diff options
| author | Eric Andersen <andersen@codepoet.org> | 2002-02-20 09:18:50 +0000 |
|---|---|---|
| committer | Eric Andersen <andersen@codepoet.org> | 2002-02-20 09:18:50 +0000 |
| commit | e356ea321c8098cf1a83a67e27d64c44de08a298 (patch) | |
| tree | 8be9273fb8f0e6acab47a9e09552cfbea5400b31 /libpthread/linuxthreads | |
| parent | 07ebf927b17572d92e785533d6e8ac1668cc57c6 (diff) | |
Merge in the pthread library. This is the linuxthreads library taken from
glibc 2.1.3 and ported to work with uClibc by Stefan Soucek and Erik Andersen
(me). Stefan has hacked things up such that linuxthreads runs on MMU-less
systems (tested only on arm-nommu). Erik cleaned things up and made it work
properly as a shared library.
-Erik
Diffstat (limited to 'libpthread/linuxthreads')
64 files changed, 12674 insertions, 0 deletions
diff --git a/libpthread/linuxthreads/ChangeLog b/libpthread/linuxthreads/ChangeLog new file mode 100644 index 000000000..0025b87c0 --- /dev/null +++ b/libpthread/linuxthreads/ChangeLog @@ -0,0 +1,1257 @@ +2000-02-22 Ulrich Drepper <drepper@redhat.com> + + * semaphore.h (SEM_FAILED): Use 0 not NULL. + +2000-02-14 Ulrich Drepper <drepper@redhat.com> + + * condvar.c (pthread_cond_timedwait_relative_old): Tight loop with + nanosleep does not work either. Get absolute time inside the + loop. + (pthread_cond_timedwait_relative_new): Likewise. + Patch by Kaz Kylheku <kaz@ashi.footprints.net>. + +2000-02-13 Ulrich Drepper <drepper@redhat.com> + + * condvar.c (pthread_cond_timedwait_relative_old): Undo last patch + but keep the code around. A bug in the kernel prevent us from + using the code. + (pthread_cond_timedwait_relative_new): Likewise. + (PR libc/1597 and libc/1598). + +2000-02-01 Kaz Kylheku <kaz@ashi.footprints.net> + + * condvar.c (pthread_cond_timedwait_relative_old): Do tight + loop around nanosleep calls instead of around most of the function + (pthread_cond_timedwait_relative_new): Likewise. + body. Got rid of backwards goto and one local. + +2000-01-31 Ulrich Drepper <drepper@redhat.com> + + * condvar.c (pthread_cond_timedwait_relative_old): Recompute time + before every nanosleep call to account for time spent in the rest + of the function. + (pthread_cond_timedwait_relative_new): Likewise. + Patch by khendricks@ivey.uwo.ca (PR libc/1564). + +2000-01-29 Ulrich Drepper <drepper@redhat.com> + + * condvar.c (pthread_cond_timedwait_relative_old): Get remaining time + from nanosleep call so that in case we restart we only wait for the + remaining time. + (pthread_cond_timedwait_relative_new): Likewise. + Patch by khendricks@ivey.uwo.ca (PR libc/1561). + +2000-01-18 Ulrich Drepper <drepper@cygnus.com> + + * manager.c (pthread_allocate_stack): Compute guard page address + correctly. Patch by HJ Lu. + +2000-01-12 Ulrich Drepper <drepper@cygnus.com> + + * internals.h (pthread_readlock_info): New structure. + (_pthread_descr_struct): Add p_readlock_list, p_readlock_free, and + p_untracked_readlock_count. + * pthread.c (__pthread_initial_thread, pthread_manager_thread): + Add initializers for new fields. + * manager.c (pthread_free): Free read/write lock lists. + * queue.h (queue_is_empty): New function. + * rwlock.c: Implement requirements about when readers should get + locks assigned. + * sysdeps/pthread/pthread.h + (PTHREAD_RWLOCK_PREFER_WRITER_NONRECURSIVE_NP): New definition. + * sysdeps/pthread/bits/pthreadtypes.h (struct _pthread_rwlock_t): + Define this name as well. + Patches by Kaz Kylheku <kaz@ashi.footprints.net>. + +2000-01-06 Andreas Jaeger <aj@suse.de> + + * pthread.c: Remove extra initializer. + +2000-01-05 Ulrich Drepper <drepper@cygnus.com> + + * pthread.c (__pthread_initial_thread, pthread_manager_thread): + Adjust initializers for struct _pthread_descr_struct change. + * internals.h (struct _pthread_descr_struct): Move new elements to + the end. + +2000-01-03 Kaz Kylheku <kaz@ashi.footprints.net> + + Redesigned how cancellation unblocks a thread from internal + cancellation points (sem_wait, pthread_join, + pthread_cond_{wait,timedwait}). + Cancellation won't eat a signal in any of these functions + (*required* by POSIX and Single Unix Spec!). + * condvar.c: Spontaneous wakeup on pthread_cond_timedwait won't eat a + simultaneous condition variable signal (not required by POSIX + or Single Unix Spec, but nice). + * spinlock.c: __pthread_lock queues back any received restarts + that don't belong to it instead of assuming ownership of lock + upon any restart; fastlock can no longer be acquired by two threads + simultaneously. + * restart.h: Restarts queue even on kernels that don't have + queued real time signals (2.0, early 2.1), thanks to atomic counter, + avoiding a rare race condition in pthread_cond_timedwait. + +1999-12-28 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/alpha/pt-machine.h: Move stack_pointer definition to the + beginning. + + * manager.c (__pthread_start): Add one more cast to assignment of + arg to prevent warning on 64bit machines. + +1999-12-21 Ulrich Drepper <drepper@cygnus.com> + + * manager.c (pthread_handle_create): Set p_pid of new thread + before calling the callback function to report a new thread. + +1999-12-20 Andreas Jaeger <aj@suse.de> + + * pthread.c (pthread_initialize): Move getrlimit call after + setting of errno. + +1999-12-01 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/i386/pt-machine.h: Move stack_pointer definition to the + beginning. + * sysdeps/i386/i686/pt-machine.h: Likewise. + Patches by Alan Modra <alan@SPRI.Levels.UniSA.Edu.Au>. + +1999-11-23 Ulrich Drepper <drepper@cygnus.com> + + * manager.c (pthread_start_thread_event): Initialize p_pid already + here. + +1999-11-22 Ulrich Drepper <drepper@cygnus.com> + + * internals.h: Add prototype for __pthread_manager_event. + * manager.c (__pthread_manager_event): New function. + (pthread_start_thread_event): Correct computation of self. + Use INIT_THREAD_SELF. + * pthread.c (__pthread_manager_thread): Initialize p_lock. + (__pthread_initialize_manager): Respect event flags also for creation + of the manager thread. + +1999-11-08 Ulrich Drepper <drepper@cygnus.com> + + * pthread.c (__pthread_initialize_manager): Initialize + __pthread_manager_thread.p_tid. + +1999-11-02 Ulrich Drepper <drepper@cygnus.com> + + * internals.h: Declare __pthread_last_event. + * manager.c: Define __pthread_last_event. + (pthread_handle_create): Set __pthread_last_event. + (pthread_exited): Likewise. + * join.c (pthread_exit): Likewise. + + * Makefile (libpthread-routines): Add events. + * events.c: New file. + * internals.h: Protect against multiple inclusion. + Include thread_dbP.h header. + (struct _pthread_descr_struct): Add new fields p_report_events and + p_eventbuf. + Declare event reporting functions. + * join.c (pthread_exit): Signal event if this is wanted. + * manager.c (__pthread_threads_events): New variable. + (pthread_handle_create): Take new parameters with event information. + Signal TD_CREATE event if wanted. + (__pthread_manager): Adjust pthread_handle_create call. + (pthread_start_thread_event): New function. Block until manager is + finished and then call pthread_start_thread. + (pthread_exited): Signal TD_REAP event if wanted. + +1999-10-26 Ulrich Drepper <drepper@cygnus.com> + + * restart.h (suspend_with_cancellation): Rewrite as a macro. + + * condvar.c (pthread_cond_timedwait_relative): Don't mark as inline. + +1999-10-21 Xavier Leroy <Xavier.Leroy@inria.fr> + + * linuxthreads/pthread.c: For i386, wrap pthread_handle_sigrestart + and pthread_handle_sigcancel with functions that restore + %gs from the signal context. For each signal handling function, + two wrappers are required, one for a non-RT signal and one for + a RT signal. + * linuxthreads/signal.c: For i386, add code to restore %gs + from the signal context in pthread_sighandler and + pthread_sighandler_rt. + +1999-10-09 Andreas Jaeger <aj@suse.de> + + * internals.h: Add __new_sem_post to get prototype in + manager.c; include semaphore.h for needed types. + +1999-10-08 Ulrich Drepper <drepper@cygnus.com> + + * manager.c (__pthread_manager) [REQ_POST]: Use __new_sem_post + directly instead of calling sem_post which should not be necessary + but is faster and might help in some case to work around problems. + +1999-09-25 Ulrich Drepper <drepper@cygnus.com> + + * condvar.c (pthread_cond_timedwait_relative): Never return with + EINTR. Patch by Andreas Schwab. + +1999-09-19 Ulrich Drepper <drepper@cygnus.com> + + * signals.c (sigaction): Correct last patch. Don't select + pthread_sighandler_rt based on the signal number but instead of + the SA_SIGINFO flag. + +1999-09-23 Ulrich Drepper <drepper@cygnus.com> + + * specific.c: Move definitions of struct pthread_key_struct and + destr_function to ... + * internals.h: ...here. + +1999-09-03 Andreas Schwab <schwab@suse.de> + + * ptfork.c (__fork): Renamed from fork and use __libc_fork. Add + fork as weak alias. + (__vfork): New function, alias vfork. + * Versions: Export __fork, vfork, and __vfork in libpthread. + +1999-08-23 Andreas Schwab <schwab@suse.de> + + * signals.c (pthread_sighandler): Add SIGCONTEXT_EXTRA_ARGS to + call to signal handler. + +1999-08-20 Ulrich Drepper <drepper@cygnus.com> + + * pthread.c (__pthread_reset_main_thread): Undo last change. + (__pthread_kill_other_threads_np): Reset signal handlers for the + signals we used in the thread implementation here. + +1999-08-19 Ulrich Drepper <drepper@cygnus.com> + + * pthread.c (__pthread_reset_main_thread): Reset signal handlers + for the signals we used in the thread implementation [PR libc/1234]. + + * Versions: Export __pthread_kill_other_threads_np from libpthread + for GLIBC_2.1.2. + + * signals.c: Pass sigcontext through wrapper to the user function. + +1999-08-01 Ulrich Drepper <drepper@cygnus.com> + + * Versions [ld.so] (GLIBC_2.0): Export __libc_internal_tsd_get and + __libc_internal_tsd_set. + +1999-07-29 Andreas Jaeger <aj@arthur.rhein-neckar.de> + + * manager.c: Remove inclusion of <linux/tasks.h> since it's not + needed anymore. + +1999-07-16 Andreas Jaeger <aj@arthur.rhein-neckar.de> + + * internals.h: Align _pthread_descr_struct to 32 bytes. + Reported by Tim Hockin <thockin@cobaltnet.com>, close PR libc/1206. + +1999-07-09 Ulrich Drepper <drepper@cygnus.com> + + * manager.c (pthread_handle_create): Free mmap region after stack + if clone failed. Patch by Kaz Kylheku <kaz@ashi.FootPrints.net>. + +1999-07-09 Cristian Gafton <gafton@redhat.com> + + * Makefile (libpthread-routines): Add oldsemaphore routine. + * Versions: Add sem_destroy, sem_getvalue, sem_init, sem_post, + sem_trywait, and sem_wait to GLIBC_2.1. + * oldsemaphore.c: New file. + * semaphore.c: Add default_symbol_versions for the changed functions. + (__new_sem_init): Rename from sem_init. + (__new_sem_post): Rename from sem_post. + (__new_sem_wait): Rename from sem_wait. + (__new_sem_trywait): Rename from sem_trywait. + (__new_sem_getvalue): Rename from sem_getvalue. + (__new_sem_destroy): Rename from sem_destroy. + +1999-06-23 Robey Pointer <robey@netscape.com> + + * internals.h: Added p_nextlock entry to separate queueing for a + lock from queueing for a CV (sometimes a thread queues on a lock + to serialize removing itself from a CV queue). + * pthread.c: Added p_nextlock to initializers. + * spinlock.c: Changed to use p_nextlock instead of p_nextwaiting. + +1999-05-23 Andreas Jaeger <aj@arthur.rhein-neckar.de> + + * man/pthread_cond_init.man: Correct example. + Reported by Tomas Berndtsson <tomas@nocrew.org>. + + * linuxthreads.texi (Condition Variables): Likewise. + +1999-05-18 Jakub Jelinek <jj@ultra.linux.cz> + + * sysdeps/sparc/sparc64/pt-machine.h (__compare_and_swap): Use + casx not cas, also successful casx returns the old value in rd + and not the new value. + +1999-05-16 Xavier Leroy <Xavier.Leroy@inria.fr> + + * manager.c: If pthread_create() is given a NULL attribute + and the thread manager runs with a realtime policy, set the + scheduling policy of the newly created thread back to SCHED_OTHER. + * manager.c: If the PTHREAD_INHERIT_SCHED attribute is given, + initialize the schedpolicy field of new_thread->p_start_args + to that of the calling thread. + +1999-04-29 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/sparc/sparc64/pt-machine.h (__compare_and_swap): cas + instruction does not allow memory element to use offset. + +1999-04-28 Ulrich Drepper <drepper@cygnus.com> + + * manager.c (pthread_allocate_stack): Optimize initialization of new + thread descriptor. + + * sysdeps/pthread/bits/libc-lock.h (__libc_lock_define_initialized): + Don't use initializer since it is all zeroes. + (__libc_once_define): Likewise. + +1999-04-16 Andreas Jaeger <aj@arthur.rhein-neckar.de> + + * sysdeps/arm/Implies: Removed since cmpxchg/no-cmpxchg + doesn't exist anymore. + * sysdeps/i386/Implies: Likewise. + * sysdeps/m68k/Implies: Likewise. + * sysdeps/mips/Implies: Likewise. + * sysdeps/powerpc/Implies: Likewise. + * sysdeps/sparc/sparc32/Implies: Likewise. + * sysdeps/sparc/sparc64/Implies: Likewise. + +1999-04-15 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/alpha/bits/semaphore.h: Removed. + * sysdeps/powerpc/bits/semaphore.h: Removed. + * sysdeps/pthread/cmpxchg/bits/semaphore.h: Removed. + * sysdeps/pthread/no-cmpxchg/bits/semaphore.h: Removed. + * Makefile (headers): Remove bits/semaphore.h. + + * semaphore.h: Define _pthread_descr if necessary. + Don't include limits.h. Define SEM_VALUE_MAX directly. + Define SEM_FAILED. + (sem_t): Protect element names with leading __. + Add declarations for sem_close, sem_open, and sem_unlink. + * semaphore.c: Adjust all functions for new element names. + Define sem_close, sem_open, and sem_unlink. + * Versions (libthread): Add sem_close, sem_open, and sem_unlink for + GLIBC_2.1.1. + * sysdeps/pthread/bits/pthreadtypes.h: Define _pthread_descr only if + necessary. + +1999-03-16 H.J. Lu <hjl@gnu.org> + + * specific.c (pthread_key_delete): Check th->p_terminated to see + if the thread is running. + + * Versions (__libc_internal_tsd_get, __libc_internal_tsd_set): + Added to GLIBC_2.0 for libc.so. + +1999-02-12 H.J. Lu <hjl@gnu.org> + + * Versions (__libc_current_sigrtmin, __libc_current_sigrtmax, + __libc_allocate_rtsig): Added to GLIBC_2.1. + + * internals.h (DEFAULT_SIG_RESTART): Removed. + (DEFAULT_SIG_CANCEL): Removed. + + * pthread.c (init_rtsigs, __libc_current_sigrtmin, + __libc_current_sigrtmax, __libc_allocate_rtsig): New functions. + (__pthread_sig_restart, __pthread_sig_cancel, + __pthread_sig_debug): Initialized. + (pthread_initialize): Call init_rtsigs () to initialize + real-time signals. + +1999-02-03 H.J. Lu <hjl@gnu.org> + + * manager.c (__pthread_manager): Do block __pthread_sig_debug. + Don't restart the thread which sent REQ_DEBUG. + (pthread_start_thread): Check if __pthread_sig_debug > 0 + before debugging. + + * pthread.c (__pthread_initialize_manager): Suspend ourself + after sending __pthread_sig_debug to gdb instead of + __pthread_sig_cancel. + +1999-01-24 H.J. Lu <hjl@gnu.org> + + * manager.c (__pthread_manager): Delete __pthread_sig_debug + from mask if __pthread_sig_debug > 0. + (pthread_handle_create): Increment __pthread_handles_num. + + * manager.c (pthread_handle_create): Don't pass CLONE_PTRACE to clone. + * pthread.c (__pthread_initialize_manager): Likewise. + + * pthread.c (pthread_initialize): Use __libc_allocate_rtsig (1) + instead of __libc_allocate_rtsig (2). + (__pthread_initialize_manager): Send __pthread_sig_debug to gdb + instead of __pthread_sig_cancel. + (pthread_handle_sigdebug): Fix comments. + +1999-01-21 Ulrich Drepper <drepper@cygnus.com> + + * manager.c (pthread_allocate_stack): Set + __pthread_nonstandard_stacks if user-specified stack is used. + +1999-01-16 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/unix/sysv/linux/bits/posix_opt.h: Add _LFS_ASYNCHRONOUS_IO, + _LFS_LARGEFILE, _LFS64_LARGEFILE, and _LFS64_STDIO from Unix98. + +1999-01-07 Xavier Leroy <Xavier.Leroy@inria.fr> + + * pthread.c: Use a third signal __pthread_sig_debug distinct + from __pthread_sig_cancel to notify gdb when a thread is + created + * manager.c: Likewise. + * internals.h: Likewise. + * signals.c: The implementation of sigwait(s) assumed that + all signals in s have signal handlers already attached. + This is not required by the standard, so make it work + also if some of the signals have no handlers. + +1999-01-05 Andreas Schwab <schwab@issan.cs.uni-dortmund.de> + + * linuxthreads.texi: Remove pointers from first @node. Move old + @node spec inside comment. + +1998-12-31 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/pthread/bits/stdio-lock.h: Define _IO_lock_lock and + _IO_lock_unlock. + +1998-12-29 Ulrich Drepper <drepper@cygnus.com> + + * semaphore.c (sem_trywait): Don't forget to unlock the semaphore + lock. Patch by Bernd Schmidt <crux@pool.informatik.rwth-aachen.de>. + +1998-12-21 Ulrich Drepper <drepper@cygnus.com> + + * manager.c: Threads now send __pthread_sig_cancel on termination. + Change clone call and signal masks. + * thread.c (pthread_handle_sigrestart): Remove special code for + manager. + (pthread_handle_sigcancel): In manager thread call + __pthread_manager_sighandler. + * sysdeps/i386/pt-machine.h (__compare_and_swap): Add memory clobber. + * sysdeps/i386/i686/pt-machine.h: Likewise. + Patches by Xavier Leroy. + +1998-12-14 Ulrich Drepper <drepper@cygnus.com> + + * spinlock.c (__pthread_unlock): Don't crash if called for an + untaken mutex. Reported by Ruslan V. Brushkoff <rus@Snif.Te.Net.UA>. + + * Examples/ex6.c: Unbuffer stdout and reduce sleep time to reduce + overall runtime. + +1998-12-13 Ulrich Drepper <drepper@cygnus.com> + + * Examples/ex3.c: Wait until all threads are started before + searching for the number to avoid race condition on very fast + systems. + +1998-12-08 Andreas Jaeger <aj@arthur.rhein-neckar.de> + + * sysdeps/pthread/pthread.h: Remove __pthread_setcanceltype + declaration since it's not needed. + + * sysdeps/pthread/pthread.h: Move internal functions to ... + * internals.h: ...here. + +1998-12-02 H.J. Lu <hjl@gnu.org> + + * pthread.c (__pthread_sig_restart): Initiliaze to 0 if + SIGRTMIN is defined. + (__pthread_sig_cancel): Likewise. + +1998-12-01 Andreas Jaeger <aj@arthur.rhein-neckar.de> + + * wrapsyscall.c: Include <sys/mman.h> for msync, + <stdlib.h> for system and <termios.h> for tcdrain prototype. + Correct msync declaration. + +1998-11-29 Roland McGrath <roland@baalperazim.frob.com> + + * sysdeps/pthread/bits/libc-tsd.h (__libc_tsd_define, __libc_tsd_get, + __libc_tsd_set): New macros for new interface. + * no-tsd.c: New file, provide uninitialized defns of + __libc_internal_tsd_get and __libc_internal_tsd_set. + * Makefile (routines): Add no-tsd. + +1998-10-12 Roland McGrath <roland@baalperazim.frob.com> + + * internals.h: Include <bits/libc-tsd.h>, not <bits/libc-lock.h>. + * sysdeps/pthread/bits/libc-lock.h (__libc_internal_tsd_get, + __libc_internal_tsd_set): Move decls to ... + * sysdeps/pthread/bits/libc-tsd.h: New file for __libc_internal_tsd_* + declarations. + + * sysdeps/pthread/bits/libc-lock.h (__libc_internal_tsd_get, + __libc_internal_tsd_set): Make these pointers to functions, not + functions; remove #pragma weak decls for them. + * specific.c (__libc_internal_tsd_get, __libc_internal_tsd_set): + Define static functions and initialized pointers to them. + +1998-11-18 Ulrich Drepper <drepper@cygnus.com> + + * Makefile (CFLAGS-mutex.c): Define as -D__NO_WEAK_PTHREAD_ALIASES. + (CFLAGS-specific.c): Likewise. + (CFLAGS-pthread.c): Likewise. + (CFLAGS-ptfork.c): Likewise. + (CFLAGS-cancel.c): Likewise. + * sysdeps/pthread/bits/libc-lock.h: Don't mark __pthread_* functions + as weak references if __NO_WEAK_PTHREAD_ALIASES is defined. + + * mutex.c (pthread_mutex_init): Define as strong symbol. + (pthread_mutex_destroy): Likewise. + (pthread_mutex_trylock): Likewise. + (pthread_mutex_lock): Likewise. + (pthread_mutex_unlock): Likewise. + (pthread_mutexattr_init): Likewise. + (pthread_mutexattr_destroy): Likewise. + (pthread_once): Likewise. + * ptfork.c (pthread_atfork): Likewise. + * specific.c (pthread_key_create): Likewise. + (pthread_setspecific): Likewise. + (pthread_getspecific): Likewise. + +1998-11-15 Andreas Schwab <schwab@issan.cs.uni-dortmund.de> + + * linuxthreads.texi: Fix punctuation after xref. + +1998-11-10 H.J. Lu <hjl@gnu.org> + + * sysdeps/unix/sysv/linux/bits/local_lim.h: Undefine NR_OPEN + if it is defined in <linux/limits.h>. + +1998-10-29 14:28 Ulrich Drepper <drepper@cygnus.com> + + * spinlock.h (__pthread_trylock): Define inline. + (__pthread_lock): Add extra parameter to declaration. Declare + using internal_function. + (__pthread_unlock): Declare using internal_function. + * spinlock.c (__pthread_lock): Add new parameter. Use it instead + of local variable self. Avoid recomputing self. Define using + internal_function. + (__pthread_trylock): Remove. + (__pthread_unlock): Define using internal_function. + * cancel.c: Adjust for __pthread_lock interface change. Use already + computed self value is possible. + * condvar.c: Likewise. + * join.c: Likewise. + * manager.c: Likewise. + * mutex.c: Likewise. + * pthread.c: Likewise. + * rwlock.c: Likewise. + * semaphore.c: Likewise. + * signals.c: Likewise. + +1998-10-27 13:46 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/pthread/pthread.h (struct _pthread_cleanup_buffer): Prepend + __ to field names of the struct. + * sysdeps/pthread/bits/pthreadtypes.h (struct _pthread_fastlock): + Likewise. + (pthread_attr_t): Likewise. + (pthread_cond_t): Likewise. + (pthread_condattr_t): Likewise. + (pthread_mutex_t): Likewise. + (pthread_mutexattr_t): Likewise. + (pthread_rwlock_t): Likewise. + (pthread_rwlockattr_t): Likewise. + * attr.c: Adjust for pthread.h and pthreadtypes.h change. + * cancel.c: Likewise. + * condvar.c: Likewise. + * manager.c: Likewise. + * mutex.c: Likewise. + * pthread.c: Likewise. + * ptlongjmp.c: Likewise. + * rwlock.c: Likewise. + * spinlock.c: Likewise. + +1998-10-09 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/i386/pt-machine.h (get_eflags, set_eflags): Mark these + also with PT_EI. + + * sysdeps/i386/i686/pt-machine.h: Remove unused inline + definitions. + + * Makefile (libpthread-routines): Add pt-machine. + * pt-machine.c: New file. + * sysdeps/alpha/pt-machine.h: Define PT_EI as extern inline is not + yet defined. Use PT_EI in extern inline definitions. + * sysdeps/arm/pt-machine.h: Likewise. + * sysdeps/i386/pt-machine.h: Likewise. + * sysdeps/i386/i686/pt-machine.h: Likewise. + * sysdeps/m68k/pt-machine.h: Likewise. + * sysdeps/mips/pt-machine.h: Likewise. + * sysdeps/powerpc/pt-machine.h: Likewise. + * sysdeps/sparc/sparc32/pt-machine.h: Likewise. + * sysdeps/sparc/sparc64/pt-machine.h: Likewise. + +1998-10-02 Andreas Jaeger <aj@arthur.rhein-neckar.de> + + * semaphore.h: Include <sys/types.h> so that _pthread_descr + is declared. + +1998-09-15 David S. Miller <davem@pierdol.cobaltmicro.com> + + * sysdeps/sparc/sparc32/pt-machine.h (INIT_THREAD_SELF): Add nr + argument. + * sysdeps/sparc/sparc64/pt-machine.h (INIT_THREAD_SELF): Likewise. + +1998-09-12 14:24 -0400 Zack Weinberg <zack@rabi.phys.columbia.edu> + + * linuxthreads/sysdeps/unix/sysv/linux/bits/sigthread.h: Add + multiple inclusion guard. + +1998-09-02 11:08 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> + + * signals.c (sigaction): Check that sig is less than NSIG to avoid + array index overflow. + +1998-09-06 10:56 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/pthread/semaphore.h: New file. + +1998-09-06 09:08 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/pthread/bits/libc-lock.h (enum __libc_tsd_key_t): Add + _LIBC_TSD_KEY_DL_ERROR. + +1998-08-31 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/i386/i686/pt-machine.h (testandset): Add memory clobber. + * sysdeps/i386/pt-machine.h: Likewise. + Suggested by Roland McGrath. + +1998-08-28 13:58 Ulrich Drepper <drepper@cygnus.com> + + * internals.h: Also define THREAD_GETMEM_NC and THREAD_SETMEM_NC to + access thread data with non-constant offsets. + * specific.c: Use THREAD_GETMEM_NC and THREAD_SETMEM_NC where + necessary. + + * sysdeps/i386/useldt.h: Fix typo. Add THREAD_GETMEM_NC and + THREAD_SETMEM_NC definitions. + + * sysdeps/sparc/sparc32/pt-machine.h: Define THREAD_GETMEM_NC and + THREAD_SETMEM_NC. + * sysdeps/sparc/sparc64/pt-machine.h: Likewise. + +1998-08-26 15:46 Ulrich Drepper <drepper@cygnus.com> + + * internals.h: Define THREAD_GETMEM and THREAD_SETMEM to default if + not already defined. + (struct _pthread_descr_struct): Add p_self and p_nr field. + * manager.c (__pthread_handles): Define second element to point + to manager thread. + (__pthread_handles_num): Initialize to 2. + (__pthread_manager): Use INIT_THREAD_SELF with two arguments. + (pthread_start_thread): Likewise. + (pthread_handle_create): Start search for free slot at entry 2. + Initialize new fields p_self and p_nr. + Call __clone with CLONE_PTRACE if available. + (pthread_free): Call FREE_THREAD_SELF if available. + * pthread.c (__pthread_initial_thread): Initialize new fields. + (__pthread_manager_thread): Likewise. + (__pthread_initialize_manager): Call __clone with CLONE_PTRACE. + + * cancel.c: Use THREAD_GETMEM and THREAD_SETMEM to access the + elements of the thread descriptor. + * condvar.c: Likewise. + * errno.c: Likewise. + * join.c: Likewise. + * manager.c: Likewise. + * pthread.c: Likewise. + * ptlongjmp.c: Likewise. + * semaphore.c: Likewise. + * signals.c: Likewise. + * specific.c: Likewise. + * spinlock.c: Likewise. + + * sysdeps/alpha/pt-machine.h (INIT_THREAD_SELF): Add extra parameter. + + * sysdeps/i386/useldt.h: New file. + * sysdeps/i386/i686/pt-machine.h: Show how to use this file. + + * sysdeps/sparc/sparc32/pt-machine.h: Define THREAD_GETMEM and + THREAD_SETMEM using __thread_self. + * sysdeps/sparc/sparc64/pt-machine.h: Likewise. + +1998-08-24 Geoff Keating <geoffk@ozemail.com.au> + + * spinlock.c (__pthread_lock): Reset p_nextwaiting to NULL if it + turned out that we didn't need to queue after all. + +1998-08-22 Geoff Keating <geoffk@ozemail.com.au> + + * sysdeps/powerpc/pt-machine.h: Remove testandset, it's not used + and wastes space; correct types. + +1998-08-08 11:18 H.J. Lu <hjl@gnu.org> + + * signals.c (sigaction): Handle NULL argument. + +1998-08-04 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/unix/sysv/linux/bits/sigthread.h: Use __sigset_t instead + of sigset_t. + +1998-08-02 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> + + * Makefile (linuxthreads-version): Extract correct number from + Banner. + +1998-07-29 Xavier Leroy <Xavier.Leroy@inria.fr> + + * Banner: Bump version number to 0.8 + * FAQ.html: Many updates, in particular w.r.t. debugging. + * manager.c: Support for non-default stacksize for + LinuxThreads-allocated stacks; + don't use guard pages for stacks with default size, rely on + rlimit(RLIMIT_STACK) instead (it's cheaper). + * attr.c: Likewise. + * cancel.c: Use __pthread_sig_cancel and __pthread_sig_restart + everywhere instead of PTHREAD_SIG_CANCEL and PTHREAD_SIG_RESTART. + * condvar.c: Likewise. + * internals.h: Likewise. + * restart.h: Likewise. + * signals.c: Likewise. + * pthread.c: Likewise; set rlimit(RLIMIT_STACK) as we need it. + +1998-07-23 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> + + * weaks.c: Define pthread_mutexattr_[sg]ettype instead of + __pthread_mutexattr_[sg]ettype. Add more weak aliases. + * Versions: Put __pthread_mutexattr_settype under version + GLIBC_2.0. Don't export __pthread_mutexattr_setkind_np and + __pthread_mutexattr_gettype. + +1998-07-23 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> + + * sysdeps/pthread/bits/libc-lock.h: Make + __pthread_mutexattr_settype weak. Don't make + __pthread_mutexattr_setkind_np weak. + +1998-07-16 10:52 Ulrich Drepper <drepper@cygnus.com> + + * manager.c (pthread_handle_create): Check whether sched_setscheduler + call can succeed here. + + * mutex.c: Define __pthread_mutexattr_settype and make + __pthread_mutexattr_setkind_np an alias. + Likewise for __pthread_mutexattr_gettype. + +1998-07-15 11:00 -0400 Zack Weinberg <zack@rabi.phys.columbia.edu> + + * attr.c (pthread_attr_setschedpolicy): Don't check whether caller + is root. + +1998-07-14 19:38 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/pthread/bits/libc-lock.h: Define __libc_cleanup_end. + +1998-07-11 Andreas Jaeger <aj@arthur.rhein-neckar.de> + + * Examples/ex6.c: Include <unistd.h> for usleep. + +1998-06-13 11:04 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> + + * Examples/ex4.c (main): Use exit, not pthread_exit. + +1998-07-09 13:39 Ulrich Drepper <drepper@cygnus.com> + + * Versions: Add __pthread_mutexattr_gettype and + __pthread_mutexattr_settype. + * lockfile.c: Use __pthread_mutexattr_settype instead of + __pthread_mutexattr_setkind_np. + * mutex.c: Define __pthread_mutexattr_gettype and + __pthread_mutexattr_settype. + * weak.c: Likewise. + * sysdeps/pthread/pthread.h: Declare __pthread_mutexattr_gettype and + __pthread_mutexattr_settype. + * sysdeps/pthread/bits/libc-lock.h (__libc_lock_init_recursive): + Use __pthread_mutexattr_settype. + +1998-07-08 22:26 Ulrich Drepper <drepper@cygnus.com> + + * Versions: Add pthread_mutexattr_gettype, pthread_mutexattr_settype. + * mutex.c: Define weak alias pthread_mutexattr_gettype and + pthread_mutexattr_settype. + * sysdeps/pthread/pthread.h: Declare these functions. + Move pthread_sigmask and pthread_kill declaration in separate header. + * sysdeps/unix/sysv/linux/bits/sigthread.h: New file. + +1998-07-07 15:20 Ulrich Drepper <drepper@cygnus.com> + + * Makefile: Add rules to compile and run tests. + * Examples/ex1.c: Little changes to fix warnings. + * Examples/ex2.c: Likewise. + * Examples/ex3.c: Likewise. + * Examples/ex4.c: Likewise. + * Examples/ex5.c: Likewise. + * Examples/ex6.c: New file. + +1998-07-05 11:54 Ulrich Drepper <drepper@cygnus.com> + + * Versions: Add pthread_attr_init to GLIBC_2.1 version in libc. + +1998-07-01 Andreas Jaeger <aj@arthur.rhein-neckar.de> + + * attr.c: Include <string.h>. + +1998-06-30 11:47 Ulrich Drepper <drepper@cygnus.com> + + * attr.c: Include errno.h. Use memcpy to copy sched_param. + * internals.h: Include limits.h. + * manager.c: Use memcpy to copy sched_param. + * ptfork.c: Include errno.h. + * pthread.c: Likewise. + * semaphore.c: Likewise. + * specific.c: Likewise. + * spinlock.h: Likewise. + * sysdeps/pthread/pthread.h: Include only allowed headers. Move + type definition to ... + * sysdeps/pthread/bits/pthreadtypes.h: ...here. New file. + +1998-06-29 12:34 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/pthread/pthread.h: Use __PMT not __P for function pointers. + + * sysdeps/pthread/pthread.h: Define various PTHREAD_* symbols also + as macros as demanded in POSIX.1, Annex C. + +1998-06-29 12:29 Ulrich Drepper <drepper@cygnus.com> + + * internals.h (struct pthread_request): For free use pthread_t + instead of pthread_descr. + * join.c (pthread_join): Pass thread_id, not th to manager. + (pthread_detach): Likewise. + * manager.c (__pthread_manager): Except thread ID in FREE_REQ case. + (pthread_exited): Remove detached queue code. + (pthread_handle_free): Expect thread ID parameter and use it to + validate the thread decsriptor. Don't use detached queue. + Patches by Xavier Leroy. + +1998-06-27 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> + + * libpthread.map: Export accept, longjmp, sigaction, siglongjmp, + _IO_flockfile, _IO_ftrylockfile, _IO_funlockfile, + __pthread_atfork, __pthread_key_create, __pthread_once. + * internals.h: Doc fix. + * pthread.c (__pthread_initialize): Define again. + +1998-06-26 Ulrich Drepper <drepper@cygnus.com> + + * manager.c (pthread_exited): If thread is not detached put it on + special list. + (pthread_handle_free): If thread is not on list with living threads + search on list with detached threads. + + * sysdeps/pthread/pthread.h (PTHREAD_RWLOCK_INITIALIZER): Correct + for new definition of pthread_rwlock_t. + + * spinlock.c: Correct test whether to compile + __pthread_compare_and_swap or not. + +1998-06-25 19:27 Ulrich Drepper <drepper@cygnus.com> + + * attr.c: Finish user stack support. Change locking code to be safe + in situations with different priorities. + * cancel.c: Likewise. + * condvar.c: Likewise. + * internals.h: Likewise. + * join.c: Likewise. + * manager.c: Likewise. + * mutex.c: Likewise. + * pthread.c: Likewise. + * ptlongjmp.c: Likewise. + * queue.h: Likewise. + * rwlock.c: Likewise. + * semaphore.c: Likewise. + * semaphore.h: Likewise. + * signals.c: Likewise. + * spinlock.c: Likewise. + * spinlock.h: Likewise. + * sysdeps/pthread/pthread.h: Likewise. + Patches by Xavier Leroy. + + * sysdeps/i386/i686/pt-machine.h: New file. + +1998-06-25 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/pthread/pthread.h: Make [sg]et_stacksize and + [sg]et_stackaddr prototypes always available. + + * sysdeps/unix/sysv/linux/bits/posix_opt.h: Define + _POSIX_THREAD_ATTR_STACKSIZE and _POSIX_THREAD_ATTR_STACKADDR. + +1998-06-24 Ulrich Drepper <drepper@cygnus.com> + + * manager.c (pthread_free): Undo patch from 980430. + Reported by David Wragg <dpw@doc.ic.ac.uk>. + +1998-06-09 15:07 Ulrich Drepper <drepper@cygnus.com> + + * manager.c: Define __pthread_manager_adjust_prio and use it to + increase priority when needed. + * internals.h: Add prototype for __pthread_manager_adjust_prio. + * mutex.c: Optimize mutexes to wake up only one thread. + * pthread.c: Move PID of manager for global variable in structure + element. + Patches by Xavier Leroy. + +1998-06-07 13:47 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/pthread/bits/libc-lock.h: Optimize cleanup handlers a bit. + +1998-06-03 Andreas Jaeger <aj@arthur.rhein-neckar.de> + + * attr.c: Correct typo. + +1998-05-01 Ulrich Drepper <drepper@cygnus.com> + + * manager.c (pthread_free): Unmap guard before the stack. + Patch by Matthias Urlichs. + +1998-04-30 Ulrich Drepper <drepper@cygnus.com> + + * manager.c (pthread_free): Detect already free child. + Patch by Xavier Leroy, reported by Matthias Urlichs. + +1998-04-23 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> + + * Makefile (linuxthreads-version): Renamed back from + libpthread-version. + +1998-04-21 Ulrich Drepper <drepper@cygnus.com> + + * ptlongjmp.c: Add prototypes for __libc_siglongjmp and + __libc_longjmp. + +1998-04-20 14:55 Ulrich Drepper <drepper@cygnus.com> + + * Makefile (libpthread-routines): Add ptlongjmp and spinlock. + * internals.h: Add definitions for new spinlock implementation. + * ptlongjmp.c: New file. + * spinlock.c: New file. + * spinlock.h (acquire): Don't reschedule using __sched_yield, use + new function __pthread_acquire to prevent deadlocks with thread + with different priorities. + Patches by Xavier Leroy <Xavier.Leroy@inria.fr>. + +1998-03-16 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> + + * manager.c (__pthread_manager): Reduce first argument to select + to include just the needed file descriptor. + +1998-03-17 00:06 Ulrich Drepper <drepper@cygnus.com> + + * manager.c: Fix last patch which caused core dumps. + + * pthread.c: Correctly handle missing SIGRTMIN. + +1998-03-15 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> + + * libpthread.map: Add __libc_internal_tsd_get and + __libc_internal_tsd_set. Add missing cancelable functions. Export + libc internal versions of the cancelable functions. + +1998-03-13 16:51 Ulrich Drepper <drepper@cygnus.com> + + * weaks.c: Define pthread_attr_init as GLIBC_2.0 and GLIBC_2.1. + +1998-03-13 00:46 Ulrich Drepper <drepper@cygnus.com> + + * attr.c: Implement pthread_attr_[gs]etguardsize, + pthread_attr_[gs]setstackaddr, pthread_attr_[gs]etstacksize. + Change pthread_attr_init to have two interfaces. + * internals.h (struct _pthread_descr_struct): Add new fields for + above functions. + * libpthread.map: Add names in GLIBC_2.1 section. + * manager.c (pthread_handle_create): Implement guardsize and + user stack. + (pthread_free): Likewise. + * pthread.c (pthread_create): Add new interface for changed + pthread_attr_t. + * sysdeps/pthread/pthread.h: Add prototypes for new functions. + * sysdeps/unix/sysv/linux/bits/local_lim.h: Add definition of + PTHREAD_STACK_MIN. + +1998-03-11 00:42 Wolfram Gloger <wmglo@dent.med.uni-muenchen.de> + + * manager.c: Enable resetting of the thread scheduling policy + to SCHED_OTHER when the parent thread has a different one. + +1998-02-01 13:51 Ulrich Drepper <drepper@cygnus.com> + + * sysdeps/unix/sysv/linux/bits/posix_opt.h: Define + _POSIX_ASYNCHRONOUS_IO. + + * sysdeps/pthread/pthread.h: Define bits for Unix98 variants of + mutexes. + * mutex.c: Implement new mutex types. + + * internals.h: Include <signal.h>. + + * libpthread.map: Add __erno_location and __h_errno_location. + + * errno.c: Return pointer to variable actually in use. This might + not be the one in the thread structure. + * internals.h (struct _pthread_descr_struct): Add new fields p_errnop + and p_h_errnop. + * manager.c (__pthread_manager): Set p_errnop and p_h_errnop member + of manager thread structure. + (pthread_handle_create): Set p_errnop and p_h_errnop members for new + thread. + * pthread.c: Adapt initializer for thread structures. + (__pthread_initial_thread): Set p_errnop and p_h_errnop member. + (__pthread_reset_main_thread): Reset p_errnop and p_h_errnop of + current thread to global variables. + +1998-01-31 17:27 Ulrich Drepper <drepper@cygnus.com> + + * rwlock.c: New file. + * Makefile (libpthread-routines): Add rwlock. + * sysdeps/pthread/pthread.h: Define data structures and declare + functions. + * libpthread.map: Add new functions. + +1997-12-18 13:50 Philip Blundell <pb@nexus.co.uk> + + * sysdeps/arm/pt-machine.h: New file; add ARM support. + * sysdeps/arm/Implies: likewise. + * README: Document it. + +1997-12-13 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> + + * signals.c: Remove unneeded initializer for sigwaited, saving a + warning. + +1997-04-11 01:18 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> + + * semaphore.c (sem_init): Set sem_spinlock only if available. + +1997-12-04 01:48 Ulrich Drepper <drepper@cygnus.com> + + * mutex.c: Implement PTHREAD_MUTEX_CHECKERROR. + * sysdeps/pthread/pthread.h: Define PTHREAD_MUTEX_CHECKERROR. + + * Makefile: Update from LinuxThreads 0.7. + * internals.h. Likewise. + * manager.c: Likewise. + * mutex.c: Likewise. + * pthread.c: Likewise. + * signals.c: Likewise. + * specific.c: Likewise. + * Examples/ex3.c: Likewise. + +1997-11-20 18:13 Ulrich Drepper <drepper@cygnus.com> + + * pthread.c (__pthread_reset_main_thread): Close pipe only if still + open. + +1997-10-29 05:38 Ulrich Drepper <drepper@cygnus.com> + + * wrapsyscall.c: Add socket functions which are also cancelation + points. + +1997-10-19 21:40 Wolfram Gloger <wg@wolfram.dent.med.uni-muenchen.de> + + * specific.c (__libc_internal_tsd_set, __libc_internal_tsd_get): + New functions for fast thread specific data within libc. + + * internals.h: Add new array p_libc_specific to struct + _pthread_descr_struct. + + * sysdeps/pthread/bits/libc-lock.h: Declare new functions. + +1997-10-13 05:39 Ulrich Drepper <drepper@cygnus.com> + + * semaphore.h: Add __BEGIN_DECLS/__END_DECLS. + Reported by Ralf Corsepius <corsepiu@faw.uni-ulm.de>. + +1997-08-29 03:05 Ulrich Drepper <drepper@cygnus.com> + + * internals.h (struct _pthread_descr_struct): Add definitions for + two-level specific key handling. + * manager.c (pthread_handle_create): Initialize specific memory array. + * specific.c: Implement two-level key handling. + * weaks.c: Don't provide dummy key handling. + * sysdeps/pthread/bits/libc-lock.h: Typedef __libc_lock_t (no #define). + Add definition of __libc_key_t. + * sysdeps/unix/sysv/linux/bits/local_lim.h: Define PTHREAD_KEYS_MAX + as 1024. + Add definition of _POSIX_THREAD_DESTRUCTOR_ITERATIONS and + PTHREAD_DESTRUCTOR_ITERATIONS. + + * manager.c (pthread_handle_create): Compare mmap result with + MAP_FAILED. + + * ptfork.c: Rename to __pthread_atfork and make old name a weak alias. + * sysdeps/pthread/bits/pthread.h: Add prototype for __pthread_atfork. + +1997-08-22 19:04 Richard Henderson <rth@cygnus.com> + + sysdeps/sparc -> sysdeps/sparc/sparc32 + sysdeps/sparc64 -> sysdeps/sparc/sparc64 + + * internals.h: Change definition of THREAD_SELF to be an expression, + not a statement that did a return. + * sysdeps/alpha/pt-machine.h (THREAD_SELF): Update accordingly. + * sysdeps/sparc/sparc32/pt-machine.h (THREAD_SELF, INIT_THREAD_SELF): + Follow Solaris and use a "system reserved" register (%g6) to hold + the thread descriptor. + * sysdeps/sparc/sparc64/pt-machine.h: Likewise. + +1997-08-03 00:09 Ulrich Drepper <drepper@cygnus.com> + + * mutex.c: Correct pthread_once. Patch by Xavier Leroy. + * sysdeps/pthread/pthread.h: Add prototype for __pthread_once. + * sysdeps/pthread/bits/pthread.h: Add macros for __libc_once. + + * semaphore.c: Include spinlock.h only when needed. + + * specific.c (__pthread_setsepcific, __pthread_getspecific): Reject + keys for entries not in use. + + * weaks.c: Implement key handling functions for real. + +1997-06-29 01:04 Richard Henderson <richard@gnu.ai.mit.edu> + + Initial sparc64-linux support: + * linuxthreads/sysdeps/sparc64/Implies: New file. + * linuxthreads/sysdeps/sparc64/pt-machine.h: Likewise. + +1997-06-29 00:48 Ulrich Drepper <drepper@cygnus.com> + + * semaphore.c: Include spinlock.h at correct place. + Patch by HJ Lu. + +1997-06-13 10:06 Richard Henderson <rth@tamu.edu> + + The Great Bit File Move: + * sysdeps/alpha/semaphorebits.h: -> .../bits/semaphore.h. + * sysdeps/powerpc/semaphorebits.h: Likewise. + * sysdeps/pthread/cmpxchg/semaphorebits.h: Likewise. + * sysdeps/pthread/no-cmpxchg/semaphorebits.h: Likewise. + * sysdeps/pthread/libc-lock.h: -> bits/ + * sysdeps/pthread/stdio-lock.h: Likewise. + * sysdeps/unix/sysv/linux/local_lim.h: Likewise. + * sysdeps/unix/sysv/linux/posix_opt.h: Likewise. + * semaphore.h: Likewise. + * sysdeps/pthread/pthread.h: Likewise. + + * lockfile.c: <foo.h> -> <bits/foo.h>. + * semaphore.h: Likewise. + + * Makefile: (headers): foo.h -> bits/foo.h. + * sysdeps/pthread/Makefile: Likewise. + +1997-04-11 01:18 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> + + * semaphore.c (sem_init): Set sem_spinlock only if available. + + * sysdeps/m68k/pt-machine.h (testandset, __compare_and_swap): Fix + asm constraints. + +1997-04-09 03:00 Ulrich Drepper <drepper@cygnus.com> + + Update from LinuxThreads 0.6. + + * attr.c (pthread_attr_getdetachstate): Use __sched_get_priority_max + and __sched_get_priority_min instead of names without `__'. + + * manager.c: Rewrite large parts to implement opaque pthread_t. + + * cancel.c: Adapt for opaque pthread_t type. + * condvar.c: Likewise. + * errno.c: Likewise. + * join.c: Likewise. + * mutex.c: Likewise. + * pthread.c: Likewise. + * signals.c: Likewise. + * specific.c: Likewise. + * restart.h: Likewise. + * queue.h: Likewise. + * Examples/ex3.c: Likewise. + * Examples/ex4.c: Likewise. + * sysdeps/pthread/pthread.h: Likewise. + + * pthread.c: Accumulate time for all threads in thread manager. + + * semaphore.c: Implement fallback implementation for architectures + sometimes missing compare-exchange operations. + + * cancel.c (pthread_cancel): Validate handle argument. + * join.c (pthread_join): Likewise. + (pthread_detach): Likewise. + * signals.c (pthread_kill): Likewise. + + * spinlock.h (acquire): Use __sched_yield not sched_yield. + + * queue.h (enqueue): Enqueue thread according to priority. + + * internals.c (struct pthread_start_args): New struct for passing + args to cloning function. + (struct _pthread): Rename to _pthread_descr_struct and adapt for + opaque pthread_t. + + * Examples/Makefile (clean): Pass -f option to rm. + + * sysdeps/i386/pt-machine.h: Add check for compare-exchange instruction + and define TEST_FOR_COMPARE_AND_SWAP. + * sysdeps/i386/i486/pt-machine.h: Removed. + + * sysdeps/unix/sysv/linux/local_lim.h (PTHREAD_THREADS_MAX): Increase + to 1024. + +1997-04-04 16:38 Ulrich Drepper <drepper@cygnus.com> + + * restart.h (suspend): Clear p_signal before suspending. + (suspend_with_cancellation): Likewise. + Patch by Xavier Leroy <Xavier.Leroy@inria.fr>. + + * weaks.c: Make __pthread_key_create return 1. + * sysdeps/pthread/libc-lock.h: Define __libc_key_create, + __libc_getspecific, __libc_setspecific, and __libc_key_t. + * sysdeps/pthread/stdio-lock.h: Don't care for implementation not + using libio. + +1997-03-19 15:13 Miguel de Icaza <miguel@nuclecu.unam.mx> + + * sysdeps/sparc/pt-machine (RELEASE): Fix. + +1997-03-01 07:55 Geoff Keating <geoffk@ozemail.com.au> + + * sysdeps/powerpc/Implies: Added. + * sysdeps/powerpc/pt-machine.h: Added. + * sysdeps/powerpc/semaphorebits.h: Added. + +1997-01-22 01:22 Ulrich Drepper <drepper@cygnus.com> + + * linuxtheads/pthread.c (__pthread_initial_thread): Correct + initializer. + (__pthread_manager_thread): Likewise. + Reported by Andreas Jaeger. + +1997-01-18 22:15 Richard Henderson <rth@tamu.edu> + + Since sigset_t no longer fits in a register, we can't pass in the + thread's initial mask so easily. Take this opportunity to simplify + the clone implementation by only accepting a single void* argument. + + * linuxthreads/manager.c (__pthread_manager): Put thread vitals + in the thread struct instead of as arguments through clone. + (pthread_start_thread): Look for them there. + * linuxthreads/internals.h (struct _pthread): Add p_initial_fn, + p_initial_fn_arg, p_initial_mask. Fix __pthread_manager proto. + * linuxthreads/pthread.c (pthread_initialize_manager): Revise + clone invocation. diff --git a/libpthread/linuxthreads/Changes b/libpthread/linuxthreads/Changes new file mode 100644 index 000000000..8ec26c9a6 --- /dev/null +++ b/libpthread/linuxthreads/Changes @@ -0,0 +1,73 @@ +Release 0.7: +- Destructors for thread-specific data now conform to the POSIX semantics + (call destructors again if non-NULL TSD remains after a round of + destruction). +- Implemented thread-specific data as a sparse array, allows more TSD keys + and smaller thread descriptors (Ulrich Drepper). +- Added "error checking" mutexes. +- Protect against multiple sigwait() on the same signals. +- Simplified implementation of semaphores when compare_and_swap is + not available. +- Fixed bug in fork() where stdin was closed if fork() was called before + the first pthread_create(). +- Fixed bug in the gethostby*_r functions (bad result if null bytes + in addresses). +- Typos in manual pages corrected. +- First cut at a PowerPC port (not working yet, runs into problems + with gcc and with the C library). + +Release 0.6: +- Validation of thread identifiers: no more crashes when operating on + a thread that has exited (based on Pavel Krauz's ideas). +- Added fallback implementation of semaphores for the 386 and the + Sparc. +- Fixed a bug in signal handling causing false restarts of suspended + threads. +- Fixed a bug in realtime scheduling causing all threads to have + default scheduling on Ix86 with libc5. +- With realtime scheduling, unlocking a mutex now restarts the + highest priority thread waiting on the mutex, not the + first-suspended thread (Richard Neitzel). +- Timing a process now returns cumulative times for all threads, not + just times for the initial thread (suggested by Wolfram Gloger). +- Cleaned up name space (internal defs prefixed by __, weak aliases + for non-portable extensions). +- MIPS port (contributed by Ralf Baechle). + +Release 0.5: +- Signal-safe semaphores a la POSIX 1003.1b added. +- Locking bug in pthread_mutex_trylock over recursive mutexes fixed. +- Race conditions in thread cancellation fixed. +- Sparc port (contributed by Miguel de Icaza). +- Support for getpwnam_r and getpwuid_r. +- Added pthread_kill_other_threads_np to be used in conjunction with + exec*(). + +Release 0.4: +- Manual pages for all functions. +- Synchronization bug causing accumulation of zombie processes fixed. +- Race condition in pthread_cond_timedwait fixed. +- Recursive mutexes are back by popular demand. +- Partial support for realtime scheduling (initiated by Richard Neitzel). +- pthread.h cleaned up a lot: now C++ compatible, added missing "const" + qualifiers, added short documentation, put to GNU libc standards + for name space pollution (Ulrich Drepper). +- Motorola 68k port (contributed by Andreas Schwab). +- Interaction with fork(2) cleaned up a lot. + +Release 0.3: +- Thread creation and reclaimation now performed by a centralized + "thread manager" thread. +- Removed recursive mutexes to make regular mutexes more efficient. +- Now available as a shared library (contributed by Richard Henderson). +- Alpha port (contributed by Richard Henderson). +- Fixed many small discrepancies with Posix 1003.1c. +- Put under the LGPL instead of the GPL. + +Release 0.2: +- Reentrant libc functions (adapted from libc 5.3.9 by Peeter Joot) +- pthread_cond_wait did not reacquire the mutex correctly on return +- More efficient pthread_cond_broadcast + +Release 0.1: +- First public release diff --git a/libpthread/linuxthreads/FAQ.html b/libpthread/linuxthreads/FAQ.html new file mode 100644 index 000000000..21be33ec4 --- /dev/null +++ b/libpthread/linuxthreads/FAQ.html @@ -0,0 +1,1039 @@ +<HTML> +<HEAD> +<TITLE>LinuxThreads Frequently Asked Questions</TITLE> +</HEAD> +<BODY> +<H1 ALIGN=center>LinuxThreads Frequently Asked Questions <BR> + (with answers)</H1> +<H2 ALIGN=center>[For LinuxThreads version 0.8]</H2> + +<HR><P> + +<A HREF="#A">A. The big picture</A><BR> +<A HREF="#B">B. Getting more information</A><BR> +<A HREF="#C">C. Issues related to the C library</A><BR> +<A HREF="#D">D. Problems, weird behaviors, potential bugs</A><BR> +<A HREF="#E">E. Missing functions, wrong types, etc</A><BR> +<A HREF="#F">F. C++ issues</A><BR> +<A HREF="#G">G. Debugging LinuxThreads programs</A><BR> +<A HREF="#H">H. Compiling multithreaded code; errno madness</A><BR> +<A HREF="#I">I. X-Windows and other libraries</A><BR> +<A HREF="#J">J. Signals and threads</A><BR> +<A HREF="#K">K. Internals of LinuxThreads</A><P> + +<HR> +<P> + +<H2><A NAME="A">A. The big picture</A></H2> + +<H4><A NAME="A.1">A.1: What is LinuxThreads?</A></H4> + +LinuxThreads is a Linux library for multi-threaded programming. +It implements the Posix 1003.1c API (Application Programming +Interface) for threads. It runs on any Linux system with kernel 2.0.0 +or more recent, and a suitable C library (see section <A HREF="C">C</A>). +<P> + +<H4><A NAME="A.2">A.2: What are threads?</A></H4> + +A thread is a sequential flow of control through a program. +Multi-threaded programming is, thus, a form of parallel programming +where several threads of control are executing concurrently in the +program. All threads execute in the same memory space, and can +therefore work concurrently on shared data.<P> + +Multi-threaded programming differs from Unix-style multi-processing in +that all threads share the same memory space (and a few other system +resources, such as file descriptors), instead of running in their own +memory space as is the case with Unix processes.<P> + +Threads are useful for two reasons. First, they allow a program to +exploit multi-processor machines: the threads can run in parallel on +several processors, allowing a single program to divide its work +between several processors, thus running faster than a single-threaded +program, which runs on only one processor at a time. Second, some +programs are best expressed as several threads of control that +communicate together, rather than as one big monolithic sequential +program. Examples include server programs, overlapping asynchronous +I/O, and graphical user interfaces.<P> + +<H4><A NAME="A.3">A.3: What is POSIX 1003.1c?</A></H4> + +It's an API for multi-threaded programming standardized by IEEE as +part of the POSIX standards. Most Unix vendors have endorsed the +POSIX 1003.1c standard. Implementations of the 1003.1c API are +already available under Sun Solaris 2.5, Digital Unix 4.0, +Silicon Graphics IRIX 6, and should soon be available from other +vendors such as IBM and HP. More generally, the 1003.1c API is +replacing relatively quickly the proprietary threads library that were +developed previously under Unix, such as Mach cthreads, Solaris +threads, and IRIX sprocs. Thus, multithreaded programs using the +1003.1c API are likely to run unchanged on a wide variety of Unix +platforms.<P> + +<H4><A NAME="A.4">A.4: What is the status of LinuxThreads?</A></H4> + +LinuxThreads implements almost all of Posix 1003.1c, as well as a few +extensions. The only part of LinuxThreads that does not conform yet +to Posix is signal handling (see section <A HREF="#J">J</A>). Apart +from the signal stuff, all the Posix 1003.1c base functionality, +as well as a number of optional extensions, are provided and conform +to the standard (to the best of my knowledge). +The signal stuff is hard to get right, at least without special kernel +support, and while I'm definitely looking at ways to implement the +Posix behavior for signals, this might take a long time before it's +completed.<P> + +<H4><A NAME="A.5">A.5: How stable is LinuxThreads?</A></H4> + +The basic functionality (thread creation and termination, mutexes, +conditions, semaphores) is very stable. Several industrial-strength +programs, such as the AOL multithreaded Web server, use LinuxThreads +and seem quite happy about it. There used to be some rough edges in +the LinuxThreads / C library interface with libc 5, but glibc 2 +fixes all of those problems and is now the standard C library on major +Linux distributions (see section <A HREF="#C">C</A>). <P> + +<HR> +<P> + +<H2><A NAME="B">B. Getting more information</A></H2> + +<H4><A NAME="B.1">B.1: What are good books and other sources of +information on POSIX threads?</A></H4> + +The FAQ for comp.programming.threads lists several books: +<A HREF="http://www.serpentine.com/~bos/threads-faq/">http://www.serpentine.com/~bos/threads-faq/</A>.<P> + +There are also some online tutorials. Follow the links from the +LinuxThreads web page: +<A HREF="http://pauillac.inria.fr/~xleroy/linuxthreads">http://pauillac.inria.fr/~xleroy/linuxthreads</A>.<P> + +<H4><A NAME="B.2">B.2: I'd like to be informed of future developments on +LinuxThreads. Is there a mailing list for this purpose?</A></H4> + +I post LinuxThreads-related announcements on the newsgroup +<A HREF="news:comp.os.linux.announce">comp.os.linux.announce</A>, +and also on the mailing list +<code>linux-threads@magenet.com</code>. +You can subscribe to the latter by writing +<A HREF="mailto:majordomo@magenet.com">majordomo@magenet.com</A>.<P> + +<H4><A NAME="B.3">B.3: What are good places for discussing +LinuxThreads?</A></H4> + +For questions about programming with POSIX threads in general, use +the newsgroup +<A HREF="news:comp.programming.threads">comp.programming.threads</A>. +Be sure you read the +<A HREF="http://www.serpentine.com/~bos/threads-faq/">FAQ</A> +for this group before you post.<P> + +For Linux-specific questions, use +<A +HREF="news:comp.os.linux.development.apps">comp.os.linux.development.apps</A> +and <A +HREF="news:comp.os.linux.development.kernel">comp.os.linux.development.kernel</A>. +The latter is especially appropriate for questions relative to the +interface between the kernel and LinuxThreads.<P> + +<H4><A NAME="B.4">B.4: How should I report a possible bug in +LinuxThreads?</A></H4> + +If you're using glibc 2, the best way by far is to use the +<code>glibcbug</code> script to mail a bug report to the glibc +maintainers. <P> + +If you're using an older libc, or don't have the <code>glibcbug</code> +script on your machine, then e-mail me directly +(<code>Xavier.Leroy@inria.fr</code>). <P> + +In both cases, before sending the bug report, make sure that it is not +addressed already in this FAQ. Also, try to send a short program that +reproduces the weird behavior you observed. <P> + +<H4><A NAME="B.5">B.5: I'd like to read the POSIX 1003.1c standard. Is +it available online?</A></H4> + +Unfortunately, no. POSIX standards are copyrighted by IEEE, and +IEEE does not distribute them freely. You can buy paper copies from +IEEE, but the price is fairly high ($120 or so). If you disagree with +this policy and you're an IEEE member, be sure to let them know.<P> + +On the other hand, you probably don't want to read the standard. It's +very hard to read, written in standard-ese, and targeted to +implementors who already know threads inside-out. A good book on +POSIX threads provides the same information in a much more readable form. +I can personally recommend Dave Butenhof's book, <CITE>Programming +with POSIX threads</CITE> (Addison-Wesley). Butenhof was part of the +POSIX committee and also designed the Digital Unix implementations of +POSIX threads, and it shows.<P> + +Another good source of information is the X/Open Group Single Unix +specification which is available both +<A HREF="http://www.rdg.opengroup.org/onlinepubs/7908799/index.html">on-line</A> +and as a +<A HREF="http://www.UNIX-systems.org/gosolo2/">book and CD/ROM</A>. +That specification includes pretty much all the POSIX standards, +including 1003.1c, with some extensions and clarifications.<P> + +<HR> +<P> + +<H2><A NAME="C">C. Issues related to the C library</A></H2> + +<H4><A NAME="C.1">C.1: Which version of the C library should I use +with LinuxThreads?</A></H4> + +The best choice by far is glibc 2, a.k.a. libc 6. It offers very good +support for multi-threading, and LinuxThreads has been closely +integrated with glibc 2. The glibc 2 distribution contains the +sources of a specially adapted version of LinuxThreads.<P> + +glibc 2 comes preinstalled as the default C library on several Linux +distributions, such as RedHat 5 and up, and Debian 2. +Those distributions include the version of LinuxThreads matching +glibc 2.<P> + +<H4><A NAME="C.2">C.2: My system has libc 5 preinstalled, not glibc +2. Can I still use LinuxThreads?</H4> + +Yes, but you're likely to run into some problems, as libc 5 only +offers minimal support for threads and contains some bugs that affect +multithreaded programs. <P> + +The versions of libc 5 that work best with LinuxThreads are +libc 5.2.18 on the one hand, and libc 5.4.12 or later on the other hand. +Avoid 5.3.12 and 5.4.7: these have problems with the per-thread errno +variable. <P> + +<H4><A NAME="C.3">C.3: So, should I switch to glibc 2, or stay with a +recent libc 5?</A></H4> + +I'd recommend you switch to glibc 2. Even for single-threaded +programs, glibc 2 is more solid and more standard-conformant than libc +5. And the shortcomings of libc 5 almost preclude any serious +multi-threaded programming.<P> + +Switching an already installed +system from libc 5 to glibc 2 is not completely straightforward. +See the <A HREF="http://sunsite.unc.edu/LDP/HOWTO/Glibc2-HOWTO.html">Glibc2 +HOWTO</A> for more information. Much easier is (re-)installing a +Linux distribution based on glibc 2, such as RedHat 6.<P> + +<H4><A NAME="C.4">C.4: Where can I find glibc 2 and the version of +LinuxThreads that goes with it?</A></H4> + +On <code>prep.ai.mit.edu</code> and its many, many mirrors around the world. +See <A +HREF="http://www.gnu.org/order/ftp.html">http://www.gnu.org/order/ftp.html</A> +for a list of mirrors.<P> + +<H4><A NAME="C.5">C.5: Where can I find libc 5 and the version of +LinuxThreads that goes with it?</A></H4> + +For libc 5, see <A HREF="ftp://sunsite.unc.edu/pub/Linux/devel/GCC/"><code>ftp://sunsite.unc.edu/pub/Linux/devel/GCC/</code></A>.<P> + +For the libc 5 version of LinuxThreads, see +<A HREF="ftp://ftp.inria.fr/INRIA/Projects/cristal/Xavier.Leroy/linuxthreads/">ftp://ftp.inria.fr/INRIA/Projects/cristal/Xavier.Leroy/linuxthreads/</A>.<P> + +<H4><A NAME="C.6">C.6: How can I recompile the glibc 2 version of the +LinuxThreads sources?</A></H4> + +You must transfer the whole glibc sources, then drop the LinuxThreads +sources in the <code>linuxthreads/</code> subdirectory, then recompile +glibc as a whole. There are now too many inter-dependencies between +LinuxThreads and glibc 2 to allow separate re-compilation of LinuxThreads. +<P> + +<H4><A NAME="C.7">C.7: What is the correspondence between LinuxThreads +version numbers, libc version numbers, and RedHat version +numbers?</A></H4> + +Here is a summary. (Information on Linux distributions other than +RedHat are welcome.)<P> + +<TABLE> +<TR><TD>LinuxThreads </TD> <TD>C library</TD> <TD>RedHat</TD></TR> +<TR><TD>0.7, 0.71 (for libc 5)</TD> <TD>libc 5.x</TD> <TD>RH 4.2</TD></TR> +<TR><TD>0.7, 0.71 (for glibc 2)</TD> <TD>glibc 2.0.x</TD> <TD>RH 5.x</TD></TR> +<TR><TD>0.8</TD> <TD>glibc 2.1.1</TD> <TD>RH 6.0</TD></TR> +<TR><TD>0.8</TD> <TD>glibc 2.1.2</TD> <TD>not yet released</TD></TR> +</TABLE> +<P> + +<HR> +<P> + +<H2><A NAME="D">D. Problems, weird behaviors, potential bugs</A></H2> + +<H4><A NAME="D.1">D.1: When I compile LinuxThreads, I run into problems in +file <code>libc_r/dirent.c</code></A></H4> + +You probably mean: +<PRE> + libc_r/dirent.c:94: structure has no member named `dd_lock' +</PRE> +I haven't actually seen this problem, but several users reported it. +My understanding is that something is wrong in the include files of +your Linux installation (<code>/usr/include/*</code>). Make sure +you're using a supported version of the libc 5 library. (See question <A +HREF="#C.2">C.2</A>).<P> + +<H4><A NAME="D.2">D.2: When I compile LinuxThreads, I run into problems with +<CODE>/usr/include/sched.h</CODE>: there are several occurrences of +<CODE>_p</CODE> that the C compiler does not understand</A></H4> + +Yes, <CODE>/usr/include/sched.h</CODE> that comes with libc 5.3.12 is broken. +Replace it with the <code>sched.h</code> file contained in the +LinuxThreads distribution. But really you should not be using libc +5.3.12 with LinuxThreads! (See question <A HREF="#C.2">C.1</A>.)<P> + +<H4><A NAME="D.3">D.3: My program does <CODE>fdopen()</CODE> on a file +descriptor opened on a pipe. When I link it with LinuxThreads, +<CODE>fdopen()</CODE> always returns NULL!</A></H4> + +You're using one of the buggy versions of libc (5.3.12, 5.4.7., etc). +See question <A HREF="#C.1">C.1</A> above.<P> + +<H4><A NAME="D.4">D.4: My program creates a lot of threads, and after +a while <CODE>pthread_create()</CODE> no longer returns!</A></H4> + +This is known bug in the version of LinuxThreads that comes with glibc +2.1.1. An upgrade to 2.1.2 is recommended. <P> + +<H4><A NAME="D.5">D.5: When I'm running a program that creates N +threads, <code>top</code> or <code>ps</code> +display N+2 processes that are running my program. What do all these +processes correspond to?</A></H4> + +Due to the general "one process per thread" model, there's one process +for the initial thread and N processes for the threads it created +using <CODE>pthread_create</CODE>. That leaves one process +unaccounted for. That extra process corresponds to the "thread +manager" thread, a thread created internally by LinuxThreads to handle +thread creation and thread termination. This extra thread is asleep +most of the time. + +<H4><A NAME="D.6">D.6: Scheduling seems to be very unfair when there +is strong contention on a mutex: instead of giving the mutex to each +thread in turn, it seems that it's almost always the same thread that +gets the mutex. Isn't this completely broken behavior?</A></H4> + +That behavior has mostly disappeared in recent releases of +LinuxThreads (version 0.8 and up). It was fairly common in older +releases, though. + +What happens in LinuxThreads 0.7 and before is the following: when a +thread unlocks a mutex, all other threads that were waiting on the +mutex are sent a signal which makes them runnable. However, the +kernel scheduler may or may not restart them immediately. If the +thread that unlocked the mutex tries to lock it again immediately +afterwards, it is likely that it will succeed, because the threads +haven't yet restarted. This results in an apparently very unfair +behavior, when the same thread repeatedly locks and unlocks the mutex, +while other threads can't lock the mutex.<P> + +In LinuxThreads 0.8 and up, <code>pthread_unlock</code> restarts only +one waiting thread, and pre-assign the mutex to that thread. Hence, +if the thread that unlocked the mutex tries to lock it again +immediately, it will block until other waiting threads have had a +chance to lock and unlock the mutex. This results in much fairer +scheduling.<P> + +Notice however that even the old "unfair" behavior is perfectly +acceptable with respect to the POSIX standard: for the default +scheduling policy, POSIX makes no guarantees of fairness, such as "the +thread waiting for the mutex for the longest time always acquires it +first". Properly written multithreaded code avoids that kind of heavy +contention on mutexes, and does not run into fairness problems. If +you need scheduling guarantees, you should consider using the +real-time scheduling policies <code>SCHED_RR</code> and +<code>SCHED_FIFO</code>, which have precisely defined scheduling +behaviors. <P> + +<H4><A NAME="D.7">D.7: I have a simple test program with two threads +that do nothing but <CODE>printf()</CODE> in tight loops, and from the +printout it seems that only one thread is running, the other doesn't +print anything!</A></H4> + +Again, this behavior is characteristic of old releases of LinuxThreads +(0.7 and before); more recent versions (0.8 and up) should not exhibit +this behavior.<P> + +The reason for this behavior is explained in +question <A HREF="#D.6">D.6</A> above: <CODE>printf()</CODE> performs +locking on <CODE>stdout</CODE>, and thus your two threads contend very +heavily for the mutex associated with <CODE>stdout</CODE>. But if you +do some real work between two calls to <CODE>printf()</CODE>, you'll +see that scheduling becomes much smoother.<P> + +<H4><A NAME="D.8">D.8: I've looked at <code><pthread.h></code> +and there seems to be a gross error in the <code>pthread_cleanup_push</code> +macro: it opens a block with <code>{</code> but does not close it! +Surely you forgot a <code>}</code> at the end of the macro, right? +</A></H4> + +Nope. That's the way it should be. The closing brace is provided by +the <code>pthread_cleanup_pop</code> macro. The POSIX standard +requires <code>pthread_cleanup_push</code> and +<code>pthread_cleanup_pop</code> to be used in matching pairs, at the +same level of brace nesting. This allows +<code>pthread_cleanup_push</code> to open a block in order to +stack-allocate some data structure, and +<code>pthread_cleanup_pop</code> to close that block. It's ugly, but +it's the standard way of implementing cleanup handlers.<P> + +<H4><A NAME="D.9">D.9: I tried to use real-time threads and my program +loops like crazy and freezes the whole machine!</A></H4> + +Versions of LinuxThreads prior to 0.8 are susceptible to ``livelocks'' +(one thread loops, consuming 100% of the CPU time) in conjunction with +real-time scheduling. Since real-time threads and processes have +higher priority than normal Linux processes, all other processes on +the machine, including the shell, the X server, etc, cannot run and +the machine appears frozen.<P> + +The problem is fixed in LinuxThreads 0.8.<P> + +<H4><A NAME="D.10">D.10: My application needs to create thousands of +threads, or maybe even more. Can I do this with +LinuxThreads?</A></H4> + +No. You're going to run into several hard limits: +<UL> +<LI>Each thread, from the kernel's standpoint, is one process. Stock +Linux kernels are limited to at most 512 processes for the super-user, +and half this number for regular users. This can be changed by +changing <code>NR_TASKS</code> in <code>include/linux/tasks.h</code> +and recompiling the kernel. On the x86 processors at least, +architectural constraints seem to limit <code>NR_TASKS</code> to 4090 +at most. +<LI>LinuxThreads contains a table of all active threads. This table +has room for 1024 threads at most. To increase this limit, you must +change <code>PTHREAD_THREADS_MAX</code> in the LinuxThreads sources +and recompile. +<LI>By default, each thread reserves 2M of virtual memory space for +its stack. This space is just reserved; actual memory is allocated +for the stack on demand. But still, on a 32-bit processor, the total +virtual memory space available for the stacks is on the order of 1G, +meaning that more than 500 threads will have a hard time fitting in. +You can overcome this limitation by moving to a 64-bit platform, or by +allocating smaller stacks yourself using the <code>setstackaddr</code> +attribute. +<LI>Finally, the Linux kernel contains many algorithms that run in +time proportional to the number of process table entries. Increasing +this number drastically will slow down the kernel operations +noticeably. +</UL> +(Other POSIX threads libraries have similar limitations, by the way.) +For all those reasons, you'd better restructure your application so +that it doesn't need more than, say, 100 threads. For instance, +in the case of a multithreaded server, instead of creating a new +thread for each connection, maintain a fixed-size pool of worker +threads that pick incoming connection requests from a queue.<P> + +<HR> +<P> + +<H2><A NAME="E">E. Missing functions, wrong types, etc</A></H2> + +<H4><A NAME="E.1">E.1: Where is <CODE>pthread_yield()</CODE> ? How +comes LinuxThreads does not implement it?</A></H4> + +Because it's not part of the (final) POSIX 1003.1c standard. +Several drafts of the standard contained <CODE>pthread_yield()</CODE>, +but then the POSIX guys discovered it was redundant with +<CODE>sched_yield()</CODE> and dropped it. So, just use +<CODE>sched_yield()</CODE> instead. + +<H4><A NAME="E.2">E.2: I've found some type errors in +<code><pthread.h></code>. +For instance, the second argument to <CODE>pthread_create()</CODE> +should be a <CODE>pthread_attr_t</CODE>, not a +<CODE>pthread_attr_t *</CODE>. Also, didn't you forget to declare +<CODE>pthread_attr_default</CODE>?</A></H4> + +No, I didn't. What you're describing is draft 4 of the POSIX +standard, which is used in OSF DCE threads. LinuxThreads conforms to the +final standard. Even though the functions have the same names as in +draft 4 and DCE, their calling conventions are slightly different. In +particular, attributes are passed by reference, not by value, and +default attributes are denoted by the NULL pointer. Since draft 4/DCE +will eventually disappear, you'd better port your program to use the +standard interface.<P> + +<H4><A NAME="E.3">E.3: I'm porting an application from Solaris and I +have to rename all thread functions from <code>thr_blah</code> to +<CODE>pthread_blah</CODE>. This is very annoying. Why did you change +all the function names?</A></H4> + +POSIX did it. The <code>thr_*</code> functions correspond to Solaris +threads, an older thread interface that you'll find only under +Solaris. The <CODE>pthread_*</CODE> functions correspond to POSIX +threads, an international standard available for many, many platforms. +Even Solaris 2.5 and later support the POSIX threads interface. So, +do yourself a favor and rewrite your code to use POSIX threads: this +way, it will run unchanged under Linux, Solaris, and quite a lot of +other platforms.<P> + +<H4><A NAME="E.4">E.4: How can I suspend and resume a thread from +another thread? Solaris has the <CODE>thr_suspend()</CODE> and +<CODE>thr_resume()</CODE> functions to do that; why don't you?</A></H4> + +The POSIX standard provides <B>no</B> mechanism by which a thread A can +suspend the execution of another thread B, without cooperation from B. +The only way to implement a suspend/restart mechanism is to have B +check periodically some global variable for a suspend request +and then suspend itself on a condition variable, which another thread +can signal later to restart B.<P> + +Notice that <CODE>thr_suspend()</CODE> is inherently dangerous and +prone to race conditions. For one thing, there is no control on where +the target thread stops: it can very well be stopped in the middle of +a critical section, while holding mutexes. Also, there is no +guarantee on when the target thread will actually stop. For these +reasons, you'd be much better off using mutexes and conditions +instead. The only situations that really require the ability to +suspend a thread are debuggers and some kind of garbage collectors.