->wait_for_running_timer); +#endif -+ base->expired_count = 0; } } -diff -Nur linux-4.9.28.orig/kernel/trace/Kconfig linux-4.9.28/kernel/trace/Kconfig ---- linux-4.9.28.orig/kernel/trace/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/trace/Kconfig 2017-05-19 03:37:25.202177573 +0200 -@@ -182,6 +182,24 @@ +diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig +index 2a96b063d659..812e37237eb8 100644 +--- a/kernel/trace/Kconfig ++++ b/kernel/trace/Kconfig +@@ -182,6 +182,24 @@ config IRQSOFF_TRACER enabled. This option and the preempt-off timing option can be used together or separately.) @@ -21331,7 +21325,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/Kconfig linux-4.9.28/kernel/trace/Kconf config PREEMPT_TRACER bool "Preemption-off Latency Tracer" default n -@@ -206,6 +224,24 @@ +@@ -206,6 +224,24 @@ config PREEMPT_TRACER enabled. This option and the irqs-off timing option can be used together or separately.) @@ -21356,7 +21350,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/Kconfig linux-4.9.28/kernel/trace/Kconf config SCHED_TRACER bool "Scheduling Latency Tracer" select GENERIC_TRACER -@@ -251,6 +287,74 @@ +@@ -251,6 +287,74 @@ config HWLAT_TRACER file. Every time a latency is greater than tracing_thresh, it will be recorded into the ring buffer. @@ -21431,9 +21425,26 @@ diff -Nur linux-4.9.28.orig/kernel/trace/Kconfig linux-4.9.28/kernel/trace/Kconf config ENABLE_DEFAULT_TRACERS bool "Trace process context switches and events" depends on !GENERIC_TRACER -diff -Nur linux-4.9.28.orig/kernel/trace/latency_hist.c linux-4.9.28/kernel/trace/latency_hist.c ---- linux-4.9.28.orig/kernel/trace/latency_hist.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux-4.9.28/kernel/trace/latency_hist.c 2017-05-19 03:37:25.202177573 +0200 +diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile +index e57980845549..83af000b783c 100644 +--- a/kernel/trace/Makefile ++++ b/kernel/trace/Makefile +@@ -38,6 +38,10 @@ obj-$(CONFIG_IRQSOFF_TRACER) += trace_irqsoff.o + obj-$(CONFIG_PREEMPT_TRACER) += trace_irqsoff.o + obj-$(CONFIG_SCHED_TRACER) += trace_sched_wakeup.o + obj-$(CONFIG_HWLAT_TRACER) += trace_hwlat.o ++obj-$(CONFIG_INTERRUPT_OFF_HIST) += latency_hist.o ++obj-$(CONFIG_PREEMPT_OFF_HIST) += latency_hist.o ++obj-$(CONFIG_WAKEUP_LATENCY_HIST) += latency_hist.o ++obj-$(CONFIG_MISSED_TIMER_OFFSETS_HIST) += latency_hist.o + obj-$(CONFIG_NOP_TRACER) += trace_nop.o + obj-$(CONFIG_STACK_TRACER) += trace_stack.o + obj-$(CONFIG_MMIOTRACE) += trace_mmiotrace.o +diff --git a/kernel/trace/latency_hist.c b/kernel/trace/latency_hist.c +new file mode 100644 +index 000000000000..7f6ee70dea41 +--- /dev/null ++++ b/kernel/trace/latency_hist.c @@ -0,0 +1,1178 @@ +/* + * kernel/trace/latency_hist.c @@ -22613,24 +22624,11 @@ diff -Nur linux-4.9.28.orig/kernel/trace/latency_hist.c linux-4.9.28/kernel/trac +} + +device_initcall(latency_hist_init); -diff -Nur linux-4.9.28.orig/kernel/trace/Makefile linux-4.9.28/kernel/trace/Makefile ---- linux-4.9.28.orig/kernel/trace/Makefile 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/trace/Makefile 2017-05-19 03:37:25.202177573 +0200 -@@ -38,6 +38,10 @@ - obj-$(CONFIG_PREEMPT_TRACER) += trace_irqsoff.o - obj-$(CONFIG_SCHED_TRACER) += trace_sched_wakeup.o - obj-$(CONFIG_HWLAT_TRACER) += trace_hwlat.o -+obj-$(CONFIG_INTERRUPT_OFF_HIST) += latency_hist.o -+obj-$(CONFIG_PREEMPT_OFF_HIST) += latency_hist.o -+obj-$(CONFIG_WAKEUP_LATENCY_HIST) += latency_hist.o -+obj-$(CONFIG_MISSED_TIMER_OFFSETS_HIST) += latency_hist.o - obj-$(CONFIG_NOP_TRACER) += trace_nop.o - obj-$(CONFIG_STACK_TRACER) += trace_stack.o - obj-$(CONFIG_MMIOTRACE) += trace_mmiotrace.o -diff -Nur linux-4.9.28.orig/kernel/trace/trace.c linux-4.9.28/kernel/trace/trace.c ---- linux-4.9.28.orig/kernel/trace/trace.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/trace/trace.c 2017-05-19 03:37:25.202177573 +0200 -@@ -1897,6 +1897,7 @@ +diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c +index 83c60f9013cb..6fb207964a84 100644 +--- a/kernel/trace/trace.c ++++ b/kernel/trace/trace.c +@@ -1897,6 +1897,7 @@ tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags, struct task_struct *tsk = current; entry->preempt_count = pc & 0xff; @@ -22638,7 +22636,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace.c linux-4.9.28/kernel/trace/trace entry->pid = (tsk) ? tsk->pid : 0; entry->flags = #ifdef CONFIG_TRACE_IRQFLAGS_SUPPORT -@@ -1907,8 +1908,11 @@ +@@ -1907,8 +1908,11 @@ tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags, ((pc & NMI_MASK ) ? TRACE_FLAG_NMI : 0) | ((pc & HARDIRQ_MASK) ? TRACE_FLAG_HARDIRQ : 0) | ((pc & SOFTIRQ_MASK) ? TRACE_FLAG_SOFTIRQ : 0) | @@ -22651,7 +22649,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace.c linux-4.9.28/kernel/trace/trace } EXPORT_SYMBOL_GPL(tracing_generic_entry_update); -@@ -2892,14 +2896,17 @@ +@@ -2892,14 +2896,17 @@ get_total_entries(struct trace_buffer *buf, static void print_lat_help_header(struct seq_file *m) { @@ -22677,7 +22675,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace.c linux-4.9.28/kernel/trace/trace } static void print_event_info(struct trace_buffer *buf, struct seq_file *m) -@@ -2925,11 +2932,14 @@ +@@ -2925,11 +2932,14 @@ static void print_func_help_header_irq(struct trace_buffer *buf, struct seq_file print_event_info(buf, m); seq_puts(m, "# _-----=> irqs-off\n" "# / _----=> need-resched\n" @@ -22697,22 +22695,11 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace.c linux-4.9.28/kernel/trace/trace } void -diff -Nur linux-4.9.28.orig/kernel/trace/trace_events.c linux-4.9.28/kernel/trace/trace_events.c ---- linux-4.9.28.orig/kernel/trace/trace_events.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/trace/trace_events.c 2017-05-19 03:37:25.202177573 +0200 -@@ -187,6 +187,8 @@ - __common_field(unsigned char, flags); - __common_field(unsigned char, preempt_count); - __common_field(int, pid); -+ __common_field(unsigned short, migrate_disable); -+ __common_field(unsigned short, padding); - - return ret; - } -diff -Nur linux-4.9.28.orig/kernel/trace/trace.h linux-4.9.28/kernel/trace/trace.h ---- linux-4.9.28.orig/kernel/trace/trace.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/trace/trace.h 2017-05-19 03:37:25.202177573 +0200 -@@ -124,6 +124,7 @@ +diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h +index b0d8576c27ae..702b9376b278 100644 +--- a/kernel/trace/trace.h ++++ b/kernel/trace/trace.h +@@ -124,6 +124,7 @@ struct kretprobe_trace_entry_head { * NEED_RESCHED - reschedule is requested * HARDIRQ - inside an interrupt handler * SOFTIRQ - inside a softirq handler @@ -22720,7 +22707,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace.h linux-4.9.28/kernel/trace/trace */ enum trace_flag_type { TRACE_FLAG_IRQS_OFF = 0x01, -@@ -133,6 +134,7 @@ +@@ -133,6 +134,7 @@ enum trace_flag_type { TRACE_FLAG_SOFTIRQ = 0x10, TRACE_FLAG_PREEMPT_RESCHED = 0x20, TRACE_FLAG_NMI = 0x40, @@ -22728,9 +22715,23 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace.h linux-4.9.28/kernel/trace/trace }; #define TRACE_BUF_SIZE 1024 -diff -Nur linux-4.9.28.orig/kernel/trace/trace_irqsoff.c linux-4.9.28/kernel/trace/trace_irqsoff.c ---- linux-4.9.28.orig/kernel/trace/trace_irqsoff.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/trace/trace_irqsoff.c 2017-05-19 03:37:25.202177573 +0200 +diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c +index 03c0a48c3ac4..0b85d516b491 100644 +--- a/kernel/trace/trace_events.c ++++ b/kernel/trace/trace_events.c +@@ -187,6 +187,8 @@ static int trace_define_common_fields(void) + __common_field(unsigned char, flags); + __common_field(unsigned char, preempt_count); + __common_field(int, pid); ++ __common_field(unsigned short, migrate_disable); ++ __common_field(unsigned short, padding); + + return ret; + } +diff --git a/kernel/trace/trace_irqsoff.c b/kernel/trace/trace_irqsoff.c +index 03cdff84d026..940bd10b4406 100644 +--- a/kernel/trace/trace_irqsoff.c ++++ b/kernel/trace/trace_irqsoff.c @@ -13,6 +13,7 @@ #include #include @@ -22739,7 +22740,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace_irqsoff.c linux-4.9.28/kernel/tra #include "trace.h" -@@ -424,11 +425,13 @@ +@@ -424,11 +425,13 @@ void start_critical_timings(void) { if (preempt_trace() || irq_trace()) start_critical_timing(CALLER_ADDR0, CALLER_ADDR1); @@ -22753,7 +22754,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace_irqsoff.c linux-4.9.28/kernel/tra if (preempt_trace() || irq_trace()) stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1); } -@@ -438,6 +441,7 @@ +@@ -438,6 +441,7 @@ EXPORT_SYMBOL_GPL(stop_critical_timings); #ifdef CONFIG_PROVE_LOCKING void time_hardirqs_on(unsigned long a0, unsigned long a1) { @@ -22761,7 +22762,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace_irqsoff.c linux-4.9.28/kernel/tra if (!preempt_trace() && irq_trace()) stop_critical_timing(a0, a1); } -@@ -446,6 +450,7 @@ +@@ -446,6 +450,7 @@ void time_hardirqs_off(unsigned long a0, unsigned long a1) { if (!preempt_trace() && irq_trace()) start_critical_timing(a0, a1); @@ -22769,7 +22770,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace_irqsoff.c linux-4.9.28/kernel/tra } #else /* !CONFIG_PROVE_LOCKING */ -@@ -471,6 +476,7 @@ +@@ -471,6 +476,7 @@ inline void print_irqtrace_events(struct task_struct *curr) */ void trace_hardirqs_on(void) { @@ -22777,7 +22778,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace_irqsoff.c linux-4.9.28/kernel/tra if (!preempt_trace() && irq_trace()) stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1); } -@@ -480,11 +486,13 @@ +@@ -480,11 +486,13 @@ void trace_hardirqs_off(void) { if (!preempt_trace() && irq_trace()) start_critical_timing(CALLER_ADDR0, CALLER_ADDR1); @@ -22791,7 +22792,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace_irqsoff.c linux-4.9.28/kernel/tra if (!preempt_trace() && irq_trace()) stop_critical_timing(CALLER_ADDR0, caller_addr); } -@@ -494,6 +502,7 @@ +@@ -494,6 +502,7 @@ __visible void trace_hardirqs_off_caller(unsigned long caller_addr) { if (!preempt_trace() && irq_trace()) start_critical_timing(CALLER_ADDR0, caller_addr); @@ -22799,7 +22800,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace_irqsoff.c linux-4.9.28/kernel/tra } EXPORT_SYMBOL(trace_hardirqs_off_caller); -@@ -503,12 +512,14 @@ +@@ -503,12 +512,14 @@ EXPORT_SYMBOL(trace_hardirqs_off_caller); #ifdef CONFIG_PREEMPT_TRACER void trace_preempt_on(unsigned long a0, unsigned long a1) { @@ -22814,10 +22815,11 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace_irqsoff.c linux-4.9.28/kernel/tra if (preempt_trace() && !irq_trace()) start_critical_timing(a0, a1); } -diff -Nur linux-4.9.28.orig/kernel/trace/trace_output.c linux-4.9.28/kernel/trace/trace_output.c ---- linux-4.9.28.orig/kernel/trace/trace_output.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/trace/trace_output.c 2017-05-19 03:37:25.202177573 +0200 -@@ -386,6 +386,7 @@ +diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c +index 3fc20422c166..65a6dde71a7d 100644 +--- a/kernel/trace/trace_output.c ++++ b/kernel/trace/trace_output.c +@@ -386,6 +386,7 @@ int trace_print_lat_fmt(struct trace_seq *s, struct trace_entry *entry) { char hardsoft_irq; char need_resched; @@ -22825,7 +22827,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace_output.c linux-4.9.28/kernel/trac char irqs_off; int hardirq; int softirq; -@@ -416,6 +417,9 @@ +@@ -416,6 +417,9 @@ int trace_print_lat_fmt(struct trace_seq *s, struct trace_entry *entry) break; } @@ -22835,7 +22837,7 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace_output.c linux-4.9.28/kernel/trac hardsoft_irq = (nmi && hardirq) ? 'Z' : nmi ? 'z' : -@@ -424,14 +428,25 @@ +@@ -424,14 +428,25 @@ int trace_print_lat_fmt(struct trace_seq *s, struct trace_entry *entry) softirq ? 's' : '.' ; @@ -22863,10 +22865,11 @@ diff -Nur linux-4.9.28.orig/kernel/trace/trace_output.c linux-4.9.28/kernel/trac return !trace_seq_has_overflowed(s); } -diff -Nur linux-4.9.28.orig/kernel/user.c linux-4.9.28/kernel/user.c ---- linux-4.9.28.orig/kernel/user.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/user.c 2017-05-19 03:37:25.202177573 +0200 -@@ -161,11 +161,11 @@ +diff --git a/kernel/user.c b/kernel/user.c +index b069ccbfb0b0..1a2e88e98b5e 100644 +--- a/kernel/user.c ++++ b/kernel/user.c +@@ -161,11 +161,11 @@ void free_uid(struct user_struct *up) if (!up) return; @@ -22880,10 +22883,11 @@ diff -Nur linux-4.9.28.orig/kernel/user.c linux-4.9.28/kernel/user.c } struct user_struct *alloc_uid(kuid_t uid) -diff -Nur linux-4.9.28.orig/kernel/watchdog.c linux-4.9.28/kernel/watchdog.c ---- linux-4.9.28.orig/kernel/watchdog.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/watchdog.c 2017-05-19 03:37:25.202177573 +0200 -@@ -315,6 +315,8 @@ +diff --git a/kernel/watchdog.c b/kernel/watchdog.c +index 6d1020c03d41..70c6a2f79f7e 100644 +--- a/kernel/watchdog.c ++++ b/kernel/watchdog.c +@@ -315,6 +315,8 @@ static int is_softlockup(unsigned long touch_ts) #ifdef CONFIG_HARDLOCKUP_DETECTOR @@ -22892,7 +22896,7 @@ diff -Nur linux-4.9.28.orig/kernel/watchdog.c linux-4.9.28/kernel/watchdog.c static struct perf_event_attr wd_hw_attr = { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CPU_CYCLES, -@@ -348,6 +350,13 @@ +@@ -348,6 +350,13 @@ static void watchdog_overflow_callback(struct perf_event *event, /* only print hardlockups once */ if (__this_cpu_read(hard_watchdog_warn) == true) return; @@ -22906,7 +22910,7 @@ diff -Nur linux-4.9.28.orig/kernel/watchdog.c linux-4.9.28/kernel/watchdog.c pr_emerg("Watchdog detected hard LOCKUP on cpu %d", this_cpu); print_modules(); -@@ -365,6 +374,7 @@ +@@ -365,6 +374,7 @@ static void watchdog_overflow_callback(struct perf_event *event, !test_and_set_bit(0, &hardlockup_allcpu_dumped)) trigger_allbutself_cpu_backtrace(); @@ -22914,7 +22918,7 @@ diff -Nur linux-4.9.28.orig/kernel/watchdog.c linux-4.9.28/kernel/watchdog.c if (hardlockup_panic) nmi_panic(regs, "Hard LOCKUP"); -@@ -512,6 +522,7 @@ +@@ -512,6 +522,7 @@ static void watchdog_enable(unsigned int cpu) /* kick off the timer for the hardlockup detector */ hrtimer_init(hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); hrtimer->function = watchdog_timer_fn; @@ -22922,9 +22926,10 @@ diff -Nur linux-4.9.28.orig/kernel/watchdog.c linux-4.9.28/kernel/watchdog.c /* Enable the perf event */ watchdog_nmi_enable(cpu); -diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c ---- linux-4.9.28.orig/kernel/workqueue.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/workqueue.c 2017-05-19 03:37:25.206177727 +0200 +diff --git a/kernel/workqueue.c b/kernel/workqueue.c +index 479d840db286..24eba6620a45 100644 +--- a/kernel/workqueue.c ++++ b/kernel/workqueue.c @@ -48,6 +48,8 @@ #include #include @@ -22934,7 +22939,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c #include "workqueue_internal.h" -@@ -121,11 +123,16 @@ +@@ -121,11 +123,16 @@ enum { * cpu or grabbing pool->lock is enough for read access. If * POOL_DISASSOCIATED is set, it's identical to L. * @@ -22952,7 +22957,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c * * PW: wq_pool_mutex and wq->mutex protected for writes. Either for reads. * -@@ -134,7 +141,7 @@ +@@ -134,7 +141,7 @@ enum { * * WQ: wq->mutex protected. * @@ -22961,7 +22966,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c * * MD: wq_mayday_lock protected. */ -@@ -185,7 +192,7 @@ +@@ -185,7 +192,7 @@ struct worker_pool { atomic_t nr_running ____cacheline_aligned_in_smp; /* @@ -22970,7 +22975,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c * from get_work_pool(). */ struct rcu_head rcu; -@@ -214,7 +221,7 @@ +@@ -214,7 +221,7 @@ struct pool_workqueue { /* * Release of unbound pwq is punted to system_wq. See put_pwq() * and pwq_unbound_release_workfn() for details. pool_workqueue @@ -22979,7 +22984,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c * determined without grabbing wq->mutex. */ struct work_struct unbound_release_work; -@@ -348,6 +355,8 @@ +@@ -348,6 +355,8 @@ EXPORT_SYMBOL_GPL(system_power_efficient_wq); struct workqueue_struct *system_freezable_power_efficient_wq __read_mostly; EXPORT_SYMBOL_GPL(system_freezable_power_efficient_wq); @@ -22988,7 +22993,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c static int worker_thread(void *__worker); static void workqueue_sysfs_unregister(struct workqueue_struct *wq); -@@ -355,20 +364,20 @@ +@@ -355,20 +364,20 @@ static void workqueue_sysfs_unregister(struct workqueue_struct *wq); #include #define assert_rcu_or_pool_mutex() \ @@ -23015,7 +23020,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c #define for_each_cpu_worker_pool(pool, cpu) \ for ((pool) = &per_cpu(cpu_worker_pools, cpu)[0]; \ -@@ -380,7 +389,7 @@ +@@ -380,7 +389,7 @@ static void workqueue_sysfs_unregister(struct workqueue_struct *wq); * @pool: iteration cursor * @pi: integer used for iteration * @@ -23024,7 +23029,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c * locked. If the pool needs to be used beyond the locking in effect, the * caller is responsible for guaranteeing that the pool stays online. * -@@ -412,7 +421,7 @@ +@@ -412,7 +421,7 @@ static void workqueue_sysfs_unregister(struct workqueue_struct *wq); * @pwq: iteration cursor * @wq: the target workqueue * @@ -23033,7 +23038,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c * If the pwq needs to be used beyond the locking in effect, the caller is * responsible for guaranteeing that the pwq stays online. * -@@ -424,6 +433,31 @@ +@@ -424,6 +433,31 @@ static void workqueue_sysfs_unregister(struct workqueue_struct *wq); if (({ assert_rcu_or_wq_mutex(wq); false; })) { } \ else @@ -23065,7 +23070,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c #ifdef CONFIG_DEBUG_OBJECTS_WORK static struct debug_obj_descr work_debug_descr; -@@ -548,7 +582,7 @@ +@@ -548,7 +582,7 @@ static int worker_pool_assign_id(struct worker_pool *pool) * @wq: the target workqueue * @node: the node ID * @@ -23074,7 +23079,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c * read locked. * If the pwq needs to be used beyond the locking in effect, the caller is * responsible for guaranteeing that the pwq stays online. -@@ -692,8 +726,8 @@ +@@ -692,8 +726,8 @@ static struct pool_workqueue *get_work_pwq(struct work_struct *work) * @work: the work item of interest * * Pools are created and destroyed under wq_pool_mutex, and allows read @@ -23085,7 +23090,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c * * All fields of the returned pool are accessible as long as the above * mentioned locking is in effect. If the returned pool needs to be used -@@ -830,50 +864,45 @@ +@@ -830,50 +864,45 @@ static struct worker *first_idle_worker(struct worker_pool *pool) */ static void wake_up_worker(struct worker_pool *pool) { @@ -23107,10 +23112,10 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c + * wq_worker_running - a worker is running again * @task: task waking up - * @cpu: CPU @task is waking up to -- * + * - * This function is called during try_to_wake_up() when a worker is - * being awoken. - * +- * - * CONTEXT: - * spin_lock_irq(rq->lock) + * This function is called when a worker returns from schedule() @@ -23154,7 +23159,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c struct worker_pool *pool; /* -@@ -882,29 +911,26 @@ +@@ -882,29 +911,26 @@ struct task_struct *wq_worker_sleeping(struct task_struct *task) * checking NOT_RUNNING. */ if (worker->flags & WORKER_NOT_RUNNING) @@ -23194,7 +23199,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c } /** -@@ -1098,12 +1124,14 @@ +@@ -1098,12 +1124,14 @@ static void put_pwq_unlocked(struct pool_workqueue *pwq) { if (pwq) { /* @@ -23212,7 +23217,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c } } -@@ -1207,7 +1235,7 @@ +@@ -1207,7 +1235,7 @@ static int try_to_grab_pending(struct work_struct *work, bool is_dwork, struct worker_pool *pool; struct pool_workqueue *pwq; @@ -23221,7 +23226,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c /* try to steal the timer if it exists */ if (is_dwork) { -@@ -1226,6 +1254,7 @@ +@@ -1226,6 +1254,7 @@ static int try_to_grab_pending(struct work_struct *work, bool is_dwork, if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) return 0; @@ -23229,7 +23234,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c /* * The queueing is in progress, or it is already queued. Try to * steal it from ->worklist without clearing WORK_STRUCT_PENDING. -@@ -1264,14 +1293,16 @@ +@@ -1264,14 +1293,16 @@ static int try_to_grab_pending(struct work_struct *work, bool is_dwork, set_work_pool_and_keep_pending(work, pool->id); spin_unlock(&pool->lock); @@ -23248,7 +23253,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c return -EAGAIN; } -@@ -1373,7 +1404,7 @@ +@@ -1373,7 +1404,7 @@ static void __queue_work(int cpu, struct workqueue_struct *wq, * queued or lose PENDING. Grabbing PENDING and queueing should * happen with IRQ disabled. */ @@ -23257,7 +23262,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c debug_work_activate(work); -@@ -1381,6 +1412,7 @@ +@@ -1381,6 +1412,7 @@ static void __queue_work(int cpu, struct workqueue_struct *wq, if (unlikely(wq->flags & __WQ_DRAINING) && WARN_ON_ONCE(!is_chained_work(wq))) return; @@ -23265,7 +23270,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c retry: if (req_cpu == WORK_CPU_UNBOUND) cpu = wq_select_unbound_cpu(raw_smp_processor_id()); -@@ -1437,10 +1469,8 @@ +@@ -1437,10 +1469,8 @@ static void __queue_work(int cpu, struct workqueue_struct *wq, /* pwq determined, queue */ trace_workqueue_queue_work(req_cpu, pwq, work); @@ -23278,7 +23283,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c pwq->nr_in_flight[pwq->work_color]++; work_flags = work_color_to_flags(pwq->work_color); -@@ -1458,7 +1488,9 @@ +@@ -1458,7 +1488,9 @@ static void __queue_work(int cpu, struct workqueue_struct *wq, insert_work(pwq, work, worklist, work_flags); @@ -23288,7 +23293,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c } /** -@@ -1478,14 +1510,14 @@ +@@ -1478,14 +1510,14 @@ bool queue_work_on(int cpu, struct workqueue_struct *wq, bool ret = false; unsigned long flags; @@ -23305,7 +23310,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c return ret; } EXPORT_SYMBOL(queue_work_on); -@@ -1552,14 +1584,14 @@ +@@ -1552,14 +1584,14 @@ bool queue_delayed_work_on(int cpu, struct workqueue_struct *wq, unsigned long flags; /* read the comment in __queue_work() */ @@ -23322,7 +23327,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c return ret; } EXPORT_SYMBOL(queue_delayed_work_on); -@@ -1594,7 +1626,7 @@ +@@ -1594,7 +1626,7 @@ bool mod_delayed_work_on(int cpu, struct workqueue_struct *wq, if (likely(ret >= 0)) { __queue_delayed_work(cpu, wq, dwork, delay); @@ -23331,7 +23336,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c } /* -ENOENT from try_to_grab_pending() becomes %true */ -@@ -1627,7 +1659,9 @@ +@@ -1627,7 +1659,9 @@ static void worker_enter_idle(struct worker *worker) worker->last_active = jiffies; /* idle_list is LIFO */ @@ -23341,7 +23346,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c if (too_many_workers(pool) && !timer_pending(&pool->idle_timer)) mod_timer(&pool->idle_timer, jiffies + IDLE_WORKER_TIMEOUT); -@@ -1660,7 +1694,9 @@ +@@ -1660,7 +1694,9 @@ static void worker_leave_idle(struct worker *worker) return; worker_clr_flags(worker, WORKER_IDLE); pool->nr_idle--; @@ -23351,7 +23356,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c } static struct worker *alloc_worker(int node) -@@ -1826,7 +1862,9 @@ +@@ -1826,7 +1862,9 @@ static void destroy_worker(struct worker *worker) pool->nr_workers--; pool->nr_idle--; @@ -23361,7 +23366,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c worker->flags |= WORKER_DIE; wake_up_process(worker->task); } -@@ -2785,14 +2823,14 @@ +@@ -2785,14 +2823,14 @@ static bool start_flush_work(struct work_struct *work, struct wq_barrier *barr) might_sleep(); @@ -23379,7 +23384,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c /* see the comment in try_to_grab_pending() with the same code */ pwq = get_work_pwq(work); if (pwq) { -@@ -2821,10 +2859,11 @@ +@@ -2821,10 +2859,11 @@ static bool start_flush_work(struct work_struct *work, struct wq_barrier *barr) else lock_map_acquire_read(&pwq->wq->lockdep_map); lock_map_release(&pwq->wq->lockdep_map); @@ -23392,7 +23397,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c return false; } -@@ -2911,7 +2950,7 @@ +@@ -2911,7 +2950,7 @@ static bool __cancel_work_timer(struct work_struct *work, bool is_dwork) /* tell other tasks trying to grab @work to back off */ mark_work_canceling(work); @@ -23401,7 +23406,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c flush_work(work); clear_work_data(work); -@@ -2966,10 +3005,10 @@ +@@ -2966,10 +3005,10 @@ EXPORT_SYMBOL_GPL(cancel_work_sync); */ bool flush_delayed_work(struct delayed_work *dwork) { @@ -23414,7 +23419,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c return flush_work(&dwork->work); } EXPORT_SYMBOL(flush_delayed_work); -@@ -2987,7 +3026,7 @@ +@@ -2987,7 +3026,7 @@ static bool __cancel_work(struct work_struct *work, bool is_dwork) return false; set_work_pool_and_clear_pending(work, get_work_pool_id(work)); @@ -23423,7 +23428,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c return ret; } -@@ -3245,7 +3284,7 @@ +@@ -3245,7 +3284,7 @@ static void rcu_free_pool(struct rcu_head *rcu) * put_unbound_pool - put a worker_pool * @pool: worker_pool to put * @@ -23432,7 +23437,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c * safe manner. get_unbound_pool() calls this function on its failure path * and this function should be able to release pools which went through, * successfully or not, init_worker_pool(). -@@ -3299,8 +3338,8 @@ +@@ -3299,8 +3338,8 @@ static void put_unbound_pool(struct worker_pool *pool) del_timer_sync(&pool->idle_timer); del_timer_sync(&pool->mayday_timer); @@ -23443,7 +23448,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c } /** -@@ -3407,14 +3446,14 @@ +@@ -3407,14 +3446,14 @@ static void pwq_unbound_release_workfn(struct work_struct *work) put_unbound_pool(pool); mutex_unlock(&wq_pool_mutex); @@ -23460,7 +23465,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c } /** -@@ -4064,7 +4103,7 @@ +@@ -4064,7 +4103,7 @@ void destroy_workqueue(struct workqueue_struct *wq) * The base ref is never dropped on per-cpu pwqs. Directly * schedule RCU free. */ @@ -23469,7 +23474,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c } else { /* * We're the sole accessor of @wq at this point. Directly -@@ -4157,7 +4196,8 @@ +@@ -4157,7 +4196,8 @@ bool workqueue_congested(int cpu, struct workqueue_struct *wq) struct pool_workqueue *pwq; bool ret; @@ -23479,7 +23484,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c if (cpu == WORK_CPU_UNBOUND) cpu = smp_processor_id(); -@@ -4168,7 +4208,8 @@ +@@ -4168,7 +4208,8 @@ bool workqueue_congested(int cpu, struct workqueue_struct *wq) pwq = unbound_pwq_by_node(wq, cpu_to_node(cpu)); ret = !list_empty(&pwq->delayed_works); @@ -23489,7 +23494,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c return ret; } -@@ -4194,15 +4235,15 @@ +@@ -4194,15 +4235,15 @@ unsigned int work_busy(struct work_struct *work) if (work_pending(work)) ret |= WORK_BUSY_PENDING; @@ -23509,7 +23514,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c return ret; } -@@ -4391,7 +4432,7 @@ +@@ -4391,7 +4432,7 @@ void show_workqueue_state(void) unsigned long flags; int pi; @@ -23518,7 +23523,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c pr_info("Showing busy workqueues and worker pools:\n"); -@@ -4444,7 +4485,7 @@ +@@ -4444,7 +4485,7 @@ void show_workqueue_state(void) spin_unlock_irqrestore(&pool->lock, flags); } @@ -23527,7 +23532,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c } /* -@@ -4782,16 +4823,16 @@ +@@ -4782,16 +4823,16 @@ bool freeze_workqueues_busy(void) * nr_active is monotonically decreasing. It's safe * to peek without lock. */ @@ -23547,7 +23552,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c } out_unlock: mutex_unlock(&wq_pool_mutex); -@@ -4981,7 +5022,8 @@ +@@ -4981,7 +5022,8 @@ static ssize_t wq_pool_ids_show(struct device *dev, const char *delim = ""; int node, written = 0; @@ -23557,7 +23562,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c for_each_node(node) { written += scnprintf(buf + written, PAGE_SIZE - written, "%s%d:%d", delim, node, -@@ -4989,7 +5031,8 @@ +@@ -4989,7 +5031,8 @@ static ssize_t wq_pool_ids_show(struct device *dev, delim = " "; } written += scnprintf(buf + written, PAGE_SIZE - written, "\n"); @@ -23567,10 +23572,11 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue.c linux-4.9.28/kernel/workqueue.c return written; } -diff -Nur linux-4.9.28.orig/kernel/workqueue_internal.h linux-4.9.28/kernel/workqueue_internal.h ---- linux-4.9.28.orig/kernel/workqueue_internal.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/kernel/workqueue_internal.h 2017-05-19 03:37:25.206177727 +0200 -@@ -43,6 +43,7 @@ +diff --git a/kernel/workqueue_internal.h b/kernel/workqueue_internal.h +index 8635417c587b..f000c4d6917e 100644 +--- a/kernel/workqueue_internal.h ++++ b/kernel/workqueue_internal.h +@@ -43,6 +43,7 @@ struct worker { unsigned long last_active; /* L: last active timestamp */ unsigned int flags; /* X: flags */ int id; /* I: worker id */ @@ -23578,7 +23584,7 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue_internal.h linux-4.9.28/kernel/work /* * Opaque string set with work_set_desc(). Printed out with task -@@ -68,7 +69,7 @@ +@@ -68,7 +69,7 @@ static inline struct worker *current_wq_worker(void) * Scheduler hooks for concurrency managed workqueue. Only to be used from * sched/core.c and workqueue.c. */ @@ -23588,10 +23594,23 @@ diff -Nur linux-4.9.28.orig/kernel/workqueue_internal.h linux-4.9.28/kernel/work +void wq_worker_sleeping(struct task_struct *task); #endif /* _KERNEL_WORKQUEUE_INTERNAL_H */ -diff -Nur linux-4.9.28.orig/lib/debugobjects.c linux-4.9.28/lib/debugobjects.c ---- linux-4.9.28.orig/lib/debugobjects.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/lib/debugobjects.c 2017-05-19 03:37:25.206177727 +0200 -@@ -308,7 +308,10 @@ +diff --git a/lib/Kconfig b/lib/Kconfig +index 260a80e313b9..b06becb3f477 100644 +--- a/lib/Kconfig ++++ b/lib/Kconfig +@@ -400,6 +400,7 @@ config CHECK_SIGNATURE + + config CPUMASK_OFFSTACK + bool "Force CPU masks off stack" if DEBUG_PER_CPU_MAPS ++ depends on !PREEMPT_RT_FULL + help + Use dynamic allocation for cpumask_var_t, instead of putting + them on the stack. This is a bit more expensive, but avoids +diff --git a/lib/debugobjects.c b/lib/debugobjects.c +index 056052dc8e91..d8494e126de8 100644 +--- a/lib/debugobjects.c ++++ b/lib/debugobjects.c +@@ -308,7 +308,10 @@ __debug_object_init(void *addr, struct debug_obj_descr *descr, int onstack) struct debug_obj *obj; unsigned long flags; @@ -23603,9 +23622,10 @@ diff -Nur linux-4.9.28.orig/lib/debugobjects.c linux-4.9.28/lib/debugobjects.c db = get_bucket((unsigned long) addr); -diff -Nur linux-4.9.28.orig/lib/idr.c linux-4.9.28/lib/idr.c ---- linux-4.9.28.orig/lib/idr.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/lib/idr.c 2017-05-19 03:37:25.206177727 +0200 +diff --git a/lib/idr.c b/lib/idr.c +index 6098336df267..9decbe914595 100644 +--- a/lib/idr.c ++++ b/lib/idr.c @@ -30,6 +30,7 @@ #include #include @@ -23614,7 +23634,7 @@ diff -Nur linux-4.9.28.orig/lib/idr.c linux-4.9.28/lib/idr.c #define MAX_IDR_SHIFT (sizeof(int) * 8 - 1) #define MAX_IDR_BIT (1U << MAX_IDR_SHIFT) -@@ -45,6 +46,37 @@ +@@ -45,6 +46,37 @@ static DEFINE_PER_CPU(struct idr_layer *, idr_preload_head); static DEFINE_PER_CPU(int, idr_preload_cnt); static DEFINE_SPINLOCK(simple_ida_lock); @@ -23652,7 +23672,7 @@ diff -Nur linux-4.9.28.orig/lib/idr.c linux-4.9.28/lib/idr.c /* the maximum ID which can be allocated given idr->layers */ static int idr_max(int layers) { -@@ -115,14 +147,14 @@ +@@ -115,14 +147,14 @@ static struct idr_layer *idr_layer_alloc(gfp_t gfp_mask, struct idr *layer_idr) * context. See idr_preload() for details. */ if (!in_interrupt()) { @@ -23669,7 +23689,7 @@ diff -Nur linux-4.9.28.orig/lib/idr.c linux-4.9.28/lib/idr.c if (new) return new; } -@@ -366,7 +398,6 @@ +@@ -366,7 +398,6 @@ static void idr_fill_slot(struct idr *idr, void *ptr, int id, idr_mark_full(pa, id); } @@ -23677,7 +23697,7 @@ diff -Nur linux-4.9.28.orig/lib/idr.c linux-4.9.28/lib/idr.c /** * idr_preload - preload for idr_alloc() * @gfp_mask: allocation mask to use for preloading -@@ -401,7 +432,7 @@ +@@ -401,7 +432,7 @@ void idr_preload(gfp_t gfp_mask) WARN_ON_ONCE(in_interrupt()); might_sleep_if(gfpflags_allow_blocking(gfp_mask)); @@ -23686,7 +23706,7 @@ diff -Nur linux-4.9.28.orig/lib/idr.c linux-4.9.28/lib/idr.c /* * idr_alloc() is likely to succeed w/o full idr_layer buffer and -@@ -413,9 +444,9 @@ +@@ -413,9 +444,9 @@ void idr_preload(gfp_t gfp_mask) while (__this_cpu_read(idr_preload_cnt) < MAX_IDR_FREE) { struct idr_layer *new; @@ -23698,10 +23718,11 @@ diff -Nur linux-4.9.28.orig/lib/idr.c linux-4.9.28/lib/idr.c if (!new) break; -diff -Nur linux-4.9.28.orig/lib/irq_poll.c linux-4.9.28/lib/irq_poll.c ---- linux-4.9.28.orig/lib/irq_poll.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/lib/irq_poll.c 2017-05-19 03:37:25.206177727 +0200 -@@ -36,6 +36,7 @@ +diff --git a/lib/irq_poll.c b/lib/irq_poll.c +index 1d6565e81030..b23a79761df7 100644 +--- a/lib/irq_poll.c ++++ b/lib/irq_poll.c +@@ -36,6 +36,7 @@ void irq_poll_sched(struct irq_poll *iop) list_add_tail(&iop->list, this_cpu_ptr(&blk_cpu_iopoll)); __raise_softirq_irqoff(IRQ_POLL_SOFTIRQ); local_irq_restore(flags); @@ -23709,7 +23730,7 @@ diff -Nur linux-4.9.28.orig/lib/irq_poll.c linux-4.9.28/lib/irq_poll.c } EXPORT_SYMBOL(irq_poll_sched); -@@ -71,6 +72,7 @@ +@@ -71,6 +72,7 @@ void irq_poll_complete(struct irq_poll *iop) local_irq_save(flags); __irq_poll_complete(iop); local_irq_restore(flags); @@ -23717,7 +23738,7 @@ diff -Nur linux-4.9.28.orig/lib/irq_poll.c linux-4.9.28/lib/irq_poll.c } EXPORT_SYMBOL(irq_poll_complete); -@@ -95,6 +97,7 @@ +@@ -95,6 +97,7 @@ static void __latent_entropy irq_poll_softirq(struct softirq_action *h) } local_irq_enable(); @@ -23725,7 +23746,7 @@ diff -Nur linux-4.9.28.orig/lib/irq_poll.c linux-4.9.28/lib/irq_poll.c /* Even though interrupts have been re-enabled, this * access is safe because interrupts can only add new -@@ -132,6 +135,7 @@ +@@ -132,6 +135,7 @@ static void __latent_entropy irq_poll_softirq(struct softirq_action *h) __raise_softirq_irqoff(IRQ_POLL_SOFTIRQ); local_irq_enable(); @@ -23733,7 +23754,7 @@ diff -Nur linux-4.9.28.orig/lib/irq_poll.c linux-4.9.28/lib/irq_poll.c } /** -@@ -195,6 +199,7 @@ +@@ -195,6 +199,7 @@ static int irq_poll_cpu_dead(unsigned int cpu) this_cpu_ptr(&blk_cpu_iopoll)); __raise_softirq_irqoff(IRQ_POLL_SOFTIRQ); local_irq_enable(); @@ -23741,21 +23762,11 @@ diff -Nur linux-4.9.28.orig/lib/irq_poll.c linux-4.9.28/lib/irq_poll.c return 0; } -diff -Nur linux-4.9.28.orig/lib/Kconfig linux-4.9.28/lib/Kconfig ---- linux-4.9.28.orig/lib/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/lib/Kconfig 2017-05-19 03:37:25.206177727 +0200 -@@ -400,6 +400,7 @@ - - config CPUMASK_OFFSTACK - bool "Force CPU masks off stack" if DEBUG_PER_CPU_MAPS -+ depends on !PREEMPT_RT_FULL - help - Use dynamic allocation for cpumask_var_t, instead of putting - them on the stack. This is a bit more expensive, but avoids -diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-selftest.c ---- linux-4.9.28.orig/lib/locking-selftest.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/lib/locking-selftest.c 2017-05-19 03:37:25.206177727 +0200 -@@ -590,6 +590,8 @@ +diff --git a/lib/locking-selftest.c b/lib/locking-selftest.c +index f3a217ea0388..4611b156ef79 100644 +--- a/lib/locking-selftest.c ++++ b/lib/locking-selftest.c +@@ -590,6 +590,8 @@ GENERATE_TESTCASE(init_held_rsem) #include "locking-selftest-spin-hardirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_hard_spin) @@ -23764,7 +23775,7 @@ diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-self #include "locking-selftest-rlock-hardirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_hard_rlock) -@@ -605,9 +607,12 @@ +@@ -605,9 +607,12 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_soft_rlock) #include "locking-selftest-wlock-softirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_soft_wlock) @@ -23777,7 +23788,7 @@ diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-self /* * Enabling hardirqs with a softirq-safe lock held: */ -@@ -640,6 +645,8 @@ +@@ -640,6 +645,8 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2A_rlock) #undef E1 #undef E2 @@ -23786,7 +23797,7 @@ diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-self /* * Enabling irqs with an irq-safe lock held: */ -@@ -663,6 +670,8 @@ +@@ -663,6 +670,8 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2A_rlock) #include "locking-selftest-spin-hardirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_hard_spin) @@ -23795,7 +23806,7 @@ diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-self #include "locking-selftest-rlock-hardirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_hard_rlock) -@@ -678,6 +687,8 @@ +@@ -678,6 +687,8 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_rlock) #include "locking-selftest-wlock-softirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_wlock) @@ -23804,7 +23815,7 @@ diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-self #undef E1 #undef E2 -@@ -709,6 +720,8 @@ +@@ -709,6 +720,8 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_wlock) #include "locking-selftest-spin-hardirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_hard_spin) @@ -23813,7 +23824,7 @@ diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-self #include "locking-selftest-rlock-hardirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_hard_rlock) -@@ -724,6 +737,8 @@ +@@ -724,6 +737,8 @@ GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_rlock) #include "locking-selftest-wlock-softirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_wlock) @@ -23822,7 +23833,7 @@ diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-self #undef E1 #undef E2 #undef E3 -@@ -757,6 +772,8 @@ +@@ -757,6 +772,8 @@ GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_wlock) #include "locking-selftest-spin-hardirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_hard_spin) @@ -23831,7 +23842,7 @@ diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-self #include "locking-selftest-rlock-hardirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_hard_rlock) -@@ -772,10 +789,14 @@ +@@ -772,10 +789,14 @@ GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_soft_rlock) #include "locking-selftest-wlock-softirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_soft_wlock) @@ -23846,7 +23857,7 @@ diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-self /* * read-lock / write-lock irq inversion. * -@@ -838,6 +859,10 @@ +@@ -838,6 +859,10 @@ GENERATE_PERMUTATIONS_3_EVENTS(irq_inversion_soft_wlock) #undef E2 #undef E3 @@ -23857,7 +23868,7 @@ diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-self /* * read-lock / write-lock recursion that is actually safe. */ -@@ -876,6 +901,8 @@ +@@ -876,6 +901,8 @@ GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion_soft) #undef E2 #undef E3 @@ -23866,7 +23877,7 @@ diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-self /* * read-lock / write-lock recursion that is unsafe. */ -@@ -1858,6 +1885,7 @@ +@@ -1858,6 +1885,7 @@ void locking_selftest(void) printk(" --------------------------------------------------------------------------\n"); @@ -23874,7 +23885,7 @@ diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-self /* * irq-context testcases: */ -@@ -1870,6 +1898,28 @@ +@@ -1870,6 +1898,28 @@ void locking_selftest(void) DO_TESTCASE_6x2("irq read-recursion", irq_read_recursion); // DO_TESTCASE_6x2B("irq read-recursion #2", irq_read_recursion2); @@ -23903,9 +23914,10 @@ diff -Nur linux-4.9.28.orig/lib/locking-selftest.c linux-4.9.28/lib/locking-self ww_tests(); -diff -Nur linux-4.9.28.orig/lib/percpu_ida.c linux-4.9.28/lib/percpu_ida.c ---- linux-4.9.28.orig/lib/percpu_ida.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/lib/percpu_ida.c 2017-05-19 03:37:25.206177727 +0200 +diff --git a/lib/percpu_ida.c b/lib/percpu_ida.c +index 6d40944960de..822a2c027e72 100644 +--- a/lib/percpu_ida.c ++++ b/lib/percpu_ida.c @@ -26,6 +26,9 @@ #include #include @@ -23916,7 +23928,7 @@ diff -Nur linux-4.9.28.orig/lib/percpu_ida.c linux-4.9.28/lib/percpu_ida.c struct percpu_ida_cpu { /* -@@ -148,13 +151,13 @@ +@@ -148,13 +151,13 @@ int percpu_ida_alloc(struct percpu_ida *pool, int state) unsigned long flags; int tag; @@ -23932,7 +23944,7 @@ diff -Nur linux-4.9.28.orig/lib/percpu_ida.c linux-4.9.28/lib/percpu_ida.c return tag; } -@@ -173,6 +176,7 @@ +@@ -173,6 +176,7 @@ int percpu_ida_alloc(struct percpu_ida *pool, int state) if (!tags->nr_free) alloc_global_tags(pool, tags); @@ -23940,7 +23952,7 @@ diff -Nur linux-4.9.28.orig/lib/percpu_ida.c linux-4.9.28/lib/percpu_ida.c if (!tags->nr_free) steal_tags(pool, tags); -@@ -184,7 +188,7 @@ +@@ -184,7 +188,7 @@ int percpu_ida_alloc(struct percpu_ida *pool, int state) } spin_unlock(&pool->lock); @@ -23949,7 +23961,7 @@ diff -Nur linux-4.9.28.orig/lib/percpu_ida.c linux-4.9.28/lib/percpu_ida.c if (tag >= 0 || state == TASK_RUNNING) break; -@@ -196,7 +200,7 @@ +@@ -196,7 +200,7 @@ int percpu_ida_alloc(struct percpu_ida *pool, int state) schedule(); @@ -23958,7 +23970,7 @@ diff -Nur linux-4.9.28.orig/lib/percpu_ida.c linux-4.9.28/lib/percpu_ida.c tags = this_cpu_ptr(pool->tag_cpu); } if (state != TASK_RUNNING) -@@ -221,7 +225,7 @@ +@@ -221,7 +225,7 @@ void percpu_ida_free(struct percpu_ida *pool, unsigned tag) BUG_ON(tag >= pool->nr_tags); @@ -23967,7 +23979,7 @@ diff -Nur linux-4.9.28.orig/lib/percpu_ida.c linux-4.9.28/lib/percpu_ida.c tags = this_cpu_ptr(pool->tag_cpu); spin_lock(&tags->lock); -@@ -253,7 +257,7 @@ +@@ -253,7 +257,7 @@ void percpu_ida_free(struct percpu_ida *pool, unsigned tag) spin_unlock(&pool->lock); } @@ -23976,7 +23988,7 @@ diff -Nur linux-4.9.28.orig/lib/percpu_ida.c linux-4.9.28/lib/percpu_ida.c } EXPORT_SYMBOL_GPL(percpu_ida_free); -@@ -345,7 +349,7 @@ +@@ -345,7 +349,7 @@ int percpu_ida_for_each_free(struct percpu_ida *pool, percpu_ida_cb fn, struct percpu_ida_cpu *remote; unsigned cpu, i, err = 0; @@ -23985,7 +23997,7 @@ diff -Nur linux-4.9.28.orig/lib/percpu_ida.c linux-4.9.28/lib/percpu_ida.c for_each_possible_cpu(cpu) { remote = per_cpu_ptr(pool->tag_cpu, cpu); spin_lock(&remote->lock); -@@ -367,7 +371,7 @@ +@@ -367,7 +371,7 @@ int percpu_ida_for_each_free(struct percpu_ida *pool, percpu_ida_cb fn, } spin_unlock(&pool->lock); out: @@ -23994,9 +24006,10 @@ diff -Nur linux-4.9.28.orig/lib/percpu_ida.c linux-4.9.28/lib/percpu_ida.c return err; } EXPORT_SYMBOL_GPL(percpu_ida_for_each_free); -diff -Nur linux-4.9.28.orig/lib/radix-tree.c linux-4.9.28/lib/radix-tree.c ---- linux-4.9.28.orig/lib/radix-tree.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/lib/radix-tree.c 2017-05-19 03:37:25.206177727 +0200 +diff --git a/lib/radix-tree.c b/lib/radix-tree.c +index 8e6d552c40dd..741da5a77fd5 100644 +--- a/lib/radix-tree.c ++++ b/lib/radix-tree.c @@ -36,7 +36,7 @@ #include #include @@ -24006,7 +24019,7 @@ diff -Nur linux-4.9.28.orig/lib/radix-tree.c linux-4.9.28/lib/radix-tree.c /* Number of nodes in fully populated tree of given height */ static unsigned long height_to_maxnodes[RADIX_TREE_MAX_PATH + 1] __read_mostly; -@@ -68,6 +68,7 @@ +@@ -68,6 +68,7 @@ struct radix_tree_preload { struct radix_tree_node *nodes; }; static DEFINE_PER_CPU(struct radix_tree_preload, radix_tree_preloads) = { 0, }; @@ -24014,7 +24027,7 @@ diff -Nur linux-4.9.28.orig/lib/radix-tree.c linux-4.9.28/lib/radix-tree.c static inline void *node_to_entry(void *ptr) { -@@ -290,13 +291,14 @@ +@@ -290,13 +291,14 @@ radix_tree_node_alloc(struct radix_tree_root *root) * succeed in getting a node here (and never reach * kmem_cache_alloc) */ @@ -24030,7 +24043,7 @@ diff -Nur linux-4.9.28.orig/lib/radix-tree.c linux-4.9.28/lib/radix-tree.c /* * Update the allocation stack trace as this is more useful * for debugging. -@@ -357,14 +359,14 @@ +@@ -357,14 +359,14 @@ static int __radix_tree_preload(gfp_t gfp_mask, int nr) */ gfp_mask &= ~__GFP_ACCOUNT; @@ -24048,7 +24061,7 @@ diff -Nur linux-4.9.28.orig/lib/radix-tree.c linux-4.9.28/lib/radix-tree.c rtp = this_cpu_ptr(&radix_tree_preloads); if (rtp->nr < nr) { node->private_data = rtp->nodes; -@@ -406,7 +408,7 @@ +@@ -406,7 +408,7 @@ int radix_tree_maybe_preload(gfp_t gfp_mask) if (gfpflags_allow_blocking(gfp_mask)) return __radix_tree_preload(gfp_mask, RADIX_TREE_PRELOAD_SIZE); /* Preloading doesn't help anything with this gfp mask, skip it */ @@ -24057,7 +24070,7 @@ diff -Nur linux-4.9.28.orig/lib/radix-tree.c linux-4.9.28/lib/radix-tree.c return 0; } EXPORT_SYMBOL(radix_tree_maybe_preload); -@@ -422,7 +424,7 @@ +@@ -422,7 +424,7 @@ int radix_tree_maybe_preload_order(gfp_t gfp_mask, int order) /* Preloading doesn't help anything with this gfp mask, skip it */ if (!gfpflags_allow_blocking(gfp_mask)) { @@ -24066,7 +24079,7 @@ diff -Nur linux-4.9.28.orig/lib/radix-tree.c linux-4.9.28/lib/radix-tree.c return 0; } -@@ -456,6 +458,12 @@ +@@ -456,6 +458,12 @@ int radix_tree_maybe_preload_order(gfp_t gfp_mask, int order) return __radix_tree_preload(gfp_mask, nr_nodes); } @@ -24079,10 +24092,11 @@ diff -Nur linux-4.9.28.orig/lib/radix-tree.c linux-4.9.28/lib/radix-tree.c /* * The maximum index which can be stored in a radix tree */ -diff -Nur linux-4.9.28.orig/lib/scatterlist.c linux-4.9.28/lib/scatterlist.c ---- linux-4.9.28.orig/lib/scatterlist.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/lib/scatterlist.c 2017-05-19 03:37:25.206177727 +0200 -@@ -620,7 +620,7 @@ +diff --git a/lib/scatterlist.c b/lib/scatterlist.c +index 004fc70fc56a..ccc46992a517 100644 +--- a/lib/scatterlist.c ++++ b/lib/scatterlist.c +@@ -620,7 +620,7 @@ void sg_miter_stop(struct sg_mapping_iter *miter) flush_kernel_dcache_page(miter->page); if (miter->__flags & SG_MITER_ATOMIC) { @@ -24091,7 +24105,7 @@ diff -Nur linux-4.9.28.orig/lib/scatterlist.c linux-4.9.28/lib/scatterlist.c kunmap_atomic(miter->addr); } else kunmap(miter->page); -@@ -664,7 +664,7 @@ +@@ -664,7 +664,7 @@ size_t sg_copy_buffer(struct scatterlist *sgl, unsigned int nents, void *buf, if (!sg_miter_skip(&miter, skip)) return false; @@ -24100,7 +24114,7 @@ diff -Nur linux-4.9.28.orig/lib/scatterlist.c linux-4.9.28/lib/scatterlist.c while (sg_miter_next(&miter) && offset < buflen) { unsigned int len; -@@ -681,7 +681,7 @@ +@@ -681,7 +681,7 @@ size_t sg_copy_buffer(struct scatterlist *sgl, unsigned int nents, void *buf, sg_miter_stop(&miter); @@ -24109,10 +24123,11 @@ diff -Nur linux-4.9.28.orig/lib/scatterlist.c linux-4.9.28/lib/scatterlist.c return offset; } EXPORT_SYMBOL(sg_copy_buffer); -diff -Nur linux-4.9.28.orig/lib/smp_processor_id.c linux-4.9.28/lib/smp_processor_id.c ---- linux-4.9.28.orig/lib/smp_processor_id.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/lib/smp_processor_id.c 2017-05-19 03:37:25.206177727 +0200 -@@ -39,8 +39,9 @@ +diff --git a/lib/smp_processor_id.c b/lib/smp_processor_id.c +index 1afec32de6f2..11fa431046a8 100644 +--- a/lib/smp_processor_id.c ++++ b/lib/smp_processor_id.c +@@ -39,8 +39,9 @@ notrace static unsigned int check_preemption_disabled(const char *what1, if (!printk_ratelimit()) goto out_enable; @@ -24124,37 +24139,24 @@ diff -Nur linux-4.9.28.orig/lib/smp_processor_id.c linux-4.9.28/lib/smp_processo print_symbol("caller is %s\n", (long)__builtin_return_address(0)); dump_stack(); -diff -Nur linux-4.9.28.orig/MAINTAINERS linux-4.9.28/MAINTAINERS ---- linux-4.9.28.orig/MAINTAINERS 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/MAINTAINERS 2017-05-19 03:37:25.122174217 +0200 -@@ -5196,6 +5196,23 @@ - F: include/uapi/linux/fuse.h - F: Documentation/filesystems/fuse.txt +diff --git a/mm/Kconfig b/mm/Kconfig +index 86e3e0e74d20..77e5862a1ed2 100644 +--- a/mm/Kconfig ++++ b/mm/Kconfig +@@ -410,7 +410,7 @@ config NOMMU_INITIAL_TRIM_EXCESS -+FUTEX SUBSYSTEM -+M: Thomas Gleixner -+M: Ingo Molnar -+R: Peter Zijlstra -+R: Darren Hart -+L: linux-kernel@vger.kernel.org -+T: git git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git locking/core -+S: Maintained -+F: kernel/futex.c -+F: kernel/futex_compat.c -+F: include/asm-generic/futex.h -+F: include/linux/futex.h -+F: include/uapi/linux/futex.h -+F: tools/testing/selftests/futex/ -+F: tools/perf/bench/futex* -+F: Documentation/*futex* -+ - FUTURE DOMAIN TMC-16x0 SCSI DRIVER (16-bit) - M: Rik Faith - L: linux-scsi@vger.kernel.org -diff -Nur linux-4.9.28.orig/mm/backing-dev.c linux-4.9.28/mm/backing-dev.c ---- linux-4.9.28.orig/mm/backing-dev.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/backing-dev.c 2017-05-19 03:37:25.206177727 +0200 -@@ -457,9 +457,9 @@ + config TRANSPARENT_HUGEPAGE + bool "Transparent Hugepage Support" +- depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE ++ depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE && !PREEMPT_RT_FULL + select COMPACTION + select RADIX_TREE_MULTIORDER + help +diff --git a/mm/backing-dev.c b/mm/backing-dev.c +index 6ff2d7744223..b5a91dd53b5f 100644 +--- a/mm/backing-dev.c ++++ b/mm/backing-dev.c +@@ -457,9 +457,9 @@ void wb_congested_put(struct bdi_writeback_congested *congested) { unsigned long flags; @@ -24166,10 +24168,11 @@ diff -Nur linux-4.9.28.orig/mm/backing-dev.c linux-4.9.28/mm/backing-dev.c return; } -diff -Nur linux-4.9.28.orig/mm/compaction.c linux-4.9.28/mm/compaction.c ---- linux-4.9.28.orig/mm/compaction.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/compaction.c 2017-05-19 03:37:25.206177727 +0200 -@@ -1593,10 +1593,12 @@ +diff --git a/mm/compaction.c b/mm/compaction.c +index 70e6bec46dc2..6678ed58b7c6 100644 +--- a/mm/compaction.c ++++ b/mm/compaction.c +@@ -1593,10 +1593,12 @@ static enum compact_result compact_zone(struct zone *zone, struct compact_contro block_start_pfn(cc->migrate_pfn, cc->order); if (cc->last_migrated_pfn < current_block_start) { @@ -24184,10 +24187,11 @@ diff -Nur linux-4.9.28.orig/mm/compaction.c linux-4.9.28/mm/compaction.c /* No more flushing until we migrate again */ cc->last_migrated_pfn = 0; } -diff -Nur linux-4.9.28.orig/mm/filemap.c linux-4.9.28/mm/filemap.c ---- linux-4.9.28.orig/mm/filemap.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/filemap.c 2017-05-19 03:37:25.210177882 +0200 -@@ -159,9 +159,12 @@ +diff --git a/mm/filemap.c b/mm/filemap.c +index edfb90e3830c..a8d2c7a73d54 100644 +--- a/mm/filemap.c ++++ b/mm/filemap.c +@@ -159,9 +159,12 @@ static int page_cache_tree_insert(struct address_space *mapping, * node->private_list is protected by * mapping->tree_lock. */ @@ -24202,7 +24206,7 @@ diff -Nur linux-4.9.28.orig/mm/filemap.c linux-4.9.28/mm/filemap.c } return 0; } -@@ -217,8 +220,10 @@ +@@ -217,8 +220,10 @@ static void page_cache_tree_delete(struct address_space *mapping, if (!dax_mapping(mapping) && !workingset_node_pages(node) && list_empty(&node->private_list)) { node->private_data = mapping; @@ -24215,9 +24219,10 @@ diff -Nur linux-4.9.28.orig/mm/filemap.c linux-4.9.28/mm/filemap.c } } -diff -Nur linux-4.9.28.orig/mm/highmem.c linux-4.9.28/mm/highmem.c ---- linux-4.9.28.orig/mm/highmem.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/highmem.c 2017-05-19 03:37:25.210177882 +0200 +diff --git a/mm/highmem.c b/mm/highmem.c +index 50b4ca6787f0..77518a3b35a1 100644 +--- a/mm/highmem.c ++++ b/mm/highmem.c @@ -29,10 +29,11 @@ #include #include @@ -24231,7 +24236,7 @@ diff -Nur linux-4.9.28.orig/mm/highmem.c linux-4.9.28/mm/highmem.c /* * Virtual_count is not a pure "count". -@@ -107,8 +108,9 @@ +@@ -107,8 +108,9 @@ static inline wait_queue_head_t *get_pkmap_wait_queue_head(unsigned int color) unsigned long totalhigh_pages __read_mostly; EXPORT_SYMBOL(totalhigh_pages); @@ -24242,21 +24247,10 @@ diff -Nur linux-4.9.28.orig/mm/highmem.c linux-4.9.28/mm/highmem.c unsigned int nr_free_highpages (void) { -diff -Nur linux-4.9.28.orig/mm/Kconfig linux-4.9.28/mm/Kconfig ---- linux-4.9.28.orig/mm/Kconfig 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/Kconfig 2017-05-19 03:37:25.206177727 +0200 -@@ -410,7 +410,7 @@ - - config TRANSPARENT_HUGEPAGE - bool "Transparent Hugepage Support" -- depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE -+ depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE && !PREEMPT_RT_FULL - select COMPACTION - select RADIX_TREE_MULTIORDER - help -diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c ---- linux-4.9.28.orig/mm/memcontrol.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/memcontrol.c 2017-05-19 03:37:25.210177882 +0200 +diff --git a/mm/memcontrol.c b/mm/memcontrol.c +index 47559cc0cdcc..1f2ebc924916 100644 +--- a/mm/memcontrol.c ++++ b/mm/memcontrol.c @@ -67,6 +67,7 @@ #include #include @@ -24265,7 +24259,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c #include -@@ -92,6 +93,8 @@ +@@ -92,6 +93,8 @@ int do_swap_account __read_mostly; #define do_swap_account 0 #endif @@ -24274,7 +24268,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c /* Whether legacy memory+swap accounting is active */ static bool do_memsw_account(void) { -@@ -1692,6 +1695,7 @@ +@@ -1692,6 +1695,7 @@ struct memcg_stock_pcp { #define FLUSHING_CACHED_CHARGE 0 }; static DEFINE_PER_CPU(struct memcg_stock_pcp, memcg_stock); @@ -24282,7 +24276,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c static DEFINE_MUTEX(percpu_charge_mutex); /** -@@ -1714,7 +1718,7 @@ +@@ -1714,7 +1718,7 @@ static bool consume_stock(struct mem_cgroup *memcg, unsigned int nr_pages) if (nr_pages > CHARGE_BATCH) return ret; @@ -24291,7 +24285,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c stock = this_cpu_ptr(&memcg_stock); if (memcg == stock->cached && stock->nr_pages >= nr_pages) { -@@ -1722,7 +1726,7 @@ +@@ -1722,7 +1726,7 @@ static bool consume_stock(struct mem_cgroup *memcg, unsigned int nr_pages) ret = true; } @@ -24300,7 +24294,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c return ret; } -@@ -1749,13 +1753,13 @@ +@@ -1749,13 +1753,13 @@ static void drain_local_stock(struct work_struct *dummy) struct memcg_stock_pcp *stock; unsigned long flags; @@ -24316,7 +24310,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c } /* -@@ -1767,7 +1771,7 @@ +@@ -1767,7 +1771,7 @@ static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages) struct memcg_stock_pcp *stock; unsigned long flags; @@ -24325,7 +24319,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c stock = this_cpu_ptr(&memcg_stock); if (stock->cached != memcg) { /* reset if necessary */ -@@ -1776,7 +1780,7 @@ +@@ -1776,7 +1780,7 @@ static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages) } stock->nr_pages += nr_pages; @@ -24334,7 +24328,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c } /* -@@ -1792,7 +1796,7 @@ +@@ -1792,7 +1796,7 @@ static void drain_all_stock(struct mem_cgroup *root_memcg) return; /* Notify other cpus that system-wide "drain" is running */ get_online_cpus(); @@ -24343,7 +24337,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c for_each_online_cpu(cpu) { struct memcg_stock_pcp *stock = &per_cpu(memcg_stock, cpu); struct mem_cgroup *memcg; -@@ -1809,7 +1813,7 @@ +@@ -1809,7 +1813,7 @@ static void drain_all_stock(struct mem_cgroup *root_memcg) schedule_work_on(cpu, &stock->work); } } @@ -24352,7 +24346,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c put_online_cpus(); mutex_unlock(&percpu_charge_mutex); } -@@ -4555,12 +4559,12 @@ +@@ -4555,12 +4559,12 @@ static int mem_cgroup_move_account(struct page *page, ret = 0; @@ -24367,7 +24361,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c out_unlock: unlock_page(page); out: -@@ -5435,10 +5439,10 @@ +@@ -5435,10 +5439,10 @@ void mem_cgroup_commit_charge(struct page *page, struct mem_cgroup *memcg, commit_charge(page, memcg, lrucare); @@ -24380,7 +24374,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c if (do_memsw_account() && PageSwapCache(page)) { swp_entry_t entry = { .val = page_private(page) }; -@@ -5494,14 +5498,14 @@ +@@ -5494,14 +5498,14 @@ static void uncharge_batch(struct mem_cgroup *memcg, unsigned long pgpgout, memcg_oom_recover(memcg); } @@ -24397,7 +24391,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c if (!mem_cgroup_is_root(memcg)) css_put_many(&memcg->css, nr_pages); -@@ -5656,10 +5660,10 @@ +@@ -5656,10 +5660,10 @@ void mem_cgroup_migrate(struct page *oldpage, struct page *newpage) commit_charge(newpage, memcg, false); @@ -24410,7 +24404,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c } DEFINE_STATIC_KEY_FALSE(memcg_sockets_enabled_key); -@@ -5850,6 +5854,7 @@ +@@ -5850,6 +5854,7 @@ void mem_cgroup_swapout(struct page *page, swp_entry_t entry) { struct mem_cgroup *memcg, *swap_memcg; unsigned short oldid; @@ -24418,7 +24412,7 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c VM_BUG_ON_PAGE(PageLRU(page), page); VM_BUG_ON_PAGE(page_count(page), page); -@@ -5890,12 +5895,16 @@ +@@ -5890,12 +5895,16 @@ void mem_cgroup_swapout(struct page *page, swp_entry_t entry) * important here to have the interrupts disabled because it is the * only synchronisation we have for udpating the per-CPU variables. */ @@ -24435,10 +24429,11 @@ diff -Nur linux-4.9.28.orig/mm/memcontrol.c linux-4.9.28/mm/memcontrol.c } /* -diff -Nur linux-4.9.28.orig/mm/mmu_context.c linux-4.9.28/mm/mmu_context.c ---- linux-4.9.28.orig/mm/mmu_context.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/mmu_context.c 2017-05-19 03:37:25.210177882 +0200 -@@ -23,6 +23,7 @@ +diff --git a/mm/mmu_context.c b/mm/mmu_context.c +index 6f4d27c5bb32..5cd25c745a8f 100644 +--- a/mm/mmu_context.c ++++ b/mm/mmu_context.c +@@ -23,6 +23,7 @@ void use_mm(struct mm_struct *mm) struct task_struct *tsk = current; task_lock(tsk); @@ -24446,7 +24441,7 @@ diff -Nur linux-4.9.28.orig/mm/mmu_context.c linux-4.9.28/mm/mmu_context.c active_mm = tsk->active_mm; if (active_mm != mm) { atomic_inc(&mm->mm_count); -@@ -30,6 +31,7 @@ +@@ -30,6 +31,7 @@ void use_mm(struct mm_struct *mm) } tsk->mm = mm; switch_mm(active_mm, mm, tsk); @@ -24454,9 +24449,10 @@ diff -Nur linux-4.9.28.orig/mm/mmu_context.c linux-4.9.28/mm/mmu_context.c task_unlock(tsk); #ifdef finish_arch_post_lock_switch finish_arch_post_lock_switch(); -diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c ---- linux-4.9.28.orig/mm/page_alloc.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/page_alloc.c 2017-05-19 03:37:25.210177882 +0200 +diff --git a/mm/page_alloc.c b/mm/page_alloc.c +index 5b06fb385dd7..86457a2fca20 100644 +--- a/mm/page_alloc.c ++++ b/mm/page_alloc.c @@ -61,6 +61,7 @@ #include #include @@ -24465,7 +24461,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c #include #include #include -@@ -281,6 +282,18 @@ +@@ -281,6 +282,18 @@ EXPORT_SYMBOL(nr_node_ids); EXPORT_SYMBOL(nr_online_nodes); #endif @@ -24484,7 +24480,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c int page_group_by_mobility_disabled __read_mostly; #ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT -@@ -1072,7 +1085,7 @@ +@@ -1072,7 +1085,7 @@ static bool bulkfree_pcp_prepare(struct page *page) #endif /* CONFIG_DEBUG_VM */ /* @@ -24493,7 +24489,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c * Assumes all pages on list are in same zone, and of same order. * count is the number of pages to free. * -@@ -1083,19 +1096,58 @@ +@@ -1083,19 +1096,58 @@ static bool bulkfree_pcp_prepare(struct page *page) * pinned" detection logic. */ static void free_pcppages_bulk(struct zone *zone, int count, @@ -24556,7 +24552,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c while (count) { struct page *page; struct list_head *list; -@@ -1111,7 +1163,7 @@ +@@ -1111,7 +1163,7 @@ static void free_pcppages_bulk(struct zone *zone, int count, batch_free++; if (++migratetype == MIGRATE_PCPTYPES) migratetype = 0; @@ -24565,7 +24561,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c } while (list_empty(list)); /* This is the only non-empty list. Free them all. */ -@@ -1119,27 +1171,12 @@ +@@ -1119,27 +1171,12 @@ static void free_pcppages_bulk(struct zone *zone, int count, batch_free = count; do { @@ -24594,7 +24590,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c } static void free_one_page(struct zone *zone, -@@ -1148,7 +1185,9 @@ +@@ -1148,7 +1185,9 @@ static void free_one_page(struct zone *zone, int migratetype) { unsigned long nr_scanned; @@ -24605,7 +24601,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c nr_scanned = node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED); if (nr_scanned) __mod_node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED, -nr_scanned); -@@ -1158,7 +1197,7 @@ +@@ -1158,7 +1197,7 @@ static void free_one_page(struct zone *zone, migratetype = get_pfnblock_migratetype(page, pfn); } __free_one_page(page, pfn, zone, order, migratetype); @@ -24614,7 +24610,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c } static void __meminit __init_single_page(struct page *page, unsigned long pfn, -@@ -1244,10 +1283,10 @@ +@@ -1244,10 +1283,10 @@ static void __free_pages_ok(struct page *page, unsigned int order) return; migratetype = get_pfnblock_migratetype(page, pfn); @@ -24627,7 +24623,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c } static void __init __free_pages_boot_core(struct page *page, unsigned int order) -@@ -2246,16 +2285,18 @@ +@@ -2246,16 +2285,18 @@ static int rmqueue_bulk(struct zone *zone, unsigned int order, void drain_zone_pages(struct zone *zone, struct per_cpu_pages *pcp) { unsigned long flags; @@ -24649,7 +24645,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c } #endif -@@ -2271,16 +2312,21 @@ +@@ -2271,16 +2312,21 @@ static void drain_pages_zone(unsigned int cpu, struct zone *zone) unsigned long flags; struct per_cpu_pageset *pset; struct per_cpu_pages *pcp; @@ -24675,7 +24671,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c } /* -@@ -2366,8 +2412,17 @@ +@@ -2366,8 +2412,17 @@ void drain_all_pages(struct zone *zone) else cpumask_clear_cpu(cpu, &cpus_with_pcps); } @@ -24693,7 +24689,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c } #ifdef CONFIG_HIBERNATION -@@ -2427,7 +2482,7 @@ +@@ -2427,7 +2482,7 @@ void free_hot_cold_page(struct page *page, bool cold) migratetype = get_pfnblock_migratetype(page, pfn); set_pcppage_migratetype(page, migratetype); @@ -24702,7 +24698,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c __count_vm_event(PGFREE); /* -@@ -2453,12 +2508,17 @@ +@@ -2453,12 +2508,17 @@ void free_hot_cold_page(struct page *page, bool cold) pcp->count++; if (pcp->count >= pcp->high) { unsigned long batch = READ_ONCE(pcp->batch); @@ -24722,7 +24718,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c } /* -@@ -2600,7 +2660,7 @@ +@@ -2600,7 +2660,7 @@ struct page *buffered_rmqueue(struct zone *preferred_zone, struct per_cpu_pages *pcp; struct list_head *list; @@ -24731,7 +24727,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c do { pcp = &this_cpu_ptr(zone->pageset)->pcp; list = &pcp->lists[migratetype]; -@@ -2627,7 +2687,7 @@ +@@ -2627,7 +2687,7 @@ struct page *buffered_rmqueue(struct zone *preferred_zone, * allocate greater than order-1 page units with __GFP_NOFAIL. */ WARN_ON_ONCE((gfp_flags & __GFP_NOFAIL) && (order > 1)); @@ -24740,7 +24736,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c do { page = NULL; -@@ -2639,22 +2699,24 @@ +@@ -2639,22 +2699,24 @@ struct page *buffered_rmqueue(struct zone *preferred_zone, if (!page) page = __rmqueue(zone, order, migratetype); } while (page && check_new_pages(page, order)); @@ -24769,7 +24765,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c return NULL; } -@@ -6531,7 +6593,9 @@ +@@ -6532,7 +6594,9 @@ static int page_alloc_cpu_notify(struct notifier_block *self, int cpu = (unsigned long)hcpu; if (action == CPU_DEAD || action == CPU_DEAD_FROZEN) { @@ -24779,7 +24775,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c drain_pages(cpu); /* -@@ -6557,6 +6621,7 @@ +@@ -6558,6 +6622,7 @@ static int page_alloc_cpu_notify(struct notifier_block *self, void __init page_alloc_init(void) { hotcpu_notifier(page_alloc_cpu_notify, 0); @@ -24787,7 +24783,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c } /* -@@ -7385,7 +7450,7 @@ +@@ -7386,7 +7451,7 @@ void zone_pcp_reset(struct zone *zone) struct per_cpu_pageset *pset; /* avoid races with drain_pages() */ @@ -24796,7 +24792,7 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c if (zone->pageset != &boot_pageset) { for_each_online_cpu(cpu) { pset = per_cpu_ptr(zone->pageset, cpu); -@@ -7394,7 +7459,7 @@ +@@ -7395,7 +7460,7 @@ void zone_pcp_reset(struct zone *zone) free_percpu(zone->pageset); zone->pageset = &boot_pageset; } @@ -24805,75 +24801,69 @@ diff -Nur linux-4.9.28.orig/mm/page_alloc.c linux-4.9.28/mm/page_alloc.c } #ifdef CONFIG_MEMORY_HOTREMOVE -diff -Nur linux-4.9.28.orig/mm/percpu.c linux-4.9.28/mm/percpu.c ---- linux-4.9.28.orig/mm/percpu.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/percpu.c 2017-05-19 03:37:25.210177882 +0200 -@@ -1283,18 +1283,7 @@ +diff --git a/mm/percpu.c b/mm/percpu.c +index f014cebbf405..4e739fcf91bf 100644 +--- a/mm/percpu.c ++++ b/mm/percpu.c +@@ -1283,6 +1283,31 @@ void free_percpu(void __percpu *ptr) } EXPORT_SYMBOL_GPL(free_percpu); --/** -- * is_kernel_percpu_address - test whether address is from static percpu area -- * @addr: address to test -- * -- * Test whether @addr belongs to in-kernel static percpu area. Module -- * static percpu areas are not considered. For those, use -- * is_module_percpu_address(). -- * -- * RETURNS: -- * %true if @addr is from in-kernel static percpu area, %false otherwise. -- */ --bool is_kernel_percpu_address(unsigned long addr) +bool __is_kernel_percpu_address(unsigned long addr, unsigned long *can_addr) - { - #ifdef CONFIG_SMP - const size_t static_size = __per_cpu_end - __per_cpu_start; -@@ -1303,16 +1292,39 @@ - - for_each_possible_cpu(cpu) { - void *start = per_cpu_ptr(base, cpu); ++{ ++#ifdef CONFIG_SMP ++ const size_t static_size = __per_cpu_end - __per_cpu_start; ++ void __percpu *base = __addr_to_pcpu_ptr(pcpu_base_addr); ++ unsigned int cpu; ++ ++ for_each_possible_cpu(cpu) { ++ void *start = per_cpu_ptr(base, cpu); + void *va = (void *)addr; - -- if ((void *)addr >= start && (void *)addr < start + static_size) ++ + if (va >= start && va < start + static_size) { + if (can_addr) { + *can_addr = (unsigned long) (va - start); + *can_addr += (unsigned long) + per_cpu_ptr(base, get_boot_cpu_id()); + } - return true; -- } ++ return true; + } + } - #endif - /* on UP, can't distinguish from other static vars, always false */ - return false; ++#endif ++ /* on UP, can't distinguish from other static vars, always false */ ++ return false; ++} ++ + /** + * is_kernel_percpu_address - test whether address is from static percpu area + * @addr: address to test +@@ -1296,20 +1321,7 @@ EXPORT_SYMBOL_GPL(free_percpu); + */ + bool is_kernel_percpu_address(unsigned long addr) + { +-#ifdef CONFIG_SMP +- const size_t static_size = __per_cpu_end - __per_cpu_start; +- void __percpu *base = __addr_to_pcpu_ptr(pcpu_base_addr); +- unsigned int cpu; +- +- for_each_possible_cpu(cpu) { +- void *start = per_cpu_ptr(base, cpu); +- +- if ((void *)addr >= start && (void *)addr < start + static_size) +- return true; +- } +-#endif +- /* on UP, can't distinguish from other static vars, always false */ +- return false; ++ return __is_kernel_percpu_address(addr, NULL); } /** -+ * is_kernel_percpu_address - test whether address is from static percpu area -+ * @addr: address to test -+ * -+ * Test whether @addr belongs to in-kernel static percpu area. Module -+ * static percpu areas are not considered. For those, use -+ * is_module_percpu_address(). -+ * -+ * RETURNS: -+ * %true if @addr is from in-kernel static percpu area, %false otherwise. -+ */ -+bool is_kernel_percpu_address(unsigned long addr) -+{ -+ return __is_kernel_percpu_address(addr, NULL); -+} -+ -+/** - * per_cpu_ptr_to_phys - convert translated percpu address to physical address - * @addr: the address to be converted to physical address - * -diff -Nur linux-4.9.28.orig/mm/slab.h linux-4.9.28/mm/slab.h ---- linux-4.9.28.orig/mm/slab.h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/slab.h 2017-05-19 03:37:25.210177882 +0200 -@@ -426,7 +426,11 @@ +diff --git a/mm/slab.h b/mm/slab.h +index ceb7d70cdb76..dfd281e43fbe 100644 +--- a/mm/slab.h ++++ b/mm/slab.h +@@ -426,7 +426,11 @@ static inline void slab_post_alloc_hook(struct kmem_cache *s, gfp_t flags, * The slab lists for all objects. */ struct kmem_cache_node { @@ -24885,10 +24875,11 @@ diff -Nur linux-4.9.28.orig/mm/slab.h linux-4.9.28/mm/slab.h #ifdef CONFIG_SLAB struct list_head slabs_partial; /* partial list first, better asm code */ -diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c ---- linux-4.9.28.orig/mm/slub.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/slub.c 2017-05-19 03:37:25.210177882 +0200 -@@ -1141,7 +1141,7 @@ +diff --git a/mm/slub.c b/mm/slub.c +index 58c7526f8de2..6d72b7f87129 100644 +--- a/mm/slub.c ++++ b/mm/slub.c +@@ -1141,7 +1141,7 @@ static noinline int free_debug_processing( unsigned long uninitialized_var(flags); int ret = 0; @@ -24897,7 +24888,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c slab_lock(page); if (s->flags & SLAB_CONSISTENCY_CHECKS) { -@@ -1176,7 +1176,7 @@ +@@ -1176,7 +1176,7 @@ static noinline int free_debug_processing( bulk_cnt, cnt); slab_unlock(page); @@ -24906,7 +24897,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c if (!ret) slab_fix(s, "Object at 0x%p not freed", object); return ret; -@@ -1304,6 +1304,12 @@ +@@ -1304,6 +1304,12 @@ static inline void dec_slabs_node(struct kmem_cache *s, int node, #endif /* CONFIG_SLUB_DEBUG */ @@ -24919,7 +24910,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c /* * Hooks for other subsystems that check memory allocations. In a typical * production configuration these hooks all should produce no code at all. -@@ -1527,10 +1533,17 @@ +@@ -1527,10 +1533,17 @@ static struct page *allocate_slab(struct kmem_cache *s, gfp_t flags, int node) void *start, *p; int idx, order; bool shuffle; @@ -24937,7 +24928,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c local_irq_enable(); flags |= s->allocflags; -@@ -1605,7 +1618,7 @@ +@@ -1605,7 +1618,7 @@ static struct page *allocate_slab(struct kmem_cache *s, gfp_t flags, int node) page->frozen = 1; out: @@ -24946,7 +24937,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c local_irq_disable(); if (!page) return NULL; -@@ -1664,6 +1677,16 @@ +@@ -1664,6 +1677,16 @@ static void __free_slab(struct kmem_cache *s, struct page *page) __free_pages(page, order); } @@ -24963,7 +24954,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c #define need_reserve_slab_rcu \ (sizeof(((struct page *)NULL)->lru) < sizeof(struct rcu_head)) -@@ -1695,6 +1718,12 @@ +@@ -1695,6 +1718,12 @@ static void free_slab(struct kmem_cache *s, struct page *page) } call_rcu(head, rcu_free_slab); @@ -24976,7 +24967,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c } else __free_slab(s, page); } -@@ -1802,7 +1831,7 @@ +@@ -1802,7 +1831,7 @@ static void *get_partial_node(struct kmem_cache *s, struct kmem_cache_node *n, if (!n || !n->nr_partial) return NULL; @@ -24985,7 +24976,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c list_for_each_entry_safe(page, page2, &n->partial, lru) { void *t; -@@ -1827,7 +1856,7 @@ +@@ -1827,7 +1856,7 @@ static void *get_partial_node(struct kmem_cache *s, struct kmem_cache_node *n, break; } @@ -24994,7 +24985,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c return object; } -@@ -2073,7 +2102,7 @@ +@@ -2073,7 +2102,7 @@ static void deactivate_slab(struct kmem_cache *s, struct page *page, * that acquire_slab() will see a slab page that * is frozen */ @@ -25003,7 +24994,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c } } else { m = M_FULL; -@@ -2084,7 +2113,7 @@ +@@ -2084,7 +2113,7 @@ static void deactivate_slab(struct kmem_cache *s, struct page *page, * slabs from diagnostic functions will not see * any frozen slabs. */ @@ -25012,7 +25003,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c } } -@@ -2119,7 +2148,7 @@ +@@ -2119,7 +2148,7 @@ static void deactivate_slab(struct kmem_cache *s, struct page *page, goto redo; if (lock) @@ -25021,7 +25012,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c if (m == M_FREE) { stat(s, DEACTIVATE_EMPTY); -@@ -2151,10 +2180,10 @@ +@@ -2151,10 +2180,10 @@ static void unfreeze_partials(struct kmem_cache *s, n2 = get_node(s, page_to_nid(page)); if (n != n2) { if (n) @@ -25034,7 +25025,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c } do { -@@ -2183,7 +2212,7 @@ +@@ -2183,7 +2212,7 @@ static void unfreeze_partials(struct kmem_cache *s, } if (n) @@ -25043,7 +25034,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c while (discard_page) { page = discard_page; -@@ -2222,14 +2251,21 @@ +@@ -2222,14 +2251,21 @@ static void put_cpu_partial(struct kmem_cache *s, struct page *page, int drain) pobjects = oldpage->pobjects; pages = oldpage->pages; if (drain && pobjects > s->cpu_partial) { @@ -25065,7 +25056,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c oldpage = NULL; pobjects = 0; pages = 0; -@@ -2301,7 +2337,22 @@ +@@ -2301,7 +2337,22 @@ static bool has_cpu_slab(int cpu, void *info) static void flush_all(struct kmem_cache *s) { @@ -25088,7 +25079,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c } /* -@@ -2356,10 +2407,10 @@ +@@ -2356,10 +2407,10 @@ static unsigned long count_partial(struct kmem_cache_node *n, unsigned long x = 0; struct page *page; @@ -25101,7 +25092,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c return x; } #endif /* CONFIG_SLUB_DEBUG || CONFIG_SYSFS */ -@@ -2497,8 +2548,10 @@ +@@ -2497,8 +2548,10 @@ static inline void *get_freelist(struct kmem_cache *s, struct page *page) * already disabled (which is the case for bulk allocation). */ static void *___slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, @@ -25113,7 +25104,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c void *freelist; struct page *page; -@@ -2558,6 +2611,13 @@ +@@ -2558,6 +2611,13 @@ static void *___slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, VM_BUG_ON(!c->page->frozen); c->freelist = get_freepointer(s, freelist); c->tid = next_tid(c->tid); @@ -25127,7 +25118,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c return freelist; new_slab: -@@ -2589,7 +2649,7 @@ +@@ -2589,7 +2649,7 @@ static void *___slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, deactivate_slab(s, page, get_freepointer(s, freelist)); c->page = NULL; c->freelist = NULL; @@ -25136,7 +25127,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c } /* -@@ -2601,6 +2661,7 @@ +@@ -2601,6 +2661,7 @@ static void *__slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, { void *p; unsigned long flags; @@ -25144,7 +25135,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c local_irq_save(flags); #ifdef CONFIG_PREEMPT -@@ -2612,8 +2673,9 @@ +@@ -2612,8 +2673,9 @@ static void *__slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, c = this_cpu_ptr(s->cpu_slab); #endif @@ -25155,7 +25146,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c return p; } -@@ -2799,7 +2861,7 @@ +@@ -2799,7 +2861,7 @@ static void __slab_free(struct kmem_cache *s, struct page *page, do { if (unlikely(n)) { @@ -25164,7 +25155,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c n = NULL; } prior = page->freelist; -@@ -2831,7 +2893,7 @@ +@@ -2831,7 +2893,7 @@ static void __slab_free(struct kmem_cache *s, struct page *page, * Otherwise the list_lock will synchronize with * other processors updating the list of slabs. */ @@ -25173,7 +25164,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c } } -@@ -2873,7 +2935,7 @@ +@@ -2873,7 +2935,7 @@ static void __slab_free(struct kmem_cache *s, struct page *page, add_partial(n, page, DEACTIVATE_TO_TAIL); stat(s, FREE_ADD_PARTIAL); } @@ -25182,7 +25173,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c return; slab_empty: -@@ -2888,7 +2950,7 @@ +@@ -2888,7 +2950,7 @@ static void __slab_free(struct kmem_cache *s, struct page *page, remove_full(s, n, page); } @@ -25191,7 +25182,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c stat(s, FREE_SLAB); discard_slab(s, page); } -@@ -3093,6 +3155,7 @@ +@@ -3093,6 +3155,7 @@ int kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, void **p) { struct kmem_cache_cpu *c; @@ -25199,7 +25190,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c int i; /* memcg and kmem_cache debug support */ -@@ -3116,7 +3179,7 @@ +@@ -3116,7 +3179,7 @@ int kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, * of re-populating per CPU c->freelist */ p[i] = ___slab_alloc(s, flags, NUMA_NO_NODE, @@ -25208,7 +25199,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c if (unlikely(!p[i])) goto error; -@@ -3128,6 +3191,7 @@ +@@ -3128,6 +3191,7 @@ int kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, } c->tid = next_tid(c->tid); local_irq_enable(); @@ -25216,7 +25207,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c /* Clear memory outside IRQ disabled fastpath loop */ if (unlikely(flags & __GFP_ZERO)) { -@@ -3275,7 +3339,7 @@ +@@ -3275,7 +3339,7 @@ static void init_kmem_cache_node(struct kmem_cache_node *n) { n->nr_partial = 0; @@ -25225,7 +25216,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c INIT_LIST_HEAD(&n->partial); #ifdef CONFIG_SLUB_DEBUG atomic_long_set(&n->nr_slabs, 0); -@@ -3619,6 +3683,10 @@ +@@ -3619,6 +3683,10 @@ static void list_slab_objects(struct kmem_cache *s, struct page *page, const char *text) { #ifdef CONFIG_SLUB_DEBUG @@ -25236,7 +25227,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c void *addr = page_address(page); void *p; unsigned long *map = kzalloc(BITS_TO_LONGS(page->objects) * -@@ -3639,6 +3707,7 @@ +@@ -3639,6 +3707,7 @@ static void list_slab_objects(struct kmem_cache *s, struct page *page, slab_unlock(page); kfree(map); #endif @@ -25244,7 +25235,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c } /* -@@ -3652,7 +3721,7 @@ +@@ -3652,7 +3721,7 @@ static void free_partial(struct kmem_cache *s, struct kmem_cache_node *n) struct page *page, *h; BUG_ON(irqs_disabled()); @@ -25253,7 +25244,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c list_for_each_entry_safe(page, h, &n->partial, lru) { if (!page->inuse) { remove_partial(n, page); -@@ -3662,7 +3731,7 @@ +@@ -3662,7 +3731,7 @@ static void free_partial(struct kmem_cache *s, struct kmem_cache_node *n) "Objects remaining in %s on __kmem_cache_shutdown()"); } } @@ -25262,7 +25253,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c list_for_each_entry_safe(page, h, &discard, lru) discard_slab(s, page); -@@ -3905,7 +3974,7 @@ +@@ -3905,7 +3974,7 @@ int __kmem_cache_shrink(struct kmem_cache *s) for (i = 0; i < SHRINK_PROMOTE_MAX; i++) INIT_LIST_HEAD(promote + i); @@ -25271,7 +25262,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c /* * Build lists of slabs to discard or promote. -@@ -3936,7 +4005,7 @@ +@@ -3936,7 +4005,7 @@ int __kmem_cache_shrink(struct kmem_cache *s) for (i = SHRINK_PROMOTE_MAX - 1; i >= 0; i--) list_splice(promote + i, &n->partial); @@ -25280,7 +25271,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c /* Release empty slabs */ list_for_each_entry_safe(page, t, &discard, lru) -@@ -4112,6 +4181,12 @@ +@@ -4112,6 +4181,12 @@ void __init kmem_cache_init(void) { static __initdata struct kmem_cache boot_kmem_cache, boot_kmem_cache_node; @@ -25293,7 +25284,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c if (debug_guardpage_minorder()) slub_max_order = 0; -@@ -4320,7 +4395,7 @@ +@@ -4320,7 +4395,7 @@ static int validate_slab_node(struct kmem_cache *s, struct page *page; unsigned long flags; @@ -25302,7 +25293,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c list_for_each_entry(page, &n->partial, lru) { validate_slab_slab(s, page, map); -@@ -4342,7 +4417,7 @@ +@@ -4342,7 +4417,7 @@ static int validate_slab_node(struct kmem_cache *s, s->name, count, atomic_long_read(&n->nr_slabs)); out: @@ -25311,7 +25302,7 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c return count; } -@@ -4530,12 +4605,12 @@ +@@ -4530,12 +4605,12 @@ static int list_locations(struct kmem_cache *s, char *buf, if (!atomic_long_read(&n->nr_slabs)) continue; @@ -25326,9 +25317,10 @@ diff -Nur linux-4.9.28.orig/mm/slub.c linux-4.9.28/mm/slub.c } for (i = 0; i < t.count; i++) { -diff -Nur linux-4.9.28.orig/mm/swap.c linux-4.9.28/mm/swap.c ---- linux-4.9.28.orig/mm/swap.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/swap.c 2017-05-19 03:37:25.210177882 +0200 +diff --git a/mm/swap.c b/mm/swap.c +index 4dcf852e1e6d..69c3a5b24060 100644 +--- a/mm/swap.c ++++ b/mm/swap.c @@ -32,6 +32,7 @@ #include #include @@ -25337,7 +25329,7 @@ diff -Nur linux-4.9.28.orig/mm/swap.c linux-4.9.28/mm/swap.c #include #include -@@ -50,6 +51,8 @@ +@@ -50,6 +51,8 @@ static DEFINE_PER_CPU(struct pagevec, lru_deactivate_pvecs); #ifdef CONFIG_SMP static DEFINE_PER_CPU(struct pagevec, activate_page_pvecs); #endif @@ -25346,7 +25338,7 @@ diff -Nur linux-4.9.28.orig/mm/swap.c linux-4.9.28/mm/swap.c /* * This path almost never happens for VM activity - pages are normally -@@ -240,11 +243,11 @@ +@@ -240,11 +243,11 @@ void rotate_reclaimable_page(struct page *page) unsigned long flags; get_page(page); @@ -25360,7 +25352,7 @@ diff -Nur linux-4.9.28.orig/mm/swap.c linux-4.9.28/mm/swap.c } } -@@ -294,12 +297,13 @@ +@@ -294,12 +297,13 @@ void activate_page(struct page *page) { page = compound_head(page); if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) { @@ -25376,7 +25368,7 @@ diff -Nur linux-4.9.28.orig/mm/swap.c linux-4.9.28/mm/swap.c } } -@@ -326,7 +330,7 @@ +@@ -326,7 +330,7 @@ void activate_page(struct page *page) static void __lru_cache_activate_page(struct page *page) { @@ -25385,7 +25377,7 @@ diff -Nur linux-4.9.28.orig/mm/swap.c linux-4.9.28/mm/swap.c int i; /* -@@ -348,7 +352,7 @@ +@@ -348,7 +352,7 @@ static void __lru_cache_activate_page(struct page *page) } } @@ -25394,7 +25386,7 @@ diff -Nur linux-4.9.28.orig/mm/swap.c linux-4.9.28/mm/swap.c } /* -@@ -390,12 +394,12 @@ +@@ -390,12 +394,12 @@ EXPORT_SYMBOL(mark_page_accessed); static void __lru_cache_add(struct page *page) { @@ -25409,7 +25401,7 @@ diff -Nur linux-4.9.28.orig/mm/swap.c linux-4.9.28/mm/swap.c } /** -@@ -593,9 +597,15 @@ +@@ -593,9 +597,15 @@ void lru_add_drain_cpu(int cpu) unsigned long flags; /* No harm done if a racing interrupt already did this */ @@ -25427,7 +25419,7 @@ diff -Nur linux-4.9.28.orig/mm/swap.c linux-4.9.28/mm/swap.c } pvec = &per_cpu(lru_deactivate_file_pvecs, cpu); -@@ -627,11 +637,12 @@ +@@ -627,11 +637,12 @@ void deactivate_file_page(struct page *page) return; if (likely(get_page_unless_zero(page))) { @@ -25442,7 +25434,7 @@ diff -Nur linux-4.9.28.orig/mm/swap.c linux-4.9.28/mm/swap.c } } -@@ -646,27 +657,31 @@ +@@ -646,27 +657,31 @@ void deactivate_file_page(struct page *page) void deactivate_page(struct page *page) { if (PageLRU(page) && PageActive(page) && !PageUnevictable(page)) { @@ -25481,7 +25473,7 @@ diff -Nur linux-4.9.28.orig/mm/swap.c linux-4.9.28/mm/swap.c /* * lru_add_drain_wq is used to do lru_add_drain_all() from a WQ_MEM_RECLAIM -@@ -686,6 +701,22 @@ +@@ -686,6 +701,22 @@ static int __init lru_init(void) } early_initcall(lru_init); @@ -25504,7 +25496,7 @@ diff -Nur linux-4.9.28.orig/mm/swap.c linux-4.9.28/mm/swap.c void lru_add_drain_all(void) { static DEFINE_MUTEX(lock); -@@ -697,21 +728,18 @@ +@@ -697,21 +728,18 @@ void lru_add_drain_all(void) cpumask_clear(&has_work); for_each_online_cpu(cpu) { @@ -25530,10 +25522,11 @@ diff -Nur linux-4.9.28.orig/mm/swap.c linux-4.9.28/mm/swap.c put_online_cpus(); mutex_unlock(&lock); -diff -Nur linux-4.9.28.orig/mm/truncate.c linux-4.9.28/mm/truncate.c ---- linux-4.9.28.orig/mm/truncate.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/truncate.c 2017-05-19 03:37:25.214178037 +0200 -@@ -62,9 +62,12 @@ +diff --git a/mm/truncate.c b/mm/truncate.c +index 8d8c62d89e6d..5bf1bd25d077 100644 +--- a/mm/truncate.c ++++ b/mm/truncate.c +@@ -62,9 +62,12 @@ static void clear_exceptional_entry(struct address_space *mapping, * protected by mapping->tree_lock. */ if (!workingset_node_shadows(node) && @@ -25548,10 +25541,11 @@ diff -Nur linux-4.9.28.orig/mm/truncate.c linux-4.9.28/mm/truncate.c __radix_tree_delete_node(&mapping->page_tree, node); unlock: spin_unlock_irq(&mapping->tree_lock); -diff -Nur linux-4.9.28.orig/mm/vmalloc.c linux-4.9.28/mm/vmalloc.c ---- linux-4.9.28.orig/mm/vmalloc.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/vmalloc.c 2017-05-19 03:37:25.214178037 +0200 -@@ -845,7 +845,7 @@ +diff --git a/mm/vmalloc.c b/mm/vmalloc.c +index f2481cb4e6b2..db4de08fa97c 100644 +--- a/mm/vmalloc.c ++++ b/mm/vmalloc.c +@@ -845,7 +845,7 @@ static void *new_vmap_block(unsigned int order, gfp_t gfp_mask) struct vmap_block *vb; struct vmap_area *va; unsigned long vb_idx; @@ -25560,7 +25554,7 @@ diff -Nur linux-4.9.28.orig/mm/vmalloc.c linux-4.9.28/mm/vmalloc.c void *vaddr; node = numa_node_id(); -@@ -888,11 +888,12 @@ +@@ -888,11 +888,12 @@ static void *new_vmap_block(unsigned int order, gfp_t gfp_mask) BUG_ON(err); radix_tree_preload_end(); @@ -25575,7 +25569,7 @@ diff -Nur linux-4.9.28.orig/mm/vmalloc.c linux-4.9.28/mm/vmalloc.c return vaddr; } -@@ -961,6 +962,7 @@ +@@ -961,6 +962,7 @@ static void *vb_alloc(unsigned long size, gfp_t gfp_mask) struct vmap_block *vb; void *vaddr = NULL; unsigned int order; @@ -25583,7 +25577,7 @@ diff -Nur linux-4.9.28.orig/mm/vmalloc.c linux-4.9.28/mm/vmalloc.c BUG_ON(offset_in_page(size)); BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC); -@@ -975,7 +977,8 @@ +@@ -975,7 +977,8 @@ static void *vb_alloc(unsigned long size, gfp_t gfp_mask) order = get_order(size); rcu_read_lock(); @@ -25593,7 +25587,7 @@ diff -Nur linux-4.9.28.orig/mm/vmalloc.c linux-4.9.28/mm/vmalloc.c list_for_each_entry_rcu(vb, &vbq->free, free_list) { unsigned long pages_off; -@@ -998,7 +1001,7 @@ +@@ -998,7 +1001,7 @@ static void *vb_alloc(unsigned long size, gfp_t gfp_mask) break; } @@ -25602,10 +25596,11 @@ diff -Nur linux-4.9.28.orig/mm/vmalloc.c linux-4.9.28/mm/vmalloc.c rcu_read_unlock(); /* Allocate new block if nothing was found */ -diff -Nur linux-4.9.28.orig/mm/vmstat.c linux-4.9.28/mm/vmstat.c ---- linux-4.9.28.orig/mm/vmstat.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/vmstat.c 2017-05-19 03:37:25.214178037 +0200 -@@ -245,6 +245,7 @@ +diff --git a/mm/vmstat.c b/mm/vmstat.c +index 604f26a4f696..312006d2db50 100644 +--- a/mm/vmstat.c ++++ b/mm/vmstat.c +@@ -245,6 +245,7 @@ void __mod_zone_page_state(struct zone *zone, enum zone_stat_item item, long x; long t; @@ -25613,7 +25608,7 @@ diff -Nur linux-4.9.28.orig/mm/vmstat.c linux-4.9.28/mm/vmstat.c x = delta + __this_cpu_read(*p); t = __this_cpu_read(pcp->stat_threshold); -@@ -254,6 +255,7 @@ +@@ -254,6 +255,7 @@ void __mod_zone_page_state(struct zone *zone, enum zone_stat_item item, x = 0; } __this_cpu_write(*p, x); @@ -25621,7 +25616,7 @@ diff -Nur linux-4.9.28.orig/mm/vmstat.c linux-4.9.28/mm/vmstat.c } EXPORT_SYMBOL(__mod_zone_page_state); -@@ -265,6 +267,7 @@ +@@ -265,6 +267,7 @@ void __mod_node_page_state(struct pglist_data *pgdat, enum node_stat_item item, long x; long t; @@ -25629,7 +25624,7 @@ diff -Nur linux-4.9.28.orig/mm/vmstat.c linux-4.9.28/mm/vmstat.c x = delta + __this_cpu_read(*p); t = __this_cpu_read(pcp->stat_threshold); -@@ -274,6 +277,7 @@ +@@ -274,6 +277,7 @@ void __mod_node_page_state(struct pglist_data *pgdat, enum node_stat_item item, x = 0; } __this_cpu_write(*p, x); @@ -25637,7 +25632,7 @@ diff -Nur linux-4.9.28.orig/mm/vmstat.c linux-4.9.28/mm/vmstat.c } EXPORT_SYMBOL(__mod_node_page_state); -@@ -306,6 +310,7 @@ +@@ -306,6 +310,7 @@ void __inc_zone_state(struct zone *zone, enum zone_stat_item item) s8 __percpu *p = pcp->vm_stat_diff + item; s8 v, t; @@ -25645,7 +25640,7 @@ diff -Nur linux-4.9.28.orig/mm/vmstat.c linux-4.9.28/mm/vmstat.c v = __this_cpu_inc_return(*p); t = __this_cpu_read(pcp->stat_threshold); if (unlikely(v > t)) { -@@ -314,6 +319,7 @@ +@@ -314,6 +319,7 @@ void __inc_zone_state(struct zone *zone, enum zone_stat_item item) zone_page_state_add(v + overstep, zone, item); __this_cpu_write(*p, -overstep); } @@ -25653,7 +25648,7 @@ diff -Nur linux-4.9.28.orig/mm/vmstat.c linux-4.9.28/mm/vmstat.c } void __inc_node_state(struct pglist_data *pgdat, enum node_stat_item item) -@@ -322,6 +328,7 @@ +@@ -322,6 +328,7 @@ void __inc_node_state(struct pglist_data *pgdat, enum node_stat_item item) s8 __percpu *p = pcp->vm_node_stat_diff + item; s8 v, t; @@ -25661,7 +25656,7 @@ diff -Nur linux-4.9.28.orig/mm/vmstat.c linux-4.9.28/mm/vmstat.c v = __this_cpu_inc_return(*p); t = __this_cpu_read(pcp->stat_threshold); if (unlikely(v > t)) { -@@ -330,6 +337,7 @@ +@@ -330,6 +337,7 @@ void __inc_node_state(struct pglist_data *pgdat, enum node_stat_item item) node_page_state_add(v + overstep, pgdat, item); __this_cpu_write(*p, -overstep); } @@ -25669,7 +25664,7 @@ diff -Nur linux-4.9.28.orig/mm/vmstat.c linux-4.9.28/mm/vmstat.c } void __inc_zone_page_state(struct page *page, enum zone_stat_item item) -@@ -350,6 +358,7 @@ +@@ -350,6 +358,7 @@ void __dec_zone_state(struct zone *zone, enum zone_stat_item item) s8 __percpu *p = pcp->vm_stat_diff + item; s8 v, t; @@ -25677,7 +25672,7 @@ diff -Nur linux-4.9.28.orig/mm/vmstat.c linux-4.9.28/mm/vmstat.c v = __this_cpu_dec_return(*p); t = __this_cpu_read(pcp->stat_threshold); if (unlikely(v < - t)) { -@@ -358,6 +367,7 @@ +@@ -358,6 +367,7 @@ void __dec_zone_state(struct zone *zone, enum zone_stat_item item) zone_page_state_add(v - overstep, zone, item); __this_cpu_write(*p, overstep); } @@ -25685,7 +25680,7 @@ diff -Nur linux-4.9.28.orig/mm/vmstat.c linux-4.9.28/mm/vmstat.c } void __dec_node_state(struct pglist_data *pgdat, enum node_stat_item item) -@@ -366,6 +376,7 @@ +@@ -366,6 +376,7 @@ void __dec_node_state(struct pglist_data *pgdat, enum node_stat_item item) s8 __percpu *p = pcp->vm_node_stat_diff + item; s8 v, t; @@ -25693,7 +25688,7 @@ diff -Nur linux-4.9.28.orig/mm/vmstat.c linux-4.9.28/mm/vmstat.c v = __this_cpu_dec_return(*p); t = __this_cpu_read(pcp->stat_threshold); if (unlikely(v < - t)) { -@@ -374,6 +385,7 @@ +@@ -374,6 +385,7 @@ void __dec_node_state(struct pglist_data *pgdat, enum node_stat_item item) node_page_state_add(v - overstep, pgdat, item); __this_cpu_write(*p, overstep); } @@ -25701,10 +25696,11 @@ diff -Nur linux-4.9.28.orig/mm/vmstat.c linux-4.9.28/mm/vmstat.c } void __dec_zone_page_state(struct page *page, enum zone_stat_item item) -diff -Nur linux-4.9.28.orig/mm/workingset.c linux-4.9.28/mm/workingset.c ---- linux-4.9.28.orig/mm/workingset.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/workingset.c 2017-05-19 03:37:25.214178037 +0200 -@@ -334,7 +334,8 @@ +diff --git a/mm/workingset.c b/mm/workingset.c +index 4c4f05655e6e..b97b1e87b54c 100644 +--- a/mm/workingset.c ++++ b/mm/workingset.c +@@ -334,7 +334,8 @@ void workingset_activation(struct page *page) * point where they would still be useful. */ @@ -25714,7 +25710,7 @@ diff -Nur linux-4.9.28.orig/mm/workingset.c linux-4.9.28/mm/workingset.c static unsigned long count_shadow_nodes(struct shrinker *shrinker, struct shrink_control *sc) -@@ -344,9 +345,9 @@ +@@ -344,9 +345,9 @@ static unsigned long count_shadow_nodes(struct shrinker *shrinker, unsigned long pages; /* list_lru lock nests inside IRQ-safe mapping->tree_lock */ @@ -25727,7 +25723,7 @@ diff -Nur linux-4.9.28.orig/mm/workingset.c linux-4.9.28/mm/workingset.c if (sc->memcg) { pages = mem_cgroup_node_nr_lru_pages(sc->memcg, sc->nid, -@@ -438,9 +439,9 @@ +@@ -438,9 +439,9 @@ static enum lru_status shadow_lru_isolate(struct list_head *item, spin_unlock(&mapping->tree_lock); ret = LRU_REMOVED_RETRY; out: @@ -25739,7 +25735,7 @@ diff -Nur linux-4.9.28.orig/mm/workingset.c linux-4.9.28/mm/workingset.c spin_lock(lru_lock); return ret; } -@@ -451,10 +452,10 @@ +@@ -451,10 +452,10 @@ static unsigned long scan_shadow_nodes(struct shrinker *shrinker, unsigned long ret; /* list_lru lock nests inside IRQ-safe mapping->tree_lock */ @@ -25753,7 +25749,7 @@ diff -Nur linux-4.9.28.orig/mm/workingset.c linux-4.9.28/mm/workingset.c return ret; } -@@ -492,7 +493,7 @@ +@@ -492,7 +493,7 @@ static int __init workingset_init(void) pr_info("workingset: timestamp_bits=%d max_order=%d bucket_order=%u\n", timestamp_bits, max_order, bucket_order); @@ -25762,7 +25758,7 @@ diff -Nur linux-4.9.28.orig/mm/workingset.c linux-4.9.28/mm/workingset.c if (ret) goto err; ret = register_shrinker(&workingset_shadow_shrinker); -@@ -500,7 +501,7 @@ +@@ -500,7 +501,7 @@ static int __init workingset_init(void) goto err_list_lru; return 0; err_list_lru: @@ -25771,9 +25767,10 @@ diff -Nur linux-4.9.28.orig/mm/workingset.c linux-4.9.28/mm/workingset.c err: return ret; } -diff -Nur linux-4.9.28.orig/mm/zsmalloc.c linux-4.9.28/mm/zsmalloc.c ---- linux-4.9.28.orig/mm/zsmalloc.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/mm/zsmalloc.c 2017-05-19 03:37:25.214178037 +0200 +diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c +index 1689bb58e0d1..e52a8cb6aa5a 100644 +--- a/mm/zsmalloc.c ++++ b/mm/zsmalloc.c @@ -53,6 +53,7 @@ #include #include @@ -25806,7 +25803,7 @@ diff -Nur linux-4.9.28.orig/mm/zsmalloc.c linux-4.9.28/mm/zsmalloc.c /* * Object location (, ) is encoded as * as single (unsigned long) handle value. -@@ -327,7 +341,7 @@ +@@ -327,7 +341,7 @@ static void SetZsPageMovable(struct zs_pool *pool, struct zspage *zspage) {} static int create_cache(struct zs_pool *pool) { @@ -25815,7 +25812,7 @@ diff -Nur linux-4.9.28.orig/mm/zsmalloc.c linux-4.9.28/mm/zsmalloc.c 0, 0, NULL); if (!pool->handle_cachep) return 1; -@@ -351,10 +365,27 @@ +@@ -351,10 +365,27 @@ static void destroy_cache(struct zs_pool *pool) static unsigned long cache_alloc_handle(struct zs_pool *pool, gfp_t gfp) { @@ -25845,7 +25842,7 @@ diff -Nur linux-4.9.28.orig/mm/zsmalloc.c linux-4.9.28/mm/zsmalloc.c static void cache_free_handle(struct zs_pool *pool, unsigned long handle) { kmem_cache_free(pool->handle_cachep, (void *)handle); -@@ -373,12 +404,18 @@ +@@ -373,12 +404,18 @@ static void cache_free_zspage(struct zs_pool *pool, struct zspage *zspage) static void record_obj(unsigned long handle, unsigned long obj) { @@ -25864,7 +25861,7 @@ diff -Nur linux-4.9.28.orig/mm/zsmalloc.c linux-4.9.28/mm/zsmalloc.c } /* zpool driver */ -@@ -467,6 +504,7 @@ +@@ -467,6 +504,7 @@ MODULE_ALIAS("zpool-zsmalloc"); /* per-cpu VM mapping areas for zspage accesses that cross page boundaries */ static DEFINE_PER_CPU(struct mapping_area, zs_map_area); @@ -25872,7 +25869,7 @@ diff -Nur linux-4.9.28.orig/mm/zsmalloc.c linux-4.9.28/mm/zsmalloc.c static bool is_zspage_isolated(struct zspage *zspage) { -@@ -902,7 +940,13 @@ +@@ -902,7 +940,13 @@ static unsigned long location_to_obj(struct page *page, unsigned int obj_idx) static unsigned long handle_to_obj(unsigned long handle) { @@ -25886,7 +25883,7 @@ diff -Nur linux-4.9.28.orig/mm/zsmalloc.c linux-4.9.28/mm/zsmalloc.c } static unsigned long obj_to_head(struct page *page, void *obj) -@@ -916,22 +960,46 @@ +@@ -916,22 +960,46 @@ static unsigned long obj_to_head(struct page *page, void *obj) static inline int testpin_tag(unsigned long handle) { @@ -25933,7 +25930,7 @@ diff -Nur linux-4.9.28.orig/mm/zsmalloc.c linux-4.9.28/mm/zsmalloc.c } static void reset_page(struct page *page) -@@ -1423,7 +1491,7 @@ +@@ -1423,7 +1491,7 @@ void *zs_map_object(struct zs_pool *pool, unsigned long handle, class = pool->size_class[class_idx]; off = (class->size * obj_idx) & ~PAGE_MASK; @@ -25942,7 +25939,7 @@ diff -Nur linux-4.9.28.orig/mm/zsmalloc.c linux-4.9.28/mm/zsmalloc.c area->vm_mm = mm; if (off + class->size <= PAGE_SIZE) { /* this object is contained entirely within a page */ -@@ -1477,7 +1545,7 @@ +@@ -1477,7 +1545,7 @@ void zs_unmap_object(struct zs_pool *pool, unsigned long handle) __zs_unmap_object(area, pages, off, class->size); } @@ -25951,10 +25948,11 @@ diff -Nur linux-4.9.28.orig/mm/zsmalloc.c linux-4.9.28/mm/zsmalloc.c migrate_read_unlock(zspage); unpin_tag(handle); -diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c ---- linux-4.9.28.orig/net/core/dev.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/core/dev.c 2017-05-19 03:37:25.214178037 +0200 -@@ -190,6 +190,7 @@ +diff --git a/net/core/dev.c b/net/core/dev.c +index 2e04fd188081..3ba60ef8c79e 100644 +--- a/net/core/dev.c ++++ b/net/core/dev.c +@@ -190,6 +190,7 @@ static unsigned int napi_gen_id = NR_CPUS; static DEFINE_READ_MOSTLY_HASHTABLE(napi_hash, 8); static seqcount_t devnet_rename_seq; @@ -25962,7 +25960,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c static inline void dev_base_seq_inc(struct net *net) { -@@ -211,14 +212,14 @@ +@@ -211,14 +212,14 @@ static inline struct hlist_head *dev_index_hash(struct net *net, int ifindex) static inline void rps_lock(struct softnet_data *sd) { #ifdef CONFIG_RPS @@ -25979,7 +25977,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c #endif } -@@ -888,7 +889,8 @@ +@@ -888,7 +889,8 @@ int netdev_get_name(struct net *net, char *name, int ifindex) strcpy(name, dev->name); rcu_read_unlock(); if (read_seqcount_retry(&devnet_rename_seq, seq)) { @@ -25989,7 +25987,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c goto retry; } -@@ -1157,20 +1159,17 @@ +@@ -1157,20 +1159,17 @@ int dev_change_name(struct net_device *dev, const char *newname) if (dev->flags & IFF_UP) return -EBUSY; @@ -26016,7 +26014,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c if (oldname[0] && !strchr(oldname, '%')) netdev_info(dev, "renamed from %s\n", oldname); -@@ -1183,11 +1182,12 @@ +@@ -1183,11 +1182,12 @@ int dev_change_name(struct net_device *dev, const char *newname) if (ret) { memcpy(dev->name, oldname, IFNAMSIZ); dev->name_assign_type = old_assign_type; @@ -26032,7 +26030,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c netdev_adjacent_rename_links(dev, oldname); -@@ -1208,7 +1208,8 @@ +@@ -1208,7 +1208,8 @@ int dev_change_name(struct net_device *dev, const char *newname) /* err >= 0 after dev_alloc_name() or stores the first errno */ if (err >= 0) { err = ret; @@ -26042,7 +26040,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c memcpy(dev->name, oldname, IFNAMSIZ); memcpy(oldname, newname, IFNAMSIZ); dev->name_assign_type = old_assign_type; -@@ -1221,6 +1222,11 @@ +@@ -1221,6 +1222,11 @@ int dev_change_name(struct net_device *dev, const char *newname) } return err; @@ -26054,7 +26052,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c } /** -@@ -2285,6 +2291,7 @@ +@@ -2285,6 +2291,7 @@ static void __netif_reschedule(struct Qdisc *q) sd->output_queue_tailp = &q->next_sched; raise_softirq_irqoff(NET_TX_SOFTIRQ); local_irq_restore(flags); @@ -26062,7 +26060,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c } void __netif_schedule(struct Qdisc *q) -@@ -2366,6 +2373,7 @@ +@@ -2366,6 +2373,7 @@ void __dev_kfree_skb_irq(struct sk_buff *skb, enum skb_free_reason reason) __this_cpu_write(softnet_data.completion_queue, skb); raise_softirq_irqoff(NET_TX_SOFTIRQ); local_irq_restore(flags); @@ -26070,7 +26068,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c } EXPORT_SYMBOL(__dev_kfree_skb_irq); -@@ -3100,7 +3108,11 @@ +@@ -3100,7 +3108,11 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, * This permits qdisc->running owner to get the lock more * often and dequeue packets faster. */ @@ -26082,7 +26080,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c if (unlikely(contended)) spin_lock(&q->busylock); -@@ -3163,8 +3175,10 @@ +@@ -3163,8 +3175,10 @@ static void skb_update_prio(struct sk_buff *skb) #define skb_update_prio(skb) #endif @@ -26093,7 +26091,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c /** * dev_loopback_xmit - loop back @skb -@@ -3398,8 +3412,7 @@ +@@ -3398,8 +3412,7 @@ static int __dev_queue_xmit(struct sk_buff *skb, void *accel_priv) int cpu = smp_processor_id(); /* ok because BHs are off */ if (txq->xmit_lock_owner != cpu) { @@ -26103,7 +26101,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c goto recursion_alert; skb = validate_xmit_skb(skb, dev); -@@ -3409,9 +3422,9 @@ +@@ -3409,9 +3422,9 @@ static int __dev_queue_xmit(struct sk_buff *skb, void *accel_priv) HARD_TX_LOCK(dev, txq, cpu); if (!netif_xmit_stopped(txq)) { @@ -26115,7 +26113,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c if (dev_xmit_complete(rc)) { HARD_TX_UNLOCK(dev, txq); goto out; -@@ -3785,6 +3798,7 @@ +@@ -3785,6 +3798,7 @@ static int enqueue_to_backlog(struct sk_buff *skb, int cpu, rps_unlock(sd); local_irq_restore(flags); @@ -26123,7 +26121,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c atomic_long_inc(&skb->dev->rx_dropped); kfree_skb(skb); -@@ -3803,7 +3817,7 @@ +@@ -3803,7 +3817,7 @@ static int netif_rx_internal(struct sk_buff *skb) struct rps_dev_flow voidflow, *rflow = &voidflow; int cpu; @@ -26132,7 +26130,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c rcu_read_lock(); cpu = get_rps_cpu(skb->dev, skb, &rflow); -@@ -3813,13 +3827,13 @@ +@@ -3813,13 +3827,13 @@ static int netif_rx_internal(struct sk_buff *skb) ret = enqueue_to_backlog(skb, cpu, &rflow->last_qtail); rcu_read_unlock(); @@ -26149,7 +26147,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c } return ret; } -@@ -3853,11 +3867,9 @@ +@@ -3853,11 +3867,9 @@ int netif_rx_ni(struct sk_buff *skb) trace_netif_rx_ni_entry(skb); @@ -26163,7 +26161,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c return err; } -@@ -4336,7 +4348,7 @@ +@@ -4336,7 +4348,7 @@ static void flush_backlog(struct work_struct *work) skb_queue_walk_safe(&sd->input_pkt_queue, skb, tmp) { if (skb->dev->reg_state == NETREG_UNREGISTERING) { __skb_unlink(skb, &sd->input_pkt_queue); @@ -26172,7 +26170,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c input_queue_head_incr(sd); } } -@@ -4346,11 +4358,14 @@ +@@ -4346,11 +4358,14 @@ static void flush_backlog(struct work_struct *work) skb_queue_walk_safe(&sd->process_queue, skb, tmp) { if (skb->dev->reg_state == NETREG_UNREGISTERING) { __skb_unlink(skb, &sd->process_queue); @@ -26188,7 +26186,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c } static void flush_all_backlogs(void) -@@ -4831,6 +4846,7 @@ +@@ -4831,6 +4846,7 @@ static void net_rps_action_and_irq_enable(struct softnet_data *sd) sd->rps_ipi_list = NULL; local_irq_enable(); @@ -26196,7 +26194,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c /* Send pending IPI's to kick RPS processing on remote cpus. */ while (remsd) { -@@ -4844,6 +4860,7 @@ +@@ -4844,6 +4860,7 @@ static void net_rps_action_and_irq_enable(struct softnet_data *sd) } else #endif local_irq_enable(); @@ -26204,7 +26202,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c } static bool sd_has_rps_ipi_waiting(struct softnet_data *sd) -@@ -4873,7 +4890,9 @@ +@@ -4873,7 +4890,9 @@ static int process_backlog(struct napi_struct *napi, int quota) while (again) { struct sk_buff *skb; @@ -26214,7 +26212,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c rcu_read_lock(); __netif_receive_skb(skb); rcu_read_unlock(); -@@ -4881,9 +4900,9 @@ +@@ -4881,9 +4900,9 @@ static int process_backlog(struct napi_struct *napi, int quota) if (++work >= quota) return work; @@ -26225,7 +26223,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c rps_lock(sd); if (skb_queue_empty(&sd->input_pkt_queue)) { /* -@@ -4921,9 +4940,11 @@ +@@ -4921,9 +4940,11 @@ void __napi_schedule(struct napi_struct *n) local_irq_save(flags); ____napi_schedule(this_cpu_ptr(&softnet_data), n); local_irq_restore(flags); @@ -26237,7 +26235,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c /** * __napi_schedule_irqoff - schedule for receive * @n: entry to schedule -@@ -4935,6 +4956,7 @@ +@@ -4935,6 +4956,7 @@ void __napi_schedule_irqoff(struct napi_struct *n) ____napi_schedule(this_cpu_ptr(&softnet_data), n); } EXPORT_SYMBOL(__napi_schedule_irqoff); @@ -26245,7 +26243,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c void __napi_complete(struct napi_struct *n) { -@@ -5224,13 +5246,21 @@ +@@ -5224,13 +5246,21 @@ static __latent_entropy void net_rx_action(struct softirq_action *h) struct softnet_data *sd = this_cpu_ptr(&softnet_data); unsigned long time_limit = jiffies + 2; int budget = netdev_budget; @@ -26267,7 +26265,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c for (;;) { struct napi_struct *n; -@@ -5261,7 +5291,7 @@ +@@ -5261,7 +5291,7 @@ static __latent_entropy void net_rx_action(struct softirq_action *h) list_splice_tail(&repoll, &list); list_splice(&list, &sd->poll_list); if (!list_empty(&sd->poll_list)) @@ -26276,7 +26274,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c net_rps_action_and_irq_enable(sd); } -@@ -8022,16 +8052,20 @@ +@@ -8022,16 +8052,20 @@ static int dev_cpu_callback(struct notifier_block *nfb, raise_softirq_irqoff(NET_TX_SOFTIRQ); local_irq_enable(); @@ -26298,7 +26296,7 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c return NOTIFY_OK; } -@@ -8336,8 +8370,9 @@ +@@ -8336,8 +8370,9 @@ static int __init net_dev_init(void) INIT_WORK(flush, flush_backlog); @@ -26310,10 +26308,11 @@ diff -Nur linux-4.9.28.orig/net/core/dev.c linux-4.9.28/net/core/dev.c INIT_LIST_HEAD(&sd->poll_list); sd->output_queue_tailp = &sd->output_queue; #ifdef CONFIG_RPS -diff -Nur linux-4.9.28.orig/net/core/filter.c linux-4.9.28/net/core/filter.c ---- linux-4.9.28.orig/net/core/filter.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/core/filter.c 2017-05-19 03:37:25.214178037 +0200 -@@ -1645,7 +1645,7 @@ +diff --git a/net/core/filter.c b/net/core/filter.c +index b391209838ef..b86e9681a88e 100644 +--- a/net/core/filter.c ++++ b/net/core/filter.c +@@ -1645,7 +1645,7 @@ static inline int __bpf_tx_skb(struct net_device *dev, struct sk_buff *skb) { int ret; @@ -26322,7 +26321,7 @@ diff -Nur linux-4.9.28.orig/net/core/filter.c linux-4.9.28/net/core/filter.c net_crit_ratelimited("bpf: recursion limit reached on datapath, buggy bpf program?\n"); kfree_skb(skb); return -ENETDOWN; -@@ -1653,9 +1653,9 @@ +@@ -1653,9 +1653,9 @@ static inline int __bpf_tx_skb(struct net_device *dev, struct sk_buff *skb) skb->dev = dev; @@ -26334,10 +26333,11 @@ diff -Nur linux-4.9.28.orig/net/core/filter.c linux-4.9.28/net/core/filter.c return ret; } -diff -Nur linux-4.9.28.orig/net/core/gen_estimator.c linux-4.9.28/net/core/gen_estimator.c ---- linux-4.9.28.orig/net/core/gen_estimator.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/core/gen_estimator.c 2017-05-19 03:37:25.214178037 +0200 -@@ -84,7 +84,7 @@ +diff --git a/net/core/gen_estimator.c b/net/core/gen_estimator.c +index cad8e791f28e..2a9364fe62a5 100644 +--- a/net/core/gen_estimator.c ++++ b/net/core/gen_estimator.c +@@ -84,7 +84,7 @@ struct gen_estimator struct gnet_stats_basic_packed *bstats; struct gnet_stats_rate_est64 *rate_est; spinlock_t *stats_lock; @@ -26346,7 +26346,7 @@ diff -Nur linux-4.9.28.orig/net/core/gen_estimator.c linux-4.9.28/net/core/gen_e int ewma_log; u32 last_packets; unsigned long avpps; -@@ -213,7 +213,7 @@ +@@ -213,7 +213,7 @@ int gen_new_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_basic_cpu __percpu *cpu_bstats, struct gnet_stats_rate_est64 *rate_est, spinlock_t *stats_lock, @@ -26355,7 +26355,7 @@ diff -Nur linux-4.9.28.orig/net/core/gen_estimator.c linux-4.9.28/net/core/gen_e struct nlattr *opt) { struct gen_estimator *est; -@@ -309,7 +309,7 @@ +@@ -309,7 +309,7 @@ int gen_replace_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_basic_cpu __percpu *cpu_bstats, struct gnet_stats_rate_est64 *rate_est, spinlock_t *stats_lock, @@ -26364,10 +26364,11 @@ diff -Nur linux-4.9.28.orig/net/core/gen_estimator.c linux-4.9.28/net/core/gen_e { gen_kill_estimator(bstats, rate_est); return gen_new_estimator(bstats, cpu_bstats, rate_est, stats_lock, running, opt); -diff -Nur linux-4.9.28.orig/net/core/gen_stats.c linux-4.9.28/net/core/gen_stats.c ---- linux-4.9.28.orig/net/core/gen_stats.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/core/gen_stats.c 2017-05-19 03:37:25.214178037 +0200 -@@ -130,7 +130,7 @@ +diff --git a/net/core/gen_stats.c b/net/core/gen_stats.c +index 508e051304fb..bc3b17b78c94 100644 +--- a/net/core/gen_stats.c ++++ b/net/core/gen_stats.c +@@ -130,7 +130,7 @@ __gnet_stats_copy_basic_cpu(struct gnet_stats_basic_packed *bstats, } void @@ -26376,7 +26377,7 @@ diff -Nur linux-4.9.28.orig/net/core/gen_stats.c linux-4.9.28/net/core/gen_stats struct gnet_stats_basic_packed *bstats, struct gnet_stats_basic_cpu __percpu *cpu, struct gnet_stats_basic_packed *b) -@@ -143,10 +143,10 @@ +@@ -143,10 +143,10 @@ __gnet_stats_copy_basic(const seqcount_t *running, } do { if (running) @@ -26389,7 +26390,7 @@ diff -Nur linux-4.9.28.orig/net/core/gen_stats.c linux-4.9.28/net/core/gen_stats } EXPORT_SYMBOL(__gnet_stats_copy_basic); -@@ -164,7 +164,7 @@ +@@ -164,7 +164,7 @@ EXPORT_SYMBOL(__gnet_stats_copy_basic); * if the room in the socket buffer was not sufficient. */ int @@ -26398,9 +26399,10 @@ diff -Nur linux-4.9.28.orig/net/core/gen_stats.c linux-4.9.28/net/core/gen_stats struct gnet_dump *d, struct gnet_stats_basic_cpu __percpu *cpu, struct gnet_stats_basic_packed *b) -diff -Nur linux-4.9.28.orig/net/core/skbuff.c linux-4.9.28/net/core/skbuff.c ---- linux-4.9.28.orig/net/core/skbuff.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/core/skbuff.c 2017-05-19 03:37:25.218178191 +0200 +diff --git a/net/core/skbuff.c b/net/core/skbuff.c +index fe008f1bd930..9fa6bea3dd3f 100644 +--- a/net/core/skbuff.c ++++ b/net/core/skbuff.c @@ -64,6 +64,7 @@ #include #include @@ -26409,7 +26411,7 @@ diff -Nur linux-4.9.28.orig/net/core/skbuff.c linux-4.9.28/net/core/skbuff.c #include #include -@@ -360,6 +361,8 @@ +@@ -360,6 +361,8 @@ struct napi_alloc_cache { static DEFINE_PER_CPU(struct page_frag_cache, netdev_alloc_cache); static DEFINE_PER_CPU(struct napi_alloc_cache, napi_alloc_cache); @@ -26418,7 +26420,7 @@ diff -Nur linux-4.9.28.orig/net/core/skbuff.c linux-4.9.28/net/core/skbuff.c static void *__netdev_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) { -@@ -367,10 +370,10 @@ +@@ -367,10 +370,10 @@ static void *__netdev_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) unsigned long flags; void *data; @@ -26431,7 +26433,7 @@ diff -Nur linux-4.9.28.orig/net/core/skbuff.c linux-4.9.28/net/core/skbuff.c return data; } -@@ -389,9 +392,13 @@ +@@ -389,9 +392,13 @@ EXPORT_SYMBOL(netdev_alloc_frag); static void *__napi_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) { @@ -26447,7 +26449,7 @@ diff -Nur linux-4.9.28.orig/net/core/skbuff.c linux-4.9.28/net/core/skbuff.c } void *napi_alloc_frag(unsigned int fragsz) -@@ -438,13 +445,13 @@ +@@ -438,13 +445,13 @@ struct sk_buff *__netdev_alloc_skb(struct net_device *dev, unsigned int len, if (sk_memalloc_socks()) gfp_mask |= __GFP_MEMALLOC; @@ -26463,7 +26465,7 @@ diff -Nur linux-4.9.28.orig/net/core/skbuff.c linux-4.9.28/net/core/skbuff.c if (unlikely(!data)) return NULL; -@@ -485,9 +492,10 @@ +@@ -485,9 +492,10 @@ EXPORT_SYMBOL(__netdev_alloc_skb); struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len, gfp_t gfp_mask) { @@ -26475,7 +26477,7 @@ diff -Nur linux-4.9.28.orig/net/core/skbuff.c linux-4.9.28/net/core/skbuff.c len += NET_SKB_PAD + NET_IP_ALIGN; -@@ -505,7 +513,10 @@ +@@ -505,7 +513,10 @@ struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len, if (sk_memalloc_socks()) gfp_mask |= __GFP_MEMALLOC; @@ -26486,7 +26488,7 @@ diff -Nur linux-4.9.28.orig/net/core/skbuff.c linux-4.9.28/net/core/skbuff.c if (unlikely(!data)) return NULL; -@@ -516,7 +527,7 @@ +@@ -516,7 +527,7 @@ struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len, } /* use OR instead of assignment to avoid clearing of bits in mask */ @@ -26495,7 +26497,7 @@ diff -Nur linux-4.9.28.orig/net/core/skbuff.c linux-4.9.28/net/core/skbuff.c skb->pfmemalloc = 1; skb->head_frag = 1; -@@ -760,23 +771,26 @@ +@@ -760,23 +771,26 @@ EXPORT_SYMBOL(consume_skb); void __kfree_skb_flush(void) { @@ -26524,7 +26526,7 @@ diff -Nur linux-4.9.28.orig/net/core/skbuff.c linux-4.9.28/net/core/skbuff.c /* record skb to CPU local list */ nc->skb_cache[nc->skb_count++] = skb; -@@ -791,6 +805,7 @@ +@@ -791,6 +805,7 @@ static inline void _kfree_skb_defer(struct sk_buff *skb) nc->skb_cache); nc->skb_count = 0; } @@ -26532,10 +26534,11 @@ diff -Nur linux-4.9.28.orig/net/core/skbuff.c linux-4.9.28/net/core/skbuff.c } void __kfree_skb_defer(struct sk_buff *skb) { -diff -Nur linux-4.9.28.orig/net/core/sock.c linux-4.9.28/net/core/sock.c ---- linux-4.9.28.orig/net/core/sock.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/core/sock.c 2017-05-19 03:37:25.218178191 +0200 -@@ -2499,12 +2499,11 @@ +diff --git a/net/core/sock.c b/net/core/sock.c +index 470a2043b846..2b09a5a33d8d 100644 +--- a/net/core/sock.c ++++ b/net/core/sock.c +@@ -2499,12 +2499,11 @@ void lock_sock_nested(struct sock *sk, int subclass) if (sk->sk_lock.owned) __lock_sock(sk); sk->sk_lock.owned = 1; @@ -26549,9 +26552,10 @@ diff -Nur linux-4.9.28.orig/net/core/sock.c linux-4.9.28/net/core/sock.c } EXPORT_SYMBOL(lock_sock_nested); -diff -Nur linux-4.9.28.orig/net/ipv4/icmp.c linux-4.9.28/net/ipv4/icmp.c ---- linux-4.9.28.orig/net/ipv4/icmp.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/ipv4/icmp.c 2017-05-19 03:37:25.218178191 +0200 +diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c +index 48734ee6293f..e6864ff11352 100644 +--- a/net/ipv4/icmp.c ++++ b/net/ipv4/icmp.c @@ -69,6 +69,7 @@ #include #include @@ -26568,7 +26572,7 @@ diff -Nur linux-4.9.28.orig/net/ipv4/icmp.c linux-4.9.28/net/ipv4/icmp.c #include #include #include -@@ -204,6 +206,8 @@ +@@ -204,6 +206,8 @@ static const struct icmp_control icmp_pointers[NR_ICMP_TYPES+1]; * * On SMP we have one ICMP socket per-cpu. */ @@ -26577,7 +26581,7 @@ diff -Nur linux-4.9.28.orig/net/ipv4/icmp.c linux-4.9.28/net/ipv4/icmp.c static struct sock *icmp_sk(struct net *net) { return *this_cpu_ptr(net->ipv4.icmp_sk); -@@ -215,12 +219,14 @@ +@@ -215,12 +219,14 @@ static inline struct sock *icmp_xmit_lock(struct net *net) local_bh_disable(); @@ -26592,7 +26596,7 @@ diff -Nur linux-4.9.28.orig/net/ipv4/icmp.c linux-4.9.28/net/ipv4/icmp.c local_bh_enable(); return NULL; } -@@ -230,6 +236,7 @@ +@@ -230,6 +236,7 @@ static inline struct sock *icmp_xmit_lock(struct net *net) static inline void icmp_xmit_unlock(struct sock *sk) { spin_unlock_bh(&sk->sk_lock.slock); @@ -26600,7 +26604,7 @@ diff -Nur linux-4.9.28.orig/net/ipv4/icmp.c linux-4.9.28/net/ipv4/icmp.c } int sysctl_icmp_msgs_per_sec __read_mostly = 1000; -@@ -358,6 +365,7 @@ +@@ -358,6 +365,7 @@ static void icmp_push_reply(struct icmp_bxm *icmp_param, struct sock *sk; struct sk_buff *skb; @@ -26608,7 +26612,7 @@ diff -Nur linux-4.9.28.orig/net/ipv4/icmp.c linux-4.9.28/net/ipv4/icmp.c sk = icmp_sk(dev_net((*rt)->dst.dev)); if (ip_append_data(sk, fl4, icmp_glue_bits, icmp_param, icmp_param->data_len+icmp_param->head_len, -@@ -380,6 +388,7 @@ +@@ -380,6 +388,7 @@ static void icmp_push_reply(struct icmp_bxm *icmp_param, skb->ip_summed = CHECKSUM_NONE; ip_push_pending_frames(sk, fl4); } @@ -26616,7 +26620,7 @@ diff -Nur linux-4.9.28.orig/net/ipv4/icmp.c linux-4.9.28/net/ipv4/icmp.c } /* -@@ -891,6 +900,30 @@ +@@ -891,6 +900,30 @@ static bool icmp_redirect(struct sk_buff *skb) } /* @@ -26647,7 +26651,7 @@ diff -Nur linux-4.9.28.orig/net/ipv4/icmp.c linux-4.9.28/net/ipv4/icmp.c * Handle ICMP_ECHO ("ping") requests. * * RFC 1122: 3.2.2.6 MUST have an echo server that answers ICMP echo -@@ -917,6 +950,11 @@ +@@ -917,6 +950,11 @@ static bool icmp_echo(struct sk_buff *skb) icmp_param.data_len = skb->len; icmp_param.head_len = sizeof(struct icmphdr); icmp_reply(&icmp_param, skb); @@ -26659,10 +26663,11 @@ diff -Nur linux-4.9.28.orig/net/ipv4/icmp.c linux-4.9.28/net/ipv4/icmp.c } /* should there be an ICMP stat for ignored echos? */ return true; -diff -Nur linux-4.9.28.orig/net/ipv4/sysctl_net_ipv4.c linux-4.9.28/net/ipv4/sysctl_net_ipv4.c ---- linux-4.9.28.orig/net/ipv4/sysctl_net_ipv4.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/ipv4/sysctl_net_ipv4.c 2017-05-19 03:37:25.218178191 +0200 -@@ -681,6 +681,13 @@ +diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c +index 80bc36b25de2..215b90adfb05 100644 +--- a/net/ipv4/sysctl_net_ipv4.c ++++ b/net/ipv4/sysctl_net_ipv4.c +@@ -681,6 +681,13 @@ static struct ctl_table ipv4_net_table[] = { .proc_handler = proc_dointvec }, { @@ -26676,9 +26681,10 @@ diff -Nur linux-4.9.28.orig/net/ipv4/sysctl_net_ipv4.c linux-4.9.28/net/ipv4/sys .procname = "icmp_ignore_bogus_error_responses", .data = &init_net.ipv4.sysctl_icmp_ignore_bogus_error_responses, .maxlen = sizeof(int), -diff -Nur linux-4.9.28.orig/net/ipv4/tcp_ipv4.c linux-4.9.28/net/ipv4/tcp_ipv4.c ---- linux-4.9.28.orig/net/ipv4/tcp_ipv4.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/ipv4/tcp_ipv4.c 2017-05-19 03:37:25.218178191 +0200 +diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c +index 6988566dc72f..672fffcde28c 100644 +--- a/net/ipv4/tcp_ipv4.c ++++ b/net/ipv4/tcp_ipv4.c @@ -62,6 +62,7 @@ #include #include @@ -26687,7 +26693,7 @@ diff -Nur linux-4.9.28.orig/net/ipv4/tcp_ipv4.c linux-4.9.28/net/ipv4/tcp_ipv4.c #include #include -@@ -568,6 +569,7 @@ +@@ -568,6 +569,7 @@ void tcp_v4_send_check(struct sock *sk, struct sk_buff *skb) } EXPORT_SYMBOL(tcp_v4_send_check); @@ -26695,7 +26701,7 @@ diff -Nur linux-4.9.28.orig/net/ipv4/tcp_ipv4.c linux-4.9.28/net/ipv4/tcp_ipv4.c /* * This routine will send an RST to the other tcp. * -@@ -695,6 +697,8 @@ +@@ -695,6 +697,8 @@ static void tcp_v4_send_reset(const struct sock *sk, struct sk_buff *skb) offsetof(struct inet_timewait_sock, tw_bound_dev_if)); arg.tos = ip_hdr(skb)->tos; @@ -26704,7 +26710,7 @@ diff -Nur linux-4.9.28.orig/net/ipv4/tcp_ipv4.c linux-4.9.28/net/ipv4/tcp_ipv4.c local_bh_disable(); ip_send_unicast_reply(*this_cpu_ptr(net->ipv4.tcp_sk), skb, &TCP_SKB_CB(skb)->header.h4.opt, -@@ -704,6 +708,7 @@ +@@ -704,6 +708,7 @@ static void tcp_v4_send_reset(const struct sock *sk, struct sk_buff *skb) __TCP_INC_STATS(net, TCP_MIB_OUTSEGS); __TCP_INC_STATS(net, TCP_MIB_OUTRSTS); local_bh_enable(); @@ -26712,7 +26718,7 @@ diff -Nur linux-4.9.28.orig/net/ipv4/tcp_ipv4.c linux-4.9.28/net/ipv4/tcp_ipv4.c #ifdef CONFIG_TCP_MD5SIG out: -@@ -779,6 +784,7 @@ +@@ -779,6 +784,7 @@ static void tcp_v4_send_ack(struct net *net, if (oif) arg.bound_dev_if = oif; arg.tos = tos; @@ -26720,7 +26726,7 @@ diff -Nur linux-4.9.28.orig/net/ipv4/tcp_ipv4.c linux-4.9.28/net/ipv4/tcp_ipv4.c local_bh_disable(); ip_send_unicast_reply(*this_cpu_ptr(net->ipv4.tcp_sk), skb, &TCP_SKB_CB(skb)->header.h4.opt, -@@ -787,6 +793,7 @@ +@@ -787,6 +793,7 @@ static void tcp_v4_send_ack(struct net *net, __TCP_INC_STATS(net, TCP_MIB_OUTSEGS); local_bh_enable(); @@ -26728,10 +26734,11 @@ diff -Nur linux-4.9.28.orig/net/ipv4/tcp_ipv4.c linux-4.9.28/net/ipv4/tcp_ipv4.c } static void tcp_v4_timewait_ack(struct sock *sk, struct sk_buff *skb) -diff -Nur linux-4.9.28.orig/net/mac80211/rx.c linux-4.9.28/net/mac80211/rx.c ---- linux-4.9.28.orig/net/mac80211/rx.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/mac80211/rx.c 2017-05-19 03:37:25.218178191 +0200 -@@ -4230,7 +4230,7 @@ +diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c +index acaaf616da71..09020dbcc089 100644 +--- a/net/mac80211/rx.c ++++ b/net/mac80211/rx.c +@@ -4230,7 +4230,7 @@ void ieee80211_rx_napi(struct ieee80211_hw *hw, struct ieee80211_sta *pubsta, struct ieee80211_supported_band *sband; struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(skb); @@ -26740,9 +26747,10 @@ diff -Nur linux-4.9.28.orig/net/mac80211/rx.c linux-4.9.28/net/mac80211/rx.c if (WARN_ON(status->band >= NUM_NL80211_BANDS)) goto drop; -diff -Nur linux-4.9.28.orig/net/netfilter/core.c linux-4.9.28/net/netfilter/core.c ---- linux-4.9.28.orig/net/netfilter/core.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/netfilter/core.c 2017-05-19 03:37:25.218178191 +0200 +diff --git a/net/netfilter/core.c b/net/netfilter/core.c +index 004af030ef1a..b64f751bda45 100644 +--- a/net/netfilter/core.c ++++ b/net/netfilter/core.c @@ -22,12 +22,18 @@ #include #include @@ -26762,9 +26770,10 @@ diff -Nur linux-4.9.28.orig/net/netfilter/core.c linux-4.9.28/net/netfilter/core static DEFINE_MUTEX(afinfo_mutex); const struct nf_afinfo __rcu *nf_afinfo[NFPROTO_NUMPROTO] __read_mostly; -diff -Nur linux-4.9.28.orig/net/packet/af_packet.c linux-4.9.28/net/packet/af_packet.c ---- linux-4.9.28.orig/net/packet/af_packet.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/packet/af_packet.c 2017-05-19 03:37:25.218178191 +0200 +diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c +index cb76ff3088e9..3f42c5b1af55 100644 +--- a/net/packet/af_packet.c ++++ b/net/packet/af_packet.c @@ -63,6 +63,7 @@ #include #include @@ -26773,7 +26782,7 @@ diff -Nur linux-4.9.28.orig/net/packet/af_packet.c linux-4.9.28/net/packet/af_pa #include #include #include -@@ -694,7 +695,7 @@ +@@ -694,7 +695,7 @@ static void prb_retire_rx_blk_timer_expired(unsigned long data) if (BLOCK_NUM_PKTS(pbd)) { while (atomic_read(&pkc->blk_fill_in_prog)) { /* Waiting for skb_copy_bits to finish... */ @@ -26782,7 +26791,7 @@ diff -Nur linux-4.9.28.orig/net/packet/af_packet.c linux-4.9.28/net/packet/af_pa } } -@@ -956,7 +957,7 @@ +@@ -956,7 +957,7 @@ static void prb_retire_current_block(struct tpacket_kbdq_core *pkc, if (!(status & TP_STATUS_BLK_TMO)) { while (atomic_read(&pkc->blk_fill_in_prog)) { /* Waiting for skb_copy_bits to finish... */ @@ -26791,9 +26800,10 @@ diff -Nur linux-4.9.28.orig/net/packet/af_packet.c linux-4.9.28/net/packet/af_pa } } prb_close_block(pkc, pbd, po, status); -diff -Nur linux-4.9.28.orig/net/rds/ib_rdma.c linux-4.9.28/net/rds/ib_rdma.c ---- linux-4.9.28.orig/net/rds/ib_rdma.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/rds/ib_rdma.c 2017-05-19 03:37:25.218178191 +0200 +diff --git a/net/rds/ib_rdma.c b/net/rds/ib_rdma.c +index 977f69886c00..f3e7a36b0396 100644 +--- a/net/rds/ib_rdma.c ++++ b/net/rds/ib_rdma.c @@ -34,6 +34,7 @@ #include #include @@ -26802,7 +26812,7 @@ diff -Nur linux-4.9.28.orig/net/rds/ib_rdma.c linux-4.9.28/net/rds/ib_rdma.c #include "rds_single_path.h" #include "ib_mr.h" -@@ -210,7 +211,7 @@ +@@ -210,7 +211,7 @@ static inline void wait_clean_list_grace(void) for_each_online_cpu(cpu) { flag = &per_cpu(clean_list_grace, cpu); while (test_bit(CLEAN_LIST_BUSY_BIT, flag)) @@ -26811,9 +26821,10 @@ diff -Nur linux-4.9.28.orig/net/rds/ib_rdma.c linux-4.9.28/net/rds/ib_rdma.c } } -diff -Nur linux-4.9.28.orig/net/rxrpc/security.c linux-4.9.28/net/rxrpc/security.c ---- linux-4.9.28.orig/net/rxrpc/security.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/rxrpc/security.c 2017-05-19 03:37:25.218178191 +0200 +diff --git a/net/rxrpc/security.c b/net/rxrpc/security.c +index 7d921e56e715..13df56a738e5 100644 +--- a/net/rxrpc/security.c ++++ b/net/rxrpc/security.c @@ -19,9 +19,6 @@ #include #include "ar-internal.h" @@ -26824,10 +26835,11 @@ diff -Nur linux-4.9.28.orig/net/rxrpc/security.c linux-4.9.28/net/rxrpc/security static const struct rxrpc_security *rxrpc_security_types[] = { [RXRPC_SECURITY_NONE] = &rxrpc_no_security, #ifdef CONFIG_RXKAD -diff -Nur linux-4.9.28.orig/net/sched/sch_api.c linux-4.9.28/net/sched/sch_api.c ---- linux-4.9.28.orig/net/sched/sch_api.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/sched/sch_api.c 2017-05-19 03:37:25.218178191 +0200 -@@ -981,7 +981,7 @@ +diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c +index 206dc24add3a..00ea9bde5bb3 100644 +--- a/net/sched/sch_api.c ++++ b/net/sched/sch_api.c +@@ -981,7 +981,7 @@ static struct Qdisc *qdisc_create(struct net_device *dev, rcu_assign_pointer(sch->stab, stab); } if (tca[TCA_RATE]) { @@ -26836,10 +26848,11 @@ diff -Nur linux-4.9.28.orig/net/sched/sch_api.c linux-4.9.28/net/sched/sch_api.c err = -EOPNOTSUPP; if (sch->flags & TCQ_F_MQROOT) -diff -Nur linux-4.9.28.orig/net/sched/sch_generic.c linux-4.9.28/net/sched/sch_generic.c ---- linux-4.9.28.orig/net/sched/sch_generic.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/sched/sch_generic.c 2017-05-19 03:37:25.218178191 +0200 -@@ -425,7 +425,11 @@ +diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c +index 6cfb6e9038c2..20727e1347de 100644 +--- a/net/sched/sch_generic.c ++++ b/net/sched/sch_generic.c +@@ -425,7 +425,11 @@ struct Qdisc noop_qdisc = { .ops = &noop_qdisc_ops, .q.lock = __SPIN_LOCK_UNLOCKED(noop_qdisc.q.lock), .dev_queue = &noop_netdev_queue, @@ -26851,7 +26864,7 @@ diff -Nur linux-4.9.28.orig/net/sched/sch_generic.c linux-4.9.28/net/sched/sch_g .busylock = __SPIN_LOCK_UNLOCKED(noop_qdisc.busylock), }; EXPORT_SYMBOL(noop_qdisc); -@@ -624,9 +628,17 @@ +@@ -624,9 +628,17 @@ struct Qdisc *qdisc_alloc(struct netdev_queue *dev_queue, lockdep_set_class(&sch->busylock, dev->qdisc_tx_busylock ?: &qdisc_tx_busylock); @@ -26869,7 +26882,7 @@ diff -Nur linux-4.9.28.orig/net/sched/sch_generic.c linux-4.9.28/net/sched/sch_g sch->ops = ops; sch->enqueue = ops->enqueue; -@@ -925,7 +937,7 @@ +@@ -925,7 +937,7 @@ void dev_deactivate_many(struct list_head *head) /* Wait for outstanding qdisc_run calls. */ list_for_each_entry(dev, head, close_list) while (some_qdisc_is_busy(dev)) @@ -26878,10 +26891,11 @@ diff -Nur linux-4.9.28.orig/net/sched/sch_generic.c linux-4.9.28/net/sched/sch_g } void dev_deactivate(struct net_device *dev) -diff -Nur linux-4.9.28.orig/net/sunrpc/svc_xprt.c linux-4.9.28/net/sunrpc/svc_xprt.c ---- linux-4.9.28.orig/net/sunrpc/svc_xprt.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/net/sunrpc/svc_xprt.c 2017-05-19 03:37:25.218178191 +0200 -@@ -396,7 +396,7 @@ +diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c +index 9c9db55a0c1e..e6583b018a72 100644 +--- a/net/sunrpc/svc_xprt.c ++++ b/net/sunrpc/svc_xprt.c +@@ -396,7 +396,7 @@ void svc_xprt_do_enqueue(struct svc_xprt *xprt) goto out; } @@ -26890,7 +26904,7 @@ diff -Nur linux-4.9.28.orig/net/sunrpc/svc_xprt.c linux-4.9.28/net/sunrpc/svc_xp pool = svc_pool_for_cpu(xprt->xpt_server, cpu); atomic_long_inc(&pool->sp_stats.packets); -@@ -432,7 +432,7 @@ +@@ -432,7 +432,7 @@ void svc_xprt_do_enqueue(struct svc_xprt *xprt) atomic_long_inc(&pool->sp_stats.threads_woken); wake_up_process(rqstp->rq_task); @@ -26899,7 +26913,7 @@ diff -Nur linux-4.9.28.orig/net/sunrpc/svc_xprt.c linux-4.9.28/net/sunrpc/svc_xp goto out; } rcu_read_unlock(); -@@ -453,7 +453,7 @@ +@@ -453,7 +453,7 @@ void svc_xprt_do_enqueue(struct svc_xprt *xprt) goto redo_search; } rqstp = NULL; @@ -26908,10 +26922,11 @@ diff -Nur linux-4.9.28.orig/net/sunrpc/svc_xprt.c linux-4.9.28/net/sunrpc/svc_xp out: trace_svc_xprt_do_enqueue(xprt, rqstp); } -diff -Nur linux-4.9.28.orig/scripts/mkcompile_h linux-4.9.28/scripts/mkcompile_h ---- linux-4.9.28.orig/scripts/mkcompile_h 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/scripts/mkcompile_h 2017-05-19 03:37:25.218178191 +0200 -@@ -4,7 +4,8 @@ +diff --git a/scripts/mkcompile_h b/scripts/mkcompile_h +index 6fdc97ef6023..523e0420d7f0 100755 +--- a/scripts/mkcompile_h ++++ b/scripts/mkcompile_h +@@ -4,7 +4,8 @@ TARGET=$1 ARCH=$2 SMP=$3 PREEMPT=$4 @@ -26921,7 +26936,7 @@ diff -Nur linux-4.9.28.orig/scripts/mkcompile_h linux-4.9.28/scripts/mkcompile_h vecho() { [ "${quiet}" = "silent_" ] || echo "$@" ; } -@@ -57,6 +58,7 @@ +@@ -57,6 +58,7 @@ UTS_VERSION="#$VERSION" CONFIG_FLAGS="" if [ -n "$SMP" ] ; then CONFIG_FLAGS="SMP"; fi if [ -n "$PREEMPT" ] ; then CONFIG_FLAGS="$CONFIG_FLAGS PREEMPT"; fi @@ -26929,10 +26944,11 @@ diff -Nur linux-4.9.28.orig/scripts/mkcompile_h linux-4.9.28/scripts/mkcompile_h UTS_VERSION="$UTS_VERSION $CONFIG_FLAGS $TIMESTAMP" # Truncate to maximum length -diff -Nur linux-4.9.28.orig/sound/core/pcm_native.c linux-4.9.28/sound/core/pcm_native.c ---- linux-4.9.28.orig/sound/core/pcm_native.c 2017-05-14 14:00:37.000000000 +0200 -+++ linux-4.9.28/sound/core/pcm_native.c 2017-05-19 03:37:25.222178345 +0200 -@@ -135,7 +135,7 @@ +diff --git a/sound/core/pcm_native.c b/sound/core/pcm_native.c +index 9d33c1e85c79..3d307bda86f9 100644 +--- a/sound/core/pcm_native.c ++++ b/sound/core/pcm_native.c +@@ -135,7 +135,7 @@ EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock); void snd_pcm_stream_lock_irq(struct snd_pcm_substream *substream) { if (!substream->pcm->nonatomic) @@ -26941,7 +26957,7 @@ diff -Nur linux-4.9.28.orig/sound/core/pcm_native.c linux-4.9.28/sound/core/pcm_ snd_pcm_stream_lock(substream); } EXPORT_SYMBOL_GPL(snd_pcm_stream_lock_irq); -@@ -150,7 +150,7 @@ +@@ -150,7 +150,7 @@ void snd_pcm_stream_unlock_irq(struct snd_pcm_substream *substream) { snd_pcm_stream_unlock(substream); if (!substream->pcm->nonatomic) @@ -26950,7 +26966,7 @@ diff -Nur linux-4.9.28.orig/sound/core/pcm_native.c linux-4.9.28/sound/core/pcm_ } EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock_irq); -@@ -158,7 +158,7 @@ +@@ -158,7 +158,7 @@ unsigned long _snd_pcm_stream_lock_irqsave(struct snd_pcm_substream *substream) { unsigned long flags = 0; if (!substream->pcm->nonatomic) @@ -26959,7 +26975,7 @@ diff -Nur linux-4.9.28.orig/sound/core/pcm_native.c linux-4.9.28/sound/core/pcm_ snd_pcm_stream_lock(substream); return flags; } -@@ -176,7 +176,7 @@ +@@ -176,7 +176,7 @@ void snd_pcm_stream_unlock_irqrestore(struct snd_pcm_substream *substream, { snd_pcm_stream_unlock(substream); if (!substream->pcm->nonatomic) diff --git a/target/linux/patches/72134397d72079a533c8fc742701fdc7f5ae7c5b/patch-realtime b/target/linux/patches/72134397d72079a533c8fc742701fdc7f5ae7c5b/patch-realtime deleted file mode 100644 index fe7c393a6..000000000 --- a/target/linux/patches/72134397d72079a533c8fc742701fdc7f5ae7c5b/patch-realtime +++ /dev/null @@ -1,26047 +0,0 @@ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/Documentation/sysrq.txt linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/Documentation/sysrq.txt ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/Documentation/sysrq.txt 2017-04-16 10:37:28.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/Documentation/sysrq.txt 2017-04-18 17:54:21.000000000 +0200 -@@ -59,10 +59,17 @@ - On other - If you know of the key combos for other architectures, please - let me know so I can add them to this section. - --On all - write a character to /proc/sysrq-trigger. e.g.: -- -+On all - write a character to /proc/sysrq-trigger, e.g.: - echo t > /proc/sysrq-trigger - -+On all - Enable network SysRq by writing a cookie to icmp_echo_sysrq, e.g. -+ echo 0x01020304 >/proc/sys/net/ipv4/icmp_echo_sysrq -+ Send an ICMP echo request with this pattern plus the particular -+ SysRq command key. Example: -+ # ping -c1 -s57 -p0102030468 -+ will trigger the SysRq-H (help) command. -+ -+ - * What are the 'command' keys? - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - 'b' - Will immediately reboot the system without syncing or unmounting -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/Documentation/trace/histograms.txt linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/Documentation/trace/histograms.txt ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/Documentation/trace/histograms.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/Documentation/trace/histograms.txt 2017-04-18 17:54:22.000000000 +0200 -@@ -0,0 +1,186 @@ -+ Using the Linux Kernel Latency Histograms -+ -+ -+This document gives a short explanation how to enable, configure and use -+latency histograms. Latency histograms are primarily relevant in the -+context of real-time enabled kernels (CONFIG_PREEMPT/CONFIG_PREEMPT_RT) -+and are used in the quality management of the Linux real-time -+capabilities. -+ -+ -+* Purpose of latency histograms -+ -+A latency histogram continuously accumulates the frequencies of latency -+data. There are two types of histograms -+- potential sources of latencies -+- effective latencies -+ -+ -+* Potential sources of latencies -+ -+Potential sources of latencies are code segments where interrupts, -+preemption or both are disabled (aka critical sections). To create -+histograms of potential sources of latency, the kernel stores the time -+stamp at the start of a critical section, determines the time elapsed -+when the end of the section is reached, and increments the frequency -+counter of that latency value - irrespective of whether any concurrently -+running process is affected by latency or not. -+- Configuration items (in the Kernel hacking/Tracers submenu) -+ CONFIG_INTERRUPT_OFF_LATENCY -+ CONFIG_PREEMPT_OFF_LATENCY -+ -+ -+* Effective latencies -+ -+Effective latencies are actually occuring during wakeup of a process. To -+determine effective latencies, the kernel stores the time stamp when a -+process is scheduled to be woken up, and determines the duration of the -+wakeup time shortly before control is passed over to this process. Note -+that the apparent latency in user space may be somewhat longer, since the -+process may be interrupted after control is passed over to it but before -+the execution in user space takes place. Simply measuring the interval -+between enqueuing and wakeup may also not appropriate in cases when a -+process is scheduled as a result of a timer expiration. The timer may have -+missed its deadline, e.g. due to disabled interrupts, but this latency -+would not be registered. Therefore, the offsets of missed timers are -+recorded in a separate histogram. If both wakeup latency and missed timer -+offsets are configured and enabled, a third histogram may be enabled that -+records the overall latency as a sum of the timer latency, if any, and the -+wakeup latency. This histogram is called "timerandwakeup". -+- Configuration items (in the Kernel hacking/Tracers submenu) -+ CONFIG_WAKEUP_LATENCY -+ CONFIG_MISSED_TIMER_OFSETS -+ -+ -+* Usage -+ -+The interface to the administration of the latency histograms is located -+in the debugfs file system. To mount it, either enter -+ -+mount -t sysfs nodev /sys -+mount -t debugfs nodev /sys/kernel/debug -+ -+from shell command line level, or add -+ -+nodev /sys sysfs defaults 0 0 -+nodev /sys/kernel/debug debugfs defaults 0 0 -+ -+to the file /etc/fstab. All latency histogram related files are then -+available in the directory /sys/kernel/debug/tracing/latency_hist. A -+particular histogram type is enabled by writing non-zero to the related -+variable in the /sys/kernel/debug/tracing/latency_hist/enable directory. -+Select "preemptirqsoff" for the histograms of potential sources of -+latencies and "wakeup" for histograms of effective latencies etc. The -+histogram data - one per CPU - are available in the files -+ -+/sys/kernel/debug/tracing/latency_hist/preemptoff/CPUx -+/sys/kernel/debug/tracing/latency_hist/irqsoff/CPUx -+/sys/kernel/debug/tracing/latency_hist/preemptirqsoff/CPUx -+/sys/kernel/debug/tracing/latency_hist/wakeup/CPUx -+/sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio/CPUx -+/sys/kernel/debug/tracing/latency_hist/missed_timer_offsets/CPUx -+/sys/kernel/debug/tracing/latency_hist/timerandwakeup/CPUx -+ -+The histograms are reset by writing non-zero to the file "reset" in a -+particular latency directory. To reset all latency data, use -+ -+#!/bin/sh -+ -+TRACINGDIR=/sys/kernel/debug/tracing -+HISTDIR=$TRACINGDIR/latency_hist -+ -+if test -d $HISTDIR -+then -+ cd $HISTDIR -+ for i in `find . | grep /reset$` -+ do -+ echo 1 >$i -+ done -+fi -+ -+ -+* Data format -+ -+Latency data are stored with a resolution of one microsecond. The -+maximum latency is 10,240 microseconds. The data are only valid, if the -+overflow register is empty. Every output line contains the latency in -+microseconds in the first row and the number of samples in the second -+row. To display only lines with a positive latency count, use, for -+example, -+ -+grep -v " 0$" /sys/kernel/debug/tracing/latency_hist/preemptoff/CPU0 -+ -+#Minimum latency: 0 microseconds. -+#Average latency: 0 microseconds. -+#Maximum latency: 25 microseconds. -+#Total samples: 3104770694 -+#There are 0 samples greater or equal than 10240 microseconds -+#usecs samples -+ 0 2984486876 -+ 1 49843506 -+ 2 58219047 -+ 3 5348126 -+ 4 2187960 -+ 5 3388262 -+ 6 959289 -+ 7 208294 -+ 8 40420 -+ 9 4485 -+ 10 14918 -+ 11 18340 -+ 12 25052 -+ 13 19455 -+ 14 5602 -+ 15 969 -+ 16 47 -+ 17 18 -+ 18 14 -+ 19 1 -+ 20 3 -+ 21 2 -+ 22 5 -+ 23 2 -+ 25 1 -+ -+ -+* Wakeup latency of a selected process -+ -+To only collect wakeup latency data of a particular process, write the -+PID of the requested process to -+ -+/sys/kernel/debug/tracing/latency_hist/wakeup/pid -+ -+PIDs are not considered, if this variable is set to 0. -+ -+ -+* Details of the process with the highest wakeup latency so far -+ -+Selected data of the process that suffered from the highest wakeup -+latency that occurred in a particular CPU are available in the file -+ -+/sys/kernel/debug/tracing/latency_hist/wakeup/max_latency-CPUx. -+ -+In addition, other relevant system data at the time when the -+latency occurred are given. -+ -+The format of the data is (all in one line): -+ () \ -+<- -+ -+The value of is only relevant in the combined timer -+and wakeup latency recording. In the wakeup recording, it is -+always 0, in the missed_timer_offsets recording, it is the same -+as . -+ -+When retrospectively searching for the origin of a latency and -+tracing was not enabled, it may be helpful to know the name and -+some basic data of the task that (finally) was switching to the -+late real-tlme task. In addition to the victim's data, also the -+data of the possible culprit are therefore displayed after the -+"<-" symbol. -+ -+Finally, the timestamp of the time when the latency occurred -+in . after the most recent system boot -+is provided. -+ -+These data are also reset when the wakeup histogram is reset. -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/Kconfig 2017-04-16 10:37:28.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/Kconfig 2017-04-18 17:54:20.000000000 +0200 -@@ -9,6 +9,7 @@ - tristate "OProfile system profiling" - depends on PROFILING - depends on HAVE_OPROFILE -+ depends on !PREEMPT_RT_FULL - select RING_BUFFER - select RING_BUFFER_ALLOW_SWAP - help -@@ -52,6 +53,7 @@ - config JUMP_LABEL - bool "Optimize very unlikely/likely branches" - depends on HAVE_ARCH_JUMP_LABEL -+ depends on (!INTERRUPT_OFF_HIST && !PREEMPT_OFF_HIST && !WAKEUP_LATENCY_HIST && !MISSED_TIMER_OFFSETS_HIST) - help - This option enables a transparent branch optimization that - makes certain almost-always-true or almost-always-false branch -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/Kconfig 2017-04-16 10:37:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/Kconfig 2017-04-18 17:54:19.000000000 +0200 -@@ -36,7 +36,7 @@ - select HAVE_ARCH_AUDITSYSCALL if (AEABI && !OABI_COMPAT) - select HAVE_ARCH_BITREVERSE if (CPU_32v7M || CPU_32v7) && !CPU_32v6 - select HAVE_ARCH_HARDENED_USERCOPY -- select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL && !CPU_ENDIAN_BE32 && MMU -+ select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL && !CPU_ENDIAN_BE32 && MMU && !PREEMPT_RT_BASE - select HAVE_ARCH_KGDB if !CPU_ENDIAN_BE32 && MMU - select HAVE_ARCH_MMAP_RND_BITS if MMU - select HAVE_ARCH_SECCOMP_FILTER if (AEABI && !OABI_COMPAT) -@@ -75,6 +75,7 @@ - select HAVE_PERF_EVENTS - select HAVE_PERF_REGS - select HAVE_PERF_USER_STACK_DUMP -+ select HAVE_PREEMPT_LAZY - select HAVE_RCU_TABLE_FREE if (SMP && ARM_LPAE) - select HAVE_REGS_AND_STACK_ACCESS_API - select HAVE_SYSCALL_TRACEPOINTS -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/include/asm/irq.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/include/asm/irq.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/include/asm/irq.h 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/include/asm/irq.h 2017-04-18 17:54:19.000000000 +0200 -@@ -22,6 +22,8 @@ - #endif - - #ifndef __ASSEMBLY__ -+#include -+ - struct irqaction; - struct pt_regs; - extern void migrate_irqs(void); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/include/asm/switch_to.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/include/asm/switch_to.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/include/asm/switch_to.h 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/include/asm/switch_to.h 2017-04-18 17:54:19.000000000 +0200 -@@ -3,6 +3,13 @@ - - #include - -+#if defined CONFIG_PREEMPT_RT_FULL && defined CONFIG_HIGHMEM -+void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p); -+#else -+static inline void -+switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) { } -+#endif -+ - /* - * For v7 SMP cores running a preemptible kernel we may be pre-empted - * during a TLB maintenance operation, so execute an inner-shareable dsb -@@ -25,6 +32,7 @@ - #define switch_to(prev,next,last) \ - do { \ - __complete_pending_tlbi(); \ -+ switch_kmaps(prev, next); \ - last = __switch_to(prev,task_thread_info(prev), task_thread_info(next)); \ - } while (0) - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/include/asm/thread_info.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/include/asm/thread_info.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/include/asm/thread_info.h 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/include/asm/thread_info.h 2017-04-18 17:54:19.000000000 +0200 -@@ -49,6 +49,7 @@ - struct thread_info { - unsigned long flags; /* low level flags */ - int preempt_count; /* 0 => preemptable, <0 => bug */ -+ int preempt_lazy_count; /* 0 => preemptable, <0 => bug */ - mm_segment_t addr_limit; /* address limit */ - struct task_struct *task; /* main task structure */ - __u32 cpu; /* cpu */ -@@ -142,7 +143,8 @@ - #define TIF_SYSCALL_TRACE 4 /* syscall trace active */ - #define TIF_SYSCALL_AUDIT 5 /* syscall auditing active */ - #define TIF_SYSCALL_TRACEPOINT 6 /* syscall tracepoint instrumentation */ --#define TIF_SECCOMP 7 /* seccomp syscall filtering active */ -+#define TIF_SECCOMP 8 /* seccomp syscall filtering active */ -+#define TIF_NEED_RESCHED_LAZY 7 - - #define TIF_NOHZ 12 /* in adaptive nohz mode */ - #define TIF_USING_IWMMXT 17 -@@ -152,6 +154,7 @@ - #define _TIF_SIGPENDING (1 << TIF_SIGPENDING) - #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) - #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) -+#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) - #define _TIF_UPROBE (1 << TIF_UPROBE) - #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) - #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) -@@ -167,7 +170,8 @@ - * Change these and you break ASM code in entry-common.S - */ - #define _TIF_WORK_MASK (_TIF_NEED_RESCHED | _TIF_SIGPENDING | \ -- _TIF_NOTIFY_RESUME | _TIF_UPROBE) -+ _TIF_NOTIFY_RESUME | _TIF_UPROBE | \ -+ _TIF_NEED_RESCHED_LAZY) - - #endif /* __KERNEL__ */ - #endif /* __ASM_ARM_THREAD_INFO_H */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/asm-offsets.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/asm-offsets.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/asm-offsets.c 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/asm-offsets.c 2017-04-18 17:54:19.000000000 +0200 -@@ -65,6 +65,7 @@ - BLANK(); - DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); - DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); -+ DEFINE(TI_PREEMPT_LAZY, offsetof(struct thread_info, preempt_lazy_count)); - DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit)); - DEFINE(TI_TASK, offsetof(struct thread_info, task)); - DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/entry-armv.S linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/entry-armv.S ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/entry-armv.S 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/entry-armv.S 2017-04-18 17:54:19.000000000 +0200 -@@ -220,11 +220,18 @@ - - #ifdef CONFIG_PREEMPT - ldr r8, [tsk, #TI_PREEMPT] @ get preempt count -- ldr r0, [tsk, #TI_FLAGS] @ get flags - teq r8, #0 @ if preempt count != 0 -+ bne 1f @ return from exeption -+ ldr r0, [tsk, #TI_FLAGS] @ get flags -+ tst r0, #_TIF_NEED_RESCHED @ if NEED_RESCHED is set -+ blne svc_preempt @ preempt! -+ -+ ldr r8, [tsk, #TI_PREEMPT_LAZY] @ get preempt lazy count -+ teq r8, #0 @ if preempt lazy count != 0 - movne r0, #0 @ force flags to 0 -- tst r0, #_TIF_NEED_RESCHED -+ tst r0, #_TIF_NEED_RESCHED_LAZY - blne svc_preempt -+1: - #endif - - svc_exit r5, irq = 1 @ return from exception -@@ -239,8 +246,14 @@ - 1: bl preempt_schedule_irq @ irq en/disable is done inside - ldr r0, [tsk, #TI_FLAGS] @ get new tasks TI_FLAGS - tst r0, #_TIF_NEED_RESCHED -+ bne 1b -+ tst r0, #_TIF_NEED_RESCHED_LAZY - reteq r8 @ go again -- b 1b -+ ldr r0, [tsk, #TI_PREEMPT_LAZY] @ get preempt lazy count -+ teq r0, #0 @ if preempt lazy count != 0 -+ beq 1b -+ ret r8 @ go again -+ - #endif - - __und_fault: -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/entry-common.S linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/entry-common.S ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/entry-common.S 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/entry-common.S 2017-04-18 17:54:19.000000000 +0200 -@@ -36,7 +36,9 @@ - UNWIND(.cantunwind ) - disable_irq_notrace @ disable interrupts - ldr r1, [tsk, #TI_FLAGS] @ re-check for syscall tracing -- tst r1, #_TIF_SYSCALL_WORK | _TIF_WORK_MASK -+ tst r1, #((_TIF_SYSCALL_WORK | _TIF_WORK_MASK) & ~_TIF_SECCOMP) -+ bne fast_work_pending -+ tst r1, #_TIF_SECCOMP - bne fast_work_pending - - /* perform architecture specific actions before user return */ -@@ -62,8 +64,11 @@ - str r0, [sp, #S_R0 + S_OFF]! @ save returned r0 - disable_irq_notrace @ disable interrupts - ldr r1, [tsk, #TI_FLAGS] @ re-check for syscall tracing -- tst r1, #_TIF_SYSCALL_WORK | _TIF_WORK_MASK -+ tst r1, #((_TIF_SYSCALL_WORK | _TIF_WORK_MASK) & ~_TIF_SECCOMP) -+ bne do_slower_path -+ tst r1, #_TIF_SECCOMP - beq no_work_pending -+do_slower_path: - UNWIND(.fnend ) - ENDPROC(ret_fast_syscall) - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/patch.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/patch.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/patch.c 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/patch.c 2017-04-18 17:54:19.000000000 +0200 -@@ -15,7 +15,7 @@ - unsigned int insn; - }; - --static DEFINE_SPINLOCK(patch_lock); -+static DEFINE_RAW_SPINLOCK(patch_lock); - - static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags) - __acquires(&patch_lock) -@@ -32,7 +32,7 @@ - return addr; - - if (flags) -- spin_lock_irqsave(&patch_lock, *flags); -+ raw_spin_lock_irqsave(&patch_lock, *flags); - else - __acquire(&patch_lock); - -@@ -47,7 +47,7 @@ - clear_fixmap(fixmap); - - if (flags) -- spin_unlock_irqrestore(&patch_lock, *flags); -+ raw_spin_unlock_irqrestore(&patch_lock, *flags); - else - __release(&patch_lock); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/process.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/process.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/process.c 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/process.c 2017-04-18 17:54:19.000000000 +0200 -@@ -322,6 +322,30 @@ - } - - #ifdef CONFIG_MMU -+/* -+ * CONFIG_SPLIT_PTLOCK_CPUS results in a page->ptl lock. If the lock is not -+ * initialized by pgtable_page_ctor() then a coredump of the vector page will -+ * fail. -+ */ -+static int __init vectors_user_mapping_init_page(void) -+{ -+ struct page *page; -+ unsigned long addr = 0xffff0000; -+ pgd_t *pgd; -+ pud_t *pud; -+ pmd_t *pmd; -+ -+ pgd = pgd_offset_k(addr); -+ pud = pud_offset(pgd, addr); -+ pmd = pmd_offset(pud, addr); -+ page = pmd_page(*(pmd)); -+ -+ pgtable_page_ctor(page); -+ -+ return 0; -+} -+late_initcall(vectors_user_mapping_init_page); -+ - #ifdef CONFIG_KUSER_HELPERS - /* - * The vectors page is always readable from user space for the -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/signal.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/signal.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/signal.c 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/signal.c 2017-04-18 17:54:19.000000000 +0200 -@@ -572,7 +572,8 @@ - */ - trace_hardirqs_off(); - do { -- if (likely(thread_flags & _TIF_NEED_RESCHED)) { -+ if (likely(thread_flags & (_TIF_NEED_RESCHED | -+ _TIF_NEED_RESCHED_LAZY))) { - schedule(); - } else { - if (unlikely(!user_mode(regs))) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/smp.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/smp.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/smp.c 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/smp.c 2017-04-18 17:54:19.000000000 +0200 -@@ -234,8 +234,6 @@ - flush_cache_louis(); - local_flush_tlb_all(); - -- clear_tasks_mm_cpumask(cpu); -- - return 0; - } - -@@ -251,6 +249,9 @@ - pr_err("CPU%u: cpu didn't die\n", cpu); - return; - } -+ -+ clear_tasks_mm_cpumask(cpu); -+ - pr_notice("CPU%u: shutdown\n", cpu); - - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/unwind.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/unwind.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kernel/unwind.c 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kernel/unwind.c 2017-04-18 17:54:19.000000000 +0200 -@@ -93,7 +93,7 @@ - static const struct unwind_idx *__origin_unwind_idx; - extern const struct unwind_idx __stop_unwind_idx[]; - --static DEFINE_SPINLOCK(unwind_lock); -+static DEFINE_RAW_SPINLOCK(unwind_lock); - static LIST_HEAD(unwind_tables); - - /* Convert a prel31 symbol to an absolute address */ -@@ -201,7 +201,7 @@ - /* module unwind tables */ - struct unwind_table *table; - -- spin_lock_irqsave(&unwind_lock, flags); -+ raw_spin_lock_irqsave(&unwind_lock, flags); - list_for_each_entry(table, &unwind_tables, list) { - if (addr >= table->begin_addr && - addr < table->end_addr) { -@@ -213,7 +213,7 @@ - break; - } - } -- spin_unlock_irqrestore(&unwind_lock, flags); -+ raw_spin_unlock_irqrestore(&unwind_lock, flags); - } - - pr_debug("%s: idx = %p\n", __func__, idx); -@@ -529,9 +529,9 @@ - tab->begin_addr = text_addr; - tab->end_addr = text_addr + text_size; - -- spin_lock_irqsave(&unwind_lock, flags); -+ raw_spin_lock_irqsave(&unwind_lock, flags); - list_add_tail(&tab->list, &unwind_tables); -- spin_unlock_irqrestore(&unwind_lock, flags); -+ raw_spin_unlock_irqrestore(&unwind_lock, flags); - - return tab; - } -@@ -543,9 +543,9 @@ - if (!tab) - return; - -- spin_lock_irqsave(&unwind_lock, flags); -+ raw_spin_lock_irqsave(&unwind_lock, flags); - list_del(&tab->list); -- spin_unlock_irqrestore(&unwind_lock, flags); -+ raw_spin_unlock_irqrestore(&unwind_lock, flags); - - kfree(tab); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kvm/arm.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kvm/arm.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/kvm/arm.c 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/kvm/arm.c 2017-04-18 17:54:19.000000000 +0200 -@@ -619,7 +619,7 @@ - * involves poking the GIC, which must be done in a - * non-preemptible context. - */ -- preempt_disable(); -+ migrate_disable(); - kvm_pmu_flush_hwstate(vcpu); - kvm_timer_flush_hwstate(vcpu); - kvm_vgic_flush_hwstate(vcpu); -@@ -640,7 +640,7 @@ - kvm_pmu_sync_hwstate(vcpu); - kvm_timer_sync_hwstate(vcpu); - kvm_vgic_sync_hwstate(vcpu); -- preempt_enable(); -+ migrate_enable(); - continue; - } - -@@ -696,7 +696,7 @@ - - kvm_vgic_sync_hwstate(vcpu); - -- preempt_enable(); -+ migrate_enable(); - - ret = handle_exit(vcpu, run, ret); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-exynos/platsmp.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-exynos/platsmp.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-exynos/platsmp.c 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-exynos/platsmp.c 2017-04-18 17:54:19.000000000 +0200 -@@ -229,7 +229,7 @@ - return (void __iomem *)(S5P_VA_SCU); - } - --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); - - static void exynos_secondary_init(unsigned int cpu) - { -@@ -242,8 +242,8 @@ - /* - * Synchronise with the boot thread. - */ -- spin_lock(&boot_lock); -- spin_unlock(&boot_lock); -+ raw_spin_lock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - } - - int exynos_set_boot_addr(u32 core_id, unsigned long boot_addr) -@@ -307,7 +307,7 @@ - * Set synchronisation state between this boot processor - * and the secondary one - */ -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); - - /* - * The secondary processor is waiting to be released from -@@ -334,7 +334,7 @@ - - if (timeout == 0) { - printk(KERN_ERR "cpu1 power enable failed"); -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - return -ETIMEDOUT; - } - } -@@ -380,7 +380,7 @@ - * calibrations, then wait for it to finish - */ - fail: -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - - return pen_release != -1 ? ret : 0; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-hisi/platmcpm.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-hisi/platmcpm.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-hisi/platmcpm.c 2017-04-16 10:37:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-hisi/platmcpm.c 2017-04-18 17:54:19.000000000 +0200 -@@ -61,7 +61,7 @@ - - static void __iomem *sysctrl, *fabric; - static int hip04_cpu_table[HIP04_MAX_CLUSTERS][HIP04_MAX_CPUS_PER_CLUSTER]; --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); - static u32 fabric_phys_addr; - /* - * [0]: bootwrapper physical address -@@ -113,7 +113,7 @@ - if (cluster >= HIP04_MAX_CLUSTERS || cpu >= HIP04_MAX_CPUS_PER_CLUSTER) - return -EINVAL; - -- spin_lock_irq(&boot_lock); -+ raw_spin_lock_irq(&boot_lock); - - if (hip04_cpu_table[cluster][cpu]) - goto out; -@@ -147,7 +147,7 @@ - - out: - hip04_cpu_table[cluster][cpu]++; -- spin_unlock_irq(&boot_lock); -+ raw_spin_unlock_irq(&boot_lock); - - return 0; - } -@@ -162,11 +162,11 @@ - cpu = MPIDR_AFFINITY_LEVEL(mpidr, 0); - cluster = MPIDR_AFFINITY_LEVEL(mpidr, 1); - -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); - hip04_cpu_table[cluster][cpu]--; - if (hip04_cpu_table[cluster][cpu] == 1) { - /* A power_up request went ahead of us. */ -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - return; - } else if (hip04_cpu_table[cluster][cpu] > 1) { - pr_err("Cluster %d CPU%d boots multiple times\n", cluster, cpu); -@@ -174,7 +174,7 @@ - } - - last_man = hip04_cluster_is_down(cluster); -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - if (last_man) { - /* Since it's Cortex A15, disable L2 prefetching. */ - asm volatile( -@@ -203,7 +203,7 @@ - cpu >= HIP04_MAX_CPUS_PER_CLUSTER); - - count = TIMEOUT_MSEC / POLL_MSEC; -- spin_lock_irq(&boot_lock); -+ raw_spin_lock_irq(&boot_lock); - for (tries = 0; tries < count; tries++) { - if (hip04_cpu_table[cluster][cpu]) - goto err; -@@ -211,10 +211,10 @@ - data = readl_relaxed(sysctrl + SC_CPU_RESET_STATUS(cluster)); - if (data & CORE_WFI_STATUS(cpu)) - break; -- spin_unlock_irq(&boot_lock); -+ raw_spin_unlock_irq(&boot_lock); - /* Wait for clean L2 when the whole cluster is down. */ - msleep(POLL_MSEC); -- spin_lock_irq(&boot_lock); -+ raw_spin_lock_irq(&boot_lock); - } - if (tries >= count) - goto err; -@@ -231,10 +231,10 @@ - goto err; - if (hip04_cluster_is_down(cluster)) - hip04_set_snoop_filter(cluster, 0); -- spin_unlock_irq(&boot_lock); -+ raw_spin_unlock_irq(&boot_lock); - return 1; - err: -- spin_unlock_irq(&boot_lock); -+ raw_spin_unlock_irq(&boot_lock); - return 0; - } - #endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-omap2/omap-smp.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-omap2/omap-smp.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-omap2/omap-smp.c 2017-04-16 10:37:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-omap2/omap-smp.c 2017-04-18 17:54:19.000000000 +0200 -@@ -64,7 +64,7 @@ - .startup_addr = omap5_secondary_startup, - }; - --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); - - void __iomem *omap4_get_scu_base(void) - { -@@ -131,8 +131,8 @@ - /* - * Synchronise with the boot thread. - */ -- spin_lock(&boot_lock); -- spin_unlock(&boot_lock); -+ raw_spin_lock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - } - - static int omap4_boot_secondary(unsigned int cpu, struct task_struct *idle) -@@ -146,7 +146,7 @@ - * Set synchronisation state between this boot processor - * and the secondary one - */ -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); - - /* - * Update the AuxCoreBoot0 with boot state for secondary core. -@@ -223,7 +223,7 @@ - * Now the secondary core is starting up let it run its - * calibrations, then wait for it to finish - */ -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - - return 0; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-prima2/platsmp.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-prima2/platsmp.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-prima2/platsmp.c 2017-04-16 10:37:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-prima2/platsmp.c 2017-04-18 17:54:19.000000000 +0200 -@@ -22,7 +22,7 @@ - - static void __iomem *clk_base; - --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); - - static void sirfsoc_secondary_init(unsigned int cpu) - { -@@ -36,8 +36,8 @@ - /* - * Synchronise with the boot thread. - */ -- spin_lock(&boot_lock); -- spin_unlock(&boot_lock); -+ raw_spin_lock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - } - - static const struct of_device_id clk_ids[] = { -@@ -75,7 +75,7 @@ - /* make sure write buffer is drained */ - mb(); - -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); - - /* - * The secondary processor is waiting to be released from -@@ -107,7 +107,7 @@ - * now the secondary core is starting up let it run its - * calibrations, then wait for it to finish - */ -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - - return pen_release != -1 ? -ENOSYS : 0; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-qcom/platsmp.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-qcom/platsmp.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-qcom/platsmp.c 2017-04-16 10:37:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-qcom/platsmp.c 2017-04-18 17:54:20.000000000 +0200 -@@ -46,7 +46,7 @@ - - extern void secondary_startup_arm(void); - --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); - - #ifdef CONFIG_HOTPLUG_CPU - static void qcom_cpu_die(unsigned int cpu) -@@ -60,8 +60,8 @@ - /* - * Synchronise with the boot thread. - */ -- spin_lock(&boot_lock); -- spin_unlock(&boot_lock); -+ raw_spin_lock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - } - - static int scss_release_secondary(unsigned int cpu) -@@ -284,7 +284,7 @@ - * set synchronisation state between this boot processor - * and the secondary one - */ -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); - - /* - * Send the secondary CPU a soft interrupt, thereby causing -@@ -297,7 +297,7 @@ - * now the secondary core is starting up let it run its - * calibrations, then wait for it to finish - */ -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - - return ret; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-spear/platsmp.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-spear/platsmp.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-spear/platsmp.c 2017-04-16 10:37:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-spear/platsmp.c 2017-04-18 17:54:20.000000000 +0200 -@@ -32,7 +32,7 @@ - sync_cache_w(&pen_release); - } - --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); - - static void __iomem *scu_base = IOMEM(VA_SCU_BASE); - -@@ -47,8 +47,8 @@ - /* - * Synchronise with the boot thread. - */ -- spin_lock(&boot_lock); -- spin_unlock(&boot_lock); -+ raw_spin_lock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - } - - static int spear13xx_boot_secondary(unsigned int cpu, struct task_struct *idle) -@@ -59,7 +59,7 @@ - * set synchronisation state between this boot processor - * and the secondary one - */ -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); - - /* - * The secondary processor is waiting to be released from -@@ -84,7 +84,7 @@ - * now the secondary core is starting up let it run its - * calibrations, then wait for it to finish - */ -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - - return pen_release != -1 ? -ENOSYS : 0; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-sti/platsmp.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-sti/platsmp.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mach-sti/platsmp.c 2017-04-16 10:37:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mach-sti/platsmp.c 2017-04-18 17:54:20.000000000 +0200 -@@ -35,7 +35,7 @@ - sync_cache_w(&pen_release); - } - --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); - - static void sti_secondary_init(unsigned int cpu) - { -@@ -48,8 +48,8 @@ - /* - * Synchronise with the boot thread. - */ -- spin_lock(&boot_lock); -- spin_unlock(&boot_lock); -+ raw_spin_lock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - } - - static int sti_boot_secondary(unsigned int cpu, struct task_struct *idle) -@@ -60,7 +60,7 @@ - * set synchronisation state between this boot processor - * and the secondary one - */ -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); - - /* - * The secondary processor is waiting to be released from -@@ -91,7 +91,7 @@ - * now the secondary core is starting up let it run its - * calibrations, then wait for it to finish - */ -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - - return pen_release != -1 ? -ENOSYS : 0; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mm/fault.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mm/fault.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mm/fault.c 2017-04-16 10:37:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mm/fault.c 2017-04-18 17:54:20.000000000 +0200 -@@ -430,6 +430,9 @@ - if (addr < TASK_SIZE) - return do_page_fault(addr, fsr, regs); - -+ if (interrupts_enabled(regs)) -+ local_irq_enable(); -+ - if (user_mode(regs)) - goto bad_area; - -@@ -497,6 +500,9 @@ - static int - do_sect_fault(unsigned long addr, unsigned int fsr, struct pt_regs *regs) - { -+ if (interrupts_enabled(regs)) -+ local_irq_enable(); -+ - do_bad_area(addr, fsr, regs); - return 0; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mm/highmem.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mm/highmem.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/mm/highmem.c 2017-04-16 10:37:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/mm/highmem.c 2017-04-18 17:54:20.000000000 +0200 -@@ -34,6 +34,11 @@ - return *ptep; - } - -+static unsigned int fixmap_idx(int type) -+{ -+ return FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); -+} -+ - void *kmap(struct page *page) - { - might_sleep(); -@@ -54,12 +59,13 @@ - - void *kmap_atomic(struct page *page) - { -+ pte_t pte = mk_pte(page, kmap_prot); - unsigned int idx; - unsigned long vaddr; - void *kmap; - int type; - -- preempt_disable(); -+ preempt_disable_nort(); - pagefault_disable(); - if (!PageHighMem(page)) - return page_address(page); -@@ -79,7 +85,7 @@ - - type = kmap_atomic_idx_push(); - -- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); -+ idx = fixmap_idx(type); - vaddr = __fix_to_virt(idx); - #ifdef CONFIG_DEBUG_HIGHMEM - /* -@@ -93,7 +99,10 @@ - * in place, so the contained TLB flush ensures the TLB is updated - * with the new mapping. - */ -- set_fixmap_pte(idx, mk_pte(page, kmap_prot)); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = pte; -+#endif -+ set_fixmap_pte(idx, pte); - - return (void *)vaddr; - } -@@ -106,44 +115,75 @@ - - if (kvaddr >= (void *)FIXADDR_START) { - type = kmap_atomic_idx(); -- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); -+ idx = fixmap_idx(type); - - if (cache_is_vivt()) - __cpuc_flush_dcache_area((void *)vaddr, PAGE_SIZE); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = __pte(0); -+#endif - #ifdef CONFIG_DEBUG_HIGHMEM - BUG_ON(vaddr != __fix_to_virt(idx)); -- set_fixmap_pte(idx, __pte(0)); - #else - (void) idx; /* to kill a warning */ - #endif -+ set_fixmap_pte(idx, __pte(0)); - kmap_atomic_idx_pop(); - } else if (vaddr >= PKMAP_ADDR(0) && vaddr < PKMAP_ADDR(LAST_PKMAP)) { - /* this address was obtained through kmap_high_get() */ - kunmap_high(pte_page(pkmap_page_table[PKMAP_NR(vaddr)])); - } - pagefault_enable(); -- preempt_enable(); -+ preempt_enable_nort(); - } - EXPORT_SYMBOL(__kunmap_atomic); - - void *kmap_atomic_pfn(unsigned long pfn) - { -+ pte_t pte = pfn_pte(pfn, kmap_prot); - unsigned long vaddr; - int idx, type; - struct page *page = pfn_to_page(pfn); - -- preempt_disable(); -+ preempt_disable_nort(); - pagefault_disable(); - if (!PageHighMem(page)) - return page_address(page); - - type = kmap_atomic_idx_push(); -- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); -+ idx = fixmap_idx(type); - vaddr = __fix_to_virt(idx); - #ifdef CONFIG_DEBUG_HIGHMEM - BUG_ON(!pte_none(get_fixmap_pte(vaddr))); - #endif -- set_fixmap_pte(idx, pfn_pte(pfn, kmap_prot)); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = pte; -+#endif -+ set_fixmap_pte(idx, pte); - - return (void *)vaddr; - } -+#if defined CONFIG_PREEMPT_RT_FULL -+void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) -+{ -+ int i; -+ -+ /* -+ * Clear @prev's kmap_atomic mappings -+ */ -+ for (i = 0; i < prev_p->kmap_idx; i++) { -+ int idx = fixmap_idx(i); -+ -+ set_fixmap_pte(idx, __pte(0)); -+ } -+ /* -+ * Restore @next_p's kmap_atomic mappings -+ */ -+ for (i = 0; i < next_p->kmap_idx; i++) { -+ int idx = fixmap_idx(i); -+ -+ if (!pte_none(next_p->kmap_pte[i])) -+ set_fixmap_pte(idx, next_p->kmap_pte[i]); -+ } -+} -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/plat-versatile/platsmp.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/plat-versatile/platsmp.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm/plat-versatile/platsmp.c 2017-04-16 10:37:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm/plat-versatile/platsmp.c 2017-04-18 17:54:20.000000000 +0200 -@@ -32,7 +32,7 @@ - sync_cache_w(&pen_release); - } - --static DEFINE_SPINLOCK(boot_lock); -+static DEFINE_RAW_SPINLOCK(boot_lock); - - void versatile_secondary_init(unsigned int cpu) - { -@@ -45,8 +45,8 @@ - /* - * Synchronise with the boot thread. - */ -- spin_lock(&boot_lock); -- spin_unlock(&boot_lock); -+ raw_spin_lock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - } - - int versatile_boot_secondary(unsigned int cpu, struct task_struct *idle) -@@ -57,7 +57,7 @@ - * Set synchronisation state between this boot processor - * and the secondary one - */ -- spin_lock(&boot_lock); -+ raw_spin_lock(&boot_lock); - - /* - * This is really belt and braces; we hold unintended secondary -@@ -87,7 +87,7 @@ - * now the secondary core is starting up let it run its - * calibrations, then wait for it to finish - */ -- spin_unlock(&boot_lock); -+ raw_spin_unlock(&boot_lock); - - return pen_release != -1 ? -ENOSYS : 0; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm64/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm64/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm64/Kconfig 2017-04-16 10:37:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm64/Kconfig 2017-04-18 17:54:20.000000000 +0200 -@@ -91,6 +91,7 @@ - select HAVE_PERF_EVENTS - select HAVE_PERF_REGS - select HAVE_PERF_USER_STACK_DUMP -+ select HAVE_PREEMPT_LAZY - select HAVE_REGS_AND_STACK_ACCESS_API - select HAVE_RCU_TABLE_FREE - select HAVE_SYSCALL_TRACEPOINTS -@@ -704,7 +705,7 @@ - - config XEN - bool "Xen guest support on ARM64" -- depends on ARM64 && OF -+ depends on ARM64 && OF && !PREEMPT_RT_FULL - select SWIOTLB_XEN - select PARAVIRT - help -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm64/include/asm/thread_info.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm64/include/asm/thread_info.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm64/include/asm/thread_info.h 2017-04-16 10:37:32.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm64/include/asm/thread_info.h 2017-04-18 17:54:20.000000000 +0200 -@@ -49,6 +49,7 @@ - mm_segment_t addr_limit; /* address limit */ - struct task_struct *task; /* main task structure */ - int preempt_count; /* 0 => preemptable, <0 => bug */ -+ int preempt_lazy_count; /* 0 => preemptable, <0 => bug */ - int cpu; /* cpu */ - }; - -@@ -112,6 +113,7 @@ - #define TIF_NEED_RESCHED 1 - #define TIF_NOTIFY_RESUME 2 /* callback before returning to user */ - #define TIF_FOREIGN_FPSTATE 3 /* CPU's FP state is not current's */ -+#define TIF_NEED_RESCHED_LAZY 4 - #define TIF_NOHZ 7 - #define TIF_SYSCALL_TRACE 8 - #define TIF_SYSCALL_AUDIT 9 -@@ -127,6 +129,7 @@ - #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) - #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) - #define _TIF_FOREIGN_FPSTATE (1 << TIF_FOREIGN_FPSTATE) -+#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) - #define _TIF_NOHZ (1 << TIF_NOHZ) - #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) - #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) -@@ -135,7 +138,9 @@ - #define _TIF_32BIT (1 << TIF_32BIT) - - #define _TIF_WORK_MASK (_TIF_NEED_RESCHED | _TIF_SIGPENDING | \ -- _TIF_NOTIFY_RESUME | _TIF_FOREIGN_FPSTATE) -+ _TIF_NOTIFY_RESUME | _TIF_FOREIGN_FPSTATE | \ -+ _TIF_NEED_RESCHED_LAZY) -+#define _TIF_NEED_RESCHED_MASK (_TIF_NEED_RESCHED | _TIF_NEED_RESCHED_LAZY) - - #define _TIF_SYSCALL_WORK (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | \ - _TIF_SYSCALL_TRACEPOINT | _TIF_SECCOMP | \ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm64/kernel/asm-offsets.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm64/kernel/asm-offsets.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm64/kernel/asm-offsets.c 2017-04-16 10:37:32.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm64/kernel/asm-offsets.c 2017-04-18 17:54:20.000000000 +0200 -@@ -38,6 +38,7 @@ - BLANK(); - DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); - DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); -+ DEFINE(TI_PREEMPT_LAZY, offsetof(struct thread_info, preempt_lazy_count)); - DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit)); - DEFINE(TI_TASK, offsetof(struct thread_info, task)); - DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm64/kernel/entry.S linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm64/kernel/entry.S ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm64/kernel/entry.S 2017-04-16 10:37:32.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm64/kernel/entry.S 2017-04-18 17:54:20.000000000 +0200 -@@ -428,11 +428,16 @@ - - #ifdef CONFIG_PREEMPT - ldr w24, [tsk, #TI_PREEMPT] // get preempt count -- cbnz w24, 1f // preempt count != 0 -+ cbnz w24, 2f // preempt count != 0 - ldr x0, [tsk, #TI_FLAGS] // get flags -- tbz x0, #TIF_NEED_RESCHED, 1f // needs rescheduling? -- bl el1_preempt -+ tbnz x0, #TIF_NEED_RESCHED, 1f // needs rescheduling? -+ -+ ldr w24, [tsk, #TI_PREEMPT_LAZY] // get preempt lazy count -+ cbnz w24, 2f // preempt lazy count != 0 -+ tbz x0, #TIF_NEED_RESCHED_LAZY, 2f // needs rescheduling? - 1: -+ bl el1_preempt -+2: - #endif - #ifdef CONFIG_TRACE_IRQFLAGS - bl trace_hardirqs_on -@@ -446,6 +451,7 @@ - 1: bl preempt_schedule_irq // irq en/disable is done inside - ldr x0, [tsk, #TI_FLAGS] // get new tasks TI_FLAGS - tbnz x0, #TIF_NEED_RESCHED, 1b // needs rescheduling? -+ tbnz x0, #TIF_NEED_RESCHED_LAZY, 1b // needs rescheduling? - ret x24 - #endif - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm64/kernel/signal.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm64/kernel/signal.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/arm64/kernel/signal.c 2017-04-16 10:37:32.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/arm64/kernel/signal.c 2017-04-18 17:54:20.000000000 +0200 -@@ -409,7 +409,7 @@ - */ - trace_hardirqs_off(); - do { -- if (thread_flags & _TIF_NEED_RESCHED) { -+ if (thread_flags & _TIF_NEED_RESCHED_MASK) { - schedule(); - } else { - local_irq_enable(); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/mips/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/mips/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/mips/Kconfig 2017-04-16 10:37:33.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/mips/Kconfig 2017-04-18 17:54:20.000000000 +0200 -@@ -2514,7 +2514,7 @@ - # - config HIGHMEM - bool "High Memory Support" -- depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM && !CPU_MIPS32_3_5_EVA -+ depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM && !CPU_MIPS32_3_5_EVA && !PREEMPT_RT_FULL - - config CPU_SUPPORTS_HIGHMEM - bool -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/powerpc/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/powerpc/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/powerpc/Kconfig 2017-04-16 10:37:34.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/powerpc/Kconfig 2017-04-18 17:54:20.000000000 +0200 -@@ -52,10 +52,11 @@ - - config RWSEM_GENERIC_SPINLOCK - bool -+ default y if PREEMPT_RT_FULL - - config RWSEM_XCHGADD_ALGORITHM - bool -- default y -+ default y if !PREEMPT_RT_FULL - - config GENERIC_LOCKBREAK - bool -@@ -134,6 +135,7 @@ - select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST - select GENERIC_STRNCPY_FROM_USER - select GENERIC_STRNLEN_USER -+ select HAVE_PREEMPT_LAZY - select HAVE_MOD_ARCH_SPECIFIC - select MODULES_USE_ELF_RELA - select CLONE_BACKWARDS -@@ -321,7 +323,7 @@ - - config HIGHMEM - bool "High memory support" -- depends on PPC32 -+ depends on PPC32 && !PREEMPT_RT_FULL - - source kernel/Kconfig.hz - source kernel/Kconfig.preempt -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/powerpc/include/asm/thread_info.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/powerpc/include/asm/thread_info.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/powerpc/include/asm/thread_info.h 2017-04-16 10:37:35.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/powerpc/include/asm/thread_info.h 2017-04-18 17:54:20.000000000 +0200 -@@ -43,6 +43,8 @@ - int cpu; /* cpu we're on */ - int preempt_count; /* 0 => preemptable, - <0 => BUG */ -+ int preempt_lazy_count; /* 0 => preemptable, -+ <0 => BUG */ - unsigned long local_flags; /* private flags for thread */ - #ifdef CONFIG_LIVEPATCH - unsigned long *livepatch_sp; -@@ -88,8 +90,7 @@ - #define TIF_SYSCALL_TRACE 0 /* syscall trace active */ - #define TIF_SIGPENDING 1 /* signal pending */ - #define TIF_NEED_RESCHED 2 /* rescheduling necessary */ --#define TIF_POLLING_NRFLAG 3 /* true if poll_idle() is polling -- TIF_NEED_RESCHED */ -+#define TIF_NEED_RESCHED_LAZY 3 /* lazy rescheduling necessary */ - #define TIF_32BIT 4 /* 32 bit binary */ - #define TIF_RESTORE_TM 5 /* need to restore TM FP/VEC/VSX */ - #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ -@@ -107,6 +108,8 @@ - #if defined(CONFIG_PPC64) - #define TIF_ELF2ABI 18 /* function descriptors must die! */ - #endif -+#define TIF_POLLING_NRFLAG 19 /* true if poll_idle() is polling -+ TIF_NEED_RESCHED */ - - /* as above, but as bit values */ - #define _TIF_SYSCALL_TRACE (1<flags) - set_bits(irqtp->flags, &curtp->flags); - } -+#endif - - irq_hw_number_t virq_to_hw(unsigned int virq) - { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/powerpc/kernel/misc_32.S linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/powerpc/kernel/misc_32.S ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/powerpc/kernel/misc_32.S 2017-04-16 10:37:35.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/powerpc/kernel/misc_32.S 2017-04-18 17:54:20.000000000 +0200 -@@ -41,6 +41,7 @@ - * We store the saved ksp_limit in the unused part - * of the STACK_FRAME_OVERHEAD - */ -+#ifndef CONFIG_PREEMPT_RT_FULL - _GLOBAL(call_do_softirq) - mflr r0 - stw r0,4(r1) -@@ -57,6 +58,7 @@ - stw r10,THREAD+KSP_LIMIT(r2) - mtlr r0 - blr -+#endif - - /* - * void call_do_irq(struct pt_regs *regs, struct thread_info *irqtp); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/powerpc/kernel/misc_64.S linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/powerpc/kernel/misc_64.S ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/powerpc/kernel/misc_64.S 2017-04-16 10:37:35.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/powerpc/kernel/misc_64.S 2017-04-18 17:54:20.000000000 +0200 -@@ -31,6 +31,7 @@ - - .text - -+#ifndef CONFIG_PREEMPT_RT_FULL - _GLOBAL(call_do_softirq) - mflr r0 - std r0,16(r1) -@@ -41,6 +42,7 @@ - ld r0,16(r1) - mtlr r0 - blr -+#endif - - _GLOBAL(call_do_irq) - mflr r0 -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/powerpc/kvm/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/powerpc/kvm/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/powerpc/kvm/Kconfig 2017-04-16 10:37:35.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/powerpc/kvm/Kconfig 2017-04-18 17:54:20.000000000 +0200 -@@ -175,6 +175,7 @@ - config KVM_MPIC - bool "KVM in-kernel MPIC emulation" - depends on KVM && E500 -+ depends on !PREEMPT_RT_FULL - select HAVE_KVM_IRQCHIP - select HAVE_KVM_IRQFD - select HAVE_KVM_IRQ_ROUTING -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/powerpc/platforms/ps3/device-init.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/powerpc/platforms/ps3/device-init.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/powerpc/platforms/ps3/device-init.c 2017-04-16 10:37:35.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/powerpc/platforms/ps3/device-init.c 2017-04-18 17:54:20.000000000 +0200 -@@ -752,7 +752,7 @@ - } - pr_debug("%s:%u: notification %s issued\n", __func__, __LINE__, op); - -- res = wait_event_interruptible(dev->done.wait, -+ res = swait_event_interruptible(dev->done.wait, - dev->done.done || kthread_should_stop()); - if (kthread_should_stop()) - res = -EINTR; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/sh/kernel/irq.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/sh/kernel/irq.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/sh/kernel/irq.c 2017-04-16 10:37:36.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/sh/kernel/irq.c 2017-04-18 17:54:20.000000000 +0200 -@@ -147,6 +147,7 @@ - hardirq_ctx[cpu] = NULL; - } - -+#ifndef CONFIG_PREEMPT_RT_FULL - void do_softirq_own_stack(void) - { - struct thread_info *curctx; -@@ -174,6 +175,7 @@ - "r5", "r6", "r7", "r8", "r9", "r15", "t", "pr" - ); - } -+#endif - #else - static inline void handle_one_irq(unsigned int irq) - { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/sparc/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/sparc/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/sparc/Kconfig 2017-04-16 10:37:36.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/sparc/Kconfig 2017-04-18 17:54:21.000000000 +0200 -@@ -194,12 +194,10 @@ - source kernel/Kconfig.hz - - config RWSEM_GENERIC_SPINLOCK -- bool -- default y if SPARC32 -+ def_bool PREEMPT_RT_FULL - - config RWSEM_XCHGADD_ALGORITHM -- bool -- default y if SPARC64 -+ def_bool !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL - - config GENERIC_HWEIGHT - bool -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/sparc/kernel/irq_64.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/sparc/kernel/irq_64.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/sparc/kernel/irq_64.c 2017-04-16 10:37:36.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/sparc/kernel/irq_64.c 2017-04-18 17:54:21.000000000 +0200 -@@ -854,6 +854,7 @@ - set_irq_regs(old_regs); - } - -+#ifndef CONFIG_PREEMPT_RT_FULL - void do_softirq_own_stack(void) - { - void *orig_sp, *sp = softirq_stack[smp_processor_id()]; -@@ -868,6 +869,7 @@ - __asm__ __volatile__("mov %0, %%sp" - : : "r" (orig_sp)); - } -+#endif - - #ifdef CONFIG_HOTPLUG_CPU - void fixup_irqs(void) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/Kconfig 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/Kconfig 2017-04-18 17:54:21.000000000 +0200 -@@ -17,6 +17,7 @@ - ### Arch settings - config X86 - def_bool y -+ select HAVE_PREEMPT_LAZY - select ACPI_LEGACY_TABLES_LOOKUP if ACPI - select ACPI_SYSTEM_POWER_STATES_SUPPORT if ACPI - select ANON_INODES -@@ -232,8 +233,11 @@ - def_bool y - depends on ISA_DMA_API - -+config RWSEM_GENERIC_SPINLOCK -+ def_bool PREEMPT_RT_FULL -+ - config RWSEM_XCHGADD_ALGORITHM -- def_bool y -+ def_bool !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL - - config GENERIC_CALIBRATE_DELAY - def_bool y -@@ -897,7 +901,7 @@ - config MAXSMP - bool "Enable Maximum number of SMP Processors and NUMA Nodes" - depends on X86_64 && SMP && DEBUG_KERNEL -- select CPUMASK_OFFSTACK -+ select CPUMASK_OFFSTACK if !PREEMPT_RT_FULL - ---help--- - Enable maximum number of CPUS and NUMA Nodes for this architecture. - If unsure, say N. -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/crypto/aesni-intel_glue.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/crypto/aesni-intel_glue.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/crypto/aesni-intel_glue.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/crypto/aesni-intel_glue.c 2017-04-18 17:54:21.000000000 +0200 -@@ -372,14 +372,14 @@ - err = blkcipher_walk_virt(desc, &walk); - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; - -- kernel_fpu_begin(); - while ((nbytes = walk.nbytes)) { -+ kernel_fpu_begin(); - aesni_ecb_enc(ctx, walk.dst.virt.addr, walk.src.virt.addr, -- nbytes & AES_BLOCK_MASK); -+ nbytes & AES_BLOCK_MASK); -+ kernel_fpu_end(); - nbytes &= AES_BLOCK_SIZE - 1; - err = blkcipher_walk_done(desc, &walk, nbytes); - } -- kernel_fpu_end(); - - return err; - } -@@ -396,14 +396,14 @@ - err = blkcipher_walk_virt(desc, &walk); - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; - -- kernel_fpu_begin(); - while ((nbytes = walk.nbytes)) { -+ kernel_fpu_begin(); - aesni_ecb_dec(ctx, walk.dst.virt.addr, walk.src.virt.addr, - nbytes & AES_BLOCK_MASK); -+ kernel_fpu_end(); - nbytes &= AES_BLOCK_SIZE - 1; - err = blkcipher_walk_done(desc, &walk, nbytes); - } -- kernel_fpu_end(); - - return err; - } -@@ -420,14 +420,14 @@ - err = blkcipher_walk_virt(desc, &walk); - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; - -- kernel_fpu_begin(); - while ((nbytes = walk.nbytes)) { -+ kernel_fpu_begin(); - aesni_cbc_enc(ctx, walk.dst.virt.addr, walk.src.virt.addr, - nbytes & AES_BLOCK_MASK, walk.iv); -+ kernel_fpu_end(); - nbytes &= AES_BLOCK_SIZE - 1; - err = blkcipher_walk_done(desc, &walk, nbytes); - } -- kernel_fpu_end(); - - return err; - } -@@ -444,14 +444,14 @@ - err = blkcipher_walk_virt(desc, &walk); - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; - -- kernel_fpu_begin(); - while ((nbytes = walk.nbytes)) { -+ kernel_fpu_begin(); - aesni_cbc_dec(ctx, walk.dst.virt.addr, walk.src.virt.addr, - nbytes & AES_BLOCK_MASK, walk.iv); -+ kernel_fpu_end(); - nbytes &= AES_BLOCK_SIZE - 1; - err = blkcipher_walk_done(desc, &walk, nbytes); - } -- kernel_fpu_end(); - - return err; - } -@@ -503,18 +503,20 @@ - err = blkcipher_walk_virt_block(desc, &walk, AES_BLOCK_SIZE); - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; - -- kernel_fpu_begin(); - while ((nbytes = walk.nbytes) >= AES_BLOCK_SIZE) { -+ kernel_fpu_begin(); - aesni_ctr_enc_tfm(ctx, walk.dst.virt.addr, walk.src.virt.addr, - nbytes & AES_BLOCK_MASK, walk.iv); -+ kernel_fpu_end(); - nbytes &= AES_BLOCK_SIZE - 1; - err = blkcipher_walk_done(desc, &walk, nbytes); - } - if (walk.nbytes) { -+ kernel_fpu_begin(); - ctr_crypt_final(ctx, &walk); -+ kernel_fpu_end(); - err = blkcipher_walk_done(desc, &walk, 0); - } -- kernel_fpu_end(); - - return err; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/crypto/cast5_avx_glue.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/crypto/cast5_avx_glue.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/crypto/cast5_avx_glue.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/crypto/cast5_avx_glue.c 2017-04-18 17:54:21.000000000 +0200 -@@ -59,7 +59,7 @@ - static int ecb_crypt(struct blkcipher_desc *desc, struct blkcipher_walk *walk, - bool enc) - { -- bool fpu_enabled = false; -+ bool fpu_enabled; - struct cast5_ctx *ctx = crypto_blkcipher_ctx(desc->tfm); - const unsigned int bsize = CAST5_BLOCK_SIZE; - unsigned int nbytes; -@@ -75,7 +75,7 @@ - u8 *wsrc = walk->src.virt.addr; - u8 *wdst = walk->dst.virt.addr; - -- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); -+ fpu_enabled = cast5_fpu_begin(false, nbytes); - - /* Process multi-block batch */ - if (nbytes >= bsize * CAST5_PARALLEL_BLOCKS) { -@@ -103,10 +103,9 @@ - } while (nbytes >= bsize); - - done: -+ cast5_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, walk, nbytes); - } -- -- cast5_fpu_end(fpu_enabled); - return err; - } - -@@ -227,7 +226,7 @@ - static int cbc_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst, - struct scatterlist *src, unsigned int nbytes) - { -- bool fpu_enabled = false; -+ bool fpu_enabled; - struct blkcipher_walk walk; - int err; - -@@ -236,12 +235,11 @@ - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; - - while ((nbytes = walk.nbytes)) { -- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); -+ fpu_enabled = cast5_fpu_begin(false, nbytes); - nbytes = __cbc_decrypt(desc, &walk); -+ cast5_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, &walk, nbytes); - } -- -- cast5_fpu_end(fpu_enabled); - return err; - } - -@@ -311,7 +309,7 @@ - static int ctr_crypt(struct blkcipher_desc *desc, struct scatterlist *dst, - struct scatterlist *src, unsigned int nbytes) - { -- bool fpu_enabled = false; -+ bool fpu_enabled; - struct blkcipher_walk walk; - int err; - -@@ -320,13 +318,12 @@ - desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; - - while ((nbytes = walk.nbytes) >= CAST5_BLOCK_SIZE) { -- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); -+ fpu_enabled = cast5_fpu_begin(false, nbytes); - nbytes = __ctr_crypt(desc, &walk); -+ cast5_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, &walk, nbytes); - } - -- cast5_fpu_end(fpu_enabled); -- - if (walk.nbytes) { - ctr_crypt_final(desc, &walk); - err = blkcipher_walk_done(desc, &walk, 0); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/crypto/glue_helper.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/crypto/glue_helper.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/crypto/glue_helper.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/crypto/glue_helper.c 2017-04-18 17:54:21.000000000 +0200 -@@ -39,7 +39,7 @@ - void *ctx = crypto_blkcipher_ctx(desc->tfm); - const unsigned int bsize = 128 / 8; - unsigned int nbytes, i, func_bytes; -- bool fpu_enabled = false; -+ bool fpu_enabled; - int err; - - err = blkcipher_walk_virt(desc, walk); -@@ -49,7 +49,7 @@ - u8 *wdst = walk->dst.virt.addr; - - fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, -- desc, fpu_enabled, nbytes); -+ desc, false, nbytes); - - for (i = 0; i < gctx->num_funcs; i++) { - func_bytes = bsize * gctx->funcs[i].num_blocks; -@@ -71,10 +71,10 @@ - } - - done: -+ glue_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, walk, nbytes); - } - -- glue_fpu_end(fpu_enabled); - return err; - } - -@@ -194,7 +194,7 @@ - struct scatterlist *src, unsigned int nbytes) - { - const unsigned int bsize = 128 / 8; -- bool fpu_enabled = false; -+ bool fpu_enabled; - struct blkcipher_walk walk; - int err; - -@@ -203,12 +203,12 @@ - - while ((nbytes = walk.nbytes)) { - fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, -- desc, fpu_enabled, nbytes); -+ desc, false, nbytes); - nbytes = __glue_cbc_decrypt_128bit(gctx, desc, &walk); -+ glue_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, &walk, nbytes); - } - -- glue_fpu_end(fpu_enabled); - return err; - } - EXPORT_SYMBOL_GPL(glue_cbc_decrypt_128bit); -@@ -277,7 +277,7 @@ - struct scatterlist *src, unsigned int nbytes) - { - const unsigned int bsize = 128 / 8; -- bool fpu_enabled = false; -+ bool fpu_enabled; - struct blkcipher_walk walk; - int err; - -@@ -286,13 +286,12 @@ - - while ((nbytes = walk.nbytes) >= bsize) { - fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, -- desc, fpu_enabled, nbytes); -+ desc, false, nbytes); - nbytes = __glue_ctr_crypt_128bit(gctx, desc, &walk); -+ glue_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, &walk, nbytes); - } - -- glue_fpu_end(fpu_enabled); -- - if (walk.nbytes) { - glue_ctr_crypt_final_128bit( - gctx->funcs[gctx->num_funcs - 1].fn_u.ctr, desc, &walk); -@@ -347,7 +346,7 @@ - void *tweak_ctx, void *crypt_ctx) - { - const unsigned int bsize = 128 / 8; -- bool fpu_enabled = false; -+ bool fpu_enabled; - struct blkcipher_walk walk; - int err; - -@@ -360,21 +359,21 @@ - - /* set minimum length to bsize, for tweak_fn */ - fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, -- desc, fpu_enabled, -+ desc, false, - nbytes < bsize ? bsize : nbytes); -- - /* calculate first value of T */ - tweak_fn(tweak_ctx, walk.iv, walk.iv); -+ glue_fpu_end(fpu_enabled); - - while (nbytes) { -+ fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, -+ desc, false, nbytes); - nbytes = __glue_xts_crypt_128bit(gctx, crypt_ctx, desc, &walk); - -+ glue_fpu_end(fpu_enabled); - err = blkcipher_walk_done(desc, &walk, nbytes); - nbytes = walk.nbytes; - } -- -- glue_fpu_end(fpu_enabled); -- - return err; - } - EXPORT_SYMBOL_GPL(glue_xts_crypt_128bit); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/entry/common.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/entry/common.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/entry/common.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/entry/common.c 2017-04-18 17:54:21.000000000 +0200 -@@ -129,7 +129,7 @@ - - #define EXIT_TO_USERMODE_LOOP_FLAGS \ - (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME | _TIF_UPROBE | \ -- _TIF_NEED_RESCHED | _TIF_USER_RETURN_NOTIFY) -+ _TIF_NEED_RESCHED_MASK | _TIF_USER_RETURN_NOTIFY) - - static void exit_to_usermode_loop(struct pt_regs *regs, u32 cached_flags) - { -@@ -145,9 +145,16 @@ - /* We have work to do. */ - local_irq_enable(); - -- if (cached_flags & _TIF_NEED_RESCHED) -+ if (cached_flags & _TIF_NEED_RESCHED_MASK) - schedule(); - -+#ifdef ARCH_RT_DELAYS_SIGNAL_SEND -+ if (unlikely(current->forced_info.si_signo)) { -+ struct task_struct *t = current; -+ force_sig_info(t->forced_info.si_signo, &t->forced_info, t); -+ t->forced_info.si_signo = 0; -+ } -+#endif - if (cached_flags & _TIF_UPROBE) - uprobe_notify_resume(regs); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/entry/entry_32.S linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/entry/entry_32.S ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/entry/entry_32.S 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/entry/entry_32.S 2017-04-18 17:54:21.000000000 +0200 -@@ -308,8 +308,25 @@ - ENTRY(resume_kernel) - DISABLE_INTERRUPTS(CLBR_ANY) - need_resched: -+ # preempt count == 0 + NEED_RS set? - cmpl $0, PER_CPU_VAR(__preempt_count) -+#ifndef CONFIG_PREEMPT_LAZY - jnz restore_all -+#else -+ jz test_int_off -+ -+ # atleast preempt count == 0 ? -+ cmpl $_PREEMPT_ENABLED,PER_CPU_VAR(__preempt_count) -+ jne restore_all -+ -+ movl PER_CPU_VAR(current_task), %ebp -+ cmpl $0,TASK_TI_preempt_lazy_count(%ebp) # non-zero preempt_lazy_count ? -+ jnz restore_all -+ -+ testl $_TIF_NEED_RESCHED_LAZY, TASK_TI_flags(%ebp) -+ jz restore_all -+test_int_off: -+#endif - testl $X86_EFLAGS_IF, PT_EFLAGS(%esp) # interrupts off (exception path) ? - jz restore_all - call preempt_schedule_irq -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/entry/entry_64.S linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/entry/entry_64.S ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/entry/entry_64.S 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/entry/entry_64.S 2017-04-18 17:54:21.000000000 +0200 -@@ -546,7 +546,23 @@ - bt $9, EFLAGS(%rsp) /* were interrupts off? */ - jnc 1f - 0: cmpl $0, PER_CPU_VAR(__preempt_count) -+#ifndef CONFIG_PREEMPT_LAZY - jnz 1f -+#else -+ jz do_preempt_schedule_irq -+ -+ # atleast preempt count == 0 ? -+ cmpl $_PREEMPT_ENABLED,PER_CPU_VAR(__preempt_count) -+ jnz 1f -+ -+ movq PER_CPU_VAR(current_task), %rcx -+ cmpl $0, TASK_TI_preempt_lazy_count(%rcx) -+ jnz 1f -+ -+ bt $TIF_NEED_RESCHED_LAZY,TASK_TI_flags(%rcx) -+ jnc 1f -+do_preempt_schedule_irq: -+#endif - call preempt_schedule_irq - jmp 0b - 1: -@@ -894,6 +910,7 @@ - jmp 2b - .previous - -+#ifndef CONFIG_PREEMPT_RT_FULL - /* Call softirq on interrupt stack. Interrupts are off. */ - ENTRY(do_softirq_own_stack) - pushq %rbp -@@ -906,6 +923,7 @@ - decl PER_CPU_VAR(irq_count) - ret - END(do_softirq_own_stack) -+#endif - - #ifdef CONFIG_XEN - idtentry xen_hypervisor_callback xen_do_hypervisor_callback has_error_code=0 -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/include/asm/preempt.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/include/asm/preempt.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/include/asm/preempt.h 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/include/asm/preempt.h 2017-04-18 17:54:21.000000000 +0200 -@@ -79,17 +79,46 @@ - * a decrement which hits zero means we have no preempt_count and should - * reschedule. - */ --static __always_inline bool __preempt_count_dec_and_test(void) -+static __always_inline bool ____preempt_count_dec_and_test(void) - { - GEN_UNARY_RMWcc("decl", __preempt_count, __percpu_arg(0), e); - } - -+static __always_inline bool __preempt_count_dec_and_test(void) -+{ -+ if (____preempt_count_dec_and_test()) -+ return true; -+#ifdef CONFIG_PREEMPT_LAZY -+ if (current_thread_info()->preempt_lazy_count) -+ return false; -+ return test_thread_flag(TIF_NEED_RESCHED_LAZY); -+#else -+ return false; -+#endif -+} -+ - /* - * Returns true when we need to resched and can (barring IRQ state). - */ - static __always_inline bool should_resched(int preempt_offset) - { -+#ifdef CONFIG_PREEMPT_LAZY -+ u32 tmp; -+ -+ tmp = raw_cpu_read_4(__preempt_count); -+ if (tmp == preempt_offset) -+ return true; -+ -+ /* preempt count == 0 ? */ -+ tmp &= ~PREEMPT_NEED_RESCHED; -+ if (tmp) -+ return false; -+ if (current_thread_info()->preempt_lazy_count) -+ return false; -+ return test_thread_flag(TIF_NEED_RESCHED_LAZY); -+#else - return unlikely(raw_cpu_read_4(__preempt_count) == preempt_offset); -+#endif - } - - #ifdef CONFIG_PREEMPT -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/include/asm/signal.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/include/asm/signal.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/include/asm/signal.h 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/include/asm/signal.h 2017-04-18 17:54:21.000000000 +0200 -@@ -27,6 +27,19 @@ - #define SA_IA32_ABI 0x02000000u - #define SA_X32_ABI 0x01000000u - -+/* -+ * Because some traps use the IST stack, we must keep preemption -+ * disabled while calling do_trap(), but do_trap() may call -+ * force_sig_info() which will grab the signal spin_locks for the -+ * task, which in PREEMPT_RT_FULL are mutexes. By defining -+ * ARCH_RT_DELAYS_SIGNAL_SEND the force_sig_info() will set -+ * TIF_NOTIFY_RESUME and set up the signal to be sent on exit of the -+ * trap. -+ */ -+#if defined(CONFIG_PREEMPT_RT_FULL) -+#define ARCH_RT_DELAYS_SIGNAL_SEND -+#endif -+ - #ifndef CONFIG_COMPAT - typedef sigset_t compat_sigset_t; - #endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/include/asm/stackprotector.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/include/asm/stackprotector.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/include/asm/stackprotector.h 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/include/asm/stackprotector.h 2017-04-18 17:54:21.000000000 +0200 -@@ -59,7 +59,7 @@ - */ - static __always_inline void boot_init_stack_canary(void) - { -- u64 canary; -+ u64 uninitialized_var(canary); - u64 tsc; - - #ifdef CONFIG_X86_64 -@@ -70,8 +70,15 @@ - * of randomness. The TSC only matters for very early init, - * there it already has some randomness on most systems. Later - * on during the bootup the random pool has true entropy too. -+ * -+ * For preempt-rt we need to weaken the randomness a bit, as -+ * we can't call into the random generator from atomic context -+ * due to locking constraints. We just leave canary -+ * uninitialized and use the TSC based randomness on top of it. - */ -+#ifndef CONFIG_PREEMPT_RT_FULL - get_random_bytes(&canary, sizeof(canary)); -+#endif - tsc = rdtsc(); - canary += tsc + (tsc << 32UL); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/include/asm/thread_info.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/include/asm/thread_info.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/include/asm/thread_info.h 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/include/asm/thread_info.h 2017-04-18 17:54:21.000000000 +0200 -@@ -54,11 +54,14 @@ - - struct thread_info { - unsigned long flags; /* low level flags */ -+ int preempt_lazy_count; /* 0 => lazy preemptable -+ <0 => BUG */ - }; - - #define INIT_THREAD_INFO(tsk) \ - { \ - .flags = 0, \ -+ .preempt_lazy_count = 0, \ - } - - #define init_stack (init_thread_union.stack) -@@ -67,6 +70,10 @@ - - #include - -+#define GET_THREAD_INFO(reg) \ -+ _ASM_MOV PER_CPU_VAR(cpu_current_top_of_stack),reg ; \ -+ _ASM_SUB $(THREAD_SIZE),reg ; -+ - #endif - - /* -@@ -85,6 +92,7 @@ - #define TIF_SYSCALL_EMU 6 /* syscall emulation active */ - #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ - #define TIF_SECCOMP 8 /* secure computing */ -+#define TIF_NEED_RESCHED_LAZY 9 /* lazy rescheduling necessary */ - #define TIF_USER_RETURN_NOTIFY 11 /* notify kernel of userspace return */ - #define TIF_UPROBE 12 /* breakpointed or singlestepping */ - #define TIF_NOTSC 16 /* TSC is not accessible in userland */ -@@ -108,6 +116,7 @@ - #define _TIF_SYSCALL_EMU (1 << TIF_SYSCALL_EMU) - #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) - #define _TIF_SECCOMP (1 << TIF_SECCOMP) -+#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) - #define _TIF_USER_RETURN_NOTIFY (1 << TIF_USER_RETURN_NOTIFY) - #define _TIF_UPROBE (1 << TIF_UPROBE) - #define _TIF_NOTSC (1 << TIF_NOTSC) -@@ -143,6 +152,8 @@ - #define _TIF_WORK_CTXSW_PREV (_TIF_WORK_CTXSW|_TIF_USER_RETURN_NOTIFY) - #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW) - -+#define _TIF_NEED_RESCHED_MASK (_TIF_NEED_RESCHED | _TIF_NEED_RESCHED_LAZY) -+ - #define STACK_WARN (THREAD_SIZE/8) - - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/include/asm/uv/uv_bau.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/include/asm/uv/uv_bau.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/include/asm/uv/uv_bau.h 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/include/asm/uv/uv_bau.h 2017-04-18 17:54:21.000000000 +0200 -@@ -624,9 +624,9 @@ - cycles_t send_message; - cycles_t period_end; - cycles_t period_time; -- spinlock_t uvhub_lock; -- spinlock_t queue_lock; -- spinlock_t disable_lock; -+ raw_spinlock_t uvhub_lock; -+ raw_spinlock_t queue_lock; -+ raw_spinlock_t disable_lock; - /* tunables */ - int max_concurr; - int max_concurr_const; -@@ -815,15 +815,15 @@ - * to be lowered below the current 'v'. atomic_add_unless can only stop - * on equal. - */ --static inline int atomic_inc_unless_ge(spinlock_t *lock, atomic_t *v, int u) -+static inline int atomic_inc_unless_ge(raw_spinlock_t *lock, atomic_t *v, int u) - { -- spin_lock(lock); -+ raw_spin_lock(lock); - if (atomic_read(v) >= u) { -- spin_unlock(lock); -+ raw_spin_unlock(lock); - return 0; - } - atomic_inc(v); -- spin_unlock(lock); -+ raw_spin_unlock(lock); - return 1; - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kernel/acpi/boot.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kernel/acpi/boot.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kernel/acpi/boot.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kernel/acpi/boot.c 2017-04-18 17:54:21.000000000 +0200 -@@ -87,7 +87,9 @@ - * ->ioapic_mutex - * ->ioapic_lock - */ -+#ifdef CONFIG_X86_IO_APIC - static DEFINE_MUTEX(acpi_ioapic_lock); -+#endif - - /* -------------------------------------------------------------------------- - Boot-time Configuration -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kernel/apic/io_apic.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kernel/apic/io_apic.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kernel/apic/io_apic.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kernel/apic/io_apic.c 2017-04-18 17:54:21.000000000 +0200 -@@ -1712,7 +1712,8 @@ - static inline bool ioapic_irqd_mask(struct irq_data *data) - { - /* If we are moving the irq we need to mask it */ -- if (unlikely(irqd_is_setaffinity_pending(data))) { -+ if (unlikely(irqd_is_setaffinity_pending(data) && -+ !irqd_irq_inprogress(data))) { - mask_ioapic_irq(data); - return true; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kernel/asm-offsets.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kernel/asm-offsets.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kernel/asm-offsets.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kernel/asm-offsets.c 2017-04-18 17:54:21.000000000 +0200 -@@ -36,6 +36,7 @@ - - BLANK(); - OFFSET(TASK_TI_flags, task_struct, thread_info.flags); -+ OFFSET(TASK_TI_preempt_lazy_count, task_struct, thread_info.preempt_lazy_count); - OFFSET(TASK_addr_limit, task_struct, thread.addr_limit); - - BLANK(); -@@ -91,4 +92,5 @@ - - BLANK(); - DEFINE(PTREGS_SIZE, sizeof(struct pt_regs)); -+ DEFINE(_PREEMPT_ENABLED, PREEMPT_ENABLED); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kernel/cpu/mcheck/mce.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kernel/cpu/mcheck/mce.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kernel/cpu/mcheck/mce.c 2017-04-18 17:54:21.000000000 +0200 -@@ -41,6 +41,8 @@ - #include - #include - #include -+#include -+#include - #include - - #include -@@ -1317,7 +1319,7 @@ - static unsigned long check_interval = INITIAL_CHECK_INTERVAL; - - static DEFINE_PER_CPU(unsigned long, mce_next_interval); /* in jiffies */ --static DEFINE_PER_CPU(struct timer_list, mce_timer); -+static DEFINE_PER_CPU(struct hrtimer, mce_timer); - - static unsigned long mce_adjust_timer_default(unsigned long interval) - { -@@ -1326,32 +1328,18 @@ - - static unsigned long (*mce_adjust_timer)(unsigned long interval) = mce_adjust_timer_default; - --static void __restart_timer(struct timer_list *t, unsigned long interval) -+static enum hrtimer_restart __restart_timer(struct hrtimer *timer, unsigned long interval) - { -- unsigned long when = jiffies + interval; -- unsigned long flags; -- -- local_irq_save(flags); -- -- if (timer_pending(t)) { -- if (time_before(when, t->expires)) -- mod_timer(t, when); -- } else { -- t->expires = round_jiffies(when); -- add_timer_on(t, smp_processor_id()); -- } -- -- local_irq_restore(flags); -+ if (!interval) -+ return HRTIMER_NORESTART; -+ hrtimer_forward_now(timer, ns_to_ktime(jiffies_to_nsecs(interval))); -+ return HRTIMER_RESTART; - } - --static void mce_timer_fn(unsigned long data) -+static enum hrtimer_restart mce_timer_fn(struct hrtimer *timer) - { -- struct timer_list *t = this_cpu_ptr(&mce_timer); -- int cpu = smp_processor_id(); - unsigned long iv; - -- WARN_ON(cpu != data); -- - iv = __this_cpu_read(mce_next_interval); - - if (mce_available(this_cpu_ptr(&cpu_info))) { -@@ -1374,7 +1362,7 @@ - - done: - __this_cpu_write(mce_next_interval, iv); -- __restart_timer(t, iv); -+ return __restart_timer(timer, iv); - } - - /* -@@ -1382,7 +1370,7 @@ - */ - void mce_timer_kick(unsigned long interval) - { -- struct timer_list *t = this_cpu_ptr(&mce_timer); -+ struct hrtimer *t = this_cpu_ptr(&mce_timer); - unsigned long iv = __this_cpu_read(mce_next_interval); - - __restart_timer(t, interval); -@@ -1397,7 +1385,7 @@ - int cpu; - - for_each_online_cpu(cpu) -- del_timer_sync(&per_cpu(mce_timer, cpu)); -+ hrtimer_cancel(&per_cpu(mce_timer, cpu)); - } - - static void mce_do_trigger(struct work_struct *work) -@@ -1407,6 +1395,56 @@ - - static DECLARE_WORK(mce_trigger_work, mce_do_trigger); - -+static void __mce_notify_work(struct swork_event *event) -+{ -+ /* Not more than two messages every minute */ -+ static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2); -+ -+ /* wake processes polling /dev/mcelog */ -+ wake_up_interruptible(&mce_chrdev_wait); -+ -+ /* -+ * There is no risk of missing notifications because -+ * work_pending is always cleared before the function is -+ * executed. -+ */ -+ if (mce_helper[0] && !work_pending(&mce_trigger_work)) -+ schedule_work(&mce_trigger_work); -+ -+ if (__ratelimit(&ratelimit)) -+ pr_info(HW_ERR "Machine check events logged\n"); -+} -+ -+#ifdef CONFIG_PREEMPT_RT_FULL -+static bool notify_work_ready __read_mostly; -+static struct swork_event notify_work; -+ -+static int mce_notify_work_init(void) -+{ -+ int err; -+ -+ err = swork_get(); -+ if (err) -+ return err; -+ -+ INIT_SWORK(¬ify_work, __mce_notify_work); -+ notify_work_ready = true; -+ return 0; -+} -+ -+static void mce_notify_work(void) -+{ -+ if (notify_work_ready) -+ swork_queue(¬ify_work); -+} -+#else -+static void mce_notify_work(void) -+{ -+ __mce_notify_work(NULL); -+} -+static inline int mce_notify_work_init(void) { return 0; } -+#endif -+ - /* - * Notify the user(s) about new machine check events. - * Can be called from interrupt context, but not from machine check/NMI -@@ -1414,19 +1452,8 @@ - */ - int mce_notify_irq(void) - { -- /* Not more than two messages every minute */ -- static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2); -- - if (test_and_clear_bit(0, &mce_need_notify)) { -- /* wake processes polling /dev/mcelog */ -- wake_up_interruptible(&mce_chrdev_wait); -- -- if (mce_helper[0]) -- schedule_work(&mce_trigger_work); -- -- if (__ratelimit(&ratelimit)) -- pr_info(HW_ERR "Machine check events logged\n"); -- -+ mce_notify_work(); - return 1; - } - return 0; -@@ -1732,7 +1759,7 @@ - } - } - --static void mce_start_timer(unsigned int cpu, struct timer_list *t) -+static void mce_start_timer(unsigned int cpu, struct hrtimer *t) - { - unsigned long iv = check_interval * HZ; - -@@ -1741,16 +1768,17 @@ - - per_cpu(mce_next_interval, cpu) = iv; - -- t->expires = round_jiffies(jiffies + iv); -- add_timer_on(t, cpu); -+ hrtimer_start_range_ns(t, ns_to_ktime(jiffies_to_usecs(iv) * 1000ULL), -+ 0, HRTIMER_MODE_REL_PINNED); - } - - static void __mcheck_cpu_init_timer(void) - { -- struct timer_list *t = this_cpu_ptr(&mce_timer); -+ struct hrtimer *t = this_cpu_ptr(&mce_timer); - unsigned int cpu = smp_processor_id(); - -- setup_pinned_timer(t, mce_timer_fn, cpu); -+ hrtimer_init(t, CLOCK_MONOTONIC, HRTIMER_MODE_REL); -+ t->function = mce_timer_fn; - mce_start_timer(cpu, t); - } - -@@ -2475,6 +2503,8 @@ - if (!mce_available(raw_cpu_ptr(&cpu_info))) - return; - -+ hrtimer_cancel(this_cpu_ptr(&mce_timer)); -+ - if (!(action & CPU_TASKS_FROZEN)) - cmci_clear(); - -@@ -2497,6 +2527,7 @@ - if (b->init) - wrmsrl(msr_ops.ctl(i), b->ctl); - } -+ __mcheck_cpu_init_timer(); - } - - /* Get notified when a cpu comes on/off. Be hotplug friendly. */ -@@ -2504,7 +2535,6 @@ - mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) - { - unsigned int cpu = (unsigned long)hcpu; -- struct timer_list *t = &per_cpu(mce_timer, cpu); - - switch (action & ~CPU_TASKS_FROZEN) { - case CPU_ONLINE: -@@ -2524,11 +2554,9 @@ - break; - case CPU_DOWN_PREPARE: - smp_call_function_single(cpu, mce_disable_cpu, &action, 1); -- del_timer_sync(t); - break; - case CPU_DOWN_FAILED: - smp_call_function_single(cpu, mce_reenable_cpu, &action, 1); -- mce_start_timer(cpu, t); - break; - } - -@@ -2567,6 +2595,10 @@ - goto err_out; - } - -+ err = mce_notify_work_init(); -+ if (err) -+ goto err_out; -+ - if (!zalloc_cpumask_var(&mce_device_initialized, GFP_KERNEL)) { - err = -ENOMEM; - goto err_out; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kernel/irq_32.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kernel/irq_32.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kernel/irq_32.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kernel/irq_32.c 2017-04-18 17:54:21.000000000 +0200 -@@ -127,6 +127,7 @@ - cpu, per_cpu(hardirq_stack, cpu), per_cpu(softirq_stack, cpu)); - } - -+#ifndef CONFIG_PREEMPT_RT_FULL - void do_softirq_own_stack(void) - { - struct irq_stack *irqstk; -@@ -143,6 +144,7 @@ - - call_on_stack(__do_softirq, isp); - } -+#endif - - bool handle_irq(struct irq_desc *desc, struct pt_regs *regs) - { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kernel/process_32.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kernel/process_32.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kernel/process_32.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kernel/process_32.c 2017-04-18 17:54:21.000000000 +0200 -@@ -35,6 +35,7 @@ - #include - #include - #include -+#include - - #include - #include -@@ -195,6 +196,35 @@ - } - EXPORT_SYMBOL_GPL(start_thread); - -+#ifdef CONFIG_PREEMPT_RT_FULL -+static void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) -+{ -+ int i; -+ -+ /* -+ * Clear @prev's kmap_atomic mappings -+ */ -+ for (i = 0; i < prev_p->kmap_idx; i++) { -+ int idx = i + KM_TYPE_NR * smp_processor_id(); -+ pte_t *ptep = kmap_pte - idx; -+ -+ kpte_clear_flush(ptep, __fix_to_virt(FIX_KMAP_BEGIN + idx)); -+ } -+ /* -+ * Restore @next_p's kmap_atomic mappings -+ */ -+ for (i = 0; i < next_p->kmap_idx; i++) { -+ int idx = i + KM_TYPE_NR * smp_processor_id(); -+ -+ if (!pte_none(next_p->kmap_pte[i])) -+ set_pte(kmap_pte - idx, next_p->kmap_pte[i]); -+ } -+} -+#else -+static inline void -+switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) { } -+#endif -+ - - /* - * switch_to(x,y) should switch tasks from x to y. -@@ -271,6 +301,8 @@ - task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT)) - __switch_to_xtra(prev_p, next_p, tss); - -+ switch_kmaps(prev_p, next_p); -+ - /* - * Leave lazy mode, flushing any hypercalls made here. - * This must be done before restoring TLS segments so -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kvm/lapic.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kvm/lapic.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kvm/lapic.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kvm/lapic.c 2017-04-18 17:54:21.000000000 +0200 -@@ -1939,6 +1939,7 @@ - hrtimer_init(&apic->lapic_timer.timer, CLOCK_MONOTONIC, - HRTIMER_MODE_ABS_PINNED); - apic->lapic_timer.timer.function = apic_timer_fn; -+ apic->lapic_timer.timer.irqsafe = 1; - - /* - * APIC is created enabled. This will prevent kvm_lapic_set_base from -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kvm/x86.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kvm/x86.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/kvm/x86.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/kvm/x86.c 2017-04-18 17:54:21.000000000 +0200 -@@ -5933,6 +5933,13 @@ - goto out; - } - -+#ifdef CONFIG_PREEMPT_RT_FULL -+ if (!boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) { -+ printk(KERN_ERR "RT requires X86_FEATURE_CONSTANT_TSC\n"); -+ return -EOPNOTSUPP; -+ } -+#endif -+ - r = kvm_mmu_module_init(); - if (r) - goto out_free_percpu; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/mm/highmem_32.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/mm/highmem_32.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/mm/highmem_32.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/mm/highmem_32.c 2017-04-18 17:54:21.000000000 +0200 -@@ -32,10 +32,11 @@ - */ - void *kmap_atomic_prot(struct page *page, pgprot_t prot) - { -+ pte_t pte = mk_pte(page, prot); - unsigned long vaddr; - int idx, type; - -- preempt_disable(); -+ preempt_disable_nort(); - pagefault_disable(); - - if (!PageHighMem(page)) -@@ -45,7 +46,10 @@ - idx = type + KM_TYPE_NR*smp_processor_id(); - vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); - BUG_ON(!pte_none(*(kmap_pte-idx))); -- set_pte(kmap_pte-idx, mk_pte(page, prot)); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = pte; -+#endif -+ set_pte(kmap_pte-idx, pte); - arch_flush_lazy_mmu_mode(); - - return (void *)vaddr; -@@ -88,6 +92,9 @@ - * is a bad idea also, in case the page changes cacheability - * attributes or becomes a protected page in a hypervisor. - */ -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = __pte(0); -+#endif - kpte_clear_flush(kmap_pte-idx, vaddr); - kmap_atomic_idx_pop(); - arch_flush_lazy_mmu_mode(); -@@ -100,7 +107,7 @@ - #endif - - pagefault_enable(); -- preempt_enable(); -+ preempt_enable_nort(); - } - EXPORT_SYMBOL(__kunmap_atomic); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/mm/iomap_32.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/mm/iomap_32.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/mm/iomap_32.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/mm/iomap_32.c 2017-04-18 17:54:21.000000000 +0200 -@@ -56,6 +56,7 @@ - - void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot) - { -+ pte_t pte = pfn_pte(pfn, prot); - unsigned long vaddr; - int idx, type; - -@@ -65,7 +66,12 @@ - type = kmap_atomic_idx_push(); - idx = type + KM_TYPE_NR * smp_processor_id(); - vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); -- set_pte(kmap_pte - idx, pfn_pte(pfn, prot)); -+ WARN_ON(!pte_none(*(kmap_pte - idx))); -+ -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = pte; -+#endif -+ set_pte(kmap_pte - idx, pte); - arch_flush_lazy_mmu_mode(); - - return (void *)vaddr; -@@ -113,6 +119,9 @@ - * is a bad idea also, in case the page changes cacheability - * attributes or becomes a protected page in a hypervisor. - */ -+#ifdef CONFIG_PREEMPT_RT_FULL -+ current->kmap_pte[type] = __pte(0); -+#endif - kpte_clear_flush(kmap_pte-idx, vaddr); - kmap_atomic_idx_pop(); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/mm/pageattr.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/mm/pageattr.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/mm/pageattr.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/mm/pageattr.c 2017-04-18 17:54:21.000000000 +0200 -@@ -214,7 +214,15 @@ - int in_flags, struct page **pages) - { - unsigned int i, level; -+#ifdef CONFIG_PREEMPT -+ /* -+ * Avoid wbinvd() because it causes latencies on all CPUs, -+ * regardless of any CPU isolation that may be in effect. -+ */ -+ unsigned long do_wbinvd = 0; -+#else - unsigned long do_wbinvd = cache && numpages >= 1024; /* 4M threshold */ -+#endif - - BUG_ON(irqs_disabled()); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/platform/uv/tlb_uv.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/platform/uv/tlb_uv.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/platform/uv/tlb_uv.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/platform/uv/tlb_uv.c 2017-04-18 17:54:21.000000000 +0200 -@@ -748,9 +748,9 @@ - - quiesce_local_uvhub(hmaster); - -- spin_lock(&hmaster->queue_lock); -+ raw_spin_lock(&hmaster->queue_lock); - reset_with_ipi(&bau_desc->distribution, bcp); -- spin_unlock(&hmaster->queue_lock); -+ raw_spin_unlock(&hmaster->queue_lock); - - end_uvhub_quiesce(hmaster); - -@@ -770,9 +770,9 @@ - - quiesce_local_uvhub(hmaster); - -- spin_lock(&hmaster->queue_lock); -+ raw_spin_lock(&hmaster->queue_lock); - reset_with_ipi(&bau_desc->distribution, bcp); -- spin_unlock(&hmaster->queue_lock); -+ raw_spin_unlock(&hmaster->queue_lock); - - end_uvhub_quiesce(hmaster); - -@@ -793,7 +793,7 @@ - cycles_t tm1; - - hmaster = bcp->uvhub_master; -- spin_lock(&hmaster->disable_lock); -+ raw_spin_lock(&hmaster->disable_lock); - if (!bcp->baudisabled) { - stat->s_bau_disabled++; - tm1 = get_cycles(); -@@ -806,7 +806,7 @@ - } - } - } -- spin_unlock(&hmaster->disable_lock); -+ raw_spin_unlock(&hmaster->disable_lock); - } - - static void count_max_concurr(int stat, struct bau_control *bcp, -@@ -869,7 +869,7 @@ - */ - static void uv1_throttle(struct bau_control *hmaster, struct ptc_stats *stat) - { -- spinlock_t *lock = &hmaster->uvhub_lock; -+ raw_spinlock_t *lock = &hmaster->uvhub_lock; - atomic_t *v; - - v = &hmaster->active_descriptor_count; -@@ -1002,7 +1002,7 @@ - struct bau_control *hmaster; - - hmaster = bcp->uvhub_master; -- spin_lock(&hmaster->disable_lock); -+ raw_spin_lock(&hmaster->disable_lock); - if (bcp->baudisabled && (get_cycles() >= bcp->set_bau_on_time)) { - stat->s_bau_reenabled++; - for_each_present_cpu(tcpu) { -@@ -1014,10 +1014,10 @@ - tbcp->period_giveups = 0; - } - } -- spin_unlock(&hmaster->disable_lock); -+ raw_spin_unlock(&hmaster->disable_lock); - return 0; - } -- spin_unlock(&hmaster->disable_lock); -+ raw_spin_unlock(&hmaster->disable_lock); - return -1; - } - -@@ -1940,9 +1940,9 @@ - bcp->cong_reps = congested_reps; - bcp->disabled_period = sec_2_cycles(disabled_period); - bcp->giveup_limit = giveup_limit; -- spin_lock_init(&bcp->queue_lock); -- spin_lock_init(&bcp->uvhub_lock); -- spin_lock_init(&bcp->disable_lock); -+ raw_spin_lock_init(&bcp->queue_lock); -+ raw_spin_lock_init(&bcp->uvhub_lock); -+ raw_spin_lock_init(&bcp->disable_lock); - } - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/platform/uv/uv_time.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/platform/uv/uv_time.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/arch/x86/platform/uv/uv_time.c 2017-04-16 10:37:37.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/arch/x86/platform/uv/uv_time.c 2017-04-18 17:54:21.000000000 +0200 -@@ -57,7 +57,7 @@ - - /* There is one of these allocated per node */ - struct uv_rtc_timer_head { -- spinlock_t lock; -+ raw_spinlock_t lock; - /* next cpu waiting for timer, local node relative: */ - int next_cpu; - /* number of cpus on this node: */ -@@ -177,7 +177,7 @@ - uv_rtc_deallocate_timers(); - return -ENOMEM; - } -- spin_lock_init(&head->lock); -+ raw_spin_lock_init(&head->lock); - head->ncpus = uv_blade_nr_possible_cpus(bid); - head->next_cpu = -1; - blade_info[bid] = head; -@@ -231,7 +231,7 @@ - unsigned long flags; - int next_cpu; - -- spin_lock_irqsave(&head->lock, flags); -+ raw_spin_lock_irqsave(&head->lock, flags); - - next_cpu = head->next_cpu; - *t = expires; -@@ -243,12 +243,12 @@ - if (uv_setup_intr(cpu, expires)) { - *t = ULLONG_MAX; - uv_rtc_find_next_timer(head, pnode); -- spin_unlock_irqrestore(&head->lock, flags); -+ raw_spin_unlock_irqrestore(&head->lock, flags); - return -ETIME; - } - } - -- spin_unlock_irqrestore(&head->lock, flags); -+ raw_spin_unlock_irqrestore(&head->lock, flags); - return 0; - } - -@@ -267,7 +267,7 @@ - unsigned long flags; - int rc = 0; - -- spin_lock_irqsave(&head->lock, flags); -+ raw_spin_lock_irqsave(&head->lock, flags); - - if ((head->next_cpu == bcpu && uv_read_rtc(NULL) >= *t) || force) - rc = 1; -@@ -279,7 +279,7 @@ - uv_rtc_find_next_timer(head, pnode); - } - -- spin_unlock_irqrestore(&head->lock, flags); -+ raw_spin_unlock_irqrestore(&head->lock, flags); - - return rc; - } -@@ -299,13 +299,18 @@ - static cycle_t uv_read_rtc(struct clocksource *cs) - { - unsigned long offset; -+ cycle_t cycles; - -+ preempt_disable(); - if (uv_get_min_hub_revision_id() == 1) - offset = 0; - else - offset = (uv_blade_processor_id() * L1_CACHE_BYTES) % PAGE_SIZE; - -- return (cycle_t)uv_read_local_mmr(UVH_RTC | offset); -+ cycles = (cycle_t)uv_read_local_mmr(UVH_RTC | offset); -+ preempt_enable(); -+ -+ return cycles; - } - - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/block/blk-core.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/block/blk-core.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/block/blk-core.c 2017-04-16 10:37:38.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/block/blk-core.c 2017-04-18 17:54:21.000000000 +0200 -@@ -125,6 +125,9 @@ - - INIT_LIST_HEAD(&rq->queuelist); - INIT_LIST_HEAD(&rq->timeout_list); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ INIT_WORK(&rq->work, __blk_mq_complete_request_remote_work); -+#endif - rq->cpu = -1; - rq->q = q; - rq->__sector = (sector_t) -1; -@@ -233,7 +236,7 @@ - **/ - void blk_start_queue(struct request_queue *q) - { -- WARN_ON(!irqs_disabled()); -+ WARN_ON_NONRT(!irqs_disabled()); - - queue_flag_clear(QUEUE_FLAG_STOPPED, q); - __blk_run_queue(q); -@@ -659,7 +662,7 @@ - if (nowait) - return -EBUSY; - -- ret = wait_event_interruptible(q->mq_freeze_wq, -+ ret = swait_event_interruptible(q->mq_freeze_wq, - !atomic_read(&q->mq_freeze_depth) || - blk_queue_dying(q)); - if (blk_queue_dying(q)) -@@ -679,7 +682,7 @@ - struct request_queue *q = - container_of(ref, struct request_queue, q_usage_counter); - -- wake_up_all(&q->mq_freeze_wq); -+ swake_up_all(&q->mq_freeze_wq); - } - - static void blk_rq_timed_out_timer(unsigned long data) -@@ -748,7 +751,7 @@ - q->bypass_depth = 1; - __set_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); - -- init_waitqueue_head(&q->mq_freeze_wq); -+ init_swait_queue_head(&q->mq_freeze_wq); - - /* - * Init percpu_ref in atomic mode so that it's faster to shutdown. -@@ -3200,7 +3203,7 @@ - blk_run_queue_async(q); - else - __blk_run_queue(q); -- spin_unlock(q->queue_lock); -+ spin_unlock_irq(q->queue_lock); - } - - static void flush_plug_callbacks(struct blk_plug *plug, bool from_schedule) -@@ -3248,7 +3251,6 @@ - void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) - { - struct request_queue *q; -- unsigned long flags; - struct request *rq; - LIST_HEAD(list); - unsigned int depth; -@@ -3268,11 +3270,6 @@ - q = NULL; - depth = 0; - -- /* -- * Save and disable interrupts here, to avoid doing it for every -- * queue lock we have to take. -- */ -- local_irq_save(flags); - while (!list_empty(&list)) { - rq = list_entry_rq(list.next); - list_del_init(&rq->queuelist); -@@ -3285,7 +3282,7 @@ - queue_unplugged(q, depth, from_schedule); - q = rq->q; - depth = 0; -- spin_lock(q->queue_lock); -+ spin_lock_irq(q->queue_lock); - } - - /* -@@ -3312,8 +3309,6 @@ - */ - if (q) - queue_unplugged(q, depth, from_schedule); -- -- local_irq_restore(flags); - } - - void blk_finish_plug(struct blk_plug *plug) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/block/blk-ioc.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/block/blk-ioc.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/block/blk-ioc.c 2017-04-16 10:37:38.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/block/blk-ioc.c 2017-04-18 17:54:21.000000000 +0200 -@@ -7,6 +7,7 @@ - #include - #include - #include -+#include - - #include "blk.h" - -@@ -109,7 +110,7 @@ - spin_unlock(q->queue_lock); - } else { - spin_unlock_irqrestore(&ioc->lock, flags); -- cpu_relax(); -+ cpu_chill(); - spin_lock_irqsave_nested(&ioc->lock, flags, 1); - } - } -@@ -187,7 +188,7 @@ - spin_unlock(icq->q->queue_lock); - } else { - spin_unlock_irqrestore(&ioc->lock, flags); -- cpu_relax(); -+ cpu_chill(); - goto retry; - } - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/block/blk-mq.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/block/blk-mq.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/block/blk-mq.c 2017-04-16 10:37:38.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/block/blk-mq.c 2017-04-18 17:54:21.000000000 +0200 -@@ -72,7 +72,7 @@ - - static void blk_mq_freeze_queue_wait(struct request_queue *q) - { -- wait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->q_usage_counter)); -+ swait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->q_usage_counter)); - } - - /* -@@ -110,7 +110,7 @@ - WARN_ON_ONCE(freeze_depth < 0); - if (!freeze_depth) { - percpu_ref_reinit(&q->q_usage_counter); -- wake_up_all(&q->mq_freeze_wq); -+ swake_up_all(&q->mq_freeze_wq); - } - } - EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue); -@@ -129,7 +129,7 @@ - * dying, we need to ensure that processes currently waiting on - * the queue are notified as well. - */ -- wake_up_all(&q->mq_freeze_wq); -+ swake_up_all(&q->mq_freeze_wq); - } - - bool blk_mq_can_queue(struct blk_mq_hw_ctx *hctx) -@@ -177,6 +177,9 @@ - rq->resid_len = 0; - rq->sense = NULL; - -+#ifdef CONFIG_PREEMPT_RT_FULL -+ INIT_WORK(&rq->work, __blk_mq_complete_request_remote_work); -+#endif - INIT_LIST_HEAD(&rq->timeout_list); - rq->timeout = 0; - -@@ -345,6 +348,17 @@ - } - EXPORT_SYMBOL(blk_mq_end_request); - -+#ifdef CONFIG_PREEMPT_RT_FULL -+ -+void __blk_mq_complete_request_remote_work(struct work_struct *work) -+{ -+ struct request *rq = container_of(work, struct request, work); -+ -+ rq->q->softirq_done_fn(rq); -+} -+ -+#else -+ - static void __blk_mq_complete_request_remote(void *data) - { - struct request *rq = data; -@@ -352,6 +366,8 @@ - rq->q->softirq_done_fn(rq); - } - -+#endif -+ - static void blk_mq_ipi_complete_request(struct request *rq) - { - struct blk_mq_ctx *ctx = rq->mq_ctx; -@@ -363,19 +379,23 @@ - return; - } - -- cpu = get_cpu(); -+ cpu = get_cpu_light(); - if (!test_bit(QUEUE_FLAG_SAME_FORCE, &rq->q->queue_flags)) - shared = cpus_share_cache(cpu, ctx->cpu); - - if (cpu != ctx->cpu && !shared && cpu_online(ctx->cpu)) { -+#ifdef CONFIG_PREEMPT_RT_FULL -+ schedule_work_on(ctx->cpu, &rq->work); -+#else - rq->csd.func = __blk_mq_complete_request_remote; - rq->csd.info = rq; - rq->csd.flags = 0; - smp_call_function_single_async(ctx->cpu, &rq->csd); -+#endif - } else { - rq->q->softirq_done_fn(rq); - } -- put_cpu(); -+ put_cpu_light(); - } - - static void __blk_mq_complete_request(struct request *rq) -@@ -906,14 +926,14 @@ - return; - - if (!async && !(hctx->flags & BLK_MQ_F_BLOCKING)) { -- int cpu = get_cpu(); -+ int cpu = get_cpu_light(); - if (cpumask_test_cpu(cpu, hctx->cpumask)) { - __blk_mq_run_hw_queue(hctx); -- put_cpu(); -+ put_cpu_light(); - return; - } - -- put_cpu(); -+ put_cpu_light(); - } - - kblockd_schedule_work_on(blk_mq_hctx_next_cpu(hctx), &hctx->run_work); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/block/blk-mq.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/block/blk-mq.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/block/blk-mq.h 2017-04-16 10:37:38.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/block/blk-mq.h 2017-04-18 17:54:21.000000000 +0200 -@@ -72,12 +72,12 @@ - */ - static inline struct blk_mq_ctx *blk_mq_get_ctx(struct request_queue *q) - { -- return __blk_mq_get_ctx(q, get_cpu()); -+ return __blk_mq_get_ctx(q, get_cpu_light()); - } - - static inline void blk_mq_put_ctx(struct blk_mq_ctx *ctx) - { -- put_cpu(); -+ put_cpu_light(); - } - - struct blk_mq_alloc_data { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/block/blk-softirq.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/block/blk-softirq.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/block/blk-softirq.c 2017-04-16 10:37:38.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/block/blk-softirq.c 2017-04-18 17:54:21.000000000 +0200 -@@ -51,6 +51,7 @@ - raise_softirq_irqoff(BLOCK_SOFTIRQ); - - local_irq_restore(flags); -+ preempt_check_resched_rt(); - } - - /* -@@ -89,6 +90,7 @@ - this_cpu_ptr(&blk_cpu_done)); - raise_softirq_irqoff(BLOCK_SOFTIRQ); - local_irq_enable(); -+ preempt_check_resched_rt(); - - return 0; - } -@@ -141,6 +143,7 @@ - goto do_local; - - local_irq_restore(flags); -+ preempt_check_resched_rt(); - } - - /** -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/block/bounce.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/block/bounce.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/block/bounce.c 2017-04-16 10:37:38.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/block/bounce.c 2017-04-18 17:54:21.000000000 +0200 -@@ -55,11 +55,11 @@ - unsigned long flags; - unsigned char *vto; - -- local_irq_save(flags); -+ local_irq_save_nort(flags); - vto = kmap_atomic(to->bv_page); - memcpy(vto + to->bv_offset, vfrom, to->bv_len); - kunmap_atomic(vto); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - } - - #else /* CONFIG_HIGHMEM */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/crypto/algapi.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/crypto/algapi.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/crypto/algapi.c 2017-04-16 10:37:38.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/crypto/algapi.c 2017-04-18 17:54:21.000000000 +0200 -@@ -719,13 +719,13 @@ - - int crypto_register_notifier(struct notifier_block *nb) - { -- return blocking_notifier_chain_register(&crypto_chain, nb); -+ return srcu_notifier_chain_register(&crypto_chain, nb); - } - EXPORT_SYMBOL_GPL(crypto_register_notifier); - - int crypto_unregister_notifier(struct notifier_block *nb) - { -- return blocking_notifier_chain_unregister(&crypto_chain, nb); -+ return srcu_notifier_chain_unregister(&crypto_chain, nb); - } - EXPORT_SYMBOL_GPL(crypto_unregister_notifier); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/crypto/api.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/crypto/api.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/crypto/api.c 2017-04-16 10:37:38.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/crypto/api.c 2017-04-18 17:54:21.000000000 +0200 -@@ -31,7 +31,7 @@ - DECLARE_RWSEM(crypto_alg_sem); - EXPORT_SYMBOL_GPL(crypto_alg_sem); - --BLOCKING_NOTIFIER_HEAD(crypto_chain); -+SRCU_NOTIFIER_HEAD(crypto_chain); - EXPORT_SYMBOL_GPL(crypto_chain); - - static struct crypto_alg *crypto_larval_wait(struct crypto_alg *alg); -@@ -236,10 +236,10 @@ - { - int ok; - -- ok = blocking_notifier_call_chain(&crypto_chain, val, v); -+ ok = srcu_notifier_call_chain(&crypto_chain, val, v); - if (ok == NOTIFY_DONE) { - request_module("cryptomgr"); -- ok = blocking_notifier_call_chain(&crypto_chain, val, v); -+ ok = srcu_notifier_call_chain(&crypto_chain, val, v); - } - - return ok; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/crypto/internal.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/crypto/internal.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/crypto/internal.h 2017-04-16 10:37:38.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/crypto/internal.h 2017-04-18 17:54:21.000000000 +0200 -@@ -47,7 +47,7 @@ - - extern struct list_head crypto_alg_list; - extern struct rw_semaphore crypto_alg_sem; --extern struct blocking_notifier_head crypto_chain; -+extern struct srcu_notifier_head crypto_chain; - - #ifdef CONFIG_PROC_FS - void __init crypto_init_proc(void); -@@ -146,7 +146,7 @@ - - static inline void crypto_notify(unsigned long val, void *v) - { -- blocking_notifier_call_chain(&crypto_chain, val, v); -+ srcu_notifier_call_chain(&crypto_chain, val, v); - } - - #endif /* _CRYPTO_INTERNAL_H */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/acpi/acpica/acglobal.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/acpi/acpica/acglobal.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/acpi/acpica/acglobal.h 2017-04-16 10:37:38.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/acpi/acpica/acglobal.h 2017-04-18 17:54:22.000000000 +0200 -@@ -116,7 +116,7 @@ - * interrupt level - */ - ACPI_GLOBAL(acpi_spinlock, acpi_gbl_gpe_lock); /* For GPE data structs and registers */ --ACPI_GLOBAL(acpi_spinlock, acpi_gbl_hardware_lock); /* For ACPI H/W except GPE registers */ -+ACPI_GLOBAL(acpi_raw_spinlock, acpi_gbl_hardware_lock); /* For ACPI H/W except GPE registers */ - ACPI_GLOBAL(acpi_spinlock, acpi_gbl_reference_count_lock); - - /* Mutex for _OSI support */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/acpi/acpica/hwregs.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/acpi/acpica/hwregs.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/acpi/acpica/hwregs.c 2017-04-16 10:37:38.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/acpi/acpica/hwregs.c 2017-04-18 17:54:22.000000000 +0200 -@@ -363,14 +363,14 @@ - ACPI_BITMASK_ALL_FIXED_STATUS, - ACPI_FORMAT_UINT64(acpi_gbl_xpm1a_status.address))); - -- lock_flags = acpi_os_acquire_lock(acpi_gbl_hardware_lock); -+ raw_spin_lock_irqsave(acpi_gbl_hardware_lock, lock_flags); - - /* Clear the fixed events in PM1 A/B */ - - status = acpi_hw_register_write(ACPI_REGISTER_PM1_STATUS, - ACPI_BITMASK_ALL_FIXED_STATUS); - -- acpi_os_release_lock(acpi_gbl_hardware_lock, lock_flags); -+ raw_spin_unlock_irqrestore(acpi_gbl_hardware_lock, lock_flags); - - if (ACPI_FAILURE(status)) { - goto exit; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/acpi/acpica/hwxface.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/acpi/acpica/hwxface.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/acpi/acpica/hwxface.c 2017-04-16 10:37:38.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/acpi/acpica/hwxface.c 2017-04-18 17:54:22.000000000 +0200 -@@ -373,7 +373,7 @@ - return_ACPI_STATUS(AE_BAD_PARAMETER); - } - -- lock_flags = acpi_os_acquire_lock(acpi_gbl_hardware_lock); -+ raw_spin_lock_irqsave(acpi_gbl_hardware_lock, lock_flags); - - /* - * At this point, we know that the parent register is one of the -@@ -434,7 +434,7 @@ - - unlock_and_exit: - -- acpi_os_release_lock(acpi_gbl_hardware_lock, lock_flags); -+ raw_spin_unlock_irqrestore(acpi_gbl_hardware_lock, lock_flags); - return_ACPI_STATUS(status); - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/acpi/acpica/utmutex.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/acpi/acpica/utmutex.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/acpi/acpica/utmutex.c 2017-04-16 10:37:39.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/acpi/acpica/utmutex.c 2017-04-18 17:54:22.000000000 +0200 -@@ -88,7 +88,7 @@ - return_ACPI_STATUS (status); - } - -- status = acpi_os_create_lock (&acpi_gbl_hardware_lock); -+ status = acpi_os_create_raw_lock (&acpi_gbl_hardware_lock); - if (ACPI_FAILURE (status)) { - return_ACPI_STATUS (status); - } -@@ -145,7 +145,7 @@ - /* Delete the spinlocks */ - - acpi_os_delete_lock(acpi_gbl_gpe_lock); -- acpi_os_delete_lock(acpi_gbl_hardware_lock); -+ acpi_os_delete_raw_lock(acpi_gbl_hardware_lock); - acpi_os_delete_lock(acpi_gbl_reference_count_lock); - - /* Delete the reader/writer lock */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ata/libata-sff.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ata/libata-sff.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ata/libata-sff.c 2017-04-16 10:37:39.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ata/libata-sff.c 2017-04-18 17:54:22.000000000 +0200 -@@ -678,9 +678,9 @@ - unsigned long flags; - unsigned int consumed; - -- local_irq_save(flags); -+ local_irq_save_nort(flags); - consumed = ata_sff_data_xfer32(dev, buf, buflen, rw); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - - return consumed; - } -@@ -719,7 +719,7 @@ - unsigned long flags; - - /* FIXME: use a bounce buffer */ -- local_irq_save(flags); -+ local_irq_save_nort(flags); - buf = kmap_atomic(page); - - /* do the actual data transfer */ -@@ -727,7 +727,7 @@ - do_write); - - kunmap_atomic(buf); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - } else { - buf = page_address(page); - ap->ops->sff_data_xfer(qc->dev, buf + offset, qc->sect_size, -@@ -864,7 +864,7 @@ - unsigned long flags; - - /* FIXME: use bounce buffer */ -- local_irq_save(flags); -+ local_irq_save_nort(flags); - buf = kmap_atomic(page); - - /* do the actual data transfer */ -@@ -872,7 +872,7 @@ - count, rw); - - kunmap_atomic(buf); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - } else { - buf = page_address(page); - consumed = ap->ops->sff_data_xfer(dev, buf + offset, -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/block/zram/zcomp.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/block/zram/zcomp.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/block/zram/zcomp.c 2017-04-16 10:37:39.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/block/zram/zcomp.c 2017-04-18 17:54:22.000000000 +0200 -@@ -118,12 +118,19 @@ - - struct zcomp_strm *zcomp_stream_get(struct zcomp *comp) - { -- return *get_cpu_ptr(comp->stream); -+ struct zcomp_strm *zstrm; -+ -+ zstrm = *this_cpu_ptr(comp->stream); -+ spin_lock(&zstrm->zcomp_lock); -+ return zstrm; - } - - void zcomp_stream_put(struct zcomp *comp) - { -- put_cpu_ptr(comp->stream); -+ struct zcomp_strm *zstrm; -+ -+ zstrm = *this_cpu_ptr(comp->stream); -+ spin_unlock(&zstrm->zcomp_lock); - } - - int zcomp_compress(struct zcomp_strm *zstrm, -@@ -174,6 +181,7 @@ - pr_err("Can't allocate a compression stream\n"); - return NOTIFY_BAD; - } -+ spin_lock_init(&zstrm->zcomp_lock); - *per_cpu_ptr(comp->stream, cpu) = zstrm; - break; - case CPU_DEAD: -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/block/zram/zcomp.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/block/zram/zcomp.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/block/zram/zcomp.h 2017-04-16 10:37:39.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/block/zram/zcomp.h 2017-04-18 17:54:22.000000000 +0200 -@@ -14,6 +14,7 @@ - /* compression/decompression buffer */ - void *buffer; - struct crypto_comp *tfm; -+ spinlock_t zcomp_lock; - }; - - /* dynamic per-device compression frontend */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/block/zram/zram_drv.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/block/zram/zram_drv.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/block/zram/zram_drv.c 2017-04-16 10:37:39.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/block/zram/zram_drv.c 2017-04-18 17:54:22.000000000 +0200 -@@ -528,6 +528,8 @@ - goto out_error; - } - -+ zram_meta_init_table_locks(meta, disksize); -+ - return meta; - - out_error: -@@ -575,28 +577,28 @@ - struct zram_meta *meta = zram->meta; - unsigned long handle; - unsigned int size; -+ struct zcomp_strm *zstrm; - -- bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value); -+ zram_lock_table(&meta->table[index]); - handle = meta->table[index].handle; - size = zram_get_obj_size(meta, index); - - if (!handle || zram_test_flag(meta, index, ZRAM_ZERO)) { -- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); -+ zram_unlock_table(&meta->table[index]); - clear_page(mem); - return 0; - } - -+ zstrm = zcomp_stream_get(zram->comp); - cmem = zs_map_object(meta->mem_pool, handle, ZS_MM_RO); - if (size == PAGE_SIZE) { - copy_page(mem, cmem); - } else { -- struct zcomp_strm *zstrm = zcomp_stream_get(zram->comp); -- - ret = zcomp_decompress(zstrm, cmem, size, mem); -- zcomp_stream_put(zram->comp); - } - zs_unmap_object(meta->mem_pool, handle); -- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); -+ zcomp_stream_put(zram->comp); -+ zram_unlock_table(&meta->table[index]); - - /* Should NEVER happen. Return bio error if it does. */ - if (unlikely(ret)) { -@@ -616,14 +618,14 @@ - struct zram_meta *meta = zram->meta; - page = bvec->bv_page; - -- bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value); -+ zram_lock_table(&meta->table[index]); - if (unlikely(!meta->table[index].handle) || - zram_test_flag(meta, index, ZRAM_ZERO)) { -- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); -+ zram_unlock_table(&meta->table[index]); - handle_zero_page(bvec); - return 0; - } -- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); -+ zram_unlock_table(&meta->table[index]); - - if (is_partial_io(bvec)) - /* Use a temporary buffer to decompress the page */ -@@ -700,10 +702,10 @@ - if (user_mem) - kunmap_atomic(user_mem); - /* Free memory associated with this sector now. */ -- bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value); -+ zram_lock_table(&meta->table[index]); - zram_free_page(zram, index); - zram_set_flag(meta, index, ZRAM_ZERO); -- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); -+ zram_unlock_table(&meta->table[index]); - - atomic64_inc(&zram->stats.zero_pages); - ret = 0; -@@ -794,12 +796,12 @@ - * Free memory associated with this sector - * before overwriting unused sectors. - */ -- bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value); -+ zram_lock_table(&meta->table[index]); - zram_free_page(zram, index); - - meta->table[index].handle = handle; - zram_set_obj_size(meta, index, clen); -- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); -+ zram_unlock_table(&meta->table[index]); - - /* Update stats */ - atomic64_add(clen, &zram->stats.compr_data_size); -@@ -842,9 +844,9 @@ - } - - while (n >= PAGE_SIZE) { -- bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value); -+ zram_lock_table(&meta->table[index]); - zram_free_page(zram, index); -- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); -+ zram_unlock_table(&meta->table[index]); - atomic64_inc(&zram->stats.notify_free); - index++; - n -= PAGE_SIZE; -@@ -973,9 +975,9 @@ - zram = bdev->bd_disk->private_data; - meta = zram->meta; - -- bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value); -+ zram_lock_table(&meta->table[index]); - zram_free_page(zram, index); -- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); -+ zram_unlock_table(&meta->table[index]); - atomic64_inc(&zram->stats.notify_free); - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/block/zram/zram_drv.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/block/zram/zram_drv.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/block/zram/zram_drv.h 2017-04-16 10:37:39.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/block/zram/zram_drv.h 2017-04-18 17:54:22.000000000 +0200 -@@ -73,6 +73,9 @@ - struct zram_table_entry { - unsigned long handle; - unsigned long value; -+#ifdef CONFIG_PREEMPT_RT_BASE -+ spinlock_t lock; -+#endif - }; - - struct zram_stats { -@@ -120,4 +123,42 @@ - */ - bool claim; /* Protected by bdev->bd_mutex */ - }; -+ -+#ifndef CONFIG_PREEMPT_RT_BASE -+static inline void zram_lock_table(struct zram_table_entry *table) -+{ -+ bit_spin_lock(ZRAM_ACCESS, &table->value); -+} -+ -+static inline void zram_unlock_table(struct zram_table_entry *table) -+{ -+ bit_spin_unlock(ZRAM_ACCESS, &table->value); -+} -+ -+static inline void zram_meta_init_table_locks(struct zram_meta *meta, u64 disksize) { } -+#else /* CONFIG_PREEMPT_RT_BASE */ -+static inline void zram_lock_table(struct zram_table_entry *table) -+{ -+ spin_lock(&table->lock); -+ __set_bit(ZRAM_ACCESS, &table->value); -+} -+ -+static inline void zram_unlock_table(struct zram_table_entry *table) -+{ -+ __clear_bit(ZRAM_ACCESS, &table->value); -+ spin_unlock(&table->lock); -+} -+ -+static inline void zram_meta_init_table_locks(struct zram_meta *meta, u64 disksize) -+{ -+ size_t num_pages = disksize >> PAGE_SHIFT; -+ size_t index; -+ -+ for (index = 0; index < num_pages; index++) { -+ spinlock_t *lock = &meta->table[index].lock; -+ spin_lock_init(lock); -+ } -+} -+#endif /* CONFIG_PREEMPT_RT_BASE */ -+ - #endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/char/random.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/char/random.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/char/random.c 2017-04-16 10:37:39.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/char/random.c 2017-04-18 17:54:22.000000000 +0200 -@@ -1028,8 +1028,6 @@ - } sample; - long delta, delta2, delta3; - -- preempt_disable(); -- - sample.jiffies = jiffies; - sample.cycles = random_get_entropy(); - sample.num = num; -@@ -1070,7 +1068,6 @@ - */ - credit_entropy_bits(r, min_t(int, fls(delta>>1), 11)); - } -- preempt_enable(); - } - - void add_input_randomness(unsigned int type, unsigned int code, -@@ -1123,28 +1120,27 @@ - return *(ptr + f->reg_idx++); - } - --void add_interrupt_randomness(int irq, int irq_flags) -+void add_interrupt_randomness(int irq, int irq_flags, __u64 ip) - { - struct entropy_store *r; - struct fast_pool *fast_pool = this_cpu_ptr(&irq_randomness); -- struct pt_regs *regs = get_irq_regs(); - unsigned long now = jiffies; - cycles_t cycles = random_get_entropy(); - __u32 c_high, j_high; -- __u64 ip; - unsigned long seed; - int credit = 0; - - if (cycles == 0) -- cycles = get_reg(fast_pool, regs); -+ cycles = get_reg(fast_pool, NULL); - c_high = (sizeof(cycles) > 4) ? cycles >> 32 : 0; - j_high = (sizeof(now) > 4) ? now >> 32 : 0; - fast_pool->pool[0] ^= cycles ^ j_high ^ irq; - fast_pool->pool[1] ^= now ^ c_high; -- ip = regs ? instruction_pointer(regs) : _RET_IP_; -+ if (!ip) -+ ip = _RET_IP_; - fast_pool->pool[2] ^= ip; - fast_pool->pool[3] ^= (sizeof(ip) > 4) ? ip >> 32 : -- get_reg(fast_pool, regs); -+ get_reg(fast_pool, NULL); - - fast_mix(fast_pool); - add_interrupt_bench(cycles); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/clocksource/tcb_clksrc.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/clocksource/tcb_clksrc.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/clocksource/tcb_clksrc.c 2017-04-16 10:37:40.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/clocksource/tcb_clksrc.c 2017-04-18 17:54:22.000000000 +0200 -@@ -23,8 +23,7 @@ - * this 32 bit free-running counter. the second channel is not used. - * - * - The third channel may be used to provide a 16-bit clockevent -- * source, used in either periodic or oneshot mode. This runs -- * at 32 KiHZ, and can handle delays of up to two seconds. -+ * source, used in either periodic or oneshot mode. - * - * A boot clocksource and clockevent source are also currently needed, - * unless the relevant platforms (ARM/AT91, AVR32/AT32) are changed so -@@ -74,6 +73,8 @@ - struct tc_clkevt_device { - struct clock_event_device clkevt; - struct clk *clk; -+ bool clk_enabled; -+ u32 freq; - void __iomem *regs; - }; - -@@ -82,15 +83,26 @@ - return container_of(clkevt, struct tc_clkevt_device, clkevt); - } - --/* For now, we always use the 32K clock ... this optimizes for NO_HZ, -- * because using one of the divided clocks would usually mean the -- * tick rate can never be less than several dozen Hz (vs 0.5 Hz). -- * -- * A divided clock could be good for high resolution timers, since -- * 30.5 usec resolution can seem "low". -- */ - static u32 timer_clock; - -+static void tc_clk_disable(struct clock_event_device *d) -+{ -+ struct tc_clkevt_device *tcd = to_tc_clkevt(d); -+ -+ clk_disable(tcd->clk); -+ tcd->clk_enabled = false; -+} -+ -+static void tc_clk_enable(struct clock_event_device *d) -+{ -+ struct tc_clkevt_device *tcd = to_tc_clkevt(d); -+ -+ if (tcd->clk_enabled) -+ return; -+ clk_enable(tcd->clk); -+ tcd->clk_enabled = true; -+} -+ - static int tc_shutdown(struct clock_event_device *d) - { - struct tc_clkevt_device *tcd = to_tc_clkevt(d); -@@ -98,8 +110,14 @@ - - __raw_writel(0xff, regs + ATMEL_TC_REG(2, IDR)); - __raw_writel(ATMEL_TC_CLKDIS, regs + ATMEL_TC_REG(2, CCR)); -+ return 0; -+} -+ -+static int tc_shutdown_clk_off(struct clock_event_device *d) -+{ -+ tc_shutdown(d); - if (!clockevent_state_detached(d)) -- clk_disable(tcd->clk); -+ tc_clk_disable(d); - - return 0; - } -@@ -112,9 +130,9 @@ - if (clockevent_state_oneshot(d) || clockevent_state_periodic(d)) - tc_shutdown(d); - -- clk_enable(tcd->clk); -+ tc_clk_enable(d); - -- /* slow clock, count up to RC, then irq and stop */ -+ /* count up to RC, then irq and stop */ - __raw_writel(timer_clock | ATMEL_TC_CPCSTOP | ATMEL_TC_WAVE | - ATMEL_TC_WAVESEL_UP_AUTO, regs + ATMEL_TC_REG(2, CMR)); - __raw_writel(ATMEL_TC_CPCS, regs + ATMEL_TC_REG(2, IER)); -@@ -134,12 +152,12 @@ - /* By not making the gentime core emulate periodic mode on top - * of oneshot, we get lower overhead and improved accuracy. - */ -- clk_enable(tcd->clk); -+ tc_clk_enable(d); - -- /* slow clock, count up to RC, then irq and restart */ -+ /* count up to RC, then irq and restart */ - __raw_writel(timer_clock | ATMEL_TC_WAVE | ATMEL_TC_WAVESEL_UP_AUTO, - regs + ATMEL_TC_REG(2, CMR)); -- __raw_writel((32768 + HZ / 2) / HZ, tcaddr + ATMEL_TC_REG(2, RC)); -+ __raw_writel((tcd->freq + HZ / 2) / HZ, tcaddr + ATMEL_TC_REG(2, RC)); - - /* Enable clock and interrupts on RC compare */ - __raw_writel(ATMEL_TC_CPCS, regs + ATMEL_TC_REG(2, IER)); -@@ -166,9 +184,13 @@ - .features = CLOCK_EVT_FEAT_PERIODIC | - CLOCK_EVT_FEAT_ONESHOT, - /* Should be lower than at91rm9200's system timer */ -+#ifdef CONFIG_ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK - .rating = 125, -+#else -+ .rating = 200, -+#endif - .set_next_event = tc_next_event, -- .set_state_shutdown = tc_shutdown, -+ .set_state_shutdown = tc_shutdown_clk_off, - .set_state_periodic = tc_set_periodic, - .set_state_oneshot = tc_set_oneshot, - }, -@@ -188,8 +210,9 @@ - return IRQ_NONE; - } - --static int __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx) -+static int __init setup_clkevents(struct atmel_tc *tc, int divisor_idx) - { -+ unsigned divisor = atmel_tc_divisors[divisor_idx]; - int ret; - struct clk *t2_clk = tc->clk[2]; - int irq = tc->irq[2]; -@@ -210,7 +233,11 @@ - clkevt.regs = tc->regs; - clkevt.clk = t2_clk; - -- timer_clock = clk32k_divisor_idx; -+ timer_clock = divisor_idx; -+ if (!divisor) -+ clkevt.freq = 32768; -+ else -+ clkevt.freq = clk_get_rate(t2_clk) / divisor; - - clkevt.clkevt.cpumask = cpumask_of(0); - -@@ -221,7 +248,7 @@ - return ret; - } - -- clockevents_config_and_register(&clkevt.clkevt, 32768, 1, 0xffff); -+ clockevents_config_and_register(&clkevt.clkevt, clkevt.freq, 1, 0xffff); - - return ret; - } -@@ -358,7 +385,11 @@ - goto err_disable_t1; - - /* channel 2: periodic and oneshot timer support */ -+#ifdef CONFIG_ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK - ret = setup_clkevents(tc, clk32k_divisor_idx); -+#else -+ ret = setup_clkevents(tc, best_divisor_idx); -+#endif - if (ret) - goto err_unregister_clksrc; - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/clocksource/timer-atmel-pit.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/clocksource/timer-atmel-pit.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/clocksource/timer-atmel-pit.c 2017-04-16 10:37:40.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/clocksource/timer-atmel-pit.c 2017-04-18 17:54:22.000000000 +0200 -@@ -46,6 +46,7 @@ - u32 cycle; - u32 cnt; - unsigned int irq; -+ bool irq_requested; - struct clk *mck; - }; - -@@ -96,15 +97,29 @@ - - /* disable irq, leaving the clocksource active */ - pit_write(data->base, AT91_PIT_MR, (data->cycle - 1) | AT91_PIT_PITEN); -+ if (data->irq_requested) { -+ free_irq(data->irq, data); -+ data->irq_requested = false; -+ } - return 0; - } - -+static irqreturn_t at91sam926x_pit_interrupt(int irq, void *dev_id); - /* - * Clockevent device: interrupts every 1/HZ (== pit_cycles * MCK/16) - */ - static int pit_clkevt_set_periodic(struct clock_event_device *dev) - { - struct pit_data *data = clkevt_to_pit_data(dev); -+ int ret; -+ -+ ret = request_irq(data->irq, at91sam926x_pit_interrupt, -+ IRQF_SHARED | IRQF_TIMER | IRQF_IRQPOLL, -+ "at91_tick", data); -+ if (ret) -+ panic(pr_fmt("Unable to setup IRQ\n")); -+ -+ data->irq_requested = true; - - /* update clocksource counter */ - data->cnt += data->cycle * PIT_PICNT(pit_read(data->base, AT91_PIT_PIVR)); -@@ -230,15 +245,6 @@ - return ret; - } - -- /* Set up irq handler */ -- ret = request_irq(data->irq, at91sam926x_pit_interrupt, -- IRQF_SHARED | IRQF_TIMER | IRQF_IRQPOLL, -- "at91_tick", data); -- if (ret) { -- pr_err("Unable to setup IRQ\n"); -- return ret; -- } -- - /* Set up and register clockevents */ - data->clkevt.name = "pit"; - data->clkevt.features = CLOCK_EVT_FEAT_PERIODIC; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/clocksource/timer-atmel-st.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/clocksource/timer-atmel-st.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/clocksource/timer-atmel-st.c 2017-04-16 10:37:40.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/clocksource/timer-atmel-st.c 2017-04-18 17:54:22.000000000 +0200 -@@ -115,18 +115,29 @@ - last_crtr = read_CRTR(); - } - -+static int atmel_st_irq; -+ - static int clkevt32k_shutdown(struct clock_event_device *evt) - { - clkdev32k_disable_and_flush_irq(); - irqmask = 0; - regmap_write(regmap_st, AT91_ST_IER, irqmask); -+ free_irq(atmel_st_irq, regmap_st); - return 0; - } - - static int clkevt32k_set_oneshot(struct clock_event_device *dev) - { -+ int ret; -+ - clkdev32k_disable_and_flush_irq(); - -+ ret = request_irq(atmel_st_irq, at91rm9200_timer_interrupt, -+ IRQF_SHARED | IRQF_TIMER | IRQF_IRQPOLL, -+ "at91_tick", regmap_st); -+ if (ret) -+ panic(pr_fmt("Unable to setup IRQ\n")); -+ - /* - * ALM for oneshot irqs, set by next_event() - * before 32 seconds have passed. -@@ -139,8 +150,16 @@ - - static int clkevt32k_set_periodic(struct clock_event_device *dev) - { -+ int ret; -+ - clkdev32k_disable_and_flush_irq(); - -+ ret = request_irq(atmel_st_irq, at91rm9200_timer_interrupt, -+ IRQF_SHARED | IRQF_TIMER | IRQF_IRQPOLL, -+ "at91_tick", regmap_st); -+ if (ret) -+ panic(pr_fmt("Unable to setup IRQ\n")); -+ - /* PIT for periodic irqs; fixed rate of 1/HZ */ - irqmask = AT91_ST_PITS; - regmap_write(regmap_st, AT91_ST_PIMR, timer_latch); -@@ -198,7 +217,7 @@ - { - struct clk *sclk; - unsigned int sclk_rate, val; -- int irq, ret; -+ int ret; - - regmap_st = syscon_node_to_regmap(node); - if (IS_ERR(regmap_st)) { -@@ -212,21 +231,12 @@ - regmap_read(regmap_st, AT91_ST_SR, &val); - - /* Get the interrupts property */ -- irq = irq_of_parse_and_map(node, 0); -- if (!irq) { -+ atmel_st_irq = irq_of_parse_and_map(node, 0); -+ if (!atmel_st_irq) { - pr_err("Unable to get IRQ from DT\n"); - return -EINVAL; - } - -- /* Make IRQs happen for the system timer */ -- ret = request_irq(irq, at91rm9200_timer_interrupt, -- IRQF_SHARED | IRQF_TIMER | IRQF_IRQPOLL, -- "at91_tick", regmap_st); -- if (ret) { -- pr_err("Unable to setup IRQ\n"); -- return ret; -- } -- - sclk = of_clk_get(node, 0); - if (IS_ERR(sclk)) { - pr_err("Unable to get slow clock\n"); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/connector/cn_proc.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/connector/cn_proc.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/connector/cn_proc.c 2017-04-16 10:37:40.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/connector/cn_proc.c 2017-04-18 17:54:22.000000000 +0200 -@@ -32,6 +32,7 @@ - #include - - #include -+#include - - /* - * Size of a cn_msg followed by a proc_event structure. Since the -@@ -54,10 +55,11 @@ - - /* proc_event_counts is used as the sequence number of the netlink message */ - static DEFINE_PER_CPU(__u32, proc_event_counts) = { 0 }; -+static DEFINE_LOCAL_IRQ_LOCK(send_msg_lock); - - static inline void send_msg(struct cn_msg *msg) - { -- preempt_disable(); -+ local_lock(send_msg_lock); - - msg->seq = __this_cpu_inc_return(proc_event_counts) - 1; - ((struct proc_event *)msg->data)->cpu = smp_processor_id(); -@@ -70,7 +72,7 @@ - */ - cn_netlink_send(msg, 0, CN_IDX_PROC, GFP_NOWAIT); - -- preempt_enable(); -+ local_unlock(send_msg_lock); - } - - void proc_fork_connector(struct task_struct *task) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/cpufreq/Kconfig.x86 linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/cpufreq/Kconfig.x86 ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/cpufreq/Kconfig.x86 2017-04-16 10:37:40.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/cpufreq/Kconfig.x86 2017-04-18 17:54:22.000000000 +0200 -@@ -124,7 +124,7 @@ - - config X86_POWERNOW_K8 - tristate "AMD Opteron/Athlon64 PowerNow!" -- depends on ACPI && ACPI_PROCESSOR && X86_ACPI_CPUFREQ -+ depends on ACPI && ACPI_PROCESSOR && X86_ACPI_CPUFREQ && !PREEMPT_RT_BASE - help - This adds the CPUFreq driver for K8/early Opteron/Athlon64 processors. - Support for K10 and newer processors is now in acpi-cpufreq. -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/i915/i915_gem_execbuffer.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/i915/i915_gem_execbuffer.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/i915/i915_gem_execbuffer.c 2017-04-16 10:37:42.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/i915/i915_gem_execbuffer.c 2017-04-18 17:54:22.000000000 +0200 -@@ -1537,7 +1537,9 @@ - if (ret) - return ret; - -+#ifndef CONFIG_PREEMPT_RT_BASE - trace_i915_gem_ring_dispatch(params->request, params->dispatch_flags); -+#endif - - i915_gem_execbuffer_move_to_active(vmas, params->request); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/i915/i915_gem_shrinker.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/i915/i915_gem_shrinker.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/i915/i915_gem_shrinker.c 2017-04-16 10:37:42.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/i915/i915_gem_shrinker.c 2017-04-18 17:54:22.000000000 +0200 -@@ -40,7 +40,7 @@ - if (!mutex_is_locked(mutex)) - return false; - --#if defined(CONFIG_DEBUG_MUTEXES) || defined(CONFIG_MUTEX_SPIN_ON_OWNER) -+#if (defined(CONFIG_DEBUG_MUTEXES) || defined(CONFIG_MUTEX_SPIN_ON_OWNER)) && !defined(CONFIG_PREEMPT_RT_BASE) - return mutex->owner == task; - #else - /* Since UP may be pre-empted, we cannot assume that we own the lock */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/i915/i915_irq.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/i915/i915_irq.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/i915/i915_irq.c 2017-04-16 10:37:42.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/i915/i915_irq.c 2017-04-18 17:54:22.000000000 +0200 -@@ -812,6 +812,7 @@ - spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); - - /* preempt_disable_rt() should go right here in PREEMPT_RT patchset. */ -+ preempt_disable_rt(); - - /* Get optional system timestamp before query. */ - if (stime) -@@ -863,6 +864,7 @@ - *etime = ktime_get(); - - /* preempt_enable_rt() should go right here in PREEMPT_RT patchset. */ -+ preempt_enable_rt(); - - spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/i915/intel_display.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/i915/intel_display.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/i915/intel_display.c 2017-04-16 10:37:42.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/i915/intel_display.c 2017-04-18 17:54:22.000000000 +0200 -@@ -12141,7 +12141,7 @@ - struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_flip_work *work; - -- WARN_ON(!in_interrupt()); -+ WARN_ON_NONRT(!in_interrupt()); - - if (crtc == NULL) - return; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/i915/intel_sprite.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/i915/intel_sprite.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/i915/intel_sprite.c 2017-04-16 10:37:42.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/i915/intel_sprite.c 2017-04-18 17:54:22.000000000 +0200 -@@ -35,6 +35,7 @@ - #include - #include - #include -+#include - #include "intel_drv.h" - #include "intel_frontbuffer.h" - #include -@@ -65,6 +66,8 @@ - 1000 * adjusted_mode->crtc_htotal); - } - -+static DEFINE_LOCAL_IRQ_LOCK(pipe_update_lock); -+ - /** - * intel_pipe_update_start() - start update of a set of display registers - * @crtc: the crtc of which the registers are going to be updated -@@ -95,7 +98,7 @@ - min = vblank_start - intel_usecs_to_scanlines(adjusted_mode, 100); - max = vblank_start - 1; - -- local_irq_disable(); -+ local_lock_irq(pipe_update_lock); - - if (min <= 0 || max <= 0) - return; -@@ -125,11 +128,11 @@ - break; - } - -- local_irq_enable(); -+ local_unlock_irq(pipe_update_lock); - - timeout = schedule_timeout(timeout); - -- local_irq_disable(); -+ local_lock_irq(pipe_update_lock); - } - - finish_wait(wq, &wait); -@@ -181,7 +184,7 @@ - crtc->base.state->event = NULL; - } - -- local_irq_enable(); -+ local_unlock_irq(pipe_update_lock); - - if (crtc->debug.start_vbl_count && - crtc->debug.start_vbl_count != end_vbl_count) { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/msm/msm_gem_shrinker.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/msm/msm_gem_shrinker.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/msm/msm_gem_shrinker.c 2017-04-16 10:37:42.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/msm/msm_gem_shrinker.c 2017-04-18 17:54:22.000000000 +0200 -@@ -23,7 +23,7 @@ - if (!mutex_is_locked(mutex)) - return false; - --#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_MUTEXES) -+#if (defined(CONFIG_SMP) || defined(CONFIG_DEBUG_MUTEXES)) && !defined(CONFIG_PREEMPT_RT_BASE) - return mutex->owner == task; - #else - /* Since UP may be pre-empted, we cannot assume that we own the lock */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/radeon/radeon_display.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/radeon/radeon_display.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/gpu/drm/radeon/radeon_display.c 2017-04-16 10:37:43.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/gpu/drm/radeon/radeon_display.c 2017-04-18 17:54:23.000000000 +0200 -@@ -1845,6 +1845,7 @@ - struct radeon_device *rdev = dev->dev_private; - - /* preempt_disable_rt() should go right here in PREEMPT_RT patchset. */ -+ preempt_disable_rt(); - - /* Get optional system timestamp before query. */ - if (stime) -@@ -1937,6 +1938,7 @@ - *etime = ktime_get(); - - /* preempt_enable_rt() should go right here in PREEMPT_RT patchset. */ -+ preempt_enable_rt(); - - /* Decode into vertical and horizontal scanout position. */ - *vpos = position & 0x1fff; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/hv/vmbus_drv.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/hv/vmbus_drv.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/hv/vmbus_drv.c 2017-04-16 10:37:44.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/hv/vmbus_drv.c 2017-04-18 17:54:23.000000000 +0200 -@@ -761,6 +761,8 @@ - void *page_addr; - struct hv_message *msg; - union hv_synic_event_flags *event; -+ struct pt_regs *regs = get_irq_regs(); -+ u64 ip = regs ? instruction_pointer(regs) : 0; - bool handled = false; - - page_addr = hv_context.synic_event_page[cpu]; -@@ -808,7 +810,7 @@ - tasklet_schedule(hv_context.msg_dpc[cpu]); - } - -- add_interrupt_randomness(HYPERVISOR_CALLBACK_VECTOR, 0); -+ add_interrupt_randomness(HYPERVISOR_CALLBACK_VECTOR, 0, ip); - } - - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/alim15x3.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/alim15x3.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/alim15x3.c 2017-04-16 10:37:44.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/alim15x3.c 2017-04-18 17:54:23.000000000 +0200 -@@ -234,7 +234,7 @@ - - isa_dev = pci_get_device(PCI_VENDOR_ID_AL, PCI_DEVICE_ID_AL_M1533, NULL); - -- local_irq_save(flags); -+ local_irq_save_nort(flags); - - if (m5229_revision < 0xC2) { - /* -@@ -325,7 +325,7 @@ - } - pci_dev_put(north); - pci_dev_put(isa_dev); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - return 0; - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/hpt366.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/hpt366.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/hpt366.c 2017-04-16 10:37:44.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/hpt366.c 2017-04-18 17:54:23.000000000 +0200 -@@ -1236,7 +1236,7 @@ - - dma_old = inb(base + 2); - -- local_irq_save(flags); -+ local_irq_save_nort(flags); - - dma_new = dma_old; - pci_read_config_byte(dev, hwif->channel ? 0x4b : 0x43, &masterdma); -@@ -1247,7 +1247,7 @@ - if (dma_new != dma_old) - outb(dma_new, base + 2); - -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - - printk(KERN_INFO " %s: BM-DMA at 0x%04lx-0x%04lx\n", - hwif->name, base, base + 7); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/ide-io-std.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/ide-io-std.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/ide-io-std.c 2017-04-16 10:37:44.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/ide-io-std.c 2017-04-18 17:54:23.000000000 +0200 -@@ -175,7 +175,7 @@ - unsigned long uninitialized_var(flags); - - if ((io_32bit & 2) && !mmio) { -- local_irq_save(flags); -+ local_irq_save_nort(flags); - ata_vlb_sync(io_ports->nsect_addr); - } - -@@ -186,7 +186,7 @@ - insl(data_addr, buf, words); - - if ((io_32bit & 2) && !mmio) -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - - if (((len + 1) & 3) < 2) - return; -@@ -219,7 +219,7 @@ - unsigned long uninitialized_var(flags); - - if ((io_32bit & 2) && !mmio) { -- local_irq_save(flags); -+ local_irq_save_nort(flags); - ata_vlb_sync(io_ports->nsect_addr); - } - -@@ -230,7 +230,7 @@ - outsl(data_addr, buf, words); - - if ((io_32bit & 2) && !mmio) -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - - if (((len + 1) & 3) < 2) - return; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/ide-io.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/ide-io.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/ide-io.c 2017-04-16 10:37:44.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/ide-io.c 2017-04-18 17:54:23.000000000 +0200 -@@ -659,7 +659,7 @@ - /* disable_irq_nosync ?? */ - disable_irq(hwif->irq); - /* local CPU only, as if we were handling an interrupt */ -- local_irq_disable(); -+ local_irq_disable_nort(); - if (hwif->polling) { - startstop = handler(drive); - } else if (drive_is_ready(drive)) { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/ide-iops.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/ide-iops.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/ide-iops.c 2017-04-16 10:37:44.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/ide-iops.c 2017-04-18 17:54:23.000000000 +0200 -@@ -129,12 +129,12 @@ - if ((stat & ATA_BUSY) == 0) - break; - -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - *rstat = stat; - return -EBUSY; - } - } -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - } - /* - * Allow status to settle, then read it again. -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/ide-probe.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/ide-probe.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/ide-probe.c 2017-04-16 10:37:44.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/ide-probe.c 2017-04-18 17:54:23.000000000 +0200 -@@ -196,10 +196,10 @@ - int bswap = 1; - - /* local CPU only; some systems need this */ -- local_irq_save(flags); -+ local_irq_save_nort(flags); - /* read 512 bytes of id info */ - hwif->tp_ops->input_data(drive, NULL, id, SECTOR_SIZE); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - - drive->dev_flags |= IDE_DFLAG_ID_READ; - #ifdef DEBUG -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/ide-taskfile.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/ide-taskfile.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/ide/ide-taskfile.c 2017-04-16 10:37:44.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/ide/ide-taskfile.c 2017-04-18 17:54:23.000000000 +0200 -@@ -250,7 +250,7 @@ - - page_is_high = PageHighMem(page); - if (page_is_high) -- local_irq_save(flags); -+ local_irq_save_nort(flags); - - buf = kmap_atomic(page) + offset; - -@@ -271,7 +271,7 @@ - kunmap_atomic(buf); - - if (page_is_high) -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - - len -= nr_bytes; - } -@@ -414,7 +414,7 @@ - } - - if ((drive->dev_flags & IDE_DFLAG_UNMASK) == 0) -- local_irq_disable(); -+ local_irq_disable_nort(); - - ide_set_handler(drive, &task_pio_intr, WAIT_WORSTCASE); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/infiniband/ulp/ipoib/ipoib_multicast.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/infiniband/ulp/ipoib/ipoib_multicast.c 2017-04-16 10:37:45.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c 2017-04-18 17:54:23.000000000 +0200 -@@ -902,7 +902,7 @@ - - ipoib_dbg_mcast(priv, "restarting multicast task\n"); - -- local_irq_save(flags); -+ local_irq_save_nort(flags); - netif_addr_lock(dev); - spin_lock(&priv->lock); - -@@ -984,7 +984,7 @@ - - spin_unlock(&priv->lock); - netif_addr_unlock(dev); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - - /* - * make sure the in-flight joins have finished before we attempt -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/input/gameport/gameport.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/input/gameport/gameport.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/input/gameport/gameport.c 2017-04-16 10:37:45.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/input/gameport/gameport.c 2017-04-18 17:54:23.000000000 +0200 -@@ -91,13 +91,13 @@ - tx = ~0; - - for (i = 0; i < 50; i++) { -- local_irq_save(flags); -+ local_irq_save_nort(flags); - t1 = ktime_get_ns(); - for (t = 0; t < 50; t++) - gameport_read(gameport); - t2 = ktime_get_ns(); - t3 = ktime_get_ns(); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - udelay(i * 10); - t = (t2 - t1) - (t3 - t2); - if (t < tx) -@@ -124,12 +124,12 @@ - tx = 1 << 30; - - for(i = 0; i < 50; i++) { -- local_irq_save(flags); -+ local_irq_save_nort(flags); - GET_TIME(t1); - for (t = 0; t < 50; t++) gameport_read(gameport); - GET_TIME(t2); - GET_TIME(t3); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - udelay(i * 10); - if ((t = DELTA(t2,t1) - DELTA(t3,t2)) < tx) tx = t; - } -@@ -148,11 +148,11 @@ - tx = 1 << 30; - - for(i = 0; i < 50; i++) { -- local_irq_save(flags); -+ local_irq_save_nort(flags); - t1 = rdtsc(); - for (t = 0; t < 50; t++) gameport_read(gameport); - t2 = rdtsc(); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - udelay(i * 10); - if (t2 - t1 < tx) tx = t2 - t1; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/iommu/amd_iommu.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/iommu/amd_iommu.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/iommu/amd_iommu.c 2017-04-16 10:37:46.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/iommu/amd_iommu.c 2017-04-18 17:54:23.000000000 +0200 -@@ -1923,10 +1923,10 @@ - int ret; - - /* -- * Must be called with IRQs disabled. Warn here to detect early -- * when its not. -+ * Must be called with IRQs disabled on a non RT kernel. Warn here to -+ * detect early when its not. - */ -- WARN_ON(!irqs_disabled()); -+ WARN_ON_NONRT(!irqs_disabled()); - - /* lock domain */ - spin_lock(&domain->lock); -@@ -2094,10 +2094,10 @@ - struct protection_domain *domain; - - /* -- * Must be called with IRQs disabled. Warn here to detect early -- * when its not. -+ * Must be called with IRQs disabled on a non RT kernel. Warn here to -+ * detect early when its not. - */ -- WARN_ON(!irqs_disabled()); -+ WARN_ON_NONRT(!irqs_disabled()); - - if (WARN_ON(!dev_data->domain)) - return; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/iommu/intel-iommu.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/iommu/intel-iommu.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/iommu/intel-iommu.c 2017-04-16 10:37:46.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/iommu/intel-iommu.c 2017-04-18 17:54:23.000000000 +0200 -@@ -479,7 +479,7 @@ - struct deferred_flush_table *tables; - }; - --DEFINE_PER_CPU(struct deferred_flush_data, deferred_flush); -+static DEFINE_PER_CPU(struct deferred_flush_data, deferred_flush); - - /* bitmap for indexing intel_iommus */ - static int g_num_of_iommus; -@@ -3716,10 +3716,8 @@ - struct intel_iommu *iommu; - struct deferred_flush_entry *entry; - struct deferred_flush_data *flush_data; -- unsigned int cpuid; - -- cpuid = get_cpu(); -- flush_data = per_cpu_ptr(&deferred_flush, cpuid); -+ flush_data = raw_cpu_ptr(&deferred_flush); - - /* Flush all CPUs' entries to avoid deferring too much. If - * this becomes a bottleneck, can just flush us, and rely on -@@ -3752,8 +3750,6 @@ - } - flush_data->size++; - spin_unlock_irqrestore(&flush_data->lock, flags); -- -- put_cpu(); - } - - static void intel_unmap(struct device *dev, dma_addr_t dev_addr, size_t size) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/iommu/iova.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/iommu/iova.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/iommu/iova.c 2017-04-16 10:37:46.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/iommu/iova.c 2017-04-18 17:54:23.000000000 +0200 -@@ -22,6 +22,7 @@ - #include - #include - #include -+#include - - static bool iova_rcache_insert(struct iova_domain *iovad, - unsigned long pfn, -@@ -420,10 +421,8 @@ - - /* Try replenishing IOVAs by flushing rcache. */ - flushed_rcache = true; -- preempt_disable(); - for_each_online_cpu(cpu) - free_cpu_cached_iovas(cpu, iovad); -- preempt_enable(); - goto retry; - } - -@@ -751,7 +750,7 @@ - bool can_insert = false; - unsigned long flags; - -- cpu_rcache = get_cpu_ptr(rcache->cpu_rcaches); -+ cpu_rcache = raw_cpu_ptr(rcache->cpu_rcaches); - spin_lock_irqsave(&cpu_rcache->lock, flags); - - if (!iova_magazine_full(cpu_rcache->loaded)) { -@@ -781,7 +780,6 @@ - iova_magazine_push(cpu_rcache->loaded, iova_pfn); - - spin_unlock_irqrestore(&cpu_rcache->lock, flags); -- put_cpu_ptr(rcache->cpu_rcaches); - - if (mag_to_free) { - iova_magazine_free_pfns(mag_to_free, iovad); -@@ -815,7 +813,7 @@ - bool has_pfn = false; - unsigned long flags; - -- cpu_rcache = get_cpu_ptr(rcache->cpu_rcaches); -+ cpu_rcache = raw_cpu_ptr(rcache->cpu_rcaches); - spin_lock_irqsave(&cpu_rcache->lock, flags); - - if (!iova_magazine_empty(cpu_rcache->loaded)) { -@@ -837,7 +835,6 @@ - iova_pfn = iova_magazine_pop(cpu_rcache->loaded, limit_pfn); - - spin_unlock_irqrestore(&cpu_rcache->lock, flags); -- put_cpu_ptr(rcache->cpu_rcaches); - - return iova_pfn; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/leds/trigger/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/leds/trigger/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/leds/trigger/Kconfig 2017-04-16 10:37:47.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/leds/trigger/Kconfig 2017-04-18 17:54:23.000000000 +0200 -@@ -69,7 +69,7 @@ - - config LEDS_TRIGGER_CPU - bool "LED CPU Trigger" -- depends on LEDS_TRIGGERS -+ depends on LEDS_TRIGGERS && !PREEMPT_RT_BASE - help - This allows LEDs to be controlled by active CPUs. This shows - the active CPUs across an array of LEDs so you can see which -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/md/bcache/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/md/bcache/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/md/bcache/Kconfig 2017-04-16 10:37:47.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/md/bcache/Kconfig 2017-04-18 17:54:23.000000000 +0200 -@@ -1,6 +1,7 @@ - - config BCACHE - tristate "Block device as cache" -+ depends on !PREEMPT_RT_FULL - ---help--- - Allows a block device to be used as cache for other devices; uses - a btree for indexing and the layout is optimized for SSDs. -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/md/dm-rq.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/md/dm-rq.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/md/dm-rq.c 2017-04-16 10:37:47.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/md/dm-rq.c 2017-04-18 17:54:23.000000000 +0200 -@@ -842,7 +842,7 @@ - /* Establish tio->ti before queuing work (map_tio_request) */ - tio->ti = ti; - kthread_queue_work(&md->kworker, &tio->work); -- BUG_ON(!irqs_disabled()); -+ BUG_ON_NONRT(!irqs_disabled()); - } - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/md/raid5.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/md/raid5.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/md/raid5.c 2017-04-16 10:37:47.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/md/raid5.c 2017-04-18 17:54:23.000000000 +0200 -@@ -1928,8 +1928,9 @@ - struct raid5_percpu *percpu; - unsigned long cpu; - -- cpu = get_cpu(); -+ cpu = get_cpu_light(); - percpu = per_cpu_ptr(conf->percpu, cpu); -+ spin_lock(&percpu->lock); - if (test_bit(STRIPE_OP_BIOFILL, &ops_request)) { - ops_run_biofill(sh); - overlap_clear++; -@@ -1985,7 +1986,8 @@ - if (test_and_clear_bit(R5_Overlap, &dev->flags)) - wake_up(&sh->raid_conf->wait_for_overlap); - } -- put_cpu(); -+ spin_unlock(&percpu->lock); -+ put_cpu_light(); - } - - static struct stripe_head *alloc_stripe(struct kmem_cache *sc, gfp_t gfp, -@@ -6391,6 +6393,7 @@ - __func__, cpu); - return -ENOMEM; - } -+ spin_lock_init(&per_cpu_ptr(conf->percpu, cpu)->lock); - return 0; - } - -@@ -6401,7 +6404,6 @@ - conf->percpu = alloc_percpu(struct raid5_percpu); - if (!conf->percpu) - return -ENOMEM; -- - err = cpuhp_state_add_instance(CPUHP_MD_RAID5_PREPARE, &conf->node); - if (!err) { - conf->scribble_disks = max(conf->raid_disks, -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/md/raid5.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/md/raid5.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/md/raid5.h 2017-04-16 10:37:47.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/md/raid5.h 2017-04-18 17:54:23.000000000 +0200 -@@ -504,6 +504,7 @@ - int recovery_disabled; - /* per cpu variables */ - struct raid5_percpu { -+ spinlock_t lock; /* Protection for -RT */ - struct page *spare_page; /* Used when checking P/Q in raid6 */ - struct flex_array *scribble; /* space for constructing buffer - * lists and performing address -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/misc/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/misc/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/misc/Kconfig 2017-04-16 10:37:53.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/misc/Kconfig 2017-04-18 17:54:23.000000000 +0200 -@@ -62,6 +62,7 @@ - config ATMEL_TCLIB - bool "Atmel AT32/AT91 Timer/Counter Library" - depends on (AVR32 || ARCH_AT91) -+ default y if PREEMPT_RT_FULL - help - Select this if you want a library to allocate the Timer/Counter - blocks found on many Atmel processors. This facilitates using -@@ -77,8 +78,7 @@ - are combined to make a single 32-bit timer. - - When GENERIC_CLOCKEVENTS is defined, the third timer channel -- may be used as a clock event device supporting oneshot mode -- (delays of up to two seconds) based on the 32 KiHz clock. -+ may be used as a clock event device supporting oneshot mode. - - config ATMEL_TCB_CLKSRC_BLOCK - int -@@ -92,6 +92,15 @@ - TC can be used for other purposes, such as PWM generation and - interval timing. - -+config ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK -+ bool "TC Block use 32 KiHz clock" -+ depends on ATMEL_TCB_CLKSRC -+ default y if !PREEMPT_RT_FULL -+ help -+ Select this to use 32 KiHz base clock rate as TC block clock -+ source for clock events. -+ -+ - config DUMMY_IRQ - tristate "Dummy IRQ handler" - default n -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/mmc/host/mmci.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/mmc/host/mmci.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/mmc/host/mmci.c 2017-04-16 10:37:53.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/mmc/host/mmci.c 2017-04-18 17:54:23.000000000 +0200 -@@ -1147,15 +1147,12 @@ - struct sg_mapping_iter *sg_miter = &host->sg_miter; - struct variant_data *variant = host->variant; - void __iomem *base = host->base; -- unsigned long flags; - u32 status; - - status = readl(base + MMCISTATUS); - - dev_dbg(mmc_dev(host->mmc), "irq1 (pio) %08x\n", status); - -- local_irq_save(flags); -- - do { - unsigned int remain, len; - char *buffer; -@@ -1195,8 +1192,6 @@ - - sg_miter_stop(sg_miter); - -- local_irq_restore(flags); -- - /* - * If we have less than the fifo 'half-full' threshold to transfer, - * trigger a PIO interrupt as soon as any data is available. -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/net/ethernet/3com/3c59x.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/net/ethernet/3com/3c59x.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/net/ethernet/3com/3c59x.c 2017-04-16 10:37:55.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/net/ethernet/3com/3c59x.c 2017-04-18 17:54:23.000000000 +0200 -@@ -842,9 +842,9 @@ - { - struct vortex_private *vp = netdev_priv(dev); - unsigned long flags; -- local_irq_save(flags); -+ local_irq_save_nort(flags); - (vp->full_bus_master_rx ? boomerang_interrupt:vortex_interrupt)(dev->irq,dev); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - } - #endif - -@@ -1910,12 +1910,12 @@ - * Block interrupts because vortex_interrupt does a bare spin_lock() - */ - unsigned long flags; -- local_irq_save(flags); -+ local_irq_save_nort(flags); - if (vp->full_bus_master_tx) - boomerang_interrupt(dev->irq, dev); - else - vortex_interrupt(dev->irq, dev); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - } - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/net/ethernet/realtek/8139too.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/net/ethernet/realtek/8139too.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/net/ethernet/realtek/8139too.c 2017-04-16 10:37:58.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/net/ethernet/realtek/8139too.c 2017-04-18 17:54:23.000000000 +0200 -@@ -2233,7 +2233,7 @@ - struct rtl8139_private *tp = netdev_priv(dev); - const int irq = tp->pci_dev->irq; - -- disable_irq(irq); -+ disable_irq_nosync(irq); - rtl8139_interrupt(irq, dev); - enable_irq(irq); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/net/wireless/intersil/orinoco/orinoco_usb.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/net/wireless/intersil/orinoco/orinoco_usb.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/net/wireless/intersil/orinoco/orinoco_usb.c 2017-04-16 10:38:02.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/net/wireless/intersil/orinoco/orinoco_usb.c 2017-04-18 17:54:24.000000000 +0200 -@@ -697,7 +697,7 @@ - while (!ctx->done.done && msecs--) - udelay(1000); - } else { -- wait_event_interruptible(ctx->done.wait, -+ swait_event_interruptible(ctx->done.wait, - ctx->done.done); - } - break; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/pci/access.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/pci/access.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/pci/access.c 2017-04-16 10:38:05.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/pci/access.c 2017-04-18 17:54:24.000000000 +0200 -@@ -672,7 +672,7 @@ - WARN_ON(!dev->block_cfg_access); - - dev->block_cfg_access = 0; -- wake_up_all(&pci_cfg_wait); -+ wake_up_all_locked(&pci_cfg_wait); - raw_spin_unlock_irqrestore(&pci_lock, flags); - } - EXPORT_SYMBOL_GPL(pci_cfg_access_unlock); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/pinctrl/qcom/pinctrl-msm.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/pinctrl/qcom/pinctrl-msm.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/pinctrl/qcom/pinctrl-msm.c 2017-04-16 10:38:05.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/pinctrl/qcom/pinctrl-msm.c 2017-04-18 17:54:24.000000000 +0200 -@@ -61,7 +61,7 @@ - struct notifier_block restart_nb; - int irq; - -- spinlock_t lock; -+ raw_spinlock_t lock; - - DECLARE_BITMAP(dual_edge_irqs, MAX_NR_GPIO); - DECLARE_BITMAP(enabled_irqs, MAX_NR_GPIO); -@@ -153,14 +153,14 @@ - if (WARN_ON(i == g->nfuncs)) - return -EINVAL; - -- spin_lock_irqsave(&pctrl->lock, flags); -+ raw_spin_lock_irqsave(&pctrl->lock, flags); - - val = readl(pctrl->regs + g->ctl_reg); - val &= ~mask; - val |= i << g->mux_bit; - writel(val, pctrl->regs + g->ctl_reg); - -- spin_unlock_irqrestore(&pctrl->lock, flags); -+ raw_spin_unlock_irqrestore(&pctrl->lock, flags); - - return 0; - } -@@ -323,14 +323,14 @@ - break; - case PIN_CONFIG_OUTPUT: - /* set output value */ -- spin_lock_irqsave(&pctrl->lock, flags); -+ raw_spin_lock_irqsave(&pctrl->lock, flags); - val = readl(pctrl->regs + g->io_reg); - if (arg) - val |= BIT(g->out_bit); - else - val &= ~BIT(g->out_bit); - writel(val, pctrl->regs + g->io_reg); -- spin_unlock_irqrestore(&pctrl->lock, flags); -+ raw_spin_unlock_irqrestore(&pctrl->lock, flags); - - /* enable output */ - arg = 1; -@@ -351,12 +351,12 @@ - return -EINVAL; - } - -- spin_lock_irqsave(&pctrl->lock, flags); -+ raw_spin_lock_irqsave(&pctrl->lock, flags); - val = readl(pctrl->regs + g->ctl_reg); - val &= ~(mask << bit); - val |= arg << bit; - writel(val, pctrl->regs + g->ctl_reg); -- spin_unlock_irqrestore(&pctrl->lock, flags); -+ raw_spin_unlock_irqrestore(&pctrl->lock, flags); - } - - return 0; -@@ -384,13 +384,13 @@ - - g = &pctrl->soc->groups[offset]; - -- spin_lock_irqsave(&pctrl->lock, flags); -+ raw_spin_lock_irqsave(&pctrl->lock, flags); - - val = readl(pctrl->regs + g->ctl_reg); - val &= ~BIT(g->oe_bit); - writel(val, pctrl->regs + g->ctl_reg); - -- spin_unlock_irqrestore(&pctrl->lock, flags); -+ raw_spin_unlock_irqrestore(&pctrl->lock, flags); - - return 0; - } -@@ -404,7 +404,7 @@ - - g = &pctrl->soc->groups[offset]; - -- spin_lock_irqsave(&pctrl->lock, flags); -+ raw_spin_lock_irqsave(&pctrl->lock, flags); - - val = readl(pctrl->regs + g->io_reg); - if (value) -@@ -417,7 +417,7 @@ - val |= BIT(g->oe_bit); - writel(val, pctrl->regs + g->ctl_reg); - -- spin_unlock_irqrestore(&pctrl->lock, flags); -+ raw_spin_unlock_irqrestore(&pctrl->lock, flags); - - return 0; - } -@@ -443,7 +443,7 @@ - - g = &pctrl->soc->groups[offset]; - -- spin_lock_irqsave(&pctrl->lock, flags); -+ raw_spin_lock_irqsave(&pctrl->lock, flags); - - val = readl(pctrl->regs + g->io_reg); - if (value) -@@ -452,7 +452,7 @@ - val &= ~BIT(g->out_bit); - writel(val, pctrl->regs + g->io_reg); - -- spin_unlock_irqrestore(&pctrl->lock, flags); -+ raw_spin_unlock_irqrestore(&pctrl->lock, flags); - } - - #ifdef CONFIG_DEBUG_FS -@@ -571,7 +571,7 @@ - - g = &pctrl->soc->groups[d->hwirq]; - -- spin_lock_irqsave(&pctrl->lock, flags); -+ raw_spin_lock_irqsave(&pctrl->lock, flags); - - val = readl(pctrl->regs + g->intr_cfg_reg); - val &= ~BIT(g->intr_enable_bit); -@@ -579,7 +579,7 @@ - - clear_bit(d->hwirq, pctrl->enabled_irqs); - -- spin_unlock_irqrestore(&pctrl->lock, flags); -+ raw_spin_unlock_irqrestore(&pctrl->lock, flags); - } - - static void msm_gpio_irq_unmask(struct irq_data *d) -@@ -592,7 +592,7 @@ - - g = &pctrl->soc->groups[d->hwirq]; - -- spin_lock_irqsave(&pctrl->lock, flags); -+ raw_spin_lock_irqsave(&pctrl->lock, flags); - - val = readl(pctrl->regs + g->intr_cfg_reg); - val |= BIT(g->intr_enable_bit); -@@ -600,7 +600,7 @@ - - set_bit(d->hwirq, pctrl->enabled_irqs); - -- spin_unlock_irqrestore(&pctrl->lock, flags); -+ raw_spin_unlock_irqrestore(&pctrl->lock, flags); - } - - static void msm_gpio_irq_ack(struct irq_data *d) -@@ -613,7 +613,7 @@ - - g = &pctrl->soc->groups[d->hwirq]; - -- spin_lock_irqsave(&pctrl->lock, flags); -+ raw_spin_lock_irqsave(&pctrl->lock, flags); - - val = readl(pctrl->regs + g->intr_status_reg); - if (g->intr_ack_high) -@@ -625,7 +625,7 @@ - if (test_bit(d->hwirq, pctrl->dual_edge_irqs)) - msm_gpio_update_dual_edge_pos(pctrl, g, d); - -- spin_unlock_irqrestore(&pctrl->lock, flags); -+ raw_spin_unlock_irqrestore(&pctrl->lock, flags); - } - - static int msm_gpio_irq_set_type(struct irq_data *d, unsigned int type) -@@ -638,7 +638,7 @@ - - g = &pctrl->soc->groups[d->hwirq]; - -- spin_lock_irqsave(&pctrl->lock, flags); -+ raw_spin_lock_irqsave(&pctrl->lock, flags); - - /* - * For hw without possibility of detecting both edges -@@ -712,7 +712,7 @@ - if (test_bit(d->hwirq, pctrl->dual_edge_irqs)) - msm_gpio_update_dual_edge_pos(pctrl, g, d); - -- spin_unlock_irqrestore(&pctrl->lock, flags); -+ raw_spin_unlock_irqrestore(&pctrl->lock, flags); - - if (type & (IRQ_TYPE_LEVEL_LOW | IRQ_TYPE_LEVEL_HIGH)) - irq_set_handler_locked(d, handle_level_irq); -@@ -728,11 +728,11 @@ - struct msm_pinctrl *pctrl = gpiochip_get_data(gc); - unsigned long flags; - -- spin_lock_irqsave(&pctrl->lock, flags); -+ raw_spin_lock_irqsave(&pctrl->lock, flags); - - irq_set_irq_wake(pctrl->irq, on); - -- spin_unlock_irqrestore(&pctrl->lock, flags); -+ raw_spin_unlock_irqrestore(&pctrl->lock, flags); - - return 0; - } -@@ -878,7 +878,7 @@ - pctrl->soc = soc_data; - pctrl->chip = msm_gpio_template; - -- spin_lock_init(&pctrl->lock); -+ raw_spin_lock_init(&pctrl->lock); - - res = platform_get_resource(pdev, IORESOURCE_MEM, 0); - pctrl->regs = devm_ioremap_resource(&pdev->dev, res); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/scsi/fcoe/fcoe.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/scsi/fcoe/fcoe.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/scsi/fcoe/fcoe.c 2017-04-16 10:38:09.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/scsi/fcoe/fcoe.c 2017-04-18 17:54:24.000000000 +0200 -@@ -1455,11 +1455,11 @@ - static int fcoe_alloc_paged_crc_eof(struct sk_buff *skb, int tlen) - { - struct fcoe_percpu_s *fps; -- int rc; -+ int rc, cpu = get_cpu_light(); - -- fps = &get_cpu_var(fcoe_percpu); -+ fps = &per_cpu(fcoe_percpu, cpu); - rc = fcoe_get_paged_crc_eof(skb, tlen, fps); -- put_cpu_var(fcoe_percpu); -+ put_cpu_light(); - - return rc; - } -@@ -1646,11 +1646,11 @@ - return 0; - } - -- stats = per_cpu_ptr(lport->stats, get_cpu()); -+ stats = per_cpu_ptr(lport->stats, get_cpu_light()); - stats->InvalidCRCCount++; - if (stats->InvalidCRCCount < 5) - printk(KERN_WARNING "fcoe: dropping frame with CRC error\n"); -- put_cpu(); -+ put_cpu_light(); - return -EINVAL; - } - -@@ -1693,7 +1693,7 @@ - */ - hp = (struct fcoe_hdr *) skb_network_header(skb); - -- stats = per_cpu_ptr(lport->stats, get_cpu()); -+ stats = per_cpu_ptr(lport->stats, get_cpu_light()); - if (unlikely(FC_FCOE_DECAPS_VER(hp) != FC_FCOE_VER)) { - if (stats->ErrorFrames < 5) - printk(KERN_WARNING "fcoe: FCoE version " -@@ -1725,13 +1725,13 @@ - goto drop; - - if (!fcoe_filter_frames(lport, fp)) { -- put_cpu(); -+ put_cpu_light(); - fc_exch_recv(lport, fp); - return; - } - drop: - stats->ErrorFrames++; -- put_cpu(); -+ put_cpu_light(); - kfree_skb(skb); - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/scsi/fcoe/fcoe_ctlr.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/scsi/fcoe/fcoe_ctlr.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/scsi/fcoe/fcoe_ctlr.c 2017-04-16 10:38:09.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/scsi/fcoe/fcoe_ctlr.c 2017-04-18 17:54:24.000000000 +0200 -@@ -834,7 +834,7 @@ - - INIT_LIST_HEAD(&del_list); - -- stats = per_cpu_ptr(fip->lp->stats, get_cpu()); -+ stats = per_cpu_ptr(fip->lp->stats, get_cpu_light()); - - list_for_each_entry_safe(fcf, next, &fip->fcfs, list) { - deadline = fcf->time + fcf->fka_period + fcf->fka_period / 2; -@@ -870,7 +870,7 @@ - sel_time = fcf->time; - } - } -- put_cpu(); -+ put_cpu_light(); - - list_for_each_entry_safe(fcf, next, &del_list, list) { - /* Removes fcf from current list */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/scsi/libfc/fc_exch.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/scsi/libfc/fc_exch.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/scsi/libfc/fc_exch.c 2017-04-16 10:38:09.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/scsi/libfc/fc_exch.c 2017-04-18 17:54:24.000000000 +0200 -@@ -814,10 +814,10 @@ - } - memset(ep, 0, sizeof(*ep)); - -- cpu = get_cpu(); -+ cpu = get_cpu_light(); - pool = per_cpu_ptr(mp->pool, cpu); - spin_lock_bh(&pool->lock); -- put_cpu(); -+ put_cpu_light(); - - /* peek cache of free slot */ - if (pool->left != FC_XID_UNKNOWN) { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/scsi/libsas/sas_ata.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/scsi/libsas/sas_ata.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/scsi/libsas/sas_ata.c 2017-04-16 10:38:09.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/scsi/libsas/sas_ata.c 2017-04-18 17:54:24.000000000 +0200 -@@ -190,7 +190,7 @@ - /* TODO: audit callers to ensure they are ready for qc_issue to - * unconditionally re-enable interrupts - */ -- local_irq_save(flags); -+ local_irq_save_nort(flags); - spin_unlock(ap->lock); - - /* If the device fell off, no sense in issuing commands */ -@@ -252,7 +252,7 @@ - - out: - spin_lock(ap->lock); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - return ret; - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/scsi/qla2xxx/qla_inline.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/scsi/qla2xxx/qla_inline.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/scsi/qla2xxx/qla_inline.h 2017-04-16 10:38:10.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/scsi/qla2xxx/qla_inline.h 2017-04-18 17:54:25.000000000 +0200 -@@ -59,12 +59,12 @@ - { - unsigned long flags; - struct qla_hw_data *ha = rsp->hw; -- local_irq_save(flags); -+ local_irq_save_nort(flags); - if (IS_P3P_TYPE(ha)) - qla82xx_poll(0, rsp); - else - ha->isp_ops->intr_handler(0, rsp); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - } - - static inline uint8_t * -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/scsi/qla2xxx/qla_isr.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/scsi/qla2xxx/qla_isr.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/scsi/qla2xxx/qla_isr.c 2017-04-16 10:38:10.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/scsi/qla2xxx/qla_isr.c 2017-04-18 17:54:25.000000000 +0200 -@@ -3125,7 +3125,11 @@ - * kref_put(). - */ - kref_get(&qentry->irq_notify.kref); -+#ifdef CONFIG_PREEMPT_RT_BASE -+ swork_queue(&qentry->irq_notify.swork); -+#else - schedule_work(&qentry->irq_notify.work); -+#endif - } - - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/thermal/x86_pkg_temp_thermal.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/thermal/x86_pkg_temp_thermal.c 2017-04-16 10:38:14.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/thermal/x86_pkg_temp_thermal.c 2017-04-18 17:54:25.000000000 +0200 -@@ -29,6 +29,7 @@ - #include - #include - #include -+#include - #include - #include - -@@ -353,7 +354,7 @@ - } - } - --static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) -+static void platform_thermal_notify_work(struct swork_event *event) - { - unsigned long flags; - int cpu = smp_processor_id(); -@@ -370,7 +371,7 @@ - pkg_work_scheduled[phy_id]) { - disable_pkg_thres_interrupt(); - spin_unlock_irqrestore(&pkg_work_lock, flags); -- return -EINVAL; -+ return; - } - pkg_work_scheduled[phy_id] = 1; - spin_unlock_irqrestore(&pkg_work_lock, flags); -@@ -379,9 +380,48 @@ - schedule_delayed_work_on(cpu, - &per_cpu(pkg_temp_thermal_threshold_work, cpu), - msecs_to_jiffies(notify_delay_ms)); -+} -+ -+#ifdef CONFIG_PREEMPT_RT_FULL -+static struct swork_event notify_work; -+ -+static int thermal_notify_work_init(void) -+{ -+ int err; -+ -+ err = swork_get(); -+ if (err) -+ return err; -+ -+ INIT_SWORK(¬ify_work, platform_thermal_notify_work); - return 0; - } - -+static void thermal_notify_work_cleanup(void) -+{ -+ swork_put(); -+} -+ -+static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) -+{ -+ swork_queue(¬ify_work); -+ return 0; -+} -+ -+#else /* !CONFIG_PREEMPT_RT_FULL */ -+ -+static int thermal_notify_work_init(void) { return 0; } -+ -+static void thermal_notify_work_cleanup(void) { } -+ -+static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) -+{ -+ platform_thermal_notify_work(NULL); -+ -+ return 0; -+} -+#endif /* CONFIG_PREEMPT_RT_FULL */ -+ - static int find_siblings_cpu(int cpu) - { - int i; -@@ -585,6 +625,9 @@ - if (!x86_match_cpu(pkg_temp_thermal_ids)) - return -ENODEV; - -+ if (!thermal_notify_work_init()) -+ return -ENODEV; -+ - spin_lock_init(&pkg_work_lock); - platform_thermal_package_notify = - pkg_temp_thermal_platform_thermal_notify; -@@ -609,7 +652,7 @@ - kfree(pkg_work_scheduled); - platform_thermal_package_notify = NULL; - platform_thermal_package_rate_control = NULL; -- -+ thermal_notify_work_cleanup(); - return -ENODEV; - } - -@@ -634,6 +677,7 @@ - mutex_unlock(&phy_dev_list_mutex); - platform_thermal_package_notify = NULL; - platform_thermal_package_rate_control = NULL; -+ thermal_notify_work_cleanup(); - for_each_online_cpu(i) - cancel_delayed_work_sync( - &per_cpu(pkg_temp_thermal_threshold_work, i)); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/tty/serial/8250/8250_core.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/tty/serial/8250/8250_core.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/tty/serial/8250/8250_core.c 2017-04-16 10:38:14.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/tty/serial/8250/8250_core.c 2017-04-18 17:54:25.000000000 +0200 -@@ -58,7 +58,16 @@ - - static unsigned int skip_txen_test; /* force skip of txen test at init time */ - --#define PASS_LIMIT 512 -+/* -+ * On -rt we can have a more delays, and legitimately -+ * so - so don't drop work spuriously and spam the -+ * syslog: -+ */ -+#ifdef CONFIG_PREEMPT_RT_FULL -+# define PASS_LIMIT 1000000 -+#else -+# define PASS_LIMIT 512 -+#endif - - #include - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/tty/serial/8250/8250_port.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/tty/serial/8250/8250_port.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/tty/serial/8250/8250_port.c 2017-04-16 10:38:14.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/tty/serial/8250/8250_port.c 2017-04-18 17:54:25.000000000 +0200 -@@ -35,6 +35,7 @@ - #include - #include - #include -+#include - #include - #include - #include -@@ -3144,9 +3145,9 @@ - - serial8250_rpm_get(up); - -- if (port->sysrq) -+ if (port->sysrq || oops_in_progress) - locked = 0; -- else if (oops_in_progress) -+ else if (in_kdb_printk()) - locked = spin_trylock_irqsave(&port->lock, flags); - else - spin_lock_irqsave(&port->lock, flags); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/tty/serial/amba-pl011.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/tty/serial/amba-pl011.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/tty/serial/amba-pl011.c 2017-04-16 10:38:14.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/tty/serial/amba-pl011.c 2017-04-18 17:54:25.000000000 +0200 -@@ -2211,13 +2211,19 @@ - - clk_enable(uap->clk); - -- local_irq_save(flags); -+ /* -+ * local_irq_save(flags); -+ * -+ * This local_irq_save() is nonsense. If we come in via sysrq -+ * handling then interrupts are already disabled. Aside of -+ * that the port.sysrq check is racy on SMP regardless. -+ */ - if (uap->port.sysrq) - locked = 0; - else if (oops_in_progress) -- locked = spin_trylock(&uap->port.lock); -+ locked = spin_trylock_irqsave(&uap->port.lock, flags); - else -- spin_lock(&uap->port.lock); -+ spin_lock_irqsave(&uap->port.lock, flags); - - /* - * First save the CR then disable the interrupts -@@ -2241,8 +2247,7 @@ - pl011_write(old_cr, uap, REG_CR); - - if (locked) -- spin_unlock(&uap->port.lock); -- local_irq_restore(flags); -+ spin_unlock_irqrestore(&uap->port.lock, flags); - - clk_disable(uap->clk); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/tty/serial/omap-serial.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/tty/serial/omap-serial.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/tty/serial/omap-serial.c 2017-04-16 10:38:15.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/tty/serial/omap-serial.c 2017-04-18 17:54:25.000000000 +0200 -@@ -1257,13 +1257,10 @@ - - pm_runtime_get_sync(up->dev); - -- local_irq_save(flags); -- if (up->port.sysrq) -- locked = 0; -- else if (oops_in_progress) -- locked = spin_trylock(&up->port.lock); -+ if (up->port.sysrq || oops_in_progress) -+ locked = spin_trylock_irqsave(&up->port.lock, flags); - else -- spin_lock(&up->port.lock); -+ spin_lock_irqsave(&up->port.lock, flags); - - /* - * First save the IER then disable the interrupts -@@ -1292,8 +1289,7 @@ - pm_runtime_mark_last_busy(up->dev); - pm_runtime_put_autosuspend(up->dev); - if (locked) -- spin_unlock(&up->port.lock); -- local_irq_restore(flags); -+ spin_unlock_irqrestore(&up->port.lock, flags); - } - - static int __init -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/usb/core/hcd.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/usb/core/hcd.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/usb/core/hcd.c 2017-04-16 10:38:15.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/usb/core/hcd.c 2017-04-18 17:54:25.000000000 +0200 -@@ -1764,9 +1764,9 @@ - * and no one may trigger the above deadlock situation when - * running complete() in tasklet. - */ -- local_irq_save(flags); -+ local_irq_save_nort(flags); - urb->complete(urb); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - - usb_anchor_resume_wakeups(anchor); - atomic_dec(&urb->use_count); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/usb/gadget/function/f_fs.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/usb/gadget/function/f_fs.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/usb/gadget/function/f_fs.c 2017-04-16 10:38:15.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/usb/gadget/function/f_fs.c 2017-04-18 17:54:25.000000000 +0200 -@@ -1593,7 +1593,7 @@ - pr_info("%s(): freeing\n", __func__); - ffs_data_clear(ffs); - BUG_ON(waitqueue_active(&ffs->ev.waitq) || -- waitqueue_active(&ffs->ep0req_completion.wait)); -+ swait_active(&ffs->ep0req_completion.wait)); - kfree(ffs->dev_name); - kfree(ffs); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/usb/gadget/legacy/inode.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/usb/gadget/legacy/inode.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/drivers/usb/gadget/legacy/inode.c 2017-04-16 10:38:15.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/drivers/usb/gadget/legacy/inode.c 2017-04-18 17:54:25.000000000 +0200 -@@ -346,7 +346,7 @@ - spin_unlock_irq (&epdata->dev->lock); - - if (likely (value == 0)) { -- value = wait_event_interruptible (done.wait, done.done); -+ value = swait_event_interruptible (done.wait, done.done); - if (value != 0) { - spin_lock_irq (&epdata->dev->lock); - if (likely (epdata->ep != NULL)) { -@@ -355,7 +355,7 @@ - usb_ep_dequeue (epdata->ep, epdata->req); - spin_unlock_irq (&epdata->dev->lock); - -- wait_event (done.wait, done.done); -+ swait_event (done.wait, done.done); - if (epdata->status == -ECONNRESET) - epdata->status = -EINTR; - } else { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/aio.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/aio.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/aio.c 2017-04-16 10:38:19.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/aio.c 2017-04-18 17:54:25.000000000 +0200 -@@ -40,6 +40,7 @@ - #include - #include - #include -+#include - - #include - #include -@@ -115,7 +116,7 @@ - struct page **ring_pages; - long nr_pages; - -- struct work_struct free_work; -+ struct swork_event free_work; - - /* - * signals when all in-flight requests are done -@@ -258,6 +259,7 @@ - .mount = aio_mount, - .kill_sb = kill_anon_super, - }; -+ BUG_ON(swork_get()); - aio_mnt = kern_mount(&aio_fs); - if (IS_ERR(aio_mnt)) - panic("Failed to create aio fs mount."); -@@ -581,9 +583,9 @@ - return cancel(&kiocb->common); - } - --static void free_ioctx(struct work_struct *work) -+static void free_ioctx(struct swork_event *sev) - { -- struct kioctx *ctx = container_of(work, struct kioctx, free_work); -+ struct kioctx *ctx = container_of(sev, struct kioctx, free_work); - - pr_debug("freeing %p\n", ctx); - -@@ -602,8 +604,8 @@ - if (ctx->rq_wait && atomic_dec_and_test(&ctx->rq_wait->count)) - complete(&ctx->rq_wait->comp); - -- INIT_WORK(&ctx->free_work, free_ioctx); -- schedule_work(&ctx->free_work); -+ INIT_SWORK(&ctx->free_work, free_ioctx); -+ swork_queue(&ctx->free_work); - } - - /* -@@ -611,9 +613,9 @@ - * and ctx->users has dropped to 0, so we know no more kiocbs can be submitted - - * now it's safe to cancel any that need to be. - */ --static void free_ioctx_users(struct percpu_ref *ref) -+static void free_ioctx_users_work(struct swork_event *sev) - { -- struct kioctx *ctx = container_of(ref, struct kioctx, users); -+ struct kioctx *ctx = container_of(sev, struct kioctx, free_work); - struct aio_kiocb *req; - - spin_lock_irq(&ctx->ctx_lock); -@@ -632,6 +634,14 @@ - percpu_ref_put(&ctx->reqs); - } - -+static void free_ioctx_users(struct percpu_ref *ref) -+{ -+ struct kioctx *ctx = container_of(ref, struct kioctx, users); -+ -+ INIT_SWORK(&ctx->free_work, free_ioctx_users_work); -+ swork_queue(&ctx->free_work); -+} -+ - static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm) - { - unsigned i, new_nr; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/autofs4/autofs_i.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/autofs4/autofs_i.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/autofs4/autofs_i.h 2017-04-16 10:38:19.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/autofs4/autofs_i.h 2017-04-18 17:54:25.000000000 +0200 -@@ -31,6 +31,7 @@ - #include - #include - #include -+#include - #include - #include - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/autofs4/expire.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/autofs4/expire.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/autofs4/expire.c 2017-04-16 10:38:19.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/autofs4/expire.c 2017-04-18 17:54:25.000000000 +0200 -@@ -148,7 +148,7 @@ - parent = p->d_parent; - if (!spin_trylock(&parent->d_lock)) { - spin_unlock(&p->d_lock); -- cpu_relax(); -+ cpu_chill(); - goto relock; - } - spin_unlock(&p->d_lock); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/buffer.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/buffer.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/buffer.c 2017-04-16 10:38:19.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/buffer.c 2017-04-18 17:54:25.000000000 +0200 -@@ -301,8 +301,7 @@ - * decide that the page is now completely done. - */ - first = page_buffers(page); -- local_irq_save(flags); -- bit_spin_lock(BH_Uptodate_Lock, &first->b_state); -+ flags = bh_uptodate_lock_irqsave(first); - clear_buffer_async_read(bh); - unlock_buffer(bh); - tmp = bh; -@@ -315,8 +314,7 @@ - } - tmp = tmp->b_this_page; - } while (tmp != bh); -- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); -- local_irq_restore(flags); -+ bh_uptodate_unlock_irqrestore(first, flags); - - /* - * If none of the buffers had errors and they are all -@@ -328,9 +326,7 @@ - return; - - still_busy: -- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); -- local_irq_restore(flags); -- return; -+ bh_uptodate_unlock_irqrestore(first, flags); - } - - /* -@@ -358,8 +354,7 @@ - } - - first = page_buffers(page); -- local_irq_save(flags); -- bit_spin_lock(BH_Uptodate_Lock, &first->b_state); -+ flags = bh_uptodate_lock_irqsave(first); - - clear_buffer_async_write(bh); - unlock_buffer(bh); -@@ -371,15 +366,12 @@ - } - tmp = tmp->b_this_page; - } -- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); -- local_irq_restore(flags); -+ bh_uptodate_unlock_irqrestore(first, flags); - end_page_writeback(page); - return; - - still_busy: -- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); -- local_irq_restore(flags); -- return; -+ bh_uptodate_unlock_irqrestore(first, flags); - } - EXPORT_SYMBOL(end_buffer_async_write); - -@@ -3383,6 +3375,7 @@ - struct buffer_head *ret = kmem_cache_zalloc(bh_cachep, gfp_flags); - if (ret) { - INIT_LIST_HEAD(&ret->b_assoc_buffers); -+ buffer_head_init_locks(ret); - preempt_disable(); - __this_cpu_inc(bh_accounting.nr); - recalc_bh_state(); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/cifs/readdir.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/cifs/readdir.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/cifs/readdir.c 2017-04-16 10:38:20.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/cifs/readdir.c 2017-04-18 17:54:25.000000000 +0200 -@@ -80,7 +80,7 @@ - struct inode *inode; - struct super_block *sb = parent->d_sb; - struct cifs_sb_info *cifs_sb = CIFS_SB(sb); -- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); -+ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); - - cifs_dbg(FYI, "%s: for %s\n", __func__, name->name); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/dcache.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/dcache.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/dcache.c 2017-04-16 10:38:20.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/dcache.c 2017-04-18 17:54:25.000000000 +0200 -@@ -19,6 +19,7 @@ - #include - #include - #include -+#include - #include - #include - #include -@@ -750,6 +751,8 @@ - */ - void dput(struct dentry *dentry) - { -+ struct dentry *parent; -+ - if (unlikely(!dentry)) - return; - -@@ -788,9 +791,18 @@ - return; - - kill_it: -- dentry = dentry_kill(dentry); -- if (dentry) { -- cond_resched(); -+ parent = dentry_kill(dentry); -+ if (parent) { -+ int r; -+ -+ if (parent == dentry) { -+ /* the task with the highest priority won't schedule */ -+ r = cond_resched(); -+ if (!r) -+ cpu_chill(); -+ } else { -+ dentry = parent; -+ } - goto repeat; - } - } -@@ -2324,7 +2336,7 @@ - if (dentry->d_lockref.count == 1) { - if (!spin_trylock(&inode->i_lock)) { - spin_unlock(&dentry->d_lock); -- cpu_relax(); -+ cpu_chill(); - goto again; - } - dentry->d_flags &= ~DCACHE_CANT_MOUNT; -@@ -2384,21 +2396,24 @@ - - static void d_wait_lookup(struct dentry *dentry) - { -- if (d_in_lookup(dentry)) { -- DECLARE_WAITQUEUE(wait, current); -- add_wait_queue(dentry->d_wait, &wait); -- do { -- set_current_state(TASK_UNINTERRUPTIBLE); -- spin_unlock(&dentry->d_lock); -- schedule(); -- spin_lock(&dentry->d_lock); -- } while (d_in_lookup(dentry)); -- } -+ struct swait_queue __wait; -+ -+ if (!d_in_lookup(dentry)) -+ return; -+ -+ INIT_LIST_HEAD(&__wait.task_list); -+ do { -+ prepare_to_swait(dentry->d_wait, &__wait, TASK_UNINTERRUPTIBLE); -+ spin_unlock(&dentry->d_lock); -+ schedule(); -+ spin_lock(&dentry->d_lock); -+ } while (d_in_lookup(dentry)); -+ finish_swait(dentry->d_wait, &__wait); - } - - struct dentry *d_alloc_parallel(struct dentry *parent, - const struct qstr *name, -- wait_queue_head_t *wq) -+ struct swait_queue_head *wq) - { - unsigned int hash = name->hash; - struct hlist_bl_head *b = in_lookup_hash(parent, hash); -@@ -2507,7 +2522,7 @@ - hlist_bl_lock(b); - dentry->d_flags &= ~DCACHE_PAR_LOOKUP; - __hlist_bl_del(&dentry->d_u.d_in_lookup_hash); -- wake_up_all(dentry->d_wait); -+ swake_up_all(dentry->d_wait); - dentry->d_wait = NULL; - hlist_bl_unlock(b); - INIT_HLIST_NODE(&dentry->d_u.d_alias); -@@ -3604,6 +3619,11 @@ - - void __init vfs_caches_init_early(void) - { -+ int i; -+ -+ for (i = 0; i < ARRAY_SIZE(in_lookup_hashtable); i++) -+ INIT_HLIST_BL_HEAD(&in_lookup_hashtable[i]); -+ - dcache_init_early(); - inode_init_early(); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/eventpoll.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/eventpoll.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/eventpoll.c 2017-04-16 10:38:20.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/eventpoll.c 2017-04-18 17:54:25.000000000 +0200 -@@ -510,12 +510,12 @@ - */ - static void ep_poll_safewake(wait_queue_head_t *wq) - { -- int this_cpu = get_cpu(); -+ int this_cpu = get_cpu_light(); - - ep_call_nested(&poll_safewake_ncalls, EP_MAX_NESTS, - ep_poll_wakeup_proc, NULL, wq, (void *) (long) this_cpu); - -- put_cpu(); -+ put_cpu_light(); - } - - static void ep_remove_wait_queue(struct eppoll_entry *pwq) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/exec.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/exec.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/exec.c 2017-04-16 10:38:20.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/exec.c 2017-04-18 17:54:25.000000000 +0200 -@@ -1017,12 +1017,14 @@ - } - } - task_lock(tsk); -+ preempt_disable_rt(); - active_mm = tsk->active_mm; - tsk->mm = mm; - tsk->active_mm = mm; - activate_mm(active_mm, mm); - tsk->mm->vmacache_seqnum = 0; - vmacache_flush(tsk); -+ preempt_enable_rt(); - task_unlock(tsk); - if (old_mm) { - up_read(&old_mm->mmap_sem); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/fuse/dir.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/fuse/dir.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/fuse/dir.c 2017-04-16 10:38:20.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/fuse/dir.c 2017-04-18 17:54:25.000000000 +0200 -@@ -1191,7 +1191,7 @@ - struct inode *dir = d_inode(parent); - struct fuse_conn *fc; - struct inode *inode; -- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); -+ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); - - if (!o->nodeid) { - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/jbd2/checkpoint.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/jbd2/checkpoint.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/jbd2/checkpoint.c 2017-04-16 10:38:20.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/jbd2/checkpoint.c 2017-04-18 17:54:25.000000000 +0200 -@@ -116,6 +116,8 @@ - nblocks = jbd2_space_needed(journal); - while (jbd2_log_space_left(journal) < nblocks) { - write_unlock(&journal->j_state_lock); -+ if (current->plug) -+ io_schedule(); - mutex_lock(&journal->j_checkpoint_mutex); - - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/locks.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/locks.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/locks.c 2017-04-16 10:38:21.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/locks.c 2017-04-18 17:54:25.000000000 +0200 -@@ -935,7 +935,7 @@ - return -ENOMEM; - } - -- percpu_down_read_preempt_disable(&file_rwsem); -+ percpu_down_read(&file_rwsem); - spin_lock(&ctx->flc_lock); - if (request->fl_flags & FL_ACCESS) - goto find_conflict; -@@ -976,7 +976,7 @@ - - out: - spin_unlock(&ctx->flc_lock); -- percpu_up_read_preempt_enable(&file_rwsem); -+ percpu_up_read(&file_rwsem); - if (new_fl) - locks_free_lock(new_fl); - locks_dispose_list(&dispose); -@@ -1013,7 +1013,7 @@ - new_fl2 = locks_alloc_lock(); - } - -- percpu_down_read_preempt_disable(&file_rwsem); -+ percpu_down_read(&file_rwsem); - spin_lock(&ctx->flc_lock); - /* - * New lock request. Walk all POSIX locks and look for conflicts. If -@@ -1185,7 +1185,7 @@ - } - out: - spin_unlock(&ctx->flc_lock); -- percpu_up_read_preempt_enable(&file_rwsem); -+ percpu_up_read(&file_rwsem); - /* - * Free any unused locks. - */ -@@ -1460,7 +1460,7 @@ - return error; - } - -- percpu_down_read_preempt_disable(&file_rwsem); -+ percpu_down_read(&file_rwsem); - spin_lock(&ctx->flc_lock); - - time_out_leases(inode, &dispose); -@@ -1512,13 +1512,13 @@ - locks_insert_block(fl, new_fl); - trace_break_lease_block(inode, new_fl); - spin_unlock(&ctx->flc_lock); -- percpu_up_read_preempt_enable(&file_rwsem); -+ percpu_up_read(&file_rwsem); - - locks_dispose_list(&dispose); - error = wait_event_interruptible_timeout(new_fl->fl_wait, - !new_fl->fl_next, break_time); - -- percpu_down_read_preempt_disable(&file_rwsem); -+ percpu_down_read(&file_rwsem); - spin_lock(&ctx->flc_lock); - trace_break_lease_unblock(inode, new_fl); - locks_delete_block(new_fl); -@@ -1535,7 +1535,7 @@ - } - out: - spin_unlock(&ctx->flc_lock); -- percpu_up_read_preempt_enable(&file_rwsem); -+ percpu_up_read(&file_rwsem); - locks_dispose_list(&dispose); - locks_free_lock(new_fl); - return error; -@@ -1609,7 +1609,7 @@ - - ctx = smp_load_acquire(&inode->i_flctx); - if (ctx && !list_empty_careful(&ctx->flc_lease)) { -- percpu_down_read_preempt_disable(&file_rwsem); -+ percpu_down_read(&file_rwsem); - spin_lock(&ctx->flc_lock); - time_out_leases(inode, &dispose); - list_for_each_entry(fl, &ctx->flc_lease, fl_list) { -@@ -1619,7 +1619,7 @@ - break; - } - spin_unlock(&ctx->flc_lock); -- percpu_up_read_preempt_enable(&file_rwsem); -+ percpu_up_read(&file_rwsem); - - locks_dispose_list(&dispose); - } -@@ -1694,7 +1694,7 @@ - return -EINVAL; - } - -- percpu_down_read_preempt_disable(&file_rwsem); -+ percpu_down_read(&file_rwsem); - spin_lock(&ctx->flc_lock); - time_out_leases(inode, &dispose); - error = check_conflicting_open(dentry, arg, lease->fl_flags); -@@ -1765,7 +1765,7 @@ - lease->fl_lmops->lm_setup(lease, priv); - out: - spin_unlock(&ctx->flc_lock); -- percpu_up_read_preempt_enable(&file_rwsem); -+ percpu_up_read(&file_rwsem); - locks_dispose_list(&dispose); - if (is_deleg) - inode_unlock(inode); -@@ -1788,7 +1788,7 @@ - return error; - } - -- percpu_down_read_preempt_disable(&file_rwsem); -+ percpu_down_read(&file_rwsem); - spin_lock(&ctx->flc_lock); - list_for_each_entry(fl, &ctx->flc_lease, fl_list) { - if (fl->fl_file == filp && -@@ -1801,7 +1801,7 @@ - if (victim) - error = fl->fl_lmops->lm_change(victim, F_UNLCK, &dispose); - spin_unlock(&ctx->flc_lock); -- percpu_up_read_preempt_enable(&file_rwsem); -+ percpu_up_read(&file_rwsem); - locks_dispose_list(&dispose); - return error; - } -@@ -2532,13 +2532,13 @@ - if (list_empty(&ctx->flc_lease)) - return; - -- percpu_down_read_preempt_disable(&file_rwsem); -+ percpu_down_read(&file_rwsem); - spin_lock(&ctx->flc_lock); - list_for_each_entry_safe(fl, tmp, &ctx->flc_lease, fl_list) - if (filp == fl->fl_file) - lease_modify(fl, F_UNLCK, &dispose); - spin_unlock(&ctx->flc_lock); -- percpu_up_read_preempt_enable(&file_rwsem); -+ percpu_up_read(&file_rwsem); - - locks_dispose_list(&dispose); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/namei.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/namei.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/namei.c 2017-04-16 10:38:21.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/namei.c 2017-04-18 17:54:25.000000000 +0200 -@@ -1626,7 +1626,7 @@ - { - struct dentry *dentry = ERR_PTR(-ENOENT), *old; - struct inode *inode = dir->d_inode; -- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); -+ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); - - inode_lock_shared(inode); - /* Don't go there if it's already dead */ -@@ -3083,7 +3083,7 @@ - struct dentry *dentry; - int error, create_error = 0; - umode_t mode = op->mode; -- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); -+ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); - - if (unlikely(IS_DEADDIR(dir_inode))) - return -ENOENT; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/namespace.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/namespace.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/namespace.c 2017-04-16 10:38:21.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/namespace.c 2017-04-18 17:54:25.000000000 +0200 -@@ -14,6 +14,7 @@ - #include - #include - #include -+#include - #include - #include - #include /* init_rootfs */ -@@ -356,8 +357,11 @@ - * incremented count after it has set MNT_WRITE_HOLD. - */ - smp_mb(); -- while (ACCESS_ONCE(mnt->mnt.mnt_flags) & MNT_WRITE_HOLD) -- cpu_relax(); -+ while (ACCESS_ONCE(mnt->mnt.mnt_flags) & MNT_WRITE_HOLD) { -+ preempt_enable(); -+ cpu_chill(); -+ preempt_disable(); -+ } - /* - * After the slowpath clears MNT_WRITE_HOLD, mnt_is_readonly will - * be set to match its requirements. So we must not load that until -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/delegation.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/delegation.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/delegation.c 2017-04-16 10:38:21.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/delegation.c 2017-04-18 17:54:25.000000000 +0200 -@@ -150,11 +150,11 @@ - sp = state->owner; - /* Block nfs4_proc_unlck */ - mutex_lock(&sp->so_delegreturn_mutex); -- seq = raw_seqcount_begin(&sp->so_reclaim_seqcount); -+ seq = read_seqbegin(&sp->so_reclaim_seqlock); - err = nfs4_open_delegation_recall(ctx, state, stateid, type); - if (!err) - err = nfs_delegation_claim_locks(ctx, state, stateid); -- if (!err && read_seqcount_retry(&sp->so_reclaim_seqcount, seq)) -+ if (!err && read_seqretry(&sp->so_reclaim_seqlock, seq)) - err = -EAGAIN; - mutex_unlock(&sp->so_delegreturn_mutex); - put_nfs_open_context(ctx); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/dir.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/dir.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/dir.c 2017-04-16 10:38:21.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/dir.c 2017-04-18 17:54:25.000000000 +0200 -@@ -485,7 +485,7 @@ - void nfs_prime_dcache(struct dentry *parent, struct nfs_entry *entry) - { - struct qstr filename = QSTR_INIT(entry->name, entry->len); -- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); -+ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); - struct dentry *dentry; - struct dentry *alias; - struct inode *dir = d_inode(parent); -@@ -1487,7 +1487,7 @@ - struct file *file, unsigned open_flags, - umode_t mode, int *opened) - { -- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); -+ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); - struct nfs_open_context *ctx; - struct dentry *res; - struct iattr attr = { .ia_valid = ATTR_OPEN }; -@@ -1802,7 +1802,11 @@ - - trace_nfs_rmdir_enter(dir, dentry); - if (d_really_is_positive(dentry)) { -+#ifdef CONFIG_PREEMPT_RT_BASE -+ down(&NFS_I(d_inode(dentry))->rmdir_sem); -+#else - down_write(&NFS_I(d_inode(dentry))->rmdir_sem); -+#endif - error = NFS_PROTO(dir)->rmdir(dir, &dentry->d_name); - /* Ensure the VFS deletes this inode */ - switch (error) { -@@ -1812,7 +1816,11 @@ - case -ENOENT: - nfs_dentry_handle_enoent(dentry); - } -+#ifdef CONFIG_PREEMPT_RT_BASE -+ up(&NFS_I(d_inode(dentry))->rmdir_sem); -+#else - up_write(&NFS_I(d_inode(dentry))->rmdir_sem); -+#endif - } else - error = NFS_PROTO(dir)->rmdir(dir, &dentry->d_name); - trace_nfs_rmdir_exit(dir, dentry, error); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/inode.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/inode.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/inode.c 2017-04-16 10:38:21.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/inode.c 2017-04-18 17:54:25.000000000 +0200 -@@ -1957,7 +1957,11 @@ - nfsi->nrequests = 0; - nfsi->commit_info.ncommit = 0; - atomic_set(&nfsi->commit_info.rpcs_out, 0); -+#ifdef CONFIG_PREEMPT_RT_BASE -+ sema_init(&nfsi->rmdir_sem, 1); -+#else - init_rwsem(&nfsi->rmdir_sem); -+#endif - nfs4_init_once(nfsi); - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/nfs4_fs.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/nfs4_fs.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/nfs4_fs.h 2017-04-16 10:38:21.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/nfs4_fs.h 2017-04-18 17:54:25.000000000 +0200 -@@ -111,7 +111,7 @@ - unsigned long so_flags; - struct list_head so_states; - struct nfs_seqid_counter so_seqid; -- seqcount_t so_reclaim_seqcount; -+ seqlock_t so_reclaim_seqlock; - struct mutex so_delegreturn_mutex; - }; - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/nfs4proc.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/nfs4proc.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/nfs4proc.c 2017-04-16 10:38:21.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/nfs4proc.c 2017-04-18 17:54:25.000000000 +0200 -@@ -2695,7 +2695,7 @@ - unsigned int seq; - int ret; - -- seq = raw_seqcount_begin(&sp->so_reclaim_seqcount); -+ seq = raw_seqcount_begin(&sp->so_reclaim_seqlock.seqcount); - - ret = _nfs4_proc_open(opendata); - if (ret != 0) -@@ -2733,7 +2733,7 @@ - - if (d_inode(dentry) == state->inode) { - nfs_inode_attach_open_context(ctx); -- if (read_seqcount_retry(&sp->so_reclaim_seqcount, seq)) -+ if (read_seqretry(&sp->so_reclaim_seqlock, seq)) - nfs4_schedule_stateid_recovery(server, state); - } - out: -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/nfs4state.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/nfs4state.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/nfs4state.c 2017-04-16 10:38:21.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/nfs4state.c 2017-04-18 17:54:25.000000000 +0200 -@@ -488,7 +488,7 @@ - nfs4_init_seqid_counter(&sp->so_seqid); - atomic_set(&sp->so_count, 1); - INIT_LIST_HEAD(&sp->so_lru); -- seqcount_init(&sp->so_reclaim_seqcount); -+ seqlock_init(&sp->so_reclaim_seqlock); - mutex_init(&sp->so_delegreturn_mutex); - return sp; - } -@@ -1497,8 +1497,12 @@ - * recovering after a network partition or a reboot from a - * server that doesn't support a grace period. - */ -+#ifdef CONFIG_PREEMPT_RT_FULL -+ write_seqlock(&sp->so_reclaim_seqlock); -+#else -+ write_seqcount_begin(&sp->so_reclaim_seqlock.seqcount); -+#endif - spin_lock(&sp->so_lock); -- raw_write_seqcount_begin(&sp->so_reclaim_seqcount); - restart: - list_for_each_entry(state, &sp->so_states, open_states) { - if (!test_and_clear_bit(ops->state_flag_bit, &state->flags)) -@@ -1567,14 +1571,20 @@ - spin_lock(&sp->so_lock); - goto restart; - } -- raw_write_seqcount_end(&sp->so_reclaim_seqcount); - spin_unlock(&sp->so_lock); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ write_sequnlock(&sp->so_reclaim_seqlock); -+#else -+ write_seqcount_end(&sp->so_reclaim_seqlock.seqcount); -+#endif - return 0; - out_err: - nfs4_put_open_state(state); -- spin_lock(&sp->so_lock); -- raw_write_seqcount_end(&sp->so_reclaim_seqcount); -- spin_unlock(&sp->so_lock); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ write_sequnlock(&sp->so_reclaim_seqlock); -+#else -+ write_seqcount_end(&sp->so_reclaim_seqlock.seqcount); -+#endif - return status; - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/unlink.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/unlink.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/nfs/unlink.c 2017-04-16 10:38:21.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/nfs/unlink.c 2017-04-18 17:54:25.000000000 +0200 -@@ -12,7 +12,7 @@ - #include - #include - #include --#include -+#include - #include - #include - -@@ -51,6 +51,29 @@ - rpc_restart_call_prepare(task); - } - -+#ifdef CONFIG_PREEMPT_RT_BASE -+static void nfs_down_anon(struct semaphore *sema) -+{ -+ down(sema); -+} -+ -+static void nfs_up_anon(struct semaphore *sema) -+{ -+ up(sema); -+} -+ -+#else -+static void nfs_down_anon(struct rw_semaphore *rwsem) -+{ -+ down_read_non_owner(rwsem); -+} -+ -+static void nfs_up_anon(struct rw_semaphore *rwsem) -+{ -+ up_read_non_owner(rwsem); -+} -+#endif -+ - /** - * nfs_async_unlink_release - Release the sillydelete data. - * @task: rpc_task of the sillydelete -@@ -64,7 +87,7 @@ - struct dentry *dentry = data->dentry; - struct super_block *sb = dentry->d_sb; - -- up_read_non_owner(&NFS_I(d_inode(dentry->d_parent))->rmdir_sem); -+ nfs_up_anon(&NFS_I(d_inode(dentry->d_parent))->rmdir_sem); - d_lookup_done(dentry); - nfs_free_unlinkdata(data); - dput(dentry); -@@ -117,10 +140,10 @@ - struct inode *dir = d_inode(dentry->d_parent); - struct dentry *alias; - -- down_read_non_owner(&NFS_I(dir)->rmdir_sem); -+ nfs_down_anon(&NFS_I(dir)->rmdir_sem); - alias = d_alloc_parallel(dentry->d_parent, &data->args.name, &data->wq); - if (IS_ERR(alias)) { -- up_read_non_owner(&NFS_I(dir)->rmdir_sem); -+ nfs_up_anon(&NFS_I(dir)->rmdir_sem); - return 0; - } - if (!d_in_lookup(alias)) { -@@ -142,7 +165,7 @@ - ret = 0; - spin_unlock(&alias->d_lock); - dput(alias); -- up_read_non_owner(&NFS_I(dir)->rmdir_sem); -+ nfs_up_anon(&NFS_I(dir)->rmdir_sem); - /* - * If we'd displaced old cached devname, free it. At that - * point dentry is definitely not a root, so we won't need -@@ -182,7 +205,7 @@ - goto out_free_name; - } - data->res.dir_attr = &data->dir_attr; -- init_waitqueue_head(&data->wq); -+ init_swait_queue_head(&data->wq); - - status = -EBUSY; - spin_lock(&dentry->d_lock); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/ntfs/aops.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/ntfs/aops.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/ntfs/aops.c 2017-04-16 10:38:21.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/ntfs/aops.c 2017-04-18 17:54:25.000000000 +0200 -@@ -92,13 +92,13 @@ - ofs = 0; - if (file_ofs < init_size) - ofs = init_size - file_ofs; -- local_irq_save(flags); -+ local_irq_save_nort(flags); - kaddr = kmap_atomic(page); - memset(kaddr + bh_offset(bh) + ofs, 0, - bh->b_size - ofs); - flush_dcache_page(page); - kunmap_atomic(kaddr); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - } - } else { - clear_buffer_uptodate(bh); -@@ -107,8 +107,7 @@ - "0x%llx.", (unsigned long long)bh->b_blocknr); - } - first = page_buffers(page); -- local_irq_save(flags); -- bit_spin_lock(BH_Uptodate_Lock, &first->b_state); -+ flags = bh_uptodate_lock_irqsave(first); - clear_buffer_async_read(bh); - unlock_buffer(bh); - tmp = bh; -@@ -123,8 +122,7 @@ - } - tmp = tmp->b_this_page; - } while (tmp != bh); -- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); -- local_irq_restore(flags); -+ bh_uptodate_unlock_irqrestore(first, flags); - /* - * If none of the buffers had errors then we can set the page uptodate, - * but we first have to perform the post read mst fixups, if the -@@ -145,13 +143,13 @@ - recs = PAGE_SIZE / rec_size; - /* Should have been verified before we got here... */ - BUG_ON(!recs); -- local_irq_save(flags); -+ local_irq_save_nort(flags); - kaddr = kmap_atomic(page); - for (i = 0; i < recs; i++) - post_read_mst_fixup((NTFS_RECORD*)(kaddr + - i * rec_size), rec_size); - kunmap_atomic(kaddr); -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - flush_dcache_page(page); - if (likely(page_uptodate && !PageError(page))) - SetPageUptodate(page); -@@ -159,9 +157,7 @@ - unlock_page(page); - return; - still_busy: -- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); -- local_irq_restore(flags); -- return; -+ bh_uptodate_unlock_irqrestore(first, flags); - } - - /** -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/proc/base.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/proc/base.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/proc/base.c 2017-04-16 10:38:22.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/proc/base.c 2017-04-18 17:54:25.000000000 +0200 -@@ -1834,7 +1834,7 @@ - - child = d_hash_and_lookup(dir, &qname); - if (!child) { -- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); -+ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); - child = d_alloc_parallel(dir, &qname, &wq); - if (IS_ERR(child)) - goto end_instantiate; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/proc/proc_sysctl.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/proc/proc_sysctl.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/proc/proc_sysctl.c 2017-04-16 10:38:22.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/proc/proc_sysctl.c 2017-04-18 17:54:26.000000000 +0200 -@@ -632,7 +632,7 @@ - - child = d_lookup(dir, &qname); - if (!child) { -- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); -+ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); - child = d_alloc_parallel(dir, &qname, &wq); - if (IS_ERR(child)) - return false; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/timerfd.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/timerfd.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/fs/timerfd.c 2017-04-16 10:38:22.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/fs/timerfd.c 2017-04-18 17:54:26.000000000 +0200 -@@ -460,7 +460,10 @@ - break; - } - spin_unlock_irq(&ctx->wqh.lock); -- cpu_relax(); -+ if (isalarm(ctx)) -+ hrtimer_wait_for_timer(&ctx->t.alarm.timer); -+ else -+ hrtimer_wait_for_timer(&ctx->t.tmr); - } - - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/acpi/platform/aclinux.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/acpi/platform/aclinux.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/acpi/platform/aclinux.h 2017-04-16 10:38:23.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/acpi/platform/aclinux.h 2017-04-18 17:54:26.000000000 +0200 -@@ -133,6 +133,7 @@ - - #define acpi_cache_t struct kmem_cache - #define acpi_spinlock spinlock_t * -+#define acpi_raw_spinlock raw_spinlock_t * - #define acpi_cpu_flags unsigned long - - /* Use native linux version of acpi_os_allocate_zeroed */ -@@ -151,6 +152,20 @@ - #define ACPI_USE_ALTERNATE_PROTOTYPE_acpi_os_get_thread_id - #define ACPI_USE_ALTERNATE_PROTOTYPE_acpi_os_create_lock - -+#define acpi_os_create_raw_lock(__handle) \ -+({ \ -+ raw_spinlock_t *lock = ACPI_ALLOCATE(sizeof(*lock)); \ -+ \ -+ if (lock) { \ -+ *(__handle) = lock; \ -+ raw_spin_lock_init(*(__handle)); \ -+ } \ -+ lock ? AE_OK : AE_NO_MEMORY; \ -+ }) -+ -+#define acpi_os_delete_raw_lock(__handle) kfree(__handle) -+ -+ - /* - * OSL interfaces used by debugger/disassembler - */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/asm-generic/bug.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/asm-generic/bug.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/asm-generic/bug.h 2017-04-16 10:38:23.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/asm-generic/bug.h 2017-04-18 17:54:26.000000000 +0200 -@@ -215,6 +215,20 @@ - # define WARN_ON_SMP(x) ({0;}) - #endif - -+#ifdef CONFIG_PREEMPT_RT_BASE -+# define BUG_ON_RT(c) BUG_ON(c) -+# define BUG_ON_NONRT(c) do { } while (0) -+# define WARN_ON_RT(condition) WARN_ON(condition) -+# define WARN_ON_NONRT(condition) do { } while (0) -+# define WARN_ON_ONCE_NONRT(condition) do { } while (0) -+#else -+# define BUG_ON_RT(c) do { } while (0) -+# define BUG_ON_NONRT(c) BUG_ON(c) -+# define WARN_ON_RT(condition) do { } while (0) -+# define WARN_ON_NONRT(condition) WARN_ON(condition) -+# define WARN_ON_ONCE_NONRT(condition) WARN_ON_ONCE(condition) -+#endif -+ - #endif /* __ASSEMBLY__ */ - - #endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/blk-mq.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/blk-mq.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/blk-mq.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/blk-mq.h 2017-04-18 17:54:26.000000000 +0200 -@@ -209,7 +209,7 @@ - return unique_tag & BLK_MQ_UNIQUE_TAG_MASK; - } - -- -+void __blk_mq_complete_request_remote_work(struct work_struct *work); - int blk_mq_request_started(struct request *rq); - void blk_mq_start_request(struct request *rq); - void blk_mq_end_request(struct request *rq, int error); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/blkdev.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/blkdev.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/blkdev.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/blkdev.h 2017-04-18 17:54:26.000000000 +0200 -@@ -89,6 +89,7 @@ - struct list_head queuelist; - union { - struct call_single_data csd; -+ struct work_struct work; - u64 fifo_time; - }; - -@@ -467,7 +468,7 @@ - struct throtl_data *td; - #endif - struct rcu_head rcu_head; -- wait_queue_head_t mq_freeze_wq; -+ struct swait_queue_head mq_freeze_wq; - struct percpu_ref q_usage_counter; - struct list_head all_q_node; - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/bottom_half.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/bottom_half.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/bottom_half.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/bottom_half.h 2017-04-18 17:54:26.000000000 +0200 -@@ -3,6 +3,39 @@ - - #include - -+#ifdef CONFIG_PREEMPT_RT_FULL -+ -+extern void __local_bh_disable(void); -+extern void _local_bh_enable(void); -+extern void __local_bh_enable(void); -+ -+static inline void local_bh_disable(void) -+{ -+ __local_bh_disable(); -+} -+ -+static inline void __local_bh_disable_ip(unsigned long ip, unsigned int cnt) -+{ -+ __local_bh_disable(); -+} -+ -+static inline void local_bh_enable(void) -+{ -+ __local_bh_enable(); -+} -+ -+static inline void __local_bh_enable_ip(unsigned long ip, unsigned int cnt) -+{ -+ __local_bh_enable(); -+} -+ -+static inline void local_bh_enable_ip(unsigned long ip) -+{ -+ __local_bh_enable(); -+} -+ -+#else -+ - #ifdef CONFIG_TRACE_IRQFLAGS - extern void __local_bh_disable_ip(unsigned long ip, unsigned int cnt); - #else -@@ -30,5 +63,6 @@ - { - __local_bh_enable_ip(_THIS_IP_, SOFTIRQ_DISABLE_OFFSET); - } -+#endif - - #endif /* _LINUX_BH_H */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/buffer_head.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/buffer_head.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/buffer_head.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/buffer_head.h 2017-04-18 17:54:26.000000000 +0200 -@@ -75,8 +75,50 @@ - struct address_space *b_assoc_map; /* mapping this buffer is - associated with */ - atomic_t b_count; /* users using this buffer_head */ -+#ifdef CONFIG_PREEMPT_RT_BASE -+ spinlock_t b_uptodate_lock; -+#if IS_ENABLED(CONFIG_JBD2) -+ spinlock_t b_state_lock; -+ spinlock_t b_journal_head_lock; -+#endif -+#endif - }; - -+static inline unsigned long bh_uptodate_lock_irqsave(struct buffer_head *bh) -+{ -+ unsigned long flags; -+ -+#ifndef CONFIG_PREEMPT_RT_BASE -+ local_irq_save(flags); -+ bit_spin_lock(BH_Uptodate_Lock, &bh->b_state); -+#else -+ spin_lock_irqsave(&bh->b_uptodate_lock, flags); -+#endif -+ return flags; -+} -+ -+static inline void -+bh_uptodate_unlock_irqrestore(struct buffer_head *bh, unsigned long flags) -+{ -+#ifndef CONFIG_PREEMPT_RT_BASE -+ bit_spin_unlock(BH_Uptodate_Lock, &bh->b_state); -+ local_irq_restore(flags); -+#else -+ spin_unlock_irqrestore(&bh->b_uptodate_lock, flags); -+#endif -+} -+ -+static inline void buffer_head_init_locks(struct buffer_head *bh) -+{ -+#ifdef CONFIG_PREEMPT_RT_BASE -+ spin_lock_init(&bh->b_uptodate_lock); -+#if IS_ENABLED(CONFIG_JBD2) -+ spin_lock_init(&bh->b_state_lock); -+ spin_lock_init(&bh->b_journal_head_lock); -+#endif -+#endif -+} -+ - /* - * macro tricks to expand the set_buffer_foo(), clear_buffer_foo() - * and buffer_foo() functions. -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/cgroup-defs.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/cgroup-defs.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/cgroup-defs.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/cgroup-defs.h 2017-04-18 17:54:26.000000000 +0200 -@@ -16,6 +16,7 @@ - #include - #include - #include -+#include - - #ifdef CONFIG_CGROUPS - -@@ -137,6 +138,7 @@ - /* percpu_ref killing and RCU release */ - struct rcu_head rcu_head; - struct work_struct destroy_work; -+ struct swork_event destroy_swork; - }; - - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/completion.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/completion.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/completion.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/completion.h 2017-04-18 17:54:26.000000000 +0200 -@@ -7,8 +7,7 @@ - * Atomic wait-for-completion handler data structures. - * See kernel/sched/completion.c for details. - */ -- --#include -+#include - - /* - * struct completion - structure used to maintain state for a "completion" -@@ -24,11 +23,11 @@ - */ - struct completion { - unsigned int done; -- wait_queue_head_t wait; -+ struct swait_queue_head wait; - }; - - #define COMPLETION_INITIALIZER(work) \ -- { 0, __WAIT_QUEUE_HEAD_INITIALIZER((work).wait) } -+ { 0, __SWAIT_QUEUE_HEAD_INITIALIZER((work).wait) } - - #define COMPLETION_INITIALIZER_ONSTACK(work) \ - ({ init_completion(&work); work; }) -@@ -73,7 +72,7 @@ - static inline void init_completion(struct completion *x) - { - x->done = 0; -- init_waitqueue_head(&x->wait); -+ init_swait_queue_head(&x->wait); - } - - /** -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/cpu.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/cpu.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/cpu.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/cpu.h 2017-04-18 17:54:26.000000000 +0200 -@@ -182,6 +182,8 @@ - extern void put_online_cpus(void); - extern void cpu_hotplug_disable(void); - extern void cpu_hotplug_enable(void); -+extern void pin_current_cpu(void); -+extern void unpin_current_cpu(void); - #define hotcpu_notifier(fn, pri) cpu_notifier(fn, pri) - #define __hotcpu_notifier(fn, pri) __cpu_notifier(fn, pri) - #define register_hotcpu_notifier(nb) register_cpu_notifier(nb) -@@ -199,6 +201,8 @@ - #define put_online_cpus() do { } while (0) - #define cpu_hotplug_disable() do { } while (0) - #define cpu_hotplug_enable() do { } while (0) -+static inline void pin_current_cpu(void) { } -+static inline void unpin_current_cpu(void) { } - #define hotcpu_notifier(fn, pri) do { (void)(fn); } while (0) - #define __hotcpu_notifier(fn, pri) do { (void)(fn); } while (0) - /* These aren't inline functions due to a GCC bug. */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/dcache.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/dcache.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/dcache.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/dcache.h 2017-04-18 17:54:26.000000000 +0200 -@@ -11,6 +11,7 @@ - #include - #include - #include -+#include - - struct path; - struct vfsmount; -@@ -100,7 +101,7 @@ - - union { - struct list_head d_lru; /* LRU list */ -- wait_queue_head_t *d_wait; /* in-lookup ones only */ -+ struct swait_queue_head *d_wait; /* in-lookup ones only */ - }; - struct list_head d_child; /* child of parent list */ - struct list_head d_subdirs; /* our children */ -@@ -230,7 +231,7 @@ - extern struct dentry * d_alloc(struct dentry *, const struct qstr *); - extern struct dentry * d_alloc_pseudo(struct super_block *, const struct qstr *); - extern struct dentry * d_alloc_parallel(struct dentry *, const struct qstr *, -- wait_queue_head_t *); -+ struct swait_queue_head *); - extern struct dentry * d_splice_alias(struct inode *, struct dentry *); - extern struct dentry * d_add_ci(struct dentry *, struct inode *, struct qstr *); - extern struct dentry * d_exact_alias(struct dentry *, struct inode *); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/delay.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/delay.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/delay.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/delay.h 2017-04-18 17:54:26.000000000 +0200 -@@ -52,4 +52,10 @@ - msleep(seconds * 1000); - } - -+#ifdef CONFIG_PREEMPT_RT_FULL -+extern void cpu_chill(void); -+#else -+# define cpu_chill() cpu_relax() -+#endif -+ - #endif /* defined(_LINUX_DELAY_H) */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/highmem.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/highmem.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/highmem.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/highmem.h 2017-04-18 17:54:26.000000000 +0200 -@@ -7,6 +7,7 @@ - #include - #include - #include -+#include - - #include - -@@ -65,7 +66,7 @@ - - static inline void *kmap_atomic(struct page *page) - { -- preempt_disable(); -+ preempt_disable_nort(); - pagefault_disable(); - return page_address(page); - } -@@ -74,7 +75,7 @@ - static inline void __kunmap_atomic(void *addr) - { - pagefault_enable(); -- preempt_enable(); -+ preempt_enable_nort(); - } - - #define kmap_atomic_pfn(pfn) kmap_atomic(pfn_to_page(pfn)) -@@ -86,32 +87,51 @@ - - #if defined(CONFIG_HIGHMEM) || defined(CONFIG_X86_32) - -+#ifndef CONFIG_PREEMPT_RT_FULL - DECLARE_PER_CPU(int, __kmap_atomic_idx); -+#endif - - static inline int kmap_atomic_idx_push(void) - { -+#ifndef CONFIG_PREEMPT_RT_FULL - int idx = __this_cpu_inc_return(__kmap_atomic_idx) - 1; - --#ifdef CONFIG_DEBUG_HIGHMEM -+# ifdef CONFIG_DEBUG_HIGHMEM - WARN_ON_ONCE(in_irq() && !irqs_disabled()); - BUG_ON(idx >= KM_TYPE_NR); --#endif -+# endif - return idx; -+#else -+ current->kmap_idx++; -+ BUG_ON(current->kmap_idx > KM_TYPE_NR); -+ return current->kmap_idx - 1; -+#endif - } - - static inline int kmap_atomic_idx(void) - { -+#ifndef CONFIG_PREEMPT_RT_FULL - return __this_cpu_read(__kmap_atomic_idx) - 1; -+#else -+ return current->kmap_idx - 1; -+#endif - } - - static inline void kmap_atomic_idx_pop(void) - { --#ifdef CONFIG_DEBUG_HIGHMEM -+#ifndef CONFIG_PREEMPT_RT_FULL -+# ifdef CONFIG_DEBUG_HIGHMEM - int idx = __this_cpu_dec_return(__kmap_atomic_idx); - - BUG_ON(idx < 0); --#else -+# else - __this_cpu_dec(__kmap_atomic_idx); -+# endif -+#else -+ current->kmap_idx--; -+# ifdef CONFIG_DEBUG_HIGHMEM -+ BUG_ON(current->kmap_idx < 0); -+# endif - #endif - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/hrtimer.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/hrtimer.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/hrtimer.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/hrtimer.h 2017-04-18 17:54:26.000000000 +0200 -@@ -87,6 +87,9 @@ - * @function: timer expiry callback function - * @base: pointer to the timer base (per cpu and per clock) - * @state: state information (See bit values above) -+ * @cb_entry: list entry to defer timers from hardirq context -+ * @irqsafe: timer can run in hardirq context -+ * @praecox: timer expiry time if expired at the time of programming - * @is_rel: Set if the timer was armed relative - * @start_pid: timer statistics field to store the pid of the task which - * started the timer -@@ -103,6 +106,11 @@ - enum hrtimer_restart (*function)(struct hrtimer *); - struct hrtimer_clock_base *base; - u8 state; -+ struct list_head cb_entry; -+ int irqsafe; -+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST -+ ktime_t praecox; -+#endif - u8 is_rel; - #ifdef CONFIG_TIMER_STATS - int start_pid; -@@ -123,11 +131,7 @@ - struct task_struct *task; - }; - --#ifdef CONFIG_64BIT - # define HRTIMER_CLOCK_BASE_ALIGN 64 --#else --# define HRTIMER_CLOCK_BASE_ALIGN 32 --#endif - - /** - * struct hrtimer_clock_base - the timer base for a specific clock -@@ -136,6 +140,7 @@ - * timer to a base on another cpu. - * @clockid: clock id for per_cpu support - * @active: red black tree root node for the active timers -+ * @expired: list head for deferred timers. - * @get_time: function to retrieve the current time of the clock - * @offset: offset of this clock to the monotonic base - */ -@@ -144,6 +149,7 @@ - int index; - clockid_t clockid; - struct timerqueue_head active; -+ struct list_head expired; - ktime_t (*get_time)(void); - ktime_t offset; - } __attribute__((__aligned__(HRTIMER_CLOCK_BASE_ALIGN))); -@@ -187,6 +193,7 @@ - raw_spinlock_t lock; - seqcount_t seq; - struct hrtimer *running; -+ struct hrtimer *running_soft; - unsigned int cpu; - unsigned int active_bases; - unsigned int clock_was_set_seq; -@@ -203,6 +210,9 @@ - unsigned int nr_hangs; - unsigned int max_hang_time; - #endif -+#ifdef CONFIG_PREEMPT_RT_BASE -+ wait_queue_head_t wait; -+#endif - struct hrtimer_clock_base clock_base[HRTIMER_MAX_CLOCK_BASES]; - } ____cacheline_aligned; - -@@ -412,6 +422,13 @@ - hrtimer_start_expires(timer, HRTIMER_MODE_ABS); - } - -+/* Softirq preemption could deadlock timer removal */ -+#ifdef CONFIG_PREEMPT_RT_BASE -+ extern void hrtimer_wait_for_timer(const struct hrtimer *timer); -+#else -+# define hrtimer_wait_for_timer(timer) do { cpu_relax(); } while (0) -+#endif -+ - /* Query timers: */ - extern ktime_t __hrtimer_get_remaining(const struct hrtimer *timer, bool adjust); - -@@ -436,9 +453,15 @@ - * Helper function to check, whether the timer is running the callback - * function - */ --static inline int hrtimer_callback_running(struct hrtimer *timer) -+static inline int hrtimer_callback_running(const struct hrtimer *timer) - { -- return timer->base->cpu_base->running == timer; -+ if (timer->base->cpu_base->running == timer) -+ return 1; -+#ifdef CONFIG_PREEMPT_RT_BASE -+ if (timer->base->cpu_base->running_soft == timer) -+ return 1; -+#endif -+ return 0; - } - - /* Forward a hrtimer so it expires after now: */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/idr.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/idr.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/idr.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/idr.h 2017-04-18 17:54:26.000000000 +0200 -@@ -95,10 +95,14 @@ - * Each idr_preload() should be matched with an invocation of this - * function. See idr_preload() for details. - */ -+#ifdef CONFIG_PREEMPT_RT_FULL -+void idr_preload_end(void); -+#else - static inline void idr_preload_end(void) - { - preempt_enable(); - } -+#endif - - /** - * idr_find - return pointer for given id -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/init_task.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/init_task.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/init_task.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/init_task.h 2017-04-18 17:54:26.000000000 +0200 -@@ -150,6 +150,12 @@ - # define INIT_PERF_EVENTS(tsk) - #endif - -+#ifdef CONFIG_PREEMPT_RT_BASE -+# define INIT_TIMER_LIST .posix_timer_list = NULL, -+#else -+# define INIT_TIMER_LIST -+#endif -+ - #ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN - # define INIT_VTIME(tsk) \ - .vtime_seqcount = SEQCNT_ZERO(tsk.vtime_seqcount), \ -@@ -250,6 +256,7 @@ - .cpu_timers = INIT_CPU_TIMERS(tsk.cpu_timers), \ - .pi_lock = __RAW_SPIN_LOCK_UNLOCKED(tsk.pi_lock), \ - .timer_slack_ns = 50000, /* 50 usec default slack */ \ -+ INIT_TIMER_LIST \ - .pids = { \ - [PIDTYPE_PID] = INIT_PID_LINK(PIDTYPE_PID), \ - [PIDTYPE_PGID] = INIT_PID_LINK(PIDTYPE_PGID), \ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/interrupt.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/interrupt.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/interrupt.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/interrupt.h 2017-04-18 17:54:26.000000000 +0200 -@@ -14,6 +14,7 @@ - #include - #include - #include -+#include - - #include - #include -@@ -61,6 +62,7 @@ - * interrupt handler after suspending interrupts. For system - * wakeup devices users need to implement wakeup detection in - * their interrupt handlers. -+ * IRQF_NO_SOFTIRQ_CALL - Do not process softirqs in the irq thread context (RT) - */ - #define IRQF_SHARED 0x00000080 - #define IRQF_PROBE_SHARED 0x00000100 -@@ -74,6 +76,7 @@ - #define IRQF_NO_THREAD 0x00010000 - #define IRQF_EARLY_RESUME 0x00020000 - #define IRQF_COND_SUSPEND 0x00040000 -+#define IRQF_NO_SOFTIRQ_CALL 0x00080000 - - #define IRQF_TIMER (__IRQF_TIMER | IRQF_NO_SUSPEND | IRQF_NO_THREAD) - -@@ -196,7 +199,7 @@ - #ifdef CONFIG_LOCKDEP - # define local_irq_enable_in_hardirq() do { } while (0) - #else --# define local_irq_enable_in_hardirq() local_irq_enable() -+# define local_irq_enable_in_hardirq() local_irq_enable_nort() - #endif - - extern void disable_irq_nosync(unsigned int irq); -@@ -216,6 +219,7 @@ - * struct irq_affinity_notify - context for notification of IRQ affinity changes - * @irq: Interrupt to which notification applies - * @kref: Reference count, for internal use -+ * @swork: Swork item, for internal use - * @work: Work item, for internal use - * @notify: Function to be called on change. This will be - * called in process context. -@@ -227,7 +231,11 @@ - struct irq_affinity_notify { - unsigned int irq; - struct kref kref; -+#ifdef CONFIG_PREEMPT_RT_BASE -+ struct swork_event swork; -+#else - struct work_struct work; -+#endif - void (*notify)(struct irq_affinity_notify *, const cpumask_t *mask); - void (*release)(struct kref *ref); - }; -@@ -406,9 +414,13 @@ - bool state); - - #ifdef CONFIG_IRQ_FORCED_THREADING -+# ifndef CONFIG_PREEMPT_RT_BASE - extern bool force_irqthreads; -+# else -+# define force_irqthreads (true) -+# endif - #else --#define force_irqthreads (0) -+#define force_irqthreads (false) - #endif - - #ifndef __ARCH_SET_SOFTIRQ_PENDING -@@ -465,9 +477,10 @@ - void (*action)(struct softirq_action *); - }; - -+#ifndef CONFIG_PREEMPT_RT_FULL - asmlinkage void do_softirq(void); - asmlinkage void __do_softirq(void); -- -+static inline void thread_do_softirq(void) { do_softirq(); } - #ifdef __ARCH_HAS_DO_SOFTIRQ - void do_softirq_own_stack(void); - #else -@@ -476,13 +489,25 @@ - __do_softirq(); - } - #endif -+#else -+extern void thread_do_softirq(void); -+#endif - - extern void open_softirq(int nr, void (*action)(struct softirq_action *)); - extern void softirq_init(void); - extern void __raise_softirq_irqoff(unsigned int nr); -+#ifdef CONFIG_PREEMPT_RT_FULL -+extern void __raise_softirq_irqoff_ksoft(unsigned int nr); -+#else -+static inline void __raise_softirq_irqoff_ksoft(unsigned int nr) -+{ -+ __raise_softirq_irqoff(nr); -+} -+#endif - - extern void raise_softirq_irqoff(unsigned int nr); - extern void raise_softirq(unsigned int nr); -+extern void softirq_check_pending_idle(void); - - DECLARE_PER_CPU(struct task_struct *, ksoftirqd); - -@@ -504,8 +529,9 @@ - to be executed on some cpu at least once after this. - * If the tasklet is already scheduled, but its execution is still not - started, it will be executed only once. -- * If this tasklet is already running on another CPU (or schedule is called -- from tasklet itself), it is rescheduled for later. -+ * If this tasklet is already running on another CPU, it is rescheduled -+ for later. -+ * Schedule must not be called from the tasklet itself (a lockup occurs) - * Tasklet is strictly serialized wrt itself, but not - wrt another tasklets. If client needs some intertask synchronization, - he makes it with spinlocks. -@@ -530,27 +556,36 @@ - enum - { - TASKLET_STATE_SCHED, /* Tasklet is scheduled for execution */ -- TASKLET_STATE_RUN /* Tasklet is running (SMP only) */ -+ TASKLET_STATE_RUN, /* Tasklet is running (SMP only) */ -+ TASKLET_STATE_PENDING /* Tasklet is pending */ - }; - --#ifdef CONFIG_SMP -+#define TASKLET_STATEF_SCHED (1 << TASKLET_STATE_SCHED) -+#define TASKLET_STATEF_RUN (1 << TASKLET_STATE_RUN) -+#define TASKLET_STATEF_PENDING (1 << TASKLET_STATE_PENDING) -+ -+#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL) - static inline int tasklet_trylock(struct tasklet_struct *t) - { - return !test_and_set_bit(TASKLET_STATE_RUN, &(t)->state); - } - -+static inline int tasklet_tryunlock(struct tasklet_struct *t) -+{ -+ return cmpxchg(&t->state, TASKLET_STATEF_RUN, 0) == TASKLET_STATEF_RUN; -+} -+ - static inline void tasklet_unlock(struct tasklet_struct *t) - { - smp_mb__before_atomic(); - clear_bit(TASKLET_STATE_RUN, &(t)->state); - } - --static inline void tasklet_unlock_wait(struct tasklet_struct *t) --{ -- while (test_bit(TASKLET_STATE_RUN, &(t)->state)) { barrier(); } --} -+extern void tasklet_unlock_wait(struct tasklet_struct *t); -+ - #else - #define tasklet_trylock(t) 1 -+#define tasklet_tryunlock(t) 1 - #define tasklet_unlock_wait(t) do { } while (0) - #define tasklet_unlock(t) do { } while (0) - #endif -@@ -599,12 +634,7 @@ - smp_mb(); - } - --static inline void tasklet_enable(struct tasklet_struct *t) --{ -- smp_mb__before_atomic(); -- atomic_dec(&t->count); --} -- -+extern void tasklet_enable(struct tasklet_struct *t); - extern void tasklet_kill(struct tasklet_struct *t); - extern void tasklet_kill_immediate(struct tasklet_struct *t, unsigned int cpu); - extern void tasklet_init(struct tasklet_struct *t, -@@ -635,6 +665,12 @@ - tasklet_kill(&ttimer->tasklet); - } - -+#ifdef CONFIG_PREEMPT_RT_FULL -+extern void softirq_early_init(void); -+#else -+static inline void softirq_early_init(void) { } -+#endif -+ - /* - * Autoprobing for irqs: - * -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/irq.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/irq.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/irq.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/irq.h 2017-04-18 17:54:26.000000000 +0200 -@@ -72,6 +72,7 @@ - * IRQ_IS_POLLED - Always polled by another interrupt. Exclude - * it from the spurious interrupt detection - * mechanism and from core side polling. -+ * IRQ_NO_SOFTIRQ_CALL - No softirq processing in the irq thread context (RT) - * IRQ_DISABLE_UNLAZY - Disable lazy irq disable - */ - enum { -@@ -99,13 +100,14 @@ - IRQ_PER_CPU_DEVID = (1 << 17), - IRQ_IS_POLLED = (1 << 18), - IRQ_DISABLE_UNLAZY = (1 << 19), -+ IRQ_NO_SOFTIRQ_CALL = (1 << 20), - }; - - #define IRQF_MODIFY_MASK \ - (IRQ_TYPE_SENSE_MASK | IRQ_NOPROBE | IRQ_NOREQUEST | \ - IRQ_NOAUTOEN | IRQ_MOVE_PCNTXT | IRQ_LEVEL | IRQ_NO_BALANCING | \ - IRQ_PER_CPU | IRQ_NESTED_THREAD | IRQ_NOTHREAD | IRQ_PER_CPU_DEVID | \ -- IRQ_IS_POLLED | IRQ_DISABLE_UNLAZY) -+ IRQ_IS_POLLED | IRQ_DISABLE_UNLAZY | IRQ_NO_SOFTIRQ_CALL) - - #define IRQ_NO_BALANCING_MASK (IRQ_PER_CPU | IRQ_NO_BALANCING) - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/irq_work.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/irq_work.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/irq_work.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/irq_work.h 2017-04-18 17:54:26.000000000 +0200 -@@ -16,6 +16,7 @@ - #define IRQ_WORK_BUSY 2UL - #define IRQ_WORK_FLAGS 3UL - #define IRQ_WORK_LAZY 4UL /* Doesn't want IPI, wait for tick */ -+#define IRQ_WORK_HARD_IRQ 8UL /* Run hard IRQ context, even on RT */ - - struct irq_work { - unsigned long flags; -@@ -51,4 +52,10 @@ - static inline void irq_work_run(void) { } - #endif - -+#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) -+void irq_work_tick_soft(void); -+#else -+static inline void irq_work_tick_soft(void) { } -+#endif -+ - #endif /* _LINUX_IRQ_WORK_H */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/irqdesc.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/irqdesc.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/irqdesc.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/irqdesc.h 2017-04-18 17:54:26.000000000 +0200 -@@ -66,6 +66,7 @@ - unsigned int irqs_unhandled; - atomic_t threads_handled; - int threads_handled_last; -+ u64 random_ip; - raw_spinlock_t lock; - struct cpumask *percpu_enabled; - const struct cpumask *percpu_affinity; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/irqflags.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/irqflags.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/irqflags.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/irqflags.h 2017-04-18 17:54:26.000000000 +0200 -@@ -25,8 +25,6 @@ - # define trace_softirqs_enabled(p) ((p)->softirqs_enabled) - # define trace_hardirq_enter() do { current->hardirq_context++; } while (0) - # define trace_hardirq_exit() do { current->hardirq_context--; } while (0) --# define lockdep_softirq_enter() do { current->softirq_context++; } while (0) --# define lockdep_softirq_exit() do { current->softirq_context--; } while (0) - # define INIT_TRACE_IRQFLAGS .softirqs_enabled = 1, - #else - # define trace_hardirqs_on() do { } while (0) -@@ -39,9 +37,15 @@ - # define trace_softirqs_enabled(p) 0 - # define trace_hardirq_enter() do { } while (0) - # define trace_hardirq_exit() do { } while (0) -+# define INIT_TRACE_IRQFLAGS -+#endif -+ -+#if defined(CONFIG_TRACE_IRQFLAGS) && !defined(CONFIG_PREEMPT_RT_FULL) -+# define lockdep_softirq_enter() do { current->softirq_context++; } while (0) -+# define lockdep_softirq_exit() do { current->softirq_context--; } while (0) -+#else - # define lockdep_softirq_enter() do { } while (0) - # define lockdep_softirq_exit() do { } while (0) --# define INIT_TRACE_IRQFLAGS - #endif - - #if defined(CONFIG_IRQSOFF_TRACER) || \ -@@ -148,4 +152,23 @@ - - #define irqs_disabled_flags(flags) raw_irqs_disabled_flags(flags) - -+/* -+ * local_irq* variants depending on RT/!RT -+ */ -+#ifdef CONFIG_PREEMPT_RT_FULL -+# define local_irq_disable_nort() do { } while (0) -+# define local_irq_enable_nort() do { } while (0) -+# define local_irq_save_nort(flags) local_save_flags(flags) -+# define local_irq_restore_nort(flags) (void)(flags) -+# define local_irq_disable_rt() local_irq_disable() -+# define local_irq_enable_rt() local_irq_enable() -+#else -+# define local_irq_disable_nort() local_irq_disable() -+# define local_irq_enable_nort() local_irq_enable() -+# define local_irq_save_nort(flags) local_irq_save(flags) -+# define local_irq_restore_nort(flags) local_irq_restore(flags) -+# define local_irq_disable_rt() do { } while (0) -+# define local_irq_enable_rt() do { } while (0) -+#endif -+ - #endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/jbd2.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/jbd2.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/jbd2.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/jbd2.h 2017-04-18 17:54:26.000000000 +0200 -@@ -347,32 +347,56 @@ - - static inline void jbd_lock_bh_state(struct buffer_head *bh) - { -+#ifndef CONFIG_PREEMPT_RT_BASE - bit_spin_lock(BH_State, &bh->b_state); -+#else -+ spin_lock(&bh->b_state_lock); -+#endif - } - - static inline int jbd_trylock_bh_state(struct buffer_head *bh) - { -+#ifndef CONFIG_PREEMPT_RT_BASE - return bit_spin_trylock(BH_State, &bh->b_state); -+#else -+ return spin_trylock(&bh->b_state_lock); -+#endif - } - - static inline int jbd_is_locked_bh_state(struct buffer_head *bh) - { -+#ifndef CONFIG_PREEMPT_RT_BASE - return bit_spin_is_locked(BH_State, &bh->b_state); -+#else -+ return spin_is_locked(&bh->b_state_lock); -+#endif - } - - static inline void jbd_unlock_bh_state(struct buffer_head *bh) - { -+#ifndef CONFIG_PREEMPT_RT_BASE - bit_spin_unlock(BH_State, &bh->b_state); -+#else -+ spin_unlock(&bh->b_state_lock); -+#endif - } - - static inline void jbd_lock_bh_journal_head(struct buffer_head *bh) - { -+#ifndef CONFIG_PREEMPT_RT_BASE - bit_spin_lock(BH_JournalHead, &bh->b_state); -+#else -+ spin_lock(&bh->b_journal_head_lock); -+#endif - } - - static inline void jbd_unlock_bh_journal_head(struct buffer_head *bh) - { -+#ifndef CONFIG_PREEMPT_RT_BASE - bit_spin_unlock(BH_JournalHead, &bh->b_state); -+#else -+ spin_unlock(&bh->b_journal_head_lock); -+#endif - } - - #define J_ASSERT(assert) BUG_ON(!(assert)) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/kdb.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/kdb.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/kdb.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/kdb.h 2017-04-18 17:54:26.000000000 +0200 -@@ -167,6 +167,7 @@ - extern __printf(1, 2) int kdb_printf(const char *, ...); - typedef __printf(1, 2) int (*kdb_printf_t)(const char *, ...); - -+#define in_kdb_printk() (kdb_trap_printk) - extern void kdb_init(int level); - - /* Access to kdb specific polling devices */ -@@ -201,6 +202,7 @@ - extern int kdb_unregister(char *); - #else /* ! CONFIG_KGDB_KDB */ - static inline __printf(1, 2) int kdb_printf(const char *fmt, ...) { return 0; } -+#define in_kdb_printk() (0) - static inline void kdb_init(int level) {} - static inline int kdb_register(char *cmd, kdb_func_t func, char *usage, - char *help, short minlen) { return 0; } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/kernel.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/kernel.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/kernel.h 2017-04-16 10:38:24.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/kernel.h 2017-04-18 17:54:26.000000000 +0200 -@@ -194,6 +194,9 @@ - */ - # define might_sleep() \ - do { __might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0) -+ -+# define might_sleep_no_state_check() \ -+ do { ___might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0) - # define sched_annotate_sleep() (current->task_state_change = 0) - #else - static inline void ___might_sleep(const char *file, int line, -@@ -201,6 +204,7 @@ - static inline void __might_sleep(const char *file, int line, - int preempt_offset) { } - # define might_sleep() do { might_resched(); } while (0) -+# define might_sleep_no_state_check() do { might_resched(); } while (0) - # define sched_annotate_sleep() do { } while (0) - #endif - -@@ -488,6 +492,7 @@ - SYSTEM_HALT, - SYSTEM_POWER_OFF, - SYSTEM_RESTART, -+ SYSTEM_SUSPEND, - } system_state; - - #define TAINT_PROPRIETARY_MODULE 0 -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/list_bl.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/list_bl.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/list_bl.h 2017-04-16 10:38:25.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/list_bl.h 2017-04-18 17:54:26.000000000 +0200 -@@ -2,6 +2,7 @@ - #define _LINUX_LIST_BL_H - - #include -+#include - #include - - /* -@@ -32,13 +33,24 @@ - - struct hlist_bl_head { - struct hlist_bl_node *first; -+#ifdef CONFIG_PREEMPT_RT_BASE -+ raw_spinlock_t lock; -+#endif - }; - - struct hlist_bl_node { - struct hlist_bl_node *next, **pprev; - }; --#define INIT_HLIST_BL_HEAD(ptr) \ -- ((ptr)->first = NULL) -+ -+#ifdef CONFIG_PREEMPT_RT_BASE -+#define INIT_HLIST_BL_HEAD(h) \ -+do { \ -+ (h)->first = NULL; \ -+ raw_spin_lock_init(&(h)->lock); \ -+} while (0) -+#else -+#define INIT_HLIST_BL_HEAD(h) (h)->first = NULL -+#endif - - static inline void INIT_HLIST_BL_NODE(struct hlist_bl_node *h) - { -@@ -118,12 +130,26 @@ - - static inline void hlist_bl_lock(struct hlist_bl_head *b) - { -+#ifndef CONFIG_PREEMPT_RT_BASE - bit_spin_lock(0, (unsigned long *)b); -+#else -+ raw_spin_lock(&b->lock); -+#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) -+ __set_bit(0, (unsigned long *)b); -+#endif -+#endif - } - - static inline void hlist_bl_unlock(struct hlist_bl_head *b) - { -+#ifndef CONFIG_PREEMPT_RT_BASE - __bit_spin_unlock(0, (unsigned long *)b); -+#else -+#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) -+ __clear_bit(0, (unsigned long *)b); -+#endif -+ raw_spin_unlock(&b->lock); -+#endif - } - - static inline bool hlist_bl_is_locked(struct hlist_bl_head *b) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/locallock.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/locallock.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/locallock.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/locallock.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,278 @@ -+#ifndef _LINUX_LOCALLOCK_H -+#define _LINUX_LOCALLOCK_H -+ -+#include -+#include -+ -+#ifdef CONFIG_PREEMPT_RT_BASE -+ -+#ifdef CONFIG_DEBUG_SPINLOCK -+# define LL_WARN(cond) WARN_ON(cond) -+#else -+# define LL_WARN(cond) do { } while (0) -+#endif -+ -+/* -+ * per cpu lock based substitute for local_irq_*() -+ */ -+struct local_irq_lock { -+ spinlock_t lock; -+ struct task_struct *owner; -+ int nestcnt; -+ unsigned long flags; -+}; -+ -+#define DEFINE_LOCAL_IRQ_LOCK(lvar) \ -+ DEFINE_PER_CPU(struct local_irq_lock, lvar) = { \ -+ .lock = __SPIN_LOCK_UNLOCKED((lvar).lock) } -+ -+#define DECLARE_LOCAL_IRQ_LOCK(lvar) \ -+ DECLARE_PER_CPU(struct local_irq_lock, lvar) -+ -+#define local_irq_lock_init(lvar) \ -+ do { \ -+ int __cpu; \ -+ for_each_possible_cpu(__cpu) \ -+ spin_lock_init(&per_cpu(lvar, __cpu).lock); \ -+ } while (0) -+ -+/* -+ * spin_lock|trylock|unlock_local flavour that does not migrate disable -+ * used for __local_lock|trylock|unlock where get_local_var/put_local_var -+ * already takes care of the migrate_disable/enable -+ * for CONFIG_PREEMPT_BASE map to the normal spin_* calls. -+ */ -+#ifdef CONFIG_PREEMPT_RT_FULL -+# define spin_lock_local(lock) rt_spin_lock__no_mg(lock) -+# define spin_trylock_local(lock) rt_spin_trylock__no_mg(lock) -+# define spin_unlock_local(lock) rt_spin_unlock__no_mg(lock) -+#else -+# define spin_lock_local(lock) spin_lock(lock) -+# define spin_trylock_local(lock) spin_trylock(lock) -+# define spin_unlock_local(lock) spin_unlock(lock) -+#endif -+ -+static inline void __local_lock(struct local_irq_lock *lv) -+{ -+ if (lv->owner != current) { -+ spin_lock_local(&lv->lock); -+ LL_WARN(lv->owner); -+ LL_WARN(lv->nestcnt); -+ lv->owner = current; -+ } -+ lv->nestcnt++; -+} -+ -+#define local_lock(lvar) \ -+ do { __local_lock(&get_local_var(lvar)); } while (0) -+ -+#define local_lock_on(lvar, cpu) \ -+ do { __local_lock(&per_cpu(lvar, cpu)); } while (0) -+ -+static inline int __local_trylock(struct local_irq_lock *lv) -+{ -+ if (lv->owner != current && spin_trylock_local(&lv->lock)) { -+ LL_WARN(lv->owner); -+ LL_WARN(lv->nestcnt); -+ lv->owner = current; -+ lv->nestcnt = 1; -+ return 1; -+ } -+ return 0; -+} -+ -+#define local_trylock(lvar) \ -+ ({ \ -+ int __locked; \ -+ __locked = __local_trylock(&get_local_var(lvar)); \ -+ if (!__locked) \ -+ put_local_var(lvar); \ -+ __locked; \ -+ }) -+ -+static inline void __local_unlock(struct local_irq_lock *lv) -+{ -+ LL_WARN(lv->nestcnt == 0); -+ LL_WARN(lv->owner != current); -+ if (--lv->nestcnt) -+ return; -+ -+ lv->owner = NULL; -+ spin_unlock_local(&lv->lock); -+} -+ -+#define local_unlock(lvar) \ -+ do { \ -+ __local_unlock(this_cpu_ptr(&lvar)); \ -+ put_local_var(lvar); \ -+ } while (0) -+ -+#define local_unlock_on(lvar, cpu) \ -+ do { __local_unlock(&per_cpu(lvar, cpu)); } while (0) -+ -+static inline void __local_lock_irq(struct local_irq_lock *lv) -+{ -+ spin_lock_irqsave(&lv->lock, lv->flags); -+ LL_WARN(lv->owner); -+ LL_WARN(lv->nestcnt); -+ lv->owner = current; -+ lv->nestcnt = 1; -+} -+ -+#define local_lock_irq(lvar) \ -+ do { __local_lock_irq(&get_local_var(lvar)); } while (0) -+ -+#define local_lock_irq_on(lvar, cpu) \ -+ do { __local_lock_irq(&per_cpu(lvar, cpu)); } while (0) -+ -+static inline void __local_unlock_irq(struct local_irq_lock *lv) -+{ -+ LL_WARN(!lv->nestcnt); -+ LL_WARN(lv->owner != current); -+ lv->owner = NULL; -+ lv->nestcnt = 0; -+ spin_unlock_irq(&lv->lock); -+} -+ -+#define local_unlock_irq(lvar) \ -+ do { \ -+ __local_unlock_irq(this_cpu_ptr(&lvar)); \ -+ put_local_var(lvar); \ -+ } while (0) -+ -+#define local_unlock_irq_on(lvar, cpu) \ -+ do { \ -+ __local_unlock_irq(&per_cpu(lvar, cpu)); \ -+ } while (0) -+ -+static inline int __local_lock_irqsave(struct local_irq_lock *lv) -+{ -+ if (lv->owner != current) { -+ __local_lock_irq(lv); -+ return 0; -+ } else { -+ lv->nestcnt++; -+ return 1; -+ } -+} -+ -+#define local_lock_irqsave(lvar, _flags) \ -+ do { \ -+ if (__local_lock_irqsave(&get_local_var(lvar))) \ -+ put_local_var(lvar); \ -+ _flags = __this_cpu_read(lvar.flags); \ -+ } while (0) -+ -+#define local_lock_irqsave_on(lvar, _flags, cpu) \ -+ do { \ -+ __local_lock_irqsave(&per_cpu(lvar, cpu)); \ -+ _flags = per_cpu(lvar, cpu).flags; \ -+ } while (0) -+ -+static inline int __local_unlock_irqrestore(struct local_irq_lock *lv, -+ unsigned long flags) -+{ -+ LL_WARN(!lv->nestcnt); -+ LL_WARN(lv->owner != current); -+ if (--lv->nestcnt) -+ return 0; -+ -+ lv->owner = NULL; -+ spin_unlock_irqrestore(&lv->lock, lv->flags); -+ return 1; -+} -+ -+#define local_unlock_irqrestore(lvar, flags) \ -+ do { \ -+ if (__local_unlock_irqrestore(this_cpu_ptr(&lvar), flags)) \ -+ put_local_var(lvar); \ -+ } while (0) -+ -+#define local_unlock_irqrestore_on(lvar, flags, cpu) \ -+ do { \ -+ __local_unlock_irqrestore(&per_cpu(lvar, cpu), flags); \ -+ } while (0) -+ -+#define local_spin_trylock_irq(lvar, lock) \ -+ ({ \ -+ int __locked; \ -+ local_lock_irq(lvar); \ -+ __locked = spin_trylock(lock); \ -+ if (!__locked) \ -+ local_unlock_irq(lvar); \ -+ __locked; \ -+ }) -+ -+#define local_spin_lock_irq(lvar, lock) \ -+ do { \ -+ local_lock_irq(lvar); \ -+ spin_lock(lock); \ -+ } while (0) -+ -+#define local_spin_unlock_irq(lvar, lock) \ -+ do { \ -+ spin_unlock(lock); \ -+ local_unlock_irq(lvar); \ -+ } while (0) -+ -+#define local_spin_lock_irqsave(lvar, lock, flags) \ -+ do { \ -+ local_lock_irqsave(lvar, flags); \ -+ spin_lock(lock); \ -+ } while (0) -+ -+#define local_spin_unlock_irqrestore(lvar, lock, flags) \ -+ do { \ -+ spin_unlock(lock); \ -+ local_unlock_irqrestore(lvar, flags); \ -+ } while (0) -+ -+#define get_locked_var(lvar, var) \ -+ (*({ \ -+ local_lock(lvar); \ -+ this_cpu_ptr(&var); \ -+ })) -+ -+#define put_locked_var(lvar, var) local_unlock(lvar); -+ -+#define local_lock_cpu(lvar) \ -+ ({ \ -+ local_lock(lvar); \ -+ smp_processor_id(); \ -+ }) -+ -+#define local_unlock_cpu(lvar) local_unlock(lvar) -+ -+#else /* PREEMPT_RT_BASE */ -+ -+#define DEFINE_LOCAL_IRQ_LOCK(lvar) __typeof__(const int) lvar -+#define DECLARE_LOCAL_IRQ_LOCK(lvar) extern __typeof__(const int) lvar -+ -+static inline void local_irq_lock_init(int lvar) { } -+ -+#define local_lock(lvar) preempt_disable() -+#define local_unlock(lvar) preempt_enable() -+#define local_lock_irq(lvar) local_irq_disable() -+#define local_lock_irq_on(lvar, cpu) local_irq_disable() -+#define local_unlock_irq(lvar) local_irq_enable() -+#define local_unlock_irq_on(lvar, cpu) local_irq_enable() -+#define local_lock_irqsave(lvar, flags) local_irq_save(flags) -+#define local_unlock_irqrestore(lvar, flags) local_irq_restore(flags) -+ -+#define local_spin_trylock_irq(lvar, lock) spin_trylock_irq(lock) -+#define local_spin_lock_irq(lvar, lock) spin_lock_irq(lock) -+#define local_spin_unlock_irq(lvar, lock) spin_unlock_irq(lock) -+#define local_spin_lock_irqsave(lvar, lock, flags) \ -+ spin_lock_irqsave(lock, flags) -+#define local_spin_unlock_irqrestore(lvar, lock, flags) \ -+ spin_unlock_irqrestore(lock, flags) -+ -+#define get_locked_var(lvar, var) get_cpu_var(var) -+#define put_locked_var(lvar, var) put_cpu_var(var) -+ -+#define local_lock_cpu(lvar) get_cpu() -+#define local_unlock_cpu(lvar) put_cpu() -+ -+#endif -+ -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/mm_types.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/mm_types.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/mm_types.h 2017-04-16 10:38:25.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/mm_types.h 2017-04-18 17:54:26.000000000 +0200 -@@ -11,6 +11,7 @@ - #include - #include - #include -+#include - #include - #include - #include -@@ -509,6 +510,9 @@ - bool tlb_flush_pending; - #endif - struct uprobes_state uprobes_state; -+#ifdef CONFIG_PREEMPT_RT_BASE -+ struct rcu_head delayed_drop; -+#endif - #ifdef CONFIG_X86_INTEL_MPX - /* address of the bounds directory */ - void __user *bd_addr; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/module.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/module.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/module.h 2017-04-16 10:38:25.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/module.h 2017-04-18 17:54:26.000000000 +0200 -@@ -496,6 +496,7 @@ - struct module *__module_text_address(unsigned long addr); - struct module *__module_address(unsigned long addr); - bool is_module_address(unsigned long addr); -+bool __is_module_percpu_address(unsigned long addr, unsigned long *can_addr); - bool is_module_percpu_address(unsigned long addr); - bool is_module_text_address(unsigned long addr); - -@@ -663,6 +664,11 @@ - return false; - } - -+static inline bool __is_module_percpu_address(unsigned long addr, unsigned long *can_addr) -+{ -+ return false; -+} -+ - static inline bool is_module_text_address(unsigned long addr) - { - return false; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/mutex.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/mutex.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/mutex.h 2017-04-16 10:38:25.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/mutex.h 2017-04-18 17:54:26.000000000 +0200 -@@ -19,6 +19,17 @@ - #include - #include - -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+# define __DEP_MAP_MUTEX_INITIALIZER(lockname) \ -+ , .dep_map = { .name = #lockname } -+#else -+# define __DEP_MAP_MUTEX_INITIALIZER(lockname) -+#endif -+ -+#ifdef CONFIG_PREEMPT_RT_FULL -+# include -+#else -+ - /* - * Simple, straightforward mutexes with strict semantics: - * -@@ -99,13 +110,6 @@ - static inline void mutex_destroy(struct mutex *lock) {} - #endif - --#ifdef CONFIG_DEBUG_LOCK_ALLOC --# define __DEP_MAP_MUTEX_INITIALIZER(lockname) \ -- , .dep_map = { .name = #lockname } --#else --# define __DEP_MAP_MUTEX_INITIALIZER(lockname) --#endif -- - #define __MUTEX_INITIALIZER(lockname) \ - { .count = ATOMIC_INIT(1) \ - , .wait_lock = __SPIN_LOCK_UNLOCKED(lockname.wait_lock) \ -@@ -173,6 +177,8 @@ - extern int mutex_trylock(struct mutex *lock); - extern void mutex_unlock(struct mutex *lock); - -+#endif /* !PREEMPT_RT_FULL */ -+ - extern int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock); - - #endif /* __LINUX_MUTEX_H */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/mutex_rt.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/mutex_rt.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/mutex_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/mutex_rt.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,89 @@ -+#ifndef __LINUX_MUTEX_RT_H -+#define __LINUX_MUTEX_RT_H -+ -+#ifndef __LINUX_MUTEX_H -+#error "Please include mutex.h" -+#endif -+ -+#include -+ -+/* FIXME: Just for __lockfunc */ -+#include -+ -+struct mutex { -+ struct rt_mutex lock; -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+ struct lockdep_map dep_map; -+#endif -+}; -+ -+#define __MUTEX_INITIALIZER(mutexname) \ -+ { \ -+ .lock = __RT_MUTEX_INITIALIZER(mutexname.lock) \ -+ __DEP_MAP_MUTEX_INITIALIZER(mutexname) \ -+ } -+ -+#define DEFINE_MUTEX(mutexname) \ -+ struct mutex mutexname = __MUTEX_INITIALIZER(mutexname) -+ -+extern void __mutex_do_init(struct mutex *lock, const char *name, struct lock_class_key *key); -+extern void __lockfunc _mutex_lock(struct mutex *lock); -+extern int __lockfunc _mutex_lock_interruptible(struct mutex *lock); -+extern int __lockfunc _mutex_lock_killable(struct mutex *lock); -+extern void __lockfunc _mutex_lock_nested(struct mutex *lock, int subclass); -+extern void __lockfunc _mutex_lock_nest_lock(struct mutex *lock, struct lockdep_map *nest_lock); -+extern int __lockfunc _mutex_lock_interruptible_nested(struct mutex *lock, int subclass); -+extern int __lockfunc _mutex_lock_killable_nested(struct mutex *lock, int subclass); -+extern int __lockfunc _mutex_trylock(struct mutex *lock); -+extern void __lockfunc _mutex_unlock(struct mutex *lock); -+ -+#define mutex_is_locked(l) rt_mutex_is_locked(&(l)->lock) -+#define mutex_lock(l) _mutex_lock(l) -+#define mutex_lock_interruptible(l) _mutex_lock_interruptible(l) -+#define mutex_lock_killable(l) _mutex_lock_killable(l) -+#define mutex_trylock(l) _mutex_trylock(l) -+#define mutex_unlock(l) _mutex_unlock(l) -+ -+#ifdef CONFIG_DEBUG_MUTEXES -+#define mutex_destroy(l) rt_mutex_destroy(&(l)->lock) -+#else -+static inline void mutex_destroy(struct mutex *lock) {} -+#endif -+ -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+# define mutex_lock_nested(l, s) _mutex_lock_nested(l, s) -+# define mutex_lock_interruptible_nested(l, s) \ -+ _mutex_lock_interruptible_nested(l, s) -+# define mutex_lock_killable_nested(l, s) \ -+ _mutex_lock_killable_nested(l, s) -+ -+# define mutex_lock_nest_lock(lock, nest_lock) \ -+do { \ -+ typecheck(struct lockdep_map *, &(nest_lock)->dep_map); \ -+ _mutex_lock_nest_lock(lock, &(nest_lock)->dep_map); \ -+} while (0) -+ -+#else -+# define mutex_lock_nested(l, s) _mutex_lock(l) -+# define mutex_lock_interruptible_nested(l, s) \ -+ _mutex_lock_interruptible(l) -+# define mutex_lock_killable_nested(l, s) \ -+ _mutex_lock_killable(l) -+# define mutex_lock_nest_lock(lock, nest_lock) mutex_lock(lock) -+#endif -+ -+# define mutex_init(mutex) \ -+do { \ -+ static struct lock_class_key __key; \ -+ \ -+ rt_mutex_init(&(mutex)->lock); \ -+ __mutex_do_init((mutex), #mutex, &__key); \ -+} while (0) -+ -+# define __mutex_init(mutex, name, key) \ -+do { \ -+ rt_mutex_init(&(mutex)->lock); \ -+ __mutex_do_init((mutex), name, key); \ -+} while (0) -+ -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/netdevice.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/netdevice.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/netdevice.h 2017-04-16 10:38:25.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/netdevice.h 2017-04-18 17:54:26.000000000 +0200 -@@ -396,7 +396,19 @@ - typedef rx_handler_result_t rx_handler_func_t(struct sk_buff **pskb); - - void __napi_schedule(struct napi_struct *n); -+ -+/* -+ * When PREEMPT_RT_FULL is defined, all device interrupt handlers -+ * run as threads, and they can also be preempted (without PREEMPT_RT -+ * interrupt threads can not be preempted). Which means that calling -+ * __napi_schedule_irqoff() from an interrupt handler can be preempted -+ * and can corrupt the napi->poll_list. -+ */ -+#ifdef CONFIG_PREEMPT_RT_FULL -+#define __napi_schedule_irqoff(n) __napi_schedule(n) -+#else - void __napi_schedule_irqoff(struct napi_struct *n); -+#endif - - static inline bool napi_disable_pending(struct napi_struct *n) - { -@@ -2463,14 +2475,53 @@ - void synchronize_net(void); - int init_dummy_netdev(struct net_device *dev); - --DECLARE_PER_CPU(int, xmit_recursion); - #define XMIT_RECURSION_LIMIT 10 -+#ifdef CONFIG_PREEMPT_RT_FULL -+static inline int dev_recursion_level(void) -+{ -+ return current->xmit_recursion; -+} -+ -+static inline int xmit_rec_read(void) -+{ -+ return current->xmit_recursion; -+} -+ -+static inline void xmit_rec_inc(void) -+{ -+ current->xmit_recursion++; -+} -+ -+static inline void xmit_rec_dec(void) -+{ -+ current->xmit_recursion--; -+} -+ -+#else -+ -+DECLARE_PER_CPU(int, xmit_recursion); - - static inline int dev_recursion_level(void) - { - return this_cpu_read(xmit_recursion); - } - -+static inline int xmit_rec_read(void) -+{ -+ return __this_cpu_read(xmit_recursion); -+} -+ -+static inline void xmit_rec_inc(void) -+{ -+ __this_cpu_inc(xmit_recursion); -+} -+ -+static inline void xmit_rec_dec(void) -+{ -+ __this_cpu_dec(xmit_recursion); -+} -+#endif -+ - struct net_device *dev_get_by_index(struct net *net, int ifindex); - struct net_device *__dev_get_by_index(struct net *net, int ifindex); - struct net_device *dev_get_by_index_rcu(struct net *net, int ifindex); -@@ -2855,6 +2906,7 @@ - unsigned int dropped; - struct sk_buff_head input_pkt_queue; - struct napi_struct backlog; -+ struct sk_buff_head tofree_queue; - - }; - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/netfilter/x_tables.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/netfilter/x_tables.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/netfilter/x_tables.h 2017-04-16 10:38:25.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/netfilter/x_tables.h 2017-04-18 17:54:26.000000000 +0200 -@@ -4,6 +4,7 @@ - - #include - #include -+#include - #include - - /* Test a struct->invflags and a boolean for inequality */ -@@ -300,6 +301,8 @@ - */ - DECLARE_PER_CPU(seqcount_t, xt_recseq); - -+DECLARE_LOCAL_IRQ_LOCK(xt_write_lock); -+ - /* xt_tee_enabled - true if x_tables needs to handle reentrancy - * - * Enabled if current ip(6)tables ruleset has at least one -j TEE rule. -@@ -320,6 +323,9 @@ - { - unsigned int addend; - -+ /* RT protection */ -+ local_lock(xt_write_lock); -+ - /* - * Low order bit of sequence is set if we already - * called xt_write_recseq_begin(). -@@ -350,6 +356,7 @@ - /* this is kind of a write_seqcount_end(), but addend is 0 or 1 */ - smp_wmb(); - __this_cpu_add(xt_recseq.sequence, addend); -+ local_unlock(xt_write_lock); - } - - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/nfs_fs.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/nfs_fs.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/nfs_fs.h 2017-04-16 10:38:25.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/nfs_fs.h 2017-04-18 17:54:26.000000000 +0200 -@@ -165,7 +165,11 @@ - - /* Readers: in-flight sillydelete RPC calls */ - /* Writers: rmdir */ -+#ifdef CONFIG_PREEMPT_RT_BASE -+ struct semaphore rmdir_sem; -+#else - struct rw_semaphore rmdir_sem; -+#endif - - #if IS_ENABLED(CONFIG_NFS_V4) - struct nfs4_cached_acl *nfs4_acl; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/nfs_xdr.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/nfs_xdr.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/nfs_xdr.h 2017-04-16 10:38:25.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/nfs_xdr.h 2017-04-18 17:54:26.000000000 +0200 -@@ -1490,7 +1490,7 @@ - struct nfs_removeargs args; - struct nfs_removeres res; - struct dentry *dentry; -- wait_queue_head_t wq; -+ struct swait_queue_head wq; - struct rpc_cred *cred; - struct nfs_fattr dir_attr; - long timeout; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/notifier.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/notifier.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/notifier.h 2017-04-16 10:38:25.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/notifier.h 2017-04-18 17:54:26.000000000 +0200 -@@ -6,7 +6,7 @@ - * - * Alan Cox - */ -- -+ - #ifndef _LINUX_NOTIFIER_H - #define _LINUX_NOTIFIER_H - #include -@@ -42,9 +42,7 @@ - * in srcu_notifier_call_chain(): no cache bounces and no memory barriers. - * As compensation, srcu_notifier_chain_unregister() is rather expensive. - * SRCU notifier chains should be used when the chain will be called very -- * often but notifier_blocks will seldom be removed. Also, SRCU notifier -- * chains are slightly more difficult to use because they require special -- * runtime initialization. -+ * often but notifier_blocks will seldom be removed. - */ - - struct notifier_block; -@@ -90,7 +88,7 @@ - (name)->head = NULL; \ - } while (0) - --/* srcu_notifier_heads must be initialized and cleaned up dynamically */ -+/* srcu_notifier_heads must be cleaned up dynamically */ - extern void srcu_init_notifier_head(struct srcu_notifier_head *nh); - #define srcu_cleanup_notifier_head(name) \ - cleanup_srcu_struct(&(name)->srcu); -@@ -103,7 +101,13 @@ - .head = NULL } - #define RAW_NOTIFIER_INIT(name) { \ - .head = NULL } --/* srcu_notifier_heads cannot be initialized statically */ -+ -+#define SRCU_NOTIFIER_INIT(name, pcpu) \ -+ { \ -+ .mutex = __MUTEX_INITIALIZER(name.mutex), \ -+ .head = NULL, \ -+ .srcu = __SRCU_STRUCT_INIT(name.srcu, pcpu), \ -+ } - - #define ATOMIC_NOTIFIER_HEAD(name) \ - struct atomic_notifier_head name = \ -@@ -115,6 +119,18 @@ - struct raw_notifier_head name = \ - RAW_NOTIFIER_INIT(name) - -+#define _SRCU_NOTIFIER_HEAD(name, mod) \ -+ static DEFINE_PER_CPU(struct srcu_struct_array, \ -+ name##_head_srcu_array); \ -+ mod struct srcu_notifier_head name = \ -+ SRCU_NOTIFIER_INIT(name, name##_head_srcu_array) -+ -+#define SRCU_NOTIFIER_HEAD(name) \ -+ _SRCU_NOTIFIER_HEAD(name, ) -+ -+#define SRCU_NOTIFIER_HEAD_STATIC(name) \ -+ _SRCU_NOTIFIER_HEAD(name, static) -+ - #ifdef __KERNEL__ - - extern int atomic_notifier_chain_register(struct atomic_notifier_head *nh, -@@ -184,12 +200,12 @@ - - /* - * Declared notifiers so far. I can imagine quite a few more chains -- * over time (eg laptop power reset chains, reboot chain (to clean -+ * over time (eg laptop power reset chains, reboot chain (to clean - * device units up), device [un]mount chain, module load/unload chain, -- * low memory chain, screenblank chain (for plug in modular screenblankers) -+ * low memory chain, screenblank chain (for plug in modular screenblankers) - * VC switch chains (for loadable kernel svgalib VC switch helpers) etc... - */ -- -+ - /* CPU notfiers are defined in include/linux/cpu.h. */ - - /* netdevice notifiers are defined in include/linux/netdevice.h */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/percpu-rwsem.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/percpu-rwsem.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/percpu-rwsem.h 2017-04-16 10:38:25.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/percpu-rwsem.h 2017-04-18 17:54:26.000000000 +0200 -@@ -4,7 +4,7 @@ - #include - #include - #include --#include -+#include - #include - #include - -@@ -12,7 +12,7 @@ - struct rcu_sync rss; - unsigned int __percpu *read_count; - struct rw_semaphore rw_sem; -- wait_queue_head_t writer; -+ struct swait_queue_head writer; - int readers_block; - }; - -@@ -22,13 +22,13 @@ - .rss = __RCU_SYNC_INITIALIZER(name.rss, RCU_SCHED_SYNC), \ - .read_count = &__percpu_rwsem_rc_##name, \ - .rw_sem = __RWSEM_INITIALIZER(name.rw_sem), \ -- .writer = __WAIT_QUEUE_HEAD_INITIALIZER(name.writer), \ -+ .writer = __SWAIT_QUEUE_HEAD_INITIALIZER(name.writer), \ - } - - extern int __percpu_down_read(struct percpu_rw_semaphore *, int); - extern void __percpu_up_read(struct percpu_rw_semaphore *); - --static inline void percpu_down_read_preempt_disable(struct percpu_rw_semaphore *sem) -+static inline void percpu_down_read(struct percpu_rw_semaphore *sem) - { - might_sleep(); - -@@ -46,16 +46,10 @@ - __this_cpu_inc(*sem->read_count); - if (unlikely(!rcu_sync_is_idle(&sem->rss))) - __percpu_down_read(sem, false); /* Unconditional memory barrier */ -- barrier(); - /* -- * The barrier() prevents the compiler from -+ * The preempt_enable() prevents the compiler from - * bleeding the critical section out. - */ --} -- --static inline void percpu_down_read(struct percpu_rw_semaphore *sem) --{ -- percpu_down_read_preempt_disable(sem); - preempt_enable(); - } - -@@ -82,13 +76,9 @@ - return ret; - } - --static inline void percpu_up_read_preempt_enable(struct percpu_rw_semaphore *sem) -+static inline void percpu_up_read(struct percpu_rw_semaphore *sem) - { -- /* -- * The barrier() prevents the compiler from -- * bleeding the critical section out. -- */ -- barrier(); -+ preempt_disable(); - /* - * Same as in percpu_down_read(). - */ -@@ -101,12 +91,6 @@ - rwsem_release(&sem->rw_sem.dep_map, 1, _RET_IP_); - } - --static inline void percpu_up_read(struct percpu_rw_semaphore *sem) --{ -- preempt_disable(); -- percpu_up_read_preempt_enable(sem); --} -- - extern void percpu_down_write(struct percpu_rw_semaphore *); - extern void percpu_up_write(struct percpu_rw_semaphore *); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/percpu.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/percpu.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/percpu.h 2017-04-16 10:38:25.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/percpu.h 2017-04-18 17:54:26.000000000 +0200 -@@ -18,6 +18,35 @@ - #define PERCPU_MODULE_RESERVE 0 - #endif - -+#ifdef CONFIG_PREEMPT_RT_FULL -+ -+#define get_local_var(var) (*({ \ -+ migrate_disable(); \ -+ this_cpu_ptr(&var); })) -+ -+#define put_local_var(var) do { \ -+ (void)&(var); \ -+ migrate_enable(); \ -+} while (0) -+ -+# define get_local_ptr(var) ({ \ -+ migrate_disable(); \ -+ this_cpu_ptr(var); }) -+ -+# define put_local_ptr(var) do { \ -+ (void)(var); \ -+ migrate_enable(); \ -+} while (0) -+ -+#else -+ -+#define get_local_var(var) get_cpu_var(var) -+#define put_local_var(var) put_cpu_var(var) -+#define get_local_ptr(var) get_cpu_ptr(var) -+#define put_local_ptr(var) put_cpu_ptr(var) -+ -+#endif -+ - /* minimum unit size, also is the maximum supported allocation size */ - #define PCPU_MIN_UNIT_SIZE PFN_ALIGN(32 << 10) - -@@ -110,6 +139,7 @@ - #endif - - extern void __percpu *__alloc_reserved_percpu(size_t size, size_t align); -+extern bool __is_kernel_percpu_address(unsigned long addr, unsigned long *can_addr); - extern bool is_kernel_percpu_address(unsigned long addr); - - #if !defined(CONFIG_SMP) || !defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/pid.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/pid.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/pid.h 2017-04-16 10:38:25.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/pid.h 2017-04-18 17:54:26.000000000 +0200 -@@ -2,6 +2,7 @@ - #define _LINUX_PID_H - - #include -+#include - - enum pid_type - { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/preempt.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/preempt.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/preempt.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/preempt.h 2017-04-18 17:54:26.000000000 +0200 -@@ -50,7 +50,11 @@ - #define HARDIRQ_OFFSET (1UL << HARDIRQ_SHIFT) - #define NMI_OFFSET (1UL << NMI_SHIFT) - --#define SOFTIRQ_DISABLE_OFFSET (2 * SOFTIRQ_OFFSET) -+#ifndef CONFIG_PREEMPT_RT_FULL -+# define SOFTIRQ_DISABLE_OFFSET (2 * SOFTIRQ_OFFSET) -+#else -+# define SOFTIRQ_DISABLE_OFFSET (0) -+#endif - - /* We use the MSB mostly because its available */ - #define PREEMPT_NEED_RESCHED 0x80000000 -@@ -59,9 +63,15 @@ - #include - - #define hardirq_count() (preempt_count() & HARDIRQ_MASK) --#define softirq_count() (preempt_count() & SOFTIRQ_MASK) - #define irq_count() (preempt_count() & (HARDIRQ_MASK | SOFTIRQ_MASK \ - | NMI_MASK)) -+#ifndef CONFIG_PREEMPT_RT_FULL -+# define softirq_count() (preempt_count() & SOFTIRQ_MASK) -+# define in_serving_softirq() (softirq_count() & SOFTIRQ_OFFSET) -+#else -+# define softirq_count() (0UL) -+extern int in_serving_softirq(void); -+#endif - - /* - * Are we doing bottom half or hardware interrupt processing? -@@ -72,7 +82,6 @@ - #define in_irq() (hardirq_count()) - #define in_softirq() (softirq_count()) - #define in_interrupt() (irq_count()) --#define in_serving_softirq() (softirq_count() & SOFTIRQ_OFFSET) - - /* - * Are we in NMI context? -@@ -91,7 +100,11 @@ - /* - * The preempt_count offset after spin_lock() - */ -+#if !defined(CONFIG_PREEMPT_RT_FULL) - #define PREEMPT_LOCK_OFFSET PREEMPT_DISABLE_OFFSET -+#else -+#define PREEMPT_LOCK_OFFSET 0 -+#endif - - /* - * The preempt_count offset needed for things like: -@@ -140,6 +153,20 @@ - #define preempt_count_inc() preempt_count_add(1) - #define preempt_count_dec() preempt_count_sub(1) - -+#ifdef CONFIG_PREEMPT_LAZY -+#define add_preempt_lazy_count(val) do { preempt_lazy_count() += (val); } while (0) -+#define sub_preempt_lazy_count(val) do { preempt_lazy_count() -= (val); } while (0) -+#define inc_preempt_lazy_count() add_preempt_lazy_count(1) -+#define dec_preempt_lazy_count() sub_preempt_lazy_count(1) -+#define preempt_lazy_count() (current_thread_info()->preempt_lazy_count) -+#else -+#define add_preempt_lazy_count(val) do { } while (0) -+#define sub_preempt_lazy_count(val) do { } while (0) -+#define inc_preempt_lazy_count() do { } while (0) -+#define dec_preempt_lazy_count() do { } while (0) -+#define preempt_lazy_count() (0) -+#endif -+ - #ifdef CONFIG_PREEMPT_COUNT - - #define preempt_disable() \ -@@ -148,13 +175,25 @@ - barrier(); \ - } while (0) - -+#define preempt_lazy_disable() \ -+do { \ -+ inc_preempt_lazy_count(); \ -+ barrier(); \ -+} while (0) -+ - #define sched_preempt_enable_no_resched() \ - do { \ - barrier(); \ - preempt_count_dec(); \ - } while (0) - --#define preempt_enable_no_resched() sched_preempt_enable_no_resched() -+#ifdef CONFIG_PREEMPT_RT_BASE -+# define preempt_enable_no_resched() sched_preempt_enable_no_resched() -+# define preempt_check_resched_rt() preempt_check_resched() -+#else -+# define preempt_enable_no_resched() preempt_enable() -+# define preempt_check_resched_rt() barrier(); -+#endif - - #define preemptible() (preempt_count() == 0 && !irqs_disabled()) - -@@ -179,6 +218,13 @@ - __preempt_schedule(); \ - } while (0) - -+#define preempt_lazy_enable() \ -+do { \ -+ dec_preempt_lazy_count(); \ -+ barrier(); \ -+ preempt_check_resched(); \ -+} while (0) -+ - #else /* !CONFIG_PREEMPT */ - #define preempt_enable() \ - do { \ -@@ -224,6 +270,7 @@ - #define preempt_disable_notrace() barrier() - #define preempt_enable_no_resched_notrace() barrier() - #define preempt_enable_notrace() barrier() -+#define preempt_check_resched_rt() barrier() - #define preemptible() 0 - - #endif /* CONFIG_PREEMPT_COUNT */ -@@ -244,10 +291,31 @@ - } while (0) - #define preempt_fold_need_resched() \ - do { \ -- if (tif_need_resched()) \ -+ if (tif_need_resched_now()) \ - set_preempt_need_resched(); \ - } while (0) - -+#ifdef CONFIG_PREEMPT_RT_FULL -+# define preempt_disable_rt() preempt_disable() -+# define preempt_enable_rt() preempt_enable() -+# define preempt_disable_nort() barrier() -+# define preempt_enable_nort() barrier() -+# ifdef CONFIG_SMP -+ extern void migrate_disable(void); -+ extern void migrate_enable(void); -+# else /* CONFIG_SMP */ -+# define migrate_disable() barrier() -+# define migrate_enable() barrier() -+# endif /* CONFIG_SMP */ -+#else -+# define preempt_disable_rt() barrier() -+# define preempt_enable_rt() barrier() -+# define preempt_disable_nort() preempt_disable() -+# define preempt_enable_nort() preempt_enable() -+# define migrate_disable() preempt_disable() -+# define migrate_enable() preempt_enable() -+#endif -+ - #ifdef CONFIG_PREEMPT_NOTIFIERS - - struct preempt_notifier; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/printk.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/printk.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/printk.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/printk.h 2017-04-18 17:54:26.000000000 +0200 -@@ -126,9 +126,11 @@ - #ifdef CONFIG_EARLY_PRINTK - extern asmlinkage __printf(1, 2) - void early_printk(const char *fmt, ...); -+extern void printk_kill(void); - #else - static inline __printf(1, 2) __cold - void early_printk(const char *s, ...) { } -+static inline void printk_kill(void) { } - #endif - - #ifdef CONFIG_PRINTK_NMI -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/radix-tree.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/radix-tree.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/radix-tree.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/radix-tree.h 2017-04-18 17:54:26.000000000 +0200 -@@ -292,6 +292,8 @@ - int radix_tree_preload(gfp_t gfp_mask); - int radix_tree_maybe_preload(gfp_t gfp_mask); - int radix_tree_maybe_preload_order(gfp_t gfp_mask, int order); -+void radix_tree_preload_end(void); -+ - void radix_tree_init(void); - void *radix_tree_tag_set(struct radix_tree_root *root, - unsigned long index, unsigned int tag); -@@ -314,11 +316,6 @@ - int radix_tree_tagged(struct radix_tree_root *root, unsigned int tag); - unsigned long radix_tree_locate_item(struct radix_tree_root *root, void *item); - --static inline void radix_tree_preload_end(void) --{ -- preempt_enable(); --} -- - /** - * struct radix_tree_iter - radix tree iterator state - * -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/random.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/random.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/random.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/random.h 2017-04-18 17:54:26.000000000 +0200 -@@ -31,7 +31,7 @@ - - extern void add_input_randomness(unsigned int type, unsigned int code, - unsigned int value) __latent_entropy; --extern void add_interrupt_randomness(int irq, int irq_flags) __latent_entropy; -+extern void add_interrupt_randomness(int irq, int irq_flags, __u64 ip) __latent_entropy; - - extern void get_random_bytes(void *buf, int nbytes); - extern int add_random_ready_callback(struct random_ready_callback *rdy); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rbtree.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rbtree.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rbtree.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rbtree.h 2017-04-18 17:54:26.000000000 +0200 -@@ -31,7 +31,7 @@ - - #include - #include --#include -+#include - - struct rb_node { - unsigned long __rb_parent_color; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rbtree_augmented.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rbtree_augmented.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rbtree_augmented.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rbtree_augmented.h 2017-04-18 17:54:26.000000000 +0200 -@@ -26,6 +26,7 @@ - - #include - #include -+#include - - /* - * Please note - only struct rb_augment_callbacks and the prototypes for -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rcu_assign_pointer.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rcu_assign_pointer.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rcu_assign_pointer.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rcu_assign_pointer.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,54 @@ -+#ifndef __LINUX_RCU_ASSIGN_POINTER_H__ -+#define __LINUX_RCU_ASSIGN_POINTER_H__ -+#include -+#include -+ -+/** -+ * RCU_INITIALIZER() - statically initialize an RCU-protected global variable -+ * @v: The value to statically initialize with. -+ */ -+#define RCU_INITIALIZER(v) (typeof(*(v)) __force __rcu *)(v) -+ -+/** -+ * rcu_assign_pointer() - assign to RCU-protected pointer -+ * @p: pointer to assign to -+ * @v: value to assign (publish) -+ * -+ * Assigns the specified value to the specified RCU-protected -+ * pointer, ensuring that any concurrent RCU readers will see -+ * any prior initialization. -+ * -+ * Inserts memory barriers on architectures that require them -+ * (which is most of them), and also prevents the compiler from -+ * reordering the code that initializes the structure after the pointer -+ * assignment. More importantly, this call documents which pointers -+ * will be dereferenced by RCU read-side code. -+ * -+ * In some special cases, you may use RCU_INIT_POINTER() instead -+ * of rcu_assign_pointer(). RCU_INIT_POINTER() is a bit faster due -+ * to the fact that it does not constrain either the CPU or the compiler. -+ * That said, using RCU_INIT_POINTER() when you should have used -+ * rcu_assign_pointer() is a very bad thing that results in -+ * impossible-to-diagnose memory corruption. So please be careful. -+ * See the RCU_INIT_POINTER() comment header for details. -+ * -+ * Note that rcu_assign_pointer() evaluates each of its arguments only -+ * once, appearances notwithstanding. One of the "extra" evaluations -+ * is in typeof() and the other visible only to sparse (__CHECKER__), -+ * neither of which actually execute the argument. As with most cpp -+ * macros, this execute-arguments-only-once property is important, so -+ * please be careful when making changes to rcu_assign_pointer() and the -+ * other macros that it invokes. -+ */ -+#define rcu_assign_pointer(p, v) \ -+({ \ -+ uintptr_t _r_a_p__v = (uintptr_t)(v); \ -+ \ -+ if (__builtin_constant_p(v) && (_r_a_p__v) == (uintptr_t)NULL) \ -+ WRITE_ONCE((p), (typeof(p))(_r_a_p__v)); \ -+ else \ -+ smp_store_release(&p, RCU_INITIALIZER((typeof(p))_r_a_p__v)); \ -+ _r_a_p__v; \ -+}) -+ -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rcupdate.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rcupdate.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rcupdate.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rcupdate.h 2017-04-18 17:54:26.000000000 +0200 -@@ -46,6 +46,7 @@ - #include - #include - #include -+#include - - #include - -@@ -178,6 +179,9 @@ - - #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ - -+#ifdef CONFIG_PREEMPT_RT_FULL -+#define call_rcu_bh call_rcu -+#else - /** - * call_rcu_bh() - Queue an RCU for invocation after a quicker grace period. - * @head: structure to be used for queueing the RCU updates. -@@ -201,6 +205,7 @@ - */ - void call_rcu_bh(struct rcu_head *head, - rcu_callback_t func); -+#endif - - /** - * call_rcu_sched() - Queue an RCU for invocation after sched grace period. -@@ -301,6 +306,11 @@ - * types of kernel builds, the rcu_read_lock() nesting depth is unknowable. - */ - #define rcu_preempt_depth() (current->rcu_read_lock_nesting) -+#ifndef CONFIG_PREEMPT_RT_FULL -+#define sched_rcu_preempt_depth() rcu_preempt_depth() -+#else -+static inline int sched_rcu_preempt_depth(void) { return 0; } -+#endif - - #else /* #ifdef CONFIG_PREEMPT_RCU */ - -@@ -326,6 +336,8 @@ - return 0; - } - -+#define sched_rcu_preempt_depth() rcu_preempt_depth() -+ - #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ - - /* Internal to kernel */ -@@ -505,7 +517,14 @@ - int debug_lockdep_rcu_enabled(void); - - int rcu_read_lock_held(void); -+#ifdef CONFIG_PREEMPT_RT_FULL -+static inline int rcu_read_lock_bh_held(void) -+{ -+ return rcu_read_lock_held(); -+} -+#else - int rcu_read_lock_bh_held(void); -+#endif - - /** - * rcu_read_lock_sched_held() - might we be in RCU-sched read-side critical section? -@@ -626,54 +645,6 @@ - }) - - /** -- * RCU_INITIALIZER() - statically initialize an RCU-protected global variable -- * @v: The value to statically initialize with. -- */ --#define RCU_INITIALIZER(v) (typeof(*(v)) __force __rcu *)(v) -- --/** -- * rcu_assign_pointer() - assign to RCU-protected pointer -- * @p: pointer to assign to -- * @v: value to assign (publish) -- * -- * Assigns the specified value to the specified RCU-protected -- * pointer, ensuring that any concurrent RCU readers will see -- * any prior initialization. -- * -- * Inserts memory barriers on architectures that require them -- * (which is most of them), and also prevents the compiler from -- * reordering the code that initializes the structure after the pointer -- * assignment. More importantly, this call documents which pointers -- * will be dereferenced by RCU read-side code. -- * -- * In some special cases, you may use RCU_INIT_POINTER() instead -- * of rcu_assign_pointer(). RCU_INIT_POINTER() is a bit faster due -- * to the fact that it does not constrain either the CPU or the compiler. -- * That said, using RCU_INIT_POINTER() when you should have used -- * rcu_assign_pointer() is a very bad thing that results in -- * impossible-to-diagnose memory corruption. So please be careful. -- * See the RCU_INIT_POINTER() comment header for details. -- * -- * Note that rcu_assign_pointer() evaluates each of its arguments only -- * once, appearances notwithstanding. One of the "extra" evaluations -- * is in typeof() and the other visible only to sparse (__CHECKER__), -- * neither of which actually execute the argument. As with most cpp -- * macros, this execute-arguments-only-once property is important, so -- * please be careful when making changes to rcu_assign_pointer() and the -- * other macros that it invokes. -- */ --#define rcu_assign_pointer(p, v) \ --({ \ -- uintptr_t _r_a_p__v = (uintptr_t)(v); \ -- \ -- if (__builtin_constant_p(v) && (_r_a_p__v) == (uintptr_t)NULL) \ -- WRITE_ONCE((p), (typeof(p))(_r_a_p__v)); \ -- else \ -- smp_store_release(&p, RCU_INITIALIZER((typeof(p))_r_a_p__v)); \ -- _r_a_p__v; \ --}) -- --/** - * rcu_access_pointer() - fetch RCU pointer with no dereferencing - * @p: The pointer to read - * -@@ -951,10 +922,14 @@ - static inline void rcu_read_lock_bh(void) - { - local_bh_disable(); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ rcu_read_lock(); -+#else - __acquire(RCU_BH); - rcu_lock_acquire(&rcu_bh_lock_map); - RCU_LOCKDEP_WARN(!rcu_is_watching(), - "rcu_read_lock_bh() used illegally while idle"); -+#endif - } - - /* -@@ -964,10 +939,14 @@ - */ - static inline void rcu_read_unlock_bh(void) - { -+#ifdef CONFIG_PREEMPT_RT_FULL -+ rcu_read_unlock(); -+#else - RCU_LOCKDEP_WARN(!rcu_is_watching(), - "rcu_read_unlock_bh() used illegally while idle"); - rcu_lock_release(&rcu_bh_lock_map); - __release(RCU_BH); -+#endif - local_bh_enable(); - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rcutree.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rcutree.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rcutree.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rcutree.h 2017-04-18 17:54:26.000000000 +0200 -@@ -44,7 +44,11 @@ - rcu_note_context_switch(); - } - -+#ifdef CONFIG_PREEMPT_RT_FULL -+# define synchronize_rcu_bh synchronize_rcu -+#else - void synchronize_rcu_bh(void); -+#endif - void synchronize_sched_expedited(void); - void synchronize_rcu_expedited(void); - -@@ -72,7 +76,11 @@ - } - - void rcu_barrier(void); -+#ifdef CONFIG_PREEMPT_RT_FULL -+# define rcu_barrier_bh rcu_barrier -+#else - void rcu_barrier_bh(void); -+#endif - void rcu_barrier_sched(void); - unsigned long get_state_synchronize_rcu(void); - void cond_synchronize_rcu(unsigned long oldstate); -@@ -82,17 +90,14 @@ - extern unsigned long rcutorture_testseq; - extern unsigned long rcutorture_vernum; - unsigned long rcu_batches_started(void); --unsigned long rcu_batches_started_bh(void); - unsigned long rcu_batches_started_sched(void); - unsigned long rcu_batches_completed(void); --unsigned long rcu_batches_completed_bh(void); - unsigned long rcu_batches_completed_sched(void); - unsigned long rcu_exp_batches_completed(void); - unsigned long rcu_exp_batches_completed_sched(void); - void show_rcu_gp_kthreads(void); - - void rcu_force_quiescent_state(void); --void rcu_bh_force_quiescent_state(void); - void rcu_sched_force_quiescent_state(void); - - void rcu_idle_enter(void); -@@ -109,6 +114,16 @@ - - bool rcu_is_watching(void); - -+#ifndef CONFIG_PREEMPT_RT_FULL -+void rcu_bh_force_quiescent_state(void); -+unsigned long rcu_batches_started_bh(void); -+unsigned long rcu_batches_completed_bh(void); -+#else -+# define rcu_bh_force_quiescent_state rcu_force_quiescent_state -+# define rcu_batches_completed_bh rcu_batches_completed -+# define rcu_batches_started_bh rcu_batches_completed -+#endif -+ - void rcu_all_qs(void); - - /* RCUtree hotplug events */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rtmutex.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rtmutex.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rtmutex.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rtmutex.h 2017-04-18 17:54:26.000000000 +0200 -@@ -13,11 +13,15 @@ - #define __LINUX_RT_MUTEX_H - - #include -+#include - #include --#include - - extern int max_lock_depth; /* for sysctl */ - -+#ifdef CONFIG_DEBUG_MUTEXES -+#include -+#endif -+ - /** - * The rt_mutex structure - * -@@ -31,8 +35,8 @@ - struct rb_root waiters; - struct rb_node *waiters_leftmost; - struct task_struct *owner; --#ifdef CONFIG_DEBUG_RT_MUTEXES - int save_state; -+#ifdef CONFIG_DEBUG_RT_MUTEXES - const char *name, *file; - int line; - void *magic; -@@ -55,22 +59,33 @@ - # define rt_mutex_debug_check_no_locks_held(task) do { } while (0) - #endif - -+# define rt_mutex_init(mutex) \ -+ do { \ -+ raw_spin_lock_init(&(mutex)->wait_lock); \ -+ __rt_mutex_init(mutex, #mutex); \ -+ } while (0) -+ - #ifdef CONFIG_DEBUG_RT_MUTEXES - # define __DEBUG_RT_MUTEX_INITIALIZER(mutexname) \ - , .name = #mutexname, .file = __FILE__, .line = __LINE__ --# define rt_mutex_init(mutex) __rt_mutex_init(mutex, __func__) - extern void rt_mutex_debug_task_free(struct task_struct *tsk); - #else - # define __DEBUG_RT_MUTEX_INITIALIZER(mutexname) --# define rt_mutex_init(mutex) __rt_mutex_init(mutex, NULL) - # define rt_mutex_debug_task_free(t) do { } while (0) - #endif - --#define __RT_MUTEX_INITIALIZER(mutexname) \ -- { .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(mutexname.wait_lock) \ -+#define __RT_MUTEX_INITIALIZER_PLAIN(mutexname) \ -+ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(mutexname.wait_lock) \ - , .waiters = RB_ROOT \ - , .owner = NULL \ -- __DEBUG_RT_MUTEX_INITIALIZER(mutexname)} -+ __DEBUG_RT_MUTEX_INITIALIZER(mutexname) -+ -+#define __RT_MUTEX_INITIALIZER(mutexname) \ -+ { __RT_MUTEX_INITIALIZER_PLAIN(mutexname) } -+ -+#define __RT_MUTEX_INITIALIZER_SAVE_STATE(mutexname) \ -+ { __RT_MUTEX_INITIALIZER_PLAIN(mutexname) \ -+ , .save_state = 1 } - - #define DEFINE_RT_MUTEX(mutexname) \ - struct rt_mutex mutexname = __RT_MUTEX_INITIALIZER(mutexname) -@@ -90,7 +105,9 @@ - extern void rt_mutex_destroy(struct rt_mutex *lock); - - extern void rt_mutex_lock(struct rt_mutex *lock); -+extern int rt_mutex_lock_state(struct rt_mutex *lock, int state); - extern int rt_mutex_lock_interruptible(struct rt_mutex *lock); -+extern int rt_mutex_lock_killable(struct rt_mutex *lock); - extern int rt_mutex_timed_lock(struct rt_mutex *lock, - struct hrtimer_sleeper *timeout); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rwlock_rt.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rwlock_rt.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rwlock_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rwlock_rt.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,99 @@ -+#ifndef __LINUX_RWLOCK_RT_H -+#define __LINUX_RWLOCK_RT_H -+ -+#ifndef __LINUX_SPINLOCK_H -+#error Do not include directly. Use spinlock.h -+#endif -+ -+#define rwlock_init(rwl) \ -+do { \ -+ static struct lock_class_key __key; \ -+ \ -+ rt_mutex_init(&(rwl)->lock); \ -+ __rt_rwlock_init(rwl, #rwl, &__key); \ -+} while (0) -+ -+extern void __lockfunc rt_write_lock(rwlock_t *rwlock); -+extern void __lockfunc rt_read_lock(rwlock_t *rwlock); -+extern int __lockfunc rt_write_trylock(rwlock_t *rwlock); -+extern int __lockfunc rt_write_trylock_irqsave(rwlock_t *trylock, unsigned long *flags); -+extern int __lockfunc rt_read_trylock(rwlock_t *rwlock); -+extern void __lockfunc rt_write_unlock(rwlock_t *rwlock); -+extern void __lockfunc rt_read_unlock(rwlock_t *rwlock); -+extern unsigned long __lockfunc rt_write_lock_irqsave(rwlock_t *rwlock); -+extern unsigned long __lockfunc rt_read_lock_irqsave(rwlock_t *rwlock); -+extern void __rt_rwlock_init(rwlock_t *rwlock, char *name, struct lock_class_key *key); -+ -+#define read_trylock(lock) __cond_lock(lock, rt_read_trylock(lock)) -+#define write_trylock(lock) __cond_lock(lock, rt_write_trylock(lock)) -+ -+#define write_trylock_irqsave(lock, flags) \ -+ __cond_lock(lock, rt_write_trylock_irqsave(lock, &flags)) -+ -+#define read_lock_irqsave(lock, flags) \ -+ do { \ -+ typecheck(unsigned long, flags); \ -+ flags = rt_read_lock_irqsave(lock); \ -+ } while (0) -+ -+#define write_lock_irqsave(lock, flags) \ -+ do { \ -+ typecheck(unsigned long, flags); \ -+ flags = rt_write_lock_irqsave(lock); \ -+ } while (0) -+ -+#define read_lock(lock) rt_read_lock(lock) -+ -+#define read_lock_bh(lock) \ -+ do { \ -+ local_bh_disable(); \ -+ rt_read_lock(lock); \ -+ } while (0) -+ -+#define read_lock_irq(lock) read_lock(lock) -+ -+#define write_lock(lock) rt_write_lock(lock) -+ -+#define write_lock_bh(lock) \ -+ do { \ -+ local_bh_disable(); \ -+ rt_write_lock(lock); \ -+ } while (0) -+ -+#define write_lock_irq(lock) write_lock(lock) -+ -+#define read_unlock(lock) rt_read_unlock(lock) -+ -+#define read_unlock_bh(lock) \ -+ do { \ -+ rt_read_unlock(lock); \ -+ local_bh_enable(); \ -+ } while (0) -+ -+#define read_unlock_irq(lock) read_unlock(lock) -+ -+#define write_unlock(lock) rt_write_unlock(lock) -+ -+#define write_unlock_bh(lock) \ -+ do { \ -+ rt_write_unlock(lock); \ -+ local_bh_enable(); \ -+ } while (0) -+ -+#define write_unlock_irq(lock) write_unlock(lock) -+ -+#define read_unlock_irqrestore(lock, flags) \ -+ do { \ -+ typecheck(unsigned long, flags); \ -+ (void) flags; \ -+ rt_read_unlock(lock); \ -+ } while (0) -+ -+#define write_unlock_irqrestore(lock, flags) \ -+ do { \ -+ typecheck(unsigned long, flags); \ -+ (void) flags; \ -+ rt_write_unlock(lock); \ -+ } while (0) -+ -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rwlock_types.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rwlock_types.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rwlock_types.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rwlock_types.h 2017-04-18 17:54:26.000000000 +0200 -@@ -1,6 +1,10 @@ - #ifndef __LINUX_RWLOCK_TYPES_H - #define __LINUX_RWLOCK_TYPES_H - -+#if !defined(__LINUX_SPINLOCK_TYPES_H) -+# error "Do not include directly, include spinlock_types.h" -+#endif -+ - /* - * include/linux/rwlock_types.h - generic rwlock type definitions - * and initializers -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rwlock_types_rt.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rwlock_types_rt.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rwlock_types_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rwlock_types_rt.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,33 @@ -+#ifndef __LINUX_RWLOCK_TYPES_RT_H -+#define __LINUX_RWLOCK_TYPES_RT_H -+ -+#ifndef __LINUX_SPINLOCK_TYPES_H -+#error "Do not include directly. Include spinlock_types.h instead" -+#endif -+ -+/* -+ * rwlocks - rtmutex which allows single reader recursion -+ */ -+typedef struct { -+ struct rt_mutex lock; -+ int read_depth; -+ unsigned int break_lock; -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+ struct lockdep_map dep_map; -+#endif -+} rwlock_t; -+ -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+# define RW_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname } -+#else -+# define RW_DEP_MAP_INIT(lockname) -+#endif -+ -+#define __RW_LOCK_UNLOCKED(name) \ -+ { .lock = __RT_MUTEX_INITIALIZER_SAVE_STATE(name.lock), \ -+ RW_DEP_MAP_INIT(name) } -+ -+#define DEFINE_RWLOCK(name) \ -+ rwlock_t name = __RW_LOCK_UNLOCKED(name) -+ -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rwsem.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rwsem.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rwsem.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rwsem.h 2017-04-18 17:54:26.000000000 +0200 -@@ -19,6 +19,10 @@ - #include - #endif - -+#ifdef CONFIG_PREEMPT_RT_FULL -+#include -+#else /* PREEMPT_RT_FULL */ -+ - struct rw_semaphore; - - #ifdef CONFIG_RWSEM_GENERIC_SPINLOCK -@@ -106,6 +110,13 @@ - return !list_empty(&sem->wait_list); - } - -+#endif /* !PREEMPT_RT_FULL */ -+ -+/* -+ * The functions below are the same for all rwsem implementations including -+ * the RT specific variant. -+ */ -+ - /* - * lock for reading - */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rwsem_rt.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rwsem_rt.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/rwsem_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/rwsem_rt.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,67 @@ -+#ifndef _LINUX_RWSEM_RT_H -+#define _LINUX_RWSEM_RT_H -+ -+#ifndef _LINUX_RWSEM_H -+#error "Include rwsem.h" -+#endif -+ -+#include -+#include -+ -+#define READER_BIAS (1U << 31) -+#define WRITER_BIAS (1U << 30) -+ -+struct rw_semaphore { -+ atomic_t readers; -+ struct rt_mutex rtmutex; -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+ struct lockdep_map dep_map; -+#endif -+}; -+ -+#define __RWSEM_INITIALIZER(name) \ -+{ \ -+ .readers = ATOMIC_INIT(READER_BIAS), \ -+ .rtmutex = __RT_MUTEX_INITIALIZER(name.rtmutex), \ -+ RW_DEP_MAP_INIT(name) \ -+} -+ -+#define DECLARE_RWSEM(lockname) \ -+ struct rw_semaphore lockname = __RWSEM_INITIALIZER(lockname) -+ -+extern void __rwsem_init(struct rw_semaphore *rwsem, const char *name, -+ struct lock_class_key *key); -+ -+#define __init_rwsem(sem, name, key) \ -+do { \ -+ rt_mutex_init(&(sem)->rtmutex); \ -+ __rwsem_init((sem), (name), (key)); \ -+} while (0) -+ -+#define init_rwsem(sem) \ -+do { \ -+ static struct lock_class_key __key; \ -+ \ -+ __init_rwsem((sem), #sem, &__key); \ -+} while (0) -+ -+static inline int rwsem_is_locked(struct rw_semaphore *sem) -+{ -+ return atomic_read(&sem->readers) != READER_BIAS; -+} -+ -+static inline int rwsem_is_contended(struct rw_semaphore *sem) -+{ -+ return atomic_read(&sem->readers) > 0; -+} -+ -+extern void __down_read(struct rw_semaphore *sem); -+extern int __down_read_trylock(struct rw_semaphore *sem); -+extern void __down_write(struct rw_semaphore *sem); -+extern int __must_check __down_write_killable(struct rw_semaphore *sem); -+extern int __down_write_trylock(struct rw_semaphore *sem); -+extern void __up_read(struct rw_semaphore *sem); -+extern void __up_write(struct rw_semaphore *sem); -+extern void __downgrade_write(struct rw_semaphore *sem); -+ -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/sched.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/sched.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/sched.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/sched.h 2017-04-18 17:54:26.000000000 +0200 -@@ -26,6 +26,7 @@ - #include - #include - #include -+#include - - #include - #include -@@ -243,10 +244,7 @@ - TASK_UNINTERRUPTIBLE | __TASK_STOPPED | \ - __TASK_TRACED | EXIT_ZOMBIE | EXIT_DEAD) - --#define task_is_traced(task) ((task->state & __TASK_TRACED) != 0) - #define task_is_stopped(task) ((task->state & __TASK_STOPPED) != 0) --#define task_is_stopped_or_traced(task) \ -- ((task->state & (__TASK_STOPPED | __TASK_TRACED)) != 0) - #define task_contributes_to_load(task) \ - ((task->state & TASK_UNINTERRUPTIBLE) != 0 && \ - (task->flags & PF_FROZEN) == 0 && \ -@@ -312,6 +310,11 @@ - - #endif - -+#define __set_current_state_no_track(state_value) \ -+ do { current->state = (state_value); } while (0) -+#define set_current_state_no_track(state_value) \ -+ set_mb(current->state, (state_value)) -+ - /* Task command name length */ - #define TASK_COMM_LEN 16 - -@@ -1013,8 +1016,18 @@ - struct wake_q_head name = { WAKE_Q_TAIL, &name.first } - - extern void wake_q_add(struct wake_q_head *head, -- struct task_struct *task); --extern void wake_up_q(struct wake_q_head *head); -+ struct task_struct *task); -+extern void __wake_up_q(struct wake_q_head *head, bool sleeper); -+ -+static inline void wake_up_q(struct wake_q_head *head) -+{ -+ __wake_up_q(head, false); -+} -+ -+static inline void wake_up_q_sleeper(struct wake_q_head *head) -+{ -+ __wake_up_q(head, true); -+} - - /* - * sched-domains (multiprocessor balancing) declarations: -@@ -1481,6 +1494,7 @@ - struct thread_info thread_info; - #endif - volatile long state; /* -1 unrunnable, 0 runnable, >0 stopped */ -+ volatile long saved_state; /* saved state for "spinlock sleepers" */ - void *stack; - atomic_t usage; - unsigned int flags; /* per process flags, defined below */ -@@ -1520,6 +1534,12 @@ - #endif - - unsigned int policy; -+#ifdef CONFIG_PREEMPT_RT_FULL -+ int migrate_disable; -+# ifdef CONFIG_SCHED_DEBUG -+ int migrate_disable_atomic; -+# endif -+#endif - int nr_cpus_allowed; - cpumask_t cpus_allowed; - -@@ -1654,6 +1674,9 @@ - - struct task_cputime cputime_expires; - struct list_head cpu_timers[3]; -+#ifdef CONFIG_PREEMPT_RT_BASE -+ struct task_struct *posix_timer_list; -+#endif - - /* process credentials */ - const struct cred __rcu *ptracer_cred; /* Tracer's credentials at attach */ -@@ -1685,10 +1708,15 @@ - /* signal handlers */ - struct signal_struct *signal; - struct sighand_struct *sighand; -+ struct sigqueue *sigqueue_cache; - - sigset_t blocked, real_blocked; - sigset_t saved_sigmask; /* restored if set_restore_sigmask() was used */ - struct sigpending pending; -+#ifdef CONFIG_PREEMPT_RT_FULL -+ /* TODO: move me into ->restart_block ? */ -+ struct siginfo forced_info; -+#endif - - unsigned long sas_ss_sp; - size_t sas_ss_size; -@@ -1917,6 +1945,12 @@ - /* bitmask and counter of trace recursion */ - unsigned long trace_recursion; - #endif /* CONFIG_TRACING */ -+#ifdef CONFIG_WAKEUP_LATENCY_HIST -+ u64 preempt_timestamp_hist; -+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST -+ long timer_offset; -+#endif -+#endif - #ifdef CONFIG_KCOV - /* Coverage collection mode enabled for this task (0 if disabled). */ - enum kcov_mode kcov_mode; -@@ -1942,9 +1976,23 @@ - unsigned int sequential_io; - unsigned int sequential_io_avg; - #endif -+#ifdef CONFIG_PREEMPT_RT_BASE -+ struct rcu_head put_rcu; -+ int softirq_nestcnt; -+ unsigned int softirqs_raised; -+#endif -+#ifdef CONFIG_PREEMPT_RT_FULL -+# if defined CONFIG_HIGHMEM || defined CONFIG_X86_32 -+ int kmap_idx; -+ pte_t kmap_pte[KM_TYPE_NR]; -+# endif -+#endif - #ifdef CONFIG_DEBUG_ATOMIC_SLEEP - unsigned long task_state_change; - #endif -+#ifdef CONFIG_PREEMPT_RT_FULL -+ int xmit_recursion; -+#endif - int pagefault_disabled; - #ifdef CONFIG_MMU - struct task_struct *oom_reaper_list; -@@ -1984,14 +2032,6 @@ - } - #endif - --/* Future-safe accessor for struct task_struct's cpus_allowed. */ --#define tsk_cpus_allowed(tsk) (&(tsk)->cpus_allowed) -- --static inline int tsk_nr_cpus_allowed(struct task_struct *p) --{ -- return p->nr_cpus_allowed; --} -- - #define TNF_MIGRATED 0x01 - #define TNF_NO_GROUP 0x02 - #define TNF_SHARED 0x04 -@@ -2207,6 +2247,15 @@ - extern void free_task(struct task_struct *tsk); - #define get_task_struct(tsk) do { atomic_inc(&(tsk)->usage); } while(0) - -+#ifdef CONFIG_PREEMPT_RT_BASE -+extern void __put_task_struct_cb(struct rcu_head *rhp); -+ -+static inline void put_task_struct(struct task_struct *t) -+{ -+ if (atomic_dec_and_test(&t->usage)) -+ call_rcu(&t->put_rcu, __put_task_struct_cb); -+} -+#else - extern void __put_task_struct(struct task_struct *t); - - static inline void put_task_struct(struct task_struct *t) -@@ -2214,6 +2263,7 @@ - if (atomic_dec_and_test(&t->usage)) - __put_task_struct(t); - } -+#endif - - struct task_struct *task_rcu_dereference(struct task_struct **ptask); - struct task_struct *try_get_task_struct(struct task_struct **ptask); -@@ -2255,6 +2305,7 @@ - /* - * Per process flags - */ -+#define PF_IN_SOFTIRQ 0x00000001 /* Task is serving softirq */ - #define PF_EXITING 0x00000004 /* getting shut down */ - #define PF_EXITPIDONE 0x00000008 /* pi exit done on shut down */ - #define PF_VCPU 0x00000010 /* I'm a virtual CPU */ -@@ -2423,6 +2474,10 @@ - - extern int set_cpus_allowed_ptr(struct task_struct *p, - const struct cpumask *new_mask); -+int migrate_me(void); -+void tell_sched_cpu_down_begin(int cpu); -+void tell_sched_cpu_down_done(int cpu); -+ - #else - static inline void do_set_cpus_allowed(struct task_struct *p, - const struct cpumask *new_mask) -@@ -2435,6 +2490,9 @@ - return -EINVAL; - return 0; - } -+static inline int migrate_me(void) { return 0; } -+static inline void tell_sched_cpu_down_begin(int cpu) { } -+static inline void tell_sched_cpu_down_done(int cpu) { } - #endif - - #ifdef CONFIG_NO_HZ_COMMON -@@ -2673,6 +2731,7 @@ - - extern int wake_up_state(struct task_struct *tsk, unsigned int state); - extern int wake_up_process(struct task_struct *tsk); -+extern int wake_up_lock_sleeper(struct task_struct * tsk); - extern void wake_up_new_task(struct task_struct *tsk); - #ifdef CONFIG_SMP - extern void kick_process(struct task_struct *tsk); -@@ -2881,6 +2940,17 @@ - __mmdrop(mm); - } - -+#ifdef CONFIG_PREEMPT_RT_BASE -+extern void __mmdrop_delayed(struct rcu_head *rhp); -+static inline void mmdrop_delayed(struct mm_struct *mm) -+{ -+ if (atomic_dec_and_test(&mm->mm_count)) -+ call_rcu(&mm->delayed_drop, __mmdrop_delayed); -+} -+#else -+# define mmdrop_delayed(mm) mmdrop(mm) -+#endif -+ - static inline void mmdrop_async_fn(struct work_struct *work) - { - struct mm_struct *mm = container_of(work, struct mm_struct, async_put_work); -@@ -3273,6 +3343,43 @@ - return unlikely(test_tsk_thread_flag(tsk,TIF_NEED_RESCHED)); - } - -+#ifdef CONFIG_PREEMPT_LAZY -+static inline void set_tsk_need_resched_lazy(struct task_struct *tsk) -+{ -+ set_tsk_thread_flag(tsk,TIF_NEED_RESCHED_LAZY); -+} -+ -+static inline void clear_tsk_need_resched_lazy(struct task_struct *tsk) -+{ -+ clear_tsk_thread_flag(tsk,TIF_NEED_RESCHED_LAZY); -+} -+ -+static inline int test_tsk_need_resched_lazy(struct task_struct *tsk) -+{ -+ return unlikely(test_tsk_thread_flag(tsk,TIF_NEED_RESCHED_LAZY)); -+} -+ -+static inline int need_resched_lazy(void) -+{ -+ return test_thread_flag(TIF_NEED_RESCHED_LAZY); -+} -+ -+static inline int need_resched_now(void) -+{ -+ return test_thread_flag(TIF_NEED_RESCHED); -+} -+ -+#else -+static inline void clear_tsk_need_resched_lazy(struct task_struct *tsk) { } -+static inline int need_resched_lazy(void) { return 0; } -+ -+static inline int need_resched_now(void) -+{ -+ return test_thread_flag(TIF_NEED_RESCHED); -+} -+ -+#endif -+ - static inline int restart_syscall(void) - { - set_tsk_thread_flag(current, TIF_SIGPENDING); -@@ -3304,6 +3411,51 @@ - return (state & TASK_INTERRUPTIBLE) || __fatal_signal_pending(p); - } - -+static inline bool __task_is_stopped_or_traced(struct task_struct *task) -+{ -+ if (task->state & (__TASK_STOPPED | __TASK_TRACED)) -+ return true; -+#ifdef CONFIG_PREEMPT_RT_FULL -+ if (task->saved_state & (__TASK_STOPPED | __TASK_TRACED)) -+ return true; -+#endif -+ return false; -+} -+ -+static inline bool task_is_stopped_or_traced(struct task_struct *task) -+{ -+ bool traced_stopped; -+ -+#ifdef CONFIG_PREEMPT_RT_FULL -+ unsigned long flags; -+ -+ raw_spin_lock_irqsave(&task->pi_lock, flags); -+ traced_stopped = __task_is_stopped_or_traced(task); -+ raw_spin_unlock_irqrestore(&task->pi_lock, flags); -+#else -+ traced_stopped = __task_is_stopped_or_traced(task); -+#endif -+ return traced_stopped; -+} -+ -+static inline bool task_is_traced(struct task_struct *task) -+{ -+ bool traced = false; -+ -+ if (task->state & __TASK_TRACED) -+ return true; -+#ifdef CONFIG_PREEMPT_RT_FULL -+ /* in case the task is sleeping on tasklist_lock */ -+ raw_spin_lock_irq(&task->pi_lock); -+ if (task->state & __TASK_TRACED) -+ traced = true; -+ else if (task->saved_state & __TASK_TRACED) -+ traced = true; -+ raw_spin_unlock_irq(&task->pi_lock); -+#endif -+ return traced; -+} -+ - /* - * cond_resched() and cond_resched_lock(): latency reduction via - * explicit rescheduling in places that are safe. The return -@@ -3329,12 +3481,16 @@ - __cond_resched_lock(lock); \ - }) - -+#ifndef CONFIG_PREEMPT_RT_FULL - extern int __cond_resched_softirq(void); - - #define cond_resched_softirq() ({ \ - ___might_sleep(__FILE__, __LINE__, SOFTIRQ_DISABLE_OFFSET); \ - __cond_resched_softirq(); \ - }) -+#else -+# define cond_resched_softirq() cond_resched() -+#endif - - static inline void cond_resched_rcu(void) - { -@@ -3509,6 +3665,31 @@ - - #endif /* CONFIG_SMP */ - -+static inline int __migrate_disabled(struct task_struct *p) -+{ -+#ifdef CONFIG_PREEMPT_RT_FULL -+ return p->migrate_disable; -+#else -+ return 0; -+#endif -+} -+ -+/* Future-safe accessor for struct task_struct's cpus_allowed. */ -+static inline const struct cpumask *tsk_cpus_allowed(struct task_struct *p) -+{ -+ if (__migrate_disabled(p)) -+ return cpumask_of(task_cpu(p)); -+ -+ return &p->cpus_allowed; -+} -+ -+static inline int tsk_nr_cpus_allowed(struct task_struct *p) -+{ -+ if (__migrate_disabled(p)) -+ return 1; -+ return p->nr_cpus_allowed; -+} -+ - extern long sched_setaffinity(pid_t pid, const struct cpumask *new_mask); - extern long sched_getaffinity(pid_t pid, struct cpumask *mask); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/seqlock.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/seqlock.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/seqlock.