<P> + +If you really must suspend a thread in LinuxThreads, you can send it a +<CODE>SIGSTOP</CODE> signal with <CODE>pthread_kill</CODE>. Send +<CODE>SIGCONT</CODE> for restarting it. +Beware, this is specific to LinuxThreads and entirely non-portable. +Indeed, a truly conforming POSIX threads implementation will stop all +threads when one thread receives the <CODE>SIGSTOP</CODE> signal! +One day, LinuxThreads will implement that behavior, and the +non-portable hack with <CODE>SIGSTOP</CODE> won't work anymore.<P> + +<H4><A NAME="E.5">E.5: Does LinuxThreads implement +<CODE>pthread_attr_setstacksize()</CODE> and +<CODE>pthread_attr_setstackaddr()</CODE>?</A></H4> + +These optional functions are provided in recent versions of +LinuxThreads (0.8 and up). Earlier releases did not provide these +optional components of the POSIX standard.<P> + +Even if <CODE>pthread_attr_setstacksize()</CODE> and +<CODE>pthread_attr_setstackaddr()</CODE> are now provided, we still +recommend that you do not use them unless you really have strong +reasons for doing so. The default stack allocation strategy for +LinuxThreads is nearly optimal: stacks start small (4k) and +automatically grow on demand to a fairly large limit (2M). +Moreover, there is no portable way to estimate the stack requirements +of a thread, so setting the stack size yourself makes your program +less reliable and non-portable.<P> + +<H4><A NAME="E.6">E.6: LinuxThreads does not support the +<CODE>PTHREAD_SCOPE_PROCESS</CODE> value of the "contentionscope" +attribute. Why? </A></H4> + +With a "one-to-one" model, as in LinuxThreads (one kernel execution +context per thread), there is only one scheduler for all processes and +all threads on the system. So, there is no way to obtain the behavior of +<CODE>PTHREAD_SCOPE_PROCESS</CODE>. + +<H4><A NAME="E.7">E.7: LinuxThreads does not implement process-shared +mutexes, conditions, and semaphores. Why?</A></H4> + +This is another optional component of the POSIX standard. Portable +applications should test <CODE>_POSIX_THREAD_PROCESS_SHARED</CODE> +before using this facility. +<P> +The goal of this extension is to allow different processes (with +different address spaces) to synchronize through mutexes, conditions +or semaphores allocated in shared memory (either SVR4 shared memory +segments or <CODE>mmap()</CODE>ed files). +<P> +The reason why this does not work in LinuxThreads is that mutexes, +conditions, and semaphores are not self-contained: their waiting +queues contain pointers to linked lists of thread descriptors, and +these pointers are meaningful only in one address space. +<P> +Matt Messier and I spent a significant amount of time trying to design a +suitable mechanism for sharing waiting queues between processes. We +came up with several solutions that combined two of the following +three desirable features, but none that combines all three: +<UL> +<LI>allow sharing between processes having different UIDs +<LI>supports cancellation +<LI>supports <CODE>pthread_cond_timedwait</CODE> +</UL> +We concluded that kernel support is required to share mutexes, +conditions and semaphores between processes. That's one place where +Linus Torvalds's intuition that "all we need in the kernel is +<CODE>clone()</CODE>" fails. +<P> +Until suitable kernel support is available, you'd better use +traditional interprocess communications to synchronize different +processes: System V semaphores and message queues, or pipes, or sockets. +<P> + +<HR> +<P> + +<H2><A NAME="F">F. C++ issues</A></H2> + +<H4><A NAME="F.1">F.1: Are there C++ wrappers for LinuxThreads?</A></H4> + +Douglas Schmidt's ACE library contains, among a lot of other +things, C++ wrappers for LinuxThreads and quite a number of other +thread libraries. Check out +<A HREF="http://www.cs.wustl.edu/~schmidt/ACE.html">http://www.cs.wustl.edu/~schmidt/ACE.html</A><P> + +<H4><A NAME="F.2">F.2: I'm trying to use LinuxThreads from a C++ +program, and the compiler complains about the third argument to +<CODE>pthread_create()</CODE> !</A></H4> + +You're probably trying to pass a class member function or some +other C++ thing as third argument to <CODE>pthread_create()</CODE>. +Recall that <CODE>pthread_create()</CODE> is a C function, and it must +be passed a C function as third argument.<P> + +<H4><A NAME="F.3">F.3: I'm trying to use LinuxThreads in conjunction +with libg++, and I'm having all sorts of trouble.</A></H4> + +>From what I understand, thread support in libg++ is completely broken, +especially with respect to locking of iostreams. H.J.Lu wrote: +<BLOCKQUOTE> +If you want to use thread, I can only suggest egcs and glibc. You +can find egcs at +<A HREF="http://www.cygnus.com/egcs">http://www.cygnus.com/egcs</A>. +egcs has libsdtc++, which is MT safe under glibc 2. If you really +want to use the libg++, I have a libg++ add-on for egcs. +</BLOCKQUOTE> +<HR> +<P> + +<H2><A NAME="G">G. Debugging LinuxThreads programs</A></H2> + +<H4><A NAME="G.1">G.1: Can I debug LinuxThreads program using gdb?</A></H4> + +Yes, but not with the stock gdb 4.17. You need a specially patched +version of gdb 4.17 developed by Eric Paire and colleages at The Open +Group, Grenoble. The patches against gdb 4.17 are available at +<A HREF="http://www.gr.opengroup.org/java/jdk/linux/debug.htm"><code>http://www.gr.opengroup.org/java/jdk/linux/debug.htm</code></A>. +Precompiled binaries of the patched gdb are available in RedHat's RPM +format at <A +HREF="http://odin.appliedtheory.com/"><code>http://odin.appliedtheory.com/</code></A>.<P> + +Some Linux distributions provide an already-patched version of gdb; +others don't. For instance, the gdb in RedHat 5.2 is thread-aware, +but apparently not the one in RedHat 6.0. Just ask (politely) the +makers of your Linux distributions to please make sure that they apply +the correct patches to gdb.<P> + +<H4><A NAME="G.2">G.2: Does it work with post-mortem debugging?</A></H4> + +Not very well. Generally, the core file does not correspond to the +thread that crashed. The reason is that the kernel will not dump core +for a process that shares its memory with other processes, such as the +other threads of your program. So, the thread that crashes silently +disappears without generating a core file. Then, all other threads of +your program die on the same signal that killed the crashing thread. +(This is required behavior according to the POSIX standard.) The last +one that dies is no longer sharing its memory with anyone else, so the +kernel generates a core file for that thread. Unfortunately, that's +not the thread you are interested in. + +<H4><A NAME="G.3">G.3: Any other ways to debug multithreaded programs, then?</A></H4> + +Assertions and <CODE>printf()</CODE> are your best friends. Try to debug +sequential parts in a single-threaded program first. Then, put +<CODE>printf()</CODE> statements all over the place to get execution traces. +Also, check invariants often with the <CODE>assert()</CODE> macro. In truth, +there is no other effective way (save for a full formal proof of your +program) to track down concurrency bugs. Debuggers are not really +effective for subtle concurrency problems, because they disrupt +program execution too much.<P> + +<HR> +<P> + +<H2><A NAME="H">H. Compiling multithreaded code; errno madness</A></H2> + +<H4><A NAME="H.1">H.1: You say all multithreaded code must be compiled +with <CODE>_REENTRANT</CODE> defined. What difference does it make?</A></H4> + +It affects include files in three ways: +<UL> +<LI> The include files define prototypes for the reentrant variants of +some of the standard library functions, +e.g. <CODE>gethostbyname_r()</CODE> as a reentrant equivalent to +<CODE>gethostbyname()</CODE>.<P> + +<LI> If <CODE>_REENTRANT</CODE> is defined, some +<code><stdio.h></code> functions are no longer defined as macros, +e.g. <CODE>getc()</CODE> and <CODE>putc()</CODE>. In a multithreaded +program, stdio functions require additional locking, which the macros +don't perform, so we must call functions instead.<P> + +<LI> More importantly, <code><errno.h></code> redefines errno when +<CODE>_REENTRANT</CODE> is +defined, so that errno refers to the thread-specific errno location +rather than the global errno variable. This is achieved by the +following <code>#define</code> in <code><errno.h></code>: +<PRE> + #define errno (*(__errno_location())) +</PRE> +which causes each reference to errno to call the +<CODE>__errno_location()</CODE> function for obtaining the location +where error codes are stored. libc provides a default definition of +<CODE>__errno_location()</CODE> that always returns +<code>&errno</code> (the address of the global errno variable). Thus, +for programs not linked with LinuxThreads, defining +<CODE>_REENTRANT</CODE> makes no difference w.r.t. errno processing. +But LinuxThreads redefines <CODE>__errno_location()</CODE> to return a +location in the thread descriptor reserved for holding the current +value of errno for the calling thread. Thus, each thread operates on +a different errno location. +</UL> +<P> + +<H4><A NAME="H.2">H.2: Why is it so important that each thread has its +own errno variable? </A></H4> + +If all threads were to store error codes in the same, global errno +variable, then the value of errno after a system call or library +function returns would be unpredictable: between the time a system +call stores its error code in the global errno and your code inspects +errno to see which error occurred, another thread might have stored +another error code in the same errno location. <P> + +<H4><A NAME="H.3">H.3: What happens if I link LinuxThreads with code +not compiled with <CODE>-D_REENTRANT</CODE>?</A></H4> + +Lots of trouble. If the code uses <CODE>getc()</CODE> or +<CODE>putc()</CODE>, it will perform I/O without proper interlocking +of the stdio buffers; this can cause lost output, duplicate output, or +just crash other stdio functions. If the code consults errno, it will +get back the wrong error code. The following code fragment is a +typical example: +<PRE> + do { + r = read(fd, buf, n); + if (r == -1) { + if (errno == EINTR) /* an error we can handle */ + continue; + else { /* other errors are fatal */ + perror("read failed"); + exit(100); + } + } + } while (...); +</PRE> +Assume this code is not compiled with <CODE>-D_REENTRANT</CODE>, and +linked with LinuxThreads. At run-time, <CODE>read()</CODE> is +interrupted. Since the C library was compiled with +<CODE>-D_REENTRANT</CODE>, <CODE>read()</CODE> stores its error code +in the location pointed to by <CODE>__errno_location()</CODE>, which +is the thread-local errno variable. Then, the code above sees that +<CODE>read()</CODE> returns -1 and looks up errno. Since +<CODE>_REENTRANT</CODE> is not defined, the reference to errno +accesses the global errno variable, which is most likely 0. Hence the +code concludes that it cannot handle the error and stops.<P> + +<H4><A NAME="H.4">H.4: With LinuxThreads, I can no longer use the signals +<code>SIGUSR1</code> and <code>SIGUSR2</code> in my programs! Why? </A></H4> + +The short answer is: because the Linux kernel you're using does not +support realtime signals. <P> + +LinuxThreads needs two signals for its internal operation. +One is used to suspend and restart threads blocked on mutex, condition +or semaphore operations. The other is used for thread +cancellation.<P> + +On ``old'' kernels (2.0 and early 2.1 kernels), there are only 32 +signals available and the kernel reserves all of them but two: +<code>SIGUSR1</code> and <code>SIGUSR2</code>. So, LinuxThreads has +no choice but use those two signals.<P> + +On recent kernels (2.2 and up), more than 32 signals are provided in +the form of realtime signals. When run on one of those kernels, +LinuxThreads uses two reserved realtime signals for its internal +operation, thus leaving <code>SIGUSR1</code> and <code>SIGUSR2</code> +free for user code. (This works only with glibc, not with libc 5.) <P> + +<H4><A NAME="H.5">H.5: Is the stack of one thread visible from the +other threads? Can I pass a pointer into my stack to other threads? +</A></H4> + +Yes, you can -- if you're very careful. The stacks are indeed visible +from all threads in the system. Some non-POSIX thread libraries seem +to map the stacks for all threads at the same virtual addresses and +change the memory mapping when they switch from one thread to +another. But this is not the case for LinuxThreads, as it would make +context switching between threads more expensive, and at any rate +might not conform to the POSIX standard.<P> + +So, you can take the address of an "auto" variable and pass it to +other threads via shared data structures. However, you need to make +absolutely sure that the function doing this will not return as long +as other threads need to access this address. It's the usual mistake +of returning the address of an "auto" variable, only made much worse +because of concurrency. It's much, much safer to systematically +heap-allocate all shared data structures. <P> + +<HR> +<P> + +<H2><A NAME="I">I. X-Windows and other libraries</A></H2> + +<H4><A NAME="I.1">I.1: My program uses both Xlib and LinuxThreads. +It stops very early with an "Xlib: unknown 0 error" message. What +does this mean? </A></H4> + +That's a prime example of the errno problem described in question <A +HREF="#H.2">H.2</A>. The binaries for Xlib you're using have not been +compiled with <CODE>-D_REENTRANT</CODE>. It happens Xlib contains a +piece of code very much like the one in question <A +HREF="#H.2">H.2</A>. So, your Xlib fetches the error code from the +wrong errno location and concludes that an error it cannot handle +occurred.<P> + +<H4><A NAME="I.2">I.2: So, what can I do to build a multithreaded X +Windows client? </A></H4> + +The best solution is to use X libraries that have been compiled with +multithreading options set. Linux distributions that come with glibc +2 as the main C library generally provide thread-safe X libraries. +At least, that seems to be the case for RedHat 5 and later.<P> + +You can try to recompile yourself the X libraries with multithreading +options set. They contain optional support for multithreading; it's +just that the binaries provided by your Linux distribution were built +without this support. See the file <code>README.Xfree3.3</code> in +the LinuxThreads distribution for patches and info on how to compile +thread-safe X libraries from the Xfree3.3 distribution. The Xfree3.3 +sources are readily available in most Linux distributions, e.g. as a +source RPM for RedHat. Be warned, however, that X Windows is a huge +system, and recompiling even just the libraries takes a lot of time +and disk space.<P> + +Another, less involving solution is to call X functions only from the +main thread of your program. Even if all threads have their own errno +location, the main thread uses the global errno variable for its errno +location. Thus, code not compiled with <code>-D_REENTRANT</code> +still "sees" the right error values if it executes in the main thread +only. <P> + +<H4><A NAME="I.2">This is a lot of work. Don't you have precompiled +thread-safe X libraries that you could distribute?</A></H4> + +No, I don't. Sorry. But consider installing a Linux distribution +that comes with thread-safe X libraries, such as RedHat 6.<P> + +<H4><A NAME="I.3">I.3: Can I use library FOO in a multithreaded +program?</A></H4> + +Most libraries cannot be used "as is" in a multithreaded program. +For one thing, they are not necessarily thread-safe: calling +simultaneously two functions of the library from two threads might not +work, due to internal use of global variables and the like. Second, +the libraries must have been compiled with <CODE>-D_REENTRANT</CODE> to avoid +the errno problems explained in question <A HREF="#H.2">H.2</A>. +<P> + +<H4><A NAME="I.4">I.4: What if I make sure that only one thread calls +functions in these libraries?</A></H4> + +This avoids problems with the library not being thread-safe. But +you're still vulnerable to errno problems. At the very least, a +recompile of the library with <CODE>-D_REENTRANT</CODE> is needed. +<P> + +<H4><A NAME="I.5">I.5: What if I make sure that only the main thread +calls functions in these libraries?</A></H4> + +That might actually work. As explained in question <A HREF="#I.1">I.1</A>, +the main thread uses the global errno variable, and can therefore +execute code not compiled with <CODE>-D_REENTRANT</CODE>.<P> + +<H4><A NAME="I.6">I.6: SVGAlib doesn't work with LinuxThreads. Why? +</A></H4> + +Because both LinuxThreads and SVGAlib use the signals +<code>SIGUSR1</code> and <code>SIGUSR2</code>. See question <A +HREF="#H.4">H.4</A>. +<P> + + +<HR> +<P> + +<H2><A NAME="J">J. Signals and threads</A></H2> + +<H4><A NAME="J.1">J.1: When it comes to signals, what is shared +between threads and what isn't?</A></H4> + +Signal handlers are shared between all threads: when a thread calls +<CODE>sigaction()</CODE>, it sets how the signal is handled not only +for itself, but for all other threads in the program as well.<P> + +On the other hand, signal masks are per-thread: each thread chooses +which signals it blocks independently of others. At thread creation +time, the newly created thread inherits the signal mask of the thread +calling <CODE>pthread_create()</CODE>. But afterwards, the new thread +can modify its signal mask independently of its creator thread.<P> + +<H4><A NAME="J.2">J.2: When I send a <CODE>SIGKILL</CODE> to a +particular thread using <CODE>pthread_kill</CODE>, all my threads are +killed!</A></H4> + +That's how it should be. The POSIX standard mandates that all threads +should terminate when the process (i.e. the collection of all threads +running the program) receives a signal whose effect is to +terminate the process (such as <CODE>SIGKILL</CODE> or <CODE>SIGINT</CODE> +when no handler is installed on that signal). This behavior makes a +lot of sense: when you type "ctrl-C" at the keyboard, or when a thread +crashes on a division by zero or a segmentation fault, you really want +all threads to stop immediately, not just the one that caused the +segmentation violation or that got the <CODE>SIGINT</CODE> signal. +(This assumes default behavior for those signals; see question +<A HREF="#J.3">J.3</A> if you install handlers for those signals.)<P> + +If you're trying to terminate a thread without bringing the whole +process down, use <code>pthread_cancel()</code>.<P> + +<H4><A NAME="J.3">J.3: I've installed a handler on a signal. Which +thread executes the handler when the signal is received?</A></H4> + +If the signal is generated by a thread during its execution (e.g. a +thread executes a division by zero and thus generates a +<CODE>SIGFPE</CODE> signal), then the handler is executed by that +thread. This also applies to signals generated by +<CODE>raise()</CODE>.<P> + +If the signal is sent to a particular thread using +<CODE>pthread_kill()</CODE>, then that thread executes the handler.<P> + +If the signal is sent via <CODE>kill()</CODE> or the tty interface +(e.g. by pressing ctrl-C), then the POSIX specs say that the handler +is executed by any thread in the process that does not currently block +the signal. In other terms, POSIX considers that the signal is sent +to the process (the collection of all threads) as a whole, and any +thread that is not blocking this signal can then handle it.<P> + +The latter case is where LinuxThreads departs from the POSIX specs. +In LinuxThreads, there is no real notion of ``the process as a whole'': +in the kernel, each thread is really a distinct process with a +distinct PID, and signals sent to the PID of a thread can only be +handled by that thread. As long as no thread is blocking the signal, +the behavior conforms to the standard: one (unspecified) thread of the +program handles the signal. But if the thread to which PID the signal +is sent blocks the signal, and some other thread does not block the +signal, then LinuxThreads will simply queue in +that thread and execute the handler only when that thread unblocks +the signal, instead of executing the handler immediately in the other +thread that does not block the signal.<P> + +This is to be viewed as a LinuxThreads bug, but I currently don't see +any way to implement the POSIX behavior without kernel support.<P> + +<H4><A NAME="J.3">J.3: How shall I go about mixing signals and threads +in my program? </A></H4> + +The less you mix them, the better. Notice that all +<CODE>pthread_*</CODE> functions are not async-signal safe, meaning +that you should not call them from signal handlers. This +recommendation is not to be taken lightly: your program can deadlock +if you call a <CODE>pthread_*</CODE> function from a signal handler! +<P> + +The only sensible things you can do from a signal handler is set a +global flag, or call <CODE>sem_post</CODE> on a semaphore, to record +the delivery of the signal. The remainder of the program can then +either poll the global flag, or use <CODE>sem_wait()</CODE> and +<CODE>sem_trywait()</CODE> on the semaphore.<P> + +Another option is to do nothing in the signal handler, and dedicate +one thread (preferably the initial thread) to wait synchronously for +signals, using <CODE>sigwait()</CODE>, and send messages to the other +threads accordingly. + +<H4><A NAME="J.4">J.4: When one thread is blocked in +<CODE>sigwait()</CODE>, other threads no longer receive the signals +<CODE>sigwait()</CODE> is waiting for! What happens? </A></H4> + +It's an unfortunate consequence of how LinuxThreads implements +<CODE>sigwait()</CODE>. Basically, it installs signal handlers on all +signals waited for, in order to record which signal was received. +Since signal handlers are shared with the other threads, this +temporarily deactivates any signal handlers you might have previously +installed on these signals.<P> + +Though surprising, this behavior actually seems to conform to the +POSIX standard. According to POSIX, <CODE>sigwait()</CODE> is +guaranteed to work as expected only if all other threads in the +program block the signals waited for (otherwise, the signals could be +delivered to other threads than the one doing <CODE>sigwait()</CODE>, +which would make <CODE>sigwait()</CODE> useless). In this particular +case, the problem described in this question does not appear.<P> + +One day, <CODE>sigwait()</CODE> will be implemented in the kernel, +along with others POSIX 1003.1b extensions, and <CODE>sigwait()</CODE> +will have a more natural behavior (as well as better performances).<P> + +<HR> +<P> + +<H2><A NAME="K">K. Internals of LinuxThreads</A></H2> + +<H4><A NAME="K.1">K.1: What is the implementation model for +LinuxThreads?</A></H4> + +LinuxThreads follows the so-called "one-to-one" model: each thread is +actually a separate process in the kernel. The kernel scheduler takes +care of scheduling the threads, just like it schedules regular +processes. The threads are created with the Linux +<code>clone()</code> system call, which is a generalization of +<code>fork()</code> allowing the new process to share the memory +space, file descriptors, and signal handlers of the parent.<P> + +Advantages of the "one-to-one" model include: +<UL> +<LI> minimal overhead on CPU-intensive multiprocessing (with +about one thread per processor); +<LI> minimal overhead on I/O operations; +<LI> a simple and robust implementation (the kernel scheduler does +most of the hard work for us). +</UL> +The main disadvantage is more expensive context switches on mutex and +condition operations, which must go through the kernel. This is +mitigated by the fact that context switches in the Linux kernel are +pretty efficient.<P> + +<H4><A NAME="K.2">K.2: Have you considered other implementation +models?</A></H4> + +There are basically two other models. The "many-to-one" model +relies on a user-level scheduler that context-switches between the +threads entirely in user code; viewed from the kernel, there is only +one process running. This model is completely out of the question for +me, since it does not take advantage of multiprocessors, and require +unholy magic to handle blocking I/O operations properly. There are +several user-level thread libraries available for Linux, but I found +all of them deficient in functionality, performance, and/or robustness. +<P> + +The "many-to-many" model combines both kernel-level and user-level +scheduling: several kernel-level threads run concurrently, each +executing a user-level scheduler that selects between user threads. +Most commercial Unix systems (Solaris, Digital Unix, IRIX) implement +POSIX threads this way. This model combines the advantages of both +the "many-to-one" and the "one-to-one" model, and is attractive +because it avoids the worst-case behaviors of both models -- +especially on kernels where context switches are expensive, such as +Digital Unix. Unfortunately, it is pretty complex to implement, and +requires kernel support which Linux does not provide. Linus Torvalds +and other Linux kernel developers have always been pushing the +"one-to-one" model in the name of overall simplicity, and are doing a +pretty good job of making kernel-level context switches between +threads efficient. LinuxThreads is just following the general +direction they set.<P> + +<HR> +<ADDRESS>Xavier.Leroy@inria.fr</ADDRESS> +</BODY> +</HTML> diff --git a/libpthread/linuxthreads/LICENSE b/libpthread/linuxthreads/LICENSE new file mode 100644 index 000000000..7bcca6050 --- /dev/null +++ b/libpthread/linuxthreads/LICENSE @@ -0,0 +1,501 @@ +GNU LIBRARY GENERAL PUBLIC LICENSE +********************************** + + Version 2, June 1991 + + Copyright (C) 1991 Free Software Foundation, Inc. + 59 Temple Place -- Suite 330, Boston, MA 02111-1307, USA + + Everyone is permitted to copy and distribute verbatim copies + of this license document, but changing it is not allowed. + + [This is the first released version of the library GPL. It is + numbered 2 because it goes with version 2 of the ordinary GPL.] + +Preamble +======== + + The licenses for most software are designed to take away your +freedom to share and change it. By contrast, the GNU General Public +Licenses are intended to guarantee your freedom to share and change +free software--to make sure the software is free for all its users. + + This license, the Library General Public License, applies to some +specially designated Free Software Foundation software, and to any +other libraries whose authors decide to use it. You can use it for +your libraries, too. + + When we speak of free software, we are referring to freedom, not +price. Our General Public Licenses are designed to make sure that you +have the freedom to distribute copies of free software (and charge for +this service if you wish), that you receive source code or can get it +if you want it, that you can change the software or use pieces of it in +new free programs; and that you know you can do these things. + + To protect your rights, we need to make restrictions that forbid +anyone to deny you these rights or to ask you to surrender the rights. +These restrictions translate to certain responsibilities for you if you +distribute copies of the library, or if you modify it. + + For example, if you distribute copies of the library, whether gratis +or for a fee, you must give the recipients all the rights that we gave +you. You must make sure that they, too, receive or can get the source +code. If you link a program with the library, you must provide +complete object files to the recipients so that they can relink them +with the library, after making changes to the library and recompiling +it. And you must show them these terms so they know their rights. + + Our method of protecting your rights has two steps: (1) copyright +the library, and (2) offer you this license which gives you legal +permission to copy, distribute and/or modify the library. + + Also, for each distributor's protection, we want to make certain +that everyone understands that there is no warranty for this free +library. If the library is modified by someone else and passed on, we +want its recipients to know that what they have is not the original +version, so that any problems introduced by others will not reflect on +the original authors' reputations. + + Finally, any free program is threatened constantly by software +patents. We wish to avoid the danger that companies distributing free +software will individually obtain patent licenses, thus in effect +transforming the program into proprietary software. To prevent this, +we have made it clear that any patent must be licensed for everyone's +free use or not licensed at all. + + Most GNU software, including some libraries, is covered by the +ordinary GNU General Public License, which was designed for utility +programs. This license, the GNU Library General Public License, +applies to certain designated libraries. This license is quite +different from the ordinary one; be sure to read it in full, and don't +assume that anything in it is the same as in the ordinary license. + + The reason we have a separate public license for some libraries is +that they blur the distinction we usually make between modifying or +adding to a program and simply using it. Linking a program with a +library, without changing the library, is in some sense simply using +the library, and is analogous to running a utility program or +application program. However, in a textual and legal sense, the linked +executable is a combined work, a derivative of the original library, +and the ordinary General Public License treats it as such. + + Because of this blurred distinction, using the ordinary General +Public License for libraries did not effectively promote software +sharing, because most developers did not use the libraries. We +concluded that weaker conditions might promote sharing better. + + However, unrestricted linking of non-free programs would deprive the +users of those programs of all benefit from the free status of the +libraries themselves. This Library General Public License is intended +to permit developers of non-free programs to use free libraries, while +preserving your freedom as a user of such programs to change the free +libraries that are incorporated in them. (We have not seen how to +achieve this as regards changes in header files, but we have achieved +it as regards changes in the actual functions of the Library.) The +hope is that this will lead to faster development of free libraries. + + The precise terms and conditions for copying, distribution and +modification follow. Pay close attention to the difference between a +"work based on the library" and a "work that uses the library". The +former contains code derived from the library, while the latter only +works together with the library. + + Note that it is possible for a library to be covered by the ordinary +General Public License rather than by this special one. + + TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION + + 0. This License Agreement applies to any software library which + contains a notice placed by the copyright holder or other + authorized party saying it may be distributed under the terms of + this Library General Public License (also called "this License"). + Each licensee is addressed as "you". + + A "library" means a collection of software functions and/or data + prepared so as to be conveniently linked with application programs + (which use some of those functions and data) to form executables. + + The "Library", below, refers to any such software library or work + which has been distributed under these terms. A "work based on the + Library" means either the Library or any derivative work under + copyright law: that is to say, a work containing the Library or a + portion of it, either verbatim or with modifications and/or + translated straightforwardly into another language. (Hereinafter, + translation is included without limitation in the term + "modification".) + + "Source code" for a work means the preferred form of the work for + making modifications to it. For a library, complete source code + means all the source code for all modules it contains, plus any + associated interface definition files, plus the scripts used to + control compilation and installation of the library. + + Activities other than copying, distribution and modification are + not covered by this License; they are outside its scope. The act + of running a program using the Library is not restricted, and + output from such a program is covered only if its contents + constitute a work based on the Library (independent of the use of + the Library in a tool for writing it). Whether that is true + depends on what the Library does and what the program that uses + the Library does. + + 1. You may copy and distribute verbatim copies of the Library's + complete source code as you receive it, in any medium, provided + that you conspicuously and appropriately publish on each copy an + appropriate copyright notice and disclaimer of warranty; keep + intact all the notices that refer to this License and to the + absence of any warranty; and distribute a copy of this License + along with the Library. + + You may charge a fee for the physical act of transferring a copy, + and you may at your option offer warranty protection in exchange + for a fee. + + 2. You may modify your copy or copies of the Library or any portion + of it, thus forming a work based on the Library, and copy and + distribute such modifications or work under the terms of Section 1 + above, provided that you also meet all of these conditions: + + a. The modified work must itself be a software library. + + b. You must cause the files modified to carry prominent notices + stating that you changed the files and the date of any change. + + c. You must cause the whole of the work to be licensed at no + charge to all third parties under the terms of this License. + + d. If a facility in the modified Library refers to a function or + a table of data to be supplied by an application program that + uses the facility, other than as an argument passed when the + facility is invoked, then you must make a good faith effort + to ensure that, in the event an application does not supply + such function or table, the facility still operates, and + performs whatever part of its purpose remains meaningful. + + (For example, a function in a library to compute square roots + has a purpose that is entirely well-defined independent of the + application. Therefore, Subsection 2d requires that any + application-supplied function or table used by this function + must be optional: if the application does not supply it, the + square root function must still compute square roots.) + + These requirements apply to the modified work as a whole. If + identifiable sections of that work are not derived from the + Library, and can be reasonably considered independent and separate + works in themselves, then this License, and its terms, do not + apply to those sections when you distribute them as separate + works. But when you distribute the same sections as part of a + whole which is a work based on the Library, the distribution of + the whole must be on the terms of this License, whose permissions + for other licensees extend to the entire whole, and thus to each + and every part regardless of who wrote it. + + Thus, it is not the intent of this section to claim rights or + contest your rights to work written entirely by you; rather, the + intent is to exercise the right to control the distribution of + derivative or collective works based on the Library. + + In addition, mere aggregation of another work not based on the + Library with the Library (or with a work based on the