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/seqlock.h 2017-04-18 17:54:26.000000000 +0200 -@@ -220,20 +220,30 @@ - return __read_seqcount_retry(s, start); - } - -- -- --static inline void raw_write_seqcount_begin(seqcount_t *s) -+static inline void __raw_write_seqcount_begin(seqcount_t *s) - { - s->sequence++; - smp_wmb(); - } - --static inline void raw_write_seqcount_end(seqcount_t *s) -+static inline void raw_write_seqcount_begin(seqcount_t *s) -+{ -+ preempt_disable_rt(); -+ __raw_write_seqcount_begin(s); -+} -+ -+static inline void __raw_write_seqcount_end(seqcount_t *s) - { - smp_wmb(); - s->sequence++; - } - -+static inline void raw_write_seqcount_end(seqcount_t *s) -+{ -+ __raw_write_seqcount_end(s); -+ preempt_enable_rt(); -+} -+ - /** - * raw_write_seqcount_barrier - do a seq write barrier - * @s: pointer to seqcount_t -@@ -428,10 +438,32 @@ - /* - * Read side functions for starting and finalizing a read side section. - */ -+#ifndef CONFIG_PREEMPT_RT_FULL - static inline unsigned read_seqbegin(const seqlock_t *sl) - { - return read_seqcount_begin(&sl->seqcount); - } -+#else -+/* -+ * Starvation safe read side for RT -+ */ -+static inline unsigned read_seqbegin(seqlock_t *sl) -+{ -+ unsigned ret; -+ -+repeat: -+ ret = ACCESS_ONCE(sl->seqcount.sequence); -+ if (unlikely(ret & 1)) { -+ /* -+ * Take the lock and let the writer proceed (i.e. evtl -+ * boost it), otherwise we could loop here forever. -+ */ -+ spin_unlock_wait(&sl->lock); -+ goto repeat; -+ } -+ return ret; -+} -+#endif - - static inline unsigned read_seqretry(const seqlock_t *sl, unsigned start) - { -@@ -446,36 +478,45 @@ - static inline void write_seqlock(seqlock_t *sl) - { - spin_lock(&sl->lock); -- write_seqcount_begin(&sl->seqcount); -+ __raw_write_seqcount_begin(&sl->seqcount); -+} -+ -+static inline int try_write_seqlock(seqlock_t *sl) -+{ -+ if (spin_trylock(&sl->lock)) { -+ __raw_write_seqcount_begin(&sl->seqcount); -+ return 1; -+ } -+ return 0; - } - - static inline void write_sequnlock(seqlock_t *sl) - { -- write_seqcount_end(&sl->seqcount); -+ __raw_write_seqcount_end(&sl->seqcount); - spin_unlock(&sl->lock); - } - - static inline void write_seqlock_bh(seqlock_t *sl) - { - spin_lock_bh(&sl->lock); -- write_seqcount_begin(&sl->seqcount); -+ __raw_write_seqcount_begin(&sl->seqcount); - } - - static inline void write_sequnlock_bh(seqlock_t *sl) - { -- write_seqcount_end(&sl->seqcount); -+ __raw_write_seqcount_end(&sl->seqcount); - spin_unlock_bh(&sl->lock); - } - - static inline void write_seqlock_irq(seqlock_t *sl) - { - spin_lock_irq(&sl->lock); -- write_seqcount_begin(&sl->seqcount); -+ __raw_write_seqcount_begin(&sl->seqcount); - } - - static inline void write_sequnlock_irq(seqlock_t *sl) - { -- write_seqcount_end(&sl->seqcount); -+ __raw_write_seqcount_end(&sl->seqcount); - spin_unlock_irq(&sl->lock); - } - -@@ -484,7 +525,7 @@ - unsigned long flags; - - spin_lock_irqsave(&sl->lock, flags); -- write_seqcount_begin(&sl->seqcount); -+ __raw_write_seqcount_begin(&sl->seqcount); - return flags; - } - -@@ -494,7 +535,7 @@ - static inline void - write_sequnlock_irqrestore(seqlock_t *sl, unsigned long flags) - { -- write_seqcount_end(&sl->seqcount); -+ __raw_write_seqcount_end(&sl->seqcount); - spin_unlock_irqrestore(&sl->lock, flags); - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/signal.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/signal.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/signal.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/signal.h 2017-04-18 17:54:26.000000000 +0200 -@@ -233,6 +233,7 @@ - } - - extern void flush_sigqueue(struct sigpending *queue); -+extern void flush_task_sigqueue(struct task_struct *tsk); - - /* Test if 'sig' is valid signal. Use this instead of testing _NSIG directly */ - static inline int valid_signal(unsigned long sig) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/skbuff.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/skbuff.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/skbuff.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/skbuff.h 2017-04-18 17:54:26.000000000 +0200 -@@ -284,6 +284,7 @@ - - __u32 qlen; - spinlock_t lock; -+ raw_spinlock_t raw_lock; - }; - - struct sk_buff; -@@ -1573,6 +1574,12 @@ - __skb_queue_head_init(list); - } - -+static inline void skb_queue_head_init_raw(struct sk_buff_head *list) -+{ -+ raw_spin_lock_init(&list->raw_lock); -+ __skb_queue_head_init(list); -+} -+ - static inline void skb_queue_head_init_class(struct sk_buff_head *list, - struct lock_class_key *class) - { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/smp.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/smp.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/smp.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/smp.h 2017-04-18 17:54:26.000000000 +0200 -@@ -120,6 +120,13 @@ - extern void __init setup_nr_cpu_ids(void); - extern void __init smp_init(void); - -+extern int __boot_cpu_id; -+ -+static inline int get_boot_cpu_id(void) -+{ -+ return __boot_cpu_id; -+} -+ - #else /* !SMP */ - - static inline void smp_send_stop(void) { } -@@ -158,6 +165,11 @@ - static inline void smp_init(void) { } - #endif - -+static inline int get_boot_cpu_id(void) -+{ -+ return 0; -+} -+ - #endif /* !SMP */ - - /* -@@ -185,6 +197,9 @@ - #define get_cpu() ({ preempt_disable(); smp_processor_id(); }) - #define put_cpu() preempt_enable() - -+#define get_cpu_light() ({ migrate_disable(); smp_processor_id(); }) -+#define put_cpu_light() migrate_enable() -+ - /* - * Callback to arch code if there's nosmp or maxcpus=0 on the - * boot command line: -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock.h 2017-04-18 17:54:26.000000000 +0200 -@@ -271,7 +271,11 @@ - #define raw_spin_can_lock(lock) (!raw_spin_is_locked(lock)) - - /* Include rwlock functions */ --#include -+#ifdef CONFIG_PREEMPT_RT_FULL -+# include -+#else -+# include -+#endif - - /* - * Pull the _spin_*()/_read_*()/_write_*() functions/declarations: -@@ -282,6 +286,10 @@ - # include - #endif - -+#ifdef CONFIG_PREEMPT_RT_FULL -+# include -+#else /* PREEMPT_RT_FULL */ -+ - /* - * Map the spin_lock functions to the raw variants for PREEMPT_RT=n - */ -@@ -416,4 +424,6 @@ - #define atomic_dec_and_lock(atomic, lock) \ - __cond_lock(lock, _atomic_dec_and_lock(atomic, lock)) - -+#endif /* !PREEMPT_RT_FULL */ -+ - #endif /* __LINUX_SPINLOCK_H */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock_api_smp.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock_api_smp.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock_api_smp.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock_api_smp.h 2017-04-18 17:54:26.000000000 +0200 -@@ -189,6 +189,8 @@ - return 0; - } - --#include -+#ifndef CONFIG_PREEMPT_RT_FULL -+# include -+#endif - - #endif /* __LINUX_SPINLOCK_API_SMP_H */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock_rt.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock_rt.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock_rt.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,162 @@ -+#ifndef __LINUX_SPINLOCK_RT_H -+#define __LINUX_SPINLOCK_RT_H -+ -+#ifndef __LINUX_SPINLOCK_H -+#error Do not include directly. Use spinlock.h -+#endif -+ -+#include -+ -+extern void -+__rt_spin_lock_init(spinlock_t *lock, char *name, struct lock_class_key *key); -+ -+#define spin_lock_init(slock) \ -+do { \ -+ static struct lock_class_key __key; \ -+ \ -+ rt_mutex_init(&(slock)->lock); \ -+ __rt_spin_lock_init(slock, #slock, &__key); \ -+} while (0) -+ -+void __lockfunc rt_spin_lock__no_mg(spinlock_t *lock); -+void __lockfunc rt_spin_unlock__no_mg(spinlock_t *lock); -+int __lockfunc rt_spin_trylock__no_mg(spinlock_t *lock); -+ -+extern void __lockfunc rt_spin_lock(spinlock_t *lock); -+extern unsigned long __lockfunc rt_spin_lock_trace_flags(spinlock_t *lock); -+extern void __lockfunc rt_spin_lock_nested(spinlock_t *lock, int subclass); -+extern void __lockfunc rt_spin_unlock(spinlock_t *lock); -+extern void __lockfunc rt_spin_unlock_wait(spinlock_t *lock); -+extern int __lockfunc rt_spin_trylock_irqsave(spinlock_t *lock, unsigned long *flags); -+extern int __lockfunc rt_spin_trylock_bh(spinlock_t *lock); -+extern int __lockfunc rt_spin_trylock(spinlock_t *lock); -+extern int atomic_dec_and_spin_lock(atomic_t *atomic, spinlock_t *lock); -+ -+/* -+ * lockdep-less calls, for derived types like rwlock: -+ * (for trylock they can use rt_mutex_trylock() directly. -+ */ -+extern void __lockfunc __rt_spin_lock__no_mg(struct rt_mutex *lock); -+extern void __lockfunc __rt_spin_lock(struct rt_mutex *lock); -+extern void __lockfunc __rt_spin_unlock(struct rt_mutex *lock); -+ -+#define spin_lock(lock) rt_spin_lock(lock) -+ -+#define spin_lock_bh(lock) \ -+ do { \ -+ local_bh_disable(); \ -+ rt_spin_lock(lock); \ -+ } while (0) -+ -+#define spin_lock_irq(lock) spin_lock(lock) -+ -+#define spin_do_trylock(lock) __cond_lock(lock, rt_spin_trylock(lock)) -+ -+#define spin_trylock(lock) \ -+({ \ -+ int __locked; \ -+ __locked = spin_do_trylock(lock); \ -+ __locked; \ -+}) -+ -+#ifdef CONFIG_LOCKDEP -+# define spin_lock_nested(lock, subclass) \ -+ do { \ -+ rt_spin_lock_nested(lock, subclass); \ -+ } while (0) -+ -+#define spin_lock_bh_nested(lock, subclass) \ -+ do { \ -+ local_bh_disable(); \ -+ rt_spin_lock_nested(lock, subclass); \ -+ } while (0) -+ -+# define spin_lock_irqsave_nested(lock, flags, subclass) \ -+ do { \ -+ typecheck(unsigned long, flags); \ -+ flags = 0; \ -+ rt_spin_lock_nested(lock, subclass); \ -+ } while (0) -+#else -+# define spin_lock_nested(lock, subclass) spin_lock(lock) -+# define spin_lock_bh_nested(lock, subclass) spin_lock_bh(lock) -+ -+# define spin_lock_irqsave_nested(lock, flags, subclass) \ -+ do { \ -+ typecheck(unsigned long, flags); \ -+ flags = 0; \ -+ spin_lock(lock); \ -+ } while (0) -+#endif -+ -+#define spin_lock_irqsave(lock, flags) \ -+ do { \ -+ typecheck(unsigned long, flags); \ -+ flags = 0; \ -+ spin_lock(lock); \ -+ } while (0) -+ -+static inline unsigned long spin_lock_trace_flags(spinlock_t *lock) -+{ -+ unsigned long flags = 0; -+#ifdef CONFIG_TRACE_IRQFLAGS -+ flags = rt_spin_lock_trace_flags(lock); -+#else -+ spin_lock(lock); /* lock_local */ -+#endif -+ return flags; -+} -+ -+/* FIXME: we need rt_spin_lock_nest_lock */ -+#define spin_lock_nest_lock(lock, nest_lock) spin_lock_nested(lock, 0) -+ -+#define spin_unlock(lock) rt_spin_unlock(lock) -+ -+#define spin_unlock_bh(lock) \ -+ do { \ -+ rt_spin_unlock(lock); \ -+ local_bh_enable(); \ -+ } while (0) -+ -+#define spin_unlock_irq(lock) spin_unlock(lock) -+ -+#define spin_unlock_irqrestore(lock, flags) \ -+ do { \ -+ typecheck(unsigned long, flags); \ -+ (void) flags; \ -+ spin_unlock(lock); \ -+ } while (0) -+ -+#define spin_trylock_bh(lock) __cond_lock(lock, rt_spin_trylock_bh(lock)) -+#define spin_trylock_irq(lock) spin_trylock(lock) -+ -+#define spin_trylock_irqsave(lock, flags) \ -+ rt_spin_trylock_irqsave(lock, &(flags)) -+ -+#define spin_unlock_wait(lock) rt_spin_unlock_wait(lock) -+ -+#ifdef CONFIG_GENERIC_LOCKBREAK -+# define spin_is_contended(lock) ((lock)->break_lock) -+#else -+# define spin_is_contended(lock) (((void)(lock), 0)) -+#endif -+ -+static inline int spin_can_lock(spinlock_t *lock) -+{ -+ return !rt_mutex_is_locked(&lock->lock); -+} -+ -+static inline int spin_is_locked(spinlock_t *lock) -+{ -+ return rt_mutex_is_locked(&lock->lock); -+} -+ -+static inline void assert_spin_locked(spinlock_t *lock) -+{ -+ BUG_ON(!spin_is_locked(lock)); -+} -+ -+#define atomic_dec_and_lock(atomic, lock) \ -+ atomic_dec_and_spin_lock(atomic, lock) -+ -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock_types.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock_types.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock_types.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock_types.h 2017-04-18 17:54:26.000000000 +0200 -@@ -9,80 +9,15 @@ - * Released under the General Public License (GPL). - */ - --#if defined(CONFIG_SMP) --# include --#else --# include --#endif -- --#include -- --typedef struct raw_spinlock { -- arch_spinlock_t raw_lock; --#ifdef CONFIG_GENERIC_LOCKBREAK -- unsigned int break_lock; --#endif --#ifdef CONFIG_DEBUG_SPINLOCK -- unsigned int magic, owner_cpu; -- void *owner; --#endif --#ifdef CONFIG_DEBUG_LOCK_ALLOC -- struct lockdep_map dep_map; --#endif --} raw_spinlock_t; -- --#define SPINLOCK_MAGIC 0xdead4ead -- --#define SPINLOCK_OWNER_INIT ((void *)-1L) -- --#ifdef CONFIG_DEBUG_LOCK_ALLOC --# define SPIN_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname } --#else --# define SPIN_DEP_MAP_INIT(lockname) --#endif -+#include - --#ifdef CONFIG_DEBUG_SPINLOCK --# define SPIN_DEBUG_INIT(lockname) \ -- .magic = SPINLOCK_MAGIC, \ -- .owner_cpu = -1, \ -- .owner = SPINLOCK_OWNER_INIT, -+#ifndef CONFIG_PREEMPT_RT_FULL -+# include -+# include - #else --# define SPIN_DEBUG_INIT(lockname) -+# include -+# include -+# include - #endif - --#define __RAW_SPIN_LOCK_INITIALIZER(lockname) \ -- { \ -- .raw_lock = __ARCH_SPIN_LOCK_UNLOCKED, \ -- SPIN_DEBUG_INIT(lockname) \ -- SPIN_DEP_MAP_INIT(lockname) } -- --#define __RAW_SPIN_LOCK_UNLOCKED(lockname) \ -- (raw_spinlock_t) __RAW_SPIN_LOCK_INITIALIZER(lockname) -- --#define DEFINE_RAW_SPINLOCK(x) raw_spinlock_t x = __RAW_SPIN_LOCK_UNLOCKED(x) -- --typedef struct spinlock { -- union { -- struct raw_spinlock rlock; -- --#ifdef CONFIG_DEBUG_LOCK_ALLOC --# define LOCK_PADSIZE (offsetof(struct raw_spinlock, dep_map)) -- struct { -- u8 __padding[LOCK_PADSIZE]; -- struct lockdep_map dep_map; -- }; --#endif -- }; --} spinlock_t; -- --#define __SPIN_LOCK_INITIALIZER(lockname) \ -- { { .rlock = __RAW_SPIN_LOCK_INITIALIZER(lockname) } } -- --#define __SPIN_LOCK_UNLOCKED(lockname) \ -- (spinlock_t ) __SPIN_LOCK_INITIALIZER(lockname) -- --#define DEFINE_SPINLOCK(x) spinlock_t x = __SPIN_LOCK_UNLOCKED(x) -- --#include -- - #endif /* __LINUX_SPINLOCK_TYPES_H */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock_types_nort.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock_types_nort.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock_types_nort.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock_types_nort.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,33 @@ -+#ifndef __LINUX_SPINLOCK_TYPES_NORT_H -+#define __LINUX_SPINLOCK_TYPES_NORT_H -+ -+#ifndef __LINUX_SPINLOCK_TYPES_H -+#error "Do not include directly. Include spinlock_types.h instead" -+#endif -+ -+/* -+ * The non RT version maps spinlocks to raw_spinlocks -+ */ -+typedef struct spinlock { -+ union { -+ struct raw_spinlock rlock; -+ -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+# define LOCK_PADSIZE (offsetof(struct raw_spinlock, dep_map)) -+ struct { -+ u8 __padding[LOCK_PADSIZE]; -+ struct lockdep_map dep_map; -+ }; -+#endif -+ }; -+} spinlock_t; -+ -+#define __SPIN_LOCK_INITIALIZER(lockname) \ -+ { { .rlock = __RAW_SPIN_LOCK_INITIALIZER(lockname) } } -+ -+#define __SPIN_LOCK_UNLOCKED(lockname) \ -+ (spinlock_t ) __SPIN_LOCK_INITIALIZER(lockname) -+ -+#define DEFINE_SPINLOCK(x) spinlock_t x = __SPIN_LOCK_UNLOCKED(x) -+ -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock_types_raw.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock_types_raw.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock_types_raw.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock_types_raw.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,56 @@ -+#ifndef __LINUX_SPINLOCK_TYPES_RAW_H -+#define __LINUX_SPINLOCK_TYPES_RAW_H -+ -+#if defined(CONFIG_SMP) -+# include -+#else -+# include -+#endif -+ -+#include -+ -+typedef struct raw_spinlock { -+ arch_spinlock_t raw_lock; -+#ifdef CONFIG_GENERIC_LOCKBREAK -+ unsigned int break_lock; -+#endif -+#ifdef CONFIG_DEBUG_SPINLOCK -+ unsigned int magic, owner_cpu; -+ void *owner; -+#endif -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+ struct lockdep_map dep_map; -+#endif -+} raw_spinlock_t; -+ -+#define SPINLOCK_MAGIC 0xdead4ead -+ -+#define SPINLOCK_OWNER_INIT ((void *)-1L) -+ -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+# define SPIN_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname } -+#else -+# define SPIN_DEP_MAP_INIT(lockname) -+#endif -+ -+#ifdef CONFIG_DEBUG_SPINLOCK -+# define SPIN_DEBUG_INIT(lockname) \ -+ .magic = SPINLOCK_MAGIC, \ -+ .owner_cpu = -1, \ -+ .owner = SPINLOCK_OWNER_INIT, -+#else -+# define SPIN_DEBUG_INIT(lockname) -+#endif -+ -+#define __RAW_SPIN_LOCK_INITIALIZER(lockname) \ -+ { \ -+ .raw_lock = __ARCH_SPIN_LOCK_UNLOCKED, \ -+ SPIN_DEBUG_INIT(lockname) \ -+ SPIN_DEP_MAP_INIT(lockname) } -+ -+#define __RAW_SPIN_LOCK_UNLOCKED(lockname) \ -+ (raw_spinlock_t) __RAW_SPIN_LOCK_INITIALIZER(lockname) -+ -+#define DEFINE_RAW_SPINLOCK(x) raw_spinlock_t x = __RAW_SPIN_LOCK_UNLOCKED(x) -+ -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock_types_rt.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock_types_rt.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/spinlock_types_rt.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/spinlock_types_rt.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,48 @@ -+#ifndef __LINUX_SPINLOCK_TYPES_RT_H -+#define __LINUX_SPINLOCK_TYPES_RT_H -+ -+#ifndef __LINUX_SPINLOCK_TYPES_H -+#error "Do not include directly. Include spinlock_types.h instead" -+#endif -+ -+#include -+ -+/* -+ * PREEMPT_RT: spinlocks - an RT mutex plus lock-break field: -+ */ -+typedef struct spinlock { -+ struct rt_mutex lock; -+ unsigned int break_lock; -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+ struct lockdep_map dep_map; -+#endif -+} spinlock_t; -+ -+#ifdef CONFIG_DEBUG_RT_MUTEXES -+# define __RT_SPIN_INITIALIZER(name) \ -+ { \ -+ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \ -+ .save_state = 1, \ -+ .file = __FILE__, \ -+ .line = __LINE__ , \ -+ } -+#else -+# define __RT_SPIN_INITIALIZER(name) \ -+ { \ -+ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \ -+ .save_state = 1, \ -+ } -+#endif -+ -+/* -+.wait_list = PLIST_HEAD_INIT_RAW((name).lock.wait_list, (name).lock.wait_lock) -+*/ -+ -+#define __SPIN_LOCK_UNLOCKED(name) \ -+ { .lock = __RT_SPIN_INITIALIZER(name.lock), \ -+ SPIN_DEP_MAP_INIT(name) } -+ -+#define DEFINE_SPINLOCK(name) \ -+ spinlock_t name = __SPIN_LOCK_UNLOCKED(name) -+ -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/srcu.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/srcu.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/srcu.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/srcu.h 2017-04-18 17:54:26.000000000 +0200 -@@ -84,10 +84,10 @@ - - void process_srcu(struct work_struct *work); - --#define __SRCU_STRUCT_INIT(name) \ -+#define __SRCU_STRUCT_INIT(name, pcpu_name) \ - { \ - .completed = -300, \ -- .per_cpu_ref = &name##_srcu_array, \ -+ .per_cpu_ref = &pcpu_name, \ - .queue_lock = __SPIN_LOCK_UNLOCKED(name.queue_lock), \ - .running = false, \ - .batch_queue = RCU_BATCH_INIT(name.batch_queue), \ -@@ -119,7 +119,7 @@ - */ - #define __DEFINE_SRCU(name, is_static) \ - static DEFINE_PER_CPU(struct srcu_struct_array, name##_srcu_array);\ -- is_static struct srcu_struct name = __SRCU_STRUCT_INIT(name) -+ is_static struct srcu_struct name = __SRCU_STRUCT_INIT(name, name##_srcu_array) - #define DEFINE_SRCU(name) __DEFINE_SRCU(name, /* not static */) - #define DEFINE_STATIC_SRCU(name) __DEFINE_SRCU(name, static) - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/suspend.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/suspend.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/suspend.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/suspend.h 2017-04-18 17:54:26.000000000 +0200 -@@ -193,6 +193,12 @@ - void (*end)(void); - }; - -+#if defined(CONFIG_SUSPEND) || defined(CONFIG_HIBERNATION) -+extern bool pm_in_action; -+#else -+# define pm_in_action false -+#endif -+ - #ifdef CONFIG_SUSPEND - /** - * suspend_set_ops - set platform dependent suspend operations -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/swait.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/swait.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/swait.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/swait.h 2017-04-18 17:54:26.000000000 +0200 -@@ -87,6 +87,7 @@ - extern void swake_up(struct swait_queue_head *q); - extern void swake_up_all(struct swait_queue_head *q); - extern void swake_up_locked(struct swait_queue_head *q); -+extern void swake_up_all_locked(struct swait_queue_head *q); - - extern void __prepare_to_swait(struct swait_queue_head *q, struct swait_queue *wait); - extern void prepare_to_swait(struct swait_queue_head *q, struct swait_queue *wait, int state); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/swap.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/swap.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/swap.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/swap.h 2017-04-18 17:54:26.000000000 +0200 -@@ -11,6 +11,7 @@ - #include - #include - #include -+#include - #include - - struct notifier_block; -@@ -247,7 +248,8 @@ - void *workingset_eviction(struct address_space *mapping, struct page *page); - bool workingset_refault(void *shadow); - void workingset_activation(struct page *page); --extern struct list_lru workingset_shadow_nodes; -+extern struct list_lru __workingset_shadow_nodes; -+DECLARE_LOCAL_IRQ_LOCK(workingset_shadow_lock); - - static inline unsigned int workingset_node_pages(struct radix_tree_node *node) - { -@@ -292,6 +294,7 @@ - - - /* linux/mm/swap.c */ -+DECLARE_LOCAL_IRQ_LOCK(swapvec_lock); - extern void lru_cache_add(struct page *); - extern void lru_cache_add_anon(struct page *page); - extern void lru_cache_add_file(struct page *page); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/swork.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/swork.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/swork.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/swork.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,24 @@ -+#ifndef _LINUX_SWORK_H -+#define _LINUX_SWORK_H -+ -+#include -+ -+struct swork_event { -+ struct list_head item; -+ unsigned long flags; -+ void (*func)(struct swork_event *); -+}; -+ -+static inline void INIT_SWORK(struct swork_event *event, -+ void (*func)(struct swork_event *)) -+{ -+ event->flags = 0; -+ event->func = func; -+} -+ -+bool swork_queue(struct swork_event *sev); -+ -+int swork_get(void); -+void swork_put(void); -+ -+#endif /* _LINUX_SWORK_H */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/thread_info.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/thread_info.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/thread_info.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/thread_info.h 2017-04-18 17:54:26.000000000 +0200 -@@ -107,7 +107,17 @@ - #define test_thread_flag(flag) \ - test_ti_thread_flag(current_thread_info(), flag) - --#define tif_need_resched() test_thread_flag(TIF_NEED_RESCHED) -+#ifdef CONFIG_PREEMPT_LAZY -+#define tif_need_resched() (test_thread_flag(TIF_NEED_RESCHED) || \ -+ test_thread_flag(TIF_NEED_RESCHED_LAZY)) -+#define tif_need_resched_now() (test_thread_flag(TIF_NEED_RESCHED)) -+#define tif_need_resched_lazy() test_thread_flag(TIF_NEED_RESCHED_LAZY)) -+ -+#else -+#define tif_need_resched() test_thread_flag(TIF_NEED_RESCHED) -+#define tif_need_resched_now() test_thread_flag(TIF_NEED_RESCHED) -+#define tif_need_resched_lazy() 0 -+#endif - - #ifndef CONFIG_HAVE_ARCH_WITHIN_STACK_FRAMES - static inline int arch_within_stack_frames(const void * const stack, -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/timer.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/timer.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/timer.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/timer.h 2017-04-18 17:54:26.000000000 +0200 -@@ -241,7 +241,7 @@ - - extern int try_to_del_timer_sync(struct timer_list *timer); - --#ifdef CONFIG_SMP -+#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL) - extern int del_timer_sync(struct timer_list *timer); - #else - # define del_timer_sync(t) del_timer(t) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/trace_events.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/trace_events.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/trace_events.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/trace_events.h 2017-04-18 17:54:26.000000000 +0200 -@@ -56,6 +56,9 @@ - unsigned char flags; - unsigned char preempt_count; - int pid; -+ unsigned short migrate_disable; -+ unsigned short padding; -+ unsigned char preempt_lazy_count; - }; - - #define TRACE_EVENT_TYPE_MAX \ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/uaccess.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/uaccess.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/uaccess.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/uaccess.h 2017-04-18 17:54:26.000000000 +0200 -@@ -24,6 +24,7 @@ - */ - static inline void pagefault_disable(void) - { -+ migrate_disable(); - pagefault_disabled_inc(); - /* - * make sure to have issued the store before a pagefault -@@ -40,6 +41,7 @@ - */ - barrier(); - pagefault_disabled_dec(); -+ migrate_enable(); - } - - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/uprobes.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/uprobes.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/uprobes.h 2017-04-16 10:38:26.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/uprobes.h 2017-04-18 17:54:26.000000000 +0200 -@@ -27,6 +27,7 @@ - #include - #include - #include -+#include - - struct vm_area_struct; - struct mm_struct; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/vmstat.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/vmstat.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/vmstat.h 2017-04-16 10:38:27.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/vmstat.h 2017-04-18 17:54:26.000000000 +0200 -@@ -33,7 +33,9 @@ - */ - static inline void __count_vm_event(enum vm_event_item item) - { -+ preempt_disable_rt(); - raw_cpu_inc(vm_event_states.event[item]); -+ preempt_enable_rt(); - } - - static inline void count_vm_event(enum vm_event_item item) -@@ -43,7 +45,9 @@ - - static inline void __count_vm_events(enum vm_event_item item, long delta) - { -+ preempt_disable_rt(); - raw_cpu_add(vm_event_states.event[item], delta); -+ preempt_enable_rt(); - } - - static inline void count_vm_events(enum vm_event_item item, long delta) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/wait.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/wait.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/linux/wait.h 2017-04-16 10:38:27.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/linux/wait.h 2017-04-18 17:54:26.000000000 +0200 -@@ -8,6 +8,7 @@ - #include - #include - #include -+#include - - typedef struct __wait_queue wait_queue_t; - typedef int (*wait_queue_func_t)(wait_queue_t *wait, unsigned mode, int flags, void *key); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/net/dst.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/net/dst.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/net/dst.h 2017-04-16 10:38:27.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/net/dst.h 2017-04-18 17:54:26.000000000 +0200 -@@ -446,7 +446,7 @@ - static inline int dst_neigh_output(struct dst_entry *dst, struct neighbour *n, - struct sk_buff *skb) - { -- const struct hh_cache *hh; -+ struct hh_cache *hh; - - if (dst->pending_confirm) { - unsigned long now = jiffies; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/net/gen_stats.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/net/gen_stats.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/net/gen_stats.h 2017-04-16 10:38:27.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/net/gen_stats.h 2017-04-18 17:54:26.000000000 +0200 -@@ -5,6 +5,7 @@ - #include - #include - #include -+#include - - struct gnet_stats_basic_cpu { - struct gnet_stats_basic_packed bstats; -@@ -33,11 +34,11 @@ - spinlock_t *lock, struct gnet_dump *d, - int padattr); - --int gnet_stats_copy_basic(const seqcount_t *running, -+int gnet_stats_copy_basic(net_seqlock_t *running, - struct gnet_dump *d, - struct gnet_stats_basic_cpu __percpu *cpu, - struct gnet_stats_basic_packed *b); --void __gnet_stats_copy_basic(const seqcount_t *running, -+void __gnet_stats_copy_basic(net_seqlock_t *running, - struct gnet_stats_basic_packed *bstats, - struct gnet_stats_basic_cpu __percpu *cpu, - struct gnet_stats_basic_packed *b); -@@ -55,14 +56,14 @@ - struct gnet_stats_basic_cpu __percpu *cpu_bstats, - struct gnet_stats_rate_est64 *rate_est, - spinlock_t *stats_lock, -- seqcount_t *running, struct nlattr *opt); -+ net_seqlock_t *running, struct nlattr *opt); - void gen_kill_estimator(struct gnet_stats_basic_packed *bstats, - struct gnet_stats_rate_est64 *rate_est); - int gen_replace_estimator(struct gnet_stats_basic_packed *bstats, - struct gnet_stats_basic_cpu __percpu *cpu_bstats, - struct gnet_stats_rate_est64 *rate_est, - spinlock_t *stats_lock, -- seqcount_t *running, struct nlattr *opt); -+ net_seqlock_t *running, struct nlattr *opt); - bool gen_estimator_active(const struct gnet_stats_basic_packed *bstats, - const struct gnet_stats_rate_est64 *rate_est); - #endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/net/neighbour.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/net/neighbour.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/net/neighbour.h 2017-04-16 10:38:27.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/net/neighbour.h 2017-04-18 17:54:26.000000000 +0200 -@@ -446,7 +446,7 @@ - } - #endif - --static inline int neigh_hh_output(const struct hh_cache *hh, struct sk_buff *skb) -+static inline int neigh_hh_output(struct hh_cache *hh, struct sk_buff *skb) - { - unsigned int seq; - int hh_len; -@@ -501,7 +501,7 @@ - - #define NEIGH_CB(skb) ((struct neighbour_cb *)(skb)->cb) - --static inline void neigh_ha_snapshot(char *dst, const struct neighbour *n, -+static inline void neigh_ha_snapshot(char *dst, struct neighbour *n, - const struct net_device *dev) - { - unsigned int seq; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/net/net_seq_lock.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/net/net_seq_lock.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/net/net_seq_lock.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/net/net_seq_lock.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,15 @@ -+#ifndef __NET_NET_SEQ_LOCK_H__ -+#define __NET_NET_SEQ_LOCK_H__ -+ -+#ifdef CONFIG_PREEMPT_RT_BASE -+# define net_seqlock_t seqlock_t -+# define net_seq_begin(__r) read_seqbegin(__r) -+# define net_seq_retry(__r, __s) read_seqretry(__r, __s) -+ -+#else -+# define net_seqlock_t seqcount_t -+# define net_seq_begin(__r) read_seqcount_begin(__r) -+# define net_seq_retry(__r, __s) read_seqcount_retry(__r, __s) -+#endif -+ -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/net/netns/ipv4.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/net/netns/ipv4.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/net/netns/ipv4.h 2017-04-16 10:38:27.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/net/netns/ipv4.h 2017-04-18 17:54:26.000000000 +0200 -@@ -69,6 +69,7 @@ - - int sysctl_icmp_echo_ignore_all; - int sysctl_icmp_echo_ignore_broadcasts; -+ int sysctl_icmp_echo_sysrq; - int sysctl_icmp_ignore_bogus_error_responses; - int sysctl_icmp_ratelimit; - int sysctl_icmp_ratemask; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/net/sch_generic.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/net/sch_generic.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/net/sch_generic.h 2017-04-16 10:38:27.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/net/sch_generic.h 2017-04-18 17:54:26.000000000 +0200 -@@ -10,6 +10,7 @@ - #include - #include - #include -+#include - - struct Qdisc_ops; - struct qdisc_walker; -@@ -86,7 +87,7 @@ - struct sk_buff *gso_skb ____cacheline_aligned_in_smp; - struct qdisc_skb_head q; - struct gnet_stats_basic_packed bstats; -- seqcount_t running; -+ net_seqlock_t running; - struct gnet_stats_queue qstats; - unsigned long state; - struct Qdisc *next_sched; -@@ -98,13 +99,22 @@ - spinlock_t busylock ____cacheline_aligned_in_smp; - }; - --static inline bool qdisc_is_running(const struct Qdisc *qdisc) -+static inline bool qdisc_is_running(struct Qdisc *qdisc) - { -+#ifdef CONFIG_PREEMPT_RT_BASE -+ return spin_is_locked(&qdisc->running.lock) ? true : false; -+#else - return (raw_read_seqcount(&qdisc->running) & 1) ? true : false; -+#endif - } - - static inline bool qdisc_run_begin(struct Qdisc *qdisc) - { -+#ifdef CONFIG_PREEMPT_RT_BASE -+ if (try_write_seqlock(&qdisc->running)) -+ return true; -+ return false; -+#else - if (qdisc_is_running(qdisc)) - return false; - /* Variant of write_seqcount_begin() telling lockdep a trylock -@@ -113,11 +123,16 @@ - raw_write_seqcount_begin(&qdisc->running); - seqcount_acquire(&qdisc->running.dep_map, 0, 1, _RET_IP_); - return true; -+#endif - } - - static inline void qdisc_run_end(struct Qdisc *qdisc) - { -+#ifdef CONFIG_PREEMPT_RT_BASE -+ write_sequnlock(&qdisc->running); -+#else - write_seqcount_end(&qdisc->running); -+#endif - } - - static inline bool qdisc_may_bulk(const struct Qdisc *qdisc) -@@ -308,7 +323,7 @@ - return qdisc_lock(root); - } - --static inline seqcount_t *qdisc_root_sleeping_running(const struct Qdisc *qdisc) -+static inline net_seqlock_t *qdisc_root_sleeping_running(const struct Qdisc *qdisc) - { - struct Qdisc *root = qdisc_root_sleeping(qdisc); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/trace/events/hist.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/trace/events/hist.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/trace/events/hist.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/trace/events/hist.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,73 @@ -+#undef TRACE_SYSTEM -+#define TRACE_SYSTEM hist -+ -+#if !defined(_TRACE_HIST_H) || defined(TRACE_HEADER_MULTI_READ) -+#define _TRACE_HIST_H -+ -+#include "latency_hist.h" -+#include -+ -+#if !defined(CONFIG_PREEMPT_OFF_HIST) && !defined(CONFIG_INTERRUPT_OFF_HIST) -+#define trace_preemptirqsoff_hist(a, b) -+#define trace_preemptirqsoff_hist_rcuidle(a, b) -+#else -+TRACE_EVENT(preemptirqsoff_hist, -+ -+ TP_PROTO(int reason, int starthist), -+ -+ TP_ARGS(reason, starthist), -+ -+ TP_STRUCT__entry( -+ __field(int, reason) -+ __field(int, starthist) -+ ), -+ -+ TP_fast_assign( -+ __entry->reason = reason; -+ __entry->starthist = starthist; -+ ), -+ -+ TP_printk("reason=%s starthist=%s", getaction(__entry->reason), -+ __entry->starthist ? "start" : "stop") -+); -+#endif -+ -+#ifndef CONFIG_MISSED_TIMER_OFFSETS_HIST -+#define trace_hrtimer_interrupt(a, b, c, d) -+#else -+TRACE_EVENT(hrtimer_interrupt, -+ -+ TP_PROTO(int cpu, long long offset, struct task_struct *curr, -+ struct task_struct *task), -+ -+ TP_ARGS(cpu, offset, curr, task), -+ -+ TP_STRUCT__entry( -+ __field(int, cpu) -+ __field(long long, offset) -+ __array(char, ccomm, TASK_COMM_LEN) -+ __field(int, cprio) -+ __array(char, tcomm, TASK_COMM_LEN) -+ __field(int, tprio) -+ ), -+ -+ TP_fast_assign( -+ __entry->cpu = cpu; -+ __entry->offset = offset; -+ memcpy(__entry->ccomm, curr->comm, TASK_COMM_LEN); -+ __entry->cprio = curr->prio; -+ memcpy(__entry->tcomm, task != NULL ? task->comm : "", -+ task != NULL ? TASK_COMM_LEN : 7); -+ __entry->tprio = task != NULL ? task->prio : -1; -+ ), -+ -+ TP_printk("cpu=%d offset=%lld curr=%s[%d] thread=%s[%d]", -+ __entry->cpu, __entry->offset, __entry->ccomm, -+ __entry->cprio, __entry->tcomm, __entry->tprio) -+); -+#endif -+ -+#endif /* _TRACE_HIST_H */ -+ -+/* This part must be outside protection */ -+#include -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/trace/events/latency_hist.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/trace/events/latency_hist.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/include/trace/events/latency_hist.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/include/trace/events/latency_hist.h 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,29 @@ -+#ifndef _LATENCY_HIST_H -+#define _LATENCY_HIST_H -+ -+enum hist_action { -+ IRQS_ON, -+ PREEMPT_ON, -+ TRACE_STOP, -+ IRQS_OFF, -+ PREEMPT_OFF, -+ TRACE_START, -+}; -+ -+static char *actions[] = { -+ "IRQS_ON", -+ "PREEMPT_ON", -+ "TRACE_STOP", -+ "IRQS_OFF", -+ "PREEMPT_OFF", -+ "TRACE_START", -+}; -+ -+static inline char *getaction(int action) -+{ -+ if (action >= 0 && action <= sizeof(actions)/sizeof(actions[0])) -+ return actions[action]; -+ return "unknown"; -+} -+ -+#endif /* _LATENCY_HIST_H */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/init/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/init/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/init/Kconfig 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/init/Kconfig 2017-04-18 17:54:26.000000000 +0200 -@@ -506,7 +506,7 @@ - - config RCU_EXPERT - bool "Make expert-level adjustments to RCU configuration" -- default n -+ default y if PREEMPT_RT_FULL - help - This option needs to be enabled if you wish to make - expert-level adjustments to RCU configuration. By default, -@@ -623,7 +623,7 @@ - - config RCU_FAST_NO_HZ - bool "Accelerate last non-dyntick-idle CPU's grace periods" -- depends on NO_HZ_COMMON && SMP && RCU_EXPERT -+ depends on NO_HZ_COMMON && SMP && RCU_EXPERT && !PREEMPT_RT_FULL - default n - help - This option permits CPUs to enter dynticks-idle state even if -@@ -650,7 +650,7 @@ - config RCU_BOOST - bool "Enable RCU priority boosting" - depends on RT_MUTEXES && PREEMPT_RCU && RCU_EXPERT -- default n -+ default y if PREEMPT_RT_FULL - help - This option boosts the priority of preempted RCU readers that - block the current preemptible RCU grace period for too long. -@@ -781,19 +781,6 @@ - - endchoice - --config RCU_EXPEDITE_BOOT -- bool -- default n -- help -- This option enables expedited grace periods at boot time, -- as if rcu_expedite_gp() had been invoked early in boot. -- The corresponding rcu_unexpedite_gp() is invoked from -- rcu_end_inkernel_boot(), which is intended to be invoked -- at the end of the kernel-only boot sequence, just before -- init is exec'ed. -- -- Accept the default if unsure. -- - endmenu # "RCU Subsystem" - - config BUILD_BIN2C -@@ -1064,6 +1051,7 @@ - config RT_GROUP_SCHED - bool "Group scheduling for SCHED_RR/FIFO" - depends on CGROUP_SCHED -+ depends on !PREEMPT_RT_FULL - default n - help - This feature lets you explicitly allocate real CPU bandwidth -@@ -1772,6 +1760,7 @@ - - config SLAB - bool "SLAB" -+ depends on !PREEMPT_RT_FULL - select HAVE_HARDENED_USERCOPY_ALLOCATOR - help - The regular slab allocator that is established and known to work -@@ -1792,6 +1781,7 @@ - config SLOB - depends on EXPERT - bool "SLOB (Simple Allocator)" -+ depends on !PREEMPT_RT_FULL - help - SLOB replaces the stock allocator with a drastically simpler - allocator. SLOB is generally more space efficient but -@@ -1810,7 +1800,7 @@ - - config SLUB_CPU_PARTIAL - default y -- depends on SLUB && SMP -+ depends on SLUB && SMP && !PREEMPT_RT_FULL - bool "SLUB per cpu partial cache" - help - Per cpu partial caches accellerate objects allocation and freeing -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/init/Makefile linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/init/Makefile ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/init/Makefile 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/init/Makefile 2017-04-18 17:54:26.000000000 +0200 -@@ -35,4 +35,4 @@ - include/generated/compile.h: FORCE - @$($(quiet)chk_compile.h) - $(Q)$(CONFIG_SHELL) $(srctree)/scripts/mkcompile_h $@ \ -- "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CC) $(KBUILD_CFLAGS)" -+ "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CONFIG_PREEMPT_RT_FULL)" "$(CC) $(KBUILD_CFLAGS)" -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/init/main.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/init/main.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/init/main.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/init/main.c 2017-04-18 17:54:26.000000000 +0200 -@@ -507,6 +507,7 @@ - setup_command_line(command_line); - setup_nr_cpu_ids(); - setup_per_cpu_areas(); -+ softirq_early_init(); - boot_cpu_state_init(); - smp_prepare_boot_cpu(); /* arch-specific boot-cpu hooks */ - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/ipc/sem.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/ipc/sem.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/ipc/sem.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/ipc/sem.c 2017-04-18 17:54:26.000000000 +0200 -@@ -712,6 +712,13 @@ - static void wake_up_sem_queue_prepare(struct list_head *pt, - struct sem_queue *q, int error) - { -+#ifdef CONFIG_PREEMPT_RT_BASE -+ struct task_struct *p = q->sleeper; -+ get_task_struct(p); -+ q->status = error; -+ wake_up_process(p); -+ put_task_struct(p); -+#else - if (list_empty(pt)) { - /* - * Hold preempt off so that we don't get preempted and have the -@@ -723,6 +730,7 @@ - q->pid = error; - - list_add_tail(&q->list, pt); -+#endif - } - - /** -@@ -736,6 +744,7 @@ - */ - static void wake_up_sem_queue_do(struct list_head *pt) - { -+#ifndef CONFIG_PREEMPT_RT_BASE - struct sem_queue *q, *t; - int did_something; - -@@ -748,6 +757,7 @@ - } - if (did_something) - preempt_enable(); -+#endif - } - - static void unlink_queue(struct sem_array *sma, struct sem_queue *q) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/Kconfig.locks linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/Kconfig.locks ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/Kconfig.locks 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/Kconfig.locks 2017-04-18 17:54:26.000000000 +0200 -@@ -225,11 +225,11 @@ - - config MUTEX_SPIN_ON_OWNER - def_bool y -- depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW -+ depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL - - config RWSEM_SPIN_ON_OWNER - def_bool y -- depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW -+ depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL - - config LOCK_SPIN_ON_OWNER - def_bool y -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/Kconfig.preempt linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/Kconfig.preempt ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/Kconfig.preempt 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/Kconfig.preempt 2017-04-18 17:54:26.000000000 +0200 -@@ -1,3 +1,16 @@ -+config PREEMPT -+ bool -+ select PREEMPT_COUNT -+ -+config PREEMPT_RT_BASE -+ bool -+ select PREEMPT -+ -+config HAVE_PREEMPT_LAZY -+ bool -+ -+config PREEMPT_LAZY -+ def_bool y if HAVE_PREEMPT_LAZY && PREEMPT_RT_FULL - - choice - prompt "Preemption Model" -@@ -33,9 +46,9 @@ - - Select this if you are building a kernel for a desktop system. - --config PREEMPT -+config PREEMPT__LL - bool "Preemptible Kernel (Low-Latency Desktop)" -- select PREEMPT_COUNT -+ select PREEMPT - select UNINLINE_SPIN_UNLOCK if !ARCH_INLINE_SPIN_UNLOCK - help - This option reduces the latency of the kernel by making -@@ -52,6 +65,22 @@ - embedded system with latency requirements in the milliseconds - range. - -+config PREEMPT_RTB -+ bool "Preemptible Kernel (Basic RT)" -+ select PREEMPT_RT_BASE -+ help -+ This option is basically the same as (Low-Latency Desktop) but -+ enables changes which are preliminary for the full preemptible -+ RT kernel. -+ -+config PREEMPT_RT_FULL -+ bool "Fully Preemptible Kernel (RT)" -+ depends on IRQ_FORCED_THREADING -+ select PREEMPT_RT_BASE -+ select PREEMPT_RCU -+ help -+ All and everything -+ - endchoice - - config PREEMPT_COUNT -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/cgroup.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/cgroup.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/cgroup.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/cgroup.c 2017-04-18 17:54:26.000000000 +0200 -@@ -5040,10 +5040,10 @@ - queue_work(cgroup_destroy_wq, &css->destroy_work); - } - --static void css_release_work_fn(struct work_struct *work) -+static void css_release_work_fn(struct swork_event *sev) - { - struct cgroup_subsys_state *css = -- container_of(work, struct cgroup_subsys_state, destroy_work); -+ container_of(sev, struct cgroup_subsys_state, destroy_swork); - struct cgroup_subsys *ss = css->ss; - struct cgroup *cgrp = css->cgroup; - -@@ -5086,8 +5086,8 @@ - struct cgroup_subsys_state *css = - container_of(ref, struct cgroup_subsys_state, refcnt); - -- INIT_WORK(&css->destroy_work, css_release_work_fn); -- queue_work(cgroup_destroy_wq, &css->destroy_work); -+ INIT_SWORK(&css->destroy_swork, css_release_work_fn); -+ swork_queue(&css->destroy_swork); - } - - static void init_and_link_css(struct cgroup_subsys_state *css, -@@ -5739,6 +5739,7 @@ - */ - cgroup_destroy_wq = alloc_workqueue("cgroup_destroy", 0, 1); - BUG_ON(!cgroup_destroy_wq); -+ BUG_ON(swork_get()); - - /* - * Used to destroy pidlists and separate to serve as flush domain. -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/cpu.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/cpu.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/cpu.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/cpu.c 2017-04-18 17:54:26.000000000 +0200 -@@ -239,6 +239,289 @@ - #define cpuhp_lock_acquire() lock_map_acquire(&cpu_hotplug.dep_map) - #define cpuhp_lock_release() lock_map_release(&cpu_hotplug.dep_map) - -+/** -+ * hotplug_pcp - per cpu hotplug descriptor -+ * @unplug: set when pin_current_cpu() needs to sync tasks -+ * @sync_tsk: the task that waits for tasks to finish pinned sections -+ * @refcount: counter of tasks in pinned sections -+ * @grab_lock: set when the tasks entering pinned sections should wait -+ * @synced: notifier for @sync_tsk to tell cpu_down it's finished -+ * @mutex: the mutex to make tasks wait (used when @grab_lock is true) -+ * @mutex_init: zero if the mutex hasn't been initialized yet. -+ * -+ * Although @unplug and @sync_tsk may point to the same task, the @unplug -+ * is used as a flag and still exists after @sync_tsk has exited and -+ * @sync_tsk set to NULL. -+ */ -+struct hotplug_pcp { -+ struct task_struct *unplug; -+ struct task_struct *sync_tsk; -+ int refcount; -+ int grab_lock; -+ struct completion synced; -+ struct completion unplug_wait; -+#ifdef CONFIG_PREEMPT_RT_FULL -+ /* -+ * Note, on PREEMPT_RT, the hotplug lock must save the state of -+ * the task, otherwise the mutex will cause the task to fail -+ * to sleep when required. (Because it's called from migrate_disable()) -+ * -+ * The spinlock_t on PREEMPT_RT is a mutex that saves the task's -+ * state. -+ */ -+ spinlock_t lock; -+#else -+ struct mutex mutex; -+#endif -+ int mutex_init; -+}; -+ -+#ifdef CONFIG_PREEMPT_RT_FULL -+# define hotplug_lock(hp) rt_spin_lock__no_mg(&(hp)->lock) -+# define hotplug_unlock(hp) rt_spin_unlock__no_mg(&(hp)->lock) -+#else -+# define hotplug_lock(hp) mutex_lock(&(hp)->mutex) -+# define hotplug_unlock(hp) mutex_unlock(&(hp)->mutex) -+#endif -+ -+static DEFINE_PER_CPU(struct hotplug_pcp, hotplug_pcp); -+ -+/** -+ * pin_current_cpu - Prevent the current cpu from being unplugged -+ * -+ * Lightweight version of get_online_cpus() to prevent cpu from being -+ * unplugged when code runs in a migration disabled region. -+ * -+ * Must be called with preemption disabled (preempt_count = 1)! -+ */ -+void pin_current_cpu(void) -+{ -+ struct hotplug_pcp *hp; -+ int force = 0; -+ -+retry: -+ hp = this_cpu_ptr(&hotplug_pcp); -+ -+ if (!hp->unplug || hp->refcount || force || preempt_count() > 1 || -+ hp->unplug == current) { -+ hp->refcount++; -+ return; -+ } -+ if (hp->grab_lock) { -+ preempt_enable(); -+ hotplug_lock(hp); -+ hotplug_unlock(hp); -+ } else { -+ preempt_enable(); -+ /* -+ * Try to push this task off of this CPU. -+ */ -+ if (!migrate_me()) { -+ preempt_disable(); -+ hp = this_cpu_ptr(&hotplug_pcp); -+ if (!hp->grab_lock) { -+ /* -+ * Just let it continue it's already pinned -+ * or about to sleep. -+ */ -+ force = 1; -+ goto retry; -+ } -+ preempt_enable(); -+ } -+ } -+ preempt_disable(); -+ goto retry; -+} -+ -+/** -+ * unpin_current_cpu - Allow unplug of current cpu -+ * -+ * Must be called with preemption or interrupts disabled! -+ */ -+void unpin_current_cpu(void) -+{ -+ struct hotplug_pcp *hp = this_cpu_ptr(&hotplug_pcp); -+ -+ WARN_ON(hp->refcount <= 0); -+ -+ /* This is safe. sync_unplug_thread is pinned to this cpu */ -+ if (!--hp->refcount && hp->unplug && hp->unplug != current) -+ wake_up_process(hp->unplug); -+} -+ -+static void wait_for_pinned_cpus(struct hotplug_pcp *hp) -+{ -+ set_current_state(TASK_UNINTERRUPTIBLE); -+ while (hp->refcount) { -+ schedule_preempt_disabled(); -+ set_current_state(TASK_UNINTERRUPTIBLE); -+ } -+} -+ -+static int sync_unplug_thread(void *data) -+{ -+ struct hotplug_pcp *hp = data; -+ -+ wait_for_completion(&hp->unplug_wait); -+ preempt_disable(); -+ hp->unplug = current; -+ wait_for_pinned_cpus(hp); -+ -+ /* -+ * This thread will synchronize the cpu_down() with threads -+ * that have pinned the CPU. When the pinned CPU count reaches -+ * zero, we inform the cpu_down code to continue to the next step. -+ */ -+ set_current_state(TASK_UNINTERRUPTIBLE); -+ preempt_enable(); -+ complete(&hp->synced); -+ -+ /* -+ * If all succeeds, the next step will need tasks to wait till -+ * the CPU is offline before continuing. To do this, the grab_lock -+ * is set and tasks going into pin_current_cpu() will block on the -+ * mutex. But we still need to wait for those that are already in -+ * pinned CPU sections. If the cpu_down() failed, the kthread_should_stop() -+ * will kick this thread out. -+ */ -+ while (!hp->grab_lock && !kthread_should_stop()) { -+ schedule(); -+ set_current_state(TASK_UNINTERRUPTIBLE); -+ } -+ -+ /* Make sure grab_lock is seen before we see a stale completion */ -+ smp_mb(); -+ -+ /* -+ * Now just before cpu_down() enters stop machine, we need to make -+ * sure all tasks that are in pinned CPU sections are out, and new -+ * tasks will now grab the lock, keeping them from entering pinned -+ * CPU sections. -+ */ -+ if (!kthread_should_stop()) { -+ preempt_disable(); -+ wait_for_pinned_cpus(hp); -+ preempt_enable(); -+ complete(&hp->synced); -+ } -+ -+ set_current_state(TASK_UNINTERRUPTIBLE); -+ while (!kthread_should_stop()) { -+ schedule(); -+ set_current_state(TASK_UNINTERRUPTIBLE); -+ } -+ set_current_state(TASK_RUNNING); -+ -+ /* -+ * Force this thread off this CPU as it's going down and -+ * we don't want any more work on this CPU. -+ */ -+ current->flags &= ~PF_NO_SETAFFINITY; -+ set_cpus_allowed_ptr(current, cpu_present_mask); -+ migrate_me(); -+ return 0; -+} -+ -+static void __cpu_unplug_sync(struct hotplug_pcp *hp) -+{ -+ wake_up_process(hp->sync_tsk); -+ wait_for_completion(&hp->synced); -+} -+ -+static void __cpu_unplug_wait(unsigned int cpu) -+{ -+ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu); -+ -+ complete(&hp->unplug_wait); -+ wait_for_completion(&hp->synced); -+} -+ -+/* -+ * Start the sync_unplug_thread on the target cpu and wait for it to -+ * complete. -+ */ -+static int cpu_unplug_begin(unsigned int cpu) -+{ -+ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu); -+ int err; -+ -+ /* Protected by cpu_hotplug.lock */ -+ if (!hp->mutex_init) { -+#ifdef CONFIG_PREEMPT_RT_FULL -+ spin_lock_init(&hp->lock); -+#else -+ mutex_init(&hp->mutex); -+#endif -+ hp->mutex_init = 1; -+ } -+ -+ /* Inform the scheduler to migrate tasks off this CPU */ -+ tell_sched_cpu_down_begin(cpu); -+ -+ init_completion(&hp->synced); -+ init_completion(&hp->unplug_wait); -+ -+ hp->sync_tsk = kthread_create(sync_unplug_thread, hp, "sync_unplug/%d", cpu); -+ if (IS_ERR(hp->sync_tsk)) { -+ err = PTR_ERR(hp->sync_tsk); -+ hp->sync_tsk = NULL; -+ return err; -+ } -+ kthread_bind(hp->sync_tsk, cpu); -+ -+ /* -+ * Wait for tasks to get out of the pinned sections, -+ * it's still OK if new tasks enter. Some CPU notifiers will -+ * wait for tasks that are going to enter these sections and -+ * we must not have them block. -+ */ -+ wake_up_process(hp->sync_tsk); -+ return 0; -+} -+ -+static void cpu_unplug_sync(unsigned int cpu) -+{ -+ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu); -+ -+ init_completion(&hp->synced); -+ /* The completion needs to be initialzied before setting grab_lock */ -+ smp_wmb(); -+ -+ /* Grab the mutex before setting grab_lock */ -+ hotplug_lock(hp); -+ hp->grab_lock = 1; -+ -+ /* -+ * The CPU notifiers have been completed. -+ * Wait for tasks to get out of pinned CPU sections and have new -+ * tasks block until the CPU is completely down. -+ */ -+ __cpu_unplug_sync(hp); -+ -+ /* All done with the sync thread */ -+ kthread_stop(hp->sync_tsk); -+ hp->sync_tsk = NULL; -+} -+ -+static void cpu_unplug_done(unsigned int cpu) -+{ -+ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu); -+ -+ hp->unplug = NULL; -+ /* Let all tasks know cpu unplug is finished before cleaning up */ -+ smp_wmb(); -+ -+ if (hp->sync_tsk) -+ kthread_stop(hp->sync_tsk); -+ -+ if (hp->grab_lock) { -+ hotplug_unlock(hp); -+ /* protected by cpu_hotplug.lock */ -+ hp->grab_lock = 0; -+ } -+ tell_sched_cpu_down_done(cpu); -+} - - void get_online_cpus(void) - { -@@ -789,10 +1072,14 @@ - struct cpuhp_cpu_state *st = per_cpu_ptr(&cpuhp_state, cpu); - int err; - -+ __cpu_unplug_wait(cpu); - /* Park the smpboot threads */ - kthread_park(per_cpu_ptr(&cpuhp_state, cpu)->thread); - smpboot_park_threads(cpu); - -+ /* Notifiers are done. Don't let any more tasks pin this CPU. */ -+ cpu_unplug_sync(cpu); -+ - /* - * Prevent irq alloc/free while the dying cpu reorganizes the - * interrupt affinities. -@@ -877,6 +1164,9 @@ - struct cpuhp_cpu_state *st = per_cpu_ptr(&cpuhp_state, cpu); - int prev_state, ret = 0; - bool hasdied = false; -+ int mycpu; -+ cpumask_var_t cpumask; -+ cpumask_var_t cpumask_org; - - if (num_online_cpus() == 1) - return -EBUSY; -@@ -884,7 +1174,34 @@ - if (!cpu_present(cpu)) - return -EINVAL; - -+ /* Move the downtaker off the unplug cpu */ -+ if (!alloc_cpumask_var(&cpumask, GFP_KERNEL)) -+ return -ENOMEM; -+ if (!alloc_cpumask_var(&cpumask_org, GFP_KERNEL)) { -+ free_cpumask_var(cpumask); -+ return -ENOMEM; -+ } -+ -+ cpumask_copy(cpumask_org, tsk_cpus_allowed(current)); -+ cpumask_andnot(cpumask, cpu_online_mask, cpumask_of(cpu)); -+ set_cpus_allowed_ptr(current, cpumask); -+ free_cpumask_var(cpumask); -+ migrate_disable(); -+ mycpu = smp_processor_id(); -+ if (mycpu == cpu) { -+ printk(KERN_ERR "Yuck! Still on unplug CPU\n!"); -+ migrate_enable(); -+ ret = -EBUSY; -+ goto restore_cpus; -+ } -+ -+ migrate_enable(); - cpu_hotplug_begin(); -+ ret = cpu_unplug_begin(cpu); -+ if (ret) { -+ printk("cpu_unplug_begin(%d) failed\n", cpu); -+ goto out_cancel; -+ } - - cpuhp_tasks_frozen = tasks_frozen; - -@@ -923,10 +1240,15 @@ - - hasdied = prev_state != st->state && st->state == CPUHP_OFFLINE; - out: -+ cpu_unplug_done(cpu); -+out_cancel: - cpu_hotplug_done(); - /* This post dead nonsense must die */ - if (!ret && hasdied) - cpu_notify_nofail(CPU_POST_DEAD, cpu); -+restore_cpus: -+ set_cpus_allowed_ptr(current, cpumask_org); -+ free_cpumask_var(cpumask_org); - return ret; - } - -@@ -1240,6 +1562,8 @@ - - #endif /* CONFIG_PM_SLEEP_SMP */ - -+int __boot_cpu_id; -+ - #endif /* CONFIG_SMP */ - - /* Boot processor state steps */ -@@ -1923,6 +2247,10 @@ - set_cpu_active(cpu, true); - set_cpu_present(cpu, true); - set_cpu_possible(cpu, true); -+ -+#ifdef CONFIG_SMP -+ __boot_cpu_id = cpu; -+#endif - } - - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/cpuset.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/cpuset.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/cpuset.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/cpuset.c 2017-04-18 17:54:26.000000000 +0200 -@@ -284,7 +284,7 @@ - */ - - static DEFINE_MUTEX(cpuset_mutex); --static DEFINE_SPINLOCK(callback_lock); -+static DEFINE_RAW_SPINLOCK(callback_lock); - - static struct workqueue_struct *cpuset_migrate_mm_wq; - -@@ -907,9 +907,9 @@ - continue; - rcu_read_unlock(); - -- spin_lock_irq(&callback_lock); -+ raw_spin_lock_irq(&callback_lock); - cpumask_copy(cp->effective_cpus, new_cpus); -- spin_unlock_irq(&callback_lock); -+ raw_spin_unlock_irq(&callback_lock); - - WARN_ON(!cgroup_subsys_on_dfl(cpuset_cgrp_subsys) && - !cpumask_equal(cp->cpus_allowed, cp->effective_cpus)); -@@ -974,9 +974,9 @@ - if (retval < 0) - return retval; - -- spin_lock_irq(&callback_lock); -+ raw_spin_lock_irq(&callback_lock); - cpumask_copy(cs->cpus_allowed, trialcs->cpus_allowed); -- spin_unlock_irq(&callback_lock); -+ raw_spin_unlock_irq(&callback_lock); - - /* use trialcs->cpus_allowed as a temp variable */ - update_cpumasks_hier(cs, trialcs->cpus_allowed); -@@ -1176,9 +1176,9 @@ - continue; - rcu_read_unlock(); - -- spin_lock_irq(&callback_lock); -+ raw_spin_lock_irq(&callback_lock); - cp->effective_mems = *new_mems; -- spin_unlock_irq(&callback_lock); -+ raw_spin_unlock_irq(&callback_lock); - - WARN_ON(!cgroup_subsys_on_dfl(cpuset_cgrp_subsys) && - !nodes_equal(cp->mems_allowed, cp->effective_mems)); -@@ -1246,9 +1246,9 @@ - if (retval < 0) - goto done; - -- spin_lock_irq(&callback_lock); -+ raw_spin_lock_irq(&callback_lock); - cs->mems_allowed = trialcs->mems_allowed; -- spin_unlock_irq(&callback_lock); -+ raw_spin_unlock_irq(&callback_lock); - - /* use trialcs->mems_allowed as a temp variable */ - update_nodemasks_hier(cs, &trialcs->mems_allowed); -@@ -1339,9 +1339,9 @@ - spread_flag_changed = ((is_spread_slab(cs) != is_spread_slab(trialcs)) - || (is_spread_page(cs) != is_spread_page(trialcs))); - -- spin_lock_irq(&callback_lock); -+ raw_spin_lock_irq(&callback_lock); - cs->flags = trialcs->flags; -- spin_unlock_irq(&callback_lock); -+ raw_spin_unlock_irq(&callback_lock); - - if (!cpumask_empty(trialcs->cpus_allowed) && balance_flag_changed) - rebuild_sched_domains_locked(); -@@ -1756,7 +1756,7 @@ - cpuset_filetype_t type = seq_cft(sf)->private; - int ret = 0; - -- spin_lock_irq(&callback_lock); -+ raw_spin_lock_irq(&callback_lock); - - switch (type) { - case FILE_CPULIST: -@@ -1775,7 +1775,7 @@ - ret = -EINVAL; - } - -- spin_unlock_irq(&callback_lock); -+ raw_spin_unlock_irq(&callback_lock); - return ret; - } - -@@ -1989,12 +1989,12 @@ - - cpuset_inc(); - -- spin_lock_irq(&callback_lock); -+ raw_spin_lock_irq(&callback_lock); - if (cgroup_subsys_on_dfl(cpuset_cgrp_subsys)) { - cpumask_copy(cs->effective_cpus, parent->effective_cpus); - cs->effective_mems = parent->effective_mems; - } -- spin_unlock_irq(&callback_lock); -+ raw_spin_unlock_irq(&callback_lock); - - if (!test_bit(CGRP_CPUSET_CLONE_CHILDREN, &css->cgroup->flags)) - goto out_unlock; -@@ -2021,12 +2021,12 @@ - } - rcu_read_unlock(); - -- spin_lock_irq(&callback_lock); -+ raw_spin_lock_irq(&callback_lock); - cs->mems_allowed = parent->mems_allowed; - cs->effective_mems = parent->mems_allowed; - cpumask_copy(cs->cpus_allowed, parent->cpus_allowed); - cpumask_copy(cs->effective_cpus, parent->cpus_allowed); -- spin_unlock_irq(&callback_lock); -+ raw_spin_unlock_irq(&callback_lock); - out_unlock: - mutex_unlock(&cpuset_mutex); - return 0; -@@ -2065,7 +2065,7 @@ - static void cpuset_bind(struct cgroup_subsys_state *root_css) - { - mutex_lock(&cpuset_mutex); -- spin_lock_irq(&callback_lock); -+ raw_spin_lock_irq(&callback_lock); - - if (cgroup_subsys_on_dfl(cpuset_cgrp_subsys)) { - cpumask_copy(top_cpuset.cpus_allowed, cpu_possible_mask); -@@ -2076,7 +2076,7 @@ - top_cpuset.mems_allowed = top_cpuset.effective_mems; - } - -- spin_unlock_irq(&callback_lock); -+ raw_spin_unlock_irq(&callback_lock); - mutex_unlock(&cpuset_mutex); - } - -@@ -2177,12 +2177,12 @@ - { - bool is_empty; - -- spin_lock_irq(&callback_lock); -+ raw_spin_lock_irq(&callback_lock); - cpumask_copy(cs->cpus_allowed, new_cpus); - cpumask_copy(cs->effective_cpus, new_cpus); - cs->mems_allowed = *new_mems; - cs->effective_mems = *new_mems; -- spin_unlock_irq(&callback_lock); -+ raw_spin_unlock_irq(&callback_lock); - - /* - * Don't call update_tasks_cpumask() if the cpuset becomes empty, -@@ -2219,10 +2219,10 @@ - if (nodes_empty(*new_mems)) - *new_mems = parent_cs(cs)->effective_mems; - -- spin_lock_irq(&callback_lock); -+ raw_spin_lock_irq(&callback_lock); - cpumask_copy(cs->effective_cpus, new_cpus); - cs->effective_mems = *new_mems; -- spin_unlock_irq(&callback_lock); -+ raw_spin_unlock_irq(&callback_lock); - - if (cpus_updated) - update_tasks_cpumask(cs); -@@ -2308,21 +2308,21 @@ - - /* synchronize cpus_allowed to cpu_active_mask */ - if (cpus_updated) { -- spin_lock_irq(&callback_lock); -+ raw_spin_lock_irq(&callback_lock); - if (!on_dfl) - cpumask_copy(top_cpuset.cpus_allowed, &new_cpus); - cpumask_copy(top_cpuset.effective_cpus, &new_cpus); -- spin_unlock_irq(&callback_lock); -+ raw_spin_unlock_irq(&callback_lock); - /* we don't mess with cpumasks of tasks in top_cpuset */ - } - - /* synchronize mems_allowed to N_MEMORY */ - if (mems_updated) { -- spin_lock_irq(&callback_lock); -+ raw_spin_lock_irq(&callback_lock); - if (!on_dfl) - top_cpuset.mems_allowed = new_mems; - top_cpuset.effective_mems = new_mems; -- spin_unlock_irq(&callback_lock); -+ raw_spin_unlock_irq(&callback_lock); - update_tasks_nodemask(&top_cpuset); - } - -@@ -2420,11 +2420,11 @@ - { - unsigned long flags; - -- spin_lock_irqsave(&callback_lock, flags); -+ raw_spin_lock_irqsave(&callback_lock, flags); - rcu_read_lock(); - guarantee_online_cpus(task_cs(tsk), pmask); - rcu_read_unlock(); -- spin_unlock_irqrestore(&callback_lock, flags); -+ raw_spin_unlock_irqrestore(&callback_lock, flags); - } - - void cpuset_cpus_allowed_fallback(struct task_struct *tsk) -@@ -2472,11 +2472,11 @@ - nodemask_t mask; - unsigned long flags; - -- spin_lock_irqsave(&callback_lock, flags); -+ raw_spin_lock_irqsave(&callback_lock, flags); - rcu_read_lock(); - guarantee_online_mems(task_cs(tsk), &mask); - rcu_read_unlock(); -- spin_unlock_irqrestore(&callback_lock, flags); -+ raw_spin_unlock_irqrestore(&callback_lock, flags); - - return mask; - } -@@ -2568,14 +2568,14 @@ - return true; - - /* Not hardwall and node outside mems_allowed: scan up cpusets */ -- spin_lock_irqsave(&callback_lock, flags); -+ raw_spin_lock_irqsave(&callback_lock, flags); - - rcu_read_lock(); - cs = nearest_hardwall_ancestor(task_cs(current)); - allowed = node_isset(node, cs->mems_allowed); - rcu_read_unlock(); - -- spin_unlock_irqrestore(&callback_lock, flags); -+ raw_spin_unlock_irqrestore(&callback_lock, flags); - return allowed; - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/debug/kdb/kdb_io.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/debug/kdb/kdb_io.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/debug/kdb/kdb_io.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/debug/kdb/kdb_io.c 2017-04-18 17:54:26.000000000 +0200 -@@ -554,7 +554,6 @@ - int linecount; - int colcount; - int logging, saved_loglevel = 0; -- int saved_trap_printk; - int got_printf_lock = 0; - int retlen = 0; - int fnd, len; -@@ -565,8 +564,6 @@ - unsigned long uninitialized_var(flags); - - preempt_disable(); -- saved_trap_printk = kdb_trap_printk; -- kdb_trap_printk = 0; - - /* Serialize kdb_printf if multiple cpus try to write at once. - * But if any cpu goes recursive in kdb, just print the output, -@@ -855,7 +852,6 @@ - } else { - __release(kdb_printf_lock); - } -- kdb_trap_printk = saved_trap_printk; - preempt_enable(); - return retlen; - } -@@ -865,9 +861,11 @@ - va_list ap; - int r; - -+ kdb_trap_printk++; - va_start(ap, fmt); - r = vkdb_printf(KDB_MSGSRC_INTERNAL, fmt, ap); - va_end(ap); -+ kdb_trap_printk--; - - return r; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/events/core.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/events/core.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/events/core.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/events/core.c 2017-04-18 17:54:26.000000000 +0200 -@@ -1050,6 +1050,7 @@ - raw_spin_lock_init(&cpuctx->hrtimer_lock); - hrtimer_init(timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS_PINNED); - timer->function = perf_mux_hrtimer_handler; -+ timer->irqsafe = 1; - } - - static int perf_mux_hrtimer_restart(struct perf_cpu_context *cpuctx) -@@ -8363,6 +8364,7 @@ - - hrtimer_init(&hwc->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); - hwc->hrtimer.function = perf_swevent_hrtimer; -+ hwc->hrtimer.irqsafe = 1; - - /* - * Since hrtimers have a fixed rate, we can do a static freq->period -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/exit.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/exit.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/exit.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/exit.c 2017-04-18 17:54:26.000000000 +0200 -@@ -143,7 +143,7 @@ - * Do this under ->siglock, we can race with another thread - * doing sigqueue_free() if we have SIGQUEUE_PREALLOC signals. - */ -- flush_sigqueue(&tsk->pending); -+ flush_task_sigqueue(tsk); - tsk->sighand = NULL; - spin_unlock(&sighand->siglock); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/fork.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/fork.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/fork.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/fork.c 2017-04-18 17:54:26.000000000 +0200 -@@ -76,6 +76,7 @@ - #include - #include - #include -+#include - - #include - #include -@@ -376,13 +377,24 @@ - if (atomic_dec_and_test(&sig->sigcnt)) - free_signal_struct(sig); - } -- -+#ifdef CONFIG_PREEMPT_RT_BASE -+static -+#endif - void __put_task_struct(struct task_struct *tsk) - { - WARN_ON(!tsk->exit_state); - WARN_ON(atomic_read(&tsk->usage)); - WARN_ON(tsk == current); - -+ /* -+ * Remove function-return probe instances associated with this -+ * task and put them back on the free list. -+ */ -+ kprobe_flush_task(tsk); -+ -+ /* Task is done with its stack. */ -+ put_task_stack(tsk); -+ - cgroup_free(tsk); - task_numa_free(tsk); - security_task_free(tsk); -@@ -393,7 +405,18 @@ - if (!profile_handoff_task(tsk)) - free_task(tsk); - } -+#ifndef CONFIG_PREEMPT_RT_BASE - EXPORT_SYMBOL_GPL(__put_task_struct); -+#else -+void __put_task_struct_cb(struct rcu_head *rhp) -+{ -+ struct task_struct *tsk = container_of(rhp, struct task_struct, put_rcu); -+ -+ __put_task_struct(tsk); -+ -+} -+EXPORT_SYMBOL_GPL(__put_task_struct_cb); -+#endif - - void __init __weak arch_task_cache_init(void) { } - -@@ -852,6 +875,19 @@ - } - EXPORT_SYMBOL_GPL(__mmdrop); - -+#ifdef CONFIG_PREEMPT_RT_BASE -+/* -+ * RCU callback for delayed mm drop. Not strictly rcu, but we don't -+ * want another facility to make this work. -+ */ -+void __mmdrop_delayed(struct rcu_head *rhp) -+{ -+ struct mm_struct *mm = container_of(rhp, struct mm_struct, delayed_drop); -+ -+ __mmdrop(mm); -+} -+#endif -+ - static inline void __mmput(struct mm_struct *mm) - { - VM_BUG_ON(atomic_read(&mm->mm_users)); -@@ -1426,6 +1462,9 @@ - */ - static void posix_cpu_timers_init(struct task_struct *tsk) - { -+#ifdef CONFIG_PREEMPT_RT_BASE -+ tsk->posix_timer_list = NULL; -+#endif - tsk->cputime_expires.prof_exp = 0; - tsk->cputime_expires.virt_exp = 0; - tsk->cputime_expires.sched_exp = 0; -@@ -1552,6 +1591,7 @@ - spin_lock_init(&p->alloc_lock); - - init_sigpending(&p->pending); -+ p->sigqueue_cache = NULL; - - p->utime = p->stime = p->gtime = 0; - p->utimescaled = p->stimescaled = 0; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/futex.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/futex.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/futex.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/futex.c 2017-04-18 17:54:26.000000000 +0200 -@@ -800,7 +800,7 @@ - return 0; - } - --static struct futex_pi_state * alloc_pi_state(void) -+static struct futex_pi_state *alloc_pi_state(void) - { - struct futex_pi_state *pi_state = current->pi_state_cache; - -@@ -810,6 +810,11 @@ - return pi_state; - } - -+static void get_pi_state(struct futex_pi_state *pi_state) -+{ -+ WARN_ON_ONCE(!atomic_inc_not_zero(&pi_state->refcount)); -+} -+ - /* - * Drops a reference to the pi_state object and frees or caches it - * when the last reference is gone. -@@ -854,7 +859,7 @@ - * Look up the task based on what TID userspace gave us. - * We dont trust it. - */ --static struct task_struct * futex_find_get_task(pid_t pid) -+static struct task_struct *futex_find_get_task(pid_t pid) - { - struct task_struct *p; - -@@ -904,7 +909,9 @@ - * task still owns the PI-state: - */ - if (head->next != next) { -+ raw_spin_unlock_irq(&curr->pi_lock); - spin_unlock(&hb->lock); -+ raw_spin_lock_irq(&curr->pi_lock); - continue; - } - -@@ -914,10 +921,12 @@ - pi_state->owner = NULL; - raw_spin_unlock_irq(&curr->pi_lock); - -- rt_mutex_unlock(&pi_state->pi_mutex); -- -+ get_pi_state(pi_state); - spin_unlock(&hb->lock); - -+ rt_mutex_futex_unlock(&pi_state->pi_mutex); -+ put_pi_state(pi_state); -+ - raw_spin_lock_irq(&curr->pi_lock); - } - raw_spin_unlock_irq(&curr->pi_lock); -@@ -971,6 +980,39 @@ - * - * [10] There is no transient state which leaves owner and user space - * TID out of sync. -+ * -+ * -+ * Serialization and lifetime rules: -+ * -+ * hb->lock: -+ * -+ * hb -> futex_q, relation -+ * futex_q -> pi_state, relation -+ * -+ * (cannot be raw because hb can contain arbitrary amount -+ * of futex_q's) -+ * -+ * pi_mutex->wait_lock: -+ * -+ * {uval, pi_state} -+ * -+ * (and pi_mutex 'obviously') -+ * -+ * p->pi_lock: -+ * -+ * p->pi_state_list -> pi_state->list, relation -+ * -+ * pi_state->refcount: -+ * -+ * pi_state lifetime -+ * -+ * -+ * Lock order: -+ * -+ * hb->lock -+ * pi_mutex->wait_lock -+ * p->pi_lock -+ * - */ - - /* -@@ -978,10 +1020,12 @@ - * the pi_state against the user space value. If correct, attach to - * it. - */ --static int attach_to_pi_state(u32 uval, struct futex_pi_state *pi_state, -+static int attach_to_pi_state(u32 __user *uaddr, u32 uval, -+ struct futex_pi_state *pi_state, - struct futex_pi_state **ps) - { - pid_t pid = uval & FUTEX_TID_MASK; -+ int ret, uval2; - - /* - * Userspace might have messed up non-PI and PI futexes [3] -@@ -989,9 +1033,39 @@ - if (unlikely(!pi_state)) - return -EINVAL; - -+ /* -+ * We get here with hb->lock held, and having found a -+ * futex_top_waiter(). This means that futex_lock_pi() of said futex_q -+ * has dropped the hb->lock in between queue_me() and unqueue_me_pi(), -+ * which in turn means that futex_lock_pi() still has a reference on -+ * our pi_state. -+ * -+ * The waiter holding a reference on @pi_state also protects against -+ * the unlocked put_pi_state() in futex_unlock_pi(), futex_lock_pi() -+ * and futex_wait_requeue_pi() as it cannot go to 0 and consequently -+ * free pi_state before we can take a reference ourselves. -+ */ - WARN_ON(!atomic_read(&pi_state->refcount)); - - /* -+ * Now that we have a pi_state, we can acquire wait_lock -+ * and do the state validation. -+ */ -+ raw_spin_lock_irq(&pi_state->pi_mutex.wait_lock); -+ -+ /* -+ * Since {uval, pi_state} is serialized by wait_lock, and our current -+ * uval was read without holding it, it can have changed. Verify it -+ * still is what we expect it to be, otherwise retry the entire -+ * operation. -+ */ -+ if (get_futex_value_locked(&uval2, uaddr)) -+ goto out_efault; -+ -+ if (uval != uval2) -+ goto out_eagain; -+ -+ /* - * Handle the owner died case: - */ - if (uval & FUTEX_OWNER_DIED) { -@@ -1006,11 +1080,11 @@ - * is not 0. Inconsistent state. [5] - */ - if (pid) -- return -EINVAL; -+ goto out_einval; - /* - * Take a ref on the state and return success. [4] - */ -- goto out_state; -+ goto out_attach; - } - - /* -@@ -1022,14 +1096,14 @@ - * Take a ref on the state and return success. [6] - */ - if (!pid) -- goto out_state; -+ goto out_attach; - } else { - /* - * If the owner died bit is not set, then the pi_state - * must have an owner. [7] - */ - if (!pi_state->owner) -- return -EINVAL; -+ goto out_einval; - } - - /* -@@ -1038,11 +1112,29 @@ - * user space TID. [9/10] - */ - if (pid != task_pid_vnr(pi_state->owner)) -- return -EINVAL; --out_state: -- atomic_inc(&pi_state->refcount); -+ goto out_einval; -+ -+out_attach: -+ get_pi_state(pi_state); -+ raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); - *ps = pi_state; - return 0; -+ -+out_einval: -+ ret = -EINVAL; -+ goto out_error; -+ -+out_eagain: -+ ret = -EAGAIN; -+ goto out_error; -+ -+out_efault: -+ ret = -EFAULT; -+ goto out_error; -+ -+out_error: -+ raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); -+ return ret; - } - - /* -@@ -1093,6 +1185,9 @@ - - /* - * No existing pi state. First waiter. [2] -+ * -+ * This creates pi_state, we have hb->lock held, this means nothing can -+ * observe this state, wait_lock is irrelevant. - */ - pi_state = alloc_pi_state(); - -@@ -1117,17 +1212,18 @@ - return 0; - } - --static int lookup_pi_state(u32 uval, struct futex_hash_bucket *hb, -+static int lookup_pi_state(u32 __user *uaddr, u32 uval, -+ struct futex_hash_bucket *hb, - union futex_key *key, struct futex_pi_state **ps) - { -- struct futex_q *match = futex_top_waiter(hb, key); -+ struct futex_q *top_waiter = futex_top_waiter(hb, key); - - /* - * If there is a waiter on that futex, validate it and - * attach to the pi_state when the validation succeeds. - */ -- if (match) -- return attach_to_pi_state(uval, match->pi_state, ps); -+ if (top_waiter) -+ return attach_to_pi_state(uaddr, uval, top_waiter->pi_state, ps); - - /* - * We are the first waiter - try to look up the owner based on -@@ -1146,7 +1242,7 @@ - if (unlikely(cmpxchg_futex_value_locked(&curval, uaddr, uval, newval))) - return -EFAULT; - -- /*If user space value changed, let the caller retry */ -+ /* If user space value changed, let the caller retry */ - return curval != uval ? -EAGAIN : 0; - } - -@@ -1174,7 +1270,7 @@ - struct task_struct *task, int set_waiters) - { - u32 uval, newval, vpid = task_pid_vnr(task); -- struct futex_q *match; -+ struct futex_q *top_waiter; - int ret; - - /* -@@ -1200,9 +1296,9 @@ - * Lookup existing state first. If it exists, try to attach to - * its pi_state. - */ -- match = futex_top_waiter(hb, key); -- if (match) -- return attach_to_pi_state(uval, match->pi_state, ps); -+ top_waiter = futex_top_waiter(hb, key); -+ if (top_waiter) -+ return attach_to_pi_state(uaddr, uval, top_waiter->pi_state, ps); - - /* - * No waiter and user TID is 0. We are here because the -@@ -1288,45 +1384,39 @@ - * memory barrier is required here to prevent the following - * store to lock_ptr from getting ahead of the plist_del. - */ -- smp_wmb(); -- q->lock_ptr = NULL; -+ smp_store_release(&q->lock_ptr, NULL); - } - --static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, -- struct futex_hash_bucket *hb) -+/* -+ * Caller must hold a reference on @pi_state. -+ */ -+static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_pi_state *pi_state) - { -- struct task_struct *new_owner; -- struct futex_pi_state *pi_state = this->pi_state; - u32 uninitialized_var(curval), newval; -+ struct task_struct *new_owner; -+ bool deboost = false; - WAKE_Q(wake_q); -- bool deboost; -+ WAKE_Q(wake_sleeper_q); - int ret = 0; - -- if (!pi_state) -- return -EINVAL; -- -- /* -- * If current does not own the pi_state then the futex is -- * inconsistent and user space fiddled with the futex value. -- */ -- if (pi_state->owner != current) -- return -EINVAL; -- -- raw_spin_lock_irq(&pi_state->pi_mutex.wait_lock); - new_owner = rt_mutex_next_owner(&pi_state->pi_mutex); -+ if (WARN_ON_ONCE(!new_owner)) { -+ /* -+ * As per the comment in futex_unlock_pi() this should not happen. -+ * -+ * When this happens, give up our locks and try again, giving -+ * the futex_lock_pi() instance time to complete, either by -+ * waiting on the rtmutex or removing itself from the futex -+ * queue. -+ */ -+ ret = -EAGAIN; -+ goto out_unlock; -+ } - - /* -- * It is possible that the next waiter (the one that brought -- * this owner to the kernel) timed out and is no longer -- * waiting on the lock. -- */ -- if (!new_owner) -- new_owner = this->task; -- -- /* -- * We pass it to the next owner. The WAITERS bit is always -- * kept enabled while there is PI state around. We cleanup the -- * owner died bit, because we are the owner. -+ * We pass it to the next owner. The WAITERS bit is always kept -+ * enabled while there is PI state around. We cleanup the owner -+ * died bit, because we are the owner. - */ - newval = FUTEX_WAITERS | task_pid_vnr(new_owner); - -@@ -1335,6 +1425,7 @@ - - if (cmpxchg_futex_value_locked(&curval, uaddr, uval, newval)) { - ret = -EFAULT; -+ - } else if (curval != uval) { - /* - * If a unconditional UNLOCK_PI operation (user space did not -@@ -1347,10 +1438,9 @@ - else - ret = -EINVAL; - } -- if (ret) { -- raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); -- return ret; -- } -+ -+ if (ret) -+ goto out_unlock; - - raw_spin_lock(&pi_state->owner->pi_lock); - WARN_ON(list_empty(&pi_state->list)); -@@ -1363,22 +1453,22 @@ - pi_state->owner = new_owner; - raw_spin_unlock(&new_owner->pi_lock); - -- raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); -- -- deboost = rt_mutex_futex_unlock(&pi_state->pi_mutex, &wake_q); -- - /* -- * First unlock HB so the waiter does not spin on it once he got woken -- * up. Second wake up the waiter before the priority is adjusted. If we -- * deboost first (and lose our higher priority), then the task might get -- * scheduled away before the wake up can take place. -+ * We've updated the uservalue, this unlock cannot fail. - */ -- spin_unlock(&hb->lock); -- wake_up_q(&wake_q); -- if (deboost) -+ deboost = __rt_mutex_futex_unlock(&pi_state->pi_mutex, &wake_q, -+ &wake_sleeper_q); -+ -+out_unlock: -+ raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); -+ -+ if (deboost) { -+ wake_up_q(&wake_q); -+ wake_up_q_sleeper(&wake_sleeper_q); - rt_mutex_adjust_prio(current); -+ } - -- return 0; -+ return ret; - } - - /* -@@ -1824,7 +1914,7 @@ - * If that call succeeds then we have pi_state and an - * initial refcount on it. - */ -- ret = lookup_pi_state(ret, hb2, &key2, &pi_state); -+ ret = lookup_pi_state(uaddr2, ret, hb2, &key2, &pi_state); - } - - switch (ret) { -@@ -1907,7 +1997,7 @@ - * refcount on the pi_state and store the pointer in - * the futex_q object of the waiter. - */ -- atomic_inc(&pi_state->refcount); -+ get_pi_state(pi_state); - this->pi_state = pi_state; - ret = rt_mutex_start_proxy_lock(&pi_state->pi_mutex, - this->rt_waiter, -@@ -1924,6 +2014,16 @@ - requeue_pi_wake_futex(this, &key2, hb2); - drop_count++; - continue; -+ } else if (ret == -EAGAIN) { -+ /* -+ * Waiter was woken by timeout or -+ * signal and has set pi_blocked_on to -+ * PI_WAKEUP_INPROGRESS before we -+ * tried to enqueue it on the rtmutex. -+ */ -+ this->pi_state = NULL; -+ put_pi_state(pi_state); -+ continue; - } else if (ret) { - /* - * rt_mutex_start_proxy_lock() detected a -@@ -2007,20 +2107,7 @@ - hb_waiters_dec(hb); - } - --/** -- * queue_me() - Enqueue the futex_q on the futex_hash_bucket -- * @q: The futex_q to enqueue -- * @hb: The destination hash bucket -- * -- * The hb->lock must be held by the caller, and is released here. A call to -- * queue_me() is typically paired with exactly one call to unqueue_me(). The -- * exceptions involve the PI related operations, which may use unqueue_me_pi() -- * or nothing if the unqueue is done as part of the wake process and the unqueue -- * state is implicit in the state of woken task (see futex_wait_requeue_pi() for -- * an example). -- */ --static inline void queue_me(struct futex_q *q, struct futex_hash_bucket *hb) -- __releases(&hb->lock) -+static inline void __queue_me(struct futex_q *q, struct futex_hash_bucket *hb) - { - int prio; - -@@ -2037,6 +2124,24 @@ - plist_node_init(&q->list, prio); - plist_add(&q->list, &hb->chain); - q->task = current; -+} -+ -+/** -+ * queue_me() - Enqueue the futex_q on the futex_hash_bucket -+ * @q: The futex_q to enqueue -+ * @hb: The destination hash bucket -+ * -+ * The hb->lock must be held by the caller, and is released here. A call to -+ * queue_me() is typically paired with exactly one call to unqueue_me(). The -+ * exceptions involve the PI related operations, which may use unqueue_me_pi() -+ * or nothing if the unqueue is done as part of the wake process and the unqueue -+ * state is implicit in the state of woken task (see futex_wait_requeue_pi() for -+ * an example). -+ */ -+static inline void queue_me(struct futex_q *q, struct futex_hash_bucket *hb) -+ __releases(&hb->lock) -+{ -+ __queue_me(q, hb); - spin_unlock(&hb->lock); - } - -@@ -2123,10 +2228,13 @@ - { - u32 newtid = task_pid_vnr(newowner) | FUTEX_WAITERS; - struct futex_pi_state *pi_state = q->pi_state; -- struct task_struct *oldowner = pi_state->owner; - u32 uval, uninitialized_var(curval), newval; -+ struct task_struct *oldowner; - int ret; - -+ raw_spin_lock_irq(&pi_state->pi_mutex.wait_lock); -+ -+ oldowner = pi_state->owner; - /* Owner died? */ - if (!pi_state->owner) - newtid |= FUTEX_OWNER_DIED; -@@ -2134,7 +2242,8 @@ - /* - * We are here either because we stole the rtmutex from the - * previous highest priority waiter or we are the highest priority -- * waiter but failed to get the rtmutex the first time. -+ * waiter but have failed to get the rtmutex the first time. -+ * - * We have to replace the newowner TID in the user space variable. - * This must be atomic as we have to preserve the owner died bit here. - * -@@ -2142,17 +2251,16 @@ - * because we can fault here. Imagine swapped out pages or a fork - * that marked all the anonymous memory readonly for cow. - * -- * Modifying pi_state _before_ the user space value would -- * leave the pi_state in an inconsistent state when we fault -- * here, because we need to drop the hash bucket lock to -- * handle the fault. This might be observed in the PID check -- * in lookup_pi_state. -+ * Modifying pi_state _before_ the user space value would leave the -+ * pi_state in an inconsistent state when we fault here, because we -+ * need to drop the locks to handle the fault. This might be observed -+ * in the PID check in lookup_pi_state. - */ - retry: - if (get_futex_value_locked(&uval, uaddr)) - goto handle_fault; - -- while (1) { -+ for (;;) { - newval = (uval & FUTEX_OWNER_DIED) | newtid; - - if (cmpxchg_futex_value_locked(&curval, uaddr, uval, newval)) -@@ -2167,47 +2275,60 @@ - * itself. - */ - if (pi_state->owner != NULL) { -- raw_spin_lock_irq(&pi_state->owner->pi_lock); -+ raw_spin_lock(&pi_state->owner->pi_lock); - WARN_ON(list_empty(&pi_state->list)); - list_del_init(&pi_state->list); -- raw_spin_unlock_irq(&pi_state->owner->pi_lock); -+ raw_spin_unlock(&pi_state->owner->pi_lock); - } - - pi_state->owner = newowner; - -- raw_spin_lock_irq(&newowner->pi_lock); -+ raw_spin_lock(&newowner->pi_lock); - WARN_ON(!list_empty(&pi_state->list)); - list_add(&pi_state->list, &newowner->pi_state_list); -- raw_spin_unlock_irq(&newowner->pi_lock); -+ raw_spin_unlock(&newowner->pi_lock); -+ raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); -+ - return 0; - - /* -- * To handle the page fault we need to drop the hash bucket -- * lock here. That gives the other task (either the highest priority -- * waiter itself or the task which stole the rtmutex) the -- * chance to try the fixup of the pi_state. So once we are -- * back from handling the fault we need to check the pi_state -- * after reacquiring the hash bucket lock and before trying to -- * do another fixup. When the fixup has been done already we -- * simply return. -+ * To handle the page fault we need to drop the locks here. That gives -+ * the other task (either the highest priority waiter itself or the -+ * task which stole the rtmutex) the chance to try the fixup of the -+ * pi_state. So once we are back from handling the fault we need to -+ * check the pi_state after reacquiring the locks and before trying to -+ * do another fixup. When the fixup has been done already we simply -+ * return. -+ * -+ * Note: we hold both hb->lock and pi_mutex->wait_lock. We can safely -+ * drop hb->lock since the caller owns the hb -> futex_q relation. -+ * Dropping the pi_mutex->wait_lock requires the state revalidate. - */ - handle_fault: -+ raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); - spin_unlock(q->lock_ptr); - - ret = fault_in_user_writeable(uaddr); - - spin_lock(q->lock_ptr); -+ raw_spin_lock_irq(&pi_state->pi_mutex.wait_lock); - - /* - * Check if someone else fixed it for us: - */ -- if (pi_state->owner != oldowner) -- return 0; -+ if (pi_state->owner != oldowner) { -+ ret = 0; -+ goto out_unlock; -+ } - - if (ret) -- return ret; -+ goto out_unlock; - - goto retry; -+ -+out_unlock: -+ raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); -+ return ret; - } - - static long futex_wait_restart(struct restart_block *restart); -@@ -2229,13 +2350,16 @@ - */ - static int fixup_owner(u32 __user *uaddr, struct futex_q *q, int locked) - { -- struct task_struct *owner; - int ret = 0; - - if (locked) { - /* - * Got the lock. We might not be the anticipated owner if we - * did a lock-steal - fix up the PI-state in that case: -+ * -+ * We can safely read pi_state->owner without holding wait_lock -+ * because we now own the rt_mutex, only the owner will attempt -+ * to change it. - */ - if (q->pi_state->owner != current) - ret = fixup_pi_state_owner(uaddr, q, current); -@@ -2243,43 +2367,15 @@ - } - - /* -- * Catch the rare case, where the lock was released when we were on the -- * way back before we locked the hash bucket. -- */ -- if (q->pi_state->owner == current) { -- /* -- * Try to get the rt_mutex now. This might fail as some other -- * task acquired the rt_mutex after we removed ourself from the -- * rt_mutex waiters list. -- */ -- if (rt_mutex_trylock(&q->pi_state->pi_mutex)) { -- locked = 1; -- goto out; -- } -- -- /* -- * pi_state is incorrect, some other task did a lock steal and -- * we returned due to timeout or signal without taking the -- * rt_mutex. Too late. -- */ -- raw_spin_lock_irq(&q->pi_state->pi_mutex.wait_lock); -- owner = rt_mutex_owner(&q->pi_state->pi_mutex); -- if (!owner) -- owner = rt_mutex_next_owner(&q->pi_state->pi_mutex); -- raw_spin_unlock_irq(&q->pi_state->pi_mutex.wait_lock); -- ret = fixup_pi_state_owner(uaddr, q, owner); -- goto out; -- } -- -- /* - * Paranoia check. If we did not take the lock, then we should not be - * the owner of the rt_mutex. - */ -- if (rt_mutex_owner(&q->pi_state->pi_mutex) == current) -+ if (rt_mutex_owner(&q->pi_state->pi_mutex) == current) { - printk(KERN_ERR "fixup_owner: ret = %d pi-mutex: %p " - "pi-state %p\n", ret, - q->pi_state->pi_mutex.owner, - q->pi_state->owner); -+ } - - out: - return ret ? ret : locked; -@@ -2503,6 +2599,8 @@ - ktime_t *time, int trylock) - { - struct hrtimer_sleeper timeout, *to = NULL; -+ struct futex_pi_state *pi_state = NULL; -+ struct rt_mutex_waiter rt_waiter; - struct futex_hash_bucket *hb; - struct futex_q q = futex_q_init; - int res, ret; -@@ -2555,25 +2653,77 @@ - } - } - -+ WARN_ON(!q.pi_state); -+ - /* - * Only actually queue now that the atomic ops are done: - */ -- queue_me(&q, hb); -+ __queue_me(&q, hb); - -- WARN_ON(!q.pi_state); -- /* -- * Block on the PI mutex: -- */ -- if (!trylock) { -- ret = rt_mutex_timed_futex_lock(&q.pi_state->pi_mutex, to); -- } else { -- ret = rt_mutex_trylock(&q.pi_state->pi_mutex); -+ if (trylock) { -+ ret = rt_mutex_futex_trylock(&q.pi_state->pi_mutex); - /* Fixup the trylock return value: */ - ret = ret ? 0 : -EWOULDBLOCK; -+ goto no_block; -+ } -+ -+ rt_mutex_init_waiter(&rt_waiter, false); -+ -+ /* -+ * On PREEMPT_RT_FULL, when hb->lock becomes an rt_mutex, we must not -+ * hold it while doing rt_mutex_start_proxy(), because then it will -+ * include hb->lock in the blocking chain, even through we'll not in -+ * fact hold it while blocking. This will lead it to report -EDEADLK -+ * and BUG when futex_unlock_pi() interleaves with this. -+ * -+ * Therefore acquire wait_lock while holding hb->lock, but drop the -+ * latter before calling rt_mutex_start_proxy_lock(). This still fully -+ * serializes against futex_unlock_pi() as that does the exact same -+ * lock handoff sequence. -+ */ -+ raw_spin_lock_irq(&q.pi_state->pi_mutex.wait_lock); -+ /* -+ * the migrate_disable() here disables migration in the in_atomic() fast -+ * path which is enabled again in the following spin_unlock(). We have -+ * one migrate_disable() pending in the slow-path which is reversed -+ * after the raw_spin_unlock_irq() where we leave the atomic context. -+ */ -+ migrate_disable(); -+ -+ spin_unlock(q.lock_ptr); -+ ret = __rt_mutex_start_proxy_lock(&q.pi_state->pi_mutex, &rt_waiter, current); -+ raw_spin_unlock_irq(&q.pi_state->pi_mutex.wait_lock); -+ migrate_enable(); -+ -+ if (ret) { -+ if (ret == 1) -+ ret = 0; -+ -+ spin_lock(q.lock_ptr); -+ goto no_block; - } - -+ -+ if (unlikely(to)) -+ hrtimer_start_expires(&to->timer, HRTIMER_MODE_ABS); -+ -+ ret = rt_mutex_wait_proxy_lock(&q.pi_state->pi_mutex, to, &rt_waiter); -+ - spin_lock(q.lock_ptr); - /* -+ * If we failed to acquire the lock (signal/timeout), we must -+ * first acquire the hb->lock before removing the lock from the -+ * rt_mutex waitqueue, such that we can keep the hb and rt_mutex -+ * wait lists consistent. -+ * -+ * In particular; it is important that futex_unlock_pi() can not -+ * observe this inconsistency. -+ */ -+ if (ret && !rt_mutex_cleanup_proxy_lock(&q.pi_state->pi_mutex, &rt_waiter)) -+ ret = 0; -+ -+no_block: -+ /* - * Fixup the pi_state owner and possibly acquire the lock if we - * haven't already. - */ -@@ -2589,12 +2739,19 @@ - * If fixup_owner() faulted and was unable to handle the fault, unlock - * it and return the fault to userspace. - */ -- if (ret && (rt_mutex_owner(&q.pi_state->pi_mutex) == current)) -- rt_mutex_unlock(&q.pi_state->pi_mutex); -+ if (ret && (rt_mutex_owner(&q.pi_state->pi_mutex) == current)) { -+ pi_state = q.pi_state; -+ get_pi_state(pi_state); -+ } - - /* Unqueue and drop the lock */ - unqueue_me_pi(&q); - -+ if (pi_state) { -+ rt_mutex_futex_unlock(&pi_state->pi_mutex); -+ put_pi_state(pi_state); -+ } -+ - goto out_put_key; - - out_unlock_put_key: -@@ -2631,7 +2788,7 @@ - u32 uninitialized_var(curval), uval, vpid = task_pid_vnr(current); - union futex_key key = FUTEX_KEY_INIT; - struct futex_hash_bucket *hb; -- struct futex_q *match; -+ struct futex_q *top_waiter; - int ret; - - retry: -@@ -2655,12 +2812,48 @@ - * all and we at least want to know if user space fiddled - * with the futex value instead of blindly unlocking. - */ -- match = futex_top_waiter(hb, &key); -- if (match) { -- ret = wake_futex_pi(uaddr, uval, match, hb); -+ top_waiter = futex_top_waiter(hb, &key); -+ if (top_waiter) { -+ struct futex_pi_state *pi_state = top_waiter->pi_state; -+ -+ ret = -EINVAL; -+ if (!pi_state) -+ goto out_unlock; -+ - /* -- * In case of success wake_futex_pi dropped the hash -- * bucket lock. -+ * If current does not own the pi_state then the futex is -+ * inconsistent and user space fiddled with the futex value. -+ */ -+ if (pi_state->owner != current) -+ goto out_unlock; -+ -+ get_pi_state(pi_state); -+ /* -+ * By taking wait_lock while still holding hb->lock, we ensure -+ * there is no point where we hold neither; and therefore -+ * wake_futex_pi() must observe a state consistent with what we -+ * observed. -+ */ -+ raw_spin_lock_irq(&pi_state->pi_mutex.wait_lock); -+ /* -+ * Magic trickery for now to make the RT migrate disable -+ * logic happy. The following spin_unlock() happens with -+ * interrupts disabled so the internal migrate_enable() -+ * won't undo the migrate_disable() which was issued when -+ * locking hb->lock. -+ */ -+ migrate_disable(); -+ spin_unlock(&hb->lock); -+ -+ /* Drops pi_state->pi_mutex.wait_lock */ -+ ret = wake_futex_pi(uaddr, uval, pi_state); -+ -+ migrate_enable(); -+ -+ put_pi_state(pi_state); -+ -+ /* -+ * Success, we're done! No tricky corner cases. - */ - if (!ret) - goto out_putkey; -@@ -2675,7 +2868,6 @@ - * setting the FUTEX_WAITERS bit. Try again. - */ - if (ret == -EAGAIN) { -- spin_unlock(&hb->lock); - put_futex_key(&key); - goto retry; - } -@@ -2683,7 +2875,7 @@ - * wake_futex_pi has detected invalid state. Tell user - * space. - */ -- goto out_unlock; -+ goto out_putkey; - } - - /* -@@ -2693,8 +2885,10 @@ - * preserve the WAITERS bit not the OWNER_DIED one. We are the - * owner. - */ -- if (cmpxchg_futex_value_locked(&curval, uaddr, uval, 0)) -+ if (cmpxchg_futex_value_locked(&curval, uaddr, uval, 0)) { -+ spin_unlock(&hb->lock); - goto pi_faulted; -+ } - - /* - * If uval has changed, let user space handle it. -@@ -2708,7 +2902,6 @@ - return ret; - - pi_faulted: -- spin_unlock(&hb->lock); - put_futex_key(&key); - - ret = fault_in_user_writeable(uaddr); -@@ -2812,8 +3005,9 @@ - u32 __user *uaddr2) - { - struct hrtimer_sleeper timeout, *to = NULL; -+ struct futex_pi_state *pi_state = NULL; - struct rt_mutex_waiter rt_waiter; -- struct futex_hash_bucket *hb; -+ struct futex_hash_bucket *hb, *hb2; - union futex_key key2 = FUTEX_KEY_INIT; - struct futex_q q = futex_q_init; - int res, ret; -@@ -2838,10 +3032,7 @@ - * The waiter is allocated on our stack, manipulated by the requeue - * code while we sleep on uaddr. - */ -- debug_rt_mutex_init_waiter(&rt_waiter); -- RB_CLEAR_NODE(&rt_waiter.pi_tree_entry); -- RB_CLEAR_NODE(&rt_waiter.tree_entry); -- rt_waiter.task = NULL; -+ rt_mutex_init_waiter(&rt_waiter, false); - - ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2, VERIFY_WRITE); - if (unlikely(ret != 0)) -@@ -2872,20 +3063,55 @@ - /* Queue the futex_q, drop the hb lock, wait for wakeup. */ - futex_wait_queue_me(hb, &q, to); - -- spin_lock(&hb->lock); -- ret = handle_early_requeue_pi_wakeup(hb, &q, &key2, to); -- spin_unlock(&hb->lock); -- if (ret) -- goto out_put_keys; -+ /* -+ * On RT we must avoid races with requeue and trying to block -+ * on two mutexes (hb->lock and uaddr2's rtmutex) by -+ * serializing access to pi_blocked_on with pi_lock. -+ */ -+ raw_spin_lock_irq(¤t->pi_lock); -+ if (current->pi_blocked_on) { -+ /* -+ * We have been requeued or are in the process of -+ * being requeued. -+ */ -+ raw_spin_unlock_irq(¤t->pi_lock); -+ } else { -+ /* -+ * Setting pi_blocked_on to PI_WAKEUP_INPROGRESS -+ * prevents a concurrent requeue from moving us to the -+ * uaddr2 rtmutex. After that we can safely acquire -+ * (and possibly block on) hb->lock. -+ */ -+ current->pi_blocked_on = PI_WAKEUP_INPROGRESS; -+ raw_spin_unlock_irq(¤t->pi_lock); -+ -+ spin_lock(&hb->lock); -+ -+ /* -+ * Clean up pi_blocked_on. We might leak it otherwise -+ * when we succeeded with the hb->lock in the fast -+ * path. -+ */ -+ raw_spin_lock_irq(¤t->pi_lock); -+ current->pi_blocked_on = NULL; -+ raw_spin_unlock_irq(¤t->pi_lock); -+ -+ ret = handle_early_requeue_pi_wakeup(hb, &q, &key2, to); -+ spin_unlock(&hb->lock); -+ if (ret) -+ goto out_put_keys; -+ } - - /* -- * In order for us to be here, we know our q.key == key2, and since -- * we took the hb->lock above, we also know that futex_requeue() has -- * completed and we no longer have to concern ourselves with a wakeup -- * race with the atomic proxy lock acquisition by the requeue code. The -- * futex_requeue dropped our key1 reference and incremented our key2 -- * reference count. -+ * In order to be here, we have either been requeued, are in -+ * the process of being requeued, or requeue successfully -+ * acquired uaddr2 on our behalf. If pi_blocked_on was -+ * non-null above, we may be racing with a requeue. Do not -+ * rely on q->lock_ptr to be hb2->lock until after blocking on -+ * hb->lock or hb2->lock. The futex_requeue dropped our key1 -+ * reference and incremented our key2 reference count. - */ -+ hb2 = hash_futex(&key2); - - /* Check if the requeue code acquired the second futex for us. */ - if (!q.rt_waiter) { -@@ -2894,16 +3120,19 @@ - * did a lock-steal - fix up the PI-state in that case. - */ - if (q.pi_state && (q.pi_state->owner != current)) { -- spin_lock(q.lock_ptr); -+ spin_lock(&hb2->lock); -+ BUG_ON(&hb2->lock != q.lock_ptr); - ret = fixup_pi_state_owner(uaddr2, &q, current); -- if (ret && rt_mutex_owner(&q.pi_state->pi_mutex) == current) -- rt_mutex_unlock(&q.pi_state->pi_mutex); -+ if (ret && rt_mutex_owner(&q.pi_state->pi_mutex) == current) { -+ pi_state = q.pi_state; -+ get_pi_state(pi_state); -+ } - /* - * Drop the reference to the pi state which - * the requeue_pi() code acquired for us. - */ - put_pi_state(q.pi_state); -- spin_unlock(q.lock_ptr); -+ spin_unlock(&hb2->lock); - } - } else { - struct rt_mutex *pi_mutex; -@@ -2915,10 +3144,14 @@ - */ - WARN_ON(!q.pi_state); - pi_mutex = &q.pi_state->pi_mutex; -- ret = rt_mutex_finish_proxy_lock(pi_mutex, to, &rt_waiter); -- debug_rt_mutex_free_waiter(&rt_waiter); -+ ret = rt_mutex_wait_proxy_lock(pi_mutex, to, &rt_waiter); - -- spin_lock(q.lock_ptr); -+ spin_lock(&hb2->lock); -+ BUG_ON(&hb2->lock != q.lock_ptr); -+ if (ret && !rt_mutex_cleanup_proxy_lock(pi_mutex, &rt_waiter)) -+ ret = 0; -+ -+ debug_rt_mutex_free_waiter(&rt_waiter); - /* - * Fixup the pi_state owner and possibly acquire the lock if we - * haven't already. -@@ -2936,13 +3169,20 @@ - * the fault, unlock the rt_mutex and return the fault to - * userspace. - */ -- if (ret && rt_mutex_owner(pi_mutex) == current) -- rt_mutex_unlock(pi_mutex); -+ if (ret && rt_mutex_owner(&q.pi_state->pi_mutex) == current) { -+ pi_state = q.pi_state; -+ get_pi_state(pi_state); -+ } - - /* Unqueue and drop the lock. */ - unqueue_me_pi(&q); - } - -+ if (pi_state) { -+ rt_mutex_futex_unlock(&pi_state->pi_mutex); -+ put_pi_state(pi_state); -+ } -+ - if (ret == -EINTR) { - /* - * We've already been requeued, but cannot restart by calling -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/irq/handle.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/irq/handle.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/irq/handle.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/irq/handle.c 2017-04-18 17:54:26.000000000 +0200 -@@ -181,10 +181,16 @@ - { - irqreturn_t retval; - unsigned int flags = 0; -+ struct pt_regs *regs = get_irq_regs(); -+ u64 ip = regs ? instruction_pointer(regs) : 0; - - retval = __handle_irq_event_percpu(desc, &flags); - -- add_interrupt_randomness(desc->irq_data.irq, flags); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ desc->random_ip = ip; -+#else -+ add_interrupt_randomness(desc->irq_data.irq, flags, ip); -+#endif - - if (!noirqdebug) - note_interrupt(desc, retval); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/irq/manage.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/irq/manage.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/irq/manage.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/irq/manage.c 2017-04-18 17:54:26.000000000 +0200 -@@ -22,6 +22,7 @@ - #include "internals.h" - - #ifdef CONFIG_IRQ_FORCED_THREADING -+# ifndef CONFIG_PREEMPT_RT_BASE - __read_mostly bool force_irqthreads; - - static int __init setup_forced_irqthreads(char *arg) -@@ -30,6 +31,7 @@ - return 0; - } - early_param("threadirqs", setup_forced_irqthreads); -+# endif - #endif - - static void __synchronize_hardirq(struct irq_desc *desc) -@@ -233,7 +235,12 @@ - - if (desc->affinity_notify) { - kref_get(&desc->affinity_notify->kref); -+ -+#ifdef CONFIG_PREEMPT_RT_BASE -+ swork_queue(&desc->affinity_notify->swork); -+#else - schedule_work(&desc->affinity_notify->work); -+#endif - } - irqd_set(data, IRQD_AFFINITY_SET); - -@@ -271,10 +278,8 @@ - } - EXPORT_SYMBOL_GPL(irq_set_affinity_hint); - --static void irq_affinity_notify(struct work_struct *work) -+static void _irq_affinity_notify(struct irq_affinity_notify *notify) - { -- struct irq_affinity_notify *notify = -- container_of(work, struct irq_affinity_notify, work); - struct irq_desc *desc = irq_to_desc(notify->irq); - cpumask_var_t cpumask; - unsigned long flags; -@@ -296,6 +301,35 @@ - kref_put(¬ify->kref, notify->release); - } - -+#ifdef CONFIG_PREEMPT_RT_BASE -+static void init_helper_thread(void) -+{ -+ static int init_sworker_once; -+ -+ if (init_sworker_once) -+ return; -+ if (WARN_ON(swork_get())) -+ return; -+ init_sworker_once = 1; -+} -+ -+static void irq_affinity_notify(struct swork_event *swork) -+{ -+ struct irq_affinity_notify *notify = -+ container_of(swork, struct irq_affinity_notify, swork); -+ _irq_affinity_notify(notify); -+} -+ -+#else -+ -+static void irq_affinity_notify(struct work_struct *work) -+{ -+ struct irq_affinity_notify *notify = -+ container_of(work, struct irq_affinity_notify, work); -+ _irq_affinity_notify(notify); -+} -+#endif -+ - /** - * irq_set_affinity_notifier - control notification of IRQ affinity changes - * @irq: Interrupt for which to enable/disable notification -@@ -324,7 +358,12 @@ - if (notify) { - notify->irq = irq; - kref_init(¬ify->kref); -+#ifdef CONFIG_PREEMPT_RT_BASE -+ INIT_SWORK(¬ify->swork, irq_affinity_notify); -+ init_helper_thread(); -+#else - INIT_WORK(¬ify->work, irq_affinity_notify); -+#endif - } - - raw_spin_lock_irqsave(&desc->lock, flags); -@@ -879,7 +918,15 @@ - local_bh_disable(); - ret = action->thread_fn(action->irq, action->dev_id); - irq_finalize_oneshot(desc, action); -- local_bh_enable(); -+ /* -+ * Interrupts which have real time requirements can be set up -+ * to avoid softirq processing in the thread handler. This is -+ * safe as these interrupts do not raise soft interrupts. -+ */ -+ if (irq_settings_no_softirq_call(desc)) -+ _local_bh_enable(); -+ else -+ local_bh_enable(); - return ret; - } - -@@ -976,6 +1023,12 @@ - if (action_ret == IRQ_WAKE_THREAD) - irq_wake_secondary(desc, action); - -+#ifdef CONFIG_PREEMPT_RT_FULL -+ migrate_disable(); -+ add_interrupt_randomness(action->irq, 0, -+ desc->random_ip ^ (unsigned long) action); -+ migrate_enable(); -+#endif - wake_threads_waitq(desc); - } - -@@ -1336,6 +1389,9 @@ - irqd_set(&desc->irq_data, IRQD_NO_BALANCING); - } - -+ if (new->flags & IRQF_NO_SOFTIRQ_CALL) -+ irq_settings_set_no_softirq_call(desc); -+ - /* Set default affinity mask once everything is setup */ - setup_affinity(desc, mask); - -@@ -2061,7 +2117,7 @@ - * This call sets the internal irqchip state of an interrupt, - * depending on the value of @which. - * -- * This function should be called with preemption disabled if the -+ * This function should be called with migration disabled if the - * interrupt controller has per-cpu registers. - */ - int irq_set_irqchip_state(unsigned int irq, enum irqchip_irq_state which, -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/irq/settings.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/irq/settings.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/irq/settings.h 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/irq/settings.h 2017-04-18 17:54:26.000000000 +0200 -@@ -16,6 +16,7 @@ - _IRQ_PER_CPU_DEVID = IRQ_PER_CPU_DEVID, - _IRQ_IS_POLLED = IRQ_IS_POLLED, - _IRQ_DISABLE_UNLAZY = IRQ_DISABLE_UNLAZY, -+ _IRQ_NO_SOFTIRQ_CALL = IRQ_NO_SOFTIRQ_CALL, - _IRQF_MODIFY_MASK = IRQF_MODIFY_MASK, - }; - -@@ -30,6 +31,7 @@ - #define IRQ_PER_CPU_DEVID GOT_YOU_MORON - #define IRQ_IS_POLLED GOT_YOU_MORON - #define IRQ_DISABLE_UNLAZY GOT_YOU_MORON -+#define IRQ_NO_SOFTIRQ_CALL GOT_YOU_MORON - #undef IRQF_MODIFY_MASK - #define IRQF_MODIFY_MASK GOT_YOU_MORON - -@@ -40,6 +42,16 @@ - desc->status_use_accessors |= (set & _IRQF_MODIFY_MASK); - } - -+static inline bool irq_settings_no_softirq_call(struct irq_desc *desc) -+{ -+ return desc->status_use_accessors & _IRQ_NO_SOFTIRQ_CALL; -+} -+ -+static inline void irq_settings_set_no_softirq_call(struct irq_desc *desc) -+{ -+ desc->status_use_accessors |= _IRQ_NO_SOFTIRQ_CALL; -+} -+ - static inline bool irq_settings_is_per_cpu(struct irq_desc *desc) - { - return desc->status_use_accessors & _IRQ_PER_CPU; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/irq/spurious.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/irq/spurious.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/irq/spurious.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/irq/spurious.c 2017-04-18 17:54:26.000000000 +0200 -@@ -442,6 +442,10 @@ - - static int __init irqfixup_setup(char *str) - { -+#ifdef CONFIG_PREEMPT_RT_BASE -+ pr_warn("irqfixup boot option not supported w/ CONFIG_PREEMPT_RT_BASE\n"); -+ return 1; -+#endif - irqfixup = 1; - printk(KERN_WARNING "Misrouted IRQ fixup support enabled.\n"); - printk(KERN_WARNING "This may impact system performance.\n"); -@@ -454,6 +458,10 @@ - - static int __init irqpoll_setup(char *str) - { -+#ifdef CONFIG_PREEMPT_RT_BASE -+ pr_warn("irqpoll boot option not supported w/ CONFIG_PREEMPT_RT_BASE\n"); -+ return 1; -+#endif - irqfixup = 2; - printk(KERN_WARNING "Misrouted IRQ fixup and polling support " - "enabled\n"); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/irq_work.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/irq_work.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/irq_work.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/irq_work.c 2017-04-18 17:54:26.000000000 +0200 -@@ -17,6 +17,7 @@ - #include - #include - #include -+#include - #include - - -@@ -65,6 +66,8 @@ - */ - bool irq_work_queue_on(struct irq_work *work, int cpu) - { -+ struct llist_head *list; -+ - /* All work should have been flushed before going offline */ - WARN_ON_ONCE(cpu_is_offline(cpu)); - -@@ -75,7 +78,12 @@ - if (!irq_work_claim(work)) - return false; - -- if (llist_add(&work->llnode, &per_cpu(raised_list, cpu))) -+ if (IS_ENABLED(CONFIG_PREEMPT_RT_FULL) && !(work->flags & IRQ_WORK_HARD_IRQ)) -+ list = &per_cpu(lazy_list, cpu); -+ else -+ list = &per_cpu(raised_list, cpu); -+ -+ if (llist_add(&work->llnode, list)) - arch_send_call_function_single_ipi(cpu); - - return true; -@@ -86,6 +94,9 @@ - /* Enqueue the irq work @work on the current CPU */ - bool irq_work_queue(struct irq_work *work) - { -+ struct llist_head *list; -+ bool lazy_work, realtime = IS_ENABLED(CONFIG_PREEMPT_RT_FULL); -+ - /* Only queue if not already pending */ - if (!irq_work_claim(work)) - return false; -@@ -93,13 +104,15 @@ - /* Queue the entry and raise the IPI if needed. */ - preempt_disable(); - -- /* If the work is "lazy", handle it from next tick if any */ -- if (work->flags & IRQ_WORK_LAZY) { -- if (llist_add(&work->llnode, this_cpu_ptr(&lazy_list)) && -- tick_nohz_tick_stopped()) -- arch_irq_work_raise(); -- } else { -- if (llist_add(&work->llnode, this_cpu_ptr(&raised_list))) -+ lazy_work = work->flags & IRQ_WORK_LAZY; -+ -+ if (lazy_work || (realtime && !(work->flags & IRQ_WORK_HARD_IRQ))) -+ list = this_cpu_ptr(&lazy_list); -+ else -+ list = this_cpu_ptr(&raised_list); -+ -+ if (llist_add(&work->llnode, list)) { -+ if (!lazy_work || tick_nohz_tick_stopped()) - arch_irq_work_raise(); - } - -@@ -116,9 +129,8 @@ - raised = this_cpu_ptr(&raised_list); - lazy = this_cpu_ptr(&lazy_list); - -- if (llist_empty(raised) || arch_irq_work_has_interrupt()) -- if (llist_empty(lazy)) -- return false; -+ if (llist_empty(raised) && llist_empty(lazy)) -+ return false; - - /* All work should have been flushed before going offline */ - WARN_ON_ONCE(cpu_is_offline(smp_processor_id())); -@@ -132,7 +144,7 @@ - struct irq_work *work; - struct llist_node *llnode; - -- BUG_ON(!irqs_disabled()); -+ BUG_ON_NONRT(!irqs_disabled()); - - if (llist_empty(list)) - return; -@@ -169,7 +181,16 @@ - void irq_work_run(void) - { - irq_work_run_list(this_cpu_ptr(&raised_list)); -- irq_work_run_list(this_cpu_ptr(&lazy_list)); -+ if (IS_ENABLED(CONFIG_PREEMPT_RT_FULL)) { -+ /* -+ * NOTE: we raise softirq via IPI for safety, -+ * and execute in irq_work_tick() to move the -+ * overhead from hard to soft irq context. -+ */ -+ if (!llist_empty(this_cpu_ptr(&lazy_list))) -+ raise_softirq(TIMER_SOFTIRQ); -+ } else -+ irq_work_run_list(this_cpu_ptr(&lazy_list)); - } - EXPORT_SYMBOL_GPL(irq_work_run); - -@@ -179,8 +200,17 @@ - - if (!llist_empty(raised) && !arch_irq_work_has_interrupt()) - irq_work_run_list(raised); -+ -+ if (!IS_ENABLED(CONFIG_PREEMPT_RT_FULL)) -+ irq_work_run_list(this_cpu_ptr(&lazy_list)); -+} -+ -+#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) -+void irq_work_tick_soft(void) -+{ - irq_work_run_list(this_cpu_ptr(&lazy_list)); - } -+#endif - - /* - * Synchronize against the irq_work @entry, ensures the entry is not -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/ksysfs.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/ksysfs.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/ksysfs.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/ksysfs.c 2017-04-18 17:54:26.000000000 +0200 -@@ -136,6 +136,15 @@ - - #endif /* CONFIG_KEXEC_CORE */ - -+#if defined(CONFIG_PREEMPT_RT_FULL) -+static ssize_t realtime_show(struct kobject *kobj, -+ struct kobj_attribute *attr, char *buf) -+{ -+ return sprintf(buf, "%d\n", 1); -+} -+KERNEL_ATTR_RO(realtime); -+#endif -+ - /* whether file capabilities are enabled */ - static ssize_t fscaps_show(struct kobject *kobj, - struct kobj_attribute *attr, char *buf) -@@ -225,6 +234,9 @@ - &rcu_expedited_attr.attr, - &rcu_normal_attr.attr, - #endif -+#ifdef CONFIG_PREEMPT_RT_FULL -+ &realtime_attr.attr, -+#endif - NULL - }; - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/Makefile linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/Makefile ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/Makefile 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/Makefile 2017-04-18 17:54:26.000000000 +0200 -@@ -2,7 +2,7 @@ - # and is generally not a function of system call inputs. - KCOV_INSTRUMENT := n - --obj-y += mutex.o semaphore.o rwsem.o percpu-rwsem.o -+obj-y += semaphore.o percpu-rwsem.o - - ifdef CONFIG_FUNCTION_TRACER - CFLAGS_REMOVE_lockdep.o = $(CC_FLAGS_FTRACE) -@@ -11,7 +11,11 @@ - CFLAGS_REMOVE_rtmutex-debug.o = $(CC_FLAGS_FTRACE) - endif - -+ifneq ($(CONFIG_PREEMPT_RT_FULL),y) -+obj-y += mutex.o - obj-$(CONFIG_DEBUG_MUTEXES) += mutex-debug.o -+endif -+obj-y += rwsem.o - obj-$(CONFIG_LOCKDEP) += lockdep.o - ifeq ($(CONFIG_PROC_FS),y) - obj-$(CONFIG_LOCKDEP) += lockdep_proc.o -@@ -24,7 +28,10 @@ - obj-$(CONFIG_DEBUG_RT_MUTEXES) += rtmutex-debug.o - obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock.o - obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock_debug.o -+ifneq ($(CONFIG_PREEMPT_RT_FULL),y) - obj-$(CONFIG_RWSEM_GENERIC_SPINLOCK) += rwsem-spinlock.o - obj-$(CONFIG_RWSEM_XCHGADD_ALGORITHM) += rwsem-xadd.o -+endif -+obj-$(CONFIG_PREEMPT_RT_FULL) += rt.o rwsem-rt.o - obj-$(CONFIG_QUEUED_RWLOCKS) += qrwlock.o - obj-$(CONFIG_LOCK_TORTURE_TEST) += locktorture.o -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/lockdep.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/lockdep.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/lockdep.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/lockdep.c 2017-04-18 17:54:26.000000000 +0200 -@@ -658,6 +658,7 @@ - struct lockdep_subclass_key *key; - struct hlist_head *hash_head; - struct lock_class *class; -+ bool is_static = false; - - if (unlikely(subclass >= MAX_LOCKDEP_SUBCLASSES)) { - debug_locks_off(); -@@ -671,10 +672,23 @@ - - /* - * Static locks do not have their class-keys yet - for them the key -- * is the lock object itself: -- */ -- if (unlikely(!lock->key)) -- lock->key = (void *)lock; -+ * is the lock object itself. If the lock is in the per cpu area, -+ * the canonical address of the lock (per cpu offset removed) is -+ * used. -+ */ -+ if (unlikely(!lock->key)) { -+ unsigned long can_addr, addr = (unsigned long)lock; -+ -+ if (__is_kernel_percpu_address(addr, &can_addr)) -+ lock->key = (void *)can_addr; -+ else if (__is_module_percpu_address(addr, &can_addr)) -+ lock->key = (void *)can_addr; -+ else if (static_obj(lock)) -+ lock->key = (void *)lock; -+ else -+ return ERR_PTR(-EINVAL); -+ is_static = true; -+ } - - /* - * NOTE: the class-key must be unique. For dynamic locks, a static -@@ -706,7 +720,7 @@ - } - } - -- return NULL; -+ return is_static || static_obj(lock->key) ? NULL : ERR_PTR(-EINVAL); - } - - /* -@@ -724,19 +738,18 @@ - DEBUG_LOCKS_WARN_ON(!irqs_disabled()); - - class = look_up_lock_class(lock, subclass); -- if (likely(class)) -+ if (likely(!IS_ERR_OR_NULL(class))) - goto out_set_class_cache; - - /* - * Debug-check: all keys must be persistent! -- */ -- if (!static_obj(lock->key)) { -+ */ -+ if (IS_ERR(class)) { - debug_locks_off(); - printk("INFO: trying to register non-static key.\n"); - printk("the code is fine but needs lockdep annotation.\n"); - printk("turning off the locking correctness validator.\n"); - dump_stack(); -- - return NULL; - } - -@@ -3410,7 +3423,7 @@ - * Clearly if the lock hasn't been acquired _ever_, we're not - * holding it either, so report failure. - */ -- if (!class) -+ if (IS_ERR_OR_NULL(class)) - return 0; - - /* -@@ -3689,6 +3702,7 @@ - } - } - -+#ifndef CONFIG_PREEMPT_RT_FULL - /* - * We dont accurately track softirq state in e.g. - * hardirq contexts (such as on 4KSTACKS), so only -@@ -3703,6 +3717,7 @@ - DEBUG_LOCKS_WARN_ON(!current->softirqs_enabled); - } - } -+#endif - - if (!debug_locks) - print_irqtrace_events(current); -@@ -4159,7 +4174,7 @@ - * If the class exists we look it up and zap it: - */ - class = look_up_lock_class(lock, j); -- if (class) -+ if (!IS_ERR_OR_NULL(class)) - zap_class(class); - } - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/locktorture.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/locktorture.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/locktorture.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/locktorture.c 2017-04-18 17:54:26.000000000 +0200 -@@ -26,7 +26,6 @@ - #include - #include - #include --#include - #include - #include - #include -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/percpu-rwsem.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/percpu-rwsem.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/percpu-rwsem.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/percpu-rwsem.c 2017-04-18 17:54:26.000000000 +0200 -@@ -18,7 +18,7 @@ - /* ->rw_sem represents the whole percpu_rw_semaphore for lockdep */ - rcu_sync_init(&sem->rss, RCU_SCHED_SYNC); - __init_rwsem(&sem->rw_sem, name, rwsem_key); -- init_waitqueue_head(&sem->writer); -+ init_swait_queue_head(&sem->writer); - sem->readers_block = 0; - return 0; - } -@@ -103,7 +103,7 @@ - __this_cpu_dec(*sem->read_count); - - /* Prod writer to recheck readers_active */ -- wake_up(&sem->writer); -+ swake_up(&sem->writer); - } - EXPORT_SYMBOL_GPL(__percpu_up_read); - -@@ -160,7 +160,7 @@ - */ - - /* Wait for all now active readers to complete. */ -- wait_event(sem->writer, readers_active_check(sem)); -+ swait_event(sem->writer, readers_active_check(sem)); - } - EXPORT_SYMBOL_GPL(percpu_down_write); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rt.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rt.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rt.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rt.c 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,331 @@ -+/* -+ * kernel/rt.c -+ * -+ * Real-Time Preemption Support -+ * -+ * started by Ingo Molnar: -+ * -+ * Copyright (C) 2004-2006 Red Hat, Inc., Ingo Molnar -+ * Copyright (C) 2006, Timesys Corp., Thomas Gleixner -+ * -+ * historic credit for proving that Linux spinlocks can be implemented via -+ * RT-aware mutexes goes to many people: The Pmutex project (Dirk Grambow -+ * and others) who prototyped it on 2.4 and did lots of comparative -+ * research and analysis; TimeSys, for proving that you can implement a -+ * fully preemptible kernel via the use of IRQ threading and mutexes; -+ * Bill Huey for persuasively arguing on lkml that the mutex model is the -+ * right one; and to MontaVista, who ported pmutexes to 2.6. -+ * -+ * This code is a from-scratch implementation and is not based on pmutexes, -+ * but the idea of converting spinlocks to mutexes is used here too. -+ * -+ * lock debugging, locking tree, deadlock detection: -+ * -+ * Copyright (C) 2004, LynuxWorks, Inc., Igor Manyilov, Bill Huey -+ * Released under the General Public License (GPL). -+ * -+ * Includes portions of the generic R/W semaphore implementation from: -+ * -+ * Copyright (c) 2001 David Howells (dhowells@redhat.com). -+ * - Derived partially from idea by Andrea Arcangeli -+ * - Derived also from comments by Linus -+ * -+ * Pending ownership of locks and ownership stealing: -+ * -+ * Copyright (C) 2005, Kihon Technologies Inc., Steven Rostedt -+ * -+ * (also by Steven Rostedt) -+ * - Converted single pi_lock to individual task locks. -+ * -+ * By Esben Nielsen: -+ * Doing priority inheritance with help of the scheduler. -+ * -+ * Copyright (C) 2006, Timesys Corp., Thomas Gleixner -+ * - major rework based on Esben Nielsens initial patch -+ * - replaced thread_info references by task_struct refs -+ * - removed task->pending_owner dependency -+ * - BKL drop/reacquire for semaphore style locks to avoid deadlocks -+ * in the scheduler return path as discussed with Steven Rostedt -+ * -+ * Copyright (C) 2006, Kihon Technologies Inc. -+ * Steven Rostedt -+ * - debugged and patched Thomas Gleixner's rework. -+ * - added back the cmpxchg to the rework. -+ * - turned atomic require back on for SMP. -+ */ -+ -+#include -+#include -+#include -+#include -+#include -+#include -+#include -+#include -+#include -+#include -+#include -+#include -+ -+#include "rtmutex_common.h" -+ -+/* -+ * struct mutex functions -+ */ -+void __mutex_do_init(struct mutex *mutex, const char *name, -+ struct lock_class_key *key) -+{ -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+ /* -+ * Make sure we are not reinitializing a held lock: -+ */ -+ debug_check_no_locks_freed((void *)mutex, sizeof(*mutex)); -+ lockdep_init_map(&mutex->dep_map, name, key, 0); -+#endif -+ mutex->lock.save_state = 0; -+} -+EXPORT_SYMBOL(__mutex_do_init); -+ -+void __lockfunc _mutex_lock(struct mutex *lock) -+{ -+ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_); -+ rt_mutex_lock(&lock->lock); -+} -+EXPORT_SYMBOL(_mutex_lock); -+ -+int __lockfunc _mutex_lock_interruptible(struct mutex *lock) -+{ -+ int ret; -+ -+ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_); -+ ret = rt_mutex_lock_interruptible(&lock->lock); -+ if (ret) -+ mutex_release(&lock->dep_map, 1, _RET_IP_); -+ return ret; -+} -+EXPORT_SYMBOL(_mutex_lock_interruptible); -+ -+int __lockfunc _mutex_lock_killable(struct mutex *lock) -+{ -+ int ret; -+ -+ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_); -+ ret = rt_mutex_lock_killable(&lock->lock); -+ if (ret) -+ mutex_release(&lock->dep_map, 1, _RET_IP_); -+ return ret; -+} -+EXPORT_SYMBOL(_mutex_lock_killable); -+ -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+void __lockfunc _mutex_lock_nested(struct mutex *lock, int subclass) -+{ -+ mutex_acquire_nest(&lock->dep_map, subclass, 0, NULL, _RET_IP_); -+ rt_mutex_lock(&lock->lock); -+} -+EXPORT_SYMBOL(_mutex_lock_nested); -+ -+void __lockfunc _mutex_lock_nest_lock(struct mutex *lock, struct lockdep_map *nest) -+{ -+ mutex_acquire_nest(&lock->dep_map, 0, 0, nest, _RET_IP_); -+ rt_mutex_lock(&lock->lock); -+} -+EXPORT_SYMBOL(_mutex_lock_nest_lock); -+ -+int __lockfunc _mutex_lock_interruptible_nested(struct mutex *lock, int subclass) -+{ -+ int ret; -+ -+ mutex_acquire_nest(&lock->dep_map, subclass, 0, NULL, _RET_IP_); -+ ret = rt_mutex_lock_interruptible(&lock->lock); -+ if (ret) -+ mutex_release(&lock->dep_map, 1, _RET_IP_); -+ return ret; -+} -+EXPORT_SYMBOL(_mutex_lock_interruptible_nested); -+ -+int __lockfunc _mutex_lock_killable_nested(struct mutex *lock, int subclass) -+{ -+ int ret; -+ -+ mutex_acquire(&lock->dep_map, subclass, 0, _RET_IP_); -+ ret = rt_mutex_lock_killable(&lock->lock); -+ if (ret) -+ mutex_release(&lock->dep_map, 1, _RET_IP_); -+ return ret; -+} -+EXPORT_SYMBOL(_mutex_lock_killable_nested); -+#endif -+ -+int __lockfunc _mutex_trylock(struct mutex *lock) -+{ -+ int ret = rt_mutex_trylock(&lock->lock); -+ -+ if (ret) -+ mutex_acquire(&lock->dep_map, 0, 1, _RET_IP_); -+ -+ return ret; -+} -+EXPORT_SYMBOL(_mutex_trylock); -+ -+void __lockfunc _mutex_unlock(struct mutex *lock) -+{ -+ mutex_release(&lock->dep_map, 1, _RET_IP_); -+ rt_mutex_unlock(&lock->lock); -+} -+EXPORT_SYMBOL(_mutex_unlock); -+ -+/* -+ * rwlock_t functions -+ */ -+int __lockfunc rt_write_trylock(rwlock_t *rwlock) -+{ -+ int ret; -+ -+ migrate_disable(); -+ ret = rt_mutex_trylock(&rwlock->lock); -+ if (ret) -+ rwlock_acquire(&rwlock->dep_map, 0, 1, _RET_IP_); -+ else -+ migrate_enable(); -+ -+ return ret; -+} -+EXPORT_SYMBOL(rt_write_trylock); -+ -+int __lockfunc rt_write_trylock_irqsave(rwlock_t *rwlock, unsigned long *flags) -+{ -+ int ret; -+ -+ *flags = 0; -+ ret = rt_write_trylock(rwlock); -+ return ret; -+} -+EXPORT_SYMBOL(rt_write_trylock_irqsave); -+ -+int __lockfunc rt_read_trylock(rwlock_t *rwlock) -+{ -+ struct rt_mutex *lock = &rwlock->lock; -+ int ret = 1; -+ -+ /* -+ * recursive read locks succeed when current owns the lock, -+ * but not when read_depth == 0 which means that the lock is -+ * write locked. -+ */ -+ if (rt_mutex_owner(lock) != current) { -+ migrate_disable(); -+ ret = rt_mutex_trylock(lock); -+ if (ret) -+ rwlock_acquire(&rwlock->dep_map, 0, 1, _RET_IP_); -+ else -+ migrate_enable(); -+ -+ } else if (!rwlock->read_depth) { -+ ret = 0; -+ } -+ -+ if (ret) -+ rwlock->read_depth++; -+ -+ return ret; -+} -+EXPORT_SYMBOL(rt_read_trylock); -+ -+void __lockfunc rt_write_lock(rwlock_t *rwlock) -+{ -+ rwlock_acquire(&rwlock->dep_map, 0, 0, _RET_IP_); -+ __rt_spin_lock(&rwlock->lock); -+} -+EXPORT_SYMBOL(rt_write_lock); -+ -+void __lockfunc rt_read_lock(rwlock_t *rwlock) -+{ -+ struct rt_mutex *lock = &rwlock->lock; -+ -+ -+ /* -+ * recursive read locks succeed when current owns the lock -+ */ -+ if (rt_mutex_owner(lock) != current) { -+ rwlock_acquire(&rwlock->dep_map, 0, 0, _RET_IP_); -+ __rt_spin_lock(lock); -+ } -+ rwlock->read_depth++; -+} -+ -+EXPORT_SYMBOL(rt_read_lock); -+ -+void __lockfunc rt_write_unlock(rwlock_t *rwlock) -+{ -+ /* NOTE: we always pass in '1' for nested, for simplicity */ -+ rwlock_release(&rwlock->dep_map, 1, _RET_IP_); -+ __rt_spin_unlock(&rwlock->lock); -+ migrate_enable(); -+} -+EXPORT_SYMBOL(rt_write_unlock); -+ -+void __lockfunc rt_read_unlock(rwlock_t *rwlock) -+{ -+ /* Release the lock only when read_depth is down to 0 */ -+ if (--rwlock->read_depth == 0) { -+ rwlock_release(&rwlock->dep_map, 1, _RET_IP_); -+ __rt_spin_unlock(&rwlock->lock); -+ migrate_enable(); -+ } -+} -+EXPORT_SYMBOL(rt_read_unlock); -+ -+unsigned long __lockfunc rt_write_lock_irqsave(rwlock_t *rwlock) -+{ -+ rt_write_lock(rwlock); -+ -+ return 0; -+} -+EXPORT_SYMBOL(rt_write_lock_irqsave); -+ -+unsigned long __lockfunc rt_read_lock_irqsave(rwlock_t *rwlock) -+{ -+ rt_read_lock(rwlock); -+ -+ return 0; -+} -+EXPORT_SYMBOL(rt_read_lock_irqsave); -+ -+void __rt_rwlock_init(rwlock_t *rwlock, char *name, struct lock_class_key *key) -+{ -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+ /* -+ * Make sure we are not reinitializing a held lock: -+ */ -+ debug_check_no_locks_freed((void *)rwlock, sizeof(*rwlock)); -+ lockdep_init_map(&rwlock->dep_map, name, key, 0); -+#endif -+ rwlock->lock.save_state = 1; -+ rwlock->read_depth = 0; -+} -+EXPORT_SYMBOL(__rt_rwlock_init); -+ -+/** -+ * atomic_dec_and_mutex_lock - return holding mutex if we dec to 0 -+ * @cnt: the atomic which we are to dec -+ * @lock: the mutex to return holding if we dec to 0 -+ * -+ * return true and hold lock if we dec to 0, return false otherwise -+ */ -+int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock) -+{ -+ /* dec if we can't possibly hit 0 */ -+ if (atomic_add_unless(cnt, -1, 1)) -+ return 0; -+ /* we might hit 0, so take the lock */ -+ mutex_lock(lock); -+ if (!atomic_dec_and_test(cnt)) { -+ /* when we actually did the dec, we didn't hit 0 */ -+ mutex_unlock(lock); -+ return 0; -+ } -+ /* we hit 0, and we hold the lock */ -+ return 1; -+} -+EXPORT_SYMBOL(atomic_dec_and_mutex_lock); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rtmutex-debug.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rtmutex-debug.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rtmutex-debug.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rtmutex-debug.c 2017-04-18 17:54:26.000000000 +0200 -@@ -173,12 +173,3 @@ - lock->name = name; - } - --void --rt_mutex_deadlock_account_lock(struct rt_mutex *lock, struct task_struct *task) --{ --} -- --void rt_mutex_deadlock_account_unlock(struct task_struct *task) --{ --} -- -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rtmutex-debug.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rtmutex-debug.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rtmutex-debug.h 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rtmutex-debug.h 2017-04-18 17:54:26.000000000 +0200 -@@ -9,9 +9,6 @@ - * This file contains macros used solely by rtmutex.c. Debug version. - */ - --extern void --rt_mutex_deadlock_account_lock(struct rt_mutex *lock, struct task_struct *task); --extern void rt_mutex_deadlock_account_unlock(struct task_struct *task); - extern void debug_rt_mutex_init_waiter(struct rt_mutex_waiter *waiter); - extern void debug_rt_mutex_free_waiter(struct rt_mutex_waiter *waiter); - extern void debug_rt_mutex_init(struct rt_mutex *lock, const char *name); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rtmutex.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rtmutex.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rtmutex.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rtmutex.c 2017-04-18 17:54:26.000000000 +0200 -@@ -7,6 +7,11 @@ - * Copyright (C) 2005-2006 Timesys Corp., Thomas Gleixner - * Copyright (C) 2005 Kihon Technologies Inc., Steven Rostedt - * Copyright (C) 2006 Esben Nielsen -+ * Adaptive Spinlocks: -+ * Copyright (C) 2008 Novell, Inc., Gregory Haskins, Sven Dietrich, -+ * and Peter Morreale, -+ * Adaptive Spinlocks simplification: -+ * Copyright (C) 2008 Red Hat, Inc., Steven Rostedt - * - * See Documentation/locking/rt-mutex-design.txt for details. - */ -@@ -16,6 +21,7 @@ - #include - #include - #include -+#include - - #include "rtmutex_common.h" - -@@ -133,6 +139,12 @@ - WRITE_ONCE(*p, owner & ~RT_MUTEX_HAS_WAITERS); - } - -+static int rt_mutex_real_waiter(struct rt_mutex_waiter *waiter) -+{ -+ return waiter && waiter != PI_WAKEUP_INPROGRESS && -+ waiter != PI_REQUEUE_INPROGRESS; -+} -+ - /* - * We can speed up the acquire/release, if there's no debugging state to be - * set up. -@@ -414,6 +426,14 @@ - return debug_rt_mutex_detect_deadlock(waiter, chwalk); - } - -+static void rt_mutex_wake_waiter(struct rt_mutex_waiter *waiter) -+{ -+ if (waiter->savestate) -+ wake_up_lock_sleeper(waiter->task); -+ else -+ wake_up_process(waiter->task); -+} -+ - /* - * Max number of times we'll walk the boosting chain: - */ -@@ -421,7 +441,8 @@ - - static inline struct rt_mutex *task_blocked_on_lock(struct task_struct *p) - { -- return p->pi_blocked_on ? p->pi_blocked_on->lock : NULL; -+ return rt_mutex_real_waiter(p->pi_blocked_on) ? -+ p->pi_blocked_on->lock : NULL; - } - - /* -@@ -557,7 +578,7 @@ - * reached or the state of the chain has changed while we - * dropped the locks. - */ -- if (!waiter) -+ if (!rt_mutex_real_waiter(waiter)) - goto out_unlock_pi; - - /* -@@ -719,13 +740,16 @@ - * follow here. This is the end of the chain we are walking. - */ - if (!rt_mutex_owner(lock)) { -+ struct rt_mutex_waiter *lock_top_waiter; -+ - /* - * If the requeue [7] above changed the top waiter, - * then we need to wake the new top waiter up to try - * to get the lock. - */ -- if (prerequeue_top_waiter != rt_mutex_top_waiter(lock)) -- wake_up_process(rt_mutex_top_waiter(lock)->task); -+ lock_top_waiter = rt_mutex_top_waiter(lock); -+ if (prerequeue_top_waiter != lock_top_waiter) -+ rt_mutex_wake_waiter(lock_top_waiter); - raw_spin_unlock_irq(&lock->wait_lock); - return 0; - } -@@ -818,6 +842,25 @@ - return ret; - } - -+ -+#define STEAL_NORMAL 0 -+#define STEAL_LATERAL 1 -+ -+/* -+ * Note that RT tasks are excluded from lateral-steals to prevent the -+ * introduction of an unbounded latency -+ */ -+static inline int lock_is_stealable(struct task_struct *task, -+ struct task_struct *pendowner, int mode) -+{ -+ if (mode == STEAL_NORMAL || rt_task(task)) { -+ if (task->prio >= pendowner->prio) -+ return 0; -+ } else if (task->prio > pendowner->prio) -+ return 0; -+ return 1; -+} -+ - /* - * Try to take an rt-mutex - * -@@ -828,8 +871,9 @@ - * @waiter: The waiter that is queued to the lock's wait tree if the - * callsite called task_blocked_on_lock(), otherwise NULL - */ --static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, -- struct rt_mutex_waiter *waiter) -+static int __try_to_take_rt_mutex(struct rt_mutex *lock, -+ struct task_struct *task, -+ struct rt_mutex_waiter *waiter, int mode) - { - /* - * Before testing whether we can acquire @lock, we set the -@@ -866,8 +910,10 @@ - * If waiter is not the highest priority waiter of - * @lock, give up. - */ -- if (waiter != rt_mutex_top_waiter(lock)) -+ if (waiter != rt_mutex_top_waiter(lock)) { -+ /* XXX lock_is_stealable() ? */ - return 0; -+ } - - /* - * We can acquire the lock. Remove the waiter from the -@@ -885,14 +931,10 @@ - * not need to be dequeued. - */ - if (rt_mutex_has_waiters(lock)) { -- /* -- * If @task->prio is greater than or equal to -- * the top waiter priority (kernel view), -- * @task lost. -- */ -- if (task->prio >= rt_mutex_top_waiter(lock)->prio) -- return 0; -+ struct task_struct *pown = rt_mutex_top_waiter(lock)->task; - -+ if (task != pown && !lock_is_stealable(task, pown, mode)) -+ return 0; - /* - * The current top waiter stays enqueued. We - * don't have to change anything in the lock -@@ -936,10 +978,394 @@ - */ - rt_mutex_set_owner(lock, task); - -- rt_mutex_deadlock_account_lock(lock, task); -+ return 1; -+} -+ -+#ifdef CONFIG_PREEMPT_RT_FULL -+/* -+ * preemptible spin_lock functions: -+ */ -+static inline void rt_spin_lock_fastlock(struct rt_mutex *lock, -+ void (*slowfn)(struct rt_mutex *lock, -+ bool mg_off), -+ bool do_mig_dis) -+{ -+ might_sleep_no_state_check(); -+ -+ if (do_mig_dis) -+ migrate_disable(); -+ -+ if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) -+ return; -+ else -+ slowfn(lock, do_mig_dis); -+} -+ -+static inline void rt_spin_lock_fastunlock(struct rt_mutex *lock, -+ void (*slowfn)(struct rt_mutex *lock)) -+{ -+ if (likely(rt_mutex_cmpxchg_release(lock, current, NULL))) -+ return; -+ else -+ slowfn(lock); -+} -+#ifdef CONFIG_SMP -+/* -+ * Note that owner is a speculative pointer and dereferencing relies -+ * on rcu_read_lock() and the check against the lock owner. -+ */ -+static int adaptive_wait(struct rt_mutex *lock, -+ struct task_struct *owner) -+{ -+ int res = 0; - -+ rcu_read_lock(); -+ for (;;) { -+ if (owner != rt_mutex_owner(lock)) -+ break; -+ /* -+ * Ensure that owner->on_cpu is dereferenced _after_ -+ * checking the above to be valid. -+ */ -+ barrier(); -+ if (!owner->on_cpu) { -+ res = 1; -+ break; -+ } -+ cpu_relax(); -+ } -+ rcu_read_unlock(); -+ return res; -+} -+#else -+static int adaptive_wait(struct rt_mutex *lock, -+ struct task_struct *orig_owner) -+{ - return 1; - } -+#endif -+ -+static int task_blocks_on_rt_mutex(struct rt_mutex *lock, -+ struct rt_mutex_waiter *waiter, -+ struct task_struct *task, -+ enum rtmutex_chainwalk chwalk); -+/* -+ * Slow path lock function spin_lock style: this variant is very -+ * careful not to miss any non-lock wakeups. -+ * -+ * We store the current state under p->pi_lock in p->saved_state and -+ * the try_to_wake_up() code handles this accordingly. -+ */ -+static void noinline __sched rt_spin_lock_slowlock(struct rt_mutex *lock, -+ bool mg_off) -+{ -+ struct task_struct *lock_owner, *self = current; -+ struct rt_mutex_waiter waiter, *top_waiter; -+ unsigned long flags; -+ int ret; -+ -+ rt_mutex_init_waiter(&waiter, true); -+ -+ raw_spin_lock_irqsave(&lock->wait_lock, flags); -+ -+ if (__try_to_take_rt_mutex(lock, self, NULL, STEAL_LATERAL)) { -+ raw_spin_unlock_irqrestore(&lock->wait_lock, flags); -+ return; -+ } -+ -+ BUG_ON(rt_mutex_owner(lock) == self); -+ -+ /* -+ * We save whatever state the task is in and we'll restore it -+ * after acquiring the lock taking real wakeups into account -+ * as well. We are serialized via pi_lock against wakeups. See -+ * try_to_wake_up(). -+ */ -+ raw_spin_lock(&self->pi_lock); -+ self->saved_state = self->state; -+ __set_current_state_no_track(TASK_UNINTERRUPTIBLE); -+ raw_spin_unlock(&self->pi_lock); -+ -+ ret = task_blocks_on_rt_mutex(lock, &waiter, self, RT_MUTEX_MIN_CHAINWALK); -+ BUG_ON(ret); -+ -+ for (;;) { -+ /* Try to acquire the lock again. */ -+ if (__try_to_take_rt_mutex(lock, self, &waiter, STEAL_LATERAL)) -+ break; -+ -+ top_waiter = rt_mutex_top_waiter(lock); -+ lock_owner = rt_mutex_owner(lock); -+ -+ raw_spin_unlock_irqrestore(&lock->wait_lock, flags); -+ -+ debug_rt_mutex_print_deadlock(&waiter); -+ -+ if (top_waiter != &waiter || adaptive_wait(lock, lock_owner)) { -+ if (mg_off) -+ migrate_enable(); -+ schedule(); -+ if (mg_off) -+ migrate_disable(); -+ } -+ -+ raw_spin_lock_irqsave(&lock->wait_lock, flags); -+ -+ raw_spin_lock(&self->pi_lock); -+ __set_current_state_no_track(TASK_UNINTERRUPTIBLE); -+ raw_spin_unlock(&self->pi_lock); -+ } -+ -+ /* -+ * Restore the task state to current->saved_state. We set it -+ * to the original state above and the try_to_wake_up() code -+ * has possibly updated it when a real (non-rtmutex) wakeup -+ * happened while we were blocked. Clear saved_state so -+ * try_to_wakeup() does not get confused. -+ */ -+ raw_spin_lock(&self->pi_lock); -+ __set_current_state_no_track(self->saved_state); -+ self->saved_state = TASK_RUNNING; -+ raw_spin_unlock(&self->pi_lock); -+ -+ /* -+ * try_to_take_rt_mutex() sets the waiter bit -+ * unconditionally. We might have to fix that up: -+ */ -+ fixup_rt_mutex_waiters(lock); -+ -+ BUG_ON(rt_mutex_has_waiters(lock) && &waiter == rt_mutex_top_waiter(lock)); -+ BUG_ON(!RB_EMPTY_NODE(&waiter.tree_entry)); -+ -+ raw_spin_unlock_irqrestore(&lock->wait_lock, flags); -+ -+ debug_rt_mutex_free_waiter(&waiter); -+} -+ -+static void mark_wakeup_next_waiter(struct wake_q_head *wake_q, -+ struct wake_q_head *wake_sleeper_q, -+ struct rt_mutex *lock); -+/* -+ * Slow path to release a rt_mutex spin_lock style -+ */ -+static void noinline __sched rt_spin_lock_slowunlock(struct rt_mutex *lock) -+{ -+ unsigned long flags; -+ WAKE_Q(wake_q); -+ WAKE_Q(wake_sleeper_q); -+ -+ raw_spin_lock_irqsave(&lock->wait_lock, flags); -+ -+ debug_rt_mutex_unlock(lock); -+ -+ if (!rt_mutex_has_waiters(lock)) { -+ lock->owner = NULL; -+ raw_spin_unlock_irqrestore(&lock->wait_lock, flags); -+ return; -+ } -+ -+ mark_wakeup_next_waiter(&wake_q, &wake_sleeper_q, lock); -+ -+ raw_spin_unlock_irqrestore(&lock->wait_lock, flags); -+ wake_up_q(&wake_q); -+ wake_up_q_sleeper(&wake_sleeper_q); -+ -+ /* Undo pi boosting.when necessary */ -+ rt_mutex_adjust_prio(current); -+} -+ -+void __lockfunc rt_spin_lock__no_mg(spinlock_t *lock) -+{ -+ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock, false); -+ spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); -+} -+EXPORT_SYMBOL(rt_spin_lock__no_mg); -+ -+void __lockfunc rt_spin_lock(spinlock_t *lock) -+{ -+ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock, true); -+ spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); -+} -+EXPORT_SYMBOL(rt_spin_lock); -+ -+void __lockfunc __rt_spin_lock(struct rt_mutex *lock) -+{ -+ rt_spin_lock_fastlock(lock, rt_spin_lock_slowlock, true); -+} -+EXPORT_SYMBOL(__rt_spin_lock); -+ -+void __lockfunc __rt_spin_lock__no_mg(struct rt_mutex *lock) -+{ -+ rt_spin_lock_fastlock(lock, rt_spin_lock_slowlock, false); -+} -+EXPORT_SYMBOL(__rt_spin_lock__no_mg); -+ -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+void __lockfunc rt_spin_lock_nested(spinlock_t *lock, int subclass) -+{ -+ spin_acquire(&lock->dep_map, subclass, 0, _RET_IP_); -+ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock, true); -+} -+EXPORT_SYMBOL(rt_spin_lock_nested); -+#endif -+ -+void __lockfunc rt_spin_unlock__no_mg(spinlock_t *lock) -+{ -+ /* NOTE: we always pass in '1' for nested, for simplicity */ -+ spin_release(&lock->dep_map, 1, _RET_IP_); -+ rt_spin_lock_fastunlock(&lock->lock, rt_spin_lock_slowunlock); -+} -+EXPORT_SYMBOL(rt_spin_unlock__no_mg); -+ -+void __lockfunc rt_spin_unlock(spinlock_t *lock) -+{ -+ /* NOTE: we always pass in '1' for nested, for simplicity */ -+ spin_release(&lock->dep_map, 1, _RET_IP_); -+ rt_spin_lock_fastunlock(&lock->lock, rt_spin_lock_slowunlock); -+ migrate_enable(); -+} -+EXPORT_SYMBOL(rt_spin_unlock); -+ -+void __lockfunc __rt_spin_unlock(struct rt_mutex *lock) -+{ -+ rt_spin_lock_fastunlock(lock, rt_spin_lock_slowunlock); -+} -+EXPORT_SYMBOL(__rt_spin_unlock); -+ -+/* -+ * Wait for the lock to get unlocked: instead of polling for an unlock -+ * (like raw spinlocks do), we lock and unlock, to force the kernel to -+ * schedule if there's contention: -+ */ -+void __lockfunc rt_spin_unlock_wait(spinlock_t *lock) -+{ -+ spin_lock(lock); -+ spin_unlock(lock); -+} -+EXPORT_SYMBOL(rt_spin_unlock_wait); -+ -+int __lockfunc rt_spin_trylock__no_mg(spinlock_t *lock) -+{ -+ int ret; -+ -+ ret = rt_mutex_trylock(&lock->lock); -+ if (ret) -+ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); -+ return ret; -+} -+EXPORT_SYMBOL(rt_spin_trylock__no_mg); -+ -+int __lockfunc rt_spin_trylock(spinlock_t *lock) -+{ -+ int ret; -+ -+ migrate_disable(); -+ ret = rt_mutex_trylock(&lock->lock); -+ if (ret) -+ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); -+ else -+ migrate_enable(); -+ return ret; -+} -+EXPORT_SYMBOL(rt_spin_trylock); -+ -+int __lockfunc rt_spin_trylock_bh(spinlock_t *lock) -+{ -+ int ret; -+ -+ local_bh_disable(); -+ ret = rt_mutex_trylock(&lock->lock); -+ if (ret) { -+ migrate_disable(); -+ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); -+ } else -+ local_bh_enable(); -+ return ret; -+} -+EXPORT_SYMBOL(rt_spin_trylock_bh); -+ -+int __lockfunc rt_spin_trylock_irqsave(spinlock_t *lock, unsigned long *flags) -+{ -+ int ret; -+ -+ *flags = 0; -+ ret = rt_mutex_trylock(&lock->lock); -+ if (ret) { -+ migrate_disable(); -+ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); -+ } -+ return ret; -+} -+EXPORT_SYMBOL(rt_spin_trylock_irqsave); -+ -+int atomic_dec_and_spin_lock(atomic_t *atomic, spinlock_t *lock) -+{ -+ /* Subtract 1 from counter unless that drops it to 0 (ie. it was 1) */ -+ if (atomic_add_unless(atomic, -1, 1)) -+ return 0; -+ rt_spin_lock(lock); -+ if (atomic_dec_and_test(atomic)) -+ return 1; -+ rt_spin_unlock(lock); -+ return 0; -+} -+EXPORT_SYMBOL(atomic_dec_and_spin_lock); -+ -+ void -+__rt_spin_lock_init(spinlock_t *lock, char *name, struct lock_class_key *key) -+{ -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+ /* -+ * Make sure we are not reinitializing a held lock: -+ */ -+ debug_check_no_locks_freed((void *)lock, sizeof(*lock)); -+ lockdep_init_map(&lock->dep_map, name, key, 0); -+#endif -+} -+EXPORT_SYMBOL(__rt_spin_lock_init); -+ -+#endif /* PREEMPT_RT_FULL */ -+ -+#ifdef CONFIG_PREEMPT_RT_FULL -+ static inline int __sched -+__mutex_lock_check_stamp(struct rt_mutex *lock, struct ww_acquire_ctx *ctx) -+{ -+ struct ww_mutex *ww = container_of(lock, struct ww_mutex, base.lock); -+ struct ww_acquire_ctx *hold_ctx = ACCESS_ONCE(ww->ctx); -+ -+ if (!hold_ctx) -+ return 0; -+ -+ if (unlikely(ctx == hold_ctx)) -+ return -EALREADY; -+ -+ if (ctx->stamp - hold_ctx->stamp <= LONG_MAX && -+ (ctx->stamp != hold_ctx->stamp || ctx > hold_ctx)) { -+#ifdef CONFIG_DEBUG_MUTEXES -+ DEBUG_LOCKS_WARN_ON(ctx->contending_lock); -+ ctx->contending_lock = ww; -+#endif -+ return -EDEADLK; -+ } -+ -+ return 0; -+} -+#else -+ static inline int __sched -+__mutex_lock_check_stamp(struct rt_mutex *lock, struct ww_acquire_ctx *ctx) -+{ -+ BUG(); -+ return 0; -+} -+ -+#endif -+ -+static inline int -+try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, -+ struct rt_mutex_waiter *waiter) -+{ -+ return __try_to_take_rt_mutex(lock, task, waiter, STEAL_NORMAL); -+} - - /* - * Task blocks on lock. -@@ -971,6 +1397,23 @@ - return -EDEADLK; - - raw_spin_lock(&task->pi_lock); -+ -+ /* -+ * In the case of futex requeue PI, this will be a proxy -+ * lock. The task will wake unaware that it is enqueueed on -+ * this lock. Avoid blocking on two locks and corrupting -+ * pi_blocked_on via the PI_WAKEUP_INPROGRESS -+ * flag. futex_wait_requeue_pi() sets this when it wakes up -+ * before requeue (due to a signal or timeout). Do not enqueue -+ * the task if PI_WAKEUP_INPROGRESS is set. -+ */ -+ if (task != current && task->pi_blocked_on == PI_WAKEUP_INPROGRESS) { -+ raw_spin_unlock(&task->pi_lock); -+ return -EAGAIN; -+ } -+ -+ BUG_ON(rt_mutex_real_waiter(task->pi_blocked_on)); -+ - __rt_mutex_adjust_prio(task); - waiter->task = task; - waiter->lock = lock; -@@ -994,7 +1437,7 @@ - rt_mutex_enqueue_pi(owner, waiter); - - __rt_mutex_adjust_prio(owner); -- if (owner->pi_blocked_on) -+ if (rt_mutex_real_waiter(owner->pi_blocked_on)) - chain_walk = 1; - } else if (rt_mutex_cond_detect_deadlock(waiter, chwalk)) { - chain_walk = 1; -@@ -1036,6 +1479,7 @@ - * Called with lock->wait_lock held and interrupts disabled. - */ - static void mark_wakeup_next_waiter(struct wake_q_head *wake_q, -+ struct wake_q_head *wake_sleeper_q, - struct rt_mutex *lock) - { - struct rt_mutex_waiter *waiter; -@@ -1064,7 +1508,10 @@ - - raw_spin_unlock(¤t->pi_lock); - -- wake_q_add(wake_q, waiter->task); -+ if (waiter->savestate) -+ wake_q_add(wake_sleeper_q, waiter->task); -+ else -+ wake_q_add(wake_q, waiter->task); - } - - /* -@@ -1078,7 +1525,7 @@ - { - bool is_top_waiter = (waiter == rt_mutex_top_waiter(lock)); - struct task_struct *owner = rt_mutex_owner(lock); -- struct rt_mutex *next_lock; -+ struct rt_mutex *next_lock = NULL; - - raw_spin_lock(¤t->pi_lock); - rt_mutex_dequeue(lock, waiter); -@@ -1102,7 +1549,8 @@ - __rt_mutex_adjust_prio(owner); - - /* Store the lock on which owner is blocked or NULL */ -- next_lock = task_blocked_on_lock(owner); -+ if (rt_mutex_real_waiter(owner->pi_blocked_on)) -+ next_lock = task_blocked_on_lock(owner); - - raw_spin_unlock(&owner->pi_lock); - -@@ -1138,21 +1586,30 @@ - raw_spin_lock_irqsave(&task->pi_lock, flags); - - waiter = task->pi_blocked_on; -- if (!waiter || (waiter->prio == task->prio && -+ if (!rt_mutex_real_waiter(waiter) || (waiter->prio == task->prio && - !dl_prio(task->prio))) { - raw_spin_unlock_irqrestore(&task->pi_lock, flags); - return; - } - next_lock = waiter->lock; -- raw_spin_unlock_irqrestore(&task->pi_lock, flags); - - /* gets dropped in rt_mutex_adjust_prio_chain()! */ - get_task_struct(task); - -+ raw_spin_unlock_irqrestore(&task->pi_lock, flags); - rt_mutex_adjust_prio_chain(task, RT_MUTEX_MIN_CHAINWALK, NULL, - next_lock, NULL, task); - } - -+void rt_mutex_init_waiter(struct rt_mutex_waiter *waiter, bool savestate) -+{ -+ debug_rt_mutex_init_waiter(waiter); -+ RB_CLEAR_NODE(&waiter->pi_tree_entry); -+ RB_CLEAR_NODE(&waiter->tree_entry); -+ waiter->task = NULL; -+ waiter->savestate = savestate; -+} -+ - /** - * __rt_mutex_slowlock() - Perform the wait-wake-try-to-take loop - * @lock: the rt_mutex to take -@@ -1166,7 +1623,8 @@ - static int __sched - __rt_mutex_slowlock(struct rt_mutex *lock, int state, - struct hrtimer_sleeper *timeout, -- struct rt_mutex_waiter *waiter) -+ struct rt_mutex_waiter *waiter, -+ struct ww_acquire_ctx *ww_ctx) - { - int ret = 0; - -@@ -1175,16 +1633,17 @@ - if (try_to_take_rt_mutex(lock, current, waiter)) - break; - -- /* -- * TASK_INTERRUPTIBLE checks for signals and -- * timeout. Ignored otherwise. -- */ -- if (unlikely(state == TASK_INTERRUPTIBLE)) { -- /* Signal pending? */ -- if (signal_pending(current)) -- ret = -EINTR; -- if (timeout && !timeout->task) -- ret = -ETIMEDOUT; -+ if (timeout && !timeout->task) { -+ ret = -ETIMEDOUT; -+ break; -+ } -+ if (signal_pending_state(state, current)) { -+ ret = -EINTR; -+ break; -+ } -+ -+ if (ww_ctx && ww_ctx->acquired > 0) { -+ ret = __mutex_lock_check_stamp(lock, ww_ctx); - if (ret) - break; - } -@@ -1223,35 +1682,94 @@ - } - } - --/* -- * Slow path lock function: -- */ --static int __sched --rt_mutex_slowlock(struct rt_mutex *lock, int state, -- struct hrtimer_sleeper *timeout, -- enum rtmutex_chainwalk chwalk) -+static __always_inline void ww_mutex_lock_acquired(struct ww_mutex *ww, -+ struct ww_acquire_ctx *ww_ctx) - { -- struct rt_mutex_waiter waiter; -- unsigned long flags; -- int ret = 0; -+#ifdef CONFIG_DEBUG_MUTEXES -+ /* -+ * If this WARN_ON triggers, you used ww_mutex_lock to acquire, -+ * but released with a normal mutex_unlock in this call. -+ * -+ * This should never happen, always use ww_mutex_unlock. -+ */ -+ DEBUG_LOCKS_WARN_ON(ww->ctx); - -- debug_rt_mutex_init_waiter(&waiter); -- RB_CLEAR_NODE(&waiter.pi_tree_entry); -- RB_CLEAR_NODE(&waiter.tree_entry); -+ /* -+ * Not quite done after calling ww_acquire_done() ? -+ */ -+ DEBUG_LOCKS_WARN_ON(ww_ctx->done_acquire); -+ -+ if (ww_ctx->contending_lock) { -+ /* -+ * After -EDEADLK you tried to -+ * acquire a different ww_mutex? Bad! -+ */ -+ DEBUG_LOCKS_WARN_ON(ww_ctx->contending_lock != ww); -+ -+ /* -+ * You called ww_mutex_lock after receiving -EDEADLK, -+ * but 'forgot' to unlock everything else first? -+ */ -+ DEBUG_LOCKS_WARN_ON(ww_ctx->acquired > 0); -+ ww_ctx->contending_lock = NULL; -+ } - - /* -- * Technically we could use raw_spin_[un]lock_irq() here, but this can -- * be called in early boot if the cmpxchg() fast path is disabled -- * (debug, no architecture support). In this case we will acquire the -- * rtmutex with lock->wait_lock held. But we cannot unconditionally -- * enable interrupts in that early boot case. So we need to use the -- * irqsave/restore variants. -+ * Naughty, using a different class will lead to undefined behavior! - */ -- raw_spin_lock_irqsave(&lock->wait_lock, flags); -+ DEBUG_LOCKS_WARN_ON(ww_ctx->ww_class != ww->ww_class); -+#endif -+ ww_ctx->acquired++; -+} -+ -+#ifdef CONFIG_PREEMPT_RT_FULL -+static void ww_mutex_account_lock(struct rt_mutex *lock, -+ struct ww_acquire_ctx *ww_ctx) -+{ -+ struct ww_mutex *ww = container_of(lock, struct ww_mutex, base.lock); -+ struct rt_mutex_waiter *waiter, *n; -+ -+ /* -+ * This branch gets optimized out for the common case, -+ * and is only important for ww_mutex_lock. -+ */ -+ ww_mutex_lock_acquired(ww, ww_ctx); -+ ww->ctx = ww_ctx; -+ -+ /* -+ * Give any possible sleeping processes the chance to wake up, -+ * so they can recheck if they have to back off. -+ */ -+ rbtree_postorder_for_each_entry_safe(waiter, n, &lock->waiters, -+ tree_entry) { -+ /* XXX debug rt mutex waiter wakeup */ -+ -+ BUG_ON(waiter->lock != lock); -+ rt_mutex_wake_waiter(waiter); -+ } -+} -+ -+#else -+ -+static void ww_mutex_account_lock(struct rt_mutex *lock, -+ struct ww_acquire_ctx *ww_ctx) -+{ -+ BUG(); -+} -+#endif -+ -+int __sched rt_mutex_slowlock_locked(struct rt_mutex *lock, int state, -+ struct hrtimer_sleeper *timeout, -+ enum rtmutex_chainwalk chwalk, -+ struct ww_acquire_ctx *ww_ctx, -+ struct rt_mutex_waiter *waiter) -+{ -+ int ret; - - /* Try to acquire the lock again: */ - if (try_to_take_rt_mutex(lock, current, NULL)) { -- raw_spin_unlock_irqrestore(&lock->wait_lock, flags); -+ if (ww_ctx) -+ ww_mutex_account_lock(lock, ww_ctx); - return 0; - } - -@@ -1261,17 +1779,27 @@ - if (unlikely(timeout)) - hrtimer_start_expires(&timeout->timer, HRTIMER_MODE_ABS); - -- ret = task_blocks_on_rt_mutex(lock, &waiter, current, chwalk); -+ ret = task_blocks_on_rt_mutex(lock, waiter, current, chwalk); - -- if (likely(!ret)) -+ if (likely(!ret)) { - /* sleep on the mutex */ -- ret = __rt_mutex_slowlock(lock, state, timeout, &waiter); -+ ret = __rt_mutex_slowlock(lock, state, timeout, waiter, -+ ww_ctx); -+ } else if (ww_ctx) { -+ /* ww_mutex received EDEADLK, let it become EALREADY */ -+ ret = __mutex_lock_check_stamp(lock, ww_ctx); -+ BUG_ON(!ret); -+ } - - if (unlikely(ret)) { - __set_current_state(TASK_RUNNING); - if (rt_mutex_has_waiters(lock)) -- remove_waiter(lock, &waiter); -- rt_mutex_handle_deadlock(ret, chwalk, &waiter); -+ remove_waiter(lock, waiter); -+ /* ww_mutex want to report EDEADLK/EALREADY, let them */ -+ if (!ww_ctx) -+ rt_mutex_handle_deadlock(ret, chwalk, waiter); -+ } else if (ww_ctx) { -+ ww_mutex_account_lock(lock, ww_ctx); - } - - /* -@@ -1279,6 +1807,36 @@ - * unconditionally. We might have to fix that up. - */ - fixup_rt_mutex_waiters(lock); -+ return ret; -+} -+ -+/* -+ * Slow path lock function: -+ */ -+static int __sched -+rt_mutex_slowlock(struct rt_mutex *lock, int state, -+ struct hrtimer_sleeper *timeout, -+ enum rtmutex_chainwalk chwalk, -+ struct ww_acquire_ctx *ww_ctx) -+{ -+ struct rt_mutex_waiter waiter; -+ unsigned long flags; -+ int ret = 0; -+ -+ rt_mutex_init_waiter(&waiter, false); -+ -+ /* -+ * Technically we could use raw_spin_[un]lock_irq() here, but this can -+ * be called in early boot if the cmpxchg() fast path is disabled -+ * (debug, no architecture support). In this case we will acquire the -+ * rtmutex with lock->wait_lock held. But we cannot unconditionally -+ * enable interrupts in that early boot case. So we need to use the -+ * irqsave/restore variants. -+ */ -+ raw_spin_lock_irqsave(&lock->wait_lock, flags); -+ -+ ret = rt_mutex_slowlock_locked(lock, state, timeout, chwalk, ww_ctx, -+ &waiter); - - raw_spin_unlock_irqrestore(&lock->wait_lock, flags); - -@@ -1331,7 +1889,8 @@ - * Return whether the current task needs to undo a potential priority boosting. - */ - static bool __sched rt_mutex_slowunlock(struct rt_mutex *lock, -- struct wake_q_head *wake_q) -+ struct wake_q_head *wake_q, -+ struct wake_q_head *wake_sleeper_q) - { - unsigned long flags; - -@@ -1340,8 +1899,6 @@ - - debug_rt_mutex_unlock(lock); - -- rt_mutex_deadlock_account_unlock(current); -- - /* - * We must be careful here if the fast path is enabled. If we - * have no waiters queued we cannot set owner to NULL here -@@ -1387,7 +1944,7 @@ - * - * Queue the next waiter for wakeup once we release the wait_lock. - */ -- mark_wakeup_next_waiter(wake_q, lock); -+ mark_wakeup_next_waiter(wake_q, wake_sleeper_q, lock); - - raw_spin_unlock_irqrestore(&lock->wait_lock, flags); - -@@ -1403,63 +1960,79 @@ - */ - static inline int - rt_mutex_fastlock(struct rt_mutex *lock, int state, -+ struct ww_acquire_ctx *ww_ctx, - int (*slowfn)(struct rt_mutex *lock, int state, - struct hrtimer_sleeper *timeout, -- enum rtmutex_chainwalk chwalk)) -+ enum rtmutex_chainwalk chwalk, -+ struct ww_acquire_ctx *ww_ctx)) - { -- if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) { -- rt_mutex_deadlock_account_lock(lock, current); -+ if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) - return 0; -- } else -- return slowfn(lock, state, NULL, RT_MUTEX_MIN_CHAINWALK); -+ -+ return slowfn(lock, state, NULL, RT_MUTEX_MIN_CHAINWALK, ww_ctx); - } - - static inline int - rt_mutex_timed_fastlock(struct rt_mutex *lock, int state, - struct hrtimer_sleeper *timeout, - enum rtmutex_chainwalk chwalk, -+ struct ww_acquire_ctx *ww_ctx, - int (*slowfn)(struct rt_mutex *lock, int state, - struct hrtimer_sleeper *timeout, -- enum rtmutex_chainwalk chwalk)) -+ enum rtmutex_chainwalk chwalk, -+ struct ww_acquire_ctx *ww_ctx)) - { - if (chwalk == RT_MUTEX_MIN_CHAINWALK && -- likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) { -- rt_mutex_deadlock_account_lock(lock, current); -+ likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) - return 0; -- } else -- return slowfn(lock, state, timeout, chwalk); -+ -+ return slowfn(lock, state, timeout, chwalk, ww_ctx); - } - - static inline int - rt_mutex_fasttrylock(struct rt_mutex *lock, - int (*slowfn)(struct rt_mutex *lock)) - { -- if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) { -- rt_mutex_deadlock_account_lock(lock, current); -+ if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) - return 1; -- } -+ - return slowfn(lock); - } - - static inline void - rt_mutex_fastunlock(struct rt_mutex *lock, - bool (*slowfn)(struct rt_mutex *lock, -- struct wake_q_head *wqh)) -+ struct wake_q_head *wqh, -+ struct wake_q_head *wq_sleeper)) - { - WAKE_Q(wake_q); -+ WAKE_Q(wake_sleeper_q); -+ bool deboost; - -- if (likely(rt_mutex_cmpxchg_release(lock, current, NULL))) { -- rt_mutex_deadlock_account_unlock(current); -+ if (likely(rt_mutex_cmpxchg_release(lock, current, NULL))) -+ return; - -- } else { -- bool deboost = slowfn(lock, &wake_q); -+ deboost = slowfn(lock, &wake_q, &wake_sleeper_q); - -- wake_up_q(&wake_q); -+ wake_up_q(&wake_q); -+ wake_up_q_sleeper(&wake_sleeper_q); - -- /* Undo pi boosting if necessary: */ -- if (deboost) -- rt_mutex_adjust_prio(current); -- } -+ /* Undo pi boosting if necessary: */ -+ if (deboost) -+ rt_mutex_adjust_prio(current); -+} -+ -+/** -+ * rt_mutex_lock_state - lock a rt_mutex with a given state -+ * -+ * @lock: The rt_mutex to be locked -+ * @state: The state to set when blocking on the rt_mutex -+ */ -+int __sched rt_mutex_lock_state(struct rt_mutex *lock, int state) -+{ -+ might_sleep(); -+ -+ return rt_mutex_fastlock(lock, state, NULL, rt_mutex_slowlock); - } - - /** -@@ -1469,15 +2042,13 @@ - */ - void __sched rt_mutex_lock(struct rt_mutex *lock) - { -- might_sleep(); -- -- rt_mutex_fastlock(lock, TASK_UNINTERRUPTIBLE, rt_mutex_slowlock); -+ rt_mutex_lock_state(lock, TASK_UNINTERRUPTIBLE); - } - EXPORT_SYMBOL_GPL(rt_mutex_lock); - - /** - * rt_mutex_lock_interruptible - lock a rt_mutex interruptible -- * -+ ** - * @lock: the rt_mutex to be locked - * - * Returns: -@@ -1486,23 +2057,32 @@ - */ - int __sched rt_mutex_lock_interruptible(struct rt_mutex *lock) - { -- might_sleep(); -- -- return rt_mutex_fastlock(lock, TASK_INTERRUPTIBLE, rt_mutex_slowlock); -+ return rt_mutex_lock_state(lock, TASK_INTERRUPTIBLE); - } - EXPORT_SYMBOL_GPL(rt_mutex_lock_interruptible); - --/* -- * Futex variant with full deadlock detection. -+/** -+ * rt_mutex_lock_killable - lock a rt_mutex killable -+ * -+ * @lock: the rt_mutex to be locked -+ * @detect_deadlock: deadlock detection on/off -+ * -+ * Returns: -+ * 0 on success -+ * -EINTR when interrupted by a signal - */ --int rt_mutex_timed_futex_lock(struct rt_mutex *lock, -- struct hrtimer_sleeper *timeout) -+int __sched rt_mutex_lock_killable(struct rt_mutex *lock) - { -- might_sleep(); -+ return rt_mutex_lock_state(lock, TASK_KILLABLE); -+} -+EXPORT_SYMBOL_GPL(rt_mutex_lock_killable); - -- return rt_mutex_timed_fastlock(lock, TASK_INTERRUPTIBLE, timeout, -- RT_MUTEX_FULL_CHAINWALK, -- rt_mutex_slowlock); -+/* -+ * Futex variant, must not use fastpath. -+ */ -+int __sched rt_mutex_futex_trylock(struct rt_mutex *lock) -+{ -+ return rt_mutex_slowtrylock(lock); - } - - /** -@@ -1525,6 +2105,7 @@ - - return rt_mutex_timed_fastlock(lock, TASK_INTERRUPTIBLE, timeout, - RT_MUTEX_MIN_CHAINWALK, -+ NULL, - rt_mutex_slowlock); - } - EXPORT_SYMBOL_GPL(rt_mutex_timed_lock); -@@ -1542,7 +2123,11 @@ - */ - int __sched rt_mutex_trylock(struct rt_mutex *lock) - { -+#ifdef CONFIG_PREEMPT_RT_FULL -+ if (WARN_ON_ONCE(in_irq() || in_nmi())) -+#else - if (WARN_ON_ONCE(in_irq() || in_nmi() || in_serving_softirq())) -+#endif - return 0; - - return rt_mutex_fasttrylock(lock, rt_mutex_slowtrylock); -@@ -1561,20 +2146,41 @@ - EXPORT_SYMBOL_GPL(rt_mutex_unlock); - - /** -- * rt_mutex_futex_unlock - Futex variant of rt_mutex_unlock -- * @lock: the rt_mutex to be unlocked -- * -- * Returns: true/false indicating whether priority adjustment is -- * required or not. -+ * Futex variant, that since futex variants do not use the fast-path, can be -+ * simple and will not need to retry. - */ --bool __sched rt_mutex_futex_unlock(struct rt_mutex *lock, -- struct wake_q_head *wqh) -+bool __sched __rt_mutex_futex_unlock(struct rt_mutex *lock, -+ struct wake_q_head *wake_q, -+ struct wake_q_head *wq_sleeper) -+{ -+ lockdep_assert_held(&lock->wait_lock); -+ -+ debug_rt_mutex_unlock(lock); -+ -+ if (!rt_mutex_has_waiters(lock)) { -+ lock->owner = NULL; -+ return false; /* done */ -+ } -+ -+ mark_wakeup_next_waiter(wake_q, wq_sleeper, lock); -+ return true; /* deboost and wakeups */ -+} -+ -+void __sched rt_mutex_futex_unlock(struct rt_mutex *lock) - { -- if (likely(rt_mutex_cmpxchg_release(lock, current, NULL))) { -- rt_mutex_deadlock_account_unlock(current); -- return false; -+ WAKE_Q(wake_q); -+ WAKE_Q(wake_sleeper_q); -+ bool deboost; -+ -+ raw_spin_lock_irq(&lock->wait_lock); -+ deboost = __rt_mutex_futex_unlock(lock, &wake_q, &wake_sleeper_q); -+ raw_spin_unlock_irq(&lock->wait_lock); -+ -+ if (deboost) { -+ wake_up_q(&wake_q); -+ wake_up_q_sleeper(&wake_sleeper_q); -+ rt_mutex_adjust_prio(current); - } -- return rt_mutex_slowunlock(lock, wqh); - } - - /** -@@ -1607,13 +2213,12 @@ - void __rt_mutex_init(struct rt_mutex *lock, const char *name) - { - lock->owner = NULL; -- raw_spin_lock_init(&lock->wait_lock); - lock->waiters = RB_ROOT; - lock->waiters_leftmost = NULL; - - debug_rt_mutex_init(lock, name); - } --EXPORT_SYMBOL_GPL(__rt_mutex_init); -+EXPORT_SYMBOL(__rt_mutex_init); - - /** - * rt_mutex_init_proxy_locked - initialize and lock a rt_mutex on behalf of a -@@ -1628,10 +2233,9 @@ - void rt_mutex_init_proxy_locked(struct rt_mutex *lock, - struct task_struct *proxy_owner) - { -- __rt_mutex_init(lock, NULL); -+ rt_mutex_init(lock); - debug_rt_mutex_proxy_lock(lock, proxy_owner); - rt_mutex_set_owner(lock, proxy_owner); -- rt_mutex_deadlock_account_lock(lock, proxy_owner); - } - - /** -@@ -1647,34 +2251,45 @@ - { - debug_rt_mutex_proxy_unlock(lock); - rt_mutex_set_owner(lock, NULL); -- rt_mutex_deadlock_account_unlock(proxy_owner); - } - --/** -- * rt_mutex_start_proxy_lock() - Start lock acquisition for another task -- * @lock: the rt_mutex to take -- * @waiter: the pre-initialized rt_mutex_waiter -- * @task: the task to prepare -- * -- * Returns: -- * 0 - task blocked on lock -- * 1 - acquired the lock for task, caller should wake it up -- * <0 - error -- * -- * Special API call for FUTEX_REQUEUE_PI support. -- */ --int rt_mutex_start_proxy_lock(struct rt_mutex *lock, -+int __rt_mutex_start_proxy_lock(struct rt_mutex *lock, - struct rt_mutex_waiter *waiter, - struct task_struct *task) - { - int ret; - -- raw_spin_lock_irq(&lock->wait_lock); -+ if (try_to_take_rt_mutex(lock, task, NULL)) -+ return 1; - -- if (try_to_take_rt_mutex(lock, task, NULL)) { -+#ifdef CONFIG_PREEMPT_RT_FULL -+ /* -+ * In PREEMPT_RT there's an added race. -+ * If the task, that we are about to requeue, times out, -+ * it can set the PI_WAKEUP_INPROGRESS. This tells the requeue -+ * to skip this task. But right after the task sets -+ * its pi_blocked_on to PI_WAKEUP_INPROGRESS it can then -+ * block on the spin_lock(&hb->lock), which in RT is an rtmutex. -+ * This will replace the PI_WAKEUP_INPROGRESS with the actual -+ * lock that it blocks on. We *must not* place this task -+ * on this proxy lock in that case. -+ * -+ * To prevent this race, we first take the task's pi_lock -+ * and check if it has updated its pi_blocked_on. If it has, -+ * we assume that it woke up and we return -EAGAIN. -+ * Otherwise, we set the task's pi_blocked_on to -+ * PI_REQUEUE_INPROGRESS, so that if the task is waking up -+ * it will know that we are in the process of requeuing it. -+ */ -+ raw_spin_lock(&task->pi_lock); -+ if (task->pi_blocked_on) { -+ raw_spin_unlock(&task->pi_lock); - raw_spin_unlock_irq(&lock->wait_lock); -- return 1; -+ return -EAGAIN; - } -+ task->pi_blocked_on = PI_REQUEUE_INPROGRESS; -+ raw_spin_unlock(&task->pi_lock); -+#endif - - /* We enforce deadlock detection for futexes */ - ret = task_blocks_on_rt_mutex(lock, waiter, task, -@@ -1690,17 +2305,41 @@ - ret = 0; - } - -- if (unlikely(ret)) -+ if (ret && rt_mutex_has_waiters(lock)) - remove_waiter(lock, waiter); - -- raw_spin_unlock_irq(&lock->wait_lock); -- - debug_rt_mutex_print_deadlock(waiter); - - return ret; - } - - /** -+ * rt_mutex_start_proxy_lock() - Start lock acquisition for another task -+ * @lock: the rt_mutex to take -+ * @waiter: the pre-initialized rt_mutex_waiter -+ * @task: the task to prepare -+ * -+ * Returns: -+ * 0 - task blocked on lock -+ * 1 - acquired the lock for task, caller should wake it up -+ * <0 - error -+ * -+ * Special API call for FUTEX_REQUEUE_PI support. -+ */ -+int rt_mutex_start_proxy_lock(struct rt_mutex *lock, -+ struct rt_mutex_waiter *waiter, -+ struct task_struct *task) -+{ -+ int ret; -+ -+ raw_spin_lock_irq(&lock->wait_lock); -+ ret = __rt_mutex_start_proxy_lock(lock, waiter, task); -+ raw_spin_unlock_irq(&lock->wait_lock); -+ -+ return ret; -+} -+ -+/** - * rt_mutex_next_owner - return the next owner of the lock - * - * @lock: the rt lock query -@@ -1721,21 +2360,23 @@ - } - - /** -- * rt_mutex_finish_proxy_lock() - Complete lock acquisition -+ * rt_mutex_wait_proxy_lock() - Wait for lock acquisition - * @lock: the rt_mutex we were woken on - * @to: the timeout, null if none. hrtimer should already have - * been started. - * @waiter: the pre-initialized rt_mutex_waiter - * -- * Complete the lock acquisition started our behalf by another thread. -+ * Wait for the the lock acquisition started on our behalf by -+ * rt_mutex_start_proxy_lock(). Upon failure, the caller must call -+ * rt_mutex_cleanup_proxy_lock(). - * - * Returns: - * 0 - success - * <0 - error, one of -EINTR, -ETIMEDOUT - * -- * Special API call for PI-futex requeue support -+ * Special API call for PI-futex support - */ --int rt_mutex_finish_proxy_lock(struct rt_mutex *lock, -+int rt_mutex_wait_proxy_lock(struct rt_mutex *lock, - struct hrtimer_sleeper *to, - struct rt_mutex_waiter *waiter) - { -@@ -1746,10 +2387,47 @@ - set_current_state(TASK_INTERRUPTIBLE); - - /* sleep on the mutex */ -- ret = __rt_mutex_slowlock(lock, TASK_INTERRUPTIBLE, to, waiter); -+ ret = __rt_mutex_slowlock(lock, TASK_INTERRUPTIBLE, to, waiter, NULL); - -- if (unlikely(ret)) -+ raw_spin_unlock_irq(&lock->wait_lock); -+ -+ return ret; -+} -+ -+/** -+ * rt_mutex_cleanup_proxy_lock() - Cleanup failed lock acquisition -+ * @lock: the rt_mutex we were woken on -+ * @waiter: the pre-initialized rt_mutex_waiter -+ * -+ * Attempt to clean up after a failed rt_mutex_wait_proxy_lock(). -+ * -+ * Unless we acquired the lock; we're still enqueued on the wait-list and can -+ * in fact still be granted ownership until we're removed. Therefore we can -+ * find we are in fact the owner and must disregard the -+ * rt_mutex_wait_proxy_lock() failure. -+ * -+ * Returns: -+ * true - did the cleanup, we done. -+ * false - we acquired the lock after rt_mutex_wait_proxy_lock() returned, -+ * caller should disregards its return value. -+ * -+ * Special API call for PI-futex support -+ */ -+bool rt_mutex_cleanup_proxy_lock(struct rt_mutex *lock, -+ struct rt_mutex_waiter *waiter) -+{ -+ bool cleanup = false; -+ -+ raw_spin_lock_irq(&lock->wait_lock); -+ /* -+ * Unless we're the owner; we're still enqueued on the wait_list. -+ * So check if we became owner, if not, take us off the wait_list. -+ */ -+ if (rt_mutex_owner(lock) != current) { - remove_waiter(lock, waiter); -+ fixup_rt_mutex_waiters(lock); -+ cleanup = true; -+ } - - /* - * try_to_take_rt_mutex() sets the waiter bit unconditionally. We might -@@ -1759,5 +2437,91 @@ - - raw_spin_unlock_irq(&lock->wait_lock); - -+ return cleanup; -+} -+ -+static inline int -+ww_mutex_deadlock_injection(struct ww_mutex *lock, struct ww_acquire_ctx *ctx) -+{ -+#ifdef CONFIG_DEBUG_WW_MUTEX_SLOWPATH -+ unsigned tmp; -+ -+ if (ctx->deadlock_inject_countdown-- == 0) { -+ tmp = ctx->deadlock_inject_interval; -+ if (tmp > UINT_MAX/4) -+ tmp = UINT_MAX; -+ else -+ tmp = tmp*2 + tmp + tmp/2; -+ -+ ctx->deadlock_inject_interval = tmp; -+ ctx->deadlock_inject_countdown = tmp; -+ ctx->contending_lock = lock; -+ -+ ww_mutex_unlock(lock); -+ -+ return -EDEADLK; -+ } -+#endif -+ -+ return 0; -+} -+ -+#ifdef CONFIG_PREEMPT_RT_FULL -+int __sched -+__ww_mutex_lock_interruptible(struct ww_mutex *lock, struct ww_acquire_ctx *ww_ctx) -+{ -+ int ret; -+ -+ might_sleep(); -+ -+ mutex_acquire_nest(&lock->base.dep_map, 0, 0, &ww_ctx->dep_map, _RET_IP_); -+ ret = rt_mutex_slowlock(&lock->base.lock, TASK_INTERRUPTIBLE, NULL, 0, ww_ctx); -+ if (ret) -+ mutex_release(&lock->base.dep_map, 1, _RET_IP_); -+ else if (!ret && ww_ctx->acquired > 1) -+ return ww_mutex_deadlock_injection(lock, ww_ctx); -+ - return ret; - } -+EXPORT_SYMBOL_GPL(__ww_mutex_lock_interruptible); -+ -+int __sched -+__ww_mutex_lock(struct ww_mutex *lock, struct ww_acquire_ctx *ww_ctx) -+{ -+ int ret; -+ -+ might_sleep(); -+ -+ mutex_acquire_nest(&lock->base.dep_map, 0, 0, &ww_ctx->dep_map, _RET_IP_); -+ ret = rt_mutex_slowlock(&lock->base.lock, TASK_UNINTERRUPTIBLE, NULL, 0, ww_ctx); -+ if (ret) -+ mutex_release(&lock->base.dep_map, 1, _RET_IP_); -+ else if (!ret && ww_ctx->acquired > 1) -+ return ww_mutex_deadlock_injection(lock, ww_ctx); -+ -+ return ret; -+} -+EXPORT_SYMBOL_GPL(__ww_mutex_lock); -+ -+void __sched ww_mutex_unlock(struct ww_mutex *lock) -+{ -+ int nest = !!lock->ctx; -+ -+ /* -+ * The unlocking fastpath is the 0->1 transition from 'locked' -+ * into 'unlocked' state: -+ */ -+ if (nest) { -+#ifdef CONFIG_DEBUG_MUTEXES -+ DEBUG_LOCKS_WARN_ON(!lock->ctx->acquired); -+#endif -+ if (lock->ctx->acquired > 0) -+ lock->ctx->acquired--; -+ lock->ctx = NULL; -+ } -+ -+ mutex_release(&lock->base.dep_map, nest, _RET_IP_); -+ rt_mutex_unlock(&lock->base.lock); -+} -+EXPORT_SYMBOL(ww_mutex_unlock); -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rtmutex.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rtmutex.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rtmutex.h 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rtmutex.h 2017-04-18 17:54:26.000000000 +0200 -@@ -11,8 +11,6 @@ - */ - - #define rt_mutex_deadlock_check(l) (0) --#define rt_mutex_deadlock_account_lock(m, t) do { } while (0) --#define rt_mutex_deadlock_account_unlock(l) do { } while (0) - #define debug_rt_mutex_init_waiter(w) do { } while (0) - #define debug_rt_mutex_free_waiter(w) do { } while (0) - #define debug_rt_mutex_lock(l) do { } while (0) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rtmutex_common.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rtmutex_common.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rtmutex_common.h 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rtmutex_common.h 2017-04-18 17:54:26.000000000 +0200 -@@ -27,6 +27,7 @@ - struct rb_node pi_tree_entry; - struct task_struct *task; - struct rt_mutex *lock; -+ bool savestate; - #ifdef CONFIG_DEBUG_RT_MUTEXES - unsigned long ip; - struct pid *deadlock_task_pid; -@@ -98,22 +99,45 @@ - /* - * PI-futex support (proxy locking functions, etc.): - */ -+#define PI_WAKEUP_INPROGRESS ((struct rt_mutex_waiter *) 1) -+#define PI_REQUEUE_INPROGRESS ((struct rt_mutex_waiter *) 2) -+ - extern struct task_struct *rt_mutex_next_owner(struct rt_mutex *lock); - extern void rt_mutex_init_proxy_locked(struct rt_mutex *lock, - struct task_struct *proxy_owner); - extern void rt_mutex_proxy_unlock(struct rt_mutex *lock, - struct task_struct *proxy_owner); -+extern void rt_mutex_init_waiter(struct rt_mutex_waiter *waiter, bool savetate); -+extern int __rt_mutex_start_proxy_lock(struct rt_mutex *lock, -+ struct rt_mutex_waiter *waiter, -+ struct task_struct *task); - extern int rt_mutex_start_proxy_lock(struct rt_mutex *lock, - struct rt_mutex_waiter *waiter, - struct task_struct *task); --extern int rt_mutex_finish_proxy_lock(struct rt_mutex *lock, -- struct hrtimer_sleeper *to, -- struct rt_mutex_waiter *waiter); --extern int rt_mutex_timed_futex_lock(struct rt_mutex *l, struct hrtimer_sleeper *to); --extern bool rt_mutex_futex_unlock(struct rt_mutex *lock, -- struct wake_q_head *wqh); -+extern int rt_mutex_wait_proxy_lock(struct rt_mutex *lock, -+ struct hrtimer_sleeper *to, -+ struct rt_mutex_waiter *waiter); -+extern bool rt_mutex_cleanup_proxy_lock(struct rt_mutex *lock, -+ struct rt_mutex_waiter *waiter); -+ -+extern int rt_mutex_futex_trylock(struct rt_mutex *l); -+ -+extern void rt_mutex_futex_unlock(struct rt_mutex *lock); -+extern bool __rt_mutex_futex_unlock(struct rt_mutex *lock, -+ struct wake_q_head *wqh, -+ struct wake_q_head *wq_sleeper); -+ - extern void rt_mutex_adjust_prio(struct task_struct *task); - -+/* RW semaphore special interface */ -+struct ww_acquire_ctx; -+ -+int __sched rt_mutex_slowlock_locked(struct rt_mutex *lock, int state, -+ struct hrtimer_sleeper *timeout, -+ enum rtmutex_chainwalk chwalk, -+ struct ww_acquire_ctx *ww_ctx, -+ struct rt_mutex_waiter *waiter); -+ - #ifdef CONFIG_DEBUG_RT_MUTEXES - # include "rtmutex-debug.h" - #else -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rwsem-rt.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rwsem-rt.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/rwsem-rt.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/rwsem-rt.c 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,268 @@ -+/* -+ */ -+#include -+#include -+#include -+ -+#include "rtmutex_common.h" -+ -+/* -+ * RT-specific reader/writer semaphores -+ * -+ * down_write() -+ * 1) Lock sem->rtmutex -+ * 2) Remove the reader BIAS to force readers into the slow path -+ * 3) Wait until all readers have left the critical region -+ * 4) Mark it write locked -+ * -+ * up_write() -+ * 1) Remove the write locked marker -+ * 2) Set the reader BIAS so readers can use the fast path again -+ * 3) Unlock sem->rtmutex to release blocked readers -+ * -+ * down_read() -+ * 1) Try fast path acquisition (reader BIAS is set) -+ * 2) Take sem->rtmutex.wait_lock which protects the writelocked flag -+ * 3) If !writelocked, acquire it for read -+ * 4) If writelocked, block on sem->rtmutex -+ * 5) unlock sem->rtmutex, goto 1) -+ * -+ * up_read() -+ * 1) Try fast path release (reader count != 1) -+ * 2) Wake the writer waiting in down_write()#3 -+ * -+ * down_read()#3 has the consequence, that rw semaphores on RT are not writer -+ * fair, but writers, which should be avoided in RT tasks (think mmap_sem), -+ * are subject to the rtmutex priority/DL inheritance mechanism. -+ * -+ * It's possible to make the rw semaphores writer fair by keeping a list of -+ * active readers. A blocked writer would force all newly incoming readers to -+ * block on the rtmutex, but the rtmutex would have to be proxy locked for one -+ * reader after the other. We can't use multi-reader inheritance because there -+ * is no way to support that with SCHED_DEADLINE. Implementing the one by one -+ * reader boosting/handover mechanism is a major surgery for a very dubious -+ * value. -+ * -+ * The risk of writer starvation is there, but the pathological use cases -+ * which trigger it are not necessarily the typical RT workloads. -+ */ -+ -+void __rwsem_init(struct rw_semaphore *sem, const char *name, -+ struct lock_class_key *key) -+{ -+#ifdef CONFIG_DEBUG_LOCK_ALLOC -+ /* -+ * Make sure we are not reinitializing a held semaphore: -+ */ -+ debug_check_no_locks_freed((void *)sem, sizeof(*sem)); -+ lockdep_init_map(&sem->dep_map, name, key, 0); -+#endif -+ atomic_set(&sem->readers, READER_BIAS); -+} -+EXPORT_SYMBOL(__rwsem_init); -+ -+int __down_read_trylock(struct rw_semaphore *sem) -+{ -+ int r, old; -+ -+ /* -+ * Increment reader count, if sem->readers < 0, i.e. READER_BIAS is -+ * set. -+ */ -+ for (r = atomic_read(&sem->readers); r < 0;) { -+ old = atomic_cmpxchg(&sem->readers, r, r + 1); -+ if (likely(old == r)) -+ return 1; -+ r = old; -+ } -+ return 0; -+} -+ -+void __sched __down_read(struct rw_semaphore *sem) -+{ -+ struct rt_mutex *m = &sem->rtmutex; -+ struct rt_mutex_waiter waiter; -+ -+ if (__down_read_trylock(sem)) -+ return; -+ -+ might_sleep(); -+ raw_spin_lock_irq(&m->wait_lock); -+ /* -+ * Allow readers as long as the writer has not completely -+ * acquired the semaphore for write. -+ */ -+ if (atomic_read(&sem->readers) != WRITER_BIAS) { -+ atomic_inc(&sem->readers); -+ raw_spin_unlock_irq(&m->wait_lock); -+ return; -+ } -+ -+ /* -+ * Call into the slow lock path with the rtmutex->wait_lock -+ * held, so this can't result in the following race: -+ * -+ * Reader1 Reader2 Writer -+ * down_read() -+ * down_write() -+ * rtmutex_lock(m) -+ * swait() -+ * down_read() -+ * unlock(m->wait_lock) -+ * up_read() -+ * swake() -+ * lock(m->wait_lock) -+ * sem->writelocked=true -+ * unlock(m->wait_lock) -+ * -+ * up_write() -+ * sem->writelocked=false -+ * rtmutex_unlock(m) -+ * down_read() -+ * down_write() -+ * rtmutex_lock(m) -+ * swait() -+ * rtmutex_lock(m) -+ * -+ * That would put Reader1 behind the writer waiting on -+ * Reader2 to call up_read() which might be unbound. -+ */ -+ rt_mutex_init_waiter(&waiter, false); -+ rt_mutex_slowlock_locked(m, TASK_UNINTERRUPTIBLE, NULL, -+ RT_MUTEX_MIN_CHAINWALK, NULL, -+ &waiter); -+ /* -+ * The slowlock() above is guaranteed to return with the rtmutex is -+ * now held, so there can't be a writer active. Increment the reader -+ * count and immediately drop the rtmutex again. -+ */ -+ atomic_inc(&sem->readers); -+ raw_spin_unlock_irq(&m->wait_lock); -+ rt_mutex_unlock(m); -+ -+ debug_rt_mutex_free_waiter(&waiter); -+} -+ -+void __up_read(struct rw_semaphore *sem) -+{ -+ struct rt_mutex *m = &sem->rtmutex; -+ struct task_struct *tsk; -+ -+ /* -+ * sem->readers can only hit 0 when a writer is waiting for the -+ * active readers to leave the critical region. -+ */ -+ if (!atomic_dec_and_test(&sem->readers)) -+ return; -+ -+ might_sleep(); -+ raw_spin_lock_irq(&m->wait_lock); -+ /* -+ * Wake the writer, i.e. the rtmutex owner. It might release the -+ * rtmutex concurrently in the fast path (due to a signal), but to -+ * clean up the rwsem it needs to acquire m->wait_lock. The worst -+ * case which can happen is a spurious wakeup. -+ */ -+ tsk = rt_mutex_owner(m); -+ if (tsk) -+ wake_up_process(tsk); -+ -+ raw_spin_unlock_irq(&m->wait_lock); -+} -+ -+static void __up_write_unlock(struct rw_semaphore *sem, int bias, -+ unsigned long flags) -+{ -+ struct rt_mutex *m = &sem->rtmutex; -+ -+ atomic_add(READER_BIAS - bias, &sem->readers); -+ raw_spin_unlock_irqrestore(&m->wait_lock, flags); -+ rt_mutex_unlock(m); -+} -+ -+static int __sched __down_write_common(struct rw_semaphore *sem, int state) -+{ -+ struct rt_mutex *m = &sem->rtmutex; -+ unsigned long flags; -+ -+ /* Take the rtmutex as a first step */ -+ if (rt_mutex_lock_state(m, state)) -+ return -EINTR; -+ -+ /* Force readers into slow path */ -+ atomic_sub(READER_BIAS, &sem->readers); -+ might_sleep(); -+ -+ set_current_state(state); -+ for (;;) { -+ raw_spin_lock_irqsave(&m->wait_lock, flags); -+ /* Have all readers left the critical region? */ -+ if (!atomic_read(&sem->readers)) { -+ atomic_set(&sem->readers, WRITER_BIAS); -+ __set_current_state(TASK_RUNNING); -+ raw_spin_unlock_irqrestore(&m->wait_lock, flags); -+ return 0; -+ } -+ -+ if (signal_pending_state(state, current)) { -+ __set_current_state(TASK_RUNNING); -+ __up_write_unlock(sem, 0, flags); -+ return -EINTR; -+ } -+ raw_spin_unlock_irqrestore(&m->wait_lock, flags); -+ -+ if (atomic_read(&sem->readers) != 0) { -+ schedule(); -+ set_current_state(state); -+ } -+ } -+} -+ -+void __sched __down_write(struct rw_semaphore *sem) -+{ -+ __down_write_common(sem, TASK_UNINTERRUPTIBLE); -+} -+ -+int __sched __down_write_killable(struct rw_semaphore *sem) -+{ -+ return __down_write_common(sem, TASK_KILLABLE); -+} -+ -+int __down_write_trylock(struct rw_semaphore *sem) -+{ -+ struct rt_mutex *m = &sem->rtmutex; -+ unsigned long flags; -+ -+ if (!rt_mutex_trylock(m)) -+ return 0; -+ -+ atomic_sub(READER_BIAS, &sem->readers); -+ -+ raw_spin_lock_irqsave(&m->wait_lock, flags); -+ if (!atomic_read(&sem->readers)) { -+ atomic_set(&sem->readers, WRITER_BIAS); -+ raw_spin_unlock_irqrestore(&m->wait_lock, flags); -+ return 1; -+ } -+ __up_write_unlock(sem, 0, flags); -+ return 0; -+} -+ -+void __up_write(struct rw_semaphore *sem) -+{ -+ struct rt_mutex *m = &sem->rtmutex; -+ unsigned long flags; -+ -+ raw_spin_lock_irqsave(&m->wait_lock, flags); -+ __up_write_unlock(sem, WRITER_BIAS, flags); -+} -+ -+void __downgrade_write(struct rw_semaphore *sem) -+{ -+ struct rt_mutex *m = &sem->rtmutex; -+ unsigned long flags; -+ -+ raw_spin_lock_irqsave(&m->wait_lock, flags); -+ /* Release it and account current as reader */ -+ __up_write_unlock(sem, WRITER_BIAS - 1, flags); -+} -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/spinlock.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/spinlock.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/spinlock.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/spinlock.c 2017-04-18 17:54:26.000000000 +0200 -@@ -124,8 +124,11 @@ - * __[spin|read|write]_lock_bh() - */ - BUILD_LOCK_OPS(spin, raw_spinlock); -+ -+#ifndef CONFIG_PREEMPT_RT_FULL - BUILD_LOCK_OPS(read, rwlock); - BUILD_LOCK_OPS(write, rwlock); -+#endif - - #endif - -@@ -209,6 +212,8 @@ - EXPORT_SYMBOL(_raw_spin_unlock_bh); - #endif - -+#ifndef CONFIG_PREEMPT_RT_FULL -+ - #ifndef CONFIG_INLINE_READ_TRYLOCK - int __lockfunc _raw_read_trylock(rwlock_t *lock) - { -@@ -353,6 +358,8 @@ - EXPORT_SYMBOL(_raw_write_unlock_bh); - #endif - -+#endif /* !PREEMPT_RT_FULL */ -+ - #ifdef CONFIG_DEBUG_LOCK_ALLOC - - void __lockfunc _raw_spin_lock_nested(raw_spinlock_t *lock, int subclass) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/spinlock_debug.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/spinlock_debug.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/locking/spinlock_debug.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/locking/spinlock_debug.c 2017-04-18 17:54:26.000000000 +0200 -@@ -31,6 +31,7 @@ - - EXPORT_SYMBOL(__raw_spin_lock_init); - -+#ifndef CONFIG_PREEMPT_RT_FULL - void __rwlock_init(rwlock_t *lock, const char *name, - struct lock_class_key *key) - { -@@ -48,6 +49,7 @@ - } - - EXPORT_SYMBOL(__rwlock_init); -+#endif - - static void spin_dump(raw_spinlock_t *lock, const char *msg) - { -@@ -159,6 +161,7 @@ - arch_spin_unlock(&lock->raw_lock); - } - -+#ifndef CONFIG_PREEMPT_RT_FULL - static void rwlock_bug(rwlock_t *lock, const char *msg) - { - if (!debug_locks_off()) -@@ -300,3 +303,5 @@ - debug_write_unlock(lock); - arch_write_unlock(&lock->raw_lock); - } -+ -+#endif -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/module.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/module.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/module.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/module.c 2017-04-18 17:54:26.000000000 +0200 -@@ -660,16 +660,7 @@ - memcpy(per_cpu_ptr(mod->percpu, cpu), from, size); - } - --/** -- * is_module_percpu_address - test whether address is from module static percpu -- * @addr: address to test -- * -- * Test whether @addr belongs to module static percpu area. -- * -- * RETURNS: -- * %true if @addr is from module static percpu area -- */ --bool is_module_percpu_address(unsigned long addr) -+bool __is_module_percpu_address(unsigned long addr, unsigned long *can_addr) - { - struct module *mod; - unsigned int cpu; -@@ -683,9 +674,15 @@ - continue; - for_each_possible_cpu(cpu) { - void *start = per_cpu_ptr(mod->percpu, cpu); -+ void *va = (void *)addr; - -- if ((void *)addr >= start && -- (void *)addr < start + mod->percpu_size) { -+ if (va >= start && va < start + mod->percpu_size) { -+ if (can_addr) { -+ *can_addr = (unsigned long) (va - start); -+ *can_addr += (unsigned long) -+ per_cpu_ptr(mod->percpu, -+ get_boot_cpu_id()); -+ } - preempt_enable(); - return true; - } -@@ -696,6 +693,20 @@ - return false; - } - -+/** -+ * is_module_percpu_address - test whether address is from module static percpu -+ * @addr: address to test -+ * -+ * Test whether @addr belongs to module static percpu area. -+ * -+ * RETURNS: -+ * %true if @addr is from module static percpu area -+ */ -+bool is_module_percpu_address(unsigned long addr) -+{ -+ return __is_module_percpu_address(addr, NULL); -+} -+ - #else /* ... !CONFIG_SMP */ - - static inline void __percpu *mod_percpu(struct module *mod) -@@ -727,6 +738,11 @@ - return false; - } - -+bool __is_module_percpu_address(unsigned long addr, unsigned long *can_addr) -+{ -+ return false; -+} -+ - #endif /* CONFIG_SMP */ - - #define MODINFO_ATTR(field) \ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/panic.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/panic.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/panic.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/panic.c 2017-04-18 17:54:26.000000000 +0200 -@@ -482,9 +482,11 @@ - - static int init_oops_id(void) - { -+#ifndef CONFIG_PREEMPT_RT_FULL - if (!oops_id) - get_random_bytes(&oops_id, sizeof(oops_id)); - else -+#endif - oops_id++; - - return 0; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/power/hibernate.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/power/hibernate.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/power/hibernate.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/power/hibernate.c 2017-04-18 17:54:26.000000000 +0200 -@@ -286,6 +286,8 @@ - - local_irq_disable(); - -+ system_state = SYSTEM_SUSPEND; -+ - error = syscore_suspend(); - if (error) { - printk(KERN_ERR "PM: Some system devices failed to power down, " -@@ -317,6 +319,7 @@ - syscore_resume(); - - Enable_irqs: -+ system_state = SYSTEM_RUNNING; - local_irq_enable(); - - Enable_cpus: -@@ -446,6 +449,7 @@ - goto Enable_cpus; - - local_irq_disable(); -+ system_state = SYSTEM_SUSPEND; - - error = syscore_suspend(); - if (error) -@@ -479,6 +483,7 @@ - syscore_resume(); - - Enable_irqs: -+ system_state = SYSTEM_RUNNING; - local_irq_enable(); - - Enable_cpus: -@@ -564,6 +569,7 @@ - goto Enable_cpus; - - local_irq_disable(); -+ system_state = SYSTEM_SUSPEND; - syscore_suspend(); - if (pm_wakeup_pending()) { - error = -EAGAIN; -@@ -576,6 +582,7 @@ - - Power_up: - syscore_resume(); -+ system_state = SYSTEM_RUNNING; - local_irq_enable(); - - Enable_cpus: -@@ -676,6 +683,10 @@ - return error; - } - -+#ifndef CONFIG_SUSPEND -+bool pm_in_action; -+#endif -+ - /** - * hibernate - Carry out system hibernation, including saving the image. - */ -@@ -689,6 +700,8 @@ - return -EPERM; - } - -+ pm_in_action = true; -+ - lock_system_sleep(); - /* The snapshot device should not be opened while we're running */ - if (!atomic_add_unless(&snapshot_device_available, -1, 0)) { -@@ -766,6 +779,7 @@ - atomic_inc(&snapshot_device_available); - Unlock: - unlock_system_sleep(); -+ pm_in_action = false; - return error; - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/power/suspend.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/power/suspend.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/power/suspend.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/power/suspend.c 2017-04-18 17:54:26.000000000 +0200 -@@ -369,6 +369,8 @@ - arch_suspend_disable_irqs(); - BUG_ON(!irqs_disabled()); - -+ system_state = SYSTEM_SUSPEND; -+ - error = syscore_suspend(); - if (!error) { - *wakeup = pm_wakeup_pending(); -@@ -385,6 +387,8 @@ - syscore_resume(); - } - -+ system_state = SYSTEM_RUNNING; -+ - arch_suspend_enable_irqs(); - BUG_ON(irqs_disabled()); - -@@ -527,6 +531,8 @@ - return error; - } - -+bool pm_in_action; -+ - /** - * pm_suspend - Externally visible function for suspending the system. - * @state: System sleep state to enter. -@@ -541,6 +547,8 @@ - if (state <= PM_SUSPEND_ON || state >= PM_SUSPEND_MAX) - return -EINVAL; - -+ pm_in_action = true; -+ - error = enter_state(state); - if (error) { - suspend_stats.fail++; -@@ -548,6 +556,7 @@ - } else { - suspend_stats.success++; - } -+ pm_in_action = false; - return error; - } - EXPORT_SYMBOL(pm_suspend); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/printk/printk.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/printk/printk.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/printk/printk.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/printk/printk.c 2017-04-18 17:54:26.000000000 +0200 -@@ -351,6 +351,65 @@ - */ - DEFINE_RAW_SPINLOCK(logbuf_lock); - -+#ifdef CONFIG_EARLY_PRINTK -+struct console *early_console; -+ -+static void early_vprintk(const char *fmt, va_list ap) -+{ -+ if (early_console) { -+ char buf[512]; -+ int n = vscnprintf(buf, sizeof(buf), fmt, ap); -+ -+ early_console->write(early_console, buf, n); -+ } -+} -+ -+asmlinkage void early_printk(const char *fmt, ...) -+{ -+ va_list ap; -+ -+ va_start(ap, fmt); -+ early_vprintk(fmt, ap); -+ va_end(ap); -+} -+ -+/* -+ * This is independent of any log levels - a global -+ * kill switch that turns off all of printk. -+ * -+ * Used by the NMI watchdog if early-printk is enabled. -+ */ -+static bool __read_mostly printk_killswitch; -+ -+static int __init force_early_printk_setup(char *str) -+{ -+ printk_killswitch = true; -+ return 0; -+} -+early_param("force_early_printk", force_early_printk_setup); -+ -+void printk_kill(void) -+{ -+ printk_killswitch = true; -+} -+ -+#ifdef CONFIG_PRINTK -+static int forced_early_printk(const char *fmt, va_list ap) -+{ -+ if (!printk_killswitch) -+ return 0; -+ early_vprintk(fmt, ap); -+ return 1; -+} -+#endif -+ -+#else -+static inline int forced_early_printk(const char *fmt, va_list ap) -+{ -+ return 0; -+} -+#endif -+ - #ifdef CONFIG_PRINTK - DECLARE_WAIT_QUEUE_HEAD(log_wait); - /* the next printk record to read by syslog(READ) or /proc/kmsg */ -@@ -1337,6 +1396,7 @@ - { - char *text; - int len = 0; -+ int attempts = 0; - - text = kmalloc(LOG_LINE_MAX + PREFIX_MAX, GFP_KERNEL); - if (!text) -@@ -1348,6 +1408,14 @@ - u64 seq; - u32 idx; - enum log_flags prev; -+ int num_msg; -+try_again: -+ attempts++; -+ if (attempts > 10) { -+ len = -EBUSY; -+ goto out; -+ } -+ num_msg = 0; - - /* - * Find first record that fits, including all following records, -@@ -1363,6 +1431,14 @@ - prev = msg->flags; - idx = log_next(idx); - seq++; -+ num_msg++; -+ if (num_msg > 5) { -+ num_msg = 0; -+ raw_spin_unlock_irq(&logbuf_lock); -+ raw_spin_lock_irq(&logbuf_lock); -+ if (clear_seq < log_first_seq) -+ goto try_again; -+ } - } - - /* move first record forward until length fits into the buffer */ -@@ -1376,6 +1452,14 @@ - prev = msg->flags; - idx = log_next(idx); - seq++; -+ num_msg++; -+ if (num_msg > 5) { -+ num_msg = 0; -+ raw_spin_unlock_irq(&logbuf_lock); -+ raw_spin_lock_irq(&logbuf_lock); -+ if (clear_seq < log_first_seq) -+ goto try_again; -+ } - } - - /* last message fitting into this dump */ -@@ -1416,6 +1500,7 @@ - clear_seq = log_next_seq; - clear_idx = log_next_idx; - } -+out: - raw_spin_unlock_irq(&logbuf_lock); - - kfree(text); -@@ -1569,6 +1654,12 @@ - if (!console_drivers) - return; - -+ if (IS_ENABLED(CONFIG_PREEMPT_RT_BASE)) { -+ if (in_irq() || in_nmi()) -+ return; -+ } -+ -+ migrate_disable(); - for_each_console(con) { - if (exclusive_console && con != exclusive_console) - continue; -@@ -1584,6 +1675,7 @@ - else - con->write(con, text, len); - } -+ migrate_enable(); - } - - /* -@@ -1781,6 +1873,13 @@ - /* cpu currently holding logbuf_lock in this function */ - static unsigned int logbuf_cpu = UINT_MAX; - -+ /* -+ * Fall back to early_printk if a debugging subsystem has -+ * killed printk output -+ */ -+ if (unlikely(forced_early_printk(fmt, args))) -+ return 1; -+ - if (level == LOGLEVEL_SCHED) { - level = LOGLEVEL_DEFAULT; - in_sched = true; -@@ -1885,13 +1984,23 @@ - - /* If called from the scheduler, we can not call up(). */ - if (!in_sched) { -+ int may_trylock = 1; -+ - lockdep_off(); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ /* -+ * we can't take a sleeping lock with IRQs or preeption disabled -+ * so we can't print in these contexts -+ */ -+ if (!(preempt_count() == 0 && !irqs_disabled())) -+ may_trylock = 0; -+#endif - /* - * Try to acquire and then immediately release the console - * semaphore. The release will print out buffers and wake up - * /dev/kmsg and syslog() users. - */ -- if (console_trylock()) -+ if (may_trylock && console_trylock()) - console_unlock(); - lockdep_on(); - } -@@ -2014,26 +2123,6 @@ - - #endif /* CONFIG_PRINTK */ - --#ifdef CONFIG_EARLY_PRINTK --struct console *early_console; -- --asmlinkage __visible void early_printk(const char *fmt, ...) --{ -- va_list ap; -- char buf[512]; -- int n; -- -- if (!early_console) -- return; -- -- va_start(ap, fmt); -- n = vscnprintf(buf, sizeof(buf), fmt, ap); -- va_end(ap); -- -- early_console->write(early_console, buf, n); --} --#endif -- - static int __add_preferred_console(char *name, int idx, char *options, - char *brl_options) - { -@@ -2303,11 +2392,16 @@ - goto out; - - len = cont_print_text(text, size); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ raw_spin_unlock_irqrestore(&logbuf_lock, flags); -+ call_console_drivers(cont.level, NULL, 0, text, len); -+#else - raw_spin_unlock(&logbuf_lock); - stop_critical_timings(); - call_console_drivers(cont.level, NULL, 0, text, len); - start_critical_timings(); - local_irq_restore(flags); -+#endif - return; - out: - raw_spin_unlock_irqrestore(&logbuf_lock, flags); -@@ -2431,13 +2525,17 @@ - console_idx = log_next(console_idx); - console_seq++; - console_prev = msg->flags; -+#ifdef CONFIG_PREEMPT_RT_FULL -+ raw_spin_unlock_irqrestore(&logbuf_lock, flags); -+ call_console_drivers(level, ext_text, ext_len, text, len); -+#else - raw_spin_unlock(&logbuf_lock); - - stop_critical_timings(); /* don't trace print latency */ - call_console_drivers(level, ext_text, ext_len, text, len); - start_critical_timings(); - local_irq_restore(flags); -- -+#endif - if (do_cond_resched) - cond_resched(); - } -@@ -2489,6 +2587,11 @@ - { - struct console *c; - -+ if (IS_ENABLED(CONFIG_PREEMPT_RT_BASE)) { -+ if (in_irq() || in_nmi()) -+ return; -+ } -+ - /* - * console_unblank can no longer be called in interrupt context unless - * oops_in_progress is set to 1.. -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/ptrace.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/ptrace.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/ptrace.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/ptrace.c 2017-04-18 17:54:26.000000000 +0200 -@@ -166,7 +166,14 @@ - - spin_lock_irq(&task->sighand->siglock); - if (task_is_traced(task) && !__fatal_signal_pending(task)) { -- task->state = __TASK_TRACED; -+ unsigned long flags; -+ -+ raw_spin_lock_irqsave(&task->pi_lock, flags); -+ if (task->state & __TASK_TRACED) -+ task->state = __TASK_TRACED; -+ else -+ task->saved_state = __TASK_TRACED; -+ raw_spin_unlock_irqrestore(&task->pi_lock, flags); - ret = true; - } - spin_unlock_irq(&task->sighand->siglock); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/rcu/rcutorture.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/rcu/rcutorture.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/rcu/rcutorture.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/rcu/rcutorture.c 2017-04-18 17:54:26.000000000 +0200 -@@ -404,6 +404,7 @@ - .name = "rcu" - }; - -+#ifndef CONFIG_PREEMPT_RT_FULL - /* - * Definitions for rcu_bh torture testing. - */ -@@ -443,6 +444,12 @@ - .name = "rcu_bh" - }; - -+#else -+static struct rcu_torture_ops rcu_bh_ops = { -+ .ttype = INVALID_RCU_FLAVOR, -+}; -+#endif -+ - /* - * Don't even think about trying any of these in real life!!! - * The names includes "busted", and they really means it! -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/rcu/tree.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/rcu/tree.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/rcu/tree.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/rcu/tree.c 2017-04-18 17:54:26.000000000 +0200 -@@ -55,6 +55,11 @@ - #include - #include - #include -+#include -+#include -+#include -+#include -+#include "../time/tick-internal.h" - - #include "tree.h" - #include "rcu.h" -@@ -260,6 +265,19 @@ - this_cpu_ptr(&rcu_sched_data), true); - } - -+#ifdef CONFIG_PREEMPT_RT_FULL -+static void rcu_preempt_qs(void); -+ -+void rcu_bh_qs(void) -+{ -+ unsigned long flags; -+ -+ /* Callers to this function, rcu_preempt_qs(), must disable irqs. */ -+ local_irq_save(flags); -+ rcu_preempt_qs(); -+ local_irq_restore(flags); -+} -+#else - void rcu_bh_qs(void) - { - if (__this_cpu_read(rcu_bh_data.cpu_no_qs.s)) { -@@ -269,6 +287,7 @@ - __this_cpu_write(rcu_bh_data.cpu_no_qs.b.norm, false); - } - } -+#endif - - static DEFINE_PER_CPU(int, rcu_sched_qs_mask); - -@@ -449,11 +468,13 @@ - /* - * Return the number of RCU BH batches started thus far for debug & stats. - */ -+#ifndef CONFIG_PREEMPT_RT_FULL - unsigned long rcu_batches_started_bh(void) - { - return rcu_bh_state.gpnum; - } - EXPORT_SYMBOL_GPL(rcu_batches_started_bh); -+#endif - - /* - * Return the number of RCU batches completed thus far for debug & stats. -@@ -473,6 +494,7 @@ - } - EXPORT_SYMBOL_GPL(rcu_batches_completed_sched); - -+#ifndef CONFIG_PREEMPT_RT_FULL - /* - * Return the number of RCU BH batches completed thus far for debug & stats. - */ -@@ -481,6 +503,7 @@ - return rcu_bh_state.completed; - } - EXPORT_SYMBOL_GPL(rcu_batches_completed_bh); -+#endif - - /* - * Return the number of RCU expedited batches completed thus far for -@@ -504,6 +527,7 @@ - } - EXPORT_SYMBOL_GPL(rcu_exp_batches_completed_sched); - -+#ifndef CONFIG_PREEMPT_RT_FULL - /* - * Force a quiescent state. - */ -@@ -522,6 +546,13 @@ - } - EXPORT_SYMBOL_GPL(rcu_bh_force_quiescent_state); - -+#else -+void rcu_force_quiescent_state(void) -+{ -+} -+EXPORT_SYMBOL_GPL(rcu_force_quiescent_state); -+#endif -+ - /* - * Force a quiescent state for RCU-sched. - */ -@@ -572,9 +603,11 @@ - case RCU_FLAVOR: - rsp = rcu_state_p; - break; -+#ifndef CONFIG_PREEMPT_RT_FULL - case RCU_BH_FLAVOR: - rsp = &rcu_bh_state; - break; -+#endif - case RCU_SCHED_FLAVOR: - rsp = &rcu_sched_state; - break; -@@ -3016,18 +3049,17 @@ - /* - * Do RCU core processing for the current CPU. - */ --static __latent_entropy void rcu_process_callbacks(struct softirq_action *unused) -+static __latent_entropy void rcu_process_callbacks(void) - { - struct rcu_state *rsp; - - if (cpu_is_offline(smp_processor_id())) - return; -- trace_rcu_utilization(TPS("Start RCU core")); - for_each_rcu_flavor(rsp) - __rcu_process_callbacks(rsp); -- trace_rcu_utilization(TPS("End RCU core")); - } - -+static DEFINE_PER_CPU(struct task_struct *, rcu_cpu_kthread_task); - /* - * Schedule RCU callback invocation. If the specified type of RCU - * does not support RCU priority boosting, just do a direct call, -@@ -3039,18 +3071,105 @@ - { - if (unlikely(!READ_ONCE(rcu_scheduler_fully_active))) - return; -- if (likely(!rsp->boost)) { -- rcu_do_batch(rsp, rdp); -+ rcu_do_batch(rsp, rdp); -+} -+ -+static void rcu_wake_cond(struct task_struct *t, int status) -+{ -+ /* -+ * If the thread is yielding, only wake it when this -+ * is invoked from idle -+ */ -+ if (t && (status != RCU_KTHREAD_YIELDING || is_idle_task(current))) -+ wake_up_process(t); -+} -+ -+/* -+ * Wake up this CPU's rcuc kthread to do RCU core processing. -+ */ -+static void invoke_rcu_core(void) -+{ -+ unsigned long flags; -+ struct task_struct *t; -+ -+ if (!cpu_online(smp_processor_id())) - return; -+ local_irq_save(flags); -+ __this_cpu_write(rcu_cpu_has_work, 1); -+ t = __this_cpu_read(rcu_cpu_kthread_task); -+ if (t != NULL && current != t) -+ rcu_wake_cond(t, __this_cpu_read(rcu_cpu_kthread_status)); -+ local_irq_restore(flags); -+} -+ -+static void rcu_cpu_kthread_park(unsigned int cpu) -+{ -+ per_cpu(rcu_cpu_kthread_status, cpu) = RCU_KTHREAD_OFFCPU; -+} -+ -+static int rcu_cpu_kthread_should_run(unsigned int cpu) -+{ -+ return __this_cpu_read(rcu_cpu_has_work); -+} -+ -+/* -+ * Per-CPU kernel thread that invokes RCU callbacks. This replaces the -+ * RCU softirq used in flavors and configurations of RCU that do not -+ * support RCU priority boosting. -+ */ -+static void rcu_cpu_kthread(unsigned int cpu) -+{ -+ unsigned int *statusp = this_cpu_ptr(&rcu_cpu_kthread_status); -+ char work, *workp = this_cpu_ptr(&rcu_cpu_has_work); -+ int spincnt; -+ -+ for (spincnt = 0; spincnt < 10; spincnt++) { -+ trace_rcu_utilization(TPS("Start CPU kthread@rcu_wait")); -+ local_bh_disable(); -+ *statusp = RCU_KTHREAD_RUNNING; -+ this_cpu_inc(rcu_cpu_kthread_loops); -+ local_irq_disable(); -+ work = *workp; -+ *workp = 0; -+ local_irq_enable(); -+ if (work) -+ rcu_process_callbacks(); -+ local_bh_enable(); -+ if (*workp == 0) { -+ trace_rcu_utilization(TPS("End CPU kthread@rcu_wait")); -+ *statusp = RCU_KTHREAD_WAITING; -+ return; -+ } - } -- invoke_rcu_callbacks_kthread(); -+ *statusp = RCU_KTHREAD_YIELDING; -+ trace_rcu_utilization(TPS("Start CPU kthread@rcu_yield")); -+ schedule_timeout_interruptible(2); -+ trace_rcu_utilization(TPS("End CPU kthread@rcu_yield")); -+ *statusp = RCU_KTHREAD_WAITING; - } - --static void invoke_rcu_core(void) -+static struct smp_hotplug_thread rcu_cpu_thread_spec = { -+ .store = &rcu_cpu_kthread_task, -+ .thread_should_run = rcu_cpu_kthread_should_run, -+ .thread_fn = rcu_cpu_kthread, -+ .thread_comm = "rcuc/%u", -+ .setup = rcu_cpu_kthread_setup, -+ .park = rcu_cpu_kthread_park, -+}; -+ -+/* -+ * Spawn per-CPU RCU core processing kthreads. -+ */ -+static int __init rcu_spawn_core_kthreads(void) - { -- if (cpu_online(smp_processor_id())) -- raise_softirq(RCU_SOFTIRQ); -+ int cpu; -+ -+ for_each_possible_cpu(cpu) -+ per_cpu(rcu_cpu_has_work, cpu) = 0; -+ BUG_ON(smpboot_register_percpu_thread(&rcu_cpu_thread_spec)); -+ return 0; - } -+early_initcall(rcu_spawn_core_kthreads); - - /* - * Handle any core-RCU processing required by a call_rcu() invocation. -@@ -3195,6 +3314,7 @@ - } - EXPORT_SYMBOL_GPL(call_rcu_sched); - -+#ifndef CONFIG_PREEMPT_RT_FULL - /* - * Queue an RCU callback for invocation after a quicker grace period. - */ -@@ -3203,6 +3323,7 @@ - __call_rcu(head, func, &rcu_bh_state, -1, 0); - } - EXPORT_SYMBOL_GPL(call_rcu_bh); -+#endif - - /* - * Queue an RCU callback for lazy invocation after a grace period. -@@ -3294,6 +3415,7 @@ - } - EXPORT_SYMBOL_GPL(synchronize_sched); - -+#ifndef CONFIG_PREEMPT_RT_FULL - /** - * synchronize_rcu_bh - wait until an rcu_bh grace period has elapsed. - * -@@ -3320,6 +3442,7 @@ - wait_rcu_gp(call_rcu_bh); - } - EXPORT_SYMBOL_GPL(synchronize_rcu_bh); -+#endif - - /** - * get_state_synchronize_rcu - Snapshot current RCU state -@@ -3698,6 +3821,7 @@ - mutex_unlock(&rsp->barrier_mutex); - } - -+#ifndef CONFIG_PREEMPT_RT_FULL - /** - * rcu_barrier_bh - Wait until all in-flight call_rcu_bh() callbacks complete. - */ -@@ -3706,6 +3830,7 @@ - _rcu_barrier(&rcu_bh_state); - } - EXPORT_SYMBOL_GPL(rcu_barrier_bh); -+#endif - - /** - * rcu_barrier_sched - Wait for in-flight call_rcu_sched() callbacks. -@@ -4227,12 +4352,13 @@ - - rcu_bootup_announce(); - rcu_init_geometry(); -+#ifndef CONFIG_PREEMPT_RT_FULL - rcu_init_one(&rcu_bh_state); -+#endif - rcu_init_one(&rcu_sched_state); - if (dump_tree) - rcu_dump_rcu_node_tree(&rcu_sched_state); - __rcu_init_preempt(); -- open_softirq(RCU_SOFTIRQ, rcu_process_callbacks); - - /* - * We don't need protection against CPU-hotplug here because -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/rcu/tree.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/rcu/tree.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/rcu/tree.h 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/rcu/tree.h 2017-04-18 17:54:26.000000000 +0200 -@@ -588,18 +588,18 @@ - */ - extern struct rcu_state rcu_sched_state; - -+#ifndef CONFIG_PREEMPT_RT_FULL - extern struct rcu_state rcu_bh_state; -+#endif - - #ifdef CONFIG_PREEMPT_RCU - extern struct rcu_state rcu_preempt_state; - #endif /* #ifdef CONFIG_PREEMPT_RCU */ - --#ifdef CONFIG_RCU_BOOST - DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_status); - DECLARE_PER_CPU(int, rcu_cpu_kthread_cpu); - DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_loops); - DECLARE_PER_CPU(char, rcu_cpu_has_work); --#endif /* #ifdef CONFIG_RCU_BOOST */ - - #ifndef RCU_TREE_NONCORE - -@@ -619,10 +619,9 @@ - static void __init __rcu_init_preempt(void); - static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags); - static void rcu_preempt_boost_start_gp(struct rcu_node *rnp); --static void invoke_rcu_callbacks_kthread(void); - static bool rcu_is_callbacks_kthread(void); -+static void rcu_cpu_kthread_setup(unsigned int cpu); - #ifdef CONFIG_RCU_BOOST --static void rcu_preempt_do_callbacks(void); - static int rcu_spawn_one_boost_kthread(struct rcu_state *rsp, - struct rcu_node *rnp); - #endif /* #ifdef CONFIG_RCU_BOOST */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/rcu/tree_plugin.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/rcu/tree_plugin.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/rcu/tree_plugin.h 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/rcu/tree_plugin.h 2017-04-18 17:54:26.000000000 +0200 -@@ -24,25 +24,10 @@ - * Paul E. McKenney - */ - --#include --#include --#include --#include --#include "../time/tick-internal.h" -- - #ifdef CONFIG_RCU_BOOST - - #include "../locking/rtmutex_common.h" - --/* -- * Control variables for per-CPU and per-rcu_node kthreads. These -- * handle all flavors of RCU. -- */ --static DEFINE_PER_CPU(struct task_struct *, rcu_cpu_kthread_task); --DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_status); --DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_loops); --DEFINE_PER_CPU(char, rcu_cpu_has_work); -- - #else /* #ifdef CONFIG_RCU_BOOST */ - - /* -@@ -55,6 +40,14 @@ - - #endif /* #else #ifdef CONFIG_RCU_BOOST */ - -+/* -+ * Control variables for per-CPU and per-rcu_node kthreads. These -+ * handle all flavors of RCU. -+ */ -+DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_status); -+DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_loops); -+DEFINE_PER_CPU(char, rcu_cpu_has_work); -+ - #ifdef CONFIG_RCU_NOCB_CPU - static cpumask_var_t rcu_nocb_mask; /* CPUs to have callbacks offloaded. */ - static bool have_rcu_nocb_mask; /* Was rcu_nocb_mask allocated? */ -@@ -426,7 +419,7 @@ - } - - /* Hardware IRQ handlers cannot block, complain if they get here. */ -- if (in_irq() || in_serving_softirq()) { -+ if (preempt_count() & (HARDIRQ_MASK | SOFTIRQ_OFFSET)) { - lockdep_rcu_suspicious(__FILE__, __LINE__, - "rcu_read_unlock() from irq or softirq with blocking in critical section!!!\n"); - pr_alert("->rcu_read_unlock_special: %#x (b: %d, enq: %d nq: %d)\n", -@@ -632,15 +625,6 @@ - t->rcu_read_unlock_special.b.need_qs = true; - } - --#ifdef CONFIG_RCU_BOOST -- --static void rcu_preempt_do_callbacks(void) --{ -- rcu_do_batch(rcu_state_p, this_cpu_ptr(rcu_data_p)); --} -- --#endif /* #ifdef CONFIG_RCU_BOOST */ -- - /* - * Queue a preemptible-RCU callback for invocation after a grace period. - */ -@@ -829,6 +813,19 @@ - - #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ - -+/* -+ * If boosting, set rcuc kthreads to realtime priority. -+ */ -+static void rcu_cpu_kthread_setup(unsigned int cpu) -+{ -+#ifdef CONFIG_RCU_BOOST -+ struct sched_param sp; -+ -+ sp.sched_priority = kthread_prio; -+ sched_setscheduler_nocheck(current, SCHED_FIFO, &sp); -+#endif /* #ifdef CONFIG_RCU_BOOST */ -+} -+ - #ifdef CONFIG_RCU_BOOST - - #include "../locking/rtmutex_common.h" -@@ -860,16 +857,6 @@ - - #endif /* #else #ifdef CONFIG_RCU_TRACE */ - --static void rcu_wake_cond(struct task_struct *t, int status) --{ -- /* -- * If the thread is yielding, only wake it when this -- * is invoked from idle -- */ -- if (status != RCU_KTHREAD_YIELDING || is_idle_task(current)) -- wake_up_process(t); --} -- - /* - * Carry out RCU priority boosting on the task indicated by ->exp_tasks - * or ->boost_tasks, advancing the pointer to the next task in the -@@ -1013,23 +1000,6 @@ - } - - /* -- * Wake up the per-CPU kthread to invoke RCU callbacks. -- */ --static void invoke_rcu_callbacks_kthread(void) --{ -- unsigned long flags; -- -- local_irq_save(flags); -- __this_cpu_write(rcu_cpu_has_work, 1); -- if (__this_cpu_read(rcu_cpu_kthread_task) != NULL && -- current != __this_cpu_read(rcu_cpu_kthread_task)) { -- rcu_wake_cond(__this_cpu_read(rcu_cpu_kthread_task), -- __this_cpu_read(rcu_cpu_kthread_status)); -- } -- local_irq_restore(flags); --} -- --/* - * Is the current CPU running the RCU-callbacks kthread? - * Caller must have preemption disabled. - */ -@@ -1083,67 +1053,6 @@ - return 0; - } - --static void rcu_kthread_do_work(void) --{ -- rcu_do_batch(&rcu_sched_state, this_cpu_ptr(&rcu_sched_data)); -- rcu_do_batch(&rcu_bh_state, this_cpu_ptr(&rcu_bh_data)); -- rcu_preempt_do_callbacks(); --} -- --static void rcu_cpu_kthread_setup(unsigned int cpu) --{ -- struct sched_param sp; -- -- sp.sched_priority = kthread_prio; -- sched_setscheduler_nocheck(current, SCHED_FIFO, &sp); --} -- --static void rcu_cpu_kthread_park(unsigned int cpu) --{ -- per_cpu(rcu_cpu_kthread_status, cpu) = RCU_KTHREAD_OFFCPU; --} -- --static int rcu_cpu_kthread_should_run(unsigned int cpu) --{ -- return __this_cpu_read(rcu_cpu_has_work); --} -- --/* -- * Per-CPU kernel thread that invokes RCU callbacks. This replaces the -- * RCU softirq used in flavors and configurations of RCU that do not -- * support RCU priority boosting. -- */ --static void rcu_cpu_kthread(unsigned int cpu) --{ -- unsigned int *statusp = this_cpu_ptr(&rcu_cpu_kthread_status); -- char work, *workp = this_cpu_ptr(&rcu_cpu_has_work); -- int spincnt; -- -- for (spincnt = 0; spincnt < 10; spincnt++) { -- trace_rcu_utilization(TPS("Start CPU kthread@rcu_wait")); -- local_bh_disable(); -- *statusp = RCU_KTHREAD_RUNNING; -- this_cpu_inc(rcu_cpu_kthread_loops); -- local_irq_disable(); -- work = *workp; -- *workp = 0; -- local_irq_enable(); -- if (work) -- rcu_kthread_do_work(); -- local_bh_enable(); -- if (*workp == 0) { -- trace_rcu_utilization(TPS("End CPU kthread@rcu_wait")); -- *statusp = RCU_KTHREAD_WAITING; -- return; -- } -- } -- *statusp = RCU_KTHREAD_YIELDING; -- trace_rcu_utilization(TPS("Start CPU kthread@rcu_yield")); -- schedule_timeout_interruptible(2); -- trace_rcu_utilization(TPS("End CPU kthread@rcu_yield")); -- *statusp = RCU_KTHREAD_WAITING; --} -- - /* - * Set the per-rcu_node kthread's affinity to cover all CPUs that are - * served by the rcu_node in question. The CPU hotplug lock is still -@@ -1174,26 +1083,12 @@ - free_cpumask_var(cm); - } - --static struct smp_hotplug_thread rcu_cpu_thread_spec = { -- .store = &rcu_cpu_kthread_task, -- .thread_should_run = rcu_cpu_kthread_should_run, -- .thread_fn = rcu_cpu_kthread, -- .thread_comm = "rcuc/%u", -- .setup = rcu_cpu_kthread_setup, -- .park = rcu_cpu_kthread_park, --}; -- - /* - * Spawn boost kthreads -- called as soon as the scheduler is running. - */ - static void __init rcu_spawn_boost_kthreads(void) - { - struct rcu_node *rnp; -- int cpu; -- -- for_each_possible_cpu(cpu) -- per_cpu(rcu_cpu_has_work, cpu) = 0; -- BUG_ON(smpboot_register_percpu_thread(&rcu_cpu_thread_spec)); - rcu_for_each_leaf_node(rcu_state_p, rnp) - (void)rcu_spawn_one_boost_kthread(rcu_state_p, rnp); - } -@@ -1216,11 +1111,6 @@ - raw_spin_unlock_irqrestore_rcu_node(rnp, flags); - } - --static void invoke_rcu_callbacks_kthread(void) --{ -- WARN_ON_ONCE(1); --} -- - static bool rcu_is_callbacks_kthread(void) - { - return false; -@@ -1244,7 +1134,7 @@ - - #endif /* #else #ifdef CONFIG_RCU_BOOST */ - --#if !defined(CONFIG_RCU_FAST_NO_HZ) -+#if !defined(CONFIG_RCU_FAST_NO_HZ) || defined(CONFIG_PREEMPT_RT_FULL) - - /* - * Check to see if any future RCU-related work will need to be done -@@ -1261,7 +1151,9 @@ - return IS_ENABLED(CONFIG_RCU_NOCB_CPU_ALL) - ? 0 : rcu_cpu_has_callbacks(NULL); - } -+#endif /* !defined(CONFIG_RCU_FAST_NO_HZ) || defined(CONFIG_PREEMPT_RT_FULL) */ - -+#if !defined(CONFIG_RCU_FAST_NO_HZ) - /* - * Because we do not have RCU_FAST_NO_HZ, don't bother cleaning up - * after it. -@@ -1357,6 +1249,8 @@ - return cbs_ready; - } - -+#ifndef CONFIG_PREEMPT_RT_FULL -+ - /* - * Allow the CPU to enter dyntick-idle mode unless it has callbacks ready - * to invoke. If the CPU has callbacks, try to advance them. Tell the -@@ -1402,6 +1296,7 @@ - *nextevt = basemono + dj * TICK_NSEC; - return 0; - } -+#endif /* #ifndef CONFIG_PREEMPT_RT_FULL */ - - /* - * Prepare a CPU for idle from an RCU perspective. The first major task -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/rcu/update.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/rcu/update.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/rcu/update.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/rcu/update.c 2017-04-18 17:54:26.000000000 +0200 -@@ -62,7 +62,7 @@ - #ifndef CONFIG_TINY_RCU - module_param(rcu_expedited, int, 0); - module_param(rcu_normal, int, 0); --static int rcu_normal_after_boot; -+static int rcu_normal_after_boot = IS_ENABLED(CONFIG_PREEMPT_RT_FULL); - module_param(rcu_normal_after_boot, int, 0); - #endif /* #ifndef CONFIG_TINY_RCU */ - -@@ -132,8 +132,7 @@ - } - EXPORT_SYMBOL_GPL(rcu_gp_is_normal); - --static atomic_t rcu_expedited_nesting = -- ATOMIC_INIT(IS_ENABLED(CONFIG_RCU_EXPEDITE_BOOT) ? 1 : 0); -+static atomic_t rcu_expedited_nesting = ATOMIC_INIT(1); - - /* - * Should normal grace-period primitives be expedited? Intended for -@@ -182,8 +181,7 @@ - */ - void rcu_end_inkernel_boot(void) - { -- if (IS_ENABLED(CONFIG_RCU_EXPEDITE_BOOT)) -- rcu_unexpedite_gp(); -+ rcu_unexpedite_gp(); - if (rcu_normal_after_boot) - WRITE_ONCE(rcu_normal, 1); - } -@@ -298,6 +296,7 @@ - } - EXPORT_SYMBOL_GPL(rcu_read_lock_held); - -+#ifndef CONFIG_PREEMPT_RT_FULL - /** - * rcu_read_lock_bh_held() - might we be in RCU-bh read-side critical section? - * -@@ -324,6 +323,7 @@ - return in_softirq() || irqs_disabled(); - } - EXPORT_SYMBOL_GPL(rcu_read_lock_bh_held); -+#endif - - #endif /* #ifdef CONFIG_DEBUG_LOCK_ALLOC */ - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/Makefile linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/Makefile ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/Makefile 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/Makefile 2017-04-18 17:54:26.000000000 +0200 -@@ -17,7 +17,7 @@ - - obj-y += core.o loadavg.o clock.o cputime.o - obj-y += idle_task.o fair.o rt.o deadline.o stop_task.o --obj-y += wait.o swait.o completion.o idle.o -+obj-y += wait.o swait.o swork.o completion.o idle.o - obj-$(CONFIG_SMP) += cpupri.o cpudeadline.o - obj-$(CONFIG_SCHED_AUTOGROUP) += auto_group.o - obj-$(CONFIG_SCHEDSTATS) += stats.o -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/completion.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/completion.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/completion.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/completion.c 2017-04-18 17:54:26.000000000 +0200 -@@ -30,10 +30,10 @@ - { - unsigned long flags; - -- spin_lock_irqsave(&x->wait.lock, flags); -+ raw_spin_lock_irqsave(&x->wait.lock, flags); - x->done++; -- __wake_up_locked(&x->wait, TASK_NORMAL, 1); -- spin_unlock_irqrestore(&x->wait.lock, flags); -+ swake_up_locked(&x->wait); -+ raw_spin_unlock_irqrestore(&x->wait.lock, flags); - } - EXPORT_SYMBOL(complete); - -@@ -50,10 +50,10 @@ - { - unsigned long flags; - -- spin_lock_irqsave(&x->wait.lock, flags); -+ raw_spin_lock_irqsave(&x->wait.lock, flags); - x->done += UINT_MAX/2; -- __wake_up_locked(&x->wait, TASK_NORMAL, 0); -- spin_unlock_irqrestore(&x->wait.lock, flags); -+ swake_up_all_locked(&x->wait); -+ raw_spin_unlock_irqrestore(&x->wait.lock, flags); - } - EXPORT_SYMBOL(complete_all); - -@@ -62,20 +62,20 @@ - long (*action)(long), long timeout, int state) - { - if (!x->done) { -- DECLARE_WAITQUEUE(wait, current); -+ DECLARE_SWAITQUEUE(wait); - -- __add_wait_queue_tail_exclusive(&x->wait, &wait); -+ __prepare_to_swait(&x->wait, &wait); - do { - if (signal_pending_state(state, current)) { - timeout = -ERESTARTSYS; - break; - } - __set_current_state(state); -- spin_unlock_irq(&x->wait.lock); -+ raw_spin_unlock_irq(&x->wait.lock); - timeout = action(timeout); -- spin_lock_irq(&x->wait.lock); -+ raw_spin_lock_irq(&x->wait.lock); - } while (!x->done && timeout); -- __remove_wait_queue(&x->wait, &wait); -+ __finish_swait(&x->wait, &wait); - if (!x->done) - return timeout; - } -@@ -89,9 +89,9 @@ - { - might_sleep(); - -- spin_lock_irq(&x->wait.lock); -+ raw_spin_lock_irq(&x->wait.lock); - timeout = do_wait_for_common(x, action, timeout, state); -- spin_unlock_irq(&x->wait.lock); -+ raw_spin_unlock_irq(&x->wait.lock); - return timeout; - } - -@@ -277,12 +277,12 @@ - if (!READ_ONCE(x->done)) - return 0; - -- spin_lock_irqsave(&x->wait.lock, flags); -+ raw_spin_lock_irqsave(&x->wait.lock, flags); - if (!x->done) - ret = 0; - else - x->done--; -- spin_unlock_irqrestore(&x->wait.lock, flags); -+ raw_spin_unlock_irqrestore(&x->wait.lock, flags); - return ret; - } - EXPORT_SYMBOL(try_wait_for_completion); -@@ -311,7 +311,7 @@ - * after it's acquired the lock. - */ - smp_rmb(); -- spin_unlock_wait(&x->wait.lock); -+ raw_spin_unlock_wait(&x->wait.lock); - return true; - } - EXPORT_SYMBOL(completion_done); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/core.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/core.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/core.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/core.c 2017-04-18 17:54:26.000000000 +0200 -@@ -129,7 +129,11 @@ - * Number of tasks to iterate in a single balance run. - * Limited because this is done with IRQs disabled. - */ -+#ifndef CONFIG_PREEMPT_RT_FULL - const_debug unsigned int sysctl_sched_nr_migrate = 32; -+#else -+const_debug unsigned int sysctl_sched_nr_migrate = 8; -+#endif - - /* - * period over which we average the RT time consumption, measured -@@ -345,6 +349,7 @@ - - hrtimer_init(&rq->hrtick_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); - rq->hrtick_timer.function = hrtick; -+ rq->hrtick_timer.irqsafe = 1; - } - #else /* CONFIG_SCHED_HRTICK */ - static inline void hrtick_clear(struct rq *rq) -@@ -449,7 +454,7 @@ - head->lastp = &node->next; - } - --void wake_up_q(struct wake_q_head *head) -+void __wake_up_q(struct wake_q_head *head, bool sleeper) - { - struct wake_q_node *node = head->first; - -@@ -466,7 +471,10 @@ - * wake_up_process() implies a wmb() to pair with the queueing - * in wake_q_add() so as not to miss wakeups. - */ -- wake_up_process(task); -+ if (sleeper) -+ wake_up_lock_sleeper(task); -+ else -+ wake_up_process(task); - put_task_struct(task); - } - } -@@ -502,6 +510,38 @@ - trace_sched_wake_idle_without_ipi(cpu); - } - -+#ifdef CONFIG_PREEMPT_LAZY -+void resched_curr_lazy(struct rq *rq) -+{ -+ struct task_struct *curr = rq->curr; -+ int cpu; -+ -+ if (!sched_feat(PREEMPT_LAZY)) { -+ resched_curr(rq); -+ return; -+ } -+ -+ lockdep_assert_held(&rq->lock); -+ -+ if (test_tsk_need_resched(curr)) -+ return; -+ -+ if (test_tsk_need_resched_lazy(curr)) -+ return; -+ -+ set_tsk_need_resched_lazy(curr); -+ -+ cpu = cpu_of(rq); -+ if (cpu == smp_processor_id()) -+ return; -+ -+ /* NEED_RESCHED_LAZY must be visible before we test polling */ -+ smp_mb(); -+ if (!tsk_is_polling(curr)) -+ smp_send_reschedule(cpu); -+} -+#endif -+ - void resched_cpu(int cpu) - { - struct rq *rq = cpu_rq(cpu); -@@ -525,11 +565,14 @@ - */ - int get_nohz_timer_target(void) - { -- int i, cpu = smp_processor_id(); -+ int i, cpu; - struct sched_domain *sd; - -+ preempt_disable_rt(); -+ cpu = smp_processor_id(); -+ - if (!idle_cpu(cpu) && is_housekeeping_cpu(cpu)) -- return cpu; -+ goto preempt_en_rt; - - rcu_read_lock(); - for_each_domain(cpu, sd) { -@@ -548,6 +591,8 @@ - cpu = housekeeping_any_cpu(); - unlock: - rcu_read_unlock(); -+preempt_en_rt: -+ preempt_enable_rt(); - return cpu; - } - /* -@@ -1100,6 +1145,11 @@ - - lockdep_assert_held(&p->pi_lock); - -+ if (__migrate_disabled(p)) { -+ cpumask_copy(&p->cpus_allowed, new_mask); -+ return; -+ } -+ - queued = task_on_rq_queued(p); - running = task_current(rq, p); - -@@ -1122,6 +1172,84 @@ - set_curr_task(rq, p); - } - -+static DEFINE_PER_CPU(struct cpumask, sched_cpumasks); -+static DEFINE_MUTEX(sched_down_mutex); -+static cpumask_t sched_down_cpumask; -+ -+void tell_sched_cpu_down_begin(int cpu) -+{ -+ mutex_lock(&sched_down_mutex); -+ cpumask_set_cpu(cpu, &sched_down_cpumask); -+ mutex_unlock(&sched_down_mutex); -+} -+ -+void tell_sched_cpu_down_done(int cpu) -+{ -+ mutex_lock(&sched_down_mutex); -+ cpumask_clear_cpu(cpu, &sched_down_cpumask); -+ mutex_unlock(&sched_down_mutex); -+} -+ -+/** -+ * migrate_me - try to move the current task off this cpu -+ * -+ * Used by the pin_current_cpu() code to try to get tasks -+ * to move off the current CPU as it is going down. -+ * It will only move the task if the task isn't pinned to -+ * the CPU (with migrate_disable, affinity or NO_SETAFFINITY) -+ * and the task has to be in a RUNNING state. Otherwise the -+ * movement of the task will wake it up (change its state -+ * to running) when the task did not expect it. -+ * -+ * Returns 1 if it succeeded in moving the current task -+ * 0 otherwise. -+ */ -+int migrate_me(void) -+{ -+ struct task_struct *p = current; -+ struct migration_arg arg; -+ struct cpumask *cpumask; -+ struct cpumask *mask; -+ unsigned int dest_cpu; -+ struct rq_flags rf; -+ struct rq *rq; -+ -+ /* -+ * We can not migrate tasks bounded to a CPU or tasks not -+ * running. The movement of the task will wake it up. -+ */ -+ if (p->flags & PF_NO_SETAFFINITY || p->state) -+ return 0; -+ -+ mutex_lock(&sched_down_mutex); -+ rq = task_rq_lock(p, &rf); -+ -+ cpumask = this_cpu_ptr(&sched_cpumasks); -+ mask = &p->cpus_allowed; -+ -+ cpumask_andnot(cpumask, mask, &sched_down_cpumask); -+ -+ if (!cpumask_weight(cpumask)) { -+ /* It's only on this CPU? */ -+ task_rq_unlock(rq, p, &rf); -+ mutex_unlock(&sched_down_mutex); -+ return 0; -+ } -+ -+ dest_cpu = cpumask_any_and(cpu_active_mask, cpumask); -+ -+ arg.task = p; -+ arg.dest_cpu = dest_cpu; -+ -+ task_rq_unlock(rq, p, &rf); -+ -+ stop_one_cpu(cpu_of(rq), migration_cpu_stop, &arg); -+ tlb_migrate_finish(p->mm); -+ mutex_unlock(&sched_down_mutex); -+ -+ return 1; -+} -+ - /* - * Change a given task's CPU affinity. Migrate the thread to a - * proper CPU and schedule it away if the CPU it's executing on -@@ -1179,7 +1307,7 @@ - } - - /* Can the task run on the task's current CPU? If so, we're done */ -- if (cpumask_test_cpu(task_cpu(p), new_mask)) -+ if (cpumask_test_cpu(task_cpu(p), new_mask) || __migrate_disabled(p)) - goto out; - - dest_cpu = cpumask_any_and(cpu_valid_mask, new_mask); -@@ -1366,6 +1494,18 @@ - return ret; - } - -+static bool check_task_state(struct task_struct *p, long match_state) -+{ -+ bool match = false; -+ -+ raw_spin_lock_irq(&p->pi_lock); -+ if (p->state == match_state || p->saved_state == match_state) -+ match = true; -+ raw_spin_unlock_irq(&p->pi_lock); -+ -+ return match; -+} -+ - /* - * wait_task_inactive - wait for a thread to unschedule. - * -@@ -1410,7 +1550,7 @@ - * is actually now running somewhere else! - */ - while (task_running(rq, p)) { -- if (match_state && unlikely(p->state != match_state)) -+ if (match_state && !check_task_state(p, match_state)) - return 0; - cpu_relax(); - } -@@ -1425,7 +1565,8 @@ - running = task_running(rq, p); - queued = task_on_rq_queued(p); - ncsw = 0; -- if (!match_state || p->state == match_state) -+ if (!match_state || p->state == match_state || -+ p->saved_state == match_state) - ncsw = p->nvcsw | LONG_MIN; /* sets MSB */ - task_rq_unlock(rq, p, &rf); - -@@ -1680,10 +1821,6 @@ - { - activate_task(rq, p, en_flags); - p->on_rq = TASK_ON_RQ_QUEUED; -- -- /* if a worker is waking up, notify workqueue */ -- if (p->flags & PF_WQ_WORKER) -- wq_worker_waking_up(p, cpu_of(rq)); - } - - /* -@@ -2018,8 +2155,27 @@ - */ - smp_mb__before_spinlock(); - raw_spin_lock_irqsave(&p->pi_lock, flags); -- if (!(p->state & state)) -+ if (!(p->state & state)) { -+ /* -+ * The task might be running due to a spinlock sleeper -+ * wakeup. Check the saved state and set it to running -+ * if the wakeup condition is true. -+ */ -+ if (!(wake_flags & WF_LOCK_SLEEPER)) { -+ if (p->saved_state & state) { -+ p->saved_state = TASK_RUNNING; -+ success = 1; -+ } -+ } - goto out; -+ } -+ -+ /* -+ * If this is a regular wakeup, then we can unconditionally -+ * clear the saved state of a "lock sleeper". -+ */ -+ if (!(wake_flags & WF_LOCK_SLEEPER)) -+ p->saved_state = TASK_RUNNING; - - trace_sched_waking(p); - -@@ -2102,53 +2258,6 @@ - } - - /** -- * try_to_wake_up_local - try to wake up a local task with rq lock held -- * @p: the thread to be awakened -- * @cookie: context's cookie for pinning -- * -- * Put @p on the run-queue if it's not already there. The caller must -- * ensure that this_rq() is locked, @p is bound to this_rq() and not -- * the current task. -- */ --static void try_to_wake_up_local(struct task_struct *p, struct pin_cookie cookie) --{ -- struct rq *rq = task_rq(p); -- -- if (WARN_ON_ONCE(rq != this_rq()) || -- WARN_ON_ONCE(p == current)) -- return; -- -- lockdep_assert_held(&rq->lock); -- -- if (!raw_spin_trylock(&p->pi_lock)) { -- /* -- * This is OK, because current is on_cpu, which avoids it being -- * picked for load-balance and preemption/IRQs are still -- * disabled avoiding further scheduler activity on it and we've -- * not yet picked a replacement task. -- */ -- lockdep_unpin_lock(&rq->lock, cookie); -- raw_spin_unlock(&rq->lock); -- raw_spin_lock(&p->pi_lock); -- raw_spin_lock(&rq->lock); -- lockdep_repin_lock(&rq->lock, cookie); -- } -- -- if (!(p->state & TASK_NORMAL)) -- goto out; -- -- trace_sched_waking(p); -- -- if (!task_on_rq_queued(p)) -- ttwu_activate(rq, p, ENQUEUE_WAKEUP); -- -- ttwu_do_wakeup(rq, p, 0, cookie); -- ttwu_stat(p, smp_processor_id(), 0); --out: -- raw_spin_unlock(&p->pi_lock); --} -- --/** - * wake_up_process - Wake up a specific process - * @p: The process to be woken up. - * -@@ -2166,6 +2275,18 @@ - } - EXPORT_SYMBOL(wake_up_process); - -+/** -+ * wake_up_lock_sleeper - Wake up a specific process blocked on a "sleeping lock" -+ * @p: The process to be woken up. -+ * -+ * Same as wake_up_process() above, but wake_flags=WF_LOCK_SLEEPER to indicate -+ * the nature of the wakeup. -+ */ -+int wake_up_lock_sleeper(struct task_struct *p) -+{ -+ return try_to_wake_up(p, TASK_ALL, WF_LOCK_SLEEPER); -+} -+ - int wake_up_state(struct task_struct *p, unsigned int state) - { - return try_to_wake_up(p, state, 0); -@@ -2442,6 +2563,9 @@ - p->on_cpu = 0; - #endif - init_task_preempt_count(p); -+#ifdef CONFIG_HAVE_PREEMPT_LAZY -+ task_thread_info(p)->preempt_lazy_count = 0; -+#endif - #ifdef CONFIG_SMP - plist_node_init(&p->pushable_tasks, MAX_PRIO); - RB_CLEAR_NODE(&p->pushable_dl_tasks); -@@ -2770,21 +2894,16 @@ - finish_arch_post_lock_switch(); - - fire_sched_in_preempt_notifiers(current); -+ /* -+ * We use mmdrop_delayed() here so we don't have to do the -+ * full __mmdrop() when we are the last user. -+ */ - if (mm) -- mmdrop(mm); -+ mmdrop_delayed(mm); - if (unlikely(prev_state == TASK_DEAD)) { - if (prev->sched_class->task_dead) - prev->sched_class->task_dead(prev); - -- /* -- * Remove function-return probe instances associated with this -- * task and put them back on the free list. -- */ -- kprobe_flush_task(prev); -- -- /* Task is done with its stack. */ -- put_task_stack(prev); -- - put_task_struct(prev); - } - -@@ -3252,6 +3371,77 @@ - schedstat_inc(this_rq()->sched_count); - } - -+#if defined(CONFIG_PREEMPT_RT_FULL) && defined(CONFIG_SMP) -+ -+void migrate_disable(void) -+{ -+ struct task_struct *p = current; -+ -+ if (in_atomic() || irqs_disabled()) { -+#ifdef CONFIG_SCHED_DEBUG -+ p->migrate_disable_atomic++; -+#endif -+ return; -+ } -+ -+#ifdef CONFIG_SCHED_DEBUG -+ if (unlikely(p->migrate_disable_atomic)) { -+ tracing_off(); -+ WARN_ON_ONCE(1); -+ } -+#endif -+ -+ if (p->migrate_disable) { -+ p->migrate_disable++; -+ return; -+ } -+ -+ preempt_disable(); -+ preempt_lazy_disable(); -+ pin_current_cpu(); -+ p->migrate_disable = 1; -+ preempt_enable(); -+} -+EXPORT_SYMBOL(migrate_disable); -+ -+void migrate_enable(void) -+{ -+ struct task_struct *p = current; -+ -+ if (in_atomic() || irqs_disabled()) { -+#ifdef CONFIG_SCHED_DEBUG -+ p->migrate_disable_atomic--; -+#endif -+ return; -+ } -+ -+#ifdef CONFIG_SCHED_DEBUG -+ if (unlikely(p->migrate_disable_atomic)) { -+ tracing_off(); -+ WARN_ON_ONCE(1); -+ } -+#endif -+ WARN_ON_ONCE(p->migrate_disable <= 0); -+ -+ if (p->migrate_disable > 1) { -+ p->migrate_disable--; -+ return; -+ } -+ -+ preempt_disable(); -+ /* -+ * Clearing migrate_disable causes tsk_cpus_allowed to -+ * show the tasks original cpu affinity. -+ */ -+ p->migrate_disable = 0; -+ -+ unpin_current_cpu(); -+ preempt_enable(); -+ preempt_lazy_enable(); -+} -+EXPORT_SYMBOL(migrate_enable); -+#endif -+ - /* - * Pick up the highest-prio task: - */ -@@ -3368,19 +3558,6 @@ - } else { - deactivate_task(rq, prev, DEQUEUE_SLEEP); - prev->on_rq = 0; -- -- /* -- * If a worker went to sleep, notify and ask workqueue -- * whether it wants to wake up a task to maintain -- * concurrency. -- */ -- if (prev->flags & PF_WQ_WORKER) { -- struct task_struct *to_wakeup; -- -- to_wakeup = wq_worker_sleeping(prev); -- if (to_wakeup) -- try_to_wake_up_local(to_wakeup, cookie); -- } - } - switch_count = &prev->nvcsw; - } -@@ -3390,6 +3567,7 @@ - - next = pick_next_task(rq, prev, cookie); - clear_tsk_need_resched(prev); -+ clear_tsk_need_resched_lazy(prev); - clear_preempt_need_resched(); - rq->clock_skip_update = 0; - -@@ -3437,9 +3615,20 @@ - - static inline void sched_submit_work(struct task_struct *tsk) - { -- if (!tsk->state || tsk_is_pi_blocked(tsk)) -+ if (!tsk->state) - return; - /* -+ * If a worker went to sleep, notify and ask workqueue whether -+ * it wants to wake up a task to maintain concurrency. -+ */ -+ if (tsk->flags & PF_WQ_WORKER) -+ wq_worker_sleeping(tsk); -+ -+ -+ if (tsk_is_pi_blocked(tsk)) -+ return; -+ -+ /* - * If we are going to sleep and we have plugged IO queued, - * make sure to submit it to avoid deadlocks. - */ -@@ -3447,6 +3636,12 @@ - blk_schedule_flush_plug(tsk); - } - -+static void sched_update_worker(struct task_struct *tsk) -+{ -+ if (tsk->flags & PF_WQ_WORKER) -+ wq_worker_running(tsk); -+} -+ - asmlinkage __visible void __sched schedule(void) - { - struct task_struct *tsk = current; -@@ -3457,6 +3652,7 @@ - __schedule(false); - sched_preempt_enable_no_resched(); - } while (need_resched()); -+ sched_update_worker(tsk); - } - EXPORT_SYMBOL(schedule); - -@@ -3520,6 +3716,30 @@ - } while (need_resched()); - } - -+#ifdef CONFIG_PREEMPT_LAZY -+/* -+ * If TIF_NEED_RESCHED is then we allow to be scheduled away since this is -+ * set by a RT task. Oterwise we try to avoid beeing scheduled out as long as -+ * preempt_lazy_count counter >0. -+ */ -+static __always_inline int preemptible_lazy(void) -+{ -+ if (test_thread_flag(TIF_NEED_RESCHED)) -+ return 1; -+ if (current_thread_info()->preempt_lazy_count) -+ return 0; -+ return 1; -+} -+ -+#else -+ -+static inline int preemptible_lazy(void) -+{ -+ return 1; -+} -+ -+#endif -+ - #ifdef CONFIG_PREEMPT - /* - * this is the entry point to schedule() from in-kernel preemption -@@ -3534,7 +3754,8 @@ - */ - if (likely(!preemptible())) - return; -- -+ if (!preemptible_lazy()) -+ return; - preempt_schedule_common(); - } - NOKPROBE_SYMBOL(preempt_schedule); -@@ -3561,6 +3782,9 @@ - if (likely(!preemptible())) - return; - -+ if (!preemptible_lazy()) -+ return; -+ - do { - /* - * Because the function tracer can trace preempt_count_sub() -@@ -3583,7 +3807,16 @@ - * an infinite recursion. - */ - prev_ctx = exception_enter(); -+ /* -+ * The add/subtract must not be traced by the function -+ * tracer. But we still want to account for the -+ * preempt off latency tracer. Since the _notrace versions -+ * of add/subtract skip the accounting for latency tracer -+ * we must force it manually. -+ */ -+ start_critical_timings(); - __schedule(true); -+ stop_critical_timings(); - exception_exit(prev_ctx); - - preempt_latency_stop(1); -@@ -4939,6 +5172,7 @@ - } - EXPORT_SYMBOL(__cond_resched_lock); - -+#ifndef CONFIG_PREEMPT_RT_FULL - int __sched __cond_resched_softirq(void) - { - BUG_ON(!in_softirq()); -@@ -4952,6 +5186,7 @@ - return 0; - } - EXPORT_SYMBOL(__cond_resched_softirq); -+#endif - - /** - * yield - yield the current processor to other threads. -@@ -5315,7 +5550,9 @@ - - /* Set the preempt count _outside_ the spinlocks! */ - init_idle_preempt_count(idle, cpu); -- -+#ifdef CONFIG_HAVE_PREEMPT_LAZY -+ task_thread_info(idle)->preempt_lazy_count = 0; -+#endif - /* - * The idle tasks have their own, simple scheduling class: - */ -@@ -5458,6 +5695,8 @@ - #endif /* CONFIG_NUMA_BALANCING */ - - #ifdef CONFIG_HOTPLUG_CPU -+static DEFINE_PER_CPU(struct mm_struct *, idle_last_mm); -+ - /* - * Ensures that the idle task is using init_mm right before its cpu goes - * offline. -@@ -5472,7 +5711,12 @@ - switch_mm_irqs_off(mm, &init_mm, current); - finish_arch_post_lock_switch(); - } -- mmdrop(mm); -+ /* -+ * Defer the cleanup to an alive cpu. On RT we can neither -+ * call mmdrop() nor mmdrop_delayed() from here. -+ */ -+ per_cpu(idle_last_mm, smp_processor_id()) = mm; -+ - } - - /* -@@ -7418,6 +7662,10 @@ - update_max_interval(); - nohz_balance_exit_idle(cpu); - hrtick_clear(rq); -+ if (per_cpu(idle_last_mm, cpu)) { -+ mmdrop_delayed(per_cpu(idle_last_mm, cpu)); -+ per_cpu(idle_last_mm, cpu) = NULL; -+ } - return 0; - } - #endif -@@ -7698,7 +7946,7 @@ - #ifdef CONFIG_DEBUG_ATOMIC_SLEEP - static inline int preempt_count_equals(int preempt_offset) - { -- int nested = preempt_count() + rcu_preempt_depth(); -+ int nested = preempt_count() + sched_rcu_preempt_depth(); - - return (nested == preempt_offset); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/deadline.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/deadline.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/deadline.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/deadline.c 2017-04-18 17:54:26.000000000 +0200 -@@ -687,6 +687,7 @@ - - hrtimer_init(timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); - timer->function = dl_task_timer; -+ timer->irqsafe = 1; - } - - static -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/debug.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/debug.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/debug.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/debug.c 2017-04-18 17:54:26.000000000 +0200 -@@ -558,6 +558,9 @@ - P(rt_throttled); - PN(rt_time); - PN(rt_runtime); -+#ifdef CONFIG_SMP -+ P(rt_nr_migratory); -+#endif - - #undef PN - #undef P -@@ -953,6 +956,10 @@ - #endif - P(policy); - P(prio); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ P(migrate_disable); -+#endif -+ P(nr_cpus_allowed); - #undef PN_SCHEDSTAT - #undef PN - #undef __PN -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/fair.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/fair.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/fair.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/fair.c 2017-04-18 17:54:26.000000000 +0200 -@@ -3518,7 +3518,7 @@ - ideal_runtime = sched_slice(cfs_rq, curr); - delta_exec = curr->sum_exec_runtime - curr->prev_sum_exec_runtime; - if (delta_exec > ideal_runtime) { -- resched_curr(rq_of(cfs_rq)); -+ resched_curr_lazy(rq_of(cfs_rq)); - /* - * The current task ran long enough, ensure it doesn't get - * re-elected due to buddy favours. -@@ -3542,7 +3542,7 @@ - return; - - if (delta > ideal_runtime) -- resched_curr(rq_of(cfs_rq)); -+ resched_curr_lazy(rq_of(cfs_rq)); - } - - static void -@@ -3684,7 +3684,7 @@ - * validating it and just reschedule. - */ - if (queued) { -- resched_curr(rq_of(cfs_rq)); -+ resched_curr_lazy(rq_of(cfs_rq)); - return; - } - /* -@@ -3866,7 +3866,7 @@ - * hierarchy can be throttled - */ - if (!assign_cfs_rq_runtime(cfs_rq) && likely(cfs_rq->curr)) -- resched_curr(rq_of(cfs_rq)); -+ resched_curr_lazy(rq_of(cfs_rq)); - } - - static __always_inline -@@ -4494,7 +4494,7 @@ - - if (delta < 0) { - if (rq->curr == p) -- resched_curr(rq); -+ resched_curr_lazy(rq); - return; - } - hrtick_start(rq, delta); -@@ -5905,7 +5905,7 @@ - return; - - preempt: -- resched_curr(rq); -+ resched_curr_lazy(rq); - /* - * Only set the backward buddy when the current task is still - * on the rq. This can happen when a wakeup gets interleaved -@@ -8631,7 +8631,7 @@ - * 'current' within the tree based on its new key value. - */ - swap(curr->vruntime, se->vruntime); -- resched_curr(rq); -+ resched_curr_lazy(rq); - } - - se->vruntime -= cfs_rq->min_vruntime; -@@ -8655,7 +8655,7 @@ - */ - if (rq->curr == p) { - if (p->prio > oldprio) -- resched_curr(rq); -+ resched_curr_lazy(rq); - } else - check_preempt_curr(rq, p, 0); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/features.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/features.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/features.h 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/features.h 2017-04-18 17:54:26.000000000 +0200 -@@ -45,11 +45,19 @@ - */ - SCHED_FEAT(NONTASK_CAPACITY, true) - -+#ifdef CONFIG_PREEMPT_RT_FULL -+SCHED_FEAT(TTWU_QUEUE, false) -+# ifdef CONFIG_PREEMPT_LAZY -+SCHED_FEAT(PREEMPT_LAZY, true) -+# endif -+#else -+ - /* - * Queue remote wakeups on the target CPU and process them - * using the scheduler IPI. Reduces rq->lock contention/bounces. - */ - SCHED_FEAT(TTWU_QUEUE, true) -+#endif - - #ifdef HAVE_RT_PUSH_IPI - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/rt.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/rt.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/rt.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/rt.c 2017-04-18 17:54:26.000000000 +0200 -@@ -47,6 +47,7 @@ - - hrtimer_init(&rt_b->rt_period_timer, - CLOCK_MONOTONIC, HRTIMER_MODE_REL); -+ rt_b->rt_period_timer.irqsafe = 1; - rt_b->rt_period_timer.function = sched_rt_period_timer; - } - -@@ -101,6 +102,7 @@ - rt_rq->push_cpu = nr_cpu_ids; - raw_spin_lock_init(&rt_rq->push_lock); - init_irq_work(&rt_rq->push_work, push_irq_work_func); -+ rt_rq->push_work.flags |= IRQ_WORK_HARD_IRQ; - #endif - #endif /* CONFIG_SMP */ - /* We start is dequeued state, because no RT tasks are queued */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/sched.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/sched.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/sched.h 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/sched.h 2017-04-18 17:54:26.000000000 +0200 -@@ -1163,6 +1163,7 @@ - #define WF_SYNC 0x01 /* waker goes to sleep after wakeup */ - #define WF_FORK 0x02 /* child wakeup after fork */ - #define WF_MIGRATED 0x4 /* internal use, task got migrated */ -+#define WF_LOCK_SLEEPER 0x08 /* wakeup spinlock "sleeper" */ - - /* - * To aid in avoiding the subversion of "niceness" due to uneven distribution -@@ -1346,6 +1347,15 @@ - extern void resched_curr(struct rq *rq); - extern void resched_cpu(int cpu); - -+#ifdef CONFIG_PREEMPT_LAZY -+extern void resched_curr_lazy(struct rq *rq); -+#else -+static inline void resched_curr_lazy(struct rq *rq) -+{ -+ resched_curr(rq); -+} -+#endif -+ - extern struct rt_bandwidth def_rt_bandwidth; - extern void init_rt_bandwidth(struct rt_bandwidth *rt_b, u64 period, u64 runtime); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/swait.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/swait.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/swait.c 2017-04-16 10:38:29.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/swait.c 2017-04-18 17:54:26.000000000 +0200 -@@ -1,5 +1,6 @@ - #include - #include -+#include - - void __init_swait_queue_head(struct swait_queue_head *q, const char *name, - struct lock_class_key *key) -@@ -29,6 +30,25 @@ - } - EXPORT_SYMBOL(swake_up_locked); - -+void swake_up_all_locked(struct swait_queue_head *q) -+{ -+ struct swait_queue *curr; -+ int wakes = 0; -+ -+ while (!list_empty(&q->task_list)) { -+ -+ curr = list_first_entry(&q->task_list, typeof(*curr), -+ task_list); -+ wake_up_process(curr->task); -+ list_del_init(&curr->task_list); -+ wakes++; -+ } -+ if (pm_in_action) -+ return; -+ WARN(wakes > 2, "complete_all() with %d waiters\n", wakes); -+} -+EXPORT_SYMBOL(swake_up_all_locked); -+ - void swake_up(struct swait_queue_head *q) - { - unsigned long flags; -@@ -54,6 +74,7 @@ - if (!swait_active(q)) - return; - -+ WARN_ON(irqs_disabled()); - raw_spin_lock_irq(&q->lock); - list_splice_init(&q->task_list, &tmp); - while (!list_empty(&tmp)) { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/swork.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/swork.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/sched/swork.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/sched/swork.c 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,173 @@ -+/* -+ * Copyright (C) 2014 BMW Car IT GmbH, Daniel Wagner daniel.wagner@bmw-carit.de -+ * -+ * Provides a framework for enqueuing callbacks from irq context -+ * PREEMPT_RT_FULL safe. The callbacks are executed in kthread context. -+ */ -+ -+#include -+#include -+#include -+#include -+#include -+#include -+ -+#define SWORK_EVENT_PENDING (1 << 0) -+ -+static DEFINE_MUTEX(worker_mutex); -+static struct sworker *glob_worker; -+ -+struct sworker { -+ struct list_head events; -+ struct swait_queue_head wq; -+ -+ raw_spinlock_t lock; -+ -+ struct task_struct *task; -+ int refs; -+}; -+ -+static bool swork_readable(struct sworker *worker) -+{ -+ bool r; -+ -+ if (kthread_should_stop()) -+ return true; -+ -+ raw_spin_lock_irq(&worker->lock); -+ r = !list_empty(&worker->events); -+ raw_spin_unlock_irq(&worker->lock); -+ -+ return r; -+} -+ -+static int swork_kthread(void *arg) -+{ -+ struct sworker *worker = arg; -+ -+ for (;;) { -+ swait_event_interruptible(worker->wq, -+ swork_readable(worker)); -+ if (kthread_should_stop()) -+ break; -+ -+ raw_spin_lock_irq(&worker->lock); -+ while (!list_empty(&worker->events)) { -+ struct swork_event *sev; -+ -+ sev = list_first_entry(&worker->events, -+ struct swork_event, item); -+ list_del(&sev->item); -+ raw_spin_unlock_irq(&worker->lock); -+ -+ WARN_ON_ONCE(!test_and_clear_bit(SWORK_EVENT_PENDING, -+ &sev->flags)); -+ sev->func(sev); -+ raw_spin_lock_irq(&worker->lock); -+ } -+ raw_spin_unlock_irq(&worker->lock); -+ } -+ return 0; -+} -+ -+static struct sworker *swork_create(void) -+{ -+ struct sworker *worker; -+ -+ worker = kzalloc(sizeof(*worker), GFP_KERNEL); -+ if (!worker) -+ return ERR_PTR(-ENOMEM); -+ -+ INIT_LIST_HEAD(&worker->events); -+ raw_spin_lock_init(&worker->lock); -+ init_swait_queue_head(&worker->wq); -+ -+ worker->task = kthread_run(swork_kthread, worker, "kswork"); -+ if (IS_ERR(worker->task)) { -+ kfree(worker); -+ return ERR_PTR(-ENOMEM); -+ } -+ -+ return worker; -+} -+ -+static void swork_destroy(struct sworker *worker) -+{ -+ kthread_stop(worker->task); -+ -+ WARN_ON(!list_empty(&worker->events)); -+ kfree(worker); -+} -+ -+/** -+ * swork_queue - queue swork -+ * -+ * Returns %false if @work was already on a queue, %true otherwise. -+ * -+ * The work is queued and processed on a random CPU -+ */ -+bool swork_queue(struct swork_event *sev) -+{ -+ unsigned long flags; -+ -+ if (test_and_set_bit(SWORK_EVENT_PENDING, &sev->flags)) -+ return false; -+ -+ raw_spin_lock_irqsave(&glob_worker->lock, flags); -+ list_add_tail(&sev->item, &glob_worker->events); -+ raw_spin_unlock_irqrestore(&glob_worker->lock, flags); -+ -+ swake_up(&glob_worker->wq); -+ return true; -+} -+EXPORT_SYMBOL_GPL(swork_queue); -+ -+/** -+ * swork_get - get an instance of the sworker -+ * -+ * Returns an negative error code if the initialization if the worker did not -+ * work, %0 otherwise. -+ * -+ */ -+int swork_get(void) -+{ -+ struct sworker *worker; -+ -+ mutex_lock(&worker_mutex); -+ if (!glob_worker) { -+ worker = swork_create(); -+ if (IS_ERR(worker)) { -+ mutex_unlock(&worker_mutex); -+ return -ENOMEM; -+ } -+ -+ glob_worker = worker; -+ } -+ -+ glob_worker->refs++; -+ mutex_unlock(&worker_mutex); -+ -+ return 0; -+} -+EXPORT_SYMBOL_GPL(swork_get); -+ -+/** -+ * swork_put - puts an instance of the sworker -+ * -+ * Will destroy the sworker thread. This function must not be called until all -+ * queued events have been completed. -+ */ -+void swork_put(void) -+{ -+ mutex_lock(&worker_mutex); -+ -+ glob_worker->refs--; -+ if (glob_worker->refs > 0) -+ goto out; -+ -+ swork_destroy(glob_worker); -+ glob_worker = NULL; -+out: -+ mutex_unlock(&worker_mutex); -+} -+EXPORT_SYMBOL_GPL(swork_put); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/signal.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/signal.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/signal.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/signal.c 2017-04-18 17:54:26.000000000 +0200 -@@ -14,6 +14,7 @@ - #include - #include - #include -+#include - #include - #include - #include -@@ -352,13 +353,30 @@ - return false; - } - -+static inline struct sigqueue *get_task_cache(struct task_struct *t) -+{ -+ struct sigqueue *q = t->sigqueue_cache; -+ -+ if (cmpxchg(&t->sigqueue_cache, q, NULL) != q) -+ return NULL; -+ return q; -+} -+ -+static inline int put_task_cache(struct task_struct *t, struct sigqueue *q) -+{ -+ if (cmpxchg(&t->sigqueue_cache, NULL, q) == NULL) -+ return 0; -+ return 1; -+} -+ - /* - * allocate a new signal queue record - * - this may be called without locks if and only if t == current, otherwise an - * appropriate lock must be held to stop the target task from exiting - */ - static struct sigqueue * --__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimit) -+__sigqueue_do_alloc(int sig, struct task_struct *t, gfp_t flags, -+ int override_rlimit, int fromslab) - { - struct sigqueue *q = NULL; - struct user_struct *user; -@@ -375,7 +393,10 @@ - if (override_rlimit || - atomic_read(&user->sigpending) <= - task_rlimit(t, RLIMIT_SIGPENDING)) { -- q = kmem_cache_alloc(sigqueue_cachep, flags); -+ if (!fromslab) -+ q = get_task_cache(t); -+ if (!q) -+ q = kmem_cache_alloc(sigqueue_cachep, flags); - } else { - print_dropped_signal(sig); - } -@@ -392,6 +413,13 @@ - return q; - } - -+static struct sigqueue * -+__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, -+ int override_rlimit) -+{ -+ return __sigqueue_do_alloc(sig, t, flags, override_rlimit, 0); -+} -+ - static void __sigqueue_free(struct sigqueue *q) - { - if (q->flags & SIGQUEUE_PREALLOC) -@@ -401,6 +429,21 @@ - kmem_cache_free(sigqueue_cachep, q); - } - -+static void sigqueue_free_current(struct sigqueue *q) -+{ -+ struct user_struct *up; -+ -+ if (q->flags & SIGQUEUE_PREALLOC) -+ return; -+ -+ up = q->user; -+ if (rt_prio(current->normal_prio) && !put_task_cache(current, q)) { -+ atomic_dec(&up->sigpending); -+ free_uid(up); -+ } else -+ __sigqueue_free(q); -+} -+ - void flush_sigqueue(struct sigpending *queue) - { - struct sigqueue *q; -@@ -414,6 +457,21 @@ - } - - /* -+ * Called from __exit_signal. Flush tsk->pending and -+ * tsk->sigqueue_cache -+ */ -+void flush_task_sigqueue(struct task_struct *tsk) -+{ -+ struct sigqueue *q; -+ -+ flush_sigqueue(&tsk->pending); -+ -+ q = get_task_cache(tsk); -+ if (q) -+ kmem_cache_free(sigqueue_cachep, q); -+} -+ -+/* - * Flush all pending signals for this kthread. - */ - void flush_signals(struct task_struct *t) -@@ -525,7 +583,7 @@ - still_pending: - list_del_init(&first->list); - copy_siginfo(info, &first->info); -- __sigqueue_free(first); -+ sigqueue_free_current(first); - } else { - /* - * Ok, it wasn't in the queue. This must be -@@ -560,6 +618,8 @@ - { - int signr; - -+ WARN_ON_ONCE(tsk != current); -+ - /* We only dequeue private signals from ourselves, we don't let - * signalfd steal them - */ -@@ -1156,8 +1216,8 @@ - * We don't want to have recursive SIGSEGV's etc, for example, - * that is why we also clear SIGNAL_UNKILLABLE. - */ --int --force_sig_info(int sig, struct siginfo *info, struct task_struct *t) -+static int -+do_force_sig_info(int sig, struct siginfo *info, struct task_struct *t) - { - unsigned long int flags; - int ret, blocked, ignored; -@@ -1182,6 +1242,39 @@ - return ret; - } - -+int force_sig_info(int sig, struct siginfo *info, struct task_struct *t) -+{ -+/* -+ * On some archs, PREEMPT_RT has to delay sending a signal from a trap -+ * since it can not enable preemption, and the signal code's spin_locks -+ * turn into mutexes. Instead, it must set TIF_NOTIFY_RESUME which will -+ * send the signal on exit of the trap. -+ */ -+#ifdef ARCH_RT_DELAYS_SIGNAL_SEND -+ if (in_atomic()) { -+ if (WARN_ON_ONCE(t != current)) -+ return 0; -+ if (WARN_ON_ONCE(t->forced_info.si_signo)) -+ return 0; -+ -+ if (is_si_special(info)) { -+ WARN_ON_ONCE(info != SEND_SIG_PRIV); -+ t->forced_info.si_signo = sig; -+ t->forced_info.si_errno = 0; -+ t->forced_info.si_code = SI_KERNEL; -+ t->forced_info.si_pid = 0; -+ t->forced_info.si_uid = 0; -+ } else { -+ t->forced_info = *info; -+ } -+ -+ set_tsk_thread_flag(t, TIF_NOTIFY_RESUME); -+ return 0; -+ } -+#endif -+ return do_force_sig_info(sig, info, t); -+} -+ - /* - * Nuke all other threads in the group. - */ -@@ -1216,12 +1309,12 @@ - * Disable interrupts early to avoid deadlocks. - * See rcu_read_unlock() comment header for details. - */ -- local_irq_save(*flags); -+ local_irq_save_nort(*flags); - rcu_read_lock(); - sighand = rcu_dereference(tsk->sighand); - if (unlikely(sighand == NULL)) { - rcu_read_unlock(); -- local_irq_restore(*flags); -+ local_irq_restore_nort(*flags); - break; - } - /* -@@ -1242,7 +1335,7 @@ - } - spin_unlock(&sighand->siglock); - rcu_read_unlock(); -- local_irq_restore(*flags); -+ local_irq_restore_nort(*flags); - } - - return sighand; -@@ -1485,7 +1578,8 @@ - */ - struct sigqueue *sigqueue_alloc(void) - { -- struct sigqueue *q = __sigqueue_alloc(-1, current, GFP_KERNEL, 0); -+ /* Preallocated sigqueue objects always from the slabcache ! */ -+ struct sigqueue *q = __sigqueue_do_alloc(-1, current, GFP_KERNEL, 0, 1); - - if (q) - q->flags |= SIGQUEUE_PREALLOC; -@@ -1846,15 +1940,7 @@ - if (gstop_done && ptrace_reparented(current)) - do_notify_parent_cldstop(current, false, why); - -- /* -- * Don't want to allow preemption here, because -- * sys_ptrace() needs this task to be inactive. -- * -- * XXX: implement read_unlock_no_resched(). -- */ -- preempt_disable(); - read_unlock(&tasklist_lock); -- preempt_enable_no_resched(); - freezable_schedule(); - } else { - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/softirq.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/softirq.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/softirq.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/softirq.c 2017-04-18 17:54:26.000000000 +0200 -@@ -21,10 +21,12 @@ - #include - #include - #include -+#include - #include - #include - #include - #include -+#include - #include - - #define CREATE_TRACE_POINTS -@@ -56,12 +58,108 @@ - static struct softirq_action softirq_vec[NR_SOFTIRQS] __cacheline_aligned_in_smp; - - DEFINE_PER_CPU(struct task_struct *, ksoftirqd); -+#ifdef CONFIG_PREEMPT_RT_FULL -+#define TIMER_SOFTIRQS ((1 << TIMER_SOFTIRQ) | (1 << HRTIMER_SOFTIRQ)) -+DEFINE_PER_CPU(struct task_struct *, ktimer_softirqd); -+#endif - - const char * const softirq_to_name[NR_SOFTIRQS] = { - "HI", "TIMER", "NET_TX", "NET_RX", "BLOCK", "IRQ_POLL", - "TASKLET", "SCHED", "HRTIMER", "RCU" - }; - -+#ifdef CONFIG_NO_HZ_COMMON -+# ifdef CONFIG_PREEMPT_RT_FULL -+ -+struct softirq_runner { -+ struct task_struct *runner[NR_SOFTIRQS]; -+}; -+ -+static DEFINE_PER_CPU(struct softirq_runner, softirq_runners); -+ -+static inline void softirq_set_runner(unsigned int sirq) -+{ -+ struct softirq_runner *sr = this_cpu_ptr(&softirq_runners); -+ -+ sr->runner[sirq] = current; -+} -+ -+static inline void softirq_clr_runner(unsigned int sirq) -+{ -+ struct softirq_runner *sr = this_cpu_ptr(&softirq_runners); -+ -+ sr->runner[sirq] = NULL; -+} -+ -+/* -+ * On preempt-rt a softirq running context might be blocked on a -+ * lock. There might be no other runnable task on this CPU because the -+ * lock owner runs on some other CPU. So we have to go into idle with -+ * the pending bit set. Therefor we need to check this otherwise we -+ * warn about false positives which confuses users and defeats the -+ * whole purpose of this test. -+ * -+ * This code is called with interrupts disabled. -+ */ -+void softirq_check_pending_idle(void) -+{ -+ static int rate_limit; -+ struct softirq_runner *sr = this_cpu_ptr(&softirq_runners); -+ u32 warnpending; -+ int i; -+ -+ if (rate_limit >= 10) -+ return; -+ -+ warnpending = local_softirq_pending() & SOFTIRQ_STOP_IDLE_MASK; -+ for (i = 0; i < NR_SOFTIRQS; i++) { -+ struct task_struct *tsk = sr->runner[i]; -+ -+ /* -+ * The wakeup code in rtmutex.c wakes up the task -+ * _before_ it sets pi_blocked_on to NULL under -+ * tsk->pi_lock. So we need to check for both: state -+ * and pi_blocked_on. -+ */ -+ if (tsk) { -+ raw_spin_lock(&tsk->pi_lock); -+ if (tsk->pi_blocked_on || tsk->state == TASK_RUNNING) { -+ /* Clear all bits pending in that task */ -+ warnpending &= ~(tsk->softirqs_raised); -+ warnpending &= ~(1 << i); -+ } -+ raw_spin_unlock(&tsk->pi_lock); -+ } -+ } -+ -+ if (warnpending) { -+ printk(KERN_ERR "NOHZ: local_softirq_pending %02x\n", -+ warnpending); -+ rate_limit++; -+ } -+} -+# else -+/* -+ * On !PREEMPT_RT we just printk rate limited: -+ */ -+void softirq_check_pending_idle(void) -+{ -+ static int rate_limit; -+ -+ if (rate_limit < 10 && -+ (local_softirq_pending() & SOFTIRQ_STOP_IDLE_MASK)) { -+ printk(KERN_ERR "NOHZ: local_softirq_pending %02x\n", -+ local_softirq_pending()); -+ rate_limit++; -+ } -+} -+# endif -+ -+#else /* !CONFIG_NO_HZ_COMMON */ -+static inline void softirq_set_runner(unsigned int sirq) { } -+static inline void softirq_clr_runner(unsigned int sirq) { } -+#endif -+ - /* - * we cannot loop indefinitely here to avoid userspace starvation, - * but we also don't want to introduce a worst case 1/HZ latency -@@ -77,6 +175,38 @@ - wake_up_process(tsk); - } - -+#ifdef CONFIG_PREEMPT_RT_FULL -+static void wakeup_timer_softirqd(void) -+{ -+ /* Interrupts are disabled: no need to stop preemption */ -+ struct task_struct *tsk = __this_cpu_read(ktimer_softirqd); -+ -+ if (tsk && tsk->state != TASK_RUNNING) -+ wake_up_process(tsk); -+} -+#endif -+ -+static void handle_softirq(unsigned int vec_nr) -+{ -+ struct softirq_action *h = softirq_vec + vec_nr; -+ int prev_count; -+ -+ prev_count = preempt_count(); -+ -+ kstat_incr_softirqs_this_cpu(vec_nr); -+ -+ trace_softirq_entry(vec_nr); -+ h->action(h); -+ trace_softirq_exit(vec_nr); -+ if (unlikely(prev_count != preempt_count())) { -+ pr_err("huh, entered softirq %u %s %p with preempt_count %08x, exited with %08x?\n", -+ vec_nr, softirq_to_name[vec_nr], h->action, -+ prev_count, preempt_count()); -+ preempt_count_set(prev_count); -+ } -+} -+ -+#ifndef CONFIG_PREEMPT_RT_FULL - /* - * If ksoftirqd is scheduled, we do not want to process pending softirqs - * right now. Let ksoftirqd handle this at its own rate, to get fairness. -@@ -88,6 +218,47 @@ - return tsk && (tsk->state == TASK_RUNNING); - } - -+static inline int ksoftirqd_softirq_pending(void) -+{ -+ return local_softirq_pending(); -+} -+ -+static void handle_pending_softirqs(u32 pending) -+{ -+ struct softirq_action *h = softirq_vec; -+ int softirq_bit; -+ -+ local_irq_enable(); -+ -+ h = softirq_vec; -+ -+ while ((softirq_bit = ffs(pending))) { -+ unsigned int vec_nr; -+ -+ h += softirq_bit - 1; -+ vec_nr = h - softirq_vec; -+ handle_softirq(vec_nr); -+ -+ h++; -+ pending >>= softirq_bit; -+ } -+ -+ rcu_bh_qs(); -+ local_irq_disable(); -+} -+ -+static void run_ksoftirqd(unsigned int cpu) -+{ -+ local_irq_disable(); -+ if (ksoftirqd_softirq_pending()) { -+ __do_softirq(); -+ local_irq_enable(); -+ cond_resched_rcu_qs(); -+ return; -+ } -+ local_irq_enable(); -+} -+ - /* - * preempt_count and SOFTIRQ_OFFSET usage: - * - preempt_count is changed by SOFTIRQ_OFFSET on entering or leaving -@@ -243,10 +414,8 @@ - unsigned long end = jiffies + MAX_SOFTIRQ_TIME; - unsigned long old_flags = current->flags; - int max_restart = MAX_SOFTIRQ_RESTART; -- struct softirq_action *h; - bool in_hardirq; - __u32 pending; -- int softirq_bit; - - /* - * Mask out PF_MEMALLOC s current task context is borrowed for the -@@ -265,36 +434,7 @@ - /* Reset the pending bitmask before enabling irqs */ - set_softirq_pending(0); - -- local_irq_enable(); -- -- h = softirq_vec; -- -- while ((softirq_bit = ffs(pending))) { -- unsigned int vec_nr; -- int prev_count; -- -- h += softirq_bit - 1; -- -- vec_nr = h - softirq_vec; -- prev_count = preempt_count(); -- -- kstat_incr_softirqs_this_cpu(vec_nr); -- -- trace_softirq_entry(vec_nr); -- h->action(h); -- trace_softirq_exit(vec_nr); -- if (unlikely(prev_count != preempt_count())) { -- pr_err("huh, entered softirq %u %s %p with preempt_count %08x, exited with %08x?\n", -- vec_nr, softirq_to_name[vec_nr], h->action, -- prev_count, preempt_count()); -- preempt_count_set(prev_count); -- } -- h++; -- pending >>= softirq_bit; -- } -- -- rcu_bh_qs(); -- local_irq_disable(); -+ handle_pending_softirqs(pending); - - pending = local_softirq_pending(); - if (pending) { -@@ -331,6 +471,309 @@ - } - - /* -+ * This function must run with irqs disabled! -+ */ -+void raise_softirq_irqoff(unsigned int nr) -+{ -+ __raise_softirq_irqoff(nr); -+ -+ /* -+ * If we're in an interrupt or softirq, we're done -+ * (this also catches softirq-disabled code). We will -+ * actually run the softirq once we return from -+ * the irq or softirq. -+ * -+ * Otherwise we wake up ksoftirqd to make sure we -+ * schedule the softirq soon. -+ */ -+ if (!in_interrupt()) -+ wakeup_softirqd(); -+} -+ -+void __raise_softirq_irqoff(unsigned int nr) -+{ -+ trace_softirq_raise(nr); -+ or_softirq_pending(1UL << nr); -+} -+ -+static inline void local_bh_disable_nort(void) { local_bh_disable(); } -+static inline void _local_bh_enable_nort(void) { _local_bh_enable(); } -+static void ksoftirqd_set_sched_params(unsigned int cpu) { } -+ -+#else /* !PREEMPT_RT_FULL */ -+ -+/* -+ * On RT we serialize softirq execution with a cpu local lock per softirq -+ */ -+static DEFINE_PER_CPU(struct local_irq_lock [NR_SOFTIRQS], local_softirq_locks); -+ -+void __init softirq_early_init(void) -+{ -+ int i; -+ -+ for (i = 0; i < NR_SOFTIRQS; i++) -+ local_irq_lock_init(local_softirq_locks[i]); -+} -+ -+static void lock_softirq(int which) -+{ -+ local_lock(local_softirq_locks[which]); -+} -+ -+static void unlock_softirq(int which) -+{ -+ local_unlock(local_softirq_locks[which]); -+} -+ -+static void do_single_softirq(int which) -+{ -+ unsigned long old_flags = current->flags; -+ -+ current->flags &= ~PF_MEMALLOC; -+ vtime_account_irq_enter(current); -+ current->flags |= PF_IN_SOFTIRQ; -+ lockdep_softirq_enter(); -+ local_irq_enable(); -+ handle_softirq(which); -+ local_irq_disable(); -+ lockdep_softirq_exit(); -+ current->flags &= ~PF_IN_SOFTIRQ; -+ vtime_account_irq_enter(current); -+ tsk_restore_flags(current, old_flags, PF_MEMALLOC); -+} -+ -+/* -+ * Called with interrupts disabled. Process softirqs which were raised -+ * in current context (or on behalf of ksoftirqd). -+ */ -+static void do_current_softirqs(void) -+{ -+ while (current->softirqs_raised) { -+ int i = __ffs(current->softirqs_raised); -+ unsigned int pending, mask = (1U << i); -+ -+ current->softirqs_raised &= ~mask; -+ local_irq_enable(); -+ -+ /* -+ * If the lock is contended, we boost the owner to -+ * process the softirq or leave the critical section -+ * now. -+ */ -+ lock_softirq(i); -+ local_irq_disable(); -+ softirq_set_runner(i); -+ /* -+ * Check with the local_softirq_pending() bits, -+ * whether we need to process this still or if someone -+ * else took care of it. -+ */ -+ pending = local_softirq_pending(); -+ if (pending & mask) { -+ set_softirq_pending(pending & ~mask); -+ do_single_softirq(i); -+ } -+ softirq_clr_runner(i); -+ WARN_ON(current->softirq_nestcnt != 1); -+ local_irq_enable(); -+ unlock_softirq(i); -+ local_irq_disable(); -+ } -+} -+ -+void __local_bh_disable(void) -+{ -+ if (++current->softirq_nestcnt == 1) -+ migrate_disable(); -+} -+EXPORT_SYMBOL(__local_bh_disable); -+ -+void __local_bh_enable(void) -+{ -+ if (WARN_ON(current->softirq_nestcnt == 0)) -+ return; -+ -+ local_irq_disable(); -+ if (current->softirq_nestcnt == 1 && current->softirqs_raised) -+ do_current_softirqs(); -+ local_irq_enable(); -+ -+ if (--current->softirq_nestcnt == 0) -+ migrate_enable(); -+} -+EXPORT_SYMBOL(__local_bh_enable); -+ -+void _local_bh_enable(void) -+{ -+ if (WARN_ON(current->softirq_nestcnt == 0)) -+ return; -+ if (--current->softirq_nestcnt == 0) -+ migrate_enable(); -+} -+EXPORT_SYMBOL(_local_bh_enable); -+ -+int in_serving_softirq(void) -+{ -+ return current->flags & PF_IN_SOFTIRQ; -+} -+EXPORT_SYMBOL(in_serving_softirq); -+ -+/* Called with preemption disabled */ -+static void run_ksoftirqd(unsigned int cpu) -+{ -+ local_irq_disable(); -+ current->softirq_nestcnt++; -+ -+ do_current_softirqs(); -+ current->softirq_nestcnt--; -+ local_irq_enable(); -+ cond_resched_rcu_qs(); -+} -+ -+/* -+ * Called from netif_rx_ni(). Preemption enabled, but migration -+ * disabled. So the cpu can't go away under us. -+ */ -+void thread_do_softirq(void) -+{ -+ if (!in_serving_softirq() && current->softirqs_raised) { -+ current->softirq_nestcnt++; -+ do_current_softirqs(); -+ current->softirq_nestcnt--; -+ } -+} -+ -+static void do_raise_softirq_irqoff(unsigned int nr) -+{ -+ unsigned int mask; -+ -+ mask = 1UL << nr; -+ -+ trace_softirq_raise(nr); -+ or_softirq_pending(mask); -+ -+ /* -+ * If we are not in a hard interrupt and inside a bh disabled -+ * region, we simply raise the flag on current. local_bh_enable() -+ * will make sure that the softirq is executed. Otherwise we -+ * delegate it to ksoftirqd. -+ */ -+ if (!in_irq() && current->softirq_nestcnt) -+ current->softirqs_raised |= mask; -+ else if (!__this_cpu_read(ksoftirqd) || !__this_cpu_read(ktimer_softirqd)) -+ return; -+ -+ if (mask & TIMER_SOFTIRQS) -+ __this_cpu_read(ktimer_softirqd)->softirqs_raised |= mask; -+ else -+ __this_cpu_read(ksoftirqd)->softirqs_raised |= mask; -+} -+ -+static void wakeup_proper_softirq(unsigned int nr) -+{ -+ if ((1UL << nr) & TIMER_SOFTIRQS) -+ wakeup_timer_softirqd(); -+ else -+ wakeup_softirqd(); -+} -+ -+void __raise_softirq_irqoff(unsigned int nr) -+{ -+ do_raise_softirq_irqoff(nr); -+ if (!in_irq() && !current->softirq_nestcnt) -+ wakeup_proper_softirq(nr); -+} -+ -+/* -+ * Same as __raise_softirq_irqoff() but will process them in ksoftirqd -+ */ -+void __raise_softirq_irqoff_ksoft(unsigned int nr) -+{ -+ unsigned int mask; -+ -+ if (WARN_ON_ONCE(!__this_cpu_read(ksoftirqd) || -+ !__this_cpu_read(ktimer_softirqd))) -+ return; -+ mask = 1UL << nr; -+ -+ trace_softirq_raise(nr); -+ or_softirq_pending(mask); -+ if (mask & TIMER_SOFTIRQS) -+ __this_cpu_read(ktimer_softirqd)->softirqs_raised |= mask; -+ else -+ __this_cpu_read(ksoftirqd)->softirqs_raised |= mask; -+ wakeup_proper_softirq(nr); -+} -+ -+/* -+ * This function must run with irqs disabled! -+ */ -+void raise_softirq_irqoff(unsigned int nr) -+{ -+ do_raise_softirq_irqoff(nr); -+ -+ /* -+ * If we're in an hard interrupt we let irq return code deal -+ * with the wakeup of ksoftirqd. -+ */ -+ if (in_irq()) -+ return; -+ /* -+ * If we are in thread context but outside of a bh disabled -+ * region, we need to wake ksoftirqd as well. -+ * -+ * CHECKME: Some of the places which do that could be wrapped -+ * into local_bh_disable/enable pairs. Though it's unclear -+ * whether this is worth the effort. To find those places just -+ * raise a WARN() if the condition is met. -+ */ -+ if (!current->softirq_nestcnt) -+ wakeup_proper_softirq(nr); -+} -+ -+static inline int ksoftirqd_softirq_pending(void) -+{ -+ return current->softirqs_raised; -+} -+ -+static inline void local_bh_disable_nort(void) { } -+static inline void _local_bh_enable_nort(void) { } -+ -+static inline void ksoftirqd_set_sched_params(unsigned int cpu) -+{ -+ /* Take over all but timer pending softirqs when starting */ -+ local_irq_disable(); -+ current->softirqs_raised = local_softirq_pending() & ~TIMER_SOFTIRQS; -+ local_irq_enable(); -+} -+ -+static inline void ktimer_softirqd_set_sched_params(unsigned int cpu) -+{ -+ struct sched_param param = { .sched_priority = 1 }; -+ -+ sched_setscheduler(current, SCHED_FIFO, ¶m); -+ -+ /* Take over timer pending softirqs when starting */ -+ local_irq_disable(); -+ current->softirqs_raised = local_softirq_pending() & TIMER_SOFTIRQS; -+ local_irq_enable(); -+} -+ -+static inline void ktimer_softirqd_clr_sched_params(unsigned int cpu, -+ bool online) -+{ -+ struct sched_param param = { .sched_priority = 0 }; -+ -+ sched_setscheduler(current, SCHED_NORMAL, ¶m); -+} -+ -+static int ktimer_softirqd_should_run(unsigned int cpu) -+{ -+ return current->softirqs_raised; -+} -+ -+#endif /* PREEMPT_RT_FULL */ -+/* - * Enter an interrupt context. - */ - void irq_enter(void) -@@ -341,9 +784,9 @@ - * Prevent raise_softirq from needlessly waking up ksoftirqd - * here, as softirq will be serviced on return from interrupt. - */ -- local_bh_disable(); -+ local_bh_disable_nort(); - tick_irq_enter(); -- _local_bh_enable(); -+ _local_bh_enable_nort(); - } - - __irq_enter(); -@@ -351,6 +794,7 @@ - - static inline void invoke_softirq(void) - { -+#ifndef CONFIG_PREEMPT_RT_FULL - if (ksoftirqd_running()) - return; - -@@ -373,6 +817,18 @@ - } else { - wakeup_softirqd(); - } -+#else /* PREEMPT_RT_FULL */ -+ unsigned long flags; -+ -+ local_irq_save(flags); -+ if (__this_cpu_read(ksoftirqd) && -+ __this_cpu_read(ksoftirqd)->softirqs_raised) -+ wakeup_softirqd(); -+ if (__this_cpu_read(ktimer_softirqd) && -+ __this_cpu_read(ktimer_softirqd)->softirqs_raised) -+ wakeup_timer_softirqd(); -+ local_irq_restore(flags); -+#endif - } - - static inline void tick_irq_exit(void) -@@ -409,26 +865,6 @@ - trace_hardirq_exit(); /* must be last! */ - } - --/* -- * This function must run with irqs disabled! -- */ --inline void raise_softirq_irqoff(unsigned int nr) --{ -- __raise_softirq_irqoff(nr); -- -- /* -- * If we're in an interrupt or softirq, we're done -- * (this also catches softirq-disabled code). We will -- * actually run the softirq once we return from -- * the irq or softirq. -- * -- * Otherwise we wake up ksoftirqd to make sure we -- * schedule the softirq soon. -- */ -- if (!in_interrupt()) -- wakeup_softirqd(); --} -- - void raise_softirq(unsigned int nr) - { - unsigned long flags; -@@ -438,12 +874,6 @@ - local_irq_restore(flags); - } - --void __raise_softirq_irqoff(unsigned int nr) --{ -- trace_softirq_raise(nr); -- or_softirq_pending(1UL << nr); --} -- - void open_softirq(int nr, void (*action)(struct softirq_action *)) - { - softirq_vec[nr].action = action; -@@ -460,15 +890,45 @@ - static DEFINE_PER_CPU(struct tasklet_head, tasklet_vec); - static DEFINE_PER_CPU(struct tasklet_head, tasklet_hi_vec); - -+static void inline -+__tasklet_common_schedule(struct tasklet_struct *t, struct tasklet_head *head, unsigned int nr) -+{ -+ if (tasklet_trylock(t)) { -+again: -+ /* We may have been preempted before tasklet_trylock -+ * and __tasklet_action may have already run. -+ * So double check the sched bit while the takslet -+ * is locked before adding it to the list. -+ */ -+ if (test_bit(TASKLET_STATE_SCHED, &t->state)) { -+ t->next = NULL; -+ *head->tail = t; -+ head->tail = &(t->next); -+ raise_softirq_irqoff(nr); -+ tasklet_unlock(t); -+ } else { -+ /* This is subtle. If we hit the corner case above -+ * It is possible that we get preempted right here, -+ * and another task has successfully called -+ * tasklet_schedule(), then this function, and -+ * failed on the trylock. Thus we must be sure -+ * before releasing the tasklet lock, that the -+ * SCHED_BIT is clear. Otherwise the tasklet -+ * may get its SCHED_BIT set, but not added to the -+ * list -+ */ -+ if (!tasklet_tryunlock(t)) -+ goto again; -+ } -+ } -+} -+ - void __tasklet_schedule(struct tasklet_struct *t) - { - unsigned long flags; - - local_irq_save(flags); -- t->next = NULL; -- *__this_cpu_read(tasklet_vec.tail) = t; -- __this_cpu_write(tasklet_vec.tail, &(t->next)); -- raise_softirq_irqoff(TASKLET_SOFTIRQ); -+ __tasklet_common_schedule(t, this_cpu_ptr(&tasklet_vec), TASKLET_SOFTIRQ); - local_irq_restore(flags); - } - EXPORT_SYMBOL(__tasklet_schedule); -@@ -478,10 +938,7 @@ - unsigned long flags; - - local_irq_save(flags); -- t->next = NULL; -- *__this_cpu_read(tasklet_hi_vec.tail) = t; -- __this_cpu_write(tasklet_hi_vec.tail, &(t->next)); -- raise_softirq_irqoff(HI_SOFTIRQ); -+ __tasklet_common_schedule(t, this_cpu_ptr(&tasklet_hi_vec), HI_SOFTIRQ); - local_irq_restore(flags); - } - EXPORT_SYMBOL(__tasklet_hi_schedule); -@@ -490,82 +947,122 @@ - { - BUG_ON(!irqs_disabled()); - -- t->next = __this_cpu_read(tasklet_hi_vec.head); -- __this_cpu_write(tasklet_hi_vec.head, t); -- __raise_softirq_irqoff(HI_SOFTIRQ); -+ __tasklet_hi_schedule(t); - } - EXPORT_SYMBOL(__tasklet_hi_schedule_first); - --static __latent_entropy void tasklet_action(struct softirq_action *a) -+void tasklet_enable(struct tasklet_struct *t) - { -- struct tasklet_struct *list; -+ if (!atomic_dec_and_test(&t->count)) -+ return; -+ if (test_and_clear_bit(TASKLET_STATE_PENDING, &t->state)) -+ tasklet_schedule(t); -+} -+EXPORT_SYMBOL(tasklet_enable); - -- local_irq_disable(); -- list = __this_cpu_read(tasklet_vec.head); -- __this_cpu_write(tasklet_vec.head, NULL); -- __this_cpu_write(tasklet_vec.tail, this_cpu_ptr(&tasklet_vec.head)); -- local_irq_enable(); -+static void __tasklet_action(struct softirq_action *a, -+ struct tasklet_struct *list) -+{ -+ int loops = 1000000; - - while (list) { - struct tasklet_struct *t = list; - - list = list->next; - -- if (tasklet_trylock(t)) { -- if (!atomic_read(&t->count)) { -- if (!test_and_clear_bit(TASKLET_STATE_SCHED, -- &t->state)) -- BUG(); -- t->func(t->data); -- tasklet_unlock(t); -- continue; -- } -- tasklet_unlock(t); -+ /* -+ * Should always succeed - after a tasklist got on the -+ * list (after getting the SCHED bit set from 0 to 1), -+ * nothing but the tasklet softirq it got queued to can -+ * lock it: -+ */ -+ if (!tasklet_trylock(t)) { -+ WARN_ON(1); -+ continue; - } - -- local_irq_disable(); - t->next = NULL; -- *__this_cpu_read(tasklet_vec.tail) = t; -- __this_cpu_write(tasklet_vec.tail, &(t->next)); -- __raise_softirq_irqoff(TASKLET_SOFTIRQ); -- local_irq_enable(); -+ -+ /* -+ * If we cannot handle the tasklet because it's disabled, -+ * mark it as pending. tasklet_enable() will later -+ * re-schedule the tasklet. -+ */ -+ if (unlikely(atomic_read(&t->count))) { -+out_disabled: -+ /* implicit unlock: */ -+ wmb(); -+ t->state = TASKLET_STATEF_PENDING; -+ continue; -+ } -+ -+ /* -+ * After this point on the tasklet might be rescheduled -+ * on another CPU, but it can only be added to another -+ * CPU's tasklet list if we unlock the tasklet (which we -+ * dont do yet). -+ */ -+ if (!test_and_clear_bit(TASKLET_STATE_SCHED, &t->state)) -+ WARN_ON(1); -+ -+again: -+ t->func(t->data); -+ -+ /* -+ * Try to unlock the tasklet. We must use cmpxchg, because -+ * another CPU might have scheduled or disabled the tasklet. -+ * We only allow the STATE_RUN -> 0 transition here. -+ */ -+ while (!tasklet_tryunlock(t)) { -+ /* -+ * If it got disabled meanwhile, bail out: -+ */ -+ if (atomic_read(&t->count)) -+ goto out_disabled; -+ /* -+ * If it got scheduled meanwhile, re-execute -+ * the tasklet function: -+ */ -+ if (test_and_clear_bit(TASKLET_STATE_SCHED, &t->state)) -+ goto again; -+ if (!--loops) { -+ printk("hm, tasklet state: %08lx\n", t->state); -+ WARN_ON(1); -+ tasklet_unlock(t); -+ break; -+ } -+ } - } - } - -+static void tasklet_action(struct softirq_action *a) -+{ -+ struct tasklet_struct *list; -+ -+ local_irq_disable(); -+ -+ list = __this_cpu_read(tasklet_vec.head); -+ __this_cpu_write(tasklet_vec.head, NULL); -+ __this_cpu_write(tasklet_vec.tail, this_cpu_ptr(&tasklet_vec.head)); -+ -+ local_irq_enable(); -+ -+ __tasklet_action(a, list); -+} -+ - static __latent_entropy void tasklet_hi_action(struct softirq_action *a) - { - struct tasklet_struct *list; - - local_irq_disable(); -+ - list = __this_cpu_read(tasklet_hi_vec.head); - __this_cpu_write(tasklet_hi_vec.head, NULL); - __this_cpu_write(tasklet_hi_vec.tail, this_cpu_ptr(&tasklet_hi_vec.head)); -- local_irq_enable(); - -- while (list) { -- struct tasklet_struct *t = list; -- -- list = list->next; -- -- if (tasklet_trylock(t)) { -- if (!atomic_read(&t->count)) { -- if (!test_and_clear_bit(TASKLET_STATE_SCHED, -- &t->state)) -- BUG(); -- t->func(t->data); -- tasklet_unlock(t); -- continue; -- } -- tasklet_unlock(t); -- } -+ local_irq_enable(); - -- local_irq_disable(); -- t->next = NULL; -- *__this_cpu_read(tasklet_hi_vec.tail) = t; -- __this_cpu_write(tasklet_hi_vec.tail, &(t->next)); -- __raise_softirq_irqoff(HI_SOFTIRQ); -- local_irq_enable(); -- } -+ __tasklet_action(a, list); - } - - void tasklet_init(struct tasklet_struct *t, -@@ -586,7 +1083,7 @@ - - while (test_and_set_bit(TASKLET_STATE_SCHED, &t->state)) { - do { -- yield(); -+ msleep(1); - } while (test_bit(TASKLET_STATE_SCHED, &t->state)); - } - tasklet_unlock_wait(t); -@@ -660,25 +1157,26 @@ - open_softirq(HI_SOFTIRQ, tasklet_hi_action); - } - --static int ksoftirqd_should_run(unsigned int cpu) -+#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL) -+void tasklet_unlock_wait(struct tasklet_struct *t) - { -- return local_softirq_pending(); --} -- --static void run_ksoftirqd(unsigned int cpu) --{ -- local_irq_disable(); -- if (local_softirq_pending()) { -+ while (test_bit(TASKLET_STATE_RUN, &(t)->state)) { - /* -- * We can safely run softirq on inline stack, as we are not deep -- * in the task stack here. -+ * Hack for now to avoid this busy-loop: - */ -- __do_softirq(); -- local_irq_enable(); -- cond_resched_rcu_qs(); -- return; -+#ifdef CONFIG_PREEMPT_RT_FULL -+ msleep(1); -+#else -+ barrier(); -+#endif - } -- local_irq_enable(); -+} -+EXPORT_SYMBOL(tasklet_unlock_wait); -+#endif -+ -+static int ksoftirqd_should_run(unsigned int cpu) -+{ -+ return ksoftirqd_softirq_pending(); - } - - #ifdef CONFIG_HOTPLUG_CPU -@@ -745,17 +1243,31 @@ - - static struct smp_hotplug_thread softirq_threads = { - .store = &ksoftirqd, -+ .setup = ksoftirqd_set_sched_params, - .thread_should_run = ksoftirqd_should_run, - .thread_fn = run_ksoftirqd, - .thread_comm = "ksoftirqd/%u", - }; - -+#ifdef CONFIG_PREEMPT_RT_FULL -+static struct smp_hotplug_thread softirq_timer_threads = { -+ .store = &ktimer_softirqd, -+ .setup = ktimer_softirqd_set_sched_params, -+ .cleanup = ktimer_softirqd_clr_sched_params, -+ .thread_should_run = ktimer_softirqd_should_run, -+ .thread_fn = run_ksoftirqd, -+ .thread_comm = "ktimersoftd/%u", -+}; -+#endif -+ - static __init int spawn_ksoftirqd(void) - { - cpuhp_setup_state_nocalls(CPUHP_SOFTIRQ_DEAD, "softirq:dead", NULL, - takeover_tasklets); - BUG_ON(smpboot_register_percpu_thread(&softirq_threads)); -- -+#ifdef CONFIG_PREEMPT_RT_FULL -+ BUG_ON(smpboot_register_percpu_thread(&softirq_timer_threads)); -+#endif - return 0; - } - early_initcall(spawn_ksoftirqd); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/stop_machine.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/stop_machine.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/stop_machine.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/stop_machine.c 2017-04-18 17:54:26.000000000 +0200 -@@ -36,7 +36,7 @@ - struct cpu_stopper { - struct task_struct *thread; - -- spinlock_t lock; -+ raw_spinlock_t lock; - bool enabled; /* is this stopper enabled? */ - struct list_head works; /* list of pending works */ - -@@ -78,14 +78,14 @@ - unsigned long flags; - bool enabled; - -- spin_lock_irqsave(&stopper->lock, flags); -+ raw_spin_lock_irqsave(&stopper->lock, flags); - enabled = stopper->enabled; - if (enabled) - __cpu_stop_queue_work(stopper, work); - else if (work->done) - cpu_stop_signal_done(work->done); -- spin_unlock_irqrestore(&stopper->lock, flags); - -+ raw_spin_unlock_irqrestore(&stopper->lock, flags); - return enabled; - } - -@@ -231,8 +231,8 @@ - struct cpu_stopper *stopper2 = per_cpu_ptr(&cpu_stopper, cpu2); - int err; - retry: -- spin_lock_irq(&stopper1->lock); -- spin_lock_nested(&stopper2->lock, SINGLE_DEPTH_NESTING); -+ raw_spin_lock_irq(&stopper1->lock); -+ raw_spin_lock_nested(&stopper2->lock, SINGLE_DEPTH_NESTING); - - err = -ENOENT; - if (!stopper1->enabled || !stopper2->enabled) -@@ -255,8 +255,8 @@ - __cpu_stop_queue_work(stopper1, work1); - __cpu_stop_queue_work(stopper2, work2); - unlock: -- spin_unlock(&stopper2->lock); -- spin_unlock_irq(&stopper1->lock); -+ raw_spin_unlock(&stopper2->lock); -+ raw_spin_unlock_irq(&stopper1->lock); - - if (unlikely(err == -EDEADLK)) { - while (stop_cpus_in_progress) -@@ -448,9 +448,9 @@ - unsigned long flags; - int run; - -- spin_lock_irqsave(&stopper->lock, flags); -+ raw_spin_lock_irqsave(&stopper->lock, flags); - run = !list_empty(&stopper->works); -- spin_unlock_irqrestore(&stopper->lock, flags); -+ raw_spin_unlock_irqrestore(&stopper->lock, flags); - return run; - } - -@@ -461,13 +461,13 @@ - - repeat: - work = NULL; -- spin_lock_irq(&stopper->lock); -+ raw_spin_lock_irq(&stopper->lock); - if (!list_empty(&stopper->works)) { - work = list_first_entry(&stopper->works, - struct cpu_stop_work, list); - list_del_init(&work->list); - } -- spin_unlock_irq(&stopper->lock); -+ raw_spin_unlock_irq(&stopper->lock); - - if (work) { - cpu_stop_fn_t fn = work->fn; -@@ -475,6 +475,8 @@ - struct cpu_stop_done *done = work->done; - int ret; - -+ /* XXX */ -+ - /* cpu stop callbacks must not sleep, make in_atomic() == T */ - preempt_count_inc(); - ret = fn(arg); -@@ -541,7 +543,7 @@ - for_each_possible_cpu(cpu) { - struct cpu_stopper *stopper = &per_cpu(cpu_stopper, cpu); - -- spin_lock_init(&stopper->lock); -+ raw_spin_lock_init(&stopper->lock); - INIT_LIST_HEAD(&stopper->works); - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/hrtimer.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/hrtimer.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/hrtimer.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/hrtimer.c 2017-04-18 17:54:26.000000000 +0200 -@@ -53,6 +53,7 @@ - #include - - #include -+#include - - #include "tick-internal.h" - -@@ -695,6 +696,29 @@ - retrigger_next_event(NULL); - } - -+#ifdef CONFIG_PREEMPT_RT_FULL -+ -+static struct swork_event clock_set_delay_work; -+ -+static void run_clock_set_delay(struct swork_event *event) -+{ -+ clock_was_set(); -+} -+ -+void clock_was_set_delayed(void) -+{ -+ swork_queue(&clock_set_delay_work); -+} -+ -+static __init int create_clock_set_delay_thread(void) -+{ -+ WARN_ON(swork_get()); -+ INIT_SWORK(&clock_set_delay_work, run_clock_set_delay); -+ return 0; -+} -+early_initcall(create_clock_set_delay_thread); -+#else /* PREEMPT_RT_FULL */ -+ - static void clock_was_set_work(struct work_struct *work) - { - clock_was_set(); -@@ -710,6 +734,7 @@ - { - schedule_work(&hrtimer_work); - } -+#endif - - #else - -@@ -719,11 +744,8 @@ - static inline void hrtimer_switch_to_hres(void) { } - static inline void - hrtimer_force_reprogram(struct hrtimer_cpu_base *base, int skip_equal) { } --static inline int hrtimer_reprogram(struct hrtimer *timer, -- struct hrtimer_clock_base *base) --{ -- return 0; --} -+static inline void hrtimer_reprogram(struct hrtimer *timer, -+ struct hrtimer_clock_base *base) { } - static inline void hrtimer_init_hres(struct hrtimer_cpu_base *base) { } - static inline void retrigger_next_event(void *arg) { } - -@@ -855,6 +877,32 @@ - } - EXPORT_SYMBOL_GPL(hrtimer_forward); - -+#ifdef CONFIG_PREEMPT_RT_BASE -+# define wake_up_timer_waiters(b) wake_up(&(b)->wait) -+ -+/** -+ * hrtimer_wait_for_timer - Wait for a running timer -+ * -+ * @timer: timer to wait for -+ * -+ * The function waits in case the timers callback function is -+ * currently executed on the waitqueue of the timer base. The -+ * waitqueue is woken up after the timer callback function has -+ * finished execution. -+ */ -+void hrtimer_wait_for_timer(const struct hrtimer *timer) -+{ -+ struct hrtimer_clock_base *base = timer->base; -+ -+ if (base && base->cpu_base && !timer->irqsafe) -+ wait_event(base->cpu_base->wait, -+ !(hrtimer_callback_running(timer))); -+} -+ -+#else -+# define wake_up_timer_waiters(b) do { } while (0) -+#endif -+ - /* - * enqueue_hrtimer - internal function to (re)start a timer - * -@@ -896,6 +944,11 @@ - if (!(state & HRTIMER_STATE_ENQUEUED)) - return; - -+ if (unlikely(!list_empty(&timer->cb_entry))) { -+ list_del_init(&timer->cb_entry); -+ return; -+ } -+ - if (!timerqueue_del(&base->active, &timer->node)) - cpu_base->active_bases &= ~(1 << base->index); - -@@ -991,7 +1044,16 @@ - new_base = switch_hrtimer_base(timer, base, mode & HRTIMER_MODE_PINNED); - - timer_stats_hrtimer_set_start_info(timer); -+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST -+ { -+ ktime_t now = new_base->get_time(); - -+ if (ktime_to_ns(tim) < ktime_to_ns(now)) -+ timer->praecox = now; -+ else -+ timer->praecox = ktime_set(0, 0); -+ } -+#endif - leftmost = enqueue_hrtimer(timer, new_base); - if (!leftmost) - goto unlock; -@@ -1063,7 +1125,7 @@ - - if (ret >= 0) - return ret; -- cpu_relax(); -+ hrtimer_wait_for_timer(timer); - } - } - EXPORT_SYMBOL_GPL(hrtimer_cancel); -@@ -1127,6 +1189,7 @@ - - base = hrtimer_clockid_to_base(clock_id); - timer->base = &cpu_base->clock_base[base]; -+ INIT_LIST_HEAD(&timer->cb_entry); - timerqueue_init(&timer->node); - - #ifdef CONFIG_TIMER_STATS -@@ -1167,6 +1230,7 @@ - seq = raw_read_seqcount_begin(&cpu_base->seq); - - if (timer->state != HRTIMER_STATE_INACTIVE || -+ cpu_base->running_soft == timer || - cpu_base->running == timer) - return true; - -@@ -1265,10 +1329,112 @@ - cpu_base->running = NULL; - } - -+#ifdef CONFIG_PREEMPT_RT_BASE -+static void hrtimer_rt_reprogram(int restart, struct hrtimer *timer, -+ struct hrtimer_clock_base *base) -+{ -+ int leftmost; -+ -+ if (restart != HRTIMER_NORESTART && -+ !(timer->state & HRTIMER_STATE_ENQUEUED)) { -+ -+ leftmost = enqueue_hrtimer(timer, base); -+ if (!leftmost) -+ return; -+#ifdef CONFIG_HIGH_RES_TIMERS -+ if (!hrtimer_is_hres_active(timer)) { -+ /* -+ * Kick to reschedule the next tick to handle the new timer -+ * on dynticks target. -+ */ -+ if (base->cpu_base->nohz_active) -+ wake_up_nohz_cpu(base->cpu_base->cpu); -+ } else { -+ -+ hrtimer_reprogram(timer, base); -+ } -+#endif -+ } -+} -+ -+/* -+ * The changes in mainline which removed the callback modes from -+ * hrtimer are not yet working with -rt. The non wakeup_process() -+ * based callbacks which involve sleeping locks need to be treated -+ * seperately. -+ */ -+static void hrtimer_rt_run_pending(void) -+{ -+ enum hrtimer_restart (*fn)(struct hrtimer *); -+ struct hrtimer_cpu_base *cpu_base; -+ struct hrtimer_clock_base *base; -+ struct hrtimer *timer; -+ int index, restart; -+ -+ local_irq_disable(); -+ cpu_base = &per_cpu(hrtimer_bases, smp_processor_id()); -+ -+ raw_spin_lock(&cpu_base->lock); -+ -+ for (index = 0; index < HRTIMER_MAX_CLOCK_BASES; index++) { -+ base = &cpu_base->clock_base[index]; -+ -+ while (!list_empty(&base->expired)) { -+ timer = list_first_entry(&base->expired, -+ struct hrtimer, cb_entry); -+ -+ /* -+ * Same as the above __run_hrtimer function -+ * just we run with interrupts enabled. -+ */ -+ debug_deactivate(timer); -+ cpu_base->running_soft = timer; -+ raw_write_seqcount_barrier(&cpu_base->seq); -+ -+ __remove_hrtimer(timer, base, HRTIMER_STATE_INACTIVE, 0); -+ timer_stats_account_hrtimer(timer); -+ fn = timer->function; -+ -+ raw_spin_unlock_irq(&cpu_base->lock); -+ restart = fn(timer); -+ raw_spin_lock_irq(&cpu_base->lock); -+ -+ hrtimer_rt_reprogram(restart, timer, base); -+ raw_write_seqcount_barrier(&cpu_base->seq); -+ -+ WARN_ON_ONCE(cpu_base->running_soft != timer); -+ cpu_base->running_soft = NULL; -+ } -+ } -+ -+ raw_spin_unlock_irq(&cpu_base->lock); -+ -+ wake_up_timer_waiters(cpu_base); -+} -+ -+static int hrtimer_rt_defer(struct hrtimer *timer) -+{ -+ if (timer->irqsafe) -+ return 0; -+ -+ __remove_hrtimer(timer, timer->base, timer->state, 0); -+ list_add_tail(&timer->cb_entry, &timer->base->expired); -+ return 1; -+} -+ -+#else -+ -+static inline int hrtimer_rt_defer(struct hrtimer *timer) { return 0; } -+ -+#endif -+ -+static enum hrtimer_restart hrtimer_wakeup(struct hrtimer *timer); -+ - static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now) - { - struct hrtimer_clock_base *base = cpu_base->clock_base; - unsigned int active = cpu_base->active_bases; -+ int raise = 0; - - for (; active; base++, active >>= 1) { - struct timerqueue_node *node; -@@ -1284,6 +1450,15 @@ - - timer = container_of(node, struct hrtimer, node); - -+ trace_hrtimer_interrupt(raw_smp_processor_id(), -+ ktime_to_ns(ktime_sub(ktime_to_ns(timer->praecox) ? -+ timer->praecox : hrtimer_get_expires(timer), -+ basenow)), -+ current, -+ timer->function == hrtimer_wakeup ? -+ container_of(timer, struct hrtimer_sleeper, -+ timer)->task : NULL); -+ - /* - * The immediate goal for using the softexpires is - * minimizing wakeups, not running timers at the -@@ -1299,9 +1474,14 @@ - if (basenow.tv64 < hrtimer_get_softexpires_tv64(timer)) - break; - -- __run_hrtimer(cpu_base, base, timer, &basenow); -+ if (!hrtimer_rt_defer(timer)) -+ __run_hrtimer(cpu_base, base, timer, &basenow); -+ else -+ raise = 1; - } - } -+ if (raise) -+ raise_softirq_irqoff(HRTIMER_SOFTIRQ); - } - - #ifdef CONFIG_HIGH_RES_TIMERS -@@ -1464,16 +1644,18 @@ - void hrtimer_init_sleeper(struct hrtimer_sleeper *sl, struct task_struct *task) - { - sl->timer.function = hrtimer_wakeup; -+ sl->timer.irqsafe = 1; - sl->task = task; - } - EXPORT_SYMBOL_GPL(hrtimer_init_sleeper); - --static int __sched do_nanosleep(struct hrtimer_sleeper *t, enum hrtimer_mode mode) -+static int __sched do_nanosleep(struct hrtimer_sleeper *t, enum hrtimer_mode mode, -+ unsigned long state) - { - hrtimer_init_sleeper(t, current); - - do { -- set_current_state(TASK_INTERRUPTIBLE); -+ set_current_state(state); - hrtimer_start_expires(&t->timer, mode); - - if (likely(t->task)) -@@ -1515,7 +1697,8 @@ - HRTIMER_MODE_ABS); - hrtimer_set_expires_tv64(&t.timer, restart->nanosleep.expires); - -- if (do_nanosleep(&t, HRTIMER_MODE_ABS)) -+ /* cpu_chill() does not care about restart state. */ -+ if (do_nanosleep(&t, HRTIMER_MODE_ABS, TASK_INTERRUPTIBLE)) - goto out; - - rmtp = restart->nanosleep.rmtp; -@@ -1532,8 +1715,10 @@ - return ret; - } - --long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, -- const enum hrtimer_mode mode, const clockid_t clockid) -+static long -+__hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, -+ const enum hrtimer_mode mode, const clockid_t clockid, -+ unsigned long state) - { - struct restart_block *restart; - struct hrtimer_sleeper t; -@@ -1546,7 +1731,7 @@ - - hrtimer_init_on_stack(&t.timer, clockid, mode); - hrtimer_set_expires_range_ns(&t.timer, timespec_to_ktime(*rqtp), slack); -- if (do_nanosleep(&t, mode)) -+ if (do_nanosleep(&t, mode, state)) - goto out; - - /* Absolute timers do not update the rmtp value and restart: */ -@@ -1573,6 +1758,12 @@ - return ret; - } - -+long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, -+ const enum hrtimer_mode mode, const clockid_t clockid) -+{ -+ return __hrtimer_nanosleep(rqtp, rmtp, mode, clockid, TASK_INTERRUPTIBLE); -+} -+ - SYSCALL_DEFINE2(nanosleep, struct timespec __user *, rqtp, - struct timespec __user *, rmtp) - { -@@ -1587,6 +1778,26 @@ - return hrtimer_nanosleep(&tu, rmtp, HRTIMER_MODE_REL, CLOCK_MONOTONIC); - } - -+#ifdef CONFIG_PREEMPT_RT_FULL -+/* -+ * Sleep for 1 ms in hope whoever holds what we want will let it go. -+ */ -+void cpu_chill(void) -+{ -+ struct timespec tu = { -+ .tv_nsec = NSEC_PER_MSEC, -+ }; -+ unsigned int freeze_flag = current->flags & PF_NOFREEZE; -+ -+ current->flags |= PF_NOFREEZE; -+ __hrtimer_nanosleep(&tu, NULL, HRTIMER_MODE_REL, CLOCK_MONOTONIC, -+ TASK_UNINTERRUPTIBLE); -+ if (!freeze_flag) -+ current->flags &= ~PF_NOFREEZE; -+} -+EXPORT_SYMBOL(cpu_chill); -+#endif -+ - /* - * Functions related to boot-time initialization: - */ -@@ -1598,10 +1809,14 @@ - for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) { - cpu_base->clock_base[i].cpu_base = cpu_base; - timerqueue_init_head(&cpu_base->clock_base[i].active); -+ INIT_LIST_HEAD(&cpu_base->clock_base[i].expired); - } - - cpu_base->cpu = cpu; - hrtimer_init_hres(cpu_base); -+#ifdef CONFIG_PREEMPT_RT_BASE -+ init_waitqueue_head(&cpu_base->wait); -+#endif - return 0; - } - -@@ -1671,9 +1886,26 @@ - - #endif /* CONFIG_HOTPLUG_CPU */ - -+#ifdef CONFIG_PREEMPT_RT_BASE -+ -+static void run_hrtimer_softirq(struct softirq_action *h) -+{ -+ hrtimer_rt_run_pending(); -+} -+ -+static void hrtimers_open_softirq(void) -+{ -+ open_softirq(HRTIMER_SOFTIRQ, run_hrtimer_softirq); -+} -+ -+#else -+static void hrtimers_open_softirq(void) { } -+#endif -+ - void __init hrtimers_init(void) - { - hrtimers_prepare_cpu(smp_processor_id()); -+ hrtimers_open_softirq(); - } - - /** -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/itimer.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/itimer.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/itimer.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/itimer.c 2017-04-18 17:54:26.000000000 +0200 -@@ -213,6 +213,7 @@ - /* We are sharing ->siglock with it_real_fn() */ - if (hrtimer_try_to_cancel(timer) < 0) { - spin_unlock_irq(&tsk->sighand->siglock); -+ hrtimer_wait_for_timer(&tsk->signal->real_timer); - goto again; - } - expires = timeval_to_ktime(value->it_value); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/jiffies.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/jiffies.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/jiffies.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/jiffies.c 2017-04-18 17:54:26.000000000 +0200 -@@ -74,7 +74,8 @@ - .max_cycles = 10, - }; - --__cacheline_aligned_in_smp DEFINE_SEQLOCK(jiffies_lock); -+__cacheline_aligned_in_smp DEFINE_RAW_SPINLOCK(jiffies_lock); -+__cacheline_aligned_in_smp seqcount_t jiffies_seq; - - #if (BITS_PER_LONG < 64) - u64 get_jiffies_64(void) -@@ -83,9 +84,9 @@ - u64 ret; - - do { -- seq = read_seqbegin(&jiffies_lock); -+ seq = read_seqcount_begin(&jiffies_seq); - ret = jiffies_64; -- } while (read_seqretry(&jiffies_lock, seq)); -+ } while (read_seqcount_retry(&jiffies_seq, seq)); - return ret; - } - EXPORT_SYMBOL(get_jiffies_64); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/ntp.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/ntp.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/ntp.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/ntp.c 2017-04-18 17:54:26.000000000 +0200 -@@ -17,6 +17,7 @@ - #include - #include - #include -+#include - - #include "ntp_internal.h" - #include "timekeeping_internal.h" -@@ -568,10 +569,35 @@ - &sync_cmos_work, timespec64_to_jiffies(&next)); - } - -+#ifdef CONFIG_PREEMPT_RT_FULL -+ -+static void run_clock_set_delay(struct swork_event *event) -+{ -+ queue_delayed_work(system_power_efficient_wq, &sync_cmos_work, 0); -+} -+ -+static struct swork_event ntp_cmos_swork; -+ -+void ntp_notify_cmos_timer(void) -+{ -+ swork_queue(&ntp_cmos_swork); -+} -+ -+static __init int create_cmos_delay_thread(void) -+{ -+ WARN_ON(swork_get()); -+ INIT_SWORK(&ntp_cmos_swork, run_clock_set_delay); -+ return 0; -+} -+early_initcall(create_cmos_delay_thread); -+ -+#else -+ - void ntp_notify_cmos_timer(void) - { - queue_delayed_work(system_power_efficient_wq, &sync_cmos_work, 0); - } -+#endif /* CONFIG_PREEMPT_RT_FULL */ - - #else - void ntp_notify_cmos_timer(void) { } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/posix-cpu-timers.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/posix-cpu-timers.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/posix-cpu-timers.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/posix-cpu-timers.c 2017-04-18 17:54:26.000000000 +0200 -@@ -3,6 +3,7 @@ - */ - - #include -+#include - #include - #include - #include -@@ -620,7 +621,7 @@ - /* - * Disarm any old timer after extracting its expiry time. - */ -- WARN_ON_ONCE(!irqs_disabled()); -+ WARN_ON_ONCE_NONRT(!irqs_disabled()); - - ret = 0; - old_incr = timer->it.cpu.incr; -@@ -1064,7 +1065,7 @@ - /* - * Now re-arm for the new expiry time. - */ -- WARN_ON_ONCE(!irqs_disabled()); -+ WARN_ON_ONCE_NONRT(!irqs_disabled()); - arm_timer(timer); - unlock_task_sighand(p, &flags); - -@@ -1153,13 +1154,13 @@ - * already updated our counts. We need to check if any timers fire now. - * Interrupts are disabled. - */ --void run_posix_cpu_timers(struct task_struct *tsk) -+static void __run_posix_cpu_timers(struct task_struct *tsk) - { - LIST_HEAD(firing); - struct k_itimer *timer, *next; - unsigned long flags; - -- WARN_ON_ONCE(!irqs_disabled()); -+ WARN_ON_ONCE_NONRT(!irqs_disabled()); - - /* - * The fast path checks that there are no expired thread or thread -@@ -1213,6 +1214,190 @@ - } - } - -+#ifdef CONFIG_PREEMPT_RT_BASE -+#include -+#include -+DEFINE_PER_CPU(struct task_struct *, posix_timer_task); -+DEFINE_PER_CPU(struct task_struct *, posix_timer_tasklist); -+ -+static int posix_cpu_timers_thread(void *data) -+{ -+ int cpu = (long)data; -+ -+ BUG_ON(per_cpu(posix_timer_task,cpu) != current); -+ -+ while (!kthread_should_stop()) { -+ struct task_struct *tsk = NULL; -+ struct task_struct *next = NULL; -+ -+ if (cpu_is_offline(cpu)) -+ goto wait_to_die; -+ -+ /* grab task list */ -+ raw_local_irq_disable(); -+ tsk = per_cpu(posix_timer_tasklist, cpu); -+ per_cpu(posix_timer_tasklist, cpu) = NULL; -+ raw_local_irq_enable(); -+ -+ /* its possible the list is empty, just return */ -+ if (!tsk) { -+ set_current_state(TASK_INTERRUPTIBLE); -+ schedule(); -+ __set_current_state(TASK_RUNNING); -+ continue; -+ } -+ -+ /* Process task list */ -+ while (1) { -+ /* save next */ -+ next = tsk->posix_timer_list; -+ -+ /* run the task timers, clear its ptr and -+ * unreference it -+ */ -+ __run_posix_cpu_timers(tsk); -+ tsk->posix_timer_list = NULL; -+ put_task_struct(tsk); -+ -+ /* check if this is the last on the list */ -+ if (next == tsk) -+ break; -+ tsk = next; -+ } -+ } -+ return 0; -+ -+wait_to_die: -+ /* Wait for kthread_stop */ -+ set_current_state(TASK_INTERRUPTIBLE); -+ while (!kthread_should_stop()) { -+ schedule(); -+ set_current_state(TASK_INTERRUPTIBLE); -+ } -+ __set_current_state(TASK_RUNNING); -+ return 0; -+} -+ -+static inline int __fastpath_timer_check(struct task_struct *tsk) -+{ -+ /* tsk == current, ensure it is safe to use ->signal/sighand */ -+ if (unlikely(tsk->exit_state)) -+ return 0; -+ -+ if (!task_cputime_zero(&tsk->cputime_expires)) -+ return 1; -+ -+ if (!task_cputime_zero(&tsk->signal->cputime_expires)) -+ return 1; -+ -+ return 0; -+} -+ -+void run_posix_cpu_timers(struct task_struct *tsk) -+{ -+ unsigned long cpu = smp_processor_id(); -+ struct task_struct *tasklist; -+ -+ BUG_ON(!irqs_disabled()); -+ if(!per_cpu(posix_timer_task, cpu)) -+ return; -+ /* get per-cpu references */ -+ tasklist = per_cpu(posix_timer_tasklist, cpu); -+ -+ /* check to see if we're already queued */ -+ if (!tsk->posix_timer_list && __fastpath_timer_check(tsk)) { -+ get_task_struct(tsk); -+ if (tasklist) { -+ tsk->posix_timer_list = tasklist; -+ } else { -+ /* -+ * The list is terminated by a self-pointing -+ * task_struct -+ */ -+ tsk->posix_timer_list = tsk; -+ } -+ per_cpu(posix_timer_tasklist, cpu) = tsk; -+ -+ wake_up_process(per_cpu(posix_timer_task, cpu)); -+ } -+} -+ -+/* -+ * posix_cpu_thread_call - callback that gets triggered when a CPU is added. -+ * Here we can start up the necessary migration thread for the new CPU. -+ */ -+static int posix_cpu_thread_call(struct notifier_block *nfb, -+ unsigned long action, void *hcpu) -+{ -+ int cpu = (long)hcpu; -+ struct task_struct *p; -+ struct sched_param param; -+ -+ switch (action) { -+ case CPU_UP_PREPARE: -+ p = kthread_create(posix_cpu_timers_thread, hcpu, -+ "posixcputmr/%d",cpu); -+ if (IS_ERR(p)) -+ return NOTIFY_BAD; -+ p->flags |= PF_NOFREEZE; -+ kthread_bind(p, cpu); -+ /* Must be high prio to avoid getting starved */ -+ param.sched_priority = MAX_RT_PRIO-1; -+ sched_setscheduler(p, SCHED_FIFO, ¶m); -+ per_cpu(posix_timer_task,cpu) = p; -+ break; -+ case CPU_ONLINE: -+ /* Strictly unneccessary, as first user will wake it. */ -+ wake_up_process(per_cpu(posix_timer_task,cpu)); -+ break; -+#ifdef CONFIG_HOTPLUG_CPU -+ case CPU_UP_CANCELED: -+ /* Unbind it from offline cpu so it can run. Fall thru. */ -+ kthread_bind(per_cpu(posix_timer_task, cpu), -+ cpumask_any(cpu_online_mask)); -+ kthread_stop(per_cpu(posix_timer_task,cpu)); -+ per_cpu(posix_timer_task,cpu) = NULL; -+ break; -+ case CPU_DEAD: -+ kthread_stop(per_cpu(posix_timer_task,cpu)); -+ per_cpu(posix_timer_task,cpu) = NULL; -+ break; -+#endif -+ } -+ return NOTIFY_OK; -+} -+ -+/* Register at highest priority so that task migration (migrate_all_tasks) -+ * happens before everything else. -+ */ -+static struct notifier_block posix_cpu_thread_notifier = { -+ .notifier_call = posix_cpu_thread_call, -+ .priority = 10 -+}; -+ -+static int __init posix_cpu_thread_init(void) -+{ -+ void *hcpu = (void *)(long)smp_processor_id(); -+ /* Start one for boot CPU. */ -+ unsigned long cpu; -+ -+ /* init the per-cpu posix_timer_tasklets */ -+ for_each_possible_cpu(cpu) -+ per_cpu(posix_timer_tasklist, cpu) = NULL; -+ -+ posix_cpu_thread_call(&posix_cpu_thread_notifier, CPU_UP_PREPARE, hcpu); -+ posix_cpu_thread_call(&posix_cpu_thread_notifier, CPU_ONLINE, hcpu); -+ register_cpu_notifier(&posix_cpu_thread_notifier); -+ return 0; -+} -+early_initcall(posix_cpu_thread_init); -+#else /* CONFIG_PREEMPT_RT_BASE */ -+void run_posix_cpu_timers(struct task_struct *tsk) -+{ -+ __run_posix_cpu_timers(tsk); -+} -+#endif /* CONFIG_PREEMPT_RT_BASE */ -+ - /* - * Set one of the process-wide special case CPU timers or RLIMIT_CPU. - * The tsk->sighand->siglock must be held by the caller. -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/posix-timers.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/posix-timers.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/posix-timers.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/posix-timers.c 2017-04-18 17:54:26.000000000 +0200 -@@ -506,6 +506,7 @@ - static struct pid *good_sigevent(sigevent_t * event) - { - struct task_struct *rtn = current->group_leader; -+ int sig = event->sigev_signo; - - if ((event->sigev_notify & SIGEV_THREAD_ID ) && - (!(rtn = find_task_by_vpid(event->sigev_notify_thread_id)) || -@@ -514,7 +515,8 @@ - return NULL; - - if (((event->sigev_notify & ~SIGEV_THREAD_ID) != SIGEV_NONE) && -- ((event->sigev_signo <= 0) || (event->sigev_signo > SIGRTMAX))) -+ (sig <= 0 || sig > SIGRTMAX || sig_kernel_only(sig) || -+ sig_kernel_coredump(sig))) - return NULL; - - return task_pid(rtn); -@@ -826,6 +828,20 @@ - return overrun; - } - -+/* -+ * Protected by RCU! -+ */ -+static void timer_wait_for_callback(struct k_clock *kc, struct k_itimer *timr) -+{ -+#ifdef CONFIG_PREEMPT_RT_FULL -+ if (kc->timer_set == common_timer_set) -+ hrtimer_wait_for_timer(&timr->it.real.timer); -+ else -+ /* FIXME: Whacky hack for posix-cpu-timers */ -+ schedule_timeout(1); -+#endif -+} -+ - /* Set a POSIX.1b interval timer. */ - /* timr->it_lock is taken. */ - static int -@@ -903,6 +919,7 @@ - if (!timr) - return -EINVAL; - -+ rcu_read_lock(); - kc = clockid_to_kclock(timr->it_clock); - if (WARN_ON_ONCE(!kc || !kc->timer_set)) - error = -EINVAL; -@@ -911,9 +928,12 @@ - - unlock_timer(timr, flag); - if (error == TIMER_RETRY) { -+ timer_wait_for_callback(kc, timr); - rtn = NULL; // We already got the old time... -+ rcu_read_unlock(); - goto retry; - } -+ rcu_read_unlock(); - - if (old_setting && !error && - copy_to_user(old_setting, &old_spec, sizeof (old_spec))) -@@ -951,10 +971,15 @@ - if (!timer) - return -EINVAL; - -+ rcu_read_lock(); - if (timer_delete_hook(timer) == TIMER_RETRY) { - unlock_timer(timer, flags); -+ timer_wait_for_callback(clockid_to_kclock(timer->it_clock), -+ timer); -+ rcu_read_unlock(); - goto retry_delete; - } -+ rcu_read_unlock(); - - spin_lock(¤t->sighand->siglock); - list_del(&timer->list); -@@ -980,8 +1005,18 @@ - retry_delete: - spin_lock_irqsave(&timer->it_lock, flags); - -+ /* On RT we can race with a deletion */ -+ if (!timer->it_signal) { -+ unlock_timer(timer, flags); -+ return; -+ } -+ - if (timer_delete_hook(timer) == TIMER_RETRY) { -+ rcu_read_lock(); - unlock_timer(timer, flags); -+ timer_wait_for_callback(clockid_to_kclock(timer->it_clock), -+ timer); -+ rcu_read_unlock(); - goto retry_delete; - } - list_del(&timer->list); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/tick-broadcast-hrtimer.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/tick-broadcast-hrtimer.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/tick-broadcast-hrtimer.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/tick-broadcast-hrtimer.c 2017-04-18 17:54:26.000000000 +0200 -@@ -107,5 +107,6 @@ - { - hrtimer_init(&bctimer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); - bctimer.function = bc_handler; -+ bctimer.irqsafe = true; - clockevents_register_device(&ce_broadcast_hrtimer); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/tick-common.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/tick-common.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/tick-common.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/tick-common.c 2017-04-18 17:54:26.000000000 +0200 -@@ -79,13 +79,15 @@ - static void tick_periodic(int cpu) - { - if (tick_do_timer_cpu == cpu) { -- write_seqlock(&jiffies_lock); -+ raw_spin_lock(&jiffies_lock); -+ write_seqcount_begin(&jiffies_seq); - - /* Keep track of the next tick event */ - tick_next_period = ktime_add(tick_next_period, tick_period); - - do_timer(1); -- write_sequnlock(&jiffies_lock); -+ write_seqcount_end(&jiffies_seq); -+ raw_spin_unlock(&jiffies_lock); - update_wall_time(); - } - -@@ -157,9 +159,9 @@ - ktime_t next; - - do { -- seq = read_seqbegin(&jiffies_lock); -+ seq = read_seqcount_begin(&jiffies_seq); - next = tick_next_period; -- } while (read_seqretry(&jiffies_lock, seq)); -+ } while (read_seqcount_retry(&jiffies_seq, seq)); - - clockevents_switch_state(dev, CLOCK_EVT_STATE_ONESHOT); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/tick-sched.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/tick-sched.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/tick-sched.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/tick-sched.c 2017-04-18 17:54:26.000000000 +0200 -@@ -62,7 +62,8 @@ - return; - - /* Reevaluate with jiffies_lock held */ -- write_seqlock(&jiffies_lock); -+ raw_spin_lock(&jiffies_lock); -+ write_seqcount_begin(&jiffies_seq); - - delta = ktime_sub(now, last_jiffies_update); - if (delta.tv64 >= tick_period.tv64) { -@@ -85,10 +86,12 @@ - /* Keep the tick_next_period variable up to date */ - tick_next_period = ktime_add(last_jiffies_update, tick_period); - } else { -- write_sequnlock(&jiffies_lock); -+ write_seqcount_end(&jiffies_seq); -+ raw_spin_unlock(&jiffies_lock); - return; - } -- write_sequnlock(&jiffies_lock); -+ write_seqcount_end(&jiffies_seq); -+ raw_spin_unlock(&jiffies_lock); - update_wall_time(); - } - -@@ -99,12 +102,14 @@ - { - ktime_t period; - -- write_seqlock(&jiffies_lock); -+ raw_spin_lock(&jiffies_lock); -+ write_seqcount_begin(&jiffies_seq); - /* Did we start the jiffies update yet ? */ - if (last_jiffies_update.tv64 == 0) - last_jiffies_update = tick_next_period; - period = last_jiffies_update; -- write_sequnlock(&jiffies_lock); -+ write_seqcount_end(&jiffies_seq); -+ raw_spin_unlock(&jiffies_lock); - return period; - } - -@@ -215,6 +220,7 @@ - - static DEFINE_PER_CPU(struct irq_work, nohz_full_kick_work) = { - .func = nohz_full_kick_func, -+ .flags = IRQ_WORK_HARD_IRQ, - }; - - /* -@@ -673,10 +679,10 @@ - - /* Read jiffies and the time when jiffies were updated last */ - do { -- seq = read_seqbegin(&jiffies_lock); -+ seq = read_seqcount_begin(&jiffies_seq); - basemono = last_jiffies_update.tv64; - basejiff = jiffies; -- } while (read_seqretry(&jiffies_lock, seq)); -+ } while (read_seqcount_retry(&jiffies_seq, seq)); - ts->last_jiffies = basejiff; - - if (rcu_needs_cpu(basemono, &next_rcu) || -@@ -877,14 +883,7 @@ - return false; - - if (unlikely(local_softirq_pending() && cpu_online(cpu))) { -- static int ratelimit; -- -- if (ratelimit < 10 && -- (local_softirq_pending() & SOFTIRQ_STOP_IDLE_MASK)) { -- pr_warn("NOHZ: local_softirq_pending %02x\n", -- (unsigned int) local_softirq_pending()); -- ratelimit++; -- } -+ softirq_check_pending_idle(); - return false; - } - -@@ -1193,6 +1192,7 @@ - * Emulate tick processing via per-CPU hrtimers: - */ - hrtimer_init(&ts->sched_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); -+ ts->sched_timer.irqsafe = 1; - ts->sched_timer.function = tick_sched_timer; - - /* Get the next period (per-CPU) */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/timekeeping.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/timekeeping.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/timekeeping.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/timekeeping.c 2017-04-18 17:54:26.000000000 +0200 -@@ -2328,8 +2328,10 @@ - */ - void xtime_update(unsigned long ticks) - { -- write_seqlock(&jiffies_lock); -+ raw_spin_lock(&jiffies_lock); -+ write_seqcount_begin(&jiffies_seq); - do_timer(ticks); -- write_sequnlock(&jiffies_lock); -+ write_seqcount_end(&jiffies_seq); -+ raw_spin_unlock(&jiffies_lock); - update_wall_time(); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/timekeeping.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/timekeeping.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/timekeeping.h 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/timekeeping.h 2017-04-18 17:54:26.000000000 +0200 -@@ -19,7 +19,8 @@ - extern void do_timer(unsigned long ticks); - extern void update_wall_time(void); - --extern seqlock_t jiffies_lock; -+extern raw_spinlock_t jiffies_lock; -+extern seqcount_t jiffies_seq; - - #define CS_NAME_LEN 32 - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/timer.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/timer.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/time/timer.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/time/timer.c 2017-04-18 17:54:26.000000000 +0200 -@@ -193,8 +193,11 @@ - #endif - - struct timer_base { -- spinlock_t lock; -+ raw_spinlock_t lock; - struct timer_list *running_timer; -+#ifdef CONFIG_PREEMPT_RT_FULL -+ struct swait_queue_head wait_for_running_timer; -+#endif - unsigned long clk; - unsigned long next_expiry; - unsigned int cpu; -@@ -203,6 +206,8 @@ - bool is_idle; - DECLARE_BITMAP(pending_map, WHEEL_SIZE); - struct hlist_head vectors[WHEEL_SIZE]; -+ struct hlist_head expired_lists[LVL_DEPTH]; -+ int expired_count; - } ____cacheline_aligned; - - static DEFINE_PER_CPU(struct timer_base, timer_bases[NR_BASES]); -@@ -948,10 +953,10 @@ - - if (!(tf & TIMER_MIGRATING)) { - base = get_timer_base(tf); -- spin_lock_irqsave(&base->lock, *flags); -+ raw_spin_lock_irqsave(&base->lock, *flags); - if (timer->flags == tf) - return base; -- spin_unlock_irqrestore(&base->lock, *flags); -+ raw_spin_unlock_irqrestore(&base->lock, *flags); - } - cpu_relax(); - } -@@ -1023,9 +1028,9 @@ - /* See the comment in lock_timer_base() */ - timer->flags |= TIMER_MIGRATING; - -- spin_unlock(&base->lock); -+ raw_spin_unlock(&base->lock); - base = new_base; -- spin_lock(&base->lock); -+ raw_spin_lock(&base->lock); - WRITE_ONCE(timer->flags, - (timer->flags & ~TIMER_BASEMASK) | base->cpu); - } -@@ -1050,7 +1055,7 @@ - } - - out_unlock: -- spin_unlock_irqrestore(&base->lock, flags); -+ raw_spin_unlock_irqrestore(&base->lock, flags); - - return ret; - } -@@ -1144,19 +1149,46 @@ - if (base != new_base) { - timer->flags |= TIMER_MIGRATING; - -- spin_unlock(&base->lock); -+ raw_spin_unlock(&base->lock); - base = new_base; -- spin_lock(&base->lock); -+ raw_spin_lock(&base->lock); - WRITE_ONCE(timer->flags, - (timer->flags & ~TIMER_BASEMASK) | cpu); - } - - debug_activate(timer, timer->expires); - internal_add_timer(base, timer); -- spin_unlock_irqrestore(&base->lock, flags); -+ raw_spin_unlock_irqrestore(&base->lock, flags); - } - EXPORT_SYMBOL_GPL(add_timer_on); - -+#ifdef CONFIG_PREEMPT_RT_FULL -+/* -+ * Wait for a running timer -+ */ -+static void wait_for_running_timer(struct timer_list *timer) -+{ -+ struct timer_base *base; -+ u32 tf = timer->flags; -+ -+ if (tf & TIMER_MIGRATING) -+ return; -+ -+ base = get_timer_base(tf); -+ swait_event(base->wait_for_running_timer, -+ base->running_timer != timer); -+} -+ -+# define wakeup_timer_waiters(b) swake_up_all(&(b)->wait_for_running_timer) -+#else -+static inline void wait_for_running_timer(struct timer_list *timer) -+{ -+ cpu_relax(); -+} -+ -+# define wakeup_timer_waiters(b) do { } while (0) -+#endif -+ - /** - * del_timer - deactive a timer. - * @timer: the timer to be deactivated -@@ -1180,7 +1212,7 @@ - if (timer_pending(timer)) { - base = lock_timer_base(timer, &flags); - ret = detach_if_pending(timer, base, true); -- spin_unlock_irqrestore(&base->lock, flags); -+ raw_spin_unlock_irqrestore(&base->lock, flags); - } - - return ret; -@@ -1208,13 +1240,13 @@ - timer_stats_timer_clear_start_info(timer); - ret = detach_if_pending(timer, base, true); - } -- spin_unlock_irqrestore(&base->lock, flags); -+ raw_spin_unlock_irqrestore(&base->lock, flags); - - return ret; - } - EXPORT_SYMBOL(try_to_del_timer_sync); - --#ifdef CONFIG_SMP -+#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL) - /** - * del_timer_sync - deactivate a timer and wait for the handler to finish. - * @timer: the timer to be deactivated -@@ -1274,7 +1306,7 @@ - int ret = try_to_del_timer_sync(timer); - if (ret >= 0) - return ret; -- cpu_relax(); -+ wait_for_running_timer(timer); - } - } - EXPORT_SYMBOL(del_timer_sync); -@@ -1323,7 +1355,8 @@ - } - } - --static void expire_timers(struct timer_base *base, struct hlist_head *head) -+static inline void __expire_timers(struct timer_base *base, -+ struct hlist_head *head) - { - while (!hlist_empty(head)) { - struct timer_list *timer; -@@ -1339,33 +1372,53 @@ - fn = timer->function; - data = timer->data; - -- if (timer->flags & TIMER_IRQSAFE) { -- spin_unlock(&base->lock); -+ if (!IS_ENABLED(CONFIG_PREEMPT_RT_FULL) && -+ timer->flags & TIMER_IRQSAFE) { -+ raw_spin_unlock(&base->lock); - call_timer_fn(timer, fn, data); -- spin_lock(&base->lock); -+ base->running_timer = NULL; -+ raw_spin_lock(&base->lock); - } else { -- spin_unlock_irq(&base->lock); -+ raw_spin_unlock_irq(&base->lock); - call_timer_fn(timer, fn, data); -- spin_lock_irq(&base->lock); -+ base->running_timer = NULL; -+ raw_spin_lock_irq(&base->lock); - } - } - } - --static int __collect_expired_timers(struct timer_base *base, -- struct hlist_head *heads) -+static void expire_timers(struct timer_base *base) -+{ -+ struct hlist_head *head; -+ -+ while (base->expired_count--) { -+ head = base->expired_lists + base->expired_count; -+ __expire_timers(base, head); -+ } -+ base->expired_count = 0; -+} -+ -+static void __collect_expired_timers(struct timer_base *base) - { - unsigned long clk = base->clk; - struct hlist_head *vec; -- int i, levels = 0; -+ int i; - unsigned int idx; - -+ /* -+ * expire_timers() must be called at least once before we can -+ * collect more timers -+ */ -+ if (WARN_ON(base->expired_count)) -+ return; -+ - for (i = 0; i < LVL_DEPTH; i++) { - idx = (clk & LVL_MASK) + i * LVL_SIZE; - - if (__test_and_clear_bit(idx, base->pending_map)) { - vec = base->vectors + idx; -- hlist_move_list(vec, heads++); -- levels++; -+ hlist_move_list(vec, -+ &base->expired_lists[base->expired_count++]); - } - /* Is it time to look at the next level? */ - if (clk & LVL_CLK_MASK) -@@ -1373,7 +1426,6 @@ - /* Shift clock for the next level granularity */ - clk >>= LVL_CLK_SHIFT; - } -- return levels; - } - - #ifdef CONFIG_NO_HZ_COMMON -@@ -1515,7 +1567,7 @@ - if (cpu_is_offline(smp_processor_id())) - return expires; - -- spin_lock(&base->lock); -+ raw_spin_lock(&base->lock); - nextevt = __next_timer_interrupt(base); - is_max_delta = (nextevt == base->clk + NEXT_TIMER_MAX_DELTA); - base->next_expiry = nextevt; -@@ -1543,7 +1595,7 @@ - if ((expires - basem) > TICK_NSEC) - base->is_idle = true; - } -- spin_unlock(&base->lock); -+ raw_spin_unlock(&base->lock); - - return cmp_next_hrtimer_event(basem, expires); - } -@@ -1566,8 +1618,7 @@ - base->is_idle = false; - } - --static int collect_expired_timers(struct timer_base *base, -- struct hlist_head *heads) -+static void collect_expired_timers(struct timer_base *base) - { - /* - * NOHZ optimization. After a long idle sleep we need to forward the -@@ -1584,20 +1635,49 @@ - if (time_after(next, jiffies)) { - /* The call site will increment clock! */ - base->clk = jiffies - 1; -- return 0; -+ return; - } - base->clk = next; - } -- return __collect_expired_timers(base, heads); -+ __collect_expired_timers(base); - } - #else --static inline int collect_expired_timers(struct timer_base *base, -- struct hlist_head *heads) -+static inline void collect_expired_timers(struct timer_base *base) - { -- return __collect_expired_timers(base, heads); -+ __collect_expired_timers(base); - } - #endif - -+static int find_expired_timers(struct timer_base *base) -+{ -+ const unsigned long int end_clk = jiffies; -+ -+ while (!base->expired_count && time_after_eq(end_clk, base->clk)) { -+ collect_expired_timers(base); -+ base->clk++; -+ } -+ -+ return base->expired_count; -+} -+ -+/* Called from CPU tick routine to quickly collect expired timers */ -+static int tick_find_expired(struct timer_base *base) -+{ -+ int count; -+ -+ raw_spin_lock(&base->lock); -+ -+ if (unlikely(time_after(jiffies, base->clk + HZ))) { -+ /* defer to ktimersoftd; don't spend too long in irq context */ -+ count = -1; -+ } else -+ count = find_expired_timers(base); -+ -+ raw_spin_unlock(&base->lock); -+ -+ return count; -+} -+ - /* - * Called from the timer interrupt handler to charge one tick to the current - * process. user_tick is 1 if the tick is user time, 0 for system. -@@ -1608,13 +1688,13 @@ - - /* Note: this timer irq context must be accounted for as well. */ - account_process_tick(p, user_tick); -+ scheduler_tick(); - run_local_timers(); - rcu_check_callbacks(user_tick); --#ifdef CONFIG_IRQ_WORK -+#if defined(CONFIG_IRQ_WORK) - if (in_irq()) - irq_work_tick(); - #endif -- scheduler_tick(); - run_posix_cpu_timers(p); - } - -@@ -1624,24 +1704,13 @@ - */ - static inline void __run_timers(struct timer_base *base) - { -- struct hlist_head heads[LVL_DEPTH]; -- int levels; -- -- if (!time_after_eq(jiffies, base->clk)) -- return; -+ raw_spin_lock_irq(&base->lock); - -- spin_lock_irq(&base->lock); -+ while (find_expired_timers(base)) -+ expire_timers(base); - -- while (time_after_eq(jiffies, base->clk)) { -- -- levels = collect_expired_timers(base, heads); -- base->clk++; -- -- while (levels--) -- expire_timers(base, heads + levels); -- } -- base->running_timer = NULL; -- spin_unlock_irq(&base->lock); -+ raw_spin_unlock_irq(&base->lock); -+ wakeup_timer_waiters(base); - } - - /* -@@ -1651,6 +1720,8 @@ - { - struct timer_base *base = this_cpu_ptr(&timer_bases[BASE_STD]); - -+ irq_work_tick_soft(); -+ - __run_timers(base); - if (IS_ENABLED(CONFIG_NO_HZ_COMMON) && base->nohz_active) - __run_timers(this_cpu_ptr(&timer_bases[BASE_DEF])); -@@ -1665,12 +1736,12 @@ - - hrtimer_run_queues(); - /* Raise the softirq only if required. */ -- if (time_before(jiffies, base->clk)) { -+ if (time_before(jiffies, base->clk) || !tick_find_expired(base)) { - if (!IS_ENABLED(CONFIG_NO_HZ_COMMON) || !base->nohz_active) - return; - /* CPU is awake, so check the deferrable base. */ - base++; -- if (time_before(jiffies, base->clk)) -+ if (time_before(jiffies, base->clk) || !tick_find_expired(base)) - return; - } - raise_softirq(TIMER_SOFTIRQ); -@@ -1836,16 +1907,17 @@ - * The caller is globally serialized and nobody else - * takes two locks at once, deadlock is not possible. - */ -- spin_lock_irq(&new_base->lock); -- spin_lock_nested(&old_base->lock, SINGLE_DEPTH_NESTING); -+ raw_spin_lock_irq(&new_base->lock); -+ raw_spin_lock_nested(&old_base->lock, SINGLE_DEPTH_NESTING); - - BUG_ON(old_base->running_timer); -+ BUG_ON(old_base->expired_count); - - for (i = 0; i < WHEEL_SIZE; i++) - migrate_timer_list(new_base, old_base->vectors + i); - -- spin_unlock(&old_base->lock); -- spin_unlock_irq(&new_base->lock); -+ raw_spin_unlock(&old_base->lock); -+ raw_spin_unlock_irq(&new_base->lock); - put_cpu_ptr(&timer_bases); - } - return 0; -@@ -1861,8 +1933,12 @@ - for (i = 0; i < NR_BASES; i++) { - base = per_cpu_ptr(&timer_bases[i], cpu); - base->cpu = cpu; -- spin_lock_init(&base->lock); -+ raw_spin_lock_init(&base->lock); - base->clk = jiffies; -+#ifdef CONFIG_PREEMPT_RT_FULL -+ init_swait_queue_head(&base->wait_for_running_timer); -+#endif -+ base->expired_count = 0; - } - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/Kconfig 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/Kconfig 2017-04-18 17:54:26.000000000 +0200 -@@ -182,6 +182,24 @@ - enabled. This option and the preempt-off timing option can be - used together or separately.) - -+config INTERRUPT_OFF_HIST -+ bool "Interrupts-off Latency Histogram" -+ depends on IRQSOFF_TRACER -+ help -+ This option generates continuously updated histograms (one per cpu) -+ of the duration of time periods with interrupts disabled. The -+ histograms are disabled by default. To enable them, write a non-zero -+ number to -+ -+ /sys/kernel/debug/tracing/latency_hist/enable/preemptirqsoff -+ -+ If PREEMPT_OFF_HIST is also selected, additional histograms (one -+ per cpu) are generated that accumulate the duration of time periods -+ when both interrupts and preemption are disabled. The histogram data -+ will be located in the debug file system at -+ -+ /sys/kernel/debug/tracing/latency_hist/irqsoff -+ - config PREEMPT_TRACER - bool "Preemption-off Latency Tracer" - default n -@@ -206,6 +224,24 @@ - enabled. This option and the irqs-off timing option can be - used together or separately.) - -+config PREEMPT_OFF_HIST -+ bool "Preemption-off Latency Histogram" -+ depends on PREEMPT_TRACER -+ help -+ This option generates continuously updated histograms (one per cpu) -+ of the duration of time periods with preemption disabled. The -+ histograms are disabled by default. To enable them, write a non-zero -+ number to -+ -+ /sys/kernel/debug/tracing/latency_hist/enable/preemptirqsoff -+ -+ If INTERRUPT_OFF_HIST is also selected, additional histograms (one -+ per cpu) are generated that accumulate the duration of time periods -+ when both interrupts and preemption are disabled. The histogram data -+ will be located in the debug file system at -+ -+ /sys/kernel/debug/tracing/latency_hist/preemptoff -+ - config SCHED_TRACER - bool "Scheduling Latency Tracer" - select GENERIC_TRACER -@@ -251,6 +287,74 @@ - file. Every time a latency is greater than tracing_thresh, it will - be recorded into the ring buffer. - -+config WAKEUP_LATENCY_HIST -+ bool "Scheduling Latency Histogram" -+ depends on SCHED_TRACER -+ help -+ This option generates continuously updated histograms (one per cpu) -+ of the scheduling latency of the highest priority task. -+ The histograms are disabled by default. To enable them, write a -+ non-zero number to -+ -+ /sys/kernel/debug/tracing/latency_hist/enable/wakeup -+ -+ Two different algorithms are used, one to determine the latency of -+ processes that exclusively use the highest priority of the system and -+ another one to determine the latency of processes that share the -+ highest system priority with other processes. The former is used to -+ improve hardware and system software, the latter to optimize the -+ priority design of a given system. The histogram data will be -+ located in the debug file system at -+ -+ /sys/kernel/debug/tracing/latency_hist/wakeup -+ -+ and -+ -+ /sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio -+ -+ If both Scheduling Latency Histogram and Missed Timer Offsets -+ Histogram are selected, additional histogram data will be collected -+ that contain, in addition to the wakeup latency, the timer latency, in -+ case the wakeup was triggered by an expired timer. These histograms -+ are available in the -+ -+ /sys/kernel/debug/tracing/latency_hist/timerandwakeup -+ -+ directory. They reflect the apparent interrupt and scheduling latency -+ and are best suitable to determine the worst-case latency of a given -+ system. To enable these histograms, write a non-zero number to -+ -+ /sys/kernel/debug/tracing/latency_hist/enable/timerandwakeup -+ -+config MISSED_TIMER_OFFSETS_HIST -+ depends on HIGH_RES_TIMERS -+ select GENERIC_TRACER -+ bool "Missed Timer Offsets Histogram" -+ help -+ Generate a histogram of missed timer offsets in microseconds. The -+ histograms are disabled by default. To enable them, write a non-zero -+ number to -+ -+ /sys/kernel/debug/tracing/latency_hist/enable/missed_timer_offsets -+ -+ The histogram data will be located in the debug file system at -+ -+ /sys/kernel/debug/tracing/latency_hist/missed_timer_offsets -+ -+ If both Scheduling Latency Histogram and Missed Timer Offsets -+ Histogram are selected, additional histogram data will be collected -+ that contain, in addition to the wakeup latency, the timer latency, in -+ case the wakeup was triggered by an expired timer. These histograms -+ are available in the -+ -+ /sys/kernel/debug/tracing/latency_hist/timerandwakeup -+ -+ directory. They reflect the apparent interrupt and scheduling latency -+ and are best suitable to determine the worst-case latency of a given -+ system. To enable these histograms, write a non-zero number to -+ -+ /sys/kernel/debug/tracing/latency_hist/enable/timerandwakeup -+ - config ENABLE_DEFAULT_TRACERS - bool "Trace process context switches and events" - depends on !GENERIC_TRACER -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/Makefile linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/Makefile ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/Makefile 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/Makefile 2017-04-18 17:54:26.000000000 +0200 -@@ -38,6 +38,10 @@ - obj-$(CONFIG_PREEMPT_TRACER) += trace_irqsoff.o - obj-$(CONFIG_SCHED_TRACER) += trace_sched_wakeup.o - obj-$(CONFIG_HWLAT_TRACER) += trace_hwlat.o -+obj-$(CONFIG_INTERRUPT_OFF_HIST) += latency_hist.o -+obj-$(CONFIG_PREEMPT_OFF_HIST) += latency_hist.o -+obj-$(CONFIG_WAKEUP_LATENCY_HIST) += latency_hist.o -+obj-$(CONFIG_MISSED_TIMER_OFFSETS_HIST) += latency_hist.o - obj-$(CONFIG_NOP_TRACER) += trace_nop.o - obj-$(CONFIG_STACK_TRACER) += trace_stack.o - obj-$(CONFIG_MMIOTRACE) += trace_mmiotrace.o -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/latency_hist.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/latency_hist.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/latency_hist.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/latency_hist.c 2017-04-18 17:54:26.000000000 +0200 -@@ -0,0 +1,1178 @@ -+/* -+ * kernel/trace/latency_hist.c -+ * -+ * Add support for histograms of preemption-off latency and -+ * interrupt-off latency and wakeup latency, it depends on -+ * Real-Time Preemption Support. -+ * -+ * Copyright (C) 2005 MontaVista Software, Inc. -+ * Yi Yang -+ * -+ * Converted to work with the new latency tracer. -+ * Copyright (C) 2008 Red Hat, Inc. -+ * Steven Rostedt -+ * -+ */ -+#include -+#include -+#include -+#include -+#include -+#include -+#include -+#include -+#include -+#include -+#include -+ -+#include "trace.h" -+#include -+ -+#define NSECS_PER_USECS 1000L -+ -+#define CREATE_TRACE_POINTS -+#include -+ -+enum { -+ IRQSOFF_LATENCY = 0, -+ PREEMPTOFF_LATENCY, -+ PREEMPTIRQSOFF_LATENCY, -+ WAKEUP_LATENCY, -+ WAKEUP_LATENCY_SHAREDPRIO, -+ MISSED_TIMER_OFFSETS, -+ TIMERANDWAKEUP_LATENCY, -+ MAX_LATENCY_TYPE, -+}; -+ -+#define MAX_ENTRY_NUM 10240 -+ -+struct hist_data { -+ atomic_t hist_mode; /* 0 log, 1 don't log */ -+ long offset; /* set it to MAX_ENTRY_NUM/2 for a bipolar scale */ -+ long min_lat; -+ long max_lat; -+ unsigned long long below_hist_bound_samples; -+ unsigned long long above_hist_bound_samples; -+ long long accumulate_lat; -+ unsigned long long total_samples; -+ unsigned long long hist_array[MAX_ENTRY_NUM]; -+}; -+ -+struct enable_data { -+ int latency_type; -+ int enabled; -+}; -+ -+static char *latency_hist_dir_root = "latency_hist"; -+ -+#ifdef CONFIG_INTERRUPT_OFF_HIST -+static DEFINE_PER_CPU(struct hist_data, irqsoff_hist); -+static char *irqsoff_hist_dir = "irqsoff"; -+static DEFINE_PER_CPU(cycles_t, hist_irqsoff_start); -+static DEFINE_PER_CPU(int, hist_irqsoff_counting); -+#endif -+ -+#ifdef CONFIG_PREEMPT_OFF_HIST -+static DEFINE_PER_CPU(struct hist_data, preemptoff_hist); -+static char *preemptoff_hist_dir = "preemptoff"; -+static DEFINE_PER_CPU(cycles_t, hist_preemptoff_start); -+static DEFINE_PER_CPU(int, hist_preemptoff_counting); -+#endif -+ -+#if defined(CONFIG_PREEMPT_OFF_HIST) && defined(CONFIG_INTERRUPT_OFF_HIST) -+static DEFINE_PER_CPU(struct hist_data, preemptirqsoff_hist); -+static char *preemptirqsoff_hist_dir = "preemptirqsoff"; -+static DEFINE_PER_CPU(cycles_t, hist_preemptirqsoff_start); -+static DEFINE_PER_CPU(int, hist_preemptirqsoff_counting); -+#endif -+ -+#if defined(CONFIG_PREEMPT_OFF_HIST) || defined(CONFIG_INTERRUPT_OFF_HIST) -+static notrace void probe_preemptirqsoff_hist(void *v, int reason, int start); -+static struct enable_data preemptirqsoff_enabled_data = { -+ .latency_type = PREEMPTIRQSOFF_LATENCY, -+ .enabled = 0, -+}; -+#endif -+ -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+struct maxlatproc_data { -+ char comm[FIELD_SIZEOF(struct task_struct, comm)]; -+ char current_comm[FIELD_SIZEOF(struct task_struct, comm)]; -+ int pid; -+ int current_pid; -+ int prio; -+ int current_prio; -+ long latency; -+ long timeroffset; -+ cycle_t timestamp; -+}; -+#endif -+ -+#ifdef CONFIG_WAKEUP_LATENCY_HIST -+static DEFINE_PER_CPU(struct hist_data, wakeup_latency_hist); -+static DEFINE_PER_CPU(struct hist_data, wakeup_latency_hist_sharedprio); -+static char *wakeup_latency_hist_dir = "wakeup"; -+static char *wakeup_latency_hist_dir_sharedprio = "sharedprio"; -+static notrace void probe_wakeup_latency_hist_start(void *v, -+ struct task_struct *p); -+static notrace void probe_wakeup_latency_hist_stop(void *v, -+ bool preempt, struct task_struct *prev, struct task_struct *next); -+static notrace void probe_sched_migrate_task(void *, -+ struct task_struct *task, int cpu); -+static struct enable_data wakeup_latency_enabled_data = { -+ .latency_type = WAKEUP_LATENCY, -+ .enabled = 0, -+}; -+static DEFINE_PER_CPU(struct maxlatproc_data, wakeup_maxlatproc); -+static DEFINE_PER_CPU(struct maxlatproc_data, wakeup_maxlatproc_sharedprio); -+static DEFINE_PER_CPU(struct task_struct *, wakeup_task); -+static DEFINE_PER_CPU(int, wakeup_sharedprio); -+static unsigned long wakeup_pid; -+#endif -+ -+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST -+static DEFINE_PER_CPU(struct hist_data, missed_timer_offsets); -+static char *missed_timer_offsets_dir = "missed_timer_offsets"; -+static notrace void probe_hrtimer_interrupt(void *v, int cpu, -+ long long offset, struct task_struct *curr, struct task_struct *task); -+static struct enable_data missed_timer_offsets_enabled_data = { -+ .latency_type = MISSED_TIMER_OFFSETS, -+ .enabled = 0, -+}; -+static DEFINE_PER_CPU(struct maxlatproc_data, missed_timer_offsets_maxlatproc); -+static unsigned long missed_timer_offsets_pid; -+#endif -+ -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+static DEFINE_PER_CPU(struct hist_data, timerandwakeup_latency_hist); -+static char *timerandwakeup_latency_hist_dir = "timerandwakeup"; -+static struct enable_data timerandwakeup_enabled_data = { -+ .latency_type = TIMERANDWAKEUP_LATENCY, -+ .enabled = 0, -+}; -+static DEFINE_PER_CPU(struct maxlatproc_data, timerandwakeup_maxlatproc); -+#endif -+ -+void notrace latency_hist(int latency_type, int cpu, long latency, -+ long timeroffset, cycle_t stop, -+ struct task_struct *p) -+{ -+ struct hist_data *my_hist; -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+ struct maxlatproc_data *mp = NULL; -+#endif -+ -+ if (!cpu_possible(cpu) || latency_type < 0 || -+ latency_type >= MAX_LATENCY_TYPE) -+ return; -+ -+ switch (latency_type) { -+#ifdef CONFIG_INTERRUPT_OFF_HIST -+ case IRQSOFF_LATENCY: -+ my_hist = &per_cpu(irqsoff_hist, cpu); -+ break; -+#endif -+#ifdef CONFIG_PREEMPT_OFF_HIST -+ case PREEMPTOFF_LATENCY: -+ my_hist = &per_cpu(preemptoff_hist, cpu); -+ break; -+#endif -+#if defined(CONFIG_PREEMPT_OFF_HIST) && defined(CONFIG_INTERRUPT_OFF_HIST) -+ case PREEMPTIRQSOFF_LATENCY: -+ my_hist = &per_cpu(preemptirqsoff_hist, cpu); -+ break; -+#endif -+#ifdef CONFIG_WAKEUP_LATENCY_HIST -+ case WAKEUP_LATENCY: -+ my_hist = &per_cpu(wakeup_latency_hist, cpu); -+ mp = &per_cpu(wakeup_maxlatproc, cpu); -+ break; -+ case WAKEUP_LATENCY_SHAREDPRIO: -+ my_hist = &per_cpu(wakeup_latency_hist_sharedprio, cpu); -+ mp = &per_cpu(wakeup_maxlatproc_sharedprio, cpu); -+ break; -+#endif -+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST -+ case MISSED_TIMER_OFFSETS: -+ my_hist = &per_cpu(missed_timer_offsets, cpu); -+ mp = &per_cpu(missed_timer_offsets_maxlatproc, cpu); -+ break; -+#endif -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+ case TIMERANDWAKEUP_LATENCY: -+ my_hist = &per_cpu(timerandwakeup_latency_hist, cpu); -+ mp = &per_cpu(timerandwakeup_maxlatproc, cpu); -+ break; -+#endif -+ -+ default: -+ return; -+ } -+ -+ latency += my_hist->offset; -+ -+ if (atomic_read(&my_hist->hist_mode) == 0) -+ return; -+ -+ if (latency < 0 || latency >= MAX_ENTRY_NUM) { -+ if (latency < 0) -+ my_hist->below_hist_bound_samples++; -+ else -+ my_hist->above_hist_bound_samples++; -+ } else -+ my_hist->hist_array[latency]++; -+ -+ if (unlikely(latency > my_hist->max_lat || -+ my_hist->min_lat == LONG_MAX)) { -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+ if (latency_type == WAKEUP_LATENCY || -+ latency_type == WAKEUP_LATENCY_SHAREDPRIO || -+ latency_type == MISSED_TIMER_OFFSETS || -+ latency_type == TIMERANDWAKEUP_LATENCY) { -+ strncpy(mp->comm, p->comm, sizeof(mp->comm)); -+ strncpy(mp->current_comm, current->comm, -+ sizeof(mp->current_comm)); -+ mp->pid = task_pid_nr(p); -+ mp->current_pid = task_pid_nr(current); -+ mp->prio = p->prio; -+ mp->current_prio = current->prio; -+ mp->latency = latency; -+ mp->timeroffset = timeroffset; -+ mp->timestamp = stop; -+ } -+#endif -+ my_hist->max_lat = latency; -+ } -+ if (unlikely(latency < my_hist->min_lat)) -+ my_hist->min_lat = latency; -+ my_hist->total_samples++; -+ my_hist->accumulate_lat += latency; -+} -+ -+static void *l_start(struct seq_file *m, loff_t *pos) -+{ -+ loff_t *index_ptr = NULL; -+ loff_t index = *pos; -+ struct hist_data *my_hist = m->private; -+ -+ if (index == 0) { -+ char minstr[32], avgstr[32], maxstr[32]; -+ -+ atomic_dec(&my_hist->hist_mode); -+ -+ if (likely(my_hist->total_samples)) { -+ long avg = (long) div64_s64(my_hist->accumulate_lat, -+ my_hist->total_samples); -+ snprintf(minstr, sizeof(minstr), "%ld", -+ my_hist->min_lat - my_hist->offset); -+ snprintf(avgstr, sizeof(avgstr), "%ld", -+ avg - my_hist->offset); -+ snprintf(maxstr, sizeof(maxstr), "%ld", -+ my_hist->max_lat - my_hist->offset); -+ } else { -+ strcpy(minstr, ""); -+ strcpy(avgstr, minstr); -+ strcpy(maxstr, minstr); -+ } -+ -+ seq_printf(m, "#Minimum latency: %s microseconds\n" -+ "#Average latency: %s microseconds\n" -+ "#Maximum latency: %s microseconds\n" -+ "#Total samples: %llu\n" -+ "#There are %llu samples lower than %ld" -+ " microseconds.\n" -+ "#There are %llu samples greater or equal" -+ " than %ld microseconds.\n" -+ "#usecs\t%16s\n", -+ minstr, avgstr, maxstr, -+ my_hist->total_samples, -+ my_hist->below_hist_bound_samples, -+ -my_hist->offset, -+ my_hist->above_hist_bound_samples, -+ MAX_ENTRY_NUM - my_hist->offset, -+ "samples"); -+ } -+ if (index < MAX_ENTRY_NUM) { -+ index_ptr = kmalloc(sizeof(loff_t), GFP_KERNEL); -+ if (index_ptr) -+ *index_ptr = index; -+ } -+ -+ return index_ptr; -+} -+ -+static void *l_next(struct seq_file *m, void *p, loff_t *pos) -+{ -+ loff_t *index_ptr = p; -+ struct hist_data *my_hist = m->private; -+ -+ if (++*pos >= MAX_ENTRY_NUM) { -+ atomic_inc(&my_hist->hist_mode); -+ return NULL; -+ } -+ *index_ptr = *pos; -+ return index_ptr; -+} -+ -+static void l_stop(struct seq_file *m, void *p) -+{ -+ kfree(p); -+} -+ -+static int l_show(struct seq_file *m, void *p) -+{ -+ int index = *(loff_t *) p; -+ struct hist_data *my_hist = m->private; -+ -+ seq_printf(m, "%6ld\t%16llu\n", index - my_hist->offset, -+ my_hist->hist_array[index]); -+ return 0; -+} -+ -+static const struct seq_operations latency_hist_seq_op = { -+ .start = l_start, -+ .next = l_next, -+ .stop = l_stop, -+ .show = l_show -+}; -+ -+static int latency_hist_open(struct inode *inode, struct file *file) -+{ -+ int ret; -+ -+ ret = seq_open(file, &latency_hist_seq_op); -+ if (!ret) { -+ struct seq_file *seq = file->private_data; -+ seq->private = inode->i_private; -+ } -+ return ret; -+} -+ -+static const struct file_operations latency_hist_fops = { -+ .open = latency_hist_open, -+ .read = seq_read, -+ .llseek = seq_lseek, -+ .release = seq_release, -+}; -+ -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+static void clear_maxlatprocdata(struct maxlatproc_data *mp) -+{ -+ mp->comm[0] = mp->current_comm[0] = '\0'; -+ mp->prio = mp->current_prio = mp->pid = mp->current_pid = -+ mp->latency = mp->timeroffset = -1; -+ mp->timestamp = 0; -+} -+#endif -+ -+static void hist_reset(struct hist_data *hist) -+{ -+ atomic_dec(&hist->hist_mode); -+ -+ memset(hist->hist_array, 0, sizeof(hist->hist_array)); -+ hist->below_hist_bound_samples = 0ULL; -+ hist->above_hist_bound_samples = 0ULL; -+ hist->min_lat = LONG_MAX; -+ hist->max_lat = LONG_MIN; -+ hist->total_samples = 0ULL; -+ hist->accumulate_lat = 0LL; -+ -+ atomic_inc(&hist->hist_mode); -+} -+ -+static ssize_t -+latency_hist_reset(struct file *file, const char __user *a, -+ size_t size, loff_t *off) -+{ -+ int cpu; -+ struct hist_data *hist = NULL; -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+ struct maxlatproc_data *mp = NULL; -+#endif -+ off_t latency_type = (off_t) file->private_data; -+ -+ for_each_online_cpu(cpu) { -+ -+ switch (latency_type) { -+#ifdef CONFIG_PREEMPT_OFF_HIST -+ case PREEMPTOFF_LATENCY: -+ hist = &per_cpu(preemptoff_hist, cpu); -+ break; -+#endif -+#ifdef CONFIG_INTERRUPT_OFF_HIST -+ case IRQSOFF_LATENCY: -+ hist = &per_cpu(irqsoff_hist, cpu); -+ break; -+#endif -+#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) -+ case PREEMPTIRQSOFF_LATENCY: -+ hist = &per_cpu(preemptirqsoff_hist, cpu); -+ break; -+#endif -+#ifdef CONFIG_WAKEUP_LATENCY_HIST -+ case WAKEUP_LATENCY: -+ hist = &per_cpu(wakeup_latency_hist, cpu); -+ mp = &per_cpu(wakeup_maxlatproc, cpu); -+ break; -+ case WAKEUP_LATENCY_SHAREDPRIO: -+ hist = &per_cpu(wakeup_latency_hist_sharedprio, cpu); -+ mp = &per_cpu(wakeup_maxlatproc_sharedprio, cpu); -+ break; -+#endif -+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST -+ case MISSED_TIMER_OFFSETS: -+ hist = &per_cpu(missed_timer_offsets, cpu); -+ mp = &per_cpu(missed_timer_offsets_maxlatproc, cpu); -+ break; -+#endif -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+ case TIMERANDWAKEUP_LATENCY: -+ hist = &per_cpu(timerandwakeup_latency_hist, cpu); -+ mp = &per_cpu(timerandwakeup_maxlatproc, cpu); -+ break; -+#endif -+ } -+ -+ hist_reset(hist); -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+ if (latency_type == WAKEUP_LATENCY || -+ latency_type == WAKEUP_LATENCY_SHAREDPRIO || -+ latency_type == MISSED_TIMER_OFFSETS || -+ latency_type == TIMERANDWAKEUP_LATENCY) -+ clear_maxlatprocdata(mp); -+#endif -+ } -+ -+ return size; -+} -+ -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+static ssize_t -+show_pid(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos) -+{ -+ char buf[64]; -+ int r; -+ unsigned long *this_pid = file->private_data; -+ -+ r = snprintf(buf, sizeof(buf), "%lu\n", *this_pid); -+ return simple_read_from_buffer(ubuf, cnt, ppos, buf, r); -+} -+ -+static ssize_t do_pid(struct file *file, const char __user *ubuf, -+ size_t cnt, loff_t *ppos) -+{ -+ char buf[64]; -+ unsigned long pid; -+ unsigned long *this_pid = file->private_data; -+ -+ if (cnt >= sizeof(buf)) -+ return -EINVAL; -+ -+ if (copy_from_user(&buf, ubuf, cnt)) -+ return -EFAULT; -+ -+ buf[cnt] = '\0'; -+ -+ if (kstrtoul(buf, 10, &pid)) -+ return -EINVAL; -+ -+ *this_pid = pid; -+ -+ return cnt; -+} -+#endif -+ -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+static ssize_t -+show_maxlatproc(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos) -+{ -+ int r; -+ struct maxlatproc_data *mp = file->private_data; -+ int strmaxlen = (TASK_COMM_LEN * 2) + (8 * 8); -+ unsigned long long t; -+ unsigned long usecs, secs; -+ char *buf; -+ -+ if (mp->pid == -1 || mp->current_pid == -1) { -+ buf = "(none)\n"; -+ return simple_read_from_buffer(ubuf, cnt, ppos, buf, -+ strlen(buf)); -+ } -+ -+ buf = kmalloc(strmaxlen, GFP_KERNEL); -+ if (buf == NULL) -+ return -ENOMEM; -+ -+ t = ns2usecs(mp->timestamp); -+ usecs = do_div(t, USEC_PER_SEC); -+ secs = (unsigned long) t; -+ r = snprintf(buf, strmaxlen, -+ "%d %d %ld (%ld) %s <- %d %d %s %lu.%06lu\n", mp->pid, -+ MAX_RT_PRIO-1 - mp->prio, mp->latency, mp->timeroffset, mp->comm, -+ mp->current_pid, MAX_RT_PRIO-1 - mp->current_prio, mp->current_comm, -+ secs, usecs); -+ r = simple_read_from_buffer(ubuf, cnt, ppos, buf, r); -+ kfree(buf); -+ return r; -+} -+#endif -+ -+static ssize_t -+show_enable(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos) -+{ -+ char buf[64]; -+ struct enable_data *ed = file->private_data; -+ int r; -+ -+ r = snprintf(buf, sizeof(buf), "%d\n", ed->enabled); -+ return simple_read_from_buffer(ubuf, cnt, ppos, buf, r); -+} -+ -+static ssize_t -+do_enable(struct file *file, const char __user *ubuf, size_t cnt, loff_t *ppos) -+{ -+ char buf[64]; -+ long enable; -+ struct enable_data *ed = file->private_data; -+ -+ if (cnt >= sizeof(buf)) -+ return -EINVAL; -+ -+ if (copy_from_user(&buf, ubuf, cnt)) -+ return -EFAULT; -+ -+ buf[cnt] = 0; -+ -+ if (kstrtoul(buf, 10, &enable)) -+ return -EINVAL; -+ -+ if ((enable && ed->enabled) || (!enable && !ed->enabled)) -+ return cnt; -+ -+ if (enable) { -+ int ret; -+ -+ switch (ed->latency_type) { -+#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST) -+ case PREEMPTIRQSOFF_LATENCY: -+ ret = register_trace_preemptirqsoff_hist( -+ probe_preemptirqsoff_hist, NULL); -+ if (ret) { -+ pr_info("wakeup trace: Couldn't assign " -+ "probe_preemptirqsoff_hist " -+ "to trace_preemptirqsoff_hist\n"); -+ return ret; -+ } -+ break; -+#endif -+#ifdef CONFIG_WAKEUP_LATENCY_HIST -+ case WAKEUP_LATENCY: -+ ret = register_trace_sched_wakeup( -+ probe_wakeup_latency_hist_start, NULL); -+ if (ret) { -+ pr_info("wakeup trace: Couldn't assign " -+ "probe_wakeup_latency_hist_start " -+ "to trace_sched_wakeup\n"); -+ return ret; -+ } -+ ret = register_trace_sched_wakeup_new( -+ probe_wakeup_latency_hist_start, NULL); -+ if (ret) { -+ pr_info("wakeup trace: Couldn't assign " -+ "probe_wakeup_latency_hist_start " -+ "to trace_sched_wakeup_new\n"); -+ unregister_trace_sched_wakeup( -+ probe_wakeup_latency_hist_start, NULL); -+ return ret; -+ } -+ ret = register_trace_sched_switch( -+ probe_wakeup_latency_hist_stop, NULL); -+ if (ret) { -+ pr_info("wakeup trace: Couldn't assign " -+ "probe_wakeup_latency_hist_stop " -+ "to trace_sched_switch\n"); -+ unregister_trace_sched_wakeup( -+ probe_wakeup_latency_hist_start, NULL); -+ unregister_trace_sched_wakeup_new( -+ probe_wakeup_latency_hist_start, NULL); -+ return ret; -+ } -+ ret = register_trace_sched_migrate_task( -+ probe_sched_migrate_task, NULL); -+ if (ret) { -+ pr_info("wakeup trace: Couldn't assign " -+ "probe_sched_migrate_task " -+ "to trace_sched_migrate_task\n"); -+ unregister_trace_sched_wakeup( -+ probe_wakeup_latency_hist_start, NULL); -+ unregister_trace_sched_wakeup_new( -+ probe_wakeup_latency_hist_start, NULL); -+ unregister_trace_sched_switch( -+ probe_wakeup_latency_hist_stop, NULL); -+ return ret; -+ } -+ break; -+#endif -+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST -+ case MISSED_TIMER_OFFSETS: -+ ret = register_trace_hrtimer_interrupt( -+ probe_hrtimer_interrupt, NULL); -+ if (ret) { -+ pr_info("wakeup trace: Couldn't assign " -+ "probe_hrtimer_interrupt " -+ "to trace_hrtimer_interrupt\n"); -+ return ret; -+ } -+ break; -+#endif -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+ case TIMERANDWAKEUP_LATENCY: -+ if (!wakeup_latency_enabled_data.enabled || -+ !missed_timer_offsets_enabled_data.enabled) -+ return -EINVAL; -+ break; -+#endif -+ default: -+ break; -+ } -+ } else { -+ switch (ed->latency_type) { -+#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST) -+ case PREEMPTIRQSOFF_LATENCY: -+ { -+ int cpu; -+ -+ unregister_trace_preemptirqsoff_hist( -+ probe_preemptirqsoff_hist, NULL); -+ for_each_online_cpu(cpu) { -+#ifdef CONFIG_INTERRUPT_OFF_HIST -+ per_cpu(hist_irqsoff_counting, -+ cpu) = 0; -+#endif -+#ifdef CONFIG_PREEMPT_OFF_HIST -+ per_cpu(hist_preemptoff_counting, -+ cpu) = 0; -+#endif -+#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) -+ per_cpu(hist_preemptirqsoff_counting, -+ cpu) = 0; -+#endif -+ } -+ } -+ break; -+#endif -+#ifdef CONFIG_WAKEUP_LATENCY_HIST -+ case WAKEUP_LATENCY: -+ { -+ int cpu; -+ -+ unregister_trace_sched_wakeup( -+ probe_wakeup_latency_hist_start, NULL); -+ unregister_trace_sched_wakeup_new( -+ probe_wakeup_latency_hist_start, NULL); -+ unregister_trace_sched_switch( -+ probe_wakeup_latency_hist_stop, NULL); -+ unregister_trace_sched_migrate_task( -+ probe_sched_migrate_task, NULL); -+ -+ for_each_online_cpu(cpu) { -+ per_cpu(wakeup_task, cpu) = NULL; -+ per_cpu(wakeup_sharedprio, cpu) = 0; -+ } -+ } -+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST -+ timerandwakeup_enabled_data.enabled = 0; -+#endif -+ break; -+#endif -+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST -+ case MISSED_TIMER_OFFSETS: -+ unregister_trace_hrtimer_interrupt( -+ probe_hrtimer_interrupt, NULL); -+#ifdef CONFIG_WAKEUP_LATENCY_HIST -+ timerandwakeup_enabled_data.enabled = 0; -+#endif -+ break; -+#endif -+ default: -+ break; -+ } -+ } -+ ed->enabled = enable; -+ return cnt; -+} -+ -+static const struct file_operations latency_hist_reset_fops = { -+ .open = tracing_open_generic, -+ .write = latency_hist_reset, -+}; -+ -+static const struct file_operations enable_fops = { -+ .open = tracing_open_generic, -+ .read = show_enable, -+ .write = do_enable, -+}; -+ -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+static const struct file_operations pid_fops = { -+ .open = tracing_open_generic, -+ .read = show_pid, -+ .write = do_pid, -+}; -+ -+static const struct file_operations maxlatproc_fops = { -+ .open = tracing_open_generic, -+ .read = show_maxlatproc, -+}; -+#endif -+ -+#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST) -+static notrace void probe_preemptirqsoff_hist(void *v, int reason, -+ int starthist) -+{ -+ int cpu = raw_smp_processor_id(); -+ int time_set = 0; -+ -+ if (starthist) { -+ cycle_t uninitialized_var(start); -+ -+ if (!preempt_count() && !irqs_disabled()) -+ return; -+ -+#ifdef CONFIG_INTERRUPT_OFF_HIST -+ if ((reason == IRQS_OFF || reason == TRACE_START) && -+ !per_cpu(hist_irqsoff_counting, cpu)) { -+ per_cpu(hist_irqsoff_counting, cpu) = 1; -+ start = ftrace_now(cpu); -+ time_set++; -+ per_cpu(hist_irqsoff_start, cpu) = start; -+ } -+#endif -+ -+#ifdef CONFIG_PREEMPT_OFF_HIST -+ if ((reason == PREEMPT_OFF || reason == TRACE_START) && -+ !per_cpu(hist_preemptoff_counting, cpu)) { -+ per_cpu(hist_preemptoff_counting, cpu) = 1; -+ if (!(time_set++)) -+ start = ftrace_now(cpu); -+ per_cpu(hist_preemptoff_start, cpu) = start; -+ } -+#endif -+ -+#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) -+ if (per_cpu(hist_irqsoff_counting, cpu) && -+ per_cpu(hist_preemptoff_counting, cpu) && -+ !per_cpu(hist_preemptirqsoff_counting, cpu)) { -+ per_cpu(hist_preemptirqsoff_counting, cpu) = 1; -+ if (!time_set) -+ start = ftrace_now(cpu); -+ per_cpu(hist_preemptirqsoff_start, cpu) = start; -+ } -+#endif -+ } else { -+ cycle_t uninitialized_var(stop); -+ -+#ifdef CONFIG_INTERRUPT_OFF_HIST -+ if ((reason == IRQS_ON || reason == TRACE_STOP) && -+ per_cpu(hist_irqsoff_counting, cpu)) { -+ cycle_t start = per_cpu(hist_irqsoff_start, cpu); -+ -+ stop = ftrace_now(cpu); -+ time_set++; -+ if (start) { -+ long latency = ((long) (stop - start)) / -+ NSECS_PER_USECS; -+ -+ latency_hist(IRQSOFF_LATENCY, cpu, latency, 0, -+ stop, NULL); -+ } -+ per_cpu(hist_irqsoff_counting, cpu) = 0; -+ } -+#endif -+ -+#ifdef CONFIG_PREEMPT_OFF_HIST -+ if ((reason == PREEMPT_ON || reason == TRACE_STOP) && -+ per_cpu(hist_preemptoff_counting, cpu)) { -+ cycle_t start = per_cpu(hist_preemptoff_start, cpu); -+ -+ if (!(time_set++)) -+ stop = ftrace_now(cpu); -+ if (start) { -+ long latency = ((long) (stop - start)) / -+ NSECS_PER_USECS; -+ -+ latency_hist(PREEMPTOFF_LATENCY, cpu, latency, -+ 0, stop, NULL); -+ } -+ per_cpu(hist_preemptoff_counting, cpu) = 0; -+ } -+#endif -+ -+#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) -+ if ((!per_cpu(hist_irqsoff_counting, cpu) || -+ !per_cpu(hist_preemptoff_counting, cpu)) && -+ per_cpu(hist_preemptirqsoff_counting, cpu)) { -+ cycle_t start = per_cpu(hist_preemptirqsoff_start, cpu); -+ -+ if (!time_set) -+ stop = ftrace_now(cpu); -+ if (start) { -+ long latency = ((long) (stop - start)) / -+ NSECS_PER_USECS; -+ -+ latency_hist(PREEMPTIRQSOFF_LATENCY, cpu, -+ latency, 0, stop, NULL); -+ } -+ per_cpu(hist_preemptirqsoff_counting, cpu) = 0; -+ } -+#endif -+ } -+} -+#endif -+ -+#ifdef CONFIG_WAKEUP_LATENCY_HIST -+static DEFINE_RAW_SPINLOCK(wakeup_lock); -+static notrace void probe_sched_migrate_task(void *v, struct task_struct *task, -+ int cpu) -+{ -+ int old_cpu = task_cpu(task); -+ -+ if (cpu != old_cpu) { -+ unsigned long flags; -+ struct task_struct *cpu_wakeup_task; -+ -+ raw_spin_lock_irqsave(&wakeup_lock, flags); -+ -+ cpu_wakeup_task = per_cpu(wakeup_task, old_cpu); -+ if (task == cpu_wakeup_task) { -+ put_task_struct(cpu_wakeup_task); -+ per_cpu(wakeup_task, old_cpu) = NULL; -+ cpu_wakeup_task = per_cpu(wakeup_task, cpu) = task; -+ get_task_struct(cpu_wakeup_task); -+ } -+ -+ raw_spin_unlock_irqrestore(&wakeup_lock, flags); -+ } -+} -+ -+static notrace void probe_wakeup_latency_hist_start(void *v, -+ struct task_struct *p) -+{ -+ unsigned long flags; -+ struct task_struct *curr = current; -+ int cpu = task_cpu(p); -+ struct task_struct *cpu_wakeup_task; -+ -+ raw_spin_lock_irqsave(&wakeup_lock, flags); -+ -+ cpu_wakeup_task = per_cpu(wakeup_task, cpu); -+ -+ if (wakeup_pid) { -+ if ((cpu_wakeup_task && p->prio == cpu_wakeup_task->prio) || -+ p->prio == curr->prio) -+ per_cpu(wakeup_sharedprio, cpu) = 1; -+ if (likely(wakeup_pid != task_pid_nr(p))) -+ goto out; -+ } else { -+ if (likely(!rt_task(p)) || -+ (cpu_wakeup_task && p->prio > cpu_wakeup_task->prio) || -+ p->prio > curr->prio) -+ goto out; -+ if ((cpu_wakeup_task && p->prio == cpu_wakeup_task->prio) || -+ p->prio == curr->prio) -+ per_cpu(wakeup_sharedprio, cpu) = 1; -+ } -+ -+ if (cpu_wakeup_task) -+ put_task_struct(cpu_wakeup_task); -+ cpu_wakeup_task = per_cpu(wakeup_task, cpu) = p; -+ get_task_struct(cpu_wakeup_task); -+ cpu_wakeup_task->preempt_timestamp_hist = -+ ftrace_now(raw_smp_processor_id()); -+out: -+ raw_spin_unlock_irqrestore(&wakeup_lock, flags); -+} -+ -+static notrace void probe_wakeup_latency_hist_stop(void *v, -+ bool preempt, struct task_struct *prev, struct task_struct *next) -+{ -+ unsigned long flags; -+ int cpu = task_cpu(next); -+ long latency; -+ cycle_t stop; -+ struct task_struct *cpu_wakeup_task; -+ -+ raw_spin_lock_irqsave(&wakeup_lock, flags); -+ -+ cpu_wakeup_task = per_cpu(wakeup_task, cpu); -+ -+ if (cpu_wakeup_task == NULL) -+ goto out; -+ -+ /* Already running? */ -+ if (unlikely(current == cpu_wakeup_task)) -+ goto out_reset; -+ -+ if (next != cpu_wakeup_task) { -+ if (next->prio < cpu_wakeup_task->prio) -+ goto out_reset; -+ -+ if (next->prio == cpu_wakeup_task->prio) -+ per_cpu(wakeup_sharedprio, cpu) = 1; -+ -+ goto out; -+ } -+ -+ if (current->prio == cpu_wakeup_task->prio) -+ per_cpu(wakeup_sharedprio, cpu) = 1; -+ -+ /* -+ * The task we are waiting for is about to be switched to. -+ * Calculate latency and store it in histogram. -+ */ -+ stop = ftrace_now(raw_smp_processor_id()); -+ -+ latency = ((long) (stop - next->preempt_timestamp_hist)) / -+ NSECS_PER_USECS; -+ -+ if (per_cpu(wakeup_sharedprio, cpu)) { -+ latency_hist(WAKEUP_LATENCY_SHAREDPRIO, cpu, latency, 0, stop, -+ next); -+ per_cpu(wakeup_sharedprio, cpu) = 0; -+ } else { -+ latency_hist(WAKEUP_LATENCY, cpu, latency, 0, stop, next); -+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST -+ if (timerandwakeup_enabled_data.enabled) { -+ latency_hist(TIMERANDWAKEUP_LATENCY, cpu, -+ next->timer_offset + latency, next->timer_offset, -+ stop, next); -+ } -+#endif -+ } -+ -+out_reset: -+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST -+ next->timer_offset = 0; -+#endif -+ put_task_struct(cpu_wakeup_task); -+ per_cpu(wakeup_task, cpu) = NULL; -+out: -+ raw_spin_unlock_irqrestore(&wakeup_lock, flags); -+} -+#endif -+ -+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST -+static notrace void probe_hrtimer_interrupt(void *v, int cpu, -+ long long latency_ns, struct task_struct *curr, -+ struct task_struct *task) -+{ -+ if (latency_ns <= 0 && task != NULL && rt_task(task) && -+ (task->prio < curr->prio || -+ (task->prio == curr->prio && -+ !cpumask_test_cpu(cpu, &task->cpus_allowed)))) { -+ long latency; -+ cycle_t now; -+ -+ if (missed_timer_offsets_pid) { -+ if (likely(missed_timer_offsets_pid != -+ task_pid_nr(task))) -+ return; -+ } -+ -+ now = ftrace_now(cpu); -+ latency = (long) div_s64(-latency_ns, NSECS_PER_USECS); -+ latency_hist(MISSED_TIMER_OFFSETS, cpu, latency, latency, now, -+ task); -+#ifdef CONFIG_WAKEUP_LATENCY_HIST -+ task->timer_offset = latency; -+#endif -+ } -+} -+#endif -+ -+static __init int latency_hist_init(void) -+{ -+ struct dentry *latency_hist_root = NULL; -+ struct dentry *dentry; -+#ifdef CONFIG_WAKEUP_LATENCY_HIST -+ struct dentry *dentry_sharedprio; -+#endif -+ struct dentry *entry; -+ struct dentry *enable_root; -+ int i = 0; -+ struct hist_data *my_hist; -+ char name[64]; -+ char *cpufmt = "CPU%d"; -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+ char *cpufmt_maxlatproc = "max_latency-CPU%d"; -+ struct maxlatproc_data *mp = NULL; -+#endif -+ -+ dentry = tracing_init_dentry(); -+ latency_hist_root = debugfs_create_dir(latency_hist_dir_root, dentry); -+ enable_root = debugfs_create_dir("enable", latency_hist_root); -+ -+#ifdef CONFIG_INTERRUPT_OFF_HIST -+ dentry = debugfs_create_dir(irqsoff_hist_dir, latency_hist_root); -+ for_each_possible_cpu(i) { -+ sprintf(name, cpufmt, i); -+ entry = debugfs_create_file(name, 0444, dentry, -+ &per_cpu(irqsoff_hist, i), &latency_hist_fops); -+ my_hist = &per_cpu(irqsoff_hist, i); -+ atomic_set(&my_hist->hist_mode, 1); -+ my_hist->min_lat = LONG_MAX; -+ } -+ entry = debugfs_create_file("reset", 0644, dentry, -+ (void *)IRQSOFF_LATENCY, &latency_hist_reset_fops); -+#endif -+ -+#ifdef CONFIG_PREEMPT_OFF_HIST -+ dentry = debugfs_create_dir(preemptoff_hist_dir, -+ latency_hist_root); -+ for_each_possible_cpu(i) { -+ sprintf(name, cpufmt, i); -+ entry = debugfs_create_file(name, 0444, dentry, -+ &per_cpu(preemptoff_hist, i), &latency_hist_fops); -+ my_hist = &per_cpu(preemptoff_hist, i); -+ atomic_set(&my_hist->hist_mode, 1); -+ my_hist->min_lat = LONG_MAX; -+ } -+ entry = debugfs_create_file("reset", 0644, dentry, -+ (void *)PREEMPTOFF_LATENCY, &latency_hist_reset_fops); -+#endif -+ -+#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) -+ dentry = debugfs_create_dir(preemptirqsoff_hist_dir, -+ latency_hist_root); -+ for_each_possible_cpu(i) { -+ sprintf(name, cpufmt, i); -+ entry = debugfs_create_file(name, 0444, dentry, -+ &per_cpu(preemptirqsoff_hist, i), &latency_hist_fops); -+ my_hist = &per_cpu(preemptirqsoff_hist, i); -+ atomic_set(&my_hist->hist_mode, 1); -+ my_hist->min_lat = LONG_MAX; -+ } -+ entry = debugfs_create_file("reset", 0644, dentry, -+ (void *)PREEMPTIRQSOFF_LATENCY, &latency_hist_reset_fops); -+#endif -+ -+#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST) -+ entry = debugfs_create_file("preemptirqsoff", 0644, -+ enable_root, (void *)&preemptirqsoff_enabled_data, -+ &enable_fops); -+#endif -+ -+#ifdef CONFIG_WAKEUP_LATENCY_HIST -+ dentry = debugfs_create_dir(wakeup_latency_hist_dir, -+ latency_hist_root); -+ dentry_sharedprio = debugfs_create_dir( -+ wakeup_latency_hist_dir_sharedprio, dentry); -+ for_each_possible_cpu(i) { -+ sprintf(name, cpufmt, i); -+ -+ entry = debugfs_create_file(name, 0444, dentry, -+ &per_cpu(wakeup_latency_hist, i), -+ &latency_hist_fops); -+ my_hist = &per_cpu(wakeup_latency_hist, i); -+ atomic_set(&my_hist->hist_mode, 1); -+ my_hist->min_lat = LONG_MAX; -+ -+ entry = debugfs_create_file(name, 0444, dentry_sharedprio, -+ &per_cpu(wakeup_latency_hist_sharedprio, i), -+ &latency_hist_fops); -+ my_hist = &per_cpu(wakeup_latency_hist_sharedprio, i); -+ atomic_set(&my_hist->hist_mode, 1); -+ my_hist->min_lat = LONG_MAX; -+ -+ sprintf(name, cpufmt_maxlatproc, i); -+ -+ mp = &per_cpu(wakeup_maxlatproc, i); -+ entry = debugfs_create_file(name, 0444, dentry, mp, -+ &maxlatproc_fops); -+ clear_maxlatprocdata(mp); -+ -+ mp = &per_cpu(wakeup_maxlatproc_sharedprio, i); -+ entry = debugfs_create_file(name, 0444, dentry_sharedprio, mp, -+ &maxlatproc_fops); -+ clear_maxlatprocdata(mp); -+ } -+ entry = debugfs_create_file("pid", 0644, dentry, -+ (void *)&wakeup_pid, &pid_fops); -+ entry = debugfs_create_file("reset", 0644, dentry, -+ (void *)WAKEUP_LATENCY, &latency_hist_reset_fops); -+ entry = debugfs_create_file("reset", 0644, dentry_sharedprio, -+ (void *)WAKEUP_LATENCY_SHAREDPRIO, &latency_hist_reset_fops); -+ entry = debugfs_create_file("wakeup", 0644, -+ enable_root, (void *)&wakeup_latency_enabled_data, -+ &enable_fops); -+#endif -+ -+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST -+ dentry = debugfs_create_dir(missed_timer_offsets_dir, -+ latency_hist_root); -+ for_each_possible_cpu(i) { -+ sprintf(name, cpufmt, i); -+ entry = debugfs_create_file(name, 0444, dentry, -+ &per_cpu(missed_timer_offsets, i), &latency_hist_fops); -+ my_hist = &per_cpu(missed_timer_offsets, i); -+ atomic_set(&my_hist->hist_mode, 1); -+ my_hist->min_lat = LONG_MAX; -+ -+ sprintf(name, cpufmt_maxlatproc, i); -+ mp = &per_cpu(missed_timer_offsets_maxlatproc, i); -+ entry = debugfs_create_file(name, 0444, dentry, mp, -+ &maxlatproc_fops); -+ clear_maxlatprocdata(mp); -+ } -+ entry = debugfs_create_file("pid", 0644, dentry, -+ (void *)&missed_timer_offsets_pid, &pid_fops); -+ entry = debugfs_create_file("reset", 0644, dentry, -+ (void *)MISSED_TIMER_OFFSETS, &latency_hist_reset_fops); -+ entry = debugfs_create_file("missed_timer_offsets", 0644, -+ enable_root, (void *)&missed_timer_offsets_enabled_data, -+ &enable_fops); -+#endif -+ -+#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ -+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) -+ dentry = debugfs_create_dir(timerandwakeup_latency_hist_dir, -+ latency_hist_root); -+ for_each_possible_cpu(i) { -+ sprintf(name, cpufmt, i); -+ entry = debugfs_create_file(name, 0444, dentry, -+ &per_cpu(timerandwakeup_latency_hist, i), -+ &latency_hist_fops); -+ my_hist = &per_cpu(timerandwakeup_latency_hist, i); -+ atomic_set(&my_hist->hist_mode, 1); -+ my_hist->min_lat = LONG_MAX; -+ -+ sprintf(name, cpufmt_maxlatproc, i); -+ mp = &per_cpu(timerandwakeup_maxlatproc, i); -+ entry = debugfs_create_file(name, 0444, dentry, mp, -+ &maxlatproc_fops); -+ clear_maxlatprocdata(mp); -+ } -+ entry = debugfs_create_file("reset", 0644, dentry, -+ (void *)TIMERANDWAKEUP_LATENCY, &latency_hist_reset_fops); -+ entry = debugfs_create_file("timerandwakeup", 0644, -+ enable_root, (void *)&timerandwakeup_enabled_data, -+ &enable_fops); -+#endif -+ return 0; -+} -+ -+device_initcall(latency_hist_init); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/trace.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/trace.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/trace.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/trace.c 2017-04-18 17:54:26.000000000 +0200 -@@ -1897,6 +1897,7 @@ - struct task_struct *tsk = current; - - entry->preempt_count = pc & 0xff; -+ entry->preempt_lazy_count = preempt_lazy_count(); - entry->pid = (tsk) ? tsk->pid : 0; - entry->flags = - #ifdef CONFIG_TRACE_IRQFLAGS_SUPPORT -@@ -1907,8 +1908,11 @@ - ((pc & NMI_MASK ) ? TRACE_FLAG_NMI : 0) | - ((pc & HARDIRQ_MASK) ? TRACE_FLAG_HARDIRQ : 0) | - ((pc & SOFTIRQ_MASK) ? TRACE_FLAG_SOFTIRQ : 0) | -- (tif_need_resched() ? TRACE_FLAG_NEED_RESCHED : 0) | -+ (tif_need_resched_now() ? TRACE_FLAG_NEED_RESCHED : 0) | -+ (need_resched_lazy() ? TRACE_FLAG_NEED_RESCHED_LAZY : 0) | - (test_preempt_need_resched() ? TRACE_FLAG_PREEMPT_RESCHED : 0); -+ -+ entry->migrate_disable = (tsk) ? __migrate_disabled(tsk) & 0xFF : 0; - } - EXPORT_SYMBOL_GPL(tracing_generic_entry_update); - -@@ -2892,14 +2896,17 @@ - - static void print_lat_help_header(struct seq_file *m) - { -- seq_puts(m, "# _------=> CPU# \n" -- "# / _-----=> irqs-off \n" -- "# | / _----=> need-resched \n" -- "# || / _---=> hardirq/softirq \n" -- "# ||| / _--=> preempt-depth \n" -- "# |||| / delay \n" -- "# cmd pid ||||| time | caller \n" -- "# \\ / ||||| \\ | / \n"); -+ seq_puts(m, "# _--------=> CPU# \n" -+ "# / _-------=> irqs-off \n" -+ "# | / _------=> need-resched \n" -+ "# || / _-----=> need-resched_lazy \n" -+ "# ||| / _----=> hardirq/softirq \n" -+ "# |||| / _---=> preempt-depth \n" -+ "# ||||| / _--=> preempt-lazy-depth\n" -+ "# |||||| / _-=> migrate-disable \n" -+ "# ||||||| / delay \n" -+ "# cmd pid |||||||| time | caller \n" -+ "# \\ / |||||||| \\ | / \n"); - } - - static void print_event_info(struct trace_buffer *buf, struct seq_file *m) -@@ -2925,11 +2932,14 @@ - print_event_info(buf, m); - seq_puts(m, "# _-----=> irqs-off\n" - "# / _----=> need-resched\n" -- "# | / _---=> hardirq/softirq\n" -- "# || / _--=> preempt-depth\n" -- "# ||| / delay\n" -- "# TASK-PID CPU# |||| TIMESTAMP FUNCTION\n" -- "# | | | |||| | |\n"); -+ "# |/ _-----=> need-resched_lazy\n" -+ "# || / _---=> hardirq/softirq\n" -+ "# ||| / _--=> preempt-depth\n" -+ "# |||| / _-=> preempt-lazy-depth\n" -+ "# ||||| / _-=> migrate-disable \n" -+ "# |||||| / delay\n" -+ "# TASK-PID CPU# ||||||| TIMESTAMP FUNCTION\n" -+ "# | | | ||||||| | |\n"); - } - - void -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/trace.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/trace.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/trace.h 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/trace.h 2017-04-18 17:54:26.000000000 +0200 -@@ -124,6 +124,7 @@ - * NEED_RESCHED - reschedule is requested - * HARDIRQ - inside an interrupt handler - * SOFTIRQ - inside a softirq handler -+ * NEED_RESCHED_LAZY - lazy reschedule is requested - */ - enum trace_flag_type { - TRACE_FLAG_IRQS_OFF = 0x01, -@@ -133,6 +134,7 @@ - TRACE_FLAG_SOFTIRQ = 0x10, - TRACE_FLAG_PREEMPT_RESCHED = 0x20, - TRACE_FLAG_NMI = 0x40, -+ TRACE_FLAG_NEED_RESCHED_LAZY = 0x80, - }; - - #define TRACE_BUF_SIZE 1024 -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/trace_events.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/trace_events.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/trace_events.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/trace_events.c 2017-04-18 17:54:26.000000000 +0200 -@@ -187,6 +187,8 @@ - __common_field(unsigned char, flags); - __common_field(unsigned char, preempt_count); - __common_field(int, pid); -+ __common_field(unsigned short, migrate_disable); -+ __common_field(unsigned short, padding); - - return ret; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/trace_irqsoff.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/trace_irqsoff.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/trace_irqsoff.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/trace_irqsoff.c 2017-04-18 17:54:26.000000000 +0200 -@@ -13,6 +13,7 @@ - #include - #include - #include -+#include - - #include "trace.h" - -@@ -424,11 +425,13 @@ - { - if (preempt_trace() || irq_trace()) - start_critical_timing(CALLER_ADDR0, CALLER_ADDR1); -+ trace_preemptirqsoff_hist_rcuidle(TRACE_START, 1); - } - EXPORT_SYMBOL_GPL(start_critical_timings); - - void stop_critical_timings(void) - { -+ trace_preemptirqsoff_hist_rcuidle(TRACE_STOP, 0); - if (preempt_trace() || irq_trace()) - stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1); - } -@@ -438,6 +441,7 @@ - #ifdef CONFIG_PROVE_LOCKING - void time_hardirqs_on(unsigned long a0, unsigned long a1) - { -+ trace_preemptirqsoff_hist_rcuidle(IRQS_ON, 0); - if (!preempt_trace() && irq_trace()) - stop_critical_timing(a0, a1); - } -@@ -446,6 +450,7 @@ - { - if (!preempt_trace() && irq_trace()) - start_critical_timing(a0, a1); -+ trace_preemptirqsoff_hist_rcuidle(IRQS_OFF, 1); - } - - #else /* !CONFIG_PROVE_LOCKING */ -@@ -471,6 +476,7 @@ - */ - void trace_hardirqs_on(void) - { -+ trace_preemptirqsoff_hist(IRQS_ON, 0); - if (!preempt_trace() && irq_trace()) - stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1); - } -@@ -480,11 +486,13 @@ - { - if (!preempt_trace() && irq_trace()) - start_critical_timing(CALLER_ADDR0, CALLER_ADDR1); -+ trace_preemptirqsoff_hist(IRQS_OFF, 1); - } - EXPORT_SYMBOL(trace_hardirqs_off); - - __visible void trace_hardirqs_on_caller(unsigned long caller_addr) - { -+ trace_preemptirqsoff_hist(IRQS_ON, 0); - if (!preempt_trace() && irq_trace()) - stop_critical_timing(CALLER_ADDR0, caller_addr); - } -@@ -494,6 +502,7 @@ - { - if (!preempt_trace() && irq_trace()) - start_critical_timing(CALLER_ADDR0, caller_addr); -+ trace_preemptirqsoff_hist(IRQS_OFF, 1); - } - EXPORT_SYMBOL(trace_hardirqs_off_caller); - -@@ -503,12 +512,14 @@ - #ifdef CONFIG_PREEMPT_TRACER - void trace_preempt_on(unsigned long a0, unsigned long a1) - { -+ trace_preemptirqsoff_hist(PREEMPT_ON, 0); - if (preempt_trace() && !irq_trace()) - stop_critical_timing(a0, a1); - } - - void trace_preempt_off(unsigned long a0, unsigned long a1) - { -+ trace_preemptirqsoff_hist(PREEMPT_ON, 1); - if (preempt_trace() && !irq_trace()) - start_critical_timing(a0, a1); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/trace_output.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/trace_output.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/trace/trace_output.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/trace/trace_output.c 2017-04-18 17:54:26.000000000 +0200 -@@ -386,6 +386,7 @@ - { - char hardsoft_irq; - char need_resched; -+ char need_resched_lazy; - char irqs_off; - int hardirq; - int softirq; -@@ -416,6 +417,9 @@ - break; - } - -+ need_resched_lazy = -+ (entry->flags & TRACE_FLAG_NEED_RESCHED_LAZY) ? 'L' : '.'; -+ - hardsoft_irq = - (nmi && hardirq) ? 'Z' : - nmi ? 'z' : -@@ -424,14 +428,25 @@ - softirq ? 's' : - '.' ; - -- trace_seq_printf(s, "%c%c%c", -- irqs_off, need_resched, hardsoft_irq); -+ trace_seq_printf(s, "%c%c%c%c", -+ irqs_off, need_resched, need_resched_lazy, -+ hardsoft_irq); - - if (entry->preempt_count) - trace_seq_printf(s, "%x", entry->preempt_count); - else - trace_seq_putc(s, '.'); - -+ if (entry->preempt_lazy_count) -+ trace_seq_printf(s, "%x", entry->preempt_lazy_count); -+ else -+ trace_seq_putc(s, '.'); -+ -+ if (entry->migrate_disable) -+ trace_seq_printf(s, "%x", entry->migrate_disable); -+ else -+ trace_seq_putc(s, '.'); -+ - return !trace_seq_has_overflowed(s); - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/user.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/user.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/user.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/user.c 2017-04-18 17:54:26.000000000 +0200 -@@ -161,11 +161,11 @@ - if (!up) - return; - -- local_irq_save(flags); -+ local_irq_save_nort(flags); - if (atomic_dec_and_lock(&up->__count, &uidhash_lock)) - free_user(up, flags); - else -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - } - - struct user_struct *alloc_uid(kuid_t uid) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/watchdog.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/watchdog.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/watchdog.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/watchdog.c 2017-04-18 17:54:26.000000000 +0200 -@@ -315,6 +315,8 @@ - - #ifdef CONFIG_HARDLOCKUP_DETECTOR - -+static DEFINE_RAW_SPINLOCK(watchdog_output_lock); -+ - static struct perf_event_attr wd_hw_attr = { - .type = PERF_TYPE_HARDWARE, - .config = PERF_COUNT_HW_CPU_CYCLES, -@@ -348,6 +350,13 @@ - /* only print hardlockups once */ - if (__this_cpu_read(hard_watchdog_warn) == true) - return; -+ /* -+ * If early-printk is enabled then make sure we do not -+ * lock up in printk() and kill console logging: -+ */ -+ printk_kill(); -+ -+ raw_spin_lock(&watchdog_output_lock); - - pr_emerg("Watchdog detected hard LOCKUP on cpu %d", this_cpu); - print_modules(); -@@ -365,6 +374,7 @@ - !test_and_set_bit(0, &hardlockup_allcpu_dumped)) - trigger_allbutself_cpu_backtrace(); - -+ raw_spin_unlock(&watchdog_output_lock); - if (hardlockup_panic) - nmi_panic(regs, "Hard LOCKUP"); - -@@ -512,6 +522,7 @@ - /* kick off the timer for the hardlockup detector */ - hrtimer_init(hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); - hrtimer->function = watchdog_timer_fn; -+ hrtimer->irqsafe = 1; - - /* Enable the perf event */ - watchdog_nmi_enable(cpu); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/workqueue.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/workqueue.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/workqueue.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/workqueue.c 2017-04-18 17:54:26.000000000 +0200 -@@ -48,6 +48,8 @@ - #include - #include - #include -+#include -+#include - - #include "workqueue_internal.h" - -@@ -121,11 +123,16 @@ - * cpu or grabbing pool->lock is enough for read access. If - * POOL_DISASSOCIATED is set, it's identical to L. - * -+ * On RT we need the extra protection via rt_lock_idle_list() for -+ * the list manipulations against read access from -+ * wq_worker_sleeping(). All other places are nicely serialized via -+ * pool->lock. -+ * - * A: pool->attach_mutex protected. - * - * PL: wq_pool_mutex protected. - * -- * PR: wq_pool_mutex protected for writes. Sched-RCU protected for reads. -+ * PR: wq_pool_mutex protected for writes. RCU protected for reads. - * - * PW: wq_pool_mutex and wq->mutex protected for writes. Either for reads. - * -@@ -134,7 +141,7 @@ - * - * WQ: wq->mutex protected. - * -- * WR: wq->mutex protected for writes. Sched-RCU protected for reads. -+ * WR: wq->mutex protected for writes. RCU protected for reads. - * - * MD: wq_mayday_lock protected. - */ -@@ -185,7 +192,7 @@ - atomic_t nr_running ____cacheline_aligned_in_smp; - - /* -- * Destruction of pool is sched-RCU protected to allow dereferences -+ * Destruction of pool is RCU protected to allow dereferences - * from get_work_pool(). - */ - struct rcu_head rcu; -@@ -214,7 +221,7 @@ - /* - * Release of unbound pwq is punted to system_wq. See put_pwq() - * and pwq_unbound_release_workfn() for details. pool_workqueue -- * itself is also sched-RCU protected so that the first pwq can be -+ * itself is also RCU protected so that the first pwq can be - * determined without grabbing wq->mutex. - */ - struct work_struct unbound_release_work; -@@ -348,6 +355,8 @@ - struct workqueue_struct *system_freezable_power_efficient_wq __read_mostly; - EXPORT_SYMBOL_GPL(system_freezable_power_efficient_wq); - -+static DEFINE_LOCAL_IRQ_LOCK(pendingb_lock); -+ - static int worker_thread(void *__worker); - static void workqueue_sysfs_unregister(struct workqueue_struct *wq); - -@@ -355,20 +364,20 @@ - #include - - #define assert_rcu_or_pool_mutex() \ -- RCU_LOCKDEP_WARN(!rcu_read_lock_sched_held() && \ -+ RCU_LOCKDEP_WARN(!rcu_read_lock_held() && \ - !lockdep_is_held(&wq_pool_mutex), \ -- "sched RCU or wq_pool_mutex should be held") -+ "RCU or wq_pool_mutex should be held") - - #define assert_rcu_or_wq_mutex(wq) \ -- RCU_LOCKDEP_WARN(!rcu_read_lock_sched_held() && \ -+ RCU_LOCKDEP_WARN(!rcu_read_lock_held() && \ - !lockdep_is_held(&wq->mutex), \ -- "sched RCU or wq->mutex should be held") -+ "RCU or wq->mutex should be held") - - #define assert_rcu_or_wq_mutex_or_pool_mutex(wq) \ -- RCU_LOCKDEP_WARN(!rcu_read_lock_sched_held() && \ -+ RCU_LOCKDEP_WARN(!rcu_read_lock_held() && \ - !lockdep_is_held(&wq->mutex) && \ - !lockdep_is_held(&wq_pool_mutex), \ -- "sched RCU, wq->mutex or wq_pool_mutex should be held") -+ "RCU, wq->mutex or wq_pool_mutex should be held") - - #define for_each_cpu_worker_pool(pool, cpu) \ - for ((pool) = &per_cpu(cpu_worker_pools, cpu)[0]; \ -@@ -380,7 +389,7 @@ - * @pool: iteration cursor - * @pi: integer used for iteration - * -- * This must be called either with wq_pool_mutex held or sched RCU read -+ * This must be called either with wq_pool_mutex held or RCU read - * locked. If the pool needs to be used beyond the locking in effect, the - * caller is responsible for guaranteeing that the pool stays online. - * -@@ -412,7 +421,7 @@ - * @pwq: iteration cursor - * @wq: the target workqueue - * -- * This must be called either with wq->mutex held or sched RCU read locked. -+ * This must be called either with wq->mutex held or RCU read locked. - * If the pwq needs to be used beyond the locking in effect, the caller is - * responsible for guaranteeing that the pwq stays online. - * -@@ -424,6 +433,31 @@ - if (({ assert_rcu_or_wq_mutex(wq); false; })) { } \ - else - -+#ifdef CONFIG_PREEMPT_RT_BASE -+static inline void rt_lock_idle_list(struct worker_pool *pool) -+{ -+ preempt_disable(); -+} -+static inline void rt_unlock_idle_list(struct worker_pool *pool) -+{ -+ preempt_enable(); -+} -+static inline void sched_lock_idle_list(struct worker_pool *pool) { } -+static inline void sched_unlock_idle_list(struct worker_pool *pool) { } -+#else -+static inline void rt_lock_idle_list(struct worker_pool *pool) { } -+static inline void rt_unlock_idle_list(struct worker_pool *pool) { } -+static inline void sched_lock_idle_list(struct worker_pool *pool) -+{ -+ spin_lock_irq(&pool->lock); -+} -+static inline void sched_unlock_idle_list(struct worker_pool *pool) -+{ -+ spin_unlock_irq(&pool->lock); -+} -+#endif -+ -+ - #ifdef CONFIG_DEBUG_OBJECTS_WORK - - static struct debug_obj_descr work_debug_descr; -@@ -548,7 +582,7 @@ - * @wq: the target workqueue - * @node: the node ID - * -- * This must be called with any of wq_pool_mutex, wq->mutex or sched RCU -+ * This must be called with any of wq_pool_mutex, wq->mutex or RCU - * read locked. - * If the pwq needs to be used beyond the locking in effect, the caller is - * responsible for guaranteeing that the pwq stays online. -@@ -692,8 +726,8 @@ - * @work: the work item of interest - * - * Pools are created and destroyed under wq_pool_mutex, and allows read -- * access under sched-RCU read lock. As such, this function should be -- * called under wq_pool_mutex or with preemption disabled. -+ * access under RCU read lock. As such, this function should be -+ * called under wq_pool_mutex or inside of a rcu_read_lock() region. - * - * All fields of the returned pool are accessible as long as the above - * mentioned locking is in effect. If the returned pool needs to be used -@@ -830,50 +864,45 @@ - */ - static void wake_up_worker(struct worker_pool *pool) - { -- struct worker *worker = first_idle_worker(pool); -+ struct worker *worker; -+ -+ rt_lock_idle_list(pool); -+ -+ worker = first_idle_worker(pool); - - if (likely(worker)) - wake_up_process(worker->task); -+ -+ rt_unlock_idle_list(pool); - } - - /** -- * wq_worker_waking_up - a worker is waking up -+ * wq_worker_running - a worker is running again - * @task: task waking up -- * @cpu: CPU @task is waking up to - * -- * This function is called during try_to_wake_up() when a worker is -- * being awoken. -- * -- * CONTEXT: -- * spin_lock_irq(rq->lock) -+ * This function is called when a worker returns from schedule() - */ --void wq_worker_waking_up(struct task_struct *task, int cpu) -+void wq_worker_running(struct task_struct *task) - { - struct worker *worker = kthread_data(task); - -- if (!(worker->flags & WORKER_NOT_RUNNING)) { -- WARN_ON_ONCE(worker->pool->cpu != cpu); -+ if (!worker->sleeping) -+ return; -+ if (!(worker->flags & WORKER_NOT_RUNNING)) - atomic_inc(&worker->pool->nr_running); -- } -+ worker->sleeping = 0; - } - - /** - * wq_worker_sleeping - a worker is going to sleep - * @task: task going to sleep - * -- * This function is called during schedule() when a busy worker is -- * going to sleep. Worker on the same cpu can be woken up by -- * returning pointer to its task. -- * -- * CONTEXT: -- * spin_lock_irq(rq->lock) -- * -- * Return: -- * Worker task on @cpu to wake up, %NULL if none. -+ * This function is called from schedule() when a busy worker is -+ * going to sleep. - */ --struct task_struct *wq_worker_sleeping(struct task_struct *task) -+void wq_worker_sleeping(struct task_struct *task) - { -- struct worker *worker = kthread_data(task), *to_wakeup = NULL; -+ struct worker *worker = kthread_data(task); - struct worker_pool *pool; - - /* -@@ -882,29 +911,26 @@ - * checking NOT_RUNNING. - */ - if (worker->flags & WORKER_NOT_RUNNING) -- return NULL; -+ return; - - pool = worker->pool; - -- /* this can only happen on the local cpu */ -- if (WARN_ON_ONCE(pool->cpu != raw_smp_processor_id())) -- return NULL; -+ if (WARN_ON_ONCE(worker->sleeping)) -+ return; -+ -+ worker->sleeping = 1; - - /* - * The counterpart of the following dec_and_test, implied mb, - * worklist not empty test sequence is in insert_work(). - * Please read comment there. -- * -- * NOT_RUNNING is clear. This means that we're bound to and -- * running on the local cpu w/ rq lock held and preemption -- * disabled, which in turn means that none else could be -- * manipulating idle_list, so dereferencing idle_list without pool -- * lock is safe. - */ - if (atomic_dec_and_test(&pool->nr_running) && -- !list_empty(&pool->worklist)) -- to_wakeup = first_idle_worker(pool); -- return to_wakeup ? to_wakeup->task : NULL; -+ !list_empty(&pool->worklist)) { -+ sched_lock_idle_list(pool); -+ wake_up_worker(pool); -+ sched_unlock_idle_list(pool); -+ } - } - - /** -@@ -1098,12 +1124,14 @@ - { - if (pwq) { - /* -- * As both pwqs and pools are sched-RCU protected, the -+ * As both pwqs and pools are RCU protected, the - * following lock operations are safe. - */ -- spin_lock_irq(&pwq->pool->lock); -+ rcu_read_lock(); -+ local_spin_lock_irq(pendingb_lock, &pwq->pool->lock); - put_pwq(pwq); -- spin_unlock_irq(&pwq->pool->lock); -+ local_spin_unlock_irq(pendingb_lock, &pwq->pool->lock); -+ rcu_read_unlock(); - } - } - -@@ -1207,7 +1235,7 @@ - struct worker_pool *pool; - struct pool_workqueue *pwq; - -- local_irq_save(*flags); -+ local_lock_irqsave(pendingb_lock, *flags); - - /* try to steal the timer if it exists */ - if (is_dwork) { -@@ -1226,6 +1254,7 @@ - if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) - return 0; - -+ rcu_read_lock(); - /* - * The queueing is in progress, or it is already queued. Try to - * steal it from ->worklist without clearing WORK_STRUCT_PENDING. -@@ -1264,14 +1293,16 @@ - set_work_pool_and_keep_pending(work, pool->id); - - spin_unlock(&pool->lock); -+ rcu_read_unlock(); - return 1; - } - spin_unlock(&pool->lock); - fail: -- local_irq_restore(*flags); -+ rcu_read_unlock(); -+ local_unlock_irqrestore(pendingb_lock, *flags); - if (work_is_canceling(work)) - return -ENOENT; -- cpu_relax(); -+ cpu_chill(); - return -EAGAIN; - } - -@@ -1373,7 +1404,7 @@ - * queued or lose PENDING. Grabbing PENDING and queueing should - * happen with IRQ disabled. - */ -- WARN_ON_ONCE(!irqs_disabled()); -+ WARN_ON_ONCE_NONRT(!irqs_disabled()); - - debug_work_activate(work); - -@@ -1381,6 +1412,7 @@ - if (unlikely(wq->flags & __WQ_DRAINING) && - WARN_ON_ONCE(!is_chained_work(wq))) - return; -+ rcu_read_lock(); - retry: - if (req_cpu == WORK_CPU_UNBOUND) - cpu = wq_select_unbound_cpu(raw_smp_processor_id()); -@@ -1437,10 +1469,8 @@ - /* pwq determined, queue */ - trace_workqueue_queue_work(req_cpu, pwq, work); - -- if (WARN_ON(!list_empty(&work->entry))) { -- spin_unlock(&pwq->pool->lock); -- return; -- } -+ if (WARN_ON(!list_empty(&work->entry))) -+ goto out; - - pwq->nr_in_flight[pwq->work_color]++; - work_flags = work_color_to_flags(pwq->work_color); -@@ -1458,7 +1488,9 @@ - - insert_work(pwq, work, worklist, work_flags); - -+out: - spin_unlock(&pwq->pool->lock); -+ rcu_read_unlock(); - } - - /** -@@ -1478,14 +1510,14 @@ - bool ret = false; - unsigned long flags; - -- local_irq_save(flags); -+ local_lock_irqsave(pendingb_lock,flags); - - if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) { - __queue_work(cpu, wq, work); - ret = true; - } - -- local_irq_restore(flags); -+ local_unlock_irqrestore(pendingb_lock, flags); - return ret; - } - EXPORT_SYMBOL(queue_work_on); -@@ -1552,14 +1584,14 @@ - unsigned long flags; - - /* read the comment in __queue_work() */ -- local_irq_save(flags); -+ local_lock_irqsave(pendingb_lock, flags); - - if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) { - __queue_delayed_work(cpu, wq, dwork, delay); - ret = true; - } - -- local_irq_restore(flags); -+ local_unlock_irqrestore(pendingb_lock, flags); - return ret; - } - EXPORT_SYMBOL(queue_delayed_work_on); -@@ -1594,7 +1626,7 @@ - - if (likely(ret >= 0)) { - __queue_delayed_work(cpu, wq, dwork, delay); -- local_irq_restore(flags); -+ local_unlock_irqrestore(pendingb_lock, flags); - } - - /* -ENOENT from try_to_grab_pending() becomes %true */ -@@ -1627,7 +1659,9 @@ - worker->last_active = jiffies; - - /* idle_list is LIFO */ -+ rt_lock_idle_list(pool); - list_add(&worker->entry, &pool->idle_list); -+ rt_unlock_idle_list(pool); - - if (too_many_workers(pool) && !timer_pending(&pool->idle_timer)) - mod_timer(&pool->idle_timer, jiffies + IDLE_WORKER_TIMEOUT); -@@ -1660,7 +1694,9 @@ - return; - worker_clr_flags(worker, WORKER_IDLE); - pool->nr_idle--; -+ rt_lock_idle_list(pool); - list_del_init(&worker->entry); -+ rt_unlock_idle_list(pool); - } - - static struct worker *alloc_worker(int node) -@@ -1826,7 +1862,9 @@ - pool->nr_workers--; - pool->nr_idle--; - -+ rt_lock_idle_list(pool); - list_del_init(&worker->entry); -+ rt_unlock_idle_list(pool); - worker->flags |= WORKER_DIE; - wake_up_process(worker->task); - } -@@ -2785,14 +2823,14 @@ - - might_sleep(); - -- local_irq_disable(); -+ rcu_read_lock(); - pool = get_work_pool(work); - if (!pool) { -- local_irq_enable(); -+ rcu_read_unlock(); - return false; - } - -- spin_lock(&pool->lock); -+ spin_lock_irq(&pool->lock); - /* see the comment in try_to_grab_pending() with the same code */ - pwq = get_work_pwq(work); - if (pwq) { -@@ -2821,10 +2859,11 @@ - else - lock_map_acquire_read(&pwq->wq->lockdep_map); - lock_map_release(&pwq->wq->lockdep_map); -- -+ rcu_read_unlock(); - return true; - already_gone: - spin_unlock_irq(&pool->lock); -+ rcu_read_unlock(); - return false; - } - -@@ -2911,7 +2950,7 @@ - - /* tell other tasks trying to grab @work to back off */ - mark_work_canceling(work); -- local_irq_restore(flags); -+ local_unlock_irqrestore(pendingb_lock, flags); - - flush_work(work); - clear_work_data(work); -@@ -2966,10 +3005,10 @@ - */ - bool flush_delayed_work(struct delayed_work *dwork) - { -- local_irq_disable(); -+ local_lock_irq(pendingb_lock); - if (del_timer_sync(&dwork->timer)) - __queue_work(dwork->cpu, dwork->wq, &dwork->work); -- local_irq_enable(); -+ local_unlock_irq(pendingb_lock); - return flush_work(&dwork->work); - } - EXPORT_SYMBOL(flush_delayed_work); -@@ -2987,7 +3026,7 @@ - return false; - - set_work_pool_and_clear_pending(work, get_work_pool_id(work)); -- local_irq_restore(flags); -+ local_unlock_irqrestore(pendingb_lock, flags); - return ret; - } - -@@ -3245,7 +3284,7 @@ - * put_unbound_pool - put a worker_pool - * @pool: worker_pool to put - * -- * Put @pool. If its refcnt reaches zero, it gets destroyed in sched-RCU -+ * Put @pool. If its refcnt reaches zero, it gets destroyed in RCU - * safe manner. get_unbound_pool() calls this function on its failure path - * and this function should be able to release pools which went through, - * successfully or not, init_worker_pool(). -@@ -3299,8 +3338,8 @@ - del_timer_sync(&pool->idle_timer); - del_timer_sync(&pool->mayday_timer); - -- /* sched-RCU protected to allow dereferences from get_work_pool() */ -- call_rcu_sched(&pool->rcu, rcu_free_pool); -+ /* RCU protected to allow dereferences from get_work_pool() */ -+ call_rcu(&pool->rcu, rcu_free_pool); - } - - /** -@@ -3407,14 +3446,14 @@ - put_unbound_pool(pool); - mutex_unlock(&wq_pool_mutex); - -- call_rcu_sched(&pwq->rcu, rcu_free_pwq); -+ call_rcu(&pwq->rcu, rcu_free_pwq); - - /* - * If we're the last pwq going away, @wq is already dead and no one - * is gonna access it anymore. Schedule RCU free. - */ - if (is_last) -- call_rcu_sched(&wq->rcu, rcu_free_wq); -+ call_rcu(&wq->rcu, rcu_free_wq); - } - - /** -@@ -4064,7 +4103,7 @@ - * The base ref is never dropped on per-cpu pwqs. Directly - * schedule RCU free. - */ -- call_rcu_sched(&wq->rcu, rcu_free_wq); -+ call_rcu(&wq->rcu, rcu_free_wq); - } else { - /* - * We're the sole accessor of @wq at this point. Directly -@@ -4157,7 +4196,8 @@ - struct pool_workqueue *pwq; - bool ret; - -- rcu_read_lock_sched(); -+ rcu_read_lock(); -+ preempt_disable(); - - if (cpu == WORK_CPU_UNBOUND) - cpu = smp_processor_id(); -@@ -4168,7 +4208,8 @@ - pwq = unbound_pwq_by_node(wq, cpu_to_node(cpu)); - - ret = !list_empty(&pwq->delayed_works); -- rcu_read_unlock_sched(); -+ preempt_enable(); -+ rcu_read_unlock(); - - return ret; - } -@@ -4194,15 +4235,15 @@ - if (work_pending(work)) - ret |= WORK_BUSY_PENDING; - -- local_irq_save(flags); -+ rcu_read_lock(); - pool = get_work_pool(work); - if (pool) { -- spin_lock(&pool->lock); -+ spin_lock_irqsave(&pool->lock, flags); - if (find_worker_executing_work(pool, work)) - ret |= WORK_BUSY_RUNNING; -- spin_unlock(&pool->lock); -+ spin_unlock_irqrestore(&pool->lock, flags); - } -- local_irq_restore(flags); -+ rcu_read_unlock(); - - return ret; - } -@@ -4391,7 +4432,7 @@ - unsigned long flags; - int pi; - -- rcu_read_lock_sched(); -+ rcu_read_lock(); - - pr_info("Showing busy workqueues and worker pools:\n"); - -@@ -4444,7 +4485,7 @@ - spin_unlock_irqrestore(&pool->lock, flags); - } - -- rcu_read_unlock_sched(); -+ rcu_read_unlock(); - } - - /* -@@ -4782,16 +4823,16 @@ - * nr_active is monotonically decreasing. It's safe - * to peek without lock. - */ -- rcu_read_lock_sched(); -+ rcu_read_lock(); - for_each_pwq(pwq, wq) { - WARN_ON_ONCE(pwq->nr_active < 0); - if (pwq->nr_active) { - busy = true; -- rcu_read_unlock_sched(); -+ rcu_read_unlock(); - goto out_unlock; - } - } -- rcu_read_unlock_sched(); -+ rcu_read_unlock(); - } - out_unlock: - mutex_unlock(&wq_pool_mutex); -@@ -4981,7 +5022,8 @@ - const char *delim = ""; - int node, written = 0; - -- rcu_read_lock_sched(); -+ get_online_cpus(); -+ rcu_read_lock(); - for_each_node(node) { - written += scnprintf(buf + written, PAGE_SIZE - written, - "%s%d:%d", delim, node, -@@ -4989,7 +5031,8 @@ - delim = " "; - } - written += scnprintf(buf + written, PAGE_SIZE - written, "\n"); -- rcu_read_unlock_sched(); -+ rcu_read_unlock(); -+ put_online_cpus(); - - return written; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/workqueue_internal.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/workqueue_internal.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/kernel/workqueue_internal.h 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/kernel/workqueue_internal.h 2017-04-18 17:54:26.000000000 +0200 -@@ -43,6 +43,7 @@ - unsigned long last_active; /* L: last active timestamp */ - unsigned int flags; /* X: flags */ - int id; /* I: worker id */ -+ int sleeping; /* None */ - - /* - * Opaque string set with work_set_desc(). Printed out with task -@@ -68,7 +69,7 @@ - * Scheduler hooks for concurrency managed workqueue. Only to be used from - * sched/core.c and workqueue.c. - */ --void wq_worker_waking_up(struct task_struct *task, int cpu); --struct task_struct *wq_worker_sleeping(struct task_struct *task); -+void wq_worker_running(struct task_struct *task); -+void wq_worker_sleeping(struct task_struct *task); - - #endif /* _KERNEL_WORKQUEUE_INTERNAL_H */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/Kconfig 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/Kconfig 2017-04-18 17:54:26.000000000 +0200 -@@ -400,6 +400,7 @@ - - config CPUMASK_OFFSTACK - bool "Force CPU masks off stack" if DEBUG_PER_CPU_MAPS -+ depends on !PREEMPT_RT_FULL - help - Use dynamic allocation for cpumask_var_t, instead of putting - them on the stack. This is a bit more expensive, but avoids -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/debugobjects.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/debugobjects.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/debugobjects.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/debugobjects.c 2017-04-18 17:54:26.000000000 +0200 -@@ -308,7 +308,10 @@ - struct debug_obj *obj; - unsigned long flags; - -- fill_pool(); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ if (preempt_count() == 0 && !irqs_disabled()) -+#endif -+ fill_pool(); - - db = get_bucket((unsigned long) addr); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/idr.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/idr.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/idr.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/idr.c 2017-04-18 17:54:26.000000000 +0200 -@@ -30,6 +30,7 @@ - #include - #include - #include -+#include - - #define MAX_IDR_SHIFT (sizeof(int) * 8 - 1) - #define MAX_IDR_BIT (1U << MAX_IDR_SHIFT) -@@ -45,6 +46,37 @@ - static DEFINE_PER_CPU(int, idr_preload_cnt); - static DEFINE_SPINLOCK(simple_ida_lock); - -+#ifdef CONFIG_PREEMPT_RT_FULL -+static DEFINE_LOCAL_IRQ_LOCK(idr_lock); -+ -+static inline void idr_preload_lock(void) -+{ -+ local_lock(idr_lock); -+} -+ -+static inline void idr_preload_unlock(void) -+{ -+ local_unlock(idr_lock); -+} -+ -+void idr_preload_end(void) -+{ -+ idr_preload_unlock(); -+} -+EXPORT_SYMBOL(idr_preload_end); -+#else -+static inline void idr_preload_lock(void) -+{ -+ preempt_disable(); -+} -+ -+static inline void idr_preload_unlock(void) -+{ -+ preempt_enable(); -+} -+#endif -+ -+ - /* the maximum ID which can be allocated given idr->layers */ - static int idr_max(int layers) - { -@@ -115,14 +147,14 @@ - * context. See idr_preload() for details. - */ - if (!in_interrupt()) { -- preempt_disable(); -+ idr_preload_lock(); - new = __this_cpu_read(idr_preload_head); - if (new) { - __this_cpu_write(idr_preload_head, new->ary[0]); - __this_cpu_dec(idr_preload_cnt); - new->ary[0] = NULL; - } -- preempt_enable(); -+ idr_preload_unlock(); - if (new) - return new; - } -@@ -366,7 +398,6 @@ - idr_mark_full(pa, id); - } - -- - /** - * idr_preload - preload for idr_alloc() - * @gfp_mask: allocation mask to use for preloading -@@ -401,7 +432,7 @@ - WARN_ON_ONCE(in_interrupt()); - might_sleep_if(gfpflags_allow_blocking(gfp_mask)); - -- preempt_disable(); -+ idr_preload_lock(); - - /* - * idr_alloc() is likely to succeed w/o full idr_layer buffer and -@@ -413,9 +444,9 @@ - while (__this_cpu_read(idr_preload_cnt) < MAX_IDR_FREE) { - struct idr_layer *new; - -- preempt_enable(); -+ idr_preload_unlock(); - new = kmem_cache_zalloc(idr_layer_cache, gfp_mask); -- preempt_disable(); -+ idr_preload_lock(); - if (!new) - break; - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/irq_poll.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/irq_poll.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/irq_poll.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/irq_poll.c 2017-04-18 17:54:26.000000000 +0200 -@@ -36,6 +36,7 @@ - list_add_tail(&iop->list, this_cpu_ptr(&blk_cpu_iopoll)); - __raise_softirq_irqoff(IRQ_POLL_SOFTIRQ); - local_irq_restore(flags); -+ preempt_check_resched_rt(); - } - EXPORT_SYMBOL(irq_poll_sched); - -@@ -71,6 +72,7 @@ - local_irq_save(flags); - __irq_poll_complete(iop); - local_irq_restore(flags); -+ preempt_check_resched_rt(); - } - EXPORT_SYMBOL(irq_poll_complete); - -@@ -95,6 +97,7 @@ - } - - local_irq_enable(); -+ preempt_check_resched_rt(); - - /* Even though interrupts have been re-enabled, this - * access is safe because interrupts can only add new -@@ -132,6 +135,7 @@ - __raise_softirq_irqoff(IRQ_POLL_SOFTIRQ); - - local_irq_enable(); -+ preempt_check_resched_rt(); - } - - /** -@@ -195,6 +199,7 @@ - this_cpu_ptr(&blk_cpu_iopoll)); - __raise_softirq_irqoff(IRQ_POLL_SOFTIRQ); - local_irq_enable(); -+ preempt_check_resched_rt(); - - return 0; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/locking-selftest.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/locking-selftest.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/locking-selftest.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/locking-selftest.c 2017-04-18 17:54:26.000000000 +0200 -@@ -590,6 +590,8 @@ - #include "locking-selftest-spin-hardirq.h" - GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_hard_spin) - -+#ifndef CONFIG_PREEMPT_RT_FULL -+ - #include "locking-selftest-rlock-hardirq.h" - GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_hard_rlock) - -@@ -605,9 +607,12 @@ - #include "locking-selftest-wlock-softirq.h" - GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_soft_wlock) - -+#endif -+ - #undef E1 - #undef E2 - -+#ifndef CONFIG_PREEMPT_RT_FULL - /* - * Enabling hardirqs with a softirq-safe lock held: - */ -@@ -640,6 +645,8 @@ - #undef E1 - #undef E2 - -+#endif -+ - /* - * Enabling irqs with an irq-safe lock held: - */ -@@ -663,6 +670,8 @@ - #include "locking-selftest-spin-hardirq.h" - GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_hard_spin) - -+#ifndef CONFIG_PREEMPT_RT_FULL -+ - #include "locking-selftest-rlock-hardirq.h" - GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_hard_rlock) - -@@ -678,6 +687,8 @@ - #include "locking-selftest-wlock-softirq.h" - GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_wlock) - -+#endif -+ - #undef E1 - #undef E2 - -@@ -709,6 +720,8 @@ - #include "locking-selftest-spin-hardirq.h" - GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_hard_spin) - -+#ifndef CONFIG_PREEMPT_RT_FULL -+ - #include "locking-selftest-rlock-hardirq.h" - GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_hard_rlock) - -@@ -724,6 +737,8 @@ - #include "locking-selftest-wlock-softirq.h" - GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_wlock) - -+#endif -+ - #undef E1 - #undef E2 - #undef E3 -@@ -757,6 +772,8 @@ - #include "locking-selftest-spin-hardirq.h" - GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_hard_spin) - -+#ifndef CONFIG_PREEMPT_RT_FULL -+ - #include "locking-selftest-rlock-hardirq.h" - GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_hard_rlock) - -@@ -772,10 +789,14 @@ - #include "locking-selftest-wlock-softirq.h" - GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_soft_wlock) - -+#endif -+ - #undef E1 - #undef E2 - #undef E3 - -+#ifndef CONFIG_PREEMPT_RT_FULL -+ - /* - * read-lock / write-lock irq inversion. - * -@@ -838,6 +859,10 @@ - #undef E2 - #undef E3 - -+#endif -+ -+#ifndef CONFIG_PREEMPT_RT_FULL -+ - /* - * read-lock / write-lock recursion that is actually safe. - */ -@@ -876,6 +901,8 @@ - #undef E2 - #undef E3 - -+#endif -+ - /* - * read-lock / write-lock recursion that is unsafe. - */ -@@ -1858,6 +1885,7 @@ - - printk(" --------------------------------------------------------------------------\n"); - -+#ifndef CONFIG_PREEMPT_RT_FULL - /* - * irq-context testcases: - */ -@@ -1870,6 +1898,28 @@ - - DO_TESTCASE_6x2("irq read-recursion", irq_read_recursion); - // DO_TESTCASE_6x2B("irq read-recursion #2", irq_read_recursion2); -+#else -+ /* On -rt, we only do hardirq context test for raw spinlock */ -+ DO_TESTCASE_1B("hard-irqs-on + irq-safe-A", irqsafe1_hard_spin, 12); -+ DO_TESTCASE_1B("hard-irqs-on + irq-safe-A", irqsafe1_hard_spin, 21); -+ -+ DO_TESTCASE_1B("hard-safe-A + irqs-on", irqsafe2B_hard_spin, 12); -+ DO_TESTCASE_1B("hard-safe-A + irqs-on", irqsafe2B_hard_spin, 21); -+ -+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 123); -+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 132); -+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 213); -+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 231); -+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 312); -+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 321); -+ -+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 123); -+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 132); -+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 213); -+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 231); -+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 312); -+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 321); -+#endif - - ww_tests(); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/percpu_ida.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/percpu_ida.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/percpu_ida.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/percpu_ida.c 2017-04-18 17:54:26.000000000 +0200 -@@ -26,6 +26,9 @@ - #include - #include - #include -+#include -+ -+static DEFINE_LOCAL_IRQ_LOCK(irq_off_lock); - - struct percpu_ida_cpu { - /* -@@ -148,13 +151,13 @@ - unsigned long flags; - int tag; - -- local_irq_save(flags); -+ local_lock_irqsave(irq_off_lock, flags); - tags = this_cpu_ptr(pool->tag_cpu); - - /* Fastpath */ - tag = alloc_local_tag(tags); - if (likely(tag >= 0)) { -- local_irq_restore(flags); -+ local_unlock_irqrestore(irq_off_lock, flags); - return tag; - } - -@@ -173,6 +176,7 @@ - - if (!tags->nr_free) - alloc_global_tags(pool, tags); -+ - if (!tags->nr_free) - steal_tags(pool, tags); - -@@ -184,7 +188,7 @@ - } - - spin_unlock(&pool->lock); -- local_irq_restore(flags); -+ local_unlock_irqrestore(irq_off_lock, flags); - - if (tag >= 0 || state == TASK_RUNNING) - break; -@@ -196,7 +200,7 @@ - - schedule(); - -- local_irq_save(flags); -+ local_lock_irqsave(irq_off_lock, flags); - tags = this_cpu_ptr(pool->tag_cpu); - } - if (state != TASK_RUNNING) -@@ -221,7 +225,7 @@ - - BUG_ON(tag >= pool->nr_tags); - -- local_irq_save(flags); -+ local_lock_irqsave(irq_off_lock, flags); - tags = this_cpu_ptr(pool->tag_cpu); - - spin_lock(&tags->lock); -@@ -253,7 +257,7 @@ - spin_unlock(&pool->lock); - } - -- local_irq_restore(flags); -+ local_unlock_irqrestore(irq_off_lock, flags); - } - EXPORT_SYMBOL_GPL(percpu_ida_free); - -@@ -345,7 +349,7 @@ - struct percpu_ida_cpu *remote; - unsigned cpu, i, err = 0; - -- local_irq_save(flags); -+ local_lock_irqsave(irq_off_lock, flags); - for_each_possible_cpu(cpu) { - remote = per_cpu_ptr(pool->tag_cpu, cpu); - spin_lock(&remote->lock); -@@ -367,7 +371,7 @@ - } - spin_unlock(&pool->lock); - out: -- local_irq_restore(flags); -+ local_unlock_irqrestore(irq_off_lock, flags); - return err; - } - EXPORT_SYMBOL_GPL(percpu_ida_for_each_free); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/radix-tree.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/radix-tree.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/radix-tree.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/radix-tree.c 2017-04-18 17:54:26.000000000 +0200 -@@ -36,7 +36,7 @@ - #include - #include - #include /* in_interrupt() */ -- -+#include - - /* Number of nodes in fully populated tree of given height */ - static unsigned long height_to_maxnodes[RADIX_TREE_MAX_PATH + 1] __read_mostly; -@@ -68,6 +68,7 @@ - struct radix_tree_node *nodes; - }; - static DEFINE_PER_CPU(struct radix_tree_preload, radix_tree_preloads) = { 0, }; -+static DEFINE_LOCAL_IRQ_LOCK(radix_tree_preloads_lock); - - static inline void *node_to_entry(void *ptr) - { -@@ -290,13 +291,14 @@ - * succeed in getting a node here (and never reach - * kmem_cache_alloc) - */ -- rtp = this_cpu_ptr(&radix_tree_preloads); -+ rtp = &get_locked_var(radix_tree_preloads_lock, radix_tree_preloads); - if (rtp->nr) { - ret = rtp->nodes; - rtp->nodes = ret->private_data; - ret->private_data = NULL; - rtp->nr--; - } -+ put_locked_var(radix_tree_preloads_lock, radix_tree_preloads); - /* - * Update the allocation stack trace as this is more useful - * for debugging. -@@ -357,14 +359,14 @@ - */ - gfp_mask &= ~__GFP_ACCOUNT; - -- preempt_disable(); -+ local_lock(radix_tree_preloads_lock); - rtp = this_cpu_ptr(&radix_tree_preloads); - while (rtp->nr < nr) { -- preempt_enable(); -+ local_unlock(radix_tree_preloads_lock); - node = kmem_cache_alloc(radix_tree_node_cachep, gfp_mask); - if (node == NULL) - goto out; -- preempt_disable(); -+ local_lock(radix_tree_preloads_lock); - rtp = this_cpu_ptr(&radix_tree_preloads); - if (rtp->nr < nr) { - node->private_data = rtp->nodes; -@@ -406,7 +408,7 @@ - if (gfpflags_allow_blocking(gfp_mask)) - return __radix_tree_preload(gfp_mask, RADIX_TREE_PRELOAD_SIZE); - /* Preloading doesn't help anything with this gfp mask, skip it */ -- preempt_disable(); -+ local_lock(radix_tree_preloads_lock); - return 0; - } - EXPORT_SYMBOL(radix_tree_maybe_preload); -@@ -422,7 +424,7 @@ - - /* Preloading doesn't help anything with this gfp mask, skip it */ - if (!gfpflags_allow_blocking(gfp_mask)) { -- preempt_disable(); -+ local_lock(radix_tree_preloads_lock); - return 0; - } - -@@ -456,6 +458,12 @@ - return __radix_tree_preload(gfp_mask, nr_nodes); - } - -+void radix_tree_preload_end(void) -+{ -+ local_unlock(radix_tree_preloads_lock); -+} -+EXPORT_SYMBOL(radix_tree_preload_end); -+ - /* - * The maximum index which can be stored in a radix tree - */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/scatterlist.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/scatterlist.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/scatterlist.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/scatterlist.c 2017-04-18 17:54:26.000000000 +0200 -@@ -620,7 +620,7 @@ - flush_kernel_dcache_page(miter->page); - - if (miter->__flags & SG_MITER_ATOMIC) { -- WARN_ON_ONCE(preemptible()); -+ WARN_ON_ONCE(!pagefault_disabled()); - kunmap_atomic(miter->addr); - } else - kunmap(miter->page); -@@ -664,7 +664,7 @@ - if (!sg_miter_skip(&miter, skip)) - return false; - -- local_irq_save(flags); -+ local_irq_save_nort(flags); - - while (sg_miter_next(&miter) && offset < buflen) { - unsigned int len; -@@ -681,7 +681,7 @@ - - sg_miter_stop(&miter); - -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - return offset; - } - EXPORT_SYMBOL(sg_copy_buffer); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/smp_processor_id.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/smp_processor_id.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/lib/smp_processor_id.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/lib/smp_processor_id.c 2017-04-18 17:54:26.000000000 +0200 -@@ -39,8 +39,9 @@ - if (!printk_ratelimit()) - goto out_enable; - -- printk(KERN_ERR "BUG: using %s%s() in preemptible [%08x] code: %s/%d\n", -- what1, what2, preempt_count() - 1, current->comm, current->pid); -+ printk(KERN_ERR "BUG: using %s%s() in preemptible [%08x %08x] code: %s/%d\n", -+ what1, what2, preempt_count() - 1, __migrate_disabled(current), -+ current->comm, current->pid); - - print_symbol("caller is %s\n", (long)__builtin_return_address(0)); - dump_stack(); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/Kconfig linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/Kconfig ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/Kconfig 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/Kconfig 2017-04-18 17:54:26.000000000 +0200 -@@ -410,7 +410,7 @@ - - config TRANSPARENT_HUGEPAGE - bool "Transparent Hugepage Support" -- depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE -+ depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE && !PREEMPT_RT_FULL - select COMPACTION - select RADIX_TREE_MULTIORDER - help -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/backing-dev.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/backing-dev.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/backing-dev.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/backing-dev.c 2017-04-18 17:54:26.000000000 +0200 -@@ -457,9 +457,9 @@ - { - unsigned long flags; - -- local_irq_save(flags); -+ local_irq_save_nort(flags); - if (!atomic_dec_and_lock(&congested->refcnt, &cgwb_lock)) { -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - return; - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/compaction.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/compaction.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/compaction.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/compaction.c 2017-04-18 17:54:26.000000000 +0200 -@@ -1593,10 +1593,12 @@ - block_start_pfn(cc->migrate_pfn, cc->order); - - if (cc->last_migrated_pfn < current_block_start) { -- cpu = get_cpu(); -+ cpu = get_cpu_light(); -+ local_lock_irq(swapvec_lock); - lru_add_drain_cpu(cpu); -+ local_unlock_irq(swapvec_lock); - drain_local_pages(zone); -- put_cpu(); -+ put_cpu_light(); - /* No more flushing until we migrate again */ - cc->last_migrated_pfn = 0; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/filemap.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/filemap.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/filemap.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/filemap.c 2017-04-18 17:54:26.000000000 +0200 -@@ -159,9 +159,12 @@ - * node->private_list is protected by - * mapping->tree_lock. - */ -- if (!list_empty(&node->private_list)) -- list_lru_del(&workingset_shadow_nodes, -+ if (!list_empty(&node->private_list)) { -+ local_lock(workingset_shadow_lock); -+ list_lru_del(&__workingset_shadow_nodes, - &node->private_list); -+ local_unlock(workingset_shadow_lock); -+ } - } - return 0; - } -@@ -217,8 +220,10 @@ - if (!dax_mapping(mapping) && !workingset_node_pages(node) && - list_empty(&node->private_list)) { - node->private_data = mapping; -- list_lru_add(&workingset_shadow_nodes, -- &node->private_list); -+ local_lock(workingset_shadow_lock); -+ list_lru_add(&__workingset_shadow_nodes, -+ &node->private_list); -+ local_unlock(workingset_shadow_lock); - } - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/highmem.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/highmem.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/highmem.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/highmem.c 2017-04-18 17:54:26.000000000 +0200 -@@ -29,10 +29,11 @@ - #include - #include - -- -+#ifndef CONFIG_PREEMPT_RT_FULL - #if defined(CONFIG_HIGHMEM) || defined(CONFIG_X86_32) - DEFINE_PER_CPU(int, __kmap_atomic_idx); - #endif -+#endif - - /* - * Virtual_count is not a pure "count". -@@ -107,8 +108,9 @@ - unsigned long totalhigh_pages __read_mostly; - EXPORT_SYMBOL(totalhigh_pages); - -- -+#ifndef CONFIG_PREEMPT_RT_FULL - EXPORT_PER_CPU_SYMBOL(__kmap_atomic_idx); -+#endif - - unsigned int nr_free_highpages (void) - { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/memcontrol.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/memcontrol.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/memcontrol.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/memcontrol.c 2017-04-18 17:54:26.000000000 +0200 -@@ -67,6 +67,7 @@ - #include - #include - #include "slab.h" -+#include - - #include - -@@ -92,6 +93,8 @@ - #define do_swap_account 0 - #endif - -+static DEFINE_LOCAL_IRQ_LOCK(event_lock); -+ - /* Whether legacy memory+swap accounting is active */ - static bool do_memsw_account(void) - { -@@ -1692,6 +1695,7 @@ - #define FLUSHING_CACHED_CHARGE 0 - }; - static DEFINE_PER_CPU(struct memcg_stock_pcp, memcg_stock); -+static DEFINE_LOCAL_IRQ_LOCK(memcg_stock_ll); - static DEFINE_MUTEX(percpu_charge_mutex); - - /** -@@ -1714,7 +1718,7 @@ - if (nr_pages > CHARGE_BATCH) - return ret; - -- local_irq_save(flags); -+ local_lock_irqsave(memcg_stock_ll, flags); - - stock = this_cpu_ptr(&memcg_stock); - if (memcg == stock->cached && stock->nr_pages >= nr_pages) { -@@ -1722,7 +1726,7 @@ - ret = true; - } - -- local_irq_restore(flags); -+ local_unlock_irqrestore(memcg_stock_ll, flags); - - return ret; - } -@@ -1749,13 +1753,13 @@ - struct memcg_stock_pcp *stock; - unsigned long flags; - -- local_irq_save(flags); -+ local_lock_irqsave(memcg_stock_ll, flags); - - stock = this_cpu_ptr(&memcg_stock); - drain_stock(stock); - clear_bit(FLUSHING_CACHED_CHARGE, &stock->flags); - -- local_irq_restore(flags); -+ local_unlock_irqrestore(memcg_stock_ll, flags); - } - - /* -@@ -1767,7 +1771,7 @@ - struct memcg_stock_pcp *stock; - unsigned long flags; - -- local_irq_save(flags); -+ local_lock_irqsave(memcg_stock_ll, flags); - - stock = this_cpu_ptr(&memcg_stock); - if (stock->cached != memcg) { /* reset if necessary */ -@@ -1776,7 +1780,7 @@ - } - stock->nr_pages += nr_pages; - -- local_irq_restore(flags); -+ local_unlock_irqrestore(memcg_stock_ll, flags); - } - - /* -@@ -1792,7 +1796,7 @@ - return; - /* Notify other cpus that system-wide "drain" is running */ - get_online_cpus(); -- curcpu = get_cpu(); -+ curcpu = get_cpu_light(); - for_each_online_cpu(cpu) { - struct memcg_stock_pcp *stock = &per_cpu(memcg_stock, cpu); - struct mem_cgroup *memcg; -@@ -1809,7 +1813,7 @@ - schedule_work_on(cpu, &stock->work); - } - } -- put_cpu(); -+ put_cpu_light(); - put_online_cpus(); - mutex_unlock(&percpu_charge_mutex); - } -@@ -4553,12 +4557,12 @@ - - ret = 0; - -- local_irq_disable(); -+ local_lock_irq(event_lock); - mem_cgroup_charge_statistics(to, page, compound, nr_pages); - memcg_check_events(to, page); - mem_cgroup_charge_statistics(from, page, compound, -nr_pages); - memcg_check_events(from, page); -- local_irq_enable(); -+ local_unlock_irq(event_lock); - out_unlock: - unlock_page(page); - out: -@@ -5433,10 +5437,10 @@ - - commit_charge(page, memcg, lrucare); - -- local_irq_disable(); -+ local_lock_irq(event_lock); - mem_cgroup_charge_statistics(memcg, page, compound, nr_pages); - memcg_check_events(memcg, page); -- local_irq_enable(); -+ local_unlock_irq(event_lock); - - if (do_memsw_account() && PageSwapCache(page)) { - swp_entry_t entry = { .val = page_private(page) }; -@@ -5492,14 +5496,14 @@ - memcg_oom_recover(memcg); - } - -- local_irq_save(flags); -+ local_lock_irqsave(event_lock, flags); - __this_cpu_sub(memcg->stat->count[MEM_CGROUP_STAT_RSS], nr_anon); - __this_cpu_sub(memcg->stat->count[MEM_CGROUP_STAT_CACHE], nr_file); - __this_cpu_sub(memcg->stat->count[MEM_CGROUP_STAT_RSS_HUGE], nr_huge); - __this_cpu_add(memcg->stat->events[MEM_CGROUP_EVENTS_PGPGOUT], pgpgout); - __this_cpu_add(memcg->stat->nr_page_events, nr_pages); - memcg_check_events(memcg, dummy_page); -- local_irq_restore(flags); -+ local_unlock_irqrestore(event_lock, flags); - - if (!mem_cgroup_is_root(memcg)) - css_put_many(&memcg->css, nr_pages); -@@ -5654,10 +5658,10 @@ - - commit_charge(newpage, memcg, false); - -- local_irq_save(flags); -+ local_lock_irqsave(event_lock, flags); - mem_cgroup_charge_statistics(memcg, newpage, compound, nr_pages); - memcg_check_events(memcg, newpage); -- local_irq_restore(flags); -+ local_unlock_irqrestore(event_lock, flags); - } - - DEFINE_STATIC_KEY_FALSE(memcg_sockets_enabled_key); -@@ -5837,6 +5841,7 @@ - { - struct mem_cgroup *memcg, *swap_memcg; - unsigned short oldid; -+ unsigned long flags; - - VM_BUG_ON_PAGE(PageLRU(page), page); - VM_BUG_ON_PAGE(page_count(page), page); -@@ -5877,12 +5882,16 @@ - * important here to have the interrupts disabled because it is the - * only synchronisation we have for udpating the per-CPU variables. - */ -+ local_lock_irqsave(event_lock, flags); -+#ifndef CONFIG_PREEMPT_RT_BASE - VM_BUG_ON(!irqs_disabled()); -+#endif - mem_cgroup_charge_statistics(memcg, page, false, -1); - memcg_check_events(memcg, page); - - if (!mem_cgroup_is_root(memcg)) - css_put(&memcg->css); -+ local_unlock_irqrestore(event_lock, flags); - } - - /* -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/mmu_context.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/mmu_context.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/mmu_context.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/mmu_context.c 2017-04-18 17:54:26.000000000 +0200 -@@ -23,6 +23,7 @@ - struct task_struct *tsk = current; - - task_lock(tsk); -+ preempt_disable_rt(); - active_mm = tsk->active_mm; - if (active_mm != mm) { - atomic_inc(&mm->mm_count); -@@ -30,6 +31,7 @@ - } - tsk->mm = mm; - switch_mm(active_mm, mm, tsk); -+ preempt_enable_rt(); - task_unlock(tsk); - #ifdef finish_arch_post_lock_switch - finish_arch_post_lock_switch(); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/page_alloc.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/page_alloc.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/page_alloc.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/page_alloc.c 2017-04-18 17:54:26.000000000 +0200 -@@ -61,6 +61,7 @@ - #include - #include - #include -+#include - #include - #include - #include -@@ -281,6 +282,18 @@ - EXPORT_SYMBOL(nr_online_nodes); - #endif - -+static DEFINE_LOCAL_IRQ_LOCK(pa_lock); -+ -+#ifdef CONFIG_PREEMPT_RT_BASE -+# define cpu_lock_irqsave(cpu, flags) \ -+ local_lock_irqsave_on(pa_lock, flags, cpu) -+# define cpu_unlock_irqrestore(cpu, flags) \ -+ local_unlock_irqrestore_on(pa_lock, flags, cpu) -+#else -+# define cpu_lock_irqsave(cpu, flags) local_irq_save(flags) -+# define cpu_unlock_irqrestore(cpu, flags) local_irq_restore(flags) -+#endif -+ - int page_group_by_mobility_disabled __read_mostly; - - #ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT -@@ -1072,7 +1085,7 @@ - #endif /* CONFIG_DEBUG_VM */ - - /* -- * Frees a number of pages from the PCP lists -+ * Frees a number of pages which have been collected from the pcp lists. - * Assumes all pages on list are in same zone, and of same order. - * count is the number of pages to free. - * -@@ -1083,19 +1096,58 @@ - * pinned" detection logic. - */ - static void free_pcppages_bulk(struct zone *zone, int count, -- struct per_cpu_pages *pcp) -+ struct list_head *list) - { -- int migratetype = 0; -- int batch_free = 0; - unsigned long nr_scanned; - bool isolated_pageblocks; -+ unsigned long flags; -+ -+ spin_lock_irqsave(&zone->lock, flags); - -- spin_lock(&zone->lock); - isolated_pageblocks = has_isolate_pageblock(zone); - nr_scanned = node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED); - if (nr_scanned) - __mod_node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED, -nr_scanned); - -+ while (!list_empty(list)) { -+ struct page *page; -+ int mt; /* migratetype of the to-be-freed page */ -+ -+ page = list_first_entry(list, struct page, lru); -+ /* must delete as __free_one_page list manipulates */ -+ list_del(&page->lru); -+ -+ mt = get_pcppage_migratetype(page); -+ /* MIGRATE_ISOLATE page should not go to pcplists */ -+ VM_BUG_ON_PAGE(is_migrate_isolate(mt), page); -+ /* Pageblock could have been isolated meanwhile */ -+ if (unlikely(isolated_pageblocks)) -+ mt = get_pageblock_migratetype(page); -+ -+ if (bulkfree_pcp_prepare(page)) -+ continue; -+ -+ __free_one_page(page, page_to_pfn(page), zone, 0, mt); -+ trace_mm_page_pcpu_drain(page, 0, mt); -+ count--; -+ } -+ WARN_ON(count != 0); -+ spin_unlock_irqrestore(&zone->lock, flags); -+} -+ -+/* -+ * Moves a number of pages from the PCP lists to free list which -+ * is freed outside of the locked region. -+ * -+ * Assumes all pages on list are in same zone, and of same order. -+ * count is the number of pages to free. -+ */ -+static void isolate_pcp_pages(int count, struct per_cpu_pages *src, -+ struct list_head *dst) -+{ -+ int migratetype = 0; -+ int batch_free = 0; -+ - while (count) { - struct page *page; - struct list_head *list; -@@ -1111,7 +1163,7 @@ - batch_free++; - if (++migratetype == MIGRATE_PCPTYPES) - migratetype = 0; -- list = &pcp->lists[migratetype]; -+ list = &src->lists[migratetype]; - } while (list_empty(list)); - - /* This is the only non-empty list. Free them all. */ -@@ -1119,27 +1171,12 @@ - batch_free = count; - - do { -- int mt; /* migratetype of the to-be-freed page */ -- - page = list_last_entry(list, struct page, lru); -- /* must delete as __free_one_page list manipulates */ - list_del(&page->lru); - -- mt = get_pcppage_migratetype(page); -- /* MIGRATE_ISOLATE page should not go to pcplists */ -- VM_BUG_ON_PAGE(is_migrate_isolate(mt), page); -- /* Pageblock could have been isolated meanwhile */ -- if (unlikely(isolated_pageblocks)) -- mt = get_pageblock_migratetype(page); -- -- if (bulkfree_pcp_prepare(page)) -- continue; -- -- __free_one_page(page, page_to_pfn(page), zone, 0, mt); -- trace_mm_page_pcpu_drain(page, 0, mt); -+ list_add(&page->lru, dst); - } while (--count && --batch_free && !list_empty(list)); - } -- spin_unlock(&zone->lock); - } - - static void free_one_page(struct zone *zone, -@@ -1148,7 +1185,9 @@ - int migratetype) - { - unsigned long nr_scanned; -- spin_lock(&zone->lock); -+ unsigned long flags; -+ -+ spin_lock_irqsave(&zone->lock, flags); - nr_scanned = node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED); - if (nr_scanned) - __mod_node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED, -nr_scanned); -@@ -1158,7 +1197,7 @@ - migratetype = get_pfnblock_migratetype(page, pfn); - } - __free_one_page(page, pfn, zone, order, migratetype); -- spin_unlock(&zone->lock); -+ spin_unlock_irqrestore(&zone->lock, flags); - } - - static void __meminit __init_single_page(struct page *page, unsigned long pfn, -@@ -1244,10 +1283,10 @@ - return; - - migratetype = get_pfnblock_migratetype(page, pfn); -- local_irq_save(flags); -+ local_lock_irqsave(pa_lock, flags); - __count_vm_events(PGFREE, 1 << order); - free_one_page(page_zone(page), page, pfn, order, migratetype); -- local_irq_restore(flags); -+ local_unlock_irqrestore(pa_lock, flags); - } - - static void __init __free_pages_boot_core(struct page *page, unsigned int order) -@@ -2246,16 +2285,18 @@ - void drain_zone_pages(struct zone *zone, struct per_cpu_pages *pcp) - { - unsigned long flags; -+ LIST_HEAD(dst); - int to_drain, batch; - -- local_irq_save(flags); -+ local_lock_irqsave(pa_lock, flags); - batch = READ_ONCE(pcp->batch); - to_drain = min(pcp->count, batch); - if (to_drain > 0) { -- free_pcppages_bulk(zone, to_drain, pcp); -+ isolate_pcp_pages(to_drain, pcp, &dst); - pcp->count -= to_drain; - } -- local_irq_restore(flags); -+ local_unlock_irqrestore(pa_lock, flags); -+ free_pcppages_bulk(zone, to_drain, &dst); - } - #endif - -@@ -2271,16 +2312,21 @@ - unsigned long flags; - struct per_cpu_pageset *pset; - struct per_cpu_pages *pcp; -+ LIST_HEAD(dst); -+ int count; - -- local_irq_save(flags); -+ cpu_lock_irqsave(cpu, flags); - pset = per_cpu_ptr(zone->pageset, cpu); - - pcp = &pset->pcp; -- if (pcp->count) { -- free_pcppages_bulk(zone, pcp->count, pcp); -+ count = pcp->count; -+ if (count) { -+ isolate_pcp_pages(count, pcp, &dst); - pcp->count = 0; - } -- local_irq_restore(flags); -+ cpu_unlock_irqrestore(cpu, flags); -+ if (count) -+ free_pcppages_bulk(zone, count, &dst); - } - - /* -@@ -2366,8 +2412,17 @@ - else - cpumask_clear_cpu(cpu, &cpus_with_pcps); - } -+#ifndef CONFIG_PREEMPT_RT_BASE - on_each_cpu_mask(&cpus_with_pcps, (smp_call_func_t) drain_local_pages, - zone, 1); -+#else -+ for_each_cpu(cpu, &cpus_with_pcps) { -+ if (zone) -+ drain_pages_zone(cpu, zone); -+ else -+ drain_pages(cpu); -+ } -+#endif - } - - #ifdef CONFIG_HIBERNATION -@@ -2427,7 +2482,7 @@ - - migratetype = get_pfnblock_migratetype(page, pfn); - set_pcppage_migratetype(page, migratetype); -- local_irq_save(flags); -+ local_lock_irqsave(pa_lock, flags); - __count_vm_event(PGFREE); - - /* -@@ -2453,12 +2508,17 @@ - pcp->count++; - if (pcp->count >= pcp->high) { - unsigned long batch = READ_ONCE(pcp->batch); -- free_pcppages_bulk(zone, batch, pcp); -+ LIST_HEAD(dst); -+ -+ isolate_pcp_pages(batch, pcp, &dst); - pcp->count -= batch; -+ local_unlock_irqrestore(pa_lock, flags); -+ free_pcppages_bulk(zone, batch, &dst); -+ return; - } - - out: -- local_irq_restore(flags); -+ local_unlock_irqrestore(pa_lock, flags); - } - - /* -@@ -2600,7 +2660,7 @@ - struct per_cpu_pages *pcp; - struct list_head *list; - -- local_irq_save(flags); -+ local_lock_irqsave(pa_lock, flags); - do { - pcp = &this_cpu_ptr(zone->pageset)->pcp; - list = &pcp->lists[migratetype]; -@@ -2627,7 +2687,7 @@ - * allocate greater than order-1 page units with __GFP_NOFAIL. - */ - WARN_ON_ONCE((gfp_flags & __GFP_NOFAIL) && (order > 1)); -- spin_lock_irqsave(&zone->lock, flags); -+ local_spin_lock_irqsave(pa_lock, &zone->lock, flags); - - do { - page = NULL; -@@ -2639,22 +2699,24 @@ - if (!page) - page = __rmqueue(zone, order, migratetype); - } while (page && check_new_pages(page, order)); -- spin_unlock(&zone->lock); -- if (!page) -+ if (!page) { -+ spin_unlock(&zone->lock); - goto failed; -+ } - __mod_zone_freepage_state(zone, -(1 << order), - get_pcppage_migratetype(page)); -+ spin_unlock(&zone->lock); - } - - __count_zid_vm_events(PGALLOC, page_zonenum(page), 1 << order); - zone_statistics(preferred_zone, zone, gfp_flags); -- local_irq_restore(flags); -+ local_unlock_irqrestore(pa_lock, flags); - - VM_BUG_ON_PAGE(bad_range(zone, page), page); - return page; - - failed: -- local_irq_restore(flags); -+ local_unlock_irqrestore(pa_lock, flags); - return NULL; - } - -@@ -6531,7 +6593,9 @@ - int cpu = (unsigned long)hcpu; - - if (action == CPU_DEAD || action == CPU_DEAD_FROZEN) { -+ local_lock_irq_on(swapvec_lock, cpu); - lru_add_drain_cpu(cpu); -+ local_unlock_irq_on(swapvec_lock, cpu); - drain_pages(cpu); - - /* -@@ -6557,6 +6621,7 @@ - void __init page_alloc_init(void) - { - hotcpu_notifier(page_alloc_cpu_notify, 0); -+ local_irq_lock_init(pa_lock); - } - - /* -@@ -7383,7 +7448,7 @@ - struct per_cpu_pageset *pset; - - /* avoid races with drain_pages() */ -- local_irq_save(flags); -+ local_lock_irqsave(pa_lock, flags); - if (zone->pageset != &boot_pageset) { - for_each_online_cpu(cpu) { - pset = per_cpu_ptr(zone->pageset, cpu); -@@ -7392,7 +7457,7 @@ - free_percpu(zone->pageset); - zone->pageset = &boot_pageset; - } -- local_irq_restore(flags); -+ local_unlock_irqrestore(pa_lock, flags); - } - - #ifdef CONFIG_MEMORY_HOTREMOVE -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/percpu.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/percpu.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/percpu.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/percpu.c 2017-04-18 17:54:26.000000000 +0200 -@@ -1283,18 +1283,7 @@ - } - EXPORT_SYMBOL_GPL(free_percpu); - --/** -- * is_kernel_percpu_address - test whether address is from static percpu area -- * @addr: address to test -- * -- * Test whether @addr belongs to in-kernel static percpu area. Module -- * static percpu areas are not considered. For those, use -- * is_module_percpu_address(). -- * -- * RETURNS: -- * %true if @addr is from in-kernel static percpu area, %false otherwise. -- */ --bool is_kernel_percpu_address(unsigned long addr) -+bool __is_kernel_percpu_address(unsigned long addr, unsigned long *can_addr) - { - #ifdef CONFIG_SMP - const size_t static_size = __per_cpu_end - __per_cpu_start; -@@ -1303,16 +1292,39 @@ - - for_each_possible_cpu(cpu) { - void *start = per_cpu_ptr(base, cpu); -+ void *va = (void *)addr; - -- if ((void *)addr >= start && (void *)addr < start + static_size) -+ if (va >= start && va < start + static_size) { -+ if (can_addr) { -+ *can_addr = (unsigned long) (va - start); -+ *can_addr += (unsigned long) -+ per_cpu_ptr(base, get_boot_cpu_id()); -+ } - return true; -- } -+ } -+ } - #endif - /* on UP, can't distinguish from other static vars, always false */ - return false; - } - - /** -+ * is_kernel_percpu_address - test whether address is from static percpu area -+ * @addr: address to test -+ * -+ * Test whether @addr belongs to in-kernel static percpu area. Module -+ * static percpu areas are not considered. For those, use -+ * is_module_percpu_address(). -+ * -+ * RETURNS: -+ * %true if @addr is from in-kernel static percpu area, %false otherwise. -+ */ -+bool is_kernel_percpu_address(unsigned long addr) -+{ -+ return __is_kernel_percpu_address(addr, NULL); -+} -+ -+/** - * per_cpu_ptr_to_phys - convert translated percpu address to physical address - * @addr: the address to be converted to physical address - * -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/slab.h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/slab.h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/slab.h 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/slab.h 2017-04-18 17:54:26.000000000 +0200 -@@ -426,7 +426,11 @@ - * The slab lists for all objects. - */ - struct kmem_cache_node { -+#ifdef CONFIG_SLUB -+ raw_spinlock_t list_lock; -+#else - spinlock_t list_lock; -+#endif - - #ifdef CONFIG_SLAB - struct list_head slabs_partial; /* partial list first, better asm code */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/slub.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/slub.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/slub.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/slub.c 2017-04-18 17:54:26.000000000 +0200 -@@ -1141,7 +1141,7 @@ - unsigned long uninitialized_var(flags); - int ret = 0; - -- spin_lock_irqsave(&n->list_lock, flags); -+ raw_spin_lock_irqsave(&n->list_lock, flags); - slab_lock(page); - - if (s->flags & SLAB_CONSISTENCY_CHECKS) { -@@ -1176,7 +1176,7 @@ - bulk_cnt, cnt); - - slab_unlock(page); -- spin_unlock_irqrestore(&n->list_lock, flags); -+ raw_spin_unlock_irqrestore(&n->list_lock, flags); - if (!ret) - slab_fix(s, "Object at 0x%p not freed", object); - return ret; -@@ -1304,6 +1304,12 @@ - - #endif /* CONFIG_SLUB_DEBUG */ - -+struct slub_free_list { -+ raw_spinlock_t lock; -+ struct list_head list; -+}; -+static DEFINE_PER_CPU(struct slub_free_list, slub_free_list); -+ - /* - * Hooks for other subsystems that check memory allocations. In a typical - * production configuration these hooks all should produce no code at all. -@@ -1527,10 +1533,17 @@ - void *start, *p; - int idx, order; - bool shuffle; -+ bool enableirqs = false; - - flags &= gfp_allowed_mask; - - if (gfpflags_allow_blocking(flags)) -+ enableirqs = true; -+#ifdef CONFIG_PREEMPT_RT_FULL -+ if (system_state == SYSTEM_RUNNING) -+ enableirqs = true; -+#endif -+ if (enableirqs) - local_irq_enable(); - - flags |= s->allocflags; -@@ -1605,7 +1618,7 @@ - page->frozen = 1; - - out: -- if (gfpflags_allow_blocking(flags)) -+ if (enableirqs) - local_irq_disable(); - if (!page) - return NULL; -@@ -1664,6 +1677,16 @@ - __free_pages(page, order); - } - -+static void free_delayed(struct list_head *h) -+{ -+ while(!list_empty(h)) { -+ struct page *page = list_first_entry(h, struct page, lru); -+ -+ list_del(&page->lru); -+ __free_slab(page->slab_cache, page); -+ } -+} -+ - #define need_reserve_slab_rcu \ - (sizeof(((struct page *)NULL)->lru) < sizeof(struct rcu_head)) - -@@ -1695,6 +1718,12 @@ - } - - call_rcu(head, rcu_free_slab); -+ } else if (irqs_disabled()) { -+ struct slub_free_list *f = this_cpu_ptr(&slub_free_list); -+ -+ raw_spin_lock(&f->lock); -+ list_add(&page->lru, &f->list); -+ raw_spin_unlock(&f->lock); - } else - __free_slab(s, page); - } -@@ -1802,7 +1831,7 @@ - if (!n || !n->nr_partial) - return NULL; - -- spin_lock(&n->list_lock); -+ raw_spin_lock(&n->list_lock); - list_for_each_entry_safe(page, page2, &n->partial, lru) { - void *t; - -@@ -1827,7 +1856,7 @@ - break; - - } -- spin_unlock(&n->list_lock); -+ raw_spin_unlock(&n->list_lock); - return object; - } - -@@ -2073,7 +2102,7 @@ - * that acquire_slab() will see a slab page that - * is frozen - */ -- spin_lock(&n->list_lock); -+ raw_spin_lock(&n->list_lock); - } - } else { - m = M_FULL; -@@ -2084,7 +2113,7 @@ - * slabs from diagnostic functions will not see - * any frozen slabs. - */ -- spin_lock(&n->list_lock); -+ raw_spin_lock(&n->list_lock); - } - } - -@@ -2119,7 +2148,7 @@ - goto redo; - - if (lock) -- spin_unlock(&n->list_lock); -+ raw_spin_unlock(&n->list_lock); - - if (m == M_FREE) { - stat(s, DEACTIVATE_EMPTY); -@@ -2151,10 +2180,10 @@ - n2 = get_node(s, page_to_nid(page)); - if (n != n2) { - if (n) -- spin_unlock(&n->list_lock); -+ raw_spin_unlock(&n->list_lock); - - n = n2; -- spin_lock(&n->list_lock); -+ raw_spin_lock(&n->list_lock); - } - - do { -@@ -2183,7 +2212,7 @@ - } - - if (n) -- spin_unlock(&n->list_lock); -+ raw_spin_unlock(&n->list_lock); - - while (discard_page) { - page = discard_page; -@@ -2222,14 +2251,21 @@ - pobjects = oldpage->pobjects; - pages = oldpage->pages; - if (drain && pobjects > s->cpu_partial) { -+ struct slub_free_list *f; - unsigned long flags; -+ LIST_HEAD(tofree); - /* - * partial array is full. Move the existing - * set to the per node partial list. - */ - local_irq_save(flags); - unfreeze_partials(s, this_cpu_ptr(s->cpu_slab)); -+ f = this_cpu_ptr(&slub_free_list); -+ raw_spin_lock(&f->lock); -+ list_splice_init(&f->list, &tofree); -+ raw_spin_unlock(&f->lock); - local_irq_restore(flags); -+ free_delayed(&tofree); - oldpage = NULL; - pobjects = 0; - pages = 0; -@@ -2301,7 +2337,22 @@ - - static void flush_all(struct kmem_cache *s) - { -+ LIST_HEAD(tofree); -+ int cpu; -+ - on_each_cpu_cond(has_cpu_slab, flush_cpu_slab, s, 1, GFP_ATOMIC); -+ for_each_online_cpu(cpu) { -+ struct slub_free_list *f; -+ -+ if (!has_cpu_slab(cpu, s)) -+ continue; -+ -+ f = &per_cpu(slub_free_list, cpu); -+ raw_spin_lock_irq(&f->lock); -+ list_splice_init(&f->list, &tofree); -+ raw_spin_unlock_irq(&f->lock); -+ free_delayed(&tofree); -+ } - } - - /* -@@ -2356,10 +2407,10 @@ - unsigned long x = 0; - struct page *page; - -- spin_lock_irqsave(&n->list_lock, flags); -+ raw_spin_lock_irqsave(&n->list_lock, flags); - list_for_each_entry(page, &n->partial, lru) - x += get_count(page); -- spin_unlock_irqrestore(&n->list_lock, flags); -+ raw_spin_unlock_irqrestore(&n->list_lock, flags); - return x; - } - #endif /* CONFIG_SLUB_DEBUG || CONFIG_SYSFS */ -@@ -2497,8 +2548,10 @@ - * already disabled (which is the case for bulk allocation). - */ - static void *___slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, -- unsigned long addr, struct kmem_cache_cpu *c) -+ unsigned long addr, struct kmem_cache_cpu *c, -+ struct list_head *to_free) - { -+ struct slub_free_list *f; - void *freelist; - struct page *page; - -@@ -2558,6 +2611,13 @@ - VM_BUG_ON(!c->page->frozen); - c->freelist = get_freepointer(s, freelist); - c->tid = next_tid(c->tid); -+ -+out: -+ f = this_cpu_ptr(&slub_free_list); -+ raw_spin_lock(&f->lock); -+ list_splice_init(&f->list, to_free); -+ raw_spin_unlock(&f->lock); -+ - return freelist; - - new_slab: -@@ -2589,7 +2649,7 @@ - deactivate_slab(s, page, get_freepointer(s, freelist)); - c->page = NULL; - c->freelist = NULL; -- return freelist; -+ goto out; - } - - /* -@@ -2601,6 +2661,7 @@ - { - void *p; - unsigned long flags; -+ LIST_HEAD(tofree); - - local_irq_save(flags); - #ifdef CONFIG_PREEMPT -@@ -2612,8 +2673,9 @@ - c = this_cpu_ptr(s->cpu_slab); - #endif - -- p = ___slab_alloc(s, gfpflags, node, addr, c); -+ p = ___slab_alloc(s, gfpflags, node, addr, c, &tofree); - local_irq_restore(flags); -+ free_delayed(&tofree); - return p; - } - -@@ -2799,7 +2861,7 @@ - - do { - if (unlikely(n)) { -- spin_unlock_irqrestore(&n->list_lock, flags); -+ raw_spin_unlock_irqrestore(&n->list_lock, flags); - n = NULL; - } - prior = page->freelist; -@@ -2831,7 +2893,7 @@ - * Otherwise the list_lock will synchronize with - * other processors updating the list of slabs. - */ -- spin_lock_irqsave(&n->list_lock, flags); -+ raw_spin_lock_irqsave(&n->list_lock, flags); - - } - } -@@ -2873,7 +2935,7 @@ - add_partial(n, page, DEACTIVATE_TO_TAIL); - stat(s, FREE_ADD_PARTIAL); - } -- spin_unlock_irqrestore(&n->list_lock, flags); -+ raw_spin_unlock_irqrestore(&n->list_lock, flags); - return; - - slab_empty: -@@ -2888,7 +2950,7 @@ - remove_full(s, n, page); - } - -- spin_unlock_irqrestore(&n->list_lock, flags); -+ raw_spin_unlock_irqrestore(&n->list_lock, flags); - stat(s, FREE_SLAB); - discard_slab(s, page); - } -@@ -3093,6 +3155,7 @@ - void **p) - { - struct kmem_cache_cpu *c; -+ LIST_HEAD(to_free); - int i; - - /* memcg and kmem_cache debug support */ -@@ -3116,7 +3179,7 @@ - * of re-populating per CPU c->freelist - */ - p[i] = ___slab_alloc(s, flags, NUMA_NO_NODE, -- _RET_IP_, c); -+ _RET_IP_, c, &to_free); - if (unlikely(!p[i])) - goto error; - -@@ -3128,6 +3191,7 @@ - } - c->tid = next_tid(c->tid); - local_irq_enable(); -+ free_delayed(&to_free); - - /* Clear memory outside IRQ disabled fastpath loop */ - if (unlikely(flags & __GFP_ZERO)) { -@@ -3275,7 +3339,7 @@ - init_kmem_cache_node(struct kmem_cache_node *n) - { - n->nr_partial = 0; -- spin_lock_init(&n->list_lock); -+ raw_spin_lock_init(&n->list_lock); - INIT_LIST_HEAD(&n->partial); - #ifdef CONFIG_SLUB_DEBUG - atomic_long_set(&n->nr_slabs, 0); -@@ -3619,6 +3683,10 @@ - const char *text) - { - #ifdef CONFIG_SLUB_DEBUG -+#ifdef CONFIG_PREEMPT_RT_BASE -+ /* XXX move out of irq-off section */ -+ slab_err(s, page, text, s->name); -+#else - void *addr = page_address(page); - void *p; - unsigned long *map = kzalloc(BITS_TO_LONGS(page->objects) * -@@ -3639,6 +3707,7 @@ - slab_unlock(page); - kfree(map); - #endif -+#endif - } - - /* -@@ -3652,7 +3721,7 @@ - struct page *page, *h; - - BUG_ON(irqs_disabled()); -- spin_lock_irq(&n->list_lock); -+ raw_spin_lock_irq(&n->list_lock); - list_for_each_entry_safe(page, h, &n->partial, lru) { - if (!page->inuse) { - remove_partial(n, page); -@@ -3662,7 +3731,7 @@ - "Objects remaining in %s on __kmem_cache_shutdown()"); - } - } -- spin_unlock_irq(&n->list_lock); -+ raw_spin_unlock_irq(&n->list_lock); - - list_for_each_entry_safe(page, h, &discard, lru) - discard_slab(s, page); -@@ -3905,7 +3974,7 @@ - for (i = 0; i < SHRINK_PROMOTE_MAX; i++) - INIT_LIST_HEAD(promote + i); - -- spin_lock_irqsave(&n->list_lock, flags); -+ raw_spin_lock_irqsave(&n->list_lock, flags); - - /* - * Build lists of slabs to discard or promote. -@@ -3936,7 +4005,7 @@ - for (i = SHRINK_PROMOTE_MAX - 1; i >= 0; i--) - list_splice(promote + i, &n->partial); - -- spin_unlock_irqrestore(&n->list_lock, flags); -+ raw_spin_unlock_irqrestore(&n->list_lock, flags); - - /* Release empty slabs */ - list_for_each_entry_safe(page, t, &discard, lru) -@@ -4112,6 +4181,12 @@ - { - static __initdata struct kmem_cache boot_kmem_cache, - boot_kmem_cache_node; -+ int cpu; -+ -+ for_each_possible_cpu(cpu) { -+ raw_spin_lock_init(&per_cpu(slub_free_list, cpu).lock); -+ INIT_LIST_HEAD(&per_cpu(slub_free_list, cpu).list); -+ } - - if (debug_guardpage_minorder()) - slub_max_order = 0; -@@ -4320,7 +4395,7 @@ - struct page *page; - unsigned long flags; - -- spin_lock_irqsave(&n->list_lock, flags); -+ raw_spin_lock_irqsave(&n->list_lock, flags); - - list_for_each_entry(page, &n->partial, lru) { - validate_slab_slab(s, page, map); -@@ -4342,7 +4417,7 @@ - s->name, count, atomic_long_read(&n->nr_slabs)); - - out: -- spin_unlock_irqrestore(&n->list_lock, flags); -+ raw_spin_unlock_irqrestore(&n->list_lock, flags); - return count; - } - -@@ -4530,12 +4605,12 @@ - if (!atomic_long_read(&n->nr_slabs)) - continue; - -- spin_lock_irqsave(&n->list_lock, flags); -+ raw_spin_lock_irqsave(&n->list_lock, flags); - list_for_each_entry(page, &n->partial, lru) - process_slab(&t, s, page, alloc, map); - list_for_each_entry(page, &n->full, lru) - process_slab(&t, s, page, alloc, map); -- spin_unlock_irqrestore(&n->list_lock, flags); -+ raw_spin_unlock_irqrestore(&n->list_lock, flags); - } - - for (i = 0; i < t.count; i++) { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/swap.