Library) on + a volume of a storage or distribution medium does not bring the + other work under the scope of this License. + + 3. You may opt to apply the terms of the ordinary GNU General Public + License instead of this License to a given copy of the Library. + To do this, you must alter all the notices that refer to this + License, so that they refer to the ordinary GNU General Public + License, version 2, instead of to this License. (If a newer + version than version 2 of the ordinary GNU General Public License + has appeared, then you can specify that version instead if you + wish.) Do not make any other change in these notices. + + Once this change is made in a given copy, it is irreversible for + that copy, so the ordinary GNU General Public License applies to + all subsequent copies and derivative works made from that copy. + + This option is useful when you wish to copy part of the code of + the Library into a program that is not a library. + + 4. You may copy and distribute the Library (or a portion or + derivative of it, under Section 2) in object code or executable + form under the terms of Sections 1 and 2 above provided that you + accompany it with the complete corresponding machine-readable + source code, which must be distributed under the terms of Sections + 1 and 2 above on a medium customarily used for software + interchange. + + If distribution of object code is made by offering access to copy + from a designated place, then offering equivalent access to copy + the source code from the same place satisfies the requirement to + distribute the source code, even though third parties are not + compelled to copy the source along with the object code. + + 5. A program that contains no derivative of any portion of the + Library, but is designed to work with the Library by being + compiled or linked with it, is called a "work that uses the + Library". Such a work, in isolation, is not a derivative work of + the Library, and therefore falls outside the scope of this License. + + However, linking a "work that uses the Library" with the Library + creates an executable that is a derivative of the Library (because + it contains portions of the Library), rather than a "work that + uses the library". The executable is therefore covered by this + License. Section 6 states terms for distribution of such + executables. + + When a "work that uses the Library" uses material from a header + file that is part of the Library, the object code for the work may + be a derivative work of the Library even though the source code is + not. Whether this is true is especially significant if the work + can be linked without the Library, or if the work is itself a + library. The threshold for this to be true is not precisely + defined by law. + + If such an object file uses only numerical parameters, data + structure layouts and accessors, and small macros and small inline + functions (ten lines or less in length), then the use of the object + file is unrestricted, regardless of whether it is legally a + derivative work. (Executables containing this object code plus + portions of the Library will still fall under Section 6.) + + Otherwise, if the work is a derivative of the Library, you may + distribute the object code for the work under the terms of Section + 6. Any executables containing that work also fall under Section 6, + whether or not they are linked directly with the Library itself. + + 6. As an exception to the Sections above, you may also compile or + link a "work that uses the Library" with the Library to produce a + work containing portions of the Library, and distribute that work + under terms of your choice, provided that the terms permit + modification of the work for the customer's own use and reverse + engineering for debugging such modifications. + + You must give prominent notice with each copy of the work that the + Library is used in it and that the Library and its use are covered + by this License. You must supply a copy of this License. If the + work during execution displays copyright notices, you must include + the copyright notice for the Library among them, as well as a + reference directing the user to the copy of this License. Also, + you must do one of these things: + + a. Accompany the work with the complete corresponding + machine-readable source code for the Library including + whatever changes were used in the work (which must be + distributed under Sections 1 and 2 above); and, if the work + is an executable linked with the Library, with the complete + machine-readable "work that uses the Library", as object code + and/or source code, so that the user can modify the Library + and then relink to produce a modified executable containing + the modified Library. (It is understood that the user who + changes the contents of definitions files in the Library will + not necessarily be able to recompile the application to use + the modified definitions.) + + b. Accompany the work with a written offer, valid for at least + three years, to give the same user the materials specified in + Subsection 6a, above, for a charge no more than the cost of + performing this distribution. + + c. If distribution of the work is made by offering access to copy + from a designated place, offer equivalent access to copy the + above specified materials from the same place. + + d. Verify that the user has already received a copy of these + materials or that you have already sent this user a copy. + + For an executable, the required form of the "work that uses the + Library" must include any data and utility programs needed for + reproducing the executable from it. However, as a special + exception, the source code distributed need not include anything + that is normally distributed (in either source or binary form) + with the major components (compiler, kernel, and so on) of the + operating system on which the executable runs, unless that + component itself accompanies the executable. + + It may happen that this requirement contradicts the license + restrictions of other proprietary libraries that do not normally + accompany the operating system. Such a contradiction means you + cannot use both them and the Library together in an executable + that you distribute. + + 7. You may place library facilities that are a work based on the + Library side-by-side in a single library together with other + library facilities not covered by this License, and distribute + such a combined library, provided that the separate distribution + of the work based on the Library and of the other library + facilities is otherwise permitted, and provided that you do these + two things: + + a. Accompany the combined library with a copy of the same work + based on the Library, uncombined with any other library + facilities. This must be distributed under the terms of the + Sections above. + + b. Give prominent notice with the combined library of the fact + that part of it is a work based on the Library, and explaining + where to find the accompanying uncombined form of the same + work. + + 8. You may not copy, modify, sublicense, link with, or distribute the + Library except as expressly provided under this License. Any + attempt otherwise to copy, modify, sublicense, link with, or + distribute the Library is void, and will automatically terminate + your rights under this License. However, parties who have + received copies, or rights, from you under this License will not + have their licenses terminated so long as such parties remain in + full compliance. + + 9. You are not required to accept this License, since you have not + signed it. However, nothing else grants you permission to modify + or distribute the Library or its derivative works. These actions + are prohibited by law if you do not accept this License. + Therefore, by modifying or distributing the Library (or any work + based on the Library), you indicate your acceptance of this + License to do so, and all its terms and conditions for copying, + distributing or modifying the Library or works based on it. + + 10. Each time you redistribute the Library (or any work based on the + Library), the recipient automatically receives a license from the + original licensor to copy, distribute, link with or modify the + Library subject to these terms and conditions. You may not impose + any further restrictions on the recipients' exercise of the rights + granted herein. You are not responsible for enforcing compliance + by third parties to this License. + + 11. If, as a consequence of a court judgment or allegation of patent + infringement or for any other reason (not limited to patent + issues), conditions are imposed on you (whether by court order, + agreement or otherwise) that contradict the conditions of this + License, they do not excuse you from the conditions of this + License. If you cannot distribute so as to satisfy simultaneously + your obligations under this License and any other pertinent + obligations, then as a consequence you may not distribute the + Library at all. For example, if a patent license would not permit + royalty-free redistribution of the Library by all those who + receive copies directly or indirectly through you, then the only + way you could satisfy both it and this License would be to refrain + entirely from distribution of the Library. + + If any portion of this section is held invalid or unenforceable + under any particular circumstance, the balance of the section is + intended to apply, and the section as a whole is intended to apply + in other circumstances. + + It is not the purpose of this section to induce you to infringe any + patents or other property right claims or to contest validity of + any such claims; this section has the sole purpose of protecting + the integrity of the free software distribution system which is + implemented by public license practices. Many people have made + generous contributions to the wide range of software distributed + through that system in reliance on consistent application of that + system; it is up to the author/donor to decide if he or she is + willing to distribute software through any other system and a + licensee cannot impose that choice. + + This section is intended to make thoroughly clear what is believed + to be a consequence of the rest of this License. + + 12. If the distribution and/or use of the Library is restricted in + certain countries either by patents or by copyrighted interfaces, + the original copyright holder who places the Library under this + License may add an explicit geographical distribution limitation + excluding those countries, so that distribution is permitted only + in or among countries not thus excluded. In such case, this + License incorporates the limitation as if written in the body of + this License. + + 13. The Free Software Foundation may publish revised and/or new + versions of the Library General Public License from time to time. + Such new versions will be similar in spirit to the present version, + but may differ in detail to address new problems or concerns. + + Each version is given a distinguishing version number. If the + Library specifies a version number of this License which applies + to it and "any later version", you have the option of following + the terms and conditions either of that version or of any later + version published by the Free Software Foundation. If the Library + does not specify a license version number, you may choose any + version ever published by the Free Software Foundation. + + 14. If you wish to incorporate parts of the Library into other free + programs whose distribution conditions are incompatible with these, + write to the author to ask for permission. For software which is + copyrighted by the Free Software Foundation, write to the Free + Software Foundation; we sometimes make exceptions for this. Our + decision will be guided by the two goals of preserving the free + status of all derivatives of our free software and of promoting + the sharing and reuse of software generally. + + NO WARRANTY + + 15. BECAUSE THE LIBRARY IS LICENSED FREE OF CHARGE, THERE IS NO + WARRANTY FOR THE LIBRARY, TO THE EXTENT PERMITTED BY APPLICABLE + LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT + HOLDERS AND/OR OTHER PARTIES PROVIDE THE LIBRARY "AS IS" WITHOUT + WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT + NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND + FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE + QUALITY AND PERFORMANCE OF THE LIBRARY IS WITH YOU. SHOULD THE + LIBRARY PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY + SERVICING, REPAIR OR CORRECTION. + + 16. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN + WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY + MODIFY AND/OR REDISTRIBUTE THE LIBRARY AS PERMITTED ABOVE, BE + LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, + INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR + INABILITY TO USE THE LIBRARY (INCLUDING BUT NOT LIMITED TO LOSS OF + DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU + OR THIRD PARTIES OR A FAILURE OF THE LIBRARY TO OPERATE WITH ANY + OTHER SOFTWARE), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN + ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. + + END OF TERMS AND CONDITIONS + +How to Apply These Terms to Your New Libraries +============================================== + + If you develop a new library, and you want it to be of the greatest +possible use to the public, we recommend making it free software that +everyone can redistribute and change. You can do so by permitting +redistribution under these terms (or, alternatively, under the terms of +the ordinary General Public License). + + To apply these terms, attach the following notices to the library. +It is safest to attach them to the start of each source file to most +effectively convey the exclusion of warranty; and each file should have +at least the "copyright" line and a pointer to where the full notice is +found. + + ONE LINE TO GIVE THE LIBRARY'S NAME AND AN IDEA OF WHAT IT DOES. + Copyright (C) YEAR NAME OF AUTHOR + + This library is free software; you can redistribute it and/or modify it + under the terms of the GNU Library General Public License as published + by the Free Software Foundation; either version 2 of the License, or (at + your option) any later version. + + This library is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, Inc., + 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA. + + Also add information on how to contact you by electronic and paper +mail. + + You should also get your employer (if you work as a programmer) or +your school, if any, to sign a "copyright disclaimer" for the library, +if necessary. Here is a sample; alter the names: + + Yoyodyne, Inc., hereby disclaims all copyright interest in the library + `Frob' (a library for tweaking knobs) written by James Random Hacker. + + SIGNATURE OF TY COON, 1 April 1990 + Ty Coon, President of Vice + + That's all there is to it! + diff --git a/libpthread/linuxthreads/Makefile b/libpthread/linuxthreads/Makefile new file mode 100644 index 000000000..eba08a25d --- /dev/null +++ b/libpthread/linuxthreads/Makefile @@ -0,0 +1,59 @@ +# Makefile for uClibc's pthread library +# +# Copyright (C) 2002 Erik Andersen <andersen@uclibc.org> +# +# This program is free software; you can redistribute it and/or modify it under +# the terms of the GNU Library General Public License as published by the Free +# Software Foundation; either version 2 of the License, or (at your option) any +# later version. +# +# This program is distributed in the hope that it will be useful, but WITHOUT +# ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS +# FOR A PARTICULAR PURPOSE. See the GNU Library General Public License for more +# details. +# +# You should have received a copy of the GNU Library General Public License +# along with this program; if not, write to the Free Software Foundation, Inc., +# 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +# Makefile for uClibc + +TOPDIR=../../ +include $(TOPDIR)Rules.mak + +#Adjust the soname version to avoid namespace collisions with glibc's libpthread +PT_VERSION=$(MAJOR_VERSION).$(MINOR_VERSION) +LIBPTHREAD=../libpthread.a + +# set up system dependencies include dirs (NOTE: order matters!) +PTDIR = $(TOPDIR)libpthread/linuxthreads/ +SYSDEPINC = -I$(PTDIR)sysdeps/unix/sysv/linux \ + -I$(PTDIR)sysdeps/pthread \ + -I$(PTDIR)sysdeps/unix/sysv \ + -I$(PTDIR)sysdeps/unix/unix \ + -I$(PTDIR)sysdeps/$(TARGET_ARCH) \ + -I$(PTDIR)sysdeps \ + -I$(TOPDIR)libc/sysdeps/linux/$(TARGET_ARCH) +CFLAGS += $(SYSDEPINC) -DLIBPTHREAD_SO="\"libpthread.so.$(PT_VERSION)\"" -D_GNU_SOURCE + +CSRC=attr.c cancel.c condvar.c errno.c events.c join.c lockfile.c manager.c \ + mutex.c oldsemaphore.c pt-machine.c ptfork.c pthread.c \ + ptlongjmp.c rwlock.c semaphore.c signals.c specific.c spinlock.c \ + wrapsyscall.c #weaks.c no-tsd.c +COBJS=$(patsubst %.c,%.o, $(CSRC)) +OBJS=$(COBJS) + +all: $(OBJS) $(LIBPTHREAD) + +$(LIBPTHREAD): ar-target + +ar-target: $(OBJS) + $(AR) $(ARFLAGS) $(LIBPTHREAD) $(OBJS) + +$(COBJS): %.o : %.c + $(CC) $(CFLAGS) -c $< -o $@ + $(STRIPTOOL) -x -R .note -R .comment $*.o + +clean: + rm -f *.[oa] *~ core + + diff --git a/libpthread/linuxthreads/README b/libpthread/linuxthreads/README new file mode 100644 index 000000000..955bd59e7 --- /dev/null +++ b/libpthread/linuxthreads/README @@ -0,0 +1,166 @@ + Linuxthreads - POSIX 1003.1c kernel threads for Linux + + Copyright 1996, 1997 Xavier Leroy (Xavier.Leroy@inria.fr) + + +DESCRIPTION: + +This is release 0.7 (late beta) of LinuxThreads, a BiCapitalized +implementation of the Posix 1003.1c "pthread" interface for Linux. + +LinuxThreads provides kernel-level threads: each thread is a separate +Unix process, sharing its address space with the other threads through +the new system call clone(). Scheduling between threads is handled by +the kernel scheduler, just like scheduling between Unix processes. + + +REQUIREMENTS: + +- Linux version 2.0 and up (requires the new clone() system call + and the new realtime scheduler). + +- For Intel platforms: libc 5.2.18 or later is required. + 5.2.18 or 5.4.12 or later are recommended; + 5.3.12 and 5.4.7 have problems (see the FAQ.html file for more info). + +- Also supports glibc 2 (a.k.a. libc 6), which actually comes with + a specially-adapted version of this library. + +- Currently supports Intel, Alpha, Sparc, Motorola 68k, ARM and MIPS + platforms. + +- Multiprocessors are supported. + + +INSTALLATION: + +- Edit the Makefile, set the variables in the "Configuration" section. + +- Do "make". + +- Do "make install". + + +USING LINUXTHREADS: + + gcc -D_REENTRANT ... -lpthread + +A complete set of manual pages is included. Also see the subdirectory +Examples/ for some sample programs. + + +STATUS: + +- All functions in the Posix 1003.1c base interface implemented. + Also supports priority scheduling. + +- For users of libc 5 (H.J.Lu's libc), a number of C library functions + are reimplemented or wrapped to make them thread-safe, including: + * malloc functions + * stdio functions (define _REENTRANT before including <stdio.h>) + * per-thread errno variable (define _REENTRANT before including <errno.h>) + * directory reading functions (opendir(), etc) + * sleep() + * gmtime(), localtime() + + New library functions provided: + * flockfile(), funlockfile(), ftrylockfile() + * reentrant versions of network database functions (gethostbyname_r(), etc) + and password functions (getpwnam_r(), etc). + +- libc 6 (glibc 2) provides much better thread support than libc 5, + and comes with a specially-adapted version of LinuxThreads. + For serious multithreaded programming, you should consider switching + to glibc 2. It is available from ftp.gnu.org:/pub/gnu and its mirrors. + + +WARNING: + +Many existing libraries are not compatible with LinuxThreads, +either because they are not inherently thread-safe, or because they +have not been compiled with the -D_REENTRANT. For more info, see the +FAQ.html file in this directory. + +A prime example of the latter is Xlib. If you link it with +LinuxThreads, you'll probably get an "unknown 0 error" very +early. This is just a consequence of the Xlib binaries using the +global variable "errno" to fetch error codes, while LinuxThreads and +the C library use the per-thread "errno" location. + +See the file README.Xfree3.3 for info on how to compile the Xfree 3.3 +libraries to make them compatible with LinuxThreads. + + +KNOWN BUGS AND LIMITATIONS: + +- Threads share pretty much everything they should share according + to the standard: memory space, file descriptors, signal handlers, + current working directory, etc. One thing that they do not share + is their pid's and parent pid's. According to the standard, they + should have the same, but that's one thing we cannot achieve + in this implementation (until the CLONE_PID flag to clone() becomes + usable). + +- The current implementation uses the two signals SIGUSR1 and SIGUSR2, + so user-level code cannot employ them. Ideally, there should be two + signals reserved for this library. One signal is used for restarting + threads blocked on mutexes or conditions; the other is for thread + cancellation. + + *** This is not anymore true when the application runs on a kernel + newer than approximately 2.1.60. + +- The stacks for the threads are allocated high in the memory space, + below the stack of the initial process, and spaced 2M apart. + Stacks are allocated with the "grow on demand" flag, so they don't + use much virtual space initially (4k, currently), but can grow + up to 2M if needed. + + Reserving such a large address space for each thread means that, + on a 32-bit architecture, no more than about 1000 threads can + coexist (assuming a 2Gb address space for user processes), + but this is reasonable, since each thread uses up one entry in the + kernel's process table, which is usually limited to 512 processes. + + Another potential problem of the "grow on demand" scheme is that + nothing prevents the user from mmap'ing something in the 2M address + window reserved for a thread stack, possibly causing later extensions of + that stack to fail. Mapping at fixed addresses should be avoided + when using this library. + +- Signal handling does not fully conform to the Posix standard, + due to the fact that threads are here distinct processes that can be + sent signals individually, so there's no notion of sending a signal + to "the" process (the collection of all threads). + More precisely, here is a summary of the standard requirements + and how they are met by the implementation: + + 1- Synchronous signals (generated by the thread execution, e.g. SIGFPE) + are delivered to the thread that raised them. + (OK.) + + 2- A fatal asynchronous signal terminates all threads in the process. + (OK. The thread manager notices when a thread dies on a signal + and kills all other threads with the same signal.) + + 3- An asynchronous signal will be delivered to one of the threads + of the program which does not block the signal (it is unspecified + which). + (No, the signal is delivered to the thread it's been sent to, + based on the pid of the thread. If that thread is currently + blocking the signal, the signal remains pending.) + + 4- The signal will be delivered to at most one thread. + (OK, except for signals generated from the terminal or sent to + the process group, which will be delivered to all threads.) + +- The current implementation of the MIPS support assumes a MIPS ISA II + processor or better. These processors support atomic operations by + ll/sc instructions. Older R2000/R3000 series processors are not + supported yet; support for these will have higher overhead. + +- The current implementation of the ARM support assumes that the SWP + (atomic swap register with memory) instruction is available. This is + the case for all processors except for the ARM1 and ARM2. On StrongARM, + the SWP instruction does not bypass the cache, so multi-processor support + will be more troublesome. diff --git a/libpthread/linuxthreads/README.Xfree3.2 b/libpthread/linuxthreads/README.Xfree3.2 new file mode 100644 index 000000000..ac08e1583 --- /dev/null +++ b/libpthread/linuxthreads/README.Xfree3.2 @@ -0,0 +1,352 @@ +This file describes how to make a threaded X11R6. + +You need the source-code of XFree-3.2. I used the sources of X11R6.1 +(files: xc-1.tar.gz xc-2.tar.gz xc-3.tar.gz) and the patches to +XFree-3.2 (files: README.X11.patch R6.1pl1-3.2.diff.gz cfont32.tgz). + +Untar the xc-?.tar.gz files in a directory called XF3.2 and apply +the XFree-3.2 patches as described in README.X11.patch or use the +whole XFree86 source. + +Now apply the thread patch with + +patch -p0 < XF3.2.xc.diff + +Go to the XF3.2/xc directory and make the whole thing: +nice make World >& world.log & +tail -f world.log + +Wait a few hours or interrupt the process after the shared libs +are made. The shared libs are: + +XF3.2/xc/lib/ICE/libICE.so.6.0* +XF3.2/xc/lib/PEX5/libPEX5.so.6.0* +XF3.2/xc/lib/SM/libSM.so.6.0* +XF3.2/xc/lib/X11/libX11.so.6.1* +XF3.2/xc/lib/XIE/libXIE.so.6.0* +XF3.2/xc/lib/XThrStub/libXThrStub.so.6.0* +XF3.2/xc/lib/Xaw/libXaw.so.6.1* +XF3.2/xc/lib/Xext/libXext.so.6.1* +XF3.2/xc/lib/Xi/libXi.so.6.0* +XF3.2/xc/lib/Xmu/libXmu.so.6.0* +XF3.2/xc/lib/Xt/libXt.so.6.0* +XF3.2/xc/lib/Xtst/libXtst.so.6.1* + +(The Program dga didn't compile, but I have not check out why.) + +Now you can copy the resulting libs + +cp XF3.2/xc/lib/*/*.so.?.? /usr/X11R6/lib/ + +and create some links + +cd /usr/X11R6/lib/ +ln -s libXThrStub.so.6.0 libXThrStub.so.6 +ln -s libXThrStub.so.6 libXThrStub.so + +or use make install (not tested, and needs new configuration). + +It is possible with the libXThrSub to compile X11 programs without linking +libpthread to them and not necessary to recompile already installed +unthreaded X11 programs, because libXThrSub keeps the dynamic linker quit. +On the other hand you can link libpthread to a X11 program to use threads. + +I used linux 2.0.23 and libc 5.4.7 . + +Hans-Helmut Bühmann hans@expmech.ing.tu-bs.de + +---------------------------------------------------------------------------- + +XF3.2.xc.diff: +----------------------------------------------------------------------------- +diff -u --recursive XF3.2.orig/xc/config/cf/linux.cf XF3.2/xc/config/cf/linux.cf +--- XF3.2.orig/xc/config/cf/linux.cf Sun Nov 10 17:05:30 1996 ++++ XF3.2/xc/config/cf/linux.cf Sun Nov 10 16:30:55 1996 +@@ -61,6 +61,14 @@ + #define HasSnprintf YES + #endif + ++#define HasPosixThreads YES ++#define ThreadedX YES ++#define BuildThreadStubLibrary YES ++#define NeedUIThrStubs YES ++#define HasThreadSafeAPI NO ++#define SystemMTDefines -D_REENTRANT ++#define ThreadsLibraries -lpthread ++ + #define AvoidNullMakeCommand YES + #define StripInstalledPrograms YES + #define CompressAllFonts YES +@@ -158,7 +166,7 @@ + #define LdPostLib /* Never needed */ + + #ifdef i386Architecture +-#define OptimizedCDebugFlags DefaultGcc2i386Opt -m486 ++#define OptimizedCDebugFlags DefaultGcc2i386Opt -m486 -pipe + #define StandardDefines -Dlinux -D__i386__ -D_POSIX_SOURCE \ + -D_BSD_SOURCE -D_SVID_SOURCE -DX_LOCALE + #define XawI18nDefines -DUSE_XWCHAR_STRING -DUSE_XMBTOWC +diff -u --recursive XF3.2.orig/xc/config/cf/lnxLib.tmpl XF3.2/xc/config/cf/lnxLib.tmpl +--- XF3.2.orig/xc/config/cf/lnxLib.tmpl Sun Nov 10 17:05:30 1996 ++++ XF3.2/xc/config/cf/lnxLib.tmpl Sat Nov 9 14:52:39 1996 +@@ -19,7 +19,7 @@ + + #define CplusplusLibC + +-#define SharedX11Reqs ++#define SharedX11Reqs -L$(BUILDLIBDIR) -lXThrStub + #define SharedOldXReqs $(LDPRELIB) $(XLIBONLY) + #define SharedXtReqs $(LDPRELIB) $(XLIBONLY) $(SMLIB) $(ICELIB) + #define SharedXawReqs $(LDPRELIB) $(XMULIB) $(XTOOLLIB) $(XLIB) +diff -u --recursive XF3.2.orig/xc/include/Xthreads.h XF3.2/xc/include/Xthreads.h +--- XF3.2.orig/xc/include/Xthreads.h Thu Dec 7 02:19:09 1995 ++++ XF3.2/xc/include/Xthreads.h Sat Nov 9 01:04:55 1996 +@@ -229,12 +229,12 @@ + #define xcondition_wait(c,m) pthread_cond_wait(c,m) + #define xcondition_signal(c) pthread_cond_signal(c) + #define xcondition_broadcast(c) pthread_cond_broadcast(c) +-#ifdef _DECTHREADS_ ++#if defined(_DECTHREADS_) || defined(linux) + static xthread_t _X_no_thread_id; + #define xthread_have_id(id) !pthread_equal(id, _X_no_thread_id) + #define xthread_clear_id(id) id = _X_no_thread_id + #define xthread_equal(id1,id2) pthread_equal(id1, id2) +-#endif /* _DECTHREADS_ */ ++#endif /* _DECTHREADS_ || linux */ + #if _CMA_VENDOR_ == _CMA__IBM + #ifdef DEBUG /* too much of a hack to enable normally */ + /* see also cma__obj_set_name() */ +diff -u --recursive XF3.2.orig/xc/lib/X11/util/makekeys.c XF3.2/xc/lib/X11/util/makekeys.c +--- XF3.2.orig/xc/lib/X11/util/makekeys.c Mon Apr 18 02:22:22 1994 ++++ XF3.2/xc/lib/X11/util/makekeys.c Sat Nov 9 00:44:14 1996 +@@ -73,7 +73,7 @@ + register char c; + int first; + int best_max_rehash; +- int best_z; ++ int best_z = 0; + int num_found; + KeySym val; + +diff -u --recursive XF3.2.orig/xc/lib/XThrStub/Imakefile XF3.2/xc/lib/XThrStub/Imakefile +--- XF3.2.orig/xc/lib/XThrStub/Imakefile Sun Nov 10 17:08:12 1996 ++++ XF3.2/xc/lib/XThrStub/Imakefile Sat Nov 9 19:04:51 1996 +@@ -25,7 +25,7 @@ + DEFINES = $(ALLOC_DEFINES) + INCLUDES = + SRCS = $(STUBSRCS) +- OBJS = $(STUBOBJS ++ OBJS = $(STUBOBJS) + LINTLIBS = $(LINTXLIB) + + #include <Library.tmpl> +diff -u --recursive XF3.2.orig/xc/lib/XThrStub/UIThrStubs.c XF3.2/xc/lib/XThrStub/UIThrStubs.c +--- XF3.2.orig/xc/lib/XThrStub/UIThrStubs.c Sun Nov 10 17:08:12 1996 ++++ XF3.2/xc/lib/XThrStub/UIThrStubs.c Sun Nov 10 15:14:55 1996 +@@ -37,16 +37,43 @@ + * specificies the thread library on the link line. + */ + ++#if defined(linux) ++#include <pthread.h> ++#else + #include <thread.h> + #include <synch.h> ++#endif + ++#if defined(linux) ++static pthread_t no_thread_id; ++#endif /* defined(linux) */ ++ ++#if defined(linux) ++#pragma weak pthread_self = _Xthr_self_stub_ ++pthread_t ++_Xthr_self_stub_() ++{ ++ return(no_thread_id); ++} ++#else /* defined(linux) */ + #pragma weak thr_self = _Xthr_self_stub_ + thread_t + _Xthr_self_stub_() + { + return((thread_t)0); + } ++#endif /* defined(linux) */ + ++#if defined(linux) ++#pragma weak pthread_mutex_init = _Xmutex_init_stub_ ++int ++_Xmutex_init_stub_(m, a) ++ pthread_mutex_t *m; ++ __const pthread_mutexattr_t *a; ++{ ++ return(0); ++} ++#else /* defined(linux) */ + #pragma weak mutex_init = _Xmutex_init_stub_ + int + _Xmutex_init_stub_(m, t, a) +@@ -56,7 +83,17 @@ + { + return(0); + } ++#endif /* defined(linux) */ + ++#if defined(linux) ++#pragma weak pthread_mutex_destroy = _Xmutex_destroy_stub_ ++int ++_Xmutex_destroy_stub_(m) ++ pthread_mutex_t *m; ++{ ++ return(0); ++} ++#else /* defined(linux) */ + #pragma weak mutex_destroy = _Xmutex_destroy_stub_ + int + _Xmutex_destroy_stub_(m) +@@ -64,7 +101,17 @@ + { + return(0); + } ++#endif /* defined(linux) */ + ++#if defined(linux) ++#pragma weak pthread_mutex_lock = _Xmutex_lock_stub_ ++int ++_Xmutex_lock_stub_(m) ++ pthread_mutex_t *m; ++{ ++ return(0); ++} ++#else /* defined(linux) */ + #pragma weak mutex_lock = _Xmutex_lock_stub_ + int + _Xmutex_lock_stub_(m) +@@ -72,7 +119,17 @@ + { + return(0); + } ++#endif /* defined(linux) */ + ++#if defined(linux) ++#pragma weak pthread_mutex_unlock = _Xmutex_unlock_stub_ ++int ++_Xmutex_unlock_stub_(m) ++ pthread_mutex_t *m; ++{ ++ return(0); ++} ++#else /* defined(linux) */ + #pragma weak mutex_unlock = _Xmutex_unlock_stub_ + int + _Xmutex_unlock_stub_(m) +@@ -80,7 +137,18 @@ + { + return(0); + } ++#endif /* defined(linux) */ + ++#if defined(linux) ++#pragma weak pthread_cond_init = _Xcond_init_stub_ ++int ++_Xcond_init_stub_(c, a) ++ pthread_cond_t *c; ++ __const pthread_condattr_t *a; ++{ ++ return(0); ++} ++#else /* defined(linux) */ + #pragma weak cond_init = _Xcond_init_stub_ + int + _Xcond_init_stub_(c, t, a) +@@ -90,7 +158,17 @@ + { + return(0); + } ++#endif /* defined(linux) */ + ++#if defined(linux) ++#pragma weak pthread_cond_destroy = _Xcond_destroy_stub_ ++int ++_Xcond_destroy_stub_(c) ++ pthread_cond_t *c; ++{ ++ return(0); ++} ++#else /* defined(linux) */ + #pragma weak cond_destroy = _Xcond_destroy_stub_ + int + _Xcond_destroy_stub_(c) +@@ -98,7 +176,18 @@ + { + return(0); + } ++#endif /* defined(linux) */ + ++#if defined(linux) ++#pragma weak pthread_cond_wait = _Xcond_wait_stub_ ++int ++_Xcond_wait_stub_(c,m) ++ pthread_cond_t *c; ++ pthread_mutex_t *m; ++{ ++ return(0); ++} ++#else /* defined(linux) */ + #pragma weak cond_wait = _Xcond_wait_stub_ + int + _Xcond_wait_stub_(c,m) +@@ -107,7 +196,17 @@ + { + return(0); + } ++#endif /* defined(linux) */ + ++#if defined(linux) ++#pragma weak pthread_cond_signal = _Xcond_signal_stub_ ++int ++_Xcond_signal_stub_(c) ++ pthread_cond_t *c; ++{ ++ return(0); ++} ++#else /* defined(linux) */ + #pragma weak cond_signal = _Xcond_signal_stub_ + int + _Xcond_signal_stub_(c) +@@ -115,7 +214,17 @@ + { + return(0); + } ++#endif /* defined(linux) */ + ++#if defined(linux) ++#pragma weak pthread_cond_broadcast = _Xcond_broadcast_stub_ ++int ++_Xcond_broadcast_stub_(c) ++ pthread_cond_t *c; ++{ ++ return(0); ++} ++#else /* defined(linux) */ + #pragma weak cond_broadcast = _Xcond_broadcast_stub_ + int + _Xcond_broadcast_stub_(c) +@@ -123,3 +232,15 @@ + { + return(0); + } ++#endif /* defined(linux) */ ++ ++#if defined(linux) ++#pragma weak pthread_equal = _Xthr_equal_stub_ ++int ++_Xthr_equal_stub_(t1, t2) ++ pthread_t t1; ++ pthread_t t2; ++{ ++ return(1); ++} ++#endif /* defined(linux) */ +------------------------------------------------------------------------- diff --git a/libpthread/linuxthreads/Versions b/libpthread/linuxthreads/Versions new file mode 100644 index 000000000..c0ec79238 --- /dev/null +++ b/libpthread/linuxthreads/Versions @@ -0,0 +1,121 @@ +libc { + GLIBC_2.0 { + pthread_attr_destroy; pthread_attr_getdetachstate; + pthread_attr_getinheritsched; pthread_attr_getschedparam; + pthread_attr_getschedpolicy; pthread_attr_getscope; pthread_attr_init; + pthread_attr_setdetachstate; pthread_attr_setinheritsched; + pthread_attr_setschedparam; pthread_attr_setschedpolicy; + pthread_attr_setscope; pthread_cond_broadcast; pthread_cond_destroy; + pthread_cond_init; pthread_cond_signal; pthread_cond_wait; + pthread_condattr_destroy; pthread_condattr_init; pthread_equal; + pthread_exit; pthread_getschedparam; pthread_mutex_destroy; + pthread_mutex_init; pthread_mutex_lock; pthread_mutex_unlock; + pthread_mutexattr_getkind_np; pthread_mutexattr_setkind_np; + pthread_self; pthread_setcancelstate; pthread_setcanceltype; + pthread_setschedparam; + + # Internal libc interface to libpthread + __libc_internal_tsd_get; __libc_internal_tsd_set; + } + GLIBC_2.1 { + pthread_attr_init; + } +} + +ld.so { + GLIBC_2.0 { + # Internal libc interface to libpthread + __libc_internal_tsd_get; __libc_internal_tsd_set; + } +} + +libpthread { + GLIBC_2.0 { + # Hidden entry point (through macros). + _pthread_cleanup_pop; _pthread_cleanup_pop_restore; _pthread_cleanup_push; + _pthread_cleanup_push_defer; + + # Internal libc interface to libpthread + __libc_internal_tsd_get; __libc_internal_tsd_set; + + # Overwritten libc functions. + accept; close; connect; fcntl; fork; fsync; longjmp; lseek; msync; + nanosleep; open; pause; raise; read; recv; recvfrom; recvmsg; send; + sendmsg; sendto; sigaction; siglongjmp; system; tcdrain; wait; + waitpid; write; + __close; __connect; __fcntl; __lseek; __open; __read; __send; __wait; + __write; + _IO_flockfile; _IO_ftrylockfile; _IO_funlockfile; + vfork; __fork; + + # POSIX.1c extensions to libc. + flockfile; funlockfile; ftrylockfile; + + # Non-standard POSIX1.x functions. + pthread_kill_other_threads_np; pthread_mutexattr_getkind_np; + pthread_mutexattr_setkind_np; + + # Real POSIX.1c