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/swap.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/swap.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/swap.c 2017-04-18 17:54:26.000000000 +0200 -@@ -32,6 +32,7 @@ - #include - #include - #include -+#include - #include - #include - -@@ -50,6 +51,8 @@ - #ifdef CONFIG_SMP - static DEFINE_PER_CPU(struct pagevec, activate_page_pvecs); - #endif -+static DEFINE_LOCAL_IRQ_LOCK(rotate_lock); -+DEFINE_LOCAL_IRQ_LOCK(swapvec_lock); - - /* - * This path almost never happens for VM activity - pages are normally -@@ -240,11 +243,11 @@ - unsigned long flags; - - get_page(page); -- local_irq_save(flags); -+ local_lock_irqsave(rotate_lock, flags); - pvec = this_cpu_ptr(&lru_rotate_pvecs); - if (!pagevec_add(pvec, page) || PageCompound(page)) - pagevec_move_tail(pvec); -- local_irq_restore(flags); -+ local_unlock_irqrestore(rotate_lock, flags); - } - } - -@@ -294,12 +297,13 @@ - { - page = compound_head(page); - if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) { -- struct pagevec *pvec = &get_cpu_var(activate_page_pvecs); -+ struct pagevec *pvec = &get_locked_var(swapvec_lock, -+ activate_page_pvecs); - - get_page(page); - if (!pagevec_add(pvec, page) || PageCompound(page)) - pagevec_lru_move_fn(pvec, __activate_page, NULL); -- put_cpu_var(activate_page_pvecs); -+ put_locked_var(swapvec_lock, activate_page_pvecs); - } - } - -@@ -326,7 +330,7 @@ - - static void __lru_cache_activate_page(struct page *page) - { -- struct pagevec *pvec = &get_cpu_var(lru_add_pvec); -+ struct pagevec *pvec = &get_locked_var(swapvec_lock, lru_add_pvec); - int i; - - /* -@@ -348,7 +352,7 @@ - } - } - -- put_cpu_var(lru_add_pvec); -+ put_locked_var(swapvec_lock, lru_add_pvec); - } - - /* -@@ -390,12 +394,12 @@ - - static void __lru_cache_add(struct page *page) - { -- struct pagevec *pvec = &get_cpu_var(lru_add_pvec); -+ struct pagevec *pvec = &get_locked_var(swapvec_lock, lru_add_pvec); - - get_page(page); - if (!pagevec_add(pvec, page) || PageCompound(page)) - __pagevec_lru_add(pvec); -- put_cpu_var(lru_add_pvec); -+ put_locked_var(swapvec_lock, lru_add_pvec); - } - - /** -@@ -593,9 +597,15 @@ - unsigned long flags; - - /* No harm done if a racing interrupt already did this */ -- local_irq_save(flags); -+#ifdef CONFIG_PREEMPT_RT_BASE -+ local_lock_irqsave_on(rotate_lock, flags, cpu); - pagevec_move_tail(pvec); -- local_irq_restore(flags); -+ local_unlock_irqrestore_on(rotate_lock, flags, cpu); -+#else -+ local_lock_irqsave(rotate_lock, flags); -+ pagevec_move_tail(pvec); -+ local_unlock_irqrestore(rotate_lock, flags); -+#endif - } - - pvec = &per_cpu(lru_deactivate_file_pvecs, cpu); -@@ -627,11 +637,12 @@ - return; - - if (likely(get_page_unless_zero(page))) { -- struct pagevec *pvec = &get_cpu_var(lru_deactivate_file_pvecs); -+ struct pagevec *pvec = &get_locked_var(swapvec_lock, -+ lru_deactivate_file_pvecs); - - if (!pagevec_add(pvec, page) || PageCompound(page)) - pagevec_lru_move_fn(pvec, lru_deactivate_file_fn, NULL); -- put_cpu_var(lru_deactivate_file_pvecs); -+ put_locked_var(swapvec_lock, lru_deactivate_file_pvecs); - } - } - -@@ -646,27 +657,31 @@ - void deactivate_page(struct page *page) - { - if (PageLRU(page) && PageActive(page) && !PageUnevictable(page)) { -- struct pagevec *pvec = &get_cpu_var(lru_deactivate_pvecs); -+ struct pagevec *pvec = &get_locked_var(swapvec_lock, -+ lru_deactivate_pvecs); - - get_page(page); - if (!pagevec_add(pvec, page) || PageCompound(page)) - pagevec_lru_move_fn(pvec, lru_deactivate_fn, NULL); -- put_cpu_var(lru_deactivate_pvecs); -+ put_locked_var(swapvec_lock, lru_deactivate_pvecs); - } - } - - void lru_add_drain(void) - { -- lru_add_drain_cpu(get_cpu()); -- put_cpu(); -+ lru_add_drain_cpu(local_lock_cpu(swapvec_lock)); -+ local_unlock_cpu(swapvec_lock); - } - --static void lru_add_drain_per_cpu(struct work_struct *dummy) -+#ifdef CONFIG_PREEMPT_RT_BASE -+static inline void remote_lru_add_drain(int cpu, struct cpumask *has_work) - { -- lru_add_drain(); -+ local_lock_on(swapvec_lock, cpu); -+ lru_add_drain_cpu(cpu); -+ local_unlock_on(swapvec_lock, cpu); - } - --static DEFINE_PER_CPU(struct work_struct, lru_add_drain_work); -+#else - - /* - * lru_add_drain_wq is used to do lru_add_drain_all() from a WQ_MEM_RECLAIM -@@ -686,6 +701,22 @@ - } - early_initcall(lru_init); - -+static void lru_add_drain_per_cpu(struct work_struct *dummy) -+{ -+ lru_add_drain(); -+} -+ -+static DEFINE_PER_CPU(struct work_struct, lru_add_drain_work); -+static inline void remote_lru_add_drain(int cpu, struct cpumask *has_work) -+{ -+ struct work_struct *work = &per_cpu(lru_add_drain_work, cpu); -+ -+ INIT_WORK(work, lru_add_drain_per_cpu); -+ queue_work_on(cpu, lru_add_drain_wq, work); -+ cpumask_set_cpu(cpu, has_work); -+} -+#endif -+ - void lru_add_drain_all(void) - { - static DEFINE_MUTEX(lock); -@@ -697,21 +728,18 @@ - cpumask_clear(&has_work); - - for_each_online_cpu(cpu) { -- struct work_struct *work = &per_cpu(lru_add_drain_work, cpu); -- - if (pagevec_count(&per_cpu(lru_add_pvec, cpu)) || - pagevec_count(&per_cpu(lru_rotate_pvecs, cpu)) || - pagevec_count(&per_cpu(lru_deactivate_file_pvecs, cpu)) || - pagevec_count(&per_cpu(lru_deactivate_pvecs, cpu)) || -- need_activate_page_drain(cpu)) { -- INIT_WORK(work, lru_add_drain_per_cpu); -- queue_work_on(cpu, lru_add_drain_wq, work); -- cpumask_set_cpu(cpu, &has_work); -- } -+ need_activate_page_drain(cpu)) -+ remote_lru_add_drain(cpu, &has_work); - } - -+#ifndef CONFIG_PREEMPT_RT_BASE - for_each_cpu(cpu, &has_work) - flush_work(&per_cpu(lru_add_drain_work, cpu)); -+#endif - - put_online_cpus(); - mutex_unlock(&lock); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/truncate.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/truncate.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/truncate.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/truncate.c 2017-04-18 17:54:26.000000000 +0200 -@@ -62,9 +62,12 @@ - * protected by mapping->tree_lock. - */ - if (!workingset_node_shadows(node) && -- !list_empty(&node->private_list)) -- list_lru_del(&workingset_shadow_nodes, -+ !list_empty(&node->private_list)) { -+ local_lock(workingset_shadow_lock); -+ list_lru_del(&__workingset_shadow_nodes, - &node->private_list); -+ local_unlock(workingset_shadow_lock); -+ } - __radix_tree_delete_node(&mapping->page_tree, node); - unlock: - spin_unlock_irq(&mapping->tree_lock); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/vmalloc.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/vmalloc.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/vmalloc.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/vmalloc.c 2017-04-18 17:54:26.000000000 +0200 -@@ -845,7 +845,7 @@ - struct vmap_block *vb; - struct vmap_area *va; - unsigned long vb_idx; -- int node, err; -+ int node, err, cpu; - void *vaddr; - - node = numa_node_id(); -@@ -888,11 +888,12 @@ - BUG_ON(err); - radix_tree_preload_end(); - -- vbq = &get_cpu_var(vmap_block_queue); -+ cpu = get_cpu_light(); -+ vbq = this_cpu_ptr(&vmap_block_queue); - spin_lock(&vbq->lock); - list_add_tail_rcu(&vb->free_list, &vbq->free); - spin_unlock(&vbq->lock); -- put_cpu_var(vmap_block_queue); -+ put_cpu_light(); - - return vaddr; - } -@@ -961,6 +962,7 @@ - struct vmap_block *vb; - void *vaddr = NULL; - unsigned int order; -+ int cpu; - - BUG_ON(offset_in_page(size)); - BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC); -@@ -975,7 +977,8 @@ - order = get_order(size); - - rcu_read_lock(); -- vbq = &get_cpu_var(vmap_block_queue); -+ cpu = get_cpu_light(); -+ vbq = this_cpu_ptr(&vmap_block_queue); - list_for_each_entry_rcu(vb, &vbq->free, free_list) { - unsigned long pages_off; - -@@ -998,7 +1001,7 @@ - break; - } - -- put_cpu_var(vmap_block_queue); -+ put_cpu_light(); - rcu_read_unlock(); - - /* Allocate new block if nothing was found */ -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/vmstat.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/vmstat.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/vmstat.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/vmstat.c 2017-04-18 17:54:26.000000000 +0200 -@@ -245,6 +245,7 @@ - long x; - long t; - -+ preempt_disable_rt(); - x = delta + __this_cpu_read(*p); - - t = __this_cpu_read(pcp->stat_threshold); -@@ -254,6 +255,7 @@ - x = 0; - } - __this_cpu_write(*p, x); -+ preempt_enable_rt(); - } - EXPORT_SYMBOL(__mod_zone_page_state); - -@@ -265,6 +267,7 @@ - long x; - long t; - -+ preempt_disable_rt(); - x = delta + __this_cpu_read(*p); - - t = __this_cpu_read(pcp->stat_threshold); -@@ -274,6 +277,7 @@ - x = 0; - } - __this_cpu_write(*p, x); -+ preempt_enable_rt(); - } - EXPORT_SYMBOL(__mod_node_page_state); - -@@ -306,6 +310,7 @@ - s8 __percpu *p = pcp->vm_stat_diff + item; - s8 v, t; - -+ preempt_disable_rt(); - v = __this_cpu_inc_return(*p); - t = __this_cpu_read(pcp->stat_threshold); - if (unlikely(v > t)) { -@@ -314,6 +319,7 @@ - zone_page_state_add(v + overstep, zone, item); - __this_cpu_write(*p, -overstep); - } -+ preempt_enable_rt(); - } - - void __inc_node_state(struct pglist_data *pgdat, enum node_stat_item item) -@@ -322,6 +328,7 @@ - s8 __percpu *p = pcp->vm_node_stat_diff + item; - s8 v, t; - -+ preempt_disable_rt(); - v = __this_cpu_inc_return(*p); - t = __this_cpu_read(pcp->stat_threshold); - if (unlikely(v > t)) { -@@ -330,6 +337,7 @@ - node_page_state_add(v + overstep, pgdat, item); - __this_cpu_write(*p, -overstep); - } -+ preempt_enable_rt(); - } - - void __inc_zone_page_state(struct page *page, enum zone_stat_item item) -@@ -350,6 +358,7 @@ - s8 __percpu *p = pcp->vm_stat_diff + item; - s8 v, t; - -+ preempt_disable_rt(); - v = __this_cpu_dec_return(*p); - t = __this_cpu_read(pcp->stat_threshold); - if (unlikely(v < - t)) { -@@ -358,6 +367,7 @@ - zone_page_state_add(v - overstep, zone, item); - __this_cpu_write(*p, overstep); - } -+ preempt_enable_rt(); - } - - void __dec_node_state(struct pglist_data *pgdat, enum node_stat_item item) -@@ -366,6 +376,7 @@ - s8 __percpu *p = pcp->vm_node_stat_diff + item; - s8 v, t; - -+ preempt_disable_rt(); - v = __this_cpu_dec_return(*p); - t = __this_cpu_read(pcp->stat_threshold); - if (unlikely(v < - t)) { -@@ -374,6 +385,7 @@ - node_page_state_add(v - overstep, pgdat, item); - __this_cpu_write(*p, overstep); - } -+ preempt_enable_rt(); - } - - void __dec_zone_page_state(struct page *page, enum zone_stat_item item) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/workingset.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/workingset.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/workingset.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/workingset.c 2017-04-18 17:54:26.000000000 +0200 -@@ -334,7 +334,8 @@ - * point where they would still be useful. - */ - --struct list_lru workingset_shadow_nodes; -+struct list_lru __workingset_shadow_nodes; -+DEFINE_LOCAL_IRQ_LOCK(workingset_shadow_lock); - - static unsigned long count_shadow_nodes(struct shrinker *shrinker, - struct shrink_control *sc) -@@ -344,9 +345,9 @@ - unsigned long pages; - - /* list_lru lock nests inside IRQ-safe mapping->tree_lock */ -- local_irq_disable(); -- shadow_nodes = list_lru_shrink_count(&workingset_shadow_nodes, sc); -- local_irq_enable(); -+ local_lock_irq(workingset_shadow_lock); -+ shadow_nodes = list_lru_shrink_count(&__workingset_shadow_nodes, sc); -+ local_unlock_irq(workingset_shadow_lock); - - if (sc->memcg) { - pages = mem_cgroup_node_nr_lru_pages(sc->memcg, sc->nid, -@@ -438,9 +439,9 @@ - spin_unlock(&mapping->tree_lock); - ret = LRU_REMOVED_RETRY; - out: -- local_irq_enable(); -+ local_unlock_irq(workingset_shadow_lock); - cond_resched(); -- local_irq_disable(); -+ local_lock_irq(workingset_shadow_lock); - spin_lock(lru_lock); - return ret; - } -@@ -451,10 +452,10 @@ - unsigned long ret; - - /* list_lru lock nests inside IRQ-safe mapping->tree_lock */ -- local_irq_disable(); -- ret = list_lru_shrink_walk(&workingset_shadow_nodes, sc, -+ local_lock_irq(workingset_shadow_lock); -+ ret = list_lru_shrink_walk(&__workingset_shadow_nodes, sc, - shadow_lru_isolate, NULL); -- local_irq_enable(); -+ local_unlock_irq(workingset_shadow_lock); - return ret; - } - -@@ -492,7 +493,7 @@ - pr_info("workingset: timestamp_bits=%d max_order=%d bucket_order=%u\n", - timestamp_bits, max_order, bucket_order); - -- ret = __list_lru_init(&workingset_shadow_nodes, true, &shadow_nodes_key); -+ ret = __list_lru_init(&__workingset_shadow_nodes, true, &shadow_nodes_key); - if (ret) - goto err; - ret = register_shrinker(&workingset_shadow_shrinker); -@@ -500,7 +501,7 @@ - goto err_list_lru; - return 0; - err_list_lru: -- list_lru_destroy(&workingset_shadow_nodes); -+ list_lru_destroy(&__workingset_shadow_nodes); - err: - return ret; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/zsmalloc.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/zsmalloc.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/mm/zsmalloc.c 2017-04-16 10:38:30.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/mm/zsmalloc.c 2017-04-18 17:54:26.000000000 +0200 -@@ -53,6 +53,7 @@ - #include - #include - #include -+#include - - #define ZSPAGE_MAGIC 0x58 - -@@ -70,9 +71,22 @@ - */ - #define ZS_MAX_ZSPAGE_ORDER 2 - #define ZS_MAX_PAGES_PER_ZSPAGE (_AC(1, UL) << ZS_MAX_ZSPAGE_ORDER) -- - #define ZS_HANDLE_SIZE (sizeof(unsigned long)) - -+#ifdef CONFIG_PREEMPT_RT_FULL -+ -+struct zsmalloc_handle { -+ unsigned long addr; -+ struct mutex lock; -+}; -+ -+#define ZS_HANDLE_ALLOC_SIZE (sizeof(struct zsmalloc_handle)) -+ -+#else -+ -+#define ZS_HANDLE_ALLOC_SIZE (sizeof(unsigned long)) -+#endif -+ - /* - * Object location (, ) is encoded as - * as single (unsigned long) handle value. -@@ -327,7 +341,7 @@ - - static int create_cache(struct zs_pool *pool) - { -- pool->handle_cachep = kmem_cache_create("zs_handle", ZS_HANDLE_SIZE, -+ pool->handle_cachep = kmem_cache_create("zs_handle", ZS_HANDLE_ALLOC_SIZE, - 0, 0, NULL); - if (!pool->handle_cachep) - return 1; -@@ -351,10 +365,27 @@ - - static unsigned long cache_alloc_handle(struct zs_pool *pool, gfp_t gfp) - { -- return (unsigned long)kmem_cache_alloc(pool->handle_cachep, -- gfp & ~(__GFP_HIGHMEM|__GFP_MOVABLE)); -+ void *p; -+ -+ p = kmem_cache_alloc(pool->handle_cachep, -+ gfp & ~(__GFP_HIGHMEM|__GFP_MOVABLE)); -+#ifdef CONFIG_PREEMPT_RT_FULL -+ if (p) { -+ struct zsmalloc_handle *zh = p; -+ -+ mutex_init(&zh->lock); -+ } -+#endif -+ return (unsigned long)p; - } - -+#ifdef CONFIG_PREEMPT_RT_FULL -+static struct zsmalloc_handle *zs_get_pure_handle(unsigned long handle) -+{ -+ return (void *)(handle &~((1 << OBJ_TAG_BITS) - 1)); -+} -+#endif -+ - static void cache_free_handle(struct zs_pool *pool, unsigned long handle) - { - kmem_cache_free(pool->handle_cachep, (void *)handle); -@@ -373,12 +404,18 @@ - - static void record_obj(unsigned long handle, unsigned long obj) - { -+#ifdef CONFIG_PREEMPT_RT_FULL -+ struct zsmalloc_handle *zh = zs_get_pure_handle(handle); -+ -+ WRITE_ONCE(zh->addr, obj); -+#else - /* - * lsb of @obj represents handle lock while other bits - * represent object value the handle is pointing so - * updating shouldn't do store tearing. - */ - WRITE_ONCE(*(unsigned long *)handle, obj); -+#endif - } - - /* zpool driver */ -@@ -467,6 +504,7 @@ - - /* per-cpu VM mapping areas for zspage accesses that cross page boundaries */ - static DEFINE_PER_CPU(struct mapping_area, zs_map_area); -+static DEFINE_LOCAL_IRQ_LOCK(zs_map_area_lock); - - static bool is_zspage_isolated(struct zspage *zspage) - { -@@ -902,7 +940,13 @@ - - static unsigned long handle_to_obj(unsigned long handle) - { -+#ifdef CONFIG_PREEMPT_RT_FULL -+ struct zsmalloc_handle *zh = zs_get_pure_handle(handle); -+ -+ return zh->addr; -+#else - return *(unsigned long *)handle; -+#endif - } - - static unsigned long obj_to_head(struct page *page, void *obj) -@@ -916,22 +960,46 @@ - - static inline int testpin_tag(unsigned long handle) - { -+#ifdef CONFIG_PREEMPT_RT_FULL -+ struct zsmalloc_handle *zh = zs_get_pure_handle(handle); -+ -+ return mutex_is_locked(&zh->lock); -+#else - return bit_spin_is_locked(HANDLE_PIN_BIT, (unsigned long *)handle); -+#endif - } - - static inline int trypin_tag(unsigned long handle) - { -+#ifdef CONFIG_PREEMPT_RT_FULL -+ struct zsmalloc_handle *zh = zs_get_pure_handle(handle); -+ -+ return mutex_trylock(&zh->lock); -+#else - return bit_spin_trylock(HANDLE_PIN_BIT, (unsigned long *)handle); -+#endif - } - - static void pin_tag(unsigned long handle) - { -+#ifdef CONFIG_PREEMPT_RT_FULL -+ struct zsmalloc_handle *zh = zs_get_pure_handle(handle); -+ -+ return mutex_lock(&zh->lock); -+#else - bit_spin_lock(HANDLE_PIN_BIT, (unsigned long *)handle); -+#endif - } - - static void unpin_tag(unsigned long handle) - { -+#ifdef CONFIG_PREEMPT_RT_FULL -+ struct zsmalloc_handle *zh = zs_get_pure_handle(handle); -+ -+ return mutex_unlock(&zh->lock); -+#else - bit_spin_unlock(HANDLE_PIN_BIT, (unsigned long *)handle); -+#endif - } - - static void reset_page(struct page *page) -@@ -1423,7 +1491,7 @@ - class = pool->size_class[class_idx]; - off = (class->size * obj_idx) & ~PAGE_MASK; - -- area = &get_cpu_var(zs_map_area); -+ area = &get_locked_var(zs_map_area_lock, zs_map_area); - area->vm_mm = mm; - if (off + class->size <= PAGE_SIZE) { - /* this object is contained entirely within a page */ -@@ -1477,7 +1545,7 @@ - - __zs_unmap_object(area, pages, off, class->size); - } -- put_cpu_var(zs_map_area); -+ put_locked_var(zs_map_area_lock, zs_map_area); - - migrate_read_unlock(zspage); - unpin_tag(handle); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/core/dev.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/core/dev.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/core/dev.c 2017-04-16 10:38:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/core/dev.c 2017-04-18 17:54:26.000000000 +0200 -@@ -190,6 +190,7 @@ - static DEFINE_READ_MOSTLY_HASHTABLE(napi_hash, 8); - - static seqcount_t devnet_rename_seq; -+static DEFINE_MUTEX(devnet_rename_mutex); - - static inline void dev_base_seq_inc(struct net *net) - { -@@ -211,14 +212,14 @@ - static inline void rps_lock(struct softnet_data *sd) - { - #ifdef CONFIG_RPS -- spin_lock(&sd->input_pkt_queue.lock); -+ raw_spin_lock(&sd->input_pkt_queue.raw_lock); - #endif - } - - static inline void rps_unlock(struct softnet_data *sd) - { - #ifdef CONFIG_RPS -- spin_unlock(&sd->input_pkt_queue.lock); -+ raw_spin_unlock(&sd->input_pkt_queue.raw_lock); - #endif - } - -@@ -888,7 +889,8 @@ - strcpy(name, dev->name); - rcu_read_unlock(); - if (read_seqcount_retry(&devnet_rename_seq, seq)) { -- cond_resched(); -+ mutex_lock(&devnet_rename_mutex); -+ mutex_unlock(&devnet_rename_mutex); - goto retry; - } - -@@ -1157,20 +1159,17 @@ - if (dev->flags & IFF_UP) - return -EBUSY; - -- write_seqcount_begin(&devnet_rename_seq); -+ mutex_lock(&devnet_rename_mutex); -+ __raw_write_seqcount_begin(&devnet_rename_seq); - -- if (strncmp(newname, dev->name, IFNAMSIZ) == 0) { -- write_seqcount_end(&devnet_rename_seq); -- return 0; -- } -+ if (strncmp(newname, dev->name, IFNAMSIZ) == 0) -+ goto outunlock; - - memcpy(oldname, dev->name, IFNAMSIZ); - - err = dev_get_valid_name(net, dev, newname); -- if (err < 0) { -- write_seqcount_end(&devnet_rename_seq); -- return err; -- } -+ if (err < 0) -+ goto outunlock; - - if (oldname[0] && !strchr(oldname, '%')) - netdev_info(dev, "renamed from %s\n", oldname); -@@ -1183,11 +1182,12 @@ - if (ret) { - memcpy(dev->name, oldname, IFNAMSIZ); - dev->name_assign_type = old_assign_type; -- write_seqcount_end(&devnet_rename_seq); -- return ret; -+ err = ret; -+ goto outunlock; - } - -- write_seqcount_end(&devnet_rename_seq); -+ __raw_write_seqcount_end(&devnet_rename_seq); -+ mutex_unlock(&devnet_rename_mutex); - - netdev_adjacent_rename_links(dev, oldname); - -@@ -1208,7 +1208,8 @@ - /* err >= 0 after dev_alloc_name() or stores the first errno */ - if (err >= 0) { - err = ret; -- write_seqcount_begin(&devnet_rename_seq); -+ mutex_lock(&devnet_rename_mutex); -+ __raw_write_seqcount_begin(&devnet_rename_seq); - memcpy(dev->name, oldname, IFNAMSIZ); - memcpy(oldname, newname, IFNAMSIZ); - dev->name_assign_type = old_assign_type; -@@ -1221,6 +1222,11 @@ - } - - return err; -+ -+outunlock: -+ __raw_write_seqcount_end(&devnet_rename_seq); -+ mutex_unlock(&devnet_rename_mutex); -+ return err; - } - - /** -@@ -2285,6 +2291,7 @@ - sd->output_queue_tailp = &q->next_sched; - raise_softirq_irqoff(NET_TX_SOFTIRQ); - local_irq_restore(flags); -+ preempt_check_resched_rt(); - } - - void __netif_schedule(struct Qdisc *q) -@@ -2366,6 +2373,7 @@ - __this_cpu_write(softnet_data.completion_queue, skb); - raise_softirq_irqoff(NET_TX_SOFTIRQ); - local_irq_restore(flags); -+ preempt_check_resched_rt(); - } - EXPORT_SYMBOL(__dev_kfree_skb_irq); - -@@ -3100,7 +3108,11 @@ - * This permits qdisc->running owner to get the lock more - * often and dequeue packets faster. - */ -+#ifdef CONFIG_PREEMPT_RT_FULL -+ contended = true; -+#else - contended = qdisc_is_running(q); -+#endif - if (unlikely(contended)) - spin_lock(&q->busylock); - -@@ -3163,8 +3175,10 @@ - #define skb_update_prio(skb) - #endif - -+#ifndef CONFIG_PREEMPT_RT_FULL - DEFINE_PER_CPU(int, xmit_recursion); - EXPORT_SYMBOL(xmit_recursion); -+#endif - - /** - * dev_loopback_xmit - loop back @skb -@@ -3398,8 +3412,7 @@ - int cpu = smp_processor_id(); /* ok because BHs are off */ - - if (txq->xmit_lock_owner != cpu) { -- if (unlikely(__this_cpu_read(xmit_recursion) > -- XMIT_RECURSION_LIMIT)) -+ if (unlikely(xmit_rec_read() > XMIT_RECURSION_LIMIT)) - goto recursion_alert; - - skb = validate_xmit_skb(skb, dev); -@@ -3409,9 +3422,9 @@ - HARD_TX_LOCK(dev, txq, cpu); - - if (!netif_xmit_stopped(txq)) { -- __this_cpu_inc(xmit_recursion); -+ xmit_rec_inc(); - skb = dev_hard_start_xmit(skb, dev, txq, &rc); -- __this_cpu_dec(xmit_recursion); -+ xmit_rec_dec(); - if (dev_xmit_complete(rc)) { - HARD_TX_UNLOCK(dev, txq); - goto out; -@@ -3785,6 +3798,7 @@ - rps_unlock(sd); - - local_irq_restore(flags); -+ preempt_check_resched_rt(); - - atomic_long_inc(&skb->dev->rx_dropped); - kfree_skb(skb); -@@ -3803,7 +3817,7 @@ - struct rps_dev_flow voidflow, *rflow = &voidflow; - int cpu; - -- preempt_disable(); -+ migrate_disable(); - rcu_read_lock(); - - cpu = get_rps_cpu(skb->dev, skb, &rflow); -@@ -3813,13 +3827,13 @@ - ret = enqueue_to_backlog(skb, cpu, &rflow->last_qtail); - - rcu_read_unlock(); -- preempt_enable(); -+ migrate_enable(); - } else - #endif - { - unsigned int qtail; -- ret = enqueue_to_backlog(skb, get_cpu(), &qtail); -- put_cpu(); -+ ret = enqueue_to_backlog(skb, get_cpu_light(), &qtail); -+ put_cpu_light(); - } - return ret; - } -@@ -3853,11 +3867,9 @@ - - trace_netif_rx_ni_entry(skb); - -- preempt_disable(); -+ local_bh_disable(); - err = netif_rx_internal(skb); -- if (local_softirq_pending()) -- do_softirq(); -- preempt_enable(); -+ local_bh_enable(); - - return err; - } -@@ -4336,7 +4348,7 @@ - skb_queue_walk_safe(&sd->input_pkt_queue, skb, tmp) { - if (skb->dev->reg_state == NETREG_UNREGISTERING) { - __skb_unlink(skb, &sd->input_pkt_queue); -- kfree_skb(skb); -+ __skb_queue_tail(&sd->tofree_queue, skb); - input_queue_head_incr(sd); - } - } -@@ -4346,11 +4358,14 @@ - skb_queue_walk_safe(&sd->process_queue, skb, tmp) { - if (skb->dev->reg_state == NETREG_UNREGISTERING) { - __skb_unlink(skb, &sd->process_queue); -- kfree_skb(skb); -+ __skb_queue_tail(&sd->tofree_queue, skb); - input_queue_head_incr(sd); - } - } -+ if (!skb_queue_empty(&sd->tofree_queue)) -+ raise_softirq_irqoff(NET_RX_SOFTIRQ); - local_bh_enable(); -+ - } - - static void flush_all_backlogs(void) -@@ -4831,6 +4846,7 @@ - sd->rps_ipi_list = NULL; - - local_irq_enable(); -+ preempt_check_resched_rt(); - - /* Send pending IPI's to kick RPS processing on remote cpus. */ - while (remsd) { -@@ -4844,6 +4860,7 @@ - } else - #endif - local_irq_enable(); -+ preempt_check_resched_rt(); - } - - static bool sd_has_rps_ipi_waiting(struct softnet_data *sd) -@@ -4873,7 +4890,9 @@ - while (again) { - struct sk_buff *skb; - -+ local_irq_disable(); - while ((skb = __skb_dequeue(&sd->process_queue))) { -+ local_irq_enable(); - rcu_read_lock(); - __netif_receive_skb(skb); - rcu_read_unlock(); -@@ -4881,9 +4900,9 @@ - if (++work >= quota) - return work; - -+ local_irq_disable(); - } - -- local_irq_disable(); - rps_lock(sd); - if (skb_queue_empty(&sd->input_pkt_queue)) { - /* -@@ -4921,9 +4940,11 @@ - local_irq_save(flags); - ____napi_schedule(this_cpu_ptr(&softnet_data), n); - local_irq_restore(flags); -+ preempt_check_resched_rt(); - } - EXPORT_SYMBOL(__napi_schedule); - -+#ifndef CONFIG_PREEMPT_RT_FULL - /** - * __napi_schedule_irqoff - schedule for receive - * @n: entry to schedule -@@ -4935,6 +4956,7 @@ - ____napi_schedule(this_cpu_ptr(&softnet_data), n); - } - EXPORT_SYMBOL(__napi_schedule_irqoff); -+#endif - - void __napi_complete(struct napi_struct *n) - { -@@ -5224,13 +5246,21 @@ - struct softnet_data *sd = this_cpu_ptr(&softnet_data); - unsigned long time_limit = jiffies + 2; - int budget = netdev_budget; -+ struct sk_buff_head tofree_q; -+ struct sk_buff *skb; - LIST_HEAD(list); - LIST_HEAD(repoll); - -+ __skb_queue_head_init(&tofree_q); -+ - local_irq_disable(); -+ skb_queue_splice_init(&sd->tofree_queue, &tofree_q); - list_splice_init(&sd->poll_list, &list); - local_irq_enable(); - -+ while ((skb = __skb_dequeue(&tofree_q))) -+ kfree_skb(skb); -+ - for (;;) { - struct napi_struct *n; - -@@ -5261,7 +5291,7 @@ - list_splice_tail(&repoll, &list); - list_splice(&list, &sd->poll_list); - if (!list_empty(&sd->poll_list)) -- __raise_softirq_irqoff(NET_RX_SOFTIRQ); -+ __raise_softirq_irqoff_ksoft(NET_RX_SOFTIRQ); - - net_rps_action_and_irq_enable(sd); - } -@@ -8022,16 +8052,20 @@ - - raise_softirq_irqoff(NET_TX_SOFTIRQ); - local_irq_enable(); -+ preempt_check_resched_rt(); - - /* Process offline CPU's input_pkt_queue */ - while ((skb = __skb_dequeue(&oldsd->process_queue))) { - netif_rx_ni(skb); - input_queue_head_incr(oldsd); - } -- while ((skb = skb_dequeue(&oldsd->input_pkt_queue))) { -+ while ((skb = __skb_dequeue(&oldsd->input_pkt_queue))) { - netif_rx_ni(skb); - input_queue_head_incr(oldsd); - } -+ while ((skb = __skb_dequeue(&oldsd->tofree_queue))) { -+ kfree_skb(skb); -+ } - - return NOTIFY_OK; - } -@@ -8336,8 +8370,9 @@ - - INIT_WORK(flush, flush_backlog); - -- skb_queue_head_init(&sd->input_pkt_queue); -- skb_queue_head_init(&sd->process_queue); -+ skb_queue_head_init_raw(&sd->input_pkt_queue); -+ skb_queue_head_init_raw(&sd->process_queue); -+ skb_queue_head_init_raw(&sd->tofree_queue); - INIT_LIST_HEAD(&sd->poll_list); - sd->output_queue_tailp = &sd->output_queue; - #ifdef CONFIG_RPS -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/core/filter.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/core/filter.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/core/filter.c 2017-04-16 10:38:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/core/filter.c 2017-04-18 17:54:26.000000000 +0200 -@@ -1645,7 +1645,7 @@ - { - int ret; - -- if (unlikely(__this_cpu_read(xmit_recursion) > XMIT_RECURSION_LIMIT)) { -+ if (unlikely(xmit_rec_read() > XMIT_RECURSION_LIMIT)) { - net_crit_ratelimited("bpf: recursion limit reached on datapath, buggy bpf program?\n"); - kfree_skb(skb); - return -ENETDOWN; -@@ -1653,9 +1653,9 @@ - - skb->dev = dev; - -- __this_cpu_inc(xmit_recursion); -+ xmit_rec_inc(); - ret = dev_queue_xmit(skb); -- __this_cpu_dec(xmit_recursion); -+ xmit_rec_dec(); - - return ret; - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/core/gen_estimator.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/core/gen_estimator.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/core/gen_estimator.c 2017-04-16 10:38:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/core/gen_estimator.c 2017-04-18 17:54:26.000000000 +0200 -@@ -84,7 +84,7 @@ - struct gnet_stats_basic_packed *bstats; - struct gnet_stats_rate_est64 *rate_est; - spinlock_t *stats_lock; -- seqcount_t *running; -+ net_seqlock_t *running; - int ewma_log; - u32 last_packets; - unsigned long avpps; -@@ -213,7 +213,7 @@ - struct gnet_stats_basic_cpu __percpu *cpu_bstats, - struct gnet_stats_rate_est64 *rate_est, - spinlock_t *stats_lock, -- seqcount_t *running, -+ net_seqlock_t *running, - struct nlattr *opt) - { - struct gen_estimator *est; -@@ -309,7 +309,7 @@ - struct gnet_stats_basic_cpu __percpu *cpu_bstats, - struct gnet_stats_rate_est64 *rate_est, - spinlock_t *stats_lock, -- seqcount_t *running, struct nlattr *opt) -+ net_seqlock_t *running, struct nlattr *opt) - { - gen_kill_estimator(bstats, rate_est); - return gen_new_estimator(bstats, cpu_bstats, rate_est, stats_lock, running, opt); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/core/gen_stats.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/core/gen_stats.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/core/gen_stats.c 2017-04-16 10:38:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/core/gen_stats.c 2017-04-18 17:54:26.000000000 +0200 -@@ -130,7 +130,7 @@ - } - - void --__gnet_stats_copy_basic(const seqcount_t *running, -+__gnet_stats_copy_basic(net_seqlock_t *running, - struct gnet_stats_basic_packed *bstats, - struct gnet_stats_basic_cpu __percpu *cpu, - struct gnet_stats_basic_packed *b) -@@ -143,10 +143,10 @@ - } - do { - if (running) -- seq = read_seqcount_begin(running); -+ seq = net_seq_begin(running); - bstats->bytes = b->bytes; - bstats->packets = b->packets; -- } while (running && read_seqcount_retry(running, seq)); -+ } while (running && net_seq_retry(running, seq)); - } - EXPORT_SYMBOL(__gnet_stats_copy_basic); - -@@ -164,7 +164,7 @@ - * if the room in the socket buffer was not sufficient. - */ - int --gnet_stats_copy_basic(const seqcount_t *running, -+gnet_stats_copy_basic(net_seqlock_t *running, - struct gnet_dump *d, - struct gnet_stats_basic_cpu __percpu *cpu, - struct gnet_stats_basic_packed *b) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/core/skbuff.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/core/skbuff.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/core/skbuff.c 2017-04-16 10:38:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/core/skbuff.c 2017-04-18 17:54:26.000000000 +0200 -@@ -64,6 +64,7 @@ - #include - #include - #include -+#include - - #include - #include -@@ -360,6 +361,8 @@ - - static DEFINE_PER_CPU(struct page_frag_cache, netdev_alloc_cache); - static DEFINE_PER_CPU(struct napi_alloc_cache, napi_alloc_cache); -+static DEFINE_LOCAL_IRQ_LOCK(netdev_alloc_lock); -+static DEFINE_LOCAL_IRQ_LOCK(napi_alloc_cache_lock); - - static void *__netdev_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) - { -@@ -367,10 +370,10 @@ - unsigned long flags; - void *data; - -- local_irq_save(flags); -+ local_lock_irqsave(netdev_alloc_lock, flags); - nc = this_cpu_ptr(&netdev_alloc_cache); - data = __alloc_page_frag(nc, fragsz, gfp_mask); -- local_irq_restore(flags); -+ local_unlock_irqrestore(netdev_alloc_lock, flags); - return data; - } - -@@ -389,9 +392,13 @@ - - static void *__napi_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) - { -- struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); -+ struct napi_alloc_cache *nc; -+ void *data; - -- return __alloc_page_frag(&nc->page, fragsz, gfp_mask); -+ nc = &get_locked_var(napi_alloc_cache_lock, napi_alloc_cache); -+ data = __alloc_page_frag(&nc->page, fragsz, gfp_mask); -+ put_locked_var(napi_alloc_cache_lock, napi_alloc_cache); -+ return data; - } - - void *napi_alloc_frag(unsigned int fragsz) -@@ -438,13 +445,13 @@ - if (sk_memalloc_socks()) - gfp_mask |= __GFP_MEMALLOC; - -- local_irq_save(flags); -+ local_lock_irqsave(netdev_alloc_lock, flags); - - nc = this_cpu_ptr(&netdev_alloc_cache); - data = __alloc_page_frag(nc, len, gfp_mask); - pfmemalloc = nc->pfmemalloc; - -- local_irq_restore(flags); -+ local_unlock_irqrestore(netdev_alloc_lock, flags); - - if (unlikely(!data)) - return NULL; -@@ -485,9 +492,10 @@ - struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len, - gfp_t gfp_mask) - { -- struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); -+ struct napi_alloc_cache *nc; - struct sk_buff *skb; - void *data; -+ bool pfmemalloc; - - len += NET_SKB_PAD + NET_IP_ALIGN; - -@@ -505,7 +513,10 @@ - if (sk_memalloc_socks()) - gfp_mask |= __GFP_MEMALLOC; - -+ nc = &get_locked_var(napi_alloc_cache_lock, napi_alloc_cache); - data = __alloc_page_frag(&nc->page, len, gfp_mask); -+ pfmemalloc = nc->page.pfmemalloc; -+ put_locked_var(napi_alloc_cache_lock, napi_alloc_cache); - if (unlikely(!data)) - return NULL; - -@@ -516,7 +527,7 @@ - } - - /* use OR instead of assignment to avoid clearing of bits in mask */ -- if (nc->page.pfmemalloc) -+ if (pfmemalloc) - skb->pfmemalloc = 1; - skb->head_frag = 1; - -@@ -760,23 +771,26 @@ - - void __kfree_skb_flush(void) - { -- struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); -+ struct napi_alloc_cache *nc; - -+ nc = &get_locked_var(napi_alloc_cache_lock, napi_alloc_cache); - /* flush skb_cache if containing objects */ - if (nc->skb_count) { - kmem_cache_free_bulk(skbuff_head_cache, nc->skb_count, - nc->skb_cache); - nc->skb_count = 0; - } -+ put_locked_var(napi_alloc_cache_lock, napi_alloc_cache); - } - - static inline void _kfree_skb_defer(struct sk_buff *skb) - { -- struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); -+ struct napi_alloc_cache *nc; - - /* drop skb->head and call any destructors for packet */ - skb_release_all(skb); - -+ nc = &get_locked_var(napi_alloc_cache_lock, napi_alloc_cache); - /* record skb to CPU local list */ - nc->skb_cache[nc->skb_count++] = skb; - -@@ -791,6 +805,7 @@ - nc->skb_cache); - nc->skb_count = 0; - } -+ put_locked_var(napi_alloc_cache_lock, napi_alloc_cache); - } - void __kfree_skb_defer(struct sk_buff *skb) - { -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/core/sock.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/core/sock.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/core/sock.c 2017-04-16 10:38:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/core/sock.c 2017-04-18 17:54:26.000000000 +0200 -@@ -2499,12 +2499,11 @@ - if (sk->sk_lock.owned) - __lock_sock(sk); - sk->sk_lock.owned = 1; -- spin_unlock(&sk->sk_lock.slock); -+ spin_unlock_bh(&sk->sk_lock.slock); - /* - * The sk_lock has mutex_lock() semantics here: - */ - mutex_acquire(&sk->sk_lock.dep_map, subclass, 0, _RET_IP_); -- local_bh_enable(); - } - EXPORT_SYMBOL(lock_sock_nested); - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/ipv4/icmp.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/ipv4/icmp.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/ipv4/icmp.c 2017-04-16 10:38:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/ipv4/icmp.c 2017-04-18 17:54:26.000000000 +0200 -@@ -69,6 +69,7 @@ - #include - #include - #include -+#include - #include - #include - #include -@@ -77,6 +78,7 @@ - #include - #include - #include -+#include - #include - #include - #include -@@ -204,6 +206,8 @@ - * - * On SMP we have one ICMP socket per-cpu. - */ -+static DEFINE_LOCAL_IRQ_LOCK(icmp_sk_lock); -+ - static struct sock *icmp_sk(struct net *net) - { - return *this_cpu_ptr(net->ipv4.icmp_sk); -@@ -215,12 +219,14 @@ - - local_bh_disable(); - -+ local_lock(icmp_sk_lock); - sk = icmp_sk(net); - - if (unlikely(!spin_trylock(&sk->sk_lock.slock))) { - /* This can happen if the output path signals a - * dst_link_failure() for an outgoing ICMP packet. - */ -+ local_unlock(icmp_sk_lock); - local_bh_enable(); - return NULL; - } -@@ -230,6 +236,7 @@ - static inline void icmp_xmit_unlock(struct sock *sk) - { - spin_unlock_bh(&sk->sk_lock.slock); -+ local_unlock(icmp_sk_lock); - } - - int sysctl_icmp_msgs_per_sec __read_mostly = 1000; -@@ -358,6 +365,7 @@ - struct sock *sk; - struct sk_buff *skb; - -+ local_lock(icmp_sk_lock); - sk = icmp_sk(dev_net((*rt)->dst.dev)); - if (ip_append_data(sk, fl4, icmp_glue_bits, icmp_param, - icmp_param->data_len+icmp_param->head_len, -@@ -380,6 +388,7 @@ - skb->ip_summed = CHECKSUM_NONE; - ip_push_pending_frames(sk, fl4); - } -+ local_unlock(icmp_sk_lock); - } - - /* -@@ -891,6 +900,30 @@ - } - - /* -+ * 32bit and 64bit have different timestamp length, so we check for -+ * the cookie at offset 20 and verify it is repeated at offset 50 -+ */ -+#define CO_POS0 20 -+#define CO_POS1 50 -+#define CO_SIZE sizeof(int) -+#define ICMP_SYSRQ_SIZE 57 -+ -+/* -+ * We got a ICMP_SYSRQ_SIZE sized ping request. Check for the cookie -+ * pattern and if it matches send the next byte as a trigger to sysrq. -+ */ -+static void icmp_check_sysrq(struct net *net, struct sk_buff *skb) -+{ -+ int cookie = htonl(net->ipv4.sysctl_icmp_echo_sysrq); -+ char *p = skb->data; -+ -+ if (!memcmp(&cookie, p + CO_POS0, CO_SIZE) && -+ !memcmp(&cookie, p + CO_POS1, CO_SIZE) && -+ p[CO_POS0 + CO_SIZE] == p[CO_POS1 + CO_SIZE]) -+ handle_sysrq(p[CO_POS0 + CO_SIZE]); -+} -+ -+/* - * Handle ICMP_ECHO ("ping") requests. - * - * RFC 1122: 3.2.2.6 MUST have an echo server that answers ICMP echo -@@ -917,6 +950,11 @@ - icmp_param.data_len = skb->len; - icmp_param.head_len = sizeof(struct icmphdr); - icmp_reply(&icmp_param, skb); -+ -+ if (skb->len == ICMP_SYSRQ_SIZE && -+ net->ipv4.sysctl_icmp_echo_sysrq) { -+ icmp_check_sysrq(net, skb); -+ } - } - /* should there be an ICMP stat for ignored echos? */ - return true; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/ipv4/sysctl_net_ipv4.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/ipv4/sysctl_net_ipv4.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/ipv4/sysctl_net_ipv4.c 2017-04-16 10:38:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/ipv4/sysctl_net_ipv4.c 2017-04-18 17:54:27.000000000 +0200 -@@ -681,6 +681,13 @@ - .proc_handler = proc_dointvec - }, - { -+ .procname = "icmp_echo_sysrq", -+ .data = &init_net.ipv4.sysctl_icmp_echo_sysrq, -+ .maxlen = sizeof(int), -+ .mode = 0644, -+ .proc_handler = proc_dointvec -+ }, -+ { - .procname = "icmp_ignore_bogus_error_responses", - .data = &init_net.ipv4.sysctl_icmp_ignore_bogus_error_responses, - .maxlen = sizeof(int), -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/ipv4/tcp_ipv4.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/ipv4/tcp_ipv4.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/ipv4/tcp_ipv4.c 2017-04-16 10:38:31.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/ipv4/tcp_ipv4.c 2017-04-18 17:54:27.000000000 +0200 -@@ -62,6 +62,7 @@ - #include - #include - #include -+#include - - #include - #include -@@ -568,6 +569,7 @@ - } - EXPORT_SYMBOL(tcp_v4_send_check); - -+static DEFINE_LOCAL_IRQ_LOCK(tcp_sk_lock); - /* - * This routine will send an RST to the other tcp. - * -@@ -695,6 +697,8 @@ - offsetof(struct inet_timewait_sock, tw_bound_dev_if)); - - arg.tos = ip_hdr(skb)->tos; -+ -+ local_lock(tcp_sk_lock); - local_bh_disable(); - ip_send_unicast_reply(*this_cpu_ptr(net->ipv4.tcp_sk), - skb, &TCP_SKB_CB(skb)->header.h4.opt, -@@ -704,6 +708,7 @@ - __TCP_INC_STATS(net, TCP_MIB_OUTSEGS); - __TCP_INC_STATS(net, TCP_MIB_OUTRSTS); - local_bh_enable(); -+ local_unlock(tcp_sk_lock); - - #ifdef CONFIG_TCP_MD5SIG - out: -@@ -779,6 +784,7 @@ - if (oif) - arg.bound_dev_if = oif; - arg.tos = tos; -+ local_lock(tcp_sk_lock); - local_bh_disable(); - ip_send_unicast_reply(*this_cpu_ptr(net->ipv4.tcp_sk), - skb, &TCP_SKB_CB(skb)->header.h4.opt, -@@ -787,6 +793,7 @@ - - __TCP_INC_STATS(net, TCP_MIB_OUTSEGS); - local_bh_enable(); -+ local_unlock(tcp_sk_lock); - } - - static void tcp_v4_timewait_ack(struct sock *sk, struct sk_buff *skb) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/mac80211/rx.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/mac80211/rx.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/mac80211/rx.c 2017-04-16 10:38:32.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/mac80211/rx.c 2017-04-18 17:54:27.000000000 +0200 -@@ -4180,7 +4180,7 @@ - struct ieee80211_supported_band *sband; - struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(skb); - -- WARN_ON_ONCE(softirq_count() == 0); -+ WARN_ON_ONCE_NONRT(softirq_count() == 0); - - if (WARN_ON(status->band >= NUM_NL80211_BANDS)) - goto drop; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/netfilter/core.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/netfilter/core.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/netfilter/core.c 2017-04-16 10:38:32.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/netfilter/core.c 2017-04-18 17:54:27.000000000 +0200 -@@ -22,12 +22,18 @@ - #include - #include - #include -+#include - #include - #include - #include - - #include "nf_internals.h" - -+#ifdef CONFIG_PREEMPT_RT_BASE -+DEFINE_LOCAL_IRQ_LOCK(xt_write_lock); -+EXPORT_PER_CPU_SYMBOL(xt_write_lock); -+#endif -+ - static DEFINE_MUTEX(afinfo_mutex); - - const struct nf_afinfo __rcu *nf_afinfo[NFPROTO_NUMPROTO] __read_mostly; -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/packet/af_packet.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/packet/af_packet.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/packet/af_packet.c 2017-04-16 10:38:33.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/packet/af_packet.c 2017-04-18 17:54:27.000000000 +0200 -@@ -63,6 +63,7 @@ - #include - #include - #include -+#include - #include - #include - #include -@@ -694,7 +695,7 @@ - if (BLOCK_NUM_PKTS(pbd)) { - while (atomic_read(&pkc->blk_fill_in_prog)) { - /* Waiting for skb_copy_bits to finish... */ -- cpu_relax(); -+ cpu_chill(); - } - } - -@@ -956,7 +957,7 @@ - if (!(status & TP_STATUS_BLK_TMO)) { - while (atomic_read(&pkc->blk_fill_in_prog)) { - /* Waiting for skb_copy_bits to finish... */ -- cpu_relax(); -+ cpu_chill(); - } - } - prb_close_block(pkc, pbd, po, status); -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/rds/ib_rdma.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/rds/ib_rdma.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/rds/ib_rdma.c 2017-04-16 10:38:33.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/rds/ib_rdma.c 2017-04-18 17:54:27.000000000 +0200 -@@ -34,6 +34,7 @@ - #include - #include - #include -+#include - - #include "rds_single_path.h" - #include "ib_mr.h" -@@ -210,7 +211,7 @@ - for_each_online_cpu(cpu) { - flag = &per_cpu(clean_list_grace, cpu); - while (test_bit(CLEAN_LIST_BUSY_BIT, flag)) -- cpu_relax(); -+ cpu_chill(); - } - } - -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/rxrpc/security.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/rxrpc/security.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/rxrpc/security.c 2017-04-16 10:38:33.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/rxrpc/security.c 2017-04-18 17:54:27.000000000 +0200 -@@ -19,9 +19,6 @@ - #include - #include "ar-internal.h" - --static LIST_HEAD(rxrpc_security_methods); --static DECLARE_RWSEM(rxrpc_security_sem); -- - static const struct rxrpc_security *rxrpc_security_types[] = { - [RXRPC_SECURITY_NONE] = &rxrpc_no_security, - #ifdef CONFIG_RXKAD -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/sched/sch_api.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/sched/sch_api.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/sched/sch_api.c 2017-04-16 10:38:33.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/sched/sch_api.c 2017-04-18 17:54:27.000000000 +0200 -@@ -981,7 +981,7 @@ - rcu_assign_pointer(sch->stab, stab); - } - if (tca[TCA_RATE]) { -- seqcount_t *running; -+ net_seqlock_t *running; - - err = -EOPNOTSUPP; - if (sch->flags & TCQ_F_MQROOT) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/sched/sch_generic.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/sched/sch_generic.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/sched/sch_generic.c 2017-04-16 10:38:33.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/sched/sch_generic.c 2017-04-18 17:54:27.000000000 +0200 -@@ -425,7 +425,11 @@ - .ops = &noop_qdisc_ops, - .q.lock = __SPIN_LOCK_UNLOCKED(noop_qdisc.q.lock), - .dev_queue = &noop_netdev_queue, -+#ifdef CONFIG_PREEMPT_RT_BASE -+ .running = __SEQLOCK_UNLOCKED(noop_qdisc.running), -+#else - .running = SEQCNT_ZERO(noop_qdisc.running), -+#endif - .busylock = __SPIN_LOCK_UNLOCKED(noop_qdisc.busylock), - }; - EXPORT_SYMBOL(noop_qdisc); -@@ -624,9 +628,17 @@ - lockdep_set_class(&sch->busylock, - dev->qdisc_tx_busylock ?: &qdisc_tx_busylock); - -+#ifdef CONFIG_PREEMPT_RT_BASE -+ seqlock_init(&sch->running); -+ lockdep_set_class(&sch->running.seqcount, -+ dev->qdisc_running_key ?: &qdisc_running_key); -+ lockdep_set_class(&sch->running.lock, -+ dev->qdisc_running_key ?: &qdisc_running_key); -+#else - seqcount_init(&sch->running); - lockdep_set_class(&sch->running, - dev->qdisc_running_key ?: &qdisc_running_key); -+#endif - - sch->ops = ops; - sch->enqueue = ops->enqueue; -@@ -925,7 +937,7 @@ - /* Wait for outstanding qdisc_run calls. */ - list_for_each_entry(dev, head, close_list) - while (some_qdisc_is_busy(dev)) -- yield(); -+ msleep(1); - } - - void dev_deactivate(struct net_device *dev) -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/sunrpc/svc_xprt.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/sunrpc/svc_xprt.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/net/sunrpc/svc_xprt.c 2017-04-16 10:38:34.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/net/sunrpc/svc_xprt.c 2017-04-18 17:54:27.000000000 +0200 -@@ -396,7 +396,7 @@ - goto out; - } - -- cpu = get_cpu(); -+ cpu = get_cpu_light(); - pool = svc_pool_for_cpu(xprt->xpt_server, cpu); - - atomic_long_inc(&pool->sp_stats.packets); -@@ -432,7 +432,7 @@ - - atomic_long_inc(&pool->sp_stats.threads_woken); - wake_up_process(rqstp->rq_task); -- put_cpu(); -+ put_cpu_light(); - goto out; - } - rcu_read_unlock(); -@@ -453,7 +453,7 @@ - goto redo_search; - } - rqstp = NULL; -- put_cpu(); -+ put_cpu_light(); - out: - trace_svc_xprt_do_enqueue(xprt, rqstp); - } -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/scripts/mkcompile_h linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/scripts/mkcompile_h ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/scripts/mkcompile_h 2017-04-16 10:38:34.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/scripts/mkcompile_h 2017-04-18 17:54:27.000000000 +0200 -@@ -4,7 +4,8 @@ - ARCH=$2 - SMP=$3 - PREEMPT=$4 --CC=$5 -+RT=$5 -+CC=$6 - - vecho() { [ "${quiet}" = "silent_" ] || echo "$@" ; } - -@@ -57,6 +58,7 @@ - CONFIG_FLAGS="" - if [ -n "$SMP" ] ; then CONFIG_FLAGS="SMP"; fi - if [ -n "$PREEMPT" ] ; then CONFIG_FLAGS="$CONFIG_FLAGS PREEMPT"; fi -+if [ -n "$RT" ] ; then CONFIG_FLAGS="$CONFIG_FLAGS RT"; fi - UTS_VERSION="$UTS_VERSION $CONFIG_FLAGS $TIMESTAMP" - - # Truncate to maximum length -diff -Nur linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/sound/core/pcm_native.c linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/sound/core/pcm_native.c ---- linux-72134397d72079a533c8fc742701fdc7f5ae7c5b.orig/sound/core/pcm_native.c 2017-04-16 10:38:35.000000000 +0200 -+++ linux-72134397d72079a533c8fc742701fdc7f5ae7c5b/sound/core/pcm_native.c 2017-04-18 17:54:27.000000000 +0200 -@@ -135,7 +135,7 @@ - void snd_pcm_stream_lock_irq(struct snd_pcm_substream *substream) - { - if (!substream->pcm->nonatomic) -- local_irq_disable(); -+ local_irq_disable_nort(); - snd_pcm_stream_lock(substream); - } - EXPORT_SYMBOL_GPL(snd_pcm_stream_lock_irq); -@@ -150,7 +150,7 @@ - { - snd_pcm_stream_unlock(substream); - if (!substream->pcm->nonatomic) -- local_irq_enable(); -+ local_irq_enable_nort(); - } - EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock_irq); - -@@ -158,7 +158,7 @@ - { - unsigned long flags = 0; - if (!substream->pcm->nonatomic) -- local_irq_save(flags); -+ local_irq_save_nort(flags); - snd_pcm_stream_lock(substream); - return flags; - } -@@ -176,7 +176,7 @@ - { - snd_pcm_stream_unlock(substream); - if (!substream->pcm->nonatomic) -- local_irq_restore(flags); -+ local_irq_restore_nort(flags); - } - EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock_irqrestore); - diff --git a/target/linux/patches/8327f281c8d5083b13b3a4e6e2a36ee7c3d67913/patch-realtime b/target/linux/patches/8327f281c8d5083b13b3a4e6e2a36ee7c3d67913/patch-realtime new file mode 100644 index 000000000..7abba16fe --- /dev/null +++ b/target/linux/patches/8327f281c8d5083b13b3a4e6e2a36ee7c3d67913/patch-realtime @@ -0,0 +1,26986 @@ +diff --git a/Documentation/sysrq.txt b/Documentation/sysrq.txt +index 3a3b30ac2a75..9e0745cafbd8 100644 +--- a/Documentation/sysrq.txt ++++ b/Documentation/sysrq.txt +@@ -59,10 +59,17 @@ On PowerPC - Press 'ALT - Print Screen (or F13) - , + On other - If you know of the key combos for other architectures, please + let me know so I can add them to this section. + +-On all - write a character to /proc/sysrq-trigger. e.g.: +- ++On all - write a character to /proc/sysrq-trigger, e.g.: + echo t > /proc/sysrq-trigger + ++On all - Enable network SysRq by writing a cookie to icmp_echo_sysrq, e.g. ++ echo 0x01020304 >/proc/sys/net/ipv4/icmp_echo_sysrq ++ Send an ICMP echo request with this pattern plus the particular ++ SysRq command key. Example: ++ # ping -c1 -s57 -p0102030468 ++ will trigger the SysRq-H (help) command. ++ ++ + * What are the 'command' keys? + ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + 'b' - Will immediately reboot the system without syncing or unmounting +diff --git a/Documentation/trace/histograms.txt b/Documentation/trace/histograms.txt +new file mode 100644 +index 000000000000..6f2aeabf7faa +--- /dev/null ++++ b/Documentation/trace/histograms.txt +@@ -0,0 +1,186 @@ ++ Using the Linux Kernel Latency Histograms ++ ++ ++This document gives a short explanation how to enable, configure and use ++latency histograms. Latency histograms are primarily relevant in the ++context of real-time enabled kernels (CONFIG_PREEMPT/CONFIG_PREEMPT_RT) ++and are used in the quality management of the Linux real-time ++capabilities. ++ ++ ++* Purpose of latency histograms ++ ++A latency histogram continuously accumulates the frequencies of latency ++data. There are two types of histograms ++- potential sources of latencies ++- effective latencies ++ ++ ++* Potential sources of latencies ++ ++Potential sources of latencies are code segments where interrupts, ++preemption or both are disabled (aka critical sections). To create ++histograms of potential sources of latency, the kernel stores the time ++stamp at the start of a critical section, determines the time elapsed ++when the end of the section is reached, and increments the frequency ++counter of that latency value - irrespective of whether any concurrently ++running process is affected by latency or not. ++- Configuration items (in the Kernel hacking/Tracers submenu) ++ CONFIG_INTERRUPT_OFF_LATENCY ++ CONFIG_PREEMPT_OFF_LATENCY ++ ++ ++* Effective latencies ++ ++Effective latencies are actually occuring during wakeup of a process. To ++determine effective latencies, the kernel stores the time stamp when a ++process is scheduled to be woken up, and determines the duration of the ++wakeup time shortly before control is passed over to this process. Note ++that the apparent latency in user space may be somewhat longer, since the ++process may be interrupted after control is passed over to it but before ++the execution in user space takes place. Simply measuring the interval ++between enqueuing and wakeup may also not appropriate in cases when a ++process is scheduled as a result of a timer expiration. The timer may have ++missed its deadline, e.g. due to disabled interrupts, but this latency ++would not be registered. Therefore, the offsets of missed timers are ++recorded in a separate histogram. If both wakeup latency and missed timer ++offsets are configured and enabled, a third histogram may be enabled that ++records the overall latency as a sum of the timer latency, if any, and the ++wakeup latency. This histogram is called "timerandwakeup". ++- Configuration items (in the Kernel hacking/Tracers submenu) ++ CONFIG_WAKEUP_LATENCY ++ CONFIG_MISSED_TIMER_OFSETS ++ ++ ++* Usage ++ ++The interface to the administration of the latency histograms is located ++in the debugfs file system. To mount it, either enter ++ ++mount -t sysfs nodev /sys ++mount -t debugfs nodev /sys/kernel/debug ++ ++from shell command line level, or add ++ ++nodev /sys sysfs defaults 0 0 ++nodev /sys/kernel/debug debugfs defaults 0 0 ++ ++to the file /etc/fstab. All latency histogram related files are then ++available in the directory /sys/kernel/debug/tracing/latency_hist. A ++particular histogram type is enabled by writing non-zero to the related ++variable in the /sys/kernel/debug/tracing/latency_hist/enable directory. ++Select "preemptirqsoff" for the histograms of potential sources of ++latencies and "wakeup" for histograms of effective latencies etc. The ++histogram data - one per CPU - are available in the files ++ ++/sys/kernel/debug/tracing/latency_hist/preemptoff/CPUx ++/sys/kernel/debug/tracing/latency_hist/irqsoff/CPUx ++/sys/kernel/debug/tracing/latency_hist/preemptirqsoff/CPUx ++/sys/kernel/debug/tracing/latency_hist/wakeup/CPUx ++/sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio/CPUx ++/sys/kernel/debug/tracing/latency_hist/missed_timer_offsets/CPUx ++/sys/kernel/debug/tracing/latency_hist/timerandwakeup/CPUx ++ ++The histograms are reset by writing non-zero to the file "reset" in a ++particular latency directory. To reset all latency data, use ++ ++#!/bin/sh ++ ++TRACINGDIR=/sys/kernel/debug/tracing ++HISTDIR=$TRACINGDIR/latency_hist ++ ++if test -d $HISTDIR ++then ++ cd $HISTDIR ++ for i in `find . | grep /reset$` ++ do ++ echo 1 >$i ++ done ++fi ++ ++ ++* Data format ++ ++Latency data are stored with a resolution of one microsecond. The ++maximum latency is 10,240 microseconds. The data are only valid, if the ++overflow register is empty. Every output line contains the latency in ++microseconds in the first row and the number of samples in the second ++row. To display only lines with a positive latency count, use, for ++example, ++ ++grep -v " 0$" /sys/kernel/debug/tracing/latency_hist/preemptoff/CPU0 ++ ++#Minimum latency: 0 microseconds. ++#Average latency: 0 microseconds. ++#Maximum latency: 25 microseconds. ++#Total samples: 3104770694 ++#There are 0 samples greater or equal than 10240 microseconds ++#usecs samples ++ 0 2984486876 ++ 1 49843506 ++ 2 58219047 ++ 3 5348126 ++ 4 2187960 ++ 5 3388262 ++ 6 959289 ++ 7 208294 ++ 8 40420 ++ 9 4485 ++ 10 14918 ++ 11 18340 ++ 12 25052 ++ 13 19455 ++ 14 5602 ++ 15 969 ++ 16 47 ++ 17 18 ++ 18 14 ++ 19 1 ++ 20 3 ++ 21 2 ++ 22 5 ++ 23 2 ++ 25 1 ++ ++ ++* Wakeup latency of a selected process ++ ++To only collect wakeup latency data of a particular process, write the ++PID of the requested process to ++ ++/sys/kernel/debug/tracing/latency_hist/wakeup/pid ++ ++PIDs are not considered, if this variable is set to 0. ++ ++ ++* Details of the process with the highest wakeup latency so far ++ ++Selected data of the process that suffered from the highest wakeup ++latency that occurred in a particular CPU are available in the file ++ ++/sys/kernel/debug/tracing/latency_hist/wakeup/max_latency-CPUx. ++ ++In addition, other relevant system data at the time when the ++latency occurred are given. ++ ++The format of the data is (all in one line): ++ () \ ++<- ++ ++The value of is only relevant in the combined timer ++and wakeup latency recording. In the wakeup recording, it is ++always 0, in the missed_timer_offsets recording, it is the same ++as . ++ ++When retrospectively searching for the origin of a latency and ++tracing was not enabled, it may be helpful to know the name and ++some basic data of the task that (finally) was switching to the ++late real-tlme task. In addition to the victim's data, also the ++data of the possible culprit are therefore displayed after the ++"<-" symbol. ++ ++Finally, the timestamp of the time when the latency occurred ++in . after the most recent system boot ++is provided. ++ ++These data are also reset when the wakeup histogram is reset. +diff --git a/MAINTAINERS b/MAINTAINERS +index 63cefa62324c..be0ea1e5c4cc 100644 +--- a/MAINTAINERS ++++ b/MAINTAINERS +@@ -5196,6 +5196,23 @@ F: fs/fuse/ + F: include/uapi/linux/fuse.h + F: Documentation/filesystems/fuse.txt + ++FUTEX SUBSYSTEM ++M: Thomas Gleixner ++M: Ingo Molnar ++R: Peter Zijlstra ++R: Darren Hart ++L: linux-kernel@vger.kernel.org ++T: git git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git locking/core ++S: Maintained ++F: kernel/futex.c ++F: kernel/futex_compat.c ++F: include/asm-generic/futex.h ++F: include/linux/futex.h ++F: include/uapi/linux/futex.h ++F: tools/testing/selftests/futex/ ++F: tools/perf/bench/futex* ++F: Documentation/*futex* ++ + FUTURE DOMAIN TMC-16x0 SCSI DRIVER (16-bit) + M: Rik Faith + L: linux-scsi@vger.kernel.org +diff --git a/arch/Kconfig b/arch/Kconfig +index 659bdd079277..099fc0f5155e 100644 +--- a/arch/Kconfig ++++ b/arch/Kconfig +@@ -9,6 +9,7 @@ config OPROFILE + tristate "OProfile system profiling" + depends on PROFILING + depends on HAVE_OPROFILE ++ depends on !PREEMPT_RT_FULL + select RING_BUFFER + select RING_BUFFER_ALLOW_SWAP + help +@@ -52,6 +53,7 @@ config KPROBES + config JUMP_LABEL + bool "Optimize very unlikely/likely branches" + depends on HAVE_ARCH_JUMP_LABEL ++ depends on (!INTERRUPT_OFF_HIST && !PREEMPT_OFF_HIST && !WAKEUP_LATENCY_HIST && !MISSED_TIMER_OFFSETS_HIST) + help + This option enables a transparent branch optimization that + makes certain almost-always-true or almost-always-false branch +diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig +index b5d529fdffab..5715844e83e3 100644 +--- a/arch/arm/Kconfig ++++ b/arch/arm/Kconfig +@@ -36,7 +36,7 @@ config ARM + select HAVE_ARCH_AUDITSYSCALL if (AEABI && !OABI_COMPAT) + select HAVE_ARCH_BITREVERSE if (CPU_32v7M || CPU_32v7) && !CPU_32v6 + select HAVE_ARCH_HARDENED_USERCOPY +- select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL && !CPU_ENDIAN_BE32 && MMU ++ select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL && !CPU_ENDIAN_BE32 && MMU && !PREEMPT_RT_BASE + select HAVE_ARCH_KGDB if !CPU_ENDIAN_BE32 && MMU + select HAVE_ARCH_MMAP_RND_BITS if MMU + select HAVE_ARCH_SECCOMP_FILTER if (AEABI && !OABI_COMPAT) +@@ -75,6 +75,7 @@ config ARM + select HAVE_PERF_EVENTS + select HAVE_PERF_REGS + select HAVE_PERF_USER_STACK_DUMP ++ select HAVE_PREEMPT_LAZY + select HAVE_RCU_TABLE_FREE if (SMP && ARM_LPAE) + select HAVE_REGS_AND_STACK_ACCESS_API + select HAVE_SYSCALL_TRACEPOINTS +diff --git a/arch/arm/include/asm/irq.h b/arch/arm/include/asm/irq.h +index e53638c8ed8a..6095a1649865 100644 +--- a/arch/arm/include/asm/irq.h ++++ b/arch/arm/include/asm/irq.h +@@ -22,6 +22,8 @@ + #endif + + #ifndef __ASSEMBLY__ ++#include ++ + struct irqaction; + struct pt_regs; + extern void migrate_irqs(void); +diff --git a/arch/arm/include/asm/switch_to.h b/arch/arm/include/asm/switch_to.h +index 12ebfcc1d539..c962084605bc 100644 +--- a/arch/arm/include/asm/switch_to.h ++++ b/arch/arm/include/asm/switch_to.h +@@ -3,6 +3,13 @@ + + #include + ++#if defined CONFIG_PREEMPT_RT_FULL && defined CONFIG_HIGHMEM ++void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p); ++#else ++static inline void ++switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) { } ++#endif ++ + /* + * For v7 SMP cores running a preemptible kernel we may be pre-empted + * during a TLB maintenance operation, so execute an inner-shareable dsb +@@ -25,6 +32,7 @@ extern struct task_struct *__switch_to(struct task_struct *, struct thread_info + #define switch_to(prev,next,last) \ + do { \ + __complete_pending_tlbi(); \ ++ switch_kmaps(prev, next); \ + last = __switch_to(prev,task_thread_info(prev), task_thread_info(next)); \ + } while (0) + +diff --git a/arch/arm/include/asm/thread_info.h b/arch/arm/include/asm/thread_info.h +index 776757d1604a..1f36a4eccc72 100644 +--- a/arch/arm/include/asm/thread_info.h ++++ b/arch/arm/include/asm/thread_info.h +@@ -49,6 +49,7 @@ struct cpu_context_save { + struct thread_info { + unsigned long flags; /* low level flags */ + int preempt_count; /* 0 => preemptable, <0 => bug */ ++ int preempt_lazy_count; /* 0 => preemptable, <0 => bug */ + mm_segment_t addr_limit; /* address limit */ + struct task_struct *task; /* main task structure */ + __u32 cpu; /* cpu */ +@@ -142,7 +143,8 @@ extern int vfp_restore_user_hwstate(struct user_vfp __user *, + #define TIF_SYSCALL_TRACE 4 /* syscall trace active */ + #define TIF_SYSCALL_AUDIT 5 /* syscall auditing active */ + #define TIF_SYSCALL_TRACEPOINT 6 /* syscall tracepoint instrumentation */ +-#define TIF_SECCOMP 7 /* seccomp syscall filtering active */ ++#define TIF_SECCOMP 8 /* seccomp syscall filtering active */ ++#define TIF_NEED_RESCHED_LAZY 7 + + #define TIF_NOHZ 12 /* in adaptive nohz mode */ + #define TIF_USING_IWMMXT 17 +@@ -152,6 +154,7 @@ extern int vfp_restore_user_hwstate(struct user_vfp __user *, + #define _TIF_SIGPENDING (1 << TIF_SIGPENDING) + #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) + #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) ++#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) + #define _TIF_UPROBE (1 << TIF_UPROBE) + #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) + #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) +@@ -167,7 +170,8 @@ extern int vfp_restore_user_hwstate(struct user_vfp __user *, + * Change these and you break ASM code in entry-common.S + */ + #define _TIF_WORK_MASK (_TIF_NEED_RESCHED | _TIF_SIGPENDING | \ +- _TIF_NOTIFY_RESUME | _TIF_UPROBE) ++ _TIF_NOTIFY_RESUME | _TIF_UPROBE | \ ++ _TIF_NEED_RESCHED_LAZY) + + #endif /* __KERNEL__ */ + #endif /* __ASM_ARM_THREAD_INFO_H */ +diff --git a/arch/arm/kernel/asm-offsets.c b/arch/arm/kernel/asm-offsets.c +index 608008229c7d..3866da3f7bb7 100644 +--- a/arch/arm/kernel/asm-offsets.c ++++ b/arch/arm/kernel/asm-offsets.c +@@ -65,6 +65,7 @@ int main(void) + BLANK(); + DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); + DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); ++ DEFINE(TI_PREEMPT_LAZY, offsetof(struct thread_info, preempt_lazy_count)); + DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit)); + DEFINE(TI_TASK, offsetof(struct thread_info, task)); + DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); +diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S +index 9f157e7c51e7..468e224d76aa 100644 +--- a/arch/arm/kernel/entry-armv.S ++++ b/arch/arm/kernel/entry-armv.S +@@ -220,11 +220,18 @@ ENDPROC(__dabt_svc) + + #ifdef CONFIG_PREEMPT + ldr r8, [tsk, #TI_PREEMPT] @ get preempt count +- ldr r0, [tsk, #TI_FLAGS] @ get flags + teq r8, #0 @ if preempt count != 0 ++ bne 1f @ return from exeption ++ ldr r0, [tsk, #TI_FLAGS] @ get flags ++ tst r0, #_TIF_NEED_RESCHED @ if NEED_RESCHED is set ++ blne svc_preempt @ preempt! ++ ++ ldr r8, [tsk, #TI_PREEMPT_LAZY] @ get preempt lazy count ++ teq r8, #0 @ if preempt lazy count != 0 + movne r0, #0 @ force flags to 0 +- tst r0, #_TIF_NEED_RESCHED ++ tst r0, #_TIF_NEED_RESCHED_LAZY + blne svc_preempt ++1: + #endif + + svc_exit r5, irq = 1 @ return from exception +@@ -239,8 +246,14 @@ ENDPROC(__irq_svc) + 1: bl preempt_schedule_irq @ irq en/disable is done inside + ldr r0, [tsk, #TI_FLAGS] @ get new tasks TI_FLAGS + tst r0, #_TIF_NEED_RESCHED ++ bne 1b ++ tst r0, #_TIF_NEED_RESCHED_LAZY + reteq r8 @ go again +- b 1b ++ ldr r0, [tsk, #TI_PREEMPT_LAZY] @ get preempt lazy count ++ teq r0, #0 @ if preempt lazy count != 0 ++ beq 1b ++ ret r8 @ go again ++ + #endif + + __und_fault: +diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S +index 10c3283d6c19..8872937862cc 100644 +--- a/arch/arm/kernel/entry-common.S ++++ b/arch/arm/kernel/entry-common.S +@@ -36,7 +36,9 @@ + UNWIND(.cantunwind ) + disable_irq_notrace @ disable interrupts + ldr r1, [tsk, #TI_FLAGS] @ re-check for syscall tracing +- tst r1, #_TIF_SYSCALL_WORK | _TIF_WORK_MASK ++ tst r1, #((_TIF_SYSCALL_WORK | _TIF_WORK_MASK) & ~_TIF_SECCOMP) ++ bne fast_work_pending ++ tst r1, #_TIF_SECCOMP + bne fast_work_pending + + /* perform architecture specific actions before user return */ +@@ -62,8 +64,11 @@ ENDPROC(ret_fast_syscall) + str r0, [sp, #S_R0 + S_OFF]! @ save returned r0 + disable_irq_notrace @ disable interrupts + ldr r1, [tsk, #TI_FLAGS] @ re-check for syscall tracing +- tst r1, #_TIF_SYSCALL_WORK | _TIF_WORK_MASK ++ tst r1, #((_TIF_SYSCALL_WORK | _TIF_WORK_MASK) & ~_TIF_SECCOMP) ++ bne do_slower_path ++ tst r1, #_TIF_SECCOMP + beq no_work_pending ++do_slower_path: + UNWIND(.fnend ) + ENDPROC(ret_fast_syscall) + +diff --git a/arch/arm/kernel/patch.c b/arch/arm/kernel/patch.c +index 69bda1a5707e..1f665acaa6a9 100644 +--- a/arch/arm/kernel/patch.c ++++ b/arch/arm/kernel/patch.c +@@ -15,7 +15,7 @@ struct patch { + unsigned int insn; + }; + +-static DEFINE_SPINLOCK(patch_lock); ++static DEFINE_RAW_SPINLOCK(patch_lock); + + static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags) + __acquires(&patch_lock) +@@ -32,7 +32,7 @@ static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags) + return addr; + + if (flags) +- spin_lock_irqsave(&patch_lock, *flags); ++ raw_spin_lock_irqsave(&patch_lock, *flags); + else + __acquire(&patch_lock); + +@@ -47,7 +47,7 @@ static void __kprobes patch_unmap(int fixmap, unsigned long *flags) + clear_fixmap(fixmap); + + if (flags) +- spin_unlock_irqrestore(&patch_lock, *flags); ++ raw_spin_unlock_irqrestore(&patch_lock, *flags); + else + __release(&patch_lock); + } +diff --git a/arch/arm/kernel/process.c b/arch/arm/kernel/process.c +index 91d2d5b01414..750550098b59 100644 +--- a/arch/arm/kernel/process.c ++++ b/arch/arm/kernel/process.c +@@ -322,6 +322,30 @@ unsigned long arch_randomize_brk(struct mm_struct *mm) + } + + #ifdef CONFIG_MMU ++/* ++ * CONFIG_SPLIT_PTLOCK_CPUS results in a page->ptl lock. If the lock is not ++ * initialized by pgtable_page_ctor() then a coredump of the vector page will ++ * fail. ++ */ ++static int __init vectors_user_mapping_init_page(void) ++{ ++ struct page *page; ++ unsigned long addr = 0xffff0000; ++ pgd_t *pgd; ++ pud_t *pud; ++ pmd_t *pmd; ++ ++ pgd = pgd_offset_k(addr); ++ pud = pud_offset(pgd, addr); ++ pmd = pmd_offset(pud, addr); ++ page = pmd_page(*(pmd)); ++ ++ pgtable_page_ctor(page); ++ ++ return 0; ++} ++late_initcall(vectors_user_mapping_init_page); ++ + #ifdef CONFIG_KUSER_HELPERS + /* + * The vectors page is always readable from user space for the +diff --git a/arch/arm/kernel/signal.c b/arch/arm/kernel/signal.c +index 7b8f2141427b..96541e00b74a 100644 +--- a/arch/arm/kernel/signal.c ++++ b/arch/arm/kernel/signal.c +@@ -572,7 +572,8 @@ do_work_pending(struct pt_regs *regs, unsigned int thread_flags, int syscall) + */ + trace_hardirqs_off(); + do { +- if (likely(thread_flags & _TIF_NEED_RESCHED)) { ++ if (likely(thread_flags & (_TIF_NEED_RESCHED | ++ _TIF_NEED_RESCHED_LAZY))) { + schedule(); + } else { + if (unlikely(!user_mode(regs))) +diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c +index 7dd14e8395e6..4cd7e3d98035 100644 +--- a/arch/arm/kernel/smp.c ++++ b/arch/arm/kernel/smp.c +@@ -234,8 +234,6 @@ int __cpu_disable(void) + flush_cache_louis(); + local_flush_tlb_all(); + +- clear_tasks_mm_cpumask(cpu); +- + return 0; + } + +@@ -251,6 +249,9 @@ void __cpu_die(unsigned int cpu) + pr_err("CPU%u: cpu didn't die\n", cpu); + return; + } ++ ++ clear_tasks_mm_cpumask(cpu); ++ + pr_notice("CPU%u: shutdown\n", cpu); + + /* +diff --git a/arch/arm/kernel/unwind.c b/arch/arm/kernel/unwind.c +index 0bee233fef9a..314cfb232a63 100644 +--- a/arch/arm/kernel/unwind.c ++++ b/arch/arm/kernel/unwind.c +@@ -93,7 +93,7 @@ extern const struct unwind_idx __start_unwind_idx[]; + static const struct unwind_idx *__origin_unwind_idx; + extern const struct unwind_idx __stop_unwind_idx[]; + +-static DEFINE_SPINLOCK(unwind_lock); ++static DEFINE_RAW_SPINLOCK(unwind_lock); + static LIST_HEAD(unwind_tables); + + /* Convert a prel31 symbol to an absolute address */ +@@ -201,7 +201,7 @@ static const struct unwind_idx *unwind_find_idx(unsigned long addr) + /* module unwind tables */ + struct unwind_table *table; + +- spin_lock_irqsave(&unwind_lock, flags); ++ raw_spin_lock_irqsave(&unwind_lock, flags); + list_for_each_entry(table, &unwind_tables, list) { + if (addr >= table->begin_addr && + addr < table->end_addr) { +@@ -213,7 +213,7 @@ static const struct unwind_idx *unwind_find_idx(unsigned long addr) + break; + } + } +- spin_unlock_irqrestore(&unwind_lock, flags); ++ raw_spin_unlock_irqrestore(&unwind_lock, flags); + } + + pr_debug("%s: idx = %p\n", __func__, idx); +@@ -529,9 +529,9 @@ struct unwind_table *unwind_table_add(unsigned long start, unsigned long size, + tab->begin_addr = text_addr; + tab->end_addr = text_addr + text_size; + +- spin_lock_irqsave(&unwind_lock, flags); ++ raw_spin_lock_irqsave(&unwind_lock, flags); + list_add_tail(&tab->list, &unwind_tables); +- spin_unlock_irqrestore(&unwind_lock, flags); ++ raw_spin_unlock_irqrestore(&unwind_lock, flags); + + return tab; + } +@@ -543,9 +543,9 @@ void unwind_table_del(struct unwind_table *tab) + if (!tab) + return; + +- spin_lock_irqsave(&unwind_lock, flags); ++ raw_spin_lock_irqsave(&unwind_lock, flags); + list_del(&tab->list); +- spin_unlock_irqrestore(&unwind_lock, flags); ++ raw_spin_unlock_irqrestore(&unwind_lock, flags); + + kfree(tab); + } +diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c +index 19b5f5c1c0ff..82aa639e6737 100644 +--- a/arch/arm/kvm/arm.c ++++ b/arch/arm/kvm/arm.c +@@ -619,7 +619,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) + * involves poking the GIC, which must be done in a + * non-preemptible context. + */ +- preempt_disable(); ++ migrate_disable(); + kvm_pmu_flush_hwstate(vcpu); + kvm_timer_flush_hwstate(vcpu); + kvm_vgic_flush_hwstate(vcpu); +@@ -640,7 +640,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) + kvm_pmu_sync_hwstate(vcpu); + kvm_timer_sync_hwstate(vcpu); + kvm_vgic_sync_hwstate(vcpu); +- preempt_enable(); ++ migrate_enable(); + continue; + } + +@@ -696,7 +696,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) + + kvm_vgic_sync_hwstate(vcpu); + +- preempt_enable(); ++ migrate_enable(); + + ret = handle_exit(vcpu, run, ret); + } +diff --git a/arch/arm/mach-exynos/platsmp.c b/arch/arm/mach-exynos/platsmp.c +index 98ffe1e62ad5..df9769ddece5 100644 +--- a/arch/arm/mach-exynos/platsmp.c ++++ b/arch/arm/mach-exynos/platsmp.c +@@ -229,7 +229,7 @@ static void __iomem *scu_base_addr(void) + return (void __iomem *)(S5P_VA_SCU); + } + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + static void exynos_secondary_init(unsigned int cpu) + { +@@ -242,8 +242,8 @@ static void exynos_secondary_init(unsigned int cpu) + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + int exynos_set_boot_addr(u32 core_id, unsigned long boot_addr) +@@ -307,7 +307,7 @@ static int exynos_boot_secondary(unsigned int cpu, struct task_struct *idle) + * Set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * The secondary processor is waiting to be released from +@@ -334,7 +334,7 @@ static int exynos_boot_secondary(unsigned int cpu, struct task_struct *idle) + + if (timeout == 0) { + printk(KERN_ERR "cpu1 power enable failed"); +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + return -ETIMEDOUT; + } + } +@@ -380,7 +380,7 @@ static int exynos_boot_secondary(unsigned int cpu, struct task_struct *idle) + * calibrations, then wait for it to finish + */ + fail: +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return pen_release != -1 ? ret : 0; + } +diff --git a/arch/arm/mach-hisi/platmcpm.c b/arch/arm/mach-hisi/platmcpm.c +index 4b653a8cb75c..b03d5a922cb1 100644 +--- a/arch/arm/mach-hisi/platmcpm.c ++++ b/arch/arm/mach-hisi/platmcpm.c +@@ -61,7 +61,7 @@ + + static void __iomem *sysctrl, *fabric; + static int hip04_cpu_table[HIP04_MAX_CLUSTERS][HIP04_MAX_CPUS_PER_CLUSTER]; +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + static u32 fabric_phys_addr; + /* + * [0]: bootwrapper physical address +@@ -113,7 +113,7 @@ static int hip04_boot_secondary(unsigned int l_cpu, struct task_struct *idle) + if (cluster >= HIP04_MAX_CLUSTERS || cpu >= HIP04_MAX_CPUS_PER_CLUSTER) + return -EINVAL; + +- spin_lock_irq(&boot_lock); ++ raw_spin_lock_irq(&boot_lock); + + if (hip04_cpu_table[cluster][cpu]) + goto out; +@@ -147,7 +147,7 @@ static int hip04_boot_secondary(unsigned int l_cpu, struct task_struct *idle) + + out: + hip04_cpu_table[cluster][cpu]++; +- spin_unlock_irq(&boot_lock); ++ raw_spin_unlock_irq(&boot_lock); + + return 0; + } +@@ -162,11 +162,11 @@ static void hip04_cpu_die(unsigned int l_cpu) + cpu = MPIDR_AFFINITY_LEVEL(mpidr, 0); + cluster = MPIDR_AFFINITY_LEVEL(mpidr, 1); + +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + hip04_cpu_table[cluster][cpu]--; + if (hip04_cpu_table[cluster][cpu] == 1) { + /* A power_up request went ahead of us. */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + return; + } else if (hip04_cpu_table[cluster][cpu] > 1) { + pr_err("Cluster %d CPU%d boots multiple times\n", cluster, cpu); +@@ -174,7 +174,7 @@ static void hip04_cpu_die(unsigned int l_cpu) + } + + last_man = hip04_cluster_is_down(cluster); +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + if (last_man) { + /* Since it's Cortex A15, disable L2 prefetching. */ + asm volatile( +@@ -203,7 +203,7 @@ static int hip04_cpu_kill(unsigned int l_cpu) + cpu >= HIP04_MAX_CPUS_PER_CLUSTER); + + count = TIMEOUT_MSEC / POLL_MSEC; +- spin_lock_irq(&boot_lock); ++ raw_spin_lock_irq(&boot_lock); + for (tries = 0; tries < count; tries++) { + if (hip04_cpu_table[cluster][cpu]) + goto err; +@@ -211,10 +211,10 @@ static int hip04_cpu_kill(unsigned int l_cpu) + data = readl_relaxed(sysctrl + SC_CPU_RESET_STATUS(cluster)); + if (data & CORE_WFI_STATUS(cpu)) + break; +- spin_unlock_irq(&boot_lock); ++ raw_spin_unlock_irq(&boot_lock); + /* Wait for clean L2 when the whole cluster is down. */ + msleep(POLL_MSEC); +- spin_lock_irq(&boot_lock); ++ raw_spin_lock_irq(&boot_lock); + } + if (tries >= count) + goto err; +@@ -231,10 +231,10 @@ static int hip04_cpu_kill(unsigned int l_cpu) + goto err; + if (hip04_cluster_is_down(cluster)) + hip04_set_snoop_filter(cluster, 0); +- spin_unlock_irq(&boot_lock); ++ raw_spin_unlock_irq(&boot_lock); + return 1; + err: +- spin_unlock_irq(&boot_lock); ++ raw_spin_unlock_irq(&boot_lock); + return 0; + } + #endif +diff --git a/arch/arm/mach-omap2/omap-smp.c b/arch/arm/mach-omap2/omap-smp.c +index b4de3da6dffa..b52893319d75 100644 +--- a/arch/arm/mach-omap2/omap-smp.c ++++ b/arch/arm/mach-omap2/omap-smp.c +@@ -64,7 +64,7 @@ static const struct omap_smp_config omap5_cfg __initconst = { + .startup_addr = omap5_secondary_startup, + }; + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + void __iomem *omap4_get_scu_base(void) + { +@@ -131,8 +131,8 @@ static void omap4_secondary_init(unsigned int cpu) + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + static int omap4_boot_secondary(unsigned int cpu, struct task_struct *idle) +@@ -146,7 +146,7 @@ static int omap4_boot_secondary(unsigned int cpu, struct task_struct *idle) + * Set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * Update the AuxCoreBoot0 with boot state for secondary core. +@@ -223,7 +223,7 @@ static int omap4_boot_secondary(unsigned int cpu, struct task_struct *idle) + * Now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return 0; + } +diff --git a/arch/arm/mach-prima2/platsmp.c b/arch/arm/mach-prima2/platsmp.c +index 0875b99add18..18b6d98d2581 100644 +--- a/arch/arm/mach-prima2/platsmp.c ++++ b/arch/arm/mach-prima2/platsmp.c +@@ -22,7 +22,7 @@ + + static void __iomem *clk_base; + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + static void sirfsoc_secondary_init(unsigned int cpu) + { +@@ -36,8 +36,8 @@ static void sirfsoc_secondary_init(unsigned int cpu) + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + static const struct of_device_id clk_ids[] = { +@@ -75,7 +75,7 @@ static int sirfsoc_boot_secondary(unsigned int cpu, struct task_struct *idle) + /* make sure write buffer is drained */ + mb(); + +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * The secondary processor is waiting to be released from +@@ -107,7 +107,7 @@ static int sirfsoc_boot_secondary(unsigned int cpu, struct task_struct *idle) + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return pen_release != -1 ? -ENOSYS : 0; + } +diff --git a/arch/arm/mach-qcom/platsmp.c b/arch/arm/mach-qcom/platsmp.c +index 5494c9e0c909..e8ce157d3548 100644 +--- a/arch/arm/mach-qcom/platsmp.c ++++ b/arch/arm/mach-qcom/platsmp.c +@@ -46,7 +46,7 @@ + + extern void secondary_startup_arm(void); + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + #ifdef CONFIG_HOTPLUG_CPU + static void qcom_cpu_die(unsigned int cpu) +@@ -60,8 +60,8 @@ static void qcom_secondary_init(unsigned int cpu) + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + static int scss_release_secondary(unsigned int cpu) +@@ -284,7 +284,7 @@ static int qcom_boot_secondary(unsigned int cpu, int (*func)(unsigned int)) + * set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * Send the secondary CPU a soft interrupt, thereby causing +@@ -297,7 +297,7 @@ static int qcom_boot_secondary(unsigned int cpu, int (*func)(unsigned int)) + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return ret; + } +diff --git a/arch/arm/mach-spear/platsmp.c b/arch/arm/mach-spear/platsmp.c +index 8d1e2d551786..7fa56cc78118 100644 +--- a/arch/arm/mach-spear/platsmp.c ++++ b/arch/arm/mach-spear/platsmp.c +@@ -32,7 +32,7 @@ static void write_pen_release(int val) + sync_cache_w(&pen_release); + } + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + static void __iomem *scu_base = IOMEM(VA_SCU_BASE); + +@@ -47,8 +47,8 @@ static void spear13xx_secondary_init(unsigned int cpu) + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + static int spear13xx_boot_secondary(unsigned int cpu, struct task_struct *idle) +@@ -59,7 +59,7 @@ static int spear13xx_boot_secondary(unsigned int cpu, struct task_struct *idle) + * set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * The secondary processor is waiting to be released from +@@ -84,7 +84,7 @@ static int spear13xx_boot_secondary(unsigned int cpu, struct task_struct *idle) + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return pen_release != -1 ? -ENOSYS : 0; + } +diff --git a/arch/arm/mach-sti/platsmp.c b/arch/arm/mach-sti/platsmp.c +index ea5a2277ee46..b988e081ac79 100644 +--- a/arch/arm/mach-sti/platsmp.c ++++ b/arch/arm/mach-sti/platsmp.c +@@ -35,7 +35,7 @@ static void write_pen_release(int val) + sync_cache_w(&pen_release); + } + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + static void sti_secondary_init(unsigned int cpu) + { +@@ -48,8 +48,8 @@ static void sti_secondary_init(unsigned int cpu) + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + static int sti_boot_secondary(unsigned int cpu, struct task_struct *idle) +@@ -60,7 +60,7 @@ static int sti_boot_secondary(unsigned int cpu, struct task_struct *idle) + * set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * The secondary processor is waiting to be released from +@@ -91,7 +91,7 @@ static int sti_boot_secondary(unsigned int cpu, struct task_struct *idle) + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return pen_release != -1 ? -ENOSYS : 0; + } +diff --git a/arch/arm/mm/fault.c b/arch/arm/mm/fault.c +index 0122ad1a6027..926b1be48043 100644 +--- a/arch/arm/mm/fault.c ++++ b/arch/arm/mm/fault.c +@@ -430,6 +430,9 @@ do_translation_fault(unsigned long addr, unsigned int fsr, + if (addr < TASK_SIZE) + return do_page_fault(addr, fsr, regs); + ++ if (interrupts_enabled(regs)) ++ local_irq_enable(); ++ + if (user_mode(regs)) + goto bad_area; + +@@ -497,6 +500,9 @@ do_translation_fault(unsigned long addr, unsigned int fsr, + static int + do_sect_fault(unsigned long addr, unsigned int fsr, struct pt_regs *regs) + { ++ if (interrupts_enabled(regs)) ++ local_irq_enable(); ++ + do_bad_area(addr, fsr, regs); + return 0; + } +diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c +index d02f8187b1cc..542692dbd40a 100644 +--- a/arch/arm/mm/highmem.c ++++ b/arch/arm/mm/highmem.c +@@ -34,6 +34,11 @@ static inline pte_t get_fixmap_pte(unsigned long vaddr) + return *ptep; + } + ++static unsigned int fixmap_idx(int type) ++{ ++ return FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); ++} ++ + void *kmap(struct page *page) + { + might_sleep(); +@@ -54,12 +59,13 @@ EXPORT_SYMBOL(kunmap); + + void *kmap_atomic(struct page *page) + { ++ pte_t pte = mk_pte(page, kmap_prot); + unsigned int idx; + unsigned long vaddr; + void *kmap; + int type; + +- preempt_disable(); ++ preempt_disable_nort(); + pagefault_disable(); + if (!PageHighMem(page)) + return page_address(page); +@@ -79,7 +85,7 @@ void *kmap_atomic(struct page *page) + + type = kmap_atomic_idx_push(); + +- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); ++ idx = fixmap_idx(type); + vaddr = __fix_to_virt(idx); + #ifdef CONFIG_DEBUG_HIGHMEM + /* +@@ -93,7 +99,10 @@ void *kmap_atomic(struct page *page) + * in place, so the contained TLB flush ensures the TLB is updated + * with the new mapping. + */ +- set_fixmap_pte(idx, mk_pte(page, kmap_prot)); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = pte; ++#endif ++ set_fixmap_pte(idx, pte); + + return (void *)vaddr; + } +@@ -106,44 +115,75 @@ void __kunmap_atomic(void *kvaddr) + + if (kvaddr >= (void *)FIXADDR_START) { + type = kmap_atomic_idx(); +- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); ++ idx = fixmap_idx(type); + + if (cache_is_vivt()) + __cpuc_flush_dcache_area((void *)vaddr, PAGE_SIZE); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = __pte(0); ++#endif + #ifdef CONFIG_DEBUG_HIGHMEM + BUG_ON(vaddr != __fix_to_virt(idx)); +- set_fixmap_pte(idx, __pte(0)); + #else + (void) idx; /* to kill a warning */ + #endif ++ set_fixmap_pte(idx, __pte(0)); + kmap_atomic_idx_pop(); + } else if (vaddr >= PKMAP_ADDR(0) && vaddr < PKMAP_ADDR(LAST_PKMAP)) { + /* this address was obtained through kmap_high_get() */ + kunmap_high(pte_page(pkmap_page_table[PKMAP_NR(vaddr)])); + } + pagefault_enable(); +- preempt_enable(); ++ preempt_enable_nort(); + } + EXPORT_SYMBOL(__kunmap_atomic); + + void *kmap_atomic_pfn(unsigned long pfn) + { ++ pte_t pte = pfn_pte(pfn, kmap_prot); + unsigned long vaddr; + int idx, type; + struct page *page = pfn_to_page(pfn); + +- preempt_disable(); ++ preempt_disable_nort(); + pagefault_disable(); + if (!PageHighMem(page)) + return page_address(page); + + type = kmap_atomic_idx_push(); +- idx = FIX_KMAP_BEGIN + type + KM_TYPE_NR * smp_processor_id(); ++ idx = fixmap_idx(type); + vaddr = __fix_to_virt(idx); + #ifdef CONFIG_DEBUG_HIGHMEM + BUG_ON(!pte_none(get_fixmap_pte(vaddr))); + #endif +- set_fixmap_pte(idx, pfn_pte(pfn, kmap_prot)); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = pte; ++#endif ++ set_fixmap_pte(idx, pte); + + return (void *)vaddr; + } ++#if defined CONFIG_PREEMPT_RT_FULL ++void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) ++{ ++ int i; ++ ++ /* ++ * Clear @prev's kmap_atomic mappings ++ */ ++ for (i = 0; i < prev_p->kmap_idx; i++) { ++ int idx = fixmap_idx(i); ++ ++ set_fixmap_pte(idx, __pte(0)); ++ } ++ /* ++ * Restore @next_p's kmap_atomic mappings ++ */ ++ for (i = 0; i < next_p->kmap_idx; i++) { ++ int idx = fixmap_idx(i); ++ ++ if (!pte_none(next_p->kmap_pte[i])) ++ set_fixmap_pte(idx, next_p->kmap_pte[i]); ++ } ++} ++#endif +diff --git a/arch/arm/plat-versatile/platsmp.c b/arch/arm/plat-versatile/platsmp.c +index c2366510187a..6b60f582b738 100644 +--- a/arch/arm/plat-versatile/platsmp.c ++++ b/arch/arm/plat-versatile/platsmp.c +@@ -32,7 +32,7 @@ static void write_pen_release(int val) + sync_cache_w(&pen_release); + } + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + void versatile_secondary_init(unsigned int cpu) + { +@@ -45,8 +45,8 @@ void versatile_secondary_init(unsigned int cpu) + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + int versatile_boot_secondary(unsigned int cpu, struct task_struct *idle) +@@ -57,7 +57,7 @@ int versatile_boot_secondary(unsigned int cpu, struct task_struct *idle) + * Set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * This is really belt and braces; we hold unintended secondary +@@ -87,7 +87,7 @@ int versatile_boot_secondary(unsigned int cpu, struct task_struct *idle) + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return pen_release != -1 ? -ENOSYS : 0; + } +diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig +index cf57a7799a0f..78d1b49fbed5 100644 +--- a/arch/arm64/Kconfig ++++ b/arch/arm64/Kconfig +@@ -91,6 +91,7 @@ config ARM64 + select HAVE_PERF_EVENTS + select HAVE_PERF_REGS + select HAVE_PERF_USER_STACK_DUMP ++ select HAVE_PREEMPT_LAZY + select HAVE_REGS_AND_STACK_ACCESS_API + select HAVE_RCU_TABLE_FREE + select HAVE_SYSCALL_TRACEPOINTS +@@ -704,7 +705,7 @@ config XEN_DOM0 + + config XEN + bool "Xen guest support on ARM64" +- depends on ARM64 && OF ++ depends on ARM64 && OF && !PREEMPT_RT_FULL + select SWIOTLB_XEN + select PARAVIRT + help +diff --git a/arch/arm64/include/asm/thread_info.h b/arch/arm64/include/asm/thread_info.h +index e9ea5a6bd449..6c500ad63c6a 100644 +--- a/arch/arm64/include/asm/thread_info.h ++++ b/arch/arm64/include/asm/thread_info.h +@@ -49,6 +49,7 @@ struct thread_info { + mm_segment_t addr_limit; /* address limit */ + struct task_struct *task; /* main task structure */ + int preempt_count; /* 0 => preemptable, <0 => bug */ ++ int preempt_lazy_count; /* 0 => preemptable, <0 => bug */ + int cpu; /* cpu */ + }; + +@@ -112,6 +113,7 @@ static inline struct thread_info *current_thread_info(void) + #define TIF_NEED_RESCHED 1 + #define TIF_NOTIFY_RESUME 2 /* callback before returning to user */ + #define TIF_FOREIGN_FPSTATE 3 /* CPU's FP state is not current's */ ++#define TIF_NEED_RESCHED_LAZY 4 + #define TIF_NOHZ 7 + #define TIF_SYSCALL_TRACE 8 + #define TIF_SYSCALL_AUDIT 9 +@@ -127,6 +129,7 @@ static inline struct thread_info *current_thread_info(void) + #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) + #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) + #define _TIF_FOREIGN_FPSTATE (1 << TIF_FOREIGN_FPSTATE) ++#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) + #define _TIF_NOHZ (1 << TIF_NOHZ) + #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) + #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) +@@ -135,7 +138,9 @@ static inline struct thread_info *current_thread_info(void) + #define _TIF_32BIT (1 << TIF_32BIT) + + #define _TIF_WORK_MASK (_TIF_NEED_RESCHED | _TIF_SIGPENDING | \ +- _TIF_NOTIFY_RESUME | _TIF_FOREIGN_FPSTATE) ++ _TIF_NOTIFY_RESUME | _TIF_FOREIGN_FPSTATE | \ ++ _TIF_NEED_RESCHED_LAZY) ++#define _TIF_NEED_RESCHED_MASK (_TIF_NEED_RESCHED | _TIF_NEED_RESCHED_LAZY) + + #define _TIF_SYSCALL_WORK (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | \ + _TIF_SYSCALL_TRACEPOINT | _TIF_SECCOMP | \ +diff --git a/arch/arm64/kernel/asm-offsets.c b/arch/arm64/kernel/asm-offsets.c +index c58ddf8c4062..a8f2f7c1fe12 100644 +--- a/arch/arm64/kernel/asm-offsets.c ++++ b/arch/arm64/kernel/asm-offsets.c +@@ -38,6 +38,7 @@ int main(void) + BLANK(); + DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); + DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); ++ DEFINE(TI_PREEMPT_LAZY, offsetof(struct thread_info, preempt_lazy_count)); + DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit)); + DEFINE(TI_TASK, offsetof(struct thread_info, task)); + DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); +diff --git a/arch/arm64/kernel/entry.S b/arch/arm64/kernel/entry.S +index 79b0fe24d5b7..f3c959ade308 100644 +--- a/arch/arm64/kernel/entry.S ++++ b/arch/arm64/kernel/entry.S +@@ -428,11 +428,16 @@ ENDPROC(el1_sync) + + #ifdef CONFIG_PREEMPT + ldr w24, [tsk, #TI_PREEMPT] // get preempt count +- cbnz w24, 1f // preempt count != 0 ++ cbnz w24, 2f // preempt count != 0 + ldr x0, [tsk, #TI_FLAGS] // get flags +- tbz x0, #TIF_NEED_RESCHED, 1f // needs rescheduling? +- bl el1_preempt ++ tbnz x0, #TIF_NEED_RESCHED, 1f // needs rescheduling? ++ ++ ldr w24, [tsk, #TI_PREEMPT_LAZY] // get preempt lazy count ++ cbnz w24, 2f // preempt lazy count != 0 ++ tbz x0, #TIF_NEED_RESCHED_LAZY, 2f // needs rescheduling? + 1: ++ bl el1_preempt ++2: + #endif + #ifdef CONFIG_TRACE_IRQFLAGS + bl trace_hardirqs_on +@@ -446,6 +451,7 @@ ENDPROC(el1_irq) + 1: bl preempt_schedule_irq // irq en/disable is done inside + ldr x0, [tsk, #TI_FLAGS] // get new tasks TI_FLAGS + tbnz x0, #TIF_NEED_RESCHED, 1b // needs rescheduling? ++ tbnz x0, #TIF_NEED_RESCHED_LAZY, 1b // needs rescheduling? + ret x24 + #endif + +diff --git a/arch/arm64/kernel/signal.c b/arch/arm64/kernel/signal.c +index 404dd67080b9..639dc6d12e72 100644 +--- a/arch/arm64/kernel/signal.c ++++ b/arch/arm64/kernel/signal.c +@@ -409,7 +409,7 @@ asmlinkage void do_notify_resume(struct pt_regs *regs, + */ + trace_hardirqs_off(); + do { +- if (thread_flags & _TIF_NEED_RESCHED) { ++ if (thread_flags & _TIF_NEED_RESCHED_MASK) { + schedule(); + } else { + local_irq_enable(); +diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig +index 5e844f68e847..dc613cc10f54 100644 +--- a/arch/mips/Kconfig ++++ b/arch/mips/Kconfig +@@ -2516,7 +2516,7 @@ config MIPS_ASID_BITS_VARIABLE + # + config HIGHMEM + bool "High Memory Support" +- depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM && !CPU_MIPS32_3_5_EVA ++ depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM && !CPU_MIPS32_3_5_EVA && !PREEMPT_RT_FULL + + config CPU_SUPPORTS_HIGHMEM + bool +diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig +index 8f01f21e78f1..619485b777d2 100644 +--- a/arch/powerpc/Kconfig ++++ b/arch/powerpc/Kconfig +@@ -52,10 +52,11 @@ config LOCKDEP_SUPPORT + + config RWSEM_GENERIC_SPINLOCK + bool ++ default y if PREEMPT_RT_FULL + + config RWSEM_XCHGADD_ALGORITHM + bool +- default y ++ default y if !PREEMPT_RT_FULL + + config GENERIC_LOCKBREAK + bool +@@ -134,6 +135,7 @@ config PPC + select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST + select GENERIC_STRNCPY_FROM_USER + select GENERIC_STRNLEN_USER ++ select HAVE_PREEMPT_LAZY + select HAVE_MOD_ARCH_SPECIFIC + select MODULES_USE_ELF_RELA + select CLONE_BACKWARDS +@@ -321,7 +323,7 @@ menu "Kernel options" + + config HIGHMEM + bool "High memory support" +- depends on PPC32 ++ depends on PPC32 && !PREEMPT_RT_FULL + + source kernel/Kconfig.hz + source kernel/Kconfig.preempt +diff --git a/arch/powerpc/include/asm/thread_info.h b/arch/powerpc/include/asm/thread_info.h +index 87e4b2d8dcd4..981e501a4359 100644 +--- a/arch/powerpc/include/asm/thread_info.h ++++ b/arch/powerpc/include/asm/thread_info.h +@@ -43,6 +43,8 @@ struct thread_info { + int cpu; /* cpu we're on */ + int preempt_count; /* 0 => preemptable, + <0 => BUG */ ++ int preempt_lazy_count; /* 0 => preemptable, ++ <0 => BUG */ + unsigned long local_flags; /* private flags for thread */ + #ifdef CONFIG_LIVEPATCH + unsigned long *livepatch_sp; +@@ -88,8 +90,7 @@ static inline struct thread_info *current_thread_info(void) + #define TIF_SYSCALL_TRACE 0 /* syscall trace active */ + #define TIF_SIGPENDING 1 /* signal pending */ + #define TIF_NEED_RESCHED 2 /* rescheduling necessary */ +-#define TIF_POLLING_NRFLAG 3 /* true if poll_idle() is polling +- TIF_NEED_RESCHED */ ++#define TIF_NEED_RESCHED_LAZY 3 /* lazy rescheduling necessary */ + #define TIF_32BIT 4 /* 32 bit binary */ + #define TIF_RESTORE_TM 5 /* need to restore TM FP/VEC/VSX */ + #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ +@@ -107,6 +108,8 @@ static inline struct thread_info *current_thread_info(void) + #if defined(CONFIG_PPC64) + #define TIF_ELF2ABI 18 /* function descriptors must die! */ + #endif ++#define TIF_POLLING_NRFLAG 19 /* true if poll_idle() is polling ++ TIF_NEED_RESCHED */ + + /* as above, but as bit values */ + #define _TIF_SYSCALL_TRACE (1<flags) + set_bits(irqtp->flags, &curtp->flags); + } ++#endif + + irq_hw_number_t virq_to_hw(unsigned int virq) + { +diff --git a/arch/powerpc/kernel/misc_32.S b/arch/powerpc/kernel/misc_32.S +index 030d72df5dd5..b471a709e100 100644 +--- a/arch/powerpc/kernel/misc_32.S ++++ b/arch/powerpc/kernel/misc_32.S +@@ -41,6 +41,7 @@ + * We store the saved ksp_limit in the unused part + * of the STACK_FRAME_OVERHEAD + */ ++#ifndef CONFIG_PREEMPT_RT_FULL + _GLOBAL(call_do_softirq) + mflr r0 + stw r0,4(r1) +@@ -57,6 +58,7 @@ _GLOBAL(call_do_softirq) + stw r10,THREAD+KSP_LIMIT(r2) + mtlr r0 + blr ++#endif + + /* + * void call_do_irq(struct pt_regs *regs, struct thread_info *irqtp); +diff --git a/arch/powerpc/kernel/misc_64.S b/arch/powerpc/kernel/misc_64.S +index 4cefe6888b18..cb2ee4be999a 100644 +--- a/arch/powerpc/kernel/misc_64.S ++++ b/arch/powerpc/kernel/misc_64.S +@@ -31,6 +31,7 @@ + + .text + ++#ifndef CONFIG_PREEMPT_RT_FULL + _GLOBAL(call_do_softirq) + mflr r0 + std r0,16(r1) +@@ -41,6 +42,7 @@ _GLOBAL(call_do_softirq) + ld r0,16(r1) + mtlr r0 + blr ++#endif + + _GLOBAL(call_do_irq) + mflr r0 +diff --git a/arch/powerpc/kvm/Kconfig b/arch/powerpc/kvm/Kconfig +index 029be26b5a17..9528089ea142 100644 +--- a/arch/powerpc/kvm/Kconfig ++++ b/arch/powerpc/kvm/Kconfig +@@ -175,6 +175,7 @@ config KVM_E500MC + config KVM_MPIC + bool "KVM in-kernel MPIC emulation" + depends on KVM && E500 ++ depends on !PREEMPT_RT_FULL + select HAVE_KVM_IRQCHIP + select HAVE_KVM_IRQFD + select HAVE_KVM_IRQ_ROUTING +diff --git a/arch/powerpc/platforms/ps3/device-init.c b/arch/powerpc/platforms/ps3/device-init.c +index e48462447ff0..2670cee66064 100644 +--- a/arch/powerpc/platforms/ps3/device-init.c ++++ b/arch/powerpc/platforms/ps3/device-init.c +@@ -752,7 +752,7 @@ static int ps3_notification_read_write(struct ps3_notification_device *dev, + } + pr_debug("%s:%u: notification %s issued\n", __func__, __LINE__, op); + +- res = wait_event_interruptible(dev->done.wait, ++ res = swait_event_interruptible(dev->done.wait, + dev->done.done || kthread_should_stop()); + if (kthread_should_stop()) + res = -EINTR; +diff --git a/arch/sh/kernel/irq.c b/arch/sh/kernel/irq.c +index 6c0378c0b8b5..abd58b4dff97 100644 +--- a/arch/sh/kernel/irq.c ++++ b/arch/sh/kernel/irq.c +@@ -147,6 +147,7 @@ void irq_ctx_exit(int cpu) + hardirq_ctx[cpu] = NULL; + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + void do_softirq_own_stack(void) + { + struct thread_info *curctx; +@@ -174,6 +175,7 @@ void do_softirq_own_stack(void) + "r5", "r6", "r7", "r8", "r9", "r15", "t", "pr" + ); + } ++#endif + #else + static inline void handle_one_irq(unsigned int irq) + { +diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig +index 165ecdd24d22..b68a464a22be 100644 +--- a/arch/sparc/Kconfig ++++ b/arch/sparc/Kconfig +@@ -194,12 +194,10 @@ config NR_CPUS + source kernel/Kconfig.hz + + config RWSEM_GENERIC_SPINLOCK +- bool +- default y if SPARC32 ++ def_bool PREEMPT_RT_FULL + + config RWSEM_XCHGADD_ALGORITHM +- bool +- default y if SPARC64 ++ def_bool !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL + + config GENERIC_HWEIGHT + bool +diff --git a/arch/sparc/kernel/irq_64.c b/arch/sparc/kernel/irq_64.c +index 34a7930b76ef..773740521008 100644 +--- a/arch/sparc/kernel/irq_64.c ++++ b/arch/sparc/kernel/irq_64.c +@@ -854,6 +854,7 @@ void __irq_entry handler_irq(int pil, struct pt_regs *regs) + set_irq_regs(old_regs); + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + void do_softirq_own_stack(void) + { + void *orig_sp, *sp = softirq_stack[smp_processor_id()]; +@@ -868,6 +869,7 @@ void do_softirq_own_stack(void) + __asm__ __volatile__("mov %0, %%sp" + : : "r" (orig_sp)); + } ++#endif + + #ifdef CONFIG_HOTPLUG_CPU + void fixup_irqs(void) +diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig +index bada636d1065..f8a995c90c01 100644 +--- a/arch/x86/Kconfig ++++ b/arch/x86/Kconfig +@@ -17,6 +17,7 @@ config X86_64 + ### Arch settings + config X86 + def_bool y ++ select HAVE_PREEMPT_LAZY + select ACPI_LEGACY_TABLES_LOOKUP if ACPI + select ACPI_SYSTEM_POWER_STATES_SUPPORT if ACPI + select ANON_INODES +@@ -232,8 +233,11 @@ config ARCH_MAY_HAVE_PC_FDC + def_bool y + depends on ISA_DMA_API + ++config RWSEM_GENERIC_SPINLOCK ++ def_bool PREEMPT_RT_FULL ++ + config RWSEM_XCHGADD_ALGORITHM +- def_bool y ++ def_bool !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL + + config GENERIC_CALIBRATE_DELAY + def_bool y +@@ -897,7 +901,7 @@ config IOMMU_HELPER + config MAXSMP + bool "Enable Maximum number of SMP Processors and NUMA Nodes" + depends on X86_64 && SMP && DEBUG_KERNEL +- select CPUMASK_OFFSTACK ++ select CPUMASK_OFFSTACK if !PREEMPT_RT_FULL + ---help--- + Enable maximum number of CPUS and NUMA Nodes for this architecture. + If unsure, say N. +diff --git a/arch/x86/crypto/aesni-intel_glue.c b/arch/x86/crypto/aesni-intel_glue.c +index aa8b0672f87a..2429414bfc71 100644 +--- a/arch/x86/crypto/aesni-intel_glue.c ++++ b/arch/x86/crypto/aesni-intel_glue.c +@@ -372,14 +372,14 @@ static int ecb_encrypt(struct blkcipher_desc *desc, + err = blkcipher_walk_virt(desc, &walk); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes)) { ++ kernel_fpu_begin(); + aesni_ecb_enc(ctx, walk.dst.virt.addr, walk.src.virt.addr, +- nbytes & AES_BLOCK_MASK); ++ nbytes & AES_BLOCK_MASK); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- kernel_fpu_end(); + + return err; + } +@@ -396,14 +396,14 @@ static int ecb_decrypt(struct blkcipher_desc *desc, + err = blkcipher_walk_virt(desc, &walk); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes)) { ++ kernel_fpu_begin(); + aesni_ecb_dec(ctx, walk.dst.virt.addr, walk.src.virt.addr, + nbytes & AES_BLOCK_MASK); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- kernel_fpu_end(); + + return err; + } +@@ -420,14 +420,14 @@ static int cbc_encrypt(struct blkcipher_desc *desc, + err = blkcipher_walk_virt(desc, &walk); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes)) { ++ kernel_fpu_begin(); + aesni_cbc_enc(ctx, walk.dst.virt.addr, walk.src.virt.addr, + nbytes & AES_BLOCK_MASK, walk.iv); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- kernel_fpu_end(); + + return err; + } +@@ -444,14 +444,14 @@ static int cbc_decrypt(struct blkcipher_desc *desc, + err = blkcipher_walk_virt(desc, &walk); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes)) { ++ kernel_fpu_begin(); + aesni_cbc_dec(ctx, walk.dst.virt.addr, walk.src.virt.addr, + nbytes & AES_BLOCK_MASK, walk.iv); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- kernel_fpu_end(); + + return err; + } +@@ -503,18 +503,20 @@ static int ctr_crypt(struct blkcipher_desc *desc, + err = blkcipher_walk_virt_block(desc, &walk, AES_BLOCK_SIZE); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes) >= AES_BLOCK_SIZE) { ++ kernel_fpu_begin(); + aesni_ctr_enc_tfm(ctx, walk.dst.virt.addr, walk.src.virt.addr, + nbytes & AES_BLOCK_MASK, walk.iv); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } + if (walk.nbytes) { ++ kernel_fpu_begin(); + ctr_crypt_final(ctx, &walk); ++ kernel_fpu_end(); + err = blkcipher_walk_done(desc, &walk, 0); + } +- kernel_fpu_end(); + + return err; + } +diff --git a/arch/x86/crypto/cast5_avx_glue.c b/arch/x86/crypto/cast5_avx_glue.c +index 8648158f3916..d7699130ee36 100644 +--- a/arch/x86/crypto/cast5_avx_glue.c ++++ b/arch/x86/crypto/cast5_avx_glue.c +@@ -59,7 +59,7 @@ static inline void cast5_fpu_end(bool fpu_enabled) + static int ecb_crypt(struct blkcipher_desc *desc, struct blkcipher_walk *walk, + bool enc) + { +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct cast5_ctx *ctx = crypto_blkcipher_ctx(desc->tfm); + const unsigned int bsize = CAST5_BLOCK_SIZE; + unsigned int nbytes; +@@ -75,7 +75,7 @@ static int ecb_crypt(struct blkcipher_desc *desc, struct blkcipher_walk *walk, + u8 *wsrc = walk->src.virt.addr; + u8 *wdst = walk->dst.virt.addr; + +- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); ++ fpu_enabled = cast5_fpu_begin(false, nbytes); + + /* Process multi-block batch */ + if (nbytes >= bsize * CAST5_PARALLEL_BLOCKS) { +@@ -103,10 +103,9 @@ static int ecb_crypt(struct blkcipher_desc *desc, struct blkcipher_walk *walk, + } while (nbytes >= bsize); + + done: ++ cast5_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, walk, nbytes); + } +- +- cast5_fpu_end(fpu_enabled); + return err; + } + +@@ -227,7 +226,7 @@ static unsigned int __cbc_decrypt(struct blkcipher_desc *desc, + static int cbc_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst, + struct scatterlist *src, unsigned int nbytes) + { +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; + +@@ -236,12 +235,11 @@ static int cbc_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst, + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + + while ((nbytes = walk.nbytes)) { +- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); ++ fpu_enabled = cast5_fpu_begin(false, nbytes); + nbytes = __cbc_decrypt(desc, &walk); ++ cast5_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- +- cast5_fpu_end(fpu_enabled); + return err; + } + +@@ -311,7 +309,7 @@ static unsigned int __ctr_crypt(struct blkcipher_desc *desc, + static int ctr_crypt(struct blkcipher_desc *desc, struct scatterlist *dst, + struct scatterlist *src, unsigned int nbytes) + { +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; + +@@ -320,13 +318,12 @@ static int ctr_crypt(struct blkcipher_desc *desc, struct scatterlist *dst, + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + + while ((nbytes = walk.nbytes) >= CAST5_BLOCK_SIZE) { +- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); ++ fpu_enabled = cast5_fpu_begin(false, nbytes); + nbytes = __ctr_crypt(desc, &walk); ++ cast5_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); + } + +- cast5_fpu_end(fpu_enabled); +- + if (walk.nbytes) { + ctr_crypt_final(desc, &walk); + err = blkcipher_walk_done(desc, &walk, 0); +diff --git a/arch/x86/crypto/glue_helper.c b/arch/x86/crypto/glue_helper.c +index 6a85598931b5..3a506ce7ed93 100644 +--- a/arch/x86/crypto/glue_helper.c ++++ b/arch/x86/crypto/glue_helper.c +@@ -39,7 +39,7 @@ static int __glue_ecb_crypt_128bit(const struct common_glue_ctx *gctx, + void *ctx = crypto_blkcipher_ctx(desc->tfm); + const unsigned int bsize = 128 / 8; + unsigned int nbytes, i, func_bytes; +- bool fpu_enabled = false; ++ bool fpu_enabled; + int err; + + err = blkcipher_walk_virt(desc, walk); +@@ -49,7 +49,7 @@ static int __glue_ecb_crypt_128bit(const struct common_glue_ctx *gctx, + u8 *wdst = walk->dst.virt.addr; + + fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, +- desc, fpu_enabled, nbytes); ++ desc, false, nbytes); + + for (i = 0; i < gctx->num_funcs; i++) { + func_bytes = bsize * gctx->funcs[i].num_blocks; +@@ -71,10 +71,10 @@ static int __glue_ecb_crypt_128bit(const struct common_glue_ctx *gctx, + } + + done: ++ glue_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, walk, nbytes); + } + +- glue_fpu_end(fpu_enabled); + return err; + } + +@@ -194,7 +194,7 @@ int glue_cbc_decrypt_128bit(const struct common_glue_ctx *gctx, + struct scatterlist *src, unsigned int nbytes) + { + const unsigned int bsize = 128 / 8; +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; + +@@ -203,12 +203,12 @@ int glue_cbc_decrypt_128bit(const struct common_glue_ctx *gctx, + + while ((nbytes = walk.nbytes)) { + fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, +- desc, fpu_enabled, nbytes); ++ desc, false, nbytes); + nbytes = __glue_cbc_decrypt_128bit(gctx, desc, &walk); ++ glue_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); + } + +- glue_fpu_end(fpu_enabled); + return err; + } + EXPORT_SYMBOL_GPL(glue_cbc_decrypt_128bit); +@@ -277,7 +277,7 @@ int glue_ctr_crypt_128bit(const struct common_glue_ctx *gctx, + struct scatterlist *src, unsigned int nbytes) + { + const unsigned int bsize = 128 / 8; +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; + +@@ -286,13 +286,12 @@ int glue_ctr_crypt_128bit(const struct common_glue_ctx *gctx, + + while ((nbytes = walk.nbytes) >= bsize) { + fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, +- desc, fpu_enabled, nbytes); ++ desc, false, nbytes); + nbytes = __glue_ctr_crypt_128bit(gctx, desc, &walk); ++ glue_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); + } + +- glue_fpu_end(fpu_enabled); +- + if (walk.nbytes) { + glue_ctr_crypt_final_128bit( + gctx->funcs[gctx->num_funcs - 1].fn_u.ctr, desc, &walk); +@@ -347,7 +346,7 @@ int glue_xts_crypt_128bit(const struct common_glue_ctx *gctx, + void *tweak_ctx, void *crypt_ctx) + { + const unsigned int bsize = 128 / 8; +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; + +@@ -360,21 +359,21 @@ int glue_xts_crypt_128bit(const struct common_glue_ctx *gctx, + + /* set minimum length to bsize, for tweak_fn */ + fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, +- desc, fpu_enabled, ++ desc, false, + nbytes < bsize ? bsize : nbytes); +- + /* calculate first value of T */ + tweak_fn(tweak_ctx, walk.iv, walk.iv); ++ glue_fpu_end(fpu_enabled); + + while (nbytes) { ++ fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, ++ desc, false, nbytes); + nbytes = __glue_xts_crypt_128bit(gctx, crypt_ctx, desc, &walk); + ++ glue_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); + nbytes = walk.nbytes; + } +- +- glue_fpu_end(fpu_enabled); +- + return err; + } + EXPORT_SYMBOL_GPL(glue_xts_crypt_128bit); +diff --git a/arch/x86/entry/common.c b/arch/x86/entry/common.c +index bdd9cc59d20f..56d01a339ba4 100644 +--- a/arch/x86/entry/common.c ++++ b/arch/x86/entry/common.c +@@ -129,7 +129,7 @@ static long syscall_trace_enter(struct pt_regs *regs) + + #define EXIT_TO_USERMODE_LOOP_FLAGS \ + (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME | _TIF_UPROBE | \ +- _TIF_NEED_RESCHED | _TIF_USER_RETURN_NOTIFY) ++ _TIF_NEED_RESCHED_MASK | _TIF_USER_RETURN_NOTIFY) + + static void exit_to_usermode_loop(struct pt_regs *regs, u32 cached_flags) + { +@@ -145,9 +145,16 @@ static void exit_to_usermode_loop(struct pt_regs *regs, u32 cached_flags) + /* We have work to do. */ + local_irq_enable(); + +- if (cached_flags & _TIF_NEED_RESCHED) ++ if (cached_flags & _TIF_NEED_RESCHED_MASK) + schedule(); + ++#ifdef ARCH_RT_DELAYS_SIGNAL_SEND ++ if (unlikely(current->forced_info.si_signo)) { ++ struct task_struct *t = current; ++ force_sig_info(t->forced_info.si_signo, &t->forced_info, t); ++ t->forced_info.si_signo = 0; ++ } ++#endif + if (cached_flags & _TIF_UPROBE) + uprobe_notify_resume(regs); + +diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S +index edba8606b99a..4a3389535fc6 100644 +--- a/arch/x86/entry/entry_32.S ++++ b/arch/x86/entry/entry_32.S +@@ -308,8 +308,25 @@ END(ret_from_exception) + ENTRY(resume_kernel) + DISABLE_INTERRUPTS(CLBR_ANY) + need_resched: ++ # preempt count == 0 + NEED_RS set? + cmpl $0, PER_CPU_VAR(__preempt_count) ++#ifndef CONFIG_PREEMPT_LAZY + jnz restore_all ++#else ++ jz test_int_off ++ ++ # atleast preempt count == 0 ? ++ cmpl $_PREEMPT_ENABLED,PER_CPU_VAR(__preempt_count) ++ jne restore_all ++ ++ movl PER_CPU_VAR(current_task), %ebp ++ cmpl $0,TASK_TI_preempt_lazy_count(%ebp) # non-zero preempt_lazy_count ? ++ jnz restore_all ++ ++ testl $_TIF_NEED_RESCHED_LAZY, TASK_TI_flags(%ebp) ++ jz restore_all ++test_int_off: ++#endif + testl $X86_EFLAGS_IF, PT_EFLAGS(%esp) # interrupts off (exception path) ? + jz restore_all + call preempt_schedule_irq +diff --git a/arch/x86/entry/entry_64.S b/arch/x86/entry/entry_64.S +index ef766a358b37..28401f826ab1 100644 +--- a/arch/x86/entry/entry_64.S ++++ b/arch/x86/entry/entry_64.S +@@ -546,7 +546,23 @@ GLOBAL(retint_user) + bt $9, EFLAGS(%rsp) /* were interrupts off? */ + jnc 1f + 0: cmpl $0, PER_CPU_VAR(__preempt_count) ++#ifndef CONFIG_PREEMPT_LAZY + jnz 1f ++#else ++ jz do_preempt_schedule_irq ++ ++ # atleast preempt count == 0 ? ++ cmpl $_PREEMPT_ENABLED,PER_CPU_VAR(__preempt_count) ++ jnz 1f ++ ++ movq PER_CPU_VAR(current_task), %rcx ++ cmpl $0, TASK_TI_preempt_lazy_count(%rcx) ++ jnz 1f ++ ++ bt $TIF_NEED_RESCHED_LAZY,TASK_TI_flags(%rcx) ++ jnc 1f ++do_preempt_schedule_irq: ++#endif + call preempt_schedule_irq + jmp 0b + 1: +@@ -894,6 +910,7 @@ EXPORT_SYMBOL(native_load_gs_index) + jmp 2b + .previous + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* Call softirq on interrupt stack. Interrupts are off. */ + ENTRY(do_softirq_own_stack) + pushq %rbp +@@ -906,6 +923,7 @@ ENTRY(do_softirq_own_stack) + decl PER_CPU_VAR(irq_count) + ret + END(do_softirq_own_stack) ++#endif + + #ifdef CONFIG_XEN + idtentry xen_hypervisor_callback xen_do_hypervisor_callback has_error_code=0 +diff --git a/arch/x86/include/asm/preempt.h b/arch/x86/include/asm/preempt.h +index 17f218645701..11bd1b7ee6eb 100644 +--- a/arch/x86/include/asm/preempt.h ++++ b/arch/x86/include/asm/preempt.h +@@ -79,17 +79,46 @@ static __always_inline void __preempt_count_sub(int val) + * a decrement which hits zero means we have no preempt_count and should + * reschedule. + */ +-static __always_inline bool __preempt_count_dec_and_test(void) ++static __always_inline bool ____preempt_count_dec_and_test(void) + { + GEN_UNARY_RMWcc("decl", __preempt_count, __percpu_arg(0), e); + } + ++static __always_inline bool __preempt_count_dec_and_test(void) ++{ ++ if (____preempt_count_dec_and_test()) ++ return true; ++#ifdef CONFIG_PREEMPT_LAZY ++ if (current_thread_info()->preempt_lazy_count) ++ return false; ++ return test_thread_flag(TIF_NEED_RESCHED_LAZY); ++#else ++ return false; ++#endif ++} ++ + /* + * Returns true when we need to resched and can (barring IRQ state). + */ + static __always_inline bool should_resched(int preempt_offset) + { ++#ifdef CONFIG_PREEMPT_LAZY ++ u32 tmp; ++ ++ tmp = raw_cpu_read_4(__preempt_count); ++ if (tmp == preempt_offset) ++ return true; ++ ++ /* preempt count == 0 ? */ ++ tmp &= ~PREEMPT_NEED_RESCHED; ++ if (tmp) ++ return false; ++ if (current_thread_info()->preempt_lazy_count) ++ return false; ++ return test_thread_flag(TIF_NEED_RESCHED_LAZY); ++#else + return unlikely(raw_cpu_read_4(__preempt_count) == preempt_offset); ++#endif + } + + #ifdef CONFIG_PREEMPT +diff --git a/arch/x86/include/asm/signal.h b/arch/x86/include/asm/signal.h +index 8af22be0fe61..d1328789b759 100644 +--- a/arch/x86/include/asm/signal.h ++++ b/arch/x86/include/asm/signal.h +@@ -27,6 +27,19 @@ typedef struct { + #define SA_IA32_ABI 0x02000000u + #define SA_X32_ABI 0x01000000u + ++/* ++ * Because some traps use the IST stack, we must keep preemption ++ * disabled while calling do_trap(), but do_trap() may call ++ * force_sig_info() which will grab the signal spin_locks for the ++ * task, which in PREEMPT_RT_FULL are mutexes. By defining ++ * ARCH_RT_DELAYS_SIGNAL_SEND the force_sig_info() will set ++ * TIF_NOTIFY_RESUME and set up the signal to be sent on exit of the ++ * trap. ++ */ ++#if defined(CONFIG_PREEMPT_RT_FULL) ++#define ARCH_RT_DELAYS_SIGNAL_SEND ++#endif ++ + #ifndef CONFIG_COMPAT + typedef sigset_t compat_sigset_t; + #endif +diff --git a/arch/x86/include/asm/stackprotector.h b/arch/x86/include/asm/stackprotector.h +index 58505f01962f..02fa39652cd6 100644 +--- a/arch/x86/include/asm/stackprotector.h ++++ b/arch/x86/include/asm/stackprotector.h +@@ -59,7 +59,7 @@ + */ + static __always_inline void boot_init_stack_canary(void) + { +- u64 canary; ++ u64 uninitialized_var(canary); + u64 tsc; + + #ifdef CONFIG_X86_64 +@@ -70,8 +70,15 @@ static __always_inline void boot_init_stack_canary(void) + * of randomness. The TSC only matters for very early init, + * there it already has some randomness on most systems. Later + * on during the bootup the random pool has true entropy too. ++ * ++ * For preempt-rt we need to weaken the randomness a bit, as ++ * we can't call into the random generator from atomic context ++ * due to locking constraints. We just leave canary ++ * uninitialized and use the TSC based randomness on top of it. + */ ++#ifndef CONFIG_PREEMPT_RT_FULL + get_random_bytes(&canary, sizeof(canary)); ++#endif + tsc = rdtsc(); + canary += tsc + (tsc << 32UL); + +diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h +index ad6f5eb07a95..5ceb3a1c2b1a 100644 +--- a/arch/x86/include/asm/thread_info.h ++++ b/arch/x86/include/asm/thread_info.h +@@ -54,11 +54,14 @@ struct task_struct; + + struct thread_info { + unsigned long flags; /* low level flags */ ++ int preempt_lazy_count; /* 0 => lazy preemptable ++ <0 => BUG */ + }; + + #define INIT_THREAD_INFO(tsk) \ + { \ + .flags = 0, \ ++ .preempt_lazy_count = 0, \ + } + + #define init_stack (init_thread_union.stack) +@@ -67,6 +70,10 @@ struct thread_info { + + #include + ++#define GET_THREAD_INFO(reg) \ ++ _ASM_MOV PER_CPU_VAR(cpu_current_top_of_stack),reg ; \ ++ _ASM_SUB $(THREAD_SIZE),reg ; ++ + #endif + + /* +@@ -85,6 +92,7 @@ struct thread_info { + #define TIF_SYSCALL_EMU 6 /* syscall emulation active */ + #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ + #define TIF_SECCOMP 8 /* secure computing */ ++#define TIF_NEED_RESCHED_LAZY 9 /* lazy rescheduling necessary */ + #define TIF_USER_RETURN_NOTIFY 11 /* notify kernel of userspace return */ + #define TIF_UPROBE 12 /* breakpointed or singlestepping */ + #define TIF_NOTSC 16 /* TSC is not accessible in userland */ +@@ -108,6 +116,7 @@ struct thread_info { + #define _TIF_SYSCALL_EMU (1 << TIF_SYSCALL_EMU) + #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) + #define _TIF_SECCOMP (1 << TIF_SECCOMP) ++#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) + #define _TIF_USER_RETURN_NOTIFY (1 << TIF_USER_RETURN_NOTIFY) + #define _TIF_UPROBE (1 << TIF_UPROBE) + #define _TIF_NOTSC (1 << TIF_NOTSC) +@@ -143,6 +152,8 @@ struct thread_info { + #define _TIF_WORK_CTXSW_PREV (_TIF_WORK_CTXSW|_TIF_USER_RETURN_NOTIFY) + #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW) + ++#define _TIF_NEED_RESCHED_MASK (_TIF_NEED_RESCHED | _TIF_NEED_RESCHED_LAZY) ++ + #define STACK_WARN (THREAD_SIZE/8) + + /* +diff --git a/arch/x86/include/asm/uv/uv_bau.h b/arch/x86/include/asm/uv/uv_bau.h +index 57ab86d94d64..35d25e27180f 100644 +--- a/arch/x86/include/asm/uv/uv_bau.h ++++ b/arch/x86/include/asm/uv/uv_bau.h +@@ -624,9 +624,9 @@ struct bau_control { + cycles_t send_message; + cycles_t period_end; + cycles_t period_time; +- spinlock_t uvhub_lock; +- spinlock_t queue_lock; +- spinlock_t disable_lock; ++ raw_spinlock_t uvhub_lock; ++ raw_spinlock_t queue_lock; ++ raw_spinlock_t disable_lock; + /* tunables */ + int max_concurr; + int max_concurr_const; +@@ -815,15 +815,15 @@ static inline int atom_asr(short i, struct atomic_short *v) + * to be lowered below the current 'v'. atomic_add_unless can only stop + * on equal. + */ +-static inline int atomic_inc_unless_ge(spinlock_t *lock, atomic_t *v, int u) ++static inline int atomic_inc_unless_ge(raw_spinlock_t *lock, atomic_t *v, int u) + { +- spin_lock(lock); ++ raw_spin_lock(lock); + if (atomic_read(v) >= u) { +- spin_unlock(lock); ++ raw_spin_unlock(lock); + return 0; + } + atomic_inc(v); +- spin_unlock(lock); ++ raw_spin_unlock(lock); + return 1; + } + +diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c +index 931ced8ca345..167975ac8af7 100644 +--- a/arch/x86/kernel/acpi/boot.c ++++ b/arch/x86/kernel/acpi/boot.c +@@ -87,7 +87,9 @@ static u64 acpi_lapic_addr __initdata = APIC_DEFAULT_PHYS_BASE; + * ->ioapic_mutex + * ->ioapic_lock + */ ++#ifdef CONFIG_X86_IO_APIC + static DEFINE_MUTEX(acpi_ioapic_lock); ++#endif + + /* -------------------------------------------------------------------------- + Boot-time Configuration +diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c +index 7249f1500bcb..a79d5c224004 100644 +--- a/arch/x86/kernel/apic/io_apic.c ++++ b/arch/x86/kernel/apic/io_apic.c +@@ -1712,7 +1712,8 @@ static bool io_apic_level_ack_pending(struct mp_chip_data *data) + static inline bool ioapic_irqd_mask(struct irq_data *data) + { + /* If we are moving the irq we need to mask it */ +- if (unlikely(irqd_is_setaffinity_pending(data))) { ++ if (unlikely(irqd_is_setaffinity_pending(data) && ++ !irqd_irq_inprogress(data))) { + mask_ioapic_irq(data); + return true; + } +diff --git a/arch/x86/kernel/asm-offsets.c b/arch/x86/kernel/asm-offsets.c +index c62e015b126c..0cc71257fca6 100644 +--- a/arch/x86/kernel/asm-offsets.c ++++ b/arch/x86/kernel/asm-offsets.c +@@ -36,6 +36,7 @@ void common(void) { + + BLANK(); + OFFSET(TASK_TI_flags, task_struct, thread_info.flags); ++ OFFSET(TASK_TI_preempt_lazy_count, task_struct, thread_info.preempt_lazy_count); + OFFSET(TASK_addr_limit, task_struct, thread.addr_limit); + + BLANK(); +@@ -91,4 +92,5 @@ void common(void) { + + BLANK(); + DEFINE(PTREGS_SIZE, sizeof(struct pt_regs)); ++ DEFINE(_PREEMPT_ENABLED, PREEMPT_ENABLED); + } +diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c +index 22cda29d654e..57c85e3af092 100644 +--- a/arch/x86/kernel/cpu/mcheck/mce.c ++++ b/arch/x86/kernel/cpu/mcheck/mce.c +@@ -41,6 +41,8 @@ + #include + #include + #include ++#include ++#include + #include + + #include +@@ -1307,7 +1309,7 @@ void mce_log_therm_throt_event(__u64 status) + static unsigned long check_interval = INITIAL_CHECK_INTERVAL; + + static DEFINE_PER_CPU(unsigned long, mce_next_interval); /* in jiffies */ +-static DEFINE_PER_CPU(struct timer_list, mce_timer); ++static DEFINE_PER_CPU(struct hrtimer, mce_timer); + + static unsigned long mce_adjust_timer_default(unsigned long interval) + { +@@ -1316,32 +1318,18 @@ static unsigned long mce_adjust_timer_default(unsigned long interval) + + static unsigned long (*mce_adjust_timer)(unsigned long interval) = mce_adjust_timer_default; + +-static void __restart_timer(struct timer_list *t, unsigned long interval) ++static enum hrtimer_restart __restart_timer(struct hrtimer *timer, unsigned long interval) + { +- unsigned long when = jiffies + interval; +- unsigned long flags; +- +- local_irq_save(flags); +- +- if (timer_pending(t)) { +- if (time_before(when, t->expires)) +- mod_timer(t, when); +- } else { +- t->expires = round_jiffies(when); +- add_timer_on(t, smp_processor_id()); +- } +- +- local_irq_restore(flags); ++ if (!interval) ++ return HRTIMER_NORESTART; ++ hrtimer_forward_now(timer, ns_to_ktime(jiffies_to_nsecs(interval))); ++ return HRTIMER_RESTART; + } + +-static void mce_timer_fn(unsigned long data) ++static enum hrtimer_restart mce_timer_fn(struct hrtimer *timer) + { +- struct timer_list *t = this_cpu_ptr(&mce_timer); +- int cpu = smp_processor_id(); + unsigned long iv; + +- WARN_ON(cpu != data); +- + iv = __this_cpu_read(mce_next_interval); + + if (mce_available(this_cpu_ptr(&cpu_info))) { +@@ -1364,7 +1352,7 @@ static void mce_timer_fn(unsigned long data) + + done: + __this_cpu_write(mce_next_interval, iv); +- __restart_timer(t, iv); ++ return __restart_timer(timer, iv); + } + + /* +@@ -1372,7 +1360,7 @@ static void mce_timer_fn(unsigned long data) + */ + void mce_timer_kick(unsigned long interval) + { +- struct timer_list *t = this_cpu_ptr(&mce_timer); ++ struct hrtimer *t = this_cpu_ptr(&mce_timer); + unsigned long iv = __this_cpu_read(mce_next_interval); + + __restart_timer(t, interval); +@@ -1387,7 +1375,7 @@ static void mce_timer_delete_all(void) + int cpu; + + for_each_online_cpu(cpu) +- del_timer_sync(&per_cpu(mce_timer, cpu)); ++ hrtimer_cancel(&per_cpu(mce_timer, cpu)); + } + + static void mce_do_trigger(struct work_struct *work) +@@ -1397,6 +1385,56 @@ static void mce_do_trigger(struct work_struct *work) + + static DECLARE_WORK(mce_trigger_work, mce_do_trigger); + ++static void __mce_notify_work(struct swork_event *event) ++{ ++ /* Not more than two messages every minute */ ++ static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2); ++ ++ /* wake processes polling /dev/mcelog */ ++ wake_up_interruptible(&mce_chrdev_wait); ++ ++ /* ++ * There is no risk of missing notifications because ++ * work_pending is always cleared before the function is ++ * executed. ++ */ ++ if (mce_helper[0] && !work_pending(&mce_trigger_work)) ++ schedule_work(&mce_trigger_work); ++ ++ if (__ratelimit(&ratelimit)) ++ pr_info(HW_ERR "Machine check events logged\n"); ++} ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++static bool notify_work_ready __read_mostly; ++static struct swork_event notify_work; ++ ++static int mce_notify_work_init(void) ++{ ++ int err; ++ ++ err = swork_get(); ++ if (err) ++ return err; ++ ++ INIT_SWORK(¬ify_work, __mce_notify_work); ++ notify_work_ready = true; ++ return 0; ++} ++ ++static void mce_notify_work(void) ++{ ++ if (notify_work_ready) ++ swork_queue(¬ify_work); ++} ++#else ++static void mce_notify_work(void) ++{ ++ __mce_notify_work(NULL); ++} ++static inline int mce_notify_work_init(void) { return 0; } ++#endif ++ + /* + * Notify the user(s) about new machine check events. + * Can be called from interrupt context, but not from machine check/NMI +@@ -1404,19 +1442,8 @@ static DECLARE_WORK(mce_trigger_work, mce_do_trigger); + */ + int mce_notify_irq(void) + { +- /* Not more than two messages every minute */ +- static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2); +- + if (test_and_clear_bit(0, &mce_need_notify)) { +- /* wake processes polling /dev/mcelog */ +- wake_up_interruptible(&mce_chrdev_wait); +- +- if (mce_helper[0]) +- schedule_work(&mce_trigger_work); +- +- if (__ratelimit(&ratelimit)) +- pr_info(HW_ERR "Machine check events logged\n"); +- ++ mce_notify_work(); + return 1; + } + return 0; +@@ -1722,7 +1749,7 @@ static void __mcheck_cpu_clear_vendor(struct cpuinfo_x86 *c) + } + } + +-static void mce_start_timer(unsigned int cpu, struct timer_list *t) ++static void mce_start_timer(unsigned int cpu, struct hrtimer *t) + { + unsigned long iv = check_interval * HZ; + +@@ -1731,16 +1758,17 @@ static void mce_start_timer(unsigned int cpu, struct timer_list *t) + + per_cpu(mce_next_interval, cpu) = iv; + +- t->expires = round_jiffies(jiffies + iv); +- add_timer_on(t, cpu); ++ hrtimer_start_range_ns(t, ns_to_ktime(jiffies_to_usecs(iv) * 1000ULL), ++ 0, HRTIMER_MODE_REL_PINNED); + } + + static void __mcheck_cpu_init_timer(void) + { +- struct timer_list *t = this_cpu_ptr(&mce_timer); ++ struct hrtimer *t = this_cpu_ptr(&mce_timer); + unsigned int cpu = smp_processor_id(); + +- setup_pinned_timer(t, mce_timer_fn, cpu); ++ hrtimer_init(t, CLOCK_MONOTONIC, HRTIMER_MODE_REL); ++ t->function = mce_timer_fn; + mce_start_timer(cpu, t); + } + +@@ -2465,6 +2493,8 @@ static void mce_disable_cpu(void *h) + if (!mce_available(raw_cpu_ptr(&cpu_info))) + return; + ++ hrtimer_cancel(this_cpu_ptr(&mce_timer)); ++ + if (!(action & CPU_TASKS_FROZEN)) + cmci_clear(); + +@@ -2487,6 +2517,7 @@ static void mce_reenable_cpu(void *h) + if (b->init) + wrmsrl(msr_ops.ctl(i), b->ctl); + } ++ __mcheck_cpu_init_timer(); + } + + /* Get notified when a cpu comes on/off. Be hotplug friendly. */ +@@ -2494,7 +2525,6 @@ static int + mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) + { + unsigned int cpu = (unsigned long)hcpu; +- struct timer_list *t = &per_cpu(mce_timer, cpu); + + switch (action & ~CPU_TASKS_FROZEN) { + case CPU_ONLINE: +@@ -2514,11 +2544,9 @@ mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) + break; + case CPU_DOWN_PREPARE: + smp_call_function_single(cpu, mce_disable_cpu, &action, 1); +- del_timer_sync(t); + break; + case CPU_DOWN_FAILED: + smp_call_function_single(cpu, mce_reenable_cpu, &action, 1); +- mce_start_timer(cpu, t); + break; + } + +@@ -2557,6 +2585,10 @@ static __init int mcheck_init_device(void) + goto err_out; + } + ++ err = mce_notify_work_init(); ++ if (err) ++ goto err_out; ++ + if (!zalloc_cpumask_var(&mce_device_initialized, GFP_KERNEL)) { + err = -ENOMEM; + goto err_out; +diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c +index 1f38d9a4d9de..053bf3b2ef39 100644 +--- a/arch/x86/kernel/irq_32.c ++++ b/arch/x86/kernel/irq_32.c +@@ -127,6 +127,7 @@ void irq_ctx_init(int cpu) + cpu, per_cpu(hardirq_stack, cpu), per_cpu(softirq_stack, cpu)); + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + void do_softirq_own_stack(void) + { + struct irq_stack *irqstk; +@@ -143,6 +144,7 @@ void do_softirq_own_stack(void) + + call_on_stack(__do_softirq, isp); + } ++#endif + + bool handle_irq(struct irq_desc *desc, struct pt_regs *regs) + { +diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c +index bd7be8efdc4c..b3b0a7f7b1ca 100644 +--- a/arch/x86/kernel/process_32.c ++++ b/arch/x86/kernel/process_32.c +@@ -35,6 +35,7 @@ + #include + #include + #include ++#include + + #include + #include +@@ -195,6 +196,35 @@ start_thread(struct pt_regs *regs, unsigned long new_ip, unsigned long new_sp) + } + EXPORT_SYMBOL_GPL(start_thread); + ++#ifdef CONFIG_PREEMPT_RT_FULL ++static void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) ++{ ++ int i; ++ ++ /* ++ * Clear @prev's kmap_atomic mappings ++ */ ++ for (i = 0; i < prev_p->kmap_idx; i++) { ++ int idx = i + KM_TYPE_NR * smp_processor_id(); ++ pte_t *ptep = kmap_pte - idx; ++ ++ kpte_clear_flush(ptep, __fix_to_virt(FIX_KMAP_BEGIN + idx)); ++ } ++ /* ++ * Restore @next_p's kmap_atomic mappings ++ */ ++ for (i = 0; i < next_p->kmap_idx; i++) { ++ int idx = i + KM_TYPE_NR * smp_processor_id(); ++ ++ if (!pte_none(next_p->kmap_pte[i])) ++ set_pte(kmap_pte - idx, next_p->kmap_pte[i]); ++ } ++} ++#else ++static inline void ++switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) { } ++#endif ++ + + /* + * switch_to(x,y) should switch tasks from x to y. +@@ -271,6 +301,8 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) + task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT)) + __switch_to_xtra(prev_p, next_p, tss); + ++ switch_kmaps(prev_p, next_p); ++ + /* + * Leave lazy mode, flushing any hypercalls made here. + * This must be done before restoring TLS segments so +diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c +index 3f05c044720b..fe68afd37162 100644 +--- a/arch/x86/kvm/lapic.c ++++ b/arch/x86/kvm/lapic.c +@@ -1939,6 +1939,7 @@ int kvm_create_lapic(struct kvm_vcpu *vcpu) + hrtimer_init(&apic->lapic_timer.timer, CLOCK_MONOTONIC, + HRTIMER_MODE_ABS_PINNED); + apic->lapic_timer.timer.function = apic_timer_fn; ++ apic->lapic_timer.timer.irqsafe = 1; + + /* + * APIC is created enabled. This will prevent kvm_lapic_set_base from +diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c +index 81bba3c2137d..fcb84512e85d 100644 +--- a/arch/x86/kvm/x86.c ++++ b/arch/x86/kvm/x86.c +@@ -5958,6 +5958,13 @@ int kvm_arch_init(void *opaque) + goto out; + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ if (!boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) { ++ printk(KERN_ERR "RT requires X86_FEATURE_CONSTANT_TSC\n"); ++ return -EOPNOTSUPP; ++ } ++#endif ++ + r = kvm_mmu_module_init(); + if (r) + goto out_free_percpu; +diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c +index 6d18b70ed5a9..f752724c22e8 100644 +--- a/arch/x86/mm/highmem_32.c ++++ b/arch/x86/mm/highmem_32.c +@@ -32,10 +32,11 @@ EXPORT_SYMBOL(kunmap); + */ + void *kmap_atomic_prot(struct page *page, pgprot_t prot) + { ++ pte_t pte = mk_pte(page, prot); + unsigned long vaddr; + int idx, type; + +- preempt_disable(); ++ preempt_disable_nort(); + pagefault_disable(); + + if (!PageHighMem(page)) +@@ -45,7 +46,10 @@ void *kmap_atomic_prot(struct page *page, pgprot_t prot) + idx = type + KM_TYPE_NR*smp_processor_id(); + vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); + BUG_ON(!pte_none(*(kmap_pte-idx))); +- set_pte(kmap_pte-idx, mk_pte(page, prot)); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = pte; ++#endif ++ set_pte(kmap_pte-idx, pte); + arch_flush_lazy_mmu_mode(); + + return (void *)vaddr; +@@ -88,6 +92,9 @@ void __kunmap_atomic(void *kvaddr) + * is a bad idea also, in case the page changes cacheability + * attributes or becomes a protected page in a hypervisor. + */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = __pte(0); ++#endif + kpte_clear_flush(kmap_pte-idx, vaddr); + kmap_atomic_idx_pop(); + arch_flush_lazy_mmu_mode(); +@@ -100,7 +107,7 @@ void __kunmap_atomic(void *kvaddr) + #endif + + pagefault_enable(); +- preempt_enable(); ++ preempt_enable_nort(); + } + EXPORT_SYMBOL(__kunmap_atomic); + +diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c +index ada98b39b8ad..585f6829653b 100644 +--- a/arch/x86/mm/iomap_32.c ++++ b/arch/x86/mm/iomap_32.c +@@ -56,6 +56,7 @@ EXPORT_SYMBOL_GPL(iomap_free); + + void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot) + { ++ pte_t pte = pfn_pte(pfn, prot); + unsigned long vaddr; + int idx, type; + +@@ -65,7 +66,12 @@ void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot) + type = kmap_atomic_idx_push(); + idx = type + KM_TYPE_NR * smp_processor_id(); + vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); +- set_pte(kmap_pte - idx, pfn_pte(pfn, prot)); ++ WARN_ON(!pte_none(*(kmap_pte - idx))); ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = pte; ++#endif ++ set_pte(kmap_pte - idx, pte); + arch_flush_lazy_mmu_mode(); + + return (void *)vaddr; +@@ -113,6 +119,9 @@ iounmap_atomic(void __iomem *kvaddr) + * is a bad idea also, in case the page changes cacheability + * attributes or becomes a protected page in a hypervisor. + */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = __pte(0); ++#endif + kpte_clear_flush(kmap_pte-idx, vaddr); + kmap_atomic_idx_pop(); + } +diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c +index e3353c97d086..01664968555c 100644 +--- a/arch/x86/mm/pageattr.c ++++ b/arch/x86/mm/pageattr.c +@@ -214,7 +214,15 @@ static void cpa_flush_array(unsigned long *start, int numpages, int cache, + int in_flags, struct page **pages) + { + unsigned int i, level; ++#ifdef CONFIG_PREEMPT ++ /* ++ * Avoid wbinvd() because it causes latencies on all CPUs, ++ * regardless of any CPU isolation that may be in effect. ++ */ ++ unsigned long do_wbinvd = 0; ++#else + unsigned long do_wbinvd = cache && numpages >= 1024; /* 4M threshold */ ++#endif + + BUG_ON(irqs_disabled()); + +diff --git a/arch/x86/platform/uv/tlb_uv.c b/arch/x86/platform/uv/tlb_uv.c +index 9e42842e924a..5398f97172f9 100644 +--- a/arch/x86/platform/uv/tlb_uv.c ++++ b/arch/x86/platform/uv/tlb_uv.c +@@ -748,9 +748,9 @@ static void destination_plugged(struct bau_desc *bau_desc, + + quiesce_local_uvhub(hmaster); + +- spin_lock(&hmaster->queue_lock); ++ raw_spin_lock(&hmaster->queue_lock); + reset_with_ipi(&bau_desc->distribution, bcp); +- spin_unlock(&hmaster->queue_lock); ++ raw_spin_unlock(&hmaster->queue_lock); + + end_uvhub_quiesce(hmaster); + +@@ -770,9 +770,9 @@ static void destination_timeout(struct bau_desc *bau_desc, + + quiesce_local_uvhub(hmaster); + +- spin_lock(&hmaster->queue_lock); ++ raw_spin_lock(&hmaster->queue_lock); + reset_with_ipi(&bau_desc->distribution, bcp); +- spin_unlock(&hmaster->queue_lock); ++ raw_spin_unlock(&hmaster->queue_lock); + + end_uvhub_quiesce(hmaster); + +@@ -793,7 +793,7 @@ static void disable_for_period(struct bau_control *bcp, struct ptc_stats *stat) + cycles_t tm1; + + hmaster = bcp->uvhub_master; +- spin_lock(&hmaster->disable_lock); ++ raw_spin_lock(&hmaster->disable_lock); + if (!bcp->baudisabled) { + stat->s_bau_disabled++; + tm1 = get_cycles(); +@@ -806,7 +806,7 @@ static void disable_for_period(struct bau_control *bcp, struct ptc_stats *stat) + } + } + } +- spin_unlock(&hmaster->disable_lock); ++ raw_spin_unlock(&hmaster->disable_lock); + } + + static void count_max_concurr(int stat, struct bau_control *bcp, +@@ -869,7 +869,7 @@ static void record_send_stats(cycles_t time1, cycles_t time2, + */ + static void uv1_throttle(struct bau_control *hmaster, struct ptc_stats *stat) + { +- spinlock_t *lock = &hmaster->uvhub_lock; ++ raw_spinlock_t *lock = &hmaster->uvhub_lock; + atomic_t *v; + + v = &hmaster->active_descriptor_count; +@@ -1002,7 +1002,7 @@ static int check_enable(struct bau_control *bcp, struct ptc_stats *stat) + struct bau_control *hmaster; + + hmaster = bcp->uvhub_master; +- spin_lock(&hmaster->disable_lock); ++ raw_spin_lock(&hmaster->disable_lock); + if (bcp->baudisabled && (get_cycles() >= bcp->set_bau_on_time)) { + stat->s_bau_reenabled++; + for_each_present_cpu(tcpu) { +@@ -1014,10 +1014,10 @@ static int check_enable(struct bau_control *bcp, struct ptc_stats *stat) + tbcp->period_giveups = 0; + } + } +- spin_unlock(&hmaster->disable_lock); ++ raw_spin_unlock(&hmaster->disable_lock); + return 0; + } +- spin_unlock(&hmaster->disable_lock); ++ raw_spin_unlock(&hmaster->disable_lock); + return -1; + } + +@@ -1940,9 +1940,9 @@ static void __init init_per_cpu_tunables(void) + bcp->cong_reps = congested_reps; + bcp->disabled_period = sec_2_cycles(disabled_period); + bcp->giveup_limit = giveup_limit; +- spin_lock_init(&bcp->queue_lock); +- spin_lock_init(&bcp->uvhub_lock); +- spin_lock_init(&bcp->disable_lock); ++ raw_spin_lock_init(&bcp->queue_lock); ++ raw_spin_lock_init(&bcp->uvhub_lock); ++ raw_spin_lock_init(&bcp->disable_lock); + } + } + +diff --git a/arch/x86/platform/uv/uv_time.c b/arch/x86/platform/uv/uv_time.c +index b333fc45f9ec..8b85916e6986 100644 +--- a/arch/x86/platform/uv/uv_time.c ++++ b/arch/x86/platform/uv/uv_time.c +@@ -57,7 +57,7 @@ static DEFINE_PER_CPU(struct clock_event_device, cpu_ced); + + /* There is one of these allocated per node */ + struct uv_rtc_timer_head { +- spinlock_t lock; ++ raw_spinlock_t lock; + /* next cpu waiting for timer, local node relative: */ + int next_cpu; + /* number of cpus on this node: */ +@@ -177,7 +177,7 @@ static __init int uv_rtc_allocate_timers(void) + uv_rtc_deallocate_timers(); + return -ENOMEM; + } +- spin_lock_init(&head->lock); ++ raw_spin_lock_init(&head->lock); + head->ncpus = uv_blade_nr_possible_cpus(bid); + head->next_cpu = -1; + blade_info[bid] = head; +@@ -231,7 +231,7 @@ static int uv_rtc_set_timer(int cpu, u64 expires) + unsigned long flags; + int next_cpu; + +- spin_lock_irqsave(&head->lock, flags); ++ raw_spin_lock_irqsave(&head->lock, flags); + + next_cpu = head->next_cpu; + *t = expires; +@@ -243,12 +243,12 @@ static int uv_rtc_set_timer(int cpu, u64 expires) + if (uv_setup_intr(cpu, expires)) { + *t = ULLONG_MAX; + uv_rtc_find_next_timer(head, pnode); +- spin_unlock_irqrestore(&head->lock, flags); ++ raw_spin_unlock_irqrestore(&head->lock, flags); + return -ETIME; + } + } + +- spin_unlock_irqrestore(&head->lock, flags); ++ raw_spin_unlock_irqrestore(&head->lock, flags); + return 0; + } + +@@ -267,7 +267,7 @@ static int uv_rtc_unset_timer(int cpu, int force) + unsigned long flags; + int rc = 0; + +- spin_lock_irqsave(&head->lock, flags); ++ raw_spin_lock_irqsave(&head->lock, flags); + + if ((head->next_cpu == bcpu && uv_read_rtc(NULL) >= *t) || force) + rc = 1; +@@ -279,7 +279,7 @@ static int uv_rtc_unset_timer(int cpu, int force) + uv_rtc_find_next_timer(head, pnode); + } + +- spin_unlock_irqrestore(&head->lock, flags); ++ raw_spin_unlock_irqrestore(&head->lock, flags); + + return rc; + } +@@ -299,13 +299,18 @@ static int uv_rtc_unset_timer(int cpu, int force) + static cycle_t uv_read_rtc(struct clocksource *cs) + { + unsigned long offset; ++ cycle_t cycles; + ++ preempt_disable(); + if (uv_get_min_hub_revision_id() == 1) + offset = 0; + else + offset = (uv_blade_processor_id() * L1_CACHE_BYTES) % PAGE_SIZE; + +- return (cycle_t)uv_read_local_mmr(UVH_RTC | offset); ++ cycles = (cycle_t)uv_read_local_mmr(UVH_RTC | offset); ++ preempt_enable(); ++ ++ return cycles; + } + + /* +diff --git a/block/blk-core.c b/block/blk-core.c +index d1f2801ce836..6f945bb0fa1a 100644 +--- a/block/blk-core.c ++++ b/block/blk-core.c +@@ -125,6 +125,9 @@ void blk_rq_init(struct request_queue *q, struct request *rq) + + INIT_LIST_HEAD(&rq->queuelist); + INIT_LIST_HEAD(&rq->timeout_list); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ INIT_WORK(&rq->work, __blk_mq_complete_request_remote_work); ++#endif + rq->cpu = -1; + rq->q = q; + rq->__sector = (sector_t) -1; +@@ -233,7 +236,7 @@ EXPORT_SYMBOL(blk_start_queue_async); + **/ + void blk_start_queue(struct request_queue *q) + { +- WARN_ON(!irqs_disabled()); ++ WARN_ON_NONRT(!irqs_disabled()); + + queue_flag_clear(QUEUE_FLAG_STOPPED, q); + __blk_run_queue(q); +@@ -659,7 +662,7 @@ int blk_queue_enter(struct request_queue *q, bool nowait) + if (nowait) + return -EBUSY; + +- ret = wait_event_interruptible(q->mq_freeze_wq, ++ ret = swait_event_interruptible(q->mq_freeze_wq, + !atomic_read(&q->mq_freeze_depth) || + blk_queue_dying(q)); + if (blk_queue_dying(q)) +@@ -679,7 +682,7 @@ static void blk_queue_usage_counter_release(struct percpu_ref *ref) + struct request_queue *q = + container_of(ref, struct request_queue, q_usage_counter); + +- wake_up_all(&q->mq_freeze_wq); ++ swake_up_all(&q->mq_freeze_wq); + } + + static void blk_rq_timed_out_timer(unsigned long data) +@@ -748,7 +751,7 @@ struct request_queue *blk_alloc_queue_node(gfp_t gfp_mask, int node_id) + q->bypass_depth = 1; + __set_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); + +- init_waitqueue_head(&q->mq_freeze_wq); ++ init_swait_queue_head(&q->mq_freeze_wq); + + /* + * Init percpu_ref in atomic mode so that it's faster to shutdown. +@@ -3200,7 +3203,7 @@ static void queue_unplugged(struct request_queue *q, unsigned int depth, + blk_run_queue_async(q); + else + __blk_run_queue(q); +- spin_unlock(q->queue_lock); ++ spin_unlock_irq(q->queue_lock); + } + + static void flush_plug_callbacks(struct blk_plug *plug, bool from_schedule) +@@ -3248,7 +3251,6 @@ EXPORT_SYMBOL(blk_check_plugged); + void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) + { + struct request_queue *q; +- unsigned long flags; + struct request *rq; + LIST_HEAD(list); + unsigned int depth; +@@ -3268,11 +3270,6 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) + q = NULL; + depth = 0; + +- /* +- * Save and disable interrupts here, to avoid doing it for every +- * queue lock we have to take. +- */ +- local_irq_save(flags); + while (!list_empty(&list)) { + rq = list_entry_rq(list.next); + list_del_init(&rq->queuelist); +@@ -3285,7 +3282,7 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) + queue_unplugged(q, depth, from_schedule); + q = rq->q; + depth = 0; +- spin_lock(q->queue_lock); ++ spin_lock_irq(q->queue_lock); + } + + /* +@@ -3312,8 +3309,6 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) + */ + if (q) + queue_unplugged(q, depth, from_schedule); +- +- local_irq_restore(flags); + } + + void blk_finish_plug(struct blk_plug *plug) +diff --git a/block/blk-ioc.c b/block/blk-ioc.c +index 381cb50a673c..dc8785233d94 100644 +--- a/block/blk-ioc.c ++++ b/block/blk-ioc.c +@@ -7,6 +7,7 @@ + #include + #include + #include ++#include + + #include "blk.h" + +@@ -109,7 +110,7 @@ static void ioc_release_fn(struct work_struct *work) + spin_unlock(q->queue_lock); + } else { + spin_unlock_irqrestore(&ioc->lock, flags); +- cpu_relax(); ++ cpu_chill(); + spin_lock_irqsave_nested(&ioc->lock, flags, 1); + } + } +@@ -187,7 +188,7 @@ void put_io_context_active(struct io_context *ioc) + spin_unlock(icq->q->queue_lock); + } else { + spin_unlock_irqrestore(&ioc->lock, flags); +- cpu_relax(); ++ cpu_chill(); + goto retry; + } + } +diff --git a/block/blk-mq.c b/block/blk-mq.c +index 7b597ec4e9c5..48c9652a701c 100644 +--- a/block/blk-mq.c ++++ b/block/blk-mq.c +@@ -72,7 +72,7 @@ EXPORT_SYMBOL_GPL(blk_mq_freeze_queue_start); + + static void blk_mq_freeze_queue_wait(struct request_queue *q) + { +- wait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->q_usage_counter)); ++ swait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->q_usage_counter)); + } + + /* +@@ -110,7 +110,7 @@ void blk_mq_unfreeze_queue(struct request_queue *q) + WARN_ON_ONCE(freeze_depth < 0); + if (!freeze_depth) { + percpu_ref_reinit(&q->q_usage_counter); +- wake_up_all(&q->mq_freeze_wq); ++ swake_up_all(&q->mq_freeze_wq); + } + } + EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue); +@@ -129,7 +129,7 @@ void blk_mq_wake_waiters(struct request_queue *q) + * dying, we need to ensure that processes currently waiting on + * the queue are notified as well. + */ +- wake_up_all(&q->mq_freeze_wq); ++ swake_up_all(&q->mq_freeze_wq); + } + + bool blk_mq_can_queue(struct blk_mq_hw_ctx *hctx) +@@ -177,6 +177,9 @@ static void blk_mq_rq_ctx_init(struct request_queue *q, struct blk_mq_ctx *ctx, + rq->resid_len = 0; + rq->sense = NULL; + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ INIT_WORK(&rq->work, __blk_mq_complete_request_remote_work); ++#endif + INIT_LIST_HEAD(&rq->timeout_list); + rq->timeout = 0; + +@@ -345,6 +348,17 @@ void blk_mq_end_request(struct request *rq, int error) + } + EXPORT_SYMBOL(blk_mq_end_request); + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ ++void __blk_mq_complete_request_remote_work(struct work_struct *work) ++{ ++ struct request *rq = container_of(work, struct request, work); ++ ++ rq->q->softirq_done_fn(rq); ++} ++ ++#else ++ + static void __blk_mq_complete_request_remote(void *data) + { + struct request *rq = data; +@@ -352,6 +366,8 @@ static void __blk_mq_complete_request_remote(void *data) + rq->q->softirq_done_fn(rq); + } + ++#endif ++ + static void blk_mq_ipi_complete_request(struct request *rq) + { + struct blk_mq_ctx *ctx = rq->mq_ctx; +@@ -363,19 +379,23 @@ static void blk_mq_ipi_complete_request(struct request *rq) + return; + } + +- cpu = get_cpu(); ++ cpu = get_cpu_light(); + if (!test_bit(QUEUE_FLAG_SAME_FORCE, &rq->q->queue_flags)) + shared = cpus_share_cache(cpu, ctx->cpu); + + if (cpu != ctx->cpu && !shared && cpu_online(ctx->cpu)) { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ schedule_work_on(ctx->cpu, &rq->work); ++#else + rq->csd.func = __blk_mq_complete_request_remote; + rq->csd.info = rq; + rq->csd.flags = 0; + smp_call_function_single_async(ctx->cpu, &rq->csd); ++#endif + } else { + rq->q->softirq_done_fn(rq); + } +- put_cpu(); ++ put_cpu_light(); + } + + static void __blk_mq_complete_request(struct request *rq) +@@ -906,14 +926,14 @@ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async) + return; + + if (!async && !(hctx->flags & BLK_MQ_F_BLOCKING)) { +- int cpu = get_cpu(); ++ int cpu = get_cpu_light(); + if (cpumask_test_cpu(cpu, hctx->cpumask)) { + __blk_mq_run_hw_queue(hctx); +- put_cpu(); ++ put_cpu_light(); + return; + } + +- put_cpu(); ++ put_cpu_light(); + } + + kblockd_schedule_work_on(blk_mq_hctx_next_cpu(hctx), &hctx->run_work); +diff --git a/block/blk-mq.h b/block/blk-mq.h +index e5d25249028c..1e846b842eab 100644 +--- a/block/blk-mq.h ++++ b/block/blk-mq.h +@@ -72,12 +72,12 @@ static inline struct blk_mq_ctx *__blk_mq_get_ctx(struct request_queue *q, + */ + static inline struct blk_mq_ctx *blk_mq_get_ctx(struct request_queue *q) + { +- return __blk_mq_get_ctx(q, get_cpu()); ++ return __blk_mq_get_ctx(q, get_cpu_light()); + } + + static inline void blk_mq_put_ctx(struct blk_mq_ctx *ctx) + { +- put_cpu(); ++ put_cpu_light(); + } + + struct blk_mq_alloc_data { +diff --git a/block/blk-softirq.c b/block/blk-softirq.c +index 06cf9807f49a..c40342643ca0 100644 +--- a/block/blk-softirq.c ++++ b/block/blk-softirq.c +@@ -51,6 +51,7 @@ static void trigger_softirq(void *data) + raise_softirq_irqoff(BLOCK_SOFTIRQ); + + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + + /* +@@ -89,6 +90,7 @@ static int blk_softirq_cpu_dead(unsigned int cpu) + this_cpu_ptr(&blk_cpu_done)); + raise_softirq_irqoff(BLOCK_SOFTIRQ); + local_irq_enable(); ++ preempt_check_resched_rt(); + + return 0; + } +@@ -141,6 +143,7 @@ void __blk_complete_request(struct request *req) + goto do_local; + + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + + /** +diff --git a/block/bounce.c b/block/bounce.c +index 1cb5dd3a5da1..2f1ec8a67cbe 100644 +--- a/block/bounce.c ++++ b/block/bounce.c +@@ -55,11 +55,11 @@ static void bounce_copy_vec(struct bio_vec *to, unsigned char *vfrom) + unsigned long flags; + unsigned char *vto; + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + vto = kmap_atomic(to->bv_page); + memcpy(vto + to->bv_offset, vfrom, to->bv_len); + kunmap_atomic(vto); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + + #else /* CONFIG_HIGHMEM */ +diff --git a/crypto/algapi.c b/crypto/algapi.c +index 1fad2a6b3bbb..ecb7315426a9 100644 +--- a/crypto/algapi.c ++++ b/crypto/algapi.c +@@ -719,13 +719,13 @@ EXPORT_SYMBOL_GPL(crypto_spawn_tfm2); + + int crypto_register_notifier(struct notifier_block *nb) + { +- return blocking_notifier_chain_register(&crypto_chain, nb); ++ return srcu_notifier_chain_register(&crypto_chain, nb); + } + EXPORT_SYMBOL_GPL(crypto_register_notifier); + + int crypto_unregister_notifier(struct notifier_block *nb) + { +- return blocking_notifier_chain_unregister(&crypto_chain, nb); ++ return srcu_notifier_chain_unregister(&crypto_chain, nb); + } + EXPORT_SYMBOL_GPL(crypto_unregister_notifier); + +diff --git a/crypto/api.c b/crypto/api.c +index bbc147cb5dec..bc1a848f02ec 100644 +--- a/crypto/api.c ++++ b/crypto/api.c +@@ -31,7 +31,7 @@ EXPORT_SYMBOL_GPL(crypto_alg_list); + DECLARE_RWSEM(crypto_alg_sem); + EXPORT_SYMBOL_GPL(crypto_alg_sem); + +-BLOCKING_NOTIFIER_HEAD(crypto_chain); ++SRCU_NOTIFIER_HEAD(crypto_chain); + EXPORT_SYMBOL_GPL(crypto_chain); + + static struct crypto_alg *crypto_larval_wait(struct crypto_alg *alg); +@@ -236,10 +236,10 @@ int crypto_probing_notify(unsigned long val, void *v) + { + int ok; + +- ok = blocking_notifier_call_chain(&crypto_chain, val, v); ++ ok = srcu_notifier_call_chain(&crypto_chain, val, v); + if (ok == NOTIFY_DONE) { + request_module("cryptomgr"); +- ok = blocking_notifier_call_chain(&crypto_chain, val, v); ++ ok = srcu_notifier_call_chain(&crypto_chain, val, v); + } + + return ok; +diff --git a/crypto/internal.h b/crypto/internal.h +index 7eefcdb00227..0ecc7f5a2f40 100644 +--- a/crypto/internal.h ++++ b/crypto/internal.h +@@ -47,7 +47,7 @@ struct crypto_larval { + + extern struct list_head crypto_alg_list; + extern struct rw_semaphore crypto_alg_sem; +-extern struct blocking_notifier_head crypto_chain; ++extern struct srcu_notifier_head crypto_chain; + + #ifdef CONFIG_PROC_FS + void __init crypto_init_proc(void); +@@ -146,7 +146,7 @@ static inline int crypto_is_moribund(struct crypto_alg *alg) + + static inline void crypto_notify(unsigned long val, void *v) + { +- blocking_notifier_call_chain(&crypto_chain, val, v); ++ srcu_notifier_call_chain(&crypto_chain, val, v); + } + + #endif /* _CRYPTO_INTERNAL_H */ +diff --git a/drivers/acpi/acpica/acglobal.h b/drivers/acpi/acpica/acglobal.h +index 750fa824d42c..441edf51484a 100644 +--- a/drivers/acpi/acpica/acglobal.h ++++ b/drivers/acpi/acpica/acglobal.h +@@ -116,7 +116,7 @@ ACPI_GLOBAL(u8, acpi_gbl_global_lock_pending); + * interrupt level + */ + ACPI_GLOBAL(acpi_spinlock, acpi_gbl_gpe_lock); /* For GPE data structs and registers */ +-ACPI_GLOBAL(acpi_spinlock, acpi_gbl_hardware_lock); /* For ACPI H/W except GPE registers */ ++ACPI_GLOBAL(acpi_raw_spinlock, acpi_gbl_hardware_lock); /* For ACPI H/W except GPE registers */ + ACPI_GLOBAL(acpi_spinlock, acpi_gbl_reference_count_lock); + + /* Mutex for _OSI support */ +diff --git a/drivers/acpi/acpica/hwregs.c b/drivers/acpi/acpica/hwregs.c +index 3b7fb99362b6..696bf8e62afb 100644 +--- a/drivers/acpi/acpica/hwregs.c ++++ b/drivers/acpi/acpica/hwregs.c +@@ -363,14 +363,14 @@ acpi_status acpi_hw_clear_acpi_status(void) + ACPI_BITMASK_ALL_FIXED_STATUS, + ACPI_FORMAT_UINT64(acpi_gbl_xpm1a_status.address))); + +- lock_flags = acpi_os_acquire_lock(acpi_gbl_hardware_lock); ++ raw_spin_lock_irqsave(acpi_gbl_hardware_lock, lock_flags); + + /* Clear the fixed events in PM1 A/B */ + + status = acpi_hw_register_write(ACPI_REGISTER_PM1_STATUS, + ACPI_BITMASK_ALL_FIXED_STATUS); + +- acpi_os_release_lock(acpi_gbl_hardware_lock, lock_flags); ++ raw_spin_unlock_irqrestore(acpi_gbl_hardware_lock, lock_flags); + + if (ACPI_FAILURE(status)) { + goto exit; +diff --git a/drivers/acpi/acpica/hwxface.c b/drivers/acpi/acpica/hwxface.c +index 98c26ff39409..6e236f2ea791 100644 +--- a/drivers/acpi/acpica/hwxface.c ++++ b/drivers/acpi/acpica/hwxface.c +@@ -373,7 +373,7 @@ acpi_status acpi_write_bit_register(u32 register_id, u32 value) + return_ACPI_STATUS(AE_BAD_PARAMETER); + } + +- lock_flags = acpi_os_acquire_lock(acpi_gbl_hardware_lock); ++ raw_spin_lock_irqsave(acpi_gbl_hardware_lock, lock_flags); + + /* + * At this point, we know that the parent register is one of the +@@ -434,7 +434,7 @@ acpi_status acpi_write_bit_register(u32 register_id, u32 value) + + unlock_and_exit: + +- acpi_os_release_lock(acpi_gbl_hardware_lock, lock_flags); ++ raw_spin_unlock_irqrestore(acpi_gbl_hardware_lock, lock_flags); + return_ACPI_STATUS(status); + } + +diff --git a/drivers/acpi/acpica/utmutex.c b/drivers/acpi/acpica/utmutex.c +index 15073375bd00..357e7ca5a587 100644 +--- a/drivers/acpi/acpica/utmutex.c ++++ b/drivers/acpi/acpica/utmutex.c +@@ -88,7 +88,7 @@ acpi_status acpi_ut_mutex_initialize(void) + return_ACPI_STATUS (status); + } + +- status = acpi_os_create_lock (&acpi_gbl_hardware_lock); ++ status = acpi_os_create_raw_lock (&acpi_gbl_hardware_lock); + if (ACPI_FAILURE (status)) { + return_ACPI_STATUS (status); + } +@@ -145,7 +145,7 @@ void acpi_ut_mutex_terminate(void) + /* Delete the spinlocks */ + + acpi_os_delete_lock(acpi_gbl_gpe_lock); +- acpi_os_delete_lock(acpi_gbl_hardware_lock); ++ acpi_os_delete_raw_lock(acpi_gbl_hardware_lock); + acpi_os_delete_lock(acpi_gbl_reference_count_lock); + + /* Delete the reader/writer lock */ +diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c +index 051b6158d1b7..7ad293bef6ed 100644 +--- a/drivers/ata/libata-sff.c ++++ b/drivers/ata/libata-sff.c +@@ -678,9 +678,9 @@ unsigned int ata_sff_data_xfer_noirq(struct ata_device *dev, unsigned char *buf, + unsigned long flags; + unsigned int consumed; + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + consumed = ata_sff_data_xfer32(dev, buf, buflen, rw); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + return consumed; + } +@@ -719,7 +719,7 @@ static void ata_pio_sector(struct ata_queued_cmd *qc) + unsigned long flags; + + /* FIXME: use a bounce buffer */ +- local_irq_save(flags); ++ local_irq_save_nort(flags); + buf = kmap_atomic(page); + + /* do the actual data transfer */ +@@ -727,7 +727,7 @@ static void ata_pio_sector(struct ata_queued_cmd *qc) + do_write); + + kunmap_atomic(buf); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } else { + buf = page_address(page); + ap->ops->sff_data_xfer(qc->dev, buf + offset, qc->sect_size, +@@ -864,7 +864,7 @@ static int __atapi_pio_bytes(struct ata_queued_cmd *qc, unsigned int bytes) + unsigned long flags; + + /* FIXME: use bounce buffer */ +- local_irq_save(flags); ++ local_irq_save_nort(flags); + buf = kmap_atomic(page); + + /* do the actual data transfer */ +@@ -872,7 +872,7 @@ static int __atapi_pio_bytes(struct ata_queued_cmd *qc, unsigned int bytes) + count, rw); + + kunmap_atomic(buf); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } else { + buf = page_address(page); + consumed = ap->ops->sff_data_xfer(dev, buf + offset, +diff --git a/drivers/block/zram/zcomp.c b/drivers/block/zram/zcomp.c +index 4b5cd3a7b2b6..fa8329ad79fd 100644 +--- a/drivers/block/zram/zcomp.c ++++ b/drivers/block/zram/zcomp.c +@@ -118,12 +118,19 @@ ssize_t zcomp_available_show(const char *comp, char *buf) + + struct zcomp_strm *zcomp_stream_get(struct zcomp *comp) + { +- return *get_cpu_ptr(comp->stream); ++ struct zcomp_strm *zstrm; ++ ++ zstrm = *this_cpu_ptr(comp->stream); ++ spin_lock(&zstrm->zcomp_lock); ++ return zstrm; + } + + void zcomp_stream_put(struct zcomp *comp) + { +- put_cpu_ptr(comp->stream); ++ struct zcomp_strm *zstrm; ++ ++ zstrm = *this_cpu_ptr(comp->stream); ++ spin_unlock(&zstrm->zcomp_lock); + } + + int zcomp_compress(struct zcomp_strm *zstrm, +@@ -174,6 +181,7 @@ static int __zcomp_cpu_notifier(struct zcomp *comp, + pr_err("Can't allocate a compression stream\n"); + return NOTIFY_BAD; + } ++ spin_lock_init(&zstrm->zcomp_lock); + *per_cpu_ptr(comp->stream, cpu) = zstrm; + break; + case CPU_DEAD: +diff --git a/drivers/block/zram/zcomp.h b/drivers/block/zram/zcomp.h +index 478cac2ed465..f7a6efdc3285 100644 +--- a/drivers/block/zram/zcomp.h ++++ b/drivers/block/zram/zcomp.h +@@ -14,6 +14,7 @@ struct zcomp_strm { + /* compression/decompression buffer */ + void *buffer; + struct crypto_comp *tfm; ++ spinlock_t zcomp_lock; + }; + + /* dynamic per-device compression frontend */ +diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c +index c9914d653968..2038d138f286 100644 +--- a/drivers/block/zram/zram_drv.c ++++ b/drivers/block/zram/zram_drv.c +@@ -528,6 +528,8 @@ static struct zram_meta *zram_meta_alloc(char *pool_name, u64 disksize) + goto out_error; + } + ++ zram_meta_init_table_locks(meta, disksize); ++ + return meta; + + out_error: +@@ -575,28 +577,28 @@ static int zram_decompress_page(struct zram *zram, char *mem, u32 index) + struct zram_meta *meta = zram->meta; + unsigned long handle; + unsigned int size; ++ struct zcomp_strm *zstrm; + +- bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value); ++ zram_lock_table(&meta->table[index]); + handle = meta->table[index].handle; + size = zram_get_obj_size(meta, index); + + if (!handle || zram_test_flag(meta, index, ZRAM_ZERO)) { +- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); ++ zram_unlock_table(&meta->table[index]); + memset(mem, 0, PAGE_SIZE); + return 0; + } + ++ zstrm = zcomp_stream_get(zram->comp); + cmem = zs_map_object(meta->mem_pool, handle, ZS_MM_RO); + if (size == PAGE_SIZE) { + memcpy(mem, cmem, PAGE_SIZE); + } else { +- struct zcomp_strm *zstrm = zcomp_stream_get(zram->comp); +- + ret = zcomp_decompress(zstrm, cmem, size, mem); +- zcomp_stream_put(zram->comp); + } + zs_unmap_object(meta->mem_pool, handle); +- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); ++ zcomp_stream_put(zram->comp); ++ zram_unlock_table(&meta->table[index]); + + /* Should NEVER happen. Return bio error if it does. */ + if (unlikely(ret)) { +@@ -616,14 +618,14 @@ static int zram_bvec_read(struct zram *zram, struct bio_vec *bvec, + struct zram_meta *meta = zram->meta; + page = bvec->bv_page; + +- bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value); ++ zram_lock_table(&meta->table[index]); + if (unlikely(!meta->table[index].handle) || + zram_test_flag(meta, index, ZRAM_ZERO)) { +- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); ++ zram_unlock_table(&meta->table[index]); + handle_zero_page(bvec); + return 0; + } +- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); ++ zram_unlock_table(&meta->table[index]); + + if (is_partial_io(bvec)) + /* Use a temporary buffer to decompress the page */ +@@ -700,10 +702,10 @@ static int zram_bvec_write(struct zram *zram, struct bio_vec *bvec, u32 index, + if (user_mem) + kunmap_atomic(user_mem); + /* Free memory associated with this sector now. */ +- bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value); ++ zram_lock_table(&meta->table[index]); + zram_free_page(zram, index); + zram_set_flag(meta, index, ZRAM_ZERO); +- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); ++ zram_unlock_table(&meta->table[index]); + + atomic64_inc(&zram->stats.zero_pages); + ret = 0; +@@ -794,12 +796,12 @@ static int zram_bvec_write(struct zram *zram, struct bio_vec *bvec, u32 index, + * Free memory associated with this sector + * before overwriting unused sectors. + */ +- bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value); ++ zram_lock_table(&meta->table[index]); + zram_free_page(zram, index); + + meta->table[index].handle = handle; + zram_set_obj_size(meta, index, clen); +- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); ++ zram_unlock_table(&meta->table[index]); + + /* Update stats */ + atomic64_add(clen, &zram->stats.compr_data_size); +@@ -842,9 +844,9 @@ static void zram_bio_discard(struct zram *zram, u32 index, + } + + while (n >= PAGE_SIZE) { +- bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value); ++ zram_lock_table(&meta->table[index]); + zram_free_page(zram, index); +- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); ++ zram_unlock_table(&meta->table[index]); + atomic64_inc(&zram->stats.notify_free); + index++; + n -= PAGE_SIZE; +@@ -973,9 +975,9 @@ static void zram_slot_free_notify(struct block_device *bdev, + zram = bdev->bd_disk->private_data; + meta = zram->meta; + +- bit_spin_lock(ZRAM_ACCESS, &meta->table[index].value); ++ zram_lock_table(&meta->table[index]); + zram_free_page(zram, index); +- bit_spin_unlock(ZRAM_ACCESS, &meta->table[index].value); ++ zram_unlock_table(&meta->table[index]); + atomic64_inc(&zram->stats.notify_free); + } + +diff --git a/drivers/block/zram/zram_drv.h b/drivers/block/zram/zram_drv.h +index 74fcf10da374..fd4020c99b9e 100644 +--- a/drivers/block/zram/zram_drv.h ++++ b/drivers/block/zram/zram_drv.h +@@ -73,6 +73,9 @@ enum zram_pageflags { + struct zram_table_entry { + unsigned long handle; + unsigned long value; ++#ifdef CONFIG_PREEMPT_RT_BASE ++ spinlock_t lock; ++#endif + }; + + struct zram_stats { +@@ -120,4 +123,42 @@ struct zram { + */ + bool claim; /* Protected by bdev->bd_mutex */ + }; ++ ++#ifndef CONFIG_PREEMPT_RT_BASE ++static inline void zram_lock_table(struct zram_table_entry *table) ++{ ++ bit_spin_lock(ZRAM_ACCESS, &table->value); ++} ++ ++static inline void zram_unlock_table(struct zram_table_entry *table) ++{ ++ bit_spin_unlock(ZRAM_ACCESS, &table->value); ++} ++ ++static inline void zram_meta_init_table_locks(struct zram_meta *meta, u64 disksize) { } ++#else /* CONFIG_PREEMPT_RT_BASE */ ++static inline void zram_lock_table(struct zram_table_entry *table) ++{ ++ spin_lock(&table->lock); ++ __set_bit(ZRAM_ACCESS, &table->value); ++} ++ ++static inline void zram_unlock_table(struct zram_table_entry *table) ++{ ++ __clear_bit(ZRAM_ACCESS, &table->value); ++ spin_unlock(&table->lock); ++} ++ ++static inline void zram_meta_init_table_locks(struct zram_meta *meta, u64 disksize) ++{ ++ size_t num_pages = disksize >> PAGE_SHIFT; ++ size_t index; ++ ++ for (index = 0; index < num_pages; index++) { ++ spinlock_t *lock = &meta->table[index].lock; ++ spin_lock_init(lock); ++ } ++} ++#endif /* CONFIG_PREEMPT_RT_BASE */ ++ + #endif +diff --git a/drivers/char/random.c b/drivers/char/random.c +index 08d1dd58c0d2..25ee319dc8e3 100644 +--- a/drivers/char/random.c ++++ b/drivers/char/random.c +@@ -262,6 +262,7 @@ + #include + #include + #include ++#include + #include + + #include +@@ -1028,8 +1029,6 @@ static void add_timer_randomness(struct timer_rand_state *state, unsigned num) + } sample; + long delta, delta2, delta3; + +- preempt_disable(); +- + sample.jiffies = jiffies; + sample.cycles = random_get_entropy(); + sample.num = num; +@@ -1070,7 +1069,6 @@ static void add_timer_randomness(struct timer_rand_state *state, unsigned num) + */ + credit_entropy_bits(r, min_t(int, fls(delta>>1), 11)); + } +- preempt_enable(); + } + + void add_input_randomness(unsigned int type, unsigned int code, +@@ -1123,28 +1121,27 @@ static __u32 get_reg(struct fast_pool *f, struct pt_regs *regs) + return *(ptr + f->reg_idx++); + } + +-void add_interrupt_randomness(int irq, int irq_flags) ++void add_interrupt_randomness(int irq, int irq_flags, __u64 ip) + { + struct entropy_store *r; + struct fast_pool *fast_pool = this_cpu_ptr(&irq_randomness); +- struct pt_regs *regs = get_irq_regs(); + unsigned long now = jiffies; + cycles_t cycles = random_get_entropy(); + __u32 c_high, j_high; +- __u64 ip; + unsigned long seed; + int credit = 0; + + if (cycles == 0) +- cycles = get_reg(fast_pool, regs); ++ cycles = get_reg(fast_pool, NULL); + c_high = (sizeof(cycles) > 4) ? cycles >> 32 : 0; + j_high = (sizeof(now) > 4) ? now >> 32 : 0; + fast_pool->pool[0] ^= cycles ^ j_high ^ irq; + fast_pool->pool[1] ^= now ^ c_high; +- ip = regs ? instruction_pointer(regs) : _RET_IP_; ++ if (!ip) ++ ip = _RET_IP_; + fast_pool->pool[2] ^= ip; + fast_pool->pool[3] ^= (sizeof(ip) > 4) ? ip >> 32 : +- get_reg(fast_pool, regs); ++ get_reg(fast_pool, NULL); + + fast_mix(fast_pool); + add_interrupt_bench(cycles); +@@ -2056,6 +2053,7 @@ struct batched_entropy { + * goal of being quite fast and not depleting entropy. + */ + static DEFINE_PER_CPU(struct batched_entropy, batched_entropy_long); ++static DEFINE_LOCAL_IRQ_LOCK(batched_entropy_long_lock); + unsigned long get_random_long(void) + { + unsigned long ret; +@@ -2064,13 +2062,13 @@ unsigned long get_random_long(void) + if (arch_get_random_long(&ret)) + return ret; + +- batch = &get_cpu_var(batched_entropy_long); ++ batch = &get_locked_var(batched_entropy_long_lock, batched_entropy_long); + if (batch->position % ARRAY_SIZE(batch->entropy_long) == 0) { + extract_crng((u8 *)batch->entropy_long); + batch->position = 0; + } + ret = batch->entropy_long[batch->position++]; +- put_cpu_var(batched_entropy_long); ++ put_locked_var(batched_entropy_long_lock, batched_entropy_long); + return ret; + } + EXPORT_SYMBOL(get_random_long); +@@ -2082,6 +2080,8 @@ unsigned int get_random_int(void) + } + #else + static DEFINE_PER_CPU(struct batched_entropy, batched_entropy_int); ++static DEFINE_LOCAL_IRQ_LOCK(batched_entropy_int_lock); ++ + unsigned int get_random_int(void) + { + unsigned int ret; +@@ -2090,13 +2090,13 @@ unsigned int get_random_int(void) + if (arch_get_random_int(&ret)) + return ret; + +- batch = &get_cpu_var(batched_entropy_int); ++ batch = &get_locked_var(batched_entropy_int_lock, batched_entropy_int); + if (batch->position % ARRAY_SIZE(batch->entropy_int) == 0) { + extract_crng((u8 *)batch->entropy_int); + batch->position = 0; + } + ret = batch->entropy_int[batch->position++]; +- put_cpu_var(batched_entropy_int); ++ put_locked_var(batched_entropy_int_lock, batched_entropy_int); + return ret; + } + #endif +diff --git a/drivers/clocksource/tcb_clksrc.c b/drivers/clocksource/tcb_clksrc.c +index 4da2af9694a2..5b6f57f500b8 100644 +--- a/drivers/clocksource/tcb_clksrc.c ++++ b/drivers/clocksource/tcb_clksrc.c +@@ -23,8 +23,7 @@ + * this 32 bit free-running counter. the second channel is not used. + * + * - The third channel may be used to provide a 16-bit clockevent +- * source, used in either periodic or oneshot mode. This runs +- * at 32 KiHZ, and can handle delays of up to two seconds. ++ * source, used in either periodic or oneshot mode. + * + * A boot clocksource and clockevent source are also currently needed, + * unless the relevant platforms (ARM/AT91, AVR32/AT32) are changed so +@@ -74,6 +73,8 @@ static struct clocksource clksrc = { + struct tc_clkevt_device { + struct clock_event_device clkevt; + struct clk *clk; ++ bool clk_enabled; ++ u32 freq; + void __iomem *regs; + }; + +@@ -82,15 +83,26 @@ static struct tc_clkevt_device *to_tc_clkevt(struct clock_event_device *clkevt) + return container_of(clkevt, struct tc_clkevt_device, clkevt); + } + +-/* For now, we always use the 32K clock ... this optimizes for NO_HZ, +- * because using one of the divided clocks would usually mean the +- * tick rate can never be less than several dozen Hz (vs 0.5 Hz). +- * +- * A divided clock could be good for high resolution timers, since +- * 30.5 usec resolution can seem "low". +- */ + static u32 timer_clock; + ++static void tc_clk_disable(struct clock_event_device *d) ++{ ++ struct tc_clkevt_device *tcd = to_tc_clkevt(d); ++ ++ clk_disable(tcd->clk); ++ tcd->clk_enabled = false; ++} ++ ++static void tc_clk_enable(struct clock_event_device *d) ++{ ++ struct tc_clkevt_device *tcd = to_tc_clkevt(d); ++ ++ if (tcd->clk_enabled) ++ return; ++ clk_enable(tcd->clk); ++ tcd->clk_enabled = true; ++} ++ + static int tc_shutdown(struct clock_event_device *d) + { + struct tc_clkevt_device *tcd = to_tc_clkevt(d); +@@ -98,8 +110,14 @@ static int tc_shutdown(struct clock_event_device *d) + + __raw_writel(0xff, regs + ATMEL_TC_REG(2, IDR)); + __raw_writel(ATMEL_TC_CLKDIS, regs + ATMEL_TC_REG(2, CCR)); ++ return 0; ++} ++ ++static int tc_shutdown_clk_off(struct clock_event_device *d) ++{ ++ tc_shutdown(d); + if (!clockevent_state_detached(d)) +- clk_disable(tcd->clk); ++ tc_clk_disable(d); + + return 0; + } +@@ -112,9 +130,9 @@ static int tc_set_oneshot(struct clock_event_device *d) + if (clockevent_state_oneshot(d) || clockevent_state_periodic(d)) + tc_shutdown(d); + +- clk_enable(tcd->clk); ++ tc_clk_enable(d); + +- /* slow clock, count up to RC, then irq and stop */ ++ /* count up to RC, then irq and stop */ + __raw_writel(timer_clock | ATMEL_TC_CPCSTOP | ATMEL_TC_WAVE | + ATMEL_TC_WAVESEL_UP_AUTO, regs + ATMEL_TC_REG(2, CMR)); + __raw_writel(ATMEL_TC_CPCS, regs + ATMEL_TC_REG(2, IER)); +@@ -134,12 +152,12 @@ static int tc_set_periodic(struct clock_event_device *d) + /* By not making the gentime core emulate periodic mode on top + * of oneshot, we get lower overhead and improved accuracy. + */ +- clk_enable(tcd->clk); ++ tc_clk_enable(d); + +- /* slow clock, count up to RC, then irq and restart */ ++ /* count up to RC, then irq and restart */ + __raw_writel(timer_clock | ATMEL_TC_WAVE | ATMEL_TC_WAVESEL_UP_AUTO, + regs + ATMEL_TC_REG(2, CMR)); +- __raw_writel((32768 + HZ / 2) / HZ, tcaddr + ATMEL_TC_REG(2, RC)); ++ __raw_writel((tcd->freq + HZ / 2) / HZ, tcaddr + ATMEL_TC_REG(2, RC)); + + /* Enable clock and interrupts on RC compare */ + __raw_writel(ATMEL_TC_CPCS, regs + ATMEL_TC_REG(2, IER)); +@@ -166,9 +184,13 @@ static struct tc_clkevt_device clkevt = { + .features = CLOCK_EVT_FEAT_PERIODIC | + CLOCK_EVT_FEAT_ONESHOT, + /* Should be lower than at91rm9200's system timer */ ++#ifdef CONFIG_ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK + .rating = 125, ++#else ++ .rating = 200, ++#endif + .set_next_event = tc_next_event, +- .set_state_shutdown = tc_shutdown, ++ .set_state_shutdown = tc_shutdown_clk_off, + .set_state_periodic = tc_set_periodic, + .set_state_oneshot = tc_set_oneshot, + }, +@@ -188,8 +210,9 @@ static irqreturn_t ch2_irq(int irq, void *handle) + return IRQ_NONE; + } + +-static int __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx) ++static int __init setup_clkevents(struct atmel_tc *tc, int divisor_idx) + { ++ unsigned divisor = atmel_tc_divisors[divisor_idx]; + int ret; + struct clk *t2_clk = tc->clk[2]; + int irq = tc->irq[2]; +@@ -210,7 +233,11 @@ static int __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx) + clkevt.regs = tc->regs; + clkevt.clk = t2_clk; + +- timer_clock = clk32k_divisor_idx; ++ timer_clock = divisor_idx; ++ if (!divisor) ++ clkevt.freq = 32768; ++ else ++ clkevt.freq = clk_get_rate(t2_clk) / divisor; + + clkevt.clkevt.cpumask = cpumask_of(0); + +@@ -221,7 +248,7 @@ static int __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx) + return ret; + } + +- clockevents_config_and_register(&clkevt.clkevt, 32768, 1, 0xffff); ++ clockevents_config_and_register(&clkevt.clkevt, clkevt.freq, 1, 0xffff); + + return ret; + } +@@ -358,7 +385,11 @@ static int __init tcb_clksrc_init(void) + goto err_disable_t1; + + /* channel 2: periodic and oneshot timer support */ ++#ifdef CONFIG_ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK + ret = setup_clkevents(tc, clk32k_divisor_idx); ++#else ++ ret = setup_clkevents(tc, best_divisor_idx); ++#endif + if (ret) + goto err_unregister_clksrc; + +diff --git a/drivers/clocksource/timer-atmel-pit.c b/drivers/clocksource/timer-atmel-pit.c +index 6555821bbdae..93288849b2bd 100644 +--- a/drivers/clocksource/timer-atmel-pit.c ++++ b/drivers/clocksource/timer-atmel-pit.c +@@ -46,6 +46,7 @@ struct pit_data { + u32 cycle; + u32 cnt; + unsigned int irq; ++ bool irq_requested; + struct clk *mck; + }; + +@@ -96,15 +97,29 @@ static int pit_clkevt_shutdown(struct clock_event_device *dev) + + /* disable irq, leaving the clocksource active */ + pit_write(data->base, AT91_PIT_MR, (data->cycle - 1) | AT91_PIT_PITEN); ++ if (data->irq_requested) { ++ free_irq(data->irq, data); ++ data->irq_requested = false; ++ } + return 0; + } + ++static irqreturn_t at91sam926x_pit_interrupt(int irq, void *dev_id); + /* + * Clockevent device: interrupts every 1/HZ (== pit_cycles * MCK/16) + */ + static int pit_clkevt_set_periodic(struct clock_event_device *dev) + { + struct pit_data *data = clkevt_to_pit_data(dev); ++ int ret; ++ ++ ret = request_irq(data->irq, at91sam926x_pit_interrupt, ++ IRQF_SHARED | IRQF_TIMER | IRQF_IRQPOLL, ++ "at91_tick", data); ++ if (ret) ++ panic(pr_fmt("Unable to setup IRQ\n")); ++ ++ data->irq_requested = true; + + /* update clocksource counter */ + data->cnt += data->cycle * PIT_PICNT(pit_read(data->base, AT91_PIT_PIVR)); +@@ -230,15 +245,6 @@ static int __init at91sam926x_pit_dt_init(struct device_node *node) + return ret; + } + +- /* Set up irq handler */ +- ret = request_irq(data->irq, at91sam926x_pit_interrupt, +- IRQF_SHARED | IRQF_TIMER | IRQF_IRQPOLL, +- "at91_tick", data); +- if (ret) { +- pr_err("Unable to setup IRQ\n"); +- return ret; +- } +- + /* Set up and register clockevents */ + data->clkevt.name = "pit"; + data->clkevt.features = CLOCK_EVT_FEAT_PERIODIC; +diff --git a/drivers/clocksource/timer-atmel-st.c b/drivers/clocksource/timer-atmel-st.c +index e90ab5b63a90..9e124087c55f 100644 +--- a/drivers/clocksource/timer-atmel-st.c ++++ b/drivers/clocksource/timer-atmel-st.c +@@ -115,18 +115,29 @@ static void clkdev32k_disable_and_flush_irq(void) + last_crtr = read_CRTR(); + } + ++static int atmel_st_irq; ++ + static int clkevt32k_shutdown(struct clock_event_device *evt) + { + clkdev32k_disable_and_flush_irq(); + irqmask = 0; + regmap_write(regmap_st, AT91_ST_IER, irqmask); ++ free_irq(atmel_st_irq, regmap_st); + return 0; + } + + static int clkevt32k_set_oneshot(struct clock_event_device *dev) + { ++ int ret; ++ + clkdev32k_disable_and_flush_irq(); + ++ ret = request_irq(atmel_st_irq, at91rm9200_timer_interrupt, ++ IRQF_SHARED | IRQF_TIMER | IRQF_IRQPOLL, ++ "at91_tick", regmap_st); ++ if (ret) ++ panic(pr_fmt("Unable to setup IRQ\n")); ++ + /* + * ALM for oneshot irqs, set by next_event() + * before 32 seconds have passed. +@@ -139,8 +150,16 @@ static int clkevt32k_set_oneshot(struct clock_event_device *dev) + + static int clkevt32k_set_periodic(struct clock_event_device *dev) + { ++ int ret; ++ + clkdev32k_disable_and_flush_irq(); + ++ ret = request_irq(atmel_st_irq, at91rm9200_timer_interrupt, ++ IRQF_SHARED | IRQF_TIMER | IRQF_IRQPOLL, ++ "at91_tick", regmap_st); ++ if (ret) ++ panic(pr_fmt("Unable to setup IRQ\n")); ++ + /* PIT for periodic irqs; fixed rate of 1/HZ */ + irqmask = AT91_ST_PITS; + regmap_write(regmap_st, AT91_ST_PIMR, timer_latch); +@@ -198,7 +217,7 @@ static int __init atmel_st_timer_init(struct device_node *node) + { + struct clk *sclk; + unsigned int sclk_rate, val; +- int irq, ret; ++ int ret; + + regmap_st = syscon_node_to_regmap(node); + if (IS_ERR(regmap_st)) { +@@ -212,21 +231,12 @@ static int __init atmel_st_timer_init(struct device_node *node) + regmap_read(regmap_st, AT91_ST_SR, &val); + + /* Get the interrupts property */ +- irq = irq_of_parse_and_map(node, 0); +- if (!irq) { ++ atmel_st_irq = irq_of_parse_and_map(node, 0); ++ if (!atmel_st_irq) { + pr_err("Unable to get IRQ from DT\n"); + return -EINVAL; + } + +- /* Make IRQs happen for the system timer */ +- ret = request_irq(irq, at91rm9200_timer_interrupt, +- IRQF_SHARED | IRQF_TIMER | IRQF_IRQPOLL, +- "at91_tick", regmap_st); +- if (ret) { +- pr_err("Unable to setup IRQ\n"); +- return ret; +- } +- + sclk = of_clk_get(node, 0); + if (IS_ERR(sclk)) { + pr_err("Unable to get slow clock\n"); +diff --git a/drivers/connector/cn_proc.c b/drivers/connector/cn_proc.c +index a782ce87715c..19d265948526 100644 +--- a/drivers/connector/cn_proc.c ++++ b/drivers/connector/cn_proc.c +@@ -32,6 +32,7 @@ + #include + + #include ++#include + + /* + * Size of a cn_msg followed by a proc_event structure. Since the +@@ -54,10 +55,11 @@ static struct cb_id cn_proc_event_id = { CN_IDX_PROC, CN_VAL_PROC }; + + /* proc_event_counts is used as the sequence number of the netlink message */ + static DEFINE_PER_CPU(__u32, proc_event_counts) = { 0 }; ++static DEFINE_LOCAL_IRQ_LOCK(send_msg_lock); + + static inline void send_msg(struct cn_msg *msg) + { +- preempt_disable(); ++ local_lock(send_msg_lock); + + msg->seq = __this_cpu_inc_return(proc_event_counts) - 1; + ((struct proc_event *)msg->data)->cpu = smp_processor_id(); +@@ -70,7 +72,7 @@ static inline void send_msg(struct cn_msg *msg) + */ + cn_netlink_send(msg, 0, CN_IDX_PROC, GFP_NOWAIT); + +- preempt_enable(); ++ local_unlock(send_msg_lock); + } + + void proc_fork_connector(struct task_struct *task) +diff --git a/drivers/cpufreq/Kconfig.x86 b/drivers/cpufreq/Kconfig.x86 +index adbd1de1cea5..1fac5074f2cf 100644 +--- a/drivers/cpufreq/Kconfig.x86 ++++ b/drivers/cpufreq/Kconfig.x86 +@@ -124,7 +124,7 @@ config X86_POWERNOW_K7_ACPI + + config X86_POWERNOW_K8 + tristate "AMD Opteron/Athlon64 PowerNow!" +- depends on ACPI && ACPI_PROCESSOR && X86_ACPI_CPUFREQ ++ depends on ACPI && ACPI_PROCESSOR && X86_ACPI_CPUFREQ && !PREEMPT_RT_BASE + help + This adds the CPUFreq driver for K8/early Opteron/Athlon64 processors. + Support for K10 and newer processors is now in acpi-cpufreq. +diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c +index 2117f172d7a2..96c15501b0c8 100644 +--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c ++++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c +@@ -1489,7 +1489,9 @@ execbuf_submit(struct i915_execbuffer_params *params, + if (ret) + return ret; + ++#ifndef CONFIG_PREEMPT_RT_BASE + trace_i915_gem_ring_dispatch(params->request, params->dispatch_flags); ++#endif + + i915_gem_execbuffer_move_to_active(vmas, params->request); + +diff --git a/drivers/gpu/drm/i915/i915_gem_shrinker.c b/drivers/gpu/drm/i915/i915_gem_shrinker.c +index 755d78832a66..97fb03dc4971 100644 +--- a/drivers/gpu/drm/i915/i915_gem_shrinker.c ++++ b/drivers/gpu/drm/i915/i915_gem_shrinker.c +@@ -40,7 +40,7 @@ static bool mutex_is_locked_by(struct mutex *mutex, struct task_struct *task) + if (!mutex_is_locked(mutex)) + return false; + +-#if defined(CONFIG_DEBUG_MUTEXES) || defined(CONFIG_MUTEX_SPIN_ON_OWNER) ++#if (defined(CONFIG_DEBUG_MUTEXES) || defined(CONFIG_MUTEX_SPIN_ON_OWNER)) && !defined(CONFIG_PREEMPT_RT_BASE) + return mutex->owner == task; + #else + /* Since UP may be pre-empted, we cannot assume that we own the lock */ +diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c +index 02908e37c228..05c0480576e1 100644 +--- a/drivers/gpu/drm/i915/i915_irq.c ++++ b/drivers/gpu/drm/i915/i915_irq.c +@@ -812,6 +812,7 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, + spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); + + /* preempt_disable_rt() should go right here in PREEMPT_RT patchset. */ ++ preempt_disable_rt(); + + /* Get optional system timestamp before query. */ + if (stime) +@@ -863,6 +864,7 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, + *etime = ktime_get(); + + /* preempt_enable_rt() should go right here in PREEMPT_RT patchset. */ ++ preempt_enable_rt(); + + spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); + +diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c +index 5dc6082639db..c32458fb3be2 100644 +--- a/drivers/gpu/drm/i915/intel_display.c ++++ b/drivers/gpu/drm/i915/intel_display.c +@@ -12131,7 +12131,7 @@ void intel_check_page_flip(struct drm_i915_private *dev_priv, int pipe) + struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + struct intel_flip_work *work; + +- WARN_ON(!in_interrupt()); ++ WARN_ON_NONRT(!in_interrupt()); + + if (crtc == NULL) + return; +diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c +index dbed12c484c9..5c540b78e8b5 100644 +--- a/drivers/gpu/drm/i915/intel_sprite.c ++++ b/drivers/gpu/drm/i915/intel_sprite.c +@@ -35,6 +35,7 @@ + #include + #include + #include ++#include + #include "intel_drv.h" + #include "intel_frontbuffer.h" + #include +@@ -65,6 +66,8 @@ int intel_usecs_to_scanlines(const struct drm_display_mode *adjusted_mode, + 1000 * adjusted_mode->crtc_htotal); + } + ++static DEFINE_LOCAL_IRQ_LOCK(pipe_update_lock); ++ + /** + * intel_pipe_update_start() - start update of a set of display registers + * @crtc: the crtc of which the registers are going to be updated +@@ -95,7 +98,7 @@ void intel_pipe_update_start(struct intel_crtc *crtc) + min = vblank_start - intel_usecs_to_scanlines(adjusted_mode, 100); + max = vblank_start - 1; + +- local_irq_disable(); ++ local_lock_irq(pipe_update_lock); + + if (min <= 0 || max <= 0) + return; +@@ -125,11 +128,11 @@ void intel_pipe_update_start(struct intel_crtc *crtc) + break; + } + +- local_irq_enable(); ++ local_unlock_irq(pipe_update_lock); + + timeout = schedule_timeout(timeout); + +- local_irq_disable(); ++ local_lock_irq(pipe_update_lock); + } + + finish_wait(wq, &wait); +@@ -181,7 +184,7 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work + crtc->base.state->event = NULL; + } + +- local_irq_enable(); ++ local_unlock_irq(pipe_update_lock); + + if (crtc->debug.start_vbl_count && + crtc->debug.start_vbl_count != end_vbl_count) { +diff --git a/drivers/gpu/drm/msm/msm_gem_shrinker.c b/drivers/gpu/drm/msm/msm_gem_shrinker.c +index 192b2d3a79cb..d5372a207326 100644 +--- a/drivers/gpu/drm/msm/msm_gem_shrinker.c ++++ b/drivers/gpu/drm/msm/msm_gem_shrinker.c +@@ -23,7 +23,7 @@ static bool mutex_is_locked_by(struct mutex *mutex, struct task_struct *task) + if (!mutex_is_locked(mutex)) + return false; + +-#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_MUTEXES) ++#if (defined(CONFIG_SMP) || defined(CONFIG_DEBUG_MUTEXES)) && !defined(CONFIG_PREEMPT_RT_BASE) + return mutex->owner == task; + #else + /* Since UP may be pre-empted, we cannot assume that we own the lock */ +diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c +index cdb8cb568c15..b6d7fd964cbc 100644 +--- a/drivers/gpu/drm/radeon/radeon_display.c ++++ b/drivers/gpu/drm/radeon/radeon_display.c +@@ -1845,6 +1845,7 @@ int radeon_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, + struct radeon_device *rdev = dev->dev_private; + + /* preempt_disable_rt() should go right here in PREEMPT_RT patchset. */ ++ preempt_disable_rt(); + + /* Get optional system timestamp before query. */ + if (stime) +@@ -1937,6 +1938,7 @@ int radeon_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, + *etime = ktime_get(); + + /* preempt_enable_rt() should go right here in PREEMPT_RT patchset. */ ++ preempt_enable_rt(); + + /* Decode into vertical and horizontal scanout position. */ + *vpos = position & 0x1fff; +diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c +index 0276d2ef06ee..8868045eabde 100644 +--- a/drivers/hv/vmbus_drv.c ++++ b/drivers/hv/vmbus_drv.c +@@ -761,6 +761,8 @@ static void vmbus_isr(void) + void *page_addr; + struct hv_message *msg; + union hv_synic_event_flags *event; ++ struct pt_regs *regs = get_irq_regs(); ++ u64 ip = regs ? instruction_pointer(regs) : 0; + bool handled = false; + + page_addr = hv_context.synic_event_page[cpu]; +@@ -808,7 +810,7 @@ static void vmbus_isr(void) + tasklet_schedule(hv_context.msg_dpc[cpu]); + } + +- add_interrupt_randomness(HYPERVISOR_CALLBACK_VECTOR, 0); ++ add_interrupt_randomness(HYPERVISOR_CALLBACK_VECTOR, 0, ip); + } + + +diff --git a/drivers/ide/alim15x3.c b/drivers/ide/alim15x3.c +index 36f76e28a0bf..394f142f90c7 100644 +--- a/drivers/ide/alim15x3.c ++++ b/drivers/ide/alim15x3.c +@@ -234,7 +234,7 @@ static int init_chipset_ali15x3(struct pci_dev *dev) + + isa_dev = pci_get_device(PCI_VENDOR_ID_AL, PCI_DEVICE_ID_AL_M1533, NULL); + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + + if (m5229_revision < 0xC2) { + /* +@@ -325,7 +325,7 @@ static int init_chipset_ali15x3(struct pci_dev *dev) + } + pci_dev_put(north); + pci_dev_put(isa_dev); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + return 0; + } + +diff --git a/drivers/ide/hpt366.c b/drivers/ide/hpt366.c +index 0ceae5cbd89a..c212e85d7f3e 100644 +--- a/drivers/ide/hpt366.c ++++ b/drivers/ide/hpt366.c +@@ -1236,7 +1236,7 @@ static int init_dma_hpt366(ide_hwif_t *hwif, + + dma_old = inb(base + 2); + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + + dma_new = dma_old; + pci_read_config_byte(dev, hwif->channel ? 0x4b : 0x43, &masterdma); +@@ -1247,7 +1247,7 @@ static int init_dma_hpt366(ide_hwif_t *hwif, + if (dma_new != dma_old) + outb(dma_new, base + 2); + +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + printk(KERN_INFO " %s: BM-DMA at 0x%04lx-0x%04lx\n", + hwif->name, base, base + 7); +diff --git a/drivers/ide/ide-io-std.c b/drivers/ide/ide-io-std.c +index 19763977568c..4169433faab5 100644 +--- a/drivers/ide/ide-io-std.c ++++ b/drivers/ide/ide-io-std.c +@@ -175,7 +175,7 @@ void ide_input_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf, + unsigned long uninitialized_var(flags); + + if ((io_32bit & 2) && !mmio) { +- local_irq_save(flags); ++ local_irq_save_nort(flags); + ata_vlb_sync(io_ports->nsect_addr); + } + +@@ -186,7 +186,7 @@ void ide_input_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf, + insl(data_addr, buf, words); + + if ((io_32bit & 2) && !mmio) +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + if (((len + 1) & 3) < 2) + return; +@@ -219,7 +219,7 @@ void ide_output_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf, + unsigned long uninitialized_var(flags); + + if ((io_32bit & 2) && !mmio) { +- local_irq_save(flags); ++ local_irq_save_nort(flags); + ata_vlb_sync(io_ports->nsect_addr); + } + +@@ -230,7 +230,7 @@ void ide_output_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf, + outsl(data_addr, buf, words); + + if ((io_32bit & 2) && !mmio) +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + if (((len + 1) & 3) < 2) + return; +diff --git a/drivers/ide/ide-io.c b/drivers/ide/ide-io.c +index 669ea1e45795..e12e43e62245 100644 +--- a/drivers/ide/ide-io.c ++++ b/drivers/ide/ide-io.c +@@ -659,7 +659,7 @@ void ide_timer_expiry (unsigned long data) + /* disable_irq_nosync ?? */ + disable_irq(hwif->irq); + /* local CPU only, as if we were handling an interrupt */ +- local_irq_disable(); ++ local_irq_disable_nort(); + if (hwif->polling) { + startstop = handler(drive); + } else if (drive_is_ready(drive)) { +diff --git a/drivers/ide/ide-iops.c b/drivers/ide/ide-iops.c +index 376f2dc410c5..f014dd1b73dc 100644 +--- a/drivers/ide/ide-iops.c ++++ b/drivers/ide/ide-iops.c +@@ -129,12 +129,12 @@ int __ide_wait_stat(ide_drive_t *drive, u8 good, u8 bad, + if ((stat & ATA_BUSY) == 0) + break; + +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + *rstat = stat; + return -EBUSY; + } + } +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + /* + * Allow status to settle, then read it again. +diff --git a/drivers/ide/ide-probe.c b/drivers/ide/ide-probe.c +index 0b63facd1d87..4ceba37afc0c 100644 +--- a/drivers/ide/ide-probe.c ++++ b/drivers/ide/ide-probe.c +@@ -196,10 +196,10 @@ static void do_identify(ide_drive_t *drive, u8 cmd, u16 *id) + int bswap = 1; + + /* local CPU only; some systems need this */ +- local_irq_save(flags); ++ local_irq_save_nort(flags); + /* read 512 bytes of id info */ + hwif->tp_ops->input_data(drive, NULL, id, SECTOR_SIZE); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + drive->dev_flags |= IDE_DFLAG_ID_READ; + #ifdef DEBUG +diff --git a/drivers/ide/ide-taskfile.c b/drivers/ide/ide-taskfile.c +index a716693417a3..be0568c722d6 100644 +--- a/drivers/ide/ide-taskfile.c ++++ b/drivers/ide/ide-taskfile.c +@@ -250,7 +250,7 @@ void ide_pio_bytes(ide_drive_t *drive, struct ide_cmd *cmd, + + page_is_high = PageHighMem(page); + if (page_is_high) +- local_irq_save(flags); ++ local_irq_save_nort(flags); + + buf = kmap_atomic(page) + offset; + +@@ -271,7 +271,7 @@ void ide_pio_bytes(ide_drive_t *drive, struct ide_cmd *cmd, + kunmap_atomic(buf); + + if (page_is_high) +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + len -= nr_bytes; + } +@@ -414,7 +414,7 @@ static ide_startstop_t pre_task_out_intr(ide_drive_t *drive, + } + + if ((drive->dev_flags & IDE_DFLAG_UNMASK) == 0) +- local_irq_disable(); ++ local_irq_disable_nort(); + + ide_set_handler(drive, &task_pio_intr, WAIT_WORSTCASE); + +diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +index fddff403d5d2..cca1bb4fbfe3 100644 +--- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c ++++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +@@ -902,7 +902,7 @@ void ipoib_mcast_restart_task(struct work_struct *work) + + ipoib_dbg_mcast(priv, "restarting multicast task\n"); + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + netif_addr_lock(dev); + spin_lock(&priv->lock); + +@@ -984,7 +984,7 @@ void ipoib_mcast_restart_task(struct work_struct *work) + + spin_unlock(&priv->lock); + netif_addr_unlock(dev); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + /* + * make sure the in-flight joins have finished before we attempt +diff --git a/drivers/input/gameport/gameport.c b/drivers/input/gameport/gameport.c +index 4a2a9e370be7..e970d9afd179 100644 +--- a/drivers/input/gameport/gameport.c ++++ b/drivers/input/gameport/gameport.c +@@ -91,13 +91,13 @@ static int gameport_measure_speed(struct gameport *gameport) + tx = ~0; + + for (i = 0; i < 50; i++) { +- local_irq_save(flags); ++ local_irq_save_nort(flags); + t1 = ktime_get_ns(); + for (t = 0; t < 50; t++) + gameport_read(gameport); + t2 = ktime_get_ns(); + t3 = ktime_get_ns(); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + udelay(i * 10); + t = (t2 - t1) - (t3 - t2); + if (t < tx) +@@ -124,12 +124,12 @@ static int old_gameport_measure_speed(struct gameport *gameport) + tx = 1 << 30; + + for(i = 0; i < 50; i++) { +- local_irq_save(flags); ++ local_irq_save_nort(flags); + GET_TIME(t1); + for (t = 0; t < 50; t++) gameport_read(gameport); + GET_TIME(t2); + GET_TIME(t3); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + udelay(i * 10); + if ((t = DELTA(t2,t1) - DELTA(t3,t2)) < tx) tx = t; + } +@@ -148,11 +148,11 @@ static int old_gameport_measure_speed(struct gameport *gameport) + tx = 1 << 30; + + for(i = 0; i < 50; i++) { +- local_irq_save(flags); ++ local_irq_save_nort(flags); + t1 = rdtsc(); + for (t = 0; t < 50; t++) gameport_read(gameport); + t2 = rdtsc(); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + udelay(i * 10); + if (t2 - t1 < tx) tx = t2 - t1; + } +diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c +index 11a13b5be73a..baaed0ac274b 100644 +--- a/drivers/iommu/amd_iommu.c ++++ b/drivers/iommu/amd_iommu.c +@@ -1923,10 +1923,10 @@ static int __attach_device(struct iommu_dev_data *dev_data, + int ret; + + /* +- * Must be called with IRQs disabled. Warn here to detect early +- * when its not. ++ * Must be called with IRQs disabled on a non RT kernel. Warn here to ++ * detect early when its not. + */ +- WARN_ON(!irqs_disabled()); ++ WARN_ON_NONRT(!irqs_disabled()); + + /* lock domain */ + spin_lock(&domain->lock); +@@ -2094,10 +2094,10 @@ static void __detach_device(struct iommu_dev_data *dev_data) + struct protection_domain *domain; + + /* +- * Must be called with IRQs disabled. Warn here to detect early +- * when its not. ++ * Must be called with IRQs disabled on a non RT kernel. Warn here to ++ * detect early when its not. + */ +- WARN_ON(!irqs_disabled()); ++ WARN_ON_NONRT(!irqs_disabled()); + + if (WARN_ON(!dev_data->domain)) + return; +diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c +index 87fcbf71b85a..674c82b61f36 100644 +--- a/drivers/iommu/intel-iommu.c ++++ b/drivers/iommu/intel-iommu.c +@@ -479,7 +479,7 @@ struct deferred_flush_data { + struct deferred_flush_table *tables; + }; + +-DEFINE_PER_CPU(struct deferred_flush_data, deferred_flush); ++static DEFINE_PER_CPU(struct deferred_flush_data, deferred_flush); + + /* bitmap for indexing intel_iommus */ + static int g_num_of_iommus; +@@ -3719,10 +3719,8 @@ static void add_unmap(struct dmar_domain *dom, unsigned long iova_pfn, + struct intel_iommu *iommu; + struct deferred_flush_entry *entry; + struct deferred_flush_data *flush_data; +- unsigned int cpuid; + +- cpuid = get_cpu(); +- flush_data = per_cpu_ptr(&deferred_flush, cpuid); ++ flush_data = raw_cpu_ptr(&deferred_flush); + + /* Flush all CPUs' entries to avoid deferring too much. If + * this becomes a bottleneck, can just flush us, and rely on +@@ -3755,8 +3753,6 @@ static void add_unmap(struct dmar_domain *dom, unsigned long iova_pfn, + } + flush_data->size++; + spin_unlock_irqrestore(&flush_data->lock, flags); +- +- put_cpu(); + } + + static void intel_unmap(struct device *dev, dma_addr_t dev_addr, size_t size) +diff --git a/drivers/iommu/iova.c b/drivers/iommu/iova.c +index e23001bfcfee..359d5d169ec0 100644 +--- a/drivers/iommu/iova.c ++++ b/drivers/iommu/iova.c +@@ -22,6 +22,7 @@ + #include + #include + #include ++#include + + static bool iova_rcache_insert(struct iova_domain *iovad, + unsigned long pfn, +@@ -420,10 +421,8 @@ alloc_iova_fast(struct iova_domain *iovad, unsigned long size, + + /* Try replenishing IOVAs by flushing rcache. */ + flushed_rcache = true; +- preempt_disable(); + for_each_online_cpu(cpu) + free_cpu_cached_iovas(cpu, iovad); +- preempt_enable(); + goto retry; + } + +@@ -751,7 +750,7 @@ static bool __iova_rcache_insert(struct iova_domain *iovad, + bool can_insert = false; + unsigned long flags; + +- cpu_rcache = get_cpu_ptr(rcache->cpu_rcaches); ++ cpu_rcache = raw_cpu_ptr(rcache->cpu_rcaches); + spin_lock_irqsave(&cpu_rcache->lock, flags); + + if (!iova_magazine_full(cpu_rcache->loaded)) { +@@ -781,7 +780,6 @@ static bool __iova_rcache_insert(struct iova_domain *iovad, + iova_magazine_push(cpu_rcache->loaded, iova_pfn); + + spin_unlock_irqrestore(&cpu_rcache->lock, flags); +- put_cpu_ptr(rcache->cpu_rcaches); + + if (mag_to_free) { + iova_magazine_free_pfns(mag_to_free, iovad); +@@ -815,7 +813,7 @@ static unsigned long __iova_rcache_get(struct iova_rcache *rcache, + bool has_pfn = false; + unsigned long flags; + +- cpu_rcache = get_cpu_ptr(rcache->cpu_rcaches); ++ cpu_rcache = raw_cpu_ptr(rcache->cpu_rcaches); + spin_lock_irqsave(&cpu_rcache->lock, flags); + + if (!iova_magazine_empty(cpu_rcache->loaded)) { +@@ -837,7 +835,6 @@ static unsigned long __iova_rcache_get(struct iova_rcache *rcache, + iova_pfn = iova_magazine_pop(cpu_rcache->loaded, limit_pfn); + + spin_unlock_irqrestore(&cpu_rcache->lock, flags); +- put_cpu_ptr(rcache->cpu_rcaches); + + return iova_pfn; + } +diff --git a/drivers/leds/trigger/Kconfig b/drivers/leds/trigger/Kconfig +index 3f9ddb9fafa7..09da5b6b44a1 100644 +--- a/drivers/leds/trigger/Kconfig ++++ b/drivers/leds/trigger/Kconfig +@@ -69,7 +69,7 @@ config LEDS_TRIGGER_BACKLIGHT + + config LEDS_TRIGGER_CPU + bool "LED CPU Trigger" +- depends on LEDS_TRIGGERS ++ depends on LEDS_TRIGGERS && !PREEMPT_RT_BASE + help + This allows LEDs to be controlled by active CPUs. This shows + the active CPUs across an array of LEDs so you can see which +diff --git a/drivers/md/bcache/Kconfig b/drivers/md/bcache/Kconfig +index 4d200883c505..98b64ed5cb81 100644 +--- a/drivers/md/bcache/Kconfig ++++ b/drivers/md/bcache/Kconfig +@@ -1,6 +1,7 @@ + + config BCACHE + tristate "Block device as cache" ++ depends on !PREEMPT_RT_FULL + ---help--- + Allows a block device to be used as cache for other devices; uses + a btree for indexing and the layout is optimized for SSDs. +diff --git a/drivers/md/dm-rq.c b/drivers/md/dm-rq.c +index ba7c4c685db3..834ec328f217 100644 +--- a/drivers/md/dm-rq.c ++++ b/drivers/md/dm-rq.c +@@ -842,7 +842,7 @@ static void dm_old_request_fn(struct request_queue *q) + /* Establish tio->ti before queuing work (map_tio_request) */ + tio->ti = ti; + kthread_queue_work(&md->kworker, &tio->work); +- BUG_ON(!irqs_disabled()); ++ BUG_ON_NONRT(!irqs_disabled()); + } + } + +diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c +index f34ad2be66a1..123469f7b560 100644 +--- a/drivers/md/raid5.c ++++ b/drivers/md/raid5.c +@@ -1928,8 +1928,9 @@ static void raid_run_ops(struct stripe_head *sh, unsigned long ops_request) + struct raid5_percpu *percpu; + unsigned long cpu; + +- cpu = get_cpu(); ++ cpu = get_cpu_light(); + percpu = per_cpu_ptr(conf->percpu, cpu); ++ spin_lock(&percpu->lock); + if (test_bit(STRIPE_OP_BIOFILL, &ops_request)) { + ops_run_biofill(sh); + overlap_clear++; +@@ -1985,7 +1986,8 @@ static void raid_run_ops(struct stripe_head *sh, unsigned long ops_request) + if (test_and_clear_bit(R5_Overlap, &dev->flags)) + wake_up(&sh->raid_conf->wait_for_overlap); + } +- put_cpu(); ++ spin_unlock(&percpu->lock); ++ put_cpu_light(); + } + + static struct stripe_head *alloc_stripe(struct kmem_cache *sc, gfp_t gfp, +@@ -6393,6 +6395,7 @@ static int raid456_cpu_up_prepare(unsigned int cpu, struct hlist_node *node) + __func__, cpu); + return -ENOMEM; + } ++ spin_lock_init(&per_cpu_ptr(conf->percpu, cpu)->lock); + return 0; + } + +@@ -6403,7 +6406,6 @@ static int raid5_alloc_percpu(struct r5conf *conf) + conf->percpu = alloc_percpu(struct raid5_percpu); + if (!conf->percpu) + return -ENOMEM; +- + err = cpuhp_state_add_instance(CPUHP_MD_RAID5_PREPARE, &conf->node); + if (!err) { + conf->scribble_disks = max(conf->raid_disks, +diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h +index 57ec49f0839e..0739604990b7 100644 +--- a/drivers/md/raid5.h ++++ b/drivers/md/raid5.h +@@ -504,6 +504,7 @@ struct r5conf { + int recovery_disabled; + /* per cpu variables */ + struct raid5_percpu { ++ spinlock_t lock; /* Protection for -RT */ + struct page *spare_page; /* Used when checking P/Q in raid6 */ + struct flex_array *scribble; /* space for constructing buffer + * lists and performing address +diff --git a/drivers/misc/Kconfig b/drivers/misc/Kconfig +index 64971baf11fa..215e91e36198 100644 +--- a/drivers/misc/Kconfig ++++ b/drivers/misc/Kconfig +@@ -54,6 +54,7 @@ config AD525X_DPOT_SPI + config ATMEL_TCLIB + bool "Atmel AT32/AT91 Timer/Counter Library" + depends on (AVR32 || ARCH_AT91) ++ default y if PREEMPT_RT_FULL + help + Select this if you want a library to allocate the Timer/Counter + blocks found on many Atmel processors. This facilitates using +@@ -69,8 +70,7 @@ config ATMEL_TCB_CLKSRC + are combined to make a single 32-bit timer. + + When GENERIC_CLOCKEVENTS is defined, the third timer channel +- may be used as a clock event device supporting oneshot mode +- (delays of up to two seconds) based on the 32 KiHz clock. ++ may be used as a clock event device supporting oneshot mode. + + config ATMEL_TCB_CLKSRC_BLOCK + int +@@ -84,6 +84,15 @@ config ATMEL_TCB_CLKSRC_BLOCK + TC can be used for other purposes, such as PWM generation and + interval timing. + ++config ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK ++ bool "TC Block use 32 KiHz clock" ++ depends on ATMEL_TCB_CLKSRC ++ default y if !PREEMPT_RT_FULL ++ help ++ Select this to use 32 KiHz base clock rate as TC block clock ++ source for clock events. ++ ++ + config DUMMY_IRQ + tristate "Dummy IRQ handler" + default n +diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c +index df990bb8c873..1a162709a85e 100644 +--- a/drivers/mmc/host/mmci.c ++++ b/drivers/mmc/host/mmci.c +@@ -1147,15 +1147,12 @@ static irqreturn_t mmci_pio_irq(int irq, void *dev_id) + struct sg_mapping_iter *sg_miter = &host->sg_miter; + struct variant_data *variant = host->variant; + void __iomem *base = host->base; +- unsigned long flags; + u32 status; + + status = readl(base + MMCISTATUS); + + dev_dbg(mmc_dev(host->mmc), "irq1 (pio) %08x\n", status); + +- local_irq_save(flags); +- + do { + unsigned int remain, len; + char *buffer; +@@ -1195,8 +1192,6 @@ static irqreturn_t mmci_pio_irq(int irq, void *dev_id) + + sg_miter_stop(sg_miter); + +- local_irq_restore(flags); +- + /* + * If we have less than the fifo 'half-full' threshold to transfer, + * trigger a PIO interrupt as soon as any data is available. +diff --git a/drivers/net/ethernet/3com/3c59x.c b/drivers/net/ethernet/3com/3c59x.c +index 9133e7926da5..63afb921ed40 100644 +--- a/drivers/net/ethernet/3com/3c59x.c ++++ b/drivers/net/ethernet/3com/3c59x.c +@@ -842,9 +842,9 @@ static void poll_vortex(struct net_device *dev) + { + struct vortex_private *vp = netdev_priv(dev); + unsigned long flags; +- local_irq_save(flags); ++ local_irq_save_nort(flags); + (vp->full_bus_master_rx ? boomerang_interrupt:vortex_interrupt)(dev->irq,dev); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + #endif + +@@ -1910,12 +1910,12 @@ static void vortex_tx_timeout(struct net_device *dev) + * Block interrupts because vortex_interrupt does a bare spin_lock() + */ + unsigned long flags; +- local_irq_save(flags); ++ local_irq_save_nort(flags); + if (vp->full_bus_master_tx) + boomerang_interrupt(dev->irq, dev); + else + vortex_interrupt(dev->irq, dev); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + } + +diff --git a/drivers/net/ethernet/realtek/8139too.c b/drivers/net/ethernet/realtek/8139too.c +index da4c2d8a4173..1420dfb56bac 100644 +--- a/drivers/net/ethernet/realtek/8139too.c ++++ b/drivers/net/ethernet/realtek/8139too.c +@@ -2233,7 +2233,7 @@ static void rtl8139_poll_controller(struct net_device *dev) + struct rtl8139_private *tp = netdev_priv(dev); + const int irq = tp->pci_dev->irq; + +- disable_irq(irq); ++ disable_irq_nosync(irq); + rtl8139_interrupt(irq, dev); + enable_irq(irq); + } +diff --git a/drivers/net/wireless/intersil/orinoco/orinoco_usb.c b/drivers/net/wireless/intersil/orinoco/orinoco_usb.c +index bca6935a94db..d7a35ee34d03 100644 +--- a/drivers/net/wireless/intersil/orinoco/orinoco_usb.c ++++ b/drivers/net/wireless/intersil/orinoco/orinoco_usb.c +@@ -697,7 +697,7 @@ static void ezusb_req_ctx_wait(struct ezusb_priv *upriv, + while (!ctx->done.done && msecs--) + udelay(1000); + } else { +- wait_event_interruptible(ctx->done.wait, ++ swait_event_interruptible(ctx->done.wait, + ctx->done.done); + } + break; +diff --git a/drivers/pci/access.c b/drivers/pci/access.c +index d11cdbb8fba3..223bbb9acb03 100644 +--- a/drivers/pci/access.c ++++ b/drivers/pci/access.c +@@ -672,7 +672,7 @@ void pci_cfg_access_unlock(struct pci_dev *dev) + WARN_ON(!dev->block_cfg_access); + + dev->block_cfg_access = 0; +- wake_up_all(&pci_cfg_wait); ++ wake_up_all_locked(&pci_cfg_wait); + raw_spin_unlock_irqrestore(&pci_lock, flags); + } + EXPORT_SYMBOL_GPL(pci_cfg_access_unlock); +diff --git a/drivers/pinctrl/qcom/pinctrl-msm.c b/drivers/pinctrl/qcom/pinctrl-msm.c +index bedce3453dd3..faf038978650 100644 +--- a/drivers/pinctrl/qcom/pinctrl-msm.c ++++ b/drivers/pinctrl/qcom/pinctrl-msm.c +@@ -61,7 +61,7 @@ struct msm_pinctrl { + struct notifier_block restart_nb; + int irq; + +- spinlock_t lock; ++ raw_spinlock_t lock; + + DECLARE_BITMAP(dual_edge_irqs, MAX_NR_GPIO); + DECLARE_BITMAP(enabled_irqs, MAX_NR_GPIO); +@@ -153,14 +153,14 @@ static int msm_pinmux_set_mux(struct pinctrl_dev *pctldev, + if (WARN_ON(i == g->nfuncs)) + return -EINVAL; + +- spin_lock_irqsave(&pctrl->lock, flags); ++ raw_spin_lock_irqsave(&pctrl->lock, flags); + + val = readl(pctrl->regs + g->ctl_reg); + val &= ~mask; + val |= i << g->mux_bit; + writel(val, pctrl->regs + g->ctl_reg); + +- spin_unlock_irqrestore(&pctrl->lock, flags); ++ raw_spin_unlock_irqrestore(&pctrl->lock, flags); + + return 0; + } +@@ -323,14 +323,14 @@ static int msm_config_group_set(struct pinctrl_dev *pctldev, + break; + case PIN_CONFIG_OUTPUT: + /* set output value */ +- spin_lock_irqsave(&pctrl->lock, flags); ++ raw_spin_lock_irqsave(&pctrl->lock, flags); + val = readl(pctrl->regs + g->io_reg); + if (arg) + val |= BIT(g->out_bit); + else + val &= ~BIT(g->out_bit); + writel(val, pctrl->regs + g->io_reg); +- spin_unlock_irqrestore(&pctrl->lock, flags); ++ raw_spin_unlock_irqrestore(&pctrl->lock, flags); + + /* enable output */ + arg = 1; +@@ -351,12 +351,12 @@ static int msm_config_group_set(struct pinctrl_dev *pctldev, + return -EINVAL; + } + +- spin_lock_irqsave(&pctrl->lock, flags); ++ raw_spin_lock_irqsave(&pctrl->lock, flags); + val = readl(pctrl->regs + g->ctl_reg); + val &= ~(mask << bit); + val |= arg << bit; + writel(val, pctrl->regs + g->ctl_reg); +- spin_unlock_irqrestore(&pctrl->lock, flags); ++ raw_spin_unlock_irqrestore(&pctrl->lock, flags); + } + + return 0; +@@ -384,13 +384,13 @@ static int msm_gpio_direction_input(struct gpio_chip *chip, unsigned offset) + + g = &pctrl->soc->groups[offset]; + +- spin_lock_irqsave(&pctrl->lock, flags); ++ raw_spin_lock_irqsave(&pctrl->lock, flags); + + val = readl(pctrl->regs + g->ctl_reg); + val &= ~BIT(g->oe_bit); + writel(val, pctrl->regs + g->ctl_reg); + +- spin_unlock_irqrestore(&pctrl->lock, flags); ++ raw_spin_unlock_irqrestore(&pctrl->lock, flags); + + return 0; + } +@@ -404,7 +404,7 @@ static int msm_gpio_direction_output(struct gpio_chip *chip, unsigned offset, in + + g = &pctrl->soc->groups[offset]; + +- spin_lock_irqsave(&pctrl->lock, flags); ++ raw_spin_lock_irqsave(&pctrl->lock, flags); + + val = readl(pctrl->regs + g->io_reg); + if (value) +@@ -417,7 +417,7 @@ static int msm_gpio_direction_output(struct gpio_chip *chip, unsigned offset, in + val |= BIT(g->oe_bit); + writel(val, pctrl->regs + g->ctl_reg); + +- spin_unlock_irqrestore(&pctrl->lock, flags); ++ raw_spin_unlock_irqrestore(&pctrl->lock, flags); + + return 0; + } +@@ -443,7 +443,7 @@ static void msm_gpio_set(struct gpio_chip *chip, unsigned offset, int value) + + g = &pctrl->soc->groups[offset]; + +- spin_lock_irqsave(&pctrl->lock, flags); ++ raw_spin_lock_irqsave(&pctrl->lock, flags); + + val = readl(pctrl->regs + g->io_reg); + if (value) +@@ -452,7 +452,7 @@ static void msm_gpio_set(struct gpio_chip *chip, unsigned offset, int value) + val &= ~BIT(g->out_bit); + writel(val, pctrl->regs + g->io_reg); + +- spin_unlock_irqrestore(&pctrl->lock, flags); ++ raw_spin_unlock_irqrestore(&pctrl->lock, flags); + } + + #ifdef CONFIG_DEBUG_FS +@@ -571,7 +571,7 @@ static void msm_gpio_irq_mask(struct irq_data *d) + + g = &pctrl->soc->groups[d->hwirq]; + +- spin_lock_irqsave(&pctrl->lock, flags); ++ raw_spin_lock_irqsave(&pctrl->lock, flags); + + val = readl(pctrl->regs + g->intr_cfg_reg); + val &= ~BIT(g->intr_enable_bit); +@@ -579,7 +579,7 @@ static void msm_gpio_irq_mask(struct irq_data *d) + + clear_bit(d->hwirq, pctrl->enabled_irqs); + +- spin_unlock_irqrestore(&pctrl->lock, flags); ++ raw_spin_unlock_irqrestore(&pctrl->lock, flags); + } + + static void msm_gpio_irq_unmask(struct irq_data *d) +@@ -592,7 +592,7 @@ static void msm_gpio_irq_unmask(struct irq_data *d) + + g = &pctrl->soc->groups[d->hwirq]; + +- spin_lock_irqsave(&pctrl->lock, flags); ++ raw_spin_lock_irqsave(&pctrl->lock, flags); + + val = readl(pctrl->regs + g->intr_cfg_reg); + val |= BIT(g->intr_enable_bit); +@@ -600,7 +600,7 @@ static void msm_gpio_irq_unmask(struct irq_data *d) + + set_bit(d->hwirq, pctrl->enabled_irqs); + +- spin_unlock_irqrestore(&pctrl->lock, flags); ++ raw_spin_unlock_irqrestore(&pctrl->lock, flags); + } + + static void msm_gpio_irq_ack(struct irq_data *d) +@@ -613,7 +613,7 @@ static void msm_gpio_irq_ack(struct irq_data *d) + + g = &pctrl->soc->groups[d->hwirq]; + +- spin_lock_irqsave(&pctrl->lock, flags); ++ raw_spin_lock_irqsave(&pctrl->lock, flags); + + val = readl(pctrl->regs + g->intr_status_reg); + if (g->intr_ack_high) +@@ -625,7 +625,7 @@ static void msm_gpio_irq_ack(struct irq_data *d) + if (test_bit(d->hwirq, pctrl->dual_edge_irqs)) + msm_gpio_update_dual_edge_pos(pctrl, g, d); + +- spin_unlock_irqrestore(&pctrl->lock, flags); ++ raw_spin_unlock_irqrestore(&pctrl->lock, flags); + } + + static int msm_gpio_irq_set_type(struct irq_data *d, unsigned int type) +@@ -638,7 +638,7 @@ static int msm_gpio_irq_set_type(struct irq_data *d, unsigned int type) + + g = &pctrl->soc->groups[d->hwirq]; + +- spin_lock_irqsave(&pctrl->lock, flags); ++ raw_spin_lock_irqsave(&pctrl->lock, flags); + + /* + * For hw without possibility of detecting both edges +@@ -712,7 +712,7 @@ static int msm_gpio_irq_set_type(struct irq_data *d, unsigned int type) + if (test_bit(d->hwirq, pctrl->dual_edge_irqs)) + msm_gpio_update_dual_edge_pos(pctrl, g, d); + +- spin_unlock_irqrestore(&pctrl->lock, flags); ++ raw_spin_unlock_irqrestore(&pctrl->lock, flags); + + if (type & (IRQ_TYPE_LEVEL_LOW | IRQ_TYPE_LEVEL_HIGH)) + irq_set_handler_locked(d, handle_level_irq); +@@ -728,11 +728,11 @@ static int msm_gpio_irq_set_wake(struct irq_data *d, unsigned int on) + struct msm_pinctrl *pctrl = gpiochip_get_data(gc); + unsigned long flags; + +- spin_lock_irqsave(&pctrl->lock, flags); ++ raw_spin_lock_irqsave(&pctrl->lock, flags); + + irq_set_irq_wake(pctrl->irq, on); + +- spin_unlock_irqrestore(&pctrl->lock, flags); ++ raw_spin_unlock_irqrestore(&pctrl->lock, flags); + + return 0; + } +@@ -878,7 +878,7 @@ int msm_pinctrl_probe(struct platform_device *pdev, + pctrl->soc = soc_data; + pctrl->chip = msm_gpio_template; + +- spin_lock_init(&pctrl->lock); ++ raw_spin_lock_init(&pctrl->lock); + + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + pctrl->regs = devm_ioremap_resource(&pdev->dev, res); +diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c +index 9bd41a35a78a..8e2d436c2e3f 100644 +--- a/drivers/scsi/fcoe/fcoe.c ++++ b/drivers/scsi/fcoe/fcoe.c +@@ -1455,11 +1455,11 @@ static int fcoe_rcv(struct sk_buff *skb, struct net_device *netdev, + static int fcoe_alloc_paged_crc_eof(struct sk_buff *skb, int tlen) + { + struct fcoe_percpu_s *fps; +- int rc; ++ int rc, cpu = get_cpu_light(); + +- fps = &get_cpu_var(fcoe_percpu); ++ fps = &per_cpu(fcoe_percpu, cpu); + rc = fcoe_get_paged_crc_eof(skb, tlen, fps); +- put_cpu_var(fcoe_percpu); ++ put_cpu_light(); + + return rc; + } +@@ -1646,11 +1646,11 @@ static inline int fcoe_filter_frames(struct fc_lport *lport, + return 0; + } + +- stats = per_cpu_ptr(lport->stats, get_cpu()); ++ stats = per_cpu_ptr(lport->stats, get_cpu_light()); + stats->InvalidCRCCount++; + if (stats->InvalidCRCCount < 5) + printk(KERN_WARNING "fcoe: dropping frame with CRC error\n"); +- put_cpu(); ++ put_cpu_light(); + return -EINVAL; + } + +@@ -1693,7 +1693,7 @@ static void fcoe_recv_frame(struct sk_buff *skb) + */ + hp = (struct fcoe_hdr *) skb_network_header(skb); + +- stats = per_cpu_ptr(lport->stats, get_cpu()); ++ stats = per_cpu_ptr(lport->stats, get_cpu_light()); + if (unlikely(FC_FCOE_DECAPS_VER(hp) != FC_FCOE_VER)) { + if (stats->ErrorFrames < 5) + printk(KERN_WARNING "fcoe: FCoE version " +@@ -1725,13 +1725,13 @@ static void fcoe_recv_frame(struct sk_buff *skb) + goto drop; + + if (!fcoe_filter_frames(lport, fp)) { +- put_cpu(); ++ put_cpu_light(); + fc_exch_recv(lport, fp); + return; + } + drop: + stats->ErrorFrames++; +- put_cpu(); ++ put_cpu_light(); + kfree_skb(skb); + } + +diff --git a/drivers/scsi/fcoe/fcoe_ctlr.c b/drivers/scsi/fcoe/fcoe_ctlr.c +index dcf36537a767..1a1f2e46452c 100644 +--- a/drivers/scsi/fcoe/fcoe_ctlr.c ++++ b/drivers/scsi/fcoe/fcoe_ctlr.c +@@ -834,7 +834,7 @@ static unsigned long fcoe_ctlr_age_fcfs(struct fcoe_ctlr *fip) + + INIT_LIST_HEAD(&del_list); + +- stats = per_cpu_ptr(fip->lp->stats, get_cpu()); ++ stats = per_cpu_ptr(fip->lp->stats, get_cpu_light()); + + list_for_each_entry_safe(fcf, next, &fip->fcfs, list) { + deadline = fcf->time + fcf->fka_period + fcf->fka_period / 2; +@@ -870,7 +870,7 @@ static unsigned long fcoe_ctlr_age_fcfs(struct fcoe_ctlr *fip) + sel_time = fcf->time; + } + } +- put_cpu(); ++ put_cpu_light(); + + list_for_each_entry_safe(fcf, next, &del_list, list) { + /* Removes fcf from current list */ +diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c +index 16ca31ad5ec0..c3987347e762 100644 +--- a/drivers/scsi/libfc/fc_exch.c ++++ b/drivers/scsi/libfc/fc_exch.c +@@ -814,10 +814,10 @@ static struct fc_exch *fc_exch_em_alloc(struct fc_lport *lport, + } + memset(ep, 0, sizeof(*ep)); + +- cpu = get_cpu(); ++ cpu = get_cpu_light(); + pool = per_cpu_ptr(mp->pool, cpu); + spin_lock_bh(&pool->lock); +- put_cpu(); ++ put_cpu_light(); + + /* peek cache of free slot */ + if (pool->left != FC_XID_UNKNOWN) { +diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c +index 87f5e694dbed..23c0a50fb6aa 100644 +--- a/drivers/scsi/libsas/sas_ata.c ++++ b/drivers/scsi/libsas/sas_ata.c +@@ -190,7 +190,7 @@ static unsigned int sas_ata_qc_issue(struct ata_queued_cmd *qc) + /* TODO: audit callers to ensure they are ready for qc_issue to + * unconditionally re-enable interrupts + */ +- local_irq_save(flags); ++ local_irq_save_nort(flags); + spin_unlock(ap->lock); + + /* If the device fell off, no sense in issuing commands */ +@@ -252,7 +252,7 @@ static unsigned int sas_ata_qc_issue(struct ata_queued_cmd *qc) + + out: + spin_lock(ap->lock); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + return ret; + } + +diff --git a/drivers/scsi/qla2xxx/qla_inline.h b/drivers/scsi/qla2xxx/qla_inline.h +index edc48f3b8230..ee5c6f9dfb6f 100644 +--- a/drivers/scsi/qla2xxx/qla_inline.h ++++ b/drivers/scsi/qla2xxx/qla_inline.h +@@ -59,12 +59,12 @@ qla2x00_poll(struct rsp_que *rsp) + { + unsigned long flags; + struct qla_hw_data *ha = rsp->hw; +- local_irq_save(flags); ++ local_irq_save_nort(flags); + if (IS_P3P_TYPE(ha)) + qla82xx_poll(0, rsp); + else + ha->isp_ops->intr_handler(0, rsp); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + + static inline uint8_t * +diff --git a/drivers/scsi/qla2xxx/qla_isr.c b/drivers/scsi/qla2xxx/qla_isr.c +index 068c4e47fac9..a2090f640397 100644 +--- a/drivers/scsi/qla2xxx/qla_isr.c ++++ b/drivers/scsi/qla2xxx/qla_isr.c +@@ -3125,7 +3125,11 @@ qla24xx_enable_msix(struct qla_hw_data *ha, struct rsp_que *rsp) + * kref_put(). + */ + kref_get(&qentry->irq_notify.kref); ++#ifdef CONFIG_PREEMPT_RT_BASE ++ swork_queue(&qentry->irq_notify.swork); ++#else + schedule_work(&qentry->irq_notify.work); ++#endif + } + + /* +diff --git a/drivers/thermal/x86_pkg_temp_thermal.c b/drivers/thermal/x86_pkg_temp_thermal.c +index 95f4c1bcdb4c..0be934799bff 100644 +--- a/drivers/thermal/x86_pkg_temp_thermal.c ++++ b/drivers/thermal/x86_pkg_temp_thermal.c +@@ -29,6 +29,7 @@ + #include + #include + #include ++#include + #include + #include + +@@ -353,7 +354,7 @@ static void pkg_temp_thermal_threshold_work_fn(struct work_struct *work) + } + } + +-static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) ++static void platform_thermal_notify_work(struct swork_event *event) + { + unsigned long flags; + int cpu = smp_processor_id(); +@@ -370,7 +371,7 @@ static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) + pkg_work_scheduled[phy_id]) { + disable_pkg_thres_interrupt(); + spin_unlock_irqrestore(&pkg_work_lock, flags); +- return -EINVAL; ++ return; + } + pkg_work_scheduled[phy_id] = 1; + spin_unlock_irqrestore(&pkg_work_lock, flags); +@@ -379,9 +380,48 @@ static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) + schedule_delayed_work_on(cpu, + &per_cpu(pkg_temp_thermal_threshold_work, cpu), + msecs_to_jiffies(notify_delay_ms)); ++} ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++static struct swork_event notify_work; ++ ++static int thermal_notify_work_init(void) ++{ ++ int err; ++ ++ err = swork_get(); ++ if (err) ++ return err; ++ ++ INIT_SWORK(¬ify_work, platform_thermal_notify_work); + return 0; + } + ++static void thermal_notify_work_cleanup(void) ++{ ++ swork_put(); ++} ++ ++static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) ++{ ++ swork_queue(¬ify_work); ++ return 0; ++} ++ ++#else /* !CONFIG_PREEMPT_RT_FULL */ ++ ++static int thermal_notify_work_init(void) { return 0; } ++ ++static void thermal_notify_work_cleanup(void) { } ++ ++static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) ++{ ++ platform_thermal_notify_work(NULL); ++ ++ return 0; ++} ++#endif /* CONFIG_PREEMPT_RT_FULL */ ++ + static int find_siblings_cpu(int cpu) + { + int i; +@@ -585,6 +625,9 @@ static int __init pkg_temp_thermal_init(void) + if (!x86_match_cpu(pkg_temp_thermal_ids)) + return -ENODEV; + ++ if (!thermal_notify_work_init()) ++ return -ENODEV; ++ + spin_lock_init(&pkg_work_lock); + platform_thermal_package_notify = + pkg_temp_thermal_platform_thermal_notify; +@@ -609,7 +652,7 @@ static int __init pkg_temp_thermal_init(void) + kfree(pkg_work_scheduled); + platform_thermal_package_notify = NULL; + platform_thermal_package_rate_control = NULL; +- ++ thermal_notify_work_cleanup(); + return -ENODEV; + } + +@@ -634,6 +677,7 @@ static void __exit pkg_temp_thermal_exit(void) + mutex_unlock(&phy_dev_list_mutex); + platform_thermal_package_notify = NULL; + platform_thermal_package_rate_control = NULL; ++ thermal_notify_work_cleanup(); + for_each_online_cpu(i) + cancel_delayed_work_sync( + &per_cpu(pkg_temp_thermal_threshold_work, i)); +diff --git a/drivers/tty/serial/8250/8250_core.c b/drivers/tty/serial/8250/8250_core.c +index e8819aa20415..dd7f9bf45d6c 100644 +--- a/drivers/tty/serial/8250/8250_core.c ++++ b/drivers/tty/serial/8250/8250_core.c +@@ -58,7 +58,16 @@ static struct uart_driver serial8250_reg; + + static unsigned int skip_txen_test; /* force skip of txen test at init time */ + +-#define PASS_LIMIT 512 ++/* ++ * On -rt we can have a more delays, and legitimately ++ * so - so don't drop work spuriously and spam the ++ * syslog: ++ */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define PASS_LIMIT 1000000 ++#else ++# define PASS_LIMIT 512 ++#endif + + #include + /* +diff --git a/drivers/tty/serial/8250/8250_port.c b/drivers/tty/serial/8250/8250_port.c +index 080d5a59d0a7..eecc4f111473 100644 +--- a/drivers/tty/serial/8250/8250_port.c ++++ b/drivers/tty/serial/8250/8250_port.c +@@ -35,6 +35,7 @@ + #include + #include + #include ++#include + #include + #include + #include +@@ -3144,9 +3145,9 @@ void serial8250_console_write(struct uart_8250_port *up, const char *s, + + serial8250_rpm_get(up); + +- if (port->sysrq) ++ if (port->sysrq || oops_in_progress) + locked = 0; +- else if (oops_in_progress) ++ else if (in_kdb_printk()) + locked = spin_trylock_irqsave(&port->lock, flags); + else + spin_lock_irqsave(&port->lock, flags); +diff --git a/drivers/tty/serial/amba-pl011.c b/drivers/tty/serial/amba-pl011.c +index e2c33b9528d8..53af53c43e8c 100644 +--- a/drivers/tty/serial/amba-pl011.c ++++ b/drivers/tty/serial/amba-pl011.c +@@ -2194,13 +2194,19 @@ pl011_console_write(struct console *co, const char *s, unsigned int count) + + clk_enable(uap->clk); + +- local_irq_save(flags); ++ /* ++ * local_irq_save(flags); ++ * ++ * This local_irq_save() is nonsense. If we come in via sysrq ++ * handling then interrupts are already disabled. Aside of ++ * that the port.sysrq check is racy on SMP regardless. ++ */ + if (uap->port.sysrq) + locked = 0; + else if (oops_in_progress) +- locked = spin_trylock(&uap->port.lock); ++ locked = spin_trylock_irqsave(&uap->port.lock, flags); + else +- spin_lock(&uap->port.lock); ++ spin_lock_irqsave(&uap->port.lock, flags); + + /* + * First save the CR then disable the interrupts +@@ -2224,8 +2230,7 @@ pl011_console_write(struct console *co, const char *s, unsigned int count) + pl011_write(old_cr, uap, REG_CR); + + if (locked) +- spin_unlock(&uap->port.lock); +- local_irq_restore(flags); ++ spin_unlock_irqrestore(&uap->port.lock, flags); + + clk_disable(uap->clk); + } +diff --git a/drivers/tty/serial/omap-serial.c b/drivers/tty/serial/omap-serial.c +index 44e5b5bf713b..400140d1dfff 100644 +--- a/drivers/tty/serial/omap-serial.c ++++ b/drivers/tty/serial/omap-serial.c +@@ -1257,13 +1257,10 @@ serial_omap_console_write(struct console *co, const char *s, + + pm_runtime_get_sync(up->dev); + +- local_irq_save(flags); +- if (up->port.sysrq) +- locked = 0; +- else if (oops_in_progress) +- locked = spin_trylock(&up->port.lock); ++ if (up->port.sysrq || oops_in_progress) ++ locked = spin_trylock_irqsave(&up->port.lock, flags); + else +- spin_lock(&up->port.lock); ++ spin_lock_irqsave(&up->port.lock, flags); + + /* + * First save the IER then disable the interrupts +@@ -1292,8 +1289,7 @@ serial_omap_console_write(struct console *co, const char *s, + pm_runtime_mark_last_busy(up->dev); + pm_runtime_put_autosuspend(up->dev); + if (locked) +- spin_unlock(&up->port.lock); +- local_irq_restore(flags); ++ spin_unlock_irqrestore(&up->port.lock, flags); + } + + static int __init +diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c +index 3b9735abf2e0..73ba3239869e 100644 +--- a/drivers/usb/core/hcd.c ++++ b/drivers/usb/core/hcd.c +@@ -1764,9 +1764,9 @@ static void __usb_hcd_giveback_urb(struct urb *urb) + * and no one may trigger the above deadlock situation when + * running complete() in tasklet. + */ +- local_irq_save(flags); ++ local_irq_save_nort(flags); + urb->complete(urb); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + usb_anchor_resume_wakeups(anchor); + atomic_dec(&urb->use_count); +diff --git a/drivers/usb/gadget/function/f_fs.c b/drivers/usb/gadget/function/f_fs.c +index 89081b834615..90b231b7ad0a 100644 +--- a/drivers/usb/gadget/function/f_fs.c ++++ b/drivers/usb/gadget/function/f_fs.c +@@ -1593,7 +1593,7 @@ static void ffs_data_put(struct ffs_data *ffs) + pr_info("%s(): freeing\n", __func__); + ffs_data_clear(ffs); + BUG_ON(waitqueue_active(&ffs->ev.waitq) || +- waitqueue_active(&ffs->ep0req_completion.wait)); ++ swait_active(&ffs->ep0req_completion.wait)); + kfree(ffs->dev_name); + kfree(ffs); + } +diff --git a/drivers/usb/gadget/legacy/inode.c b/drivers/usb/gadget/legacy/inode.c +index 1468d8f085a3..6aae3ae25c18 100644 +--- a/drivers/usb/gadget/legacy/inode.c ++++ b/drivers/usb/gadget/legacy/inode.c +@@ -346,7 +346,7 @@ ep_io (struct ep_data *epdata, void *buf, unsigned len) + spin_unlock_irq (&epdata->dev->lock); + + if (likely (value == 0)) { +- value = wait_event_interruptible (done.wait, done.done); ++ value = swait_event_interruptible (done.wait, done.done); + if (value != 0) { + spin_lock_irq (&epdata->dev->lock); + if (likely (epdata->ep != NULL)) { +@@ -355,7 +355,7 @@ ep_io (struct ep_data *epdata, void *buf, unsigned len) + usb_ep_dequeue (epdata->ep, epdata->req); + spin_unlock_irq (&epdata->dev->lock); + +- wait_event (done.wait, done.done); ++ swait_event (done.wait, done.done); + if (epdata->status == -ECONNRESET) + epdata->status = -EINTR; + } else { +diff --git a/fs/aio.c b/fs/aio.c +index 428484f2f841..2b02e2eb2158 100644 +--- a/fs/aio.c ++++ b/fs/aio.c +@@ -40,6 +40,7 @@ + #include + #include + #include ++#include + + #include + #include +@@ -115,7 +116,7 @@ struct kioctx { + struct page **ring_pages; + long nr_pages; + +- struct work_struct free_work; ++ struct swork_event free_work; + + /* + * signals when all in-flight requests are done +@@ -258,6 +259,7 @@ static int __init aio_setup(void) + .mount = aio_mount, + .kill_sb = kill_anon_super, + }; ++ BUG_ON(swork_get()); + aio_mnt = kern_mount(&aio_fs); + if (IS_ERR(aio_mnt)) + panic("Failed to create aio fs mount."); +@@ -581,9 +583,9 @@ static int kiocb_cancel(struct aio_kiocb *kiocb) + return cancel(&kiocb->common); + } + +-static void free_ioctx(struct work_struct *work) ++static void free_ioctx(struct swork_event *sev) + { +- struct kioctx *ctx = container_of(work, struct kioctx, free_work); ++ struct kioctx *ctx = container_of(sev, struct kioctx, free_work); + + pr_debug("freeing %p\n", ctx); + +@@ -602,8 +604,8 @@ static void free_ioctx_reqs(struct percpu_ref *ref) + if (ctx->rq_wait && atomic_dec_and_test(&ctx->rq_wait->count)) + complete(&ctx->rq_wait->comp); + +- INIT_WORK(&ctx->free_work, free_ioctx); +- schedule_work(&ctx->free_work); ++ INIT_SWORK(&ctx->free_work, free_ioctx); ++ swork_queue(&ctx->free_work); + } + + /* +@@ -611,9 +613,9 @@ static void free_ioctx_reqs(struct percpu_ref *ref) + * and ctx->users has dropped to 0, so we know no more kiocbs can be submitted - + * now it's safe to cancel any that need to be. + */ +-static void free_ioctx_users(struct percpu_ref *ref) ++static void free_ioctx_users_work(struct swork_event *sev) + { +- struct kioctx *ctx = container_of(ref, struct kioctx, users); ++ struct kioctx *ctx = container_of(sev, struct kioctx, free_work); + struct aio_kiocb *req; + + spin_lock_irq(&ctx->ctx_lock); +@@ -632,6 +634,14 @@ static void free_ioctx_users(struct percpu_ref *ref) + percpu_ref_put(&ctx->reqs); + } + ++static void free_ioctx_users(struct percpu_ref *ref) ++{ ++ struct kioctx *ctx = container_of(ref, struct kioctx, users); ++ ++ INIT_SWORK(&ctx->free_work, free_ioctx_users_work); ++ swork_queue(&ctx->free_work); ++} ++ + static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm) + { + unsigned i, new_nr; +diff --git a/fs/autofs4/autofs_i.h b/fs/autofs4/autofs_i.h +index a1fba4285277..3796769b4cd1 100644 +--- a/fs/autofs4/autofs_i.h ++++ b/fs/autofs4/autofs_i.h +@@ -31,6 +31,7 @@ + #include + #include + #include ++#include + #include + #include + +diff --git a/fs/autofs4/expire.c b/fs/autofs4/expire.c +index d8e6d421c27f..2e689ab1306b 100644 +--- a/fs/autofs4/expire.c ++++ b/fs/autofs4/expire.c +@@ -148,7 +148,7 @@ static struct dentry *get_next_positive_dentry(struct dentry *prev, + parent = p->d_parent; + if (!spin_trylock(&parent->d_lock)) { + spin_unlock(&p->d_lock); +- cpu_relax(); ++ cpu_chill(); + goto relock; + } + spin_unlock(&p->d_lock); +diff --git a/fs/buffer.c b/fs/buffer.c +index b205a629001d..5646afc022ba 100644 +--- a/fs/buffer.c ++++ b/fs/buffer.c +@@ -301,8 +301,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate) + * decide that the page is now completely done. + */ + first = page_buffers(page); +- local_irq_save(flags); +- bit_spin_lock(BH_Uptodate_Lock, &first->b_state); ++ flags = bh_uptodate_lock_irqsave(first); + clear_buffer_async_read(bh); + unlock_buffer(bh); + tmp = bh; +@@ -315,8 +314,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate) + } + tmp = tmp->b_this_page; + } while (tmp != bh); +- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); +- local_irq_restore(flags); ++ bh_uptodate_unlock_irqrestore(first, flags); + + /* + * If none of the buffers had errors and they are all +@@ -328,9 +326,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate) + return; + + still_busy: +- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); +- local_irq_restore(flags); +- return; ++ bh_uptodate_unlock_irqrestore(first, flags); + } + + /* +@@ -358,8 +354,7 @@ void end_buffer_async_write(struct buffer_head *bh, int uptodate) + } + + first = page_buffers(page); +- local_irq_save(flags); +- bit_spin_lock(BH_Uptodate_Lock, &first->b_state); ++ flags = bh_uptodate_lock_irqsave(first); + + clear_buffer_async_write(bh); + unlock_buffer(bh); +@@ -371,15 +366,12 @@ void end_buffer_async_write(struct buffer_head *bh, int uptodate) + } + tmp = tmp->b_this_page; + } +- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); +- local_irq_restore(flags); ++ bh_uptodate_unlock_irqrestore(first, flags); + end_page_writeback(page); + return; + + still_busy: +- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); +- local_irq_restore(flags); +- return; ++ bh_uptodate_unlock_irqrestore(first, flags); + } + EXPORT_SYMBOL(end_buffer_async_write); + +@@ -3383,6 +3375,7 @@ struct buffer_head *alloc_buffer_head(gfp_t gfp_flags) + struct buffer_head *ret = kmem_cache_zalloc(bh_cachep, gfp_flags); + if (ret) { + INIT_LIST_HEAD(&ret->b_assoc_buffers); ++ buffer_head_init_locks(ret); + preempt_disable(); + __this_cpu_inc(bh_accounting.nr); + recalc_bh_state(); +diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c +index a27fc8791551..791aecb7c1ac 100644 +--- a/fs/cifs/readdir.c ++++ b/fs/cifs/readdir.c +@@ -80,7 +80,7 @@ cifs_prime_dcache(struct dentry *parent, struct qstr *name, + struct inode *inode; + struct super_block *sb = parent->d_sb; + struct cifs_sb_info *cifs_sb = CIFS_SB(sb); +- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); ++ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); + + cifs_dbg(FYI, "%s: for %s\n", __func__, name->name); + +diff --git a/fs/dcache.c b/fs/dcache.c +index 4485a48f4091..691039a6a872 100644 +--- a/fs/dcache.c ++++ b/fs/dcache.c +@@ -19,6 +19,7 @@ + #include + #include + #include ++#include + #include + #include + #include +@@ -750,6 +751,8 @@ static inline bool fast_dput(struct dentry *dentry) + */ + void dput(struct dentry *dentry) + { ++ struct dentry *parent; ++ + if (unlikely(!dentry)) + return; + +@@ -788,9 +791,18 @@ void dput(struct dentry *dentry) + return; + + kill_it: +- dentry = dentry_kill(dentry); +- if (dentry) { +- cond_resched(); ++ parent = dentry_kill(dentry); ++ if (parent) { ++ int r; ++ ++ if (parent == dentry) { ++ /* the task with the highest priority won't schedule */ ++ r = cond_resched(); ++ if (!r) ++ cpu_chill(); ++ } else { ++ dentry = parent; ++ } + goto repeat; + } + } +@@ -2324,7 +2336,7 @@ void d_delete(struct dentry * dentry) + if (dentry->d_lockref.count == 1) { + if (!spin_trylock(&inode->i_lock)) { + spin_unlock(&dentry->d_lock); +- cpu_relax(); ++ cpu_chill(); + goto again; + } + dentry->d_flags &= ~DCACHE_CANT_MOUNT; +@@ -2384,21 +2396,24 @@ static inline void end_dir_add(struct inode *dir, unsigned n) + + static void d_wait_lookup(struct dentry *dentry) + { +- if (d_in_lookup(dentry)) { +- DECLARE_WAITQUEUE(wait, current); +- add_wait_queue(dentry->d_wait, &wait); +- do { +- set_current_state(TASK_UNINTERRUPTIBLE); +- spin_unlock(&dentry->d_lock); +- schedule(); +- spin_lock(&dentry->d_lock); +- } while (d_in_lookup(dentry)); +- } ++ struct swait_queue __wait; ++ ++ if (!d_in_lookup(dentry)) ++ return; ++ ++ INIT_LIST_HEAD(&__wait.task_list); ++ do { ++ prepare_to_swait(dentry->d_wait, &__wait, TASK_UNINTERRUPTIBLE); ++ spin_unlock(&dentry->d_lock); ++ schedule(); ++ spin_lock(&dentry->d_lock); ++ } while (d_in_lookup(dentry)); ++ finish_swait(dentry->d_wait, &__wait); + } + + struct dentry *d_alloc_parallel(struct dentry *parent, + const struct qstr *name, +- wait_queue_head_t *wq) ++ struct swait_queue_head *wq) + { + unsigned int hash = name->hash; + struct hlist_bl_head *b = in_lookup_hash(parent, hash); +@@ -2507,7 +2522,7 @@ void __d_lookup_done(struct dentry *dentry) + hlist_bl_lock(b); + dentry->d_flags &= ~DCACHE_PAR_LOOKUP; + __hlist_bl_del(&dentry->d_u.d_in_lookup_hash); +- wake_up_all(dentry->d_wait); ++ swake_up_all(dentry->d_wait); + dentry->d_wait = NULL; + hlist_bl_unlock(b); + INIT_HLIST_NODE(&dentry->d_u.d_alias); +@@ -3604,6 +3619,11 @@ EXPORT_SYMBOL(d_genocide); + + void __init vfs_caches_init_early(void) + { ++ int i; ++ ++ for (i = 0; i < ARRAY_SIZE(in_lookup_hashtable); i++) ++ INIT_HLIST_BL_HEAD(&in_lookup_hashtable[i]); ++ + dcache_init_early(); + inode_init_early(); + } +diff --git a/fs/eventpoll.c b/fs/eventpoll.c +index 10db91218933..42af0a06f657 100644 +--- a/fs/eventpoll.c ++++ b/fs/eventpoll.c +@@ -510,12 +510,12 @@ static int ep_poll_wakeup_proc(void *priv, void *cookie, int call_nests) + */ + static void ep_poll_safewake(wait_queue_head_t *wq) + { +- int this_cpu = get_cpu(); ++ int this_cpu = get_cpu_light(); + + ep_call_nested(&poll_safewake_ncalls, EP_MAX_NESTS, + ep_poll_wakeup_proc, NULL, wq, (void *) (long) this_cpu); + +- put_cpu(); ++ put_cpu_light(); + } + + static void ep_remove_wait_queue(struct eppoll_entry *pwq) +diff --git a/fs/exec.c b/fs/exec.c +index 67e86571685a..fe14cdd84016 100644 +--- a/fs/exec.c ++++ b/fs/exec.c +@@ -1017,12 +1017,14 @@ static int exec_mmap(struct mm_struct *mm) + } + } + task_lock(tsk); ++ preempt_disable_rt(); + active_mm = tsk->active_mm; + tsk->mm = mm; + tsk->active_mm = mm; + activate_mm(active_mm, mm); + tsk->mm->vmacache_seqnum = 0; + vmacache_flush(tsk); ++ preempt_enable_rt(); + task_unlock(tsk); + if (old_mm) { + up_read(&old_mm->mmap_sem); +diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c +index 642c57b8de7b..8494b9308333 100644 +--- a/fs/fuse/dir.c ++++ b/fs/fuse/dir.c +@@ -1191,7 +1191,7 @@ static int fuse_direntplus_link(struct file *file, + struct inode *dir = d_inode(parent); + struct fuse_conn *fc; + struct inode *inode; +- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); ++ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); + + if (!o->nodeid) { + /* +diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c +index 684996c8a3a4..6e18a06aaabe 100644 +--- a/fs/jbd2/checkpoint.c ++++ b/fs/jbd2/checkpoint.c +@@ -116,6 +116,8 @@ void __jbd2_log_wait_for_space(journal_t *journal) + nblocks = jbd2_space_needed(journal); + while (jbd2_log_space_left(journal) < nblocks) { + write_unlock(&journal->j_state_lock); ++ if (current->plug) ++ io_schedule(); + mutex_lock(&journal->j_checkpoint_mutex); + + /* +diff --git a/fs/locks.c b/fs/locks.c +index 22c5b4aa4961..269c6a44449a 100644 +--- a/fs/locks.c ++++ b/fs/locks.c +@@ -935,7 +935,7 @@ static int flock_lock_inode(struct inode *inode, struct file_lock *request) + return -ENOMEM; + } + +- percpu_down_read_preempt_disable(&file_rwsem); ++ percpu_down_read(&file_rwsem); + spin_lock(&ctx->flc_lock); + if (request->fl_flags & FL_ACCESS) + goto find_conflict; +@@ -976,7 +976,7 @@ static int flock_lock_inode(struct inode *inode, struct file_lock *request) + + out: + spin_unlock(&ctx->flc_lock); +- percpu_up_read_preempt_enable(&file_rwsem); ++ percpu_up_read(&file_rwsem); + if (new_fl) + locks_free_lock(new_fl); + locks_dispose_list(&dispose); +@@ -1013,7 +1013,7 @@ static int posix_lock_inode(struct inode *inode, struct file_lock *request, + new_fl2 = locks_alloc_lock(); + } + +- percpu_down_read_preempt_disable(&file_rwsem); ++ percpu_down_read(&file_rwsem); + spin_lock(&ctx->flc_lock); + /* + * New lock request. Walk all POSIX locks and look for conflicts. If +@@ -1185,7 +1185,7 @@ static int posix_lock_inode(struct inode *inode, struct file_lock *request, + } + out: + spin_unlock(&ctx->flc_lock); +- percpu_up_read_preempt_enable(&file_rwsem); ++ percpu_up_read(&file_rwsem); + /* + * Free any unused locks. + */ +@@ -1460,7 +1460,7 @@ int __break_lease(struct inode *inode, unsigned int mode, unsigned int type) + return error; + } + +- percpu_down_read_preempt_disable(&file_rwsem); ++ percpu_down_read(&file_rwsem); + spin_lock(&ctx->flc_lock); + + time_out_leases(inode, &dispose); +@@ -1512,13 +1512,13 @@ int __break_lease(struct inode *inode, unsigned int mode, unsigned int type) + locks_insert_block(fl, new_fl); + trace_break_lease_block(inode, new_fl); + spin_unlock(&ctx->flc_lock); +- percpu_up_read_preempt_enable(&file_rwsem); ++ percpu_up_read(&file_rwsem); + + locks_dispose_list(&dispose); + error = wait_event_interruptible_timeout(new_fl->fl_wait, + !new_fl->fl_next, break_time); + +- percpu_down_read_preempt_disable(&file_rwsem); ++ percpu_down_read(&file_rwsem); + spin_lock(&ctx->flc_lock); + trace_break_lease_unblock(inode, new_fl); + locks_delete_block(new_fl); +@@ -1535,7 +1535,7 @@ int __break_lease(struct inode *inode, unsigned int mode, unsigned int type) + } + out: + spin_unlock(&ctx->flc_lock); +- percpu_up_read_preempt_enable(&file_rwsem); ++ percpu_up_read(&file_rwsem); + locks_dispose_list(&dispose); + locks_free_lock(new_fl); + return error; +@@ -1609,7 +1609,7 @@ int fcntl_getlease(struct file *filp) + + ctx = smp_load_acquire(&inode->i_flctx); + if (ctx && !list_empty_careful(&ctx->flc_lease)) { +- percpu_down_read_preempt_disable(&file_rwsem); ++ percpu_down_read(&file_rwsem); + spin_lock(&ctx->flc_lock); + time_out_leases(inode, &dispose); + list_for_each_entry(fl, &ctx->flc_lease, fl_list) { +@@ -1619,7 +1619,7 @@ int fcntl_getlease(struct file *filp) + break; + } + spin_unlock(&ctx->flc_lock); +- percpu_up_read_preempt_enable(&file_rwsem); ++ percpu_up_read(&file_rwsem); + + locks_dispose_list(&dispose); + } +@@ -1694,7 +1694,7 @@ generic_add_lease(struct file *filp, long arg, struct file_lock **flp, void **pr + return -EINVAL; + } + +- percpu_down_read_preempt_disable(&file_rwsem); ++ percpu_down_read(&file_rwsem); + spin_lock(&ctx->flc_lock); + time_out_leases(inode, &dispose); + error = check_conflicting_open(dentry, arg, lease->fl_flags); +@@ -1765,7 +1765,7 @@ generic_add_lease(struct file *filp, long arg, struct file_lock **flp, void **pr + lease->fl_lmops->lm_setup(lease, priv); + out: + spin_unlock(&ctx->flc_lock); +- percpu_up_read_preempt_enable(&file_rwsem); ++ percpu_up_read(&file_rwsem); + locks_dispose_list(&dispose); + if (is_deleg) + inode_unlock(inode); +@@ -1788,7 +1788,7 @@ static int generic_delete_lease(struct file *filp, void *owner) + return error; + } + +- percpu_down_read_preempt_disable(&file_rwsem); ++ percpu_down_read(&file_rwsem); + spin_lock(&ctx->flc_lock); + list_for_each_entry(fl, &ctx->flc_lease, fl_list) { + if (fl->fl_file == filp && +@@ -1801,7 +1801,7 @@ static int generic_delete_lease(struct file *filp, void *owner) + if (victim) + error = fl->fl_lmops->lm_change(victim, F_UNLCK, &dispose); + spin_unlock(&ctx->flc_lock); +- percpu_up_read_preempt_enable(&file_rwsem); ++ percpu_up_read(&file_rwsem); + locks_dispose_list(&dispose); + return error; + } +@@ -2532,13 +2532,13 @@ locks_remove_lease(struct file *filp, struct file_lock_context *ctx) + if (list_empty(&ctx->flc_lease)) + return; + +- percpu_down_read_preempt_disable(&file_rwsem); ++ percpu_down_read(&file_rwsem); + spin_lock(&ctx->flc_lock); + list_for_each_entry_safe(fl, tmp, &ctx->flc_lease, fl_list) + if (filp == fl->fl_file) + lease_modify(fl, F_UNLCK, &dispose); + spin_unlock(&ctx->flc_lock); +- percpu_up_read_preempt_enable(&file_rwsem); ++ percpu_up_read(&file_rwsem); + + locks_dispose_list(&dispose); + } +diff --git a/fs/namei.c b/fs/namei.c +index d5e5140c1045..150fbdd8e04c 100644 +--- a/fs/namei.c ++++ b/fs/namei.c +@@ -1626,7 +1626,7 @@ static struct dentry *lookup_slow(const struct qstr *name, + { + struct dentry *dentry = ERR_PTR(-ENOENT), *old; + struct inode *inode = dir->d_inode; +- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); ++ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); + + inode_lock_shared(inode); + /* Don't go there if it's already dead */ +@@ -3083,7 +3083,7 @@ static int lookup_open(struct nameidata *nd, struct path *path, + struct dentry *dentry; + int error, create_error = 0; + umode_t mode = op->mode; +- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); ++ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); + + if (unlikely(IS_DEADDIR(dir_inode))) + return -ENOENT; +diff --git a/fs/namespace.c b/fs/namespace.c +index 5e35057f07ac..843d274ba167 100644 +--- a/fs/namespace.c ++++ b/fs/namespace.c +@@ -14,6 +14,7 @@ + #include + #include + #include ++#include + #include + #include + #include /* init_rootfs */ +@@ -356,8 +357,11 @@ int __mnt_want_write(struct vfsmount *m) + * incremented count after it has set MNT_WRITE_HOLD. + */ + smp_mb(); +- while (ACCESS_ONCE(mnt->mnt.mnt_flags) & MNT_WRITE_HOLD) +- cpu_relax(); ++ while (ACCESS_ONCE(mnt->mnt.mnt_flags) & MNT_WRITE_HOLD) { ++ preempt_enable(); ++ cpu_chill(); ++ preempt_disable(); ++ } + /* + * After the slowpath clears MNT_WRITE_HOLD, mnt_is_readonly will + * be set to match its requirements. So we must not load that until +diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c +index dff600ae0d74..d726d2e09353 100644 +--- a/fs/nfs/delegation.c ++++ b/fs/nfs/delegation.c +@@ -150,11 +150,11 @@ static int nfs_delegation_claim_opens(struct inode *inode, + sp = state->owner; + /* Block nfs4_proc_unlck */ + mutex_lock(&sp->so_delegreturn_mutex); +- seq = raw_seqcount_begin(&sp->so_reclaim_seqcount); ++ seq = read_seqbegin(&sp->so_reclaim_seqlock); + err = nfs4_open_delegation_recall(ctx, state, stateid, type); + if (!err) + err = nfs_delegation_claim_locks(ctx, state, stateid); +- if (!err && read_seqcount_retry(&sp->so_reclaim_seqcount, seq)) ++ if (!err && read_seqretry(&sp->so_reclaim_seqlock, seq)) + err = -EAGAIN; + mutex_unlock(&sp->so_delegreturn_mutex); + put_nfs_open_context(ctx); +diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c +index 53e02b8bd9bd..a66e7d77cfbb 100644 +--- a/fs/nfs/dir.c ++++ b/fs/nfs/dir.c +@@ -485,7 +485,7 @@ static + void nfs_prime_dcache(struct dentry *parent, struct nfs_entry *entry) + { + struct qstr filename = QSTR_INIT(entry->name, entry->len); +- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); ++ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); + struct dentry *dentry; + struct dentry *alias; + struct inode *dir = d_inode(parent); +@@ -1487,7 +1487,7 @@ int nfs_atomic_open(struct inode *dir, struct dentry *dentry, + struct file *file, unsigned open_flags, + umode_t mode, int *opened) + { +- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); ++ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); + struct nfs_open_context *ctx; + struct dentry *res; + struct iattr attr = { .ia_valid = ATTR_OPEN }; +@@ -1802,7 +1802,11 @@ int nfs_rmdir(struct inode *dir, struct dentry *dentry) + + trace_nfs_rmdir_enter(dir, dentry); + if (d_really_is_positive(dentry)) { ++#ifdef CONFIG_PREEMPT_RT_BASE ++ down(&NFS_I(d_inode(dentry))->rmdir_sem); ++#else + down_write(&NFS_I(d_inode(dentry))->rmdir_sem); ++#endif + error = NFS_PROTO(dir)->rmdir(dir, &dentry->d_name); + /* Ensure the VFS deletes this inode */ + switch (error) { +@@ -1812,7 +1816,11 @@ int nfs_rmdir(struct inode *dir, struct dentry *dentry) + case -ENOENT: + nfs_dentry_handle_enoent(dentry); + } ++#ifdef CONFIG_PREEMPT_RT_BASE ++ up(&NFS_I(d_inode(dentry))->rmdir_sem); ++#else + up_write(&NFS_I(d_inode(dentry))->rmdir_sem); ++#endif + } else + error = NFS_PROTO(dir)->rmdir(dir, &dentry->d_name); + trace_nfs_rmdir_exit(dir, dentry, error); +diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c +index bf4ec5ecc97e..36cd5fc9192c 100644 +--- a/fs/nfs/inode.c ++++ b/fs/nfs/inode.c +@@ -1957,7 +1957,11 @@ static void init_once(void *foo) + nfsi->nrequests = 0; + nfsi->commit_info.ncommit = 0; + atomic_set(&nfsi->commit_info.rpcs_out, 0); ++#ifdef CONFIG_PREEMPT_RT_BASE ++ sema_init(&nfsi->rmdir_sem, 1); ++#else + init_rwsem(&nfsi->rmdir_sem); ++#endif + nfs4_init_once(nfsi); + } + +diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h +index 1452177c822d..f43b01d54c59 100644 +--- a/fs/nfs/nfs4_fs.h ++++ b/fs/nfs/nfs4_fs.h +@@ -111,7 +111,7 @@ struct nfs4_state_owner { + unsigned long so_flags; + struct list_head so_states; + struct nfs_seqid_counter so_seqid; +- seqcount_t so_reclaim_seqcount; ++ seqlock_t so_reclaim_seqlock; + struct mutex so_delegreturn_mutex; + }; + +diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c +index fc9b04941739..7c9bc1c7efe7 100644 +--- a/fs/nfs/nfs4proc.c ++++ b/fs/nfs/nfs4proc.c +@@ -2697,7 +2697,7 @@ static int _nfs4_open_and_get_state(struct nfs4_opendata *opendata, + unsigned int seq; + int ret; + +- seq = raw_seqcount_begin(&sp->so_reclaim_seqcount); ++ seq = raw_seqcount_begin(&sp->so_reclaim_seqlock.seqcount); + + ret = _nfs4_proc_open(opendata); + if (ret != 0) +@@ -2735,7 +2735,7 @@ static int _nfs4_open_and_get_state(struct nfs4_opendata *opendata, + + if (d_inode(dentry) == state->inode) { + nfs_inode_attach_open_context(ctx); +- if (read_seqcount_retry(&sp->so_reclaim_seqcount, seq)) ++ if (read_seqretry(&sp->so_reclaim_seqlock, seq)) + nfs4_schedule_stateid_recovery(server, state); + } + out: +diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c +index 0959c9661662..dabd834d7686 100644 +--- a/fs/nfs/nfs4state.c ++++ b/fs/nfs/nfs4state.c +@@ -488,7 +488,7 @@ nfs4_alloc_state_owner(struct nfs_server *server, + nfs4_init_seqid_counter(&sp->so_seqid); + atomic_set(&sp->so_count, 1); + INIT_LIST_HEAD(&sp->so_lru); +- seqcount_init(&sp->so_reclaim_seqcount); ++ seqlock_init(&sp->so_reclaim_seqlock); + mutex_init(&sp->so_delegreturn_mutex); + return sp; + } +@@ -1497,8 +1497,12 @@ static int nfs4_reclaim_open_state(struct nfs4_state_owner *sp, const struct nfs + * recovering after a network partition or a reboot from a + * server that doesn't support a grace period. + */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ write_seqlock(&sp->so_reclaim_seqlock); ++#else ++ write_seqcount_begin(&sp->so_reclaim_seqlock.seqcount); ++#endif + spin_lock(&sp->so_lock); +- raw_write_seqcount_begin(&sp->so_reclaim_seqcount); + restart: + list_for_each_entry(state, &sp->so_states, open_states) { + if (!test_and_clear_bit(ops->state_flag_bit, &state->flags)) +@@ -1567,14 +1571,20 @@ static int nfs4_reclaim_open_state(struct nfs4_state_owner *sp, const struct nfs + spin_lock(&sp->so_lock); + goto restart; + } +- raw_write_seqcount_end(&sp->so_reclaim_seqcount); + spin_unlock(&sp->so_lock); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ write_sequnlock(&sp->so_reclaim_seqlock); ++#else ++ write_seqcount_end(&sp->so_reclaim_seqlock.seqcount); ++#endif + return 0; + out_err: + nfs4_put_open_state(state); +- spin_lock(&sp->so_lock); +- raw_write_seqcount_end(&sp->so_reclaim_seqcount); +- spin_unlock(&sp->so_lock); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ write_sequnlock(&sp->so_reclaim_seqlock); ++#else ++ write_seqcount_end(&sp->so_reclaim_seqlock.seqcount); ++#endif + return status; + } + +diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c +index 191aa577dd1f..58990c8f52e0 100644 +--- a/fs/nfs/unlink.c ++++ b/fs/nfs/unlink.c +@@ -12,7 +12,7 @@ + #include + #include + #include +-#include ++#include + #include + #include + +@@ -51,6 +51,29 @@ static void nfs_async_unlink_done(struct rpc_task *task, void *calldata) + rpc_restart_call_prepare(task); + } + ++#ifdef CONFIG_PREEMPT_RT_BASE ++static void nfs_down_anon(struct semaphore *sema) ++{ ++ down(sema); ++} ++ ++static void nfs_up_anon(struct semaphore *sema) ++{ ++ up(sema); ++} ++ ++#else ++static void nfs_down_anon(struct rw_semaphore *rwsem) ++{ ++ down_read_non_owner(rwsem); ++} ++ ++static void nfs_up_anon(struct rw_semaphore *rwsem) ++{ ++ up_read_non_owner(rwsem); ++} ++#endif ++ + /** + * nfs_async_unlink_release - Release the sillydelete data. + * @task: rpc_task of the sillydelete +@@ -64,7 +87,7 @@ static void nfs_async_unlink_release(void *calldata) + struct dentry *dentry = data->dentry; + struct super_block *sb = dentry->d_sb; + +- up_read_non_owner(&NFS_I(d_inode(dentry->d_parent))->rmdir_sem); ++ nfs_up_anon(&NFS_I(d_inode(dentry->d_parent))->rmdir_sem); + d_lookup_done(dentry); + nfs_free_unlinkdata(data); + dput(dentry); +@@ -117,10 +140,10 @@ static int nfs_call_unlink(struct dentry *dentry, struct nfs_unlinkdata *data) + struct inode *dir = d_inode(dentry->d_parent); + struct dentry *alias; + +- down_read_non_owner(&NFS_I(dir)->rmdir_sem); ++ nfs_down_anon(&NFS_I(dir)->rmdir_sem); + alias = d_alloc_parallel(dentry->d_parent, &data->args.name, &data->wq); + if (IS_ERR(alias)) { +- up_read_non_owner(&NFS_I(dir)->rmdir_sem); ++ nfs_up_anon(&NFS_I(dir)->rmdir_sem); + return 0; + } + if (!d_in_lookup(alias)) { +@@ -142,7 +165,7 @@ static int nfs_call_unlink(struct dentry *dentry, struct nfs_unlinkdata *data) + ret = 0; + spin_unlock(&alias->d_lock); + dput(alias); +- up_read_non_owner(&NFS_I(dir)->rmdir_sem); ++ nfs_up_anon(&NFS_I(dir)->rmdir_sem); + /* + * If we'd displaced old cached devname, free it. At that + * point dentry is definitely not a root, so we won't need +@@ -182,7 +205,7 @@ nfs_async_unlink(struct dentry *dentry, const struct qstr *name) + goto out_free_name; + } + data->res.dir_attr = &data->dir_attr; +- init_waitqueue_head(&data->wq); ++ init_swait_queue_head(&data->wq); + + status = -EBUSY; + spin_lock(&dentry->d_lock); +diff --git a/fs/ntfs/aops.c b/fs/ntfs/aops.c +index fe251f187ff8..e89da4fb14c2 100644 +--- a/fs/ntfs/aops.c ++++ b/fs/ntfs/aops.c +@@ -92,13 +92,13 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) + ofs = 0; + if (file_ofs < init_size) + ofs = init_size - file_ofs; +- local_irq_save(flags); ++ local_irq_save_nort(flags); + kaddr = kmap_atomic(page); + memset(kaddr + bh_offset(bh) + ofs, 0, + bh->b_size - ofs); + flush_dcache_page(page); + kunmap_atomic(kaddr); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + } else { + clear_buffer_uptodate(bh); +@@ -107,8 +107,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) + "0x%llx.", (unsigned long long)bh->b_blocknr); + } + first = page_buffers(page); +- local_irq_save(flags); +- bit_spin_lock(BH_Uptodate_Lock, &first->b_state); ++ flags = bh_uptodate_lock_irqsave(first); + clear_buffer_async_read(bh); + unlock_buffer(bh); + tmp = bh; +@@ -123,8 +122,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) + } + tmp = tmp->b_this_page; + } while (tmp != bh); +- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); +- local_irq_restore(flags); ++ bh_uptodate_unlock_irqrestore(first, flags); + /* + * If none of the buffers had errors then we can set the page uptodate, + * but we first have to perform the post read mst fixups, if the +@@ -145,13 +143,13 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) + recs = PAGE_SIZE / rec_size; + /* Should have been verified before we got here... */ + BUG_ON(!recs); +- local_irq_save(flags); ++ local_irq_save_nort(flags); + kaddr = kmap_atomic(page); + for (i = 0; i < recs; i++) + post_read_mst_fixup((NTFS_RECORD*)(kaddr + + i * rec_size), rec_size); + kunmap_atomic(kaddr); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + flush_dcache_page(page); + if (likely(page_uptodate && !PageError(page))) + SetPageUptodate(page); +@@ -159,9 +157,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) + unlock_page(page); + return; + still_busy: +- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); +- local_irq_restore(flags); +- return; ++ bh_uptodate_unlock_irqrestore(first, flags); + } + + /** +diff --git a/fs/proc/base.c b/fs/proc/base.c +index ca651ac00660..41d9dc789285 100644 +--- a/fs/proc/base.c ++++ b/fs/proc/base.c +@@ -1834,7 +1834,7 @@ bool proc_fill_cache(struct file *file, struct dir_context *ctx, + + child = d_hash_and_lookup(dir, &qname); + if (!child) { +- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); ++ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); + child = d_alloc_parallel(dir, &qname, &wq); + if (IS_ERR(child)) + goto end_instantiate; +diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c +index d4e37acd4821..000cea46434a 100644 +--- a/fs/proc/proc_sysctl.c ++++ b/fs/proc/proc_sysctl.c +@@ -632,7 +632,7 @@ static bool proc_sys_fill_cache(struct file *file, + + child = d_lookup(dir, &qname); + if (!child) { +- DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq); ++ DECLARE_SWAIT_QUEUE_HEAD_ONSTACK(wq); + child = d_alloc_parallel(dir, &qname, &wq); + if (IS_ERR(child)) + return false; +diff --git a/fs/timerfd.c b/fs/timerfd.c +index ab8dd1538381..5580853f57dd 100644 +--- a/fs/timerfd.c ++++ b/fs/timerfd.c +@@ -471,7 +471,10 @@ static int do_timerfd_settime(int ufd, int flags, + break; + } + spin_unlock_irq(&ctx->wqh.lock); +- cpu_relax(); ++ if (isalarm(ctx)) ++ hrtimer_wait_for_timer(&ctx->t.alarm.timer); ++ else ++ hrtimer_wait_for_timer(&ctx->t.tmr); + } + + /* +diff --git a/include/acpi/platform/aclinux.h b/include/acpi/platform/aclinux.h +index e861a24f06f2..b5c97d3059c7 100644 +--- a/include/acpi/platform/aclinux.h ++++ b/include/acpi/platform/aclinux.h +@@ -133,6 +133,7 @@ + + #define acpi_cache_t struct kmem_cache + #define acpi_spinlock spinlock_t * ++#define acpi_raw_spinlock raw_spinlock_t * + #define acpi_cpu_flags unsigned long + + /* Use native linux version of acpi_os_allocate_zeroed */ +@@ -151,6 +152,20 @@ + #define ACPI_USE_ALTERNATE_PROTOTYPE_acpi_os_get_thread_id + #define ACPI_USE_ALTERNATE_PROTOTYPE_acpi_os_create_lock + ++#define acpi_os_create_raw_lock(__handle) \ ++({ \ ++ raw_spinlock_t *lock = ACPI_ALLOCATE(sizeof(*lock)); \ ++ \ ++ if (lock) { \ ++ *(__handle) = lock; \ ++ raw_spin_lock_init(*(__handle)); \ ++ } \ ++ lock ? AE_OK : AE_NO_MEMORY; \ ++ }) ++ ++#define acpi_os_delete_raw_lock(__handle) kfree(__handle) ++ ++ + /* + * OSL interfaces used by debugger/disassembler + */ +diff --git a/include/asm-generic/bug.h b/include/asm-generic/bug.h +index 6f96247226a4..fa53a21263c2 100644 +--- a/include/asm-generic/bug.h ++++ b/include/asm-generic/bug.h +@@ -215,6 +215,20 @@ void __warn(const char *file, int line, void *caller, unsigned taint, + # define WARN_ON_SMP(x) ({0;}) + #endif + ++#ifdef CONFIG_PREEMPT_RT_BASE ++# define BUG_ON_RT(c) BUG_ON(c) ++# define BUG_ON_NONRT(c) do { } while (0) ++# define WARN_ON_RT(condition) WARN_ON(condition) ++# define WARN_ON_NONRT(condition) do { } while (0) ++# define WARN_ON_ONCE_NONRT(condition) do { } while (0) ++#else ++# define BUG_ON_RT(c) do { } while (0) ++# define BUG_ON_NONRT(c) BUG_ON(c) ++# define WARN_ON_RT(condition) do { } while (0) ++# define WARN_ON_NONRT(condition) WARN_ON(condition) ++# define WARN_ON_ONCE_NONRT(condition) WARN_ON_ONCE(condition) ++#endif ++ + #endif /* __ASSEMBLY__ */ + + #endif +diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h +index 535ab2e13d2e..cfc246899473 100644 +--- a/include/linux/blk-mq.h ++++ b/include/linux/blk-mq.h +@@ -209,7 +209,7 @@ static inline u16 blk_mq_unique_tag_to_tag(u32 unique_tag) + return unique_tag & BLK_MQ_UNIQUE_TAG_MASK; + } + +- ++void __blk_mq_complete_request_remote_work(struct work_struct *work); + int blk_mq_request_started(struct request *rq); + void blk_mq_start_request(struct request *rq); + void blk_mq_end_request(struct request *rq, int error); +diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h +index f6a816129856..ec7a4676f8a8 100644 +--- a/include/linux/blkdev.h ++++ b/include/linux/blkdev.h +@@ -89,6 +89,7 @@ struct request { + struct list_head queuelist; + union { + struct call_single_data csd; ++ struct work_struct work; + u64 fifo_time; + }; + +@@ -467,7 +468,7 @@ struct request_queue { + struct throtl_data *td; + #endif + struct rcu_head rcu_head; +- wait_queue_head_t mq_freeze_wq; ++ struct swait_queue_head mq_freeze_wq; + struct percpu_ref q_usage_counter; + struct list_head all_q_node; + +diff --git a/include/linux/bottom_half.h b/include/linux/bottom_half.h +index 8fdcb783197d..d07dbeec7bc1 100644 +--- a/include/linux/bottom_half.h ++++ b/include/linux/bottom_half.h +@@ -3,6 +3,39 @@ + + #include + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ ++extern void __local_bh_disable(void); ++extern void _local_bh_enable(void); ++extern void __local_bh_enable(void); ++ ++static inline void local_bh_disable(void) ++{ ++ __local_bh_disable(); ++} ++ ++static inline void __local_bh_disable_ip(unsigned long ip, unsigned int cnt) ++{ ++ __local_bh_disable(); ++} ++ ++static inline void local_bh_enable(void) ++{ ++ __local_bh_enable(); ++} ++ ++static inline void __local_bh_enable_ip(unsigned long ip, unsigned int cnt) ++{ ++ __local_bh_enable(); ++} ++ ++static inline void local_bh_enable_ip(unsigned long ip) ++{ ++ __local_bh_enable(); ++} ++ ++#else ++ + #ifdef CONFIG_TRACE_IRQFLAGS + extern void __local_bh_disable_ip(unsigned long ip, unsigned int cnt); + #else +@@ -30,5 +63,6 @@ static inline void local_bh_enable(void) + { + __local_bh_enable_ip(_THIS_IP_, SOFTIRQ_DISABLE_OFFSET); + } ++#endif + + #endif /* _LINUX_BH_H */ +diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h +index ebbacd14d450..be5e87f6360a 100644 +--- a/include/linux/buffer_head.h ++++ b/include/linux/buffer_head.h +@@ -75,8 +75,50 @@ struct buffer_head { + struct address_space *b_assoc_map; /* mapping this buffer is + associated with */ + atomic_t b_count; /* users using this buffer_head */ ++#ifdef CONFIG_PREEMPT_RT_BASE ++ spinlock_t b_uptodate_lock; ++#if IS_ENABLED(CONFIG_JBD2) ++ spinlock_t b_state_lock; ++ spinlock_t b_journal_head_lock; ++#endif ++#endif + }; + ++static inline unsigned long bh_uptodate_lock_irqsave(struct buffer_head *bh) ++{ ++ unsigned long flags; ++ ++#ifndef CONFIG_PREEMPT_RT_BASE ++ local_irq_save(flags); ++ bit_spin_lock(BH_Uptodate_Lock, &bh->b_state); ++#else ++ spin_lock_irqsave(&bh->b_uptodate_lock, flags); ++#endif ++ return flags; ++} ++ ++static inline void ++bh_uptodate_unlock_irqrestore(struct buffer_head *bh, unsigned long flags) ++{ ++#ifndef CONFIG_PREEMPT_RT_BASE ++ bit_spin_unlock(BH_Uptodate_Lock, &bh->b_state); ++ local_irq_restore(flags); ++#else ++ spin_unlock_irqrestore(&bh->b_uptodate_lock, flags); ++#endif ++} ++ ++static inline void buffer_head_init_locks(struct buffer_head *bh) ++{ ++#ifdef CONFIG_PREEMPT_RT_BASE ++ spin_lock_init(&bh->b_uptodate_lock); ++#if IS_ENABLED(CONFIG_JBD2) ++ spin_lock_init(&bh->b_state_lock); ++ spin_lock_init(&bh->b_journal_head_lock); ++#endif ++#endif ++} ++ + /* + * macro tricks to expand the set_buffer_foo(), clear_buffer_foo() + * and buffer_foo() functions. +diff --git a/include/linux/cgroup-defs.h b/include/linux/cgroup-defs.h +index 5b17de62c962..56027cc01a56 100644 +--- a/include/linux/cgroup-defs.h ++++ b/include/linux/cgroup-defs.h +@@ -16,6 +16,7 @@ + #include + #include + #include ++#include + + #ifdef CONFIG_CGROUPS + +@@ -137,6 +138,7 @@ struct cgroup_subsys_state { + /* percpu_ref killing and RCU release */ + struct rcu_head rcu_head; + struct work_struct destroy_work; ++ struct swork_event destroy_swork; + }; + + /* +diff --git a/include/linux/completion.h b/include/linux/completion.h +index 5d5aaae3af43..3bca1590e29f 100644 +--- a/include/linux/completion.h ++++ b/include/linux/completion.h +@@ -7,8 +7,7 @@ + * Atomic wait-for-completion handler data structures. + * See kernel/sched/completion.c for details. + */ +- +-#include ++#include + + /* + * struct completion - structure used to maintain state for a "completion" +@@ -24,11 +23,11 @@ + */ + struct completion { + unsigned int done; +- wait_queue_head_t wait; ++ struct swait_queue_head wait; + }; + + #define COMPLETION_INITIALIZER(work) \ +- { 0, __WAIT_QUEUE_HEAD_INITIALIZER((work).wait) } ++ { 0, __SWAIT_QUEUE_HEAD_INITIALIZER((work).wait) } + + #define COMPLETION_INITIALIZER_ONSTACK(work) \ + ({ init_completion(&work); work; }) +@@ -73,7 +72,7 @@ struct completion { + static inline void init_completion(struct completion *x) + { + x->done = 0; +- init_waitqueue_head(&x->wait); ++ init_swait_queue_head(&x->wait); + } + + /** +diff --git a/include/linux/cpu.h b/include/linux/cpu.h +index e571128ad99a..5e52d28c20c1 100644 +--- a/include/linux/cpu.h ++++ b/include/linux/cpu.h +@@ -182,6 +182,8 @@ extern void get_online_cpus(void); + extern void put_online_cpus(void); + extern void cpu_hotplug_disable(void); + extern void cpu_hotplug_enable(void); ++extern void pin_current_cpu(void); ++extern void unpin_current_cpu(void); + #define hotcpu_notifier(fn, pri) cpu_notifier(fn, pri) + #define __hotcpu_notifier(fn, pri) __cpu_notifier(fn, pri) + #define register_hotcpu_notifier(nb) register_cpu_notifier(nb) +@@ -199,6 +201,8 @@ static inline void cpu_hotplug_done(void) {} + #define put_online_cpus() do { } while (0) + #define cpu_hotplug_disable() do { } while (0) + #define cpu_hotplug_enable() do { } while (0) ++static inline void pin_current_cpu(void) { } ++static inline void unpin_current_cpu(void) { } + #define hotcpu_notifier(fn, pri) do { (void)(fn); } while (0) + #define __hotcpu_notifier(fn, pri) do { (void)(fn); } while (0) + /* These aren't inline functions due to a GCC bug. */ +diff --git a/include/linux/dcache.h b/include/linux/dcache.h +index 5beed7b30561..61cab7ef458e 100644 +--- a/include/linux/dcache.h ++++ b/include/linux/dcache.h +@@ -11,6 +11,7 @@ + #include + #include + #include ++#include + + struct path; + struct vfsmount; +@@ -100,7 +101,7 @@ struct dentry { + + union { + struct list_head d_lru; /* LRU list */ +- wait_queue_head_t *d_wait; /* in-lookup ones only */ ++ struct swait_queue_head *d_wait; /* in-lookup ones only */ + }; + struct list_head d_child; /* child of parent list */ + struct list_head d_subdirs; /* our children */ +@@ -230,7 +231,7 @@ extern void d_set_d_op(struct dentry *dentry, const struct dentry_operations *op + extern struct dentry * d_alloc(struct dentry *, const struct qstr *); + extern struct dentry * d_alloc_pseudo(struct super_block *, const struct qstr *); + extern struct dentry * d_alloc_parallel(struct dentry *, const struct qstr *, +- wait_queue_head_t *); ++ struct swait_queue_head *); + extern struct dentry * d_splice_alias(struct inode *, struct dentry *); + extern struct dentry * d_add_ci(struct dentry *, struct inode *, struct qstr *); + extern struct dentry * d_exact_alias(struct dentry *, struct inode *); +diff --git a/include/linux/delay.h b/include/linux/delay.h +index a6ecb34cf547..37caab306336 100644 +--- a/include/linux/delay.h ++++ b/include/linux/delay.h +@@ -52,4 +52,10 @@ static inline void ssleep(unsigned int seconds) + msleep(seconds * 1000); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++extern void cpu_chill(void); ++#else ++# define cpu_chill() cpu_relax() ++#endif ++ + #endif /* defined(_LINUX_DELAY_H) */ +diff --git a/include/linux/highmem.h b/include/linux/highmem.h +index bb3f3297062a..a117a33ef72c 100644 +--- a/include/linux/highmem.h ++++ b/include/linux/highmem.h +@@ -7,6 +7,7 @@ + #include + #include + #include ++#include + + #include + +@@ -65,7 +66,7 @@ static inline void kunmap(struct page *page) + + static inline void *kmap_atomic(struct page *page) + { +- preempt_disable(); ++ preempt_disable_nort(); + pagefault_disable(); + return page_address(page); + } +@@ -74,7 +75,7 @@ static inline void *kmap_atomic(struct page *page) + static inline void __kunmap_atomic(void *addr) + { + pagefault_enable(); +- preempt_enable(); ++ preempt_enable_nort(); + } + + #define kmap_atomic_pfn(pfn) kmap_atomic(pfn_to_page(pfn)) +@@ -86,32 +87,51 @@ static inline void __kunmap_atomic(void *addr) + + #if defined(CONFIG_HIGHMEM) || defined(CONFIG_X86_32) + ++#ifndef CONFIG_PREEMPT_RT_FULL + DECLARE_PER_CPU(int, __kmap_atomic_idx); ++#endif + + static inline int kmap_atomic_idx_push(void) + { ++#ifndef CONFIG_PREEMPT_RT_FULL + int idx = __this_cpu_inc_return(__kmap_atomic_idx) - 1; + +-#ifdef CONFIG_DEBUG_HIGHMEM ++# ifdef CONFIG_DEBUG_HIGHMEM + WARN_ON_ONCE(in_irq() && !irqs_disabled()); + BUG_ON(idx >= KM_TYPE_NR); +-#endif ++# endif + return idx; ++#else ++ current->kmap_idx++; ++ BUG_ON(current->kmap_idx > KM_TYPE_NR); ++ return current->kmap_idx - 1; ++#endif + } + + static inline int kmap_atomic_idx(void) + { ++#ifndef CONFIG_PREEMPT_RT_FULL + return __this_cpu_read(__kmap_atomic_idx) - 1; ++#else ++ return current->kmap_idx - 1; ++#endif + } + + static inline void kmap_atomic_idx_pop(void) + { +-#ifdef CONFIG_DEBUG_HIGHMEM ++#ifndef CONFIG_PREEMPT_RT_FULL ++# ifdef CONFIG_DEBUG_HIGHMEM + int idx = __this_cpu_dec_return(__kmap_atomic_idx); + + BUG_ON(idx < 0); +-#else ++# else + __this_cpu_dec(__kmap_atomic_idx); ++# endif ++#else ++ current->kmap_idx--; ++# ifdef CONFIG_DEBUG_HIGHMEM ++ BUG_ON(current->kmap_idx < 0); ++# endif + #endif + } + +diff --git a/include/linux/hrtimer.h b/include/linux/hrtimer.h +index 5e00f80b1535..a34e10b55cde 100644 +--- a/include/linux/hrtimer.h ++++ b/include/linux/hrtimer.h +@@ -87,6 +87,9 @@ enum hrtimer_restart { + * @function: timer expiry callback function + * @base: pointer to the timer base (per cpu and per clock) + * @state: state information (See bit values above) ++ * @cb_entry: list entry to defer timers from hardirq context ++ * @irqsafe: timer can run in hardirq context ++ * @praecox: timer expiry time if expired at the time of programming + * @is_rel: Set if the timer was armed relative + * @start_pid: timer statistics field to store the pid of the task which + * started the timer +@@ -103,6 +106,11 @@ struct hrtimer { + enum hrtimer_restart (*function)(struct hrtimer *); + struct hrtimer_clock_base *base; + u8 state; ++ struct list_head cb_entry; ++ int irqsafe; ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ ktime_t praecox; ++#endif + u8 is_rel; + #ifdef CONFIG_TIMER_STATS + int start_pid; +@@ -123,11 +131,7 @@ struct hrtimer_sleeper { + struct task_struct *task; + }; + +-#ifdef CONFIG_64BIT + # define HRTIMER_CLOCK_BASE_ALIGN 64 +-#else +-# define HRTIMER_CLOCK_BASE_ALIGN 32 +-#endif + + /** + * struct hrtimer_clock_base - the timer base for a specific clock +@@ -136,6 +140,7 @@ struct hrtimer_sleeper { + * timer to a base on another cpu. + * @clockid: clock id for per_cpu support + * @active: red black tree root node for the active timers ++ * @expired: list head for deferred timers. + * @get_time: function to retrieve the current time of the clock + * @offset: offset of this clock to the monotonic base + */ +@@ -144,6 +149,7 @@ struct hrtimer_clock_base { + int index; + clockid_t clockid; + struct timerqueue_head active; ++ struct list_head expired; + ktime_t (*get_time)(void); + ktime_t offset; + } __attribute__((__aligned__(HRTIMER_CLOCK_BASE_ALIGN))); +@@ -187,6 +193,7 @@ struct hrtimer_cpu_base { + raw_spinlock_t lock; + seqcount_t seq; + struct hrtimer *running; ++ struct hrtimer *running_soft; + unsigned int cpu; + unsigned int active_bases; + unsigned int clock_was_set_seq; +@@ -203,6 +210,9 @@ struct hrtimer_cpu_base { + unsigned int nr_hangs; + unsigned int max_hang_time; + #endif ++#ifdef CONFIG_PREEMPT_RT_BASE ++ wait_queue_head_t wait; ++#endif + struct hrtimer_clock_base clock_base[HRTIMER_MAX_CLOCK_BASES]; + } ____cacheline_aligned; + +@@ -412,6 +422,13 @@ static inline void hrtimer_restart(struct hrtimer *timer) + hrtimer_start_expires(timer, HRTIMER_MODE_ABS); + } + ++/* Softirq preemption could deadlock timer removal */ ++#ifdef CONFIG_PREEMPT_RT_BASE ++ extern void hrtimer_wait_for_timer(const struct hrtimer *timer); ++#else ++# define hrtimer_wait_for_timer(timer) do { cpu_relax(); } while (0) ++#endif ++ + /* Query timers: */ + extern ktime_t __hrtimer_get_remaining(const struct hrtimer *timer, bool adjust); + +@@ -436,9 +453,15 @@ static inline int hrtimer_is_queued(struct hrtimer *timer) + * Helper function to check, whether the timer is running the callback + * function + */ +-static inline int hrtimer_callback_running(struct hrtimer *timer) ++static inline int hrtimer_callback_running(const struct hrtimer *timer) + { +- return timer->base->cpu_base->running == timer; ++ if (timer->base->cpu_base->running == timer) ++ return 1; ++#ifdef CONFIG_PREEMPT_RT_BASE ++ if (timer->base->cpu_base->running_soft == timer) ++ return 1; ++#endif ++ return 0; + } + + /* Forward a hrtimer so it expires after now: */ +diff --git a/include/linux/idr.h b/include/linux/idr.h +index 083d61e92706..5899796f50cb 100644 +--- a/include/linux/idr.h ++++ b/include/linux/idr.h +@@ -95,10 +95,14 @@ bool idr_is_empty(struct idr *idp); + * Each idr_preload() should be matched with an invocation of this + * function. See idr_preload() for details. + */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++void idr_preload_end(void); ++#else + static inline void idr_preload_end(void) + { + preempt_enable(); + } ++#endif + + /** + * idr_find - return pointer for given id +diff --git a/include/linux/init_task.h b/include/linux/init_task.h +index 325f649d77ff..a56e263f5005 100644 +--- a/include/linux/init_task.h ++++ b/include/linux/init_task.h +@@ -150,6 +150,12 @@ extern struct task_group root_task_group; + # define INIT_PERF_EVENTS(tsk) + #endif + ++#ifdef CONFIG_PREEMPT_RT_BASE ++# define INIT_TIMER_LIST .posix_timer_list = NULL, ++#else ++# define INIT_TIMER_LIST ++#endif ++ + #ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN + # define INIT_VTIME(tsk) \ + .vtime_seqcount = SEQCNT_ZERO(tsk.vtime_seqcount), \ +@@ -164,6 +170,7 @@ extern struct task_group root_task_group; + #ifdef CONFIG_RT_MUTEXES + # define INIT_RT_MUTEXES(tsk) \ + .pi_waiters = RB_ROOT, \ ++ .pi_top_task = NULL, \ + .pi_waiters_leftmost = NULL, + #else + # define INIT_RT_MUTEXES(tsk) +@@ -250,6 +257,7 @@ extern struct task_group root_task_group; + .cpu_timers = INIT_CPU_TIMERS(tsk.cpu_timers), \ + .pi_lock = __RAW_SPIN_LOCK_UNLOCKED(tsk.pi_lock), \ + .timer_slack_ns = 50000, /* 50 usec default slack */ \ ++ INIT_TIMER_LIST \ + .pids = { \ + [PIDTYPE_PID] = INIT_PID_LINK(PIDTYPE_PID), \ + [PIDTYPE_PGID] = INIT_PID_LINK(PIDTYPE_PGID), \ +diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h +index 72f0721f75e7..480972ae47d3 100644 +--- a/include/linux/interrupt.h ++++ b/include/linux/interrupt.h +@@ -14,6 +14,7 @@ + #include + #include + #include ++#include + + #include + #include +@@ -61,6 +62,7 @@ + * interrupt handler after suspending interrupts. For system + * wakeup devices users need to implement wakeup detection in + * their interrupt handlers. ++ * IRQF_NO_SOFTIRQ_CALL - Do not process softirqs in the irq thread context (RT) + */ + #define IRQF_SHARED 0x00000080 + #define IRQF_PROBE_SHARED 0x00000100 +@@ -74,6 +76,7 @@ + #define IRQF_NO_THREAD 0x00010000 + #define IRQF_EARLY_RESUME 0x00020000 + #define IRQF_COND_SUSPEND 0x00040000 ++#define IRQF_NO_SOFTIRQ_CALL 0x00080000 + + #define IRQF_TIMER (__IRQF_TIMER | IRQF_NO_SUSPEND | IRQF_NO_THREAD) + +@@ -196,7 +199,7 @@ extern void devm_free_irq(struct device *dev, unsigned int irq, void *dev_id); + #ifdef CONFIG_LOCKDEP + # define local_irq_enable_in_hardirq() do { } while (0) + #else +-# define local_irq_enable_in_hardirq() local_irq_enable() ++# define local_irq_enable_in_hardirq() local_irq_enable_nort() + #endif + + extern void disable_irq_nosync(unsigned int irq); +@@ -216,6 +219,7 @@ extern void resume_device_irqs(void); + * struct irq_affinity_notify - context for notification of IRQ affinity changes + * @irq: Interrupt to which notification applies + * @kref: Reference count, for internal use ++ * @swork: Swork item, for internal use + * @work: Work item, for internal use + * @notify: Function to be called on change. This will be + * called in process context. +@@ -227,7 +231,11 @@ extern void resume_device_irqs(void); + struct irq_affinity_notify { + unsigned int irq; + struct kref kref; ++#ifdef CONFIG_PREEMPT_RT_BASE ++ struct swork_event swork; ++#else + struct work_struct work; ++#endif + void (*notify)(struct irq_affinity_notify *, const cpumask_t *mask); + void (*release)(struct kref *ref); + }; +@@ -406,9 +414,13 @@ extern int irq_set_irqchip_state(unsigned int irq, enum irqchip_irq_state which, + bool state); + + #ifdef CONFIG_IRQ_FORCED_THREADING ++# ifndef CONFIG_PREEMPT_RT_BASE + extern bool force_irqthreads; ++# else ++# define force_irqthreads (true) ++# endif + #else +-#define force_irqthreads (0) ++#define force_irqthreads (false) + #endif + + #ifndef __ARCH_SET_SOFTIRQ_PENDING +@@ -465,9 +477,10 @@ struct softirq_action + void (*action)(struct softirq_action *); + }; + ++#ifndef CONFIG_PREEMPT_RT_FULL + asmlinkage void do_softirq(void); + asmlinkage void __do_softirq(void); +- ++static inline void thread_do_softirq(void) { do_softirq(); } + #ifdef __ARCH_HAS_DO_SOFTIRQ + void do_softirq_own_stack(void); + #else +@@ -476,13 +489,25 @@ static inline void do_softirq_own_stack(void) + __do_softirq(); + } + #endif ++#else ++extern void thread_do_softirq(void); ++#endif + + extern void open_softirq(int nr, void (*action)(struct softirq_action *)); + extern void softirq_init(void); + extern void __raise_softirq_irqoff(unsigned int nr); ++#ifdef CONFIG_PREEMPT_RT_FULL ++extern void __raise_softirq_irqoff_ksoft(unsigned int nr); ++#else ++static inline void __raise_softirq_irqoff_ksoft(unsigned int nr) ++{ ++ __raise_softirq_irqoff(nr); ++} ++#endif + + extern void raise_softirq_irqoff(unsigned int nr); + extern void raise_softirq(unsigned int nr); ++extern void softirq_check_pending_idle(void); + + DECLARE_PER_CPU(struct task_struct *, ksoftirqd); + +@@ -504,8 +529,9 @@ static inline struct task_struct *this_cpu_ksoftirqd(void) + to be executed on some cpu at least once after this. + * If the tasklet is already scheduled, but its execution is still not + started, it will be executed only once. +- * If this tasklet is already running on another CPU (or schedule is called +- from tasklet itself), it is rescheduled for later. ++ * If this tasklet is already running on another CPU, it is rescheduled ++ for later. ++ * Schedule must not be called from the tasklet itself (a lockup occurs) + * Tasklet is strictly serialized wrt itself, but not + wrt another tasklets. If client needs some intertask synchronization, + he makes it with spinlocks. +@@ -530,27 +556,36 @@ struct tasklet_struct name = { NULL, 0, ATOMIC_INIT(1), func, data } + enum + { + TASKLET_STATE_SCHED, /* Tasklet is scheduled for execution */ +- TASKLET_STATE_RUN /* Tasklet is running (SMP only) */ ++ TASKLET_STATE_RUN, /* Tasklet is running (SMP only) */ ++ TASKLET_STATE_PENDING /* Tasklet is pending */ + }; + +-#ifdef CONFIG_SMP ++#define TASKLET_STATEF_SCHED (1 << TASKLET_STATE_SCHED) ++#define TASKLET_STATEF_RUN (1 << TASKLET_STATE_RUN) ++#define TASKLET_STATEF_PENDING (1 << TASKLET_STATE_PENDING) ++ ++#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL) + static inline int tasklet_trylock(struct tasklet_struct *t) + { + return !test_and_set_bit(TASKLET_STATE_RUN, &(t)->state); + } + ++static inline int tasklet_tryunlock(struct tasklet_struct *t) ++{ ++ return cmpxchg(&t->state, TASKLET_STATEF_RUN, 0) == TASKLET_STATEF_RUN; ++} ++ + static inline void tasklet_unlock(struct tasklet_struct *t) + { + smp_mb__before_atomic(); + clear_bit(TASKLET_STATE_RUN, &(t)->state); + } + +-static inline void tasklet_unlock_wait(struct tasklet_struct *t) +-{ +- while (test_bit(TASKLET_STATE_RUN, &(t)->state)) { barrier(); } +-} ++extern void tasklet_unlock_wait(struct tasklet_struct *t); ++ + #else + #define tasklet_trylock(t) 1 ++#define tasklet_tryunlock(t) 1 + #define tasklet_unlock_wait(t) do { } while (0) + #define tasklet_unlock(t) do { } while (0) + #endif +@@ -599,12 +634,7 @@ static inline void tasklet_disable(struct tasklet_struct *t) + smp_mb(); + } + +-static inline void tasklet_enable(struct tasklet_struct *t) +-{ +- smp_mb__before_atomic(); +- atomic_dec(&t->count); +-} +- ++extern void tasklet_enable(struct tasklet_struct *t); + extern void tasklet_kill(struct tasklet_struct *t); + extern void tasklet_kill_immediate(struct tasklet_struct *t, unsigned int cpu); + extern void tasklet_init(struct tasklet_struct *t, +@@ -635,6 +665,12 @@ void tasklet_hrtimer_cancel(struct tasklet_hrtimer *ttimer) + tasklet_kill(&ttimer->tasklet); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++extern void softirq_early_init(void); ++#else ++static inline void softirq_early_init(void) { } ++#endif ++ + /* + * Autoprobing for irqs: + * +diff --git a/include/linux/irq.h b/include/linux/irq.h +index 39e3254e5769..8ebac94fbb9f 100644 +--- a/include/linux/irq.h ++++ b/include/linux/irq.h +@@ -72,6 +72,7 @@ enum irqchip_irq_state; + * IRQ_IS_POLLED - Always polled by another interrupt. Exclude + * it from the spurious interrupt detection + * mechanism and from core side polling. ++ * IRQ_NO_SOFTIRQ_CALL - No softirq processing in the irq thread context (RT) + * IRQ_DISABLE_UNLAZY - Disable lazy irq disable + */ + enum { +@@ -99,13 +100,14 @@ enum { + IRQ_PER_CPU_DEVID = (1 << 17), + IRQ_IS_POLLED = (1 << 18), + IRQ_DISABLE_UNLAZY = (1 << 19), ++ IRQ_NO_SOFTIRQ_CALL = (1 << 20), + }; + + #define IRQF_MODIFY_MASK \ + (IRQ_TYPE_SENSE_MASK | IRQ_NOPROBE | IRQ_NOREQUEST | \ + IRQ_NOAUTOEN | IRQ_MOVE_PCNTXT | IRQ_LEVEL | IRQ_NO_BALANCING | \ + IRQ_PER_CPU | IRQ_NESTED_THREAD | IRQ_NOTHREAD | IRQ_PER_CPU_DEVID | \ +- IRQ_IS_POLLED | IRQ_DISABLE_UNLAZY) ++ IRQ_IS_POLLED | IRQ_DISABLE_UNLAZY | IRQ_NO_SOFTIRQ_CALL) + + #define IRQ_NO_BALANCING_MASK (IRQ_PER_CPU | IRQ_NO_BALANCING) + +diff --git a/include/linux/irq_work.h b/include/linux/irq_work.h +index 47b9ebd4a74f..2543aab05daa 100644 +--- a/include/linux/irq_work.h ++++ b/include/linux/irq_work.h +@@ -16,6 +16,7 @@ + #define IRQ_WORK_BUSY 2UL + #define IRQ_WORK_FLAGS 3UL + #define IRQ_WORK_LAZY 4UL /* Doesn't want IPI, wait for tick */ ++#define IRQ_WORK_HARD_IRQ 8UL /* Run hard IRQ context, even on RT */ + + struct irq_work { + unsigned long flags; +@@ -51,4 +52,10 @@ static inline bool irq_work_needs_cpu(void) { return false; } + static inline void irq_work_run(void) { } + #endif + ++#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) ++void irq_work_tick_soft(void); ++#else ++static inline void irq_work_tick_soft(void) { } ++#endif ++ + #endif /* _LINUX_IRQ_WORK_H */ +diff --git a/include/linux/irqdesc.h b/include/linux/irqdesc.h +index c9be57931b58..eeeb540971ae 100644 +--- a/include/linux/irqdesc.h ++++ b/include/linux/irqdesc.h +@@ -66,6 +66,7 @@ struct irq_desc { + unsigned int irqs_unhandled; + atomic_t threads_handled; + int threads_handled_last; ++ u64 random_ip; + raw_spinlock_t lock; + struct cpumask *percpu_enabled; + const struct cpumask *percpu_affinity; +diff --git a/include/linux/irqflags.h b/include/linux/irqflags.h +index 5dd1272d1ab2..9b77034f7c5e 100644 +--- a/include/linux/irqflags.h ++++ b/include/linux/irqflags.h +@@ -25,8 +25,6 @@ + # define trace_softirqs_enabled(p) ((p)->softirqs_enabled) + # define trace_hardirq_enter() do { current->hardirq_context++; } while (0) + # define trace_hardirq_exit() do { current->hardirq_context--; } while (0) +-# define lockdep_softirq_enter() do { current->softirq_context++; } while (0) +-# define lockdep_softirq_exit() do { current->softirq_context--; } while (0) + # define INIT_TRACE_IRQFLAGS .softirqs_enabled = 1, + #else + # define trace_hardirqs_on() do { } while (0) +@@ -39,9 +37,15 @@ + # define trace_softirqs_enabled(p) 0 + # define trace_hardirq_enter() do { } while (0) + # define trace_hardirq_exit() do { } while (0) ++# define INIT_TRACE_IRQFLAGS ++#endif ++ ++#if defined(CONFIG_TRACE_IRQFLAGS) && !defined(CONFIG_PREEMPT_RT_FULL) ++# define lockdep_softirq_enter() do { current->softirq_context++; } while (0) ++# define lockdep_softirq_exit() do { current->softirq_context--; } while (0) ++#else + # define lockdep_softirq_enter() do { } while (0) + # define lockdep_softirq_exit() do { } while (0) +-# define INIT_TRACE_IRQFLAGS + #endif + + #if defined(CONFIG_IRQSOFF_TRACER) || \ +@@ -148,4 +152,23 @@ + + #define irqs_disabled_flags(flags) raw_irqs_disabled_flags(flags) + ++/* ++ * local_irq* variants depending on RT/!RT ++ */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define local_irq_disable_nort() do { } while (0) ++# define local_irq_enable_nort() do { } while (0) ++# define local_irq_save_nort(flags) local_save_flags(flags) ++# define local_irq_restore_nort(flags) (void)(flags) ++# define local_irq_disable_rt() local_irq_disable() ++# define local_irq_enable_rt() local_irq_enable() ++#else ++# define local_irq_disable_nort() local_irq_disable() ++# define local_irq_enable_nort() local_irq_enable() ++# define local_irq_save_nort(flags) local_irq_save(flags) ++# define local_irq_restore_nort(flags) local_irq_restore(flags) ++# define local_irq_disable_rt() do { } while (0) ++# define local_irq_enable_rt() do { } while (0) ++#endif ++ + #endif +diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h +index dfaa1f4dcb0c..d57dd06544a1 100644 +--- a/include/linux/jbd2.h ++++ b/include/linux/jbd2.h +@@ -347,32 +347,56 @@ static inline struct journal_head *bh2jh(struct buffer_head *bh) + + static inline void jbd_lock_bh_state(struct buffer_head *bh) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + bit_spin_lock(BH_State, &bh->b_state); ++#else ++ spin_lock(&bh->b_state_lock); ++#endif + } + + static inline int jbd_trylock_bh_state(struct buffer_head *bh) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + return bit_spin_trylock(BH_State, &bh->b_state); ++#else ++ return spin_trylock(&bh->b_state_lock); ++#endif + } + + static inline int jbd_is_locked_bh_state(struct buffer_head *bh) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + return bit_spin_is_locked(BH_State, &bh->b_state); ++#else ++ return spin_is_locked(&bh->b_state_lock); ++#endif + } + + static inline void jbd_unlock_bh_state(struct buffer_head *bh) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + bit_spin_unlock(BH_State, &bh->b_state); ++#else ++ spin_unlock(&bh->b_state_lock); ++#endif + } + + static inline void jbd_lock_bh_journal_head(struct buffer_head *bh) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + bit_spin_lock(BH_JournalHead, &bh->b_state); ++#else ++ spin_lock(&bh->b_journal_head_lock); ++#endif + } + + static inline void jbd_unlock_bh_journal_head(struct buffer_head *bh) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + bit_spin_unlock(BH_JournalHead, &bh->b_state); ++#else ++ spin_unlock(&bh->b_journal_head_lock); ++#endif + } + + #define J_ASSERT(assert) BUG_ON(!(assert)) +diff --git a/include/linux/kdb.h b/include/linux/kdb.h +index 410decacff8f..0861bebfc188 100644 +--- a/include/linux/kdb.h ++++ b/include/linux/kdb.h +@@ -167,6 +167,7 @@ extern __printf(2, 0) int vkdb_printf(enum kdb_msgsrc src, const char *fmt, + extern __printf(1, 2) int kdb_printf(const char *, ...); + typedef __printf(1, 2) int (*kdb_printf_t)(const char *, ...); + ++#define in_kdb_printk() (kdb_trap_printk) + extern void kdb_init(int level); + + /* Access to kdb specific polling devices */ +@@ -201,6 +202,7 @@ extern int kdb_register_flags(char *, kdb_func_t, char *, char *, + extern int kdb_unregister(char *); + #else /* ! CONFIG_KGDB_KDB */ + static inline __printf(1, 2) int kdb_printf(const char *fmt, ...) { return 0; } ++#define in_kdb_printk() (0) + static inline void kdb_init(int level) {} + static inline int kdb_register(char *cmd, kdb_func_t func, char *usage, + char *help, short minlen) { return 0; } +diff --git a/include/linux/kernel.h b/include/linux/kernel.h +index bc6ed52a39b9..7894d55e4998 100644 +--- a/include/linux/kernel.h ++++ b/include/linux/kernel.h +@@ -194,6 +194,9 @@ extern int _cond_resched(void); + */ + # define might_sleep() \ + do { __might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0) ++ ++# define might_sleep_no_state_check() \ ++ do { ___might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0) + # define sched_annotate_sleep() (current->task_state_change = 0) + #else + static inline void ___might_sleep(const char *file, int line, +@@ -201,6 +204,7 @@ extern int _cond_resched(void); + static inline void __might_sleep(const char *file, int line, + int preempt_offset) { } + # define might_sleep() do { might_resched(); } while (0) ++# define might_sleep_no_state_check() do { might_resched(); } while (0) + # define sched_annotate_sleep() do { } while (0) + #endif + +@@ -488,6 +492,7 @@ extern enum system_states { + SYSTEM_HALT, + SYSTEM_POWER_OFF, + SYSTEM_RESTART, ++ SYSTEM_SUSPEND, + } system_state; + + #define TAINT_PROPRIETARY_MODULE 0 +diff --git a/include/linux/list_bl.h b/include/linux/list_bl.h +index cb483305e1f5..4e5062316bb6 100644 +--- a/include/linux/list_bl.h ++++ b/include/linux/list_bl.h +@@ -2,6 +2,7 @@ + #define _LINUX_LIST_BL_H + + #include ++#include + #include + + /* +@@ -32,13 +33,24 @@ + + struct hlist_bl_head { + struct hlist_bl_node *first; ++#ifdef CONFIG_PREEMPT_RT_BASE ++ raw_spinlock_t lock; ++#endif + }; + + struct hlist_bl_node { + struct hlist_bl_node *next, **pprev; + }; +-#define INIT_HLIST_BL_HEAD(ptr) \ +- ((ptr)->first = NULL) ++ ++#ifdef CONFIG_PREEMPT_RT_BASE ++#define INIT_HLIST_BL_HEAD(h) \ ++do { \ ++ (h)->first = NULL; \ ++ raw_spin_lock_init(&(h)->lock); \ ++} while (0) ++#else ++#define INIT_HLIST_BL_HEAD(h) (h)->first = NULL ++#endif + + static inline void INIT_HLIST_BL_NODE(struct hlist_bl_node *h) + { +@@ -118,12 +130,26 @@ static inline void hlist_bl_del_init(struct hlist_bl_node *n) + + static inline void hlist_bl_lock(struct hlist_bl_head *b) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + bit_spin_lock(0, (unsigned long *)b); ++#else ++ raw_spin_lock(&b->lock); ++#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) ++ __set_bit(0, (unsigned long *)b); ++#endif ++#endif + } + + static inline void hlist_bl_unlock(struct hlist_bl_head *b) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + __bit_spin_unlock(0, (unsigned long *)b); ++#else ++#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) ++ __clear_bit(0, (unsigned long *)b); ++#endif ++ raw_spin_unlock(&b->lock); ++#endif + } + + static inline bool hlist_bl_is_locked(struct hlist_bl_head *b) +diff --git a/include/linux/locallock.h b/include/linux/locallock.h +new file mode 100644 +index 000000000000..845c77f1a5ca +--- /dev/null ++++ b/include/linux/locallock.h +@@ -0,0 +1,278 @@ ++#ifndef _LINUX_LOCALLOCK_H ++#define _LINUX_LOCALLOCK_H ++ ++#include ++#include ++ ++#ifdef CONFIG_PREEMPT_RT_BASE ++ ++#ifdef CONFIG_DEBUG_SPINLOCK ++# define LL_WARN(cond) WARN_ON(cond) ++#else ++# define LL_WARN(cond) do { } while (0) ++#endif ++ ++/* ++ * per cpu lock based substitute for local_irq_*() ++ */ ++struct local_irq_lock { ++ spinlock_t lock; ++ struct task_struct *owner; ++ int nestcnt; ++ unsigned long flags; ++}; ++ ++#define DEFINE_LOCAL_IRQ_LOCK(lvar) \ ++ DEFINE_PER_CPU(struct local_irq_lock, lvar) = { \ ++ .lock = __SPIN_LOCK_UNLOCKED((lvar).lock) } ++ ++#define DECLARE_LOCAL_IRQ_LOCK(lvar) \ ++ DECLARE_PER_CPU(struct local_irq_lock, lvar) ++ ++#define local_irq_lock_init(lvar) \ ++ do { \ ++ int __cpu; \ ++ for_each_possible_cpu(__cpu) \ ++ spin_lock_init(&per_cpu(lvar, __cpu).lock); \ ++ } while (0) ++ ++/* ++ * spin_lock|trylock|unlock_local flavour that does not migrate disable ++ * used for __local_lock|trylock|unlock where get_local_var/put_local_var ++ * already takes care of the migrate_disable/enable ++ * for CONFIG_PREEMPT_BASE map to the normal spin_* calls. ++ */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define spin_lock_local(lock) rt_spin_lock__no_mg(lock) ++# define spin_trylock_local(lock) rt_spin_trylock__no_mg(lock) ++# define spin_unlock_local(lock) rt_spin_unlock__no_mg(lock) ++#else ++# define spin_lock_local(lock) spin_lock(lock) ++# define spin_trylock_local(lock) spin_trylock(lock) ++# define spin_unlock_local(lock) spin_unlock(lock) ++#endif ++ ++static inline void __local_lock(struct local_irq_lock *lv) ++{ ++ if (lv->owner != current) { ++ spin_lock_local(&lv->lock); ++ LL_WARN(lv->owner); ++ LL_WARN(lv->nestcnt); ++ lv->owner = current; ++ } ++ lv->nestcnt++; ++} ++ ++#define local_lock(lvar) \ ++ do { __local_lock(&get_local_var(lvar)); } while (0) ++ ++#define local_lock_on(lvar, cpu) \ ++ do { __local_lock(&per_cpu(lvar, cpu)); } while (0) ++ ++static inline int __local_trylock(struct local_irq_lock *lv) ++{ ++ if (lv->owner != current && spin_trylock_local(&lv->lock)) { ++ LL_WARN(lv->owner); ++ LL_WARN(lv->nestcnt); ++ lv->owner = current; ++ lv->nestcnt = 1; ++ return 1; ++ } ++ return 0; ++} ++ ++#define local_trylock(lvar) \ ++ ({ \ ++ int __locked; \ ++ __locked = __local_trylock(&get_local_var(lvar)); \ ++ if (!__locked) \ ++ put_local_var(lvar); \ ++ __locked; \ ++ }) ++ ++static inline void __local_unlock(struct local_irq_lock *lv) ++{ ++ LL_WARN(lv->nestcnt == 0); ++ LL_WARN(lv->owner != current); ++ if (--lv->nestcnt) ++ return; ++ ++ lv->owner = NULL; ++ spin_unlock_local(&lv->lock); ++} ++ ++#define local_unlock(lvar) \ ++ do { \ ++ __local_unlock(this_cpu_ptr(&lvar)); \ ++ put_local_var(lvar); \ ++ } while (0) ++ ++#define local_unlock_on(lvar, cpu) \ ++ do { __local_unlock(&per_cpu(lvar, cpu)); } while (0) ++ ++static inline void __local_lock_irq(struct local_irq_lock *lv) ++{ ++ spin_lock_irqsave(&lv->lock, lv->flags); ++ LL_WARN(lv->owner); ++ LL_WARN(lv->nestcnt); ++ lv->owner = current; ++ lv->nestcnt = 1; ++} ++ ++#define local_lock_irq(lvar) \ ++ do { __local_lock_irq(&get_local_var(lvar)); } while (0) ++ ++#define local_lock_irq_on(lvar, cpu) \ ++ do { __local_lock_irq(&per_cpu(lvar, cpu)); } while (0) ++ ++static inline void __local_unlock_irq(struct local_irq_lock *lv) ++{ ++ LL_WARN(!lv->nestcnt); ++ LL_WARN(lv->owner != current); ++ lv->owner = NULL; ++ lv->nestcnt = 0; ++ spin_unlock_irq(&lv->lock); ++} ++ ++#define local_unlock_irq(lvar) \ ++ do { \ ++ __local_unlock_irq(this_cpu_ptr(&lvar)); \ ++ put_local_var(lvar); \ ++ } while (0) ++ ++#define local_unlock_irq_on(lvar, cpu) \ ++ do { \ ++ __local_unlock_irq(&per_cpu(lvar, cpu)); \ ++ } while (0) ++ ++static inline int __local_lock_irqsave(struct local_irq_lock *lv) ++{ ++ if (lv->owner != current) { ++ __local_lock_irq(lv); ++ return 0; ++ } else { ++ lv->nestcnt++; ++ return 1; ++ } ++} ++ ++#define local_lock_irqsave(lvar, _flags) \ ++ do { \ ++ if (__local_lock_irqsave(&get_local_var(lvar))) \ ++ put_local_var(lvar); \ ++ _flags = __this_cpu_read(lvar.flags); \ ++ } while (0) ++ ++#define local_lock_irqsave_on(lvar, _flags, cpu) \ ++ do { \ ++ __local_lock_irqsave(&per_cpu(lvar, cpu)); \ ++ _flags = per_cpu(lvar, cpu).flags; \ ++ } while (0) ++ ++static inline int __local_unlock_irqrestore(struct local_irq_lock *lv, ++ unsigned long flags) ++{ ++ LL_WARN(!lv->nestcnt); ++ LL_WARN(lv->owner != current); ++ if (--lv->nestcnt) ++ return 0; ++ ++ lv->owner = NULL; ++ spin_unlock_irqrestore(&lv->lock, lv->flags); ++ return 1; ++} ++ ++#define local_unlock_irqrestore(lvar, flags) \ ++ do { \ ++ if (__local_unlock_irqrestore(this_cpu_ptr(&lvar), flags)) \ ++ put_local_var(lvar); \ ++ } while (0) ++ ++#define local_unlock_irqrestore_on(lvar, flags, cpu) \ ++ do { \ ++ __local_unlock_irqrestore(&per_cpu(lvar, cpu), flags); \ ++ } while (0) ++ ++#define local_spin_trylock_irq(lvar, lock) \ ++ ({ \ ++ int __locked; \ ++ local_lock_irq(lvar); \ ++ __locked = spin_trylock(lock); \ ++ if (!__locked) \ ++ local_unlock_irq(lvar); \ ++ __locked; \ ++ }) ++ ++#define local_spin_lock_irq(lvar, lock) \ ++ do { \ ++ local_lock_irq(lvar); \ ++ spin_lock(lock); \ ++ } while (0) ++ ++#define local_spin_unlock_irq(lvar, lock) \ ++ do { \ ++ spin_unlock(lock); \ ++ local_unlock_irq(lvar); \ ++ } while (0) ++ ++#define local_spin_lock_irqsave(lvar, lock, flags) \ ++ do { \ ++ local_lock_irqsave(lvar, flags); \ ++ spin_lock(lock); \ ++ } while (0) ++ ++#define local_spin_unlock_irqrestore(lvar, lock, flags) \ ++ do { \ ++ spin_unlock(lock); \ ++ local_unlock_irqrestore(lvar, flags); \ ++ } while (0) ++ ++#define get_locked_var(lvar, var) \ ++ (*({ \ ++ local_lock(lvar); \ ++ this_cpu_ptr(&var); \ ++ })) ++ ++#define put_locked_var(lvar, var) local_unlock(lvar); ++ ++#define local_lock_cpu(lvar) \ ++ ({ \ ++ local_lock(lvar); \ ++ smp_processor_id(); \ ++ }) ++ ++#define local_unlock_cpu(lvar) local_unlock(lvar) ++ ++#else /* PREEMPT_RT_BASE */ ++ ++#define DEFINE_LOCAL_IRQ_LOCK(lvar) __typeof__(const int) lvar ++#define DECLARE_LOCAL_IRQ_LOCK(lvar) extern __typeof__(const int) lvar ++ ++static inline void local_irq_lock_init(int lvar) { } ++ ++#define local_lock(lvar) preempt_disable() ++#define local_unlock(lvar) preempt_enable() ++#define local_lock_irq(lvar) local_irq_disable() ++#define local_lock_irq_on(lvar, cpu) local_irq_disable() ++#define local_unlock_irq(lvar) local_irq_enable() ++#define local_unlock_irq_on(lvar, cpu) local_irq_enable() ++#define local_lock_irqsave(lvar, flags) local_irq_save(flags) ++#define local_unlock_irqrestore(lvar, flags) local_irq_restore(flags) ++ ++#define local_spin_trylock_irq(lvar, lock) spin_trylock_irq(lock) ++#define local_spin_lock_irq(lvar, lock) spin_lock_irq(lock) ++#define local_spin_unlock_irq(lvar, lock) spin_unlock_irq(lock) ++#define local_spin_lock_irqsave(lvar, lock, flags) \ ++ spin_lock_irqsave(lock, flags) ++#define local_spin_unlock_irqrestore(lvar, lock, flags) \ ++ spin_unlock_irqrestore(lock, flags) ++ ++#define get_locked_var(lvar, var) get_cpu_var(var) ++#define put_locked_var(lvar, var) put_cpu_var(var) ++ ++#define local_lock_cpu(lvar) get_cpu() ++#define local_unlock_cpu(lvar) put_cpu() ++ ++#endif ++ ++#endif +diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h +index 08d947fc4c59..705fb564a605 100644 +--- a/include/linux/mm_types.h ++++ b/include/linux/mm_types.h +@@ -11,6 +11,7 @@ + #include + #include + #include ++#include + #include + #include + #include +@@ -509,6 +510,9 @@ struct mm_struct { + bool tlb_flush_pending; + #endif + struct uprobes_state uprobes_state; ++#ifdef CONFIG_PREEMPT_RT_BASE ++ struct rcu_head delayed_drop; ++#endif + #ifdef CONFIG_X86_INTEL_MPX + /* address of the bounds directory */ + void __user *bd_addr; +diff --git a/include/linux/module.h b/include/linux/module.h +index 0c3207d26ac0..5944baaa3f28 100644 +--- a/include/linux/module.h ++++ b/include/linux/module.h +@@ -496,6 +496,7 @@ static inline int module_is_live(struct module *mod) + struct module *__module_text_address(unsigned long addr); + struct module *__module_address(unsigned long addr); + bool is_module_address(unsigned long addr); ++bool __is_module_percpu_address(unsigned long addr, unsigned long *can_addr); + bool is_module_percpu_address(unsigned long addr); + bool is_module_text_address(unsigned long addr); + +@@ -663,6 +664,11 @@ static inline bool is_module_percpu_address(unsigned long addr) + return false; + } + ++static inline bool __is_module_percpu_address(unsigned long addr, unsigned long *can_addr) ++{ ++ return false; ++} ++ + static inline bool is_module_text_address(unsigned long addr) + { + return false; +diff --git a/include/linux/mutex.h b/include/linux/mutex.h +index 2cb7531e7d7a..b3fdfc820216 100644 +--- a/include/linux/mutex.h ++++ b/include/linux/mutex.h +@@ -19,6 +19,17 @@ + #include + #include + ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++# define __DEP_MAP_MUTEX_INITIALIZER(lockname) \ ++ , .dep_map = { .name = #lockname } ++#else ++# define __DEP_MAP_MUTEX_INITIALIZER(lockname) ++#endif ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++# include ++#else ++ + /* + * Simple, straightforward mutexes with strict semantics: + * +@@ -99,13 +110,6 @@ do { \ + static inline void mutex_destroy(struct mutex *lock) {} + #endif + +-#ifdef CONFIG_DEBUG_LOCK_ALLOC +-# define __DEP_MAP_MUTEX_INITIALIZER(lockname) \ +- , .dep_map = { .name = #lockname } +-#else +-# define __DEP_MAP_MUTEX_INITIALIZER(lockname) +-#endif +- + #define __MUTEX_INITIALIZER(lockname) \ + { .count = ATOMIC_INIT(1) \ + , .wait_lock = __SPIN_LOCK_UNLOCKED(lockname.wait_lock) \ +@@ -173,6 +177,8 @@ extern int __must_check mutex_lock_killable(struct mutex *lock); + extern int mutex_trylock(struct mutex *lock); + extern void mutex_unlock(struct mutex *lock); + ++#endif /* !PREEMPT_RT_FULL */ ++ + extern int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock); + + #endif /* __LINUX_MUTEX_H */ +diff --git a/include/linux/mutex_rt.h b/include/linux/mutex_rt.h +new file mode 100644 +index 000000000000..e0284edec655 +--- /dev/null ++++ b/include/linux/mutex_rt.h +@@ -0,0 +1,89 @@ ++#ifndef __LINUX_MUTEX_RT_H ++#define __LINUX_MUTEX_RT_H ++ ++#ifndef __LINUX_MUTEX_H ++#error "Please include mutex.h" ++#endif ++ ++#include ++ ++/* FIXME: Just for __lockfunc */ ++#include ++ ++struct mutex { ++ struct rt_mutex lock; ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ struct lockdep_map dep_map; ++#endif ++}; ++ ++#define __MUTEX_INITIALIZER(mutexname) \ ++ { \ ++ .lock = __RT_MUTEX_INITIALIZER(mutexname.lock) \ ++ __DEP_MAP_MUTEX_INITIALIZER(mutexname) \ ++ } ++ ++#define DEFINE_MUTEX(mutexname) \ ++ struct mutex mutexname = __MUTEX_INITIALIZER(mutexname) ++ ++extern void __mutex_do_init(struct mutex *lock, const char *name, struct lock_class_key *key); ++extern void __lockfunc _mutex_lock(struct mutex *lock); ++extern int __lockfunc _mutex_lock_interruptible(struct mutex *lock); ++extern int __lockfunc _mutex_lock_killable(struct mutex *lock); ++extern void __lockfunc _mutex_lock_nested(struct mutex *lock, int subclass); ++extern void __lockfunc _mutex_lock_nest_lock(struct mutex *lock, struct lockdep_map *nest_lock); ++extern int __lockfunc _mutex_lock_interruptible_nested(struct mutex *lock, int subclass); ++extern int __lockfunc _mutex_lock_killable_nested(struct mutex *lock, int subclass); ++extern int __lockfunc _mutex_trylock(struct mutex *lock); ++extern void __lockfunc _mutex_unlock(struct mutex *lock); ++ ++#define mutex_is_locked(l) rt_mutex_is_locked(&(l)->lock) ++#define mutex_lock(l) _mutex_lock(l) ++#define mutex_lock_interruptible(l) _mutex_lock_interruptible(l) ++#define mutex_lock_killable(l) _mutex_lock_killable(l) ++#define mutex_trylock(l) _mutex_trylock(l) ++#define mutex_unlock(l) _mutex_unlock(l) ++ ++#ifdef CONFIG_DEBUG_MUTEXES ++#define mutex_destroy(l) rt_mutex_destroy(&(l)->lock) ++#else ++static inline void mutex_destroy(struct mutex *lock) {} ++#endif ++ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++# define mutex_lock_nested(l, s) _mutex_lock_nested(l, s) ++# define mutex_lock_interruptible_nested(l, s) \ ++ _mutex_lock_interruptible_nested(l, s) ++# define mutex_lock_killable_nested(l, s) \ ++ _mutex_lock_killable_nested(l, s) ++ ++# define mutex_lock_nest_lock(lock, nest_lock) \ ++do { \ ++ typecheck(struct lockdep_map *, &(nest_lock)->dep_map); \ ++ _mutex_lock_nest_lock(lock, &(nest_lock)->dep_map); \ ++} while (0) ++ ++#else ++# define mutex_lock_nested(l, s) _mutex_lock(l) ++# define mutex_lock_interruptible_nested(l, s) \ ++ _mutex_lock_interruptible(l) ++# define mutex_lock_killable_nested(l, s) \ ++ _mutex_lock_killable(l) ++# define mutex_lock_nest_lock(lock, nest_lock) mutex_lock(lock) ++#endif ++ ++# define mutex_init(mutex) \ ++do { \ ++ static struct lock_class_key __key; \ ++ \ ++ rt_mutex_init(&(mutex)->lock); \ ++ __mutex_do_init((mutex), #mutex, &__key); \ ++} while (0) ++ ++# define __mutex_init(mutex, name, key) \ ++do { \ ++ rt_mutex_init(&(mutex)->lock); \ ++ __mutex_do_init((mutex), name, key); \ ++} while (0) ++ ++#endif +diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h +index bb9b102c15cd..a5b12b8ad196 100644 +--- a/include/linux/netdevice.h ++++ b/include/linux/netdevice.h +@@ -396,7 +396,19 @@ typedef enum rx_handler_result rx_handler_result_t; + typedef rx_handler_result_t rx_handler_func_t(struct sk_buff **pskb); + + void __napi_schedule(struct napi_struct *n); ++ ++/* ++ * When PREEMPT_RT_FULL is defined, all device interrupt handlers ++ * run as threads, and they can also be preempted (without PREEMPT_RT ++ * interrupt threads can not be preempted). Which means that calling ++ * __napi_schedule_irqoff() from an interrupt handler can be preempted ++ * and can corrupt the napi->poll_list. ++ */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++#define __napi_schedule_irqoff(n) __napi_schedule(n) ++#else + void __napi_schedule_irqoff(struct napi_struct *n); ++#endif + + static inline bool napi_disable_pending(struct napi_struct *n) + { +@@ -2463,14 +2475,53 @@ void netdev_freemem(struct net_device *dev); + void synchronize_net(void); + int init_dummy_netdev(struct net_device *dev); + +-DECLARE_PER_CPU(int, xmit_recursion); + #define XMIT_RECURSION_LIMIT 10 ++#ifdef CONFIG_PREEMPT_RT_FULL ++static inline int dev_recursion_level(void) ++{ ++ return current->xmit_recursion; ++} ++ ++static inline int xmit_rec_read(void) ++{ ++ return current->xmit_recursion; ++} ++ ++static inline void xmit_rec_inc(void) ++{ ++ current->xmit_recursion++; ++} ++ ++static inline void xmit_rec_dec(void) ++{ ++ current->xmit_recursion--; ++} ++ ++#else ++ ++DECLARE_PER_CPU(int, xmit_recursion); + + static inline int dev_recursion_level(void) + { + return this_cpu_read(xmit_recursion); + } + ++static inline int xmit_rec_read(void) ++{ ++ return __this_cpu_read(xmit_recursion); ++} ++ ++static inline void xmit_rec_inc(void) ++{ ++ __this_cpu_inc(xmit_recursion); ++} ++ ++static inline void xmit_rec_dec(void) ++{ ++ __this_cpu_dec(xmit_recursion); ++} ++#endif ++ + struct net_device *dev_get_by_index(struct net *net, int ifindex); + struct net_device *__dev_get_by_index(struct net *net, int ifindex); + struct net_device *dev_get_by_index_rcu(struct net *net, int ifindex); +@@ -2855,6 +2906,7 @@ struct softnet_data { + unsigned int dropped; + struct sk_buff_head input_pkt_queue; + struct napi_struct backlog; ++ struct sk_buff_head tofree_queue; + + }; + +diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h +index 2ad1a2b289b5..b4d10155af54 100644 +--- a/include/linux/netfilter/x_tables.h ++++ b/include/linux/netfilter/x_tables.h +@@ -4,6 +4,7 @@ + + #include + #include ++#include + #include + + /* Test a struct->invflags and a boolean for inequality */ +@@ -300,6 +301,8 @@ void xt_free_table_info(struct xt_table_info *info); + */ + DECLARE_PER_CPU(seqcount_t, xt_recseq); + ++DECLARE_LOCAL_IRQ_LOCK(xt_write_lock); ++ + /* xt_tee_enabled - true if x_tables needs to handle reentrancy + * + * Enabled if current ip(6)tables ruleset has at least one -j TEE rule. +@@ -320,6 +323,9 @@ static inline unsigned int xt_write_recseq_begin(void) + { + unsigned int addend; + ++ /* RT protection */ ++ local_lock(xt_write_lock); ++ + /* + * Low order bit of sequence is set if we already + * called xt_write_recseq_begin(). +@@ -350,6 +356,7 @@ static inline void xt_write_recseq_end(unsigned int addend) + /* this is kind of a write_seqcount_end(), but addend is 0 or 1 */ + smp_wmb(); + __this_cpu_add(xt_recseq.sequence, addend); ++ local_unlock(xt_write_lock); + } + + /* +diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h +index 810124b33327..d54ca43d571f 100644 +--- a/include/linux/nfs_fs.h ++++ b/include/linux/nfs_fs.h +@@ -165,7 +165,11 @@ struct nfs_inode { + + /* Readers: in-flight sillydelete RPC calls */ + /* Writers: rmdir */ ++#ifdef CONFIG_PREEMPT_RT_BASE ++ struct semaphore rmdir_sem; ++#else + struct rw_semaphore rmdir_sem; ++#endif + + #if IS_ENABLED(CONFIG_NFS_V4) + struct nfs4_cached_acl *nfs4_acl; +diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h +index beb1e10f446e..ebaf2e7bfe29 100644 +--- a/include/linux/nfs_xdr.h ++++ b/include/linux/nfs_xdr.h +@@ -1490,7 +1490,7 @@ struct nfs_unlinkdata { + struct nfs_removeargs args; + struct nfs_removeres res; + struct dentry *dentry; +- wait_queue_head_t wq; ++ struct swait_queue_head wq; + struct rpc_cred *cred; + struct nfs_fattr dir_attr; + long timeout; +diff --git a/include/linux/notifier.h b/include/linux/notifier.h +index 4149868de4e6..babe5b9bcb91 100644 +--- a/include/linux/notifier.h ++++ b/include/linux/notifier.h +@@ -6,7 +6,7 @@ + * + * Alan Cox + */ +- ++ + #ifndef _LINUX_NOTIFIER_H + #define _LINUX_NOTIFIER_H + #include +@@ -42,9 +42,7 @@ + * in srcu_notifier_call_chain(): no cache bounces and no memory barriers. + * As compensation, srcu_notifier_chain_unregister() is rather expensive. + * SRCU notifier chains should be used when the chain will be called very +- * often but notifier_blocks will seldom be removed. Also, SRCU notifier +- * chains are slightly more difficult to use because they require special +- * runtime initialization. ++ * often but notifier_blocks will seldom be removed. + */ + + struct notifier_block; +@@ -90,7 +88,7 @@ struct srcu_notifier_head { + (name)->head = NULL; \ + } while (0) + +-/* srcu_notifier_heads must be initialized and cleaned up dynamically */ ++/* srcu_notifier_heads must be cleaned up dynamically */ + extern void srcu_init_notifier_head(struct srcu_notifier_head *nh); + #define srcu_cleanup_notifier_head(name) \ + cleanup_srcu_struct(&(name)->srcu); +@@ -103,7 +101,13 @@ extern void srcu_init_notifier_head(struct srcu_notifier_head *nh); + .head = NULL } + #define RAW_NOTIFIER_INIT(name) { \ + .head = NULL } +-/* srcu_notifier_heads cannot be initialized statically */ ++ ++#define SRCU_NOTIFIER_INIT(name, pcpu) \ ++ { \ ++ .mutex = __MUTEX_INITIALIZER(name.mutex), \ ++ .head = NULL, \ ++ .srcu = __SRCU_STRUCT_INIT(name.srcu, pcpu), \ ++ } + + #define ATOMIC_NOTIFIER_HEAD(name) \ + struct atomic_notifier_head name = \ +@@ -115,6 +119,18 @@ extern void srcu_init_notifier_head(struct srcu_notifier_head *nh); + struct raw_notifier_head name = \ + RAW_NOTIFIER_INIT(name) + ++#define _SRCU_NOTIFIER_HEAD(name, mod) \ ++ static DEFINE_PER_CPU(struct srcu_struct_array, \ ++ name##_head_srcu_array); \ ++ mod struct srcu_notifier_head name = \ ++ SRCU_NOTIFIER_INIT(name, name##_head_srcu_array) ++ ++#define SRCU_NOTIFIER_HEAD(name) \ ++ _SRCU_NOTIFIER_HEAD(name, ) ++ ++#define SRCU_NOTIFIER_HEAD_STATIC(name) \ ++ _SRCU_NOTIFIER_HEAD(name, static) ++ + #ifdef __KERNEL__ + + extern int atomic_notifier_chain_register(struct atomic_notifier_head *nh, +@@ -184,12 +200,12 @@ static inline int notifier_to_errno(int ret) + + /* + * Declared notifiers so far. I can imagine quite a few more chains +- * over time (eg laptop power reset chains, reboot chain (to clean ++ * over time (eg laptop power reset chains, reboot chain (to clean + * device units up), device [un]mount chain, module load/unload chain, +- * low memory chain, screenblank chain (for plug in modular screenblankers) ++ * low memory chain, screenblank chain (for plug in modular screenblankers) + * VC switch chains (for loadable kernel svgalib VC switch helpers) etc... + */ +- ++ + /* CPU notfiers are defined in include/linux/cpu.h. */ + + /* netdevice notifiers are defined in include/linux/netdevice.h */ +diff --git a/include/linux/percpu-rwsem.h b/include/linux/percpu-rwsem.h +index 5b2e6159b744..ea940f451606 100644 +--- a/include/linux/percpu-rwsem.h ++++ b/include/linux/percpu-rwsem.h +@@ -4,7 +4,7 @@ + #include + #include + #include +-#include ++#include + #include + #include + +@@ -12,7 +12,7 @@ struct percpu_rw_semaphore { + struct rcu_sync rss; + unsigned int __percpu *read_count; + struct rw_semaphore rw_sem; +- wait_queue_head_t writer; ++ struct swait_queue_head writer; + int readers_block; + }; + +@@ -22,13 +22,13 @@ static struct percpu_rw_semaphore name = { \ + .rss = __RCU_SYNC_INITIALIZER(name.rss, RCU_SCHED_SYNC), \ + .read_count = &__percpu_rwsem_rc_##name, \ + .rw_sem = __RWSEM_INITIALIZER(name.rw_sem), \ +- .writer = __WAIT_QUEUE_HEAD_INITIALIZER(name.writer), \ ++ .writer = __SWAIT_QUEUE_HEAD_INITIALIZER(name.writer), \ + } + + extern int __percpu_down_read(struct percpu_rw_semaphore *, int); + extern void __percpu_up_read(struct percpu_rw_semaphore *); + +-static inline void percpu_down_read_preempt_disable(struct percpu_rw_semaphore *sem) ++static inline void percpu_down_read(struct percpu_rw_semaphore *sem) + { + might_sleep(); + +@@ -46,16 +46,10 @@ static inline void percpu_down_read_preempt_disable(struct percpu_rw_semaphore * + __this_cpu_inc(*sem->read_count); + if (unlikely(!rcu_sync_is_idle(&sem->rss))) + __percpu_down_read(sem, false); /* Unconditional memory barrier */ +- barrier(); + /* +- * The barrier() prevents the compiler from ++ * The preempt_enable() prevents the compiler from + * bleeding the critical section out. + */ +-} +- +-static inline void percpu_down_read(struct percpu_rw_semaphore *sem) +-{ +- percpu_down_read_preempt_disable(sem); + preempt_enable(); + } + +@@ -82,13 +76,9 @@ static inline int percpu_down_read_trylock(struct percpu_rw_semaphore *sem) + return ret; + } + +-static inline void percpu_up_read_preempt_enable(struct percpu_rw_semaphore *sem) ++static inline void percpu_up_read(struct percpu_rw_semaphore *sem) + { +- /* +- * The barrier() prevents the compiler from +- * bleeding the critical section out. +- */ +- barrier(); ++ preempt_disable(); + /* + * Same as in percpu_down_read(). + */ +@@ -101,12 +91,6 @@ static inline void percpu_up_read_preempt_enable(struct percpu_rw_semaphore *sem + rwsem_release(&sem->rw_sem.dep_map, 1, _RET_IP_); + } + +-static inline void percpu_up_read(struct percpu_rw_semaphore *sem) +-{ +- preempt_disable(); +- percpu_up_read_preempt_enable(sem); +-} +- + extern void percpu_down_write(struct percpu_rw_semaphore *); + extern void percpu_up_write(struct percpu_rw_semaphore *); + +diff --git a/include/linux/percpu.h b/include/linux/percpu.h +index 56939d3f6e53..b988bf40ad3e 100644 +--- a/include/linux/percpu.h ++++ b/include/linux/percpu.h +@@ -18,6 +18,35 @@ + #define PERCPU_MODULE_RESERVE 0 + #endif + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ ++#define get_local_var(var) (*({ \ ++ migrate_disable(); \ ++ this_cpu_ptr(&var); })) ++ ++#define put_local_var(var) do { \ ++ (void)&(var); \ ++ migrate_enable(); \ ++} while (0) ++ ++# define get_local_ptr(var) ({ \ ++ migrate_disable(); \ ++ this_cpu_ptr(var); }) ++ ++# define put_local_ptr(var) do { \ ++ (void)(var); \ ++ migrate_enable(); \ ++} while (0) ++ ++#else ++ ++#define get_local_var(var) get_cpu_var(var) ++#define put_local_var(var) put_cpu_var(var) ++#define get_local_ptr(var) get_cpu_ptr(var) ++#define put_local_ptr(var) put_cpu_ptr(var) ++ ++#endif ++ + /* minimum unit size, also is the maximum supported allocation size */ + #define PCPU_MIN_UNIT_SIZE PFN_ALIGN(32 << 10) + +@@ -110,6 +139,7 @@ extern int __init pcpu_page_first_chunk(size_t reserved_size, + #endif + + extern void __percpu *__alloc_reserved_percpu(size_t size, size_t align); ++extern bool __is_kernel_percpu_address(unsigned long addr, unsigned long *can_addr); + extern bool is_kernel_percpu_address(unsigned long addr); + + #if !defined(CONFIG_SMP) || !defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) +diff --git a/include/linux/pid.h b/include/linux/pid.h +index 23705a53abba..2cc64b779f03 100644 +--- a/include/linux/pid.h ++++ b/include/linux/pid.h +@@ -2,6 +2,7 @@ + #define _LINUX_PID_H + + #include ++#include + + enum pid_type + { +diff --git a/include/linux/preempt.h b/include/linux/preempt.h +index 75e4e30677f1..1cfb1cb72354 100644 +--- a/include/linux/preempt.h ++++ b/include/linux/preempt.h +@@ -50,7 +50,11 @@ + #define HARDIRQ_OFFSET (1UL << HARDIRQ_SHIFT) + #define NMI_OFFSET (1UL << NMI_SHIFT) + +-#define SOFTIRQ_DISABLE_OFFSET (2 * SOFTIRQ_OFFSET) ++#ifndef CONFIG_PREEMPT_RT_FULL ++# define SOFTIRQ_DISABLE_OFFSET (2 * SOFTIRQ_OFFSET) ++#else ++# define SOFTIRQ_DISABLE_OFFSET (0) ++#endif + + /* We use the MSB mostly because its available */ + #define PREEMPT_NEED_RESCHED 0x80000000 +@@ -59,9 +63,15 @@ + #include + + #define hardirq_count() (preempt_count() & HARDIRQ_MASK) +-#define softirq_count() (preempt_count() & SOFTIRQ_MASK) + #define irq_count() (preempt_count() & (HARDIRQ_MASK | SOFTIRQ_MASK \ + | NMI_MASK)) ++#ifndef CONFIG_PREEMPT_RT_FULL ++# define softirq_count() (preempt_count() & SOFTIRQ_MASK) ++# define in_serving_softirq() (softirq_count() & SOFTIRQ_OFFSET) ++#else ++# define softirq_count() (0UL) ++extern int in_serving_softirq(void); ++#endif + + /* + * Are we doing bottom half or hardware interrupt processing? +@@ -72,7 +82,6 @@ + #define in_irq() (hardirq_count()) + #define in_softirq() (softirq_count()) + #define in_interrupt() (irq_count()) +-#define in_serving_softirq() (softirq_count() & SOFTIRQ_OFFSET) + + /* + * Are we in NMI context? +@@ -91,7 +100,11 @@ + /* + * The preempt_count offset after spin_lock() + */ ++#if !defined(CONFIG_PREEMPT_RT_FULL) + #define PREEMPT_LOCK_OFFSET PREEMPT_DISABLE_OFFSET ++#else ++#define PREEMPT_LOCK_OFFSET 0 ++#endif + + /* + * The preempt_count offset needed for things like: +@@ -140,6 +153,20 @@ extern void preempt_count_sub(int val); + #define preempt_count_inc() preempt_count_add(1) + #define preempt_count_dec() preempt_count_sub(1) + ++#ifdef CONFIG_PREEMPT_LAZY ++#define add_preempt_lazy_count(val) do { preempt_lazy_count() += (val); } while (0) ++#define sub_preempt_lazy_count(val) do { preempt_lazy_count() -= (val); } while (0) ++#define inc_preempt_lazy_count() add_preempt_lazy_count(1) ++#define dec_preempt_lazy_count() sub_preempt_lazy_count(1) ++#define preempt_lazy_count() (current_thread_info()->preempt_lazy_count) ++#else ++#define add_preempt_lazy_count(val) do { } while (0) ++#define sub_preempt_lazy_count(val) do { } while (0) ++#define inc_preempt_lazy_count() do { } while (0) ++#define dec_preempt_lazy_count() do { } while (0) ++#define preempt_lazy_count() (0) ++#endif ++ + #ifdef CONFIG_PREEMPT_COUNT + + #define preempt_disable() \ +@@ -148,13 +175,25 @@ do { \ + barrier(); \ + } while (0) + ++#define preempt_lazy_disable() \ ++do { \ ++ inc_preempt_lazy_count(); \ ++ barrier(); \ ++} while (0) ++ + #define sched_preempt_enable_no_resched() \ + do { \ + barrier(); \ + preempt_count_dec(); \ + } while (0) + +-#define preempt_enable_no_resched() sched_preempt_enable_no_resched() ++#ifdef CONFIG_PREEMPT_RT_BASE ++# define preempt_enable_no_resched() sched_preempt_enable_no_resched() ++# define preempt_check_resched_rt() preempt_check_resched() ++#else ++# define preempt_enable_no_resched() preempt_enable() ++# define preempt_check_resched_rt() barrier(); ++#endif + + #define preemptible() (preempt_count() == 0 && !irqs_disabled()) + +@@ -179,6 +218,13 @@ do { \ + __preempt_schedule(); \ + } while (0) + ++#define preempt_lazy_enable() \ ++do { \ ++ dec_preempt_lazy_count(); \ ++ barrier(); \ ++ preempt_check_resched(); \ ++} while (0) ++ + #else /* !CONFIG_PREEMPT */ + #define preempt_enable() \ + do { \ +@@ -224,6 +270,7 @@ do { \ + #define preempt_disable_notrace() barrier() + #define preempt_enable_no_resched_notrace() barrier() + #define preempt_enable_notrace() barrier() ++#define preempt_check_resched_rt() barrier() + #define preemptible() 0 + + #endif /* CONFIG_PREEMPT_COUNT */ +@@ -244,10 +291,31 @@ do { \ + } while (0) + #define preempt_fold_need_resched() \ + do { \ +- if (tif_need_resched()) \ ++ if (tif_need_resched_now()) \ + set_preempt_need_resched(); \ + } while (0) + ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define preempt_disable_rt() preempt_disable() ++# define preempt_enable_rt() preempt_enable() ++# define preempt_disable_nort() barrier() ++# define preempt_enable_nort() barrier() ++# ifdef CONFIG_SMP ++ extern void migrate_disable(void); ++ extern void migrate_enable(void); ++# else /* CONFIG_SMP */ ++# define migrate_disable() barrier() ++# define migrate_enable() barrier() ++# endif /* CONFIG_SMP */ ++#else ++# define preempt_disable_rt() barrier() ++# define preempt_enable_rt() barrier() ++# define preempt_disable_nort() preempt_disable() ++# define preempt_enable_nort() preempt_enable() ++# define migrate_disable() preempt_disable() ++# define migrate_enable() preempt_enable() ++#endif ++ + #ifdef CONFIG_PREEMPT_NOTIFIERS + + struct preempt_notifier; +diff --git a/include/linux/printk.h b/include/linux/printk.h +index eac1af8502bb..37e647af0b0b 100644 +--- a/include/linux/printk.h ++++ b/include/linux/printk.h +@@ -126,9 +126,11 @@ struct va_format { + #ifdef CONFIG_EARLY_PRINTK + extern asmlinkage __printf(1, 2) + void early_printk(const char *fmt, ...); ++extern void printk_kill(void); + #else + static inline __printf(1, 2) __cold + void early_printk(const char *s, ...) { } ++static inline void printk_kill(void) { } + #endif + + #ifdef CONFIG_PRINTK_NMI +diff --git a/include/linux/radix-tree.h b/include/linux/radix-tree.h +index af3581b8a451..277295039c8f 100644 +--- a/include/linux/radix-tree.h ++++ b/include/linux/radix-tree.h +@@ -292,6 +292,8 @@ unsigned int radix_tree_gang_lookup_slot(struct radix_tree_root *root, + int radix_tree_preload(gfp_t gfp_mask); + int radix_tree_maybe_preload(gfp_t gfp_mask); + int radix_tree_maybe_preload_order(gfp_t gfp_mask, int order); ++void radix_tree_preload_end(void); ++ + void radix_tree_init(void); + void *radix_tree_tag_set(struct radix_tree_root *root, + unsigned long index, unsigned int tag); +@@ -314,11 +316,6 @@ unsigned long radix_tree_range_tag_if_tagged(struct radix_tree_root *root, + int radix_tree_tagged(struct radix_tree_root *root, unsigned int tag); + unsigned long radix_tree_locate_item(struct radix_tree_root *root, void *item); + +-static inline void radix_tree_preload_end(void) +-{ +- preempt_enable(); +-} +- + /** + * struct radix_tree_iter - radix tree iterator state + * +diff --git a/include/linux/random.h b/include/linux/random.h +index 16ab429735a7..9d0fecb5b6c2 100644 +--- a/include/linux/random.h ++++ b/include/linux/random.h +@@ -31,7 +31,7 @@ static inline void add_latent_entropy(void) {} + + extern void add_input_randomness(unsigned int type, unsigned int code, + unsigned int value) __latent_entropy; +-extern void add_interrupt_randomness(int irq, int irq_flags) __latent_entropy; ++extern void add_interrupt_randomness(int irq, int irq_flags, __u64 ip) __latent_entropy; + + extern void get_random_bytes(void *buf, int nbytes); + extern int add_random_ready_callback(struct random_ready_callback *rdy); +diff --git a/include/linux/rbtree.h b/include/linux/rbtree.h +index e585018498d5..25c64474fc27 100644 +--- a/include/linux/rbtree.h ++++ b/include/linux/rbtree.h +@@ -31,7 +31,7 @@ + + #include + #include +-#include ++#include + + struct rb_node { + unsigned long __rb_parent_color; +diff --git a/include/linux/rbtree_augmented.h b/include/linux/rbtree_augmented.h +index d076183e49be..36bfb4dd57ae 100644 +--- a/include/linux/rbtree_augmented.h ++++ b/include/linux/rbtree_augmented.h +@@ -26,6 +26,7 @@ + + #include + #include ++#include + + /* + * Please note - only struct rb_augment_callbacks and the prototypes for +diff --git a/include/linux/rcu_assign_pointer.h b/include/linux/rcu_assign_pointer.h +new file mode 100644 +index 000000000000..7066962a4379 +--- /dev/null ++++ b/include/linux/rcu_assign_pointer.h +@@ -0,0 +1,54 @@ ++#ifndef __LINUX_RCU_ASSIGN_POINTER_H__ ++#define __LINUX_RCU_ASSIGN_POINTER_H__ ++#include ++#include ++ ++/** ++ * RCU_INITIALIZER() - statically initialize an RCU-protected global variable ++ * @v: The value to statically initialize with. ++ */ ++#define RCU_INITIALIZER(v) (typeof(*(v)) __force __rcu *)(v) ++ ++/** ++ * rcu_assign_pointer() - assign to RCU-protected pointer ++ * @p: pointer to assign to ++ * @v: value to assign (publish) ++ * ++ * Assigns the specified value to the specified RCU-protected ++ * pointer, ensuring that any concurrent RCU readers will see ++ * any prior initialization. ++ * ++ * Inserts memory barriers on architectures that require them ++ * (which is most of them), and also prevents the compiler from ++ * reordering the code that initializes the structure after the pointer ++ * assignment. More importantly, this call documents which pointers ++ * will be dereferenced by RCU read-side code. ++ * ++ * In some special cases, you may use RCU_INIT_POINTER() instead ++ * of rcu_assign_pointer(). RCU_INIT_POINTER() is a bit faster due ++ * to the fact that it does not constrain either the CPU or the compiler. ++ * That said, using RCU_INIT_POINTER() when you should have used ++ * rcu_assign_pointer() is a very bad thing that results in ++ * impossible-to-diagnose memory corruption. So please be careful. ++ * See the RCU_INIT_POINTER() comment header for details. ++ * ++ * Note that rcu_assign_pointer() evaluates each of its arguments only ++ * once, appearances notwithstanding. One of the "extra" evaluations ++ * is in typeof() and the other visible only to sparse (__CHECKER__), ++ * neither of which actually execute the argument. As with most cpp ++ * macros, this execute-arguments-only-once property is important, so ++ * please be careful when making changes to rcu_assign_pointer() and the ++ * other macros that it invokes. ++ */ ++#define rcu_assign_pointer(p, v) \ ++({ \ ++ uintptr_t _r_a_p__v = (uintptr_t)(v); \ ++ \ ++ if (__builtin_constant_p(v) && (_r_a_p__v) == (uintptr_t)NULL) \ ++ WRITE_ONCE((p), (typeof(p))(_r_a_p__v)); \ ++ else \ ++ smp_store_release(&p, RCU_INITIALIZER((typeof(p))_r_a_p__v)); \ ++ _r_a_p__v; \ ++}) ++ ++#endif +diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h +index 01f71e1d2e94..30cc001d0d5a 100644 +--- a/include/linux/rcupdate.h ++++ b/include/linux/rcupdate.h +@@ -46,6 +46,7 @@ + #include + #include + #include ++#include + + #include + +@@ -178,6 +179,9 @@ void call_rcu(struct rcu_head *head, + + #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ + ++#ifdef CONFIG_PREEMPT_RT_FULL ++#define call_rcu_bh call_rcu ++#else + /** + * call_rcu_bh() - Queue an RCU for invocation after a quicker grace period. + * @head: structure to be used for queueing the RCU updates. +@@ -201,6 +205,7 @@ void call_rcu(struct rcu_head *head, + */ + void call_rcu_bh(struct rcu_head *head, + rcu_callback_t func); ++#endif + + /** + * call_rcu_sched() - Queue an RCU for invocation after sched grace period. +@@ -301,6 +306,11 @@ void synchronize_rcu(void); + * types of kernel builds, the rcu_read_lock() nesting depth is unknowable. + */ + #define rcu_preempt_depth() (current->rcu_read_lock_nesting) ++#ifndef CONFIG_PREEMPT_RT_FULL ++#define sched_rcu_preempt_depth() rcu_preempt_depth() ++#else ++static inline int sched_rcu_preempt_depth(void) { return 0; } ++#endif + + #else /* #ifdef CONFIG_PREEMPT_RCU */ + +@@ -326,6 +336,8 @@ static inline int rcu_preempt_depth(void) + return 0; + } + ++#define sched_rcu_preempt_depth() rcu_preempt_depth() ++ + #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ + + /* Internal to kernel */ +@@ -505,7 +517,14 @@ extern struct lockdep_map rcu_callback_map; + int debug_lockdep_rcu_enabled(void); + + int rcu_read_lock_held(void); ++#ifdef CONFIG_PREEMPT_RT_FULL ++static inline int rcu_read_lock_bh_held(void) ++{ ++ return rcu_read_lock_held(); ++} ++#else + int rcu_read_lock_bh_held(void); ++#endif + + /** + * rcu_read_lock_sched_held() - might we be in RCU-sched read-side critical section? +@@ -626,54 +645,6 @@ static inline void rcu_preempt_sleep_check(void) + }) + + /** +- * RCU_INITIALIZER() - statically initialize an RCU-protected global variable +- * @v: The value to statically initialize with. +- */ +-#define RCU_INITIALIZER(v) (typeof(*(v)) __force __rcu *)(v) +- +-/** +- * rcu_assign_pointer() - assign to RCU-protected pointer +- * @p: pointer to assign to +- * @v: value to assign (publish) +- * +- * Assigns the specified value to the specified RCU-protected +- * pointer, ensuring that any concurrent RCU readers will see +- * any prior initialization. +- * +- * Inserts memory barriers on architectures that require them +- * (which is most of them), and also prevents the compiler from +- * reordering the code that initializes the structure after the pointer +- * assignment. More importantly, this call documents which pointers +- * will be dereferenced by RCU read-side code. +- * +- * In some special cases, you may use RCU_INIT_POINTER() instead +- * of rcu_assign_pointer(). RCU_INIT_POINTER() is a bit faster due +- * to the fact that it does not constrain either the CPU or the compiler. +- * That said, using RCU_INIT_POINTER() when you should have used +- * rcu_assign_pointer() is a very bad thing that results in +- * impossible-to-diagnose memory corruption. So please be careful. +- * See the RCU_INIT_POINTER() comment header for details. +- * +- * Note that rcu_assign_pointer() evaluates each of its arguments only +- * once, appearances notwithstanding. One of the "extra" evaluations +- * is in typeof() and the other visible only to sparse (__CHECKER__), +- * neither of which actually execute the argument. As with most cpp +- * macros, this execute-arguments-only-once property is important, so +- * please be careful when making changes to rcu_assign_pointer() and the +- * other macros that it invokes. +- */ +-#define rcu_assign_pointer(p, v) \ +-({ \ +- uintptr_t _r_a_p__v = (uintptr_t)(v); \ +- \ +- if (__builtin_constant_p(v) && (_r_a_p__v) == (uintptr_t)NULL) \ +- WRITE_ONCE((p), (typeof(p))(_r_a_p__v)); \ +- else \ +- smp_store_release(&p, RCU_INITIALIZER((typeof(p))_r_a_p__v)); \ +- _r_a_p__v; \ +-}) +- +-/** + * rcu_access_pointer() - fetch RCU pointer with no dereferencing + * @p: The pointer to read + * +@@ -951,10 +922,14 @@ static inline void rcu_read_unlock(void) + static inline void rcu_read_lock_bh(void) + { + local_bh_disable(); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ rcu_read_lock(); ++#else + __acquire(RCU_BH); + rcu_lock_acquire(&rcu_bh_lock_map); + RCU_LOCKDEP_WARN(!rcu_is_watching(), + "rcu_read_lock_bh() used illegally while idle"); ++#endif + } + + /* +@@ -964,10 +939,14 @@ static inline void rcu_read_lock_bh(void) + */ + static inline void rcu_read_unlock_bh(void) + { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ rcu_read_unlock(); ++#else + RCU_LOCKDEP_WARN(!rcu_is_watching(), + "rcu_read_unlock_bh() used illegally while idle"); + rcu_lock_release(&rcu_bh_lock_map); + __release(RCU_BH); ++#endif + local_bh_enable(); + } + +diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h +index 63a4e4cf40a5..08ab12df2863 100644 +--- a/include/linux/rcutree.h ++++ b/include/linux/rcutree.h +@@ -44,7 +44,11 @@ static inline void rcu_virt_note_context_switch(int cpu) + rcu_note_context_switch(); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define synchronize_rcu_bh synchronize_rcu ++#else + void synchronize_rcu_bh(void); ++#endif + void synchronize_sched_expedited(void); + void synchronize_rcu_expedited(void); + +@@ -72,7 +76,11 @@ static inline void synchronize_rcu_bh_expedited(void) + } + + void rcu_barrier(void); ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define rcu_barrier_bh rcu_barrier ++#else + void rcu_barrier_bh(void); ++#endif + void rcu_barrier_sched(void); + unsigned long get_state_synchronize_rcu(void); + void cond_synchronize_rcu(unsigned long oldstate); +@@ -82,17 +90,14 @@ void cond_synchronize_sched(unsigned long oldstate); + extern unsigned long rcutorture_testseq; + extern unsigned long rcutorture_vernum; + unsigned long rcu_batches_started(void); +-unsigned long rcu_batches_started_bh(void); + unsigned long rcu_batches_started_sched(void); + unsigned long rcu_batches_completed(void); +-unsigned long rcu_batches_completed_bh(void); + unsigned long rcu_batches_completed_sched(void); + unsigned long rcu_exp_batches_completed(void); + unsigned long rcu_exp_batches_completed_sched(void); + void show_rcu_gp_kthreads(void); + + void rcu_force_quiescent_state(void); +-void rcu_bh_force_quiescent_state(void); + void rcu_sched_force_quiescent_state(void); + + void rcu_idle_enter(void); +@@ -109,6 +114,16 @@ extern int rcu_scheduler_active __read_mostly; + + bool rcu_is_watching(void); + ++#ifndef CONFIG_PREEMPT_RT_FULL ++void rcu_bh_force_quiescent_state(void); ++unsigned long rcu_batches_started_bh(void); ++unsigned long rcu_batches_completed_bh(void); ++#else ++# define rcu_bh_force_quiescent_state rcu_force_quiescent_state ++# define rcu_batches_completed_bh rcu_batches_completed ++# define rcu_batches_started_bh rcu_batches_completed ++#endif ++ + void rcu_all_qs(void); + + /* RCUtree hotplug events */ +diff --git a/include/linux/rtmutex.h b/include/linux/rtmutex.h +index 1abba5ce2a2f..294a8b4875f1 100644 +--- a/include/linux/rtmutex.h ++++ b/include/linux/rtmutex.h +@@ -13,11 +13,15 @@ + #define __LINUX_RT_MUTEX_H + + #include ++#include + #include +-#include + + extern int max_lock_depth; /* for sysctl */ + ++#ifdef CONFIG_DEBUG_MUTEXES ++#include ++#endif ++ + /** + * The rt_mutex structure + * +@@ -31,8 +35,8 @@ struct rt_mutex { + struct rb_root waiters; + struct rb_node *waiters_leftmost; + struct task_struct *owner; +-#ifdef CONFIG_DEBUG_RT_MUTEXES + int save_state; ++#ifdef CONFIG_DEBUG_RT_MUTEXES + const char *name, *file; + int line; + void *magic; +@@ -55,22 +59,33 @@ struct hrtimer_sleeper; + # define rt_mutex_debug_check_no_locks_held(task) do { } while (0) + #endif + ++# define rt_mutex_init(mutex) \ ++ do { \ ++ raw_spin_lock_init(&(mutex)->wait_lock); \ ++ __rt_mutex_init(mutex, #mutex); \ ++ } while (0) ++ + #ifdef CONFIG_DEBUG_RT_MUTEXES + # define __DEBUG_RT_MUTEX_INITIALIZER(mutexname) \ + , .name = #mutexname, .file = __FILE__, .line = __LINE__ +-# define rt_mutex_init(mutex) __rt_mutex_init(mutex, __func__) + extern void rt_mutex_debug_task_free(struct task_struct *tsk); + #else + # define __DEBUG_RT_MUTEX_INITIALIZER(mutexname) +-# define rt_mutex_init(mutex) __rt_mutex_init(mutex, NULL) + # define rt_mutex_debug_task_free(t) do { } while (0) + #endif + +-#define __RT_MUTEX_INITIALIZER(mutexname) \ +- { .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(mutexname.wait_lock) \ ++#define __RT_MUTEX_INITIALIZER_PLAIN(mutexname) \ ++ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(mutexname.wait_lock) \ + , .waiters = RB_ROOT \ + , .owner = NULL \ +- __DEBUG_RT_MUTEX_INITIALIZER(mutexname)} ++ __DEBUG_RT_MUTEX_INITIALIZER(mutexname) ++ ++#define __RT_MUTEX_INITIALIZER(mutexname) \ ++ { __RT_MUTEX_INITIALIZER_PLAIN(mutexname) } ++ ++#define __RT_MUTEX_INITIALIZER_SAVE_STATE(mutexname) \ ++ { __RT_MUTEX_INITIALIZER_PLAIN(mutexname) \ ++ , .save_state = 1 } + + #define DEFINE_RT_MUTEX(mutexname) \ + struct rt_mutex mutexname = __RT_MUTEX_INITIALIZER(mutexname) +@@ -90,7 +105,9 @@ extern void __rt_mutex_init(struct rt_mutex *lock, const char *name); + extern void rt_mutex_destroy(struct rt_mutex *lock); + + extern void rt_mutex_lock(struct rt_mutex *lock); ++extern int rt_mutex_lock_state(struct rt_mutex *lock, int state); + extern int rt_mutex_lock_interruptible(struct rt_mutex *lock); ++extern int rt_mutex_lock_killable(struct rt_mutex *lock); + extern int rt_mutex_timed_lock(struct rt_mutex *lock, + struct hrtimer_sleeper *timeout); + +diff --git a/include/linux/rwlock_rt.h b/include/linux/rwlock_rt.h +new file mode 100644 +index 000000000000..49ed2d45d3be +--- /dev/null ++++ b/include/linux/rwlock_rt.h +@@ -0,0 +1,99 @@ ++#ifndef __LINUX_RWLOCK_RT_H ++#define __LINUX_RWLOCK_RT_H ++ ++#ifndef __LINUX_SPINLOCK_H ++#error Do not include directly. Use spinlock.h ++#endif ++ ++#define rwlock_init(rwl) \ ++do { \ ++ static struct lock_class_key __key; \ ++ \ ++ rt_mutex_init(&(rwl)->lock); \ ++ __rt_rwlock_init(rwl, #rwl, &__key); \ ++} while (0) ++ ++extern void __lockfunc rt_write_lock(rwlock_t *rwlock); ++extern void __lockfunc rt_read_lock(rwlock_t *rwlock); ++extern int __lockfunc rt_write_trylock(rwlock_t *rwlock); ++extern int __lockfunc rt_write_trylock_irqsave(rwlock_t *trylock, unsigned long *flags); ++extern int __lockfunc rt_read_trylock(rwlock_t *rwlock); ++extern void __lockfunc rt_write_unlock(rwlock_t *rwlock); ++extern void __lockfunc rt_read_unlock(rwlock_t *rwlock); ++extern unsigned long __lockfunc rt_write_lock_irqsave(rwlock_t *rwlock); ++extern unsigned long __lockfunc rt_read_lock_irqsave(rwlock_t *rwlock); ++extern void __rt_rwlock_init(rwlock_t *rwlock, char *name, struct lock_class_key *key); ++ ++#define read_trylock(lock) __cond_lock(lock, rt_read_trylock(lock)) ++#define write_trylock(lock) __cond_lock(lock, rt_write_trylock(lock)) ++ ++#define write_trylock_irqsave(lock, flags) \ ++ __cond_lock(lock, rt_write_trylock_irqsave(lock, &flags)) ++ ++#define read_lock_irqsave(lock, flags) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ flags = rt_read_lock_irqsave(lock); \ ++ } while (0) ++ ++#define write_lock_irqsave(lock, flags) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ flags = rt_write_lock_irqsave(lock); \ ++ } while (0) ++ ++#define read_lock(lock) rt_read_lock(lock) ++ ++#define read_lock_bh(lock) \ ++ do { \ ++ local_bh_disable(); \ ++ rt_read_lock(lock); \ ++ } while (0) ++ ++#define read_lock_irq(lock) read_lock(lock) ++ ++#define write_lock(lock) rt_write_lock(lock) ++ ++#define write_lock_bh(lock) \ ++ do { \ ++ local_bh_disable(); \ ++ rt_write_lock(lock); \ ++ } while (0) ++ ++#define write_lock_irq(lock) write_lock(lock) ++ ++#define read_unlock(lock) rt_read_unlock(lock) ++ ++#define read_unlock_bh(lock) \ ++ do { \ ++ rt_read_unlock(lock); \ ++ local_bh_enable(); \ ++ } while (0) ++ ++#define read_unlock_irq(lock) read_unlock(lock) ++ ++#define write_unlock(lock) rt_write_unlock(lock) ++ ++#define write_unlock_bh(lock) \ ++ do { \ ++ rt_write_unlock(lock); \ ++ local_bh_enable(); \ ++ } while (0) ++ ++#define write_unlock_irq(lock) write_unlock(lock) ++ ++#define read_unlock_irqrestore(lock, flags) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ (void) flags; \ ++ rt_read_unlock(lock); \ ++ } while (0) ++ ++#define write_unlock_irqrestore(lock, flags) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ (void) flags; \ ++ rt_write_unlock(lock); \ ++ } while (0) ++ ++#endif +diff --git a/include/linux/rwlock_types.h b/include/linux/rwlock_types.h +index cc0072e93e36..5317cd957292 100644 +--- a/include/linux/rwlock_types.h ++++ b/include/linux/rwlock_types.h +@@ -1,6 +1,10 @@ + #ifndef __LINUX_RWLOCK_TYPES_H + #define __LINUX_RWLOCK_TYPES_H + ++#if !defined(__LINUX_SPINLOCK_TYPES_H) ++# error "Do not include directly, include spinlock_types.h" ++#endif ++ + /* + * include/linux/rwlock_types.h - generic rwlock type definitions + * and initializers +diff --git a/include/linux/rwlock_types_rt.h b/include/linux/rwlock_types_rt.h +new file mode 100644 +index 000000000000..51b28d775fe1 +--- /dev/null ++++ b/include/linux/rwlock_types_rt.h +@@ -0,0 +1,33 @@ ++#ifndef __LINUX_RWLOCK_TYPES_RT_H ++#define __LINUX_RWLOCK_TYPES_RT_H ++ ++#ifndef __LINUX_SPINLOCK_TYPES_H ++#error "Do not include directly. Include spinlock_types.h instead" ++#endif ++ ++/* ++ * rwlocks - rtmutex which allows single reader recursion ++ */ ++typedef struct { ++ struct rt_mutex lock; ++ int read_depth; ++ unsigned int break_lock; ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ struct lockdep_map dep_map; ++#endif ++} rwlock_t; ++ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++# define RW_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname } ++#else ++# define RW_DEP_MAP_INIT(lockname) ++#endif ++ ++#define __RW_LOCK_UNLOCKED(name) \ ++ { .lock = __RT_MUTEX_INITIALIZER_SAVE_STATE(name.lock), \ ++ RW_DEP_MAP_INIT(name) } ++ ++#define DEFINE_RWLOCK(name) \ ++ rwlock_t name = __RW_LOCK_UNLOCKED(name) ++ ++#endif +diff --git a/include/linux/rwsem.h b/include/linux/rwsem.h +index dd1d14250340..aa2ac1f65c2d 100644 +--- a/include/linux/rwsem.h ++++ b/include/linux/rwsem.h +@@ -19,6 +19,10 @@ + #include + #endif + ++#ifdef CONFIG_PREEMPT_RT_FULL ++#include ++#else /* PREEMPT_RT_FULL */ ++ + struct rw_semaphore; + + #ifdef CONFIG_RWSEM_GENERIC_SPINLOCK +@@ -106,6 +110,13 @@ static inline int rwsem_is_contended(struct rw_semaphore *sem) + return !list_empty(&sem->wait_list); + } + ++#endif /* !PREEMPT_RT_FULL */ ++ ++/* ++ * The functions below are the same for all rwsem implementations including ++ * the RT specific variant. ++ */ ++ + /* + * lock for reading + */ +diff --git a/include/linux/rwsem_rt.h b/include/linux/rwsem_rt.h +new file mode 100644 +index 000000000000..2ffbf093ae92 +--- /dev/null ++++ b/include/linux/rwsem_rt.h +@@ -0,0 +1,67 @@ ++#ifndef _LINUX_RWSEM_RT_H ++#define _LINUX_RWSEM_RT_H ++ ++#ifndef _LINUX_RWSEM_H ++#error "Include rwsem.h" ++#endif ++ ++#include ++#include ++ ++#define READER_BIAS (1U << 31) ++#define WRITER_BIAS (1U << 30) ++ ++struct rw_semaphore { ++ atomic_t readers; ++ struct rt_mutex rtmutex; ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ struct lockdep_map dep_map; ++#endif ++}; ++ ++#define __RWSEM_INITIALIZER(name) \ ++{ \ ++ .readers = ATOMIC_INIT(READER_BIAS), \ ++ .rtmutex = __RT_MUTEX_INITIALIZER(name.rtmutex), \ ++ RW_DEP_MAP_INIT(name) \ ++} ++ ++#define DECLARE_RWSEM(lockname) \ ++ struct rw_semaphore lockname = __RWSEM_INITIALIZER(lockname) ++ ++extern void __rwsem_init(struct rw_semaphore *rwsem, const char *name, ++ struct lock_class_key *key); ++ ++#define __init_rwsem(sem, name, key) \ ++do { \ ++ rt_mutex_init(&(sem)->rtmutex); \ ++ __rwsem_init((sem), (name), (key)); \ ++} while (0) ++ ++#define init_rwsem(sem) \ ++do { \ ++ static struct lock_class_key __key; \ ++ \ ++ __init_rwsem((sem), #sem, &__key); \ ++} while (0) ++ ++static inline int rwsem_is_locked(struct rw_semaphore *sem) ++{ ++ return atomic_read(&sem->readers) != READER_BIAS; ++} ++ ++static inline int rwsem_is_contended(struct rw_semaphore *sem) ++{ ++ return atomic_read(&sem->readers) > 0; ++} ++ ++extern void __down_read(struct rw_semaphore *sem); ++extern int __down_read_trylock(struct rw_semaphore *sem); ++extern void __down_write(struct rw_semaphore *sem); ++extern int __must_check __down_write_killable(struct rw_semaphore *sem); ++extern int __down_write_trylock(struct rw_semaphore *sem); ++extern void __up_read(struct rw_semaphore *sem); ++extern void __up_write(struct rw_semaphore *sem); ++extern void __downgrade_write(struct rw_semaphore *sem); ++ ++#endif +diff --git a/include/linux/sched.h b/include/linux/sched.h +index f425eb3318ab..4d779486ad6b 100644 +--- a/include/linux/sched.h ++++ b/include/linux/sched.h +@@ -26,6 +26,7 @@ struct sched_param { + #include + #include + #include ++#include + + #include + #include +@@ -236,17 +237,13 @@ extern char ___assert_task_state[1 - 2*!!( + + /* Convenience macros for the sake of wake_up */ + #define TASK_NORMAL (TASK_INTERRUPTIBLE | TASK_UNINTERRUPTIBLE) +-#define TASK_ALL (TASK_NORMAL | __TASK_STOPPED | __TASK_TRACED) + + /* get_task_state() */ + #define TASK_REPORT (TASK_RUNNING | TASK_INTERRUPTIBLE | \ + TASK_UNINTERRUPTIBLE | __TASK_STOPPED | \ + __TASK_TRACED | EXIT_ZOMBIE | EXIT_DEAD) + +-#define task_is_traced(task) ((task->state & __TASK_TRACED) != 0) + #define task_is_stopped(task) ((task->state & __TASK_STOPPED) != 0) +-#define task_is_stopped_or_traced(task) \ +- ((task->state & (__TASK_STOPPED | __TASK_TRACED)) != 0) + #define task_contributes_to_load(task) \ + ((task->state & TASK_UNINTERRUPTIBLE) != 0 && \ + (task->flags & PF_FROZEN) == 0 && \ +@@ -312,6 +309,11 @@ extern char ___assert_task_state[1 - 2*!!( + + #endif + ++#define __set_current_state_no_track(state_value) \ ++ do { current->state = (state_value); } while (0) ++#define set_current_state_no_track(state_value) \ ++ set_mb(current->state, (state_value)) ++ + /* Task command name length */ + #define TASK_COMM_LEN 16 + +@@ -1013,8 +1015,18 @@ struct wake_q_head { + struct wake_q_head name = { WAKE_Q_TAIL, &name.first } + + extern void wake_q_add(struct wake_q_head *head, +- struct task_struct *task); +-extern void wake_up_q(struct wake_q_head *head); ++ struct task_struct *task); ++extern void __wake_up_q(struct wake_q_head *head, bool sleeper); ++ ++static inline void wake_up_q(struct wake_q_head *head) ++{ ++ __wake_up_q(head, false); ++} ++ ++static inline void wake_up_q_sleeper(struct wake_q_head *head) ++{ ++ __wake_up_q(head, true); ++} + + /* + * sched-domains (multiprocessor balancing) declarations: +@@ -1481,6 +1493,7 @@ struct task_struct { + struct thread_info thread_info; + #endif + volatile long state; /* -1 unrunnable, 0 runnable, >0 stopped */ ++ volatile long saved_state; /* saved state for "spinlock sleepers" */ + void *stack; + atomic_t usage; + unsigned int flags; /* per process flags, defined below */ +@@ -1520,6 +1533,12 @@ struct task_struct { + #endif + + unsigned int policy; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ int migrate_disable; ++# ifdef CONFIG_SCHED_DEBUG ++ int migrate_disable_atomic; ++# endif ++#endif + int nr_cpus_allowed; + cpumask_t cpus_allowed; + +@@ -1658,6 +1677,9 @@ struct task_struct { + + struct task_cputime cputime_expires; + struct list_head cpu_timers[3]; ++#ifdef CONFIG_PREEMPT_RT_BASE ++ struct task_struct *posix_timer_list; ++#endif + + /* process credentials */ + const struct cred __rcu *ptracer_cred; /* Tracer's credentials at attach */ +@@ -1689,10 +1711,15 @@ struct task_struct { + /* signal handlers */ + struct signal_struct *signal; + struct sighand_struct *sighand; ++ struct sigqueue *sigqueue_cache; + + sigset_t blocked, real_blocked; + sigset_t saved_sigmask; /* restored if set_restore_sigmask() was used */ + struct sigpending pending; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ /* TODO: move me into ->restart_block ? */ ++ struct siginfo forced_info; ++#endif + + unsigned long sas_ss_sp; + size_t sas_ss_size; +@@ -1723,6 +1750,8 @@ struct task_struct { + /* PI waiters blocked on a rt_mutex held by this task */ + struct rb_root pi_waiters; + struct rb_node *pi_waiters_leftmost; ++ /* Updated under owner's pi_lock and rq lock */ ++ struct task_struct *pi_top_task; + /* Deadlock detection and priority inheritance handling */ + struct rt_mutex_waiter *pi_blocked_on; + #endif +@@ -1921,6 +1950,12 @@ struct task_struct { + /* bitmask and counter of trace recursion */ + unsigned long trace_recursion; + #endif /* CONFIG_TRACING */ ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ u64 preempt_timestamp_hist; ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ long timer_offset; ++#endif ++#endif + #ifdef CONFIG_KCOV + /* Coverage collection mode enabled for this task (0 if disabled). */ + enum kcov_mode kcov_mode; +@@ -1946,9 +1981,23 @@ struct task_struct { + unsigned int sequential_io; + unsigned int sequential_io_avg; + #endif ++#ifdef CONFIG_PREEMPT_RT_BASE ++ struct rcu_head put_rcu; ++ int softirq_nestcnt; ++ unsigned int softirqs_raised; ++#endif ++#ifdef CONFIG_PREEMPT_RT_FULL ++# if defined CONFIG_HIGHMEM || defined CONFIG_X86_32 ++ int kmap_idx; ++ pte_t kmap_pte[KM_TYPE_NR]; ++# endif ++#endif + #ifdef CONFIG_DEBUG_ATOMIC_SLEEP + unsigned long task_state_change; + #endif ++#ifdef CONFIG_PREEMPT_RT_FULL ++ int xmit_recursion; ++#endif + int pagefault_disabled; + #ifdef CONFIG_MMU + struct task_struct *oom_reaper_list; +@@ -1988,14 +2037,6 @@ static inline struct vm_struct *task_stack_vm_area(const struct task_struct *t) + } + #endif + +-/* Future-safe accessor for struct task_struct's cpus_allowed. */ +-#define tsk_cpus_allowed(tsk) (&(tsk)->cpus_allowed) +- +-static inline int tsk_nr_cpus_allowed(struct task_struct *p) +-{ +- return p->nr_cpus_allowed; +-} +- + #define TNF_MIGRATED 0x01 + #define TNF_NO_GROUP 0x02 + #define TNF_SHARED 0x04 +@@ -2211,6 +2252,15 @@ extern struct pid *cad_pid; + extern void free_task(struct task_struct *tsk); + #define get_task_struct(tsk) do { atomic_inc(&(tsk)->usage); } while(0) + ++#ifdef CONFIG_PREEMPT_RT_BASE ++extern void __put_task_struct_cb(struct rcu_head *rhp); ++ ++static inline void put_task_struct(struct task_struct *t) ++{ ++ if (atomic_dec_and_test(&t->usage)) ++ call_rcu(&t->put_rcu, __put_task_struct_cb); ++} ++#else + extern void __put_task_struct(struct task_struct *t); + + static inline void put_task_struct(struct task_struct *t) +@@ -2218,6 +2268,7 @@ static inline void put_task_struct(struct task_struct *t) + if (atomic_dec_and_test(&t->usage)) + __put_task_struct(t); + } ++#endif + + struct task_struct *task_rcu_dereference(struct task_struct **ptask); + struct task_struct *try_get_task_struct(struct task_struct **ptask); +@@ -2259,6 +2310,7 @@ extern void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, + /* + * Per process flags + */ ++#define PF_IN_SOFTIRQ 0x00000001 /* Task is serving softirq */ + #define PF_EXITING 0x00000004 /* getting shut down */ + #define PF_EXITPIDONE 0x00000008 /* pi exit done on shut down */ + #define PF_VCPU 0x00000010 /* I'm a virtual CPU */ +@@ -2427,6 +2479,10 @@ extern void do_set_cpus_allowed(struct task_struct *p, + + extern int set_cpus_allowed_ptr(struct task_struct *p, + const struct cpumask *new_mask); ++int migrate_me(void); ++void tell_sched_cpu_down_begin(int cpu); ++void tell_sched_cpu_down_done(int cpu); ++ + #else + static inline void do_set_cpus_allowed(struct task_struct *p, + const struct cpumask *new_mask) +@@ -2439,6 +2495,9 @@ static inline int set_cpus_allowed_ptr(struct task_struct *p, + return -EINVAL; + return 0; + } ++static inline int migrate_me(void) { return 0; } ++static inline void tell_sched_cpu_down_begin(int cpu) { } ++static inline void tell_sched_cpu_down_done(int cpu) { } + #endif + + #ifdef CONFIG_NO_HZ_COMMON +@@ -2677,6 +2736,7 @@ extern void xtime_update(unsigned long ticks); + + extern int wake_up_state(struct task_struct *tsk, unsigned int state); + extern int wake_up_process(struct task_struct *tsk); ++extern int wake_up_lock_sleeper(struct task_struct * tsk); + extern void wake_up_new_task(struct task_struct *tsk); + #ifdef CONFIG_SMP + extern void kick_process(struct task_struct *tsk); +@@ -2885,6 +2945,17 @@ static inline void mmdrop(struct mm_struct *mm) + __mmdrop(mm); + } + ++#ifdef CONFIG_PREEMPT_RT_BASE ++extern void __mmdrop_delayed(struct rcu_head *rhp); ++static inline void mmdrop_delayed(struct mm_struct *mm) ++{ ++ if (atomic_dec_and_test(&mm->mm_count)) ++ call_rcu(&mm->delayed_drop, __mmdrop_delayed); ++} ++#else ++# define mmdrop_delayed(mm) mmdrop(mm) ++#endif ++ + static inline void mmdrop_async_fn(struct work_struct *work) + { + struct mm_struct *mm = container_of(work, struct mm_struct, async_put_work); +@@ -3277,6 +3348,43 @@ static inline int test_tsk_need_resched(struct task_struct *tsk) + return unlikely(test_tsk_thread_flag(tsk,TIF_NEED_RESCHED)); + } + ++#ifdef CONFIG_PREEMPT_LAZY ++static inline void set_tsk_need_resched_lazy(struct task_struct *tsk) ++{ ++ set_tsk_thread_flag(tsk,TIF_NEED_RESCHED_LAZY); ++} ++ ++static inline void clear_tsk_need_resched_lazy(struct task_struct *tsk) ++{ ++ clear_tsk_thread_flag(tsk,TIF_NEED_RESCHED_LAZY); ++} ++ ++static inline int test_tsk_need_resched_lazy(struct task_struct *tsk) ++{ ++ return unlikely(test_tsk_thread_flag(tsk,TIF_NEED_RESCHED_LAZY)); ++} ++ ++static inline int need_resched_lazy(void) ++{ ++ return test_thread_flag(TIF_NEED_RESCHED_LAZY); ++} ++ ++static inline int need_resched_now(void) ++{ ++ return test_thread_flag(TIF_NEED_RESCHED); ++} ++ ++#else ++static inline void clear_tsk_need_resched_lazy(struct task_struct *tsk) { } ++static inline int need_resched_lazy(void) { return 0; } ++ ++static inline int need_resched_now(void) ++{ ++ return test_thread_flag(TIF_NEED_RESCHED); ++} ++ ++#endif ++ + static inline int restart_syscall(void) + { + set_tsk_thread_flag(current, TIF_SIGPENDING); +@@ -3308,6 +3416,51 @@ static inline int signal_pending_state(long state, struct task_struct *p) + return (state & TASK_INTERRUPTIBLE) || __fatal_signal_pending(p); + } + ++static inline bool __task_is_stopped_or_traced(struct task_struct *task) ++{ ++ if (task->state & (__TASK_STOPPED | __TASK_TRACED)) ++ return true; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ if (task->saved_state & (__TASK_STOPPED | __TASK_TRACED)) ++ return true; ++#endif ++ return false; ++} ++ ++static inline bool task_is_stopped_or_traced(struct task_struct *task) ++{ ++ bool traced_stopped; ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ unsigned long flags; ++ ++ raw_spin_lock_irqsave(&task->pi_lock, flags); ++ traced_stopped = __task_is_stopped_or_traced(task); ++ raw_spin_unlock_irqrestore(&task->pi_lock, flags); ++#else ++ traced_stopped = __task_is_stopped_or_traced(task); ++#endif ++ return traced_stopped; ++} ++ ++static inline bool task_is_traced(struct task_struct *task) ++{ ++ bool traced = false; ++ ++ if (task->state & __TASK_TRACED) ++ return true; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ /* in case the task is sleeping on tasklist_lock */ ++ raw_spin_lock_irq(&task->pi_lock); ++ if (task->state & __TASK_TRACED) ++ traced = true; ++ else if (task->saved_state & __TASK_TRACED) ++ traced = true; ++ raw_spin_unlock_irq(&task->pi_lock); ++#endif ++ return traced; ++} ++ + /* + * cond_resched() and cond_resched_lock(): latency reduction via + * explicit rescheduling in places that are safe. The return +@@ -3333,12 +3486,16 @@ extern int __cond_resched_lock(spinlock_t *lock); + __cond_resched_lock(lock); \ + }) + ++#ifndef CONFIG_PREEMPT_RT_FULL + extern int __cond_resched_softirq(void); + + #define cond_resched_softirq() ({ \ + ___might_sleep(__FILE__, __LINE__, SOFTIRQ_DISABLE_OFFSET); \ + __cond_resched_softirq(); \ + }) ++#else ++# define cond_resched_softirq() cond_resched() ++#endif + + static inline void cond_resched_rcu(void) + { +@@ -3513,6 +3670,31 @@ static inline void set_task_cpu(struct task_struct *p, unsigned int cpu) + + #endif /* CONFIG_SMP */ + ++static inline int __migrate_disabled(struct task_struct *p) ++{ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ return p->migrate_disable; ++#else ++ return 0; ++#endif ++} ++ ++/* Future-safe accessor for struct task_struct's cpus_allowed. */ ++static inline const struct cpumask *tsk_cpus_allowed(struct task_struct *p) ++{ ++ if (__migrate_disabled(p)) ++ return cpumask_of(task_cpu(p)); ++ ++ return &p->cpus_allowed; ++} ++ ++static inline int tsk_nr_cpus_allowed(struct task_struct *p) ++{ ++ if (__migrate_disabled(p)) ++ return 1; ++ return p->nr_cpus_allowed; ++} ++ + extern long sched_setaffinity(pid_t pid, const struct cpumask *new_mask); + extern long sched_getaffinity(pid_t pid, struct cpumask *mask); + +diff --git a/include/linux/sched/rt.h b/include/linux/sched/rt.h +index a30b172df6e1..db3e91f2bc03 100644 +--- a/include/linux/sched/rt.h ++++ b/include/linux/sched/rt.h +@@ -16,27 +16,20 @@ static inline int rt_task(struct task_struct *p) + } + + #ifdef CONFIG_RT_MUTEXES +-extern int rt_mutex_getprio(struct task_struct *p); +-extern void rt_mutex_setprio(struct task_struct *p, int prio); +-extern int rt_mutex_get_effective_prio(struct task_struct *task, int newprio); +-extern struct task_struct *rt_mutex_get_top_task(struct task_struct *task); ++/* ++ * Must hold either p->pi_lock or task_rq(p)->lock. ++ */ ++static inline struct task_struct *rt_mutex_get_top_task(struct task_struct *p) ++{ ++ return p->pi_top_task; ++} ++extern void rt_mutex_setprio(struct task_struct *p, struct task_struct *pi_task); + extern void rt_mutex_adjust_pi(struct task_struct *p); + static inline bool tsk_is_pi_blocked(struct task_struct *tsk) + { + return tsk->pi_blocked_on != NULL; + } + #else +-static inline int rt_mutex_getprio(struct task_struct *p) +-{ +- return p->normal_prio; +-} +- +-static inline int rt_mutex_get_effective_prio(struct task_struct *task, +- int newprio) +-{ +- return newprio; +-} +- + static inline struct task_struct *rt_mutex_get_top_task(struct task_struct *task) + { + return NULL; +diff --git a/include/linux/seqlock.h b/include/linux/seqlock.h +index ead97654c4e9..3d7223ffdd3b 100644 +--- a/include/linux/seqlock.h ++++ b/include/linux/seqlock.h +@@ -220,20 +220,30 @@ static inline int read_seqcount_retry(const seqcount_t *s, unsigned start) + return __read_seqcount_retry(s, start); + } + +- +- +-static inline void raw_write_seqcount_begin(seqcount_t *s) ++static inline void __raw_write_seqcount_begin(seqcount_t *s) + { + s->sequence++; + smp_wmb(); + } + +-static inline void raw_write_seqcount_end(seqcount_t *s) ++static inline void raw_write_seqcount_begin(seqcount_t *s) ++{ ++ preempt_disable_rt(); ++ __raw_write_seqcount_begin(s); ++} ++ ++static inline void __raw_write_seqcount_end(seqcount_t *s) + { + smp_wmb(); + s->sequence++; + } + ++static inline void raw_write_seqcount_end(seqcount_t *s) ++{ ++ __raw_write_seqcount_end(s); ++ preempt_enable_rt(); ++} ++ + /** + * raw_write_seqcount_barrier - do a seq write barrier + * @s: pointer to seqcount_t +@@ -428,10 +438,32 @@ typedef struct { + /* + * Read side functions for starting and finalizing a read side section. + */ ++#ifndef CONFIG_PREEMPT_RT_FULL + static inline unsigned read_seqbegin(const seqlock_t *sl) + { + return read_seqcount_begin(&sl->seqcount); + } ++#else ++/* ++ * Starvation safe read side for RT ++ */ ++static inline unsigned read_seqbegin(seqlock_t *sl) ++{ ++ unsigned ret; ++ ++repeat: ++ ret = ACCESS_ONCE(sl->seqcount.sequence); ++ if (unlikely(ret & 1)) { ++ /* ++ * Take the lock and let the writer proceed (i.e. evtl ++ * boost it), otherwise we could loop here forever. ++ */ ++ spin_unlock_wait(&sl->lock); ++ goto repeat; ++ } ++ return ret; ++} ++#endif + + static inline unsigned read_seqretry(const seqlock_t *sl, unsigned start) + { +@@ -446,36 +478,45 @@ static inline unsigned read_seqretry(const seqlock_t *sl, unsigned start) + static inline void write_seqlock(seqlock_t *sl) + { + spin_lock(&sl->lock); +- write_seqcount_begin(&sl->seqcount); ++ __raw_write_seqcount_begin(&sl->seqcount); ++} ++ ++static inline int try_write_seqlock(seqlock_t *sl) ++{ ++ if (spin_trylock(&sl->lock)) { ++ __raw_write_seqcount_begin(&sl->seqcount); ++ return 1; ++ } ++ return 0; + } + + static inline void write_sequnlock(seqlock_t *sl) + { +- write_seqcount_end(&sl->seqcount); ++ __raw_write_seqcount_end(&sl->seqcount); + spin_unlock(&sl->lock); + } + + static inline void write_seqlock_bh(seqlock_t *sl) + { + spin_lock_bh(&sl->lock); +- write_seqcount_begin(&sl->seqcount); ++ __raw_write_seqcount_begin(&sl->seqcount); + } + + static inline void write_sequnlock_bh(seqlock_t *sl) + { +- write_seqcount_end(&sl->seqcount); ++ __raw_write_seqcount_end(&sl->seqcount); + spin_unlock_bh(&sl->lock); + } + + static inline void write_seqlock_irq(seqlock_t *sl) + { + spin_lock_irq(&sl->lock); +- write_seqcount_begin(&sl->seqcount); ++ __raw_write_seqcount_begin(&sl->seqcount); + } + + static inline void write_sequnlock_irq(seqlock_t *sl) + { +- write_seqcount_end(&sl->seqcount); ++ __raw_write_seqcount_end(&sl->seqcount); + spin_unlock_irq(&sl->lock); + } + +@@ -484,7 +525,7 @@ static inline unsigned long __write_seqlock_irqsave(seqlock_t *sl) + unsigned long flags; + + spin_lock_irqsave(&sl->lock, flags); +- write_seqcount_begin(&sl->seqcount); ++ __raw_write_seqcount_begin(&sl->seqcount); + return flags; + } + +@@ -494,7 +535,7 @@ static inline unsigned long __write_seqlock_irqsave(seqlock_t *sl) + static inline void + write_sequnlock_irqrestore(seqlock_t *sl, unsigned long flags) + { +- write_seqcount_end(&sl->seqcount); ++ __raw_write_seqcount_end(&sl->seqcount); + spin_unlock_irqrestore(&sl->lock, flags); + } + +diff --git a/include/linux/signal.h b/include/linux/signal.h +index b63f63eaa39c..295540fdfc72 100644 +--- a/include/linux/signal.h ++++ b/include/linux/signal.h +@@ -233,6 +233,7 @@ static inline void init_sigpending(struct sigpending *sig) + } + + extern void flush_sigqueue(struct sigpending *queue); ++extern void flush_task_sigqueue(struct task_struct *tsk); + + /* Test if 'sig' is valid signal. Use this instead of testing _NSIG directly */ + static inline int valid_signal(unsigned long sig) +diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h +index 32810f279f8e..0db6e31161f6 100644 +--- a/include/linux/skbuff.h ++++ b/include/linux/skbuff.h +@@ -284,6 +284,7 @@ struct sk_buff_head { + + __u32 qlen; + spinlock_t lock; ++ raw_spinlock_t raw_lock; + }; + + struct sk_buff; +@@ -1573,6 +1574,12 @@ static inline void skb_queue_head_init(struct sk_buff_head *list) + __skb_queue_head_init(list); + } + ++static inline void skb_queue_head_init_raw(struct sk_buff_head *list) ++{ ++ raw_spin_lock_init(&list->raw_lock); ++ __skb_queue_head_init(list); ++} ++ + static inline void skb_queue_head_init_class(struct sk_buff_head *list, + struct lock_class_key *class) + { +diff --git a/include/linux/smp.h b/include/linux/smp.h +index 8e0cb7a0f836..891c533724f5 100644 +--- a/include/linux/smp.h ++++ b/include/linux/smp.h +@@ -120,6 +120,13 @@ extern unsigned int setup_max_cpus; + extern void __init setup_nr_cpu_ids(void); + extern void __init smp_init(void); + ++extern int __boot_cpu_id; ++ ++static inline int get_boot_cpu_id(void) ++{ ++ return __boot_cpu_id; ++} ++ + #else /* !SMP */ + + static inline void smp_send_stop(void) { } +@@ -158,6 +165,11 @@ static inline void smp_init(void) { up_late_init(); } + static inline void smp_init(void) { } + #endif + ++static inline int get_boot_cpu_id(void) ++{ ++ return 0; ++} ++ + #endif /* !SMP */ + + /* +@@ -185,6 +197,9 @@ static inline void smp_init(void) { } + #define get_cpu() ({ preempt_disable(); smp_processor_id(); }) + #define put_cpu() preempt_enable() + ++#define get_cpu_light() ({ migrate_disable(); smp_processor_id(); }) ++#define put_cpu_light() migrate_enable() ++ + /* + * Callback to arch code if there's nosmp or maxcpus=0 on the + * boot command line: +diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h +index 47dd0cebd204..b241cc044bd3 100644 +--- a/include/linux/spinlock.h ++++ b/include/linux/spinlock.h +@@ -271,7 +271,11 @@ static inline void do_raw_spin_unlock(raw_spinlock_t *lock) __releases(lock) + #define raw_spin_can_lock(lock) (!raw_spin_is_locked(lock)) + + /* Include rwlock functions */ +-#include ++#ifdef CONFIG_PREEMPT_RT_FULL ++# include ++#else ++# include ++#endif + + /* + * Pull the _spin_*()/_read_*()/_write_*() functions/declarations: +@@ -282,6 +286,10 @@ static inline void do_raw_spin_unlock(raw_spinlock_t *lock) __releases(lock) + # include + #endif + ++#ifdef CONFIG_PREEMPT_RT_FULL ++# include ++#else /* PREEMPT_RT_FULL */ ++ + /* + * Map the spin_lock functions to the raw variants for PREEMPT_RT=n + */ +@@ -416,4 +424,6 @@ extern int _atomic_dec_and_lock(atomic_t *atomic, spinlock_t *lock); + #define atomic_dec_and_lock(atomic, lock) \ + __cond_lock(lock, _atomic_dec_and_lock(atomic, lock)) + ++#endif /* !PREEMPT_RT_FULL */ ++ + #endif /* __LINUX_SPINLOCK_H */ +diff --git a/include/linux/spinlock_api_smp.h b/include/linux/spinlock_api_smp.h +index 5344268e6e62..043263f30e81 100644 +--- a/include/linux/spinlock_api_smp.h ++++ b/include/linux/spinlock_api_smp.h +@@ -189,6 +189,8 @@ static inline int __raw_spin_trylock_bh(raw_spinlock_t *lock) + return 0; + } + +-#include ++#ifndef CONFIG_PREEMPT_RT_FULL ++# include ++#endif + + #endif /* __LINUX_SPINLOCK_API_SMP_H */ +diff --git a/include/linux/spinlock_rt.h b/include/linux/spinlock_rt.h +new file mode 100644 +index 000000000000..43ca841b913a +--- /dev/null ++++ b/include/linux/spinlock_rt.h +@@ -0,0 +1,162 @@ ++#ifndef __LINUX_SPINLOCK_RT_H ++#define __LINUX_SPINLOCK_RT_H ++ ++#ifndef __LINUX_SPINLOCK_H ++#error Do not include directly. Use spinlock.h ++#endif ++ ++#include ++ ++extern void ++__rt_spin_lock_init(spinlock_t *lock, char *name, struct lock_class_key *key); ++ ++#define spin_lock_init(slock) \ ++do { \ ++ static struct lock_class_key __key; \ ++ \ ++ rt_mutex_init(&(slock)->lock); \ ++ __rt_spin_lock_init(slock, #slock, &__key); \ ++} while (0) ++ ++void __lockfunc rt_spin_lock__no_mg(spinlock_t *lock); ++void __lockfunc rt_spin_unlock__no_mg(spinlock_t *lock); ++int __lockfunc rt_spin_trylock__no_mg(spinlock_t *lock); ++ ++extern void __lockfunc rt_spin_lock(spinlock_t *lock); ++extern unsigned long __lockfunc rt_spin_lock_trace_flags(spinlock_t *lock); ++extern void __lockfunc rt_spin_lock_nested(spinlock_t *lock, int subclass); ++extern void __lockfunc rt_spin_unlock(spinlock_t *lock); ++extern void __lockfunc rt_spin_unlock_wait(spinlock_t *lock); ++extern int __lockfunc rt_spin_trylock_irqsave(spinlock_t *lock, unsigned long *flags); ++extern int __lockfunc rt_spin_trylock_bh(spinlock_t *lock); ++extern int __lockfunc rt_spin_trylock(spinlock_t *lock); ++extern int atomic_dec_and_spin_lock(atomic_t *atomic, spinlock_t *lock); ++ ++/* ++ * lockdep-less calls, for derived types like rwlock: ++ * (for trylock they can use rt_mutex_trylock() directly. ++ */ ++extern void __lockfunc __rt_spin_lock__no_mg(struct rt_mutex *lock); ++extern void __lockfunc __rt_spin_lock(struct rt_mutex *lock); ++extern void __lockfunc __rt_spin_unlock(struct rt_mutex *lock); ++ ++#define spin_lock(lock) rt_spin_lock(lock) ++ ++#define spin_lock_bh(lock) \ ++ do { \ ++ local_bh_disable(); \ ++ rt_spin_lock(lock); \ ++ } while (0) ++ ++#define spin_lock_irq(lock) spin_lock(lock) ++ ++#define spin_do_trylock(lock) __cond_lock(lock, rt_spin_trylock(lock)) ++ ++#define spin_trylock(lock) \ ++({ \ ++ int __locked; \ ++ __locked = spin_do_trylock(lock); \ ++ __locked; \ ++}) ++ ++#ifdef CONFIG_LOCKDEP ++# define spin_lock_nested(lock, subclass) \ ++ do { \ ++ rt_spin_lock_nested(lock, subclass); \ ++ } while (0) ++ ++#define spin_lock_bh_nested(lock, subclass) \ ++ do { \ ++ local_bh_disable(); \ ++ rt_spin_lock_nested(lock, subclass); \ ++ } while (0) ++ ++# define spin_lock_irqsave_nested(lock, flags, subclass) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ flags = 0; \ ++ rt_spin_lock_nested(lock, subclass); \ ++ } while (0) ++#else ++# define spin_lock_nested(lock, subclass) spin_lock(lock) ++# define spin_lock_bh_nested(lock, subclass) spin_lock_bh(lock) ++ ++# define spin_lock_irqsave_nested(lock, flags, subclass) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ flags = 0; \ ++ spin_lock(lock); \ ++ } while (0) ++#endif ++ ++#define spin_lock_irqsave(lock, flags) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ flags = 0; \ ++ spin_lock(lock); \ ++ } while (0) ++ ++static inline unsigned long spin_lock_trace_flags(spinlock_t *lock) ++{ ++ unsigned long flags = 0; ++#ifdef CONFIG_TRACE_IRQFLAGS ++ flags = rt_spin_lock_trace_flags(lock); ++#else ++ spin_lock(lock); /* lock_local */ ++#endif ++ return flags; ++} ++ ++/* FIXME: we need rt_spin_lock_nest_lock */ ++#define spin_lock_nest_lock(lock, nest_lock) spin_lock_nested(lock, 0) ++ ++#define spin_unlock(lock) rt_spin_unlock(lock) ++ ++#define spin_unlock_bh(lock) \ ++ do { \ ++ rt_spin_unlock(lock); \ ++ local_bh_enable(); \ ++ } while (0) ++ ++#define spin_unlock_irq(lock) spin_unlock(lock) ++ ++#define spin_unlock_irqrestore(lock, flags) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ (void) flags; \ ++ spin_unlock(lock); \ ++ } while (0) ++ ++#define spin_trylock_bh(lock) __cond_lock(lock, rt_spin_trylock_bh(lock)) ++#define spin_trylock_irq(lock) spin_trylock(lock) ++ ++#define spin_trylock_irqsave(lock, flags) \ ++ rt_spin_trylock_irqsave(lock, &(flags)) ++ ++#define spin_unlock_wait(lock) rt_spin_unlock_wait(lock) ++ ++#ifdef CONFIG_GENERIC_LOCKBREAK ++# define spin_is_contended(lock) ((lock)->break_lock) ++#else ++# define spin_is_contended(lock) (((void)(lock), 0)) ++#endif ++ ++static inline int spin_can_lock(spinlock_t *lock) ++{ ++ return !rt_mutex_is_locked(&lock->lock); ++} ++ ++static inline int spin_is_locked(spinlock_t *lock) ++{ ++ return rt_mutex_is_locked(&lock->lock); ++} ++ ++static inline void assert_spin_locked(spinlock_t *lock) ++{ ++ BUG_ON(!spin_is_locked(lock)); ++} ++ ++#define atomic_dec_and_lock(atomic, lock) \ ++ atomic_dec_and_spin_lock(atomic, lock) ++ ++#endif +diff --git a/include/linux/spinlock_types.h b/include/linux/spinlock_types.h +index 73548eb13a5d..10bac715ea96 100644 +--- a/include/linux/spinlock_types.h ++++ b/include/linux/spinlock_types.h +@@ -9,80 +9,15 @@ + * Released under the General Public License (GPL). + */ + +-#if defined(CONFIG_SMP) +-# include ++#include ++ ++#ifndef CONFIG_PREEMPT_RT_FULL ++# include ++# include + #else +-# include ++# include ++# include ++# include + #endif + +-#include +- +-typedef struct raw_spinlock { +- arch_spinlock_t raw_lock; +-#ifdef CONFIG_GENERIC_LOCKBREAK +- unsigned int break_lock; +-#endif +-#ifdef CONFIG_DEBUG_SPINLOCK +- unsigned int magic, owner_cpu; +- void *owner; +-#endif +-#ifdef CONFIG_DEBUG_LOCK_ALLOC +- struct lockdep_map dep_map; +-#endif +-} raw_spinlock_t; +- +-#define SPINLOCK_MAGIC 0xdead4ead +- +-#define SPINLOCK_OWNER_INIT ((void *)-1L) +- +-#ifdef CONFIG_DEBUG_LOCK_ALLOC +-# define SPIN_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname } +-#else +-# define SPIN_DEP_MAP_INIT(lockname) +-#endif +- +-#ifdef CONFIG_DEBUG_SPINLOCK +-# define SPIN_DEBUG_INIT(lockname) \ +- .magic = SPINLOCK_MAGIC, \ +- .owner_cpu = -1, \ +- .owner = SPINLOCK_OWNER_INIT, +-#else +-# define SPIN_DEBUG_INIT(lockname) +-#endif +- +-#define __RAW_SPIN_LOCK_INITIALIZER(lockname) \ +- { \ +- .raw_lock = __ARCH_SPIN_LOCK_UNLOCKED, \ +- SPIN_DEBUG_INIT(lockname) \ +- SPIN_DEP_MAP_INIT(lockname) } +- +-#define __RAW_SPIN_LOCK_UNLOCKED(lockname) \ +- (raw_spinlock_t) __RAW_SPIN_LOCK_INITIALIZER(lockname) +- +-#define DEFINE_RAW_SPINLOCK(x) raw_spinlock_t x = __RAW_SPIN_LOCK_UNLOCKED(x) +- +-typedef struct spinlock { +- union { +- struct raw_spinlock rlock; +- +-#ifdef CONFIG_DEBUG_LOCK_ALLOC +-# define LOCK_PADSIZE (offsetof(struct raw_spinlock, dep_map)) +- struct { +- u8 __padding[LOCK_PADSIZE]; +- struct lockdep_map dep_map; +- }; +-#endif +- }; +-} spinlock_t; +- +-#define __SPIN_LOCK_INITIALIZER(lockname) \ +- { { .rlock = __RAW_SPIN_LOCK_INITIALIZER(lockname) } } +- +-#define __SPIN_LOCK_UNLOCKED(lockname) \ +- (spinlock_t ) __SPIN_LOCK_INITIALIZER(lockname) +- +-#define DEFINE_SPINLOCK(x) spinlock_t x = __SPIN_LOCK_UNLOCKED(x) +- +-#include +- + #endif /* __LINUX_SPINLOCK_TYPES_H */ +diff --git a/include/linux/spinlock_types_nort.h b/include/linux/spinlock_types_nort.h +new file mode 100644 +index 000000000000..f1dac1fb1d6a +--- /dev/null ++++ b/include/linux/spinlock_types_nort.h +@@ -0,0 +1,33 @@ ++#ifndef __LINUX_SPINLOCK_TYPES_NORT_H ++#define __LINUX_SPINLOCK_TYPES_NORT_H ++ ++#ifndef __LINUX_SPINLOCK_TYPES_H ++#error "Do not include directly. Include spinlock_types.h instead" ++#endif ++ ++/* ++ * The non RT version maps spinlocks to raw_spinlocks ++ */ ++typedef struct spinlock { ++ union { ++ struct raw_spinlock rlock; ++ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++# define LOCK_PADSIZE (offsetof(struct raw_spinlock, dep_map)) ++ struct { ++ u8 __padding[LOCK_PADSIZE]; ++ struct lockdep_map dep_map; ++ }; ++#endif ++ }; ++} spinlock_t; ++ ++#define __SPIN_LOCK_INITIALIZER(lockname) \ ++ { { .rlock = __RAW_SPIN_LOCK_INITIALIZER(lockname) } } ++ ++#define __SPIN_LOCK_UNLOCKED(lockname) \ ++ (spinlock_t ) __SPIN_LOCK_INITIALIZER(lockname) ++ ++#define DEFINE_SPINLOCK(x) spinlock_t x = __SPIN_LOCK_UNLOCKED(x) ++ ++#endif +diff --git a/include/linux/spinlock_types_raw.h b/include/linux/spinlock_types_raw.h +new file mode 100644 +index 000000000000..edffc4d53fc9 +--- /dev/null ++++ b/include/linux/spinlock_types_raw.h +@@ -0,0 +1,56 @@ ++#ifndef __LINUX_SPINLOCK_TYPES_RAW_H ++#define __LINUX_SPINLOCK_TYPES_RAW_H ++ ++#if defined(CONFIG_SMP) ++# include ++#else ++# include ++#endif ++ ++#include ++ ++typedef struct raw_spinlock { ++ arch_spinlock_t raw_lock; ++#ifdef CONFIG_GENERIC_LOCKBREAK ++ unsigned int break_lock; ++#endif ++#ifdef CONFIG_DEBUG_SPINLOCK ++ unsigned int magic, owner_cpu; ++ void *owner; ++#endif ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ struct lockdep_map dep_map; ++#endif ++} raw_spinlock_t; ++ ++#define SPINLOCK_MAGIC 0xdead4ead ++ ++#define SPINLOCK_OWNER_INIT ((void *)-1L) ++ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++# define SPIN_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname } ++#else ++# define SPIN_DEP_MAP_INIT(lockname) ++#endif ++ ++#ifdef CONFIG_DEBUG_SPINLOCK ++# define SPIN_DEBUG_INIT(lockname) \ ++ .magic = SPINLOCK_MAGIC, \ ++ .owner_cpu = -1, \ ++ .owner = SPINLOCK_OWNER_INIT, ++#else ++# define SPIN_DEBUG_INIT(lockname) ++#endif ++ ++#define __RAW_SPIN_LOCK_INITIALIZER(lockname) \ ++ { \ ++ .raw_lock = __ARCH_SPIN_LOCK_UNLOCKED, \ ++ SPIN_DEBUG_INIT(lockname) \ ++ SPIN_DEP_MAP_INIT(lockname) } ++ ++#define __RAW_SPIN_LOCK_UNLOCKED(lockname) \ ++ (raw_spinlock_t) __RAW_SPIN_LOCK_INITIALIZER(lockname) ++ ++#define DEFINE_RAW_SPINLOCK(x) raw_spinlock_t x = __RAW_SPIN_LOCK_UNLOCKED(x) ++ ++#endif +diff --git a/include/linux/spinlock_types_rt.h b/include/linux/spinlock_types_rt.h +new file mode 100644 +index 000000000000..3e3d8c5f7a9a +--- /dev/null ++++ b/include/linux/spinlock_types_rt.h +@@ -0,0 +1,48 @@ ++#ifndef __LINUX_SPINLOCK_TYPES_RT_H ++#define __LINUX_SPINLOCK_TYPES_RT_H ++ ++#ifndef __LINUX_SPINLOCK_TYPES_H ++#error "Do not include directly. Include spinlock_types.h instead" ++#endif ++ ++#include ++ ++/* ++ * PREEMPT_RT: spinlocks - an RT mutex plus lock-break field: ++ */ ++typedef struct spinlock { ++ struct rt_mutex lock; ++ unsigned int break_lock; ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ struct lockdep_map dep_map; ++#endif ++} spinlock_t; ++ ++#ifdef CONFIG_DEBUG_RT_MUTEXES ++# define __RT_SPIN_INITIALIZER(name) \ ++ { \ ++ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \ ++ .save_state = 1, \ ++ .file = __FILE__, \ ++ .line = __LINE__ , \ ++ } ++#else ++# define __RT_SPIN_INITIALIZER(name) \ ++ { \ ++ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \ ++ .save_state = 1, \ ++ } ++#endif ++ ++/* ++.wait_list = PLIST_HEAD_INIT_RAW((name).lock.wait_list, (name).lock.wait_lock) ++*/ ++ ++#define __SPIN_LOCK_UNLOCKED(name) \ ++ { .lock = __RT_SPIN_INITIALIZER(name.lock), \ ++ SPIN_DEP_MAP_INIT(name) } ++ ++#define DEFINE_SPINLOCK(name) \ ++ spinlock_t name = __SPIN_LOCK_UNLOCKED(name) ++ ++#endif +diff --git a/include/linux/srcu.h b/include/linux/srcu.h +index dc8eb63c6568..e793d3a257da 100644 +--- a/include/linux/srcu.h ++++ b/include/linux/srcu.h +@@ -84,10 +84,10 @@ int init_srcu_struct(struct srcu_struct *sp); + + void process_srcu(struct work_struct *work); + +-#define __SRCU_STRUCT_INIT(name) \ ++#define __SRCU_STRUCT_INIT(name, pcpu_name) \ + { \ + .completed = -300, \ +- .per_cpu_ref = &name##_srcu_array, \ ++ .per_cpu_ref = &pcpu_name, \ + .queue_lock = __SPIN_LOCK_UNLOCKED(name.queue_lock), \ + .running = false, \ + .batch_queue = RCU_BATCH_INIT(name.batch_queue), \ +@@ -119,7 +119,7 @@ void process_srcu(struct work_struct *work); + */ + #define __DEFINE_SRCU(name, is_static) \ + static DEFINE_PER_CPU(struct srcu_struct_array, name##_srcu_array);\ +- is_static struct srcu_struct name = __SRCU_STRUCT_INIT(name) ++ is_static struct srcu_struct name = __SRCU_STRUCT_INIT(name, name##_srcu_array) + #define DEFINE_SRCU(name) __DEFINE_SRCU(name, /* not static */) + #define DEFINE_STATIC_SRCU(name) __DEFINE_SRCU(name, static) + +diff --git a/include/linux/suspend.h b/include/linux/suspend.h +index d9718378a8be..e81e6dc7dcb1 100644 +--- a/include/linux/suspend.h ++++ b/include/linux/suspend.h +@@ -193,6 +193,12 @@ struct platform_freeze_ops { + void (*end)(void); + }; + ++#if defined(CONFIG_SUSPEND) || defined(CONFIG_HIBERNATION) ++extern bool pm_in_action; ++#else ++# define pm_in_action false ++#endif ++ + #ifdef CONFIG_SUSPEND + /** + * suspend_set_ops - set platform dependent suspend operations +diff --git a/include/linux/swait.h b/include/linux/swait.h +index c1f9c62a8a50..83f004a72320 100644 +--- a/include/linux/swait.h ++++ b/include/linux/swait.h +@@ -87,6 +87,7 @@ static inline int swait_active(struct swait_queue_head *q) + extern void swake_up(struct swait_queue_head *q); + extern void swake_up_all(struct swait_queue_head *q); + extern void swake_up_locked(struct swait_queue_head *q); ++extern void swake_up_all_locked(struct swait_queue_head *q); + + extern void __prepare_to_swait(struct swait_queue_head *q, struct swait_queue *wait); + extern void prepare_to_swait(struct swait_queue_head *q, struct swait_queue *wait, int state); +diff --git a/include/linux/swap.h b/include/linux/swap.h +index 55ff5593c193..52bf5477dc92 100644 +--- a/include/linux/swap.h ++++ b/include/linux/swap.h +@@ -11,6 +11,7 @@ + #include + #include + #include ++#include + #include + + struct notifier_block; +@@ -247,7 +248,8 @@ struct swap_info_struct { + void *workingset_eviction(struct address_space *mapping, struct page *page); + bool workingset_refault(void *shadow); + void workingset_activation(struct page *page); +-extern struct list_lru workingset_shadow_nodes; ++extern struct list_lru __workingset_shadow_nodes; ++DECLARE_LOCAL_IRQ_LOCK(workingset_shadow_lock); + + static inline unsigned int workingset_node_pages(struct radix_tree_node *node) + { +@@ -292,6 +294,7 @@ extern unsigned long nr_free_pagecache_pages(void); + + + /* linux/mm/swap.c */ ++DECLARE_LOCAL_IRQ_LOCK(swapvec_lock); + extern void lru_cache_add(struct page *); + extern void lru_cache_add_anon(struct page *page); + extern void lru_cache_add_file(struct page *page); +diff --git a/include/linux/swork.h b/include/linux/swork.h +new file mode 100644 +index 000000000000..f175fa9a6016 +--- /dev/null ++++ b/include/linux/swork.h +@@ -0,0 +1,24 @@ ++#ifndef _LINUX_SWORK_H ++#define _LINUX_SWORK_H ++ ++#include ++ ++struct swork_event { ++ struct list_head item; ++ unsigned long flags; ++ void (*func)(struct swork_event *); ++}; ++ ++static inline void INIT_SWORK(struct swork_event *event, ++ void (*func)(struct swork_event *)) ++{ ++ event->flags = 0; ++ event->func = func; ++} ++ ++bool swork_queue(struct swork_event *sev); ++ ++int swork_get(void); ++void swork_put(void); ++ ++#endif /* _LINUX_SWORK_H */ +diff --git a/include/linux/thread_info.h b/include/linux/thread_info.h +index 2873baf5372a..eb1a108f17ca 100644 +--- a/include/linux/thread_info.h ++++ b/include/linux/thread_info.h +@@ -107,7 +107,17 @@ static inline int test_ti_thread_flag(struct thread_info *ti, int flag) + #define test_thread_flag(flag) \ + test_ti_thread_flag(current_thread_info(), flag) + +-#define tif_need_resched() test_thread_flag(TIF_NEED_RESCHED) ++#ifdef CONFIG_PREEMPT_LAZY ++#define tif_need_resched() (test_thread_flag(TIF_NEED_RESCHED) || \ ++ test_thread_flag(TIF_NEED_RESCHED_LAZY)) ++#define tif_need_resched_now() (test_thread_flag(TIF_NEED_RESCHED)) ++#define tif_need_resched_lazy() test_thread_flag(TIF_NEED_RESCHED_LAZY)) ++ ++#else ++#define tif_need_resched() test_thread_flag(TIF_NEED_RESCHED) ++#define tif_need_resched_now() test_thread_flag(TIF_NEED_RESCHED) ++#define tif_need_resched_lazy() 0 ++#endif + + #ifndef CONFIG_HAVE_ARCH_WITHIN_STACK_FRAMES + static inline int arch_within_stack_frames(const void * const stack, +diff --git a/include/linux/timer.h b/include/linux/timer.h +index 51d601f192d4..83cea629efe1 100644 +--- a/include/linux/timer.h ++++ b/include/linux/timer.h +@@ -241,7 +241,7 @@ extern void add_timer(struct timer_list *timer); + + extern int try_to_del_timer_sync(struct timer_list *timer); + +-#ifdef CONFIG_SMP ++#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL) + extern int del_timer_sync(struct timer_list *timer); + #else + # define del_timer_sync(t) del_timer(t) +diff --git a/include/linux/trace_events.h b/include/linux/trace_events.h +index be007610ceb0..15154b13a53b 100644 +--- a/include/linux/trace_events.h ++++ b/include/linux/trace_events.h +@@ -56,6 +56,9 @@ struct trace_entry { + unsigned char flags; + unsigned char preempt_count; + int pid; ++ unsigned short migrate_disable; ++ unsigned short padding; ++ unsigned char preempt_lazy_count; + }; + + #define TRACE_EVENT_TYPE_MAX \ +diff --git a/include/linux/uaccess.h b/include/linux/uaccess.h +index f30c187ed785..83bf0f798426 100644 +--- a/include/linux/uaccess.h ++++ b/include/linux/uaccess.h +@@ -24,6 +24,7 @@ static __always_inline void pagefault_disabled_dec(void) + */ + static inline void pagefault_disable(void) + { ++ migrate_disable(); + pagefault_disabled_inc(); + /* + * make sure to have issued the store before a pagefault +@@ -40,6 +41,7 @@ static inline void pagefault_enable(void) + */ + barrier(); + pagefault_disabled_dec(); ++ migrate_enable(); + } + + /* +diff --git a/include/linux/uprobes.h b/include/linux/uprobes.h +index 4a29c75b146e..0a294e950df8 100644 +--- a/include/linux/uprobes.h ++++ b/include/linux/uprobes.h +@@ -27,6 +27,7 @@ + #include + #include + #include ++#include + + struct vm_area_struct; + struct mm_struct; +diff --git a/include/linux/vmstat.h b/include/linux/vmstat.h +index 613771909b6e..e28c5a43229d 100644 +--- a/include/linux/vmstat.h ++++ b/include/linux/vmstat.h +@@ -33,7 +33,9 @@ DECLARE_PER_CPU(struct vm_event_state, vm_event_states); + */ + static inline void __count_vm_event(enum vm_event_item item) + { ++ preempt_disable_rt(); + raw_cpu_inc(vm_event_states.event[item]); ++ preempt_enable_rt(); + } + + static inline void count_vm_event(enum vm_event_item item) +@@ -43,7 +45,9 @@ static inline void count_vm_event(enum vm_event_item item) + + static inline void __count_vm_events(enum vm_event_item item, long delta) + { ++ preempt_disable_rt(); + raw_cpu_add(vm_event_states.event[item], delta); ++ preempt_enable_rt(); + } + + static inline void count_vm_events(enum vm_event_item item, long delta) +diff --git a/include/linux/wait.h b/include/linux/wait.h +index 2408e8d5c05c..db50d6609195 100644 +--- a/include/linux/wait.h ++++ b/include/linux/wait.h +@@ -8,6 +8,7 @@ + #include + #include + #include ++#include + + typedef struct __wait_queue wait_queue_t; + typedef int (*wait_queue_func_t)(wait_queue_t *wait, unsigned mode, int flags, void *key); +diff --git a/include/net/dst.h b/include/net/dst.h +index 6835d224d47b..55a5a9698f14 100644 +--- a/include/net/dst.h ++++ b/include/net/dst.h +@@ -446,7 +446,7 @@ static inline void dst_confirm(struct dst_entry *dst) + static inline int dst_neigh_output(struct dst_entry *dst, struct neighbour *n, + struct sk_buff *skb) + { +- const struct hh_cache *hh; ++ struct hh_cache *hh; + + if (dst->pending_confirm) { + unsigned long now = jiffies; +diff --git a/include/net/gen_stats.h b/include/net/gen_stats.h +index 231e121cc7d9..d125222b979d 100644 +--- a/include/net/gen_stats.h ++++ b/include/net/gen_stats.h +@@ -5,6 +5,7 @@ + #include + #include + #include ++#include + + struct gnet_stats_basic_cpu { + struct gnet_stats_basic_packed bstats; +@@ -33,11 +34,11 @@ int gnet_stats_start_copy_compat(struct sk_buff *skb, int type, + spinlock_t *lock, struct gnet_dump *d, + int padattr); + +-int gnet_stats_copy_basic(const seqcount_t *running, ++int gnet_stats_copy_basic(net_seqlock_t *running, + struct gnet_dump *d, + struct gnet_stats_basic_cpu __percpu *cpu, + struct gnet_stats_basic_packed *b); +-void __gnet_stats_copy_basic(const seqcount_t *running, ++void __gnet_stats_copy_basic(net_seqlock_t *running, + struct gnet_stats_basic_packed *bstats, + struct gnet_stats_basic_cpu __percpu *cpu, + struct gnet_stats_basic_packed *b); +@@ -55,14 +56,14 @@ int gen_new_estimator(struct gnet_stats_basic_packed *bstats, + struct gnet_stats_basic_cpu __percpu *cpu_bstats, + struct gnet_stats_rate_est64 *rate_est, + spinlock_t *stats_lock, +- seqcount_t *running, struct nlattr *opt); ++ net_seqlock_t *running, struct nlattr *opt); + void gen_kill_estimator(struct gnet_stats_basic_packed *bstats, + struct gnet_stats_rate_est64 *rate_est); + int gen_replace_estimator(struct gnet_stats_basic_packed *bstats, + struct gnet_stats_basic_cpu __percpu *cpu_bstats, + struct gnet_stats_rate_est64 *rate_est, + spinlock_t *stats_lock, +- seqcount_t *running, struct nlattr *opt); ++ net_seqlock_t *running, struct nlattr *opt); + bool gen_estimator_active(const struct gnet_stats_basic_packed *bstats, + const struct gnet_stats_rate_est64 *rate_est); + #endif +diff --git a/include/net/neighbour.h b/include/net/neighbour.h +index 8b683841e574..bf656008f6e7 100644 +--- a/include/net/neighbour.h ++++ b/include/net/neighbour.h +@@ -446,7 +446,7 @@ static inline int neigh_hh_bridge(struct hh_cache *hh, struct sk_buff *skb) + } + #endif + +-static inline int neigh_hh_output(const struct hh_cache *hh, struct sk_buff *skb) ++static inline int neigh_hh_output(struct hh_cache *hh, struct sk_buff *skb) + { + unsigned int seq; + int hh_len; +@@ -501,7 +501,7 @@ struct neighbour_cb { + + #define NEIGH_CB(skb) ((struct neighbour_cb *)(skb)->cb) + +-static inline void neigh_ha_snapshot(char *dst, const struct neighbour *n, ++static inline void neigh_ha_snapshot(char *dst, struct neighbour *n, + const struct net_device *dev) + { + unsigned int seq; +diff --git a/include/net/net_seq_lock.h b/include/net/net_seq_lock.h +new file mode 100644 +index 000000000000..a7034298a82a +--- /dev/null ++++ b/include/net/net_seq_lock.h +@@ -0,0 +1,15 @@ ++#ifndef __NET_NET_SEQ_LOCK_H__ ++#define __NET_NET_SEQ_LOCK_H__ ++ ++#ifdef CONFIG_PREEMPT_RT_BASE ++# define net_seqlock_t seqlock_t ++# define net_seq_begin(__r) read_seqbegin(__r) ++# define net_seq_retry(__r, __s) read_seqretry(__r, __s) ++ ++#else ++# define net_seqlock_t seqcount_t ++# define net_seq_begin(__r) read_seqcount_begin(__r) ++# define net_seq_retry(__r, __s) read_seqcount_retry(__r, __s) ++#endif ++ ++#endif +diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h +index 7adf4386ac8f..d3fd5c357268 100644 +--- a/include/net/netns/ipv4.h ++++ b/include/net/netns/ipv4.h +@@ -69,6 +69,7 @@ struct netns_ipv4 { + + int sysctl_icmp_echo_ignore_all; + int sysctl_icmp_echo_ignore_broadcasts; ++ int sysctl_icmp_echo_sysrq; + int sysctl_icmp_ignore_bogus_error_responses; + int sysctl_icmp_ratelimit; + int sysctl_icmp_ratemask; +diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h +index e6aa0a249672..b57736f2a8a3 100644 +--- a/include/net/sch_generic.h ++++ b/include/net/sch_generic.h +@@ -10,6 +10,7 @@ + #include + #include + #include ++#include + + struct Qdisc_ops; + struct qdisc_walker; +@@ -86,7 +87,7 @@ struct Qdisc { + struct sk_buff *gso_skb ____cacheline_aligned_in_smp; + struct qdisc_skb_head q; + struct gnet_stats_basic_packed bstats; +- seqcount_t running; ++ net_seqlock_t running; + struct gnet_stats_queue qstats; + unsigned long state; + struct Qdisc *next_sched; +@@ -98,13 +99,22 @@ struct Qdisc { + spinlock_t busylock ____cacheline_aligned_in_smp; + }; + +-static inline bool qdisc_is_running(const struct Qdisc *qdisc) ++static inline bool qdisc_is_running(struct Qdisc *qdisc) + { ++#ifdef CONFIG_PREEMPT_RT_BASE ++ return spin_is_locked(&qdisc->running.lock) ? true : false; ++#else + return (raw_read_seqcount(&qdisc->running) & 1) ? true : false; ++#endif + } + + static inline bool qdisc_run_begin(struct Qdisc *qdisc) + { ++#ifdef CONFIG_PREEMPT_RT_BASE ++ if (try_write_seqlock(&qdisc->running)) ++ return true; ++ return false; ++#else + if (qdisc_is_running(qdisc)) + return false; + /* Variant of write_seqcount_begin() telling lockdep a trylock +@@ -113,11 +123,16 @@ static inline bool qdisc_run_begin(struct Qdisc *qdisc) + raw_write_seqcount_begin(&qdisc->running); + seqcount_acquire(&qdisc->running.dep_map, 0, 1, _RET_IP_); + return true; ++#endif + } + + static inline void qdisc_run_end(struct Qdisc *qdisc) + { ++#ifdef CONFIG_PREEMPT_RT_BASE ++ write_sequnlock(&qdisc->running); ++#else + write_seqcount_end(&qdisc->running); ++#endif + } + + static inline bool qdisc_may_bulk(const struct Qdisc *qdisc) +@@ -308,7 +323,7 @@ static inline spinlock_t *qdisc_root_sleeping_lock(const struct Qdisc *qdisc) + return qdisc_lock(root); + } + +-static inline seqcount_t *qdisc_root_sleeping_running(const struct Qdisc *qdisc) ++static inline net_seqlock_t *qdisc_root_sleeping_running(const struct Qdisc *qdisc) + { + struct Qdisc *root = qdisc_root_sleeping(qdisc); + +diff --git a/include/trace/events/hist.h b/include/trace/events/hist.h +new file mode 100644 +index 000000000000..f7710de1b1f3 +--- /dev/null ++++ b/include/trace/events/hist.h +@@ -0,0 +1,73 @@ ++#undef TRACE_SYSTEM ++#define TRACE_SYSTEM hist ++ ++#if !defined(_TRACE_HIST_H) || defined(TRACE_HEADER_MULTI_READ) ++#define _TRACE_HIST_H ++ ++#include "latency_hist.h" ++#include ++ ++#if !defined(CONFIG_PREEMPT_OFF_HIST) && !defined(CONFIG_INTERRUPT_OFF_HIST) ++#define trace_preemptirqsoff_hist(a, b) ++#define trace_preemptirqsoff_hist_rcuidle(a, b) ++#else ++TRACE_EVENT(preemptirqsoff_hist, ++ ++ TP_PROTO(int reason, int starthist), ++ ++ TP_ARGS(reason, starthist), ++ ++ TP_STRUCT__entry( ++ __field(int, reason) ++ __field(int, starthist) ++ ), ++ ++ TP_fast_assign( ++ __entry->reason = reason; ++ __entry->starthist = starthist; ++ ), ++ ++ TP_printk("reason=%s starthist=%s", getaction(__entry->reason), ++ __entry->starthist ? "start" : "stop") ++); ++#endif ++ ++#ifndef CONFIG_MISSED_TIMER_OFFSETS_HIST ++#define trace_hrtimer_interrupt(a, b, c, d) ++#else ++TRACE_EVENT(hrtimer_interrupt, ++ ++ TP_PROTO(int cpu, long long offset, struct task_struct *curr, ++ struct task_struct *task), ++ ++ TP_ARGS(cpu, offset, curr, task), ++ ++ TP_STRUCT__entry( ++ __field(int, cpu) ++ __field(long long, offset) ++ __array(char, ccomm, TASK_COMM_LEN) ++ __field(int, cprio) ++ __array(char, tcomm, TASK_COMM_LEN) ++ __field(int, tprio) ++ ), ++ ++ TP_fast_assign( ++ __entry->cpu = cpu; ++ __entry->offset = offset; ++ memcpy(__entry->ccomm, curr->comm, TASK_COMM_LEN); ++ __entry->cprio = curr->prio; ++ memcpy(__entry->tcomm, task != NULL ? task->comm : "", ++ task != NULL ? TASK_COMM_LEN : 7); ++ __entry->tprio = task != NULL ? task->prio : -1; ++ ), ++ ++ TP_printk("cpu=%d offset=%lld curr=%s[%d] thread=%s[%d]", ++ __entry->cpu, __entry->offset, __entry->ccomm, ++ __entry->cprio, __entry->tcomm, __entry->tprio) ++); ++#endif ++ ++#endif /* _TRACE_HIST_H */ ++ ++/* This part must be outside protection */ ++#include +diff --git a/include/trace/events/latency_hist.h b/include/trace/events/latency_hist.h +new file mode 100644 +index 000000000000..d3f2fbd560b1 +--- /dev/null ++++ b/include/trace/events/latency_hist.h +@@ -0,0 +1,29 @@ ++#ifndef _LATENCY_HIST_H ++#define _LATENCY_HIST_H ++ ++enum hist_action { ++ IRQS_ON, ++ PREEMPT_ON, ++ TRACE_STOP, ++ IRQS_OFF, ++ PREEMPT_OFF, ++ TRACE_START, ++}; ++ ++static char *actions[] = { ++ "IRQS_ON", ++ "PREEMPT_ON", ++ "TRACE_STOP", ++ "IRQS_OFF", ++ "PREEMPT_OFF", ++ "TRACE_START", ++}; ++ ++static inline char *getaction(int action) ++{ ++ if (action >= 0 && action <= sizeof(actions)/sizeof(actions[0])) ++ return actions[action]; ++ return "unknown"; ++} ++ ++#endif /* _LATENCY_HIST_H */ +diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h +index 9b90c57517a9..516ae88cddf4 100644 +--- a/include/trace/events/sched.h ++++ b/include/trace/events/sched.h +@@ -70,7 +70,7 @@ DECLARE_EVENT_CLASS(sched_wakeup_template, + TP_fast_assign( + memcpy(__entry->comm, p->comm, TASK_COMM_LEN); + __entry->pid = p->pid; +- __entry->prio = p->prio; ++ __entry->prio = p->prio; /* XXX SCHED_DEADLINE */ + __entry->success = 1; /* rudiment, kill when possible */ + __entry->target_cpu = task_cpu(p); + ), +@@ -147,6 +147,7 @@ TRACE_EVENT(sched_switch, + memcpy(__entry->prev_comm, prev->comm, TASK_COMM_LEN); + __entry->next_pid = next->pid; + __entry->next_prio = next->prio; ++ /* XXX SCHED_DEADLINE */ + ), + + TP_printk("prev_comm=%s prev_pid=%d prev_prio=%d prev_state=%s%s ==> next_comm=%s next_pid=%d next_prio=%d", +@@ -181,7 +182,7 @@ TRACE_EVENT(sched_migrate_task, + TP_fast_assign( + memcpy(__entry->comm, p->comm, TASK_COMM_LEN); + __entry->pid = p->pid; +- __entry->prio = p->prio; ++ __entry->prio = p->prio; /* XXX SCHED_DEADLINE */ + __entry->orig_cpu = task_cpu(p); + __entry->dest_cpu = dest_cpu; + ), +@@ -206,7 +207,7 @@ DECLARE_EVENT_CLASS(sched_process_template, + TP_fast_assign( + memcpy(__entry->comm, p->comm, TASK_COMM_LEN); + __entry->pid = p->pid; +- __entry->prio = p->prio; ++ __entry->prio = p->prio; /* XXX SCHED_DEADLINE */ + ), + + TP_printk("comm=%s pid=%d prio=%d", +@@ -253,7 +254,7 @@ TRACE_EVENT(sched_process_wait, + TP_fast_assign( + memcpy(__entry->comm, current->comm, TASK_COMM_LEN); + __entry->pid = pid_nr(pid); +- __entry->prio = current->prio; ++ __entry->prio = current->prio; /* XXX SCHED_DEADLINE */ + ), + + TP_printk("comm=%s pid=%d prio=%d", +@@ -413,9 +414,9 @@ DEFINE_EVENT(sched_stat_runtime, sched_stat_runtime, + */ + TRACE_EVENT(sched_pi_setprio, + +- TP_PROTO(struct task_struct *tsk, int newprio), ++ TP_PROTO(struct task_struct *tsk, struct task_struct *pi_task), + +- TP_ARGS(tsk, newprio), ++ TP_ARGS(tsk, pi_task), + + TP_STRUCT__entry( + __array( char, comm, TASK_COMM_LEN ) +@@ -428,7 +429,8 @@ TRACE_EVENT(sched_pi_setprio, + memcpy(__entry->comm, tsk->comm, TASK_COMM_LEN); + __entry->pid = tsk->pid; + __entry->oldprio = tsk->prio; +- __entry->newprio = newprio; ++ __entry->newprio = pi_task ? pi_task->prio : tsk->prio; ++ /* XXX SCHED_DEADLINE bits missing */ + ), + + TP_printk("comm=%s pid=%d oldprio=%d newprio=%d", +diff --git a/init/Kconfig b/init/Kconfig +index 34407f15e6d3..2ce33a32e65d 100644 +--- a/init/Kconfig ++++ b/init/Kconfig +@@ -506,7 +506,7 @@ config TINY_RCU + + config RCU_EXPERT + bool "Make expert-level adjustments to RCU configuration" +- default n ++ default y if PREEMPT_RT_FULL + help + This option needs to be enabled if you wish to make + expert-level adjustments to RCU configuration. By default, +@@ -623,7 +623,7 @@ config RCU_FANOUT_LEAF + + config RCU_FAST_NO_HZ + bool "Accelerate last non-dyntick-idle CPU's grace periods" +- depends on NO_HZ_COMMON && SMP && RCU_EXPERT ++ depends on NO_HZ_COMMON && SMP && RCU_EXPERT && !PREEMPT_RT_FULL + default n + help + This option permits CPUs to enter dynticks-idle state even if +@@ -650,7 +650,7 @@ config TREE_RCU_TRACE + config RCU_BOOST + bool "Enable RCU priority boosting" + depends on RT_MUTEXES && PREEMPT_RCU && RCU_EXPERT +- default n ++ default y if PREEMPT_RT_FULL + help + This option boosts the priority of preempted RCU readers that + block the current preemptible RCU grace period for too long. +@@ -781,19 +781,6 @@ config RCU_NOCB_CPU_ALL + + endchoice + +-config RCU_EXPEDITE_BOOT +- bool +- default n +- help +- This option enables expedited grace periods at boot time, +- as if rcu_expedite_gp() had been invoked early in boot. +- The corresponding rcu_unexpedite_gp() is invoked from +- rcu_end_inkernel_boot(), which is intended to be invoked +- at the end of the kernel-only boot sequence, just before +- init is exec'ed. +- +- Accept the default if unsure. +- + endmenu # "RCU Subsystem" + + config BUILD_BIN2C +@@ -1064,6 +1051,7 @@ config CFS_BANDWIDTH + config RT_GROUP_SCHED + bool "Group scheduling for SCHED_RR/FIFO" + depends on CGROUP_SCHED ++ depends on !PREEMPT_RT_FULL + default n + help + This feature lets you explicitly allocate real CPU bandwidth +@@ -1772,6 +1760,7 @@ choice + + config SLAB + bool "SLAB" ++ depends on !PREEMPT_RT_FULL + select HAVE_HARDENED_USERCOPY_ALLOCATOR + help + The regular slab allocator that is established and known to work +@@ -1792,6 +1781,7 @@ config SLUB + config SLOB + depends on EXPERT + bool "SLOB (Simple Allocator)" ++ depends on !PREEMPT_RT_FULL + help + SLOB replaces the stock allocator with a drastically simpler + allocator. SLOB is generally more space efficient but +@@ -1810,7 +1800,7 @@ config SLAB_FREELIST_RANDOM + + config SLUB_CPU_PARTIAL + default y +- depends on SLUB && SMP ++ depends on SLUB && SMP && !PREEMPT_RT_FULL + bool "SLUB per cpu partial cache" + help + Per cpu partial caches accellerate objects allocation and freeing +diff --git a/init/Makefile b/init/Makefile +index c4fb45525d08..821190dfaa75 100644 +--- a/init/Makefile ++++ b/init/Makefile +@@ -35,4 +35,4 @@ $(obj)/version.o: include/generated/compile.h + include/generated/compile.h: FORCE + @$($(quiet)chk_compile.h) + $(Q)$(CONFIG_SHELL) $(srctree)/scripts/mkcompile_h $@ \ +- "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CC) $(KBUILD_CFLAGS)" ++ "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CONFIG_PREEMPT_RT_FULL)" "$(CC) $(KBUILD_CFLAGS)" +diff --git a/init/main.c b/init/main.c +index ae3996ae9bac..6470deef01c9 100644 +--- a/init/main.c ++++ b/init/main.c +@@ -507,6 +507,7 @@ asmlinkage __visible void __init start_kernel(void) + setup_command_line(command_line); + setup_nr_cpu_ids(); + setup_per_cpu_areas(); ++ softirq_early_init(); + boot_cpu_state_init(); + smp_prepare_boot_cpu(); /* arch-specific boot-cpu hooks */ + +diff --git a/ipc/sem.c b/ipc/sem.c +index 10b94bc59d4a..b8360eaacc7a 100644 +--- a/ipc/sem.c ++++ b/ipc/sem.c +@@ -712,6 +712,13 @@ static int perform_atomic_semop(struct sem_array *sma, struct sem_queue *q) + static void wake_up_sem_queue_prepare(struct list_head *pt, + struct sem_queue *q, int error) + { ++#ifdef CONFIG_PREEMPT_RT_BASE ++ struct task_struct *p = q->sleeper; ++ get_task_struct(p); ++ q->status = error; ++ wake_up_process(p); ++ put_task_struct(p); ++#else + if (list_empty(pt)) { + /* + * Hold preempt off so that we don't get preempted and have the +@@ -723,6 +730,7 @@ static void wake_up_sem_queue_prepare(struct list_head *pt, + q->pid = error; + + list_add_tail(&q->list, pt); ++#endif + } + + /** +@@ -736,6 +744,7 @@ static void wake_up_sem_queue_prepare(struct list_head *pt, + */ + static void wake_up_sem_queue_do(struct list_head *pt) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + struct sem_queue *q, *t; + int did_something; + +@@ -748,6 +757,7 @@ static void wake_up_sem_queue_do(struct list_head *pt) + } + if (did_something) + preempt_enable(); ++#endif + } + + static void unlink_queue(struct sem_array *sma, struct sem_queue *q) +diff --git a/kernel/Kconfig.locks b/kernel/Kconfig.locks +index ebdb0043203a..b9e6aa7e5aa6 100644 +--- a/kernel/Kconfig.locks ++++ b/kernel/Kconfig.locks +@@ -225,11 +225,11 @@ config ARCH_SUPPORTS_ATOMIC_RMW + + config MUTEX_SPIN_ON_OWNER + def_bool y +- depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW ++ depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL + + config RWSEM_SPIN_ON_OWNER + def_bool y +- depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW ++ depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL + + config LOCK_SPIN_ON_OWNER + def_bool y +diff --git a/kernel/Kconfig.preempt b/kernel/Kconfig.preempt +index 3f9c97419f02..11dbe26a8279 100644 +--- a/kernel/Kconfig.preempt ++++ b/kernel/Kconfig.preempt +@@ -1,3 +1,16 @@ ++config PREEMPT ++ bool ++ select PREEMPT_COUNT ++ ++config PREEMPT_RT_BASE ++ bool ++ select PREEMPT ++ ++config HAVE_PREEMPT_LAZY ++ bool ++ ++config PREEMPT_LAZY ++ def_bool y if HAVE_PREEMPT_LAZY && PREEMPT_RT_FULL + + choice + prompt "Preemption Model" +@@ -33,9 +46,9 @@ config PREEMPT_VOLUNTARY + + Select this if you are building a kernel for a desktop system. + +-config PREEMPT ++config PREEMPT__LL + bool "Preemptible Kernel (Low-Latency Desktop)" +- select PREEMPT_COUNT ++ select PREEMPT + select UNINLINE_SPIN_UNLOCK if !ARCH_INLINE_SPIN_UNLOCK + help + This option reduces the latency of the kernel by making +@@ -52,6 +65,22 @@ config PREEMPT + embedded system with latency requirements in the milliseconds + range. + ++config PREEMPT_RTB ++ bool "Preemptible Kernel (Basic RT)" ++ select PREEMPT_RT_BASE ++ help ++ This option is basically the same as (Low-Latency Desktop) but ++ enables changes which are preliminary for the full preemptible ++ RT kernel. ++ ++config PREEMPT_RT_FULL ++ bool "Fully Preemptible Kernel (RT)" ++ depends on IRQ_FORCED_THREADING ++ select PREEMPT_RT_BASE ++ select PREEMPT_RCU ++ help ++ All and everything ++ + endchoice + + config PREEMPT_COUNT +diff --git a/kernel/cgroup.c b/kernel/cgroup.c +index a3d2aad2443f..bb6b252648ff 100644 +--- a/kernel/cgroup.c ++++ b/kernel/cgroup.c +@@ -5041,10 +5041,10 @@ static void css_free_rcu_fn(struct rcu_head *rcu_head) + queue_work(cgroup_destroy_wq, &css->destroy_work); + } + +-static void css_release_work_fn(struct work_struct *work) ++static void css_release_work_fn(struct swork_event *sev) + { + struct cgroup_subsys_state *css = +- container_of(work, struct cgroup_subsys_state, destroy_work); ++ container_of(sev, struct cgroup_subsys_state, destroy_swork); + struct cgroup_subsys *ss = css->ss; + struct cgroup *cgrp = css->cgroup; + +@@ -5087,8 +5087,8 @@ static void css_release(struct percpu_ref *ref) + struct cgroup_subsys_state *css = + container_of(ref, struct cgroup_subsys_state, refcnt); + +- INIT_WORK(&css->destroy_work, css_release_work_fn); +- queue_work(cgroup_destroy_wq, &css->destroy_work); ++ INIT_SWORK(&css->destroy_swork, css_release_work_fn); ++ swork_queue(&css->destroy_swork); + } + + static void init_and_link_css(struct cgroup_subsys_state *css, +@@ -5740,6 +5740,7 @@ static int __init cgroup_wq_init(void) + */ + cgroup_destroy_wq = alloc_workqueue("cgroup_destroy", 0, 1); + BUG_ON(!cgroup_destroy_wq); ++ BUG_ON(swork_get()); + + /* + * Used to destroy pidlists and separate to serve as flush domain. +diff --git a/kernel/cpu.c b/kernel/cpu.c +index 99c6c568bc55..f1c64e563970 100644 +--- a/kernel/cpu.c ++++ b/kernel/cpu.c +@@ -239,6 +239,289 @@ static struct { + #define cpuhp_lock_acquire() lock_map_acquire(&cpu_hotplug.dep_map) + #define cpuhp_lock_release() lock_map_release(&cpu_hotplug.dep_map) + ++/** ++ * hotplug_pcp - per cpu hotplug descriptor ++ * @unplug: set when pin_current_cpu() needs to sync tasks ++ * @sync_tsk: the task that waits for tasks to finish pinned sections ++ * @refcount: counter of tasks in pinned sections ++ * @grab_lock: set when the tasks entering pinned sections should wait ++ * @synced: notifier for @sync_tsk to tell cpu_down it's finished ++ * @mutex: the mutex to make tasks wait (used when @grab_lock is true) ++ * @mutex_init: zero if the mutex hasn't been initialized yet. ++ * ++ * Although @unplug and @sync_tsk may point to the same task, the @unplug ++ * is used as a flag and still exists after @sync_tsk has exited and ++ * @sync_tsk set to NULL. ++ */ ++struct hotplug_pcp { ++ struct task_struct *unplug; ++ struct task_struct *sync_tsk; ++ int refcount; ++ int grab_lock; ++ struct completion synced; ++ struct completion unplug_wait; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ /* ++ * Note, on PREEMPT_RT, the hotplug lock must save the state of ++ * the task, otherwise the mutex will cause the task to fail ++ * to sleep when required. (Because it's called from migrate_disable()) ++ * ++ * The spinlock_t on PREEMPT_RT is a mutex that saves the task's ++ * state. ++ */ ++ spinlock_t lock; ++#else ++ struct mutex mutex; ++#endif ++ int mutex_init; ++}; ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define hotplug_lock(hp) rt_spin_lock__no_mg(&(hp)->lock) ++# define hotplug_unlock(hp) rt_spin_unlock__no_mg(&(hp)->lock) ++#else ++# define hotplug_lock(hp) mutex_lock(&(hp)->mutex) ++# define hotplug_unlock(hp) mutex_unlock(&(hp)->mutex) ++#endif ++ ++static DEFINE_PER_CPU(struct hotplug_pcp, hotplug_pcp); ++ ++/** ++ * pin_current_cpu - Prevent the current cpu from being unplugged ++ * ++ * Lightweight version of get_online_cpus() to prevent cpu from being ++ * unplugged when code runs in a migration disabled region. ++ * ++ * Must be called with preemption disabled (preempt_count = 1)! ++ */ ++void pin_current_cpu(void) ++{ ++ struct hotplug_pcp *hp; ++ int force = 0; ++ ++retry: ++ hp = this_cpu_ptr(&hotplug_pcp); ++ ++ if (!hp->unplug || hp->refcount || force || preempt_count() > 1 || ++ hp->unplug == current) { ++ hp->refcount++; ++ return; ++ } ++ if (hp->grab_lock) { ++ preempt_enable(); ++ hotplug_lock(hp); ++ hotplug_unlock(hp); ++ } else { ++ preempt_enable(); ++ /* ++ * Try to push this task off of this CPU. ++ */ ++ if (!migrate_me()) { ++ preempt_disable(); ++ hp = this_cpu_ptr(&hotplug_pcp); ++ if (!hp->grab_lock) { ++ /* ++ * Just let it continue it's already pinned ++ * or about to sleep. ++ */ ++ force = 1; ++ goto retry; ++ } ++ preempt_enable(); ++ } ++ } ++ preempt_disable(); ++ goto retry; ++} ++ ++/** ++ * unpin_current_cpu - Allow unplug of current cpu ++ * ++ * Must be called with preemption or interrupts disabled! ++ */ ++void unpin_current_cpu(void) ++{ ++ struct hotplug_pcp *hp = this_cpu_ptr(&hotplug_pcp); ++ ++ WARN_ON(hp->refcount <= 0); ++ ++ /* This is safe. sync_unplug_thread is pinned to this cpu */ ++ if (!--hp->refcount && hp->unplug && hp->unplug != current) ++ wake_up_process(hp->unplug); ++} ++ ++static void wait_for_pinned_cpus(struct hotplug_pcp *hp) ++{ ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ while (hp->refcount) { ++ schedule_preempt_disabled(); ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ } ++} ++ ++static int sync_unplug_thread(void *data) ++{ ++ struct hotplug_pcp *hp = data; ++ ++ wait_for_completion(&hp->unplug_wait); ++ preempt_disable(); ++ hp->unplug = current; ++ wait_for_pinned_cpus(hp); ++ ++ /* ++ * This thread will synchronize the cpu_down() with threads ++ * that have pinned the CPU. When the pinned CPU count reaches ++ * zero, we inform the cpu_down code to continue to the next step. ++ */ ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ preempt_enable(); ++ complete(&hp->synced); ++ ++ /* ++ * If all succeeds, the next step will need tasks to wait till ++ * the CPU is offline before continuing. To do this, the grab_lock ++ * is set and tasks going into pin_current_cpu() will block on the ++ * mutex. But we still need to wait for those that are already in ++ * pinned CPU sections. If the cpu_down() failed, the kthread_should_stop() ++ * will kick this thread out. ++ */ ++ while (!hp->grab_lock && !kthread_should_stop()) { ++ schedule(); ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ } ++ ++ /* Make sure grab_lock is seen before we see a stale completion */ ++ smp_mb(); ++ ++ /* ++ * Now just before cpu_down() enters stop machine, we need to make ++ * sure all tasks that are in pinned CPU sections are out, and new ++ * tasks will now grab the lock, keeping them from entering pinned ++ * CPU sections. ++ */ ++ if (!kthread_should_stop()) { ++ preempt_disable(); ++ wait_for_pinned_cpus(hp); ++ preempt_enable(); ++ complete(&hp->synced); ++ } ++ ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ while (!kthread_should_stop()) { ++ schedule(); ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ } ++ set_current_state(TASK_RUNNING); ++ ++ /* ++ * Force this thread off this CPU as it's going down and ++ * we don't want any more work on this CPU. ++ */ ++ current->flags &= ~PF_NO_SETAFFINITY; ++ set_cpus_allowed_ptr(current, cpu_present_mask); ++ migrate_me(); ++ return 0; ++} ++ ++static void __cpu_unplug_sync(struct hotplug_pcp *hp) ++{ ++ wake_up_process(hp->sync_tsk); ++ wait_for_completion(&hp->synced); ++} ++ ++static void __cpu_unplug_wait(unsigned int cpu) ++{ ++ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu); ++ ++ complete(&hp->unplug_wait); ++ wait_for_completion(&hp->synced); ++} ++ ++/* ++ * Start the sync_unplug_thread on the target cpu and wait for it to ++ * complete. ++ */ ++static int cpu_unplug_begin(unsigned int cpu) ++{ ++ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu); ++ int err; ++ ++ /* Protected by cpu_hotplug.lock */ ++ if (!hp->mutex_init) { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ spin_lock_init(&hp->lock); ++#else ++ mutex_init(&hp->mutex); ++#endif ++ hp->mutex_init = 1; ++ } ++ ++ /* Inform the scheduler to migrate tasks off this CPU */ ++ tell_sched_cpu_down_begin(cpu); ++ ++ init_completion(&hp->synced); ++ init_completion(&hp->unplug_wait); ++ ++ hp->sync_tsk = kthread_create(sync_unplug_thread, hp, "sync_unplug/%d", cpu); ++ if (IS_ERR(hp->sync_tsk)) { ++ err = PTR_ERR(hp->sync_tsk); ++ hp->sync_tsk = NULL; ++ return err; ++ } ++ kthread_bind(hp->sync_tsk, cpu); ++ ++ /* ++ * Wait for tasks to get out of the pinned sections, ++ * it's still OK if new tasks enter. Some CPU notifiers will ++ * wait for tasks that are going to enter these sections and ++ * we must not have them block. ++ */ ++ wake_up_process(hp->sync_tsk); ++ return 0; ++} ++ ++static void cpu_unplug_sync(unsigned int cpu) ++{ ++ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu); ++ ++ init_completion(&hp->synced); ++ /* The completion needs to be initialzied before setting grab_lock */ ++ smp_wmb(); ++ ++ /* Grab the mutex before setting grab_lock */ ++ hotplug_lock(hp); ++ hp->grab_lock = 1; ++ ++ /* ++ * The CPU notifiers have been completed. ++ * Wait for tasks to get out of pinned CPU sections and have new ++ * tasks block until the CPU is completely down. ++ */ ++ __cpu_unplug_sync(hp); ++ ++ /* All done with the sync thread */ ++ kthread_stop(hp->sync_tsk); ++ hp->sync_tsk = NULL; ++} ++ ++static void cpu_unplug_done(unsigned int cpu) ++{ ++ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu); ++ ++ hp->unplug = NULL; ++ /* Let all tasks know cpu unplug is finished before cleaning up */ ++ smp_wmb(); ++ ++ if (hp->sync_tsk) ++ kthread_stop(hp->sync_tsk); ++ ++ if (hp->grab_lock) { ++ hotplug_unlock(hp); ++ /* protected by cpu_hotplug.lock */ ++ hp->grab_lock = 0; ++ } ++ tell_sched_cpu_down_done(cpu); ++} + + void get_online_cpus(void) + { +@@ -789,10 +1072,14 @@ static int takedown_cpu(unsigned int cpu) + struct cpuhp_cpu_state *st = per_cpu_ptr(&cpuhp_state, cpu); + int err; + ++ __cpu_unplug_wait(cpu); + /* Park the smpboot threads */ + kthread_park(per_cpu_ptr(&cpuhp_state, cpu)->thread); + smpboot_park_threads(cpu); + ++ /* Notifiers are done. Don't let any more tasks pin this CPU. */ ++ cpu_unplug_sync(cpu); ++ + /* + * Prevent irq alloc/free while the dying cpu reorganizes the + * interrupt affinities. +@@ -877,6 +1164,9 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen, + struct cpuhp_cpu_state *st = per_cpu_ptr(&cpuhp_state, cpu); + int prev_state, ret = 0; + bool hasdied = false; ++ int mycpu; ++ cpumask_var_t cpumask; ++ cpumask_var_t cpumask_org; + + if (num_online_cpus() == 1) + return -EBUSY; +@@ -884,7 +1174,34 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen, + if (!cpu_present(cpu)) + return -EINVAL; + ++ /* Move the downtaker off the unplug cpu */ ++ if (!alloc_cpumask_var(&cpumask, GFP_KERNEL)) ++ return -ENOMEM; ++ if (!alloc_cpumask_var(&cpumask_org, GFP_KERNEL)) { ++ free_cpumask_var(cpumask); ++ return -ENOMEM; ++ } ++ ++ cpumask_copy(cpumask_org, tsk_cpus_allowed(current)); ++ cpumask_andnot(cpumask, cpu_online_mask, cpumask_of(cpu)); ++ set_cpus_allowed_ptr(current, cpumask); ++ free_cpumask_var(cpumask); ++ migrate_disable(); ++ mycpu = smp_processor_id(); ++ if (mycpu == cpu) { ++ printk(KERN_ERR "Yuck! Still on unplug CPU\n!"); ++ migrate_enable(); ++ ret = -EBUSY; ++ goto restore_cpus; ++ } ++ ++ migrate_enable(); + cpu_hotplug_begin(); ++ ret = cpu_unplug_begin(cpu); ++ if (ret) { ++ printk("cpu_unplug_begin(%d) failed\n", cpu); ++ goto out_cancel; ++ } + + cpuhp_tasks_frozen = tasks_frozen; + +@@ -923,10 +1240,15 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen, + + hasdied = prev_state != st->state && st->state == CPUHP_OFFLINE; + out: ++ cpu_unplug_done(cpu); ++out_cancel: + cpu_hotplug_done(); + /* This post dead nonsense must die */ + if (!ret && hasdied) + cpu_notify_nofail(CPU_POST_DEAD, cpu); ++restore_cpus: ++ set_cpus_allowed_ptr(current, cpumask_org); ++ free_cpumask_var(cpumask_org); + return ret; + } + +@@ -1240,6 +1562,8 @@ core_initcall(cpu_hotplug_pm_sync_init); + + #endif /* CONFIG_PM_SLEEP_SMP */ + ++int __boot_cpu_id; ++ + #endif /* CONFIG_SMP */ + + /* Boot processor state steps */ +@@ -1924,6 +2248,10 @@ void __init boot_cpu_init(void) + set_cpu_active(cpu, true); + set_cpu_present(cpu, true); + set_cpu_possible(cpu, true); ++ ++#ifdef CONFIG_SMP ++ __boot_cpu_id = cpu; ++#endif + } + + /* +diff --git a/kernel/cpuset.c b/kernel/cpuset.c +index 29f815d2ef7e..341b17f24f95 100644 +--- a/kernel/cpuset.c ++++ b/kernel/cpuset.c +@@ -284,7 +284,7 @@ static struct cpuset top_cpuset = { + */ + + static DEFINE_MUTEX(cpuset_mutex); +-static DEFINE_SPINLOCK(callback_lock); ++static DEFINE_RAW_SPINLOCK(callback_lock); + + static struct workqueue_struct *cpuset_migrate_mm_wq; + +@@ -907,9 +907,9 @@ static void update_cpumasks_hier(struct cpuset *cs, struct cpumask *new_cpus) + continue; + rcu_read_unlock(); + +- spin_lock_irq(&callback_lock); ++ raw_spin_lock_irq(&callback_lock); + cpumask_copy(cp->effective_cpus, new_cpus); +- spin_unlock_irq(&callback_lock); ++ raw_spin_unlock_irq(&callback_lock); + + WARN_ON(!cgroup_subsys_on_dfl(cpuset_cgrp_subsys) && + !cpumask_equal(cp->cpus_allowed, cp->effective_cpus)); +@@ -974,9 +974,9 @@ static int update_cpumask(struct cpuset *cs, struct cpuset *trialcs, + if (retval < 0) + return retval; + +- spin_lock_irq(&callback_lock); ++ raw_spin_lock_irq(&callback_lock); + cpumask_copy(cs->cpus_allowed, trialcs->cpus_allowed); +- spin_unlock_irq(&callback_lock); ++ raw_spin_unlock_irq(&callback_lock); + + /* use trialcs->cpus_allowed as a temp variable */ + update_cpumasks_hier(cs, trialcs->cpus_allowed); +@@ -1176,9 +1176,9 @@ static void update_nodemasks_hier(struct cpuset *cs, nodemask_t *new_mems) + continue; + rcu_read_unlock(); + +- spin_lock_irq(&callback_lock); ++ raw_spin_lock_irq(&callback_lock); + cp->effective_mems = *new_mems; +- spin_unlock_irq(&callback_lock); ++ raw_spin_unlock_irq(&callback_lock); + + WARN_ON(!cgroup_subsys_on_dfl(cpuset_cgrp_subsys) && + !nodes_equal(cp->mems_allowed, cp->effective_mems)); +@@ -1246,9 +1246,9 @@ static int update_nodemask(struct cpuset *cs, struct cpuset *trialcs, + if (retval < 0) + goto done; + +- spin_lock_irq(&callback_lock); ++ raw_spin_lock_irq(&callback_lock); + cs->mems_allowed = trialcs->mems_allowed; +- spin_unlock_irq(&callback_lock); ++ raw_spin_unlock_irq(&callback_lock); + + /* use trialcs->mems_allowed as a temp variable */ + update_nodemasks_hier(cs, &trialcs->mems_allowed); +@@ -1339,9 +1339,9 @@ static int update_flag(cpuset_flagbits_t bit, struct cpuset *cs, + spread_flag_changed = ((is_spread_slab(cs) != is_spread_slab(trialcs)) + || (is_spread_page(cs) != is_spread_page(trialcs))); + +- spin_lock_irq(&callback_lock); ++ raw_spin_lock_irq(&callback_lock); + cs->flags = trialcs->flags; +- spin_unlock_irq(&callback_lock); ++ raw_spin_unlock_irq(&callback_lock); + + if (!cpumask_empty(trialcs->cpus_allowed) && balance_flag_changed) + rebuild_sched_domains_locked(); +@@ -1756,7 +1756,7 @@ static int cpuset_common_seq_show(struct seq_file *sf, void *v) + cpuset_filetype_t type = seq_cft(sf)->private; + int ret = 0; + +- spin_lock_irq(&callback_lock); ++ raw_spin_lock_irq(&callback_lock); + + switch (type) { + case FILE_CPULIST: +@@ -1775,7 +1775,7 @@ static int cpuset_common_seq_show(struct seq_file *sf, void *v) + ret = -EINVAL; + } + +- spin_unlock_irq(&callback_lock); ++ raw_spin_unlock_irq(&callback_lock); + return ret; + } + +@@ -1989,12 +1989,12 @@ static int cpuset_css_online(struct cgroup_subsys_state *css) + + cpuset_inc(); + +- spin_lock_irq(&callback_lock); ++ raw_spin_lock_irq(&callback_lock); + if (cgroup_subsys_on_dfl(cpuset_cgrp_subsys)) { + cpumask_copy(cs->effective_cpus, parent->effective_cpus); + cs->effective_mems = parent->effective_mems; + } +- spin_unlock_irq(&callback_lock); ++ raw_spin_unlock_irq(&callback_lock); + + if (!test_bit(CGRP_CPUSET_CLONE_CHILDREN, &css->cgroup->flags)) + goto out_unlock; +@@ -2021,12 +2021,12 @@ static int cpuset_css_online(struct cgroup_subsys_state *css) + } + rcu_read_unlock(); + +- spin_lock_irq(&callback_lock); ++ raw_spin_lock_irq(&callback_lock); + cs->mems_allowed = parent->mems_allowed; + cs->effective_mems = parent->mems_allowed; + cpumask_copy(cs->cpus_allowed, parent->cpus_allowed); + cpumask_copy(cs->effective_cpus, parent->cpus_allowed); +- spin_unlock_irq(&callback_lock); ++ raw_spin_unlock_irq(&callback_lock); + out_unlock: + mutex_unlock(&cpuset_mutex); + return 0; +@@ -2065,7 +2065,7 @@ static void cpuset_css_free(struct cgroup_subsys_state *css) + static void cpuset_bind(struct cgroup_subsys_state *root_css) + { + mutex_lock(&cpuset_mutex); +- spin_lock_irq(&callback_lock); ++ raw_spin_lock_irq(&callback_lock); + + if (cgroup_subsys_on_dfl(cpuset_cgrp_subsys)) { + cpumask_copy(top_cpuset.cpus_allowed, cpu_possible_mask); +@@ -2076,7 +2076,7 @@ static void cpuset_bind(struct cgroup_subsys_state *root_css) + top_cpuset.mems_allowed = top_cpuset.effective_mems; + } + +- spin_unlock_irq(&callback_lock); ++ raw_spin_unlock_irq(&callback_lock); + mutex_unlock(&cpuset_mutex); + } + +@@ -2177,12 +2177,12 @@ hotplug_update_tasks_legacy(struct cpuset *cs, + { + bool is_empty; + +- spin_lock_irq(&callback_lock); ++ raw_spin_lock_irq(&callback_lock); + cpumask_copy(cs->cpus_allowed, new_cpus); + cpumask_copy(cs->effective_cpus, new_cpus); + cs->mems_allowed = *new_mems; + cs->effective_mems = *new_mems; +- spin_unlock_irq(&callback_lock); ++ raw_spin_unlock_irq(&callback_lock); + + /* + * Don't call update_tasks_cpumask() if the cpuset becomes empty, +@@ -2219,10 +2219,10 @@ hotplug_update_tasks(struct cpuset *cs, + if (nodes_empty(*new_mems)) + *new_mems = parent_cs(cs)->effective_mems; + +- spin_lock_irq(&callback_lock); ++ raw_spin_lock_irq(&callback_lock); + cpumask_copy(cs->effective_cpus, new_cpus); + cs->effective_mems = *new_mems; +- spin_unlock_irq(&callback_lock); ++ raw_spin_unlock_irq(&callback_lock); + + if (cpus_updated) + update_tasks_cpumask(cs); +@@ -2308,21 +2308,21 @@ static void cpuset_hotplug_workfn(struct work_struct *work) + + /* synchronize cpus_allowed to cpu_active_mask */ + if (cpus_updated) { +- spin_lock_irq(&callback_lock); ++ raw_spin_lock_irq(&callback_lock); + if (!on_dfl) + cpumask_copy(top_cpuset.cpus_allowed, &new_cpus); + cpumask_copy(top_cpuset.effective_cpus, &new_cpus); +- spin_unlock_irq(&callback_lock); ++ raw_spin_unlock_irq(&callback_lock); + /* we don't mess with cpumasks of tasks in top_cpuset */ + } + + /* synchronize mems_allowed to N_MEMORY */ + if (mems_updated) { +- spin_lock_irq(&callback_lock); ++ raw_spin_lock_irq(&callback_lock); + if (!on_dfl) + top_cpuset.mems_allowed = new_mems; + top_cpuset.effective_mems = new_mems; +- spin_unlock_irq(&callback_lock); ++ raw_spin_unlock_irq(&callback_lock); + update_tasks_nodemask(&top_cpuset); + } + +@@ -2420,11 +2420,11 @@ void cpuset_cpus_allowed(struct task_struct *tsk, struct cpumask *pmask) + { + unsigned long flags; + +- spin_lock_irqsave(&callback_lock, flags); ++ raw_spin_lock_irqsave(&callback_lock, flags); + rcu_read_lock(); + guarantee_online_cpus(task_cs(tsk), pmask); + rcu_read_unlock(); +- spin_unlock_irqrestore(&callback_lock, flags); ++ raw_spin_unlock_irqrestore(&callback_lock, flags); + } + + void cpuset_cpus_allowed_fallback(struct task_struct *tsk) +@@ -2472,11 +2472,11 @@ nodemask_t cpuset_mems_allowed(struct task_struct *tsk) + nodemask_t mask; + unsigned long flags; + +- spin_lock_irqsave(&callback_lock, flags); ++ raw_spin_lock_irqsave(&callback_lock, flags); + rcu_read_lock(); + guarantee_online_mems(task_cs(tsk), &mask); + rcu_read_unlock(); +- spin_unlock_irqrestore(&callback_lock, flags); ++ raw_spin_unlock_irqrestore(&callback_lock, flags); + + return mask; + } +@@ -2568,14 +2568,14 @@ bool __cpuset_node_allowed(int node, gfp_t gfp_mask) + return true; + + /* Not hardwall and node outside mems_allowed: scan up cpusets */ +- spin_lock_irqsave(&callback_lock, flags); ++ raw_spin_lock_irqsave(&callback_lock, flags); + + rcu_read_lock(); + cs = nearest_hardwall_ancestor(task_cs(current)); + allowed = node_isset(node, cs->mems_allowed); + rcu_read_unlock(); + +- spin_unlock_irqrestore(&callback_lock, flags); ++ raw_spin_unlock_irqrestore(&callback_lock, flags); + return allowed; + } + +diff --git a/kernel/debug/kdb/kdb_io.c b/kernel/debug/kdb/kdb_io.c +index fc1ef736253c..83c666537a7a 100644 +--- a/kernel/debug/kdb/kdb_io.c ++++ b/kernel/debug/kdb/kdb_io.c +@@ -554,7 +554,6 @@ int vkdb_printf(enum kdb_msgsrc src, const char *fmt, va_list ap) + int linecount; + int colcount; + int logging, saved_loglevel = 0; +- int saved_trap_printk; + int got_printf_lock = 0; + int retlen = 0; + int fnd, len; +@@ -565,8 +564,6 @@ int vkdb_printf(enum kdb_msgsrc src, const char *fmt, va_list ap) + unsigned long uninitialized_var(flags); + + preempt_disable(); +- saved_trap_printk = kdb_trap_printk; +- kdb_trap_printk = 0; + + /* Serialize kdb_printf if multiple cpus try to write at once. + * But if any cpu goes recursive in kdb, just print the output, +@@ -855,7 +852,6 @@ int vkdb_printf(enum kdb_msgsrc src, const char *fmt, va_list ap) + } else { + __release(kdb_printf_lock); + } +- kdb_trap_printk = saved_trap_printk; + preempt_enable(); + return retlen; + } +@@ -865,9 +861,11 @@ int kdb_printf(const char *fmt, ...) + va_list ap; + int r; + ++ kdb_trap_printk++; + va_start(ap, fmt); + r = vkdb_printf(KDB_MSGSRC_INTERNAL, fmt, ap); + va_end(ap); ++ kdb_trap_printk--; + + return r; + } +diff --git a/kernel/events/core.c b/kernel/events/core.c +index 07c0dc806dfc..baf1a2867d74 100644 +--- a/kernel/events/core.c ++++ b/kernel/events/core.c +@@ -1050,6 +1050,7 @@ static void __perf_mux_hrtimer_init(struct perf_cpu_context *cpuctx, int cpu) + raw_spin_lock_init(&cpuctx->hrtimer_lock); + hrtimer_init(timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS_PINNED); + timer->function = perf_mux_hrtimer_handler; ++ timer->irqsafe = 1; + } + + static int perf_mux_hrtimer_restart(struct perf_cpu_context *cpuctx) +@@ -8363,6 +8364,7 @@ static void perf_swevent_init_hrtimer(struct perf_event *event) + + hrtimer_init(&hwc->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + hwc->hrtimer.function = perf_swevent_hrtimer; ++ hwc->hrtimer.irqsafe = 1; + + /* + * Since hrtimers have a fixed rate, we can do a static freq->period +diff --git a/kernel/exit.c b/kernel/exit.c +index 3076f3089919..fb2ebcf3ca7c 100644 +--- a/kernel/exit.c ++++ b/kernel/exit.c +@@ -143,7 +143,7 @@ static void __exit_signal(struct task_struct *tsk) + * Do this under ->siglock, we can race with another thread + * doing sigqueue_free() if we have SIGQUEUE_PREALLOC signals. + */ +- flush_sigqueue(&tsk->pending); ++ flush_task_sigqueue(tsk); + tsk->sighand = NULL; + spin_unlock(&sighand->siglock); + +diff --git a/kernel/fork.c b/kernel/fork.c +index 59faac4de181..0edb0f3c1db8 100644 +--- a/kernel/fork.c ++++ b/kernel/fork.c +@@ -76,6 +76,7 @@ + #include + #include + #include ++#include + + #include + #include +@@ -376,13 +377,24 @@ static inline void put_signal_struct(struct signal_struct *sig) + if (atomic_dec_and_test(&sig->sigcnt)) + free_signal_struct(sig); + } +- ++#ifdef CONFIG_PREEMPT_RT_BASE ++static ++#endif + void __put_task_struct(struct task_struct *tsk) + { + WARN_ON(!tsk->exit_state); + WARN_ON(atomic_read(&tsk->usage)); + WARN_ON(tsk == current); + ++ /* ++ * Remove function-return probe instances associated with this ++ * task and put them back on the free list. ++ */ ++ kprobe_flush_task(tsk); ++ ++ /* Task is done with its stack. */ ++ put_task_stack(tsk); ++ + cgroup_free(tsk); + task_numa_free(tsk); + security_task_free(tsk); +@@ -393,7 +405,18 @@ void __put_task_struct(struct task_struct *tsk) + if (!profile_handoff_task(tsk)) + free_task(tsk); + } ++#ifndef CONFIG_PREEMPT_RT_BASE + EXPORT_SYMBOL_GPL(__put_task_struct); ++#else ++void __put_task_struct_cb(struct rcu_head *rhp) ++{ ++ struct task_struct *tsk = container_of(rhp, struct task_struct, put_rcu); ++ ++ __put_task_struct(tsk); ++ ++} ++EXPORT_SYMBOL_GPL(__put_task_struct_cb); ++#endif + + void __init __weak arch_task_cache_init(void) { } + +@@ -852,6 +875,19 @@ void __mmdrop(struct mm_struct *mm) + } + EXPORT_SYMBOL_GPL(__mmdrop); + ++#ifdef CONFIG_PREEMPT_RT_BASE ++/* ++ * RCU callback for delayed mm drop. Not strictly rcu, but we don't ++ * want another facility to make this work. ++ */ ++void __mmdrop_delayed(struct rcu_head *rhp) ++{ ++ struct mm_struct *mm = container_of(rhp, struct mm_struct, delayed_drop); ++ ++ __mmdrop(mm); ++} ++#endif ++ + static inline void __mmput(struct mm_struct *mm) + { + VM_BUG_ON(atomic_read(&mm->mm_users)); +@@ -1417,6 +1453,7 @@ static void rt_mutex_init_task(struct task_struct *p) + #ifdef CONFIG_RT_MUTEXES + p->pi_waiters = RB_ROOT; + p->pi_waiters_leftmost = NULL; ++ p->pi_top_task = NULL; + p->pi_blocked_on = NULL; + #endif + } +@@ -1426,6 +1463,9 @@ static void rt_mutex_init_task(struct task_struct *p) + */ + static void posix_cpu_timers_init(struct task_struct *tsk) + { ++#ifdef CONFIG_PREEMPT_RT_BASE ++ tsk->posix_timer_list = NULL; ++#endif + tsk->cputime_expires.prof_exp = 0; + tsk->cputime_expires.virt_exp = 0; + tsk->cputime_expires.sched_exp = 0; +@@ -1552,6 +1592,7 @@ static __latent_entropy struct task_struct *copy_process( + spin_lock_init(&p->alloc_lock); + + init_sigpending(&p->pending); ++ p->sigqueue_cache = NULL; + + p->utime = p->stime = p->gtime = 0; + p->utimescaled = p->stimescaled = 0; +diff --git a/kernel/futex.c b/kernel/futex.c +index 4c6b6e697b73..d9bab63efccb 100644 +--- a/kernel/futex.c ++++ b/kernel/futex.c +@@ -800,7 +800,7 @@ static int refill_pi_state_cache(void) + return 0; + } + +-static struct futex_pi_state * alloc_pi_state(void) ++static struct futex_pi_state *alloc_pi_state(void) + { + struct futex_pi_state *pi_state = current->pi_state_cache; + +@@ -810,6 +810,11 @@ static struct futex_pi_state * alloc_pi_state(void) + return pi_state; + } + ++static void get_pi_state(struct futex_pi_state *pi_state) ++{ ++ WARN_ON_ONCE(!atomic_inc_not_zero(&pi_state->refcount)); ++} ++ + /* + * Drops a reference to the pi_state object and frees or caches it + * when the last reference is gone. +@@ -854,7 +859,7 @@ static void put_pi_state(struct futex_pi_state *pi_state) + * Look up the task based on what TID userspace gave us. + * We dont trust it. + */ +-static struct task_struct * futex_find_get_task(pid_t pid) ++static struct task_struct *futex_find_get_task(pid_t pid) + { + struct task_struct *p; + +@@ -904,7 +909,9 @@ void exit_pi_state_list(struct task_struct *curr) + * task still owns the PI-state: + */ + if (head->next != next) { ++ raw_spin_unlock_irq(&curr->pi_lock); + spin_unlock(&hb->lock); ++ raw_spin_lock_irq(&curr->pi_lock); + continue; + } + +@@ -914,10 +921,12 @@ void exit_pi_state_list(struct task_struct *curr) + pi_state->owner = NULL; + raw_spin_unlock_irq(&curr->pi_lock); + +- rt_mutex_unlock(&pi_state->pi_mutex); +- ++ get_pi_state(pi_state); + spin_unlock(&hb->lock); + ++ rt_mutex_futex_unlock(&pi_state->pi_mutex); ++ put_pi_state(pi_state); ++ + raw_spin_lock_irq(&curr->pi_lock); + } + raw_spin_unlock_irq(&curr->pi_lock); +@@ -971,6 +980,39 @@ void exit_pi_state_list(struct task_struct *curr) + * + * [10] There is no transient state which leaves owner and user space + * TID out of sync. ++ * ++ * ++ * Serialization and lifetime rules: ++ * ++ * hb->lock: ++ * ++ * hb -> futex_q, relation ++ * futex_q -> pi_state, relation ++ * ++ * (cannot be raw because hb can contain arbitrary amount ++ * of futex_q's) ++ * ++ * pi_mutex->wait_lock: ++ * ++ * {uval, pi_state} ++ * ++ * (and pi_mutex 'obviously') ++ * ++ * p->pi_lock: ++ * ++ * p->pi_state_list -> pi_state->list, relation ++ * ++ * pi_state->refcount: ++ * ++ * pi_state lifetime ++ * ++ * ++ * Lock order: ++ * ++ * hb->lock ++ * pi_mutex->wait_lock ++ * p->pi_lock ++ * + */ + + /* +@@ -978,10 +1020,13 @@ void exit_pi_state_list(struct task_struct *curr) + * the pi_state against the user space value. If correct, attach to + * it. + */ +-static int attach_to_pi_state(u32 uval, struct futex_pi_state *pi_state, ++static int attach_to_pi_state(u32 __user *uaddr, u32 uval, ++ struct futex_pi_state *pi_state, + struct futex_pi_state **ps) + { + pid_t pid = uval & FUTEX_TID_MASK; ++ u32 uval2; ++ int ret; + + /* + * Userspace might have messed up non-PI and PI futexes [3] +@@ -989,9 +1034,39 @@ static int attach_to_pi_state(u32 uval, struct futex_pi_state *pi_state, + if (unlikely(!pi_state)) + return -EINVAL; + ++ /* ++ * We get here with hb->lock held, and having found a ++ * futex_top_waiter(). This means that futex_lock_pi() of said futex_q ++ * has dropped the hb->lock in between queue_me() and unqueue_me_pi(), ++ * which in turn means that futex_lock_pi() still has a reference on ++ * our pi_state. ++ * ++ * The waiter holding a reference on @pi_state also protects against ++ * the unlocked put_pi_state() in futex_unlock_pi(), futex_lock_pi() ++ * and futex_wait_requeue_pi() as it cannot go to 0 and consequently ++ * free pi_state before we can take a reference ourselves. ++ */ + WARN_ON(!atomic_read(&pi_state->refcount)); + + /* ++ * Now that we have a pi_state, we can acquire wait_lock ++ * and do the state validation. ++ */ ++ raw_spin_lock_irq(&pi_state->pi_mutex.wait_lock); ++ ++ /* ++ * Since {uval, pi_state} is serialized by wait_lock, and our current ++ * uval was read without holding it, it can have changed. Verify it ++ * still is what we expect it to be, otherwise retry the entire ++ * operation. ++ */ ++ if (get_futex_value_locked(&uval2, uaddr)) ++ goto out_efault; ++ ++ if (uval != uval2) ++ goto out_eagain; ++ ++ /* + * Handle the owner died case: + */ + if (uval & FUTEX_OWNER_DIED) { +@@ -1006,11 +1081,11 @@ static int attach_to_pi_state(u32 uval, struct futex_pi_state *pi_state, + * is not 0. Inconsistent state. [5] + */ + if (pid) +- return -EINVAL; ++ goto out_einval; + /* + * Take a ref on the state and return success. [4] + */ +- goto out_state; ++ goto out_attach; + } + + /* +@@ -1022,14 +1097,14 @@ static int attach_to_pi_state(u32 uval, struct futex_pi_state *pi_state, + * Take a ref on the state and return success. [6] + */ + if (!pid) +- goto out_state; ++ goto out_attach; + } else { + /* + * If the owner died bit is not set, then the pi_state + * must have an owner. [7] + */ + if (!pi_state->owner) +- return -EINVAL; ++ goto out_einval; + } + + /* +@@ -1038,11 +1113,29 @@ static int attach_to_pi_state(u32 uval, struct futex_pi_state *pi_state, + * user space TID. [9/10] + */ + if (pid != task_pid_vnr(pi_state->owner)) +- return -EINVAL; +-out_state: +- atomic_inc(&pi_state->refcount); ++ goto out_einval; ++ ++out_attach: ++ get_pi_state(pi_state); ++ raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); + *ps = pi_state; + return 0; ++ ++out_einval: ++ ret = -EINVAL; ++ goto out_error; ++ ++out_eagain: ++ ret = -EAGAIN; ++ goto out_error; ++ ++out_efault: ++ ret = -EFAULT; ++ goto out_error; ++ ++out_error: ++ raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); ++ return ret; + } + + /* +@@ -1093,6 +1186,9 @@ static int attach_to_pi_owner(u32 uval, union futex_key *key, + + /* + * No existing pi state. First waiter. [2] ++ * ++ * This creates pi_state, we have hb->lock held, this means nothing can ++ * observe this state, wait_lock is irrelevant. + */ + pi_state = alloc_pi_state(); + +@@ -1117,17 +1213,18 @@ static int attach_to_pi_owner(u32 uval, union futex_key *key, + return 0; + } + +-static int lookup_pi_state(u32 uval, struct futex_hash_bucket *hb, ++static int lookup_pi_state(u32 __user *uaddr, u32 uval, ++ struct futex_hash_bucket *hb, + union futex_key *key, struct futex_pi_state **ps) + { +- struct futex_q *match = futex_top_waiter(hb, key); ++ struct futex_q *top_waiter = futex_top_waiter(hb, key); + + /* + * If there is a waiter on that futex, validate it and + * attach to the pi_state when the validation succeeds. + */ +- if (match) +- return attach_to_pi_state(uval, match->pi_state, ps); ++ if (top_waiter) ++ return attach_to_pi_state(uaddr, uval, top_waiter->pi_state, ps); + + /* + * We are the first waiter - try to look up the owner based on +@@ -1146,7 +1243,7 @@ static int lock_pi_update_atomic(u32 __user *uaddr, u32 uval, u32 newval) + if (unlikely(cmpxchg_futex_value_locked(&curval, uaddr, uval, newval))) + return -EFAULT; + +- /*If user space value changed, let the caller retry */ ++ /* If user space value changed, let the caller retry */ + return curval != uval ? -EAGAIN : 0; + } + +@@ -1174,7 +1271,7 @@ static int futex_lock_pi_atomic(u32 __user *uaddr, struct futex_hash_bucket *hb, + struct task_struct *task, int set_waiters) + { + u32 uval, newval, vpid = task_pid_vnr(task); +- struct futex_q *match; ++ struct futex_q *top_waiter; + int ret; + + /* +@@ -1200,9 +1297,9 @@ static int futex_lock_pi_atomic(u32 __user *uaddr, struct futex_hash_bucket *hb, + * Lookup existing state first. If it exists, try to attach to + * its pi_state. + */ +- match = futex_top_waiter(hb, key); +- if (match) +- return attach_to_pi_state(uval, match->pi_state, ps); ++ top_waiter = futex_top_waiter(hb, key); ++ if (top_waiter) ++ return attach_to_pi_state(uaddr, uval, top_waiter->pi_state, ps); + + /* + * No waiter and user TID is 0. We are here because the +@@ -1283,50 +1380,45 @@ static void mark_wake_futex(struct wake_q_head *wake_q, struct futex_q *q) + wake_q_add(wake_q, p); + __unqueue_futex(q); + /* +- * The waiting task can free the futex_q as soon as +- * q->lock_ptr = NULL is written, without taking any locks. A +- * memory barrier is required here to prevent the following +- * store to lock_ptr from getting ahead of the plist_del. ++ * The waiting task can free the futex_q as soon as q->lock_ptr = NULL ++ * is written, without taking any locks. This is possible in the event ++ * of a spurious wakeup, for example. A memory barrier is required here ++ * to prevent the following store to lock_ptr from getting ahead of the ++ * plist_del in __unqueue_futex(). + */ +- smp_wmb(); +- q->lock_ptr = NULL; ++ smp_store_release(&q->lock_ptr, NULL); + } + +-static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, +- struct futex_hash_bucket *hb) ++/* ++ * Caller must hold a reference on @pi_state. ++ */ ++static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_pi_state *pi_state) + { +- struct task_struct *new_owner; +- struct futex_pi_state *pi_state = this->pi_state; + u32 uninitialized_var(curval), newval; ++ struct task_struct *new_owner; ++ bool postunlock = false; + WAKE_Q(wake_q); +- bool deboost; ++ WAKE_Q(wake_sleeper_q); + int ret = 0; + +- if (!pi_state) +- return -EINVAL; +- +- /* +- * If current does not own the pi_state then the futex is +- * inconsistent and user space fiddled with the futex value. +- */ +- if (pi_state->owner != current) +- return -EINVAL; +- +- raw_spin_lock_irq(&pi_state->pi_mutex.wait_lock); + new_owner = rt_mutex_next_owner(&pi_state->pi_mutex); ++ if (WARN_ON_ONCE(!new_owner)) { ++ /* ++ * As per the comment in futex_unlock_pi() this should not happen. ++ * ++ * When this happens, give up our locks and try again, giving ++ * the futex_lock_pi() instance time to complete, either by ++ * waiting on the rtmutex or removing itself from the futex ++ * queue. ++ */ ++ ret = -EAGAIN; ++ goto out_unlock; ++ } + + /* +- * It is possible that the next waiter (the one that brought +- * this owner to the kernel) timed out and is no longer +- * waiting on the lock. +- */ +- if (!new_owner) +- new_owner = this->task; +- +- /* +- * We pass it to the next owner. The WAITERS bit is always +- * kept enabled while there is PI state around. We cleanup the +- * owner died bit, because we are the owner. ++ * We pass it to the next owner. The WAITERS bit is always kept ++ * enabled while there is PI state around. We cleanup the owner ++ * died bit, because we are the owner. + */ + newval = FUTEX_WAITERS | task_pid_vnr(new_owner); + +@@ -1335,6 +1427,7 @@ static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, + + if (cmpxchg_futex_value_locked(&curval, uaddr, uval, newval)) { + ret = -EFAULT; ++ + } else if (curval != uval) { + /* + * If a unconditional UNLOCK_PI operation (user space did not +@@ -1347,10 +1440,14 @@ static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, + else + ret = -EINVAL; + } +- if (ret) { +- raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); +- return ret; +- } ++ ++ if (ret) ++ goto out_unlock; ++ ++ /* ++ * This is a point of no return; once we modify the uval there is no ++ * going back and subsequent operations must not fail. ++ */ + + raw_spin_lock(&pi_state->owner->pi_lock); + WARN_ON(list_empty(&pi_state->list)); +@@ -1363,22 +1460,15 @@ static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, + pi_state->owner = new_owner; + raw_spin_unlock(&new_owner->pi_lock); + ++ postunlock = __rt_mutex_futex_unlock(&pi_state->pi_mutex, &wake_q, ++ &wake_sleeper_q); ++out_unlock: + raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); + +- deboost = rt_mutex_futex_unlock(&pi_state->pi_mutex, &wake_q); ++ if (postunlock) ++ rt_mutex_postunlock(&wake_q, &wake_sleeper_q); + +- /* +- * First unlock HB so the waiter does not spin on it once he got woken +- * up. Second wake up the waiter before the priority is adjusted. If we +- * deboost first (and lose our higher priority), then the task might get +- * scheduled away before the wake up can take place. +- */ +- spin_unlock(&hb->lock); +- wake_up_q(&wake_q); +- if (deboost) +- rt_mutex_adjust_prio(current); +- +- return 0; ++ return ret; + } + + /* +@@ -1824,7 +1914,7 @@ static int futex_requeue(u32 __user *uaddr1, unsigned int flags, + * If that call succeeds then we have pi_state and an + * initial refcount on it. + */ +- ret = lookup_pi_state(ret, hb2, &key2, &pi_state); ++ ret = lookup_pi_state(uaddr2, ret, hb2, &key2, &pi_state); + } + + switch (ret) { +@@ -1907,7 +1997,7 @@ static int futex_requeue(u32 __user *uaddr1, unsigned int flags, + * refcount on the pi_state and store the pointer in + * the futex_q object of the waiter. + */ +- atomic_inc(&pi_state->refcount); ++ get_pi_state(pi_state); + this->pi_state = pi_state; + ret = rt_mutex_start_proxy_lock(&pi_state->pi_mutex, + this->rt_waiter, +@@ -1924,6 +2014,16 @@ static int futex_requeue(u32 __user *uaddr1, unsigned int flags, + requeue_pi_wake_futex(this, &key2, hb2); + drop_count++; + continue; ++ } else if (ret == -EAGAIN) { ++ /* ++ * Waiter was woken by timeout or ++ * signal and has set pi_blocked_on to ++ * PI_WAKEUP_INPROGRESS before we ++ * tried to enqueue it on the rtmutex. ++ */ ++ this->pi_state = NULL; ++ put_pi_state(pi_state); ++ continue; + } else if (ret) { + /* + * rt_mutex_start_proxy_lock() detected a +@@ -2007,20 +2107,7 @@ queue_unlock(struct futex_hash_bucket *hb) + hb_waiters_dec(hb); + } + +-/** +- * queue_me() - Enqueue the futex_q on the futex_hash_bucket +- * @q: The futex_q to enqueue +- * @hb: The destination hash bucket +- * +- * The hb->lock must be held by the caller, and is released here. A call to +- * queue_me() is typically paired with exactly one call to unqueue_me(). The +- * exceptions involve the PI related operations, which may use unqueue_me_pi() +- * or nothing if the unqueue is done as part of the wake process and the unqueue +- * state is implicit in the state of woken task (see futex_wait_requeue_pi() for +- * an example). +- */ +-static inline void queue_me(struct futex_q *q, struct futex_hash_bucket *hb) +- __releases(&hb->lock) ++static inline void __queue_me(struct futex_q *q, struct futex_hash_bucket *hb) + { + int prio; + +@@ -2037,6 +2124,24 @@ static inline void queue_me(struct futex_q *q, struct futex_hash_bucket *hb) + plist_node_init(&q->list, prio); + plist_add(&q->list, &hb->chain); + q->task = current; ++} ++ ++/** ++ * queue_me() - Enqueue the futex_q on the futex_hash_bucket ++ * @q: The futex_q to enqueue ++ * @hb: The destination hash bucket ++ * ++ * The hb->lock must be held by the caller, and is released here. A call to ++ * queue_me() is typically paired with exactly one call to unqueue_me(). The ++ * exceptions involve the PI related operations, which may use unqueue_me_pi() ++ * or nothing if the unqueue is done as part of the wake process and the unqueue ++ * state is implicit in the state of woken task (see futex_wait_requeue_pi() for ++ * an example). ++ */ ++static inline void queue_me(struct futex_q *q, struct futex_hash_bucket *hb) ++ __releases(&hb->lock) ++{ ++ __queue_me(q, hb); + spin_unlock(&hb->lock); + } + +@@ -2123,10 +2228,13 @@ static int fixup_pi_state_owner(u32 __user *uaddr, struct futex_q *q, + { + u32 newtid = task_pid_vnr(newowner) | FUTEX_WAITERS; + struct futex_pi_state *pi_state = q->pi_state; +- struct task_struct *oldowner = pi_state->owner; + u32 uval, uninitialized_var(curval), newval; ++ struct task_struct *oldowner; + int ret; + ++ raw_spin_lock_irq(&pi_state->pi_mutex.wait_lock); ++ ++ oldowner = pi_state->owner; + /* Owner died? */ + if (!pi_state->owner) + newtid |= FUTEX_OWNER_DIED; +@@ -2134,7 +2242,8 @@ static int fixup_pi_state_owner(u32 __user *uaddr, struct futex_q *q, + /* + * We are here either because we stole the rtmutex from the + * previous highest priority waiter or we are the highest priority +- * waiter but failed to get the rtmutex the first time. ++ * waiter but have failed to get the rtmutex the first time. ++ * + * We have to replace the newowner TID in the user space variable. + * This must be atomic as we have to preserve the owner died bit here. + * +@@ -2142,17 +2251,16 @@ static int fixup_pi_state_owner(u32 __user *uaddr, struct futex_q *q, + * because we can fault here. Imagine swapped out pages or a fork + * that marked all the anonymous memory readonly for cow. + * +- * Modifying pi_state _before_ the user space value would +- * leave the pi_state in an inconsistent state when we fault +- * here, because we need to drop the hash bucket lock to +- * handle the fault. This might be observed in the PID check +- * in lookup_pi_state. ++ * Modifying pi_state _before_ the user space value would leave the ++ * pi_state in an inconsistent state when we fault here, because we ++ * need to drop the locks to handle the fault. This might be observed ++ * in the PID check in lookup_pi_state. + */ + retry: + if (get_futex_value_locked(&uval, uaddr)) + goto handle_fault; + +- while (1) { ++ for (;;) { + newval = (uval & FUTEX_OWNER_DIED) | newtid; + + if (cmpxchg_futex_value_locked(&curval, uaddr, uval, newval)) +@@ -2167,47 +2275,60 @@ static int fixup_pi_state_owner(u32 __user *uaddr, struct futex_q *q, + * itself. + */ + if (pi_state->owner != NULL) { +- raw_spin_lock_irq(&pi_state->owner->pi_lock); ++ raw_spin_lock(&pi_state->owner->pi_lock); + WARN_ON(list_empty(&pi_state->list)); + list_del_init(&pi_state->list); +- raw_spin_unlock_irq(&pi_state->owner->pi_lock); ++ raw_spin_unlock(&pi_state->owner->pi_lock); + } + + pi_state->owner = newowner; + +- raw_spin_lock_irq(&newowner->pi_lock); ++ raw_spin_lock(&newowner->pi_lock); + WARN_ON(!list_empty(&pi_state->list)); + list_add(&pi_state->list, &newowner->pi_state_list); +- raw_spin_unlock_irq(&newowner->pi_lock); ++ raw_spin_unlock(&newowner->pi_lock); ++ raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); ++ + return 0; + + /* +- * To handle the page fault we need to drop the hash bucket +- * lock here. That gives the other task (either the highest priority +- * waiter itself or the task which stole the rtmutex) the +- * chance to try the fixup of the pi_state. So once we are +- * back from handling the fault we need to check the pi_state +- * after reacquiring the hash bucket lock and before trying to +- * do another fixup. When the fixup has been done already we +- * simply return. ++ * To handle the page fault we need to drop the locks here. That gives ++ * the other task (either the highest priority waiter itself or the ++ * task which stole the rtmutex) the chance to try the fixup of the ++ * pi_state. So once we are back from handling the fault we need to ++ * check the pi_state after reacquiring the locks and before trying to ++ * do another fixup. When the fixup has been done already we simply ++ * return. ++ * ++ * Note: we hold both hb->lock and pi_mutex->wait_lock. We can safely ++ * drop hb->lock since the caller owns the hb -> futex_q relation. ++ * Dropping the pi_mutex->wait_lock requires the state revalidate. + */ + handle_fault: ++ raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); + spin_unlock(q->lock_ptr); + + ret = fault_in_user_writeable(uaddr); + + spin_lock(q->lock_ptr); ++ raw_spin_lock_irq(&pi_state->pi_mutex.wait_lock); + + /* + * Check if someone else fixed it for us: + */ +- if (pi_state->owner != oldowner) +- return 0; ++ if (pi_state->owner != oldowner) { ++ ret = 0; ++ goto out_unlock; ++ } + + if (ret) +- return ret; ++ goto out_unlock; + + goto retry; ++ ++out_unlock: ++ raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); ++ return ret; + } + + static long futex_wait_restart(struct restart_block *restart); +@@ -2229,13 +2350,16 @@ static long futex_wait_restart(struct restart_block *restart); + */ + static int fixup_owner(u32 __user *uaddr, struct futex_q *q, int locked) + { +- struct task_struct *owner; + int ret = 0; + + if (locked) { + /* + * Got the lock. We might not be the anticipated owner if we + * did a lock-steal - fix up the PI-state in that case: ++ * ++ * We can safely read pi_state->owner without holding wait_lock ++ * because we now own the rt_mutex, only the owner will attempt ++ * to change it. + */ + if (q->pi_state->owner != current) + ret = fixup_pi_state_owner(uaddr, q, current); +@@ -2243,43 +2367,15 @@ static int fixup_owner(u32 __user *uaddr, struct futex_q *q, int locked) + } + + /* +- * Catch the rare case, where the lock was released when we were on the +- * way back before we locked the hash bucket. +- */ +- if (q->pi_state->owner == current) { +- /* +- * Try to get the rt_mutex now. This might fail as some other +- * task acquired the rt_mutex after we removed ourself from the +- * rt_mutex waiters list. +- */ +- if (rt_mutex_trylock(&q->pi_state->pi_mutex)) { +- locked = 1; +- goto out; +- } +- +- /* +- * pi_state is incorrect, some other task did a lock steal and +- * we returned due to timeout or signal without taking the +- * rt_mutex. Too late. +- */ +- raw_spin_lock_irq(&q->pi_state->pi_mutex.wait_lock); +- owner = rt_mutex_owner(&q->pi_state->pi_mutex); +- if (!owner) +- owner = rt_mutex_next_owner(&q->pi_state->pi_mutex); +- raw_spin_unlock_irq(&q->pi_state->pi_mutex.wait_lock); +- ret = fixup_pi_state_owner(uaddr, q, owner); +- goto out; +- } +- +- /* + * Paranoia check. If we did not take the lock, then we should not be + * the owner of the rt_mutex. + */ +- if (rt_mutex_owner(&q->pi_state->pi_mutex) == current) ++ if (rt_mutex_owner(&q->pi_state->pi_mutex) == current) { + printk(KERN_ERR "fixup_owner: ret = %d pi-mutex: %p " + "pi-state %p\n", ret, + q->pi_state->pi_mutex.owner, + q->pi_state->owner); ++ } + + out: + return ret ? ret : locked; +@@ -2503,6 +2599,8 @@ static int futex_lock_pi(u32 __user *uaddr, unsigned int flags, + ktime_t *time, int trylock) + { + struct hrtimer_sleeper timeout, *to = NULL; ++ struct futex_pi_state *pi_state = NULL; ++ struct rt_mutex_waiter rt_waiter; + struct futex_hash_bucket *hb; + struct futex_q q = futex_q_init; + int res, ret; +@@ -2555,25 +2653,77 @@ static int futex_lock_pi(u32 __user *uaddr, unsigned int flags, + } + } + ++ WARN_ON(!q.pi_state); ++ + /* + * Only actually queue now that the atomic ops are done: + */ +- queue_me(&q, hb); ++ __queue_me(&q, hb); + +- WARN_ON(!q.pi_state); +- /* +- * Block on the PI mutex: +- */ +- if (!trylock) { +- ret = rt_mutex_timed_futex_lock(&q.pi_state->pi_mutex, to); +- } else { +- ret = rt_mutex_trylock(&q.pi_state->pi_mutex); ++ if (trylock) { ++ ret = rt_mutex_futex_trylock(&q.pi_state->pi_mutex); + /* Fixup the trylock return value: */ + ret = ret ? 0 : -EWOULDBLOCK; ++ goto no_block; + } + ++ rt_mutex_init_waiter(&rt_waiter, false); ++ ++ /* ++ * On PREEMPT_RT_FULL, when hb->lock becomes an rt_mutex, we must not ++ * hold it while doing rt_mutex_start_proxy(), because then it will ++ * include hb->lock in the blocking chain, even through we'll not in ++ * fact hold it while blocking. This will lead it to report -EDEADLK ++ * and BUG when futex_unlock_pi() interleaves with this. ++ * ++ * Therefore acquire wait_lock while holding hb->lock, but drop the ++ * latter before calling rt_mutex_start_proxy_lock(). This still fully ++ * serializes against futex_unlock_pi() as that does the exact same ++ * lock handoff sequence. ++ */ ++ raw_spin_lock_irq(&q.pi_state->pi_mutex.wait_lock); ++ /* ++ * the migrate_disable() here disables migration in the in_atomic() fast ++ * path which is enabled again in the following spin_unlock(). We have ++ * one migrate_disable() pending in the slow-path which is reversed ++ * after the raw_spin_unlock_irq() where we leave the atomic context. ++ */ ++ migrate_disable(); ++ ++ spin_unlock(q.lock_ptr); ++ ret = __rt_mutex_start_proxy_lock(&q.pi_state->pi_mutex, &rt_waiter, current); ++ raw_spin_unlock_irq(&q.pi_state->pi_mutex.wait_lock); ++ migrate_enable(); ++ ++ if (ret) { ++ if (ret == 1) ++ ret = 0; ++ ++ spin_lock(q.lock_ptr); ++ goto no_block; ++ } ++ ++ ++ if (unlikely(to)) ++ hrtimer_start_expires(&to->timer, HRTIMER_MODE_ABS); ++ ++ ret = rt_mutex_wait_proxy_lock(&q.pi_state->pi_mutex, to, &rt_waiter); ++ + spin_lock(q.lock_ptr); + /* ++ * If we failed to acquire the lock (signal/timeout), we must ++ * first acquire the hb->lock before removing the lock from the ++ * rt_mutex waitqueue, such that we can keep the hb and rt_mutex ++ * wait lists consistent. ++ * ++ * In particular; it is important that futex_unlock_pi() can not ++ * observe this inconsistency. ++ */ ++ if (ret && !rt_mutex_cleanup_proxy_lock(&q.pi_state->pi_mutex, &rt_waiter)) ++ ret = 0; ++ ++no_block: ++ /* + * Fixup the pi_state owner and possibly acquire the lock if we + * haven't already. + */ +@@ -2589,12 +2739,19 @@ static int futex_lock_pi(u32 __user *uaddr, unsigned int flags, + * If fixup_owner() faulted and was unable to handle the fault, unlock + * it and return the fault to userspace. + */ +- if (ret && (rt_mutex_owner(&q.pi_state->pi_mutex) == current)) +- rt_mutex_unlock(&q.pi_state->pi_mutex); ++ if (ret && (rt_mutex_owner(&q.pi_state->pi_mutex) == current)) { ++ pi_state = q.pi_state; ++ get_pi_state(pi_state); ++ } + + /* Unqueue and drop the lock */ + unqueue_me_pi(&q); + ++ if (pi_state) { ++ rt_mutex_futex_unlock(&pi_state->pi_mutex); ++ put_pi_state(pi_state); ++ } ++ + goto out_put_key; + + out_unlock_put_key: +@@ -2603,8 +2760,10 @@ static int futex_lock_pi(u32 __user *uaddr, unsigned int flags, + out_put_key: + put_futex_key(&q.key); + out: +- if (to) ++ if (to) { ++ hrtimer_cancel(&to->timer); + destroy_hrtimer_on_stack(&to->timer); ++ } + return ret != -EINTR ? ret : -ERESTARTNOINTR; + + uaddr_faulted: +@@ -2631,7 +2790,7 @@ static int futex_unlock_pi(u32 __user *uaddr, unsigned int flags) + u32 uninitialized_var(curval), uval, vpid = task_pid_vnr(current); + union futex_key key = FUTEX_KEY_INIT; + struct futex_hash_bucket *hb; +- struct futex_q *match; ++ struct futex_q *top_waiter; + int ret; + + retry: +@@ -2655,12 +2814,48 @@ static int futex_unlock_pi(u32 __user *uaddr, unsigned int flags) + * all and we at least want to know if user space fiddled + * with the futex value instead of blindly unlocking. + */ +- match = futex_top_waiter(hb, &key); +- if (match) { +- ret = wake_futex_pi(uaddr, uval, match, hb); ++ top_waiter = futex_top_waiter(hb, &key); ++ if (top_waiter) { ++ struct futex_pi_state *pi_state = top_waiter->pi_state; ++ ++ ret = -EINVAL; ++ if (!pi_state) ++ goto out_unlock; ++ + /* +- * In case of success wake_futex_pi dropped the hash +- * bucket lock. ++ * If current does not own the pi_state then the futex is ++ * inconsistent and user space fiddled with the futex value. ++ */ ++ if (pi_state->owner != current) ++ goto out_unlock; ++ ++ get_pi_state(pi_state); ++ /* ++ * By taking wait_lock while still holding hb->lock, we ensure ++ * there is no point where we hold neither; and therefore ++ * wake_futex_pi() must observe a state consistent with what we ++ * observed. ++ */ ++ raw_spin_lock_irq(&pi_state->pi_mutex.wait_lock); ++ /* ++ * Magic trickery for now to make the RT migrate disable ++ * logic happy. The following spin_unlock() happens with ++ * interrupts disabled so the internal migrate_enable() ++ * won't undo the migrate_disable() which was issued when ++ * locking hb->lock. ++ */ ++ migrate_disable(); ++ spin_unlock(&hb->lock); ++ ++ /* Drops pi_state->pi_mutex.wait_lock */ ++ ret = wake_futex_pi(uaddr, uval, pi_state); ++ ++ migrate_enable(); ++ ++ put_pi_state(pi_state); ++ ++ /* ++ * Success, we're done! No tricky corner cases. + */ + if (!ret) + goto out_putkey; +@@ -2675,7 +2870,6 @@ static int futex_unlock_pi(u32 __user *uaddr, unsigned int flags) + * setting the FUTEX_WAITERS bit. Try again. + */ + if (ret == -EAGAIN) { +- spin_unlock(&hb->lock); + put_futex_key(&key); + goto retry; + } +@@ -2683,7 +2877,7 @@ static int futex_unlock_pi(u32 __user *uaddr, unsigned int flags) + * wake_futex_pi has detected invalid state. Tell user + * space. + */ +- goto out_unlock; ++ goto out_putkey; + } + + /* +@@ -2693,8 +2887,10 @@ static int futex_unlock_pi(u32 __user *uaddr, unsigned int flags) + * preserve the WAITERS bit not the OWNER_DIED one. We are the + * owner. + */ +- if (cmpxchg_futex_value_locked(&curval, uaddr, uval, 0)) ++ if (cmpxchg_futex_value_locked(&curval, uaddr, uval, 0)) { ++ spin_unlock(&hb->lock); + goto pi_faulted; ++ } + + /* + * If uval has changed, let user space handle it. +@@ -2708,7 +2904,6 @@ static int futex_unlock_pi(u32 __user *uaddr, unsigned int flags) + return ret; + + pi_faulted: +- spin_unlock(&hb->lock); + put_futex_key(&key); + + ret = fault_in_user_writeable(uaddr); +@@ -2812,8 +3007,9 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, + u32 __user *uaddr2) + { + struct hrtimer_sleeper timeout, *to = NULL; ++ struct futex_pi_state *pi_state = NULL; + struct rt_mutex_waiter rt_waiter; +- struct futex_hash_bucket *hb; ++ struct futex_hash_bucket *hb, *hb2; + union futex_key key2 = FUTEX_KEY_INIT; + struct futex_q q = futex_q_init; + int res, ret; +@@ -2838,10 +3034,7 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, + * The waiter is allocated on our stack, manipulated by the requeue + * code while we sleep on uaddr. + */ +- debug_rt_mutex_init_waiter(&rt_waiter); +- RB_CLEAR_NODE(&rt_waiter.pi_tree_entry); +- RB_CLEAR_NODE(&rt_waiter.tree_entry); +- rt_waiter.task = NULL; ++ rt_mutex_init_waiter(&rt_waiter, false); + + ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2, VERIFY_WRITE); + if (unlikely(ret != 0)) +@@ -2872,20 +3065,55 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, + /* Queue the futex_q, drop the hb lock, wait for wakeup. */ + futex_wait_queue_me(hb, &q, to); + +- spin_lock(&hb->lock); +- ret = handle_early_requeue_pi_wakeup(hb, &q, &key2, to); +- spin_unlock(&hb->lock); +- if (ret) +- goto out_put_keys; ++ /* ++ * On RT we must avoid races with requeue and trying to block ++ * on two mutexes (hb->lock and uaddr2's rtmutex) by ++ * serializing access to pi_blocked_on with pi_lock. ++ */ ++ raw_spin_lock_irq(¤t->pi_lock); ++ if (current->pi_blocked_on) { ++ /* ++ * We have been requeued or are in the process of ++ * being requeued. ++ */ ++ raw_spin_unlock_irq(¤t->pi_lock); ++ } else { ++ /* ++ * Setting pi_blocked_on to PI_WAKEUP_INPROGRESS ++ * prevents a concurrent requeue from moving us to the ++ * uaddr2 rtmutex. After that we can safely acquire ++ * (and possibly block on) hb->lock. ++ */ ++ current->pi_blocked_on = PI_WAKEUP_INPROGRESS; ++ raw_spin_unlock_irq(¤t->pi_lock); ++ ++ spin_lock(&hb->lock); ++ ++ /* ++ * Clean up pi_blocked_on. We might leak it otherwise ++ * when we succeeded with the hb->lock in the fast ++ * path. ++ */ ++ raw_spin_lock_irq(¤t->pi_lock); ++ current->pi_blocked_on = NULL; ++ raw_spin_unlock_irq(¤t->pi_lock); ++ ++ ret = handle_early_requeue_pi_wakeup(hb, &q, &key2, to); ++ spin_unlock(&hb->lock); ++ if (ret) ++ goto out_put_keys; ++ } + + /* +- * In order for us to be here, we know our q.key == key2, and since +- * we took the hb->lock above, we also know that futex_requeue() has +- * completed and we no longer have to concern ourselves with a wakeup +- * race with the atomic proxy lock acquisition by the requeue code. The +- * futex_requeue dropped our key1 reference and incremented our key2 +- * reference count. ++ * In order to be here, we have either been requeued, are in ++ * the process of being requeued, or requeue successfully ++ * acquired uaddr2 on our behalf. If pi_blocked_on was ++ * non-null above, we may be racing with a requeue. Do not ++ * rely on q->lock_ptr to be hb2->lock until after blocking on ++ * hb->lock or hb2->lock. The futex_requeue dropped our key1 ++ * reference and incremented our key2 reference count. + */ ++ hb2 = hash_futex(&key2); + + /* Check if the requeue code acquired the second futex for us. */ + if (!q.rt_waiter) { +@@ -2894,16 +3122,19 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, + * did a lock-steal - fix up the PI-state in that case. + */ + if (q.pi_state && (q.pi_state->owner != current)) { +- spin_lock(q.lock_ptr); ++ spin_lock(&hb2->lock); ++ BUG_ON(&hb2->lock != q.lock_ptr); + ret = fixup_pi_state_owner(uaddr2, &q, current); +- if (ret && rt_mutex_owner(&q.pi_state->pi_mutex) == current) +- rt_mutex_unlock(&q.pi_state->pi_mutex); ++ if (ret && rt_mutex_owner(&q.pi_state->pi_mutex) == current) { ++ pi_state = q.pi_state; ++ get_pi_state(pi_state); ++ } + /* + * Drop the reference to the pi state which + * the requeue_pi() code acquired for us. + */ + put_pi_state(q.pi_state); +- spin_unlock(q.lock_ptr); ++ spin_unlock(&hb2->lock); + } + } else { + struct rt_mutex *pi_mutex; +@@ -2915,10 +3146,14 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, + */ + WARN_ON(!q.pi_state); + pi_mutex = &q.pi_state->pi_mutex; +- ret = rt_mutex_finish_proxy_lock(pi_mutex, to, &rt_waiter); +- debug_rt_mutex_free_waiter(&rt_waiter); ++ ret = rt_mutex_wait_proxy_lock(pi_mutex, to, &rt_waiter); + +- spin_lock(q.lock_ptr); ++ spin_lock(&hb2->lock); ++ BUG_ON(&hb2->lock != q.lock_ptr); ++ if (ret && !rt_mutex_cleanup_proxy_lock(pi_mutex, &rt_waiter)) ++ ret = 0; ++ ++ debug_rt_mutex_free_waiter(&rt_waiter); + /* + * Fixup the pi_state owner and possibly acquire the lock if we + * haven't already. +@@ -2936,13 +3171,20 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, + * the fault, unlock the rt_mutex and return the fault to + * userspace. + */ +- if (ret && rt_mutex_owner(pi_mutex) == current) +- rt_mutex_unlock(pi_mutex); ++ if (ret && rt_mutex_owner(&q.pi_state->pi_mutex) == current) { ++ pi_state = q.pi_state; ++ get_pi_state(pi_state); ++ } + + /* Unqueue and drop the lock. */ + unqueue_me_pi(&q); + } + ++ if (pi_state) { ++ rt_mutex_futex_unlock(&pi_state->pi_mutex); ++ put_pi_state(pi_state); ++ } ++ + if (ret == -EINTR) { + /* + * We've already been requeued, but cannot restart by calling +diff --git a/kernel/irq/handle.c b/kernel/irq/handle.c +index d3f24905852c..f87aa8fdcc51 100644 +--- a/kernel/irq/handle.c ++++ b/kernel/irq/handle.c +@@ -181,10 +181,16 @@ irqreturn_t handle_irq_event_percpu(struct irq_desc *desc) + { + irqreturn_t retval; + unsigned int flags = 0; ++ struct pt_regs *regs = get_irq_regs(); ++ u64 ip = regs ? instruction_pointer(regs) : 0; + + retval = __handle_irq_event_percpu(desc, &flags); + +- add_interrupt_randomness(desc->irq_data.irq, flags); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ desc->random_ip = ip; ++#else ++ add_interrupt_randomness(desc->irq_data.irq, flags, ip); ++#endif + + if (!noirqdebug) + note_interrupt(desc, retval); +diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c +index 6b669593e7eb..e357bf6c59d5 100644 +--- a/kernel/irq/manage.c ++++ b/kernel/irq/manage.c +@@ -22,6 +22,7 @@ + #include "internals.h" + + #ifdef CONFIG_IRQ_FORCED_THREADING ++# ifndef CONFIG_PREEMPT_RT_BASE + __read_mostly bool force_irqthreads; + + static int __init setup_forced_irqthreads(char *arg) +@@ -30,6 +31,7 @@ static int __init setup_forced_irqthreads(char *arg) + return 0; + } + early_param("threadirqs", setup_forced_irqthreads); ++# endif + #endif + + static void __synchronize_hardirq(struct irq_desc *desc) +@@ -233,7 +235,12 @@ int irq_set_affinity_locked(struct irq_data *data, const struct cpumask *mask, + + if (desc->affinity_notify) { + kref_get(&desc->affinity_notify->kref); ++ ++#ifdef CONFIG_PREEMPT_RT_BASE ++ swork_queue(&desc->affinity_notify->swork); ++#else + schedule_work(&desc->affinity_notify->work); ++#endif + } + irqd_set(data, IRQD_AFFINITY_SET); + +@@ -271,10 +278,8 @@ int irq_set_affinity_hint(unsigned int irq, const struct cpumask *m) + } + EXPORT_SYMBOL_GPL(irq_set_affinity_hint); + +-static void irq_affinity_notify(struct work_struct *work) ++static void _irq_affinity_notify(struct irq_affinity_notify *notify) + { +- struct irq_affinity_notify *notify = +- container_of(work, struct irq_affinity_notify, work); + struct irq_desc *desc = irq_to_desc(notify->irq); + cpumask_var_t cpumask; + unsigned long flags; +@@ -296,6 +301,35 @@ static void irq_affinity_notify(struct work_struct *work) + kref_put(¬ify->kref, notify->release); + } + ++#ifdef CONFIG_PREEMPT_RT_BASE ++static void init_helper_thread(void) ++{ ++ static int init_sworker_once; ++ ++ if (init_sworker_once) ++ return; ++ if (WARN_ON(swork_get())) ++ return; ++ init_sworker_once = 1; ++} ++ ++static void irq_affinity_notify(struct swork_event *swork) ++{ ++ struct irq_affinity_notify *notify = ++ container_of(swork, struct irq_affinity_notify, swork); ++ _irq_affinity_notify(notify); ++} ++ ++#else ++ ++static void irq_affinity_notify(struct work_struct *work) ++{ ++ struct irq_affinity_notify *notify = ++ container_of(work, struct irq_affinity_notify, work); ++ _irq_affinity_notify(notify); ++} ++#endif ++ + /** + * irq_set_affinity_notifier - control notification of IRQ affinity changes + * @irq: Interrupt for which to enable/disable notification +@@ -324,7 +358,12 @@ irq_set_affinity_notifier(unsigned int irq, struct irq_affinity_notify *notify) + if (notify) { + notify->irq = irq; + kref_init(¬ify->kref); ++#ifdef CONFIG_PREEMPT_RT_BASE ++ INIT_SWORK(¬ify->swork, irq_affinity_notify); ++ init_helper_thread(); ++#else + INIT_WORK(¬ify->work, irq_affinity_notify); ++#endif + } + + raw_spin_lock_irqsave(&desc->lock, flags); +@@ -879,7 +918,15 @@ irq_forced_thread_fn(struct irq_desc *desc, struct irqaction *action) + local_bh_disable(); + ret = action->thread_fn(action->irq, action->dev_id); + irq_finalize_oneshot(desc, action); +- local_bh_enable(); ++ /* ++ * Interrupts which have real time requirements can be set up ++ * to avoid softirq processing in the thread handler. This is ++ * safe as these interrupts do not raise soft interrupts. ++ */ ++ if (irq_settings_no_softirq_call(desc)) ++ _local_bh_enable(); ++ else ++ local_bh_enable(); + return ret; + } + +@@ -976,6 +1023,12 @@ static int irq_thread(void *data) + if (action_ret == IRQ_WAKE_THREAD) + irq_wake_secondary(desc, action); + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ migrate_disable(); ++ add_interrupt_randomness(action->irq, 0, ++ desc->random_ip ^ (unsigned long) action); ++ migrate_enable(); ++#endif + wake_threads_waitq(desc); + } + +@@ -1336,6 +1389,9 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new) + irqd_set(&desc->irq_data, IRQD_NO_BALANCING); + } + ++ if (new->flags & IRQF_NO_SOFTIRQ_CALL) ++ irq_settings_set_no_softirq_call(desc); ++ + /* Set default affinity mask once everything is setup */ + setup_affinity(desc, mask); + +@@ -2061,7 +2117,7 @@ EXPORT_SYMBOL_GPL(irq_get_irqchip_state); + * This call sets the internal irqchip state of an interrupt, + * depending on the value of @which. + * +- * This function should be called with preemption disabled if the ++ * This function should be called with migration disabled if the + * interrupt controller has per-cpu registers. + */ + int irq_set_irqchip_state(unsigned int irq, enum irqchip_irq_state which, +diff --git a/kernel/irq/settings.h b/kernel/irq/settings.h +index 320579d89091..2df2d4445b1e 100644 +--- a/kernel/irq/settings.h ++++ b/kernel/irq/settings.h +@@ -16,6 +16,7 @@ enum { + _IRQ_PER_CPU_DEVID = IRQ_PER_CPU_DEVID, + _IRQ_IS_POLLED = IRQ_IS_POLLED, + _IRQ_DISABLE_UNLAZY = IRQ_DISABLE_UNLAZY, ++ _IRQ_NO_SOFTIRQ_CALL = IRQ_NO_SOFTIRQ_CALL, + _IRQF_MODIFY_MASK = IRQF_MODIFY_MASK, + }; + +@@ -30,6 +31,7 @@ enum { + #define IRQ_PER_CPU_DEVID GOT_YOU_MORON + #define IRQ_IS_POLLED GOT_YOU_MORON + #define IRQ_DISABLE_UNLAZY GOT_YOU_MORON ++#define IRQ_NO_SOFTIRQ_CALL GOT_YOU_MORON + #undef IRQF_MODIFY_MASK + #define IRQF_MODIFY_MASK GOT_YOU_MORON + +@@ -40,6 +42,16 @@ irq_settings_clr_and_set(struct irq_desc *desc, u32 clr, u32 set) + desc->status_use_accessors |= (set & _IRQF_MODIFY_MASK); + } + ++static inline bool irq_settings_no_softirq_call(struct irq_desc *desc) ++{ ++ return desc->status_use_accessors & _IRQ_NO_SOFTIRQ_CALL; ++} ++ ++static inline void irq_settings_set_no_softirq_call(struct irq_desc *desc) ++{ ++ desc->status_use_accessors |= _IRQ_NO_SOFTIRQ_CALL; ++} ++ + static inline bool irq_settings_is_per_cpu(struct irq_desc *desc) + { + return desc->status_use_accessors & _IRQ_PER_CPU; +diff --git a/kernel/irq/spurious.c b/kernel/irq/spurious.c +index 5707f97a3e6a..73f38dc7a7fb 100644 +--- a/kernel/irq/spurious.c ++++ b/kernel/irq/spurious.c +@@ -442,6 +442,10 @@ MODULE_PARM_DESC(noirqdebug, "Disable irq lockup detection when true"); + + static int __init irqfixup_setup(char *str) + { ++#ifdef CONFIG_PREEMPT_RT_BASE ++ pr_warn("irqfixup boot option not supported w/ CONFIG_PREEMPT_RT_BASE\n"); ++ return 1; ++#endif + irqfixup = 1; + printk(KERN_WARNING "Misrouted IRQ fixup support enabled.\n"); + printk(KERN_WARNING "This may impact system performance.\n"); +@@ -454,6 +458,10 @@ module_param(irqfixup, int, 0644); + + static int __init irqpoll_setup(char *str) + { ++#ifdef CONFIG_PREEMPT_RT_BASE ++ pr_warn("irqpoll boot option not supported w/ CONFIG_PREEMPT_RT_BASE\n"); ++ return 1; ++#endif + irqfixup = 2; + printk(KERN_WARNING "Misrouted IRQ fixup and polling support " + "enabled\n"); +diff --git a/kernel/irq_work.c b/kernel/irq_work.c +index bcf107ce0854..2899ba0d23d1 100644 +--- a/kernel/irq_work.c ++++ b/kernel/irq_work.c +@@ -17,6 +17,7 @@ + #include + #include + #include ++#include + #include + + +@@ -65,6 +66,8 @@ void __weak arch_irq_work_raise(void) + */ + bool irq_work_queue_on(struct irq_work *work, int cpu) + { ++ struct llist_head *list; ++ + /* All work should have been flushed before going offline */ + WARN_ON_ONCE(cpu_is_offline(cpu)); + +@@ -75,7 +78,12 @@ bool irq_work_queue_on(struct irq_work *work, int cpu) + if (!irq_work_claim(work)) + return false; + +- if (llist_add(&work->llnode, &per_cpu(raised_list, cpu))) ++ if (IS_ENABLED(CONFIG_PREEMPT_RT_FULL) && !(work->flags & IRQ_WORK_HARD_IRQ)) ++ list = &per_cpu(lazy_list, cpu); ++ else ++ list = &per_cpu(raised_list, cpu); ++ ++ if (llist_add(&work->llnode, list)) + arch_send_call_function_single_ipi(cpu); + + return true; +@@ -86,6 +94,9 @@ EXPORT_SYMBOL_GPL(irq_work_queue_on); + /* Enqueue the irq work @work on the current CPU */ + bool irq_work_queue(struct irq_work *work) + { ++ struct llist_head *list; ++ bool lazy_work, realtime = IS_ENABLED(CONFIG_PREEMPT_RT_FULL); ++ + /* Only queue if not already pending */ + if (!irq_work_claim(work)) + return false; +@@ -93,13 +104,15 @@ bool irq_work_queue(struct irq_work *work) + /* Queue the entry and raise the IPI if needed. */ + preempt_disable(); + +- /* If the work is "lazy", handle it from next tick if any */ +- if (work->flags & IRQ_WORK_LAZY) { +- if (llist_add(&work->llnode, this_cpu_ptr(&lazy_list)) && +- tick_nohz_tick_stopped()) +- arch_irq_work_raise(); +- } else { +- if (llist_add(&work->llnode, this_cpu_ptr(&raised_list))) ++ lazy_work = work->flags & IRQ_WORK_LAZY; ++ ++ if (lazy_work || (realtime && !(work->flags & IRQ_WORK_HARD_IRQ))) ++ list = this_cpu_ptr(&lazy_list); ++ else ++ list = this_cpu_ptr(&raised_list); ++ ++ if (llist_add(&work->llnode, list)) { ++ if (!lazy_work || tick_nohz_tick_stopped()) + arch_irq_work_raise(); + } + +@@ -116,9 +129,8 @@ bool irq_work_needs_cpu(void) + raised = this_cpu_ptr(&raised_list); + lazy = this_cpu_ptr(&lazy_list); + +- if (llist_empty(raised) || arch_irq_work_has_interrupt()) +- if (llist_empty(lazy)) +- return false; ++ if (llist_empty(raised) && llist_empty(lazy)) ++ return false; + + /* All work should have been flushed before going offline */ + WARN_ON_ONCE(cpu_is_offline(smp_processor_id())); +@@ -132,7 +144,7 @@ static void irq_work_run_list(struct llist_head *list) + struct irq_work *work; + struct llist_node *llnode; + +- BUG_ON(!irqs_disabled()); ++ BUG_ON_NONRT(!irqs_disabled()); + + if (llist_empty(list)) + return; +@@ -169,7 +181,16 @@ static void irq_work_run_list(struct llist_head *list) + void irq_work_run(void) + { + irq_work_run_list(this_cpu_ptr(&raised_list)); +- irq_work_run_list(this_cpu_ptr(&lazy_list)); ++ if (IS_ENABLED(CONFIG_PREEMPT_RT_FULL)) { ++ /* ++ * NOTE: we raise softirq via IPI for safety, ++ * and execute in irq_work_tick() to move the ++ * overhead from hard to soft irq context. ++ */ ++ if (!llist_empty(this_cpu_ptr(&lazy_list))) ++ raise_softirq(TIMER_SOFTIRQ); ++ } else ++ irq_work_run_list(this_cpu_ptr(&lazy_list)); + } + EXPORT_SYMBOL_GPL(irq_work_run); + +@@ -179,8 +200,17 @@ void irq_work_tick(void) + + if (!llist_empty(raised) && !arch_irq_work_has_interrupt()) + irq_work_run_list(raised); ++ ++ if (!IS_ENABLED(CONFIG_PREEMPT_RT_FULL)) ++ irq_work_run_list(this_cpu_ptr(&lazy_list)); ++} ++ ++#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) ++void irq_work_tick_soft(void) ++{ + irq_work_run_list(this_cpu_ptr(&lazy_list)); + } ++#endif + + /* + * Synchronize against the irq_work @entry, ensures the entry is not +diff --git a/kernel/ksysfs.c b/kernel/ksysfs.c +index ee1bc1bb8feb..ddef07958840 100644 +--- a/kernel/ksysfs.c ++++ b/kernel/ksysfs.c +@@ -136,6 +136,15 @@ KERNEL_ATTR_RO(vmcoreinfo); + + #endif /* CONFIG_KEXEC_CORE */ + ++#if defined(CONFIG_PREEMPT_RT_FULL) ++static ssize_t realtime_show(struct kobject *kobj, ++ struct kobj_attribute *attr, char *buf) ++{ ++ return sprintf(buf, "%d\n", 1); ++} ++KERNEL_ATTR_RO(realtime); ++#endif ++ + /* whether file capabilities are enabled */ + static ssize_t fscaps_show(struct kobject *kobj, + struct kobj_attribute *attr, char *buf) +@@ -225,6 +234,9 @@ static struct attribute * kernel_attrs[] = { + &rcu_expedited_attr.attr, + &rcu_normal_attr.attr, + #endif ++#ifdef CONFIG_PREEMPT_RT_FULL ++ &realtime_attr.attr, ++#endif + NULL + }; + +diff --git a/kernel/locking/Makefile b/kernel/locking/Makefile +index 6f88e352cd4f..6ff9e8011dd0 100644 +--- a/kernel/locking/Makefile ++++ b/kernel/locking/Makefile +@@ -2,7 +2,7 @@ + # and is generally not a function of system call inputs. + KCOV_INSTRUMENT := n + +-obj-y += mutex.o semaphore.o rwsem.o percpu-rwsem.o ++obj-y += semaphore.o percpu-rwsem.o + + ifdef CONFIG_FUNCTION_TRACER + CFLAGS_REMOVE_lockdep.o = $(CC_FLAGS_FTRACE) +@@ -11,7 +11,11 @@ CFLAGS_REMOVE_mutex-debug.o = $(CC_FLAGS_FTRACE) + CFLAGS_REMOVE_rtmutex-debug.o = $(CC_FLAGS_FTRACE) + endif + ++ifneq ($(CONFIG_PREEMPT_RT_FULL),y) ++obj-y += mutex.o + obj-$(CONFIG_DEBUG_MUTEXES) += mutex-debug.o ++endif ++obj-y += rwsem.o + obj-$(CONFIG_LOCKDEP) += lockdep.o + ifeq ($(CONFIG_PROC_FS),y) + obj-$(CONFIG_LOCKDEP) += lockdep_proc.o +@@ -24,7 +28,10 @@ obj-$(CONFIG_RT_MUTEXES) += rtmutex.o + obj-$(CONFIG_DEBUG_RT_MUTEXES) += rtmutex-debug.o + obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock.o + obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock_debug.o ++ifneq ($(CONFIG_PREEMPT_RT_FULL),y) + obj-$(CONFIG_RWSEM_GENERIC_SPINLOCK) += rwsem-spinlock.o + obj-$(CONFIG_RWSEM_XCHGADD_ALGORITHM) += rwsem-xadd.o ++endif ++obj-$(CONFIG_PREEMPT_RT_FULL) += rt.o rwsem-rt.o + obj-$(CONFIG_QUEUED_RWLOCKS) += qrwlock.o + obj-$(CONFIG_LOCK_TORTURE_TEST) += locktorture.o +diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c +index 4d7ffc0a0d00..3d157b3128eb 100644 +--- a/kernel/locking/lockdep.c ++++ b/kernel/locking/lockdep.c +@@ -658,6 +658,7 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) + struct lockdep_subclass_key *key; + struct hlist_head *hash_head; + struct lock_class *class; ++ bool is_static = false; + + if (unlikely(subclass >= MAX_LOCKDEP_SUBCLASSES)) { + debug_locks_off(); +@@ -671,10 +672,23 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) + + /* + * Static locks do not have their class-keys yet - for them the key +- * is the lock object itself: ++ * is the lock object itself. If the lock is in the per cpu area, ++ * the canonical address of the lock (per cpu offset removed) is ++ * used. + */ +- if (unlikely(!lock->key)) +- lock->key = (void *)lock; ++ if (unlikely(!lock->key)) { ++ unsigned long can_addr, addr = (unsigned long)lock; ++ ++ if (__is_kernel_percpu_address(addr, &can_addr)) ++ lock->key = (void *)can_addr; ++ else if (__is_module_percpu_address(addr, &can_addr)) ++ lock->key = (void *)can_addr; ++ else if (static_obj(lock)) ++ lock->key = (void *)lock; ++ else ++ return ERR_PTR(-EINVAL); ++ is_static = true; ++ } + + /* + * NOTE: the class-key must be unique. For dynamic locks, a static +@@ -706,7 +720,7 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) + } + } + +- return NULL; ++ return is_static || static_obj(lock->key) ? NULL : ERR_PTR(-EINVAL); + } + + /* +@@ -724,19 +738,18 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) + DEBUG_LOCKS_WARN_ON(!irqs_disabled()); + + class = look_up_lock_class(lock, subclass); +- if (likely(class)) ++ if (likely(!IS_ERR_OR_NULL(class))) + goto out_set_class_cache; + + /* + * Debug-check: all keys must be persistent! +- */ +- if (!static_obj(lock->key)) { ++ */ ++ if (IS_ERR(class)) { + debug_locks_off(); + printk("INFO: trying to register non-static key.\n"); + printk("the code is fine but needs lockdep annotation.\n"); + printk("turning off the locking correctness validator.\n"); + dump_stack(); +- + return NULL; + } + +@@ -3410,7 +3423,7 @@ static int match_held_lock(struct held_lock *hlock, struct lockdep_map *lock) + * Clearly if the lock hasn't been acquired _ever_, we're not + * holding it either, so report failure. + */ +- if (!class) ++ if (IS_ERR_OR_NULL(class)) + return 0; + + /* +@@ -3689,6 +3702,7 @@ static void check_flags(unsigned long flags) + } + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * We dont accurately track softirq state in e.g. + * hardirq contexts (such as on 4KSTACKS), so only +@@ -3703,6 +3717,7 @@ static void check_flags(unsigned long flags) + DEBUG_LOCKS_WARN_ON(!current->softirqs_enabled); + } + } ++#endif + + if (!debug_locks) + print_irqtrace_events(current); +@@ -4159,7 +4174,7 @@ void lockdep_reset_lock(struct lockdep_map *lock) + * If the class exists we look it up and zap it: + */ + class = look_up_lock_class(lock, j); +- if (class) ++ if (!IS_ERR_OR_NULL(class)) + zap_class(class); + } + /* +diff --git a/kernel/locking/locktorture.c b/kernel/locking/locktorture.c +index f8c5af52a131..788068773e61 100644 +--- a/kernel/locking/locktorture.c ++++ b/kernel/locking/locktorture.c +@@ -26,7 +26,6 @@ + #include + #include + #include +-#include + #include + #include + #include +diff --git a/kernel/locking/percpu-rwsem.c b/kernel/locking/percpu-rwsem.c +index ce182599cf2e..2ad3a1e8344c 100644 +--- a/kernel/locking/percpu-rwsem.c ++++ b/kernel/locking/percpu-rwsem.c +@@ -18,7 +18,7 @@ int __percpu_init_rwsem(struct percpu_rw_semaphore *sem, + /* ->rw_sem represents the whole percpu_rw_semaphore for lockdep */ + rcu_sync_init(&sem->rss, RCU_SCHED_SYNC); + __init_rwsem(&sem->rw_sem, name, rwsem_key); +- init_waitqueue_head(&sem->writer); ++ init_swait_queue_head(&sem->writer); + sem->readers_block = 0; + return 0; + } +@@ -103,7 +103,7 @@ void __percpu_up_read(struct percpu_rw_semaphore *sem) + __this_cpu_dec(*sem->read_count); + + /* Prod writer to recheck readers_active */ +- wake_up(&sem->writer); ++ swake_up(&sem->writer); + } + EXPORT_SYMBOL_GPL(__percpu_up_read); + +@@ -160,7 +160,7 @@ void percpu_down_write(struct percpu_rw_semaphore *sem) + */ + + /* Wait for all now active readers to complete. */ +- wait_event(sem->writer, readers_active_check(sem)); ++ swait_event(sem->writer, readers_active_check(sem)); + } + EXPORT_SYMBOL_GPL(percpu_down_write); + +diff --git a/kernel/locking/rt.c b/kernel/locking/rt.c +new file mode 100644 +index 000000000000..6284e3b15091 +--- /dev/null ++++ b/kernel/locking/rt.c +@@ -0,0 +1,331 @@ ++/* ++ * kernel/rt.c ++ * ++ * Real-Time Preemption Support ++ * ++ * started by Ingo Molnar: ++ * ++ * Copyright (C) 2004-2006 Red Hat, Inc., Ingo Molnar ++ * Copyright (C) 2006, Timesys Corp., Thomas Gleixner ++ * ++ * historic credit for proving that Linux spinlocks can be implemented via ++ * RT-aware mutexes goes to many people: The Pmutex project (Dirk Grambow ++ * and others) who prototyped it on 2.4 and did lots of comparative ++ * research and analysis; TimeSys, for proving that you can implement a ++ * fully preemptible kernel via the use of IRQ threading and mutexes; ++ * Bill Huey for persuasively arguing on lkml that the mutex model is the ++ * right one; and to MontaVista, who ported pmutexes to 2.6. ++ * ++ * This code is a from-scratch implementation and is not based on pmutexes, ++ * but the idea of converting spinlocks to mutexes is used here too. ++ * ++ * lock debugging, locking tree, deadlock detection: ++ * ++ * Copyright (C) 2004, LynuxWorks, Inc., Igor Manyilov, Bill Huey ++ * Released under the General Public License (GPL). ++ * ++ * Includes portions of the generic R/W semaphore implementation from: ++ * ++ * Copyright (c) 2001 David Howells (dhowells@redhat.com). ++ * - Derived partially from idea by Andrea Arcangeli ++ * - Derived also from comments by Linus ++ * ++ * Pending ownership of locks and ownership stealing: ++ * ++ * Copyright (C) 2005, Kihon Technologies Inc., Steven Rostedt ++ * ++ * (also by Steven Rostedt) ++ * - Converted single pi_lock to individual task locks. ++ * ++ * By Esben Nielsen: ++ * Doing priority inheritance with help of the scheduler. ++ * ++ * Copyright (C) 2006, Timesys Corp., Thomas Gleixner ++ * - major rework based on Esben Nielsens initial patch ++ * - replaced thread_info references by task_struct refs ++ * - removed task->pending_owner dependency ++ * - BKL drop/reacquire for semaphore style locks to avoid deadlocks ++ * in the scheduler return path as discussed with Steven Rostedt ++ * ++ * Copyright (C) 2006, Kihon Technologies Inc. ++ * Steven Rostedt ++ * - debugged and patched Thomas Gleixner's rework. ++ * - added back the cmpxchg to the rework. ++ * - turned atomic require back on for SMP. ++ */ ++ ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++ ++#include "rtmutex_common.h" ++ ++/* ++ * struct mutex functions ++ */ ++void __mutex_do_init(struct mutex *mutex, const char *name, ++ struct lock_class_key *key) ++{ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ /* ++ * Make sure we are not reinitializing a held lock: ++ */ ++ debug_check_no_locks_freed((void *)mutex, sizeof(*mutex)); ++ lockdep_init_map(&mutex->dep_map, name, key, 0); ++#endif ++ mutex->lock.save_state = 0; ++} ++EXPORT_SYMBOL(__mutex_do_init); ++ ++void __lockfunc _mutex_lock(struct mutex *lock) ++{ ++ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_); ++ rt_mutex_lock(&lock->lock); ++} ++EXPORT_SYMBOL(_mutex_lock); ++ ++int __lockfunc _mutex_lock_interruptible(struct mutex *lock) ++{ ++ int ret; ++ ++ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_); ++ ret = rt_mutex_lock_interruptible(&lock->lock); ++ if (ret) ++ mutex_release(&lock->dep_map, 1, _RET_IP_); ++ return ret; ++} ++EXPORT_SYMBOL(_mutex_lock_interruptible); ++ ++int __lockfunc _mutex_lock_killable(struct mutex *lock) ++{ ++ int ret; ++ ++ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_); ++ ret = rt_mutex_lock_killable(&lock->lock); ++ if (ret) ++ mutex_release(&lock->dep_map, 1, _RET_IP_); ++ return ret; ++} ++EXPORT_SYMBOL(_mutex_lock_killable); ++ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++void __lockfunc _mutex_lock_nested(struct mutex *lock, int subclass) ++{ ++ mutex_acquire_nest(&lock->dep_map, subclass, 0, NULL, _RET_IP_); ++ rt_mutex_lock(&lock->lock); ++} ++EXPORT_SYMBOL(_mutex_lock_nested); ++ ++void __lockfunc _mutex_lock_nest_lock(struct mutex *lock, struct lockdep_map *nest) ++{ ++ mutex_acquire_nest(&lock->dep_map, 0, 0, nest, _RET_IP_); ++ rt_mutex_lock(&lock->lock); ++} ++EXPORT_SYMBOL(_mutex_lock_nest_lock); ++ ++int __lockfunc _mutex_lock_interruptible_nested(struct mutex *lock, int subclass) ++{ ++ int ret; ++ ++ mutex_acquire_nest(&lock->dep_map, subclass, 0, NULL, _RET_IP_); ++ ret = rt_mutex_lock_interruptible(&lock->lock); ++ if (ret) ++ mutex_release(&lock->dep_map, 1, _RET_IP_); ++ return ret; ++} ++EXPORT_SYMBOL(_mutex_lock_interruptible_nested); ++ ++int __lockfunc _mutex_lock_killable_nested(struct mutex *lock, int subclass) ++{ ++ int ret; ++ ++ mutex_acquire(&lock->dep_map, subclass, 0, _RET_IP_); ++ ret = rt_mutex_lock_killable(&lock->lock); ++ if (ret) ++ mutex_release(&lock->dep_map, 1, _RET_IP_); ++ return ret; ++} ++EXPORT_SYMBOL(_mutex_lock_killable_nested); ++#endif ++ ++int __lockfunc _mutex_trylock(struct mutex *lock) ++{ ++ int ret = rt_mutex_trylock(&lock->lock); ++ ++ if (ret) ++ mutex_acquire(&lock->dep_map, 0, 1, _RET_IP_); ++ ++ return ret; ++} ++EXPORT_SYMBOL(_mutex_trylock); ++ ++void __lockfunc _mutex_unlock(struct mutex *lock) ++{ ++ mutex_release(&lock->dep_map, 1, _RET_IP_); ++ rt_mutex_unlock(&lock->lock); ++} ++EXPORT_SYMBOL(_mutex_unlock); ++ ++/* ++ * rwlock_t functions ++ */ ++int __lockfunc rt_write_trylock(rwlock_t *rwlock) ++{ ++ int ret; ++ ++ migrate_disable(); ++ ret = rt_mutex_trylock(&rwlock->lock); ++ if (ret) ++ rwlock_acquire(&rwlock->dep_map, 0, 1, _RET_IP_); ++ else ++ migrate_enable(); ++ ++ return ret; ++} ++EXPORT_SYMBOL(rt_write_trylock); ++ ++int __lockfunc rt_write_trylock_irqsave(rwlock_t *rwlock, unsigned long *flags) ++{ ++ int ret; ++ ++ *flags = 0; ++ ret = rt_write_trylock(rwlock); ++ return ret; ++} ++EXPORT_SYMBOL(rt_write_trylock_irqsave); ++ ++int __lockfunc rt_read_trylock(rwlock_t *rwlock) ++{ ++ struct rt_mutex *lock = &rwlock->lock; ++ int ret = 1; ++ ++ /* ++ * recursive read locks succeed when current owns the lock, ++ * but not when read_depth == 0 which means that the lock is ++ * write locked. ++ */ ++ if (rt_mutex_owner(lock) != current) { ++ migrate_disable(); ++ ret = rt_mutex_trylock(lock); ++ if (ret) ++ rwlock_acquire(&rwlock->dep_map, 0, 1, _RET_IP_); ++ else ++ migrate_enable(); ++ ++ } else if (!rwlock->read_depth) { ++ ret = 0; ++ } ++ ++ if (ret) ++ rwlock->read_depth++; ++ ++ return ret; ++} ++EXPORT_SYMBOL(rt_read_trylock); ++ ++void __lockfunc rt_write_lock(rwlock_t *rwlock) ++{ ++ rwlock_acquire(&rwlock->dep_map, 0, 0, _RET_IP_); ++ __rt_spin_lock(&rwlock->lock); ++} ++EXPORT_SYMBOL(rt_write_lock); ++ ++void __lockfunc rt_read_lock(rwlock_t *rwlock) ++{ ++ struct rt_mutex *lock = &rwlock->lock; ++ ++ ++ /* ++ * recursive read locks succeed when current owns the lock ++ */ ++ if (rt_mutex_owner(lock) != current) { ++ rwlock_acquire(&rwlock->dep_map, 0, 0, _RET_IP_); ++ __rt_spin_lock(lock); ++ } ++ rwlock->read_depth++; ++} ++ ++EXPORT_SYMBOL(rt_read_lock); ++ ++void __lockfunc rt_write_unlock(rwlock_t *rwlock) ++{ ++ /* NOTE: we always pass in '1' for nested, for simplicity */ ++ rwlock_release(&rwlock->dep_map, 1, _RET_IP_); ++ __rt_spin_unlock(&rwlock->lock); ++ migrate_enable(); ++} ++EXPORT_SYMBOL(rt_write_unlock); ++ ++void __lockfunc rt_read_unlock(rwlock_t *rwlock) ++{ ++ /* Release the lock only when read_depth is down to 0 */ ++ if (--rwlock->read_depth == 0) { ++ rwlock_release(&rwlock->dep_map, 1, _RET_IP_); ++ __rt_spin_unlock(&rwlock->lock); ++ migrate_enable(); ++ } ++} ++EXPORT_SYMBOL(rt_read_unlock); ++ ++unsigned long __lockfunc rt_write_lock_irqsave(rwlock_t *rwlock) ++{ ++ rt_write_lock(rwlock); ++ ++ return 0; ++} ++EXPORT_SYMBOL(rt_write_lock_irqsave); ++ ++unsigned long __lockfunc rt_read_lock_irqsave(rwlock_t *rwlock) ++{ ++ rt_read_lock(rwlock); ++ ++ return 0; ++} ++EXPORT_SYMBOL(rt_read_lock_irqsave); ++ ++void __rt_rwlock_init(rwlock_t *rwlock, char *name, struct lock_class_key *key) ++{ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ /* ++ * Make sure we are not reinitializing a held lock: ++ */ ++ debug_check_no_locks_freed((void *)rwlock, sizeof(*rwlock)); ++ lockdep_init_map(&rwlock->dep_map, name, key, 0); ++#endif ++ rwlock->lock.save_state = 1; ++ rwlock->read_depth = 0; ++} ++EXPORT_SYMBOL(__rt_rwlock_init); ++ ++/** ++ * atomic_dec_and_mutex_lock - return holding mutex if we dec to 0 ++ * @cnt: the atomic which we are to dec ++ * @lock: the mutex to return holding if we dec to 0 ++ * ++ * return true and hold lock if we dec to 0, return false otherwise ++ */ ++int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock) ++{ ++ /* dec if we can't possibly hit 0 */ ++ if (atomic_add_unless(cnt, -1, 1)) ++ return 0; ++ /* we might hit 0, so take the lock */ ++ mutex_lock(lock); ++ if (!atomic_dec_and_test(cnt)) { ++ /* when we actually did the dec, we didn't hit 0 */ ++ mutex_unlock(lock); ++ return 0; ++ } ++ /* we hit 0, and we hold the lock */ ++ return 1; ++} ++EXPORT_SYMBOL(atomic_dec_and_mutex_lock); +diff --git a/kernel/locking/rtmutex-debug.c b/kernel/locking/rtmutex-debug.c +index 62b6cee8ea7f..0613c4b1d059 100644 +--- a/kernel/locking/rtmutex-debug.c ++++ b/kernel/locking/rtmutex-debug.c +@@ -173,12 +173,3 @@ void debug_rt_mutex_init(struct rt_mutex *lock, const char *name) + lock->name = name; + } + +-void +-rt_mutex_deadlock_account_lock(struct rt_mutex *lock, struct task_struct *task) +-{ +-} +- +-void rt_mutex_deadlock_account_unlock(struct task_struct *task) +-{ +-} +- +diff --git a/kernel/locking/rtmutex-debug.h b/kernel/locking/rtmutex-debug.h +index d0519c3432b6..b585af9a1b50 100644 +--- a/kernel/locking/rtmutex-debug.h ++++ b/kernel/locking/rtmutex-debug.h +@@ -9,9 +9,6 @@ + * This file contains macros used solely by rtmutex.c. Debug version. + */ + +-extern void +-rt_mutex_deadlock_account_lock(struct rt_mutex *lock, struct task_struct *task); +-extern void rt_mutex_deadlock_account_unlock(struct task_struct *task); + extern void debug_rt_mutex_init_waiter(struct rt_mutex_waiter *waiter); + extern void debug_rt_mutex_free_waiter(struct rt_mutex_waiter *waiter); + extern void debug_rt_mutex_init(struct rt_mutex *lock, const char *name); +diff --git a/kernel/locking/rtmutex.c b/kernel/locking/rtmutex.c +index 2c49d76f96c3..218f1d26afe7 100644 +--- a/kernel/locking/rtmutex.c ++++ b/kernel/locking/rtmutex.c +@@ -7,6 +7,11 @@ + * Copyright (C) 2005-2006 Timesys Corp., Thomas Gleixner + * Copyright (C) 2005 Kihon Technologies Inc., Steven Rostedt + * Copyright (C) 2006 Esben Nielsen ++ * Adaptive Spinlocks: ++ * Copyright (C) 2008 Novell, Inc., Gregory Haskins, Sven Dietrich, ++ * and Peter Morreale, ++ * Adaptive Spinlocks simplification: ++ * Copyright (C) 2008 Red Hat, Inc., Steven Rostedt + * + * See Documentation/locking/rt-mutex-design.txt for details. + */ +@@ -16,6 +21,7 @@ + #include + #include + #include ++#include + + #include "rtmutex_common.h" + +@@ -133,6 +139,12 @@ static void fixup_rt_mutex_waiters(struct rt_mutex *lock) + WRITE_ONCE(*p, owner & ~RT_MUTEX_HAS_WAITERS); + } + ++static int rt_mutex_real_waiter(struct rt_mutex_waiter *waiter) ++{ ++ return waiter && waiter != PI_WAKEUP_INPROGRESS && ++ waiter != PI_REQUEUE_INPROGRESS; ++} ++ + /* + * We can speed up the acquire/release, if there's no debugging state to be + * set up. +@@ -222,12 +234,25 @@ static inline bool unlock_rt_mutex_safe(struct rt_mutex *lock, + } + #endif + ++#define STEAL_NORMAL 0 ++#define STEAL_LATERAL 1 ++/* ++ * Only use with rt_mutex_waiter_{less,equal}() ++ */ ++#define task_to_waiter(p) \ ++ &(struct rt_mutex_waiter){ .prio = (p)->prio, .deadline = (p)->dl.deadline } ++ + static inline int + rt_mutex_waiter_less(struct rt_mutex_waiter *left, +- struct rt_mutex_waiter *right) ++ struct rt_mutex_waiter *right, int mode) + { +- if (left->prio < right->prio) +- return 1; ++ if (mode == STEAL_NORMAL) { ++ if (left->prio < right->prio) ++ return 1; ++ } else { ++ if (left->prio <= right->prio) ++ return 1; ++ } + + /* + * If both waiters have dl_prio(), we check the deadlines of the +@@ -236,12 +261,30 @@ rt_mutex_waiter_less(struct rt_mutex_waiter *left, + * then right waiter has a dl_prio() too. + */ + if (dl_prio(left->prio)) +- return dl_time_before(left->task->dl.deadline, +- right->task->dl.deadline); ++ return dl_time_before(left->deadline, right->deadline); + + return 0; + } + ++static inline int ++rt_mutex_waiter_equal(struct rt_mutex_waiter *left, ++ struct rt_mutex_waiter *right) ++{ ++ if (left->prio != right->prio) ++ return 0; ++ ++ /* ++ * If both waiters have dl_prio(), we check the deadlines of the ++ * associated tasks. ++ * If left waiter has a dl_prio(), and we didn't return 0 above, ++ * then right waiter has a dl_prio() too. ++ */ ++ if (dl_prio(left->prio)) ++ return left->deadline == right->deadline; ++ ++ return 1; ++} ++ + static void + rt_mutex_enqueue(struct rt_mutex *lock, struct rt_mutex_waiter *waiter) + { +@@ -253,7 +296,7 @@ rt_mutex_enqueue(struct rt_mutex *lock, struct rt_mutex_waiter *waiter) + while (*link) { + parent = *link; + entry = rb_entry(parent, struct rt_mutex_waiter, tree_entry); +- if (rt_mutex_waiter_less(waiter, entry)) { ++ if (rt_mutex_waiter_less(waiter, entry, STEAL_NORMAL)) { + link = &parent->rb_left; + } else { + link = &parent->rb_right; +@@ -292,7 +335,7 @@ rt_mutex_enqueue_pi(struct task_struct *task, struct rt_mutex_waiter *waiter) + while (*link) { + parent = *link; + entry = rb_entry(parent, struct rt_mutex_waiter, pi_tree_entry); +- if (rt_mutex_waiter_less(waiter, entry)) { ++ if (rt_mutex_waiter_less(waiter, entry, STEAL_NORMAL)) { + link = &parent->rb_left; + } else { + link = &parent->rb_right; +@@ -320,72 +363,16 @@ rt_mutex_dequeue_pi(struct task_struct *task, struct rt_mutex_waiter *waiter) + RB_CLEAR_NODE(&waiter->pi_tree_entry); + } + +-/* +- * Calculate task priority from the waiter tree priority +- * +- * Return task->normal_prio when the waiter tree is empty or when +- * the waiter is not allowed to do priority boosting +- */ +-int rt_mutex_getprio(struct task_struct *task) ++static void rt_mutex_adjust_prio(struct task_struct *p) + { +- if (likely(!task_has_pi_waiters(task))) +- return task->normal_prio; ++ struct task_struct *pi_task = NULL; + +- return min(task_top_pi_waiter(task)->prio, +- task->normal_prio); +-} ++ lockdep_assert_held(&p->pi_lock); + +-struct task_struct *rt_mutex_get_top_task(struct task_struct *task) +-{ +- if (likely(!task_has_pi_waiters(task))) +- return NULL; ++ if (task_has_pi_waiters(p)) ++ pi_task = task_top_pi_waiter(p)->task; + +- return task_top_pi_waiter(task)->task; +-} +- +-/* +- * Called by sched_setscheduler() to get the priority which will be +- * effective after the change. +- */ +-int rt_mutex_get_effective_prio(struct task_struct *task, int newprio) +-{ +- if (!task_has_pi_waiters(task)) +- return newprio; +- +- if (task_top_pi_waiter(task)->task->prio <= newprio) +- return task_top_pi_waiter(task)->task->prio; +- return newprio; +-} +- +-/* +- * Adjust the priority of a task, after its pi_waiters got modified. +- * +- * This can be both boosting and unboosting. task->pi_lock must be held. +- */ +-static void __rt_mutex_adjust_prio(struct task_struct *task) +-{ +- int prio = rt_mutex_getprio(task); +- +- if (task->prio != prio || dl_prio(prio)) +- rt_mutex_setprio(task, prio); +-} +- +-/* +- * Adjust task priority (undo boosting). Called from the exit path of +- * rt_mutex_slowunlock() and rt_mutex_slowlock(). +- * +- * (Note: We do this outside of the protection of lock->wait_lock to +- * allow the lock to be taken while or before we readjust the priority +- * of task. We do not use the spin_xx_mutex() variants here as we are +- * outside of the debug path.) +- */ +-void rt_mutex_adjust_prio(struct task_struct *task) +-{ +- unsigned long flags; +- +- raw_spin_lock_irqsave(&task->pi_lock, flags); +- __rt_mutex_adjust_prio(task); +- raw_spin_unlock_irqrestore(&task->pi_lock, flags); ++ rt_mutex_setprio(p, pi_task); + } + + /* +@@ -414,6 +401,14 @@ static bool rt_mutex_cond_detect_deadlock(struct rt_mutex_waiter *waiter, + return debug_rt_mutex_detect_deadlock(waiter, chwalk); + } + ++static void rt_mutex_wake_waiter(struct rt_mutex_waiter *waiter) ++{ ++ if (waiter->savestate) ++ wake_up_lock_sleeper(waiter->task); ++ else ++ wake_up_process(waiter->task); ++} ++ + /* + * Max number of times we'll walk the boosting chain: + */ +@@ -421,7 +416,8 @@ int max_lock_depth = 1024; + + static inline struct rt_mutex *task_blocked_on_lock(struct task_struct *p) + { +- return p->pi_blocked_on ? p->pi_blocked_on->lock : NULL; ++ return rt_mutex_real_waiter(p->pi_blocked_on) ? ++ p->pi_blocked_on->lock : NULL; + } + + /* +@@ -557,7 +553,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, + * reached or the state of the chain has changed while we + * dropped the locks. + */ +- if (!waiter) ++ if (!rt_mutex_real_waiter(waiter)) + goto out_unlock_pi; + + /* +@@ -608,7 +604,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, + * enabled we continue, but stop the requeueing in the chain + * walk. + */ +- if (waiter->prio == task->prio) { ++ if (rt_mutex_waiter_equal(waiter, task_to_waiter(task))) { + if (!detect_deadlock) + goto out_unlock_pi; + else +@@ -704,7 +700,26 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, + + /* [7] Requeue the waiter in the lock waiter tree. */ + rt_mutex_dequeue(lock, waiter); ++ ++ /* ++ * Update the waiter prio fields now that we're dequeued. ++ * ++ * These values can have changed through either: ++ * ++ * sys_sched_set_scheduler() / sys_sched_setattr() ++ * ++ * or ++ * ++ * DL CBS enforcement advancing the effective deadline. ++ * ++ * Even though pi_waiters also uses these fields, and that tree is only ++ * updated in [11], we can do this here, since we hold [L], which ++ * serializes all pi_waiters access and rb_erase() does not care about ++ * the values of the node being removed. ++ */ + waiter->prio = task->prio; ++ waiter->deadline = task->dl.deadline; ++ + rt_mutex_enqueue(lock, waiter); + + /* [8] Release the task */ +@@ -719,13 +734,16 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, + * follow here. This is the end of the chain we are walking. + */ + if (!rt_mutex_owner(lock)) { ++ struct rt_mutex_waiter *lock_top_waiter; ++ + /* + * If the requeue [7] above changed the top waiter, + * then we need to wake the new top waiter up to try + * to get the lock. + */ +- if (prerequeue_top_waiter != rt_mutex_top_waiter(lock)) +- wake_up_process(rt_mutex_top_waiter(lock)->task); ++ lock_top_waiter = rt_mutex_top_waiter(lock); ++ if (prerequeue_top_waiter != lock_top_waiter) ++ rt_mutex_wake_waiter(lock_top_waiter); + raw_spin_unlock_irq(&lock->wait_lock); + return 0; + } +@@ -745,7 +763,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, + */ + rt_mutex_dequeue_pi(task, prerequeue_top_waiter); + rt_mutex_enqueue_pi(task, waiter); +- __rt_mutex_adjust_prio(task); ++ rt_mutex_adjust_prio(task); + + } else if (prerequeue_top_waiter == waiter) { + /* +@@ -761,7 +779,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, + rt_mutex_dequeue_pi(task, waiter); + waiter = rt_mutex_top_waiter(lock); + rt_mutex_enqueue_pi(task, waiter); +- __rt_mutex_adjust_prio(task); ++ rt_mutex_adjust_prio(task); + } else { + /* + * Nothing changed. No need to do any priority +@@ -818,6 +836,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, + return ret; + } + ++ + /* + * Try to take an rt-mutex + * +@@ -828,9 +847,12 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, + * @waiter: The waiter that is queued to the lock's wait tree if the + * callsite called task_blocked_on_lock(), otherwise NULL + */ +-static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, +- struct rt_mutex_waiter *waiter) ++static int __try_to_take_rt_mutex(struct rt_mutex *lock, ++ struct task_struct *task, ++ struct rt_mutex_waiter *waiter, int mode) + { ++ lockdep_assert_held(&lock->wait_lock); ++ + /* + * Before testing whether we can acquire @lock, we set the + * RT_MUTEX_HAS_WAITERS bit in @lock->owner. This forces all +@@ -866,8 +888,10 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, + * If waiter is not the highest priority waiter of + * @lock, give up. + */ +- if (waiter != rt_mutex_top_waiter(lock)) ++ if (waiter != rt_mutex_top_waiter(lock)) { ++ /* XXX rt_mutex_waiter_less() ? */ + return 0; ++ } + + /* + * We can acquire the lock. Remove the waiter from the +@@ -885,14 +909,26 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, + * not need to be dequeued. + */ + if (rt_mutex_has_waiters(lock)) { ++ struct task_struct *pown = rt_mutex_top_waiter(lock)->task; ++ ++ if (task != pown) ++ return 0; ++ ++ /* ++ * Note that RT tasks are excluded from lateral-steals ++ * to prevent the introduction of an unbounded latency. ++ */ ++ if (rt_task(task)) ++ mode = STEAL_NORMAL; + /* + * If @task->prio is greater than or equal to + * the top waiter priority (kernel view), + * @task lost. + */ +- if (task->prio >= rt_mutex_top_waiter(lock)->prio) ++ if (!rt_mutex_waiter_less(task_to_waiter(task), ++ rt_mutex_top_waiter(lock), ++ mode)) + return 0; +- + /* + * The current top waiter stays enqueued. We + * don't have to change anything in the lock +@@ -936,11 +972,384 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, + */ + rt_mutex_set_owner(lock, task); + +- rt_mutex_deadlock_account_lock(lock, task); +- + return 1; + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++/* ++ * preemptible spin_lock functions: ++ */ ++static inline void rt_spin_lock_fastlock(struct rt_mutex *lock, ++ void (*slowfn)(struct rt_mutex *lock, ++ bool mg_off), ++ bool do_mig_dis) ++{ ++ might_sleep_no_state_check(); ++ ++ if (do_mig_dis) ++ migrate_disable(); ++ ++ if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) ++ return; ++ else ++ slowfn(lock, do_mig_dis); ++} ++ ++static inline void rt_spin_lock_fastunlock(struct rt_mutex *lock, ++ void (*slowfn)(struct rt_mutex *lock)) ++{ ++ if (likely(rt_mutex_cmpxchg_release(lock, current, NULL))) ++ return; ++ else ++ slowfn(lock); ++} ++#ifdef CONFIG_SMP ++/* ++ * Note that owner is a speculative pointer and dereferencing relies ++ * on rcu_read_lock() and the check against the lock owner. ++ */ ++static int adaptive_wait(struct rt_mutex *lock, ++ struct task_struct *owner) ++{ ++ int res = 0; ++ ++ rcu_read_lock(); ++ for (;;) { ++ if (owner != rt_mutex_owner(lock)) ++ break; ++ /* ++ * Ensure that owner->on_cpu is dereferenced _after_ ++ * checking the above to be valid. ++ */ ++ barrier(); ++ if (!owner->on_cpu) { ++ res = 1; ++ break; ++ } ++ cpu_relax(); ++ } ++ rcu_read_unlock(); ++ return res; ++} ++#else ++static int adaptive_wait(struct rt_mutex *lock, ++ struct task_struct *orig_owner) ++{ ++ return 1; ++} ++#endif ++ ++static int task_blocks_on_rt_mutex(struct rt_mutex *lock, ++ struct rt_mutex_waiter *waiter, ++ struct task_struct *task, ++ enum rtmutex_chainwalk chwalk); ++/* ++ * Slow path lock function spin_lock style: this variant is very ++ * careful not to miss any non-lock wakeups. ++ * ++ * We store the current state under p->pi_lock in p->saved_state and ++ * the try_to_wake_up() code handles this accordingly. ++ */ ++static void noinline __sched rt_spin_lock_slowlock(struct rt_mutex *lock, ++ bool mg_off) ++{ ++ struct task_struct *lock_owner, *self = current; ++ struct rt_mutex_waiter waiter, *top_waiter; ++ unsigned long flags; ++ int ret; ++ ++ rt_mutex_init_waiter(&waiter, true); ++ ++ raw_spin_lock_irqsave(&lock->wait_lock, flags); ++ ++ if (__try_to_take_rt_mutex(lock, self, NULL, STEAL_LATERAL)) { ++ raw_spin_unlock_irqrestore(&lock->wait_lock, flags); ++ return; ++ } ++ ++ BUG_ON(rt_mutex_owner(lock) == self); ++ ++ /* ++ * We save whatever state the task is in and we'll restore it ++ * after acquiring the lock taking real wakeups into account ++ * as well. We are serialized via pi_lock against wakeups. See ++ * try_to_wake_up(). ++ */ ++ raw_spin_lock(&self->pi_lock); ++ self->saved_state = self->state; ++ __set_current_state_no_track(TASK_UNINTERRUPTIBLE); ++ raw_spin_unlock(&self->pi_lock); ++ ++ ret = task_blocks_on_rt_mutex(lock, &waiter, self, RT_MUTEX_MIN_CHAINWALK); ++ BUG_ON(ret); ++ ++ for (;;) { ++ /* Try to acquire the lock again. */ ++ if (__try_to_take_rt_mutex(lock, self, &waiter, STEAL_LATERAL)) ++ break; ++ ++ top_waiter = rt_mutex_top_waiter(lock); ++ lock_owner = rt_mutex_owner(lock); ++ ++ raw_spin_unlock_irqrestore(&lock->wait_lock, flags); ++ ++ debug_rt_mutex_print_deadlock(&waiter); ++ ++ if (top_waiter != &waiter || adaptive_wait(lock, lock_owner)) { ++ if (mg_off) ++ migrate_enable(); ++ schedule(); ++ if (mg_off) ++ migrate_disable(); ++ } ++ ++ raw_spin_lock_irqsave(&lock->wait_lock, flags); ++ ++ raw_spin_lock(&self->pi_lock); ++ __set_current_state_no_track(TASK_UNINTERRUPTIBLE); ++ raw_spin_unlock(&self->pi_lock); ++ } ++ ++ /* ++ * Restore the task state to current->saved_state. We set it ++ * to the original state above and the try_to_wake_up() code ++ * has possibly updated it when a real (non-rtmutex) wakeup ++ * happened while we were blocked. Clear saved_state so ++ * try_to_wakeup() does not get confused. ++ */ ++ raw_spin_lock(&self->pi_lock); ++ __set_current_state_no_track(self->saved_state); ++ self->saved_state = TASK_RUNNING; ++ raw_spin_unlock(&self->pi_lock); ++ ++ /* ++ * try_to_take_rt_mutex() sets the waiter bit ++ * unconditionally. We might have to fix that up: ++ */ ++ fixup_rt_mutex_waiters(lock); ++ ++ BUG_ON(rt_mutex_has_waiters(lock) && &waiter == rt_mutex_top_waiter(lock)); ++ BUG_ON(!RB_EMPTY_NODE(&waiter.tree_entry)); ++ ++ raw_spin_unlock_irqrestore(&lock->wait_lock, flags); ++ ++ debug_rt_mutex_free_waiter(&waiter); ++} ++ ++static bool __sched __rt_mutex_unlock_common(struct rt_mutex *lock, ++ struct wake_q_head *wake_q, ++ struct wake_q_head *wq_sleeper); ++/* ++ * Slow path to release a rt_mutex spin_lock style ++ */ ++static void noinline __sched rt_spin_lock_slowunlock(struct rt_mutex *lock) ++{ ++ unsigned long flags; ++ WAKE_Q(wake_q); ++ WAKE_Q(wake_sleeper_q); ++ bool postunlock; ++ ++ raw_spin_lock_irqsave(&lock->wait_lock, flags); ++ postunlock = __rt_mutex_unlock_common(lock, &wake_q, &wake_sleeper_q); ++ raw_spin_unlock_irqrestore(&lock->wait_lock, flags); ++ ++ if (postunlock) ++ rt_mutex_postunlock(&wake_q, &wake_sleeper_q); ++} ++ ++void __lockfunc rt_spin_lock__no_mg(spinlock_t *lock) ++{ ++ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock, false); ++ spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); ++} ++EXPORT_SYMBOL(rt_spin_lock__no_mg); ++ ++void __lockfunc rt_spin_lock(spinlock_t *lock) ++{ ++ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock, true); ++ spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); ++} ++EXPORT_SYMBOL(rt_spin_lock); ++ ++void __lockfunc __rt_spin_lock(struct rt_mutex *lock) ++{ ++ rt_spin_lock_fastlock(lock, rt_spin_lock_slowlock, true); ++} ++EXPORT_SYMBOL(__rt_spin_lock); ++ ++void __lockfunc __rt_spin_lock__no_mg(struct rt_mutex *lock) ++{ ++ rt_spin_lock_fastlock(lock, rt_spin_lock_slowlock, false); ++} ++EXPORT_SYMBOL(__rt_spin_lock__no_mg); ++ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++void __lockfunc rt_spin_lock_nested(spinlock_t *lock, int subclass) ++{ ++ spin_acquire(&lock->dep_map, subclass, 0, _RET_IP_); ++ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock, true); ++} ++EXPORT_SYMBOL(rt_spin_lock_nested); ++#endif ++ ++void __lockfunc rt_spin_unlock__no_mg(spinlock_t *lock) ++{ ++ /* NOTE: we always pass in '1' for nested, for simplicity */ ++ spin_release(&lock->dep_map, 1, _RET_IP_); ++ rt_spin_lock_fastunlock(&lock->lock, rt_spin_lock_slowunlock); ++} ++EXPORT_SYMBOL(rt_spin_unlock__no_mg); ++ ++void __lockfunc rt_spin_unlock(spinlock_t *lock) ++{ ++ /* NOTE: we always pass in '1' for nested, for simplicity */ ++ spin_release(&lock->dep_map, 1, _RET_IP_); ++ rt_spin_lock_fastunlock(&lock->lock, rt_spin_lock_slowunlock); ++ migrate_enable(); ++} ++EXPORT_SYMBOL(rt_spin_unlock); ++ ++void __lockfunc __rt_spin_unlock(struct rt_mutex *lock) ++{ ++ rt_spin_lock_fastunlock(lock, rt_spin_lock_slowunlock); ++} ++EXPORT_SYMBOL(__rt_spin_unlock); ++ ++/* ++ * Wait for the lock to get unlocked: instead of polling for an unlock ++ * (like raw spinlocks do), we lock and unlock, to force the kernel to ++ * schedule if there's contention: ++ */ ++void __lockfunc rt_spin_unlock_wait(spinlock_t *lock) ++{ ++ spin_lock(lock); ++ spin_unlock(lock); ++} ++EXPORT_SYMBOL(rt_spin_unlock_wait); ++ ++int __lockfunc rt_spin_trylock__no_mg(spinlock_t *lock) ++{ ++ int ret; ++ ++ ret = rt_mutex_trylock(&lock->lock); ++ if (ret) ++ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); ++ return ret; ++} ++EXPORT_SYMBOL(rt_spin_trylock__no_mg); ++ ++int __lockfunc rt_spin_trylock(spinlock_t *lock) ++{ ++ int ret; ++ ++ migrate_disable(); ++ ret = rt_mutex_trylock(&lock->lock); ++ if (ret) ++ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); ++ else ++ migrate_enable(); ++ return ret; ++} ++EXPORT_SYMBOL(rt_spin_trylock); ++ ++int __lockfunc rt_spin_trylock_bh(spinlock_t *lock) ++{ ++ int ret; ++ ++ local_bh_disable(); ++ ret = rt_mutex_trylock(&lock->lock); ++ if (ret) { ++ migrate_disable(); ++ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); ++ } else ++ local_bh_enable(); ++ return ret; ++} ++EXPORT_SYMBOL(rt_spin_trylock_bh); ++ ++int __lockfunc rt_spin_trylock_irqsave(spinlock_t *lock, unsigned long *flags) ++{ ++ int ret; ++ ++ *flags = 0; ++ ret = rt_mutex_trylock(&lock->lock); ++ if (ret) { ++ migrate_disable(); ++ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); ++ } ++ return ret; ++} ++EXPORT_SYMBOL(rt_spin_trylock_irqsave); ++ ++int atomic_dec_and_spin_lock(atomic_t *atomic, spinlock_t *lock) ++{ ++ /* Subtract 1 from counter unless that drops it to 0 (ie. it was 1) */ ++ if (atomic_add_unless(atomic, -1, 1)) ++ return 0; ++ rt_spin_lock(lock); ++ if (atomic_dec_and_test(atomic)) ++ return 1; ++ rt_spin_unlock(lock); ++ return 0; ++} ++EXPORT_SYMBOL(atomic_dec_and_spin_lock); ++ ++ void ++__rt_spin_lock_init(spinlock_t *lock, char *name, struct lock_class_key *key) ++{ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ /* ++ * Make sure we are not reinitializing a held lock: ++ */ ++ debug_check_no_locks_freed((void *)lock, sizeof(*lock)); ++ lockdep_init_map(&lock->dep_map, name, key, 0); ++#endif ++} ++EXPORT_SYMBOL(__rt_spin_lock_init); ++ ++#endif /* PREEMPT_RT_FULL */ ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ static inline int __sched ++__mutex_lock_check_stamp(struct rt_mutex *lock, struct ww_acquire_ctx *ctx) ++{ ++ struct ww_mutex *ww = container_of(lock, struct ww_mutex, base.lock); ++ struct ww_acquire_ctx *hold_ctx = ACCESS_ONCE(ww->ctx); ++ ++ if (!hold_ctx) ++ return 0; ++ ++ if (unlikely(ctx == hold_ctx)) ++ return -EALREADY; ++ ++ if (ctx->stamp - hold_ctx->stamp <= LONG_MAX && ++ (ctx->stamp != hold_ctx->stamp || ctx > hold_ctx)) { ++#ifdef CONFIG_DEBUG_MUTEXES ++ DEBUG_LOCKS_WARN_ON(ctx->contending_lock); ++ ctx->contending_lock = ww; ++#endif ++ return -EDEADLK; ++ } ++ ++ return 0; ++} ++#else ++ static inline int __sched ++__mutex_lock_check_stamp(struct rt_mutex *lock, struct ww_acquire_ctx *ctx) ++{ ++ BUG(); ++ return 0; ++} ++ ++#endif ++ ++static inline int ++try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, ++ struct rt_mutex_waiter *waiter) ++{ ++ return __try_to_take_rt_mutex(lock, task, waiter, STEAL_NORMAL); ++} ++ + /* + * Task blocks on lock. + * +@@ -958,6 +1367,8 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, + struct rt_mutex *next_lock; + int chain_walk = 0, res; + ++ lockdep_assert_held(&lock->wait_lock); ++ + /* + * Early deadlock detection. We really don't want the task to + * enqueue on itself just to untangle the mess later. It's not +@@ -971,10 +1382,28 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, + return -EDEADLK; + + raw_spin_lock(&task->pi_lock); +- __rt_mutex_adjust_prio(task); ++ ++ /* ++ * In the case of futex requeue PI, this will be a proxy ++ * lock. The task will wake unaware that it is enqueueed on ++ * this lock. Avoid blocking on two locks and corrupting ++ * pi_blocked_on via the PI_WAKEUP_INPROGRESS ++ * flag. futex_wait_requeue_pi() sets this when it wakes up ++ * before requeue (due to a signal or timeout). Do not enqueue ++ * the task if PI_WAKEUP_INPROGRESS is set. ++ */ ++ if (task != current && task->pi_blocked_on == PI_WAKEUP_INPROGRESS) { ++ raw_spin_unlock(&task->pi_lock); ++ return -EAGAIN; ++ } ++ ++ BUG_ON(rt_mutex_real_waiter(task->pi_blocked_on)); ++ ++ rt_mutex_adjust_prio(task); + waiter->task = task; + waiter->lock = lock; + waiter->prio = task->prio; ++ waiter->deadline = task->dl.deadline; + + /* Get the top priority waiter on the lock */ + if (rt_mutex_has_waiters(lock)) +@@ -993,8 +1422,8 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, + rt_mutex_dequeue_pi(owner, top_waiter); + rt_mutex_enqueue_pi(owner, waiter); + +- __rt_mutex_adjust_prio(owner); +- if (owner->pi_blocked_on) ++ rt_mutex_adjust_prio(owner); ++ if (rt_mutex_real_waiter(owner->pi_blocked_on)) + chain_walk = 1; + } else if (rt_mutex_cond_detect_deadlock(waiter, chwalk)) { + chain_walk = 1; +@@ -1036,6 +1465,7 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, + * Called with lock->wait_lock held and interrupts disabled. + */ + static void mark_wakeup_next_waiter(struct wake_q_head *wake_q, ++ struct wake_q_head *wake_sleeper_q, + struct rt_mutex *lock) + { + struct rt_mutex_waiter *waiter; +@@ -1045,12 +1475,14 @@ static void mark_wakeup_next_waiter(struct wake_q_head *wake_q, + waiter = rt_mutex_top_waiter(lock); + + /* +- * Remove it from current->pi_waiters. We do not adjust a +- * possible priority boost right now. We execute wakeup in the +- * boosted mode and go back to normal after releasing +- * lock->wait_lock. ++ * Remove it from current->pi_waiters and deboost. ++ * ++ * We must in fact deboost here in order to ensure we call ++ * rt_mutex_setprio() to update p->pi_top_task before the ++ * task unblocks. + */ + rt_mutex_dequeue_pi(current, waiter); ++ rt_mutex_adjust_prio(current); + + /* + * As we are waking up the top waiter, and the waiter stays +@@ -1062,9 +1494,22 @@ static void mark_wakeup_next_waiter(struct wake_q_head *wake_q, + */ + lock->owner = (void *) RT_MUTEX_HAS_WAITERS; + ++ /* ++ * We deboosted before waking the top waiter task such that we don't ++ * run two tasks with the 'same' priority (and ensure the ++ * p->pi_top_task pointer points to a blocked task). This however can ++ * lead to priority inversion if we would get preempted after the ++ * deboost but before waking our donor task, hence the preempt_disable() ++ * before unlock. ++ * ++ * Pairs with preempt_enable() in rt_mutex_postunlock(); ++ */ ++ preempt_disable(); ++ if (waiter->savestate) ++ wake_q_add(wake_sleeper_q, waiter->task); ++ else ++ wake_q_add(wake_q, waiter->task); + raw_spin_unlock(¤t->pi_lock); +- +- wake_q_add(wake_q, waiter->task); + } + + /* +@@ -1078,7 +1523,9 @@ static void remove_waiter(struct rt_mutex *lock, + { + bool is_top_waiter = (waiter == rt_mutex_top_waiter(lock)); + struct task_struct *owner = rt_mutex_owner(lock); +- struct rt_mutex *next_lock; ++ struct rt_mutex *next_lock = NULL; ++ ++ lockdep_assert_held(&lock->wait_lock); + + raw_spin_lock(¤t->pi_lock); + rt_mutex_dequeue(lock, waiter); +@@ -1099,10 +1546,11 @@ static void remove_waiter(struct rt_mutex *lock, + if (rt_mutex_has_waiters(lock)) + rt_mutex_enqueue_pi(owner, rt_mutex_top_waiter(lock)); + +- __rt_mutex_adjust_prio(owner); ++ rt_mutex_adjust_prio(owner); + + /* Store the lock on which owner is blocked or NULL */ +- next_lock = task_blocked_on_lock(owner); ++ if (rt_mutex_real_waiter(owner->pi_blocked_on)) ++ next_lock = task_blocked_on_lock(owner); + + raw_spin_unlock(&owner->pi_lock); + +@@ -1138,21 +1586,30 @@ void rt_mutex_adjust_pi(struct task_struct *task) + raw_spin_lock_irqsave(&task->pi_lock, flags); + + waiter = task->pi_blocked_on; +- if (!waiter || (waiter->prio == task->prio && +- !dl_prio(task->prio))) { ++ if (!rt_mutex_real_waiter(waiter) || ++ rt_mutex_waiter_equal(waiter, task_to_waiter(task))) { + raw_spin_unlock_irqrestore(&task->pi_lock, flags); + return; + } + next_lock = waiter->lock; +- raw_spin_unlock_irqrestore(&task->pi_lock, flags); + + /* gets dropped in rt_mutex_adjust_prio_chain()! */ + get_task_struct(task); + ++ raw_spin_unlock_irqrestore(&task->pi_lock, flags); + rt_mutex_adjust_prio_chain(task, RT_MUTEX_MIN_CHAINWALK, NULL, + next_lock, NULL, task); + } + ++void rt_mutex_init_waiter(struct rt_mutex_waiter *waiter, bool savestate) ++{ ++ debug_rt_mutex_init_waiter(waiter); ++ RB_CLEAR_NODE(&waiter->pi_tree_entry); ++ RB_CLEAR_NODE(&waiter->tree_entry); ++ waiter->task = NULL; ++ waiter->savestate = savestate; ++} ++ + /** + * __rt_mutex_slowlock() - Perform the wait-wake-try-to-take loop + * @lock: the rt_mutex to take +@@ -1166,7 +1623,8 @@ void rt_mutex_adjust_pi(struct task_struct *task) + static int __sched + __rt_mutex_slowlock(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, +- struct rt_mutex_waiter *waiter) ++ struct rt_mutex_waiter *waiter, ++ struct ww_acquire_ctx *ww_ctx) + { + int ret = 0; + +@@ -1175,16 +1633,17 @@ __rt_mutex_slowlock(struct rt_mutex *lock, int state, + if (try_to_take_rt_mutex(lock, current, waiter)) + break; + +- /* +- * TASK_INTERRUPTIBLE checks for signals and +- * timeout. Ignored otherwise. +- */ +- if (unlikely(state == TASK_INTERRUPTIBLE)) { +- /* Signal pending? */ +- if (signal_pending(current)) +- ret = -EINTR; +- if (timeout && !timeout->task) +- ret = -ETIMEDOUT; ++ if (timeout && !timeout->task) { ++ ret = -ETIMEDOUT; ++ break; ++ } ++ if (signal_pending_state(state, current)) { ++ ret = -EINTR; ++ break; ++ } ++ ++ if (ww_ctx && ww_ctx->acquired > 0) { ++ ret = __mutex_lock_check_stamp(lock, ww_ctx); + if (ret) + break; + } +@@ -1223,21 +1682,148 @@ static void rt_mutex_handle_deadlock(int res, int detect_deadlock, + } + } + ++static __always_inline void ww_mutex_lock_acquired(struct ww_mutex *ww, ++ struct ww_acquire_ctx *ww_ctx) ++{ ++#ifdef CONFIG_DEBUG_MUTEXES ++ /* ++ * If this WARN_ON triggers, you used ww_mutex_lock to acquire, ++ * but released with a normal mutex_unlock in this call. ++ * ++ * This should never happen, always use ww_mutex_unlock. ++ */ ++ DEBUG_LOCKS_WARN_ON(ww->ctx); ++ ++ /* ++ * Not quite done after calling ww_acquire_done() ? ++ */ ++ DEBUG_LOCKS_WARN_ON(ww_ctx->done_acquire); ++ ++ if (ww_ctx->contending_lock) { ++ /* ++ * After -EDEADLK you tried to ++ * acquire a different ww_mutex? Bad! ++ */ ++ DEBUG_LOCKS_WARN_ON(ww_ctx->contending_lock != ww); ++ ++ /* ++ * You called ww_mutex_lock after receiving -EDEADLK, ++ * but 'forgot' to unlock everything else first? ++ */ ++ DEBUG_LOCKS_WARN_ON(ww_ctx->acquired > 0); ++ ww_ctx->contending_lock = NULL; ++ } ++ ++ /* ++ * Naughty, using a different class will lead to undefined behavior! ++ */ ++ DEBUG_LOCKS_WARN_ON(ww_ctx->ww_class != ww->ww_class); ++#endif ++ ww_ctx->acquired++; ++} ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++static void ww_mutex_account_lock(struct rt_mutex *lock, ++ struct ww_acquire_ctx *ww_ctx) ++{ ++ struct ww_mutex *ww = container_of(lock, struct ww_mutex, base.lock); ++ struct rt_mutex_waiter *waiter, *n; ++ ++ /* ++ * This branch gets optimized out for the common case, ++ * and is only important for ww_mutex_lock. ++ */ ++ ww_mutex_lock_acquired(ww, ww_ctx); ++ ww->ctx = ww_ctx; ++ ++ /* ++ * Give any possible sleeping processes the chance to wake up, ++ * so they can recheck if they have to back off. ++ */ ++ rbtree_postorder_for_each_entry_safe(waiter, n, &lock->waiters, ++ tree_entry) { ++ /* XXX debug rt mutex waiter wakeup */ ++ ++ BUG_ON(waiter->lock != lock); ++ rt_mutex_wake_waiter(waiter); ++ } ++} ++ ++#else ++ ++static void ww_mutex_account_lock(struct rt_mutex *lock, ++ struct ww_acquire_ctx *ww_ctx) ++{ ++ BUG(); ++} ++#endif ++ ++int __sched rt_mutex_slowlock_locked(struct rt_mutex *lock, int state, ++ struct hrtimer_sleeper *timeout, ++ enum rtmutex_chainwalk chwalk, ++ struct ww_acquire_ctx *ww_ctx, ++ struct rt_mutex_waiter *waiter) ++{ ++ int ret; ++ ++ /* Try to acquire the lock again: */ ++ if (try_to_take_rt_mutex(lock, current, NULL)) { ++ if (ww_ctx) ++ ww_mutex_account_lock(lock, ww_ctx); ++ return 0; ++ } ++ ++ set_current_state(state); ++ ++ /* Setup the timer, when timeout != NULL */ ++ if (unlikely(timeout)) ++ hrtimer_start_expires(&timeout->timer, HRTIMER_MODE_ABS); ++ ++ ret = task_blocks_on_rt_mutex(lock, waiter, current, chwalk); ++ ++ if (likely(!ret)) { ++ /* sleep on the mutex */ ++ ret = __rt_mutex_slowlock(lock, state, timeout, waiter, ++ ww_ctx); ++ } else if (ww_ctx) { ++ /* ww_mutex received EDEADLK, let it become EALREADY */ ++ ret = __mutex_lock_check_stamp(lock, ww_ctx); ++ BUG_ON(!ret); ++ } ++ ++ if (unlikely(ret)) { ++ __set_current_state(TASK_RUNNING); ++ if (rt_mutex_has_waiters(lock)) ++ remove_waiter(lock, waiter); ++ /* ww_mutex want to report EDEADLK/EALREADY, let them */ ++ if (!ww_ctx) ++ rt_mutex_handle_deadlock(ret, chwalk, waiter); ++ } else if (ww_ctx) { ++ ww_mutex_account_lock(lock, ww_ctx); ++ } ++ ++ /* ++ * try_to_take_rt_mutex() sets the waiter bit ++ * unconditionally. We might have to fix that up. ++ */ ++ fixup_rt_mutex_waiters(lock); ++ return ret; ++} ++ + /* + * Slow path lock function: + */ + static int __sched + rt_mutex_slowlock(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, +- enum rtmutex_chainwalk chwalk) ++ enum rtmutex_chainwalk chwalk, ++ struct ww_acquire_ctx *ww_ctx) + { + struct rt_mutex_waiter waiter; + unsigned long flags; + int ret = 0; + +- debug_rt_mutex_init_waiter(&waiter); +- RB_CLEAR_NODE(&waiter.pi_tree_entry); +- RB_CLEAR_NODE(&waiter.tree_entry); ++ rt_mutex_init_waiter(&waiter, false); + + /* + * Technically we could use raw_spin_[un]lock_irq() here, but this can +@@ -1249,36 +1835,8 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state, + */ + raw_spin_lock_irqsave(&lock->wait_lock, flags); + +- /* Try to acquire the lock again: */ +- if (try_to_take_rt_mutex(lock, current, NULL)) { +- raw_spin_unlock_irqrestore(&lock->wait_lock, flags); +- return 0; +- } +- +- set_current_state(state); +- +- /* Setup the timer, when timeout != NULL */ +- if (unlikely(timeout)) +- hrtimer_start_expires(&timeout->timer, HRTIMER_MODE_ABS); +- +- ret = task_blocks_on_rt_mutex(lock, &waiter, current, chwalk); +- +- if (likely(!ret)) +- /* sleep on the mutex */ +- ret = __rt_mutex_slowlock(lock, state, timeout, &waiter); +- +- if (unlikely(ret)) { +- __set_current_state(TASK_RUNNING); +- if (rt_mutex_has_waiters(lock)) +- remove_waiter(lock, &waiter); +- rt_mutex_handle_deadlock(ret, chwalk, &waiter); +- } +- +- /* +- * try_to_take_rt_mutex() sets the waiter bit +- * unconditionally. We might have to fix that up. +- */ +- fixup_rt_mutex_waiters(lock); ++ ret = rt_mutex_slowlock_locked(lock, state, timeout, chwalk, ww_ctx, ++ &waiter); + + raw_spin_unlock_irqrestore(&lock->wait_lock, flags); + +@@ -1328,10 +1886,12 @@ static inline int rt_mutex_slowtrylock(struct rt_mutex *lock) + + /* + * Slow path to release a rt-mutex. +- * Return whether the current task needs to undo a potential priority boosting. ++ * ++ * Return whether the current task needs to call rt_mutex_postunlock(). + */ + static bool __sched rt_mutex_slowunlock(struct rt_mutex *lock, +- struct wake_q_head *wake_q) ++ struct wake_q_head *wake_q, ++ struct wake_q_head *wake_sleeper_q) + { + unsigned long flags; + +@@ -1340,8 +1900,6 @@ static bool __sched rt_mutex_slowunlock(struct rt_mutex *lock, + + debug_rt_mutex_unlock(lock); + +- rt_mutex_deadlock_account_unlock(current); +- + /* + * We must be careful here if the fast path is enabled. If we + * have no waiters queued we cannot set owner to NULL here +@@ -1387,12 +1945,10 @@ static bool __sched rt_mutex_slowunlock(struct rt_mutex *lock, + * + * Queue the next waiter for wakeup once we release the wait_lock. + */ +- mark_wakeup_next_waiter(wake_q, lock); +- ++ mark_wakeup_next_waiter(wake_q, wake_sleeper_q, lock); + raw_spin_unlock_irqrestore(&lock->wait_lock, flags); + +- /* check PI boosting */ +- return true; ++ return true; /* call rt_mutex_postunlock() */ + } + + /* +@@ -1403,63 +1959,85 @@ static bool __sched rt_mutex_slowunlock(struct rt_mutex *lock, + */ + static inline int + rt_mutex_fastlock(struct rt_mutex *lock, int state, ++ struct ww_acquire_ctx *ww_ctx, + int (*slowfn)(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, +- enum rtmutex_chainwalk chwalk)) ++ enum rtmutex_chainwalk chwalk, ++ struct ww_acquire_ctx *ww_ctx)) + { +- if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) { +- rt_mutex_deadlock_account_lock(lock, current); ++ if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) + return 0; +- } else +- return slowfn(lock, state, NULL, RT_MUTEX_MIN_CHAINWALK); ++ ++ return slowfn(lock, state, NULL, RT_MUTEX_MIN_CHAINWALK, ww_ctx); + } + + static inline int + rt_mutex_timed_fastlock(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, + enum rtmutex_chainwalk chwalk, ++ struct ww_acquire_ctx *ww_ctx, + int (*slowfn)(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, +- enum rtmutex_chainwalk chwalk)) ++ enum rtmutex_chainwalk chwalk, ++ struct ww_acquire_ctx *ww_ctx)) + { + if (chwalk == RT_MUTEX_MIN_CHAINWALK && +- likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) { +- rt_mutex_deadlock_account_lock(lock, current); ++ likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) + return 0; +- } else +- return slowfn(lock, state, timeout, chwalk); ++ ++ return slowfn(lock, state, timeout, chwalk, ww_ctx); + } + + static inline int + rt_mutex_fasttrylock(struct rt_mutex *lock, + int (*slowfn)(struct rt_mutex *lock)) + { +- if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) { +- rt_mutex_deadlock_account_lock(lock, current); ++ if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) + return 1; +- } ++ + return slowfn(lock); + } + ++/* ++ * Performs the wakeup of the the top-waiter and re-enables preemption. ++ */ ++void rt_mutex_postunlock(struct wake_q_head *wake_q, ++ struct wake_q_head *wq_sleeper) ++{ ++ wake_up_q(wake_q); ++ wake_up_q_sleeper(wq_sleeper); ++ ++ /* Pairs with preempt_disable() in rt_mutex_slowunlock() */ ++ preempt_enable(); ++} ++ + static inline void + rt_mutex_fastunlock(struct rt_mutex *lock, + bool (*slowfn)(struct rt_mutex *lock, +- struct wake_q_head *wqh)) ++ struct wake_q_head *wqh, ++ struct wake_q_head *wq_sleeper)) + { + WAKE_Q(wake_q); ++ WAKE_Q(wake_sleeper_q); + +- if (likely(rt_mutex_cmpxchg_release(lock, current, NULL))) { +- rt_mutex_deadlock_account_unlock(current); ++ if (likely(rt_mutex_cmpxchg_release(lock, current, NULL))) ++ return; + +- } else { +- bool deboost = slowfn(lock, &wake_q); ++ if (slowfn(lock, &wake_q, &wake_sleeper_q)) ++ rt_mutex_postunlock(&wake_q, &wake_sleeper_q); ++} + +- wake_up_q(&wake_q); ++/** ++ * rt_mutex_lock_state - lock a rt_mutex with a given state ++ * ++ * @lock: The rt_mutex to be locked ++ * @state: The state to set when blocking on the rt_mutex ++ */ ++int __sched rt_mutex_lock_state(struct rt_mutex *lock, int state) ++{ ++ might_sleep(); + +- /* Undo pi boosting if necessary: */ +- if (deboost) +- rt_mutex_adjust_prio(current); +- } ++ return rt_mutex_fastlock(lock, state, NULL, rt_mutex_slowlock); + } + + /** +@@ -1469,15 +2047,13 @@ rt_mutex_fastunlock(struct rt_mutex *lock, + */ + void __sched rt_mutex_lock(struct rt_mutex *lock) + { +- might_sleep(); +- +- rt_mutex_fastlock(lock, TASK_UNINTERRUPTIBLE, rt_mutex_slowlock); ++ rt_mutex_lock_state(lock, TASK_UNINTERRUPTIBLE); + } + EXPORT_SYMBOL_GPL(rt_mutex_lock); + + /** + * rt_mutex_lock_interruptible - lock a rt_mutex interruptible +- * ++ ** + * @lock: the rt_mutex to be locked + * + * Returns: +@@ -1486,23 +2062,32 @@ EXPORT_SYMBOL_GPL(rt_mutex_lock); + */ + int __sched rt_mutex_lock_interruptible(struct rt_mutex *lock) + { +- might_sleep(); +- +- return rt_mutex_fastlock(lock, TASK_INTERRUPTIBLE, rt_mutex_slowlock); ++ return rt_mutex_lock_state(lock, TASK_INTERRUPTIBLE); + } + EXPORT_SYMBOL_GPL(rt_mutex_lock_interruptible); + +-/* +- * Futex variant with full deadlock detection. ++/** ++ * rt_mutex_lock_killable - lock a rt_mutex killable ++ * ++ * @lock: the rt_mutex to be locked ++ * @detect_deadlock: deadlock detection on/off ++ * ++ * Returns: ++ * 0 on success ++ * -EINTR when interrupted by a signal + */ +-int rt_mutex_timed_futex_lock(struct rt_mutex *lock, +- struct hrtimer_sleeper *timeout) ++int __sched rt_mutex_lock_killable(struct rt_mutex *lock) + { +- might_sleep(); ++ return rt_mutex_lock_state(lock, TASK_KILLABLE); ++} ++EXPORT_SYMBOL_GPL(rt_mutex_lock_killable); + +- return rt_mutex_timed_fastlock(lock, TASK_INTERRUPTIBLE, timeout, +- RT_MUTEX_FULL_CHAINWALK, +- rt_mutex_slowlock); ++/* ++ * Futex variant, must not use fastpath. ++ */ ++int __sched rt_mutex_futex_trylock(struct rt_mutex *lock) ++{ ++ return rt_mutex_slowtrylock(lock); + } + + /** +@@ -1525,6 +2110,7 @@ rt_mutex_timed_lock(struct rt_mutex *lock, struct hrtimer_sleeper *timeout) + + return rt_mutex_timed_fastlock(lock, TASK_INTERRUPTIBLE, timeout, + RT_MUTEX_MIN_CHAINWALK, ++ NULL, + rt_mutex_slowlock); + } + EXPORT_SYMBOL_GPL(rt_mutex_timed_lock); +@@ -1542,7 +2128,11 @@ EXPORT_SYMBOL_GPL(rt_mutex_timed_lock); + */ + int __sched rt_mutex_trylock(struct rt_mutex *lock) + { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ if (WARN_ON_ONCE(in_irq() || in_nmi())) ++#else + if (WARN_ON_ONCE(in_irq() || in_nmi() || in_serving_softirq())) ++#endif + return 0; + + return rt_mutex_fasttrylock(lock, rt_mutex_slowtrylock); +@@ -1560,21 +2150,53 @@ void __sched rt_mutex_unlock(struct rt_mutex *lock) + } + EXPORT_SYMBOL_GPL(rt_mutex_unlock); + +-/** +- * rt_mutex_futex_unlock - Futex variant of rt_mutex_unlock +- * @lock: the rt_mutex to be unlocked +- * +- * Returns: true/false indicating whether priority adjustment is +- * required or not. +- */ +-bool __sched rt_mutex_futex_unlock(struct rt_mutex *lock, +- struct wake_q_head *wqh) ++static bool __sched __rt_mutex_unlock_common(struct rt_mutex *lock, ++ struct wake_q_head *wake_q, ++ struct wake_q_head *wq_sleeper) + { +- if (likely(rt_mutex_cmpxchg_release(lock, current, NULL))) { +- rt_mutex_deadlock_account_unlock(current); +- return false; ++ lockdep_assert_held(&lock->wait_lock); ++ ++ debug_rt_mutex_unlock(lock); ++ ++ if (!rt_mutex_has_waiters(lock)) { ++ lock->owner = NULL; ++ return false; /* done */ + } +- return rt_mutex_slowunlock(lock, wqh); ++ ++ /* ++ * We've already deboosted, mark_wakeup_next_waiter() will ++ * retain preempt_disabled when we drop the wait_lock, to ++ * avoid inversion prior to the wakeup. preempt_disable() ++ * therein pairs with rt_mutex_postunlock(). ++ */ ++ mark_wakeup_next_waiter(wake_q, wq_sleeper, lock); ++ ++ return true; /* call postunlock() */ ++} ++ ++/** ++ * Futex variant, that since futex variants do not use the fast-path, can be ++ * simple and will not need to retry. ++ */ ++bool __sched __rt_mutex_futex_unlock(struct rt_mutex *lock, ++ struct wake_q_head *wake_q, ++ struct wake_q_head *wq_sleeper) ++{ ++ return __rt_mutex_unlock_common(lock, wake_q, wq_sleeper); ++} ++ ++void __sched rt_mutex_futex_unlock(struct rt_mutex *lock) ++{ ++ WAKE_Q(wake_q); ++ WAKE_Q(wake_sleeper_q); ++ bool postunlock; ++ ++ raw_spin_lock_irq(&lock->wait_lock); ++ postunlock = __rt_mutex_futex_unlock(lock, &wake_q, &wake_sleeper_q); ++ raw_spin_unlock_irq(&lock->wait_lock); ++ ++ if (postunlock) ++ rt_mutex_postunlock(&wake_q, &wake_sleeper_q); + } + + /** +@@ -1607,13 +2229,12 @@ EXPORT_SYMBOL_GPL(rt_mutex_destroy); + void __rt_mutex_init(struct rt_mutex *lock, const char *name) + { + lock->owner = NULL; +- raw_spin_lock_init(&lock->wait_lock); + lock->waiters = RB_ROOT; + lock->waiters_leftmost = NULL; + + debug_rt_mutex_init(lock, name); + } +-EXPORT_SYMBOL_GPL(__rt_mutex_init); ++EXPORT_SYMBOL(__rt_mutex_init); + + /** + * rt_mutex_init_proxy_locked - initialize and lock a rt_mutex on behalf of a +@@ -1628,10 +2249,9 @@ EXPORT_SYMBOL_GPL(__rt_mutex_init); + void rt_mutex_init_proxy_locked(struct rt_mutex *lock, + struct task_struct *proxy_owner) + { +- __rt_mutex_init(lock, NULL); ++ rt_mutex_init(lock); + debug_rt_mutex_proxy_lock(lock, proxy_owner); + rt_mutex_set_owner(lock, proxy_owner); +- rt_mutex_deadlock_account_lock(lock, proxy_owner); + } + + /** +@@ -1647,7 +2267,66 @@ void rt_mutex_proxy_unlock(struct rt_mutex *lock, + { + debug_rt_mutex_proxy_unlock(lock); + rt_mutex_set_owner(lock, NULL); +- rt_mutex_deadlock_account_unlock(proxy_owner); ++} ++ ++int __rt_mutex_start_proxy_lock(struct rt_mutex *lock, ++ struct rt_mutex_waiter *waiter, ++ struct task_struct *task) ++{ ++ int ret; ++ ++ if (try_to_take_rt_mutex(lock, task, NULL)) ++ return 1; ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ /* ++ * In PREEMPT_RT there's an added race. ++ * If the task, that we are about to requeue, times out, ++ * it can set the PI_WAKEUP_INPROGRESS. This tells the requeue ++ * to skip this task. But right after the task sets ++ * its pi_blocked_on to PI_WAKEUP_INPROGRESS it can then ++ * block on the spin_lock(&hb->lock), which in RT is an rtmutex. ++ * This will replace the PI_WAKEUP_INPROGRESS with the actual ++ * lock that it blocks on. We *must not* place this task ++ * on this proxy lock in that case. ++ * ++ * To prevent this race, we first take the task's pi_lock ++ * and check if it has updated its pi_blocked_on. If it has, ++ * we assume that it woke up and we return -EAGAIN. ++ * Otherwise, we set the task's pi_blocked_on to ++ * PI_REQUEUE_INPROGRESS, so that if the task is waking up ++ * it will know that we are in the process of requeuing it. ++ */ ++ raw_spin_lock(&task->pi_lock); ++ if (task->pi_blocked_on) { ++ raw_spin_unlock(&task->pi_lock); ++ raw_spin_unlock_irq(&lock->wait_lock); ++ return -EAGAIN; ++ } ++ task->pi_blocked_on = PI_REQUEUE_INPROGRESS; ++ raw_spin_unlock(&task->pi_lock); ++#endif ++ ++ /* We enforce deadlock detection for futexes */ ++ ret = task_blocks_on_rt_mutex(lock, waiter, task, ++ RT_MUTEX_FULL_CHAINWALK); ++ ++ if (ret && !rt_mutex_owner(lock)) { ++ /* ++ * Reset the return value. We might have ++ * returned with -EDEADLK and the owner ++ * released the lock while we were walking the ++ * pi chain. Let the waiter sort it out. ++ */ ++ ret = 0; ++ } ++ ++ if (ret && rt_mutex_has_waiters(lock)) ++ remove_waiter(lock, waiter); ++ ++ debug_rt_mutex_print_deadlock(waiter); ++ ++ return ret; + } + + /** +@@ -1670,33 +2349,9 @@ int rt_mutex_start_proxy_lock(struct rt_mutex *lock, + int ret; + + raw_spin_lock_irq(&lock->wait_lock); +- +- if (try_to_take_rt_mutex(lock, task, NULL)) { +- raw_spin_unlock_irq(&lock->wait_lock); +- return 1; +- } +- +- /* We enforce deadlock detection for futexes */ +- ret = task_blocks_on_rt_mutex(lock, waiter, task, +- RT_MUTEX_FULL_CHAINWALK); +- +- if (ret && !rt_mutex_owner(lock)) { +- /* +- * Reset the return value. We might have +- * returned with -EDEADLK and the owner +- * released the lock while we were walking the +- * pi chain. Let the waiter sort it out. +- */ +- ret = 0; +- } +- +- if (unlikely(ret)) +- remove_waiter(lock, waiter); +- ++ ret = __rt_mutex_start_proxy_lock(lock, waiter, task); + raw_spin_unlock_irq(&lock->wait_lock); + +- debug_rt_mutex_print_deadlock(waiter); +- + return ret; + } + +@@ -1721,36 +2376,106 @@ struct task_struct *rt_mutex_next_owner(struct rt_mutex *lock) + } + + /** +- * rt_mutex_finish_proxy_lock() - Complete lock acquisition ++ * rt_mutex_wait_proxy_lock() - Wait for lock acquisition + * @lock: the rt_mutex we were woken on + * @to: the timeout, null if none. hrtimer should already have + * been started. + * @waiter: the pre-initialized rt_mutex_waiter + * +- * Complete the lock acquisition started our behalf by another thread. ++ * Wait for the the lock acquisition started on our behalf by ++ * rt_mutex_start_proxy_lock(). Upon failure, the caller must call ++ * rt_mutex_cleanup_proxy_lock(). + * + * Returns: + * 0 - success + * <0 - error, one of -EINTR, -ETIMEDOUT + * +- * Special API call for PI-futex requeue support ++ * Special API call for PI-futex support + */ +-int rt_mutex_finish_proxy_lock(struct rt_mutex *lock, ++int rt_mutex_wait_proxy_lock(struct rt_mutex *lock, + struct hrtimer_sleeper *to, + struct rt_mutex_waiter *waiter) + { ++ struct task_struct *tsk = current; + int ret; + + raw_spin_lock_irq(&lock->wait_lock); +- +- set_current_state(TASK_INTERRUPTIBLE); +- + /* sleep on the mutex */ +- ret = __rt_mutex_slowlock(lock, TASK_INTERRUPTIBLE, to, waiter); ++ set_current_state(TASK_INTERRUPTIBLE); ++ ret = __rt_mutex_slowlock(lock, TASK_INTERRUPTIBLE, to, waiter, NULL); ++ /* ++ * try_to_take_rt_mutex() sets the waiter bit unconditionally. We might ++ * have to fix that up. ++ */ ++ fixup_rt_mutex_waiters(lock); + +- if (unlikely(ret)) ++ /* ++ * RT has a problem here when the wait got interrupted by a timeout ++ * or a signal. task->pi_blocked_on is still set. The task must ++ * acquire the hash bucket lock when returning from this function. ++ * ++ * If the hash bucket lock is contended then the ++ * BUG_ON(rt_mutex_real_waiter(task->pi_blocked_on)) in ++ * task_blocks_on_rt_mutex() will trigger. This can be avoided by ++ * clearing task->pi_blocked_on which removes the task from the ++ * boosting chain of the rtmutex. That's correct because the task ++ * is not longer blocked on it. ++ */ ++ if (ret) { ++ raw_spin_lock(&tsk->pi_lock); ++ tsk->pi_blocked_on = NULL; ++ raw_spin_unlock(&tsk->pi_lock); ++ } ++ raw_spin_unlock_irq(&lock->wait_lock); ++ ++ return ret; ++} ++ ++/** ++ * rt_mutex_cleanup_proxy_lock() - Cleanup failed lock acquisition ++ * @lock: the rt_mutex we were woken on ++ * @waiter: the pre-initialized rt_mutex_waiter ++ * ++ * Attempt to clean up after a failed rt_mutex_wait_proxy_lock(). ++ * ++ * Unless we acquired the lock; we're still enqueued on the wait-list and can ++ * in fact still be granted ownership until we're removed. Therefore we can ++ * find we are in fact the owner and must disregard the ++ * rt_mutex_wait_proxy_lock() failure. ++ * ++ * Returns: ++ * true - did the cleanup, we done. ++ * false - we acquired the lock after rt_mutex_wait_proxy_lock() returned, ++ * caller should disregards its return value. ++ * ++ * Special API call for PI-futex support ++ */ ++bool rt_mutex_cleanup_proxy_lock(struct rt_mutex *lock, ++ struct rt_mutex_waiter *waiter) ++{ ++ bool cleanup = false; ++ ++ raw_spin_lock_irq(&lock->wait_lock); ++ /* ++ * Do an unconditional try-lock, this deals with the lock stealing ++ * state where __rt_mutex_futex_unlock() -> mark_wakeup_next_waiter() ++ * sets a NULL owner. ++ * ++ * We're not interested in the return value, because the subsequent ++ * test on rt_mutex_owner() will infer that. If the trylock succeeded, ++ * we will own the lock and it will have removed the waiter. If we ++ * failed the trylock, we're still not owner and we need to remove ++ * ourselves. ++ */ ++ try_to_take_rt_mutex(lock, current, waiter); ++ /* ++ * Unless we're the owner; we're still enqueued on the wait_list. ++ * So check if we became owner, if not, take us off the wait_list. ++ */ ++ if (rt_mutex_owner(lock) != current) { + remove_waiter(lock, waiter); +- ++ cleanup = true; ++ } + /* + * try_to_take_rt_mutex() sets the waiter bit unconditionally. We might + * have to fix that up. +@@ -1759,5 +2484,91 @@ int rt_mutex_finish_proxy_lock(struct rt_mutex *lock, + + raw_spin_unlock_irq(&lock->wait_lock); + ++ return cleanup; ++} ++ ++static inline int ++ww_mutex_deadlock_injection(struct ww_mutex *lock, struct ww_acquire_ctx *ctx) ++{ ++#ifdef CONFIG_DEBUG_WW_MUTEX_SLOWPATH ++ unsigned tmp; ++ ++ if (ctx->deadlock_inject_countdown-- == 0) { ++ tmp = ctx->deadlock_inject_interval; ++ if (tmp > UINT_MAX/4) ++ tmp = UINT_MAX; ++ else ++ tmp = tmp*2 + tmp + tmp/2; ++ ++ ctx->deadlock_inject_interval = tmp; ++ ctx->deadlock_inject_countdown = tmp; ++ ctx->contending_lock = lock; ++ ++ ww_mutex_unlock(lock); ++ ++ return -EDEADLK; ++ } ++#endif ++ ++ return 0; ++} ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++int __sched ++__ww_mutex_lock_interruptible(struct ww_mutex *lock, struct ww_acquire_ctx *ww_ctx) ++{ ++ int ret; ++ ++ might_sleep(); ++ ++ mutex_acquire_nest(&lock->base.dep_map, 0, 0, &ww_ctx->dep_map, _RET_IP_); ++ ret = rt_mutex_slowlock(&lock->base.lock, TASK_INTERRUPTIBLE, NULL, 0, ww_ctx); ++ if (ret) ++ mutex_release(&lock->base.dep_map, 1, _RET_IP_); ++ else if (!ret && ww_ctx->acquired > 1) ++ return ww_mutex_deadlock_injection(lock, ww_ctx); ++ + return ret; + } ++EXPORT_SYMBOL_GPL(__ww_mutex_lock_interruptible); ++ ++int __sched ++__ww_mutex_lock(struct ww_mutex *lock, struct ww_acquire_ctx *ww_ctx) ++{ ++ int ret; ++ ++ might_sleep(); ++ ++ mutex_acquire_nest(&lock->base.dep_map, 0, 0, &ww_ctx->dep_map, _RET_IP_); ++ ret = rt_mutex_slowlock(&lock->base.lock, TASK_UNINTERRUPTIBLE, NULL, 0, ww_ctx); ++ if (ret) ++ mutex_release(&lock->base.dep_map, 1, _RET_IP_); ++ else if (!ret && ww_ctx->acquired > 1) ++ return ww_mutex_deadlock_injection(lock, ww_ctx); ++ ++ return ret; ++} ++EXPORT_SYMBOL_GPL(__ww_mutex_lock); ++ ++void __sched ww_mutex_unlock(struct ww_mutex *lock) ++{ ++ int nest = !!lock->ctx; ++ ++ /* ++ * The unlocking fastpath is the 0->1 transition from 'locked' ++ * into 'unlocked' state: ++ */ ++ if (nest) { ++#ifdef CONFIG_DEBUG_MUTEXES ++ DEBUG_LOCKS_WARN_ON(!lock->ctx->acquired); ++#endif ++ if (lock->ctx->acquired > 0) ++ lock->ctx->acquired--; ++ lock->ctx = NULL; ++ } ++ ++ mutex_release(&lock->base.dep_map, nest, _RET_IP_); ++ rt_mutex_unlock(&lock->base.lock); ++} ++EXPORT_SYMBOL(ww_mutex_unlock); ++#endif +diff --git a/kernel/locking/rtmutex.h b/kernel/locking/rtmutex.h +index c4060584c407..6607802efa8b 100644 +--- a/kernel/locking/rtmutex.h ++++ b/kernel/locking/rtmutex.h +@@ -11,8 +11,6 @@ + */ + + #define rt_mutex_deadlock_check(l) (0) +-#define rt_mutex_deadlock_account_lock(m, t) do { } while (0) +-#define rt_mutex_deadlock_account_unlock(l) do { } while (0) + #define debug_rt_mutex_init_waiter(w) do { } while (0) + #define debug_rt_mutex_free_waiter(w) do { } while (0) + #define debug_rt_mutex_lock(l) do { } while (0) +diff --git a/kernel/locking/rtmutex_common.h b/kernel/locking/rtmutex_common.h +index e317e1cbb3eb..64d89d780059 100644 +--- a/kernel/locking/rtmutex_common.h ++++ b/kernel/locking/rtmutex_common.h +@@ -27,12 +27,14 @@ struct rt_mutex_waiter { + struct rb_node pi_tree_entry; + struct task_struct *task; + struct rt_mutex *lock; ++ bool savestate; + #ifdef CONFIG_DEBUG_RT_MUTEXES + unsigned long ip; + struct pid *deadlock_task_pid; + struct rt_mutex *deadlock_lock; + #endif + int prio; ++ u64 deadline; + }; + + /* +@@ -98,21 +100,45 @@ enum rtmutex_chainwalk { + /* + * PI-futex support (proxy locking functions, etc.): + */ ++#define PI_WAKEUP_INPROGRESS ((struct rt_mutex_waiter *) 1) ++#define PI_REQUEUE_INPROGRESS ((struct rt_mutex_waiter *) 2) ++ + extern struct task_struct *rt_mutex_next_owner(struct rt_mutex *lock); + extern void rt_mutex_init_proxy_locked(struct rt_mutex *lock, + struct task_struct *proxy_owner); + extern void rt_mutex_proxy_unlock(struct rt_mutex *lock, + struct task_struct *proxy_owner); ++extern void rt_mutex_init_waiter(struct rt_mutex_waiter *waiter, bool savetate); ++extern int __rt_mutex_start_proxy_lock(struct rt_mutex *lock, ++ struct rt_mutex_waiter *waiter, ++ struct task_struct *task); + extern int rt_mutex_start_proxy_lock(struct rt_mutex *lock, + struct rt_mutex_waiter *waiter, + struct task_struct *task); +-extern int rt_mutex_finish_proxy_lock(struct rt_mutex *lock, +- struct hrtimer_sleeper *to, +- struct rt_mutex_waiter *waiter); +-extern int rt_mutex_timed_futex_lock(struct rt_mutex *l, struct hrtimer_sleeper *to); +-extern bool rt_mutex_futex_unlock(struct rt_mutex *lock, +- struct wake_q_head *wqh); +-extern void rt_mutex_adjust_prio(struct task_struct *task); ++extern int rt_mutex_wait_proxy_lock(struct rt_mutex *lock, ++ struct hrtimer_sleeper *to, ++ struct rt_mutex_waiter *waiter); ++extern bool rt_mutex_cleanup_proxy_lock(struct rt_mutex *lock, ++ struct rt_mutex_waiter *waiter); ++ ++extern int rt_mutex_futex_trylock(struct rt_mutex *l); ++ ++extern void rt_mutex_futex_unlock(struct rt_mutex *lock); ++extern bool __rt_mutex_futex_unlock(struct rt_mutex *lock, ++ struct wake_q_head *wqh, ++ struct wake_q_head *wq_sleeper); ++ ++extern void rt_mutex_postunlock(struct wake_q_head *wake_q, ++ struct wake_q_head *wq_sleeper); ++ ++/* RW semaphore special interface */ ++struct ww_acquire_ctx; ++ ++int __sched rt_mutex_slowlock_locked(struct rt_mutex *lock, int state, ++ struct hrtimer_sleeper *timeout, ++ enum rtmutex_chainwalk chwalk, ++ struct ww_acquire_ctx *ww_ctx, ++ struct rt_mutex_waiter *waiter); + + #ifdef CONFIG_DEBUG_RT_MUTEXES + # include "rtmutex-debug.h" +diff --git a/kernel/locking/rwsem-rt.c b/kernel/locking/rwsem-rt.c +new file mode 100644 +index 000000000000..4a708ffcded6 +--- /dev/null ++++ b/kernel/locking/rwsem-rt.c +@@ -0,0 +1,268 @@ ++/* ++ */ ++#include ++#include ++#include ++ ++#include "rtmutex_common.h" ++ ++/* ++ * RT-specific reader/writer semaphores ++ * ++ * down_write() ++ * 1) Lock sem->rtmutex ++ * 2) Remove the reader BIAS to force readers into the slow path ++ * 3) Wait until all readers have left the critical region ++ * 4) Mark it write locked ++ * ++ * up_write() ++ * 1) Remove the write locked marker ++ * 2) Set the reader BIAS so readers can use the fast path again ++ * 3) Unlock sem->rtmutex to release blocked readers ++ * ++ * down_read() ++ * 1) Try fast path acquisition (reader BIAS is set) ++ * 2) Take sem->rtmutex.wait_lock which protects the writelocked flag ++ * 3) If !writelocked, acquire it for read ++ * 4) If writelocked, block on sem->rtmutex ++ * 5) unlock sem->rtmutex, goto 1) ++ * ++ * up_read() ++ * 1) Try fast path release (reader count != 1) ++ * 2) Wake the writer waiting in down_write()#3 ++ * ++ * down_read()#3 has the consequence, that rw semaphores on RT are not writer ++ * fair, but writers, which should be avoided in RT tasks (think mmap_sem), ++ * are subject to the rtmutex priority/DL inheritance mechanism. ++ * ++ * It's possible to make the rw semaphores writer fair by keeping a list of ++ * active readers. A blocked writer would force all newly incoming readers to ++ * block on the rtmutex, but the rtmutex would have to be proxy locked for one ++ * reader after the other. We can't use multi-reader inheritance because there ++ * is no way to support that with SCHED_DEADLINE. Implementing the one by one ++ * reader boosting/handover mechanism is a major surgery for a very dubious ++ * value. ++ * ++ * The risk of writer starvation is there, but the pathological use cases ++ * which trigger it are not necessarily the typical RT workloads. ++ */ ++ ++void __rwsem_init(struct rw_semaphore *sem, const char *name, ++ struct lock_class_key *key) ++{ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ /* ++ * Make sure we are not reinitializing a held semaphore: ++ */ ++ debug_check_no_locks_freed((void *)sem, sizeof(*sem)); ++ lockdep_init_map(&sem->dep_map, name, key, 0); ++#endif ++ atomic_set(&sem->readers, READER_BIAS); ++} ++EXPORT_SYMBOL(__rwsem_init); ++ ++int __down_read_trylock(struct rw_semaphore *sem) ++{ ++ int r, old; ++ ++ /* ++ * Increment reader count, if sem->readers < 0, i.e. READER_BIAS is ++ * set. ++ */ ++ for (r = atomic_read(&sem->readers); r < 0;) { ++ old = atomic_cmpxchg(&sem->readers, r, r + 1); ++ if (likely(old == r)) ++ return 1; ++ r = old; ++ } ++ return 0; ++} ++ ++void __sched __down_read(struct rw_semaphore *sem) ++{ ++ struct rt_mutex *m = &sem->rtmutex; ++ struct rt_mutex_waiter waiter; ++ ++ if (__down_read_trylock(sem)) ++ return; ++ ++ might_sleep(); ++ raw_spin_lock_irq(&m->wait_lock); ++ /* ++ * Allow readers as long as the writer has not completely ++ * acquired the semaphore for write. ++ */ ++ if (atomic_read(&sem->readers) != WRITER_BIAS) { ++ atomic_inc(&sem->readers); ++ raw_spin_unlock_irq(&m->wait_lock); ++ return; ++ } ++ ++ /* ++ * Call into the slow lock path with the rtmutex->wait_lock ++ * held, so this can't result in the following race: ++ * ++ * Reader1 Reader2 Writer ++ * down_read() ++ * down_write() ++ * rtmutex_lock(m) ++ * swait() ++ * down_read() ++ * unlock(m->wait_lock) ++ * up_read() ++ * swake() ++ * lock(m->wait_lock) ++ * sem->writelocked=true ++ * unlock(m->wait_lock) ++ * ++ * up_write() ++ * sem->writelocked=false ++ * rtmutex_unlock(m) ++ * down_read() ++ * down_write() ++ * rtmutex_lock(m) ++ * swait() ++ * rtmutex_lock(m) ++ * ++ * That would put Reader1 behind the writer waiting on ++ * Reader2 to call up_read() which might be unbound. ++ */ ++ rt_mutex_init_waiter(&waiter, false); ++ rt_mutex_slowlock_locked(m, TASK_UNINTERRUPTIBLE, NULL, ++ RT_MUTEX_MIN_CHAINWALK, NULL, ++ &waiter); ++ /* ++ * The slowlock() above is guaranteed to return with the rtmutex is ++ * now held, so there can't be a writer active. Increment the reader ++ * count and immediately drop the rtmutex again. ++ */ ++ atomic_inc(&sem->readers); ++ raw_spin_unlock_irq(&m->wait_lock); ++ rt_mutex_unlock(m); ++ ++ debug_rt_mutex_free_waiter(&waiter); ++} ++ ++void __up_read(struct rw_semaphore *sem) ++{ ++ struct rt_mutex *m = &sem->rtmutex; ++ struct task_struct *tsk; ++ ++ /* ++ * sem->readers can only hit 0 when a writer is waiting for the ++ * active readers to leave the critical region. ++ */ ++ if (!atomic_dec_and_test(&sem->readers)) ++ return; ++ ++ might_sleep(); ++ raw_spin_lock_irq(&m->wait_lock); ++ /* ++ * Wake the writer, i.e. the rtmutex owner. It might release the ++ * rtmutex concurrently in the fast path (due to a signal), but to ++ * clean up the rwsem it needs to acquire m->wait_lock. The worst ++ * case which can happen is a spurious wakeup. ++ */ ++ tsk = rt_mutex_owner(m); ++ if (tsk) ++ wake_up_process(tsk); ++ ++ raw_spin_unlock_irq(&m->wait_lock); ++} ++ ++static void __up_write_unlock(struct rw_semaphore *sem, int bias, ++ unsigned long flags) ++{ ++ struct rt_mutex *m = &sem->rtmutex; ++ ++ atomic_add(READER_BIAS - bias, &sem->readers); ++ raw_spin_unlock_irqrestore(&m->wait_lock, flags); ++ rt_mutex_unlock(m); ++} ++ ++static int __sched __down_write_common(struct rw_semaphore *sem, int state) ++{ ++ struct rt_mutex *m = &sem->rtmutex; ++ unsigned long flags; ++ ++ /* Take the rtmutex as a first step */ ++ if (rt_mutex_lock_state(m, state)) ++ return -EINTR; ++ ++ /* Force readers into slow path */ ++ atomic_sub(READER_BIAS, &sem->readers); ++ might_sleep(); ++ ++ set_current_state(state); ++ for (;;) { ++ raw_spin_lock_irqsave(&m->wait_lock, flags); ++ /* Have all readers left the critical region? */ ++ if (!atomic_read(&sem->readers)) { ++ atomic_set(&sem->readers, WRITER_BIAS); ++ __set_current_state(TASK_RUNNING); ++ raw_spin_unlock_irqrestore(&m->wait_lock, flags); ++ return 0; ++ } ++ ++ if (signal_pending_state(state, current)) { ++ __set_current_state(TASK_RUNNING); ++ __up_write_unlock(sem, 0, flags); ++ return -EINTR; ++ } ++ raw_spin_unlock_irqrestore(&m->wait_lock, flags); ++ ++ if (atomic_read(&sem->readers) != 0) { ++ schedule(); ++ set_current_state(state); ++ } ++ } ++} ++ ++void __sched __down_write(struct rw_semaphore *sem) ++{ ++ __down_write_common(sem, TASK_UNINTERRUPTIBLE); ++} ++ ++int __sched __down_write_killable(struct rw_semaphore *sem) ++{ ++ return __down_write_common(sem, TASK_KILLABLE); ++} ++ ++int __down_write_trylock(struct rw_semaphore *sem) ++{ ++ struct rt_mutex *m = &sem->rtmutex; ++ unsigned long flags; ++ ++ if (!rt_mutex_trylock(m)) ++ return 0; ++ ++ atomic_sub(READER_BIAS, &sem->readers); ++ ++ raw_spin_lock_irqsave(&m->wait_lock, flags); ++ if (!atomic_read(&sem->readers)) { ++ atomic_set(&sem->readers, WRITER_BIAS); ++ raw_spin_unlock_irqrestore(&m->wait_lock, flags); ++ return 1; ++ } ++ __up_write_unlock(sem, 0, flags); ++ return 0; ++} ++ ++void __up_write(struct rw_semaphore *sem) ++{ ++ struct rt_mutex *m = &sem->rtmutex; ++ unsigned long flags; ++ ++ raw_spin_lock_irqsave(&m->wait_lock, flags); ++ __up_write_unlock(sem, WRITER_BIAS, flags); ++} ++ ++void __downgrade_write(struct rw_semaphore *sem) ++{ ++ struct rt_mutex *m = &sem->rtmutex; ++ unsigned long flags; ++ ++ raw_spin_lock_irqsave(&m->wait_lock, flags); ++ /* Release it and account current as reader */ ++ __up_write_unlock(sem, WRITER_BIAS - 1, flags); ++} +diff --git a/kernel/locking/spinlock.c b/kernel/locking/spinlock.c +index db3ccb1dd614..909779647bd1 100644 +--- a/kernel/locking/spinlock.c ++++ b/kernel/locking/spinlock.c +@@ -124,8 +124,11 @@ void __lockfunc __raw_##op##_lock_bh(locktype##_t *lock) \ + * __[spin|read|write]_lock_bh() + */ + BUILD_LOCK_OPS(spin, raw_spinlock); ++ ++#ifndef CONFIG_PREEMPT_RT_FULL + BUILD_LOCK_OPS(read, rwlock); + BUILD_LOCK_OPS(write, rwlock); ++#endif + + #endif + +@@ -209,6 +212,8 @@ void __lockfunc _raw_spin_unlock_bh(raw_spinlock_t *lock) + EXPORT_SYMBOL(_raw_spin_unlock_bh); + #endif + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + #ifndef CONFIG_INLINE_READ_TRYLOCK + int __lockfunc _raw_read_trylock(rwlock_t *lock) + { +@@ -353,6 +358,8 @@ void __lockfunc _raw_write_unlock_bh(rwlock_t *lock) + EXPORT_SYMBOL(_raw_write_unlock_bh); + #endif + ++#endif /* !PREEMPT_RT_FULL */ ++ + #ifdef CONFIG_DEBUG_LOCK_ALLOC + + void __lockfunc _raw_spin_lock_nested(raw_spinlock_t *lock, int subclass) +diff --git a/kernel/locking/spinlock_debug.c b/kernel/locking/spinlock_debug.c +index 0374a596cffa..94970338d518 100644 +--- a/kernel/locking/spinlock_debug.c ++++ b/kernel/locking/spinlock_debug.c +@@ -31,6 +31,7 @@ void __raw_spin_lock_init(raw_spinlock_t *lock, const char *name, + + EXPORT_SYMBOL(__raw_spin_lock_init); + ++#ifndef CONFIG_PREEMPT_RT_FULL + void __rwlock_init(rwlock_t *lock, const char *name, + struct lock_class_key *key) + { +@@ -48,6 +49,7 @@ void __rwlock_init(rwlock_t *lock, const char *name, + } + + EXPORT_SYMBOL(__rwlock_init); ++#endif + + static void spin_dump(raw_spinlock_t *lock, const char *msg) + { +@@ -159,6 +161,7 @@ void do_raw_spin_unlock(raw_spinlock_t *lock) + arch_spin_unlock(&lock->raw_lock); + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + static void rwlock_bug(rwlock_t *lock, const char *msg) + { + if (!debug_locks_off()) +@@ -300,3 +303,5 @@ void do_raw_write_unlock(rwlock_t *lock) + debug_write_unlock(lock); + arch_write_unlock(&lock->raw_lock); + } ++ ++#endif +diff --git a/kernel/module.c b/kernel/module.c +index 0e54d5bf0097..f27764fbfa24 100644 +--- a/kernel/module.c ++++ b/kernel/module.c +@@ -660,16 +660,7 @@ static void percpu_modcopy(struct module *mod, + memcpy(per_cpu_ptr(mod->percpu, cpu), from, size); + } + +-/** +- * is_module_percpu_address - test whether address is from module static percpu +- * @addr: address to test +- * +- * Test whether @addr belongs to module static percpu area. +- * +- * RETURNS: +- * %true if @addr is from module static percpu area +- */ +-bool is_module_percpu_address(unsigned long addr) ++bool __is_module_percpu_address(unsigned long addr, unsigned long *can_addr) + { + struct module *mod; + unsigned int cpu; +@@ -683,9 +674,15 @@ bool is_module_percpu_address(unsigned long addr) + continue; + for_each_possible_cpu(cpu) { + void *start = per_cpu_ptr(mod->percpu, cpu); ++ void *va = (void *)addr; + +- if ((void *)addr >= start && +- (void *)addr < start + mod->percpu_size) { ++ if (va >= start && va < start + mod->percpu_size) { ++ if (can_addr) { ++ *can_addr = (unsigned long) (va - start); ++ *can_addr += (unsigned long) ++ per_cpu_ptr(mod->percpu, ++ get_boot_cpu_id()); ++ } + preempt_enable(); + return true; + } +@@ -696,6 +693,20 @@ bool is_module_percpu_address(unsigned long addr) + return false; + } + ++/** ++ * is_module_percpu_address - test whether address is from module static percpu ++ * @addr: address to test ++ * ++ * Test whether @addr belongs to module static percpu area. ++ * ++ * RETURNS: ++ * %true if @addr is from module static percpu area ++ */ ++bool is_module_percpu_address(unsigned long addr) ++{ ++ return __is_module_percpu_address(addr, NULL); ++} ++ + #else /* ... !CONFIG_SMP */ + + static inline void __percpu *mod_percpu(struct module *mod) +@@ -727,6 +738,11 @@ bool is_module_percpu_address(unsigned long addr) + return false; + } + ++bool __is_module_percpu_address(unsigned long addr, unsigned long *can_addr) ++{ ++ return false; ++} ++ + #endif /* CONFIG_SMP */ + + #define MODINFO_ATTR(field) \ +diff --git a/kernel/panic.c b/kernel/panic.c +index e6480e20379e..7e9c1918a94e 100644 +--- a/kernel/panic.c ++++ b/kernel/panic.c +@@ -482,9 +482,11 @@ static u64 oops_id; + + static int init_oops_id(void) + { ++#ifndef CONFIG_PREEMPT_RT_FULL + if (!oops_id) + get_random_bytes(&oops_id, sizeof(oops_id)); + else ++#endif + oops_id++; + + return 0; +diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c +index b26dbc48c75b..968255f27a33 100644 +--- a/kernel/power/hibernate.c ++++ b/kernel/power/hibernate.c +@@ -286,6 +286,8 @@ static int create_image(int platform_mode) + + local_irq_disable(); + ++ system_state = SYSTEM_SUSPEND; ++ + error = syscore_suspend(); + if (error) { + printk(KERN_ERR "PM: Some system devices failed to power down, " +@@ -317,6 +319,7 @@ static int create_image(int platform_mode) + syscore_resume(); + + Enable_irqs: ++ system_state = SYSTEM_RUNNING; + local_irq_enable(); + + Enable_cpus: +@@ -446,6 +449,7 @@ static int resume_target_kernel(bool platform_mode) + goto Enable_cpus; + + local_irq_disable(); ++ system_state = SYSTEM_SUSPEND; + + error = syscore_suspend(); + if (error) +@@ -479,6 +483,7 @@ static int resume_target_kernel(bool platform_mode) + syscore_resume(); + + Enable_irqs: ++ system_state = SYSTEM_RUNNING; + local_irq_enable(); + + Enable_cpus: +@@ -564,6 +569,7 @@ int hibernation_platform_enter(void) + goto Enable_cpus; + + local_irq_disable(); ++ system_state = SYSTEM_SUSPEND; + syscore_suspend(); + if (pm_wakeup_pending()) { + error = -EAGAIN; +@@ -576,6 +582,7 @@ int hibernation_platform_enter(void) + + Power_up: + syscore_resume(); ++ system_state = SYSTEM_RUNNING; + local_irq_enable(); + + Enable_cpus: +@@ -676,6 +683,10 @@ static int load_image_and_restore(void) + return error; + } + ++#ifndef CONFIG_SUSPEND ++bool pm_in_action; ++#endif ++ + /** + * hibernate - Carry out system hibernation, including saving the image. + */ +@@ -689,6 +700,8 @@ int hibernate(void) + return -EPERM; + } + ++ pm_in_action = true; ++ + lock_system_sleep(); + /* The snapshot device should not be opened while we're running */ + if (!atomic_add_unless(&snapshot_device_available, -1, 0)) { +@@ -766,6 +779,7 @@ int hibernate(void) + atomic_inc(&snapshot_device_available); + Unlock: + unlock_system_sleep(); ++ pm_in_action = false; + return error; + } + +diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c +index 6ccb08f57fcb..c8cbb5ed2fe3 100644 +--- a/kernel/power/suspend.c ++++ b/kernel/power/suspend.c +@@ -369,6 +369,8 @@ static int suspend_enter(suspend_state_t state, bool *wakeup) + arch_suspend_disable_irqs(); + BUG_ON(!irqs_disabled()); + ++ system_state = SYSTEM_SUSPEND; ++ + error = syscore_suspend(); + if (!error) { + *wakeup = pm_wakeup_pending(); +@@ -385,6 +387,8 @@ static int suspend_enter(suspend_state_t state, bool *wakeup) + syscore_resume(); + } + ++ system_state = SYSTEM_RUNNING; ++ + arch_suspend_enable_irqs(); + BUG_ON(irqs_disabled()); + +@@ -527,6 +531,8 @@ static int enter_state(suspend_state_t state) + return error; + } + ++bool pm_in_action; ++ + /** + * pm_suspend - Externally visible function for suspending the system. + * @state: System sleep state to enter. +@@ -541,6 +547,8 @@ int pm_suspend(suspend_state_t state) + if (state <= PM_SUSPEND_ON || state >= PM_SUSPEND_MAX) + return -EINVAL; + ++ pm_in_action = true; ++ + error = enter_state(state); + if (error) { + suspend_stats.fail++; +@@ -548,6 +556,7 @@ int pm_suspend(suspend_state_t state) + } else { + suspend_stats.success++; + } ++ pm_in_action = false; + return error; + } + EXPORT_SYMBOL(pm_suspend); +diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c +index 9c5b231684d0..cf15bdb6855b 100644 +--- a/kernel/printk/printk.c ++++ b/kernel/printk/printk.c +@@ -351,6 +351,65 @@ __packed __aligned(4) + */ + DEFINE_RAW_SPINLOCK(logbuf_lock); + ++#ifdef CONFIG_EARLY_PRINTK ++struct console *early_console; ++ ++static void early_vprintk(const char *fmt, va_list ap) ++{ ++ if (early_console) { ++ char buf[512]; ++ int n = vscnprintf(buf, sizeof(buf), fmt, ap); ++ ++ early_console->write(early_console, buf, n); ++ } ++} ++ ++asmlinkage void early_printk(const char *fmt, ...) ++{ ++ va_list ap; ++ ++ va_start(ap, fmt); ++ early_vprintk(fmt, ap); ++ va_end(ap); ++} ++ ++/* ++ * This is independent of any log levels - a global ++ * kill switch that turns off all of printk. ++ * ++ * Used by the NMI watchdog if early-printk is enabled. ++ */ ++static bool __read_mostly printk_killswitch; ++ ++static int __init force_early_printk_setup(char *str) ++{ ++ printk_killswitch = true; ++ return 0; ++} ++early_param("force_early_printk", force_early_printk_setup); ++ ++void printk_kill(void) ++{ ++ printk_killswitch = true; ++} ++ ++#ifdef CONFIG_PRINTK ++static int forced_early_printk(const char *fmt, va_list ap) ++{ ++ if (!printk_killswitch) ++ return 0; ++ early_vprintk(fmt, ap); ++ return 1; ++} ++#endif ++ ++#else ++static inline int forced_early_printk(const char *fmt, va_list ap) ++{ ++ return 0; ++} ++#endif ++ + #ifdef CONFIG_PRINTK + DECLARE_WAIT_QUEUE_HEAD(log_wait); + /* the next printk record to read by syslog(READ) or /proc/kmsg */ +@@ -1337,6 +1396,7 @@ static int syslog_print_all(char __user *buf, int size, bool clear) + { + char *text; + int len = 0; ++ int attempts = 0; + + text = kmalloc(LOG_LINE_MAX + PREFIX_MAX, GFP_KERNEL); + if (!text) +@@ -1348,6 +1408,14 @@ static int syslog_print_all(char __user *buf, int size, bool clear) + u64 seq; + u32 idx; + enum log_flags prev; ++ int num_msg; ++try_again: ++ attempts++; ++ if (attempts > 10) { ++ len = -EBUSY; ++ goto out; ++ } ++ num_msg = 0; + + /* + * Find first record that fits, including all following records, +@@ -1363,6 +1431,14 @@ static int syslog_print_all(char __user *buf, int size, bool clear) + prev = msg->flags; + idx = log_next(idx); + seq++; ++ num_msg++; ++ if (num_msg > 5) { ++ num_msg = 0; ++ raw_spin_unlock_irq(&logbuf_lock); ++ raw_spin_lock_irq(&logbuf_lock); ++ if (clear_seq < log_first_seq) ++ goto try_again; ++ } + } + + /* move first record forward until length fits into the buffer */ +@@ -1376,6 +1452,14 @@ static int syslog_print_all(char __user *buf, int size, bool clear) + prev = msg->flags; + idx = log_next(idx); + seq++; ++ num_msg++; ++ if (num_msg > 5) { ++ num_msg = 0; ++ raw_spin_unlock_irq(&logbuf_lock); ++ raw_spin_lock_irq(&logbuf_lock); ++ if (clear_seq < log_first_seq) ++ goto try_again; ++ } + } + + /* last message fitting into this dump */ +@@ -1416,6 +1500,7 @@ static int syslog_print_all(char __user *buf, int size, bool clear) + clear_seq = log_next_seq; + clear_idx = log_next_idx; + } ++out: + raw_spin_unlock_irq(&logbuf_lock); + + kfree(text); +@@ -1569,6 +1654,12 @@ static void call_console_drivers(int level, + if (!console_drivers) + return; + ++ if (IS_ENABLED(CONFIG_PREEMPT_RT_BASE)) { ++ if (in_irq() || in_nmi()) ++ return; ++ } ++ ++ migrate_disable(); + for_each_console(con) { + if (exclusive_console && con != exclusive_console) + continue; +@@ -1584,6 +1675,7 @@ static void call_console_drivers(int level, + else + con->write(con, text, len); + } ++ migrate_enable(); + } + + /* +@@ -1781,6 +1873,13 @@ asmlinkage int vprintk_emit(int facility, int level, + /* cpu currently holding logbuf_lock in this function */ + static unsigned int logbuf_cpu = UINT_MAX; + ++ /* ++ * Fall back to early_printk if a debugging subsystem has ++ * killed printk output ++ */ ++ if (unlikely(forced_early_printk(fmt, args))) ++ return 1; ++ + if (level == LOGLEVEL_SCHED) { + level = LOGLEVEL_DEFAULT; + in_sched = true; +@@ -1885,13 +1984,23 @@ asmlinkage int vprintk_emit(int facility, int level, + + /* If called from the scheduler, we can not call up(). */ + if (!in_sched) { ++ int may_trylock = 1; ++ + lockdep_off(); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ /* ++ * we can't take a sleeping lock with IRQs or preeption disabled ++ * so we can't print in these contexts ++ */ ++ if (!(preempt_count() == 0 && !irqs_disabled())) ++ may_trylock = 0; ++#endif + /* + * Try to acquire and then immediately release the console + * semaphore. The release will print out buffers and wake up + * /dev/kmsg and syslog() users. + */ +- if (console_trylock()) ++ if (may_trylock && console_trylock()) + console_unlock(); + lockdep_on(); + } +@@ -2014,26 +2123,6 @@ DEFINE_PER_CPU(printk_func_t, printk_func); + + #endif /* CONFIG_PRINTK */ + +-#ifdef CONFIG_EARLY_PRINTK +-struct console *early_console; +- +-asmlinkage __visible void early_printk(const char *fmt, ...) +-{ +- va_list ap; +- char buf[512]; +- int n; +- +- if (!early_console) +- return; +- +- va_start(ap, fmt); +- n = vscnprintf(buf, sizeof(buf), fmt, ap); +- va_end(ap); +- +- early_console->write(early_console, buf, n); +-} +-#endif +- + static int __add_preferred_console(char *name, int idx, char *options, + char *brl_options) + { +@@ -2303,11 +2392,16 @@ static void console_cont_flush(char *text, size_t size) + goto out; + + len = cont_print_text(text, size); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ raw_spin_unlock_irqrestore(&logbuf_lock, flags); ++ call_console_drivers(cont.level, NULL, 0, text, len); ++#else + raw_spin_unlock(&logbuf_lock); + stop_critical_timings(); + call_console_drivers(cont.level, NULL, 0, text, len); + start_critical_timings(); + local_irq_restore(flags); ++#endif + return; + out: + raw_spin_unlock_irqrestore(&logbuf_lock, flags); +@@ -2431,13 +2525,17 @@ void console_unlock(void) + console_idx = log_next(console_idx); + console_seq++; + console_prev = msg->flags; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ raw_spin_unlock_irqrestore(&logbuf_lock, flags); ++ call_console_drivers(level, ext_text, ext_len, text, len); ++#else + raw_spin_unlock(&logbuf_lock); + + stop_critical_timings(); /* don't trace print latency */ + call_console_drivers(level, ext_text, ext_len, text, len); + start_critical_timings(); + local_irq_restore(flags); +- ++#endif + if (do_cond_resched) + cond_resched(); + } +@@ -2489,6 +2587,11 @@ void console_unblank(void) + { + struct console *c; + ++ if (IS_ENABLED(CONFIG_PREEMPT_RT_BASE)) { ++ if (in_irq() || in_nmi()) ++ return; ++ } ++ + /* + * console_unblank can no longer be called in interrupt context unless + * oops_in_progress is set to 1.. +diff --git a/kernel/ptrace.c b/kernel/ptrace.c +index a5caecef88be..61e7c5e2183c 100644 +--- a/kernel/ptrace.c ++++ b/kernel/ptrace.c +@@ -166,7 +166,14 @@ static bool ptrace_freeze_traced(struct task_struct *task) + + spin_lock_irq(&task->sighand->siglock); + if (task_is_traced(task) && !__fatal_signal_pending(task)) { +- task->state = __TASK_TRACED; ++ unsigned long flags; ++ ++ raw_spin_lock_irqsave(&task->pi_lock, flags); ++ if (task->state & __TASK_TRACED) ++ task->state = __TASK_TRACED; ++ else ++ task->saved_state = __TASK_TRACED; ++ raw_spin_unlock_irqrestore(&task->pi_lock, flags); + ret = true; + } + spin_unlock_irq(&task->sighand->siglock); +diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c +index bf08fee53dc7..eeb8ce4ad7b6 100644 +--- a/kernel/rcu/rcutorture.c ++++ b/kernel/rcu/rcutorture.c +@@ -404,6 +404,7 @@ static struct rcu_torture_ops rcu_ops = { + .name = "rcu" + }; + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * Definitions for rcu_bh torture testing. + */ +@@ -443,6 +444,12 @@ static struct rcu_torture_ops rcu_bh_ops = { + .name = "rcu_bh" + }; + ++#else ++static struct rcu_torture_ops rcu_bh_ops = { ++ .ttype = INVALID_RCU_FLAVOR, ++}; ++#endif ++ + /* + * Don't even think about trying any of these in real life!!! + * The names includes "busted", and they really means it! +diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c +index 10f62c6f48e7..dbee19478f09 100644 +--- a/kernel/rcu/tree.c ++++ b/kernel/rcu/tree.c +@@ -55,6 +55,11 @@ + #include + #include + #include ++#include ++#include ++#include ++#include ++#include "../time/tick-internal.h" + + #include "tree.h" + #include "rcu.h" +@@ -260,6 +265,19 @@ void rcu_sched_qs(void) + this_cpu_ptr(&rcu_sched_data), true); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++static void rcu_preempt_qs(void); ++ ++void rcu_bh_qs(void) ++{ ++ unsigned long flags; ++ ++ /* Callers to this function, rcu_preempt_qs(), must disable irqs. */ ++ local_irq_save(flags); ++ rcu_preempt_qs(); ++ local_irq_restore(flags); ++} ++#else + void rcu_bh_qs(void) + { + if (__this_cpu_read(rcu_bh_data.cpu_no_qs.s)) { +@@ -269,6 +287,7 @@ void rcu_bh_qs(void) + __this_cpu_write(rcu_bh_data.cpu_no_qs.b.norm, false); + } + } ++#endif + + static DEFINE_PER_CPU(int, rcu_sched_qs_mask); + +@@ -449,11 +468,13 @@ EXPORT_SYMBOL_GPL(rcu_batches_started_sched); + /* + * Return the number of RCU BH batches started thus far for debug & stats. + */ ++#ifndef CONFIG_PREEMPT_RT_FULL + unsigned long rcu_batches_started_bh(void) + { + return rcu_bh_state.gpnum; + } + EXPORT_SYMBOL_GPL(rcu_batches_started_bh); ++#endif + + /* + * Return the number of RCU batches completed thus far for debug & stats. +@@ -473,6 +494,7 @@ unsigned long rcu_batches_completed_sched(void) + } + EXPORT_SYMBOL_GPL(rcu_batches_completed_sched); + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * Return the number of RCU BH batches completed thus far for debug & stats. + */ +@@ -481,6 +503,7 @@ unsigned long rcu_batches_completed_bh(void) + return rcu_bh_state.completed; + } + EXPORT_SYMBOL_GPL(rcu_batches_completed_bh); ++#endif + + /* + * Return the number of RCU expedited batches completed thus far for +@@ -504,6 +527,7 @@ unsigned long rcu_exp_batches_completed_sched(void) + } + EXPORT_SYMBOL_GPL(rcu_exp_batches_completed_sched); + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * Force a quiescent state. + */ +@@ -522,6 +546,13 @@ void rcu_bh_force_quiescent_state(void) + } + EXPORT_SYMBOL_GPL(rcu_bh_force_quiescent_state); + ++#else ++void rcu_force_quiescent_state(void) ++{ ++} ++EXPORT_SYMBOL_GPL(rcu_force_quiescent_state); ++#endif ++ + /* + * Force a quiescent state for RCU-sched. + */ +@@ -572,9 +603,11 @@ void rcutorture_get_gp_data(enum rcutorture_type test_type, int *flags, + case RCU_FLAVOR: + rsp = rcu_state_p; + break; ++#ifndef CONFIG_PREEMPT_RT_FULL + case RCU_BH_FLAVOR: + rsp = &rcu_bh_state; + break; ++#endif + case RCU_SCHED_FLAVOR: + rsp = &rcu_sched_state; + break; +@@ -3016,18 +3049,17 @@ __rcu_process_callbacks(struct rcu_state *rsp) + /* + * Do RCU core processing for the current CPU. + */ +-static __latent_entropy void rcu_process_callbacks(struct softirq_action *unused) ++static __latent_entropy void rcu_process_callbacks(void) + { + struct rcu_state *rsp; + + if (cpu_is_offline(smp_processor_id())) + return; +- trace_rcu_utilization(TPS("Start RCU core")); + for_each_rcu_flavor(rsp) + __rcu_process_callbacks(rsp); +- trace_rcu_utilization(TPS("End RCU core")); + } + ++static DEFINE_PER_CPU(struct task_struct *, rcu_cpu_kthread_task); + /* + * Schedule RCU callback invocation. If the specified type of RCU + * does not support RCU priority boosting, just do a direct call, +@@ -3039,19 +3071,106 @@ static void invoke_rcu_callbacks(struct rcu_state *rsp, struct rcu_data *rdp) + { + if (unlikely(!READ_ONCE(rcu_scheduler_fully_active))) + return; +- if (likely(!rsp->boost)) { +- rcu_do_batch(rsp, rdp); +- return; +- } +- invoke_rcu_callbacks_kthread(); ++ rcu_do_batch(rsp, rdp); + } + ++static void rcu_wake_cond(struct task_struct *t, int status) ++{ ++ /* ++ * If the thread is yielding, only wake it when this ++ * is invoked from idle ++ */ ++ if (t && (status != RCU_KTHREAD_YIELDING || is_idle_task(current))) ++ wake_up_process(t); ++} ++ ++/* ++ * Wake up this CPU's rcuc kthread to do RCU core processing. ++ */ + static void invoke_rcu_core(void) + { +- if (cpu_online(smp_processor_id())) +- raise_softirq(RCU_SOFTIRQ); ++ unsigned long flags; ++ struct task_struct *t; ++ ++ if (!cpu_online(smp_processor_id())) ++ return; ++ local_irq_save(flags); ++ __this_cpu_write(rcu_cpu_has_work, 1); ++ t = __this_cpu_read(rcu_cpu_kthread_task); ++ if (t != NULL && current != t) ++ rcu_wake_cond(t, __this_cpu_read(rcu_cpu_kthread_status)); ++ local_irq_restore(flags); + } + ++static void rcu_cpu_kthread_park(unsigned int cpu) ++{ ++ per_cpu(rcu_cpu_kthread_status, cpu) = RCU_KTHREAD_OFFCPU; ++} ++ ++static int rcu_cpu_kthread_should_run(unsigned int cpu) ++{ ++ return __this_cpu_read(rcu_cpu_has_work); ++} ++ ++/* ++ * Per-CPU kernel thread that invokes RCU callbacks. This replaces the ++ * RCU softirq used in flavors and configurations of RCU that do not ++ * support RCU priority boosting. ++ */ ++static void rcu_cpu_kthread(unsigned int cpu) ++{ ++ unsigned int *statusp = this_cpu_ptr(&rcu_cpu_kthread_status); ++ char work, *workp = this_cpu_ptr(&rcu_cpu_has_work); ++ int spincnt; ++ ++ for (spincnt = 0; spincnt < 10; spincnt++) { ++ trace_rcu_utilization(TPS("Start CPU kthread@rcu_wait")); ++ local_bh_disable(); ++ *statusp = RCU_KTHREAD_RUNNING; ++ this_cpu_inc(rcu_cpu_kthread_loops); ++ local_irq_disable(); ++ work = *workp; ++ *workp = 0; ++ local_irq_enable(); ++ if (work) ++ rcu_process_callbacks(); ++ local_bh_enable(); ++ if (*workp == 0) { ++ trace_rcu_utilization(TPS("End CPU kthread@rcu_wait")); ++ *statusp = RCU_KTHREAD_WAITING; ++ return; ++ } ++ } ++ *statusp = RCU_KTHREAD_YIELDING; ++ trace_rcu_utilization(TPS("Start CPU kthread@rcu_yield")); ++ schedule_timeout_interruptible(2); ++ trace_rcu_utilization(TPS("End CPU kthread@rcu_yield")); ++ *statusp = RCU_KTHREAD_WAITING; ++} ++ ++static struct smp_hotplug_thread rcu_cpu_thread_spec = { ++ .store = &rcu_cpu_kthread_task, ++ .thread_should_run = rcu_cpu_kthread_should_run, ++ .thread_fn = rcu_cpu_kthread, ++ .thread_comm = "rcuc/%u", ++ .setup = rcu_cpu_kthread_setup, ++ .park = rcu_cpu_kthread_park, ++}; ++ ++/* ++ * Spawn per-CPU RCU core processing kthreads. ++ */ ++static int __init rcu_spawn_core_kthreads(void) ++{ ++ int cpu; ++ ++ for_each_possible_cpu(cpu) ++ per_cpu(rcu_cpu_has_work, cpu) = 0; ++ BUG_ON(smpboot_register_percpu_thread(&rcu_cpu_thread_spec)); ++ return 0; ++} ++early_initcall(rcu_spawn_core_kthreads); ++ + /* + * Handle any core-RCU processing required by a call_rcu() invocation. + */ +@@ -3195,6 +3314,7 @@ void call_rcu_sched(struct rcu_head *head, rcu_callback_t func) + } + EXPORT_SYMBOL_GPL(call_rcu_sched); + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * Queue an RCU callback for invocation after a quicker grace period. + */ +@@ -3203,6 +3323,7 @@ void call_rcu_bh(struct rcu_head *head, rcu_callback_t func) + __call_rcu(head, func, &rcu_bh_state, -1, 0); + } + EXPORT_SYMBOL_GPL(call_rcu_bh); ++#endif + + /* + * Queue an RCU callback for lazy invocation after a grace period. +@@ -3294,6 +3415,7 @@ void synchronize_sched(void) + } + EXPORT_SYMBOL_GPL(synchronize_sched); + ++#ifndef CONFIG_PREEMPT_RT_FULL + /** + * synchronize_rcu_bh - wait until an rcu_bh grace period has elapsed. + * +@@ -3320,6 +3442,7 @@ void synchronize_rcu_bh(void) + wait_rcu_gp(call_rcu_bh); + } + EXPORT_SYMBOL_GPL(synchronize_rcu_bh); ++#endif + + /** + * get_state_synchronize_rcu - Snapshot current RCU state +@@ -3698,6 +3821,7 @@ static void _rcu_barrier(struct rcu_state *rsp) + mutex_unlock(&rsp->barrier_mutex); + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + /** + * rcu_barrier_bh - Wait until all in-flight call_rcu_bh() callbacks complete. + */ +@@ -3706,6 +3830,7 @@ void rcu_barrier_bh(void) + _rcu_barrier(&rcu_bh_state); + } + EXPORT_SYMBOL_GPL(rcu_barrier_bh); ++#endif + + /** + * rcu_barrier_sched - Wait for in-flight call_rcu_sched() callbacks. +@@ -4227,12 +4352,13 @@ void __init rcu_init(void) + + rcu_bootup_announce(); + rcu_init_geometry(); ++#ifndef CONFIG_PREEMPT_RT_FULL + rcu_init_one(&rcu_bh_state); ++#endif + rcu_init_one(&rcu_sched_state); + if (dump_tree) + rcu_dump_rcu_node_tree(&rcu_sched_state); + __rcu_init_preempt(); +- open_softirq(RCU_SOFTIRQ, rcu_process_callbacks); + + /* + * We don't need protection against CPU-hotplug here because +diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h +index e99a5234d9ed..958ac107062c 100644 +--- a/kernel/rcu/tree.h ++++ b/kernel/rcu/tree.h +@@ -588,18 +588,18 @@ extern struct list_head rcu_struct_flavors; + */ + extern struct rcu_state rcu_sched_state; + ++#ifndef CONFIG_PREEMPT_RT_FULL + extern struct rcu_state rcu_bh_state; ++#endif + + #ifdef CONFIG_PREEMPT_RCU + extern struct rcu_state rcu_preempt_state; + #endif /* #ifdef CONFIG_PREEMPT_RCU */ + +-#ifdef CONFIG_RCU_BOOST + DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_status); + DECLARE_PER_CPU(int, rcu_cpu_kthread_cpu); + DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_loops); + DECLARE_PER_CPU(char, rcu_cpu_has_work); +-#endif /* #ifdef CONFIG_RCU_BOOST */ + + #ifndef RCU_TREE_NONCORE + +@@ -619,10 +619,9 @@ void call_rcu(struct rcu_head *head, rcu_callback_t func); + static void __init __rcu_init_preempt(void); + static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags); + static void rcu_preempt_boost_start_gp(struct rcu_node *rnp); +-static void invoke_rcu_callbacks_kthread(void); + static bool rcu_is_callbacks_kthread(void); ++static void rcu_cpu_kthread_setup(unsigned int cpu); + #ifdef CONFIG_RCU_BOOST +-static void rcu_preempt_do_callbacks(void); + static int rcu_spawn_one_boost_kthread(struct rcu_state *rsp, + struct rcu_node *rnp); + #endif /* #ifdef CONFIG_RCU_BOOST */ +diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h +index 56583e764ebf..7c656f8e192f 100644 +--- a/kernel/rcu/tree_plugin.h ++++ b/kernel/rcu/tree_plugin.h +@@ -24,25 +24,10 @@ + * Paul E. McKenney + */ + +-#include +-#include +-#include +-#include +-#include "../time/tick-internal.h" +- + #ifdef CONFIG_RCU_BOOST + + #include "../locking/rtmutex_common.h" + +-/* +- * Control variables for per-CPU and per-rcu_node kthreads. These +- * handle all flavors of RCU. +- */ +-static DEFINE_PER_CPU(struct task_struct *, rcu_cpu_kthread_task); +-DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_status); +-DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_loops); +-DEFINE_PER_CPU(char, rcu_cpu_has_work); +- + #else /* #ifdef CONFIG_RCU_BOOST */ + + /* +@@ -55,6 +40,14 @@ DEFINE_PER_CPU(char, rcu_cpu_has_work); + + #endif /* #else #ifdef CONFIG_RCU_BOOST */ + ++/* ++ * Control variables for per-CPU and per-rcu_node kthreads. These ++ * handle all flavors of RCU. ++ */ ++DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_status); ++DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_loops); ++DEFINE_PER_CPU(char, rcu_cpu_has_work); ++ + #ifdef CONFIG_RCU_NOCB_CPU + static cpumask_var_t rcu_nocb_mask; /* CPUs to have callbacks offloaded. */ + static bool have_rcu_nocb_mask; /* Was rcu_nocb_mask allocated? */ +@@ -426,7 +419,7 @@ void rcu_read_unlock_special(struct task_struct *t) + } + + /* Hardware IRQ handlers cannot block, complain if they get here. */ +- if (in_irq() || in_serving_softirq()) { ++ if (preempt_count() & (HARDIRQ_MASK | SOFTIRQ_OFFSET)) { + lockdep_rcu_suspicious(__FILE__, __LINE__, + "rcu_read_unlock() from irq or softirq with blocking in critical section!!!\n"); + pr_alert("->rcu_read_unlock_special: %#x (b: %d, enq: %d nq: %d)\n", +@@ -632,15 +625,6 @@ static void rcu_preempt_check_callbacks(void) + t->rcu_read_unlock_special.b.need_qs = true; + } + +-#ifdef CONFIG_RCU_BOOST +- +-static void rcu_preempt_do_callbacks(void) +-{ +- rcu_do_batch(rcu_state_p, this_cpu_ptr(rcu_data_p)); +-} +- +-#endif /* #ifdef CONFIG_RCU_BOOST */ +- + /* + * Queue a preemptible-RCU callback for invocation after a grace period. + */ +@@ -829,6 +813,19 @@ void exit_rcu(void) + + #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ + ++/* ++ * If boosting, set rcuc kthreads to realtime priority. ++ */ ++static void rcu_cpu_kthread_setup(unsigned int cpu) ++{ ++#ifdef CONFIG_RCU_BOOST ++ struct sched_param sp; ++ ++ sp.sched_priority = kthread_prio; ++ sched_setscheduler_nocheck(current, SCHED_FIFO, &sp); ++#endif /* #ifdef CONFIG_RCU_BOOST */ ++} ++ + #ifdef CONFIG_RCU_BOOST + + #include "../locking/rtmutex_common.h" +@@ -860,16 +857,6 @@ static void rcu_initiate_boost_trace(struct rcu_node *rnp) + + #endif /* #else #ifdef CONFIG_RCU_TRACE */ + +-static void rcu_wake_cond(struct task_struct *t, int status) +-{ +- /* +- * If the thread is yielding, only wake it when this +- * is invoked from idle +- */ +- if (status != RCU_KTHREAD_YIELDING || is_idle_task(current)) +- wake_up_process(t); +-} +- + /* + * Carry out RCU priority boosting on the task indicated by ->exp_tasks + * or ->boost_tasks, advancing the pointer to the next task in the +@@ -1013,23 +1000,6 @@ static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags) + } + + /* +- * Wake up the per-CPU kthread to invoke RCU callbacks. +- */ +-static void invoke_rcu_callbacks_kthread(void) +-{ +- unsigned long flags; +- +- local_irq_save(flags); +- __this_cpu_write(rcu_cpu_has_work, 1); +- if (__this_cpu_read(rcu_cpu_kthread_task) != NULL && +- current != __this_cpu_read(rcu_cpu_kthread_task)) { +- rcu_wake_cond(__this_cpu_read(rcu_cpu_kthread_task), +- __this_cpu_read(rcu_cpu_kthread_status)); +- } +- local_irq_restore(flags); +-} +- +-/* + * Is the current CPU running the RCU-callbacks kthread? + * Caller must have preemption disabled. + */ +@@ -1083,67 +1053,6 @@ static int rcu_spawn_one_boost_kthread(struct rcu_state *rsp, + return 0; + } + +-static void rcu_kthread_do_work(void) +-{ +- rcu_do_batch(&rcu_sched_state, this_cpu_ptr(&rcu_sched_data)); +- rcu_do_batch(&rcu_bh_state, this_cpu_ptr(&rcu_bh_data)); +- rcu_preempt_do_callbacks(); +-} +- +-static void rcu_cpu_kthread_setup(unsigned int cpu) +-{ +- struct sched_param sp; +- +- sp.sched_priority = kthread_prio; +- sched_setscheduler_nocheck(current, SCHED_FIFO, &sp); +-} +- +-static void rcu_cpu_kthread_park(unsigned int cpu) +-{ +- per_cpu(rcu_cpu_kthread_status, cpu) = RCU_KTHREAD_OFFCPU; +-} +- +-static int rcu_cpu_kthread_should_run(unsigned int cpu) +-{ +- return __this_cpu_read(rcu_cpu_has_work); +-} +- +-/* +- * Per-CPU kernel thread that invokes RCU callbacks. This replaces the +- * RCU softirq used in flavors and configurations of RCU that do not +- * support RCU priority boosting. +- */ +-static void rcu_cpu_kthread(unsigned int cpu) +-{ +- unsigned int *statusp = this_cpu_ptr(&rcu_cpu_kthread_status); +- char work, *workp = this_cpu_ptr(&rcu_cpu_has_work); +- int spincnt; +- +- for (spincnt = 0; spincnt < 10; spincnt++) { +- trace_rcu_utilization(TPS("Start CPU kthread@rcu_wait")); +- local_bh_disable(); +- *statusp = RCU_KTHREAD_RUNNING; +- this_cpu_inc(rcu_cpu_kthread_loops); +- local_irq_disable(); +- work = *workp; +- *workp = 0; +- local_irq_enable(); +- if (work) +- rcu_kthread_do_work(); +- local_bh_enable(); +- if (*workp == 0) { +- trace_rcu_utilization(TPS("End CPU kthread@rcu_wait")); +- *statusp = RCU_KTHREAD_WAITING; +- return; +- } +- } +- *statusp = RCU_KTHREAD_YIELDING; +- trace_rcu_utilization(TPS("Start CPU kthread@rcu_yield")); +- schedule_timeout_interruptible(2); +- trace_rcu_utilization(TPS("End CPU kthread@rcu_yield")); +- *statusp = RCU_KTHREAD_WAITING; +-} +- + /* + * Set the per-rcu_node kthread's affinity to cover all CPUs that are + * served by the rcu_node in question. The CPU hotplug lock is still +@@ -1174,26 +1083,12 @@ static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu) + free_cpumask_var(cm); + } + +-static struct smp_hotplug_thread rcu_cpu_thread_spec = { +- .store = &rcu_cpu_kthread_task, +- .thread_should_run = rcu_cpu_kthread_should_run, +- .thread_fn = rcu_cpu_kthread, +- .thread_comm = "rcuc/%u", +- .setup = rcu_cpu_kthread_setup, +- .park = rcu_cpu_kthread_park, +-}; +- + /* + * Spawn boost kthreads -- called as soon as the scheduler is running. + */ + static void __init rcu_spawn_boost_kthreads(void) + { + struct rcu_node *rnp; +- int cpu; +- +- for_each_possible_cpu(cpu) +- per_cpu(rcu_cpu_has_work, cpu) = 0; +- BUG_ON(smpboot_register_percpu_thread(&rcu_cpu_thread_spec)); + rcu_for_each_leaf_node(rcu_state_p, rnp) + (void)rcu_spawn_one_boost_kthread(rcu_state_p, rnp); + } +@@ -1216,11 +1111,6 @@ static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags) + raw_spin_unlock_irqrestore_rcu_node(rnp, flags); + } + +-static void invoke_rcu_callbacks_kthread(void) +-{ +- WARN_ON_ONCE(1); +-} +- + static bool rcu_is_callbacks_kthread(void) + { + return false; +@@ -1244,7 +1134,7 @@ static void rcu_prepare_kthreads(int cpu) + + #endif /* #else #ifdef CONFIG_RCU_BOOST */ + +-#if !defined(CONFIG_RCU_FAST_NO_HZ) ++#if !defined(CONFIG_RCU_FAST_NO_HZ) || defined(CONFIG_PREEMPT_RT_FULL) + + /* + * Check to see if any future RCU-related work will need to be done +@@ -1261,7 +1151,9 @@ int rcu_needs_cpu(u64 basemono, u64 *nextevt) + return IS_ENABLED(CONFIG_RCU_NOCB_CPU_ALL) + ? 0 : rcu_cpu_has_callbacks(NULL); + } ++#endif /* !defined(CONFIG_RCU_FAST_NO_HZ) || defined(CONFIG_PREEMPT_RT_FULL) */ + ++#if !defined(CONFIG_RCU_FAST_NO_HZ) + /* + * Because we do not have RCU_FAST_NO_HZ, don't bother cleaning up + * after it. +@@ -1357,6 +1249,8 @@ static bool __maybe_unused rcu_try_advance_all_cbs(void) + return cbs_ready; + } + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + /* + * Allow the CPU to enter dyntick-idle mode unless it has callbacks ready + * to invoke. If the CPU has callbacks, try to advance them. Tell the +@@ -1402,6 +1296,7 @@ int rcu_needs_cpu(u64 basemono, u64 *nextevt) + *nextevt = basemono + dj * TICK_NSEC; + return 0; + } ++#endif /* #ifndef CONFIG_PREEMPT_RT_FULL */ + + /* + * Prepare a CPU for idle from an RCU perspective. The first major task +diff --git a/kernel/rcu/update.c b/kernel/rcu/update.c +index 4f6db7e6a117..ee02e1e1b3e5 100644 +--- a/kernel/rcu/update.c ++++ b/kernel/rcu/update.c +@@ -62,7 +62,7 @@ + #ifndef CONFIG_TINY_RCU + module_param(rcu_expedited, int, 0); + module_param(rcu_normal, int, 0); +-static int rcu_normal_after_boot; ++static int rcu_normal_after_boot = IS_ENABLED(CONFIG_PREEMPT_RT_FULL); + module_param(rcu_normal_after_boot, int, 0); + #endif /* #ifndef CONFIG_TINY_RCU */ + +@@ -132,8 +132,7 @@ bool rcu_gp_is_normal(void) + } + EXPORT_SYMBOL_GPL(rcu_gp_is_normal); + +-static atomic_t rcu_expedited_nesting = +- ATOMIC_INIT(IS_ENABLED(CONFIG_RCU_EXPEDITE_BOOT) ? 1 : 0); ++static atomic_t rcu_expedited_nesting = ATOMIC_INIT(1); + + /* + * Should normal grace-period primitives be expedited? Intended for +@@ -182,8 +181,7 @@ EXPORT_SYMBOL_GPL(rcu_unexpedite_gp); + */ + void rcu_end_inkernel_boot(void) + { +- if (IS_ENABLED(CONFIG_RCU_EXPEDITE_BOOT)) +- rcu_unexpedite_gp(); ++ rcu_unexpedite_gp(); + if (rcu_normal_after_boot) + WRITE_ONCE(rcu_normal, 1); + } +@@ -298,6 +296,7 @@ int rcu_read_lock_held(void) + } + EXPORT_SYMBOL_GPL(rcu_read_lock_held); + ++#ifndef CONFIG_PREEMPT_RT_FULL + /** + * rcu_read_lock_bh_held() - might we be in RCU-bh read-side critical section? + * +@@ -324,6 +323,7 @@ int rcu_read_lock_bh_held(void) + return in_softirq() || irqs_disabled(); + } + EXPORT_SYMBOL_GPL(rcu_read_lock_bh_held); ++#endif + + #endif /* #ifdef CONFIG_DEBUG_LOCK_ALLOC */ + +diff --git a/kernel/sched/Makefile b/kernel/sched/Makefile +index 5e59b832ae2b..7337a7f60e3f 100644 +--- a/kernel/sched/Makefile ++++ b/kernel/sched/Makefile +@@ -17,7 +17,7 @@ endif + + obj-y += core.o loadavg.o clock.o cputime.o + obj-y += idle_task.o fair.o rt.o deadline.o stop_task.o +-obj-y += wait.o swait.o completion.o idle.o ++obj-y += wait.o swait.o swork.o completion.o idle.o + obj-$(CONFIG_SMP) += cpupri.o cpudeadline.o + obj-$(CONFIG_SCHED_AUTOGROUP) += auto_group.o + obj-$(CONFIG_SCHEDSTATS) += stats.o +diff --git a/kernel/sched/completion.c b/kernel/sched/completion.c +index 8d0f35debf35..b62cf6400fe0 100644 +--- a/kernel/sched/completion.c ++++ b/kernel/sched/completion.c +@@ -30,10 +30,10 @@ void complete(struct completion *x) + { + unsigned long flags; + +- spin_lock_irqsave(&x->wait.lock, flags); ++ raw_spin_lock_irqsave(&x->wait.lock, flags); + x->done++; +- __wake_up_locked(&x->wait, TASK_NORMAL, 1); +- spin_unlock_irqrestore(&x->wait.lock, flags); ++ swake_up_locked(&x->wait); ++ raw_spin_unlock_irqrestore(&x->wait.lock, flags); + } + EXPORT_SYMBOL(complete); + +@@ -50,10 +50,10 @@ void complete_all(struct completion *x) + { + unsigned long flags; + +- spin_lock_irqsave(&x->wait.lock, flags); ++ raw_spin_lock_irqsave(&x->wait.lock, flags); + x->done += UINT_MAX/2; +- __wake_up_locked(&x->wait, TASK_NORMAL, 0); +- spin_unlock_irqrestore(&x->wait.lock, flags); ++ swake_up_all_locked(&x->wait); ++ raw_spin_unlock_irqrestore(&x->wait.lock, flags); + } + EXPORT_SYMBOL(complete_all); + +@@ -62,20 +62,20 @@ do_wait_for_common(struct completion *x, + long (*action)(long), long timeout, int state) + { + if (!x->done) { +- DECLARE_WAITQUEUE(wait, current); ++ DECLARE_SWAITQUEUE(wait); + +- __add_wait_queue_tail_exclusive(&x->wait, &wait); ++ __prepare_to_swait(&x->wait, &wait); + do { + if (signal_pending_state(state, current)) { + timeout = -ERESTARTSYS; + break; + } + __set_current_state(state); +- spin_unlock_irq(&x->wait.lock); ++ raw_spin_unlock_irq(&x->wait.lock); + timeout = action(timeout); +- spin_lock_irq(&x->wait.lock); ++ raw_spin_lock_irq(&x->wait.lock); + } while (!x->done && timeout); +- __remove_wait_queue(&x->wait, &wait); ++ __finish_swait(&x->wait, &wait); + if (!x->done) + return timeout; + } +@@ -89,9 +89,9 @@ __wait_for_common(struct completion *x, + { + might_sleep(); + +- spin_lock_irq(&x->wait.lock); ++ raw_spin_lock_irq(&x->wait.lock); + timeout = do_wait_for_common(x, action, timeout, state); +- spin_unlock_irq(&x->wait.lock); ++ raw_spin_unlock_irq(&x->wait.lock); + return timeout; + } + +@@ -277,12 +277,12 @@ bool try_wait_for_completion(struct completion *x) + if (!READ_ONCE(x->done)) + return 0; + +- spin_lock_irqsave(&x->wait.lock, flags); ++ raw_spin_lock_irqsave(&x->wait.lock, flags); + if (!x->done) + ret = 0; + else + x->done--; +- spin_unlock_irqrestore(&x->wait.lock, flags); ++ raw_spin_unlock_irqrestore(&x->wait.lock, flags); + return ret; + } + EXPORT_SYMBOL(try_wait_for_completion); +@@ -311,7 +311,7 @@ bool completion_done(struct completion *x) + * after it's acquired the lock. + */ + smp_rmb(); +- spin_unlock_wait(&x->wait.lock); ++ raw_spin_unlock_wait(&x->wait.lock); + return true; + } + EXPORT_SYMBOL(completion_done); +diff --git a/kernel/sched/core.c b/kernel/sched/core.c +index 154fd689fe02..10e832da70b6 100644 +--- a/kernel/sched/core.c ++++ b/kernel/sched/core.c +@@ -129,7 +129,11 @@ const_debug unsigned int sysctl_sched_features = + * Number of tasks to iterate in a single balance run. + * Limited because this is done with IRQs disabled. + */ ++#ifndef CONFIG_PREEMPT_RT_FULL + const_debug unsigned int sysctl_sched_nr_migrate = 32; ++#else ++const_debug unsigned int sysctl_sched_nr_migrate = 8; ++#endif + + /* + * period over which we average the RT time consumption, measured +@@ -345,6 +349,7 @@ static void init_rq_hrtick(struct rq *rq) + + hrtimer_init(&rq->hrtick_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + rq->hrtick_timer.function = hrtick; ++ rq->hrtick_timer.irqsafe = 1; + } + #else /* CONFIG_SCHED_HRTICK */ + static inline void hrtick_clear(struct rq *rq) +@@ -449,7 +454,7 @@ void wake_q_add(struct wake_q_head *head, struct task_struct *task) + head->lastp = &node->next; + } + +-void wake_up_q(struct wake_q_head *head) ++void __wake_up_q(struct wake_q_head *head, bool sleeper) + { + struct wake_q_node *node = head->first; + +@@ -466,7 +471,10 @@ void wake_up_q(struct wake_q_head *head) + * wake_up_process() implies a wmb() to pair with the queueing + * in wake_q_add() so as not to miss wakeups. + */ +- wake_up_process(task); ++ if (sleeper) ++ wake_up_lock_sleeper(task); ++ else ++ wake_up_process(task); + put_task_struct(task); + } + } +@@ -502,6 +510,38 @@ void resched_curr(struct rq *rq) + trace_sched_wake_idle_without_ipi(cpu); + } + ++#ifdef CONFIG_PREEMPT_LAZY ++void resched_curr_lazy(struct rq *rq) ++{ ++ struct task_struct *curr = rq->curr; ++ int cpu; ++ ++ if (!sched_feat(PREEMPT_LAZY)) { ++ resched_curr(rq); ++ return; ++ } ++ ++ lockdep_assert_held(&rq->lock); ++ ++ if (test_tsk_need_resched(curr)) ++ return; ++ ++ if (test_tsk_need_resched_lazy(curr)) ++ return; ++ ++ set_tsk_need_resched_lazy(curr); ++ ++ cpu = cpu_of(rq); ++ if (cpu == smp_processor_id()) ++ return; ++ ++ /* NEED_RESCHED_LAZY must be visible before we test polling */ ++ smp_mb(); ++ if (!tsk_is_polling(curr)) ++ smp_send_reschedule(cpu); ++} ++#endif ++ + void resched_cpu(int cpu) + { + struct rq *rq = cpu_rq(cpu); +@@ -525,11 +565,14 @@ void resched_cpu(int cpu) + */ + int get_nohz_timer_target(void) + { +- int i, cpu = smp_processor_id(); ++ int i, cpu; + struct sched_domain *sd; + ++ preempt_disable_rt(); ++ cpu = smp_processor_id(); ++ + if (!idle_cpu(cpu) && is_housekeeping_cpu(cpu)) +- return cpu; ++ goto preempt_en_rt; + + rcu_read_lock(); + for_each_domain(cpu, sd) { +@@ -548,6 +591,8 @@ int get_nohz_timer_target(void) + cpu = housekeeping_any_cpu(); + unlock: + rcu_read_unlock(); ++preempt_en_rt: ++ preempt_enable_rt(); + return cpu; + } + /* +@@ -1100,6 +1145,11 @@ void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask) + + lockdep_assert_held(&p->pi_lock); + ++ if (__migrate_disabled(p)) { ++ cpumask_copy(&p->cpus_allowed, new_mask); ++ return; ++ } ++ + queued = task_on_rq_queued(p); + running = task_current(rq, p); + +@@ -1122,6 +1172,84 @@ void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask) + set_curr_task(rq, p); + } + ++static DEFINE_PER_CPU(struct cpumask, sched_cpumasks); ++static DEFINE_MUTEX(sched_down_mutex); ++static cpumask_t sched_down_cpumask; ++ ++void tell_sched_cpu_down_begin(int cpu) ++{ ++ mutex_lock(&sched_down_mutex); ++ cpumask_set_cpu(cpu, &sched_down_cpumask); ++ mutex_unlock(&sched_down_mutex); ++} ++ ++void tell_sched_cpu_down_done(int cpu) ++{ ++ mutex_lock(&sched_down_mutex); ++ cpumask_clear_cpu(cpu, &sched_down_cpumask); ++ mutex_unlock(&sched_down_mutex); ++} ++ ++/** ++ * migrate_me - try to move the current task off this cpu ++ * ++ * Used by the pin_current_cpu() code to try to get tasks ++ * to move off the current CPU as it is going down. ++ * It will only move the task if the task isn't pinned to ++ * the CPU (with migrate_disable, affinity or NO_SETAFFINITY) ++ * and the task has to be in a RUNNING state. Otherwise the ++ * movement of the task will wake it up (change its state ++ * to running) when the task did not expect it. ++ * ++ * Returns 1 if it succeeded in moving the current task ++ * 0 otherwise. ++ */ ++int migrate_me(void) ++{ ++ struct task_struct *p = current; ++ struct migration_arg arg; ++ struct cpumask *cpumask; ++ struct cpumask *mask; ++ unsigned int dest_cpu; ++ struct rq_flags rf; ++ struct rq *rq; ++ ++ /* ++ * We can not migrate tasks bounded to a CPU or tasks not ++ * running. The movement of the task will wake it up. ++ */ ++ if (p->flags & PF_NO_SETAFFINITY || p->state) ++ return 0; ++ ++ mutex_lock(&sched_down_mutex); ++ rq = task_rq_lock(p, &rf); ++ ++ cpumask = this_cpu_ptr(&sched_cpumasks); ++ mask = &p->cpus_allowed; ++ ++ cpumask_andnot(cpumask, mask, &sched_down_cpumask); ++ ++ if (!cpumask_weight(cpumask)) { ++ /* It's only on this CPU? */ ++ task_rq_unlock(rq, p, &rf); ++ mutex_unlock(&sched_down_mutex); ++ return 0; ++ } ++ ++ dest_cpu = cpumask_any_and(cpu_active_mask, cpumask); ++ ++ arg.task = p; ++ arg.dest_cpu = dest_cpu; ++ ++ task_rq_unlock(rq, p, &rf); ++ ++ stop_one_cpu(cpu_of(rq), migration_cpu_stop, &arg); ++ tlb_migrate_finish(p->mm); ++ mutex_unlock(&sched_down_mutex); ++ ++ return 1; ++} ++ + /* + * Change a given task's CPU affinity. Migrate the thread to a + * proper CPU and schedule it away if the CPU it's executing on +@@ -1179,7 +1307,7 @@ static int __set_cpus_allowed_ptr(struct task_struct *p, + } + + /* Can the task run on the task's current CPU? If so, we're done */ +- if (cpumask_test_cpu(task_cpu(p), new_mask)) ++ if (cpumask_test_cpu(task_cpu(p), new_mask) || __migrate_disabled(p)) + goto out; + + dest_cpu = cpumask_any_and(cpu_valid_mask, new_mask); +@@ -1366,6 +1494,18 @@ int migrate_swap(struct task_struct *cur, struct task_struct *p) + return ret; + } + ++static bool check_task_state(struct task_struct *p, long match_state) ++{ ++ bool match = false; ++ ++ raw_spin_lock_irq(&p->pi_lock); ++ if (p->state == match_state || p->saved_state == match_state) ++ match = true; ++ raw_spin_unlock_irq(&p->pi_lock); ++ ++ return match; ++} ++ + /* + * wait_task_inactive - wait for a thread to unschedule. + * +@@ -1410,7 +1550,7 @@ unsigned long wait_task_inactive(struct task_struct *p, long match_state) + * is actually now running somewhere else! + */ + while (task_running(rq, p)) { +- if (match_state && unlikely(p->state != match_state)) ++ if (match_state && !check_task_state(p, match_state)) + return 0; + cpu_relax(); + } +@@ -1425,7 +1565,8 @@ unsigned long wait_task_inactive(struct task_struct *p, long match_state) + running = task_running(rq, p); + queued = task_on_rq_queued(p); + ncsw = 0; +- if (!match_state || p->state == match_state) ++ if (!match_state || p->state == match_state || ++ p->saved_state == match_state) + ncsw = p->nvcsw | LONG_MIN; /* sets MSB */ + task_rq_unlock(rq, p, &rf); + +@@ -1680,10 +1821,6 @@ static inline void ttwu_activate(struct rq *rq, struct task_struct *p, int en_fl + { + activate_task(rq, p, en_flags); + p->on_rq = TASK_ON_RQ_QUEUED; +- +- /* if a worker is waking up, notify workqueue */ +- if (p->flags & PF_WQ_WORKER) +- wq_worker_waking_up(p, cpu_of(rq)); + } + + /* +@@ -2018,8 +2155,27 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags) + */ + smp_mb__before_spinlock(); + raw_spin_lock_irqsave(&p->pi_lock, flags); +- if (!(p->state & state)) ++ if (!(p->state & state)) { ++ /* ++ * The task might be running due to a spinlock sleeper ++ * wakeup. Check the saved state and set it to running ++ * if the wakeup condition is true. ++ */ ++ if (!(wake_flags & WF_LOCK_SLEEPER)) { ++ if (p->saved_state & state) { ++ p->saved_state = TASK_RUNNING; ++ success = 1; ++ } ++ } + goto out; ++ } ++ ++ /* ++ * If this is a regular wakeup, then we can unconditionally ++ * clear the saved state of a "lock sleeper". ++ */ ++ if (!(wake_flags & WF_LOCK_SLEEPER)) ++ p->saved_state = TASK_RUNNING; + + trace_sched_waking(p); + +@@ -2102,53 +2258,6 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags) + } + + /** +- * try_to_wake_up_local - try to wake up a local task with rq lock held +- * @p: the thread to be awakened +- * @cookie: context's cookie for pinning +- * +- * Put @p on the run-queue if it's not already there. The caller must +- * ensure that this_rq() is locked, @p is bound to this_rq() and not +- * the current task. +- */ +-static void try_to_wake_up_local(struct task_struct *p, struct pin_cookie cookie) +-{ +- struct rq *rq = task_rq(p); +- +- if (WARN_ON_ONCE(rq != this_rq()) || +- WARN_ON_ONCE(p == current)) +- return; +- +- lockdep_assert_held(&rq->lock); +- +- if (!raw_spin_trylock(&p->pi_lock)) { +- /* +- * This is OK, because current is on_cpu, which avoids it being +- * picked for load-balance and preemption/IRQs are still +- * disabled avoiding further scheduler activity on it and we've +- * not yet picked a replacement task. +- */ +- lockdep_unpin_lock(&rq->lock, cookie); +- raw_spin_unlock(&rq->lock); +- raw_spin_lock(&p->pi_lock); +- raw_spin_lock(&rq->lock); +- lockdep_repin_lock(&rq->lock, cookie); +- } +- +- if (!(p->state & TASK_NORMAL)) +- goto out; +- +- trace_sched_waking(p); +- +- if (!task_on_rq_queued(p)) +- ttwu_activate(rq, p, ENQUEUE_WAKEUP); +- +- ttwu_do_wakeup(rq, p, 0, cookie); +- ttwu_stat(p, smp_processor_id(), 0); +-out: +- raw_spin_unlock(&p->pi_lock); +-} +- +-/** + * wake_up_process - Wake up a specific process + * @p: The process to be woken up. + * +@@ -2166,6 +2275,18 @@ int wake_up_process(struct task_struct *p) + } + EXPORT_SYMBOL(wake_up_process); + ++/** ++ * wake_up_lock_sleeper - Wake up a specific process blocked on a "sleeping lock" ++ * @p: The process to be woken up. ++ * ++ * Same as wake_up_process() above, but wake_flags=WF_LOCK_SLEEPER to indicate ++ * the nature of the wakeup. ++ */ ++int wake_up_lock_sleeper(struct task_struct *p) ++{ ++ return try_to_wake_up(p, TASK_UNINTERRUPTIBLE, WF_LOCK_SLEEPER); ++} ++ + int wake_up_state(struct task_struct *p, unsigned int state) + { + return try_to_wake_up(p, state, 0); +@@ -2442,6 +2563,9 @@ int sched_fork(unsigned long clone_flags, struct task_struct *p) + p->on_cpu = 0; + #endif + init_task_preempt_count(p); ++#ifdef CONFIG_HAVE_PREEMPT_LAZY ++ task_thread_info(p)->preempt_lazy_count = 0; ++#endif + #ifdef CONFIG_SMP + plist_node_init(&p->pushable_tasks, MAX_PRIO); + RB_CLEAR_NODE(&p->pushable_dl_tasks); +@@ -2770,21 +2894,16 @@ static struct rq *finish_task_switch(struct task_struct *prev) + finish_arch_post_lock_switch(); + + fire_sched_in_preempt_notifiers(current); ++ /* ++ * We use mmdrop_delayed() here so we don't have to do the ++ * full __mmdrop() when we are the last user. ++ */ + if (mm) +- mmdrop(mm); ++ mmdrop_delayed(mm); + if (unlikely(prev_state == TASK_DEAD)) { + if (prev->sched_class->task_dead) + prev->sched_class->task_dead(prev); + +- /* +- * Remove function-return probe instances associated with this +- * task and put them back on the free list. +- */ +- kprobe_flush_task(prev); +- +- /* Task is done with its stack. */ +- put_task_stack(prev); +- + put_task_struct(prev); + } + +@@ -3252,6 +3371,77 @@ static inline void schedule_debug(struct task_struct *prev) + schedstat_inc(this_rq()->sched_count); + } + ++#if defined(CONFIG_PREEMPT_RT_FULL) && defined(CONFIG_SMP) ++ ++void migrate_disable(void) ++{ ++ struct task_struct *p = current; ++ ++ if (in_atomic() || irqs_disabled()) { ++#ifdef CONFIG_SCHED_DEBUG ++ p->migrate_disable_atomic++; ++#endif ++ return; ++ } ++ ++#ifdef CONFIG_SCHED_DEBUG ++ if (unlikely(p->migrate_disable_atomic)) { ++ tracing_off(); ++ WARN_ON_ONCE(1); ++ } ++#endif ++ ++ if (p->migrate_disable) { ++ p->migrate_disable++; ++ return; ++ } ++ ++ preempt_disable(); ++ preempt_lazy_disable(); ++ pin_current_cpu(); ++ p->migrate_disable = 1; ++ preempt_enable(); ++} ++EXPORT_SYMBOL(migrate_disable); ++ ++void migrate_enable(void) ++{ ++ struct task_struct *p = current; ++ ++ if (in_atomic() || irqs_disabled()) { ++#ifdef CONFIG_SCHED_DEBUG ++ p->migrate_disable_atomic--; ++#endif ++ return; ++ } ++ ++#ifdef CONFIG_SCHED_DEBUG ++ if (unlikely(p->migrate_disable_atomic)) { ++ tracing_off(); ++ WARN_ON_ONCE(1); ++ } ++#endif ++ WARN_ON_ONCE(p->migrate_disable <= 0); ++ ++ if (p->migrate_disable > 1) { ++ p->migrate_disable--; ++ return; ++ } ++ ++ preempt_disable(); ++ /* ++ * Clearing migrate_disable causes tsk_cpus_allowed to ++ * show the tasks original cpu affinity. ++ */ ++ p->migrate_disable = 0; ++ ++ unpin_current_cpu(); ++ preempt_enable(); ++ preempt_lazy_enable(); ++} ++EXPORT_SYMBOL(migrate_enable); ++#endif ++ + /* + * Pick up the highest-prio task: + */ +@@ -3368,19 +3558,6 @@ static void __sched notrace __schedule(bool preempt) + } else { + deactivate_task(rq, prev, DEQUEUE_SLEEP); + prev->on_rq = 0; +- +- /* +- * If a worker went to sleep, notify and ask workqueue +- * whether it wants to wake up a task to maintain +- * concurrency. +- */ +- if (prev->flags & PF_WQ_WORKER) { +- struct task_struct *to_wakeup; +- +- to_wakeup = wq_worker_sleeping(prev); +- if (to_wakeup) +- try_to_wake_up_local(to_wakeup, cookie); +- } + } + switch_count = &prev->nvcsw; + } +@@ -3390,6 +3567,7 @@ static void __sched notrace __schedule(bool preempt) + + next = pick_next_task(rq, prev, cookie); + clear_tsk_need_resched(prev); ++ clear_tsk_need_resched_lazy(prev); + clear_preempt_need_resched(); + rq->clock_skip_update = 0; + +@@ -3437,9 +3615,20 @@ void __noreturn do_task_dead(void) + + static inline void sched_submit_work(struct task_struct *tsk) + { +- if (!tsk->state || tsk_is_pi_blocked(tsk)) ++ if (!tsk->state) + return; + /* ++ * If a worker went to sleep, notify and ask workqueue whether ++ * it wants to wake up a task to maintain concurrency. ++ */ ++ if (tsk->flags & PF_WQ_WORKER) ++ wq_worker_sleeping(tsk); ++ ++ ++ if (tsk_is_pi_blocked(tsk)) ++ return; ++ ++ /* + * If we are going to sleep and we have plugged IO queued, + * make sure to submit it to avoid deadlocks. + */ +@@ -3447,6 +3636,12 @@ static inline void sched_submit_work(struct task_struct *tsk) + blk_schedule_flush_plug(tsk); + } + ++static void sched_update_worker(struct task_struct *tsk) ++{ ++ if (tsk->flags & PF_WQ_WORKER) ++ wq_worker_running(tsk); ++} ++ + asmlinkage __visible void __sched schedule(void) + { + struct task_struct *tsk = current; +@@ -3457,6 +3652,7 @@ asmlinkage __visible void __sched schedule(void) + __schedule(false); + sched_preempt_enable_no_resched(); + } while (need_resched()); ++ sched_update_worker(tsk); + } + EXPORT_SYMBOL(schedule); + +@@ -3520,6 +3716,30 @@ static void __sched notrace preempt_schedule_common(void) + } while (need_resched()); + } + ++#ifdef CONFIG_PREEMPT_LAZY ++/* ++ * If TIF_NEED_RESCHED is then we allow to be scheduled away since this is ++ * set by a RT task. Oterwise we try to avoid beeing scheduled out as long as ++ * preempt_lazy_count counter >0. ++ */ ++static __always_inline int preemptible_lazy(void) ++{ ++ if (test_thread_flag(TIF_NEED_RESCHED)) ++ return 1; ++ if (current_thread_info()->preempt_lazy_count) ++ return 0; ++ return 1; ++} ++ ++#else ++ ++static inline int preemptible_lazy(void) ++{ ++ return 1; ++} ++ ++#endif ++ + #ifdef CONFIG_PREEMPT + /* + * this is the entry point to schedule() from in-kernel preemption +@@ -3534,7 +3754,8 @@ asmlinkage __visible void __sched notrace preempt_schedule(void) + */ + if (likely(!preemptible())) + return; +- ++ if (!preemptible_lazy()) ++ return; + preempt_schedule_common(); + } + NOKPROBE_SYMBOL(preempt_schedule); +@@ -3561,6 +3782,9 @@ asmlinkage __visible void __sched notrace preempt_schedule_notrace(void) + if (likely(!preemptible())) + return; + ++ if (!preemptible_lazy()) ++ return; ++ + do { + /* + * Because the function tracer can trace preempt_count_sub() +@@ -3583,7 +3807,16 @@ asmlinkage __visible void __sched notrace preempt_schedule_notrace(void) + * an infinite recursion. + */ + prev_ctx = exception_enter(); ++ /* ++ * The add/subtract must not be traced by the function ++ * tracer. But we still want to account for the ++ * preempt off latency tracer. Since the _notrace versions ++ * of add/subtract skip the accounting for latency tracer ++ * we must force it manually. ++ */ ++ start_critical_timings(); + __schedule(true); ++ stop_critical_timings(); + exception_exit(prev_ctx); + + preempt_latency_stop(1); +@@ -3629,10 +3862,25 @@ EXPORT_SYMBOL(default_wake_function); + + #ifdef CONFIG_RT_MUTEXES + ++static inline int __rt_effective_prio(struct task_struct *pi_task, int prio) ++{ ++ if (pi_task) ++ prio = min(prio, pi_task->prio); ++ ++ return prio; ++} ++ ++static inline int rt_effective_prio(struct task_struct *p, int prio) ++{ ++ struct task_struct *pi_task = rt_mutex_get_top_task(p); ++ ++ return __rt_effective_prio(pi_task, prio); ++} ++ + /* + * rt_mutex_setprio - set the current priority of a task +- * @p: task +- * @prio: prio value (kernel-internal form) ++ * @p: task to boost ++ * @pi_task: donor task + * + * This function changes the 'effective' priority of a task. It does + * not touch ->normal_prio like __setscheduler(). +@@ -3640,16 +3888,40 @@ EXPORT_SYMBOL(default_wake_function); + * Used by the rt_mutex code to implement priority inheritance + * logic. Call site only calls if the priority of the task changed. + */ +-void rt_mutex_setprio(struct task_struct *p, int prio) ++void rt_mutex_setprio(struct task_struct *p, struct task_struct *pi_task) + { +- int oldprio, queued, running, queue_flag = DEQUEUE_SAVE | DEQUEUE_MOVE; ++ int prio, oldprio, queued, running, queue_flag = DEQUEUE_SAVE | DEQUEUE_MOVE; + const struct sched_class *prev_class; + struct rq_flags rf; + struct rq *rq; + +- BUG_ON(prio > MAX_PRIO); ++ /* XXX used to be waiter->prio, not waiter->task->prio */ ++ prio = __rt_effective_prio(pi_task, p->normal_prio); ++ ++ /* ++ * If nothing changed; bail early. ++ */ ++ if (p->pi_top_task == pi_task && prio == p->prio && !dl_prio(prio)) ++ return; + + rq = __task_rq_lock(p, &rf); ++ /* ++ * Set under pi_lock && rq->lock, such that the value can be used under ++ * either lock. ++ * ++ * Note that there is loads of tricky to make this pointer cache work ++ * right. rt_mutex_slowunlock()+rt_mutex_postunlock() work together to ++ * ensure a task is de-boosted (pi_task is set to NULL) before the ++ * task is allowed to run again (and can exit). This ensures the pointer ++ * points to a blocked task -- which guaratees the task is present. ++ */ ++ p->pi_top_task = pi_task; ++ ++ /* ++ * For FIFO/RR we only need to set prio, if that matches we're done. ++ */ ++ if (prio == p->prio && !dl_prio(prio)) ++ goto out_unlock; + + /* + * Idle task boosting is a nono in general. There is one +@@ -3669,7 +3941,7 @@ void rt_mutex_setprio(struct task_struct *p, int prio) + goto out_unlock; + } + +- trace_sched_pi_setprio(p, prio); ++ trace_sched_pi_setprio(p, pi_task); + oldprio = p->prio; + + if (oldprio == prio) +@@ -3693,7 +3965,6 @@ void rt_mutex_setprio(struct task_struct *p, int prio) + * running task + */ + if (dl_prio(prio)) { +- struct task_struct *pi_task = rt_mutex_get_top_task(p); + if (!dl_prio(p->normal_prio) || + (pi_task && dl_entity_preempt(&pi_task->dl, &p->dl))) { + p->dl.dl_boosted = 1; +@@ -3730,6 +4001,11 @@ void rt_mutex_setprio(struct task_struct *p, int prio) + balance_callback(rq); + preempt_enable(); + } ++#else ++static inline int rt_effective_prio(struct task_struct *p, int prio) ++{ ++ return prio; ++} + #endif + + void set_user_nice(struct task_struct *p, long nice) +@@ -3974,10 +4250,9 @@ static void __setscheduler(struct rq *rq, struct task_struct *p, + * Keep a potential priority boosting if called from + * sched_setscheduler(). + */ ++ p->prio = normal_prio(p); + if (keep_boost) +- p->prio = rt_mutex_get_effective_prio(p, normal_prio(p)); +- else +- p->prio = normal_prio(p); ++ p->prio = rt_effective_prio(p, p->prio); + + if (dl_prio(p->prio)) + p->sched_class = &dl_sched_class; +@@ -4264,7 +4539,7 @@ static int __sched_setscheduler(struct task_struct *p, + * the runqueue. This will be done when the task deboost + * itself. + */ +- new_effective_prio = rt_mutex_get_effective_prio(p, newprio); ++ new_effective_prio = rt_effective_prio(p, newprio); + if (new_effective_prio == oldprio) + queue_flags &= ~DEQUEUE_MOVE; + } +@@ -4939,6 +5214,7 @@ int __cond_resched_lock(spinlock_t *lock) + } + EXPORT_SYMBOL(__cond_resched_lock); + ++#ifndef CONFIG_PREEMPT_RT_FULL + int __sched __cond_resched_softirq(void) + { + BUG_ON(!in_softirq()); +@@ -4952,6 +5228,7 @@ int __sched __cond_resched_softirq(void) + return 0; + } + EXPORT_SYMBOL(__cond_resched_softirq); ++#endif + + /** + * yield - yield the current processor to other threads. +@@ -5315,7 +5592,9 @@ void init_idle(struct task_struct *idle, int cpu) + + /* Set the preempt count _outside_ the spinlocks! */ + init_idle_preempt_count(idle, cpu); +- ++#ifdef CONFIG_HAVE_PREEMPT_LAZY ++ task_thread_info(idle)->preempt_lazy_count = 0; ++#endif + /* + * The idle tasks have their own, simple scheduling class: + */ +@@ -5458,6 +5737,8 @@ void sched_setnuma(struct task_struct *p, int nid) + #endif /* CONFIG_NUMA_BALANCING */ + + #ifdef CONFIG_HOTPLUG_CPU ++static DEFINE_PER_CPU(struct mm_struct *, idle_last_mm); ++ + /* + * Ensures that the idle task is using init_mm right before its cpu goes + * offline. +@@ -5472,7 +5753,12 @@ void idle_task_exit(void) + switch_mm_irqs_off(mm, &init_mm, current); + finish_arch_post_lock_switch(); + } +- mmdrop(mm); ++ /* ++ * Defer the cleanup to an alive cpu. On RT we can neither ++ * call mmdrop() nor mmdrop_delayed() from here. ++ */ ++ per_cpu(idle_last_mm, smp_processor_id()) = mm; ++ + } + + /* +@@ -7418,6 +7704,10 @@ int sched_cpu_dying(unsigned int cpu) + update_max_interval(); + nohz_balance_exit_idle(cpu); + hrtick_clear(rq); ++ if (per_cpu(idle_last_mm, cpu)) { ++ mmdrop_delayed(per_cpu(idle_last_mm, cpu)); ++ per_cpu(idle_last_mm, cpu) = NULL; ++ } + return 0; + } + #endif +@@ -7698,7 +7988,7 @@ void __init sched_init(void) + #ifdef CONFIG_DEBUG_ATOMIC_SLEEP + static inline int preempt_count_equals(int preempt_offset) + { +- int nested = preempt_count() + rcu_preempt_depth(); ++ int nested = preempt_count() + sched_rcu_preempt_depth(); + + return (nested == preempt_offset); + } +diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c +index c95c5122b105..e00accf92a4b 100644 +--- a/kernel/sched/deadline.c ++++ b/kernel/sched/deadline.c +@@ -687,6 +687,7 @@ void init_dl_task_timer(struct sched_dl_entity *dl_se) + + hrtimer_init(timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + timer->function = dl_task_timer; ++ timer->irqsafe = 1; + } + + static +diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c +index fa178b62ea79..935224123441 100644 +--- a/kernel/sched/debug.c ++++ b/kernel/sched/debug.c +@@ -558,6 +558,9 @@ void print_rt_rq(struct seq_file *m, int cpu, struct rt_rq *rt_rq) + P(rt_throttled); + PN(rt_time); + PN(rt_runtime); ++#ifdef CONFIG_SMP ++ P(rt_nr_migratory); ++#endif + + #undef PN + #undef P +@@ -953,6 +956,10 @@ void proc_sched_show_task(struct task_struct *p, struct seq_file *m) + #endif + P(policy); + P(prio); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ P(migrate_disable); ++#endif ++ P(nr_cpus_allowed); + #undef PN_SCHEDSTAT + #undef PN + #undef __PN +diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c +index c242944f5cbd..4aeb2e2e41bc 100644 +--- a/kernel/sched/fair.c ++++ b/kernel/sched/fair.c +@@ -3518,7 +3518,7 @@ check_preempt_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr) + ideal_runtime = sched_slice(cfs_rq, curr); + delta_exec = curr->sum_exec_runtime - curr->prev_sum_exec_runtime; + if (delta_exec > ideal_runtime) { +- resched_curr(rq_of(cfs_rq)); ++ resched_curr_lazy(rq_of(cfs_rq)); + /* + * The current task ran long enough, ensure it doesn't get + * re-elected due to buddy favours. +@@ -3542,7 +3542,7 @@ check_preempt_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr) + return; + + if (delta > ideal_runtime) +- resched_curr(rq_of(cfs_rq)); ++ resched_curr_lazy(rq_of(cfs_rq)); + } + + static void +@@ -3684,7 +3684,7 @@ entity_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr, int queued) + * validating it and just reschedule. + */ + if (queued) { +- resched_curr(rq_of(cfs_rq)); ++ resched_curr_lazy(rq_of(cfs_rq)); + return; + } + /* +@@ -3866,7 +3866,7 @@ static void __account_cfs_rq_runtime(struct cfs_rq *cfs_rq, u64 delta_exec) + * hierarchy can be throttled + */ + if (!assign_cfs_rq_runtime(cfs_rq) && likely(cfs_rq->curr)) +- resched_curr(rq_of(cfs_rq)); ++ resched_curr_lazy(rq_of(cfs_rq)); + } + + static __always_inline +@@ -4494,7 +4494,7 @@ static void hrtick_start_fair(struct rq *rq, struct task_struct *p) + + if (delta < 0) { + if (rq->curr == p) +- resched_curr(rq); ++ resched_curr_lazy(rq); + return; + } + hrtick_start(rq, delta); +@@ -5905,7 +5905,7 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_ + return; + + preempt: +- resched_curr(rq); ++ resched_curr_lazy(rq); + /* + * Only set the backward buddy when the current task is still + * on the rq. This can happen when a wakeup gets interleaved +@@ -8631,7 +8631,7 @@ static void task_fork_fair(struct task_struct *p) + * 'current' within the tree based on its new key value. + */ + swap(curr->vruntime, se->vruntime); +- resched_curr(rq); ++ resched_curr_lazy(rq); + } + + se->vruntime -= cfs_rq->min_vruntime; +@@ -8655,7 +8655,7 @@ prio_changed_fair(struct rq *rq, struct task_struct *p, int oldprio) + */ + if (rq->curr == p) { + if (p->prio > oldprio) +- resched_curr(rq); ++ resched_curr_lazy(rq); + } else + check_preempt_curr(rq, p, 0); + } +diff --git a/kernel/sched/features.h b/kernel/sched/features.h +index 69631fa46c2f..6d28fcd08872 100644 +--- a/kernel/sched/features.h ++++ b/kernel/sched/features.h +@@ -45,11 +45,19 @@ SCHED_FEAT(LB_BIAS, true) + */ + SCHED_FEAT(NONTASK_CAPACITY, true) + ++#ifdef CONFIG_PREEMPT_RT_FULL ++SCHED_FEAT(TTWU_QUEUE, false) ++# ifdef CONFIG_PREEMPT_LAZY ++SCHED_FEAT(PREEMPT_LAZY, true) ++# endif ++#else ++ + /* + * Queue remote wakeups on the target CPU and process them + * using the scheduler IPI. Reduces rq->lock contention/bounces. + */ + SCHED_FEAT(TTWU_QUEUE, true) ++#endif + + #ifdef HAVE_RT_PUSH_IPI + /* +diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c +index f139f22ce30d..b0691f4e7d49 100644 +--- a/kernel/sched/rt.c ++++ b/kernel/sched/rt.c +@@ -47,6 +47,7 @@ void init_rt_bandwidth(struct rt_bandwidth *rt_b, u64 period, u64 runtime) + + hrtimer_init(&rt_b->rt_period_timer, + CLOCK_MONOTONIC, HRTIMER_MODE_REL); ++ rt_b->rt_period_timer.irqsafe = 1; + rt_b->rt_period_timer.function = sched_rt_period_timer; + } + +@@ -101,6 +102,7 @@ void init_rt_rq(struct rt_rq *rt_rq) + rt_rq->push_cpu = nr_cpu_ids; + raw_spin_lock_init(&rt_rq->push_lock); + init_irq_work(&rt_rq->push_work, push_irq_work_func); ++ rt_rq->push_work.flags |= IRQ_WORK_HARD_IRQ; + #endif + #endif /* CONFIG_SMP */ + /* We start is dequeued state, because no RT tasks are queued */ +diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h +index 055f935d4421..19324ac27026 100644 +--- a/kernel/sched/sched.h ++++ b/kernel/sched/sched.h +@@ -1163,6 +1163,7 @@ static inline void finish_lock_switch(struct rq *rq, struct task_struct *prev) + #define WF_SYNC 0x01 /* waker goes to sleep after wakeup */ + #define WF_FORK 0x02 /* child wakeup after fork */ + #define WF_MIGRATED 0x4 /* internal use, task got migrated */ ++#define WF_LOCK_SLEEPER 0x08 /* wakeup spinlock "sleeper" */ + + /* + * To aid in avoiding the subversion of "niceness" due to uneven distribution +@@ -1346,6 +1347,15 @@ extern void init_sched_fair_class(void); + extern void resched_curr(struct rq *rq); + extern void resched_cpu(int cpu); + ++#ifdef CONFIG_PREEMPT_LAZY ++extern void resched_curr_lazy(struct rq *rq); ++#else ++static inline void resched_curr_lazy(struct rq *rq) ++{ ++ resched_curr(rq); ++} ++#endif ++ + extern struct rt_bandwidth def_rt_bandwidth; + extern void init_rt_bandwidth(struct rt_bandwidth *rt_b, u64 period, u64 runtime); + +diff --git a/kernel/sched/swait.c b/kernel/sched/swait.c +index 82f0dff90030..ef027ff3250a 100644 +--- a/kernel/sched/swait.c ++++ b/kernel/sched/swait.c +@@ -1,5 +1,6 @@ + #include + #include ++#include + + void __init_swait_queue_head(struct swait_queue_head *q, const char *name, + struct lock_class_key *key) +@@ -29,6 +30,25 @@ void swake_up_locked(struct swait_queue_head *q) + } + EXPORT_SYMBOL(swake_up_locked); + ++void swake_up_all_locked(struct swait_queue_head *q) ++{ ++ struct swait_queue *curr; ++ int wakes = 0; ++ ++ while (!list_empty(&q->task_list)) { ++ ++ curr = list_first_entry(&q->task_list, typeof(*curr), ++ task_list); ++ wake_up_process(curr->task); ++ list_del_init(&curr->task_list); ++ wakes++; ++ } ++ if (pm_in_action) ++ return; ++ WARN(wakes > 2, "complete_all() with %d waiters\n", wakes); ++} ++EXPORT_SYMBOL(swake_up_all_locked); ++ + void swake_up(struct swait_queue_head *q) + { + unsigned long flags; +@@ -54,6 +74,7 @@ void swake_up_all(struct swait_queue_head *q) + if (!swait_active(q)) + return; + ++ WARN_ON(irqs_disabled()); + raw_spin_lock_irq(&q->lock); + list_splice_init(&q->task_list, &tmp); + while (!list_empty(&tmp)) { +diff --git a/kernel/sched/swork.c b/kernel/sched/swork.c +new file mode 100644 +index 000000000000..1950f40ca725 +--- /dev/null ++++ b/kernel/sched/swork.c +@@ -0,0 +1,173 @@ ++/* ++ * Copyright (C) 2014 BMW Car IT GmbH, Daniel Wagner daniel.wagner@bmw-carit.de ++ * ++ * Provides a framework for enqueuing callbacks from irq context ++ * PREEMPT_RT_FULL safe. The callbacks are executed in kthread context. ++ */ ++ ++#include ++#include ++#include ++#include ++#include ++#include ++ ++#define SWORK_EVENT_PENDING (1 << 0) ++ ++static DEFINE_MUTEX(worker_mutex); ++static struct sworker *glob_worker; ++ ++struct sworker { ++ struct list_head events; ++ struct swait_queue_head wq; ++ ++ raw_spinlock_t lock; ++ ++ struct task_struct *task; ++ int refs; ++}; ++ ++static bool swork_readable(struct sworker *worker) ++{ ++ bool r; ++ ++ if (kthread_should_stop()) ++ return true; ++ ++ raw_spin_lock_irq(&worker->lock); ++ r = !list_empty(&worker->events); ++ raw_spin_unlock_irq(&worker->lock); ++ ++ return r; ++} ++ ++static int swork_kthread(void *arg) ++{ ++ struct sworker *worker = arg; ++ ++ for (;;) { ++ swait_event_interruptible(worker->wq, ++ swork_readable(worker)); ++ if (kthread_should_stop()) ++ break; ++ ++ raw_spin_lock_irq(&worker->lock); ++ while (!list_empty(&worker->events)) { ++ struct swork_event *sev; ++ ++ sev = list_first_entry(&worker->events, ++ struct swork_event, item); ++ list_del(&sev->item); ++ raw_spin_unlock_irq(&worker->lock); ++ ++ WARN_ON_ONCE(!test_and_clear_bit(SWORK_EVENT_PENDING, ++ &sev->flags)); ++ sev->func(sev); ++ raw_spin_lock_irq(&worker->lock); ++ } ++ raw_spin_unlock_irq(&worker->lock); ++ } ++ return 0; ++} ++ ++static struct sworker *swork_create(void) ++{ ++ struct sworker *worker; ++ ++ worker = kzalloc(sizeof(*worker), GFP_KERNEL); ++ if (!worker) ++ return ERR_PTR(-ENOMEM); ++ ++ INIT_LIST_HEAD(&worker->events); ++ raw_spin_lock_init(&worker->lock); ++ init_swait_queue_head(&worker->wq); ++ ++ worker->task = kthread_run(swork_kthread, worker, "kswork"); ++ if (IS_ERR(worker->task)) { ++ kfree(worker); ++ return ERR_PTR(-ENOMEM); ++ } ++ ++ return worker; ++} ++ ++static void swork_destroy(struct sworker *worker) ++{ ++ kthread_stop(worker->task); ++ ++ WARN_ON(!list_empty(&worker->events)); ++ kfree(worker); ++} ++ ++/** ++ * swork_queue - queue swork ++ * ++ * Returns %false if @work was already on a queue, %true otherwise. ++ * ++ * The work is queued and processed on a random CPU ++ */ ++bool swork_queue(struct swork_event *sev) ++{ ++ unsigned long flags; ++ ++ if (test_and_set_bit(SWORK_EVENT_PENDING, &sev->flags)) ++ return false; ++ ++ raw_spin_lock_irqsave(&glob_worker->lock, flags); ++ list_add_tail(&sev->item, &glob_worker->events); ++ raw_spin_unlock_irqrestore(&glob_worker->lock, flags); ++ ++ swake_up(&glob_worker->wq); ++ return true; ++} ++EXPORT_SYMBOL_GPL(swork_queue); ++ ++/** ++ * swork_get - get an instance of the sworker ++ * ++ * Returns an negative error code if the initialization if the worker did not ++ * work, %0 otherwise. ++ * ++ */ ++int swork_get(void) ++{ ++ struct sworker *worker; ++ ++ mutex_lock(&worker_mutex); ++ if (!glob_worker) { ++ worker = swork_create(); ++ if (IS_ERR(worker)) { ++ mutex_unlock(&worker_mutex); ++ return -ENOMEM; ++ } ++ ++ glob_worker = worker; ++ } ++ ++ glob_worker->refs++; ++ mutex_unlock(&worker_mutex); ++ ++ return 0; ++} ++EXPORT_SYMBOL_GPL(swork_get); ++ ++/** ++ * swork_put - puts an instance of the sworker ++ * ++ * Will destroy the sworker thread. This function must not be called until all ++ * queued events have been completed. ++ */ ++void swork_put(void) ++{ ++ mutex_lock(&worker_mutex); ++ ++ glob_worker->refs--; ++ if (glob_worker->refs > 0) ++ goto out; ++ ++ swork_destroy(glob_worker); ++ glob_worker = NULL; ++out: ++ mutex_unlock(&worker_mutex); ++} ++EXPORT_SYMBOL_GPL(swork_put); +diff --git a/kernel/signal.c b/kernel/signal.c +index 0b1415720a15..c884647951f7 100644 +--- a/kernel/signal.c ++++ b/kernel/signal.c +@@ -14,6 +14,7 @@ + #include + #include + #include ++#include + #include + #include + #include +@@ -352,13 +353,30 @@ static bool task_participate_group_stop(struct task_struct *task) + return false; + } + ++static inline struct sigqueue *get_task_cache(struct task_struct *t) ++{ ++ struct sigqueue *q = t->sigqueue_cache; ++ ++ if (cmpxchg(&t->sigqueue_cache, q, NULL) != q) ++ return NULL; ++ return q; ++} ++ ++static inline int put_task_cache(struct task_struct *t, struct sigqueue *q) ++{ ++ if (cmpxchg(&t->sigqueue_cache, NULL, q) == NULL) ++ return 0; ++ return 1; ++} ++ + /* + * allocate a new signal queue record + * - this may be called without locks if and only if t == current, otherwise an + * appropriate lock must be held to stop the target task from exiting + */ + static struct sigqueue * +-__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimit) ++__sigqueue_do_alloc(int sig, struct task_struct *t, gfp_t flags, ++ int override_rlimit, int fromslab) + { + struct sigqueue *q = NULL; + struct user_struct *user; +@@ -375,7 +393,10 @@ __sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimi + if (override_rlimit || + atomic_read(&user->sigpending) <= + task_rlimit(t, RLIMIT_SIGPENDING)) { +- q = kmem_cache_alloc(sigqueue_cachep, flags); ++ if (!fromslab) ++ q = get_task_cache(t); ++ if (!q) ++ q = kmem_cache_alloc(sigqueue_cachep, flags); + } else { + print_dropped_signal(sig); + } +@@ -392,6 +413,13 @@ __sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimi + return q; + } + ++static struct sigqueue * ++__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, ++ int override_rlimit) ++{ ++ return __sigqueue_do_alloc(sig, t, flags, override_rlimit, 0); ++} ++ + static void __sigqueue_free(struct sigqueue *q) + { + if (q->flags & SIGQUEUE_PREALLOC) +@@ -401,6 +429,21 @@ static void __sigqueue_free(struct sigqueue *q) + kmem_cache_free(sigqueue_cachep, q); + } + ++static void sigqueue_free_current(struct sigqueue *q) ++{ ++ struct user_struct *up; ++ ++ if (q->flags & SIGQUEUE_PREALLOC) ++ return; ++ ++ up = q->user; ++ if (rt_prio(current->normal_prio) && !put_task_cache(current, q)) { ++ atomic_dec(&up->sigpending); ++ free_uid(up); ++ } else ++ __sigqueue_free(q); ++} ++ + void flush_sigqueue(struct sigpending *queue) + { + struct sigqueue *q; +@@ -414,6 +457,21 @@ void flush_sigqueue(struct sigpending *queue) + } + + /* ++ * Called from __exit_signal. Flush tsk->pending and ++ * tsk->sigqueue_cache ++ */ ++void flush_task_sigqueue(struct task_struct *tsk) ++{ ++ struct sigqueue *q; ++ ++ flush_sigqueue(&tsk->pending); ++ ++ q = get_task_cache(tsk); ++ if (q) ++ kmem_cache_free(sigqueue_cachep, q); ++} ++ ++/* + * Flush all pending signals for this kthread. + */ + void flush_signals(struct task_struct *t) +@@ -525,7 +583,7 @@ static void collect_signal(int sig, struct sigpending *list, siginfo_t *info) + still_pending: + list_del_init(&first->list); + copy_siginfo(info, &first->info); +- __sigqueue_free(first); ++ sigqueue_free_current(first); + } else { + /* + * Ok, it wasn't in the queue. This must be +@@ -560,6 +618,8 @@ int dequeue_signal(struct task_struct *tsk, sigset_t *mask, siginfo_t *info) + { + int signr; + ++ WARN_ON_ONCE(tsk != current); ++ + /* We only dequeue private signals from ourselves, we don't let + * signalfd steal them + */ +@@ -1156,8 +1216,8 @@ int do_send_sig_info(int sig, struct siginfo *info, struct task_struct *p, + * We don't want to have recursive SIGSEGV's etc, for example, + * that is why we also clear SIGNAL_UNKILLABLE. + */ +-int +-force_sig_info(int sig, struct siginfo *info, struct task_struct *t) ++static int ++do_force_sig_info(int sig, struct siginfo *info, struct task_struct *t) + { + unsigned long int flags; + int ret, blocked, ignored; +@@ -1182,6 +1242,39 @@ force_sig_info(int sig, struct siginfo *info, struct task_struct *t) + return ret; + } + ++int force_sig_info(int sig, struct siginfo *info, struct task_struct *t) ++{ ++/* ++ * On some archs, PREEMPT_RT has to delay sending a signal from a trap ++ * since it can not enable preemption, and the signal code's spin_locks ++ * turn into mutexes. Instead, it must set TIF_NOTIFY_RESUME which will ++ * send the signal on exit of the trap. ++ */ ++#ifdef ARCH_RT_DELAYS_SIGNAL_SEND ++ if (in_atomic()) { ++ if (WARN_ON_ONCE(t != current)) ++ return 0; ++ if (WARN_ON_ONCE(t->forced_info.si_signo)) ++ return 0; ++ ++ if (is_si_special(info)) { ++ WARN_ON_ONCE(info != SEND_SIG_PRIV); ++ t->forced_info.si_signo = sig; ++ t->forced_info.si_errno = 0; ++ t->forced_info.si_code = SI_KERNEL; ++ t->forced_info.si_pid = 0; ++ t->forced_info.si_uid = 0; ++ } else { ++ t->forced_info = *info; ++ } ++ ++ set_tsk_thread_flag(t, TIF_NOTIFY_RESUME); ++ return 0; ++ } ++#endif ++ return do_force_sig_info(sig, info, t); ++} ++ + /* + * Nuke all other threads in the group. + */ +@@ -1216,12 +1309,12 @@ struct sighand_struct *__lock_task_sighand(struct task_struct *tsk, + * Disable interrupts early to avoid deadlocks. + * See rcu_read_unlock() comment header for details. + */ +- local_irq_save(*flags); ++ local_irq_save_nort(*flags); + rcu_read_lock(); + sighand = rcu_dereference(tsk->sighand); + if (unlikely(sighand == NULL)) { + rcu_read_unlock(); +- local_irq_restore(*flags); ++ local_irq_restore_nort(*flags); + break; + } + /* +@@ -1242,7 +1335,7 @@ struct sighand_struct *__lock_task_sighand(struct task_struct *tsk, + } + spin_unlock(&sighand->siglock); + rcu_read_unlock(); +- local_irq_restore(*flags); ++ local_irq_restore_nort(*flags); + } + + return sighand; +@@ -1485,7 +1578,8 @@ EXPORT_SYMBOL(kill_pid); + */ + struct sigqueue *sigqueue_alloc(void) + { +- struct sigqueue *q = __sigqueue_alloc(-1, current, GFP_KERNEL, 0); ++ /* Preallocated sigqueue objects always from the slabcache ! */ ++ struct sigqueue *q = __sigqueue_do_alloc(-1, current, GFP_KERNEL, 0, 1); + + if (q) + q->flags |= SIGQUEUE_PREALLOC; +@@ -1846,15 +1940,7 @@ static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info) + if (gstop_done && ptrace_reparented(current)) + do_notify_parent_cldstop(current, false, why); + +- /* +- * Don't want to allow preemption here, because +- * sys_ptrace() needs this task to be inactive. +- * +- * XXX: implement read_unlock_no_resched(). +- */ +- preempt_disable(); + read_unlock(&tasklist_lock); +- preempt_enable_no_resched(); + freezable_schedule(); + } else { + /* +diff --git a/kernel/softirq.c b/kernel/softirq.c +index 744fa611cae0..819bd7cf5ad0 100644 +--- a/kernel/softirq.c ++++ b/kernel/softirq.c +@@ -21,10 +21,12 @@ + #include + #include + #include ++#include + #include + #include + #include + #include ++#include + #include + + #define CREATE_TRACE_POINTS +@@ -56,12 +58,108 @@ EXPORT_SYMBOL(irq_stat); + static struct softirq_action softirq_vec[NR_SOFTIRQS] __cacheline_aligned_in_smp; + + DEFINE_PER_CPU(struct task_struct *, ksoftirqd); ++#ifdef CONFIG_PREEMPT_RT_FULL ++#define TIMER_SOFTIRQS ((1 << TIMER_SOFTIRQ) | (1 << HRTIMER_SOFTIRQ)) ++DEFINE_PER_CPU(struct task_struct *, ktimer_softirqd); ++#endif + + const char * const softirq_to_name[NR_SOFTIRQS] = { + "HI", "TIMER", "NET_TX", "NET_RX", "BLOCK", "IRQ_POLL", + "TASKLET", "SCHED", "HRTIMER", "RCU" + }; + ++#ifdef CONFIG_NO_HZ_COMMON ++# ifdef CONFIG_PREEMPT_RT_FULL ++ ++struct softirq_runner { ++ struct task_struct *runner[NR_SOFTIRQS]; ++}; ++ ++static DEFINE_PER_CPU(struct softirq_runner, softirq_runners); ++ ++static inline void softirq_set_runner(unsigned int sirq) ++{ ++ struct softirq_runner *sr = this_cpu_ptr(&softirq_runners); ++ ++ sr->runner[sirq] = current; ++} ++ ++static inline void softirq_clr_runner(unsigned int sirq) ++{ ++ struct softirq_runner *sr = this_cpu_ptr(&softirq_runners); ++ ++ sr->runner[sirq] = NULL; ++} ++ ++/* ++ * On preempt-rt a softirq running context might be blocked on a ++ * lock. There might be no other runnable task on this CPU because the ++ * lock owner runs on some other CPU. So we have to go into idle with ++ * the pending bit set. Therefor we need to check this otherwise we ++ * warn about false positives which confuses users and defeats the ++ * whole purpose of this test. ++ * ++ * This code is called with interrupts disabled. ++ */ ++void softirq_check_pending_idle(void) ++{ ++ static int rate_limit; ++ struct softirq_runner *sr = this_cpu_ptr(&softirq_runners); ++ u32 warnpending; ++ int i; ++ ++ if (rate_limit >= 10) ++ return; ++ ++ warnpending = local_softirq_pending() & SOFTIRQ_STOP_IDLE_MASK; ++ for (i = 0; i < NR_SOFTIRQS; i++) { ++ struct task_struct *tsk = sr->runner[i]; ++ ++ /* ++ * The wakeup code in rtmutex.c wakes up the task ++ * _before_ it sets pi_blocked_on to NULL under ++ * tsk->pi_lock. So we need to check for both: state ++ * and pi_blocked_on. ++ */ ++ if (tsk) { ++ raw_spin_lock(&tsk->pi_lock); ++ if (tsk->pi_blocked_on || tsk->state == TASK_RUNNING) { ++ /* Clear all bits pending in that task */ ++ warnpending &= ~(tsk->softirqs_raised); ++ warnpending &= ~(1 << i); ++ } ++ raw_spin_unlock(&tsk->pi_lock); ++ } ++ } ++ ++ if (warnpending) { ++ printk(KERN_ERR "NOHZ: local_softirq_pending %02x\n", ++ warnpending); ++ rate_limit++; ++ } ++} ++# else ++/* ++ * On !PREEMPT_RT we just printk rate limited: ++ */ ++void softirq_check_pending_idle(void) ++{ ++ static int rate_limit; ++ ++ if (rate_limit < 10 && ++ (local_softirq_pending() & SOFTIRQ_STOP_IDLE_MASK)) { ++ printk(KERN_ERR "NOHZ: local_softirq_pending %02x\n", ++ local_softirq_pending()); ++ rate_limit++; ++ } ++} ++# endif ++ ++#else /* !CONFIG_NO_HZ_COMMON */ ++static inline void softirq_set_runner(unsigned int sirq) { } ++static inline void softirq_clr_runner(unsigned int sirq) { } ++#endif ++ + /* + * we cannot loop indefinitely here to avoid userspace starvation, + * but we also don't want to introduce a worst case 1/HZ latency +@@ -77,6 +175,38 @@ static void wakeup_softirqd(void) + wake_up_process(tsk); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++static void wakeup_timer_softirqd(void) ++{ ++ /* Interrupts are disabled: no need to stop preemption */ ++ struct task_struct *tsk = __this_cpu_read(ktimer_softirqd); ++ ++ if (tsk && tsk->state != TASK_RUNNING) ++ wake_up_process(tsk); ++} ++#endif ++ ++static void handle_softirq(unsigned int vec_nr) ++{ ++ struct softirq_action *h = softirq_vec + vec_nr; ++ int prev_count; ++ ++ prev_count = preempt_count(); ++ ++ kstat_incr_softirqs_this_cpu(vec_nr); ++ ++ trace_softirq_entry(vec_nr); ++ h->action(h); ++ trace_softirq_exit(vec_nr); ++ if (unlikely(prev_count != preempt_count())) { ++ pr_err("huh, entered softirq %u %s %p with preempt_count %08x, exited with %08x?\n", ++ vec_nr, softirq_to_name[vec_nr], h->action, ++ prev_count, preempt_count()); ++ preempt_count_set(prev_count); ++ } ++} ++ ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * If ksoftirqd is scheduled, we do not want to process pending softirqs + * right now. Let ksoftirqd handle this at its own rate, to get fairness. +@@ -88,6 +218,47 @@ static bool ksoftirqd_running(void) + return tsk && (tsk->state == TASK_RUNNING); + } + ++static inline int ksoftirqd_softirq_pending(void) ++{ ++ return local_softirq_pending(); ++} ++ ++static void handle_pending_softirqs(u32 pending) ++{ ++ struct softirq_action *h = softirq_vec; ++ int softirq_bit; ++ ++ local_irq_enable(); ++ ++ h = softirq_vec; ++ ++ while ((softirq_bit = ffs(pending))) { ++ unsigned int vec_nr; ++ ++ h += softirq_bit - 1; ++ vec_nr = h - softirq_vec; ++ handle_softirq(vec_nr); ++ ++ h++; ++ pending >>= softirq_bit; ++ } ++ ++ rcu_bh_qs(); ++ local_irq_disable(); ++} ++ ++static void run_ksoftirqd(unsigned int cpu) ++{ ++ local_irq_disable(); ++ if (ksoftirqd_softirq_pending()) { ++ __do_softirq(); ++ local_irq_enable(); ++ cond_resched_rcu_qs(); ++ return; ++ } ++ local_irq_enable(); ++} ++ + /* + * preempt_count and SOFTIRQ_OFFSET usage: + * - preempt_count is changed by SOFTIRQ_OFFSET on entering or leaving +@@ -243,10 +414,8 @@ asmlinkage __visible void __softirq_entry __do_softirq(void) + unsigned long end = jiffies + MAX_SOFTIRQ_TIME; + unsigned long old_flags = current->flags; + int max_restart = MAX_SOFTIRQ_RESTART; +- struct softirq_action *h; + bool in_hardirq; + __u32 pending; +- int softirq_bit; + + /* + * Mask out PF_MEMALLOC s current task context is borrowed for the +@@ -265,36 +434,7 @@ asmlinkage __visible void __softirq_entry __do_softirq(void) + /* Reset the pending bitmask before enabling irqs */ + set_softirq_pending(0); + +- local_irq_enable(); +- +- h = softirq_vec; +- +- while ((softirq_bit = ffs(pending))) { +- unsigned int vec_nr; +- int prev_count; +- +- h += softirq_bit - 1; +- +- vec_nr = h - softirq_vec; +- prev_count = preempt_count(); +- +- kstat_incr_softirqs_this_cpu(vec_nr); +- +- trace_softirq_entry(vec_nr); +- h->action(h); +- trace_softirq_exit(vec_nr); +- if (unlikely(prev_count != preempt_count())) { +- pr_err("huh, entered softirq %u %s %p with preempt_count %08x, exited with %08x?\n", +- vec_nr, softirq_to_name[vec_nr], h->action, +- prev_count, preempt_count()); +- preempt_count_set(prev_count); +- } +- h++; +- pending >>= softirq_bit; +- } +- +- rcu_bh_qs(); +- local_irq_disable(); ++ handle_pending_softirqs(pending); + + pending = local_softirq_pending(); + if (pending) { +@@ -331,6 +471,309 @@ asmlinkage __visible void do_softirq(void) + } + + /* ++ * This function must run with irqs disabled! ++ */ ++void raise_softirq_irqoff(unsigned int nr) ++{ ++ __raise_softirq_irqoff(nr); ++ ++ /* ++ * If we're in an interrupt or softirq, we're done ++ * (this also catches softirq-disabled code). We will ++ * actually run the softirq once we return from ++ * the irq or softirq. ++ * ++ * Otherwise we wake up ksoftirqd to make sure we ++ * schedule the softirq soon. ++ */ ++ if (!in_interrupt()) ++ wakeup_softirqd(); ++} ++ ++void __raise_softirq_irqoff(unsigned int nr) ++{ ++ trace_softirq_raise(nr); ++ or_softirq_pending(1UL << nr); ++} ++ ++static inline void local_bh_disable_nort(void) { local_bh_disable(); } ++static inline void _local_bh_enable_nort(void) { _local_bh_enable(); } ++static void ksoftirqd_set_sched_params(unsigned int cpu) { } ++ ++#else /* !PREEMPT_RT_FULL */ ++ ++/* ++ * On RT we serialize softirq execution with a cpu local lock per softirq ++ */ ++static DEFINE_PER_CPU(struct local_irq_lock [NR_SOFTIRQS], local_softirq_locks); ++ ++void __init softirq_early_init(void) ++{ ++ int i; ++ ++ for (i = 0; i < NR_SOFTIRQS; i++) ++ local_irq_lock_init(local_softirq_locks[i]); ++} ++ ++static void lock_softirq(int which) ++{ ++ local_lock(local_softirq_locks[which]); ++} ++ ++static void unlock_softirq(int which) ++{ ++ local_unlock(local_softirq_locks[which]); ++} ++ ++static void do_single_softirq(int which) ++{ ++ unsigned long old_flags = current->flags; ++ ++ current->flags &= ~PF_MEMALLOC; ++ vtime_account_irq_enter(current); ++ current->flags |= PF_IN_SOFTIRQ; ++ lockdep_softirq_enter(); ++ local_irq_enable(); ++ handle_softirq(which); ++ local_irq_disable(); ++ lockdep_softirq_exit(); ++ current->flags &= ~PF_IN_SOFTIRQ; ++ vtime_account_irq_enter(current); ++ tsk_restore_flags(current, old_flags, PF_MEMALLOC); ++} ++ ++/* ++ * Called with interrupts disabled. Process softirqs which were raised ++ * in current context (or on behalf of ksoftirqd). ++ */ ++static void do_current_softirqs(void) ++{ ++ while (current->softirqs_raised) { ++ int i = __ffs(current->softirqs_raised); ++ unsigned int pending, mask = (1U << i); ++ ++ current->softirqs_raised &= ~mask; ++ local_irq_enable(); ++ ++ /* ++ * If the lock is contended, we boost the owner to ++ * process the softirq or leave the critical section ++ * now. ++ */ ++ lock_softirq(i); ++ local_irq_disable(); ++ softirq_set_runner(i); ++ /* ++ * Check with the local_softirq_pending() bits, ++ * whether we need to process this still or if someone ++ * else took care of it. ++ */ ++ pending = local_softirq_pending(); ++ if (pending & mask) { ++ set_softirq_pending(pending & ~mask); ++ do_single_softirq(i); ++ } ++ softirq_clr_runner(i); ++ WARN_ON(current->softirq_nestcnt != 1); ++ local_irq_enable(); ++ unlock_softirq(i); ++ local_irq_disable(); ++ } ++} ++ ++void __local_bh_disable(void) ++{ ++ if (++current->softirq_nestcnt == 1) ++ migrate_disable(); ++} ++EXPORT_SYMBOL(__local_bh_disable); ++ ++void __local_bh_enable(void) ++{ ++ if (WARN_ON(current->softirq_nestcnt == 0)) ++ return; ++ ++ local_irq_disable(); ++ if (current->softirq_nestcnt == 1 && current->softirqs_raised) ++ do_current_softirqs(); ++ local_irq_enable(); ++ ++ if (--current->softirq_nestcnt == 0) ++ migrate_enable(); ++} ++EXPORT_SYMBOL(__local_bh_enable); ++ ++void _local_bh_enable(void) ++{ ++ if (WARN_ON(current->softirq_nestcnt == 0)) ++ return; ++ if (--current->softirq_nestcnt == 0) ++ migrate_enable(); ++} ++EXPORT_SYMBOL(_local_bh_enable); ++ ++int in_serving_softirq(void) ++{ ++ return current->flags & PF_IN_SOFTIRQ; ++} ++EXPORT_SYMBOL(in_serving_softirq); ++ ++/* Called with preemption disabled */ ++static void run_ksoftirqd(unsigned int cpu) ++{ ++ local_irq_disable(); ++ current->softirq_nestcnt++; ++ ++ do_current_softirqs(); ++ current->softirq_nestcnt--; ++ local_irq_enable(); ++ cond_resched_rcu_qs(); ++} ++ ++/* ++ * Called from netif_rx_ni(). Preemption enabled, but migration ++ * disabled. So the cpu can't go away under us. ++ */ ++void thread_do_softirq(void) ++{ ++ if (!in_serving_softirq() && current->softirqs_raised) { ++ current->softirq_nestcnt++; ++ do_current_softirqs(); ++ current->softirq_nestcnt--; ++ } ++} ++ ++static void do_raise_softirq_irqoff(unsigned int nr) ++{ ++ unsigned int mask; ++ ++ mask = 1UL << nr; ++ ++ trace_softirq_raise(nr); ++ or_softirq_pending(mask); ++ ++ /* ++ * If we are not in a hard interrupt and inside a bh disabled ++ * region, we simply raise the flag on current. local_bh_enable() ++ * will make sure that the softirq is executed. Otherwise we ++ * delegate it to ksoftirqd. ++ */ ++ if (!in_irq() && current->softirq_nestcnt) ++ current->softirqs_raised |= mask; ++ else if (!__this_cpu_read(ksoftirqd) || !__this_cpu_read(ktimer_softirqd)) ++ return; ++ ++ if (mask & TIMER_SOFTIRQS) ++ __this_cpu_read(ktimer_softirqd)->softirqs_raised |= mask; ++ else ++ __this_cpu_read(ksoftirqd)->softirqs_raised |= mask; ++} ++ ++static void wakeup_proper_softirq(unsigned int nr) ++{ ++ if ((1UL << nr) & TIMER_SOFTIRQS) ++ wakeup_timer_softirqd(); ++ else ++ wakeup_softirqd(); ++} ++ ++void __raise_softirq_irqoff(unsigned int nr) ++{ ++ do_raise_softirq_irqoff(nr); ++ if (!in_irq() && !current->softirq_nestcnt) ++ wakeup_proper_softirq(nr); ++} ++ ++/* ++ * Same as __raise_softirq_irqoff() but will process them in ksoftirqd ++ */ ++void __raise_softirq_irqoff_ksoft(unsigned int nr) ++{ ++ unsigned int mask; ++ ++ if (WARN_ON_ONCE(!__this_cpu_read(ksoftirqd) || ++ !__this_cpu_read(ktimer_softirqd))) ++ return; ++ mask = 1UL << nr; ++ ++ trace_softirq_raise(nr); ++ or_softirq_pending(mask); ++ if (mask & TIMER_SOFTIRQS) ++ __this_cpu_read(ktimer_softirqd)->softirqs_raised |= mask; ++ else ++ __this_cpu_read(ksoftirqd)->softirqs_raised |= mask; ++ wakeup_proper_softirq(nr); ++} ++ ++/* ++ * This function must run with irqs disabled! ++ */ ++void raise_softirq_irqoff(unsigned int nr) ++{ ++ do_raise_softirq_irqoff(nr); ++ ++ /* ++ * If we're in an hard interrupt we let irq return code deal ++ * with the wakeup of ksoftirqd. ++ */ ++ if (in_irq()) ++ return; ++ /* ++ * If we are in thread context but outside of a bh disabled ++ * region, we need to wake ksoftirqd as well. ++ * ++ * CHECKME: Some of the places which do that could be wrapped ++ * into local_bh_disable/enable pairs. Though it's unclear ++ * whether this is worth the effort. To find those places just ++ * raise a WARN() if the condition is met. ++ */ ++ if (!current->softirq_nestcnt) ++ wakeup_proper_softirq(nr); ++} ++ ++static inline int ksoftirqd_softirq_pending(void) ++{ ++ return current->softirqs_raised; ++} ++ ++static inline void local_bh_disable_nort(void) { } ++static inline void _local_bh_enable_nort(void) { } ++ ++static inline void ksoftirqd_set_sched_params(unsigned int cpu) ++{ ++ /* Take over all but timer pending softirqs when starting */ ++ local_irq_disable(); ++ current->softirqs_raised = local_softirq_pending() & ~TIMER_SOFTIRQS; ++ local_irq_enable(); ++} ++ ++static inline void ktimer_softirqd_set_sched_params(unsigned int cpu) ++{ ++ struct sched_param param = { .sched_priority = 1 }; ++ ++ sched_setscheduler(current, SCHED_FIFO, ¶m); ++ ++ /* Take over timer pending softirqs when starting */ ++ local_irq_disable(); ++ current->softirqs_raised = local_softirq_pending() & TIMER_SOFTIRQS; ++ local_irq_enable(); ++} ++ ++static inline void ktimer_softirqd_clr_sched_params(unsigned int cpu, ++ bool online) ++{ ++ struct sched_param param = { .sched_priority = 0 }; ++ ++ sched_setscheduler(current, SCHED_NORMAL, ¶m); ++} ++ ++static int ktimer_softirqd_should_run(unsigned int cpu) ++{ ++ return current->softirqs_raised; ++} ++ ++#endif /* PREEMPT_RT_FULL */ ++/* + * Enter an interrupt context. + */ + void irq_enter(void) +@@ -341,9 +784,9 @@ void irq_enter(void) + * Prevent raise_softirq from needlessly waking up ksoftirqd + * here, as softirq will be serviced on return from interrupt. + */ +- local_bh_disable(); ++ local_bh_disable_nort(); + tick_irq_enter(); +- _local_bh_enable(); ++ _local_bh_enable_nort(); + } + + __irq_enter(); +@@ -351,6 +794,7 @@ void irq_enter(void) + + static inline void invoke_softirq(void) + { ++#ifndef CONFIG_PREEMPT_RT_FULL + if (ksoftirqd_running()) + return; + +@@ -373,6 +817,18 @@ static inline void invoke_softirq(void) + } else { + wakeup_softirqd(); + } ++#else /* PREEMPT_RT_FULL */ ++ unsigned long flags; ++ ++ local_irq_save(flags); ++ if (__this_cpu_read(ksoftirqd) && ++ __this_cpu_read(ksoftirqd)->softirqs_raised) ++ wakeup_softirqd(); ++ if (__this_cpu_read(ktimer_softirqd) && ++ __this_cpu_read(ktimer_softirqd)->softirqs_raised) ++ wakeup_timer_softirqd(); ++ local_irq_restore(flags); ++#endif + } + + static inline void tick_irq_exit(void) +@@ -409,26 +865,6 @@ void irq_exit(void) + trace_hardirq_exit(); /* must be last! */ + } + +-/* +- * This function must run with irqs disabled! +- */ +-inline void raise_softirq_irqoff(unsigned int nr) +-{ +- __raise_softirq_irqoff(nr); +- +- /* +- * If we're in an interrupt or softirq, we're done +- * (this also catches softirq-disabled code). We will +- * actually run the softirq once we return from +- * the irq or softirq. +- * +- * Otherwise we wake up ksoftirqd to make sure we +- * schedule the softirq soon. +- */ +- if (!in_interrupt()) +- wakeup_softirqd(); +-} +- + void raise_softirq(unsigned int nr) + { + unsigned long flags; +@@ -438,12 +874,6 @@ void raise_softirq(unsigned int nr) + local_irq_restore(flags); + } + +-void __raise_softirq_irqoff(unsigned int nr) +-{ +- trace_softirq_raise(nr); +- or_softirq_pending(1UL << nr); +-} +- + void open_softirq(int nr, void (*action)(struct softirq_action *)) + { + softirq_vec[nr].action = action; +@@ -460,15 +890,45 @@ struct tasklet_head { + static DEFINE_PER_CPU(struct tasklet_head, tasklet_vec); + static DEFINE_PER_CPU(struct tasklet_head, tasklet_hi_vec); + ++static void inline ++__tasklet_common_schedule(struct tasklet_struct *t, struct tasklet_head *head, unsigned int nr) ++{ ++ if (tasklet_trylock(t)) { ++again: ++ /* We may have been preempted before tasklet_trylock ++ * and __tasklet_action may have already run. ++ * So double check the sched bit while the takslet ++ * is locked before adding it to the list. ++ */ ++ if (test_bit(TASKLET_STATE_SCHED, &t->state)) { ++ t->next = NULL; ++ *head->tail = t; ++ head->tail = &(t->next); ++ raise_softirq_irqoff(nr); ++ tasklet_unlock(t); ++ } else { ++ /* This is subtle. If we hit the corner case above ++ * It is possible that we get preempted right here, ++ * and another task has successfully called ++ * tasklet_schedule(), then this function, and ++ * failed on the trylock. Thus we must be sure ++ * before releasing the tasklet lock, that the ++ * SCHED_BIT is clear. Otherwise the tasklet ++ * may get its SCHED_BIT set, but not added to the ++ * list ++ */ ++ if (!tasklet_tryunlock(t)) ++ goto again; ++ } ++ } ++} ++ + void __tasklet_schedule(struct tasklet_struct *t) + { + unsigned long flags; + + local_irq_save(flags); +- t->next = NULL; +- *__this_cpu_read(tasklet_vec.tail) = t; +- __this_cpu_write(tasklet_vec.tail, &(t->next)); +- raise_softirq_irqoff(TASKLET_SOFTIRQ); ++ __tasklet_common_schedule(t, this_cpu_ptr(&tasklet_vec), TASKLET_SOFTIRQ); + local_irq_restore(flags); + } + EXPORT_SYMBOL(__tasklet_schedule); +@@ -478,10 +938,7 @@ void __tasklet_hi_schedule(struct tasklet_struct *t) + unsigned long flags; + + local_irq_save(flags); +- t->next = NULL; +- *__this_cpu_read(tasklet_hi_vec.tail) = t; +- __this_cpu_write(tasklet_hi_vec.tail, &(t->next)); +- raise_softirq_irqoff(HI_SOFTIRQ); ++ __tasklet_common_schedule(t, this_cpu_ptr(&tasklet_hi_vec), HI_SOFTIRQ); + local_irq_restore(flags); + } + EXPORT_SYMBOL(__tasklet_hi_schedule); +@@ -490,82 +947,122 @@ void __tasklet_hi_schedule_first(struct tasklet_struct *t) + { + BUG_ON(!irqs_disabled()); + +- t->next = __this_cpu_read(tasklet_hi_vec.head); +- __this_cpu_write(tasklet_hi_vec.head, t); +- __raise_softirq_irqoff(HI_SOFTIRQ); ++ __tasklet_hi_schedule(t); + } + EXPORT_SYMBOL(__tasklet_hi_schedule_first); + +-static __latent_entropy void tasklet_action(struct softirq_action *a) ++void tasklet_enable(struct tasklet_struct *t) + { +- struct tasklet_struct *list; ++ if (!atomic_dec_and_test(&t->count)) ++ return; ++ if (test_and_clear_bit(TASKLET_STATE_PENDING, &t->state)) ++ tasklet_schedule(t); ++} ++EXPORT_SYMBOL(tasklet_enable); + +- local_irq_disable(); +- list = __this_cpu_read(tasklet_vec.head); +- __this_cpu_write(tasklet_vec.head, NULL); +- __this_cpu_write(tasklet_vec.tail, this_cpu_ptr(&tasklet_vec.head)); +- local_irq_enable(); ++static void __tasklet_action(struct softirq_action *a, ++ struct tasklet_struct *list) ++{ ++ int loops = 1000000; + + while (list) { + struct tasklet_struct *t = list; + + list = list->next; + +- if (tasklet_trylock(t)) { +- if (!atomic_read(&t->count)) { +- if (!test_and_clear_bit(TASKLET_STATE_SCHED, +- &t->state)) +- BUG(); +- t->func(t->data); +- tasklet_unlock(t); +- continue; +- } +- tasklet_unlock(t); ++ /* ++ * Should always succeed - after a tasklist got on the ++ * list (after getting the SCHED bit set from 0 to 1), ++ * nothing but the tasklet softirq it got queued to can ++ * lock it: ++ */ ++ if (!tasklet_trylock(t)) { ++ WARN_ON(1); ++ continue; + } + +- local_irq_disable(); + t->next = NULL; +- *__this_cpu_read(tasklet_vec.tail) = t; +- __this_cpu_write(tasklet_vec.tail, &(t->next)); +- __raise_softirq_irqoff(TASKLET_SOFTIRQ); +- local_irq_enable(); ++ ++ /* ++ * If we cannot handle the tasklet because it's disabled, ++ * mark it as pending. tasklet_enable() will later ++ * re-schedule the tasklet. ++ */ ++ if (unlikely(atomic_read(&t->count))) { ++out_disabled: ++ /* implicit unlock: */ ++ wmb(); ++ t->state = TASKLET_STATEF_PENDING; ++ continue; ++ } ++ ++ /* ++ * After this point on the tasklet might be rescheduled ++ * on another CPU, but it can only be added to another ++ * CPU's tasklet list if we unlock the tasklet (which we ++ * dont do yet). ++ */ ++ if (!test_and_clear_bit(TASKLET_STATE_SCHED, &t->state)) ++ WARN_ON(1); ++ ++again: ++ t->func(t->data); ++ ++ /* ++ * Try to unlock the tasklet. We must use cmpxchg, because ++ * another CPU might have scheduled or disabled the tasklet. ++ * We only allow the STATE_RUN -> 0 transition here. ++ */ ++ while (!tasklet_tryunlock(t)) { ++ /* ++ * If it got disabled meanwhile, bail out: ++ */ ++ if (atomic_read(&t->count)) ++ goto out_disabled; ++ /* ++ * If it got scheduled meanwhile, re-execute ++ * the tasklet function: ++ */ ++ if (test_and_clear_bit(TASKLET_STATE_SCHED, &t->state)) ++ goto again; ++ if (!--loops) { ++ printk("hm, tasklet state: %08lx\n", t->state); ++ WARN_ON(1); ++ tasklet_unlock(t); ++ break; ++ } ++ } + } + } + ++static void tasklet_action(struct softirq_action *a) ++{ ++ struct tasklet_struct *list; ++ ++ local_irq_disable(); ++ ++ list = __this_cpu_read(tasklet_vec.head); ++ __this_cpu_write(tasklet_vec.head, NULL); ++ __this_cpu_write(tasklet_vec.tail, this_cpu_ptr(&tasklet_vec.head)); ++ ++ local_irq_enable(); ++ ++ __tasklet_action(a, list); ++} ++ + static __latent_entropy void tasklet_hi_action(struct softirq_action *a) + { + struct tasklet_struct *list; + + local_irq_disable(); ++ + list = __this_cpu_read(tasklet_hi_vec.head); + __this_cpu_write(tasklet_hi_vec.head, NULL); + __this_cpu_write(tasklet_hi_vec.tail, this_cpu_ptr(&tasklet_hi_vec.head)); ++ + local_irq_enable(); + +- while (list) { +- struct tasklet_struct *t = list; +- +- list = list->next; +- +- if (tasklet_trylock(t)) { +- if (!atomic_read(&t->count)) { +- if (!test_and_clear_bit(TASKLET_STATE_SCHED, +- &t->state)) +- BUG(); +- t->func(t->data); +- tasklet_unlock(t); +- continue; +- } +- tasklet_unlock(t); +- } +- +- local_irq_disable(); +- t->next = NULL; +- *__this_cpu_read(tasklet_hi_vec.tail) = t; +- __this_cpu_write(tasklet_hi_vec.tail, &(t->next)); +- __raise_softirq_irqoff(HI_SOFTIRQ); +- local_irq_enable(); +- } ++ __tasklet_action(a, list); + } + + void tasklet_init(struct tasklet_struct *t, +@@ -586,7 +1083,7 @@ void tasklet_kill(struct tasklet_struct *t) + + while (test_and_set_bit(TASKLET_STATE_SCHED, &t->state)) { + do { +- yield(); ++ msleep(1); + } while (test_bit(TASKLET_STATE_SCHED, &t->state)); + } + tasklet_unlock_wait(t); +@@ -660,25 +1157,26 @@ void __init softirq_init(void) + open_softirq(HI_SOFTIRQ, tasklet_hi_action); + } + ++#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL) ++void tasklet_unlock_wait(struct tasklet_struct *t) ++{ ++ while (test_bit(TASKLET_STATE_RUN, &(t)->state)) { ++ /* ++ * Hack for now to avoid this busy-loop: ++ */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ msleep(1); ++#else ++ barrier(); ++#endif ++ } ++} ++EXPORT_SYMBOL(tasklet_unlock_wait); ++#endif ++ + static int ksoftirqd_should_run(unsigned int cpu) + { +- return local_softirq_pending(); +-} +- +-static void run_ksoftirqd(unsigned int cpu) +-{ +- local_irq_disable(); +- if (local_softirq_pending()) { +- /* +- * We can safely run softirq on inline stack, as we are not deep +- * in the task stack here. +- */ +- __do_softirq(); +- local_irq_enable(); +- cond_resched_rcu_qs(); +- return; +- } +- local_irq_enable(); ++ return ksoftirqd_softirq_pending(); + } + + #ifdef CONFIG_HOTPLUG_CPU +@@ -745,17 +1243,31 @@ static int takeover_tasklets(unsigned int cpu) + + static struct smp_hotplug_thread softirq_threads = { + .store = &ksoftirqd, ++ .setup = ksoftirqd_set_sched_params, + .thread_should_run = ksoftirqd_should_run, + .thread_fn = run_ksoftirqd, + .thread_comm = "ksoftirqd/%u", + }; + ++#ifdef CONFIG_PREEMPT_RT_FULL ++static struct smp_hotplug_thread softirq_timer_threads = { ++ .store = &ktimer_softirqd, ++ .setup = ktimer_softirqd_set_sched_params, ++ .cleanup = ktimer_softirqd_clr_sched_params, ++ .thread_should_run = ktimer_softirqd_should_run, ++ .thread_fn = run_ksoftirqd, ++ .thread_comm = "ktimersoftd/%u", ++}; ++#endif ++ + static __init int spawn_ksoftirqd(void) + { + cpuhp_setup_state_nocalls(CPUHP_SOFTIRQ_DEAD, "softirq:dead", NULL, + takeover_tasklets); + BUG_ON(smpboot_register_percpu_thread(&softirq_threads)); +- ++#ifdef CONFIG_PREEMPT_RT_FULL ++ BUG_ON(smpboot_register_percpu_thread(&softirq_timer_threads)); ++#endif + return 0; + } + early_initcall(spawn_ksoftirqd); +diff --git a/kernel/stop_machine.c b/kernel/stop_machine.c +index ec9ab2f01489..8b89dbedeaff 100644 +--- a/kernel/stop_machine.c ++++ b/kernel/stop_machine.c +@@ -36,7 +36,7 @@ struct cpu_stop_done { + struct cpu_stopper { + struct task_struct *thread; + +- spinlock_t lock; ++ raw_spinlock_t lock; + bool enabled; /* is this stopper enabled? */ + struct list_head works; /* list of pending works */ + +@@ -78,14 +78,14 @@ static bool cpu_stop_queue_work(unsigned int cpu, struct cpu_stop_work *work) + unsigned long flags; + bool enabled; + +- spin_lock_irqsave(&stopper->lock, flags); ++ raw_spin_lock_irqsave(&stopper->lock, flags); + enabled = stopper->enabled; + if (enabled) + __cpu_stop_queue_work(stopper, work); + else if (work->done) + cpu_stop_signal_done(work->done); +- spin_unlock_irqrestore(&stopper->lock, flags); + ++ raw_spin_unlock_irqrestore(&stopper->lock, flags); + return enabled; + } + +@@ -231,8 +231,8 @@ static int cpu_stop_queue_two_works(int cpu1, struct cpu_stop_work *work1, + struct cpu_stopper *stopper2 = per_cpu_ptr(&cpu_stopper, cpu2); + int err; + retry: +- spin_lock_irq(&stopper1->lock); +- spin_lock_nested(&stopper2->lock, SINGLE_DEPTH_NESTING); ++ raw_spin_lock_irq(&stopper1->lock); ++ raw_spin_lock_nested(&stopper2->lock, SINGLE_DEPTH_NESTING); + + err = -ENOENT; + if (!stopper1->enabled || !stopper2->enabled) +@@ -255,8 +255,8 @@ static int cpu_stop_queue_two_works(int cpu1, struct cpu_stop_work *work1, + __cpu_stop_queue_work(stopper1, work1); + __cpu_stop_queue_work(stopper2, work2); + unlock: +- spin_unlock(&stopper2->lock); +- spin_unlock_irq(&stopper1->lock); ++ raw_spin_unlock(&stopper2->lock); ++ raw_spin_unlock_irq(&stopper1->lock); + + if (unlikely(err == -EDEADLK)) { + while (stop_cpus_in_progress) +@@ -448,9 +448,9 @@ static int cpu_stop_should_run(unsigned int cpu) + unsigned long flags; + int run; + +- spin_lock_irqsave(&stopper->lock, flags); ++ raw_spin_lock_irqsave(&stopper->lock, flags); + run = !list_empty(&stopper->works); +- spin_unlock_irqrestore(&stopper->lock, flags); ++ raw_spin_unlock_irqrestore(&stopper->lock, flags); + return run; + } + +@@ -461,13 +461,13 @@ static void cpu_stopper_thread(unsigned int cpu) + + repeat: + work = NULL; +- spin_lock_irq(&stopper->lock); ++ raw_spin_lock_irq(&stopper->lock); + if (!list_empty(&stopper->works)) { + work = list_first_entry(&stopper->works, + struct cpu_stop_work, list); + list_del_init(&work->list); + } +- spin_unlock_irq(&stopper->lock); ++ raw_spin_unlock_irq(&stopper->lock); + + if (work) { + cpu_stop_fn_t fn = work->fn; +@@ -475,6 +475,8 @@ static void cpu_stopper_thread(unsigned int cpu) + struct cpu_stop_done *done = work->done; + int ret; + ++ /* XXX */ ++ + /* cpu stop callbacks must not sleep, make in_atomic() == T */ + preempt_count_inc(); + ret = fn(arg); +@@ -541,7 +543,7 @@ static int __init cpu_stop_init(void) + for_each_possible_cpu(cpu) { + struct cpu_stopper *stopper = &per_cpu(cpu_stopper, cpu); + +- spin_lock_init(&stopper->lock); ++ raw_spin_lock_init(&stopper->lock); + INIT_LIST_HEAD(&stopper->works); + } + +diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c +index bb5ec425dfe0..8338b14ed3a3 100644 +--- a/kernel/time/hrtimer.c ++++ b/kernel/time/hrtimer.c +@@ -53,6 +53,7 @@ + #include + + #include ++#include + + #include "tick-internal.h" + +@@ -695,6 +696,29 @@ static void hrtimer_switch_to_hres(void) + retrigger_next_event(NULL); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ ++static struct swork_event clock_set_delay_work; ++ ++static void run_clock_set_delay(struct swork_event *event) ++{ ++ clock_was_set(); ++} ++ ++void clock_was_set_delayed(void) ++{ ++ swork_queue(&clock_set_delay_work); ++} ++ ++static __init int create_clock_set_delay_thread(void) ++{ ++ WARN_ON(swork_get()); ++ INIT_SWORK(&clock_set_delay_work, run_clock_set_delay); ++ return 0; ++} ++early_initcall(create_clock_set_delay_thread); ++#else /* PREEMPT_RT_FULL */ ++ + static void clock_was_set_work(struct work_struct *work) + { + clock_was_set(); +@@ -710,6 +734,7 @@ void clock_was_set_delayed(void) + { + schedule_work(&hrtimer_work); + } ++#endif + + #else + +@@ -719,11 +744,8 @@ static inline int hrtimer_is_hres_enabled(void) { return 0; } + static inline void hrtimer_switch_to_hres(void) { } + static inline void + hrtimer_force_reprogram(struct hrtimer_cpu_base *base, int skip_equal) { } +-static inline int hrtimer_reprogram(struct hrtimer *timer, +- struct hrtimer_clock_base *base) +-{ +- return 0; +-} ++static inline void hrtimer_reprogram(struct hrtimer *timer, ++ struct hrtimer_clock_base *base) { } + static inline void hrtimer_init_hres(struct hrtimer_cpu_base *base) { } + static inline void retrigger_next_event(void *arg) { } + +@@ -855,6 +877,32 @@ u64 hrtimer_forward(struct hrtimer *timer, ktime_t now, ktime_t interval) + } + EXPORT_SYMBOL_GPL(hrtimer_forward); + ++#ifdef CONFIG_PREEMPT_RT_BASE ++# define wake_up_timer_waiters(b) wake_up(&(b)->wait) ++ ++/** ++ * hrtimer_wait_for_timer - Wait for a running timer ++ * ++ * @timer: timer to wait for ++ * ++ * The function waits in case the timers callback function is ++ * currently executed on the waitqueue of the timer base. The ++ * waitqueue is woken up after the timer callback function has ++ * finished execution. ++ */ ++void hrtimer_wait_for_timer(const struct hrtimer *timer) ++{ ++ struct hrtimer_clock_base *base = timer->base; ++ ++ if (base && base->cpu_base && !timer->irqsafe) ++ wait_event(base->cpu_base->wait, ++ !(hrtimer_callback_running(timer))); ++} ++ ++#else ++# define wake_up_timer_waiters(b) do { } while (0) ++#endif ++ + /* + * enqueue_hrtimer - internal function to (re)start a timer + * +@@ -896,6 +944,11 @@ static void __remove_hrtimer(struct hrtimer *timer, + if (!(state & HRTIMER_STATE_ENQUEUED)) + return; + ++ if (unlikely(!list_empty(&timer->cb_entry))) { ++ list_del_init(&timer->cb_entry); ++ return; ++ } ++ + if (!timerqueue_del(&base->active, &timer->node)) + cpu_base->active_bases &= ~(1 << base->index); + +@@ -991,7 +1044,16 @@ void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, + new_base = switch_hrtimer_base(timer, base, mode & HRTIMER_MODE_PINNED); + + timer_stats_hrtimer_set_start_info(timer); ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ { ++ ktime_t now = new_base->get_time(); + ++ if (ktime_to_ns(tim) < ktime_to_ns(now)) ++ timer->praecox = now; ++ else ++ timer->praecox = ktime_set(0, 0); ++ } ++#endif + leftmost = enqueue_hrtimer(timer, new_base); + if (!leftmost) + goto unlock; +@@ -1063,7 +1125,7 @@ int hrtimer_cancel(struct hrtimer *timer) + + if (ret >= 0) + return ret; +- cpu_relax(); ++ hrtimer_wait_for_timer(timer); + } + } + EXPORT_SYMBOL_GPL(hrtimer_cancel); +@@ -1127,6 +1189,7 @@ static void __hrtimer_init(struct hrtimer *timer, clockid_t clock_id, + + base = hrtimer_clockid_to_base(clock_id); + timer->base = &cpu_base->clock_base[base]; ++ INIT_LIST_HEAD(&timer->cb_entry); + timerqueue_init(&timer->node); + + #ifdef CONFIG_TIMER_STATS +@@ -1167,6 +1230,7 @@ bool hrtimer_active(const struct hrtimer *timer) + seq = raw_read_seqcount_begin(&cpu_base->seq); + + if (timer->state != HRTIMER_STATE_INACTIVE || ++ cpu_base->running_soft == timer || + cpu_base->running == timer) + return true; + +@@ -1265,10 +1329,112 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base, + cpu_base->running = NULL; + } + ++#ifdef CONFIG_PREEMPT_RT_BASE ++static void hrtimer_rt_reprogram(int restart, struct hrtimer *timer, ++ struct hrtimer_clock_base *base) ++{ ++ int leftmost; ++ ++ if (restart != HRTIMER_NORESTART && ++ !(timer->state & HRTIMER_STATE_ENQUEUED)) { ++ ++ leftmost = enqueue_hrtimer(timer, base); ++ if (!leftmost) ++ return; ++#ifdef CONFIG_HIGH_RES_TIMERS ++ if (!hrtimer_is_hres_active(timer)) { ++ /* ++ * Kick to reschedule the next tick to handle the new timer ++ * on dynticks target. ++ */ ++ if (base->cpu_base->nohz_active) ++ wake_up_nohz_cpu(base->cpu_base->cpu); ++ } else { ++ ++ hrtimer_reprogram(timer, base); ++ } ++#endif ++ } ++} ++ ++/* ++ * The changes in mainline which removed the callback modes from ++ * hrtimer are not yet working with -rt. The non wakeup_process() ++ * based callbacks which involve sleeping locks need to be treated ++ * seperately. ++ */ ++static void hrtimer_rt_run_pending(void) ++{ ++ enum hrtimer_restart (*fn)(struct hrtimer *); ++ struct hrtimer_cpu_base *cpu_base; ++ struct hrtimer_clock_base *base; ++ struct hrtimer *timer; ++ int index, restart; ++ ++ local_irq_disable(); ++ cpu_base = &per_cpu(hrtimer_bases, smp_processor_id()); ++ ++ raw_spin_lock(&cpu_base->lock); ++ ++ for (index = 0; index < HRTIMER_MAX_CLOCK_BASES; index++) { ++ base = &cpu_base->clock_base[index]; ++ ++ while (!list_empty(&base->expired)) { ++ timer = list_first_entry(&base->expired, ++ struct hrtimer, cb_entry); ++ ++ /* ++ * Same as the above __run_hrtimer function ++ * just we run with interrupts enabled. ++ */ ++ debug_deactivate(timer); ++ cpu_base->running_soft = timer; ++ raw_write_seqcount_barrier(&cpu_base->seq); ++ ++ __remove_hrtimer(timer, base, HRTIMER_STATE_INACTIVE, 0); ++ timer_stats_account_hrtimer(timer); ++ fn = timer->function; ++ ++ raw_spin_unlock_irq(&cpu_base->lock); ++ restart = fn(timer); ++ raw_spin_lock_irq(&cpu_base->lock); ++ ++ hrtimer_rt_reprogram(restart, timer, base); ++ raw_write_seqcount_barrier(&cpu_base->seq); ++ ++ WARN_ON_ONCE(cpu_base->running_soft != timer); ++ cpu_base->running_soft = NULL; ++ } ++ } ++ ++ raw_spin_unlock_irq(&cpu_base->lock); ++ ++ wake_up_timer_waiters(cpu_base); ++} ++ ++static int hrtimer_rt_defer(struct hrtimer *timer) ++{ ++ if (timer->irqsafe) ++ return 0; ++ ++ __remove_hrtimer(timer, timer->base, timer->state, 0); ++ list_add_tail(&timer->cb_entry, &timer->base->expired); ++ return 1; ++} ++ ++#else ++ ++static inline int hrtimer_rt_defer(struct hrtimer *timer) { return 0; } ++ ++#endif ++ ++static enum hrtimer_restart hrtimer_wakeup(struct hrtimer *timer); ++ + static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now) + { + struct hrtimer_clock_base *base = cpu_base->clock_base; + unsigned int active = cpu_base->active_bases; ++ int raise = 0; + + for (; active; base++, active >>= 1) { + struct timerqueue_node *node; +@@ -1284,6 +1450,15 @@ static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now) + + timer = container_of(node, struct hrtimer, node); + ++ trace_hrtimer_interrupt(raw_smp_processor_id(), ++ ktime_to_ns(ktime_sub(ktime_to_ns(timer->praecox) ? ++ timer->praecox : hrtimer_get_expires(timer), ++ basenow)), ++ current, ++ timer->function == hrtimer_wakeup ? ++ container_of(timer, struct hrtimer_sleeper, ++ timer)->task : NULL); ++ + /* + * The immediate goal for using the softexpires is + * minimizing wakeups, not running timers at the +@@ -1299,9 +1474,14 @@ static void __hrtimer_run_queues(struct hrtimer_cpu_base *cpu_base, ktime_t now) + if (basenow.tv64 < hrtimer_get_softexpires_tv64(timer)) + break; + +- __run_hrtimer(cpu_base, base, timer, &basenow); ++ if (!hrtimer_rt_defer(timer)) ++ __run_hrtimer(cpu_base, base, timer, &basenow); ++ else ++ raise = 1; + } + } ++ if (raise) ++ raise_softirq_irqoff(HRTIMER_SOFTIRQ); + } + + #ifdef CONFIG_HIGH_RES_TIMERS +@@ -1464,16 +1644,18 @@ static enum hrtimer_restart hrtimer_wakeup(struct hrtimer *timer) + void hrtimer_init_sleeper(struct hrtimer_sleeper *sl, struct task_struct *task) + { + sl->timer.function = hrtimer_wakeup; ++ sl->timer.irqsafe = 1; + sl->task = task; + } + EXPORT_SYMBOL_GPL(hrtimer_init_sleeper); + +-static int __sched do_nanosleep(struct hrtimer_sleeper *t, enum hrtimer_mode mode) ++static int __sched do_nanosleep(struct hrtimer_sleeper *t, enum hrtimer_mode mode, ++ unsigned long state) + { + hrtimer_init_sleeper(t, current); + + do { +- set_current_state(TASK_INTERRUPTIBLE); ++ set_current_state(state); + hrtimer_start_expires(&t->timer, mode); + + if (likely(t->task)) +@@ -1515,7 +1697,8 @@ long __sched hrtimer_nanosleep_restart(struct restart_block *restart) + HRTIMER_MODE_ABS); + hrtimer_set_expires_tv64(&t.timer, restart->nanosleep.expires); + +- if (do_nanosleep(&t, HRTIMER_MODE_ABS)) ++ /* cpu_chill() does not care about restart state. */ ++ if (do_nanosleep(&t, HRTIMER_MODE_ABS, TASK_INTERRUPTIBLE)) + goto out; + + rmtp = restart->nanosleep.rmtp; +@@ -1532,8 +1715,10 @@ long __sched hrtimer_nanosleep_restart(struct restart_block *restart) + return ret; + } + +-long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, +- const enum hrtimer_mode mode, const clockid_t clockid) ++static long ++__hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, ++ const enum hrtimer_mode mode, const clockid_t clockid, ++ unsigned long state) + { + struct restart_block *restart; + struct hrtimer_sleeper t; +@@ -1546,7 +1731,7 @@ long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, + + hrtimer_init_on_stack(&t.timer, clockid, mode); + hrtimer_set_expires_range_ns(&t.timer, timespec_to_ktime(*rqtp), slack); +- if (do_nanosleep(&t, mode)) ++ if (do_nanosleep(&t, mode, state)) + goto out; + + /* Absolute timers do not update the rmtp value and restart: */ +@@ -1573,6 +1758,12 @@ long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, + return ret; + } + ++long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, ++ const enum hrtimer_mode mode, const clockid_t clockid) ++{ ++ return __hrtimer_nanosleep(rqtp, rmtp, mode, clockid, TASK_INTERRUPTIBLE); ++} ++ + SYSCALL_DEFINE2(nanosleep, struct timespec __user *, rqtp, + struct timespec __user *, rmtp) + { +@@ -1587,6 +1778,26 @@ SYSCALL_DEFINE2(nanosleep, struct timespec __user *, rqtp, + return hrtimer_nanosleep(&tu, rmtp, HRTIMER_MODE_REL, CLOCK_MONOTONIC); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++/* ++ * Sleep for 1 ms in hope whoever holds what we want will let it go. ++ */ ++void cpu_chill(void) ++{ ++ struct timespec tu = { ++ .tv_nsec = NSEC_PER_MSEC, ++ }; ++ unsigned int freeze_flag = current->flags & PF_NOFREEZE; ++ ++ current->flags |= PF_NOFREEZE; ++ __hrtimer_nanosleep(&tu, NULL, HRTIMER_MODE_REL, CLOCK_MONOTONIC, ++ TASK_UNINTERRUPTIBLE); ++ if (!freeze_flag) ++ current->flags &= ~PF_NOFREEZE; ++} ++EXPORT_SYMBOL(cpu_chill); ++#endif ++ + /* + * Functions related to boot-time initialization: + */ +@@ -1598,10 +1809,14 @@ int hrtimers_prepare_cpu(unsigned int cpu) + for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) { + cpu_base->clock_base[i].cpu_base = cpu_base; + timerqueue_init_head(&cpu_base->clock_base[i].active); ++ INIT_LIST_HEAD(&cpu_base->clock_base[i].expired); + } + + cpu_base->cpu = cpu; + hrtimer_init_hres(cpu_base); ++#ifdef CONFIG_PREEMPT_RT_BASE ++ init_waitqueue_head(&cpu_base->wait); ++#endif + return 0; + } + +@@ -1671,9 +1886,26 @@ int hrtimers_dead_cpu(unsigned int scpu) + + #endif /* CONFIG_HOTPLUG_CPU */ + ++#ifdef CONFIG_PREEMPT_RT_BASE ++ ++static void run_hrtimer_softirq(struct softirq_action *h) ++{ ++ hrtimer_rt_run_pending(); ++} ++ ++static void hrtimers_open_softirq(void) ++{ ++ open_softirq(HRTIMER_SOFTIRQ, run_hrtimer_softirq); ++} ++ ++#else ++static void hrtimers_open_softirq(void) { } ++#endif ++ + void __init hrtimers_init(void) + { + hrtimers_prepare_cpu(smp_processor_id()); ++ hrtimers_open_softirq(); + } + + /** +diff --git a/kernel/time/itimer.c b/kernel/time/itimer.c +index 1d5c7204ddc9..184de6751180 100644 +--- a/kernel/time/itimer.c ++++ b/kernel/time/itimer.c +@@ -213,6 +213,7 @@ int do_setitimer(int which, struct itimerval *value, struct itimerval *ovalue) + /* We are sharing ->siglock with it_real_fn() */ + if (hrtimer_try_to_cancel(timer) < 0) { + spin_unlock_irq(&tsk->sighand->siglock); ++ hrtimer_wait_for_timer(&tsk->signal->real_timer); + goto again; + } + expires = timeval_to_ktime(value->it_value); +diff --git a/kernel/time/jiffies.c b/kernel/time/jiffies.c +index 555e21f7b966..a5d6435fabbb 100644 +--- a/kernel/time/jiffies.c ++++ b/kernel/time/jiffies.c +@@ -74,7 +74,8 @@ static struct clocksource clocksource_jiffies = { + .max_cycles = 10, + }; + +-__cacheline_aligned_in_smp DEFINE_SEQLOCK(jiffies_lock); ++__cacheline_aligned_in_smp DEFINE_RAW_SPINLOCK(jiffies_lock); ++__cacheline_aligned_in_smp seqcount_t jiffies_seq; + + #if (BITS_PER_LONG < 64) + u64 get_jiffies_64(void) +@@ -83,9 +84,9 @@ u64 get_jiffies_64(void) + u64 ret; + + do { +- seq = read_seqbegin(&jiffies_lock); ++ seq = read_seqcount_begin(&jiffies_seq); + ret = jiffies_64; +- } while (read_seqretry(&jiffies_lock, seq)); ++ } while (read_seqcount_retry(&jiffies_seq, seq)); + return ret; + } + EXPORT_SYMBOL(get_jiffies_64); +diff --git a/kernel/time/ntp.c b/kernel/time/ntp.c +index 6df8927c58a5..05b7391bf9bd 100644 +--- a/kernel/time/ntp.c ++++ b/kernel/time/ntp.c +@@ -17,6 +17,7 @@ + #include + #include + #include ++#include + + #include "ntp_internal.h" + #include "timekeeping_internal.h" +@@ -568,10 +569,35 @@ static void sync_cmos_clock(struct work_struct *work) + &sync_cmos_work, timespec64_to_jiffies(&next)); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ ++static void run_clock_set_delay(struct swork_event *event) ++{ ++ queue_delayed_work(system_power_efficient_wq, &sync_cmos_work, 0); ++} ++ ++static struct swork_event ntp_cmos_swork; ++ ++void ntp_notify_cmos_timer(void) ++{ ++ swork_queue(&ntp_cmos_swork); ++} ++ ++static __init int create_cmos_delay_thread(void) ++{ ++ WARN_ON(swork_get()); ++ INIT_SWORK(&ntp_cmos_swork, run_clock_set_delay); ++ return 0; ++} ++early_initcall(create_cmos_delay_thread); ++ ++#else ++ + void ntp_notify_cmos_timer(void) + { + queue_delayed_work(system_power_efficient_wq, &sync_cmos_work, 0); + } ++#endif /* CONFIG_PREEMPT_RT_FULL */ + + #else + void ntp_notify_cmos_timer(void) { } +diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c +index 39008d78927a..633f4eaca9e7 100644 +--- a/kernel/time/posix-cpu-timers.c ++++ b/kernel/time/posix-cpu-timers.c +@@ -3,6 +3,7 @@ + */ + + #include ++#include + #include + #include + #include +@@ -620,7 +621,7 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int timer_flags, + /* + * Disarm any old timer after extracting its expiry time. + */ +- WARN_ON_ONCE(!irqs_disabled()); ++ WARN_ON_ONCE_NONRT(!irqs_disabled()); + + ret = 0; + old_incr = timer->it.cpu.incr; +@@ -1064,7 +1065,7 @@ void posix_cpu_timer_schedule(struct k_itimer *timer) + /* + * Now re-arm for the new expiry time. + */ +- WARN_ON_ONCE(!irqs_disabled()); ++ WARN_ON_ONCE_NONRT(!irqs_disabled()); + arm_timer(timer); + unlock_task_sighand(p, &flags); + +@@ -1153,13 +1154,13 @@ static inline int fastpath_timer_check(struct task_struct *tsk) + * already updated our counts. We need to check if any timers fire now. + * Interrupts are disabled. + */ +-void run_posix_cpu_timers(struct task_struct *tsk) ++static void __run_posix_cpu_timers(struct task_struct *tsk) + { + LIST_HEAD(firing); + struct k_itimer *timer, *next; + unsigned long flags; + +- WARN_ON_ONCE(!irqs_disabled()); ++ WARN_ON_ONCE_NONRT(!irqs_disabled()); + + /* + * The fast path checks that there are no expired thread or thread +@@ -1213,6 +1214,190 @@ void run_posix_cpu_timers(struct task_struct *tsk) + } + } + ++#ifdef CONFIG_PREEMPT_RT_BASE ++#include ++#include ++DEFINE_PER_CPU(struct task_struct *, posix_timer_task); ++DEFINE_PER_CPU(struct task_struct *, posix_timer_tasklist); ++ ++static int posix_cpu_timers_thread(void *data) ++{ ++ int cpu = (long)data; ++ ++ BUG_ON(per_cpu(posix_timer_task,cpu) != current); ++ ++ while (!kthread_should_stop()) { ++ struct task_struct *tsk = NULL; ++ struct task_struct *next = NULL; ++ ++ if (cpu_is_offline(cpu)) ++ goto wait_to_die; ++ ++ /* grab task list */ ++ raw_local_irq_disable(); ++ tsk = per_cpu(posix_timer_tasklist, cpu); ++ per_cpu(posix_timer_tasklist, cpu) = NULL; ++ raw_local_irq_enable(); ++ ++ /* its possible the list is empty, just return */ ++ if (!tsk) { ++ set_current_state(TASK_INTERRUPTIBLE); ++ schedule(); ++ __set_current_state(TASK_RUNNING); ++ continue; ++ } ++ ++ /* Process task list */ ++ while (1) { ++ /* save next */ ++ next = tsk->posix_timer_list; ++ ++ /* run the task timers, clear its ptr and ++ * unreference it ++ */ ++ __run_posix_cpu_timers(tsk); ++ tsk->posix_timer_list = NULL; ++ put_task_struct(tsk); ++ ++ /* check if this is the last on the list */ ++ if (next == tsk) ++ break; ++ tsk = next; ++ } ++ } ++ return 0; ++ ++wait_to_die: ++ /* Wait for kthread_stop */ ++ set_current_state(TASK_INTERRUPTIBLE); ++ while (!kthread_should_stop()) { ++ schedule(); ++ set_current_state(TASK_INTERRUPTIBLE); ++ } ++ __set_current_state(TASK_RUNNING); ++ return 0; ++} ++ ++static inline int __fastpath_timer_check(struct task_struct *tsk) ++{ ++ /* tsk == current, ensure it is safe to use ->signal/sighand */ ++ if (unlikely(tsk->exit_state)) ++ return 0; ++ ++ if (!task_cputime_zero(&tsk->cputime_expires)) ++ return 1; ++ ++ if (!task_cputime_zero(&tsk->signal->cputime_expires)) ++ return 1; ++ ++ return 0; ++} ++ ++void run_posix_cpu_timers(struct task_struct *tsk) ++{ ++ unsigned long cpu = smp_processor_id(); ++ struct task_struct *tasklist; ++ ++ BUG_ON(!irqs_disabled()); ++ if(!per_cpu(posix_timer_task, cpu)) ++ return; ++ /* get per-cpu references */ ++ tasklist = per_cpu(posix_timer_tasklist, cpu); ++ ++ /* check to see if we're already queued */ ++ if (!tsk->posix_timer_list && __fastpath_timer_check(tsk)) { ++ get_task_struct(tsk); ++ if (tasklist) { ++ tsk->posix_timer_list = tasklist; ++ } else { ++ /* ++ * The list is terminated by a self-pointing ++ * task_struct ++ */ ++ tsk->posix_timer_list = tsk; ++ } ++ per_cpu(posix_timer_tasklist, cpu) = tsk; ++ ++ wake_up_process(per_cpu(posix_timer_task, cpu)); ++ } ++} ++ ++/* ++ * posix_cpu_thread_call - callback that gets triggered when a CPU is added. ++ * Here we can start up the necessary migration thread for the new CPU. ++ */ ++static int posix_cpu_thread_call(struct notifier_block *nfb, ++ unsigned long action, void *hcpu) ++{ ++ int cpu = (long)hcpu; ++ struct task_struct *p; ++ struct sched_param param; ++ ++ switch (action) { ++ case CPU_UP_PREPARE: ++ p = kthread_create(posix_cpu_timers_thread, hcpu, ++ "posixcputmr/%d",cpu); ++ if (IS_ERR(p)) ++ return NOTIFY_BAD; ++ p->flags |= PF_NOFREEZE; ++ kthread_bind(p, cpu); ++ /* Must be high prio to avoid getting starved */ ++ param.sched_priority = MAX_RT_PRIO-1; ++ sched_setscheduler(p, SCHED_FIFO, ¶m); ++ per_cpu(posix_timer_task,cpu) = p; ++ break; ++ case CPU_ONLINE: ++ /* Strictly unneccessary, as first user will wake it. */ ++ wake_up_process(per_cpu(posix_timer_task,cpu)); ++ break; ++#ifdef CONFIG_HOTPLUG_CPU ++ case CPU_UP_CANCELED: ++ /* Unbind it from offline cpu so it can run. Fall thru. */ ++ kthread_bind(per_cpu(posix_timer_task, cpu), ++ cpumask_any(cpu_online_mask)); ++ kthread_stop(per_cpu(posix_timer_task,cpu)); ++ per_cpu(posix_timer_task,cpu) = NULL; ++ break; ++ case CPU_DEAD: ++ kthread_stop(per_cpu(posix_timer_task,cpu)); ++ per_cpu(posix_timer_task,cpu) = NULL; ++ break; ++#endif ++ } ++ return NOTIFY_OK; ++} ++ ++/* Register at highest priority so that task migration (migrate_all_tasks) ++ * happens before everything else. ++ */ ++static struct notifier_block posix_cpu_thread_notifier = { ++ .notifier_call = posix_cpu_thread_call, ++ .priority = 10 ++}; ++ ++static int __init posix_cpu_thread_init(void) ++{ ++ void *hcpu = (void *)(long)smp_processor_id(); ++ /* Start one for boot CPU. */ ++ unsigned long cpu; ++ ++ /* init the per-cpu posix_timer_tasklets */ ++ for_each_possible_cpu(cpu) ++ per_cpu(posix_timer_tasklist, cpu) = NULL; ++ ++ posix_cpu_thread_call(&posix_cpu_thread_notifier, CPU_UP_PREPARE, hcpu); ++ posix_cpu_thread_call(&posix_cpu_thread_notifier, CPU_ONLINE, hcpu); ++ register_cpu_notifier(&posix_cpu_thread_notifier); ++ return 0; ++} ++early_initcall(posix_cpu_thread_init); ++#else /* CONFIG_PREEMPT_RT_BASE */ ++void run_posix_cpu_timers(struct task_struct *tsk) ++{ ++ __run_posix_cpu_timers(tsk); ++} ++#endif /* CONFIG_PREEMPT_RT_BASE */ ++ + /* + * Set one of the process-wide special case CPU timers or RLIMIT_CPU. + * The tsk->sighand->siglock must be held by the caller. +diff --git a/kernel/time/posix-timers.c b/kernel/time/posix-timers.c +index f2826c35e918..464a98155a0e 100644 +--- a/kernel/time/posix-timers.c ++++ b/kernel/time/posix-timers.c +@@ -506,6 +506,7 @@ static enum hrtimer_restart posix_timer_fn(struct hrtimer *timer) + static struct pid *good_sigevent(sigevent_t * event) + { + struct task_struct *rtn = current->group_leader; ++ int sig = event->sigev_signo; + + if ((event->sigev_notify & SIGEV_THREAD_ID ) && + (!(rtn = find_task_by_vpid(event->sigev_notify_thread_id)) || +@@ -514,7 +515,8 @@ static struct pid *good_sigevent(sigevent_t * event) + return NULL; + + if (((event->sigev_notify & ~SIGEV_THREAD_ID) != SIGEV_NONE) && +- ((event->sigev_signo <= 0) || (event->sigev_signo > SIGRTMAX))) ++ (sig <= 0 || sig > SIGRTMAX || sig_kernel_only(sig) || ++ sig_kernel_coredump(sig))) + return NULL; + + return task_pid(rtn); +@@ -826,6 +828,20 @@ SYSCALL_DEFINE1(timer_getoverrun, timer_t, timer_id) + return overrun; + } + ++/* ++ * Protected by RCU! ++ */ ++static void timer_wait_for_callback(struct k_clock *kc, struct k_itimer *timr) ++{ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ if (kc->timer_set == common_timer_set) ++ hrtimer_wait_for_timer(&timr->it.real.timer); ++ else ++ /* FIXME: Whacky hack for posix-cpu-timers */ ++ schedule_timeout(1); ++#endif ++} ++ + /* Set a POSIX.1b interval timer. */ + /* timr->it_lock is taken. */ + static int +@@ -903,6 +919,7 @@ SYSCALL_DEFINE4(timer_settime, timer_t, timer_id, int, flags, + if (!timr) + return -EINVAL; + ++ rcu_read_lock(); + kc = clockid_to_kclock(timr->it_clock); + if (WARN_ON_ONCE(!kc || !kc->timer_set)) + error = -EINVAL; +@@ -911,9 +928,12 @@ SYSCALL_DEFINE4(timer_settime, timer_t, timer_id, int, flags, + + unlock_timer(timr, flag); + if (error == TIMER_RETRY) { ++ timer_wait_for_callback(kc, timr); + rtn = NULL; // We already got the old time... ++ rcu_read_unlock(); + goto retry; + } ++ rcu_read_unlock(); + + if (old_setting && !error && + copy_to_user(old_setting, &old_spec, sizeof (old_spec))) +@@ -951,10 +971,15 @@ SYSCALL_DEFINE1(timer_delete, timer_t, timer_id) + if (!timer) + return -EINVAL; + ++ rcu_read_lock(); + if (timer_delete_hook(timer) == TIMER_RETRY) { + unlock_timer(timer, flags); ++ timer_wait_for_callback(clockid_to_kclock(timer->it_clock), ++ timer); ++ rcu_read_unlock(); + goto retry_delete; + } ++ rcu_read_unlock(); + + spin_lock(¤t->sighand->siglock); + list_del(&timer->list); +@@ -980,8 +1005,18 @@ static void itimer_delete(struct k_itimer *timer) + retry_delete: + spin_lock_irqsave(&timer->it_lock, flags); + +- if (timer_delete_hook(timer) == TIMER_RETRY) { ++ /* On RT we can race with a deletion */ ++ if (!timer->it_signal) { + unlock_timer(timer, flags); ++ return; ++ } ++ ++ if (timer_delete_hook(timer) == TIMER_RETRY) { ++ rcu_read_lock(); ++ unlock_timer(timer, flags); ++ timer_wait_for_callback(clockid_to_kclock(timer->it_clock), ++ timer); ++ rcu_read_unlock(); + goto retry_delete; + } + list_del(&timer->list); +diff --git a/kernel/time/tick-broadcast-hrtimer.c b/kernel/time/tick-broadcast-hrtimer.c +index 690b797f522e..fe8ba1619879 100644 +--- a/kernel/time/tick-broadcast-hrtimer.c ++++ b/kernel/time/tick-broadcast-hrtimer.c +@@ -107,5 +107,6 @@ void tick_setup_hrtimer_broadcast(void) + { + hrtimer_init(&bctimer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); + bctimer.function = bc_handler; ++ bctimer.irqsafe = true; + clockevents_register_device(&ce_broadcast_hrtimer); + } +diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c +index 4fcd99e12aa0..5a47f2e98faf 100644 +--- a/kernel/time/tick-common.c ++++ b/kernel/time/tick-common.c +@@ -79,13 +79,15 @@ int tick_is_oneshot_available(void) + static void tick_periodic(int cpu) + { + if (tick_do_timer_cpu == cpu) { +- write_seqlock(&jiffies_lock); ++ raw_spin_lock(&jiffies_lock); ++ write_seqcount_begin(&jiffies_seq); + + /* Keep track of the next tick event */ + tick_next_period = ktime_add(tick_next_period, tick_period); + + do_timer(1); +- write_sequnlock(&jiffies_lock); ++ write_seqcount_end(&jiffies_seq); ++ raw_spin_unlock(&jiffies_lock); + update_wall_time(); + } + +@@ -157,9 +159,9 @@ void tick_setup_periodic(struct clock_event_device *dev, int broadcast) + ktime_t next; + + do { +- seq = read_seqbegin(&jiffies_lock); ++ seq = read_seqcount_begin(&jiffies_seq); + next = tick_next_period; +- } while (read_seqretry(&jiffies_lock, seq)); ++ } while (read_seqcount_retry(&jiffies_seq, seq)); + + clockevents_switch_state(dev, CLOCK_EVT_STATE_ONESHOT); + +diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c +index 3bcb61b52f6c..66d85482a96e 100644 +--- a/kernel/time/tick-sched.c ++++ b/kernel/time/tick-sched.c +@@ -62,7 +62,8 @@ static void tick_do_update_jiffies64(ktime_t now) + return; + + /* Reevaluate with jiffies_lock held */ +- write_seqlock(&jiffies_lock); ++ raw_spin_lock(&jiffies_lock); ++ write_seqcount_begin(&jiffies_seq); + + delta = ktime_sub(now, last_jiffies_update); + if (delta.tv64 >= tick_period.tv64) { +@@ -85,10 +86,12 @@ static void tick_do_update_jiffies64(ktime_t now) + /* Keep the tick_next_period variable up to date */ + tick_next_period = ktime_add(last_jiffies_update, tick_period); + } else { +- write_sequnlock(&jiffies_lock); ++ write_seqcount_end(&jiffies_seq); ++ raw_spin_unlock(&jiffies_lock); + return; + } +- write_sequnlock(&jiffies_lock); ++ write_seqcount_end(&jiffies_seq); ++ raw_spin_unlock(&jiffies_lock); + update_wall_time(); + } + +@@ -99,12 +102,14 @@ static ktime_t tick_init_jiffy_update(void) + { + ktime_t period; + +- write_seqlock(&jiffies_lock); ++ raw_spin_lock(&jiffies_lock); ++ write_seqcount_begin(&jiffies_seq); + /* Did we start the jiffies update yet ? */ + if (last_jiffies_update.tv64 == 0) + last_jiffies_update = tick_next_period; + period = last_jiffies_update; +- write_sequnlock(&jiffies_lock); ++ write_seqcount_end(&jiffies_seq); ++ raw_spin_unlock(&jiffies_lock); + return period; + } + +@@ -215,6 +220,7 @@ static void nohz_full_kick_func(struct irq_work *work) + + static DEFINE_PER_CPU(struct irq_work, nohz_full_kick_work) = { + .func = nohz_full_kick_func, ++ .flags = IRQ_WORK_HARD_IRQ, + }; + + /* +@@ -673,10 +679,10 @@ static ktime_t tick_nohz_stop_sched_tick(struct tick_sched *ts, + + /* Read jiffies and the time when jiffies were updated last */ + do { +- seq = read_seqbegin(&jiffies_lock); ++ seq = read_seqcount_begin(&jiffies_seq); + basemono = last_jiffies_update.tv64; + basejiff = jiffies; +- } while (read_seqretry(&jiffies_lock, seq)); ++ } while (read_seqcount_retry(&jiffies_seq, seq)); + ts->last_jiffies = basejiff; + + if (rcu_needs_cpu(basemono, &next_rcu) || +@@ -877,14 +883,7 @@ static bool can_stop_idle_tick(int cpu, struct tick_sched *ts) + return false; + + if (unlikely(local_softirq_pending() && cpu_online(cpu))) { +- static int ratelimit; +- +- if (ratelimit < 10 && +- (local_softirq_pending() & SOFTIRQ_STOP_IDLE_MASK)) { +- pr_warn("NOHZ: local_softirq_pending %02x\n", +- (unsigned int) local_softirq_pending()); +- ratelimit++; +- } ++ softirq_check_pending_idle(); + return false; + } + +@@ -1193,6 +1192,7 @@ void tick_setup_sched_timer(void) + * Emulate tick processing via per-CPU hrtimers: + */ + hrtimer_init(&ts->sched_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); ++ ts->sched_timer.irqsafe = 1; + ts->sched_timer.function = tick_sched_timer; + + /* Get the next period (per-CPU) */ +diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c +index 46e312e9be38..fa75cf5d9253 100644 +--- a/kernel/time/timekeeping.c ++++ b/kernel/time/timekeeping.c +@@ -2328,8 +2328,10 @@ EXPORT_SYMBOL(hardpps); + */ + void xtime_update(unsigned long ticks) + { +- write_seqlock(&jiffies_lock); ++ raw_spin_lock(&jiffies_lock); ++ write_seqcount_begin(&jiffies_seq); + do_timer(ticks); +- write_sequnlock(&jiffies_lock); ++ write_seqcount_end(&jiffies_seq); ++ raw_spin_unlock(&jiffies_lock); + update_wall_time(); + } +diff --git a/kernel/time/timekeeping.h b/kernel/time/timekeeping.h +index 704f595ce83f..763a3e5121ff 100644 +--- a/kernel/time/timekeeping.h ++++ b/kernel/time/timekeeping.h +@@ -19,7 +19,8 @@ extern void timekeeping_resume(void); + extern void do_timer(unsigned long ticks); + extern void update_wall_time(void); + +-extern seqlock_t jiffies_lock; ++extern raw_spinlock_t jiffies_lock; ++extern seqcount_t jiffies_seq; + + #define CS_NAME_LEN 32 + +diff --git a/kernel/time/timer.c b/kernel/time/timer.c +index c611c47de884..08a5ab762495 100644 +--- a/kernel/time/timer.c ++++ b/kernel/time/timer.c +@@ -193,8 +193,11 @@ EXPORT_SYMBOL(jiffies_64); + #endif + + struct timer_base { +- spinlock_t lock; ++ raw_spinlock_t lock; + struct timer_list *running_timer; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ struct swait_queue_head wait_for_running_timer; ++#endif + unsigned long clk; + unsigned long next_expiry; + unsigned int cpu; +@@ -948,10 +951,10 @@ static struct timer_base *lock_timer_base(struct timer_list *timer, + + if (!(tf & TIMER_MIGRATING)) { + base = get_timer_base(tf); +- spin_lock_irqsave(&base->lock, *flags); ++ raw_spin_lock_irqsave(&base->lock, *flags); + if (timer->flags == tf) + return base; +- spin_unlock_irqrestore(&base->lock, *flags); ++ raw_spin_unlock_irqrestore(&base->lock, *flags); + } + cpu_relax(); + } +@@ -1023,9 +1026,9 @@ __mod_timer(struct timer_list *timer, unsigned long expires, bool pending_only) + /* See the comment in lock_timer_base() */ + timer->flags |= TIMER_MIGRATING; + +- spin_unlock(&base->lock); ++ raw_spin_unlock(&base->lock); + base = new_base; +- spin_lock(&base->lock); ++ raw_spin_lock(&base->lock); + WRITE_ONCE(timer->flags, + (timer->flags & ~TIMER_BASEMASK) | base->cpu); + } +@@ -1050,7 +1053,7 @@ __mod_timer(struct timer_list *timer, unsigned long expires, bool pending_only) + } + + out_unlock: +- spin_unlock_irqrestore(&base->lock, flags); ++ raw_spin_unlock_irqrestore(&base->lock, flags); + + return ret; + } +@@ -1144,19 +1147,46 @@ void add_timer_on(struct timer_list *timer, int cpu) + if (base != new_base) { + timer->flags |= TIMER_MIGRATING; + +- spin_unlock(&base->lock); ++ raw_spin_unlock(&base->lock); + base = new_base; +- spin_lock(&base->lock); ++ raw_spin_lock(&base->lock); + WRITE_ONCE(timer->flags, + (timer->flags & ~TIMER_BASEMASK) | cpu); + } + + debug_activate(timer, timer->expires); + internal_add_timer(base, timer); +- spin_unlock_irqrestore(&base->lock, flags); ++ raw_spin_unlock_irqrestore(&base->lock, flags); + } + EXPORT_SYMBOL_GPL(add_timer_on); + ++#ifdef CONFIG_PREEMPT_RT_FULL ++/* ++ * Wait for a running timer ++ */ ++static void wait_for_running_timer(struct timer_list *timer) ++{ ++ struct timer_base *base; ++ u32 tf = timer->flags; ++ ++ if (tf & TIMER_MIGRATING) ++ return; ++ ++ base = get_timer_base(tf); ++ swait_event(base->wait_for_running_timer, ++ base->running_timer != timer); ++} ++ ++# define wakeup_timer_waiters(b) swake_up_all(&(b)->wait_for_running_timer) ++#else ++static inline void wait_for_running_timer(struct timer_list *timer) ++{ ++ cpu_relax(); ++} ++ ++# define wakeup_timer_waiters(b) do { } while (0) ++#endif ++ + /** + * del_timer - deactive a timer. + * @timer: the timer to be deactivated +@@ -1180,7 +1210,7 @@ int del_timer(struct timer_list *timer) + if (timer_pending(timer)) { + base = lock_timer_base(timer, &flags); + ret = detach_if_pending(timer, base, true); +- spin_unlock_irqrestore(&base->lock, flags); ++ raw_spin_unlock_irqrestore(&base->lock, flags); + } + + return ret; +@@ -1208,13 +1238,13 @@ int try_to_del_timer_sync(struct timer_list *timer) + timer_stats_timer_clear_start_info(timer); + ret = detach_if_pending(timer, base, true); + } +- spin_unlock_irqrestore(&base->lock, flags); ++ raw_spin_unlock_irqrestore(&base->lock, flags); + + return ret; + } + EXPORT_SYMBOL(try_to_del_timer_sync); + +-#ifdef CONFIG_SMP ++#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL) + /** + * del_timer_sync - deactivate a timer and wait for the handler to finish. + * @timer: the timer to be deactivated +@@ -1274,7 +1304,7 @@ int del_timer_sync(struct timer_list *timer) + int ret = try_to_del_timer_sync(timer); + if (ret >= 0) + return ret; +- cpu_relax(); ++ wait_for_running_timer(timer); + } + } + EXPORT_SYMBOL(del_timer_sync); +@@ -1339,14 +1369,17 @@ static void expire_timers(struct timer_base *base, struct hlist_head *head) + fn = timer->function; + data = timer->data; + +- if (timer->flags & TIMER_IRQSAFE) { +- spin_unlock(&base->lock); ++ if (!IS_ENABLED(CONFIG_PREEMPT_RT_FULL) && ++ timer->flags & TIMER_IRQSAFE) { ++ raw_spin_unlock(&base->lock); + call_timer_fn(timer, fn, data); +- spin_lock(&base->lock); ++ base->running_timer = NULL; ++ raw_spin_lock(&base->lock); + } else { +- spin_unlock_irq(&base->lock); ++ raw_spin_unlock_irq(&base->lock); + call_timer_fn(timer, fn, data); +- spin_lock_irq(&base->lock); ++ base->running_timer = NULL; ++ raw_spin_lock_irq(&base->lock); + } + } + } +@@ -1515,7 +1548,7 @@ u64 get_next_timer_interrupt(unsigned long basej, u64 basem) + if (cpu_is_offline(smp_processor_id())) + return expires; + +- spin_lock(&base->lock); ++ raw_spin_lock(&base->lock); + nextevt = __next_timer_interrupt(base); + is_max_delta = (nextevt == base->clk + NEXT_TIMER_MAX_DELTA); + base->next_expiry = nextevt; +@@ -1543,7 +1576,7 @@ u64 get_next_timer_interrupt(unsigned long basej, u64 basem) + if ((expires - basem) > TICK_NSEC) + base->is_idle = true; + } +- spin_unlock(&base->lock); ++ raw_spin_unlock(&base->lock); + + return cmp_next_hrtimer_event(basem, expires); + } +@@ -1608,13 +1641,13 @@ void update_process_times(int user_tick) + + /* Note: this timer irq context must be accounted for as well. */ + account_process_tick(p, user_tick); ++ scheduler_tick(); + run_local_timers(); + rcu_check_callbacks(user_tick); +-#ifdef CONFIG_IRQ_WORK ++#if defined(CONFIG_IRQ_WORK) + if (in_irq()) + irq_work_tick(); + #endif +- scheduler_tick(); + run_posix_cpu_timers(p); + } + +@@ -1630,7 +1663,7 @@ static inline void __run_timers(struct timer_base *base) + if (!time_after_eq(jiffies, base->clk)) + return; + +- spin_lock_irq(&base->lock); ++ raw_spin_lock_irq(&base->lock); + + while (time_after_eq(jiffies, base->clk)) { + +@@ -1640,8 +1673,8 @@ static inline void __run_timers(struct timer_base *base) + while (levels--) + expire_timers(base, heads + levels); + } +- base->running_timer = NULL; +- spin_unlock_irq(&base->lock); ++ raw_spin_unlock_irq(&base->lock); ++ wakeup_timer_waiters(base); + } + + /* +@@ -1651,6 +1684,8 @@ static __latent_entropy void run_timer_softirq(struct softirq_action *h) + { + struct timer_base *base = this_cpu_ptr(&timer_bases[BASE_STD]); + ++ irq_work_tick_soft(); ++ + __run_timers(base); + if (IS_ENABLED(CONFIG_NO_HZ_COMMON) && base->nohz_active) + __run_timers(this_cpu_ptr(&timer_bases[BASE_DEF])); +@@ -1836,16 +1871,16 @@ int timers_dead_cpu(unsigned int cpu) + * The caller is globally serialized and nobody else + * takes two locks at once, deadlock is not possible. + */ +- spin_lock_irq(&new_base->lock); +- spin_lock_nested(&old_base->lock, SINGLE_DEPTH_NESTING); ++ raw_spin_lock_irq(&new_base->lock); ++ raw_spin_lock_nested(&old_base->lock, SINGLE_DEPTH_NESTING); + + BUG_ON(old_base->running_timer); + + for (i = 0; i < WHEEL_SIZE; i++) + migrate_timer_list(new_base, old_base->vectors + i); + +- spin_unlock(&old_base->lock); +- spin_unlock_irq(&new_base->lock); ++ raw_spin_unlock(&old_base->lock); ++ raw_spin_unlock_irq(&new_base->lock); + put_cpu_ptr(&timer_bases); + } + return 0; +@@ -1861,8 +1896,11 @@ static void __init init_timer_cpu(int cpu) + for (i = 0; i < NR_BASES; i++) { + base = per_cpu_ptr(&timer_bases[i], cpu); + base->cpu = cpu; +- spin_lock_init(&base->lock); ++ raw_spin_lock_init(&base->lock); + base->clk = jiffies; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ init_swait_queue_head(&base->wait_for_running_timer); ++#endif + } + } + +diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig +index 2a96b063d659..812e37237eb8 100644 +--- a/kernel/trace/Kconfig ++++ b/kernel/trace/Kconfig +@@ -182,6 +182,24 @@ config IRQSOFF_TRACER + enabled. This option and the preempt-off timing option can be + used together or separately.) + ++config INTERRUPT_OFF_HIST ++ bool "Interrupts-off Latency Histogram" ++ depends on IRQSOFF_TRACER ++ help ++ This option generates continuously updated histograms (one per cpu) ++ of the duration of time periods with interrupts disabled. The ++ histograms are disabled by default. To enable them, write a non-zero ++ number to ++ ++ /sys/kernel/debug/tracing/latency_hist/enable/preemptirqsoff ++ ++ If PREEMPT_OFF_HIST is also selected, additional histograms (one ++ per cpu) are generated that accumulate the duration of time periods ++ when both interrupts and preemption are disabled. The histogram data ++ will be located in the debug file system at ++ ++ /sys/kernel/debug/tracing/latency_hist/irqsoff ++ + config PREEMPT_TRACER + bool "Preemption-off Latency Tracer" + default n +@@ -206,6 +224,24 @@ config PREEMPT_TRACER + enabled. This option and the irqs-off timing option can be + used together or separately.) + ++config PREEMPT_OFF_HIST ++ bool "Preemption-off Latency Histogram" ++ depends on PREEMPT_TRACER ++ help ++ This option generates continuously updated histograms (one per cpu) ++ of the duration of time periods with preemption disabled. The ++ histograms are disabled by default. To enable them, write a non-zero ++ number to ++ ++ /sys/kernel/debug/tracing/latency_hist/enable/preemptirqsoff ++ ++ If INTERRUPT_OFF_HIST is also selected, additional histograms (one ++ per cpu) are generated that accumulate the duration of time periods ++ when both interrupts and preemption are disabled. The histogram data ++ will be located in the debug file system at ++ ++ /sys/kernel/debug/tracing/latency_hist/preemptoff ++ + config SCHED_TRACER + bool "Scheduling Latency Tracer" + select GENERIC_TRACER +@@ -251,6 +287,74 @@ config HWLAT_TRACER + file. Every time a latency is greater than tracing_thresh, it will + be recorded into the ring buffer. + ++config WAKEUP_LATENCY_HIST ++ bool "Scheduling Latency Histogram" ++ depends on SCHED_TRACER ++ help ++ This option generates continuously updated histograms (one per cpu) ++ of the scheduling latency of the highest priority task. ++ The histograms are disabled by default. To enable them, write a ++ non-zero number to ++ ++ /sys/kernel/debug/tracing/latency_hist/enable/wakeup ++ ++ Two different algorithms are used, one to determine the latency of ++ processes that exclusively use the highest priority of the system and ++ another one to determine the latency of processes that share the ++ highest system priority with other processes. The former is used to ++ improve hardware and system software, the latter to optimize the ++ priority design of a given system. The histogram data will be ++ located in the debug file system at ++ ++ /sys/kernel/debug/tracing/latency_hist/wakeup ++ ++ and ++ ++ /sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio ++ ++ If both Scheduling Latency Histogram and Missed Timer Offsets ++ Histogram are selected, additional histogram data will be collected ++ that contain, in addition to the wakeup latency, the timer latency, in ++ case the wakeup was triggered by an expired timer. These histograms ++ are available in the ++ ++ /sys/kernel/debug/tracing/latency_hist/timerandwakeup ++ ++ directory. They reflect the apparent interrupt and scheduling latency ++ and are best suitable to determine the worst-case latency of a given ++ system. To enable these histograms, write a non-zero number to ++ ++ /sys/kernel/debug/tracing/latency_hist/enable/timerandwakeup ++ ++config MISSED_TIMER_OFFSETS_HIST ++ depends on HIGH_RES_TIMERS ++ select GENERIC_TRACER ++ bool "Missed Timer Offsets Histogram" ++ help ++ Generate a histogram of missed timer offsets in microseconds. The ++ histograms are disabled by default. To enable them, write a non-zero ++ number to ++ ++ /sys/kernel/debug/tracing/latency_hist/enable/missed_timer_offsets ++ ++ The histogram data will be located in the debug file system at ++ ++ /sys/kernel/debug/tracing/latency_hist/missed_timer_offsets ++ ++ If both Scheduling Latency Histogram and Missed Timer Offsets ++ Histogram are selected, additional histogram data will be collected ++ that contain, in addition to the wakeup latency, the timer latency, in ++ case the wakeup was triggered by an expired timer. These histograms ++ are available in the ++ ++ /sys/kernel/debug/tracing/latency_hist/timerandwakeup ++ ++ directory. They reflect the apparent interrupt and scheduling latency ++ and are best suitable to determine the worst-case latency of a given ++ system. To enable these histograms, write a non-zero number to ++ ++ /sys/kernel/debug/tracing/latency_hist/enable/timerandwakeup ++ + config ENABLE_DEFAULT_TRACERS + bool "Trace process context switches and events" + depends on !GENERIC_TRACER +diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile +index e57980845549..83af000b783c 100644 +--- a/kernel/trace/Makefile ++++ b/kernel/trace/Makefile +@@ -38,6 +38,10 @@ obj-$(CONFIG_IRQSOFF_TRACER) += trace_irqsoff.o + obj-$(CONFIG_PREEMPT_TRACER) += trace_irqsoff.o + obj-$(CONFIG_SCHED_TRACER) += trace_sched_wakeup.o + obj-$(CONFIG_HWLAT_TRACER) += trace_hwlat.o ++obj-$(CONFIG_INTERRUPT_OFF_HIST) += latency_hist.o ++obj-$(CONFIG_PREEMPT_OFF_HIST) += latency_hist.o ++obj-$(CONFIG_WAKEUP_LATENCY_HIST) += latency_hist.o ++obj-$(CONFIG_MISSED_TIMER_OFFSETS_HIST) += latency_hist.o + obj-$(CONFIG_NOP_TRACER) += trace_nop.o + obj-$(CONFIG_STACK_TRACER) += trace_stack.o + obj-$(CONFIG_MMIOTRACE) += trace_mmiotrace.o +diff --git a/kernel/trace/latency_hist.c b/kernel/trace/latency_hist.c +new file mode 100644 +index 000000000000..7f6ee70dea41 +--- /dev/null ++++ b/kernel/trace/latency_hist.c +@@ -0,0 +1,1178 @@ ++/* ++ * kernel/trace/latency_hist.c ++ * ++ * Add support for histograms of preemption-off latency and ++ * interrupt-off latency and wakeup latency, it depends on ++ * Real-Time Preemption Support. ++ * ++ * Copyright (C) 2005 MontaVista Software, Inc. ++ * Yi Yang ++ * ++ * Converted to work with the new latency tracer. ++ * Copyright (C) 2008 Red Hat, Inc. ++ * Steven Rostedt ++ * ++ */ ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++ ++#include "trace.h" ++#include ++ ++#define NSECS_PER_USECS 1000L ++ ++#define CREATE_TRACE_POINTS ++#include ++ ++enum { ++ IRQSOFF_LATENCY = 0, ++ PREEMPTOFF_LATENCY, ++ PREEMPTIRQSOFF_LATENCY, ++ WAKEUP_LATENCY, ++ WAKEUP_LATENCY_SHAREDPRIO, ++ MISSED_TIMER_OFFSETS, ++ TIMERANDWAKEUP_LATENCY, ++ MAX_LATENCY_TYPE, ++}; ++ ++#define MAX_ENTRY_NUM 10240 ++ ++struct hist_data { ++ atomic_t hist_mode; /* 0 log, 1 don't log */ ++ long offset; /* set it to MAX_ENTRY_NUM/2 for a bipolar scale */ ++ long min_lat; ++ long max_lat; ++ unsigned long long below_hist_bound_samples; ++ unsigned long long above_hist_bound_samples; ++ long long accumulate_lat; ++ unsigned long long total_samples; ++ unsigned long long hist_array[MAX_ENTRY_NUM]; ++}; ++ ++struct enable_data { ++ int latency_type; ++ int enabled; ++}; ++ ++static char *latency_hist_dir_root = "latency_hist"; ++ ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++static DEFINE_PER_CPU(struct hist_data, irqsoff_hist); ++static char *irqsoff_hist_dir = "irqsoff"; ++static DEFINE_PER_CPU(cycles_t, hist_irqsoff_start); ++static DEFINE_PER_CPU(int, hist_irqsoff_counting); ++#endif ++ ++#ifdef CONFIG_PREEMPT_OFF_HIST ++static DEFINE_PER_CPU(struct hist_data, preemptoff_hist); ++static char *preemptoff_hist_dir = "preemptoff"; ++static DEFINE_PER_CPU(cycles_t, hist_preemptoff_start); ++static DEFINE_PER_CPU(int, hist_preemptoff_counting); ++#endif ++ ++#if defined(CONFIG_PREEMPT_OFF_HIST) && defined(CONFIG_INTERRUPT_OFF_HIST) ++static DEFINE_PER_CPU(struct hist_data, preemptirqsoff_hist); ++static char *preemptirqsoff_hist_dir = "preemptirqsoff"; ++static DEFINE_PER_CPU(cycles_t, hist_preemptirqsoff_start); ++static DEFINE_PER_CPU(int, hist_preemptirqsoff_counting); ++#endif ++ ++#if defined(CONFIG_PREEMPT_OFF_HIST) || defined(CONFIG_INTERRUPT_OFF_HIST) ++static notrace void probe_preemptirqsoff_hist(void *v, int reason, int start); ++static struct enable_data preemptirqsoff_enabled_data = { ++ .latency_type = PREEMPTIRQSOFF_LATENCY, ++ .enabled = 0, ++}; ++#endif ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++struct maxlatproc_data { ++ char comm[FIELD_SIZEOF(struct task_struct, comm)]; ++ char current_comm[FIELD_SIZEOF(struct task_struct, comm)]; ++ int pid; ++ int current_pid; ++ int prio; ++ int current_prio; ++ long latency; ++ long timeroffset; ++ cycle_t timestamp; ++}; ++#endif ++ ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++static DEFINE_PER_CPU(struct hist_data, wakeup_latency_hist); ++static DEFINE_PER_CPU(struct hist_data, wakeup_latency_hist_sharedprio); ++static char *wakeup_latency_hist_dir = "wakeup"; ++static char *wakeup_latency_hist_dir_sharedprio = "sharedprio"; ++static notrace void probe_wakeup_latency_hist_start(void *v, ++ struct task_struct *p); ++static notrace void probe_wakeup_latency_hist_stop(void *v, ++ bool preempt, struct task_struct *prev, struct task_struct *next); ++static notrace void probe_sched_migrate_task(void *, ++ struct task_struct *task, int cpu); ++static struct enable_data wakeup_latency_enabled_data = { ++ .latency_type = WAKEUP_LATENCY, ++ .enabled = 0, ++}; ++static DEFINE_PER_CPU(struct maxlatproc_data, wakeup_maxlatproc); ++static DEFINE_PER_CPU(struct maxlatproc_data, wakeup_maxlatproc_sharedprio); ++static DEFINE_PER_CPU(struct task_struct *, wakeup_task); ++static DEFINE_PER_CPU(int, wakeup_sharedprio); ++static unsigned long wakeup_pid; ++#endif ++ ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++static DEFINE_PER_CPU(struct hist_data, missed_timer_offsets); ++static char *missed_timer_offsets_dir = "missed_timer_offsets"; ++static notrace void probe_hrtimer_interrupt(void *v, int cpu, ++ long long offset, struct task_struct *curr, struct task_struct *task); ++static struct enable_data missed_timer_offsets_enabled_data = { ++ .latency_type = MISSED_TIMER_OFFSETS, ++ .enabled = 0, ++}; ++static DEFINE_PER_CPU(struct maxlatproc_data, missed_timer_offsets_maxlatproc); ++static unsigned long missed_timer_offsets_pid; ++#endif ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++static DEFINE_PER_CPU(struct hist_data, timerandwakeup_latency_hist); ++static char *timerandwakeup_latency_hist_dir = "timerandwakeup"; ++static struct enable_data timerandwakeup_enabled_data = { ++ .latency_type = TIMERANDWAKEUP_LATENCY, ++ .enabled = 0, ++}; ++static DEFINE_PER_CPU(struct maxlatproc_data, timerandwakeup_maxlatproc); ++#endif ++ ++void notrace latency_hist(int latency_type, int cpu, long latency, ++ long timeroffset, cycle_t stop, ++ struct task_struct *p) ++{ ++ struct hist_data *my_hist; ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ struct maxlatproc_data *mp = NULL; ++#endif ++ ++ if (!cpu_possible(cpu) || latency_type < 0 || ++ latency_type >= MAX_LATENCY_TYPE) ++ return; ++ ++ switch (latency_type) { ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++ case IRQSOFF_LATENCY: ++ my_hist = &per_cpu(irqsoff_hist, cpu); ++ break; ++#endif ++#ifdef CONFIG_PREEMPT_OFF_HIST ++ case PREEMPTOFF_LATENCY: ++ my_hist = &per_cpu(preemptoff_hist, cpu); ++ break; ++#endif ++#if defined(CONFIG_PREEMPT_OFF_HIST) && defined(CONFIG_INTERRUPT_OFF_HIST) ++ case PREEMPTIRQSOFF_LATENCY: ++ my_hist = &per_cpu(preemptirqsoff_hist, cpu); ++ break; ++#endif ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ case WAKEUP_LATENCY: ++ my_hist = &per_cpu(wakeup_latency_hist, cpu); ++ mp = &per_cpu(wakeup_maxlatproc, cpu); ++ break; ++ case WAKEUP_LATENCY_SHAREDPRIO: ++ my_hist = &per_cpu(wakeup_latency_hist_sharedprio, cpu); ++ mp = &per_cpu(wakeup_maxlatproc_sharedprio, cpu); ++ break; ++#endif ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ case MISSED_TIMER_OFFSETS: ++ my_hist = &per_cpu(missed_timer_offsets, cpu); ++ mp = &per_cpu(missed_timer_offsets_maxlatproc, cpu); ++ break; ++#endif ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ case TIMERANDWAKEUP_LATENCY: ++ my_hist = &per_cpu(timerandwakeup_latency_hist, cpu); ++ mp = &per_cpu(timerandwakeup_maxlatproc, cpu); ++ break; ++#endif ++ ++ default: ++ return; ++ } ++ ++ latency += my_hist->offset; ++ ++ if (atomic_read(&my_hist->hist_mode) == 0) ++ return; ++ ++ if (latency < 0 || latency >= MAX_ENTRY_NUM) { ++ if (latency < 0) ++ my_hist->below_hist_bound_samples++; ++ else ++ my_hist->above_hist_bound_samples++; ++ } else ++ my_hist->hist_array[latency]++; ++ ++ if (unlikely(latency > my_hist->max_lat || ++ my_hist->min_lat == LONG_MAX)) { ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ if (latency_type == WAKEUP_LATENCY || ++ latency_type == WAKEUP_LATENCY_SHAREDPRIO || ++ latency_type == MISSED_TIMER_OFFSETS || ++ latency_type == TIMERANDWAKEUP_LATENCY) { ++ strncpy(mp->comm, p->comm, sizeof(mp->comm)); ++ strncpy(mp->current_comm, current->comm, ++ sizeof(mp->current_comm)); ++ mp->pid = task_pid_nr(p); ++ mp->current_pid = task_pid_nr(current); ++ mp->prio = p->prio; ++ mp->current_prio = current->prio; ++ mp->latency = latency; ++ mp->timeroffset = timeroffset; ++ mp->timestamp = stop; ++ } ++#endif ++ my_hist->max_lat = latency; ++ } ++ if (unlikely(latency < my_hist->min_lat)) ++ my_hist->min_lat = latency; ++ my_hist->total_samples++; ++ my_hist->accumulate_lat += latency; ++} ++ ++static void *l_start(struct seq_file *m, loff_t *pos) ++{ ++ loff_t *index_ptr = NULL; ++ loff_t index = *pos; ++ struct hist_data *my_hist = m->private; ++ ++ if (index == 0) { ++ char minstr[32], avgstr[32], maxstr[32]; ++ ++ atomic_dec(&my_hist->hist_mode); ++ ++ if (likely(my_hist->total_samples)) { ++ long avg = (long) div64_s64(my_hist->accumulate_lat, ++ my_hist->total_samples); ++ snprintf(minstr, sizeof(minstr), "%ld", ++ my_hist->min_lat - my_hist->offset); ++ snprintf(avgstr, sizeof(avgstr), "%ld", ++ avg - my_hist->offset); ++ snprintf(maxstr, sizeof(maxstr), "%ld", ++ my_hist->max_lat - my_hist->offset); ++ } else { ++ strcpy(minstr, ""); ++ strcpy(avgstr, minstr); ++ strcpy(maxstr, minstr); ++ } ++ ++ seq_printf(m, "#Minimum latency: %s microseconds\n" ++ "#Average latency: %s microseconds\n" ++ "#Maximum latency: %s microseconds\n" ++ "#Total samples: %llu\n" ++ "#There are %llu samples lower than %ld" ++ " microseconds.\n" ++ "#There are %llu samples greater or equal" ++ " than %ld microseconds.\n" ++ "#usecs\t%16s\n", ++ minstr, avgstr, maxstr, ++ my_hist->total_samples, ++ my_hist->below_hist_bound_samples, ++ -my_hist->offset, ++ my_hist->above_hist_bound_samples, ++ MAX_ENTRY_NUM - my_hist->offset, ++ "samples"); ++ } ++ if (index < MAX_ENTRY_NUM) { ++ index_ptr = kmalloc(sizeof(loff_t), GFP_KERNEL); ++ if (index_ptr) ++ *index_ptr = index; ++ } ++ ++ return index_ptr; ++} ++ ++static void *l_next(struct seq_file *m, void *p, loff_t *pos) ++{ ++ loff_t *index_ptr = p; ++ struct hist_data *my_hist = m->private; ++ ++ if (++*pos >= MAX_ENTRY_NUM) { ++ atomic_inc(&my_hist->hist_mode); ++ return NULL; ++ } ++ *index_ptr = *pos; ++ return index_ptr; ++} ++ ++static void l_stop(struct seq_file *m, void *p) ++{ ++ kfree(p); ++} ++ ++static int l_show(struct seq_file *m, void *p) ++{ ++ int index = *(loff_t *) p; ++ struct hist_data *my_hist = m->private; ++ ++ seq_printf(m, "%6ld\t%16llu\n", index - my_hist->offset, ++ my_hist->hist_array[index]); ++ return 0; ++} ++ ++static const struct seq_operations latency_hist_seq_op = { ++ .start = l_start, ++ .next = l_next, ++ .stop = l_stop, ++ .show = l_show ++}; ++ ++static int latency_hist_open(struct inode *inode, struct file *file) ++{ ++ int ret; ++ ++ ret = seq_open(file, &latency_hist_seq_op); ++ if (!ret) { ++ struct seq_file *seq = file->private_data; ++ seq->private = inode->i_private; ++ } ++ return ret; ++} ++ ++static const struct file_operations latency_hist_fops = { ++ .open = latency_hist_open, ++ .read = seq_read, ++ .llseek = seq_lseek, ++ .release = seq_release, ++}; ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++static void clear_maxlatprocdata(struct maxlatproc_data *mp) ++{ ++ mp->comm[0] = mp->current_comm[0] = '\0'; ++ mp->prio = mp->current_prio = mp->pid = mp->current_pid = ++ mp->latency = mp->timeroffset = -1; ++ mp->timestamp = 0; ++} ++#endif ++ ++static void hist_reset(struct hist_data *hist) ++{ ++ atomic_dec(&hist->hist_mode); ++ ++ memset(hist->hist_array, 0, sizeof(hist->hist_array)); ++ hist->below_hist_bound_samples = 0ULL; ++ hist->above_hist_bound_samples = 0ULL; ++ hist->min_lat = LONG_MAX; ++ hist->max_lat = LONG_MIN; ++ hist->total_samples = 0ULL; ++ hist->accumulate_lat = 0LL; ++ ++ atomic_inc(&hist->hist_mode); ++} ++ ++static ssize_t ++latency_hist_reset(struct file *file, const char __user *a, ++ size_t size, loff_t *off) ++{ ++ int cpu; ++ struct hist_data *hist = NULL; ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ struct maxlatproc_data *mp = NULL; ++#endif ++ off_t latency_type = (off_t) file->private_data; ++ ++ for_each_online_cpu(cpu) { ++ ++ switch (latency_type) { ++#ifdef CONFIG_PREEMPT_OFF_HIST ++ case PREEMPTOFF_LATENCY: ++ hist = &per_cpu(preemptoff_hist, cpu); ++ break; ++#endif ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++ case IRQSOFF_LATENCY: ++ hist = &per_cpu(irqsoff_hist, cpu); ++ break; ++#endif ++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) ++ case PREEMPTIRQSOFF_LATENCY: ++ hist = &per_cpu(preemptirqsoff_hist, cpu); ++ break; ++#endif ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ case WAKEUP_LATENCY: ++ hist = &per_cpu(wakeup_latency_hist, cpu); ++ mp = &per_cpu(wakeup_maxlatproc, cpu); ++ break; ++ case WAKEUP_LATENCY_SHAREDPRIO: ++ hist = &per_cpu(wakeup_latency_hist_sharedprio, cpu); ++ mp = &per_cpu(wakeup_maxlatproc_sharedprio, cpu); ++ break; ++#endif ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ case MISSED_TIMER_OFFSETS: ++ hist = &per_cpu(missed_timer_offsets, cpu); ++ mp = &per_cpu(missed_timer_offsets_maxlatproc, cpu); ++ break; ++#endif ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ case TIMERANDWAKEUP_LATENCY: ++ hist = &per_cpu(timerandwakeup_latency_hist, cpu); ++ mp = &per_cpu(timerandwakeup_maxlatproc, cpu); ++ break; ++#endif ++ } ++ ++ hist_reset(hist); ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ if (latency_type == WAKEUP_LATENCY || ++ latency_type == WAKEUP_LATENCY_SHAREDPRIO || ++ latency_type == MISSED_TIMER_OFFSETS || ++ latency_type == TIMERANDWAKEUP_LATENCY) ++ clear_maxlatprocdata(mp); ++#endif ++ } ++ ++ return size; ++} ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++static ssize_t ++show_pid(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos) ++{ ++ char buf[64]; ++ int r; ++ unsigned long *this_pid = file->private_data; ++ ++ r = snprintf(buf, sizeof(buf), "%lu\n", *this_pid); ++ return simple_read_from_buffer(ubuf, cnt, ppos, buf, r); ++} ++ ++static ssize_t do_pid(struct file *file, const char __user *ubuf, ++ size_t cnt, loff_t *ppos) ++{ ++ char buf[64]; ++ unsigned long pid; ++ unsigned long *this_pid = file->private_data; ++ ++ if (cnt >= sizeof(buf)) ++ return -EINVAL; ++ ++ if (copy_from_user(&buf, ubuf, cnt)) ++ return -EFAULT; ++ ++ buf[cnt] = '\0'; ++ ++ if (kstrtoul(buf, 10, &pid)) ++ return -EINVAL; ++ ++ *this_pid = pid; ++ ++ return cnt; ++} ++#endif ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++static ssize_t ++show_maxlatproc(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos) ++{ ++ int r; ++ struct maxlatproc_data *mp = file->private_data; ++ int strmaxlen = (TASK_COMM_LEN * 2) + (8 * 8); ++ unsigned long long t; ++ unsigned long usecs, secs; ++ char *buf; ++ ++ if (mp->pid == -1 || mp->current_pid == -1) { ++ buf = "(none)\n"; ++ return simple_read_from_buffer(ubuf, cnt, ppos, buf, ++ strlen(buf)); ++ } ++ ++ buf = kmalloc(strmaxlen, GFP_KERNEL); ++ if (buf == NULL) ++ return -ENOMEM; ++ ++ t = ns2usecs(mp->timestamp); ++ usecs = do_div(t, USEC_PER_SEC); ++ secs = (unsigned long) t; ++ r = snprintf(buf, strmaxlen, ++ "%d %d %ld (%ld) %s <- %d %d %s %lu.%06lu\n", mp->pid, ++ MAX_RT_PRIO-1 - mp->prio, mp->latency, mp->timeroffset, mp->comm, ++ mp->current_pid, MAX_RT_PRIO-1 - mp->current_prio, mp->current_comm, ++ secs, usecs); ++ r = simple_read_from_buffer(ubuf, cnt, ppos, buf, r); ++ kfree(buf); ++ return r; ++} ++#endif ++ ++static ssize_t ++show_enable(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos) ++{ ++ char buf[64]; ++ struct enable_data *ed = file->private_data; ++ int r; ++ ++ r = snprintf(buf, sizeof(buf), "%d\n", ed->enabled); ++ return simple_read_from_buffer(ubuf, cnt, ppos, buf, r); ++} ++ ++static ssize_t ++do_enable(struct file *file, const char __user *ubuf, size_t cnt, loff_t *ppos) ++{ ++ char buf[64]; ++ long enable; ++ struct enable_data *ed = file->private_data; ++ ++ if (cnt >= sizeof(buf)) ++ return -EINVAL; ++ ++ if (copy_from_user(&buf, ubuf, cnt)) ++ return -EFAULT; ++ ++ buf[cnt] = 0; ++ ++ if (kstrtoul(buf, 10, &enable)) ++ return -EINVAL; ++ ++ if ((enable && ed->enabled) || (!enable && !ed->enabled)) ++ return cnt; ++ ++ if (enable) { ++ int ret; ++ ++ switch (ed->latency_type) { ++#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST) ++ case PREEMPTIRQSOFF_LATENCY: ++ ret = register_trace_preemptirqsoff_hist( ++ probe_preemptirqsoff_hist, NULL); ++ if (ret) { ++ pr_info("wakeup trace: Couldn't assign " ++ "probe_preemptirqsoff_hist " ++ "to trace_preemptirqsoff_hist\n"); ++ return ret; ++ } ++ break; ++#endif ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ case WAKEUP_LATENCY: ++ ret = register_trace_sched_wakeup( ++ probe_wakeup_latency_hist_start, NULL); ++ if (ret) { ++ pr_info("wakeup trace: Couldn't assign " ++ "probe_wakeup_latency_hist_start " ++ "to trace_sched_wakeup\n"); ++ return ret; ++ } ++ ret = register_trace_sched_wakeup_new( ++ probe_wakeup_latency_hist_start, NULL); ++ if (ret) { ++ pr_info("wakeup trace: Couldn't assign " ++ "probe_wakeup_latency_hist_start " ++ "to trace_sched_wakeup_new\n"); ++ unregister_trace_sched_wakeup( ++ probe_wakeup_latency_hist_start, NULL); ++ return ret; ++ } ++ ret = register_trace_sched_switch( ++ probe_wakeup_latency_hist_stop, NULL); ++ if (ret) { ++ pr_info("wakeup trace: Couldn't assign " ++ "probe_wakeup_latency_hist_stop " ++ "to trace_sched_switch\n"); ++ unregister_trace_sched_wakeup( ++ probe_wakeup_latency_hist_start, NULL); ++ unregister_trace_sched_wakeup_new( ++ probe_wakeup_latency_hist_start, NULL); ++ return ret; ++ } ++ ret = register_trace_sched_migrate_task( ++ probe_sched_migrate_task, NULL); ++ if (ret) { ++ pr_info("wakeup trace: Couldn't assign " ++ "probe_sched_migrate_task " ++ "to trace_sched_migrate_task\n"); ++ unregister_trace_sched_wakeup( ++ probe_wakeup_latency_hist_start, NULL); ++ unregister_trace_sched_wakeup_new( ++ probe_wakeup_latency_hist_start, NULL); ++ unregister_trace_sched_switch( ++ probe_wakeup_latency_hist_stop, NULL); ++ return ret; ++ } ++ break; ++#endif ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ case MISSED_TIMER_OFFSETS: ++ ret = register_trace_hrtimer_interrupt( ++ probe_hrtimer_interrupt, NULL); ++ if (ret) { ++ pr_info("wakeup trace: Couldn't assign " ++ "probe_hrtimer_interrupt " ++ "to trace_hrtimer_interrupt\n"); ++ return ret; ++ } ++ break; ++#endif ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ case TIMERANDWAKEUP_LATENCY: ++ if (!wakeup_latency_enabled_data.enabled || ++ !missed_timer_offsets_enabled_data.enabled) ++ return -EINVAL; ++ break; ++#endif ++ default: ++ break; ++ } ++ } else { ++ switch (ed->latency_type) { ++#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST) ++ case PREEMPTIRQSOFF_LATENCY: ++ { ++ int cpu; ++ ++ unregister_trace_preemptirqsoff_hist( ++ probe_preemptirqsoff_hist, NULL); ++ for_each_online_cpu(cpu) { ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++ per_cpu(hist_irqsoff_counting, ++ cpu) = 0; ++#endif ++#ifdef CONFIG_PREEMPT_OFF_HIST ++ per_cpu(hist_preemptoff_counting, ++ cpu) = 0; ++#endif ++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) ++ per_cpu(hist_preemptirqsoff_counting, ++ cpu) = 0; ++#endif ++ } ++ } ++ break; ++#endif ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ case WAKEUP_LATENCY: ++ { ++ int cpu; ++ ++ unregister_trace_sched_wakeup( ++ probe_wakeup_latency_hist_start, NULL); ++ unregister_trace_sched_wakeup_new( ++ probe_wakeup_latency_hist_start, NULL); ++ unregister_trace_sched_switch( ++ probe_wakeup_latency_hist_stop, NULL); ++ unregister_trace_sched_migrate_task( ++ probe_sched_migrate_task, NULL); ++ ++ for_each_online_cpu(cpu) { ++ per_cpu(wakeup_task, cpu) = NULL; ++ per_cpu(wakeup_sharedprio, cpu) = 0; ++ } ++ } ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ timerandwakeup_enabled_data.enabled = 0; ++#endif ++ break; ++#endif ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ case MISSED_TIMER_OFFSETS: ++ unregister_trace_hrtimer_interrupt( ++ probe_hrtimer_interrupt, NULL); ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ timerandwakeup_enabled_data.enabled = 0; ++#endif ++ break; ++#endif ++ default: ++ break; ++ } ++ } ++ ed->enabled = enable; ++ return cnt; ++} ++ ++static const struct file_operations latency_hist_reset_fops = { ++ .open = tracing_open_generic, ++ .write = latency_hist_reset, ++}; ++ ++static const struct file_operations enable_fops = { ++ .open = tracing_open_generic, ++ .read = show_enable, ++ .write = do_enable, ++}; ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++static const struct file_operations pid_fops = { ++ .open = tracing_open_generic, ++ .read = show_pid, ++ .write = do_pid, ++}; ++ ++static const struct file_operations maxlatproc_fops = { ++ .open = tracing_open_generic, ++ .read = show_maxlatproc, ++}; ++#endif ++ ++#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST) ++static notrace void probe_preemptirqsoff_hist(void *v, int reason, ++ int starthist) ++{ ++ int cpu = raw_smp_processor_id(); ++ int time_set = 0; ++ ++ if (starthist) { ++ cycle_t uninitialized_var(start); ++ ++ if (!preempt_count() && !irqs_disabled()) ++ return; ++ ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++ if ((reason == IRQS_OFF || reason == TRACE_START) && ++ !per_cpu(hist_irqsoff_counting, cpu)) { ++ per_cpu(hist_irqsoff_counting, cpu) = 1; ++ start = ftrace_now(cpu); ++ time_set++; ++ per_cpu(hist_irqsoff_start, cpu) = start; ++ } ++#endif ++ ++#ifdef CONFIG_PREEMPT_OFF_HIST ++ if ((reason == PREEMPT_OFF || reason == TRACE_START) && ++ !per_cpu(hist_preemptoff_counting, cpu)) { ++ per_cpu(hist_preemptoff_counting, cpu) = 1; ++ if (!(time_set++)) ++ start = ftrace_now(cpu); ++ per_cpu(hist_preemptoff_start, cpu) = start; ++ } ++#endif ++ ++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) ++ if (per_cpu(hist_irqsoff_counting, cpu) && ++ per_cpu(hist_preemptoff_counting, cpu) && ++ !per_cpu(hist_preemptirqsoff_counting, cpu)) { ++ per_cpu(hist_preemptirqsoff_counting, cpu) = 1; ++ if (!time_set) ++ start = ftrace_now(cpu); ++ per_cpu(hist_preemptirqsoff_start, cpu) = start; ++ } ++#endif ++ } else { ++ cycle_t uninitialized_var(stop); ++ ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++ if ((reason == IRQS_ON || reason == TRACE_STOP) && ++ per_cpu(hist_irqsoff_counting, cpu)) { ++ cycle_t start = per_cpu(hist_irqsoff_start, cpu); ++ ++ stop = ftrace_now(cpu); ++ time_set++; ++ if (start) { ++ long latency = ((long) (stop - start)) / ++ NSECS_PER_USECS; ++ ++ latency_hist(IRQSOFF_LATENCY, cpu, latency, 0, ++ stop, NULL); ++ } ++ per_cpu(hist_irqsoff_counting, cpu) = 0; ++ } ++#endif ++ ++#ifdef CONFIG_PREEMPT_OFF_HIST ++ if ((reason == PREEMPT_ON || reason == TRACE_STOP) && ++ per_cpu(hist_preemptoff_counting, cpu)) { ++ cycle_t start = per_cpu(hist_preemptoff_start, cpu); ++ ++ if (!(time_set++)) ++ stop = ftrace_now(cpu); ++ if (start) { ++ long latency = ((long) (stop - start)) / ++ NSECS_PER_USECS; ++ ++ latency_hist(PREEMPTOFF_LATENCY, cpu, latency, ++ 0, stop, NULL); ++ } ++ per_cpu(hist_preemptoff_counting, cpu) = 0; ++ } ++#endif ++ ++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) ++ if ((!per_cpu(hist_irqsoff_counting, cpu) || ++ !per_cpu(hist_preemptoff_counting, cpu)) && ++ per_cpu(hist_preemptirqsoff_counting, cpu)) { ++ cycle_t start = per_cpu(hist_preemptirqsoff_start, cpu); ++ ++ if (!time_set) ++ stop = ftrace_now(cpu); ++ if (start) { ++ long latency = ((long) (stop - start)) / ++ NSECS_PER_USECS; ++ ++ latency_hist(PREEMPTIRQSOFF_LATENCY, cpu, ++ latency, 0, stop, NULL); ++ } ++ per_cpu(hist_preemptirqsoff_counting, cpu) = 0; ++ } ++#endif ++ } ++} ++#endif ++ ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++static DEFINE_RAW_SPINLOCK(wakeup_lock); ++static notrace void probe_sched_migrate_task(void *v, struct task_struct *task, ++ int cpu) ++{ ++ int old_cpu = task_cpu(task); ++ ++ if (cpu != old_cpu) { ++ unsigned long flags; ++ struct task_struct *cpu_wakeup_task; ++ ++ raw_spin_lock_irqsave(&wakeup_lock, flags); ++ ++ cpu_wakeup_task = per_cpu(wakeup_task, old_cpu); ++ if (task == cpu_wakeup_task) { ++ put_task_struct(cpu_wakeup_task); ++ per_cpu(wakeup_task, old_cpu) = NULL; ++ cpu_wakeup_task = per_cpu(wakeup_task, cpu) = task; ++ get_task_struct(cpu_wakeup_task); ++ } ++ ++ raw_spin_unlock_irqrestore(&wakeup_lock, flags); ++ } ++} ++ ++static notrace void probe_wakeup_latency_hist_start(void *v, ++ struct task_struct *p) ++{ ++ unsigned long flags; ++ struct task_struct *curr = current; ++ int cpu = task_cpu(p); ++ struct task_struct *cpu_wakeup_task; ++ ++ raw_spin_lock_irqsave(&wakeup_lock, flags); ++ ++ cpu_wakeup_task = per_cpu(wakeup_task, cpu); ++ ++ if (wakeup_pid) { ++ if ((cpu_wakeup_task && p->prio == cpu_wakeup_task->prio) || ++ p->prio == curr->prio) ++ per_cpu(wakeup_sharedprio, cpu) = 1; ++ if (likely(wakeup_pid != task_pid_nr(p))) ++ goto out; ++ } else { ++ if (likely(!rt_task(p)) || ++ (cpu_wakeup_task && p->prio > cpu_wakeup_task->prio) || ++ p->prio > curr->prio) ++ goto out; ++ if ((cpu_wakeup_task && p->prio == cpu_wakeup_task->prio) || ++ p->prio == curr->prio) ++ per_cpu(wakeup_sharedprio, cpu) = 1; ++ } ++ ++ if (cpu_wakeup_task) ++ put_task_struct(cpu_wakeup_task); ++ cpu_wakeup_task = per_cpu(wakeup_task, cpu) = p; ++ get_task_struct(cpu_wakeup_task); ++ cpu_wakeup_task->preempt_timestamp_hist = ++ ftrace_now(raw_smp_processor_id()); ++out: ++ raw_spin_unlock_irqrestore(&wakeup_lock, flags); ++} ++ ++static notrace void probe_wakeup_latency_hist_stop(void *v, ++ bool preempt, struct task_struct *prev, struct task_struct *next) ++{ ++ unsigned long flags; ++ int cpu = task_cpu(next); ++ long latency; ++ cycle_t stop; ++ struct task_struct *cpu_wakeup_task; ++ ++ raw_spin_lock_irqsave(&wakeup_lock, flags); ++ ++ cpu_wakeup_task = per_cpu(wakeup_task, cpu); ++ ++ if (cpu_wakeup_task == NULL) ++ goto out; ++ ++ /* Already running? */ ++ if (unlikely(current == cpu_wakeup_task)) ++ goto out_reset; ++ ++ if (next != cpu_wakeup_task) { ++ if (next->prio < cpu_wakeup_task->prio) ++ goto out_reset; ++ ++ if (next->prio == cpu_wakeup_task->prio) ++ per_cpu(wakeup_sharedprio, cpu) = 1; ++ ++ goto out; ++ } ++ ++ if (current->prio == cpu_wakeup_task->prio) ++ per_cpu(wakeup_sharedprio, cpu) = 1; ++ ++ /* ++ * The task we are waiting for is about to be switched to. ++ * Calculate latency and store it in histogram. ++ */ ++ stop = ftrace_now(raw_smp_processor_id()); ++ ++ latency = ((long) (stop - next->preempt_timestamp_hist)) / ++ NSECS_PER_USECS; ++ ++ if (per_cpu(wakeup_sharedprio, cpu)) { ++ latency_hist(WAKEUP_LATENCY_SHAREDPRIO, cpu, latency, 0, stop, ++ next); ++ per_cpu(wakeup_sharedprio, cpu) = 0; ++ } else { ++ latency_hist(WAKEUP_LATENCY, cpu, latency, 0, stop, next); ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ if (timerandwakeup_enabled_data.enabled) { ++ latency_hist(TIMERANDWAKEUP_LATENCY, cpu, ++ next->timer_offset + latency, next->timer_offset, ++ stop, next); ++ } ++#endif ++ } ++ ++out_reset: ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ next->timer_offset = 0; ++#endif ++ put_task_struct(cpu_wakeup_task); ++ per_cpu(wakeup_task, cpu) = NULL; ++out: ++ raw_spin_unlock_irqrestore(&wakeup_lock, flags); ++} ++#endif ++ ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++static notrace void probe_hrtimer_interrupt(void *v, int cpu, ++ long long latency_ns, struct task_struct *curr, ++ struct task_struct *task) ++{ ++ if (latency_ns <= 0 && task != NULL && rt_task(task) && ++ (task->prio < curr->prio || ++ (task->prio == curr->prio && ++ !cpumask_test_cpu(cpu, &task->cpus_allowed)))) { ++ long latency; ++ cycle_t now; ++ ++ if (missed_timer_offsets_pid) { ++ if (likely(missed_timer_offsets_pid != ++ task_pid_nr(task))) ++ return; ++ } ++ ++ now = ftrace_now(cpu); ++ latency = (long) div_s64(-latency_ns, NSECS_PER_USECS); ++ latency_hist(MISSED_TIMER_OFFSETS, cpu, latency, latency, now, ++ task); ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ task->timer_offset = latency; ++#endif ++ } ++} ++#endif ++ ++static __init int latency_hist_init(void) ++{ ++ struct dentry *latency_hist_root = NULL; ++ struct dentry *dentry; ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ struct dentry *dentry_sharedprio; ++#endif ++ struct dentry *entry; ++ struct dentry *enable_root; ++ int i = 0; ++ struct hist_data *my_hist; ++ char name[64]; ++ char *cpufmt = "CPU%d"; ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ char *cpufmt_maxlatproc = "max_latency-CPU%d"; ++ struct maxlatproc_data *mp = NULL; ++#endif ++ ++ dentry = tracing_init_dentry(); ++ latency_hist_root = debugfs_create_dir(latency_hist_dir_root, dentry); ++ enable_root = debugfs_create_dir("enable", latency_hist_root); ++ ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++ dentry = debugfs_create_dir(irqsoff_hist_dir, latency_hist_root); ++ for_each_possible_cpu(i) { ++ sprintf(name, cpufmt, i); ++ entry = debugfs_create_file(name, 0444, dentry, ++ &per_cpu(irqsoff_hist, i), &latency_hist_fops); ++ my_hist = &per_cpu(irqsoff_hist, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ } ++ entry = debugfs_create_file("reset", 0644, dentry, ++ (void *)IRQSOFF_LATENCY, &latency_hist_reset_fops); ++#endif ++ ++#ifdef CONFIG_PREEMPT_OFF_HIST ++ dentry = debugfs_create_dir(preemptoff_hist_dir, ++ latency_hist_root); ++ for_each_possible_cpu(i) { ++ sprintf(name, cpufmt, i); ++ entry = debugfs_create_file(name, 0444, dentry, ++ &per_cpu(preemptoff_hist, i), &latency_hist_fops); ++ my_hist = &per_cpu(preemptoff_hist, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ } ++ entry = debugfs_create_file("reset", 0644, dentry, ++ (void *)PREEMPTOFF_LATENCY, &latency_hist_reset_fops); ++#endif ++ ++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) ++ dentry = debugfs_create_dir(preemptirqsoff_hist_dir, ++ latency_hist_root); ++ for_each_possible_cpu(i) { ++ sprintf(name, cpufmt, i); ++ entry = debugfs_create_file(name, 0444, dentry, ++ &per_cpu(preemptirqsoff_hist, i), &latency_hist_fops); ++ my_hist = &per_cpu(preemptirqsoff_hist, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ } ++ entry = debugfs_create_file("reset", 0644, dentry, ++ (void *)PREEMPTIRQSOFF_LATENCY, &latency_hist_reset_fops); ++#endif ++ ++#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST) ++ entry = debugfs_create_file("preemptirqsoff", 0644, ++ enable_root, (void *)&preemptirqsoff_enabled_data, ++ &enable_fops); ++#endif ++ ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ dentry = debugfs_create_dir(wakeup_latency_hist_dir, ++ latency_hist_root); ++ dentry_sharedprio = debugfs_create_dir( ++ wakeup_latency_hist_dir_sharedprio, dentry); ++ for_each_possible_cpu(i) { ++ sprintf(name, cpufmt, i); ++ ++ entry = debugfs_create_file(name, 0444, dentry, ++ &per_cpu(wakeup_latency_hist, i), ++ &latency_hist_fops); ++ my_hist = &per_cpu(wakeup_latency_hist, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ ++ entry = debugfs_create_file(name, 0444, dentry_sharedprio, ++ &per_cpu(wakeup_latency_hist_sharedprio, i), ++ &latency_hist_fops); ++ my_hist = &per_cpu(wakeup_latency_hist_sharedprio, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ ++ sprintf(name, cpufmt_maxlatproc, i); ++ ++ mp = &per_cpu(wakeup_maxlatproc, i); ++ entry = debugfs_create_file(name, 0444, dentry, mp, ++ &maxlatproc_fops); ++ clear_maxlatprocdata(mp); ++ ++ mp = &per_cpu(wakeup_maxlatproc_sharedprio, i); ++ entry = debugfs_create_file(name, 0444, dentry_sharedprio, mp, ++ &maxlatproc_fops); ++ clear_maxlatprocdata(mp); ++ } ++ entry = debugfs_create_file("pid", 0644, dentry, ++ (void *)&wakeup_pid, &pid_fops); ++ entry = debugfs_create_file("reset", 0644, dentry, ++ (void *)WAKEUP_LATENCY, &latency_hist_reset_fops); ++ entry = debugfs_create_file("reset", 0644, dentry_sharedprio, ++ (void *)WAKEUP_LATENCY_SHAREDPRIO, &latency_hist_reset_fops); ++ entry = debugfs_create_file("wakeup", 0644, ++ enable_root, (void *)&wakeup_latency_enabled_data, ++ &enable_fops); ++#endif ++ ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ dentry = debugfs_create_dir(missed_timer_offsets_dir, ++ latency_hist_root); ++ for_each_possible_cpu(i) { ++ sprintf(name, cpufmt, i); ++ entry = debugfs_create_file(name, 0444, dentry, ++ &per_cpu(missed_timer_offsets, i), &latency_hist_fops); ++ my_hist = &per_cpu(missed_timer_offsets, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ ++ sprintf(name, cpufmt_maxlatproc, i); ++ mp = &per_cpu(missed_timer_offsets_maxlatproc, i); ++ entry = debugfs_create_file(name, 0444, dentry, mp, ++ &maxlatproc_fops); ++ clear_maxlatprocdata(mp); ++ } ++ entry = debugfs_create_file("pid", 0644, dentry, ++ (void *)&missed_timer_offsets_pid, &pid_fops); ++ entry = debugfs_create_file("reset", 0644, dentry, ++ (void *)MISSED_TIMER_OFFSETS, &latency_hist_reset_fops); ++ entry = debugfs_create_file("missed_timer_offsets", 0644, ++ enable_root, (void *)&missed_timer_offsets_enabled_data, ++ &enable_fops); ++#endif ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ dentry = debugfs_create_dir(timerandwakeup_latency_hist_dir, ++ latency_hist_root); ++ for_each_possible_cpu(i) { ++ sprintf(name, cpufmt, i); ++ entry = debugfs_create_file(name, 0444, dentry, ++ &per_cpu(timerandwakeup_latency_hist, i), ++ &latency_hist_fops); ++ my_hist = &per_cpu(timerandwakeup_latency_hist, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ ++ sprintf(name, cpufmt_maxlatproc, i); ++ mp = &per_cpu(timerandwakeup_maxlatproc, i); ++ entry = debugfs_create_file(name, 0444, dentry, mp, ++ &maxlatproc_fops); ++ clear_maxlatprocdata(mp); ++ } ++ entry = debugfs_create_file("reset", 0644, dentry, ++ (void *)TIMERANDWAKEUP_LATENCY, &latency_hist_reset_fops); ++ entry = debugfs_create_file("timerandwakeup", 0644, ++ enable_root, (void *)&timerandwakeup_enabled_data, ++ &enable_fops); ++#endif ++ return 0; ++} ++ ++device_initcall(latency_hist_init); +diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c +index 83c60f9013cb..6fb207964a84 100644 +--- a/kernel/trace/trace.c ++++ b/kernel/trace/trace.c +@@ -1897,6 +1897,7 @@ tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags, + struct task_struct *tsk = current; + + entry->preempt_count = pc & 0xff; ++ entry->preempt_lazy_count = preempt_lazy_count(); + entry->pid = (tsk) ? tsk->pid : 0; + entry->flags = + #ifdef CONFIG_TRACE_IRQFLAGS_SUPPORT +@@ -1907,8 +1908,11 @@ tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags, + ((pc & NMI_MASK ) ? TRACE_FLAG_NMI : 0) | + ((pc & HARDIRQ_MASK) ? TRACE_FLAG_HARDIRQ : 0) | + ((pc & SOFTIRQ_MASK) ? TRACE_FLAG_SOFTIRQ : 0) | +- (tif_need_resched() ? TRACE_FLAG_NEED_RESCHED : 0) | ++ (tif_need_resched_now() ? TRACE_FLAG_NEED_RESCHED : 0) | ++ (need_resched_lazy() ? TRACE_FLAG_NEED_RESCHED_LAZY : 0) | + (test_preempt_need_resched() ? TRACE_FLAG_PREEMPT_RESCHED : 0); ++ ++ entry->migrate_disable = (tsk) ? __migrate_disabled(tsk) & 0xFF : 0; + } + EXPORT_SYMBOL_GPL(tracing_generic_entry_update); + +@@ -2892,14 +2896,17 @@ get_total_entries(struct trace_buffer *buf, + + static void print_lat_help_header(struct seq_file *m) + { +- seq_puts(m, "# _------=> CPU# \n" +- "# / _-----=> irqs-off \n" +- "# | / _----=> need-resched \n" +- "# || / _---=> hardirq/softirq \n" +- "# ||| / _--=> preempt-depth \n" +- "# |||| / delay \n" +- "# cmd pid ||||| time | caller \n" +- "# \\ / ||||| \\ | / \n"); ++ seq_puts(m, "# _--------=> CPU# \n" ++ "# / _-------=> irqs-off \n" ++ "# | / _------=> need-resched \n" ++ "# || / _-----=> need-resched_lazy \n" ++ "# ||| / _----=> hardirq/softirq \n" ++ "# |||| / _---=> preempt-depth \n" ++ "# ||||| / _--=> preempt-lazy-depth\n" ++ "# |||||| / _-=> migrate-disable \n" ++ "# ||||||| / delay \n" ++ "# cmd pid |||||||| time | caller \n" ++ "# \\ / |||||||| \\ | / \n"); + } + + static void print_event_info(struct trace_buffer *buf, struct seq_file *m) +@@ -2925,11 +2932,14 @@ static void print_func_help_header_irq(struct trace_buffer *buf, struct seq_file + print_event_info(buf, m); + seq_puts(m, "# _-----=> irqs-off\n" + "# / _----=> need-resched\n" +- "# | / _---=> hardirq/softirq\n" +- "# || / _--=> preempt-depth\n" +- "# ||| / delay\n" +- "# TASK-PID CPU# |||| TIMESTAMP FUNCTION\n" +- "# | | | |||| | |\n"); ++ "# |/ _-----=> need-resched_lazy\n" ++ "# || / _---=> hardirq/softirq\n" ++ "# ||| / _--=> preempt-depth\n" ++ "# |||| / _-=> preempt-lazy-depth\n" ++ "# ||||| / _-=> migrate-disable \n" ++ "# |||||| / delay\n" ++ "# TASK-PID CPU# ||||||| TIMESTAMP FUNCTION\n" ++ "# | | | ||||||| | |\n"); + } + + void +diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h +index b0d8576c27ae..702b9376b278 100644 +--- a/kernel/trace/trace.h ++++ b/kernel/trace/trace.h +@@ -124,6 +124,7 @@ struct kretprobe_trace_entry_head { + * NEED_RESCHED - reschedule is requested + * HARDIRQ - inside an interrupt handler + * SOFTIRQ - inside a softirq handler ++ * NEED_RESCHED_LAZY - lazy reschedule is requested + */ + enum trace_flag_type { + TRACE_FLAG_IRQS_OFF = 0x01, +@@ -133,6 +134,7 @@ enum trace_flag_type { + TRACE_FLAG_SOFTIRQ = 0x10, + TRACE_FLAG_PREEMPT_RESCHED = 0x20, + TRACE_FLAG_NMI = 0x40, ++ TRACE_FLAG_NEED_RESCHED_LAZY = 0x80, + }; + + #define TRACE_BUF_SIZE 1024 +diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c +index 03c0a48c3ac4..0b85d516b491 100644 +--- a/kernel/trace/trace_events.c ++++ b/kernel/trace/trace_events.c +@@ -187,6 +187,8 @@ static int trace_define_common_fields(void) + __common_field(unsigned char, flags); + __common_field(unsigned char, preempt_count); + __common_field(int, pid); ++ __common_field(unsigned short, migrate_disable); ++ __common_field(unsigned short, padding); + + return ret; + } +diff --git a/kernel/trace/trace_irqsoff.c b/kernel/trace/trace_irqsoff.c +index 03cdff84d026..940bd10b4406 100644 +--- a/kernel/trace/trace_irqsoff.c ++++ b/kernel/trace/trace_irqsoff.c +@@ -13,6 +13,7 @@ + #include + #include + #include ++#include + + #include "trace.h" + +@@ -424,11 +425,13 @@ void start_critical_timings(void) + { + if (preempt_trace() || irq_trace()) + start_critical_timing(CALLER_ADDR0, CALLER_ADDR1); ++ trace_preemptirqsoff_hist_rcuidle(TRACE_START, 1); + } + EXPORT_SYMBOL_GPL(start_critical_timings); + + void stop_critical_timings(void) + { ++ trace_preemptirqsoff_hist_rcuidle(TRACE_STOP, 0); + if (preempt_trace() || irq_trace()) + stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1); + } +@@ -438,6 +441,7 @@ EXPORT_SYMBOL_GPL(stop_critical_timings); + #ifdef CONFIG_PROVE_LOCKING + void time_hardirqs_on(unsigned long a0, unsigned long a1) + { ++ trace_preemptirqsoff_hist_rcuidle(IRQS_ON, 0); + if (!preempt_trace() && irq_trace()) + stop_critical_timing(a0, a1); + } +@@ -446,6 +450,7 @@ void time_hardirqs_off(unsigned long a0, unsigned long a1) + { + if (!preempt_trace() && irq_trace()) + start_critical_timing(a0, a1); ++ trace_preemptirqsoff_hist_rcuidle(IRQS_OFF, 1); + } + + #else /* !CONFIG_PROVE_LOCKING */ +@@ -471,6 +476,7 @@ inline void print_irqtrace_events(struct task_struct *curr) + */ + void trace_hardirqs_on(void) + { ++ trace_preemptirqsoff_hist(IRQS_ON, 0); + if (!preempt_trace() && irq_trace()) + stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1); + } +@@ -480,11 +486,13 @@ void trace_hardirqs_off(void) + { + if (!preempt_trace() && irq_trace()) + start_critical_timing(CALLER_ADDR0, CALLER_ADDR1); ++ trace_preemptirqsoff_hist(IRQS_OFF, 1); + } + EXPORT_SYMBOL(trace_hardirqs_off); + + __visible void trace_hardirqs_on_caller(unsigned long caller_addr) + { ++ trace_preemptirqsoff_hist(IRQS_ON, 0); + if (!preempt_trace() && irq_trace()) + stop_critical_timing(CALLER_ADDR0, caller_addr); + } +@@ -494,6 +502,7 @@ __visible void trace_hardirqs_off_caller(unsigned long caller_addr) + { + if (!preempt_trace() && irq_trace()) + start_critical_timing(CALLER_ADDR0, caller_addr); ++ trace_preemptirqsoff_hist(IRQS_OFF, 1); + } + EXPORT_SYMBOL(trace_hardirqs_off_caller); + +@@ -503,12 +512,14 @@ EXPORT_SYMBOL(trace_hardirqs_off_caller); + #ifdef CONFIG_PREEMPT_TRACER + void trace_preempt_on(unsigned long a0, unsigned long a1) + { ++ trace_preemptirqsoff_hist(PREEMPT_ON, 0); + if (preempt_trace() && !irq_trace()) + stop_critical_timing(a0, a1); + } + + void trace_preempt_off(unsigned long a0, unsigned long a1) + { ++ trace_preemptirqsoff_hist(PREEMPT_ON, 1); + if (preempt_trace() && !irq_trace()) + start_critical_timing(a0, a1); + } +diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c +index 3fc20422c166..65a6dde71a7d 100644 +--- a/kernel/trace/trace_output.c ++++ b/kernel/trace/trace_output.c +@@ -386,6 +386,7 @@ int trace_print_lat_fmt(struct trace_seq *s, struct trace_entry *entry) + { + char hardsoft_irq; + char need_resched; ++ char need_resched_lazy; + char irqs_off; + int hardirq; + int softirq; +@@ -416,6 +417,9 @@ int trace_print_lat_fmt(struct trace_seq *s, struct trace_entry *entry) + break; + } + ++ need_resched_lazy = ++ (entry->flags & TRACE_FLAG_NEED_RESCHED_LAZY) ? 'L' : '.'; ++ + hardsoft_irq = + (nmi && hardirq) ? 'Z' : + nmi ? 'z' : +@@ -424,14 +428,25 @@ int trace_print_lat_fmt(struct trace_seq *s, struct trace_entry *entry) + softirq ? 's' : + '.' ; + +- trace_seq_printf(s, "%c%c%c", +- irqs_off, need_resched, hardsoft_irq); ++ trace_seq_printf(s, "%c%c%c%c", ++ irqs_off, need_resched, need_resched_lazy, ++ hardsoft_irq); + + if (entry->preempt_count) + trace_seq_printf(s, "%x", entry->preempt_count); + else + trace_seq_putc(s, '.'); + ++ if (entry->preempt_lazy_count) ++ trace_seq_printf(s, "%x", entry->preempt_lazy_count); ++ else ++ trace_seq_putc(s, '.'); ++ ++ if (entry->migrate_disable) ++ trace_seq_printf(s, "%x", entry->migrate_disable); ++ else ++ trace_seq_putc(s, '.'); ++ + return !trace_seq_has_overflowed(s); + } + +diff --git a/kernel/user.c b/kernel/user.c +index b069ccbfb0b0..1a2e88e98b5e 100644 +--- a/kernel/user.c ++++ b/kernel/user.c +@@ -161,11 +161,11 @@ void free_uid(struct user_struct *up) + if (!up) + return; + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + if (atomic_dec_and_lock(&up->__count, &uidhash_lock)) + free_user(up, flags); + else +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + + struct user_struct *alloc_uid(kuid_t uid) +diff --git a/kernel/watchdog.c b/kernel/watchdog.c +index 6d1020c03d41..70c6a2f79f7e 100644 +--- a/kernel/watchdog.c ++++ b/kernel/watchdog.c +@@ -315,6 +315,8 @@ static int is_softlockup(unsigned long touch_ts) + + #ifdef CONFIG_HARDLOCKUP_DETECTOR + ++static DEFINE_RAW_SPINLOCK(watchdog_output_lock); ++ + static struct perf_event_attr wd_hw_attr = { + .type = PERF_TYPE_HARDWARE, + .config = PERF_COUNT_HW_CPU_CYCLES, +@@ -348,6 +350,13 @@ static void watchdog_overflow_callback(struct perf_event *event, + /* only print hardlockups once */ + if (__this_cpu_read(hard_watchdog_warn) == true) + return; ++ /* ++ * If early-printk is enabled then make sure we do not ++ * lock up in printk() and kill console logging: ++ */ ++ printk_kill(); ++ ++ raw_spin_lock(&watchdog_output_lock); + + pr_emerg("Watchdog detected hard LOCKUP on cpu %d", this_cpu); + print_modules(); +@@ -365,6 +374,7 @@ static void watchdog_overflow_callback(struct perf_event *event, + !test_and_set_bit(0, &hardlockup_allcpu_dumped)) + trigger_allbutself_cpu_backtrace(); + ++ raw_spin_unlock(&watchdog_output_lock); + if (hardlockup_panic) + nmi_panic(regs, "Hard LOCKUP"); + +@@ -512,6 +522,7 @@ static void watchdog_enable(unsigned int cpu) + /* kick off the timer for the hardlockup detector */ + hrtimer_init(hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + hrtimer->function = watchdog_timer_fn; ++ hrtimer->irqsafe = 1; + + /* Enable the perf event */ + watchdog_nmi_enable(cpu); +diff --git a/kernel/workqueue.c b/kernel/workqueue.c +index 479d840db286..24eba6620a45 100644 +--- a/kernel/workqueue.c ++++ b/kernel/workqueue.c +@@ -48,6 +48,8 @@ + #include + #include + #include ++#include ++#include + + #include "workqueue_internal.h" + +@@ -121,11 +123,16 @@ enum { + * cpu or grabbing pool->lock is enough for read access. If + * POOL_DISASSOCIATED is set, it's identical to L. + * ++ * On RT we need the extra protection via rt_lock_idle_list() for ++ * the list manipulations against read access from ++ * wq_worker_sleeping(). All other places are nicely serialized via ++ * pool->lock. ++ * + * A: pool->attach_mutex protected. + * + * PL: wq_pool_mutex protected. + * +- * PR: wq_pool_mutex protected for writes. Sched-RCU protected for reads. ++ * PR: wq_pool_mutex protected for writes. RCU protected for reads. + * + * PW: wq_pool_mutex and wq->mutex protected for writes. Either for reads. + * +@@ -134,7 +141,7 @@ enum { + * + * WQ: wq->mutex protected. + * +- * WR: wq->mutex protected for writes. Sched-RCU protected for reads. ++ * WR: wq->mutex protected for writes. RCU protected for reads. + * + * MD: wq_mayday_lock protected. + */ +@@ -185,7 +192,7 @@ struct worker_pool { + atomic_t nr_running ____cacheline_aligned_in_smp; + + /* +- * Destruction of pool is sched-RCU protected to allow dereferences ++ * Destruction of pool is RCU protected to allow dereferences + * from get_work_pool(). + */ + struct rcu_head rcu; +@@ -214,7 +221,7 @@ struct pool_workqueue { + /* + * Release of unbound pwq is punted to system_wq. See put_pwq() + * and pwq_unbound_release_workfn() for details. pool_workqueue +- * itself is also sched-RCU protected so that the first pwq can be ++ * itself is also RCU protected so that the first pwq can be + * determined without grabbing wq->mutex. + */ + struct work_struct unbound_release_work; +@@ -348,6 +355,8 @@ EXPORT_SYMBOL_GPL(system_power_efficient_wq); + struct workqueue_struct *system_freezable_power_efficient_wq __read_mostly; + EXPORT_SYMBOL_GPL(system_freezable_power_efficient_wq); + ++static DEFINE_LOCAL_IRQ_LOCK(pendingb_lock); ++ + static int worker_thread(void *__worker); + static void workqueue_sysfs_unregister(struct workqueue_struct *wq); + +@@ -355,20 +364,20 @@ static void workqueue_sysfs_unregister(struct workqueue_struct *wq); + #include + + #define assert_rcu_or_pool_mutex() \ +- RCU_LOCKDEP_WARN(!rcu_read_lock_sched_held() && \ ++ RCU_LOCKDEP_WARN(!rcu_read_lock_held() && \ + !lockdep_is_held(&wq_pool_mutex), \ +- "sched RCU or wq_pool_mutex should be held") ++ "RCU or wq_pool_mutex should be held") + + #define assert_rcu_or_wq_mutex(wq) \ +- RCU_LOCKDEP_WARN(!rcu_read_lock_sched_held() && \ ++ RCU_LOCKDEP_WARN(!rcu_read_lock_held() && \ + !lockdep_is_held(&wq->mutex), \ +- "sched RCU or wq->mutex should be held") ++ "RCU or wq->mutex should be held") + + #define assert_rcu_or_wq_mutex_or_pool_mutex(wq) \ +- RCU_LOCKDEP_WARN(!rcu_read_lock_sched_held() && \ ++ RCU_LOCKDEP_WARN(!rcu_read_lock_held() && \ + !lockdep_is_held(&wq->mutex) && \ + !lockdep_is_held(&wq_pool_mutex), \ +- "sched RCU, wq->mutex or wq_pool_mutex should be held") ++ "RCU, wq->mutex or wq_pool_mutex should be held") + + #define for_each_cpu_worker_pool(pool, cpu) \ + for ((pool) = &per_cpu(cpu_worker_pools, cpu)[0]; \ +@@ -380,7 +389,7 @@ static void workqueue_sysfs_unregister(struct workqueue_struct *wq); + * @pool: iteration cursor + * @pi: integer used for iteration + * +- * This must be called either with wq_pool_mutex held or sched RCU read ++ * This must be called either with wq_pool_mutex held or RCU read + * locked. If the pool needs to be used beyond the locking in effect, the + * caller is responsible for guaranteeing that the pool stays online. + * +@@ -412,7 +421,7 @@ static void workqueue_sysfs_unregister(struct workqueue_struct *wq); + * @pwq: iteration cursor + * @wq: the target workqueue + * +- * This must be called either with wq->mutex held or sched RCU read locked. ++ * This must be called either with wq->mutex held or RCU read locked. + * If the pwq needs to be used beyond the locking in effect, the caller is + * responsible for guaranteeing that the pwq stays online. + * +@@ -424,6 +433,31 @@ static void workqueue_sysfs_unregister(struct workqueue_struct *wq); + if (({ assert_rcu_or_wq_mutex(wq); false; })) { } \ + else + ++#ifdef CONFIG_PREEMPT_RT_BASE ++static inline void rt_lock_idle_list(struct worker_pool *pool) ++{ ++ preempt_disable(); ++} ++static inline void rt_unlock_idle_list(struct worker_pool *pool) ++{ ++ preempt_enable(); ++} ++static inline void sched_lock_idle_list(struct worker_pool *pool) { } ++static inline void sched_unlock_idle_list(struct worker_pool *pool) { } ++#else ++static inline void rt_lock_idle_list(struct worker_pool *pool) { } ++static inline void rt_unlock_idle_list(struct worker_pool *pool) { } ++static inline void sched_lock_idle_list(struct worker_pool *pool) ++{ ++ spin_lock_irq(&pool->lock); ++} ++static inline void sched_unlock_idle_list(struct worker_pool *pool) ++{ ++ spin_unlock_irq(&pool->lock); ++} ++#endif ++ ++ + #ifdef CONFIG_DEBUG_OBJECTS_WORK + + static struct debug_obj_descr work_debug_descr; +@@ -548,7 +582,7 @@ static int worker_pool_assign_id(struct worker_pool *pool) + * @wq: the target workqueue + * @node: the node ID + * +- * This must be called with any of wq_pool_mutex, wq->mutex or sched RCU ++ * This must be called with any of wq_pool_mutex, wq->mutex or RCU + * read locked. + * If the pwq needs to be used beyond the locking in effect, the caller is + * responsible for guaranteeing that the pwq stays online. +@@ -692,8 +726,8 @@ static struct pool_workqueue *get_work_pwq(struct work_struct *work) + * @work: the work item of interest + * + * Pools are created and destroyed under wq_pool_mutex, and allows read +- * access under sched-RCU read lock. As such, this function should be +- * called under wq_pool_mutex or with preemption disabled. ++ * access under RCU read lock. As such, this function should be ++ * called under wq_pool_mutex or inside of a rcu_read_lock() region. + * + * All fields of the returned pool are accessible as long as the above + * mentioned locking is in effect. If the returned pool needs to be used +@@ -830,50 +864,45 @@ static struct worker *first_idle_worker(struct worker_pool *pool) + */ + static void wake_up_worker(struct worker_pool *pool) + { +- struct worker *worker = first_idle_worker(pool); ++ struct worker *worker; ++ ++ rt_lock_idle_list(pool); ++ ++ worker = first_idle_worker(pool); + + if (likely(worker)) + wake_up_process(worker->task); ++ ++ rt_unlock_idle_list(pool); + } + + /** +- * wq_worker_waking_up - a worker is waking up ++ * wq_worker_running - a worker is running again + * @task: task waking up +- * @cpu: CPU @task is waking up to + * +- * This function is called during try_to_wake_up() when a worker is +- * being awoken. +- * +- * CONTEXT: +- * spin_lock_irq(rq->lock) ++ * This function is called when a worker returns from schedule() + */ +-void wq_worker_waking_up(struct task_struct *task, int cpu) ++void wq_worker_running(struct task_struct *task) + { + struct worker *worker = kthread_data(task); + +- if (!(worker->flags & WORKER_NOT_RUNNING)) { +- WARN_ON_ONCE(worker->pool->cpu != cpu); ++ if (!worker->sleeping) ++ return; ++ if (!(worker->flags & WORKER_NOT_RUNNING)) + atomic_inc(&worker->pool->nr_running); +- } ++ worker->sleeping = 0; + } + + /** + * wq_worker_sleeping - a worker is going to sleep + * @task: task going to sleep + * +- * This function is called during schedule() when a busy worker is +- * going to sleep. Worker on the same cpu can be woken up by +- * returning pointer to its task. +- * +- * CONTEXT: +- * spin_lock_irq(rq->lock) +- * +- * Return: +- * Worker task on @cpu to wake up, %NULL if none. ++ * This function is called from schedule() when a busy worker is ++ * going to sleep. + */ +-struct task_struct *wq_worker_sleeping(struct task_struct *task) ++void wq_worker_sleeping(struct task_struct *task) + { +- struct worker *worker = kthread_data(task), *to_wakeup = NULL; ++ struct worker *worker = kthread_data(task); + struct worker_pool *pool; + + /* +@@ -882,29 +911,26 @@ struct task_struct *wq_worker_sleeping(struct task_struct *task) + * checking NOT_RUNNING. + */ + if (worker->flags & WORKER_NOT_RUNNING) +- return NULL; ++ return; + + pool = worker->pool; + +- /* this can only happen on the local cpu */ +- if (WARN_ON_ONCE(pool->cpu != raw_smp_processor_id())) +- return NULL; ++ if (WARN_ON_ONCE(worker->sleeping)) ++ return; ++ ++ worker->sleeping = 1; + + /* + * The counterpart of the following dec_and_test, implied mb, + * worklist not empty test sequence is in insert_work(). + * Please read comment there. +- * +- * NOT_RUNNING is clear. This means that we're bound to and +- * running on the local cpu w/ rq lock held and preemption +- * disabled, which in turn means that none else could be +- * manipulating idle_list, so dereferencing idle_list without pool +- * lock is safe. + */ + if (atomic_dec_and_test(&pool->nr_running) && +- !list_empty(&pool->worklist)) +- to_wakeup = first_idle_worker(pool); +- return to_wakeup ? to_wakeup->task : NULL; ++ !list_empty(&pool->worklist)) { ++ sched_lock_idle_list(pool); ++ wake_up_worker(pool); ++ sched_unlock_idle_list(pool); ++ } + } + + /** +@@ -1098,12 +1124,14 @@ static void put_pwq_unlocked(struct pool_workqueue *pwq) + { + if (pwq) { + /* +- * As both pwqs and pools are sched-RCU protected, the ++ * As both pwqs and pools are RCU protected, the + * following lock operations are safe. + */ +- spin_lock_irq(&pwq->pool->lock); ++ rcu_read_lock(); ++ local_spin_lock_irq(pendingb_lock, &pwq->pool->lock); + put_pwq(pwq); +- spin_unlock_irq(&pwq->pool->lock); ++ local_spin_unlock_irq(pendingb_lock, &pwq->pool->lock); ++ rcu_read_unlock(); + } + } + +@@ -1207,7 +1235,7 @@ static int try_to_grab_pending(struct work_struct *work, bool is_dwork, + struct worker_pool *pool; + struct pool_workqueue *pwq; + +- local_irq_save(*flags); ++ local_lock_irqsave(pendingb_lock, *flags); + + /* try to steal the timer if it exists */ + if (is_dwork) { +@@ -1226,6 +1254,7 @@ static int try_to_grab_pending(struct work_struct *work, bool is_dwork, + if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) + return 0; + ++ rcu_read_lock(); + /* + * The queueing is in progress, or it is already queued. Try to + * steal it from ->worklist without clearing WORK_STRUCT_PENDING. +@@ -1264,14 +1293,16 @@ static int try_to_grab_pending(struct work_struct *work, bool is_dwork, + set_work_pool_and_keep_pending(work, pool->id); + + spin_unlock(&pool->lock); ++ rcu_read_unlock(); + return 1; + } + spin_unlock(&pool->lock); + fail: +- local_irq_restore(*flags); ++ rcu_read_unlock(); ++ local_unlock_irqrestore(pendingb_lock, *flags); + if (work_is_canceling(work)) + return -ENOENT; +- cpu_relax(); ++ cpu_chill(); + return -EAGAIN; + } + +@@ -1373,7 +1404,7 @@ static void __queue_work(int cpu, struct workqueue_struct *wq, + * queued or lose PENDING. Grabbing PENDING and queueing should + * happen with IRQ disabled. + */ +- WARN_ON_ONCE(!irqs_disabled()); ++ WARN_ON_ONCE_NONRT(!irqs_disabled()); + + debug_work_activate(work); + +@@ -1381,6 +1412,7 @@ static void __queue_work(int cpu, struct workqueue_struct *wq, + if (unlikely(wq->flags & __WQ_DRAINING) && + WARN_ON_ONCE(!is_chained_work(wq))) + return; ++ rcu_read_lock(); + retry: + if (req_cpu == WORK_CPU_UNBOUND) + cpu = wq_select_unbound_cpu(raw_smp_processor_id()); +@@ -1437,10 +1469,8 @@ static void __queue_work(int cpu, struct workqueue_struct *wq, + /* pwq determined, queue */ + trace_workqueue_queue_work(req_cpu, pwq, work); + +- if (WARN_ON(!list_empty(&work->entry))) { +- spin_unlock(&pwq->pool->lock); +- return; +- } ++ if (WARN_ON(!list_empty(&work->entry))) ++ goto out; + + pwq->nr_in_flight[pwq->work_color]++; + work_flags = work_color_to_flags(pwq->work_color); +@@ -1458,7 +1488,9 @@ static void __queue_work(int cpu, struct workqueue_struct *wq, + + insert_work(pwq, work, worklist, work_flags); + ++out: + spin_unlock(&pwq->pool->lock); ++ rcu_read_unlock(); + } + + /** +@@ -1478,14 +1510,14 @@ bool queue_work_on(int cpu, struct workqueue_struct *wq, + bool ret = false; + unsigned long flags; + +- local_irq_save(flags); ++ local_lock_irqsave(pendingb_lock,flags); + + if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) { + __queue_work(cpu, wq, work); + ret = true; + } + +- local_irq_restore(flags); ++ local_unlock_irqrestore(pendingb_lock, flags); + return ret; + } + EXPORT_SYMBOL(queue_work_on); +@@ -1552,14 +1584,14 @@ bool queue_delayed_work_on(int cpu, struct workqueue_struct *wq, + unsigned long flags; + + /* read the comment in __queue_work() */ +- local_irq_save(flags); ++ local_lock_irqsave(pendingb_lock, flags); + + if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) { + __queue_delayed_work(cpu, wq, dwork, delay); + ret = true; + } + +- local_irq_restore(flags); ++ local_unlock_irqrestore(pendingb_lock, flags); + return ret; + } + EXPORT_SYMBOL(queue_delayed_work_on); +@@ -1594,7 +1626,7 @@ bool mod_delayed_work_on(int cpu, struct workqueue_struct *wq, + + if (likely(ret >= 0)) { + __queue_delayed_work(cpu, wq, dwork, delay); +- local_irq_restore(flags); ++ local_unlock_irqrestore(pendingb_lock, flags); + } + + /* -ENOENT from try_to_grab_pending() becomes %true */ +@@ -1627,7 +1659,9 @@ static void worker_enter_idle(struct worker *worker) + worker->last_active = jiffies; + + /* idle_list is LIFO */ ++ rt_lock_idle_list(pool); + list_add(&worker->entry, &pool->idle_list); ++ rt_unlock_idle_list(pool); + + if (too_many_workers(pool) && !timer_pending(&pool->idle_timer)) + mod_timer(&pool->idle_timer, jiffies + IDLE_WORKER_TIMEOUT); +@@ -1660,7 +1694,9 @@ static void worker_leave_idle(struct worker *worker) + return; + worker_clr_flags(worker, WORKER_IDLE); + pool->nr_idle--; ++ rt_lock_idle_list(pool); + list_del_init(&worker->entry); ++ rt_unlock_idle_list(pool); + } + + static struct worker *alloc_worker(int node) +@@ -1826,7 +1862,9 @@ static void destroy_worker(struct worker *worker) + pool->nr_workers--; + pool->nr_idle--; + ++ rt_lock_idle_list(pool); + list_del_init(&worker->entry); ++ rt_unlock_idle_list(pool); + worker->flags |= WORKER_DIE; + wake_up_process(worker->task); + } +@@ -2785,14 +2823,14 @@ static bool start_flush_work(struct work_struct *work, struct wq_barrier *barr) + + might_sleep(); + +- local_irq_disable(); ++ rcu_read_lock(); + pool = get_work_pool(work); + if (!pool) { +- local_irq_enable(); ++ rcu_read_unlock(); + return false; + } + +- spin_lock(&pool->lock); ++ spin_lock_irq(&pool->lock); + /* see the comment in try_to_grab_pending() with the same code */ + pwq = get_work_pwq(work); + if (pwq) { +@@ -2821,10 +2859,11 @@ static bool start_flush_work(struct work_struct *work, struct wq_barrier *barr) + else + lock_map_acquire_read(&pwq->wq->lockdep_map); + lock_map_release(&pwq->wq->lockdep_map); +- ++ rcu_read_unlock(); + return true; + already_gone: + spin_unlock_irq(&pool->lock); ++ rcu_read_unlock(); + return false; + } + +@@ -2911,7 +2950,7 @@ static bool __cancel_work_timer(struct work_struct *work, bool is_dwork) + + /* tell other tasks trying to grab @work to back off */ + mark_work_canceling(work); +- local_irq_restore(flags); ++ local_unlock_irqrestore(pendingb_lock, flags); + + flush_work(work); + clear_work_data(work); +@@ -2966,10 +3005,10 @@ EXPORT_SYMBOL_GPL(cancel_work_sync); + */ + bool flush_delayed_work(struct delayed_work *dwork) + { +- local_irq_disable(); ++ local_lock_irq(pendingb_lock); + if (del_timer_sync(&dwork->timer)) + __queue_work(dwork->cpu, dwork->wq, &dwork->work); +- local_irq_enable(); ++ local_unlock_irq(pendingb_lock); + return flush_work(&dwork->work); + } + EXPORT_SYMBOL(flush_delayed_work); +@@ -2987,7 +3026,7 @@ static bool __cancel_work(struct work_struct *work, bool is_dwork) + return false; + + set_work_pool_and_clear_pending(work, get_work_pool_id(work)); +- local_irq_restore(flags); ++ local_unlock_irqrestore(pendingb_lock, flags); + return ret; + } + +@@ -3245,7 +3284,7 @@ static void rcu_free_pool(struct rcu_head *rcu) + * put_unbound_pool - put a worker_pool + * @pool: worker_pool to put + * +- * Put @pool. If its refcnt reaches zero, it gets destroyed in sched-RCU ++ * Put @pool. If its refcnt reaches zero, it gets destroyed in RCU + * safe manner. get_unbound_pool() calls this function on its failure path + * and this function should be able to release pools which went through, + * successfully or not, init_worker_pool(). +@@ -3299,8 +3338,8 @@ static void put_unbound_pool(struct worker_pool *pool) + del_timer_sync(&pool->idle_timer); + del_timer_sync(&pool->mayday_timer); + +- /* sched-RCU protected to allow dereferences from get_work_pool() */ +- call_rcu_sched(&pool->rcu, rcu_free_pool); ++ /* RCU protected to allow dereferences from get_work_pool() */ ++ call_rcu(&pool->rcu, rcu_free_pool); + } + + /** +@@ -3407,14 +3446,14 @@ static void pwq_unbound_release_workfn(struct work_struct *work) + put_unbound_pool(pool); + mutex_unlock(&wq_pool_mutex); + +- call_rcu_sched(&pwq->rcu, rcu_free_pwq); ++ call_rcu(&pwq->rcu, rcu_free_pwq); + + /* + * If we're the last pwq going away, @wq is already dead and no one + * is gonna access it anymore. Schedule RCU free. + */ + if (is_last) +- call_rcu_sched(&wq->rcu, rcu_free_wq); ++ call_rcu(&wq->rcu, rcu_free_wq); + } + + /** +@@ -4064,7 +4103,7 @@ void destroy_workqueue(struct workqueue_struct *wq) + * The base ref is never dropped on per-cpu pwqs. Directly + * schedule RCU free. + */ +- call_rcu_sched(&wq->rcu, rcu_free_wq); ++ call_rcu(&wq->rcu, rcu_free_wq); + } else { + /* + * We're the sole accessor of @wq at this point. Directly +@@ -4157,7 +4196,8 @@ bool workqueue_congested(int cpu, struct workqueue_struct *wq) + struct pool_workqueue *pwq; + bool ret; + +- rcu_read_lock_sched(); ++ rcu_read_lock(); ++ preempt_disable(); + + if (cpu == WORK_CPU_UNBOUND) + cpu = smp_processor_id(); +@@ -4168,7 +4208,8 @@ bool workqueue_congested(int cpu, struct workqueue_struct *wq) + pwq = unbound_pwq_by_node(wq, cpu_to_node(cpu)); + + ret = !list_empty(&pwq->delayed_works); +- rcu_read_unlock_sched(); ++ preempt_enable(); ++ rcu_read_unlock(); + + return ret; + } +@@ -4194,15 +4235,15 @@ unsigned int work_busy(struct work_struct *work) + if (work_pending(work)) + ret |= WORK_BUSY_PENDING; + +- local_irq_save(flags); ++ rcu_read_lock(); + pool = get_work_pool(work); + if (pool) { +- spin_lock(&pool->lock); ++ spin_lock_irqsave(&pool->lock, flags); + if (find_worker_executing_work(pool, work)) + ret |= WORK_BUSY_RUNNING; +- spin_unlock(&pool->lock); ++ spin_unlock_irqrestore(&pool->lock, flags); + } +- local_irq_restore(flags); ++ rcu_read_unlock(); + + return ret; + } +@@ -4391,7 +4432,7 @@ void show_workqueue_state(void) + unsigned long flags; + int pi; + +- rcu_read_lock_sched(); ++ rcu_read_lock(); + + pr_info("Showing busy workqueues and worker pools:\n"); + +@@ -4444,7 +4485,7 @@ void show_workqueue_state(void) + spin_unlock_irqrestore(&pool->lock, flags); + } + +- rcu_read_unlock_sched(); ++ rcu_read_unlock(); + } + + /* +@@ -4782,16 +4823,16 @@ bool freeze_workqueues_busy(void) + * nr_active is monotonically decreasing. It's safe + * to peek without lock. + */ +- rcu_read_lock_sched(); ++ rcu_read_lock(); + for_each_pwq(pwq, wq) { + WARN_ON_ONCE(pwq->nr_active < 0); + if (pwq->nr_active) { + busy = true; +- rcu_read_unlock_sched(); ++ rcu_read_unlock(); + goto out_unlock; + } + } +- rcu_read_unlock_sched(); ++ rcu_read_unlock(); + } + out_unlock: + mutex_unlock(&wq_pool_mutex); +@@ -4981,7 +5022,8 @@ static ssize_t wq_pool_ids_show(struct device *dev, + const char *delim = ""; + int node, written = 0; + +- rcu_read_lock_sched(); ++ get_online_cpus(); ++ rcu_read_lock(); + for_each_node(node) { + written += scnprintf(buf + written, PAGE_SIZE - written, + "%s%d:%d", delim, node, +@@ -4989,7 +5031,8 @@ static ssize_t wq_pool_ids_show(struct device *dev, + delim = " "; + } + written += scnprintf(buf + written, PAGE_SIZE - written, "\n"); +- rcu_read_unlock_sched(); ++ rcu_read_unlock(); ++ put_online_cpus(); + + return written; + } +diff --git a/kernel/workqueue_internal.h b/kernel/workqueue_internal.h +index 8635417c587b..f000c4d6917e 100644 +--- a/kernel/workqueue_internal.h ++++ b/kernel/workqueue_internal.h +@@ -43,6 +43,7 @@ struct worker { + unsigned long last_active; /* L: last active timestamp */ + unsigned int flags; /* X: flags */ + int id; /* I: worker id */ ++ int sleeping; /* None */ + + /* + * Opaque string set with work_set_desc(). Printed out with task +@@ -68,7 +69,7 @@ static inline struct worker *current_wq_worker(void) + * Scheduler hooks for concurrency managed workqueue. Only to be used from + * sched/core.c and workqueue.c. + */ +-void wq_worker_waking_up(struct task_struct *task, int cpu); +-struct task_struct *wq_worker_sleeping(struct task_struct *task); ++void wq_worker_running(struct task_struct *task); ++void wq_worker_sleeping(struct task_struct *task); + + #endif /* _KERNEL_WORKQUEUE_INTERNAL_H */ +diff --git a/lib/Kconfig b/lib/Kconfig +index 260a80e313b9..b06becb3f477 100644 +--- a/lib/Kconfig ++++ b/lib/Kconfig +@@ -400,6 +400,7 @@ config CHECK_SIGNATURE + + config CPUMASK_OFFSTACK + bool "Force CPU masks off stack" if DEBUG_PER_CPU_MAPS ++ depends on !PREEMPT_RT_FULL + help + Use dynamic allocation for cpumask_var_t, instead of putting + them on the stack. This is a bit more expensive, but avoids +diff --git a/lib/debugobjects.c b/lib/debugobjects.c +index 056052dc8e91..d8494e126de8 100644 +--- a/lib/debugobjects.c ++++ b/lib/debugobjects.c +@@ -308,7 +308,10 @@ __debug_object_init(void *addr, struct debug_obj_descr *descr, int onstack) + struct debug_obj *obj; + unsigned long flags; + +- fill_pool(); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ if (preempt_count() == 0 && !irqs_disabled()) ++#endif ++ fill_pool(); + + db = get_bucket((unsigned long) addr); + +diff --git a/lib/idr.c b/lib/idr.c +index 6098336df267..9decbe914595 100644 +--- a/lib/idr.c ++++ b/lib/idr.c +@@ -30,6 +30,7 @@ + #include + #include + #include ++#include + + #define MAX_IDR_SHIFT (sizeof(int) * 8 - 1) + #define MAX_IDR_BIT (1U << MAX_IDR_SHIFT) +@@ -45,6 +46,37 @@ static DEFINE_PER_CPU(struct idr_layer *, idr_preload_head); + static DEFINE_PER_CPU(int, idr_preload_cnt); + static DEFINE_SPINLOCK(simple_ida_lock); + ++#ifdef CONFIG_PREEMPT_RT_FULL ++static DEFINE_LOCAL_IRQ_LOCK(idr_lock); ++ ++static inline void idr_preload_lock(void) ++{ ++ local_lock(idr_lock); ++} ++ ++static inline void idr_preload_unlock(void) ++{ ++ local_unlock(idr_lock); ++} ++ ++void idr_preload_end(void) ++{ ++ idr_preload_unlock(); ++} ++EXPORT_SYMBOL(idr_preload_end); ++#else ++static inline void idr_preload_lock(void) ++{ ++ preempt_disable(); ++} ++ ++static inline void idr_preload_unlock(void) ++{ ++ preempt_enable(); ++} ++#endif ++ ++ + /* the maximum ID which can be allocated given idr->layers */ + static int idr_max(int layers) + { +@@ -115,14 +147,14 @@ static struct idr_layer *idr_layer_alloc(gfp_t gfp_mask, struct idr *layer_idr) + * context. See idr_preload() for details. + */ + if (!in_interrupt()) { +- preempt_disable(); ++ idr_preload_lock(); + new = __this_cpu_read(idr_preload_head); + if (new) { + __this_cpu_write(idr_preload_head, new->ary[0]); + __this_cpu_dec(idr_preload_cnt); + new->ary[0] = NULL; + } +- preempt_enable(); ++ idr_preload_unlock(); + if (new) + return new; + } +@@ -366,7 +398,6 @@ static void idr_fill_slot(struct idr *idr, void *ptr, int id, + idr_mark_full(pa, id); + } + +- + /** + * idr_preload - preload for idr_alloc() + * @gfp_mask: allocation mask to use for preloading +@@ -401,7 +432,7 @@ void idr_preload(gfp_t gfp_mask) + WARN_ON_ONCE(in_interrupt()); + might_sleep_if(gfpflags_allow_blocking(gfp_mask)); + +- preempt_disable(); ++ idr_preload_lock(); + + /* + * idr_alloc() is likely to succeed w/o full idr_layer buffer and +@@ -413,9 +444,9 @@ void idr_preload(gfp_t gfp_mask) + while (__this_cpu_read(idr_preload_cnt) < MAX_IDR_FREE) { + struct idr_layer *new; + +- preempt_enable(); ++ idr_preload_unlock(); + new = kmem_cache_zalloc(idr_layer_cache, gfp_mask); +- preempt_disable(); ++ idr_preload_lock(); + if (!new) + break; + +diff --git a/lib/irq_poll.c b/lib/irq_poll.c +index 1d6565e81030..b23a79761df7 100644 +--- a/lib/irq_poll.c ++++ b/lib/irq_poll.c +@@ -36,6 +36,7 @@ void irq_poll_sched(struct irq_poll *iop) + list_add_tail(&iop->list, this_cpu_ptr(&blk_cpu_iopoll)); + __raise_softirq_irqoff(IRQ_POLL_SOFTIRQ); + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + EXPORT_SYMBOL(irq_poll_sched); + +@@ -71,6 +72,7 @@ void irq_poll_complete(struct irq_poll *iop) + local_irq_save(flags); + __irq_poll_complete(iop); + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + EXPORT_SYMBOL(irq_poll_complete); + +@@ -95,6 +97,7 @@ static void __latent_entropy irq_poll_softirq(struct softirq_action *h) + } + + local_irq_enable(); ++ preempt_check_resched_rt(); + + /* Even though interrupts have been re-enabled, this + * access is safe because interrupts can only add new +@@ -132,6 +135,7 @@ static void __latent_entropy irq_poll_softirq(struct softirq_action *h) + __raise_softirq_irqoff(IRQ_POLL_SOFTIRQ); + + local_irq_enable(); ++ preempt_check_resched_rt(); + } + + /** +@@ -195,6 +199,7 @@ static int irq_poll_cpu_dead(unsigned int cpu) + this_cpu_ptr(&blk_cpu_iopoll)); + __raise_softirq_irqoff(IRQ_POLL_SOFTIRQ); + local_irq_enable(); ++ preempt_check_resched_rt(); + + return 0; + } +diff --git a/lib/locking-selftest.c b/lib/locking-selftest.c +index f3a217ea0388..4611b156ef79 100644 +--- a/lib/locking-selftest.c ++++ b/lib/locking-selftest.c +@@ -590,6 +590,8 @@ GENERATE_TESTCASE(init_held_rsem) + #include "locking-selftest-spin-hardirq.h" + GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_hard_spin) + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + #include "locking-selftest-rlock-hardirq.h" + GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_hard_rlock) + +@@ -605,9 +607,12 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_soft_rlock) + #include "locking-selftest-wlock-softirq.h" + GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_soft_wlock) + ++#endif ++ + #undef E1 + #undef E2 + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * Enabling hardirqs with a softirq-safe lock held: + */ +@@ -640,6 +645,8 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2A_rlock) + #undef E1 + #undef E2 + ++#endif ++ + /* + * Enabling irqs with an irq-safe lock held: + */ +@@ -663,6 +670,8 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2A_rlock) + #include "locking-selftest-spin-hardirq.h" + GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_hard_spin) + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + #include "locking-selftest-rlock-hardirq.h" + GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_hard_rlock) + +@@ -678,6 +687,8 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_rlock) + #include "locking-selftest-wlock-softirq.h" + GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_wlock) + ++#endif ++ + #undef E1 + #undef E2 + +@@ -709,6 +720,8 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_wlock) + #include "locking-selftest-spin-hardirq.h" + GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_hard_spin) + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + #include "locking-selftest-rlock-hardirq.h" + GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_hard_rlock) + +@@ -724,6 +737,8 @@ GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_rlock) + #include "locking-selftest-wlock-softirq.h" + GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_wlock) + ++#endif ++ + #undef E1 + #undef E2 + #undef E3 +@@ -757,6 +772,8 @@ GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_wlock) + #include "locking-selftest-spin-hardirq.h" + GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_hard_spin) + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + #include "locking-selftest-rlock-hardirq.h" + GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_hard_rlock) + +@@ -772,10 +789,14 @@ GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_soft_rlock) + #include "locking-selftest-wlock-softirq.h" + GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_soft_wlock) + ++#endif ++ + #undef E1 + #undef E2 + #undef E3 + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + /* + * read-lock / write-lock irq inversion. + * +@@ -838,6 +859,10 @@ GENERATE_PERMUTATIONS_3_EVENTS(irq_inversion_soft_wlock) + #undef E2 + #undef E3 + ++#endif ++ ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + /* + * read-lock / write-lock recursion that is actually safe. + */ +@@ -876,6 +901,8 @@ GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion_soft) + #undef E2 + #undef E3 + ++#endif ++ + /* + * read-lock / write-lock recursion that is unsafe. + */ +@@ -1858,6 +1885,7 @@ void locking_selftest(void) + + printk(" --------------------------------------------------------------------------\n"); + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * irq-context testcases: + */ +@@ -1870,6 +1898,28 @@ void locking_selftest(void) + + DO_TESTCASE_6x2("irq read-recursion", irq_read_recursion); + // DO_TESTCASE_6x2B("irq read-recursion #2", irq_read_recursion2); ++#else ++ /* On -rt, we only do hardirq context test for raw spinlock */ ++ DO_TESTCASE_1B("hard-irqs-on + irq-safe-A", irqsafe1_hard_spin, 12); ++ DO_TESTCASE_1B("hard-irqs-on + irq-safe-A", irqsafe1_hard_spin, 21); ++ ++ DO_TESTCASE_1B("hard-safe-A + irqs-on", irqsafe2B_hard_spin, 12); ++ DO_TESTCASE_1B("hard-safe-A + irqs-on", irqsafe2B_hard_spin, 21); ++ ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 123); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 132); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 213); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 231); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 312); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 321); ++ ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 123); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 132); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 213); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 231); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 312); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 321); ++#endif + + ww_tests(); + +diff --git a/lib/percpu_ida.c b/lib/percpu_ida.c +index 6d40944960de..822a2c027e72 100644 +--- a/lib/percpu_ida.c ++++ b/lib/percpu_ida.c +@@ -26,6 +26,9 @@ + #include + #include + #include ++#include ++ ++static DEFINE_LOCAL_IRQ_LOCK(irq_off_lock); + + struct percpu_ida_cpu { + /* +@@ -148,13 +151,13 @@ int percpu_ida_alloc(struct percpu_ida *pool, int state) + unsigned long flags; + int tag; + +- local_irq_save(flags); ++ local_lock_irqsave(irq_off_lock, flags); + tags = this_cpu_ptr(pool->tag_cpu); + + /* Fastpath */ + tag = alloc_local_tag(tags); + if (likely(tag >= 0)) { +- local_irq_restore(flags); ++ local_unlock_irqrestore(irq_off_lock, flags); + return tag; + } + +@@ -173,6 +176,7 @@ int percpu_ida_alloc(struct percpu_ida *pool, int state) + + if (!tags->nr_free) + alloc_global_tags(pool, tags); ++ + if (!tags->nr_free) + steal_tags(pool, tags); + +@@ -184,7 +188,7 @@ int percpu_ida_alloc(struct percpu_ida *pool, int state) + } + + spin_unlock(&pool->lock); +- local_irq_restore(flags); ++ local_unlock_irqrestore(irq_off_lock, flags); + + if (tag >= 0 || state == TASK_RUNNING) + break; +@@ -196,7 +200,7 @@ int percpu_ida_alloc(struct percpu_ida *pool, int state) + + schedule(); + +- local_irq_save(flags); ++ local_lock_irqsave(irq_off_lock, flags); + tags = this_cpu_ptr(pool->tag_cpu); + } + if (state != TASK_RUNNING) +@@ -221,7 +225,7 @@ void percpu_ida_free(struct percpu_ida *pool, unsigned tag) + + BUG_ON(tag >= pool->nr_tags); + +- local_irq_save(flags); ++ local_lock_irqsave(irq_off_lock, flags); + tags = this_cpu_ptr(pool->tag_cpu); + + spin_lock(&tags->lock); +@@ -253,7 +257,7 @@ void percpu_ida_free(struct percpu_ida *pool, unsigned tag) + spin_unlock(&pool->lock); + } + +- local_irq_restore(flags); ++ local_unlock_irqrestore(irq_off_lock, flags); + } + EXPORT_SYMBOL_GPL(percpu_ida_free); + +@@ -345,7 +349,7 @@ int percpu_ida_for_each_free(struct percpu_ida *pool, percpu_ida_cb fn, + struct percpu_ida_cpu *remote; + unsigned cpu, i, err = 0; + +- local_irq_save(flags); ++ local_lock_irqsave(irq_off_lock, flags); + for_each_possible_cpu(cpu) { + remote = per_cpu_ptr(pool->tag_cpu, cpu); + spin_lock(&remote->lock); +@@ -367,7 +371,7 @@ int percpu_ida_for_each_free(struct percpu_ida *pool, percpu_ida_cb fn, + } + spin_unlock(&pool->lock); + out: +- local_irq_restore(flags); ++ local_unlock_irqrestore(irq_off_lock, flags); + return err; + } + EXPORT_SYMBOL_GPL(percpu_ida_for_each_free); +diff --git a/lib/radix-tree.c b/lib/radix-tree.c +index 8e6d552c40dd..741da5a77fd5 100644 +--- a/lib/radix-tree.c ++++ b/lib/radix-tree.c +@@ -36,7 +36,7 @@ + #include + #include + #include /* in_interrupt() */ +- ++#include + + /* Number of nodes in fully populated tree of given height */ + static unsigned long height_to_maxnodes[RADIX_TREE_MAX_PATH + 1] __read_mostly; +@@ -68,6 +68,7 @@ struct radix_tree_preload { + struct radix_tree_node *nodes; + }; + static DEFINE_PER_CPU(struct radix_tree_preload, radix_tree_preloads) = { 0, }; ++static DEFINE_LOCAL_IRQ_LOCK(radix_tree_preloads_lock); + + static inline void *node_to_entry(void *ptr) + { +@@ -290,13 +291,14 @@ radix_tree_node_alloc(struct radix_tree_root *root) + * succeed in getting a node here (and never reach + * kmem_cache_alloc) + */ +- rtp = this_cpu_ptr(&radix_tree_preloads); ++ rtp = &get_locked_var(radix_tree_preloads_lock, radix_tree_preloads); + if (rtp->nr) { + ret = rtp->nodes; + rtp->nodes = ret->private_data; + ret->private_data = NULL; + rtp->nr--; + } ++ put_locked_var(radix_tree_preloads_lock, radix_tree_preloads); + /* + * Update the allocation stack trace as this is more useful + * for debugging. +@@ -357,14 +359,14 @@ static int __radix_tree_preload(gfp_t gfp_mask, int nr) + */ + gfp_mask &= ~__GFP_ACCOUNT; + +- preempt_disable(); ++ local_lock(radix_tree_preloads_lock); + rtp = this_cpu_ptr(&radix_tree_preloads); + while (rtp->nr < nr) { +- preempt_enable(); ++ local_unlock(radix_tree_preloads_lock); + node = kmem_cache_alloc(radix_tree_node_cachep, gfp_mask); + if (node == NULL) + goto out; +- preempt_disable(); ++ local_lock(radix_tree_preloads_lock); + rtp = this_cpu_ptr(&radix_tree_preloads); + if (rtp->nr < nr) { + node->private_data = rtp->nodes; +@@ -406,7 +408,7 @@ int radix_tree_maybe_preload(gfp_t gfp_mask) + if (gfpflags_allow_blocking(gfp_mask)) + return __radix_tree_preload(gfp_mask, RADIX_TREE_PRELOAD_SIZE); + /* Preloading doesn't help anything with this gfp mask, skip it */ +- preempt_disable(); ++ local_lock(radix_tree_preloads_lock); + return 0; + } + EXPORT_SYMBOL(radix_tree_maybe_preload); +@@ -422,7 +424,7 @@ int radix_tree_maybe_preload_order(gfp_t gfp_mask, int order) + + /* Preloading doesn't help anything with this gfp mask, skip it */ + if (!gfpflags_allow_blocking(gfp_mask)) { +- preempt_disable(); ++ local_lock(radix_tree_preloads_lock); + return 0; + } + +@@ -456,6 +458,12 @@ int radix_tree_maybe_preload_order(gfp_t gfp_mask, int order) + return __radix_tree_preload(gfp_mask, nr_nodes); + } + ++void radix_tree_preload_end(void) ++{ ++ local_unlock(radix_tree_preloads_lock); ++} ++EXPORT_SYMBOL(radix_tree_preload_end); ++ + /* + * The maximum index which can be stored in a radix tree + */ +diff --git a/lib/scatterlist.c b/lib/scatterlist.c +index 004fc70fc56a..ccc46992a517 100644 +--- a/lib/scatterlist.c ++++ b/lib/scatterlist.c +@@ -620,7 +620,7 @@ void sg_miter_stop(struct sg_mapping_iter *miter) + flush_kernel_dcache_page(miter->page); + + if (miter->__flags & SG_MITER_ATOMIC) { +- WARN_ON_ONCE(preemptible()); ++ WARN_ON_ONCE(!pagefault_disabled()); + kunmap_atomic(miter->addr); + } else + kunmap(miter->page); +@@ -664,7 +664,7 @@ size_t sg_copy_buffer(struct scatterlist *sgl, unsigned int nents, void *buf, + if (!sg_miter_skip(&miter, skip)) + return false; + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + + while (sg_miter_next(&miter) && offset < buflen) { + unsigned int len; +@@ -681,7 +681,7 @@ size_t sg_copy_buffer(struct scatterlist *sgl, unsigned int nents, void *buf, + + sg_miter_stop(&miter); + +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + return offset; + } + EXPORT_SYMBOL(sg_copy_buffer); +diff --git a/lib/smp_processor_id.c b/lib/smp_processor_id.c +index 1afec32de6f2..11fa431046a8 100644 +--- a/lib/smp_processor_id.c ++++ b/lib/smp_processor_id.c +@@ -39,8 +39,9 @@ notrace static unsigned int check_preemption_disabled(const char *what1, + if (!printk_ratelimit()) + goto out_enable; + +- printk(KERN_ERR "BUG: using %s%s() in preemptible [%08x] code: %s/%d\n", +- what1, what2, preempt_count() - 1, current->comm, current->pid); ++ printk(KERN_ERR "BUG: using %s%s() in preemptible [%08x %08x] code: %s/%d\n", ++ what1, what2, preempt_count() - 1, __migrate_disabled(current), ++ current->comm, current->pid); + + print_symbol("caller is %s\n", (long)__builtin_return_address(0)); + dump_stack(); +diff --git a/mm/Kconfig b/mm/Kconfig +index 86e3e0e74d20..77e5862a1ed2 100644 +--- a/mm/Kconfig ++++ b/mm/Kconfig +@@ -410,7 +410,7 @@ config NOMMU_INITIAL_TRIM_EXCESS + + config TRANSPARENT_HUGEPAGE + bool "Transparent Hugepage Support" +- depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE ++ depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE && !PREEMPT_RT_FULL + select COMPACTION + select RADIX_TREE_MULTIORDER + help +diff --git a/mm/backing-dev.c b/mm/backing-dev.c +index 6ff2d7744223..b5a91dd53b5f 100644 +--- a/mm/backing-dev.c ++++ b/mm/backing-dev.c +@@ -457,9 +457,9 @@ void wb_congested_put(struct bdi_writeback_congested *congested) + { + unsigned long flags; + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + if (!atomic_dec_and_lock(&congested->refcnt, &cgwb_lock)) { +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + return; + } + +diff --git a/mm/compaction.c b/mm/compaction.c +index 70e6bec46dc2..6678ed58b7c6 100644 +--- a/mm/compaction.c ++++ b/mm/compaction.c +@@ -1593,10 +1593,12 @@ static enum compact_result compact_zone(struct zone *zone, struct compact_contro + block_start_pfn(cc->migrate_pfn, cc->order); + + if (cc->last_migrated_pfn < current_block_start) { +- cpu = get_cpu(); ++ cpu = get_cpu_light(); ++ local_lock_irq(swapvec_lock); + lru_add_drain_cpu(cpu); ++ local_unlock_irq(swapvec_lock); + drain_local_pages(zone); +- put_cpu(); ++ put_cpu_light(); + /* No more flushing until we migrate again */ + cc->last_migrated_pfn = 0; + } +diff --git a/mm/filemap.c b/mm/filemap.c +index edfb90e3830c..a8d2c7a73d54 100644 +--- a/mm/filemap.c ++++ b/mm/filemap.c +@@ -159,9 +159,12 @@ static int page_cache_tree_insert(struct address_space *mapping, + * node->private_list is protected by + * mapping->tree_lock. + */ +- if (!list_empty(&node->private_list)) +- list_lru_del(&workingset_shadow_nodes, ++ if (!list_empty(&node->private_list)) { ++ local_lock(workingset_shadow_lock); ++ list_lru_del(&__workingset_shadow_nodes, + &node->private_list); ++ local_unlock(workingset_shadow_lock); ++ } + } + return 0; + } +@@ -217,8 +220,10 @@ static void page_cache_tree_delete(struct address_space *mapping, + if (!dax_mapping(mapping) && !workingset_node_pages(node) && + list_empty(&node->private_list)) { + node->private_data = mapping; +- list_lru_add(&workingset_shadow_nodes, +- &node->private_list); ++ local_lock(workingset_shadow_lock); ++ list_lru_add(&__workingset_shadow_nodes, ++ &node->private_list); ++ local_unlock(workingset_shadow_lock); + } + } + +diff --git a/mm/highmem.c b/mm/highmem.c +index 50b4ca6787f0..77518a3b35a1 100644 +--- a/mm/highmem.c ++++ b/mm/highmem.c +@@ -29,10 +29,11 @@ + #include + #include + +- ++#ifndef CONFIG_PREEMPT_RT_FULL + #if defined(CONFIG_HIGHMEM) || defined(CONFIG_X86_32) + DEFINE_PER_CPU(int, __kmap_atomic_idx); + #endif ++#endif + + /* + * Virtual_count is not a pure "count". +@@ -107,8 +108,9 @@ static inline wait_queue_head_t *get_pkmap_wait_queue_head(unsigned int color) + unsigned long totalhigh_pages __read_mostly; + EXPORT_SYMBOL(totalhigh_pages); + +- ++#ifndef CONFIG_PREEMPT_RT_FULL + EXPORT_PER_CPU_SYMBOL(__kmap_atomic_idx); ++#endif + + unsigned int nr_free_highpages (void) + { +diff --git a/mm/memcontrol.c b/mm/memcontrol.c +index 47559cc0cdcc..1f2ebc924916 100644 +--- a/mm/memcontrol.c ++++ b/mm/memcontrol.c +@@ -67,6 +67,7 @@ + #include + #include + #include "slab.h" ++#include + + #include + +@@ -92,6 +93,8 @@ int do_swap_account __read_mostly; + #define do_swap_account 0 + #endif + ++static DEFINE_LOCAL_IRQ_LOCK(event_lock); ++ + /* Whether legacy memory+swap accounting is active */ + static bool do_memsw_account(void) + { +@@ -1692,6 +1695,7 @@ struct memcg_stock_pcp { + #define FLUSHING_CACHED_CHARGE 0 + }; + static DEFINE_PER_CPU(struct memcg_stock_pcp, memcg_stock); ++static DEFINE_LOCAL_IRQ_LOCK(memcg_stock_ll); + static DEFINE_MUTEX(percpu_charge_mutex); + + /** +@@ -1714,7 +1718,7 @@ static bool consume_stock(struct mem_cgroup *memcg, unsigned int nr_pages) + if (nr_pages > CHARGE_BATCH) + return ret; + +- local_irq_save(flags); ++ local_lock_irqsave(memcg_stock_ll, flags); + + stock = this_cpu_ptr(&memcg_stock); + if (memcg == stock->cached && stock->nr_pages >= nr_pages) { +@@ -1722,7 +1726,7 @@ static bool consume_stock(struct mem_cgroup *memcg, unsigned int nr_pages) + ret = true; + } + +- local_irq_restore(flags); ++ local_unlock_irqrestore(memcg_stock_ll, flags); + + return ret; + } +@@ -1749,13 +1753,13 @@ static void drain_local_stock(struct work_struct *dummy) + struct memcg_stock_pcp *stock; + unsigned long flags; + +- local_irq_save(flags); ++ local_lock_irqsave(memcg_stock_ll, flags); + + stock = this_cpu_ptr(&memcg_stock); + drain_stock(stock); + clear_bit(FLUSHING_CACHED_CHARGE, &stock->flags); + +- local_irq_restore(flags); ++ local_unlock_irqrestore(memcg_stock_ll, flags); + } + + /* +@@ -1767,7 +1771,7 @@ static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages) + struct memcg_stock_pcp *stock; + unsigned long flags; + +- local_irq_save(flags); ++ local_lock_irqsave(memcg_stock_ll, flags); + + stock = this_cpu_ptr(&memcg_stock); + if (stock->cached != memcg) { /* reset if necessary */ +@@ -1776,7 +1780,7 @@ static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages) + } + stock->nr_pages += nr_pages; + +- local_irq_restore(flags); ++ local_unlock_irqrestore(memcg_stock_ll, flags); + } + + /* +@@ -1792,7 +1796,7 @@ static void drain_all_stock(struct mem_cgroup *root_memcg) + return; + /* Notify other cpus that system-wide "drain" is running */ + get_online_cpus(); +- curcpu = get_cpu(); ++ curcpu = get_cpu_light(); + for_each_online_cpu(cpu) { + struct memcg_stock_pcp *stock = &per_cpu(memcg_stock, cpu); + struct mem_cgroup *memcg; +@@ -1809,7 +1813,7 @@ static void drain_all_stock(struct mem_cgroup *root_memcg) + schedule_work_on(cpu, &stock->work); + } + } +- put_cpu(); ++ put_cpu_light(); + put_online_cpus(); + mutex_unlock(&percpu_charge_mutex); + } +@@ -4555,12 +4559,12 @@ static int mem_cgroup_move_account(struct page *page, + + ret = 0; + +- local_irq_disable(); ++ local_lock_irq(event_lock); + mem_cgroup_charge_statistics(to, page, compound, nr_pages); + memcg_check_events(to, page); + mem_cgroup_charge_statistics(from, page, compound, -nr_pages); + memcg_check_events(from, page); +- local_irq_enable(); ++ local_unlock_irq(event_lock); + out_unlock: + unlock_page(page); + out: +@@ -5435,10 +5439,10 @@ void mem_cgroup_commit_charge(struct page *page, struct mem_cgroup *memcg, + + commit_charge(page, memcg, lrucare); + +- local_irq_disable(); ++ local_lock_irq(event_lock); + mem_cgroup_charge_statistics(memcg, page, compound, nr_pages); + memcg_check_events(memcg, page); +- local_irq_enable(); ++ local_unlock_irq(event_lock); + + if (do_memsw_account() && PageSwapCache(page)) { + swp_entry_t entry = { .val = page_private(page) }; +@@ -5494,14 +5498,14 @@ static void uncharge_batch(struct mem_cgroup *memcg, unsigned long pgpgout, + memcg_oom_recover(memcg); + } + +- local_irq_save(flags); ++ local_lock_irqsave(event_lock, flags); + __this_cpu_sub(memcg->stat->count[MEM_CGROUP_STAT_RSS], nr_anon); + __this_cpu_sub(memcg->stat->count[MEM_CGROUP_STAT_CACHE], nr_file); + __this_cpu_sub(memcg->stat->count[MEM_CGROUP_STAT_RSS_HUGE], nr_huge); + __this_cpu_add(memcg->stat->events[MEM_CGROUP_EVENTS_PGPGOUT], pgpgout); + __this_cpu_add(memcg->stat->nr_page_events, nr_pages); + memcg_check_events(memcg, dummy_page); +- local_irq_restore(flags); ++ local_unlock_irqrestore(event_lock, flags); + + if (!mem_cgroup_is_root(memcg)) + css_put_many(&memcg->css, nr_pages); +@@ -5656,10 +5660,10 @@ void mem_cgroup_migrate(struct page *oldpage, struct page *newpage) + + commit_charge(newpage, memcg, false); + +- local_irq_save(flags); ++ local_lock_irqsave(event_lock, flags); + mem_cgroup_charge_statistics(memcg, newpage, compound, nr_pages); + memcg_check_events(memcg, newpage); +- local_irq_restore(flags); ++ local_unlock_irqrestore(event_lock, flags); + } + + DEFINE_STATIC_KEY_FALSE(memcg_sockets_enabled_key); +@@ -5850,6 +5854,7 @@ void mem_cgroup_swapout(struct page *page, swp_entry_t entry) + { + struct mem_cgroup *memcg, *swap_memcg; + unsigned short oldid; ++ unsigned long flags; + + VM_BUG_ON_PAGE(PageLRU(page), page); + VM_BUG_ON_PAGE(page_count(page), page); +@@ -5890,12 +5895,16 @@ void mem_cgroup_swapout(struct page *page, swp_entry_t entry) + * important here to have the interrupts disabled because it is the + * only synchronisation we have for udpating the per-CPU variables. + */ ++ local_lock_irqsave(event_lock, flags); ++#ifndef CONFIG_PREEMPT_RT_BASE + VM_BUG_ON(!irqs_disabled()); ++#endif + mem_cgroup_charge_statistics(memcg, page, false, -1); + memcg_check_events(memcg, page); + + if (!mem_cgroup_is_root(memcg)) + css_put(&memcg->css); ++ local_unlock_irqrestore(event_lock, flags); + } + + /* +diff --git a/mm/mmu_context.c b/mm/mmu_context.c +index 6f4d27c5bb32..5cd25c745a8f 100644 +--- a/mm/mmu_context.c ++++ b/mm/mmu_context.c +@@ -23,6 +23,7 @@ void use_mm(struct mm_struct *mm) + struct task_struct *tsk = current; + + task_lock(tsk); ++ preempt_disable_rt(); + active_mm = tsk->active_mm; + if (active_mm != mm) { + atomic_inc(&mm->mm_count); +@@ -30,6 +31,7 @@ void use_mm(struct mm_struct *mm) + } + tsk->mm = mm; + switch_mm(active_mm, mm, tsk); ++ preempt_enable_rt(); + task_unlock(tsk); + #ifdef finish_arch_post_lock_switch + finish_arch_post_lock_switch(); +diff --git a/mm/page_alloc.c b/mm/page_alloc.c +index 5b06fb385dd7..86457a2fca20 100644 +--- a/mm/page_alloc.c ++++ b/mm/page_alloc.c +@@ -61,6 +61,7 @@ + #include + #include + #include ++#include + #include + #include + #include +@@ -281,6 +282,18 @@ EXPORT_SYMBOL(nr_node_ids); + EXPORT_SYMBOL(nr_online_nodes); + #endif + ++static DEFINE_LOCAL_IRQ_LOCK(pa_lock); ++ ++#ifdef CONFIG_PREEMPT_RT_BASE ++# define cpu_lock_irqsave(cpu, flags) \ ++ local_lock_irqsave_on(pa_lock, flags, cpu) ++# define cpu_unlock_irqrestore(cpu, flags) \ ++ local_unlock_irqrestore_on(pa_lock, flags, cpu) ++#else ++# define cpu_lock_irqsave(cpu, flags) local_irq_save(flags) ++# define cpu_unlock_irqrestore(cpu, flags) local_irq_restore(flags) ++#endif ++ + int page_group_by_mobility_disabled __read_mostly; + + #ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT +@@ -1072,7 +1085,7 @@ static bool bulkfree_pcp_prepare(struct page *page) + #endif /* CONFIG_DEBUG_VM */ + + /* +- * Frees a number of pages from the PCP lists ++ * Frees a number of pages which have been collected from the pcp lists. + * Assumes all pages on list are in same zone, and of same order. + * count is the number of pages to free. + * +@@ -1083,19 +1096,58 @@ static bool bulkfree_pcp_prepare(struct page *page) + * pinned" detection logic. + */ + static void free_pcppages_bulk(struct zone *zone, int count, +- struct per_cpu_pages *pcp) ++ struct list_head *list) + { +- int migratetype = 0; +- int batch_free = 0; + unsigned long nr_scanned; + bool isolated_pageblocks; ++ unsigned long flags; ++ ++ spin_lock_irqsave(&zone->lock, flags); + +- spin_lock(&zone->lock); + isolated_pageblocks = has_isolate_pageblock(zone); + nr_scanned = node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED); + if (nr_scanned) + __mod_node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED, -nr_scanned); + ++ while (!list_empty(list)) { ++ struct page *page; ++ int mt; /* migratetype of the to-be-freed page */ ++ ++ page = list_first_entry(list, struct page, lru); ++ /* must delete as __free_one_page list manipulates */ ++ list_del(&page->lru); ++ ++ mt = get_pcppage_migratetype(page); ++ /* MIGRATE_ISOLATE page should not go to pcplists */ ++ VM_BUG_ON_PAGE(is_migrate_isolate(mt), page); ++ /* Pageblock could have been isolated meanwhile */ ++ if (unlikely(isolated_pageblocks)) ++ mt = get_pageblock_migratetype(page); ++ ++ if (bulkfree_pcp_prepare(page)) ++ continue; ++ ++ __free_one_page(page, page_to_pfn(page), zone, 0, mt); ++ trace_mm_page_pcpu_drain(page, 0, mt); ++ count--; ++ } ++ WARN_ON(count != 0); ++ spin_unlock_irqrestore(&zone->lock, flags); ++} ++ ++/* ++ * Moves a number of pages from the PCP lists to free list which ++ * is freed outside of the locked region. ++ * ++ * Assumes all pages on list are in same zone, and of same order. ++ * count is the number of pages to free. ++ */ ++static void isolate_pcp_pages(int count, struct per_cpu_pages *src, ++ struct list_head *dst) ++{ ++ int migratetype = 0; ++ int batch_free = 0; ++ + while (count) { + struct page *page; + struct list_head *list; +@@ -1111,7 +1163,7 @@ static void free_pcppages_bulk(struct zone *zone, int count, + batch_free++; + if (++migratetype == MIGRATE_PCPTYPES) + migratetype = 0; +- list = &pcp->lists[migratetype]; ++ list = &src->lists[migratetype]; + } while (list_empty(list)); + + /* This is the only non-empty list. Free them all. */ +@@ -1119,27 +1171,12 @@ static void free_pcppages_bulk(struct zone *zone, int count, + batch_free = count; + + do { +- int mt; /* migratetype of the to-be-freed page */ +- + page = list_last_entry(list, struct page, lru); +- /* must delete as __free_one_page list manipulates */ + list_del(&page->lru); + +- mt = get_pcppage_migratetype(page); +- /* MIGRATE_ISOLATE page should not go to pcplists */ +- VM_BUG_ON_PAGE(is_migrate_isolate(mt), page); +- /* Pageblock could have been isolated meanwhile */ +- if (unlikely(isolated_pageblocks)) +- mt = get_pageblock_migratetype(page); +- +- if (bulkfree_pcp_prepare(page)) +- continue; +- +- __free_one_page(page, page_to_pfn(page), zone, 0, mt); +- trace_mm_page_pcpu_drain(page, 0, mt); ++ list_add(&page->lru, dst); + } while (--count && --batch_free && !list_empty(list)); + } +- spin_unlock(&zone->lock); + } + + static void free_one_page(struct zone *zone, +@@ -1148,7 +1185,9 @@ static void free_one_page(struct zone *zone, + int migratetype) + { + unsigned long nr_scanned; +- spin_lock(&zone->lock); ++ unsigned long flags; ++ ++ spin_lock_irqsave(&zone->lock, flags); + nr_scanned = node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED); + if (nr_scanned) + __mod_node_page_state(zone->zone_pgdat, NR_PAGES_SCANNED, -nr_scanned); +@@ -1158,7 +1197,7 @@ static void free_one_page(struct zone *zone, + migratetype = get_pfnblock_migratetype(page, pfn); + } + __free_one_page(page, pfn, zone, order, migratetype); +- spin_unlock(&zone->lock); ++ spin_unlock_irqrestore(&zone->lock, flags); + } + + static void __meminit __init_single_page(struct page *page, unsigned long pfn, +@@ -1244,10 +1283,10 @@ static void __free_pages_ok(struct page *page, unsigned int order) + return; + + migratetype = get_pfnblock_migratetype(page, pfn); +- local_irq_save(flags); ++ local_lock_irqsave(pa_lock, flags); + __count_vm_events(PGFREE, 1 << order); + free_one_page(page_zone(page), page, pfn, order, migratetype); +- local_irq_restore(flags); ++ local_unlock_irqrestore(pa_lock, flags); + } + + static void __init __free_pages_boot_core(struct page *page, unsigned int order) +@@ -2246,16 +2285,18 @@ static int rmqueue_bulk(struct zone *zone, unsigned int order, + void drain_zone_pages(struct zone *zone, struct per_cpu_pages *pcp) + { + unsigned long flags; ++ LIST_HEAD(dst); + int to_drain, batch; + +- local_irq_save(flags); ++ local_lock_irqsave(pa_lock, flags); + batch = READ_ONCE(pcp->batch); + to_drain = min(pcp->count, batch); + if (to_drain > 0) { +- free_pcppages_bulk(zone, to_drain, pcp); ++ isolate_pcp_pages(to_drain, pcp, &dst); + pcp->count -= to_drain; + } +- local_irq_restore(flags); ++ local_unlock_irqrestore(pa_lock, flags); ++ free_pcppages_bulk(zone, to_drain, &dst); + } + #endif + +@@ -2271,16 +2312,21 @@ static void drain_pages_zone(unsigned int cpu, struct zone *zone) + unsigned long flags; + struct per_cpu_pageset *pset; + struct per_cpu_pages *pcp; ++ LIST_HEAD(dst); ++ int count; + +- local_irq_save(flags); ++ cpu_lock_irqsave(cpu, flags); + pset = per_cpu_ptr(zone->pageset, cpu); + + pcp = &pset->pcp; +- if (pcp->count) { +- free_pcppages_bulk(zone, pcp->count, pcp); ++ count = pcp->count; ++ if (count) { ++ isolate_pcp_pages(count, pcp, &dst); + pcp->count = 0; + } +- local_irq_restore(flags); ++ cpu_unlock_irqrestore(cpu, flags); ++ if (count) ++ free_pcppages_bulk(zone, count, &dst); + } + + /* +@@ -2366,8 +2412,17 @@ void drain_all_pages(struct zone *zone) + else + cpumask_clear_cpu(cpu, &cpus_with_pcps); + } ++#ifndef CONFIG_PREEMPT_RT_BASE + on_each_cpu_mask(&cpus_with_pcps, (smp_call_func_t) drain_local_pages, + zone, 1); ++#else ++ for_each_cpu(cpu, &cpus_with_pcps) { ++ if (zone) ++ drain_pages_zone(cpu, zone); ++ else ++ drain_pages(cpu); ++ } ++#endif + } + + #ifdef CONFIG_HIBERNATION +@@ -2427,7 +2482,7 @@ void free_hot_cold_page(struct page *page, bool cold) + + migratetype = get_pfnblock_migratetype(page, pfn); + set_pcppage_migratetype(page, migratetype); +- local_irq_save(flags); ++ local_lock_irqsave(pa_lock, flags); + __count_vm_event(PGFREE); + + /* +@@ -2453,12 +2508,17 @@ void free_hot_cold_page(struct page *page, bool cold) + pcp->count++; + if (pcp->count >= pcp->high) { + unsigned long batch = READ_ONCE(pcp->batch); +- free_pcppages_bulk(zone, batch, pcp); ++ LIST_HEAD(dst); ++ ++ isolate_pcp_pages(batch, pcp, &dst); + pcp->count -= batch; ++ local_unlock_irqrestore(pa_lock, flags); ++ free_pcppages_bulk(zone, batch, &dst); ++ return; + } + + out: +- local_irq_restore(flags); ++ local_unlock_irqrestore(pa_lock, flags); + } + + /* +@@ -2600,7 +2660,7 @@ struct page *buffered_rmqueue(struct zone *preferred_zone, + struct per_cpu_pages *pcp; + struct list_head *list; + +- local_irq_save(flags); ++ local_lock_irqsave(pa_lock, flags); + do { + pcp = &this_cpu_ptr(zone->pageset)->pcp; + list = &pcp->lists[migratetype]; +@@ -2627,7 +2687,7 @@ struct page *buffered_rmqueue(struct zone *preferred_zone, + * allocate greater than order-1 page units with __GFP_NOFAIL. + */ + WARN_ON_ONCE((gfp_flags & __GFP_NOFAIL) && (order > 1)); +- spin_lock_irqsave(&zone->lock, flags); ++ local_spin_lock_irqsave(pa_lock, &zone->lock, flags); + + do { + page = NULL; +@@ -2639,22 +2699,24 @@ struct page *buffered_rmqueue(struct zone *preferred_zone, + if (!page) + page = __rmqueue(zone, order, migratetype); + } while (page && check_new_pages(page, order)); +- spin_unlock(&zone->lock); +- if (!page) ++ if (!page) { ++ spin_unlock(&zone->lock); + goto failed; ++ } + __mod_zone_freepage_state(zone, -(1 << order), + get_pcppage_migratetype(page)); ++ spin_unlock(&zone->lock); + } + + __count_zid_vm_events(PGALLOC, page_zonenum(page), 1 << order); + zone_statistics(preferred_zone, zone, gfp_flags); +- local_irq_restore(flags); ++ local_unlock_irqrestore(pa_lock, flags); + + VM_BUG_ON_PAGE(bad_range(zone, page), page); + return page; + + failed: +- local_irq_restore(flags); ++ local_unlock_irqrestore(pa_lock, flags); + return NULL; + } + +@@ -6532,7 +6594,9 @@ static int page_alloc_cpu_notify(struct notifier_block *self, + int cpu = (unsigned long)hcpu; + + if (action == CPU_DEAD || action == CPU_DEAD_FROZEN) { ++ local_lock_irq_on(swapvec_lock, cpu); + lru_add_drain_cpu(cpu); ++ local_unlock_irq_on(swapvec_lock, cpu); + drain_pages(cpu); + + /* +@@ -6558,6 +6622,7 @@ static int page_alloc_cpu_notify(struct notifier_block *self, + void __init page_alloc_init(void) + { + hotcpu_notifier(page_alloc_cpu_notify, 0); ++ local_irq_lock_init(pa_lock); + } + + /* +@@ -7386,7 +7451,7 @@ void zone_pcp_reset(struct zone *zone) + struct per_cpu_pageset *pset; + + /* avoid races with drain_pages() */ +- local_irq_save(flags); ++ local_lock_irqsave(pa_lock, flags); + if (zone->pageset != &boot_pageset) { + for_each_online_cpu(cpu) { + pset = per_cpu_ptr(zone->pageset, cpu); +@@ -7395,7 +7460,7 @@ void zone_pcp_reset(struct zone *zone) + free_percpu(zone->pageset); + zone->pageset = &boot_pageset; + } +- local_irq_restore(flags); ++ local_unlock_irqrestore(pa_lock, flags); + } + + #ifdef CONFIG_MEMORY_HOTREMOVE +diff --git a/mm/percpu.c b/mm/percpu.c +index f014cebbf405..4e739fcf91bf 100644 +--- a/mm/percpu.c ++++ b/mm/percpu.c +@@ -1283,6 +1283,31 @@ void free_percpu(void __percpu *ptr) + } + EXPORT_SYMBOL_GPL(free_percpu); + ++bool __is_kernel_percpu_address(unsigned long addr, unsigned long *can_addr) ++{ ++#ifdef CONFIG_SMP ++ const size_t static_size = __per_cpu_end - __per_cpu_start; ++ void __percpu *base = __addr_to_pcpu_ptr(pcpu_base_addr); ++ unsigned int cpu; ++ ++ for_each_possible_cpu(cpu) { ++ void *start = per_cpu_ptr(base, cpu); ++ void *va = (void *)addr; ++ ++ if (va >= start && va < start + static_size) { ++ if (can_addr) { ++ *can_addr = (unsigned long) (va - start); ++ *can_addr += (unsigned long) ++ per_cpu_ptr(base, get_boot_cpu_id()); ++ } ++ return true; ++ } ++ } ++#endif ++ /* on UP, can't distinguish from other static vars, always false */ ++ return false; ++} ++ + /** + * is_kernel_percpu_address - test whether address is from static percpu area + * @addr: address to test +@@ -1296,20 +1321,7 @@ EXPORT_SYMBOL_GPL(free_percpu); + */ + bool is_kernel_percpu_address(unsigned long addr) + { +-#ifdef CONFIG_SMP +- const size_t static_size = __per_cpu_end - __per_cpu_start; +- void __percpu *base = __addr_to_pcpu_ptr(pcpu_base_addr); +- unsigned int cpu; +- +- for_each_possible_cpu(cpu) { +- void *start = per_cpu_ptr(base, cpu); +- +- if ((void *)addr >= start && (void *)addr < start + static_size) +- return true; +- } +-#endif +- /* on UP, can't distinguish from other static vars, always false */ +- return false; ++ return __is_kernel_percpu_address(addr, NULL); + } + + /** +diff --git a/mm/slab.h b/mm/slab.h +index ceb7d70cdb76..dfd281e43fbe 100644 +--- a/mm/slab.h ++++ b/mm/slab.h +@@ -426,7 +426,11 @@ static inline void slab_post_alloc_hook(struct kmem_cache *s, gfp_t flags, + * The slab lists for all objects. + */ + struct kmem_cache_node { ++#ifdef CONFIG_SLUB ++ raw_spinlock_t list_lock; ++#else + spinlock_t list_lock; ++#endif + + #ifdef CONFIG_SLAB + struct list_head slabs_partial; /* partial list first, better asm code */ +diff --git a/mm/slub.c b/mm/slub.c +index 58c7526f8de2..6d72b7f87129 100644 +--- a/mm/slub.c ++++ b/mm/slub.c +@@ -1141,7 +1141,7 @@ static noinline int free_debug_processing( + unsigned long uninitialized_var(flags); + int ret = 0; + +- spin_lock_irqsave(&n->list_lock, flags); ++ raw_spin_lock_irqsave(&n->list_lock, flags); + slab_lock(page); + + if (s->flags & SLAB_CONSISTENCY_CHECKS) { +@@ -1176,7 +1176,7 @@ static noinline int free_debug_processing( + bulk_cnt, cnt); + + slab_unlock(page); +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + if (!ret) + slab_fix(s, "Object at 0x%p not freed", object); + return ret; +@@ -1304,6 +1304,12 @@ static inline void dec_slabs_node(struct kmem_cache *s, int node, + + #endif /* CONFIG_SLUB_DEBUG */ + ++struct slub_free_list { ++ raw_spinlock_t lock; ++ struct list_head list; ++}; ++static DEFINE_PER_CPU(struct slub_free_list, slub_free_list); ++ + /* + * Hooks for other subsystems that check memory allocations. In a typical + * production configuration these hooks all should produce no code at all. +@@ -1527,10 +1533,17 @@ static struct page *allocate_slab(struct kmem_cache *s, gfp_t flags, int node) + void *start, *p; + int idx, order; + bool shuffle; ++ bool enableirqs = false; + + flags &= gfp_allowed_mask; + + if (gfpflags_allow_blocking(flags)) ++ enableirqs = true; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ if (system_state == SYSTEM_RUNNING) ++ enableirqs = true; ++#endif ++ if (enableirqs) + local_irq_enable(); + + flags |= s->allocflags; +@@ -1605,7 +1618,7 @@ static struct page *allocate_slab(struct kmem_cache *s, gfp_t flags, int node) + page->frozen = 1; + + out: +- if (gfpflags_allow_blocking(flags)) ++ if (enableirqs) + local_irq_disable(); + if (!page) + return NULL; +@@ -1664,6 +1677,16 @@ static void __free_slab(struct kmem_cache *s, struct page *page) + __free_pages(page, order); + } + ++static void free_delayed(struct list_head *h) ++{ ++ while(!list_empty(h)) { ++ struct page *page = list_first_entry(h, struct page, lru); ++ ++ list_del(&page->lru); ++ __free_slab(page->slab_cache, page); ++ } ++} ++ + #define need_reserve_slab_rcu \ + (sizeof(((struct page *)NULL)->lru) < sizeof(struct rcu_head)) + +@@ -1695,6 +1718,12 @@ static void free_slab(struct kmem_cache *s, struct page *page) + } + + call_rcu(head, rcu_free_slab); ++ } else if (irqs_disabled()) { ++ struct slub_free_list *f = this_cpu_ptr(&slub_free_list); ++ ++ raw_spin_lock(&f->lock); ++ list_add(&page->lru, &f->list); ++ raw_spin_unlock(&f->lock); + } else + __free_slab(s, page); + } +@@ -1802,7 +1831,7 @@ static void *get_partial_node(struct kmem_cache *s, struct kmem_cache_node *n, + if (!n || !n->nr_partial) + return NULL; + +- spin_lock(&n->list_lock); ++ raw_spin_lock(&n->list_lock); + list_for_each_entry_safe(page, page2, &n->partial, lru) { + void *t; + +@@ -1827,7 +1856,7 @@ static void *get_partial_node(struct kmem_cache *s, struct kmem_cache_node *n, + break; + + } +- spin_unlock(&n->list_lock); ++ raw_spin_unlock(&n->list_lock); + return object; + } + +@@ -2073,7 +2102,7 @@ static void deactivate_slab(struct kmem_cache *s, struct page *page, + * that acquire_slab() will see a slab page that + * is frozen + */ +- spin_lock(&n->list_lock); ++ raw_spin_lock(&n->list_lock); + } + } else { + m = M_FULL; +@@ -2084,7 +2113,7 @@ static void deactivate_slab(struct kmem_cache *s, struct page *page, + * slabs from diagnostic functions will not see + * any frozen slabs. + */ +- spin_lock(&n->list_lock); ++ raw_spin_lock(&n->list_lock); + } + } + +@@ -2119,7 +2148,7 @@ static void deactivate_slab(struct kmem_cache *s, struct page *page, + goto redo; + + if (lock) +- spin_unlock(&n->list_lock); ++ raw_spin_unlock(&n->list_lock); + + if (m == M_FREE) { + stat(s, DEACTIVATE_EMPTY); +@@ -2151,10 +2180,10 @@ static void unfreeze_partials(struct kmem_cache *s, + n2 = get_node(s, page_to_nid(page)); + if (n != n2) { + if (n) +- spin_unlock(&n->list_lock); ++ raw_spin_unlock(&n->list_lock); + + n = n2; +- spin_lock(&n->list_lock); ++ raw_spin_lock(&n->list_lock); + } + + do { +@@ -2183,7 +2212,7 @@ static void unfreeze_partials(struct kmem_cache *s, + } + + if (n) +- spin_unlock(&n->list_lock); ++ raw_spin_unlock(&n->list_lock); + + while (discard_page) { + page = discard_page; +@@ -2222,14 +2251,21 @@ static void put_cpu_partial(struct kmem_cache *s, struct page *page, int drain) + pobjects = oldpage->pobjects; + pages = oldpage->pages; + if (drain && pobjects > s->cpu_partial) { ++ struct slub_free_list *f; + unsigned long flags; ++ LIST_HEAD(tofree); + /* + * partial array is full. Move the existing + * set to the per node partial list. + */ + local_irq_save(flags); + unfreeze_partials(s, this_cpu_ptr(s->cpu_slab)); ++ f = this_cpu_ptr(&slub_free_list); ++ raw_spin_lock(&f->lock); ++ list_splice_init(&f->list, &tofree); ++ raw_spin_unlock(&f->lock); + local_irq_restore(flags); ++ free_delayed(&tofree); + oldpage = NULL; + pobjects = 0; + pages = 0; +@@ -2301,7 +2337,22 @@ static bool has_cpu_slab(int cpu, void *info) + + static void flush_all(struct kmem_cache *s) + { ++ LIST_HEAD(tofree); ++ int cpu; ++ + on_each_cpu_cond(has_cpu_slab, flush_cpu_slab, s, 1, GFP_ATOMIC); ++ for_each_online_cpu(cpu) { ++ struct slub_free_list *f; ++ ++ if (!has_cpu_slab(cpu, s)) ++ continue; ++ ++ f = &per_cpu(slub_free_list, cpu); ++ raw_spin_lock_irq(&f->lock); ++ list_splice_init(&f->list, &tofree); ++ raw_spin_unlock_irq(&f->lock); ++ free_delayed(&tofree); ++ } + } + + /* +@@ -2356,10 +2407,10 @@ static unsigned long count_partial(struct kmem_cache_node *n, + unsigned long x = 0; + struct page *page; + +- spin_lock_irqsave(&n->list_lock, flags); ++ raw_spin_lock_irqsave(&n->list_lock, flags); + list_for_each_entry(page, &n->partial, lru) + x += get_count(page); +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + return x; + } + #endif /* CONFIG_SLUB_DEBUG || CONFIG_SYSFS */ +@@ -2497,8 +2548,10 @@ static inline void *get_freelist(struct kmem_cache *s, struct page *page) + * already disabled (which is the case for bulk allocation). + */ + static void *___slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, +- unsigned long addr, struct kmem_cache_cpu *c) ++ unsigned long addr, struct kmem_cache_cpu *c, ++ struct list_head *to_free) + { ++ struct slub_free_list *f; + void *freelist; + struct page *page; + +@@ -2558,6 +2611,13 @@ static void *___slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, + VM_BUG_ON(!c->page->frozen); + c->freelist = get_freepointer(s, freelist); + c->tid = next_tid(c->tid); ++ ++out: ++ f = this_cpu_ptr(&slub_free_list); ++ raw_spin_lock(&f->lock); ++ list_splice_init(&f->list, to_free); ++ raw_spin_unlock(&f->lock); ++ + return freelist; + + new_slab: +@@ -2589,7 +2649,7 @@ static void *___slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, + deactivate_slab(s, page, get_freepointer(s, freelist)); + c->page = NULL; + c->freelist = NULL; +- return freelist; ++ goto out; + } + + /* +@@ -2601,6 +2661,7 @@ static void *__slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, + { + void *p; + unsigned long flags; ++ LIST_HEAD(tofree); + + local_irq_save(flags); + #ifdef CONFIG_PREEMPT +@@ -2612,8 +2673,9 @@ static void *__slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, + c = this_cpu_ptr(s->cpu_slab); + #endif + +- p = ___slab_alloc(s, gfpflags, node, addr, c); ++ p = ___slab_alloc(s, gfpflags, node, addr, c, &tofree); + local_irq_restore(flags); ++ free_delayed(&tofree); + return p; + } + +@@ -2799,7 +2861,7 @@ static void __slab_free(struct kmem_cache *s, struct page *page, + + do { + if (unlikely(n)) { +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + n = NULL; + } + prior = page->freelist; +@@ -2831,7 +2893,7 @@ static void __slab_free(struct kmem_cache *s, struct page *page, + * Otherwise the list_lock will synchronize with + * other processors updating the list of slabs. + */ +- spin_lock_irqsave(&n->list_lock, flags); ++ raw_spin_lock_irqsave(&n->list_lock, flags); + + } + } +@@ -2873,7 +2935,7 @@ static void __slab_free(struct kmem_cache *s, struct page *page, + add_partial(n, page, DEACTIVATE_TO_TAIL); + stat(s, FREE_ADD_PARTIAL); + } +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + return; + + slab_empty: +@@ -2888,7 +2950,7 @@ static void __slab_free(struct kmem_cache *s, struct page *page, + remove_full(s, n, page); + } + +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + stat(s, FREE_SLAB); + discard_slab(s, page); + } +@@ -3093,6 +3155,7 @@ int kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, + void **p) + { + struct kmem_cache_cpu *c; ++ LIST_HEAD(to_free); + int i; + + /* memcg and kmem_cache debug support */ +@@ -3116,7 +3179,7 @@ int kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, + * of re-populating per CPU c->freelist + */ + p[i] = ___slab_alloc(s, flags, NUMA_NO_NODE, +- _RET_IP_, c); ++ _RET_IP_, c, &to_free); + if (unlikely(!p[i])) + goto error; + +@@ -3128,6 +3191,7 @@ int kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, + } + c->tid = next_tid(c->tid); + local_irq_enable(); ++ free_delayed(&to_free); + + /* Clear memory outside IRQ disabled fastpath loop */ + if (unlikely(flags & __GFP_ZERO)) { +@@ -3275,7 +3339,7 @@ static void + init_kmem_cache_node(struct kmem_cache_node *n) + { + n->nr_partial = 0; +- spin_lock_init(&n->list_lock); ++ raw_spin_lock_init(&n->list_lock); + INIT_LIST_HEAD(&n->partial); + #ifdef CONFIG_SLUB_DEBUG + atomic_long_set(&n->nr_slabs, 0); +@@ -3619,6 +3683,10 @@ static void list_slab_objects(struct kmem_cache *s, struct page *page, + const char *text) + { + #ifdef CONFIG_SLUB_DEBUG ++#ifdef CONFIG_PREEMPT_RT_BASE ++ /* XXX move out of irq-off section */ ++ slab_err(s, page, text, s->name); ++#else + void *addr = page_address(page); + void *p; + unsigned long *map = kzalloc(BITS_TO_LONGS(page->objects) * +@@ -3639,6 +3707,7 @@ static void list_slab_objects(struct kmem_cache *s, struct page *page, + slab_unlock(page); + kfree(map); + #endif ++#endif + } + + /* +@@ -3652,7 +3721,7 @@ static void free_partial(struct kmem_cache *s, struct kmem_cache_node *n) + struct page *page, *h; + + BUG_ON(irqs_disabled()); +- spin_lock_irq(&n->list_lock); ++ raw_spin_lock_irq(&n->list_lock); + list_for_each_entry_safe(page, h, &n->partial, lru) { + if (!page->inuse) { + remove_partial(n, page); +@@ -3662,7 +3731,7 @@ static void free_partial(struct kmem_cache *s, struct kmem_cache_node *n) + "Objects remaining in %s on __kmem_cache_shutdown()"); + } + } +- spin_unlock_irq(&n->list_lock); ++ raw_spin_unlock_irq(&n->list_lock); + + list_for_each_entry_safe(page, h, &discard, lru) + discard_slab(s, page); +@@ -3905,7 +3974,7 @@ int __kmem_cache_shrink(struct kmem_cache *s) + for (i = 0; i < SHRINK_PROMOTE_MAX; i++) + INIT_LIST_HEAD(promote + i); + +- spin_lock_irqsave(&n->list_lock, flags); ++ raw_spin_lock_irqsave(&n->list_lock, flags); + + /* + * Build lists of slabs to discard or promote. +@@ -3936,7 +4005,7 @@ int __kmem_cache_shrink(struct kmem_cache *s) + for (i = SHRINK_PROMOTE_MAX - 1; i >= 0; i--) + list_splice(promote + i, &n->partial); + +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + + /* Release empty slabs */ + list_for_each_entry_safe(page, t, &discard, lru) +@@ -4112,6 +4181,12 @@ void __init kmem_cache_init(void) + { + static __initdata struct kmem_cache boot_kmem_cache, + boot_kmem_cache_node; ++ int cpu; ++ ++ for_each_possible_cpu(cpu) { ++ raw_spin_lock_init(&per_cpu(slub_free_list, cpu).lock); ++ INIT_LIST_HEAD(&per_cpu(slub_free_list, cpu).list); ++ } + + if (debug_guardpage_minorder()) + slub_max_order = 0; +@@ -4320,7 +4395,7 @@ static int validate_slab_node(struct kmem_cache *s, + struct page *page; + unsigned long flags; + +- spin_lock_irqsave(&n->list_lock, flags); ++ raw_spin_lock_irqsave(&n->list_lock, flags); + + list_for_each_entry(page, &n->partial, lru) { + validate_slab_slab(s, page, map); +@@ -4342,7 +4417,7 @@ static int validate_slab_node(struct kmem_cache *s, + s->name, count, atomic_long_read(&n->nr_slabs)); + + out: +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + return count; + } + +@@ -4530,12 +4605,12 @@ static int list_locations(struct kmem_cache *s, char *buf, + if (!atomic_long_read(&n->nr_slabs)) + continue; + +- spin_lock_irqsave(&n->list_lock, flags); ++ raw_spin_lock_irqsave(&n->list_lock, flags); + list_for_each_entry(page, &n->partial, lru) + process_slab(&t, s, page, alloc, map); + list_for_each_entry(page, &n->full, lru) + process_slab(&t, s, page, alloc, map); +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + } + + for (i = 0; i < t.count; i++) { +diff --git a/mm/swap.c b/mm/swap.c +index 4dcf852e1e6d..69c3a5b24060 100644 +--- a/mm/swap.c ++++ b/mm/swap.c +@@ -32,6 +32,7 @@ + #include + #include + #include ++#include + #include + #include + +@@ -50,6 +51,8 @@ static DEFINE_PER_CPU(struct pagevec, lru_deactivate_pvecs); + #ifdef CONFIG_SMP + static DEFINE_PER_CPU(struct pagevec, activate_page_pvecs); + #endif ++static DEFINE_LOCAL_IRQ_LOCK(rotate_lock); ++DEFINE_LOCAL_IRQ_LOCK(swapvec_lock); + + /* + * This path almost never happens for VM activity - pages are normally +@@ -240,11 +243,11 @@ void rotate_reclaimable_page(struct page *page) + unsigned long flags; + + get_page(page); +- local_irq_save(flags); ++ local_lock_irqsave(rotate_lock, flags); + pvec = this_cpu_ptr(&lru_rotate_pvecs); + if (!pagevec_add(pvec, page) || PageCompound(page)) + pagevec_move_tail(pvec); +- local_irq_restore(flags); ++ local_unlock_irqrestore(rotate_lock, flags); + } + } + +@@ -294,12 +297,13 @@ void activate_page(struct page *page) + { + page = compound_head(page); + if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) { +- struct pagevec *pvec = &get_cpu_var(activate_page_pvecs); ++ struct pagevec *pvec = &get_locked_var(swapvec_lock, ++ activate_page_pvecs); + + get_page(page); + if (!pagevec_add(pvec, page) || PageCompound(page)) + pagevec_lru_move_fn(pvec, __activate_page, NULL); +- put_cpu_var(activate_page_pvecs); ++ put_locked_var(swapvec_lock, activate_page_pvecs); + } + } + +@@ -326,7 +330,7 @@ void activate_page(struct page *page) + + static void __lru_cache_activate_page(struct page *page) + { +- struct pagevec *pvec = &get_cpu_var(lru_add_pvec); ++ struct pagevec *pvec = &get_locked_var(swapvec_lock, lru_add_pvec); + int i; + + /* +@@ -348,7 +352,7 @@ static void __lru_cache_activate_page(struct page *page) + } + } + +- put_cpu_var(lru_add_pvec); ++ put_locked_var(swapvec_lock, lru_add_pvec); + } + + /* +@@ -390,12 +394,12 @@ EXPORT_SYMBOL(mark_page_accessed); + + static void __lru_cache_add(struct page *page) + { +- struct pagevec *pvec = &get_cpu_var(lru_add_pvec); ++ struct pagevec *pvec = &get_locked_var(swapvec_lock, lru_add_pvec); + + get_page(page); + if (!pagevec_add(pvec, page) || PageCompound(page)) + __pagevec_lru_add(pvec); +- put_cpu_var(lru_add_pvec); ++ put_locked_var(swapvec_lock, lru_add_pvec); + } + + /** +@@ -593,9 +597,15 @@ void lru_add_drain_cpu(int cpu) + unsigned long flags; + + /* No harm done if a racing interrupt already did this */ +- local_irq_save(flags); ++#ifdef CONFIG_PREEMPT_RT_BASE ++ local_lock_irqsave_on(rotate_lock, flags, cpu); + pagevec_move_tail(pvec); +- local_irq_restore(flags); ++ local_unlock_irqrestore_on(rotate_lock, flags, cpu); ++#else ++ local_lock_irqsave(rotate_lock, flags); ++ pagevec_move_tail(pvec); ++ local_unlock_irqrestore(rotate_lock, flags); ++#endif + } + + pvec = &per_cpu(lru_deactivate_file_pvecs, cpu); +@@ -627,11 +637,12 @@ void deactivate_file_page(struct page *page) + return; + + if (likely(get_page_unless_zero(page))) { +- struct pagevec *pvec = &get_cpu_var(lru_deactivate_file_pvecs); ++ struct pagevec *pvec = &get_locked_var(swapvec_lock, ++ lru_deactivate_file_pvecs); + + if (!pagevec_add(pvec, page) || PageCompound(page)) + pagevec_lru_move_fn(pvec, lru_deactivate_file_fn, NULL); +- put_cpu_var(lru_deactivate_file_pvecs); ++ put_locked_var(swapvec_lock, lru_deactivate_file_pvecs); + } + } + +@@ -646,27 +657,31 @@ void deactivate_file_page(struct page *page) + void deactivate_page(struct page *page) + { + if (PageLRU(page) && PageActive(page) && !PageUnevictable(page)) { +- struct pagevec *pvec = &get_cpu_var(lru_deactivate_pvecs); ++ struct pagevec *pvec = &get_locked_var(swapvec_lock, ++ lru_deactivate_pvecs); + + get_page(page); + if (!pagevec_add(pvec, page) || PageCompound(page)) + pagevec_lru_move_fn(pvec, lru_deactivate_fn, NULL); +- put_cpu_var(lru_deactivate_pvecs); ++ put_locked_var(swapvec_lock, lru_deactivate_pvecs); + } + } + + void lru_add_drain(void) + { +- lru_add_drain_cpu(get_cpu()); +- put_cpu(); ++ lru_add_drain_cpu(local_lock_cpu(swapvec_lock)); ++ local_unlock_cpu(swapvec_lock); + } + +-static void lru_add_drain_per_cpu(struct work_struct *dummy) ++#ifdef CONFIG_PREEMPT_RT_BASE ++static inline void remote_lru_add_drain(int cpu, struct cpumask *has_work) + { +- lru_add_drain(); ++ local_lock_on(swapvec_lock, cpu); ++ lru_add_drain_cpu(cpu); ++ local_unlock_on(swapvec_lock, cpu); + } + +-static DEFINE_PER_CPU(struct work_struct, lru_add_drain_work); ++#else + + /* + * lru_add_drain_wq is used to do lru_add_drain_all() from a WQ_MEM_RECLAIM +@@ -686,6 +701,22 @@ static int __init lru_init(void) + } + early_initcall(lru_init); + ++static void lru_add_drain_per_cpu(struct work_struct *dummy) ++{ ++ lru_add_drain(); ++} ++ ++static DEFINE_PER_CPU(struct work_struct, lru_add_drain_work); ++static inline void remote_lru_add_drain(int cpu, struct cpumask *has_work) ++{ ++ struct work_struct *work = &per_cpu(lru_add_drain_work, cpu); ++ ++ INIT_WORK(work, lru_add_drain_per_cpu); ++ queue_work_on(cpu, lru_add_drain_wq, work); ++ cpumask_set_cpu(cpu, has_work); ++} ++#endif ++ + void lru_add_drain_all(void) + { + static DEFINE_MUTEX(lock); +@@ -697,21 +728,18 @@ void lru_add_drain_all(void) + cpumask_clear(&has_work); + + for_each_online_cpu(cpu) { +- struct work_struct *work = &per_cpu(lru_add_drain_work, cpu); +- + if (pagevec_count(&per_cpu(lru_add_pvec, cpu)) || + pagevec_count(&per_cpu(lru_rotate_pvecs, cpu)) || + pagevec_count(&per_cpu(lru_deactivate_file_pvecs, cpu)) || + pagevec_count(&per_cpu(lru_deactivate_pvecs, cpu)) || +- need_activate_page_drain(cpu)) { +- INIT_WORK(work, lru_add_drain_per_cpu); +- queue_work_on(cpu, lru_add_drain_wq, work); +- cpumask_set_cpu(cpu, &has_work); +- } ++ need_activate_page_drain(cpu)) ++ remote_lru_add_drain(cpu, &has_work); + } + ++#ifndef CONFIG_PREEMPT_RT_BASE + for_each_cpu(cpu, &has_work) + flush_work(&per_cpu(lru_add_drain_work, cpu)); ++#endif + + put_online_cpus(); + mutex_unlock(&lock); +diff --git a/mm/truncate.c b/mm/truncate.c +index 8d8c62d89e6d..5bf1bd25d077 100644 +--- a/mm/truncate.c ++++ b/mm/truncate.c +@@ -62,9 +62,12 @@ static void clear_exceptional_entry(struct address_space *mapping, + * protected by mapping->tree_lock. + */ + if (!workingset_node_shadows(node) && +- !list_empty(&node->private_list)) +- list_lru_del(&workingset_shadow_nodes, ++ !list_empty(&node->private_list)) { ++ local_lock(workingset_shadow_lock); ++ list_lru_del(&__workingset_shadow_nodes, + &node->private_list); ++ local_unlock(workingset_shadow_lock); ++ } + __radix_tree_delete_node(&mapping->page_tree, node); + unlock: + spin_unlock_irq(&mapping->tree_lock); +diff --git a/mm/vmalloc.c b/mm/vmalloc.c +index f2481cb4e6b2..db4de08fa97c 100644 +--- a/mm/vmalloc.c ++++ b/mm/vmalloc.c +@@ -845,7 +845,7 @@ static void *new_vmap_block(unsigned int order, gfp_t gfp_mask) + struct vmap_block *vb; + struct vmap_area *va; + unsigned long vb_idx; +- int node, err; ++ int node, err, cpu; + void *vaddr; + + node = numa_node_id(); +@@ -888,11 +888,12 @@ static void *new_vmap_block(unsigned int order, gfp_t gfp_mask) + BUG_ON(err); + radix_tree_preload_end(); + +- vbq = &get_cpu_var(vmap_block_queue); ++ cpu = get_cpu_light(); ++ vbq = this_cpu_ptr(&vmap_block_queue); + spin_lock(&vbq->lock); + list_add_tail_rcu(&vb->free_list, &vbq->free); + spin_unlock(&vbq->lock); +- put_cpu_var(vmap_block_queue); ++ put_cpu_light(); + + return vaddr; + } +@@ -961,6 +962,7 @@ static void *vb_alloc(unsigned long size, gfp_t gfp_mask) + struct vmap_block *vb; + void *vaddr = NULL; + unsigned int order; ++ int cpu; + + BUG_ON(offset_in_page(size)); + BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC); +@@ -975,7 +977,8 @@ static void *vb_alloc(unsigned long size, gfp_t gfp_mask) + order = get_order(size); + + rcu_read_lock(); +- vbq = &get_cpu_var(vmap_block_queue); ++ cpu = get_cpu_light(); ++ vbq = this_cpu_ptr(&vmap_block_queue); + list_for_each_entry_rcu(vb, &vbq->free, free_list) { + unsigned long pages_off; + +@@ -998,7 +1001,7 @@ static void *vb_alloc(unsigned long size, gfp_t gfp_mask) + break; + } + +- put_cpu_var(vmap_block_queue); ++ put_cpu_light(); + rcu_read_unlock(); + + /* Allocate new block if nothing was found */ +diff --git a/mm/vmstat.c b/mm/vmstat.c +index 604f26a4f696..312006d2db50 100644 +--- a/mm/vmstat.c ++++ b/mm/vmstat.c +@@ -245,6 +245,7 @@ void __mod_zone_page_state(struct zone *zone, enum zone_stat_item item, + long x; + long t; + ++ preempt_disable_rt(); + x = delta + __this_cpu_read(*p); + + t = __this_cpu_read(pcp->stat_threshold); +@@ -254,6 +255,7 @@ void __mod_zone_page_state(struct zone *zone, enum zone_stat_item item, + x = 0; + } + __this_cpu_write(*p, x); ++ preempt_enable_rt(); + } + EXPORT_SYMBOL(__mod_zone_page_state); + +@@ -265,6 +267,7 @@ void __mod_node_page_state(struct pglist_data *pgdat, enum node_stat_item item, + long x; + long t; + ++ preempt_disable_rt(); + x = delta + __this_cpu_read(*p); + + t = __this_cpu_read(pcp->stat_threshold); +@@ -274,6 +277,7 @@ void __mod_node_page_state(struct pglist_data *pgdat, enum node_stat_item item, + x = 0; + } + __this_cpu_write(*p, x); ++ preempt_enable_rt(); + } + EXPORT_SYMBOL(__mod_node_page_state); + +@@ -306,6 +310,7 @@ void __inc_zone_state(struct zone *zone, enum zone_stat_item item) + s8 __percpu *p = pcp->vm_stat_diff + item; + s8 v, t; + ++ preempt_disable_rt(); + v = __this_cpu_inc_return(*p); + t = __this_cpu_read(pcp->stat_threshold); + if (unlikely(v > t)) { +@@ -314,6 +319,7 @@ void __inc_zone_state(struct zone *zone, enum zone_stat_item item) + zone_page_state_add(v + overstep, zone, item); + __this_cpu_write(*p, -overstep); + } ++ preempt_enable_rt(); + } + + void __inc_node_state(struct pglist_data *pgdat, enum node_stat_item item) +@@ -322,6 +328,7 @@ void __inc_node_state(struct pglist_data *pgdat, enum node_stat_item item) + s8 __percpu *p = pcp->vm_node_stat_diff + item; + s8 v, t; + ++ preempt_disable_rt(); + v = __this_cpu_inc_return(*p); + t = __this_cpu_read(pcp->stat_threshold); + if (unlikely(v > t)) { +@@ -330,6 +337,7 @@ void __inc_node_state(struct pglist_data *pgdat, enum node_stat_item item) + node_page_state_add(v + overstep, pgdat, item); + __this_cpu_write(*p, -overstep); + } ++ preempt_enable_rt(); + } + + void __inc_zone_page_state(struct page *page, enum zone_stat_item item) +@@ -350,6 +358,7 @@ void __dec_zone_state(struct zone *zone, enum zone_stat_item item) + s8 __percpu *p = pcp->vm_stat_diff + item; + s8 v, t; + ++ preempt_disable_rt(); + v = __this_cpu_dec_return(*p); + t = __this_cpu_read(pcp->stat_threshold); + if (unlikely(v < - t)) { +@@ -358,6 +367,7 @@ void __dec_zone_state(struct zone *zone, enum zone_stat_item item) + zone_page_state_add(v - overstep, zone, item); + __this_cpu_write(*p, overstep); + } ++ preempt_enable_rt(); + } + + void __dec_node_state(struct pglist_data *pgdat, enum node_stat_item item) +@@ -366,6 +376,7 @@ void __dec_node_state(struct pglist_data *pgdat, enum node_stat_item item) + s8 __percpu *p = pcp->vm_node_stat_diff + item; + s8 v, t; + ++ preempt_disable_rt(); + v = __this_cpu_dec_return(*p); + t = __this_cpu_read(pcp->stat_threshold); + if (unlikely(v < - t)) { +@@ -374,6 +385,7 @@ void __dec_node_state(struct pglist_data *pgdat, enum node_stat_item item) + node_page_state_add(v - overstep, pgdat, item); + __this_cpu_write(*p, overstep); + } ++ preempt_enable_rt(); + } + + void __dec_zone_page_state(struct page *page, enum zone_stat_item item) +diff --git a/mm/workingset.c b/mm/workingset.c +index 4c4f05655e6e..b97b1e87b54c 100644 +--- a/mm/workingset.c ++++ b/mm/workingset.c +@@ -334,7 +334,8 @@ void workingset_activation(struct page *page) + * point where they would still be useful. + */ + +-struct list_lru workingset_shadow_nodes; ++struct list_lru __workingset_shadow_nodes; ++DEFINE_LOCAL_IRQ_LOCK(workingset_shadow_lock); + + static unsigned long count_shadow_nodes(struct shrinker *shrinker, + struct shrink_control *sc) +@@ -344,9 +345,9 @@ static unsigned long count_shadow_nodes(struct shrinker *shrinker, + unsigned long pages; + + /* list_lru lock nests inside IRQ-safe mapping->tree_lock */ +- local_irq_disable(); +- shadow_nodes = list_lru_shrink_count(&workingset_shadow_nodes, sc); +- local_irq_enable(); ++ local_lock_irq(workingset_shadow_lock); ++ shadow_nodes = list_lru_shrink_count(&__workingset_shadow_nodes, sc); ++ local_unlock_irq(workingset_shadow_lock); + + if (sc->memcg) { + pages = mem_cgroup_node_nr_lru_pages(sc->memcg, sc->nid, +@@ -438,9 +439,9 @@ static enum lru_status shadow_lru_isolate(struct list_head *item, + spin_unlock(&mapping->tree_lock); + ret = LRU_REMOVED_RETRY; + out: +- local_irq_enable(); ++ local_unlock_irq(workingset_shadow_lock); + cond_resched(); +- local_irq_disable(); ++ local_lock_irq(workingset_shadow_lock); + spin_lock(lru_lock); + return ret; + } +@@ -451,10 +452,10 @@ static unsigned long scan_shadow_nodes(struct shrinker *shrinker, + unsigned long ret; + + /* list_lru lock nests inside IRQ-safe mapping->tree_lock */ +- local_irq_disable(); +- ret = list_lru_shrink_walk(&workingset_shadow_nodes, sc, ++ local_lock_irq(workingset_shadow_lock); ++ ret = list_lru_shrink_walk(&__workingset_shadow_nodes, sc, + shadow_lru_isolate, NULL); +- local_irq_enable(); ++ local_unlock_irq(workingset_shadow_lock); + return ret; + } + +@@ -492,7 +493,7 @@ static int __init workingset_init(void) + pr_info("workingset: timestamp_bits=%d max_order=%d bucket_order=%u\n", + timestamp_bits, max_order, bucket_order); + +- ret = __list_lru_init(&workingset_shadow_nodes, true, &shadow_nodes_key); ++ ret = __list_lru_init(&__workingset_shadow_nodes, true, &shadow_nodes_key); + if (ret) + goto err; + ret = register_shrinker(&workingset_shadow_shrinker); +@@ -500,7 +501,7 @@ static int __init workingset_init(void) + goto err_list_lru; + return 0; + err_list_lru: +- list_lru_destroy(&workingset_shadow_nodes); ++ list_lru_destroy(&__workingset_shadow_nodes); + err: + return ret; + } +diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c +index 1689bb58e0d1..e52a8cb6aa5a 100644 +--- a/mm/zsmalloc.c ++++ b/mm/zsmalloc.c +@@ -53,6 +53,7 @@ + #include + #include + #include ++#include + + #define ZSPAGE_MAGIC 0x58 + +@@ -70,9 +71,22 @@ + */ + #define ZS_MAX_ZSPAGE_ORDER 2 + #define ZS_MAX_PAGES_PER_ZSPAGE (_AC(1, UL) << ZS_MAX_ZSPAGE_ORDER) +- + #define ZS_HANDLE_SIZE (sizeof(unsigned long)) + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ ++struct zsmalloc_handle { ++ unsigned long addr; ++ struct mutex lock; ++}; ++ ++#define ZS_HANDLE_ALLOC_SIZE (sizeof(struct zsmalloc_handle)) ++ ++#else ++ ++#define ZS_HANDLE_ALLOC_SIZE (sizeof(unsigned long)) ++#endif ++ + /* + * Object location (, ) is encoded as + * as single (unsigned long) handle value. +@@ -327,7 +341,7 @@ static void SetZsPageMovable(struct zs_pool *pool, struct zspage *zspage) {} + + static int create_cache(struct zs_pool *pool) + { +- pool->handle_cachep = kmem_cache_create("zs_handle", ZS_HANDLE_SIZE, ++ pool->handle_cachep = kmem_cache_create("zs_handle", ZS_HANDLE_ALLOC_SIZE, + 0, 0, NULL); + if (!pool->handle_cachep) + return 1; +@@ -351,10 +365,27 @@ static void destroy_cache(struct zs_pool *pool) + + static unsigned long cache_alloc_handle(struct zs_pool *pool, gfp_t gfp) + { +- return (unsigned long)kmem_cache_alloc(pool->handle_cachep, +- gfp & ~(__GFP_HIGHMEM|__GFP_MOVABLE)); ++ void *p; ++ ++ p = kmem_cache_alloc(pool->handle_cachep, ++ gfp & ~(__GFP_HIGHMEM|__GFP_MOVABLE)); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ if (p) { ++ struct zsmalloc_handle *zh = p; ++ ++ mutex_init(&zh->lock); ++ } ++#endif ++ return (unsigned long)p; + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++static struct zsmalloc_handle *zs_get_pure_handle(unsigned long handle) ++{ ++ return (void *)(handle &~((1 << OBJ_TAG_BITS) - 1)); ++} ++#endif ++ + static void cache_free_handle(struct zs_pool *pool, unsigned long handle) + { + kmem_cache_free(pool->handle_cachep, (void *)handle); +@@ -373,12 +404,18 @@ static void cache_free_zspage(struct zs_pool *pool, struct zspage *zspage) + + static void record_obj(unsigned long handle, unsigned long obj) + { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ struct zsmalloc_handle *zh = zs_get_pure_handle(handle); ++ ++ WRITE_ONCE(zh->addr, obj); ++#else + /* + * lsb of @obj represents handle lock while other bits + * represent object value the handle is pointing so + * updating shouldn't do store tearing. + */ + WRITE_ONCE(*(unsigned long *)handle, obj); ++#endif + } + + /* zpool driver */ +@@ -467,6 +504,7 @@ MODULE_ALIAS("zpool-zsmalloc"); + + /* per-cpu VM mapping areas for zspage accesses that cross page boundaries */ + static DEFINE_PER_CPU(struct mapping_area, zs_map_area); ++static DEFINE_LOCAL_IRQ_LOCK(zs_map_area_lock); + + static bool is_zspage_isolated(struct zspage *zspage) + { +@@ -902,7 +940,13 @@ static unsigned long location_to_obj(struct page *page, unsigned int obj_idx) + + static unsigned long handle_to_obj(unsigned long handle) + { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ struct zsmalloc_handle *zh = zs_get_pure_handle(handle); ++ ++ return zh->addr; ++#else + return *(unsigned long *)handle; ++#endif + } + + static unsigned long obj_to_head(struct page *page, void *obj) +@@ -916,22 +960,46 @@ static unsigned long obj_to_head(struct page *page, void *obj) + + static inline int testpin_tag(unsigned long handle) + { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ struct zsmalloc_handle *zh = zs_get_pure_handle(handle); ++ ++ return mutex_is_locked(&zh->lock); ++#else + return bit_spin_is_locked(HANDLE_PIN_BIT, (unsigned long *)handle); ++#endif + } + + static inline int trypin_tag(unsigned long handle) + { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ struct zsmalloc_handle *zh = zs_get_pure_handle(handle); ++ ++ return mutex_trylock(&zh->lock); ++#else + return bit_spin_trylock(HANDLE_PIN_BIT, (unsigned long *)handle); ++#endif + } + + static void pin_tag(unsigned long handle) + { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ struct zsmalloc_handle *zh = zs_get_pure_handle(handle); ++ ++ return mutex_lock(&zh->lock); ++#else + bit_spin_lock(HANDLE_PIN_BIT, (unsigned long *)handle); ++#endif + } + + static void unpin_tag(unsigned long handle) + { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ struct zsmalloc_handle *zh = zs_get_pure_handle(handle); ++ ++ return mutex_unlock(&zh->lock); ++#else + bit_spin_unlock(HANDLE_PIN_BIT, (unsigned long *)handle); ++#endif + } + + static void reset_page(struct page *page) +@@ -1423,7 +1491,7 @@ void *zs_map_object(struct zs_pool *pool, unsigned long handle, + class = pool->size_class[class_idx]; + off = (class->size * obj_idx) & ~PAGE_MASK; + +- area = &get_cpu_var(zs_map_area); ++ area = &get_locked_var(zs_map_area_lock, zs_map_area); + area->vm_mm = mm; + if (off + class->size <= PAGE_SIZE) { + /* this object is contained entirely within a page */ +@@ -1477,7 +1545,7 @@ void zs_unmap_object(struct zs_pool *pool, unsigned long handle) + + __zs_unmap_object(area, pages, off, class->size); + } +- put_cpu_var(zs_map_area); ++ put_locked_var(zs_map_area_lock, zs_map_area); + + migrate_read_unlock(zspage); + unpin_tag(handle); +diff --git a/net/core/dev.c b/net/core/dev.c +index 2e04fd188081..3ba60ef8c79e 100644 +--- a/net/core/dev.c ++++ b/net/core/dev.c +@@ -190,6 +190,7 @@ static unsigned int napi_gen_id = NR_CPUS; + static DEFINE_READ_MOSTLY_HASHTABLE(napi_hash, 8); + + static seqcount_t devnet_rename_seq; ++static DEFINE_MUTEX(devnet_rename_mutex); + + static inline void dev_base_seq_inc(struct net *net) + { +@@ -211,14 +212,14 @@ static inline struct hlist_head *dev_index_hash(struct net *net, int ifindex) + static inline void rps_lock(struct softnet_data *sd) + { + #ifdef CONFIG_RPS +- spin_lock(&sd->input_pkt_queue.lock); ++ raw_spin_lock(&sd->input_pkt_queue.raw_lock); + #endif + } + + static inline void rps_unlock(struct softnet_data *sd) + { + #ifdef CONFIG_RPS +- spin_unlock(&sd->input_pkt_queue.lock); ++ raw_spin_unlock(&sd->input_pkt_queue.raw_lock); + #endif + } + +@@ -888,7 +889,8 @@ int netdev_get_name(struct net *net, char *name, int ifindex) + strcpy(name, dev->name); + rcu_read_unlock(); + if (read_seqcount_retry(&devnet_rename_seq, seq)) { +- cond_resched(); ++ mutex_lock(&devnet_rename_mutex); ++ mutex_unlock(&devnet_rename_mutex); + goto retry; + } + +@@ -1157,20 +1159,17 @@ int dev_change_name(struct net_device *dev, const char *newname) + if (dev->flags & IFF_UP) + return -EBUSY; + +- write_seqcount_begin(&devnet_rename_seq); ++ mutex_lock(&devnet_rename_mutex); ++ __raw_write_seqcount_begin(&devnet_rename_seq); + +- if (strncmp(newname, dev->name, IFNAMSIZ) == 0) { +- write_seqcount_end(&devnet_rename_seq); +- return 0; +- } ++ if (strncmp(newname, dev->name, IFNAMSIZ) == 0) ++ goto outunlock; + + memcpy(oldname, dev->name, IFNAMSIZ); + + err = dev_get_valid_name(net, dev, newname); +- if (err < 0) { +- write_seqcount_end(&devnet_rename_seq); +- return err; +- } ++ if (err < 0) ++ goto outunlock; + + if (oldname[0] && !strchr(oldname, '%')) + netdev_info(dev, "renamed from %s\n", oldname); +@@ -1183,11 +1182,12 @@ int dev_change_name(struct net_device *dev, const char *newname) + if (ret) { + memcpy(dev->name, oldname, IFNAMSIZ); + dev->name_assign_type = old_assign_type; +- write_seqcount_end(&devnet_rename_seq); +- return ret; ++ err = ret; ++ goto outunlock; + } + +- write_seqcount_end(&devnet_rename_seq); ++ __raw_write_seqcount_end(&devnet_rename_seq); ++ mutex_unlock(&devnet_rename_mutex); + + netdev_adjacent_rename_links(dev, oldname); + +@@ -1208,7 +1208,8 @@ int dev_change_name(struct net_device *dev, const char *newname) + /* err >= 0 after dev_alloc_name() or stores the first errno */ + if (err >= 0) { + err = ret; +- write_seqcount_begin(&devnet_rename_seq); ++ mutex_lock(&devnet_rename_mutex); ++ __raw_write_seqcount_begin(&devnet_rename_seq); + memcpy(dev->name, oldname, IFNAMSIZ); + memcpy(oldname, newname, IFNAMSIZ); + dev->name_assign_type = old_assign_type; +@@ -1221,6 +1222,11 @@ int dev_change_name(struct net_device *dev, const char *newname) + } + + return err; ++ ++outunlock: ++ __raw_write_seqcount_end(&devnet_rename_seq); ++ mutex_unlock(&devnet_rename_mutex); ++ return err; + } + + /** +@@ -2285,6 +2291,7 @@ static void __netif_reschedule(struct Qdisc *q) + sd->output_queue_tailp = &q->next_sched; + raise_softirq_irqoff(NET_TX_SOFTIRQ); + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + + void __netif_schedule(struct Qdisc *q) +@@ -2366,6 +2373,7 @@ void __dev_kfree_skb_irq(struct sk_buff *skb, enum skb_free_reason reason) + __this_cpu_write(softnet_data.completion_queue, skb); + raise_softirq_irqoff(NET_TX_SOFTIRQ); + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + EXPORT_SYMBOL(__dev_kfree_skb_irq); + +@@ -3100,7 +3108,11 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, + * This permits qdisc->running owner to get the lock more + * often and dequeue packets faster. + */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ contended = true; ++#else + contended = qdisc_is_running(q); ++#endif + if (unlikely(contended)) + spin_lock(&q->busylock); + +@@ -3163,8 +3175,10 @@ static void skb_update_prio(struct sk_buff *skb) + #define skb_update_prio(skb) + #endif + ++#ifndef CONFIG_PREEMPT_RT_FULL + DEFINE_PER_CPU(int, xmit_recursion); + EXPORT_SYMBOL(xmit_recursion); ++#endif + + /** + * dev_loopback_xmit - loop back @skb +@@ -3398,8 +3412,7 @@ static int __dev_queue_xmit(struct sk_buff *skb, void *accel_priv) + int cpu = smp_processor_id(); /* ok because BHs are off */ + + if (txq->xmit_lock_owner != cpu) { +- if (unlikely(__this_cpu_read(xmit_recursion) > +- XMIT_RECURSION_LIMIT)) ++ if (unlikely(xmit_rec_read() > XMIT_RECURSION_LIMIT)) + goto recursion_alert; + + skb = validate_xmit_skb(skb, dev); +@@ -3409,9 +3422,9 @@ static int __dev_queue_xmit(struct sk_buff *skb, void *accel_priv) + HARD_TX_LOCK(dev, txq, cpu); + + if (!netif_xmit_stopped(txq)) { +- __this_cpu_inc(xmit_recursion); ++ xmit_rec_inc(); + skb = dev_hard_start_xmit(skb, dev, txq, &rc); +- __this_cpu_dec(xmit_recursion); ++ xmit_rec_dec(); + if (dev_xmit_complete(rc)) { + HARD_TX_UNLOCK(dev, txq); + goto out; +@@ -3785,6 +3798,7 @@ static int enqueue_to_backlog(struct sk_buff *skb, int cpu, + rps_unlock(sd); + + local_irq_restore(flags); ++ preempt_check_resched_rt(); + + atomic_long_inc(&skb->dev->rx_dropped); + kfree_skb(skb); +@@ -3803,7 +3817,7 @@ static int netif_rx_internal(struct sk_buff *skb) + struct rps_dev_flow voidflow, *rflow = &voidflow; + int cpu; + +- preempt_disable(); ++ migrate_disable(); + rcu_read_lock(); + + cpu = get_rps_cpu(skb->dev, skb, &rflow); +@@ -3813,13 +3827,13 @@ static int netif_rx_internal(struct sk_buff *skb) + ret = enqueue_to_backlog(skb, cpu, &rflow->last_qtail); + + rcu_read_unlock(); +- preempt_enable(); ++ migrate_enable(); + } else + #endif + { + unsigned int qtail; +- ret = enqueue_to_backlog(skb, get_cpu(), &qtail); +- put_cpu(); ++ ret = enqueue_to_backlog(skb, get_cpu_light(), &qtail); ++ put_cpu_light(); + } + return ret; + } +@@ -3853,11 +3867,9 @@ int netif_rx_ni(struct sk_buff *skb) + + trace_netif_rx_ni_entry(skb); + +- preempt_disable(); ++ local_bh_disable(); + err = netif_rx_internal(skb); +- if (local_softirq_pending()) +- do_softirq(); +- preempt_enable(); ++ local_bh_enable(); + + return err; + } +@@ -4336,7 +4348,7 @@ static void flush_backlog(struct work_struct *work) + skb_queue_walk_safe(&sd->input_pkt_queue, skb, tmp) { + if (skb->dev->reg_state == NETREG_UNREGISTERING) { + __skb_unlink(skb, &sd->input_pkt_queue); +- kfree_skb(skb); ++ __skb_queue_tail(&sd->tofree_queue, skb); + input_queue_head_incr(sd); + } + } +@@ -4346,11 +4358,14 @@ static void flush_backlog(struct work_struct *work) + skb_queue_walk_safe(&sd->process_queue, skb, tmp) { + if (skb->dev->reg_state == NETREG_UNREGISTERING) { + __skb_unlink(skb, &sd->process_queue); +- kfree_skb(skb); ++ __skb_queue_tail(&sd->tofree_queue, skb); + input_queue_head_incr(sd); + } + } ++ if (!skb_queue_empty(&sd->tofree_queue)) ++ raise_softirq_irqoff(NET_RX_SOFTIRQ); + local_bh_enable(); ++ + } + + static void flush_all_backlogs(void) +@@ -4831,6 +4846,7 @@ static void net_rps_action_and_irq_enable(struct softnet_data *sd) + sd->rps_ipi_list = NULL; + + local_irq_enable(); ++ preempt_check_resched_rt(); + + /* Send pending IPI's to kick RPS processing on remote cpus. */ + while (remsd) { +@@ -4844,6 +4860,7 @@ static void net_rps_action_and_irq_enable(struct softnet_data *sd) + } else + #endif + local_irq_enable(); ++ preempt_check_resched_rt(); + } + + static bool sd_has_rps_ipi_waiting(struct softnet_data *sd) +@@ -4873,7 +4890,9 @@ static int process_backlog(struct napi_struct *napi, int quota) + while (again) { + struct sk_buff *skb; + ++ local_irq_disable(); + while ((skb = __skb_dequeue(&sd->process_queue))) { ++ local_irq_enable(); + rcu_read_lock(); + __netif_receive_skb(skb); + rcu_read_unlock(); +@@ -4881,9 +4900,9 @@ static int process_backlog(struct napi_struct *napi, int quota) + if (++work >= quota) + return work; + ++ local_irq_disable(); + } + +- local_irq_disable(); + rps_lock(sd); + if (skb_queue_empty(&sd->input_pkt_queue)) { + /* +@@ -4921,9 +4940,11 @@ void __napi_schedule(struct napi_struct *n) + local_irq_save(flags); + ____napi_schedule(this_cpu_ptr(&softnet_data), n); + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + EXPORT_SYMBOL(__napi_schedule); + ++#ifndef CONFIG_PREEMPT_RT_FULL + /** + * __napi_schedule_irqoff - schedule for receive + * @n: entry to schedule +@@ -4935,6 +4956,7 @@ void __napi_schedule_irqoff(struct napi_struct *n) + ____napi_schedule(this_cpu_ptr(&softnet_data), n); + } + EXPORT_SYMBOL(__napi_schedule_irqoff); ++#endif + + void __napi_complete(struct napi_struct *n) + { +@@ -5224,13 +5246,21 @@ static __latent_entropy void net_rx_action(struct softirq_action *h) + struct softnet_data *sd = this_cpu_ptr(&softnet_data); + unsigned long time_limit = jiffies + 2; + int budget = netdev_budget; ++ struct sk_buff_head tofree_q; ++ struct sk_buff *skb; + LIST_HEAD(list); + LIST_HEAD(repoll); + ++ __skb_queue_head_init(&tofree_q); ++ + local_irq_disable(); ++ skb_queue_splice_init(&sd->tofree_queue, &tofree_q); + list_splice_init(&sd->poll_list, &list); + local_irq_enable(); + ++ while ((skb = __skb_dequeue(&tofree_q))) ++ kfree_skb(skb); ++ + for (;;) { + struct napi_struct *n; + +@@ -5261,7 +5291,7 @@ static __latent_entropy void net_rx_action(struct softirq_action *h) + list_splice_tail(&repoll, &list); + list_splice(&list, &sd->poll_list); + if (!list_empty(&sd->poll_list)) +- __raise_softirq_irqoff(NET_RX_SOFTIRQ); ++ __raise_softirq_irqoff_ksoft(NET_RX_SOFTIRQ); + + net_rps_action_and_irq_enable(sd); + } +@@ -8022,16 +8052,20 @@ static int dev_cpu_callback(struct notifier_block *nfb, + + raise_softirq_irqoff(NET_TX_SOFTIRQ); + local_irq_enable(); ++ preempt_check_resched_rt(); + + /* Process offline CPU's input_pkt_queue */ + while ((skb = __skb_dequeue(&oldsd->process_queue))) { + netif_rx_ni(skb); + input_queue_head_incr(oldsd); + } +- while ((skb = skb_dequeue(&oldsd->input_pkt_queue))) { ++ while ((skb = __skb_dequeue(&oldsd->input_pkt_queue))) { + netif_rx_ni(skb); + input_queue_head_incr(oldsd); + } ++ while ((skb = __skb_dequeue(&oldsd->tofree_queue))) { ++ kfree_skb(skb); ++ } + + return NOTIFY_OK; + } +@@ -8336,8 +8370,9 @@ static int __init net_dev_init(void) + + INIT_WORK(flush, flush_backlog); + +- skb_queue_head_init(&sd->input_pkt_queue); +- skb_queue_head_init(&sd->process_queue); ++ skb_queue_head_init_raw(&sd->input_pkt_queue); ++ skb_queue_head_init_raw(&sd->process_queue); ++ skb_queue_head_init_raw(&sd->tofree_queue); + INIT_LIST_HEAD(&sd->poll_list); + sd->output_queue_tailp = &sd->output_queue; + #ifdef CONFIG_RPS +diff --git a/net/core/filter.c b/net/core/filter.c +index b391209838ef..b86e9681a88e 100644 +--- a/net/core/filter.c ++++ b/net/core/filter.c +@@ -1645,7 +1645,7 @@ static inline int __bpf_tx_skb(struct net_device *dev, struct sk_buff *skb) + { + int ret; + +- if (unlikely(__this_cpu_read(xmit_recursion) > XMIT_RECURSION_LIMIT)) { ++ if (unlikely(xmit_rec_read() > XMIT_RECURSION_LIMIT)) { + net_crit_ratelimited("bpf: recursion limit reached on datapath, buggy bpf program?\n"); + kfree_skb(skb); + return -ENETDOWN; +@@ -1653,9 +1653,9 @@ static inline int __bpf_tx_skb(struct net_device *dev, struct sk_buff *skb) + + skb->dev = dev; + +- __this_cpu_inc(xmit_recursion); ++ xmit_rec_inc(); + ret = dev_queue_xmit(skb); +- __this_cpu_dec(xmit_recursion); ++ xmit_rec_dec(); + + return ret; + } +diff --git a/net/core/gen_estimator.c b/net/core/gen_estimator.c +index cad8e791f28e..2a9364fe62a5 100644 +--- a/net/core/gen_estimator.c ++++ b/net/core/gen_estimator.c +@@ -84,7 +84,7 @@ struct gen_estimator + struct gnet_stats_basic_packed *bstats; + struct gnet_stats_rate_est64 *rate_est; + spinlock_t *stats_lock; +- seqcount_t *running; ++ net_seqlock_t *running; + int ewma_log; + u32 last_packets; + unsigned long avpps; +@@ -213,7 +213,7 @@ int gen_new_estimator(struct gnet_stats_basic_packed *bstats, + struct gnet_stats_basic_cpu __percpu *cpu_bstats, + struct gnet_stats_rate_est64 *rate_est, + spinlock_t *stats_lock, +- seqcount_t *running, ++ net_seqlock_t *running, + struct nlattr *opt) + { + struct gen_estimator *est; +@@ -309,7 +309,7 @@ int gen_replace_estimator(struct gnet_stats_basic_packed *bstats, + struct gnet_stats_basic_cpu __percpu *cpu_bstats, + struct gnet_stats_rate_est64 *rate_est, + spinlock_t *stats_lock, +- seqcount_t *running, struct nlattr *opt) ++ net_seqlock_t *running, struct nlattr *opt) + { + gen_kill_estimator(bstats, rate_est); + return gen_new_estimator(bstats, cpu_bstats, rate_est, stats_lock, running, opt); +diff --git a/net/core/gen_stats.c b/net/core/gen_stats.c +index 508e051304fb..bc3b17b78c94 100644 +--- a/net/core/gen_stats.c ++++ b/net/core/gen_stats.c +@@ -130,7 +130,7 @@ __gnet_stats_copy_basic_cpu(struct gnet_stats_basic_packed *bstats, + } + + void +-__gnet_stats_copy_basic(const seqcount_t *running, ++__gnet_stats_copy_basic(net_seqlock_t *running, + struct gnet_stats_basic_packed *bstats, + struct gnet_stats_basic_cpu __percpu *cpu, + struct gnet_stats_basic_packed *b) +@@ -143,10 +143,10 @@ __gnet_stats_copy_basic(const seqcount_t *running, + } + do { + if (running) +- seq = read_seqcount_begin(running); ++ seq = net_seq_begin(running); + bstats->bytes = b->bytes; + bstats->packets = b->packets; +- } while (running && read_seqcount_retry(running, seq)); ++ } while (running && net_seq_retry(running, seq)); + } + EXPORT_SYMBOL(__gnet_stats_copy_basic); + +@@ -164,7 +164,7 @@ EXPORT_SYMBOL(__gnet_stats_copy_basic); + * if the room in the socket buffer was not sufficient. + */ + int +-gnet_stats_copy_basic(const seqcount_t *running, ++gnet_stats_copy_basic(net_seqlock_t *running, + struct gnet_dump *d, + struct gnet_stats_basic_cpu __percpu *cpu, + struct gnet_stats_basic_packed *b) +diff --git a/net/core/skbuff.c b/net/core/skbuff.c +index fe008f1bd930..9fa6bea3dd3f 100644 +--- a/net/core/skbuff.c ++++ b/net/core/skbuff.c +@@ -64,6 +64,7 @@ + #include + #include + #include ++#include + + #include + #include +@@ -360,6 +361,8 @@ struct napi_alloc_cache { + + static DEFINE_PER_CPU(struct page_frag_cache, netdev_alloc_cache); + static DEFINE_PER_CPU(struct napi_alloc_cache, napi_alloc_cache); ++static DEFINE_LOCAL_IRQ_LOCK(netdev_alloc_lock); ++static DEFINE_LOCAL_IRQ_LOCK(napi_alloc_cache_lock); + + static void *__netdev_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) + { +@@ -367,10 +370,10 @@ static void *__netdev_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) + unsigned long flags; + void *data; + +- local_irq_save(flags); ++ local_lock_irqsave(netdev_alloc_lock, flags); + nc = this_cpu_ptr(&netdev_alloc_cache); + data = __alloc_page_frag(nc, fragsz, gfp_mask); +- local_irq_restore(flags); ++ local_unlock_irqrestore(netdev_alloc_lock, flags); + return data; + } + +@@ -389,9 +392,13 @@ EXPORT_SYMBOL(netdev_alloc_frag); + + static void *__napi_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) + { +- struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); ++ struct napi_alloc_cache *nc; ++ void *data; + +- return __alloc_page_frag(&nc->page, fragsz, gfp_mask); ++ nc = &get_locked_var(napi_alloc_cache_lock, napi_alloc_cache); ++ data = __alloc_page_frag(&nc->page, fragsz, gfp_mask); ++ put_locked_var(napi_alloc_cache_lock, napi_alloc_cache); ++ return data; + } + + void *napi_alloc_frag(unsigned int fragsz) +@@ -438,13 +445,13 @@ struct sk_buff *__netdev_alloc_skb(struct net_device *dev, unsigned int len, + if (sk_memalloc_socks()) + gfp_mask |= __GFP_MEMALLOC; + +- local_irq_save(flags); ++ local_lock_irqsave(netdev_alloc_lock, flags); + + nc = this_cpu_ptr(&netdev_alloc_cache); + data = __alloc_page_frag(nc, len, gfp_mask); + pfmemalloc = nc->pfmemalloc; + +- local_irq_restore(flags); ++ local_unlock_irqrestore(netdev_alloc_lock, flags); + + if (unlikely(!data)) + return NULL; +@@ -485,9 +492,10 @@ EXPORT_SYMBOL(__netdev_alloc_skb); + struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len, + gfp_t gfp_mask) + { +- struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); ++ struct napi_alloc_cache *nc; + struct sk_buff *skb; + void *data; ++ bool pfmemalloc; + + len += NET_SKB_PAD + NET_IP_ALIGN; + +@@ -505,7 +513,10 @@ struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len, + if (sk_memalloc_socks()) + gfp_mask |= __GFP_MEMALLOC; + ++ nc = &get_locked_var(napi_alloc_cache_lock, napi_alloc_cache); + data = __alloc_page_frag(&nc->page, len, gfp_mask); ++ pfmemalloc = nc->page.pfmemalloc; ++ put_locked_var(napi_alloc_cache_lock, napi_alloc_cache); + if (unlikely(!data)) + return NULL; + +@@ -516,7 +527,7 @@ struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len, + } + + /* use OR instead of assignment to avoid clearing of bits in mask */ +- if (nc->page.pfmemalloc) ++ if (pfmemalloc) + skb->pfmemalloc = 1; + skb->head_frag = 1; + +@@ -760,23 +771,26 @@ EXPORT_SYMBOL(consume_skb); + + void __kfree_skb_flush(void) + { +- struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); ++ struct napi_alloc_cache *nc; + ++ nc = &get_locked_var(napi_alloc_cache_lock, napi_alloc_cache); + /* flush skb_cache if containing objects */ + if (nc->skb_count) { + kmem_cache_free_bulk(skbuff_head_cache, nc->skb_count, + nc->skb_cache); + nc->skb_count = 0; + } ++ put_locked_var(napi_alloc_cache_lock, napi_alloc_cache); + } + + static inline void _kfree_skb_defer(struct sk_buff *skb) + { +- struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); ++ struct napi_alloc_cache *nc; + + /* drop skb->head and call any destructors for packet */ + skb_release_all(skb); + ++ nc = &get_locked_var(napi_alloc_cache_lock, napi_alloc_cache); + /* record skb to CPU local list */ + nc->skb_cache[nc->skb_count++] = skb; + +@@ -791,6 +805,7 @@ static inline void _kfree_skb_defer(struct sk_buff *skb) + nc->skb_cache); + nc->skb_count = 0; + } ++ put_locked_var(napi_alloc_cache_lock, napi_alloc_cache); + } + void __kfree_skb_defer(struct sk_buff *skb) + { +diff --git a/net/core/sock.c b/net/core/sock.c +index 470a2043b846..2b09a5a33d8d 100644 +--- a/net/core/sock.c ++++ b/net/core/sock.c +@@ -2499,12 +2499,11 @@ void lock_sock_nested(struct sock *sk, int subclass) + if (sk->sk_lock.owned) + __lock_sock(sk); + sk->sk_lock.owned = 1; +- spin_unlock(&sk->sk_lock.slock); ++ spin_unlock_bh(&sk->sk_lock.slock); + /* + * The sk_lock has mutex_lock() semantics here: + */ + mutex_acquire(&sk->sk_lock.dep_map, subclass, 0, _RET_IP_); +- local_bh_enable(); + } + EXPORT_SYMBOL(lock_sock_nested); + +diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c +index 48734ee6293f..e6864ff11352 100644 +--- a/net/ipv4/icmp.c ++++ b/net/ipv4/icmp.c +@@ -69,6 +69,7 @@ + #include + #include + #include ++#include + #include + #include + #include +@@ -77,6 +78,7 @@ + #include + #include + #include ++#include + #include + #include + #include +@@ -204,6 +206,8 @@ static const struct icmp_control icmp_pointers[NR_ICMP_TYPES+1]; + * + * On SMP we have one ICMP socket per-cpu. + */ ++static DEFINE_LOCAL_IRQ_LOCK(icmp_sk_lock); ++ + static struct sock *icmp_sk(struct net *net) + { + return *this_cpu_ptr(net->ipv4.icmp_sk); +@@ -215,12 +219,14 @@ static inline struct sock *icmp_xmit_lock(struct net *net) + + local_bh_disable(); + ++ local_lock(icmp_sk_lock); + sk = icmp_sk(net); + + if (unlikely(!spin_trylock(&sk->sk_lock.slock))) { + /* This can happen if the output path signals a + * dst_link_failure() for an outgoing ICMP packet. + */ ++ local_unlock(icmp_sk_lock); + local_bh_enable(); + return NULL; + } +@@ -230,6 +236,7 @@ static inline struct sock *icmp_xmit_lock(struct net *net) + static inline void icmp_xmit_unlock(struct sock *sk) + { + spin_unlock_bh(&sk->sk_lock.slock); ++ local_unlock(icmp_sk_lock); + } + + int sysctl_icmp_msgs_per_sec __read_mostly = 1000; +@@ -358,6 +365,7 @@ static void icmp_push_reply(struct icmp_bxm *icmp_param, + struct sock *sk; + struct sk_buff *skb; + ++ local_lock(icmp_sk_lock); + sk = icmp_sk(dev_net((*rt)->dst.dev)); + if (ip_append_data(sk, fl4, icmp_glue_bits, icmp_param, + icmp_param->data_len+icmp_param->head_len, +@@ -380,6 +388,7 @@ static void icmp_push_reply(struct icmp_bxm *icmp_param, + skb->ip_summed = CHECKSUM_NONE; + ip_push_pending_frames(sk, fl4); + } ++ local_unlock(icmp_sk_lock); + } + + /* +@@ -891,6 +900,30 @@ static bool icmp_redirect(struct sk_buff *skb) + } + + /* ++ * 32bit and 64bit have different timestamp length, so we check for ++ * the cookie at offset 20 and verify it is repeated at offset 50 ++ */ ++#define CO_POS0 20 ++#define CO_POS1 50 ++#define CO_SIZE sizeof(int) ++#define ICMP_SYSRQ_SIZE 57 ++ ++/* ++ * We got a ICMP_SYSRQ_SIZE sized ping request. Check for the cookie ++ * pattern and if it matches send the next byte as a trigger to sysrq. ++ */ ++static void icmp_check_sysrq(struct net *net, struct sk_buff *skb) ++{ ++ int cookie = htonl(net->ipv4.sysctl_icmp_echo_sysrq); ++ char *p = skb->data; ++ ++ if (!memcmp(&cookie, p + CO_POS0, CO_SIZE) && ++ !memcmp(&cookie, p + CO_POS1, CO_SIZE) && ++ p[CO_POS0 + CO_SIZE] == p[CO_POS1 + CO_SIZE]) ++ handle_sysrq(p[CO_POS0 + CO_SIZE]); ++} ++ ++/* + * Handle ICMP_ECHO ("ping") requests. + * + * RFC 1122: 3.2.2.6 MUST have an echo server that answers ICMP echo +@@ -917,6 +950,11 @@ static bool icmp_echo(struct sk_buff *skb) + icmp_param.data_len = skb->len; + icmp_param.head_len = sizeof(struct icmphdr); + icmp_reply(&icmp_param, skb); ++ ++ if (skb->len == ICMP_SYSRQ_SIZE && ++ net->ipv4.sysctl_icmp_echo_sysrq) { ++ icmp_check_sysrq(net, skb); ++ } + } + /* should there be an ICMP stat for ignored echos? */ + return true; +diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c +index 80bc36b25de2..215b90adfb05 100644 +--- a/net/ipv4/sysctl_net_ipv4.c ++++ b/net/ipv4/sysctl_net_ipv4.c +@@ -681,6 +681,13 @@ static struct ctl_table ipv4_net_table[] = { + .proc_handler = proc_dointvec + }, + { ++ .procname = "icmp_echo_sysrq", ++ .data = &init_net.ipv4.sysctl_icmp_echo_sysrq, ++ .maxlen = sizeof(int), ++ .mode = 0644, ++ .proc_handler = proc_dointvec ++ }, ++ { + .procname = "icmp_ignore_bogus_error_responses", + .data = &init_net.ipv4.sysctl_icmp_ignore_bogus_error_responses, + .maxlen = sizeof(int), +diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c +index 6988566dc72f..672fffcde28c 100644 +--- a/net/ipv4/tcp_ipv4.c ++++ b/net/ipv4/tcp_ipv4.c +@@ -62,6 +62,7 @@ + #include + #include + #include ++#include + + #include + #include +@@ -568,6 +569,7 @@ void tcp_v4_send_check(struct sock *sk, struct sk_buff *skb) + } + EXPORT_SYMBOL(tcp_v4_send_check); + ++static DEFINE_LOCAL_IRQ_LOCK(tcp_sk_lock); + /* + * This routine will send an RST to the other tcp. + * +@@ -695,6 +697,8 @@ static void tcp_v4_send_reset(const struct sock *sk, struct sk_buff *skb) + offsetof(struct inet_timewait_sock, tw_bound_dev_if)); + + arg.tos = ip_hdr(skb)->tos; ++ ++ local_lock(tcp_sk_lock); + local_bh_disable(); + ip_send_unicast_reply(*this_cpu_ptr(net->ipv4.tcp_sk), + skb, &TCP_SKB_CB(skb)->header.h4.opt, +@@ -704,6 +708,7 @@ static void tcp_v4_send_reset(const struct sock *sk, struct sk_buff *skb) + __TCP_INC_STATS(net, TCP_MIB_OUTSEGS); + __TCP_INC_STATS(net, TCP_MIB_OUTRSTS); + local_bh_enable(); ++ local_unlock(tcp_sk_lock); + + #ifdef CONFIG_TCP_MD5SIG + out: +@@ -779,6 +784,7 @@ static void tcp_v4_send_ack(struct net *net, + if (oif) + arg.bound_dev_if = oif; + arg.tos = tos; ++ local_lock(tcp_sk_lock); + local_bh_disable(); + ip_send_unicast_reply(*this_cpu_ptr(net->ipv4.tcp_sk), + skb, &TCP_SKB_CB(skb)->header.h4.opt, +@@ -787,6 +793,7 @@ static void tcp_v4_send_ack(struct net *net, + + __TCP_INC_STATS(net, TCP_MIB_OUTSEGS); + local_bh_enable(); ++ local_unlock(tcp_sk_lock); + } + + static void tcp_v4_timewait_ack(struct sock *sk, struct sk_buff *skb) +diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c +index acaaf616da71..09020dbcc089 100644 +--- a/net/mac80211/rx.c ++++ b/net/mac80211/rx.c +@@ -4230,7 +4230,7 @@ void ieee80211_rx_napi(struct ieee80211_hw *hw, struct ieee80211_sta *pubsta, + struct ieee80211_supported_band *sband; + struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(skb); + +- WARN_ON_ONCE(softirq_count() == 0); ++ WARN_ON_ONCE_NONRT(softirq_count() == 0); + + if (WARN_ON(status->band >= NUM_NL80211_BANDS)) + goto drop; +diff --git a/net/netfilter/core.c b/net/netfilter/core.c +index 004af030ef1a..b64f751bda45 100644 +--- a/net/netfilter/core.c ++++ b/net/netfilter/core.c +@@ -22,12 +22,18 @@ + #include + #include + #include ++#include + #include + #include + #include + + #include "nf_internals.h" + ++#ifdef CONFIG_PREEMPT_RT_BASE ++DEFINE_LOCAL_IRQ_LOCK(xt_write_lock); ++EXPORT_PER_CPU_SYMBOL(xt_write_lock); ++#endif ++ + static DEFINE_MUTEX(afinfo_mutex); + + const struct nf_afinfo __rcu *nf_afinfo[NFPROTO_NUMPROTO] __read_mostly; +diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c +index cb76ff3088e9..3f42c5b1af55 100644 +--- a/net/packet/af_packet.c ++++ b/net/packet/af_packet.c +@@ -63,6 +63,7 @@ + #include + #include + #include ++#include + #include + #include + #include +@@ -694,7 +695,7 @@ static void prb_retire_rx_blk_timer_expired(unsigned long data) + if (BLOCK_NUM_PKTS(pbd)) { + while (atomic_read(&pkc->blk_fill_in_prog)) { + /* Waiting for skb_copy_bits to finish... */ +- cpu_relax(); ++ cpu_chill(); + } + } + +@@ -956,7 +957,7 @@ static void prb_retire_current_block(struct tpacket_kbdq_core *pkc, + if (!(status & TP_STATUS_BLK_TMO)) { + while (atomic_read(&pkc->blk_fill_in_prog)) { + /* Waiting for skb_copy_bits to finish... */ +- cpu_relax(); ++ cpu_chill(); + } + } + prb_close_block(pkc, pbd, po, status); +diff --git a/net/rds/ib_rdma.c b/net/rds/ib_rdma.c +index 977f69886c00..f3e7a36b0396 100644 +--- a/net/rds/ib_rdma.c ++++ b/net/rds/ib_rdma.c +@@ -34,6 +34,7 @@ + #include + #include + #include ++#include + + #include "rds_single_path.h" + #include "ib_mr.h" +@@ -210,7 +211,7 @@ static inline void wait_clean_list_grace(void) + for_each_online_cpu(cpu) { + flag = &per_cpu(clean_list_grace, cpu); + while (test_bit(CLEAN_LIST_BUSY_BIT, flag)) +- cpu_relax(); ++ cpu_chill(); + } + } + +diff --git a/net/rxrpc/security.c b/net/rxrpc/security.c +index 7d921e56e715..13df56a738e5 100644 +--- a/net/rxrpc/security.c ++++ b/net/rxrpc/security.c +@@ -19,9 +19,6 @@ + #include + #include "ar-internal.h" + +-static LIST_HEAD(rxrpc_security_methods); +-static DECLARE_RWSEM(rxrpc_security_sem); +- + static const struct rxrpc_security *rxrpc_security_types[] = { + [RXRPC_SECURITY_NONE] = &rxrpc_no_security, + #ifdef CONFIG_RXKAD +diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c +index 206dc24add3a..00ea9bde5bb3 100644 +--- a/net/sched/sch_api.c ++++ b/net/sched/sch_api.c +@@ -981,7 +981,7 @@ static struct Qdisc *qdisc_create(struct net_device *dev, + rcu_assign_pointer(sch->stab, stab); + } + if (tca[TCA_RATE]) { +- seqcount_t *running; ++ net_seqlock_t *running; + + err = -EOPNOTSUPP; + if (sch->flags & TCQ_F_MQROOT) +diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c +index 6cfb6e9038c2..20727e1347de 100644 +--- a/net/sched/sch_generic.c ++++ b/net/sched/sch_generic.c +@@ -425,7 +425,11 @@ struct Qdisc noop_qdisc = { + .ops = &noop_qdisc_ops, + .q.lock = __SPIN_LOCK_UNLOCKED(noop_qdisc.q.lock), + .dev_queue = &noop_netdev_queue, ++#ifdef CONFIG_PREEMPT_RT_BASE ++ .running = __SEQLOCK_UNLOCKED(noop_qdisc.running), ++#else + .running = SEQCNT_ZERO(noop_qdisc.running), ++#endif + .busylock = __SPIN_LOCK_UNLOCKED(noop_qdisc.busylock), + }; + EXPORT_SYMBOL(noop_qdisc); +@@ -624,9 +628,17 @@ struct Qdisc *qdisc_alloc(struct netdev_queue *dev_queue, + lockdep_set_class(&sch->busylock, + dev->qdisc_tx_busylock ?: &qdisc_tx_busylock); + ++#ifdef CONFIG_PREEMPT_RT_BASE ++ seqlock_init(&sch->running); ++ lockdep_set_class(&sch->running.seqcount, ++ dev->qdisc_running_key ?: &qdisc_running_key); ++ lockdep_set_class(&sch->running.lock, ++ dev->qdisc_running_key ?: &qdisc_running_key); ++#else + seqcount_init(&sch->running); + lockdep_set_class(&sch->running, + dev->qdisc_running_key ?: &qdisc_running_key); ++#endif + + sch->ops = ops; + sch->enqueue = ops->enqueue; +@@ -925,7 +937,7 @@ void dev_deactivate_many(struct list_head *head) + /* Wait for outstanding qdisc_run calls. */ + list_for_each_entry(dev, head, close_list) + while (some_qdisc_is_busy(dev)) +- yield(); ++ msleep(1); + } + + void dev_deactivate(struct net_device *dev) +diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c +index 9c9db55a0c1e..e6583b018a72 100644 +--- a/net/sunrpc/svc_xprt.c ++++ b/net/sunrpc/svc_xprt.c +@@ -396,7 +396,7 @@ void svc_xprt_do_enqueue(struct svc_xprt *xprt) + goto out; + } + +- cpu = get_cpu(); ++ cpu = get_cpu_light(); + pool = svc_pool_for_cpu(xprt->xpt_server, cpu); + + atomic_long_inc(&pool->sp_stats.packets); +@@ -432,7 +432,7 @@ void svc_xprt_do_enqueue(struct svc_xprt *xprt) + + atomic_long_inc(&pool->sp_stats.threads_woken); + wake_up_process(rqstp->rq_task); +- put_cpu(); ++ put_cpu_light(); + goto out; + } + rcu_read_unlock(); +@@ -453,7 +453,7 @@ void svc_xprt_do_enqueue(struct svc_xprt *xprt) + goto redo_search; + } + rqstp = NULL; +- put_cpu(); ++ put_cpu_light(); + out: + trace_svc_xprt_do_enqueue(xprt, rqstp); + } +diff --git a/scripts/mkcompile_h b/scripts/mkcompile_h +index 6fdc97ef6023..523e0420d7f0 100755 +--- a/scripts/mkcompile_h ++++ b/scripts/mkcompile_h +@@ -4,7 +4,8 @@ TARGET=$1 + ARCH=$2 + SMP=$3 + PREEMPT=$4 +-CC=$5 ++RT=$5 ++CC=$6 + + vecho() { [ "${quiet}" = "silent_" ] || echo "$@" ; } + +@@ -57,6 +58,7 @@ UTS_VERSION="#$VERSION" + CONFIG_FLAGS="" + if [ -n "$SMP" ] ; then CONFIG_FLAGS="SMP"; fi + if [ -n "$PREEMPT" ] ; then CONFIG_FLAGS="$CONFIG_FLAGS PREEMPT"; fi ++if [ -n "$RT" ] ; then CONFIG_FLAGS="$CONFIG_FLAGS RT"; fi + UTS_VERSION="$UTS_VERSION $CONFIG_FLAGS $TIMESTAMP" + + # Truncate to maximum length +diff --git a/sound/core/pcm_native.c b/sound/core/pcm_native.c +index 9d33c1e85c79..3d307bda86f9 100644 +--- a/sound/core/pcm_native.c ++++ b/sound/core/pcm_native.c +@@ -135,7 +135,7 @@ EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock); + void snd_pcm_stream_lock_irq(struct snd_pcm_substream *substream) + { + if (!substream->pcm->nonatomic) +- local_irq_disable(); ++ local_irq_disable_nort(); + snd_pcm_stream_lock(substream); + } + EXPORT_SYMBOL_GPL(snd_pcm_stream_lock_irq); +@@ -150,7 +150,7 @@ void snd_pcm_stream_unlock_irq(struct snd_pcm_substream *substream) + { + snd_pcm_stream_unlock(substream); + if (!substream->pcm->nonatomic) +- local_irq_enable(); ++ local_irq_enable_nort(); + } + EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock_irq); + +@@ -158,7 +158,7 @@ unsigned long _snd_pcm_stream_lock_irqsave(struct snd_pcm_substream *substream) + { + unsigned long flags = 0; + if (!substream->pcm->nonatomic) +- local_irq_save(flags); ++ local_irq_save_nort(flags); + snd_pcm_stream_lock(substream); + return flags; + } +@@ -176,7 +176,7 @@ void snd_pcm_stream_unlock_irqrestore(struct snd_pcm_substream *substream, + { + snd_pcm_stream_unlock(substream); + if (!substream->pcm->nonatomic) +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock_irqrestore); + -- cgit v1.2.3