functions. + pthread_atfork; pthread_attr_destroy; pthread_attr_getdetachstate; + pthread_attr_getinheritsched; pthread_attr_getschedparam; + pthread_attr_getschedpolicy; pthread_attr_getscope; pthread_attr_init; + pthread_attr_setdetachstate; pthread_attr_setinheritsched; + pthread_attr_setschedparam; pthread_attr_setschedpolicy; + pthread_attr_setscope; pthread_cancel; pthread_cond_broadcast; + pthread_cond_destroy; pthread_cond_init; pthread_cond_signal; + pthread_cond_timedwait; pthread_cond_wait; pthread_condattr_destroy; + pthread_condattr_init; pthread_create; pthread_detach; pthread_equal; + pthread_exit; pthread_getschedparam; pthread_getspecific; pthread_join; + pthread_key_create; pthread_key_delete; pthread_kill; + pthread_mutex_destroy; pthread_mutex_init; pthread_mutex_lock; + pthread_mutex_trylock; pthread_mutex_unlock; pthread_mutexattr_destroy; + pthread_mutexattr_init; pthread_once; pthread_self; pthread_setcancelstate; + pthread_setcanceltype; pthread_setschedparam; pthread_setspecific; + pthread_sigmask; pthread_testcancel; + + sem_destroy; sem_getvalue; sem_init; sem_post; sem_trywait; sem_wait; + sigwait; + + # Protected names for functions used in other shared objects. + __pthread_atfork; __pthread_initialize; __pthread_getspecific; + __pthread_key_create; __pthread_mutex_destroy; __pthread_mutex_init; + __pthread_mutex_lock; __pthread_mutex_trylock; __pthread_mutex_unlock; + __pthread_mutexattr_destroy; __pthread_mutexattr_init; + __pthread_mutexattr_settype; __pthread_once; __pthread_setspecific; + + # The error functions. + __errno_location; __h_errno_location; + } + GLIBC_2.1 { + # Functions with changed interface. + pthread_attr_init; pthread_create; + + # Unix98 extensions. + pthread_rwlock_init; pthread_rwlock_destroy; pthread_rwlock_rdlock; + pthread_rwlock_tryrdlock; pthread_rwlock_wrlock; pthread_rwlock_trywrlock; + pthread_rwlock_unlock; pthread_rwlockattr_init; pthread_rwlockattr_destroy; + pthread_rwlockattr_getpshared; pthread_rwlockattr_setpshared; + pthread_rwlockattr_getkind_np; pthread_rwlockattr_setkind_np; + + pthread_attr_getguardsize; pthread_attr_setguardsize; + pthread_attr_getstackaddr; pthread_attr_setstackaddr; + pthread_attr_getstacksize; pthread_attr_setstacksize; + + pthread_getconcurrency; pthread_setconcurrency; + + pthread_mutexattr_gettype; pthread_mutexattr_settype; + + sem_destroy; sem_getvalue; sem_init; sem_post; sem_trywait; sem_wait; + + # helper functions + __libc_current_sigrtmin; __libc_current_sigrtmax; + __libc_allocate_rtsig; + } + GLIBC_2.1.1 { + sem_close; sem_open; sem_unlink; + } + GLIBC_2.1.2 { + __pthread_kill_other_threads_np; + __vfork; + } +} diff --git a/libpthread/linuxthreads/attr.c b/libpthread/linuxthreads/attr.c new file mode 100644 index 000000000..2907a56a8 --- /dev/null +++ b/libpthread/linuxthreads/attr.c @@ -0,0 +1,214 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* changed for uClibc */ +#define __sched_get_priority_min sched_get_priority_min +#define __sched_get_priority_max sched_get_priority_max + +/* Handling of thread attributes */ + +#include <errno.h> +#include <string.h> +#include <unistd.h> +#include <sys/param.h> +#include "pthread.h" +#include "internals.h" + +extern int __getpagesize(void); + +/* NOTE: With uClibc I don't think we need this versioning stuff. + * Therefore, define the function pthread_attr_init() here using + * a strong symbol. */ + +//int __pthread_attr_init_2_1(pthread_attr_t *attr) +int pthread_attr_init(pthread_attr_t *attr) +{ + size_t ps = __getpagesize (); + + attr->__detachstate = PTHREAD_CREATE_JOINABLE; + attr->__schedpolicy = SCHED_OTHER; + attr->__schedparam.sched_priority = 0; + attr->__inheritsched = PTHREAD_EXPLICIT_SCHED; + attr->__scope = PTHREAD_SCOPE_SYSTEM; + attr->__guardsize = ps; + attr->__stackaddr = NULL; + attr->__stackaddr_set = 0; + attr->__stacksize = STACK_SIZE - ps; + return 0; +} + +/* uClibc: leave out this for now. */ +#if DO_PTHREAD_VERSIONING_WITH_UCLIBC +#if defined HAVE_ELF && defined PIC && defined DO_VERSIONING +default_symbol_version (__pthread_attr_init_2_1, pthread_attr_init, GLIBC_2.1); + +int __pthread_attr_init_2_0(pthread_attr_t *attr) +{ + attr->__detachstate = PTHREAD_CREATE_JOINABLE; + attr->__schedpolicy = SCHED_OTHER; + attr->__schedparam.sched_priority = 0; + attr->__inheritsched = PTHREAD_EXPLICIT_SCHED; + attr->__scope = PTHREAD_SCOPE_SYSTEM; + return 0; +} +symbol_version (__pthread_attr_init_2_0, pthread_attr_init, GLIBC_2.0); +#else +strong_alias (__pthread_attr_init_2_1, pthread_attr_init) +#endif +#endif /* DO_PTHREAD_VERSIONING_WITH_UCLIBC */ + +int pthread_attr_destroy(pthread_attr_t *attr) +{ + return 0; +} + +int pthread_attr_setdetachstate(pthread_attr_t *attr, int detachstate) +{ + if (detachstate < PTHREAD_CREATE_JOINABLE || + detachstate > PTHREAD_CREATE_DETACHED) + return EINVAL; + attr->__detachstate = detachstate; + return 0; +} + +int pthread_attr_getdetachstate(const pthread_attr_t *attr, int *detachstate) +{ + *detachstate = attr->__detachstate; + return 0; +} + +int pthread_attr_setschedparam(pthread_attr_t *attr, + const struct sched_param *param) +{ + int max_prio = __sched_get_priority_max(attr->__schedpolicy); + int min_prio = __sched_get_priority_min(attr->__schedpolicy); + + if (param->sched_priority < min_prio || param->sched_priority > max_prio) + return EINVAL; + memcpy (&attr->__schedparam, param, sizeof (struct sched_param)); + return 0; +} + +int pthread_attr_getschedparam(const pthread_attr_t *attr, + struct sched_param *param) +{ + memcpy (param, &attr->__schedparam, sizeof (struct sched_param)); + return 0; +} + +int pthread_attr_setschedpolicy(pthread_attr_t *attr, int policy) +{ + if (policy != SCHED_OTHER && policy != SCHED_FIFO && policy != SCHED_RR) + return EINVAL; + attr->__schedpolicy = policy; + return 0; +} + +int pthread_attr_getschedpolicy(const pthread_attr_t *attr, int *policy) +{ + *policy = attr->__schedpolicy; + return 0; +} + +int pthread_attr_setinheritsched(pthread_attr_t *attr, int inherit) +{ + if (inherit != PTHREAD_INHERIT_SCHED && inherit != PTHREAD_EXPLICIT_SCHED) + return EINVAL; + attr->__inheritsched = inherit; + return 0; +} + +int pthread_attr_getinheritsched(const pthread_attr_t *attr, int *inherit) +{ + *inherit = attr->__inheritsched; + return 0; +} + +int pthread_attr_setscope(pthread_attr_t *attr, int scope) +{ + switch (scope) { + case PTHREAD_SCOPE_SYSTEM: + attr->__scope = scope; + return 0; + case PTHREAD_SCOPE_PROCESS: + return ENOTSUP; + default: + return EINVAL; + } +} + +int pthread_attr_getscope(const pthread_attr_t *attr, int *scope) +{ + *scope = attr->__scope; + return 0; +} + +int __pthread_attr_setguardsize(pthread_attr_t *attr, size_t guardsize) +{ + size_t ps = __getpagesize (); + + /* First round up the guard size. */ + guardsize = roundup (guardsize, ps); + + /* The guard size must not be larger than the stack itself */ + if (guardsize >= attr->__stacksize) return EINVAL; + + attr->__guardsize = guardsize; + + return 0; +} +weak_alias (__pthread_attr_setguardsize, pthread_attr_setguardsize) + +int __pthread_attr_getguardsize(const pthread_attr_t *attr, size_t *guardsize) +{ + *guardsize = attr->__guardsize; + return 0; +} +weak_alias (__pthread_attr_getguardsize, pthread_attr_getguardsize) + +int __pthread_attr_setstackaddr(pthread_attr_t *attr, void *stackaddr) +{ + attr->__stackaddr = stackaddr; + attr->__stackaddr_set = 1; + return 0; +} +weak_alias (__pthread_attr_setstackaddr, pthread_attr_setstackaddr) + +int __pthread_attr_getstackaddr(const pthread_attr_t *attr, void **stackaddr) +{ + /* XXX This function has a stupid definition. The standard specifies + no error value but what is if no stack address was set? We simply + return the value we have in the member. */ + *stackaddr = attr->__stackaddr; + return 0; +} +weak_alias (__pthread_attr_getstackaddr, pthread_attr_getstackaddr) + +int __pthread_attr_setstacksize(pthread_attr_t *attr, size_t stacksize) +{ + /* We don't accept value smaller than PTHREAD_STACK_MIN. */ + if (stacksize < PTHREAD_STACK_MIN) + return EINVAL; + + attr->__stacksize = stacksize; + return 0; +} +weak_alias (__pthread_attr_setstacksize, pthread_attr_setstacksize) + +int __pthread_attr_getstacksize(const pthread_attr_t *attr, size_t *stacksize) +{ + *stacksize = attr->__stacksize; + return 0; +} +weak_alias (__pthread_attr_getstacksize, pthread_attr_getstacksize) diff --git a/libpthread/linuxthreads/cancel.c b/libpthread/linuxthreads/cancel.c new file mode 100644 index 000000000..8fd8c1e60 --- /dev/null +++ b/libpthread/linuxthreads/cancel.c @@ -0,0 +1,171 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* Thread cancellation */ + +#include <errno.h> +#include "pthread.h" +#include "internals.h" +#include "spinlock.h" +#include "restart.h" + +int pthread_setcancelstate(int state, int * oldstate) +{ + pthread_descr self = thread_self(); + if (state < PTHREAD_CANCEL_ENABLE || state > PTHREAD_CANCEL_DISABLE) + return EINVAL; + if (oldstate != NULL) *oldstate = THREAD_GETMEM(self, p_cancelstate); + THREAD_SETMEM(self, p_cancelstate, state); + if (THREAD_GETMEM(self, p_canceled) && + THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE && + THREAD_GETMEM(self, p_canceltype) == PTHREAD_CANCEL_ASYNCHRONOUS) + pthread_exit(PTHREAD_CANCELED); + return 0; +} + +int pthread_setcanceltype(int type, int * oldtype) +{ + pthread_descr self = thread_self(); + if (type < PTHREAD_CANCEL_DEFERRED || type > PTHREAD_CANCEL_ASYNCHRONOUS) + return EINVAL; + if (oldtype != NULL) *oldtype = THREAD_GETMEM(self, p_canceltype); + THREAD_SETMEM(self, p_canceltype, type); + if (THREAD_GETMEM(self, p_canceled) && + THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE && + THREAD_GETMEM(self, p_canceltype) == PTHREAD_CANCEL_ASYNCHRONOUS) + pthread_exit(PTHREAD_CANCELED); + return 0; +} + +int pthread_cancel(pthread_t thread) +{ + pthread_handle handle = thread_handle(thread); + int pid; + int dorestart = 0; + pthread_descr th; + pthread_extricate_if *pextricate; + + __pthread_lock(&handle->h_lock, NULL); + if (invalid_handle(handle, thread)) { + __pthread_unlock(&handle->h_lock); + return ESRCH; + } + + th = handle->h_descr; + + if (th->p_canceled) { + __pthread_unlock(&handle->h_lock); + return 0; + } + + pextricate = th->p_extricate; + th->p_canceled = 1; + pid = th->p_pid; + + /* If the thread has registered an extrication interface, then + invoke the interface. If it returns 1, then we succeeded in + dequeuing the thread from whatever waiting object it was enqueued + with. In that case, it is our responsibility to wake it up. + And also to set the p_woken_by_cancel flag so the woken thread + can tell that it was woken by cancellation. */ + + if (pextricate != NULL) { + dorestart = pextricate->pu_extricate_func(pextricate->pu_object, th); + th->p_woken_by_cancel = dorestart; + } + + __pthread_unlock(&handle->h_lock); + + /* If the thread has suspended or is about to, then we unblock it by + issuing a restart, instead of a cancel signal. Otherwise we send + the cancel signal to unblock the thread from a cancellation point, + or to initiate asynchronous cancellation. The restart is needed so + we have proper accounting of restarts; suspend decrements the thread's + resume count, and restart() increments it. This also means that suspend's + handling of the cancel signal is obsolete. */ + + if (dorestart) + restart(th); + else + kill(pid, __pthread_sig_cancel); + + return 0; +} + +void pthread_testcancel(void) +{ + pthread_descr self = thread_self(); + if (THREAD_GETMEM(self, p_canceled) + && THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE) + pthread_exit(PTHREAD_CANCELED); +} + +void _pthread_cleanup_push(struct _pthread_cleanup_buffer * buffer, + void (*routine)(void *), void * arg) +{ + pthread_descr self = thread_self(); + buffer->__routine = routine; + buffer->__arg = arg; + buffer->__prev = THREAD_GETMEM(self, p_cleanup); + THREAD_SETMEM(self, p_cleanup, buffer); +} + +void _pthread_cleanup_pop(struct _pthread_cleanup_buffer * buffer, + int execute) +{ + pthread_descr self = thread_self(); + if (execute) buffer->__routine(buffer->__arg); + THREAD_SETMEM(self, p_cleanup, buffer->__prev); +} + +void _pthread_cleanup_push_defer(struct _pthread_cleanup_buffer * buffer, + void (*routine)(void *), void * arg) +{ + pthread_descr self = thread_self(); + buffer->__routine = routine; + buffer->__arg = arg; + buffer->__canceltype = THREAD_GETMEM(self, p_canceltype); + buffer->__prev = THREAD_GETMEM(self, p_cleanup); + THREAD_SETMEM(self, p_canceltype, PTHREAD_CANCEL_DEFERRED); + THREAD_SETMEM(self, p_cleanup, buffer); +} + +void _pthread_cleanup_pop_restore(struct _pthread_cleanup_buffer * buffer, + int execute) +{ + pthread_descr self = thread_self(); + if (execute) buffer->__routine(buffer->__arg); + THREAD_SETMEM(self, p_cleanup, buffer->__prev); + THREAD_SETMEM(self, p_canceltype, buffer->__canceltype); + if (THREAD_GETMEM(self, p_canceled) && + THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE && + THREAD_GETMEM(self, p_canceltype) == PTHREAD_CANCEL_ASYNCHRONOUS) + pthread_exit(PTHREAD_CANCELED); +} + +void __pthread_perform_cleanup(void) +{ + pthread_descr self = thread_self(); + struct _pthread_cleanup_buffer * c; + for (c = THREAD_GETMEM(self, p_cleanup); c != NULL; c = c->__prev) + c->__routine(c->__arg); +} + +#ifndef PIC +/* We need a hook to force the cancelation wrappers to be linked in when + static libpthread is used. */ +extern const int __pthread_provide_wrappers; +static const int * const __pthread_require_wrappers = + &__pthread_provide_wrappers; +#endif diff --git a/libpthread/linuxthreads/condvar.c b/libpthread/linuxthreads/condvar.c new file mode 100644 index 000000000..85754a18f --- /dev/null +++ b/libpthread/linuxthreads/condvar.c @@ -0,0 +1,417 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* and Pavel Krauz (krauz@fsid.cvut.cz). */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* Condition variables */ + +#include <errno.h> +#include <sched.h> +#include <stddef.h> +#include <sys/time.h> +#include "pthread.h" +#include "internals.h" +#include "spinlock.h" +#include "queue.h" +#include "restart.h" + +static int pthread_cond_timedwait_relative_old(pthread_cond_t *, + pthread_mutex_t *, const struct timespec *); + +static int pthread_cond_timedwait_relative_new(pthread_cond_t *, + pthread_mutex_t *, const struct timespec *); + +static int (*pthread_cond_tw_rel)(pthread_cond_t *, pthread_mutex_t *, + const struct timespec *) = pthread_cond_timedwait_relative_old; + +/* initialize this module */ +void __pthread_init_condvar(int rt_sig_available) +{ + if (rt_sig_available) + pthread_cond_tw_rel = pthread_cond_timedwait_relative_new; +} + +int pthread_cond_init(pthread_cond_t *cond, + const pthread_condattr_t *cond_attr) +{ + __pthread_init_lock(&cond->__c_lock); + cond->__c_waiting = NULL; + return 0; +} + +int pthread_cond_destroy(pthread_cond_t *cond) +{ + if (cond->__c_waiting != NULL) return EBUSY; + return 0; +} + +/* Function called by pthread_cancel to remove the thread from + waiting on a condition variable queue. */ + +static int cond_extricate_func(void *obj, pthread_descr th) +{ + volatile pthread_descr self = thread_self(); + pthread_cond_t *cond = obj; + int did_remove = 0; + + __pthread_lock(&cond->__c_lock, self); + did_remove = remove_from_queue(&cond->__c_waiting, th); + __pthread_unlock(&cond->__c_lock); + + return did_remove; +} + +int pthread_cond_wait(pthread_cond_t *cond, pthread_mutex_t *mutex) +{ + volatile pthread_descr self = thread_self(); + pthread_extricate_if extr; + int already_canceled = 0; + + /* Set up extrication interface */ + extr.pu_object = cond; + extr.pu_extricate_func = cond_extricate_func; + + /* Register extrication interface */ + __pthread_set_own_extricate_if(self, &extr); + + /* Atomically enqueue thread for waiting, but only if it is not + canceled. If the thread is canceled, then it will fall through the + suspend call below, and then call pthread_exit without + having to worry about whether it is still on the condition variable queue. + This depends on pthread_cancel setting p_canceled before calling the + extricate function. */ + + __pthread_lock(&cond->__c_lock, self); + if (!(THREAD_GETMEM(self, p_canceled) + && THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE)) + enqueue(&cond->__c_waiting, self); + else + already_canceled = 1; + __pthread_unlock(&cond->__c_lock); + + if (already_canceled) { + __pthread_set_own_extricate_if(self, 0); + pthread_exit(PTHREAD_CANCELED); + } + + pthread_mutex_unlock(mutex); + + suspend(self); + __pthread_set_own_extricate_if(self, 0); + + /* Check for cancellation again, to provide correct cancellation + point behavior */ + + if (THREAD_GETMEM(self, p_woken_by_cancel) + && THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE) { + THREAD_SETMEM(self, p_woken_by_cancel, 0); + pthread_mutex_lock(mutex); + pthread_exit(PTHREAD_CANCELED); + } + + pthread_mutex_lock(mutex); + return 0; +} + +/* The following function is used on kernels that don't have rt signals. + SIGUSR1 is used as the restart signal. The different code is needed + because that ordinary signal does not queue. */ + +static int +pthread_cond_timedwait_relative_old(pthread_cond_t *cond, + pthread_mutex_t *mutex, + const struct timespec * abstime) +{ + volatile pthread_descr self = thread_self(); + sigset_t unblock, initial_mask; + int already_canceled = 0; + int was_signalled = 0; + sigjmp_buf jmpbuf; + pthread_extricate_if extr; + + /* Set up extrication interface */ + extr.pu_object = cond; + extr.pu_extricate_func = cond_extricate_func; + + /* Register extrication interface */ + __pthread_set_own_extricate_if(self, &extr); + + /* Enqueue to wait on the condition and check for cancellation. */ + __pthread_lock(&cond->__c_lock, self); + if (!(THREAD_GETMEM(self, p_canceled) + && THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE)) + enqueue(&cond->__c_waiting, self); + else + already_canceled = 1; + __pthread_unlock(&cond->__c_lock); + + if (already_canceled) { + __pthread_set_own_extricate_if(self, 0); + pthread_exit(PTHREAD_CANCELED); + } + + pthread_mutex_unlock(mutex); + + if (atomic_decrement(&self->p_resume_count) == 0) { + /* Set up a longjmp handler for the restart signal, unblock + the signal and sleep. */ + + if (sigsetjmp(jmpbuf, 1) == 0) { + THREAD_SETMEM(self, p_signal_jmp, &jmpbuf); + THREAD_SETMEM(self, p_signal, 0); + /* Unblock the restart signal */ + sigemptyset(&unblock); + sigaddset(&unblock, __pthread_sig_restart); + sigprocmask(SIG_UNBLOCK, &unblock, &initial_mask); + + while (1) { + struct timeval now; + struct timespec reltime; + + /* Compute a time offset relative to now. */ + gettimeofday (&now, NULL); + reltime.tv_nsec = abstime->tv_nsec - now.tv_usec * 1000; + reltime.tv_sec = abstime->tv_sec - now.tv_sec; + if (reltime.tv_nsec < 0) { + reltime.tv_nsec += 1000000000; + reltime.tv_sec -= 1; + } + + /* Sleep for the required duration. If woken by a signal, resume waiting + as required by Single Unix Specification. */ + if (reltime.tv_sec < 0 || __libc_nanosleep(&reltime, NULL) == 0) + break; + } + + /* Block the restart signal again */ + sigprocmask(SIG_SETMASK, &initial_mask, NULL); + was_signalled = 0; + } else { + was_signalled = 1; + } + THREAD_SETMEM(self, p_signal_jmp, NULL); + } + + /* Now was_signalled is true if we exited the above code + due to the delivery of a restart signal. In that case, + we know we have been dequeued and resumed and that the + resume count is balanced. Otherwise, there are some + cases to consider. First, try to bump up the resume count + back to zero. If it goes to 1, it means restart() was + invoked on this thread. The signal must be consumed + and the count bumped down and everything is cool. + Otherwise, no restart was delivered yet, so we remove + the thread from the queue. If this succeeds, it's a clear + case of timeout. If we fail to remove from the queue, then we + must wait for a restart. */ + + if (!was_signalled) { + if (atomic_increment(&self->p_resume_count) != -1) { + __pthread_wait_for_restart_signal(self); + atomic_decrement(&self->p_resume_count); /* should be zero now! */ + } else { + int was_on_queue; + __pthread_lock(&cond->__c_lock, self); + was_on_queue = remove_from_queue(&cond->__c_waiting, self); + __pthread_unlock(&cond->__c_lock); + + if (was_on_queue) { + __pthread_set_own_extricate_if(self, 0); + pthread_mutex_lock(mutex); + return ETIMEDOUT; + } + + suspend(self); + } + } + + __pthread_set_own_extricate_if(self, 0); + + /* The remaining logic is the same as in other cancellable waits, + such as pthread_join sem_wait or pthread_cond wait. */ + + if (THREAD_GETMEM(self, p_woken_by_cancel) + && THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE) { + THREAD_SETMEM(self, p_woken_by_cancel, 0); + pthread_mutex_lock(mutex); + pthread_exit(PTHREAD_CANCELED); + } + + pthread_mutex_lock(mutex); + return 0; +} + +/* The following function is used on new (late 2.1 and 2.2 and higher) kernels + that have rt signals which queue. */ + +static int +pthread_cond_timedwait_relative_new(pthread_cond_t *cond, + pthread_mutex_t *mutex, + const struct timespec * abstime) +{ + volatile pthread_descr self = thread_self(); + sigset_t unblock, initial_mask; + int already_canceled = 0; + int was_signalled = 0; + sigjmp_buf jmpbuf; + pthread_extricate_if extr; + + /* Set up extrication interface */ + extr.pu_object = cond; + extr.pu_extricate_func = cond_extricate_func; + + /* Register extrication interface */ + __pthread_set_own_extricate_if(self, &extr); + + /* Enqueue to wait on the condition and check for cancellation. */ + __pthread_lock(&cond->__c_lock, self); + if (!(THREAD_GETMEM(self, p_canceled) + && THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE)) + enqueue(&cond->__c_waiting, self); + else + already_canceled = 1; + __pthread_unlock(&cond->__c_lock); + + if (already_canceled) { + __pthread_set_own_extricate_if(self, 0); + pthread_exit(PTHREAD_CANCELED); + } + + pthread_mutex_unlock(mutex); + + /* Set up a longjmp handler for the restart signal, unblock + the signal and sleep. */ + + if (sigsetjmp(jmpbuf, 1) == 0) { + THREAD_SETMEM(self, p_signal_jmp, &jmpbuf); + THREAD_SETMEM(self, p_signal, 0); + /* Unblock the restart signal */ + sigemptyset(&unblock); + sigaddset(&unblock, __pthread_sig_restart); + sigprocmask(SIG_UNBLOCK, &unblock, &initial_mask); + + while (1) { + struct timeval now; + struct timespec reltime; + + /* Compute a time offset relative to now. */ + gettimeofday (&now, NULL); + reltime.tv_nsec = abstime->tv_nsec - now.tv_usec * 1000; + reltime.tv_sec = abstime->tv_sec - now.tv_sec; + if (reltime.tv_nsec < 0) { + reltime.tv_nsec += 1000000000; + reltime.tv_sec -= 1; + } + + /* Sleep for the required duration. If woken by a signal, + resume waiting as required by Single Unix Specification. */ + if (reltime.tv_sec < 0 || __libc_nanosleep(&reltime, NULL) == 0) + break; + } + + /* Block the restart signal again */ + sigprocmask(SIG_SETMASK, &initial_mask, NULL); + was_signalled = 0; + } else { + was_signalled = 1; + } + THREAD_SETMEM(self, p_signal_jmp, NULL); + + /* Now was_signalled is true if we exited the above code + due to the delivery of a restart signal. In that case, + everything is cool. We have been removed from the queue + by the other thread, and consumed its signal. + + Otherwise we this thread woke up spontaneously, or due to a signal other + than restart. The next thing to do is to try to remove the thread + from the queue. This may fail due to a race against another thread + trying to do the same. In the failed case, we know we were signalled, + and we may also have to consume a restart signal. */ + + if (!was_signalled) { + int was_on_queue; + + /* __pthread_lock will queue back any spurious restarts that + may happen to it. */ + + __pthread_lock(&cond->__c_lock, self); + was_on_queue = remove_from_queue(&cond->__c_waiting, self); + __pthread_unlock(&cond->__c_lock); + + if (was_on_queue) { + __pthread_set_own_extricate_if(self, 0); + pthread_mutex_lock(mutex); + return ETIMEDOUT; + } + + /* Eat the outstanding restart() from the signaller */ + suspend(self); + } + + __pthread_set_own_extricate_if(self, 0); + + /* The remaining logic is the same as in other cancellable waits, + such as pthread_join sem_wait or pthread_cond wait. */ + + if (THREAD_GETMEM(self, p_woken_by_cancel) + && THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE) { + THREAD_SETMEM(self, p_woken_by_cancel, 0); + pthread_mutex_lock(mutex); + pthread_exit(PTHREAD_CANCELED); + } + + pthread_mutex_lock(mutex); + return 0; +} + +int pthread_cond_timedwait(pthread_cond_t *cond, pthread_mutex_t *mutex, + const struct timespec * abstime) +{ + /* Indirect call through pointer! */ + return pthread_cond_tw_rel(cond, mutex, abstime); +} + +int pthread_cond_signal(pthread_cond_t *cond) +{ + pthread_descr th; + + __pthread_lock(&cond->__c_lock, NULL); + th = dequeue(&cond->__c_waiting); + __pthread_unlock(&cond->__c_lock); + if (th != NULL) restart(th); + return 0; +} + +int pthread_cond_broadcast(pthread_cond_t *cond) +{ + pthread_descr tosignal, th; + + __pthread_lock(&cond->__c_lock, NULL); + /* Copy the current state of the waiting queue and empty it */ + tosignal = cond->__c_waiting; + cond->__c_waiting = NULL; + __pthread_unlock(&cond->__c_lock); + /* Now signal each process in the queue */ + while ((th = dequeue(&tosignal)) != NULL) restart(th); + return 0; +} + +int pthread_condattr_init(pthread_condattr_t *attr) +{ + return 0; +} + +int pthread_condattr_destroy(pthread_condattr_t *attr) +{ + return 0; +} diff --git a/libpthread/linuxthreads/configure b/libpthread/linuxthreads/configure new file mode 100644 index 000000000..3eafc93f5 --- /dev/null +++ b/libpthread/linuxthreads/configure @@ -0,0 +1,5 @@ +# This is only to keep the GNU C library configure mechanism happy. +# +# Perhaps some day we need a real configuration script for different +# kernel versions or so. +exit 0 diff --git a/libpthread/linuxthreads/debug.h b/libpthread/linuxthreads/debug.h new file mode 100644 index 000000000..79acb23c0 --- /dev/null +++ b/libpthread/linuxthreads/debug.h @@ -0,0 +1,50 @@ +/**************************************************************************** +** +** NAME: +** debug.h +** +** DESCRIPTION: +** This header file defines the debug macros used in pthreads. To turn +** debugging on, add -DDEBUG_PT to CFLAGS. It was added to the original +** distribution of linuxthreads. +** +** This program is free software; you can redistribute it and/or +** modify it under the terms of the GNU Library General Public License +** as published by the Free Software Foundation; either version 2 +** of the License, or (at your option) any later version. +** +** This program is distributed in the hope that it will be useful, +** but WITHOUT ANY WARRANTY; without even the implied warranty of +** MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +** GNU Library General Public License for more details. +** +****************************************************************************/ + +#ifndef _PT_DEBUG_H +#define _PT_DEBUG_H + +/* include asserts for now */ +#define DO_ASSERT + +/* define the PDEBUG macro here */ +#undef PDEBUG +#ifdef DEBUG_PT +# define PDEBUG(fmt, args...) __pthread_message(__FUNCTION__": " fmt, ## args) +#else +# define PDEBUG(fmt, args...) /* debug switched off */ +#endif + +/* nothing; placeholder to disable a PDEBUG message but don't delete it */ +#undef PDEBUGG +#define PDEBUGG(fmt, args...) + +/* Define ASSERT to stop/warn. Should be void in production code */ +#undef ASSERT +#ifdef DO_ASSERT +# define ASSERT(x) if (!(x)) fprintf(stderr, "pt: assertion failed in %s:%i.\n",\ + __FILE__, __LINE__) +#else +# define ASSERT(x) +#endif + +#endif /* _PT_DEBUG_H */ diff --git a/libpthread/linuxthreads/errno.c b/libpthread/linuxthreads/errno.c new file mode 100644 index 000000000..ad43be47a --- /dev/null +++ b/libpthread/linuxthreads/errno.c @@ -0,0 +1,34 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* Define the location of errno for the remainder of the C library */ + +#define __FORCE_GLIBC +#include <features.h> +#include <errno.h> +#include <netdb.h> +#include "pthread.h" +#include "internals.h" + +int * __errno_location() +{ + pthread_descr self = thread_self(); + return THREAD_GETMEM (self, p_errnop); +} + +int * __h_errno_location() +{ + pthread_descr self = thread_self(); + return THREAD_GETMEM (self, p_h_errnop); +} diff --git a/libpthread/linuxthreads/events.c b/libpthread/linuxthreads/events.c new file mode 100644 index 000000000..e5be3d935 --- /dev/null +++ b/libpthread/linuxthreads/events.c @@ -0,0 +1,35 @@ +/* Event functions used while debugging. + Copyright (C) 1999 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +/* The functions contained here do nothing, they just return. */ + +void +__linuxthreads_create_event (void) +{ +} + +void +__linuxthreads_death_event (void) +{ +} + +void +__linuxthreads_reap_event (void) +{ +} diff --git a/libpthread/linuxthreads/internals.h b/libpthread/linuxthreads/internals.h new file mode 100644 index 000000000..933ccb5f3 --- /dev/null +++ b/libpthread/linuxthreads/internals.h @@ -0,0 +1,480 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +#ifndef _INTERNALS_H +#define _INTERNALS_H 1 + +/* Internal data structures */ + +/* Includes */ + +#include <bits/libc-tsd.h> /* for _LIBC_TSD_KEY_N */ +#include <limits.h> +#include <setjmp.h> +#include <signal.h> +#include <unistd.h> +#include <sys/types.h> +#include "pt-machine.h" +#include "semaphore.h" +#include "../linuxthreads_db/thread_dbP.h" + +#ifndef THREAD_GETMEM +# define THREAD_GETMEM(descr, member) descr->member +#endif +#ifndef THREAD_GETMEM_NC +# define THREAD_GETMEM_NC(descr, member) descr->member +#endif +#ifndef THREAD_SETMEM +# define THREAD_SETMEM(descr, member, value) descr->member = (value) +#endif +#ifndef THREAD_SETMEM_NC +# define THREAD_SETMEM_NC(descr, member, value) descr->member = (value) +#endif + +/* Arguments passed to thread creation routine */ + +struct pthread_start_args { + void * (*start_routine)(void *); /* function to run */ + void * arg; /* its argument */ + sigset_t mask; /* initial signal mask for thread */ + int schedpolicy; /* initial scheduling policy (if any) */ + struct sched_param schedparam; /* initial scheduling parameters (if any) */ +}; + + +/* We keep thread specific data in a special data structure, a two-level + array. The top-level array contains pointers to dynamically allocated + arrays of a certain number of data pointers. So we can implement a + sparse array. Each dynamic second-level array has + PTHREAD_KEY_2NDLEVEL_SIZE + entries. This value shouldn't be too large. */ +#define PTHREAD_KEY_2NDLEVEL_SIZE 32 + +/* We need to address PTHREAD_KEYS_MAX key with PTHREAD_KEY_2NDLEVEL_SIZE + keys in each subarray. */ +#define PTHREAD_KEY_1STLEVEL_SIZE \ + ((PTHREAD_KEYS_MAX + PTHREAD_KEY_2NDLEVEL_SIZE - 1) \ + / PTHREAD_KEY_2NDLEVEL_SIZE) + +typedef void (*destr_function)(void *); + +struct pthread_key_struct { + int in_use; /* already allocated? */ + destr_function destr; /* destruction routine */ +}; + + +#define PTHREAD_START_ARGS_INITIALIZER { NULL, NULL, {{0, }}, 0, { 0 } } + +/* The type of thread descriptors */ + +typedef struct _pthread_descr_struct * pthread_descr; + +/* Callback interface for removing the thread from waiting on an + object if it is cancelled while waiting or about to wait. + This hold a pointer to the object, and a pointer to a function + which ``extricates'' the thread from its enqueued state. + The function takes two arguments: pointer to the wait object, + and a pointer to the thread. It returns 1 if an extrication + actually occured, and hence the thread must also be signalled. + It returns 0 if the thread had already been extricated. */ + +typedef struct _pthread_extricate_struct { + void *pu_object; + int (*pu_extricate_func)(void *, pthread_descr); +} pthread_extricate_if; + +/* Atomic counter made possible by compare_and_swap */ + +struct pthread_atomic { + long p_count; + int p_spinlock; +}; + +/* Context info for read write locks. The pthread_rwlock_info structure + is information about a lock that has been read-locked by the thread + in whose list this structure appears. The pthread_rwlock_context + is embedded in the thread context and contains a pointer to the + head of the list of lock info structures, as well as a count of + read locks that are untracked, because no info structure could be + allocated for them. */ + +struct _pthread_rwlock_t; + +typedef struct _pthread_rwlock_info { + struct _pthread_rwlock_info *pr_next; + struct _pthread_rwlock_t *pr_lock; + int pr_lock_count; +} pthread_readlock_info; + +struct _pthread_descr_struct { + pthread_descr p_nextlive, p_prevlive; + /* Double chaining of active threads */ + pthread_descr p_nextwaiting; /* Next element in the queue holding the thr */ + pthread_descr p_nextlock; /* can be on a queue and waiting on a lock */ + pthread_t p_tid; /* Thread identifier */ + int p_pid; /* PID of Unix process */ + int p_priority; /* Thread priority (== 0 if not realtime) */ + struct _pthread_fastlock * p_lock; /* Spinlock for synchronized accesses */ + int p_signal; /* last signal received */ + sigjmp_buf * p_signal_jmp; /* where to siglongjmp on a signal or NULL */ + sigjmp_buf * p_cancel_jmp; /* where to siglongjmp on a cancel or NULL */ + char p_terminated; /* true if terminated e.g. by pthread_exit */ + char p_detached; /* true if detached */ + char p_exited; /* true if the assoc. process terminated */ + void * p_retval; /* placeholder for return value */ + int p_retcode; /* placeholder for return code */ + pthread_descr p_joining; /* thread joining on that thread or NULL */ + struct _pthread_cleanup_buffer * p_cleanup; /* cleanup functions */ + char p_cancelstate; /* cancellation state */ + char p_canceltype; /* cancellation type (deferred/async) */ + char p_canceled; /* cancellation request pending */ + int * p_errnop; /* pointer to used errno variable */ + int p_errno; /* error returned by last system call */ + int * p_h_errnop; /* pointer to used h_errno variable */ + int p_h_errno; /* error returned by last netdb function */ + char * p_in_sighandler; /* stack address of sighandler, or NULL */ + char p_sigwaiting; /* true if a sigwait() is in progress */ + struct pthread_start_args p_start_args; /* arguments for thread creation */ + void ** p_specific[PTHREAD_KEY_1STLEVEL_SIZE]; /* thread-specific data */ + void * p_libc_specific[_LIBC_TSD_KEY_N]; /* thread-specific data for libc */ + int p_userstack; /* nonzero if the user provided the stack */ + void *p_guardaddr; /* address of guard area or NULL */ + size_t p_guardsize; /* size of guard area */ + pthread_descr p_self; /* Pointer to this structure */ + int p_nr; /* Index of descriptor in __pthread_handles */ + int p_report_events; /* Nonzero if events must be reported. */ + td_eventbuf_t p_eventbuf; /* Data for event. */ + struct pthread_atomic p_resume_count; /* number of times restart() was + called on thread */ + char p_woken_by_cancel; /* cancellation performed wakeup */ + pthread_extricate_if *p_extricate; /* See above */ + pthread_readlock_info *p_readlock_list; /* List of readlock info structs */ + pthread_readlock_info *p_readlock_free; /* Free list of structs */ + int p_untracked_readlock_count; /* Readlocks not tracked by list */ + /* New elements must be added at the end. */ +} __attribute__ ((aligned(32))); /* We need to align the structure so that + doubles are aligned properly. This is 8 + bytes on MIPS and 16 bytes on MIPS64. + 32 bytes might give better cache + utilization. */ + +/* The type of thread handles. */ + +typedef struct pthread_handle_struct * pthread_handle; + +struct pthread_handle_struct { + struct _pthread_fastlock h_lock; /* Fast lock for sychronized access */ + pthread_descr h_descr; /* Thread descriptor or NULL if invalid */ + char * h_bottom; /* Lowest address in the stack thread */ +}; + +/* The type of messages sent to the thread manager thread */ + +struct pthread_request { + pthread_descr req_thread; /* Thread doing the request */ + enum { /* Request kind */ + REQ_CREATE, REQ_FREE, REQ_PROCESS_EXIT, REQ_MAIN_THREAD_EXIT, + REQ_POST, REQ_DEBUG + } req_kind; + union { /* Arguments for request */ + struct { /* For REQ_CREATE: */ + const pthread_attr_t * attr; /* thread attributes */ + void * (*fn)(void *); /* start function */ + void * arg; /* argument to start function */ + sigset_t mask; /* signal mask */ + } create; + struct { /* For REQ_FREE: */ + pthread_t thread_id; /* identifier of thread to free */ + } free; + struct { /* For REQ_PROCESS_EXIT: */ + int code; /* exit status */ + } exit; + void * post; /* For REQ_POST: the semaphore */ + } req_args; +}; + + +/* Signals used for suspend/restart and for cancellation notification. */ + +extern int __pthread_sig_restart; +extern int __pthread_sig_cancel; + +/* Signal used for interfacing with gdb */ + +extern int __pthread_sig_debug; + +/* Global array of thread handles, used for validating a thread id + and retrieving the corresponding thread descriptor. Also used for + mapping the available stack segments. */ + +extern struct pthread_handle_struct __pthread_handles[PTHREAD_THREADS_MAX]; + +/* Descriptor of the initial thread */ + +extern struct _pthread_descr_struct __pthread_initial_thread; + +/* Descriptor of the manager thread */ + +extern struct _pthread_descr_struct __pthread_manager_thread; + +/* Descriptor of the main thread */ + +extern pthread_descr __pthread_main_thread; + +/* Limit between the stack of the initial thread (above) and the + stacks of other threads (below). Aligned on a STACK_SIZE boundary. + Initially 0, meaning that the current thread is (by definition) + the initial thread. */ + +/* For non-MMU systems also remember to stack top of the initial thread. + * This is adapted when other stacks are malloc'ed since we don't know + * the bounds a-priori. -StS */ + +extern char *__pthread_initial_thread_bos; +#ifndef __UCLIBC_HAS_MMU__ +extern char *__pthread_initial_thread_tos; +#define NOMMU_INITIAL_THREAD_BOUNDS(tos,bos) if ((tos)>=__pthread_initial_thread_bos && (bos)<=__pthread_initial_thread_tos) __pthread_initial_thread_bos = (tos)+1 +#else +#define NOMMU_INITIAL_THREAD_BOUNDS(tos,bos) /* empty */ +#endif /* __UCLIBC_HAS_MMU__ */ + + +/* Indicate whether at least one thread has a user-defined stack (if 1), + or all threads have stacks supplied by LinuxThreads (if 0). */ + +extern int __pthread_nonstandard_stacks; + +/* File descriptor for sending requests to the thread manager. + Initially -1, meaning that __pthread_initialize_manager must be called. */ + +extern int __pthread_manager_request; + +/* Other end of the pipe for sending requests to the thread manager. */ + +extern int __pthread_manager_reader; + +/* Limits of the thread manager stack. */ + +extern char *__pthread_manager_thread_bos; +extern char *__pthread_manager_thread_tos; + +/* Pending request for a process-wide exit */ + +extern int __pthread_exit_requested, __pthread_exit_code; + +/* Set to 1 by gdb if we're debugging */ + +extern volatile int __pthread_threads_debug; + +/* Globally enabled events. */ +extern volatile td_thr_events_t __pthread_threads_events; + +/* Pointer to descriptor of thread with last event. */ +extern volatile pthread_descr __pthread_last_event; + +/* Return the handle corresponding to a thread id */ + +static inline pthread_handle thread_handle(pthread_t id) +{ + return &__pthread_handles[id % PTHREAD_THREADS_MAX]; +} + +/* Validate a thread handle. Must have acquired h->h_spinlock before. */ + +static inline int invalid_handle(pthread_handle h, pthread_t id) +{ + return h->h_descr == NULL || h->h_descr->p_tid != id; +} + +/* Fill in defaults left unspecified by pt-machine.h. */ + +/* The page size we can get from the system. This should likely not be + changed by the machine file but, you never know. */ +#ifndef PAGE_SIZE +#define PAGE_SIZE (sysconf (_SC_PAGE_SIZE)) +#endif + +/* The max size of the thread stack segments. If the default + THREAD_SELF implementation is used, this must be a power of two and + a multiple of PAGE_SIZE. */ +#ifndef STACK_SIZE +#define STACK_SIZE (2 * 1024 * 1024) +#endif + +/* The initial size of the thread stack. Must be a multiple of PAGE_SIZE. */ +#ifndef INITIAL_STACK_SIZE +#define INITIAL_STACK_SIZE (4 * PAGE_SIZE) +#endif + +/* Size of the thread manager stack. The "- 32" avoids wasting space + with some malloc() implementations. */ +#ifndef THREAD_MANAGER_STACK_SIZE +#define THREAD_MANAGER_STACK_SIZE (2 * PAGE_SIZE - 32) +#endif + +/* The base of the "array" of thread stacks. The array will grow down from + here. Defaults to the calculated bottom of the initial application + stack. */ +#ifndef THREAD_STACK_START_ADDRESS +#define THREAD_STACK_START_ADDRESS __pthread_initial_thread_bos +#endif + +/* Get some notion of the current stack. Need not be exactly the top + of the stack, just something somewhere in the current frame. */ +#ifndef CURRENT_STACK_FRAME +#define CURRENT_STACK_FRAME ({ char __csf; &__csf; }) +#endif + +/* Recover thread descriptor for the current thread */ + +extern pthread_descr __pthread_find_self (void) __attribute__ ((const)); + +static inline pthread_descr thread_self (void) __attribute__ ((const)); +static inline pthread_descr thread_self (void) +{ +#ifdef THREAD_SELF + return THREAD_SELF; +#else + char *sp = CURRENT_STACK_FRAME; +#ifdef __UCLIBC_HAS_MMU__ + if (sp >= __pthread_initial_thread_bos) + return &__pthread_initial_thread; + else if (sp >= __pthread_manager_thread_bos + && sp < __pthread_manager_thread_tos) + return &__pthread_manager_thread; + else if (__pthread_nonstandard_stacks) + return __pthread_find_self(); + else + return (pthread_descr)(((unsigned long)sp | (STACK_SIZE-1))+1) - 1; +#else + /* For non-MMU we need to be more careful about the initial thread stack. + * We refine the initial thread stack bounds dynamically as we allocate + * the other stack frame such that it doesn't overlap with them. Then + * we can be sure to pick the right thread according to the current SP */ + + /* Since we allow other stack frames to be above or below, we need to + * treat this case special. When pthread_initialize() wasn't called yet, + * only the initial thread is there. */ + if (__pthread_initial_thread_bos == NULL) { + return &__pthread_initial_thread; + } + else if (sp >= __pthread_initial_thread_bos + && sp < __pthread_initial_thread_tos) { + return &__pthread_initial_thread; + } + else if (sp >= __pthread_manager_thread_bos + && sp < __pthread_manager_thread_tos) { + return &__pthread_manager_thread; + } + else { + return __pthread_find_self(); + } +#endif /* __UCLIBC_HAS_MMU__ */ +#endif +} + +/* Max number of times we must spin on a spinlock calling sched_yield(). + After MAX_SPIN_COUNT iterations, we put the calling thread to sleep. */ + +#ifndef MAX_SPIN_COUNT +#define MAX_SPIN_COUNT 50 +#endif + +/* Duration of sleep (in nanoseconds) when we can't acquire a spinlock + after MAX_SPIN_COUNT iterations of sched_yield(). + With the 2.0 and 2.1 kernels, this MUST BE > 2ms. + (Otherwise the kernel does busy-waiting for realtime threads, + giving other threads no chance to run.) */ + +#ifndef SPIN_SLEEP_DURATION +#define SPIN_SLEEP_DURATION 2000001 +#endif + +/* Debugging */ + +#ifdef DEBUG +#include <assert.h> +#define ASSERT assert +#define MSG __pthread_message +#else +#define ASSERT(x) +#define MSG(msg,arg...) +#endif + +/* Internal global functions */ + +void __pthread_destroy_specifics(void); +void __pthread_perform_cleanup(void); +int __pthread_initialize_manager(void); +void __pthread_message(char * fmt, ...); +int __pthread_manager(void *reqfd); +int __pthread_manager_event(void *reqfd); +void __pthread_manager_sighandler(int sig); +void __pthread_reset_main_thread(void); +void __fresetlockfiles(void); +void __pthread_manager_adjust_prio(int thread_prio); +void __pthread_set_own_extricate_if(pthread_descr self, pthread_extricate_if *peif); + +extern int __pthread_attr_setguardsize __P ((pthread_attr_t *__attr, + size_t __guardsize)); +extern int __pthread_attr_getguardsize __P ((__const pthread_attr_t *__attr, + size_t *__guardsize)); +extern int __pthread_attr_setstackaddr __P ((pthread_attr_t *__attr, + void *__stackaddr)); +extern int __pthread_attr_getstackaddr __P ((__const pthread_attr_t *__attr, + void **__stackaddr)); +extern int __pthread_attr_setstacksize __P ((pthread_attr_t *__attr, + size_t __stacksize)); +extern int __pthread_attr_getstacksize __P ((__const pthread_attr_t *__attr, + size_t *__stacksize)); +extern int __pthread_getconcurrency __P ((void)); +extern int __pthread_setconcurrency __P ((int __level)); +extern int __pthread_mutexattr_gettype __P ((__const pthread_mutexattr_t *__attr, + int *__kind)); +extern void __pthread_kill_other_threads_np __P ((void)); + +void __pthread_restart_old(pthread_descr th); +void __pthread_suspend_old(pthread_descr self); + +void __pthread_restart_new(pthread_descr th); +void __pthread_suspend_new(pthread_descr self); + +void __pthread_wait_for_restart_signal(pthread_descr self); + +void __pthread_init_condvar(int rt_sig_available); + +/* Global pointers to old or new suspend functions */ + +extern void (*__pthread_restart)(pthread_descr); +extern void (*__pthread_suspend)(pthread_descr); + +/* Prototypes for the function without cancelation support when the + normal version has it. */ +extern int __libc_close (int fd); +extern int __libc_nanosleep (const struct timespec *requested_time, + struct timespec *remaining); +extern ssize_t __libc_read (int fd, void *buf, size_t count); +extern pid_t __libc_waitpid (pid_t pid, int *stat_loc, int options); +extern ssize_t __libc_write (int fd, const void *buf, size_t count); + +/* Prototypes for some of the new semaphore functions. */ +extern int __new_sem_post (sem_t * sem); + +/* The functions called the signal events. */ +extern void __linuxthreads_create_event (void); +extern void __linuxthreads_death_event (void); +extern void __linuxthreads_reap_event (void); + +#endif /* internals.h */ diff --git a/libpthread/linuxthreads/join.c b/libpthread/linuxthreads/join.c new file mode 100644 index 000000000..ccb11b124 --- /dev/null +++ b/libpthread/linuxthreads/join.c @@ -0,0 +1,213 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* Thread termination and joining */ + +#include <errno.h> +#include <sched.h> +#include <unistd.h> +#include "pthread.h" +#include "internals.h" +#include "spinlock.h" +#include "restart.h" +#include "debug.h" /* PDEBUG, added by StS */ + +void pthread_exit(void * retval) +{ + pthread_descr self = thread_self(); + pthread_descr joining; + struct pthread_request request; +PDEBUG("self=%p, pid=%d\n", self, self->p_pid); + + /* Reset the cancellation flag to avoid looping if the cleanup handlers + contain cancellation points */ + THREAD_SETMEM(self, p_canceled, 0); + /* Call cleanup functions and destroy the thread-specific data */ + __pthread_perform_cleanup(); + __pthread_destroy_specifics(); + /* Store return value */ + __pthread_lock(THREAD_GETMEM(self, p_lock), self); + THREAD_SETMEM(self, p_retval, retval); + /* Say that we've terminated */ + THREAD_SETMEM(self, p_terminated, 1); + /* See whether we have to signal the death. */ + if (THREAD_GETMEM(self, p_report_events)) + { + /* See whether TD_DEATH is in any of the mask. */ + int idx = __td_eventword (TD_DEATH); + uint32_t mask = __td_eventmask (TD_DEATH); + + if ((mask & (__pthread_threads_events.event_bits[idx] + | THREAD_GETMEM(self, + p_eventbuf.eventmask).event_bits[idx])) + != 0) + { + /* Yep, we have to signal the death. */ + THREAD_SETMEM(self, p_eventbuf.eventnum, TD_DEATH); + THREAD_SETMEM(self, p_eventbuf.eventdata, self); + __pthread_last_event = self; + + /* Now call the function to signal the event. */ + __linuxthreads_death_event(); + } + } + /* See if someone is joining on us */ + joining = THREAD_GETMEM(self, p_joining); +PDEBUG("joining = %p, pid=%d\n", joining, joining->p_pid); + __pthread_unlock(THREAD_GETMEM(self, p_lock)); + /* Restart joining thread if any */ + if (joining != NULL) restart(joining); + /* If this is the initial thread, block until all threads have terminated. + If another thread calls exit, we'll be terminated from our signal + handler. */ + if (self == __pthread_main_thread && __pthread_manager_request >= 0) { + request.req_thread = self; + request.req_kind = REQ_MAIN_THREAD_EXIT; + __libc_write(__pthread_manager_request, (char *)&request, sizeof(request)); + suspend(self); + } + /* Exit the process (but don't flush stdio streams, and don't run + atexit functions). */ + _exit(0); +} + +/* Function called by pthread_cancel to remove the thread from + waiting on a condition variable queue. */ + +static int join_extricate_func(void *obj, pthread_descr th) +{ + volatile pthread_descr self = thread_self(); + pthread_handle handle = obj; + pthread_descr jo; + int did_remove = 0; + + __pthread_lock(&handle->h_lock, self); + jo = handle->h_descr; + did_remove = jo->p_joining != NULL; + jo->p_joining = NULL; + __pthread_unlock(&handle->h_lock); + + return did_remove; +} + +int pthread_join(pthread_t thread_id, void ** thread_return) +{ + volatile pthread_descr self = thread_self(); + struct pthread_request request; + pthread_handle handle = thread_handle(thread_id); + pthread_descr th; + pthread_extricate_if extr; + int already_canceled = 0; +PDEBUG("\n"); + + /* Set up extrication interface */ + extr.pu_object = handle; + extr.pu_extricate_func = join_extricate_func; + + __pthread_lock(&handle->h_lock, self); + if (invalid_handle(handle, thread_id)) { + __pthread_unlock(&handle->h_lock); + return ESRCH; + } + th = handle->h_descr; + if (th == self) { + __pthread_unlock(&handle->h_lock); + return EDEADLK; + } + /* If detached or already joined, error */ + if (th->p_detached || th->p_joining != NULL) { + __pthread_unlock(&handle->h_lock); + return EINVAL; + } + /* If not terminated yet, suspend ourselves. */ + if (! th->p_terminated) { + /* Register extrication interface */ + __pthread_set_own_extricate_if(self, &extr); + if (!(THREAD_GETMEM(self, p_canceled) + && THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE)) + th->p_joining = self; + else + already_canceled = 1; + __pthread_unlock(&handle->h_lock); + + if (already_canceled) { + __pthread_set_own_extricate_if(self, 0); + pthread_exit(PTHREAD_CANCELED); + } + +PDEBUG("before suspend\n"); + suspend(self); +PDEBUG("after suspend\n"); + /* Deregister extrication interface */ + __pthread_set_own_extricate_if(self, 0); + + /* This is a cancellation point */ + if (THREAD_GETMEM(self, p_woken_by_cancel) + && THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE) { + THREAD_SETMEM(self, p_woken_by_cancel, 0); + pthread_exit(PTHREAD_CANCELED); + } + __pthread_lock(&handle->h_lock, self); + } + /* Get return value */ + if (thread_return != NULL) *thread_return = th->p_retval; + __pthread_unlock(&handle->h_lock); + /* Send notification to thread manager */ + if (__pthread_manager_request >= 0) { + request.req_thread = self; + request.req_kind = REQ_FREE; + request.req_args.free.thread_id = thread_id; + __libc_write(__pthread_manager_request, + (char *) &request, sizeof(request)); + } + return 0; +} + +int pthread_detach(pthread_t thread_id) +{ + int terminated; + struct pthread_request request; + pthread_handle handle = thread_handle(thread_id); + pthread_descr th; + + __pthread_lock(&handle->h_lock, NULL); + if (invalid_handle(handle, thread_id)) { + __pthread_unlock(&handle->h_lock); + return ESRCH; + } + th = handle->h_descr; + /* If already detached, error */ + if (th->p_detached) { + __pthread_unlock(&handle->h_lock); + return EINVAL; + } + /* If already joining, don't do anything. */ + if (th->p_joining != NULL) { + __pthread_unlock(&handle->h_lock); + return 0; + } + /* Mark as detached */ + th->p_detached = 1; + terminated = th->p_terminated; + __pthread_unlock(&handle->h_lock); + /* If already terminated, notify thread manager to reclaim resources */ + if (terminated && __pthread_manager_request >= 0) { + request.req_thread = thread_self(); + request.req_kind = REQ_FREE; + request.req_args.free.thread_id = thread_id; + __libc_write(__pthread_manager_request, + (char *) &request, sizeof(request)); + } + return 0; +} diff --git a/libpthread/linuxthreads/linuxthreads.texi b/libpthread/linuxthreads/linuxthreads.texi new file mode 100644 index 000000000..7a98103b3 --- /dev/null +++ b/libpthread/linuxthreads/linuxthreads.texi @@ -0,0 +1,1428 @@ +@node POSIX Threads +@c @node POSIX Threads, , Top, Top +@chapter POSIX Threads +@c %MENU% The standard threads library + +@c This chapter needs more work bigtime. -zw + +This chapter describes the pthreads (POSIX threads) library. This +library provides support functions for multithreaded programs: thread +primitives, synchronization objects, and so forth. It also implements +POSIX 1003.1b semaphores (not to be confused with System V semaphores). + +The threads operations (@samp{pthread_*}) do not use @var{errno}. +Instead they return an error code directly. The semaphore operations do +use @var{errno}. + +@menu +* Basic Thread Operations:: Creating, terminating, and waiting for threads. +* Thread Attributes:: Tuning thread scheduling. +* Cancellation:: Stopping a thread before it's done. +* Cleanup Handlers:: Deallocating resources when a thread is + cancelled. +* Mutexes:: One way to synchronize threads. +* Condition Variables:: Another way. +* POSIX Semaphores:: And a third way. +* Thread-Specific Data:: Variables with different values in + different threads. +* Threads and Signal Handling:: Why you should avoid mixing the two, and + how to do it if you must. +* Miscellaneous Thread Functions:: A grab bag of utility routines. +@end menu + +@node Basic Thread Operations +@section Basic Thread Operations + +These functions are the thread equivalents of @code{fork}, @code{exit}, +and @code{wait}. + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_create (pthread_t * @var{thread}, pthread_attr_t * @var{attr}, void * (*@var{start_routine})(void *), void * @var{arg}) +@code{pthread_create} creates a new thread of control that executes +concurrently with the calling thread. The new thread calls the +function @var{start_routine}, passing it @var{arg} as first argument. The +new thread terminates either explicitly, by calling @code{pthread_exit}, +or implicitly, by returning from the @var{start_routine} function. The +latter case is equivalent to calling @code{pthread_exit} with the result +returned by @var{start_routine} as exit code. + +The @var{attr} argument specifies thread attributes to be applied to the +new thread. @xref{Thread Attributes}, for details. The @var{attr} +argument can also be @code{NULL}, in which case default attributes are +used: the created thread is joinable (not detached) and has an ordinary +(not realtime) scheduling policy. + +On success, the identifier of the newly created thread is stored in the +location pointed by the @var{thread} argument, and a 0 is returned. On +error, a non-zero error code is returned. + +This function may return the following errors: +@table @code +@item EAGAIN +Not enough system resources to create a process for the new thread, +or more than @code{PTHREAD_THREADS_MAX} threads are already active. +@end table +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun void pthread_exit (void *@var{retval}) +@code{pthread_exit} terminates the execution of the calling thread. All +cleanup handlers (@pxref{Cleanup Handlers}) that have been set for the +calling thread with @code{pthread_cleanup_push} are executed in reverse +order (the most recently pushed handler is executed first). Finalization +functions for thread-specific data are then called for all keys that +have non-@code{NULL} values associated with them in the calling thread +(@pxref{Thread-Specific Data}). Finally, execution of the calling +thread is stopped. + +The @var{retval} argument is the return value of the thread. It can be +retrieved from another thread using @code{pthread_join}. + +The @code{pthread_exit} function never returns. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_cancel (pthread_t @var{thread}) + +@code{pthread_cancel} sends a cancellation request to the thread denoted +by the @var{thread} argument. If there is no such thread, +@code{pthread_cancel} fails and returns @code{ESRCH}. Otherwise it +returns 0. @xref{Cancellation}, for details. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_join (pthread_t @var{th}, void **thread_@var{return}) +@code{pthread_join} suspends the execution of the calling thread until +the thread identified by @var{th} terminates, either by calling +@code{pthread_exit} or by being cancelled. + +If @var{thread_return} is not @code{NULL}, the return value of @var{th} +is stored in the location pointed to by @var{thread_return}. The return +value of @var{th} is either the argument it gave to @code{pthread_exit}, +or @code{PTHREAD_CANCELED} if @var{th} was cancelled. + +The joined thread @code{th} must be in the joinable state: it must not +have been detached using @code{pthread_detach} or the +@code{PTHREAD_CREATE_DETACHED} attribute to @code{pthread_create}. + +When a joinable thread terminates, its memory resources (thread +descriptor and stack) are not deallocated until another thread performs +@code{pthread_join} on it. Therefore, @code{pthread_join} must be called +once for each joinable thread created to avoid memory leaks. + +At most one thread can wait for the termination of a given +thread. Calling @code{pthread_join} on a thread @var{th} on which +another thread is already waiting for termination returns an error. + +@code{pthread_join} is a cancellation point. If a thread is canceled +while suspended in @code{pthread_join}, the thread execution resumes +immediately and the cancellation is executed without waiting for the +@var{th} thread to terminate. If cancellation occurs during +@code{pthread_join}, the @var{th} thread remains not joined. + +On success, the return value of @var{th} is stored in the location +pointed to by @var{thread_return}, and 0 is returned. On error, one of +the following values is returned: +@table @code +@item ESRCH +No thread could be found corresponding to that specified by @var{th}. +@item EINVAL +The @var{th} thread has been detached, or another thread is already +waiting on termination of @var{th}. +@item EDEADLK +The @var{th} argument refers to the calling thread. +@end table +@end deftypefun + +@node Thread Attributes +@section Thread Attributes + +@comment pthread.h +@comment POSIX + +Threads have a number of attributes that may be set at creation time. +This is done by filling a thread attribute object @var{attr} of type +@code{pthread_attr_t}, then passing it as second argument to +@code{pthread_create}. Passing @code{NULL} is equivalent to passing a +thread attribute object with all attributes set to their default values. + +Attribute objects are consulted only when creating a new thread. The +same attribute object can be used for creating several threads. +Modifying an attribute object after a call to @code{pthread_create} does +not change the attributes of the thread previously created. + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_attr_init (pthread_attr_t *@var{attr}) +@code{pthread_attr_init} initializes the thread attribute object +@var{attr} and fills it with default values for the attributes. (The +default values are listed below for each attribute.) + +Each attribute @var{attrname} (see below for a list of all attributes) +can be individually set using the function +@code{pthread_attr_set@var{attrname}} and retrieved using the function +@code{pthread_attr_get@var{attrname}}. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_attr_destroy (pthread_attr_t *@var{attr}) +@code{pthread_attr_destroy} destroys the attribute object pointed to by +@var{attr} releasing any resources associated with it. @var{attr} is +left in an undefined state, and you must not use it again in a call to +any pthreads function until it has been reinitialized. +@end deftypefun + +@findex pthread_attr_setinheritsched +@findex pthread_attr_setschedparam +@findex pthread_attr_setschedpolicy +@findex pthread_attr_setscope +@comment pthread.h +@comment POSIX +@deftypefun int pthread_attr_set@var{attr} (pthread_attr_t *@var{obj}, int @var{value}) +Set attribute @var{attr} to @var{value} in the attribute object pointed +to by @var{obj}. See below for a list of possible attributes and the +values they can take. + +On success, these functions return 0. If @var{value} is not meaningful +for the @var{attr} being modified, they will return the error code +@code{EINVAL}. Some of the functions have other failure modes; see +below. +@end deftypefun + +@findex pthread_attr_getinheritsched +@findex pthread_attr_getschedparam +@findex pthread_attr_getschedpolicy +@findex pthread_attr_getscope +@comment pthread.h +@comment POSIX +@deftypefun int pthread_attr_get@var{attr} (const pthread_attr_t *@var{obj}, int *@var{value}) +Store the current setting of @var{attr} in @var{obj} into the variable +pointed to by @var{value}. + +These functions always return 0. +@end deftypefun + +The following thread attributes are supported: +@table @samp +@item detachstate +Choose whether the thread is created in the joinable state (value +@code{PTHREAD_CREATE_JOINABLE}) or in the detached state +(@code{PTHREAD_CREATE_DETACHED}). The default is +@code{PTHREAD_CREATE_JOINABLE}. + +In the joinable state, another thread can synchronize on the thread +termination and recover its termination code using @code{pthread_join}, +but some of the thread resources are kept allocated after the thread +terminates, and reclaimed only when another thread performs +@code{pthread_join} on that thread. + +In the detached state, the thread resources are immediately freed when +it terminates, but @code{pthread_join} cannot be used to synchronize on +the thread termination. + +A thread created in the joinable state can later be put in the detached +thread using @code{pthread_detach}. + +@item schedpolicy +Select the scheduling policy for the thread: one of @code{SCHED_OTHER} +(regular, non-realtime scheduling), @code{SCHED_RR} (realtime, +round-robin) or @code{SCHED_FIFO} (realtime, first-in first-out). +The default is @code{SCHED_OTHER}. +@c Not doc'd in our manual: FIXME. +@c See @code{sched_setpolicy} for more information on scheduling policies. + +The realtime scheduling policies @code{SCHED_RR} and @code{SCHED_FIFO} +are available only to processes with superuser privileges. +@code{pthread_attr_setschedparam} will fail and return @code{ENOTSUP} if +you try to set a realtime policy when you are unprivileged. + +The scheduling policy of a thread can be changed after creation with +@code{pthread_setschedparam}. + +@item schedparam +Change the scheduling parameter (the scheduling priority) +for the thread. The default is 0. + +This attribute is not significant if the scheduling policy is +@code{SCHED_OTHER}; it only matters for the realtime policies +@code{SCHED_RR} and @code{SCHED_FIFO}. + +The scheduling priority of a thread can be changed after creation with +@code{pthread_setschedparam}. + +@item inheritsched +Choose whether the scheduling policy and scheduling parameter for the +newly created thread are determined by the values of the +@var{schedpolicy} and @var{schedparam} attributes (value +@code{PTHREAD_EXPLICIT_SCHED}) or are inherited from the parent thread +(value @code{PTHREAD_INHERIT_SCHED}). The default is +@code{PTHREAD_EXPLICIT_SCHED}. + +@item scope +Choose the scheduling contention scope for the created thread. The +default is @code{PTHREAD_SCOPE_SYSTEM}, meaning that the threads contend +for CPU time with all processes running on the machine. In particular, +thread priorities are interpreted relative to the priorities of all +other processes on the machine. The other possibility, +@code{PTHREAD_SCOPE_PROCESS}, means that scheduling contention occurs +only between the threads of the running process: thread priorities are +interpreted relative to the priorities of the other threads of the +process, regardless of the priorities of other processes. + +@code{PTHREAD_SCOPE_PROCESS} is not supported in LinuxThreads. If you +try to set the scope to this value @code{pthread_attr_setscope} will +fail and return @code{ENOTSUP}. +@end table + +@node Cancellation +@section Cancellation + +Cancellation is the mechanism by which a thread can terminate the +execution of another thread. More precisely, a thread can send a +cancellation request to another thread. Depending on its settings, the +target thread can then either ignore the request, honor it immediately, +or defer it till it reaches a cancellation point. When threads are +first created by @code{pthread_create}, they always defer cancellation +requests. + +When a thread eventually honors a cancellation request, it behaves as if +@code{pthread_exit(PTHREAD_CANCELED)} was called. All cleanup handlers +are executed in reverse order, finalization functions for +thread-specific data are called, and finally the thread stops executing. +If the cancelled thread was joinable, the return value +@code{PTHREAD_CANCELED} is provided to whichever thread calls +@var{pthread_join} on it. See @code{pthread_exit} for more information. + +Cancellation points are the points where the thread checks for pending +cancellation requests and performs them. The POSIX threads functions +@code{pthread_join}, @code{pthread_cond_wait}, +@code{pthread_cond_timedwait}, @code{pthread_testcancel}, +@code{sem_wait}, and @code{sigwait} are cancellation points. In +addition, these system calls are cancellation points: + +@multitable @columnfractions .33 .33 .33 +@item @t{accept} @tab @t{open} @tab @t{sendmsg} +@item @t{close} @tab @t{pause} @tab @t{sendto} +@item @t{connect} @tab @t{read} @tab @t{system} +@item @t{fcntl} @tab @t{recv} @tab @t{tcdrain} +@item @t{fsync} @tab @t{recvfrom} @tab @t{wait} +@item @t{lseek} @tab @t{recvmsg} @tab @t{waitpid} +@item @t{msync} @tab @t{send} @tab @t{write} +@item @t{nanosleep} +@end multitable + +@noindent +All library functions that call these functions (such as +@code{printf}) are also cancellation points. + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_setcancelstate (int @var{state}, int *@var{oldstate}) +@code{pthread_setcancelstate} changes the cancellation state for the +calling thread -- that is, whether cancellation requests are ignored or +not. The @var{state} argument is the new cancellation state: either +@code{PTHREAD_CANCEL_ENABLE} to enable cancellation, or +@code{PTHREAD_CANCEL_DISABLE} to disable cancellation (cancellation +requests are ignored). + +If @var{oldstate} is not @code{NULL}, the previous cancellation state is +stored in the location pointed to by @var{oldstate}, and can thus be +restored later by another call to @code{pthread_setcancelstate}. + +If the @var{state} argument is not @code{PTHREAD_CANCEL_ENABLE} or +@code{PTHREAD_CANCEL_DISABLE}, @code{pthread_setcancelstate} fails and +returns @code{EINVAL}. Otherwise it returns 0. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_setcanceltype (int @var{type}, int *@var{oldtype}) +@code{pthread_setcanceltype} changes the type of responses to +cancellation requests for the calling thread: asynchronous (immediate) +or deferred. The @var{type} argument is the new cancellation type: +either @code{PTHREAD_CANCEL_ASYNCHRONOUS} to cancel the calling thread +as soon as the cancellation request is received, or +@code{PTHREAD_CANCEL_DEFERRED} to keep the cancellation request pending +until the next cancellation point. If @var{oldtype} is not @code{NULL}, +the previous cancellation state is stored in the location pointed to by +@var{oldtype}, and can thus be restored later by another call to +@code{pthread_setcanceltype}. + +If the @var{type} argument is not @code{PTHREAD_CANCEL_DEFERRED} or +@code{PTHREAD_CANCEL_ASYNCHRONOUS}, @code{pthread_setcanceltype} fails +and returns @code{EINVAL}. Otherwise it returns 0. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun void pthread_testcancel (@var{void}) +@code{pthread_testcancel} does nothing except testing for pending +cancellation and executing it. Its purpose is to introduce explicit +checks for cancellation in long sequences of code that do not call +cancellation point functions otherwise. +@end deftypefun + +@node Cleanup Handlers +@section Cleanup Handlers + +Cleanup handlers are functions that get called when a thread terminates, +either by calling @code{pthread_exit} or because of +cancellation. Cleanup handlers are installed and removed following a +stack-like discipline. + +The purpose of cleanup handlers is to free the resources that a thread +may hold at the time it terminates. In particular, if a thread exits or +is cancelled while it owns a locked mutex, the mutex will remain locked +forever and prevent other threads from executing normally. The best way +to avoid this is, just before locking the mutex, to install a cleanup +handler whose effect is to unlock the mutex. Cleanup handlers can be +used similarly to free blocks allocated with @code{malloc} or close file +descriptors on thread termination. + +Here is how to lock a mutex @var{mut} in such a way that it will be +unlocked if the thread is canceled while @var{mut} is locked: + +@smallexample +pthread_cleanup_push(pthread_mutex_unlock, (void *) &mut); +pthread_mutex_lock(&mut); +/* do some work */ +pthread_mutex_unlock(&mut); +pthread_cleanup_pop(0); +@end smallexample + +Equivalently, the last two lines can be replaced by + +@smallexample +pthread_cleanup_pop(1); +@end smallexample + +Notice that the code above is safe only in deferred cancellation mode +(see @code{pthread_setcanceltype}). In asynchronous cancellation mode, a +cancellation can occur between @code{pthread_cleanup_push} and +@code{pthread_mutex_lock}, or between @code{pthread_mutex_unlock} and +@code{pthread_cleanup_pop}, resulting in both cases in the thread trying +to unlock a mutex not locked by the current thread. This is the main +reason why asynchronous cancellation is difficult to use. + +If the code above must also work in asynchronous cancellation mode, +then it must switch to deferred mode for locking and unlocking the +mutex: + +@smallexample +pthread_setcanceltype(PTHREAD_CANCEL_DEFERRED, &oldtype); +pthread_cleanup_push(pthread_mutex_unlock, (void *) &mut); +pthread_mutex_lock(&mut); +/* do some work */ +pthread_cleanup_pop(1); +pthread_setcanceltype(oldtype, NULL); +@end smallexample + +The code above can be rewritten in a more compact and efficient way, +using the non-portable functions @code{pthread_cleanup_push_defer_np} +and @code{pthread_cleanup_pop_restore_np}: + +@smallexample +pthread_cleanup_push_defer_np(pthread_mutex_unlock, (void *) &mut); +pthread_mutex_lock(&mut); +/* do some work */ +pthread_cleanup_pop_restore_np(1); +@end smallexample + +@comment pthread.h +@comment POSIX +@deftypefun void pthread_cleanup_push (void (*@var{routine}) (void *), void *@var{arg}) + +@code{pthread_cleanup_push} installs the @var{routine} function with +argument @var{arg} as a cleanup handler. From this point on to the +matching @code{pthread_cleanup_pop}, the function @var{routine} will be +called with arguments @var{arg} when the thread terminates, either +through @code{pthread_exit} or by cancellation. If several cleanup +handlers are active at that point, they are called in LIFO order: the +most recently installed handler is called first. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun void pthread_cleanup_pop (int @var{execute}) +@code{pthread_cleanup_pop} removes the most recently installed cleanup +handler. If the @var{execute} argument is not 0, it also executes the +handler, by calling the @var{routine} function with arguments +@var{arg}. If the @var{execute} argument is 0, the handler is only +removed but not executed. +@end deftypefun + +Matching pairs of @code{pthread_cleanup_push} and +@code{pthread_cleanup_pop} must occur in the same function, at the same +level of block nesting. Actually, @code{pthread_cleanup_push} and +@code{pthread_cleanup_pop} are macros, and the expansion of +@code{pthread_cleanup_push} introduces an open brace @code{@{} with the +matching closing brace @code{@}} being introduced by the expansion of the +matching @code{pthread_cleanup_pop}. + +@comment pthread.h +@comment GNU +@deftypefun void pthread_cleanup_push_defer_np (void (*@var{routine}) (void *), void *@var{arg}) +@code{pthread_cleanup_push_defer_np} is a non-portable extension that +combines @code{pthread_cleanup_push} and @code{pthread_setcanceltype}. +It pushes a cleanup handler just as @code{pthread_cleanup_push} does, +but also saves the current cancellation type and sets it to deferred +cancellation. This ensures that the cleanup mechanism is effective even +if the thread was initially in asynchronous cancellation mode. +@end deftypefun + +@comment pthread.h +@comment GNU +@deftypefun void pthread_cleanup_pop_restore_np (int @var{execute}) +@code{pthread_cleanup_pop_restore_np} pops a cleanup handler introduced +by @code{pthread_cleanup_push_defer_np}, and restores the cancellation +type to its value at the time @code{pthread_cleanup_push_defer_np} was +called. +@end deftypefun + +@code{pthread_cleanup_push_defer_np} and +@code{pthread_cleanup_pop_restore_np} must occur in matching pairs, at +the same level of block nesting. + +The sequence + +@smallexample +pthread_cleanup_push_defer_np(routine, arg); +... +pthread_cleanup_pop_defer_np(execute); +@end smallexample + +@noindent +is functionally equivalent to (but more compact and efficient than) + +@smallexample +@{ + int oldtype; + pthread_setcanceltype(PTHREAD_CANCEL_DEFERRED, &oldtype); + pthread_cleanup_push(routine, arg); + ... + pthread_cleanup_pop(execute); + pthread_setcanceltype(oldtype, NULL); +@} +@end smallexample + + +@node Mutexes +@section Mutexes + +A mutex is a MUTual EXclusion device, and is useful for protecting +shared data structures from concurrent modifications, and implementing +critical sections and monitors. + +A mutex has two possible states: unlocked (not owned by any thread), +and locked (owned by one thread). A mutex can never be owned by two +different threads simultaneously. A thread attempting to lock a mutex +that is already locked by another thread is suspended until the owning +thread unlocks the mutex first. + +None of the mutex functions is a cancellation point, not even +@code{pthread_mutex_lock}, in spite of the fact that it can suspend a +thread for arbitrary durations. This way, the status of mutexes at +cancellation points is predictable, allowing cancellation handlers to +unlock precisely those mutexes that need to be unlocked before the +thread stops executing. Consequently, threads using deferred +cancellation should never hold a mutex for extended periods of time. + +It is not safe to call mutex functions from a signal handler. In +particular, calling @code{pthread_mutex_lock} or +@code{pthread_mutex_unlock} from a signal handler may deadlock the +calling thread. + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_mutex_init (pthread_mutex_t *@var{mutex}, const pthread_mutexattr_t *@var{mutexattr}) + +@code{pthread_mutex_init} initializes the mutex object pointed to by +@var{mutex} according to the mutex attributes specified in @var{mutexattr}. +If @var{mutexattr} is @code{NULL}, default attributes are used instead. + +The LinuxThreads implementation supports only one mutex attribute, +the @var{mutex kind}, which is either ``fast'', ``recursive'', or +``error checking''. The kind of a mutex determines whether +it can be locked again by a thread that already owns it. +The default kind is ``fast''. + +Variables of type @code{pthread_mutex_t} can also be initialized +statically, using the constants @code{PTHREAD_MUTEX_INITIALIZER} (for +fast mutexes), @code{PTHREAD_RECURSIVE_MUTEX_INITIALIZER_NP} (for +recursive mutexes), and @code{PTHREAD_ERRORCHECK_MUTEX_INITIALIZER_NP} +(for error checking mutexes). + +@code{pthread_mutex_init} always returns 0. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_mutex_lock (pthread_mutex_t *mutex)) +@code{pthread_mutex_lock} locks the given mutex. If the mutex is +currently unlocked, it becomes locked and owned by the calling thread, +and @code{pthread_mutex_lock} returns immediately. If the mutex is +already locked by another thread, @code{pthread_mutex_lock} suspends the +calling thread until the mutex is unlocked. + +If the mutex is already locked by the calling thread, the behavior of +@code{pthread_mutex_lock} depends on the kind of the mutex. If the mutex +is of the ``fast'' kind, the calling thread is suspended. It will +remain suspended forever, because no other thread can unlock the mutex. +If the mutex is of the ``error checking'' kind, @code{pthread_mutex_lock} +returns immediately with the error code @code{EDEADLK}. If the mutex is +of the ``recursive'' kind, @code{pthread_mutex_lock} succeeds and +returns immediately, recording the number of times the calling thread +has locked the mutex. An equal number of @code{pthread_mutex_unlock} +operations must be performed before the mutex returns to the unlocked +state. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_mutex_trylock (pthread_mutex_t *@var{mutex}) +@code{pthread_mutex_trylock} behaves identically to +@code{pthread_mutex_lock}, except that it does not block the calling +thread if the mutex is already locked by another thread (or by the +calling thread in the case of a ``fast'' mutex). Instead, +@code{pthread_mutex_trylock} returns immediately with the error code +@code{EBUSY}. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_mutex_unlock (pthread_mutex_t *@var{mutex}) +@code{pthread_mutex_unlock} unlocks the given mutex. The mutex is +assumed to be locked and owned by the calling thread on entrance to +@code{pthread_mutex_unlock}. If the mutex is of the ``fast'' kind, +@code{pthread_mutex_unlock} always returns it to the unlocked state. If +it is of the ``recursive'' kind, it decrements the locking count of the +mutex (number of @code{pthread_mutex_lock} operations performed on it by +the calling thread), and only when this count reaches zero is the mutex +actually unlocked. + +On ``error checking'' mutexes, @code{pthread_mutex_unlock} actually +checks at run-time that the mutex is locked on entrance, and that it was +locked by the same thread that is now calling +@code{pthread_mutex_unlock}. If these conditions are not met, +@code{pthread_mutex_unlock} returns @code{EPERM}, and the mutex remains +unchanged. ``Fast'' and ``recursive'' mutexes perform no such checks, +thus allowing a locked mutex to be unlocked by a thread other than its +owner. This is non-portable behavior and must not be relied upon. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_mutex_destroy (pthread_mutex_t *@var{mutex}) +@code{pthread_mutex_destroy} destroys a mutex object, freeing the +resources it might hold. The mutex must be unlocked on entrance. In the +LinuxThreads implementation, no resources are associated with mutex +objects, thus @code{pthread_mutex_destroy} actually does nothing except +checking that the mutex is unlocked. + +If the mutex is locked by some thread, @code{pthread_mutex_destroy} +returns @code{EBUSY}. Otherwise it returns 0. +@end deftypefun + +If any of the above functions (except @code{pthread_mutex_init}) +is applied to an uninitialized mutex, they will simply return +@code{EINVAL} and do nothing. + +A shared global variable @var{x} can be protected by a mutex as follows: + +@smallexample +int x; +pthread_mutex_t mut = PTHREAD_MUTEX_INITIALIZER; +@end smallexample + +All accesses and modifications to @var{x} should be bracketed by calls to +@code{pthread_mutex_lock} and @code{pthread_mutex_unlock} as follows: + +@smallexample +pthread_mutex_lock(&mut); +/* operate on x */ +pthread_mutex_unlock(&mut); +@end smallexample + +Mutex attributes can be specified at mutex creation time, by passing a +mutex attribute object as second argument to @code{pthread_mutex_init}. +Passing @code{NULL} is equivalent to passing a mutex attribute object +with all attributes set to their default values. + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_mutexattr_init (pthread_mutexattr_t *@var{attr}) +@code{pthread_mutexattr_init} initializes the mutex attribute object +@var{attr} and fills it with default values for the attributes. + +This function always returns 0. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_mutexattr_destroy (pthread_mutexattr_t *@var{attr}) +@code{pthread_mutexattr_destroy} destroys a mutex attribute object, +which must not be reused until it is +reinitialized. @code{pthread_mutexattr_destroy} does nothing in the +LinuxThreads implementation. + +This function always returns 0. +@end deftypefun + +LinuxThreads supports only one mutex attribute: the mutex kind, which is +either @code{PTHREAD_MUTEX_FAST_NP} for ``fast'' mutexes, +@code{PTHREAD_MUTEX_RECURSIVE_NP} for ``recursive'' mutexes, or +@code{PTHREAD_MUTEX_ERRORCHECK_NP} for ``error checking'' mutexes. As +the @code{NP} suffix indicates, this is a non-portable extension to the +POSIX standard and should not be employed in portable programs. + +The mutex kind determines what happens if a thread attempts to lock a +mutex it already owns with @code{pthread_mutex_lock}. If the mutex is of +the ``fast'' kind, @code{pthread_mutex_lock} simply suspends the calling +thread forever. If the mutex is of the ``error checking'' kind, +@code{pthread_mutex_lock} returns immediately with the error code +@code{EDEADLK}. If the mutex is of the ``recursive'' kind, the call to +@code{pthread_mutex_lock} returns immediately with a success return +code. The number of times the thread owning the mutex has locked it is +recorded in the mutex. The owning thread must call +@code{pthread_mutex_unlock} the same number of times before the mutex +returns to the unlocked state. + +The default mutex kind is ``fast'', that is, @code{PTHREAD_MUTEX_FAST_NP}. + +@comment pthread.h +@comment GNU +@deftypefun int pthread_mutexattr_setkind_np (pthread_mutexattr_t *@var{attr}, int @var{kind}) +@code{pthread_mutexattr_setkind_np} sets the mutex kind attribute in +@var{attr} to the value specified by @var{kind}. + +If @var{kind} is not @code{PTHREAD_MUTEX_FAST_NP}, +@code{PTHREAD_MUTEX_RECURSIVE_NP}, or +@code{PTHREAD_MUTEX_ERRORCHECK_NP}, this function will return +@code{EINVAL} and leave @var{attr} unchanged. +@end deftypefun + +@comment pthread.h +@comment GNU +@deftypefun int pthread_mutexattr_getkind_np (const pthread_mutexattr_t *@var{attr}, int *@var{kind}) +@code{pthread_mutexattr_getkind_np} retrieves the current value of the +mutex kind attribute in @var{attr} and stores it in the location pointed +to by @var{kind}. + +This function always returns 0. +@end deftypefun + +@node Condition Variables +@section Condition Variables + +A condition (short for ``condition variable'') is a synchronization +device that allows threads to suspend execution until some predicate on +shared data is satisfied. The basic operations on conditions are: signal +the condition (when the predicate becomes true), and wait for the +condition, suspending the thread execution until another thread signals +the condition. + +A condition variable must always be associated with a mutex, to avoid +the race condition where a thread prepares to wait on a condition +variable and another thread signals the condition just before the first +thread actually waits on it. + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_cond_init (pthread_cond_t *@var{cond}, pthread_condattr_t *cond_@var{attr}) + +@code{pthread_cond_init} initializes the condition variable @var{cond}, +using the condition attributes specified in @var{cond_attr}, or default +attributes if @var{cond_attr} is @code{NULL}. The LinuxThreads +implementation supports no attributes for conditions, hence the +@var{cond_attr} parameter is actually ignored. + +Variables of type @code{pthread_cond_t} can also be initialized +statically, using the constant @code{PTHREAD_COND_INITIALIZER}. + +This function always returns 0. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_cond_signal (pthread_cond_t *@var{cond}) +@code{pthread_cond_signal} restarts one of the threads that are waiting +on the condition variable @var{cond}. If no threads are waiting on +@var{cond}, nothing happens. If several threads are waiting on +@var{cond}, exactly one is restarted, but it is not specified which. + +This function always returns 0. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_cond_broadcast (pthread_cond_t *@var{cond}) +@code{pthread_cond_broadcast} restarts all the threads that are waiting +on the condition variable @var{cond}. Nothing happens if no threads are +waiting on @var{cond}. + +This function always returns 0. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_cond_wait (pthread_cond_t *@var{cond}, pthread_mutex_t *@var{mutex}) +@code{pthread_cond_wait} atomically unlocks the @var{mutex} (as per +@code{pthread_unlock_mutex}) and waits for the condition variable +@var{cond} to be signaled. The thread execution is suspended and does +not consume any CPU time until the condition variable is signaled. The +@var{mutex} must be locked by the calling thread on entrance to +@code{pthread_cond_wait}. Before returning to the calling thread, +@code{pthread_cond_wait} re-acquires @var{mutex} (as per +@code{pthread_lock_mutex}). + +Unlocking the mutex and suspending on the condition variable is done +atomically. Thus, if all threads always acquire the mutex before +signaling the condition, this guarantees that the condition cannot be +signaled (and thus ignored) between the time a thread locks the mutex +and the time it waits on the condition variable. + +This function always returns 0. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_cond_timedwait (pthread_cond_t *@var{cond}, pthread_mutex_t *@var{mutex}, const struct timespec *@var{abstime}) +@code{pthread_cond_timedwait} atomically unlocks @var{mutex} and waits +on @var{cond}, as @code{pthread_cond_wait} does, but it also bounds the +duration of the wait. If @var{cond} has not been signaled before time +@var{abstime}, the mutex @var{mutex} is re-acquired and +@code{pthread_cond_timedwait} returns the error code @code{ETIMEDOUT}. +The wait can also be interrupted by a signal; in that case +@code{pthread_cond_timedwait} returns @code{EINTR}. + +The @var{abstime} parameter specifies an absolute time, with the same +origin as @code{time} and @code{gettimeofday}: an @var{abstime} of 0 +corresponds to 00:00:00 GMT, January 1, 1970. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_cond_destroy (pthread_cond_t *@var{cond}) +@code{pthread_cond_destroy} destroys the condition variable @var{cond}, +freeing the resources it might hold. If any threads are waiting on the +condition variable, @code{pthread_cond_destroy} leaves @var{cond} +untouched and returns @code{EBUSY}. Otherwise it returns 0, and +@var{cond} must not be used again until it is reinitialized. + +In the LinuxThreads implementation, no resources are associated with +condition variables, so @code{pthread_cond_destroy} actually does +nothing. +@end deftypefun + +@code{pthread_cond_wait} and @code{pthread_cond_timedwait} are +cancellation points. If a thread is cancelled while suspended in one of +these functions, the thread immediately resumes execution, relocks the +mutex specified by @var{mutex}, and finally executes the cancellation. +Consequently, cleanup handlers are assured that @var{mutex} is locked +when they are called. + +It is not safe to call the condition variable functions from a signal +handler. In particular, calling @code{pthread_cond_signal} or +@code{pthread_cond_broadcast} from a signal handler may deadlock the +calling thread. + +Consider two shared variables @var{x} and @var{y}, protected by the +mutex @var{mut}, and a condition variable @var{cond} that is to be +signaled whenever @var{x} becomes greater than @var{y}. + +@smallexample +int x,y; +pthread_mutex_t mut = PTHREAD_MUTEX_INITIALIZER; +pthread_cond_t cond = PTHREAD_COND_INITIALIZER; +@end smallexample + +Waiting until @var{x} is greater than @var{y} is performed as follows: + +@smallexample +pthread_mutex_lock(&mut); +while (x <= y) @{ + pthread_cond_wait(&cond, &mut); +@} +/* operate on x and y */ +pthread_mutex_unlock(&mut); +@end smallexample + +Modifications on @var{x} and @var{y} that may cause @var{x} to become greater than +@var{y} should signal the condition if needed: + +@smallexample +pthread_mutex_lock(&mut); +/* modify x and y */ +if (x > y) pthread_cond_broadcast(&cond); +pthread_mutex_unlock(&mut); +@end smallexample + +If it can be proved that at most one waiting thread needs to be waken +up (for instance, if there are only two threads communicating through +@var{x} and @var{y}), @code{pthread_cond_signal} can be used as a slightly more +efficient alternative to @code{pthread_cond_broadcast}. In doubt, use +@code{pthread_cond_broadcast}. + +To wait for @var{x} to becomes greater than @var{y} with a timeout of 5 +seconds, do: + +@smallexample +struct timeval now; +struct timespec timeout; +int retcode; + +pthread_mutex_lock(&mut); +gettimeofday(&now); +timeout.tv_sec = now.tv_sec + 5; +timeout.tv_nsec = now.tv_usec * 1000; +retcode = 0; +while (x <= y && retcode != ETIMEDOUT) @{ + retcode = pthread_cond_timedwait(&cond, &mut, &timeout); +@} +if (retcode == ETIMEDOUT) @{ + /* timeout occurred */ +@} else @{ + /* operate on x and y */ +@} +pthread_mutex_unlock(&mut); +@end smallexample + +Condition attributes can be specified at condition creation time, by +passing a condition attribute object as second argument to +@code{pthread_cond_init}. Passing @code{NULL} is equivalent to passing +a condition attribute object with all attributes set to their default +values. + +The LinuxThreads implementation supports no attributes for +conditions. The functions on condition attributes are included only for +compliance with the POSIX standard. + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_condattr_init (pthread_condattr_t *@var{attr}) +@deftypefunx int pthread_condattr_destroy (pthread_condattr_t *@var{attr}) +@code{pthread_condattr_init} initializes the condition attribute object +@var{attr} and fills it with default values for the attributes. +@code{pthread_condattr_destroy} destroys the condition attribute object +@var{attr}. + +Both functions do nothing in the LinuxThreads implementation. + +@code{pthread_condattr_init} and @code{pthread_condattr_destroy} always +return 0. +@end deftypefun + +@node POSIX Semaphores +@section POSIX Semaphores + +@vindex SEM_VALUE_MAX +Semaphores are counters for resources shared between threads. The +basic operations on semaphores are: increment the counter atomically, +and wait until the counter is non-null and decrement it atomically. + +Semaphores have a maximum value past which they cannot be incremented. +The macro @code{SEM_VALUE_MAX} is defined to be this maximum value. In +the GNU C library, @code{SEM_VALUE_MAX} is equal to @code{INT_MAX} +(@pxref{Range of Type}), but it may be much smaller on other systems. + +The pthreads library implements POSIX 1003.1b semaphores. These should +not be confused with System V semaphores (@code{ipc}, @code{semctl} and +@code{semop}). +@c !!! SysV IPC is not doc'd at all in our manual + +All the semaphore functions and macros are defined in @file{semaphore.h}. + +@comment semaphore.h +@comment POSIX +@deftypefun int sem_init (sem_t *@var{sem}, int @var{pshared}, unsigned int @var{value}) +@code{sem_init} initializes the semaphore object pointed to by +@var{sem}. The count associated with the semaphore is set initially to +@var{value}. The @var{pshared} argument indicates whether the semaphore +is local to the current process (@var{pshared} is zero) or is to be +shared between several processes (@var{pshared} is not zero). + +On success @code{sem_init} returns 0. On failure it returns -1 and sets +@var{errno} to one of the following values: + +@table @code +@item EINVAL +@var{value} exceeds the maximal counter value @code{SEM_VALUE_MAX} + +@item ENOSYS +@var{pshared} is not zero. LinuxThreads currently does not support +process-shared semaphores. (This will eventually change.) +@end table +@end deftypefun + +@comment semaphore.h +@comment POSIX +@deftypefun int sem_destroy (sem_t * @var{sem}) +@code{sem_destroy} destroys a semaphore object, freeing the resources it +might hold. If any threads are waiting on the semaphore when +@code{sem_destroy} is called, it fails and sets @var{errno} to +@code{EBUSY}. + +In the LinuxThreads implementation, no resources are associated with +semaphore objects, thus @code{sem_destroy} actually does nothing except +checking that no thread is waiting on the semaphore. This will change +when process-shared semaphores are implemented. +@end deftypefun + +@comment semaphore.h +@comment POSIX +@deftypefun int sem_wait (sem_t * @var{sem}) +@code{sem_wait} suspends the calling thread until the semaphore pointed +to by @var{sem} has non-zero count. It then atomically decreases the +semaphore count. + +@code{sem_wait} is a cancellation point. It always returns 0. +@end deftypefun + +@comment semaphore.h +@comment POSIX +@deftypefun int sem_trywait (sem_t * @var{sem}) +@code{sem_trywait} is a non-blocking variant of @code{sem_wait}. If the +semaphore pointed to by @var{sem} has non-zero count, the count is +atomically decreased and @code{sem_trywait} immediately returns 0. If +the semaphore count is zero, @code{sem_trywait} immediately returns -1 +and sets errno to @code{EAGAIN}. +@end deftypefun + +@comment semaphore.h +@comment POSIX +@deftypefun int sem_post (sem_t * @var{sem}) +@code{sem_post} atomically increases the count of the semaphore pointed to +by @var{sem}. This function never blocks. + +@c !!! This para appears not to agree with the code. +On processors supporting atomic compare-and-swap (Intel 486, Pentium and +later, Alpha, PowerPC, MIPS II, Motorola 68k, Ultrasparc), the +@code{sem_post} function is can safely be called from signal handlers. +This is the only thread synchronization function provided by POSIX +threads that is async-signal safe. On the Intel 386 and earlier Sparc +chips, the current LinuxThreads implementation of @code{sem_post} is not +async-signal safe, because the hardware does not support the required +atomic operations. + +@code{sem_post} always succeeds and returns 0, unless the semaphore +count would exceed @code{SEM_VALUE_MAX} after being incremented. In +that case @code{sem_post} returns -1 and sets @var{errno} to +@code{EINVAL}. The semaphore count is left unchanged. +@end deftypefun + +@comment semaphore.h +@comment POSIX +@deftypefun int sem_getvalue (sem_t * @var{sem}, int * @var{sval}) +@code{sem_getvalue} stores in the location pointed to by @var{sval} the +current count of the semaphore @var{sem}. It always returns 0. +@end deftypefun + +@node Thread-Specific Data +@section Thread-Specific Data + +Programs often need global or static variables that have different +values in different threads. Since threads share one memory space, this +cannot be achieved with regular variables. Thread-specific data is the +POSIX threads answer to this need. + +Each thread possesses a private memory block, the thread-specific data +area, or TSD area for short. This area is indexed by TSD keys. The TSD +area associates values of type @code{void *} to TSD keys. TSD keys are +common to all threads, but the value associated with a given TSD key can +be different in each thread. + +For concreteness, the TSD areas can be viewed as arrays of @code{void *} +pointers, TSD keys as integer indices into these arrays, and the value +of a TSD key as the value of the corresponding array element in the +calling thread. + +When a thread is created, its TSD area initially associates @code{NULL} +with all keys. + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_key_create (pthread_key_t *@var{key}, void (*destr_function) (void *)) +@code{pthread_key_create} allocates a new TSD key. The key is stored in +the location pointed to by @var{key}. There is a limit of +@code{PTHREAD_KEYS_MAX} on the number of keys allocated at a given +time. The value initially associated with the returned key is +@code{NULL} in all currently executing threads. + +The @var{destr_function} argument, if not @code{NULL}, specifies a +destructor function associated with the key. When a thread terminates +via @code{pthread_exit} or by cancellation, @var{destr_function} is +called on the value associated with the key in that thread. The +@var{destr_function} is not called if a key is deleted with +@code{pthread_key_delete} or a value is changed with +@code{pthread_setspecific}. The order in which destructor functions are +called at thread termination time is unspecified. + +Before the destructor function is called, the @code{NULL} value is +associated with the key in the current thread. A destructor function +might, however, re-associate non-@code{NULL} values to that key or some +other key. To deal with this, if after all the destructors have been +called for all non-@code{NULL} values, there are still some +non-@code{NULL} values with associated destructors, then the process is +repeated. The LinuxThreads implementation stops the process after +@code{PTHREAD_DESTRUCTOR_ITERATIONS} iterations, even if some +non-@code{NULL} values with associated descriptors remain. Other +implementations may loop indefinitely. + +@code{pthread_key_create} returns 0 unless @code{PTHREAD_KEYS_MAX} keys +have already been allocated, in which case it fails and returns +@code{EAGAIN}. +@end deftypefun + + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_key_delete (pthread_key_t @var{key}) +@code{pthread_key_delete} deallocates a TSD key. It does not check +whether non-@code{NULL} values are associated with that key in the +currently executing threads, nor call the destructor function associated +with the key. + +If there is no such key @var{key}, it returns @code{EINVAL}. Otherwise +it returns 0. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_setspecific (pthread_key_t @var{key}, const void *@var{pointer}) +@code{pthread_setspecific} changes the value associated with @var{key} +in the calling thread, storing the given @var{pointer} instead. + +If there is no such key @var{key}, it returns @code{EINVAL}. Otherwise +it returns 0. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun {void *} pthread_getspecific (pthread_key_t @var{key}) +@code{pthread_getspecific} returns the value currently associated with +@var{key} in the calling thread. + +If there is no such key @var{key}, it returns @code{NULL}. +@end deftypefun + +The following code fragment allocates a thread-specific array of 100 +characters, with automatic reclaimation at thread exit: + +@smallexample +/* Key for the thread-specific buffer */ +static pthread_key_t buffer_key; + +/* Once-only initialisation of the key */ +static pthread_once_t buffer_key_once = PTHREAD_ONCE_INIT; + +/* Allocate the thread-specific buffer */ +void buffer_alloc(void) +@{ + pthread_once(&buffer_key_once, buffer_key_alloc); + pthread_setspecific(buffer_key, malloc(100)); +@} + +/* Return the thread-specific buffer */ +char * get_buffer(void) +@{ + return (char *) pthread_getspecific(buffer_key); +@} + +/* Allocate the key */ +static void buffer_key_alloc() +@{ + pthread_key_create(&buffer_key, buffer_destroy); +@} + +/* Free the thread-specific buffer */ +static void buffer_destroy(void * buf) +@{ + free(buf); +@} +@end smallexample + +@node Threads and Signal Handling +@section Threads and Signal Handling + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_sigmask (int @var{how}, const sigset_t *@var{newmask}, sigset_t *@var{oldmask}) +@code{pthread_sigmask} changes the signal mask for the calling thread as +described by the @var{how} and @var{newmask} arguments. If @var{oldmask} +is not @code{NULL}, the previous signal mask is stored in the location +pointed to by @var{oldmask}. + +The meaning of the @var{how} and @var{newmask} arguments is the same as +for @code{sigprocmask}. If @var{how} is @code{SIG_SETMASK}, the signal +mask is set to @var{newmask}. If @var{how} is @code{SIG_BLOCK}, the +signals specified to @var{newmask} are added to the current signal mask. +If @var{how} is @code{SIG_UNBLOCK}, the signals specified to +@var{newmask} are removed from the current signal mask. + +Recall that signal masks are set on a per-thread basis, but signal +actions and signal handlers, as set with @code{sigaction}, are shared +between all threads. + +The @code{pthread_sigmask} function returns 0 on success, and one of the +following error codes on error: +@table @code +@item EINVAL +@var{how} is not one of @code{SIG_SETMASK}, @code{SIG_BLOCK}, or @code{SIG_UNBLOCK} + +@item EFAULT +@var{newmask} or @var{oldmask} point to invalid addresses +@end table +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_kill (pthread_t @var{thread}, int @var{signo}) +@code{pthread_kill} sends signal number @var{signo} to the thread +@var{thread}. The signal is delivered and handled as described in +@ref{Signal Handling}. + +@code{pthread_kill} returns 0 on success, one of the following error codes +on error: +@table @code +@item EINVAL +@var{signo} is not a valid signal number + +@item ESRCH +The thread @var{thread} does not exist (e.g. it has already terminated) +@end table +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int sigwait (const sigset_t *@var{set}, int *@var{sig}) +@code{sigwait} suspends the calling thread until one of the signals in +@var{set} is delivered to the calling thread. It then stores the number +of the signal received in the location pointed to by @var{sig} and +returns. The signals in @var{set} must be blocked and not ignored on +entrance to @code{sigwait}. If the delivered signal has a signal handler +function attached, that function is @emph{not} called. + +@code{sigwait} is a cancellation point. It always returns 0. +@end deftypefun + +For @code{sigwait} to work reliably, the signals being waited for must be +blocked in all threads, not only in the calling thread, since +otherwise the POSIX semantics for signal delivery do not guarantee +that it's the thread doing the @code{sigwait} that will receive the signal. +The best way to achieve this is block those signals before any threads +are created, and never unblock them in the program other than by +calling @code{sigwait}. + +Signal handling in LinuxThreads departs significantly from the POSIX +standard. According to the standard, ``asynchronous'' (external) signals +are addressed to the whole process (the collection of all threads), +which then delivers them to one particular thread. The thread that +actually receives the signal is any thread that does not currently block +the signal. + +In LinuxThreads, each thread is actually a kernel process with its own +PID, so external signals are always directed to one particular thread. +If, for instance, another thread is blocked in @code{sigwait} on that +signal, it will not be restarted. + +The LinuxThreads implementation of @code{sigwait} installs dummy signal +handlers for the signals in @var{set} for the duration of the +wait. Since signal handlers are shared between all threads, other +threads must not attach their own signal handlers to these signals, or +alternatively they should all block these signals (which is recommended +anyway). + +@node Miscellaneous Thread Functions +@section Miscellaneous Thread Functions + +@comment pthread.h +@comment POSIX +@deftypefun {pthread_t} pthread_self (@var{void}) +@code{pthread_self} returns the thread identifier for the calling thread. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_equal (pthread_t thread1, pthread_t thread2) +@code{pthread_equal} determines if two thread identifiers refer to the same +thread. + +A non-zero value is returned if @var{thread1} and @var{thread2} refer to +the same thread. Otherwise, 0 is returned. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_detach (pthread_t @var{th}) +@code{pthread_detach} puts the thread @var{th} in the detached +state. This guarantees that the memory resources consumed by @var{th} +will be freed immediately when @var{th} terminates. However, this +prevents other threads from synchronizing on the termination of @var{th} +using @code{pthread_join}. + +A thread can be created initially in the detached state, using the +@code{detachstate} attribute to @code{pthread_create}. In contrast, +@code{pthread_detach} applies to threads created in the joinable state, +and which need to be put in the detached state later. + +After @code{pthread_detach} completes, subsequent attempts to perform +@code{pthread_join} on @var{th} will fail. If another thread is already +joining the thread @var{th} at the time @code{pthread_detach} is called, +@code{pthread_detach} does nothing and leaves @var{th} in the joinable +state. + +On success, 0 is returned. On error, one of the following codes is +returned: +@table @code +@item ESRCH +No thread could be found corresponding to that specified by @var{th} +@item EINVAL +The thread @var{th} is already in the detached state +@end table +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_atfork (void (*@var{prepare})(void), void (*@var{parent})(void), void (*@var{child})(void)) + +@code{pthread_atfork} registers handler functions to be called just +before and just after a new process is created with @code{fork}. The +@var{prepare} handler will be called from the parent process, just +before the new process is created. The @var{parent} handler will be +called from the parent process, just before @code{fork} returns. The +@var{child} handler will be called from the child process, just before +@code{fork} returns. + +@code{pthread_atfork} returns 0 on success and a non-zero error code on +error. + +One or more of the three handlers @var{prepare}, @var{parent} and +@var{child} can be given as @code{NULL}, meaning that no handler needs +to be called at the corresponding point. + +@code{pthread_atfork} can be called several times to install several +sets of handlers. At @code{fork} time, the @var{prepare} handlers are +called in LIFO order (last added with @code{pthread_atfork}, first +called before @code{fork}), while the @var{parent} and @var{child} +handlers are called in FIFO order (first added, first called). + +If there is insufficient memory available to register the handlers, +@code{pthread_atfork} fails and returns @code{ENOMEM}. Otherwise it +returns 0. +@end deftypefun + +To understand the purpose of @code{pthread_atfork}, recall that +@code{fork} duplicates the whole memory space, including mutexes in +their current locking state, but only the calling thread: other threads +are not running in the child process. Thus, if a mutex is locked by a +thread other than the thread calling @code{fork}, that mutex will remain +locked forever in the child process, possibly blocking the execution of +the child process. To avoid this, install handlers with +@code{pthread_atfork} as follows: the @var{prepare} handler locks the +global mutexes (in locking order), and the @var{parent} and @var{child} +handlers unlock them (in reverse order). Alternatively, @var{prepare} +and @var{parent} can be set to @code{NULL} and @var{child} to a function +that calls @code{pthread_mutex_init} on the global mutexes. + +@comment pthread.h +@comment GNU +@deftypefun void pthread_kill_other_threads_np (@var{void}) +@code{pthread_kill_other_threads_np} is a non-portable LinuxThreads extension. +It causes all threads in the program to terminate immediately, except +the calling thread which proceeds normally. It is intended to be +called just before a thread calls one of the @code{exec} functions, +e.g. @code{execve}. + +Termination of the other threads is not performed through +@code{pthread_cancel} and completely bypasses the cancellation +mechanism. Hence, the current settings for cancellation state and +cancellation type are ignored, and the cleanup handlers are not +executed in the terminated threads. + +According to POSIX 1003.1c, a successful @code{exec*} in one of the +threads should automatically terminate all other threads in the program. +This behavior is not yet implemented in LinuxThreads. Calling +@code{pthread_kill_other_threads_np} before @code{exec*} achieves much +of the same behavior, except that if @code{exec*} ultimately fails, then +all other threads are already killed. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_once (pthread_once_t *once_@var{control}, void (*@var{init_routine}) (void)) + +The purpose of @code{pthread_once} is to ensure that a piece of +initialization code is executed at most once. The @var{once_control} +argument points to a static or extern variable statically initialized +to @code{PTHREAD_ONCE_INIT}. + +The first time @code{pthread_once} is called with a given +@var{once_control} argument, it calls @var{init_routine} with no +argument and changes the value of the @var{once_control} variable to +record that initialization has been performed. Subsequent calls to +@code{pthread_once} with the same @code{once_control} argument do +nothing. + +@code{pthread_once} always returns 0. +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_setschedparam (pthread_t target_@var{thread}, int @var{policy}, const struct sched_param *@var{param}) + +@code{pthread_setschedparam} sets the scheduling parameters for the +thread @var{target_thread} as indicated by @var{policy} and +@var{param}. @var{policy} can be either @code{SCHED_OTHER} (regular, +non-realtime scheduling), @code{SCHED_RR} (realtime, round-robin) or +@code{SCHED_FIFO} (realtime, first-in first-out). @var{param} specifies +the scheduling priority for the two realtime policies. See +@code{sched_setpolicy} for more information on scheduling policies. + +The realtime scheduling policies @code{SCHED_RR} and @code{SCHED_FIFO} +are available only to processes with superuser privileges. + +On success, @code{pthread_setschedparam} returns 0. On error it returns +one of the following codes: +@table @code +@item EINVAL +@var{policy} is not one of @code{SCHED_OTHER}, @code{SCHED_RR}, +@code{SCHED_FIFO}, or the priority value specified by @var{param} is not +valid for the specified policy + +@item EPERM +Realtime scheduling was requested but the calling process does not have +sufficient privileges. + +@item ESRCH +The @var{target_thread} is invalid or has already terminated + +@item EFAULT +@var{param} points outside the process memory space +@end table +@end deftypefun + +@comment pthread.h +@comment POSIX +@deftypefun int pthread_getschedparam (pthread_t target_@var{thread}, int *@var{policy}, struct sched_param *@var{param}) + +@code{pthread_getschedparam} retrieves the scheduling policy and +scheduling parameters for the thread @var{target_thread} and stores them +in the locations pointed to by @var{policy} and @var{param}, +respectively. + +@code{pthread_getschedparam} returns 0 on success, or one of the +following error codes on failure: +@table @code +@item ESRCH +The @var{target_thread} is invalid or has already terminated. + +@item EFAULT +@var{policy} or @var{param} point outside the process memory space. + +@end table +@end deftypefun diff --git a/libpthread/linuxthreads/lockfile.c b/libpthread/linuxthreads/lockfile.c new file mode 100644 index 000000000..18c3fed8c --- /dev/null +++ b/libpthread/linuxthreads/lockfile.c @@ -0,0 +1,88 @@ +/* lockfile - Handle locking and unlocking of stream. + Copyright (C) 1996, 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#include <bits/libc-lock.h> +#include <stdio.h> +#include <pthread.h> + +#ifdef USE_IN_LIBIO +#include "../libio/libioP.h" +#endif + +void +__flockfile (FILE *stream) +{ +#ifdef USE_IN_LIBIO + __pthread_mutex_lock (stream->_lock); +#else +#endif +} +#ifdef USE_IN_LIBIO +#undef _IO_flockfile +strong_alias (__flockfile, _IO_flockfile) +#endif +weak_alias (__flockfile, flockfile); + + +void +__funlockfile (FILE *stream) +{ +#ifdef USE_IN_LIBIO + __pthread_mutex_unlock (stream->_lock); +#else +#endif +} +#ifdef USE_IN_LIBIO +#undef _IO_funlockfile +strong_alias (__funlockfile, _IO_funlockfile) +#endif +weak_alias (__funlockfile, funlockfile); + + +int +__ftrylockfile (FILE *stream) +{ +#ifdef USE_IN_LIBIO + return __pthread_mutex_trylock (stream->_lock); +#else + return 0; +#endif +} +#ifdef USE_IN_LIBIO +strong_alias (__ftrylockfile, _IO_ftrylockfile) +#endif +weak_alias (__ftrylockfile, ftrylockfile); + + +void +__fresetlockfiles (void) +{ +#ifdef USE_IN_LIBIO + _IO_FILE *fp; + pthread_mutexattr_t attr; + + __pthread_mutexattr_init (&attr); + __pthread_mutexattr_settype (&attr, PTHREAD_MUTEX_RECURSIVE_NP); + + for (fp = _IO_list_all; fp != NULL; fp = fp->_chain) + __pthread_mutex_init (fp->_lock, &attr); + + __pthread_mutexattr_destroy (&attr); +#endif +} diff --git a/libpthread/linuxthreads/manager.c b/libpthread/linuxthreads/manager.c new file mode 100644 index 000000000..5d355e107 --- /dev/null +++ b/libpthread/linuxthreads/manager.c @@ -0,0 +1,786 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* The "thread manager" thread: manages creation and termination of threads */ + +/* mods for uClibc: getpwd and getpagesize are the syscalls */ +#define __getpid getpid +#define __getpagesize getpagesize + +#include <errno.h> +#include <sched.h> +#include <stddef.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <unistd.h> +#include <sys/poll.h> /* for poll */ +#include <sys/mman.h> /* for mmap */ +#include <sys/param.h> +#include <sys/time.h> +#include <sys/wait.h> /* for waitpid macros */ + +#include "pthread.h" +#include "internals.h" +#include "spinlock.h" +#include "restart.h" +#include "semaphore.h" +#include "debug.h" /* PDEBUG, added by StS */ + +/* Array of active threads. Entry 0 is reserved for the initial thread. */ +struct pthread_handle_struct __pthread_handles[PTHREAD_THREADS_MAX] = +{ { LOCK_INITIALIZER, &__pthread_initial_thread, 0}, + { LOCK_INITIALIZER, &__pthread_manager_thread, 0}, /* All NULLs */ }; + +/* For debugging purposes put the maximum number of threads in a variable. */ +const int __linuxthreads_pthread_threads_max = PTHREAD_THREADS_MAX; + +/* Indicate whether at least one thread has a user-defined stack (if 1), + or if all threads have stacks supplied by LinuxThreads (if 0). */ +int __pthread_nonstandard_stacks; + +/* Number of active entries in __pthread_handles (used by gdb) */ +volatile int __pthread_handles_num = 2; + +/* Whether to use debugger additional actions for thread creation + (set to 1 by gdb) */ +volatile int __pthread_threads_debug; + +/* Globally enabled events. */ +volatile td_thr_events_t __pthread_threads_events; + +/* Pointer to thread descriptor with last event. */ +volatile pthread_descr __pthread_last_event; + +/* Mapping from stack segment to thread descriptor. */ +/* Stack segment numbers are also indices into the __pthread_handles array. */ +/* Stack segment number 0 is reserved for the initial thread. */ + +static inline pthread_descr thread_segment(int seg) +{ + return (pthread_descr)(THREAD_STACK_START_ADDRESS - (seg - 1) * STACK_SIZE) + - 1; +} + +/* Flag set in signal handler to record child termination */ + +static volatile int terminated_children = 0; + +/* Flag set when the initial thread is blocked on pthread_exit waiting + for all other threads to terminate */ + +static int main_thread_exiting = 0; + +/* Counter used to generate unique thread identifier. + Thread identifier is pthread_threads_counter + segment. */ + +static pthread_t pthread_threads_counter = 0; + +/* Forward declarations */ + +static int pthread_handle_create(pthread_t *thread, const pthread_attr_t *attr, + void * (*start_routine)(void *), void *arg, + sigset_t *mask, int father_pid, + int report_events, + td_thr_events_t *event_maskp); +static void pthread_handle_free(pthread_t th_id); +static void pthread_handle_exit(pthread_descr issuing_thread, int exitcode); +static void pthread_reap_children(void); +static void pthread_kill_all_threads(int sig, int main_thread_also); + +/* The server thread managing requests for thread creation and termination */ + +int __pthread_manager(void *arg) +{ + int reqfd = (int) (long int) arg; + struct pollfd ufd; + sigset_t mask; + int n; + struct pthread_request request; + + /* If we have special thread_self processing, initialize it. */ +#ifdef INIT_THREAD_SELF + INIT_THREAD_SELF(&__pthread_manager_thread, 1); +#endif + /* Set the error variable. */ + __pthread_manager_thread.p_errnop = &__pthread_manager_thread.p_errno; + __pthread_manager_thread.p_h_errnop = &__pthread_manager_thread.p_h_errno; + /* Block all signals except __pthread_sig_cancel and SIGTRAP */ + sigfillset(&mask); + sigdelset(&mask, __pthread_sig_cancel); /* for thread termination */ + sigdelset(&mask, SIGTRAP); /* for debugging purposes */ + sigprocmask(SIG_SETMASK, &mask, NULL); + /* Raise our priority to match that of main thread */ + __pthread_manager_adjust_prio(__pthread_main_thread->p_priority); + /* Synchronize debugging of the thread manager */ + n = __libc_read(reqfd, (char *)&request, sizeof(request)); + ASSERT(n == sizeof(request) && request.req_kind == REQ_DEBUG); + ufd.fd = reqfd; + ufd.events = POLLIN; + /* Enter server loop */ + while(1) { +PDEBUG("before poll\n"); + n = poll(&ufd, 1, 2000); +PDEBUG("after poll\n"); + + /* Check for termination of the main thread */ + if (getppid() == 1) { + pthread_kill_all_threads(SIGKILL, 0); + _exit(0); + } + /* Check for dead children */ + if (terminated_children) { + terminated_children = 0; + pthread_reap_children(); + } + /* Read and execute request */ + if (n == 1 && (ufd.revents & POLLIN)) { +PDEBUG("before __libc_read\n"); + n = __libc_read(reqfd, (char *)&request, sizeof(request)); +PDEBUG("after __libc_read, n=%d\n", n); + ASSERT(n == sizeof(request)); + switch(request.req_kind) { + case REQ_CREATE: +PDEBUG("got REQ_CREATE\n"); + request.req_thread->p_retcode = + pthread_handle_create((pthread_t *) &request.req_thread->p_retval, + request.req_args.create.attr, + request.req_args.create.fn, + request.req_args.create.arg, + &request.req_args.create.mask, + request.req_thread->p_pid, + request.req_thread->p_report_events, + &request.req_thread->p_eventbuf.eventmask); +PDEBUG("restarting %d\n", request.req_thread); + restart(request.req_thread); + break; + case REQ_FREE: +PDEBUG("got REQ_FREE\n"); + pthread_handle_free(request.req_args.free.thread_id); + break; + case REQ_PROCESS_EXIT: +PDEBUG("got REQ_PROCESS_EXIT from %d, exit code = %d\n", + request.req_thread, request.req_args.exit.code); + pthread_handle_exit(request.req_thread, + request.req_args.exit.code); + break; + case REQ_MAIN_THREAD_EXIT: +PDEBUG("got REQ_MAIN_THREAD_EXIT\n"); + main_thread_exiting = 1; + if (__pthread_main_thread->p_nextlive == __pthread_main_thread) { + restart(__pthread_main_thread); + return 0; + } + break; + case REQ_POST: +PDEBUG("got REQ_POST\n"); + __new_sem_post(request.req_args.post); + break; + case REQ_DEBUG: +PDEBUG("got REQ_DEBUG\n"); + /* Make gdb aware of new thread and gdb will restart the + new thread when it is ready to handle the new thread. */ + if (__pthread_threads_debug && __pthread_sig_debug > 0) +PDEBUG("about to call raise(__pthread_sig_debug)\n"); + raise(__pthread_sig_debug); + break; + } + } + } +} + +int __pthread_manager_event(void *arg) +{ + /* If we have special thread_self processing, initialize it. */ +#ifdef INIT_THREAD_SELF + INIT_THREAD_SELF(&__pthread_manager_thread, 1); +#endif + + /* Get the lock the manager will free once all is correctly set up. */ + __pthread_lock (THREAD_GETMEM((&__pthread_manager_thread), p_lock), NULL); + /* Free it immediately. */ + __pthread_unlock (THREAD_GETMEM((&__pthread_manager_thread), p_lock)); + + return __pthread_manager(arg); +} + +/* Process creation */ + +static int pthread_start_thread(void *arg) +{ + pthread_descr self = (pthread_descr) arg; + struct pthread_request request; + void * outcome; + /* Initialize special thread_self processing, if any. */ +#ifdef INIT_THREAD_SELF + INIT_THREAD_SELF(self, self->p_nr); +#endif +PDEBUG("\n"); + /* Make sure our pid field is initialized, just in case we get there + before our father has initialized it. */ + THREAD_SETMEM(self, p_pid, __getpid()); + /* Initial signal mask is that of the creating thread. (Otherwise, + we'd just inherit the mask of the thread manager.) */ + sigprocmask(SIG_SETMASK, &self->p_start_args.mask, NULL); + /* Set the scheduling policy and priority for the new thread, if needed */ + if (THREAD_GETMEM(self, p_start_args.schedpolicy) >= 0) + /* Explicit scheduling attributes were provided: apply them */ + sched_setscheduler(THREAD_GETMEM(self, p_pid), + THREAD_GETMEM(self, p_start_args.schedpolicy), + &self->p_start_args.schedparam); + else if (__pthread_manager_thread.p_priority > 0) + /* Default scheduling required, but thread manager runs in realtime + scheduling: switch new thread to SCHED_OTHER policy */ + { + struct sched_param default_params; + default_params.sched_priority = 0; + sched_setscheduler(THREAD_GETMEM(self, p_pid), + SCHED_OTHER, &default_params); + } + /* Make gdb aware of new thread */ + if (__pthread_threads_debug && __pthread_sig_debug > 0) { + request.req_thread = self; + request.req_kind = REQ_DEBUG; + __libc_write(__pthread_manager_request, + (char *) &request, sizeof(request)); + suspend(self); + } + /* Run the thread code */ + outcome = self->p_start_args.start_routine(THREAD_GETMEM(self, + p_start_args.arg)); + /* Exit with the given return value */ + pthread_exit(outcome); + return 0; +} + +static int pthread_start_thread_event(void *arg) +{ + pthread_descr self = (pthread_descr) arg; + +#ifdef INIT_THREAD_SELF + INIT_THREAD_SELF(self, self->p_nr); +#endif + /* Make sure our pid field is initialized, just in case we get there + before our father has initialized it. */ + THREAD_SETMEM(self, p_pid, __getpid()); + /* Get the lock the manager will free once all is correctly set up. */ + __pthread_lock (THREAD_GETMEM(self, p_lock), NULL); + /* Free it immediately. */ + __pthread_unlock (THREAD_GETMEM(self, p_lock)); + + /* Continue with the real function. */ + return pthread_start_thread (arg); +} + +static int pthread_allocate_stack(const pthread_attr_t *attr, + pthread_descr default_new_thread, + int pagesize, + pthread_descr * out_new_thread, + char ** out_new_thread_bottom, + char ** out_guardaddr, + size_t * out_guardsize) +{ + pthread_descr new_thread; + char * new_thread_bottom; + char * guardaddr; + size_t stacksize, guardsize; + + if (attr != NULL && attr->__stackaddr_set) + { + /* The user provided a stack. */ + new_thread = + (pthread_descr) ((long)(attr->__stackaddr) & -sizeof(void *)) - 1; + new_thread_bottom = (char *) attr->__stackaddr - attr->__stacksize; + guardaddr = NULL; + guardsize = 0; + __pthread_nonstandard_stacks = 1; + } + else + { +#ifdef __UCLIBC_HAS_MMU__ + stacksize = STACK_SIZE - pagesize; + if (attr != NULL) + stacksize = MIN (stacksize, roundup(attr->__stacksize, pagesize)); + /* Allocate space for stack and thread descriptor at default address */ + new_thread = default_new_thread; + new_thread_bottom = (char *) (new_thread + 1) - stacksize; + if (mmap((caddr_t)((char *)(new_thread + 1) - INITIAL_STACK_SIZE), + INITIAL_STACK_SIZE, PROT_READ | PROT_WRITE | PROT_EXEC, + MAP_PRIVATE | MAP_ANONYMOUS | MAP_FIXED | MAP_GROWSDOWN, + -1, 0) == MAP_FAILED) + /* Bad luck, this segment is already mapped. */ + return -1; + /* We manage to get a stack. Now see whether we need a guard + and allocate it if necessary. Notice that the default + attributes (stack_size = STACK_SIZE - pagesize) do not need + a guard page, since the RLIMIT_STACK soft limit prevents stacks + from running into one another. */ + if (stacksize == STACK_SIZE - pagesize) + { + /* We don't need a guard page. */ + guardaddr = NULL; + guardsize = 0; + } + else + { + /* Put a bad page at the bottom of the stack */ + guardsize = attr->__guardsize; + guardaddr = (void *)new_thread_bottom - guardsize; + if (mmap ((caddr_t) guardaddr, guardsize, 0, MAP_FIXED, -1, 0) + == MAP_FAILED) + { + /* We don't make this an error. */ + guardaddr = NULL; + guardsize = 0; + } + } +#else + /* We cannot mmap to this huge chunk of stack space when we don't have + * an MMU. Pretend we are using a user provided stack even if there was + * none provided by the user. Thus, we get around the mmap and reservation + * of a huge stack segment. -StS */ + + char *new_stack; + + if ((new_stack = malloc(INITIAL_STACK_SIZE)) == NULL) { + /* bad luck, we cannot malloc any more */ + return -1; + } + + PDEBUG("malloced chunk: base=%p, size=0x%04x\n", new_stack, INITIAL_STACK_SIZE); + + /* Set up the pointers. new_thread marks the TOP of the stack frame and + * the address of the pthread_descr struct at the same time. Therefore we + * must account for its size and fit it in the malloc()'ed block. The + * value of `new_thread' is then passed to clone() as the stack argument. + * + * ^ +------------------------+ + * | | pthread_descr struct | + * | +------------------------+ <- new_thread + * malloc block | | | + * | | thread stack | + * | | | + * v +------------------------+ <- new_thread_bottom + * + * Note: The calculated value of new_thread must be word aligned otherwise + * the kernel chokes on a non-aligned stack frame. Choose the lower + * available word boundary. + */ + new_thread_bottom = (pthread_descr) new_stack; + new_thread = (long)((char *) new_stack + INITIAL_STACK_SIZE - sizeof(*new_thread) - 1) + & -sizeof(void*); /* align new_thread */ + guardaddr = NULL; + guardsize = 0; + + PDEBUG("thread stack: bos=%p, tos=%p\n", new_thread_bottom, new_thread); + + /* check the initial thread stack boundaries so they don't overlap */ + NOMMU_INITIAL_THREAD_BOUNDS(new_thread, new_thread_bottom); + + PDEBUG("initial stack: bos=%p, tos=%p\n", __pthread_initial_thread_bos, + __pthread_initial_thread_tos); + + /* on non-MMU systems we always have non-standard stack frames */ + __pthread_nonstandard_stacks = 1; + +#endif /* __UCLIBC_HAS_MMU__ */ + } + + /* Clear the thread data structure. */ + memset (new_thread, '\0', sizeof (*new_thread)); + *out_new_thread = new_thread; + *out_new_thread_bottom = new_thread_bottom; + *out_guardaddr = guardaddr; + *out_guardsize = guardsize; + return 0; +} + +static int pthread_handle_create(pthread_t *thread, const pthread_attr_t *attr, + void * (*start_routine)(void *), void *arg, + sigset_t * mask, int father_pid, + int report_events, + td_thr_events_t *event_maskp) +{ + size_t sseg; + int pid; + pthread_descr new_thread; + char * new_thread_bottom; + pthread_t new_thread_id; + char *guardaddr = NULL; + size_t guardsize = 0; + int pagesize = __getpagesize(); + + /* First check whether we have to change the policy and if yes, whether + we can do this. Normally this should be done by examining the + return value of the sched_setscheduler call in pthread_start_thread + but this is hard to implement. FIXME */ + if (attr != NULL && attr->__schedpolicy != SCHED_OTHER && geteuid () != 0) + return EPERM; + /* Find a free segment for the thread, and allocate a stack if needed */ + for (sseg = 2; ; sseg++) + { + if (sseg >= PTHREAD_THREADS_MAX) + return EAGAIN; + if (__pthread_handles[sseg].h_descr != NULL) + continue; + if (pthread_allocate_stack(attr, thread_segment(sseg), pagesize, + &new_thread, &new_thread_bottom, + &guardaddr, &guardsize) == 0) + break; + } + __pthread_handles_num++; + /* Allocate new thread identifier */ + pthread_threads_counter += PTHREAD_THREADS_MAX; + new_thread_id = sseg + pthread_threads_counter; + /* Initialize the thread descriptor. Elements which have to be + initialized to zero already have this value. */ + new_thread->p_tid = new_thread_id; + new_thread->p_lock = &(__pthread_handles[sseg].h_lock); + new_thread->p_cancelstate = PTHREAD_CANCEL_ENABLE; + new_thread->p_canceltype = PTHREAD_CANCEL_DEFERRED; + new_thread->p_errnop = &new_thread->p_errno; + new_thread->p_h_errnop = &new_thread->p_h_errno; + new_thread->p_guardaddr = guardaddr; + new_thread->p_guardsize = guardsize; + new_thread->p_self = new_thread; + new_thread->p_nr = sseg; + /* Initialize the thread handle */ + __pthread_init_lock(&__pthread_handles[sseg].h_lock); + __pthread_handles[sseg].h_descr = new_thread; + __pthread_handles[sseg].h_bottom = new_thread_bottom; + /* Determine scheduling parameters for the thread */ + new_thread->p_start_args.schedpolicy = -1; + if (attr != NULL) { + new_thread->p_detached = attr->__detachstate; + new_thread->p_userstack = attr->__stackaddr_set; + + switch(attr->__inheritsched) { + case PTHREAD_EXPLICIT_SCHED: + new_thread->p_start_args.schedpolicy = attr->__schedpolicy; + memcpy (&new_thread->p_start_args.schedparam, &attr->__schedparam, + sizeof (struct sched_param)); + break; + case PTHREAD_INHERIT_SCHED: + new_thread->p_start_args.schedpolicy = sched_getscheduler(father_pid); + sched_getparam(father_pid, &new_thread->p_start_args.schedparam); + break; + } + new_thread->p_priority = + new_thread->p_start_args.schedparam.sched_priority; + } + /* Finish setting up arguments to pthread_start_thread */ + new_thread->p_start_args.start_routine = start_routine; + new_thread->p_start_args.arg = arg; + new_thread->p_start_args.mask = *mask; + /* Raise priority of thread manager if needed */ + __pthread_manager_adjust_prio(new_thread->p_priority); + /* Do the cloning. We have to use two different functions depending + on whether we are debugging or not. */ + pid = 0; /* Note that the thread never can have PID zero. */ + if (report_events) + { + /* See whether the TD_CREATE event bit is set in any of the + masks. */ + int idx = __td_eventword (TD_CREATE); + uint32_t mask = __td_eventmask (TD_CREATE); + + if ((mask & (__pthread_threads_events.event_bits[idx] + | event_maskp->event_bits[idx])) != 0) + { + /* Lock the mutex the child will use now so that it will stop. */ + __pthread_lock(new_thread->p_lock, NULL); + + /* We have to report this event. */ + pid = clone(pthread_start_thread_event, (void **) new_thread, + CLONE_VM | CLONE_FS | CLONE_FILES | CLONE_SIGHAND | + __pthread_sig_cancel, new_thread); + if (pid != -1) + { + /* Now fill in the information about the new thread in + the newly created thread's data structure. We cannot let + the new thread do this since we don't know whether it was + already scheduled when we send the event. */ + new_thread->p_eventbuf.eventdata = new_thread; + new_thread->p_eventbuf.eventnum = TD_CREATE; + __pthread_last_event = new_thread; + + /* We have to set the PID here since the callback function + in the debug library will need it and we cannot guarantee + the child got scheduled before the debugger. */ + new_thread->p_pid = pid; + + /* Now call the function which signals the event. */ + __linuxthreads_create_event (); + + /* Now restart the thread. */ + __pthread_unlock(new_thread->p_lock); + } + } + } + if (pid == 0) +PDEBUG("cloning new_thread = %p\n", new_thread); + pid = clone(pthread_start_thread, (void **) new_thread, + CLONE_VM | CLONE_FS | CLONE_FILES | CLONE_SIGHAND | + __pthread_sig_cancel, new_thread); + /* Check if cloning succeeded */ + if (pid == -1) { + /* Free the stack if we allocated it */ + if (attr == NULL || !attr->__stackaddr_set) + { +#ifdef __UCLIBC_HAS_MMU__ + if (new_thread->p_guardsize != 0) + munmap(new_thread->p_guardaddr, new_thread->p_guardsize); + munmap((caddr_t)((char *)(new_thread+1) - INITIAL_STACK_SIZE), + INITIAL_STACK_SIZE); +#else + free(new_thread_bottom); +#endif /* __UCLIBC_HAS_MMU__ */ + } + __pthread_handles[sseg].h_descr = NULL; + __pthread_handles[sseg].h_bottom = NULL; + __pthread_handles_num--; + return errno; + } +PDEBUG("new thread pid = %d\n", pid); + /* Insert new thread in doubly linked list of active threads */ + new_thread->p_prevlive = __pthread_main_thread; + new_thread->p_nextlive = __pthread_main_thread->p_nextlive; + __pthread_main_thread->p_nextlive->p_prevlive = new_thread; + __pthread_main_thread->p_nextlive = new_thread; + /* Set pid field of the new thread, in case we get there before the + child starts. */ + new_thread->p_pid = pid; + /* We're all set */ + *thread = new_thread_id; + return 0; +} + + +/* Try to free the resources of a thread when requested by pthread_join + or pthread_detach on a terminated thread. */ + +static void pthread_free(pthread_descr th) +{ + pthread_handle handle; + pthread_readlock_info *iter, *next; + char *h_bottom_save; + + ASSERT(th->p_exited); + /* Make the handle invalid */ + handle = thread_handle(th->p_tid); + __pthread_lock(&handle->h_lock, NULL); + h_bottom_save = handle->h_bottom; + handle->h_descr = NULL; + handle->h_bottom = (char *)(-1L); + __pthread_unlock(&handle->h_lock); +#ifdef FREE_THREAD_SELF + FREE_THREAD_SELF(th, th->p_nr); +#endif + /* One fewer threads in __pthread_handles */ + __pthread_handles_num--; + + /* Destroy read lock list, and list of free read lock structures. + If the former is not empty, it means the thread exited while + holding read locks! */ + + for (iter = th->p_readlock_list; iter != NULL; iter = next) + { + next = iter->pr_next; + free(iter); + } + + for (iter = th->p_readlock_free; iter != NULL; iter = next) + { + next = iter->pr_next; + free(iter); + } + + /* If initial thread, nothing to free */ + if (th == &__pthread_initial_thread) return; +#ifdef __UCLIBC_HAS_MMU__ + if (!th->p_userstack) + { + /* Free the stack and thread descriptor area */ + if (th->p_guardsize != 0) + munmap(th->p_guardaddr, th->p_guardsize); + munmap((caddr_t) ((char *)(th+1) - STACK_SIZE), STACK_SIZE); + } +#else + /* For non-MMU systems we always malloc the stack, so free it here. -StS */ + if (!th->p_userstack) { + free(h_bottom_save); + } +#endif /* __UCLIBC_HAS_MMU__ */ +} + +/* Handle threads that have exited */ + +static void pthread_exited(pid_t pid) +{ + pthread_descr th; + int detached; + /* Find thread with that pid */ + for (th = __pthread_main_thread->p_nextlive; + th != __pthread_main_thread; + th = th->p_nextlive) { + if (th->p_pid == pid) { + /* Remove thread from list of active threads */ + th->p_nextlive->p_prevlive = th->p_prevlive; + th->p_prevlive->p_nextlive = th->p_nextlive; + /* Mark thread as exited, and if detached, free its resources */ + __pthread_lock(th->p_lock, NULL); + th->p_exited = 1; + /* If we have to signal this event do it now. */ + if (th->p_report_events) + { + /* See whether TD_DEATH is in any of the mask. */ + int idx = __td_eventword (TD_REAP); + uint32_t mask = __td_eventmask (TD_REAP); + + if ((mask & (__pthread_threads_events.event_bits[idx] + | th->p_eventbuf.eventmask.event_bits[idx])) != 0) + { + /* Yep, we have to signal the death. */ + th->p_eventbuf.eventnum = TD_DEATH; + th->p_eventbuf.eventdata = th; + __pthread_last_event = th; + + /* Now call the function to signal the event. */ + __linuxthreads_reap_event(); + } + } + detached = th->p_detached; + __pthread_unlock(th->p_lock); + if (detached) + pthread_free(th); + break; + } + } + /* If all threads have exited and the main thread is pending on a + pthread_exit, wake up the main thread and terminate ourselves. */ + if (main_thread_exiting && + __pthread_main_thread->p_nextlive == __pthread_main_thread) { + restart(__pthread_main_thread); + _exit(0); + } +} + +static void pthread_reap_children(void) +{ + pid_t pid; + int status; +PDEBUG("\n"); + + while ((pid = __libc_waitpid(-1, &status, WNOHANG | __WCLONE)) > 0) { + pthread_exited(pid); + if (WIFSIGNALED(status)) { + /* If a thread died due to a signal, send the same signal to + all other threads, including the main thread. */ + pthread_kill_all_threads(WTERMSIG(status), 1); + _exit(0); + } + } +} + +/* Try to free the resources of a thread when requested by pthread_join + or pthread_detach on a terminated thread. */ + +static void pthread_handle_free(pthread_t th_id) +{ + pthread_handle handle = thread_handle(th_id); + pthread_descr th; + + __pthread_lock(&handle->h_lock, NULL); + if (invalid_handle(handle, th_id)) { + /* pthread_reap_children has deallocated the thread already, + nothing needs to be done */ + __pthread_unlock(&handle->h_lock); + return; + } + th = handle->h_descr; + if (th->p_exited) { + __pthread_unlock(&handle->h_lock); + pthread_free(th); + } else { + /* The Unix process of the thread is still running. + Mark the thread as detached so that the thread manager will + deallocate its resources when the Unix process exits. */ + th->p_detached = 1; + __pthread_unlock(&handle->h_lock); + } +} + +/* Send a signal to all running threads */ + +static void pthread_kill_all_threads(int sig, int main_thread_also) +{ + pthread_descr th; + for (th = __pthread_main_thread->p_nextlive; + th != __pthread_main_thread; + th = th->p_nextlive) { + kill(th->p_pid, sig); + } + if (main_thread_also) { + kill(__pthread_main_thread->p_pid, sig); + } +} + +/* Process-wide exit() */ + +static void pthread_handle_exit(pthread_descr issuing_thread, int exitcode) +{ + pthread_descr th; + __pthread_exit_requested = 1; + __pthread_exit_code = exitcode; + /* Send the CANCEL signal to all running threads, including the main + thread, but excluding the thread from which the exit request originated + (that thread must complete the exit, e.g. calling atexit functions + and flushing stdio buffers). */ + for (th = issuing_thread->p_nextlive; + th != issuing_thread; + th = th->p_nextlive) { + kill(th->p_pid, __pthread_sig_cancel); + } + /* Now, wait for all these threads, so that they don't become zombies + and their times are properly added to the thread manager's times. */ + for (th = issuing_thread->p_nextlive; + th != issuing_thread; + th = th->p_nextlive) { + waitpid(th->p_pid, NULL, __WCLONE); + } + restart(issuing_thread); + _exit(0); +} + +/* Handler for __pthread_sig_cancel in thread manager thread */ + +void __pthread_manager_sighandler(int sig) +{ + terminated_children = 1; +} + +/* Adjust priority of thread manager so that it always run at a priority + higher than all threads */ + +void __pthread_manager_adjust_prio(int thread_prio) +{ + struct sched_param param; + + if (thread_prio <= __pthread_manager_thread.p_priority) return; + param.sched_priority = + thread_prio < sched_get_priority_max(SCHED_FIFO) + ? thread_prio + 1 : thread_prio; + sched_setscheduler(__pthread_manager_thread.p_pid, SCHED_FIFO, ¶m); + __pthread_manager_thread.p_priority = thread_prio; +} diff --git a/libpthread/linuxthreads/mutex.c b/libpthread/linuxthreads/mutex.c new file mode 100644 index 000000000..2217a504e --- /dev/null +++ b/libpthread/linuxthreads/mutex.c @@ -0,0 +1,201 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* changes for uClibc: remove strong_alias'es and define the real symbol */ + +/* Mutexes */ + +#include <errno.h> +#include <sched.h> +#include <stddef.h> +#include "pthread.h" +#include "internals.h" +#include "spinlock.h" +#include "queue.h" +#include "restart.h" + +int pthread_mutex_init(pthread_mutex_t * mutex, + const pthread_mutexattr_t * mutex_attr) +{ + __pthread_init_lock(&mutex->__m_lock); + mutex->__m_kind = + mutex_attr == NULL ? PTHREAD_MUTEX_FAST_NP : mutex_attr->__mutexkind; + mutex->__m_count = 0; + mutex->__m_owner = NULL; + return 0; +} +//strong_alias (__pthread_mutex_init, pthread_mutex_init) + +int pthread_mutex_destroy(pthread_mutex_t * mutex) +{ + if (mutex->__m_lock.__status != 0) return EBUSY; + return 0; +} +//strong_alias (__pthread_mutex_destroy, pthread_mutex_destroy) + +int pthread_mutex_trylock(pthread_mutex_t * mutex) +{ + pthread_descr self; + int retcode; + + switch(mutex->__m_kind) { + case PTHREAD_MUTEX_FAST_NP: + retcode = __pthread_trylock(&mutex->__m_lock); + return retcode; + case PTHREAD_MUTEX_RECURSIVE_NP: + self = thread_self(); + if (mutex->__m_owner == self) { + mutex->__m_count++; + return 0; + } + retcode = __pthread_trylock(&mutex->__m_lock); + if (retcode == 0) { + mutex->__m_owner = self; + mutex->__m_count = 0; + } + return retcode; + case PTHREAD_MUTEX_ERRORCHECK_NP: + retcode = __pthread_trylock(&mutex->__m_lock); + if (retcode == 0) { + mutex->__m_owner = thread_self(); + } + return retcode; + default: + return EINVAL; + } +} +//strong_alias (__pthread_mutex_trylock, pthread_mutex_trylock) + +int pthread_mutex_lock(pthread_mutex_t * mutex) +{ + pthread_descr self; + + switch(mutex->__m_kind) { + case PTHREAD_MUTEX_FAST_NP: + __pthread_lock(&mutex->__m_lock, NULL); + return 0; + case PTHREAD_MUTEX_RECURSIVE_NP: + self = thread_self(); + if (mutex->__m_owner == self) { + mutex->__m_count++; + return 0; + } + __pthread_lock(&mutex->__m_lock, self); + mutex->__m_owner = self; + mutex->__m_count = 0; + return 0; + case PTHREAD_MUTEX_ERRORCHECK_NP: + self = thread_self(); + if (mutex->__m_owner == self) return EDEADLK; + __pthread_lock(&mutex->__m_lock, self); + mutex->__m_owner = self; + return 0; + default: + return EINVAL; + } +} +//strong_alias (__pthread_mutex_lock, pthread_mutex_lock) + +int pthread_mutex_unlock(pthread_mutex_t * mutex) +{ + switch (mutex->__m_kind) { + case PTHREAD_MUTEX_FAST_NP: + __pthread_unlock(&mutex->__m_lock); + return 0; + case PTHREAD_MUTEX_RECURSIVE_NP: + if (mutex->__m_count > 0) { + mutex->__m_count--; + return 0; + } + mutex->__m_owner = NULL; + __pthread_unlock(&mutex->__m_lock); + return 0; + case PTHREAD_MUTEX_ERRORCHECK_NP: + if (mutex->__m_owner != thread_self() || mutex->__m_lock.__status == 0) + return EPERM; + mutex->__m_owner = NULL; + __pthread_unlock(&mutex->__m_lock); + return 0; + default: + return EINVAL; + } +} +//strong_alias (__pthread_mutex_unlock, pthread_mutex_unlock) + +int pthread_mutexattr_init(pthread_mutexattr_t *attr) +{ + attr->__mutexkind = PTHREAD_MUTEX_FAST_NP; + return 0; +} +//strong_alias (__pthread_mutexattr_init, pthread_mutexattr_init) + +int pthread_mutexattr_destroy(pthread_mutexattr_t *attr) +{ + return 0; +} +//strong_alias (__pthread_mutexattr_destroy, pthread_mutexattr_destroy) + +int __pthread_mutexattr_settype(pthread_mutexattr_t *attr, int kind) +{ + if (kind != PTHREAD_MUTEX_FAST_NP + && kind != PTHREAD_MUTEX_RECURSIVE_NP + && kind != PTHREAD_MUTEX_ERRORCHECK_NP) + return EINVAL; + attr->__mutexkind = kind; + return 0; +} +weak_alias (__pthread_mutexattr_settype, pthread_mutexattr_settype) +weak_alias ( __pthread_mutexattr_settype, __pthread_mutexattr_setkind_np) +weak_alias (__pthread_mutexattr_setkind_np, pthread_mutexattr_setkind_np) + +int __pthread_mutexattr_gettype(const pthread_mutexattr_t *attr, int *kind) +{ + *kind = attr->__mutexkind; + return 0; +} +weak_alias (__pthread_mutexattr_gettype, pthread_mutexattr_gettype) +weak_alias (__pthread_mutexattr_gettype, __pthread_mutexattr_getkind_np) +weak_alias (__pthread_mutexattr_getkind_np, pthread_mutexattr_getkind_np) + +/* Once-only execution */ + +static pthread_mutex_t once_masterlock = PTHREAD_MUTEX_INITIALIZER; +static pthread_cond_t once_finished = PTHREAD_COND_INITIALIZER; + +enum { NEVER = 0, IN_PROGRESS = 1, DONE = 2 }; + +int pthread_once(pthread_once_t * once_control, void (*init_routine)(void)) +{ + /* Test without locking first for speed */ + if (*once_control == DONE) return 0; + /* Lock and test again */ + pthread_mutex_lock(&once_masterlock); + /* If init_routine is being called from another routine, wait until + it completes. */ + while (*once_control == IN_PROGRESS) { + pthread_cond_wait(&once_finished, &once_masterlock); + } + /* Here *once_control is stable and either NEVER or DONE. */ + if (*once_control == NEVER) { + *once_control = IN_PROGRESS; + pthread_mutex_unlock(&once_masterlock); + init_routine(); + pthread_mutex_lock(&once_masterlock); + *once_control = DONE; + pthread_cond_broadcast(&once_finished); + } + pthread_mutex_unlock(&once_masterlock); + return 0; +} +//strong_alias (__pthread_once, pthread_once) diff --git a/libpthread/linuxthreads/no-tsd.c b/libpthread/linuxthreads/no-tsd.c new file mode 100644 index 000000000..ef79cb832 --- /dev/null +++ b/libpthread/linuxthreads/no-tsd.c @@ -0,0 +1,33 @@ +/* libc-internal interface for thread-specific data. + Copyright (C) 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#include <bits/libc-tsd.h> + +/* This file provides uinitialized (common) definitions for the + hooks used internally by libc to access thread-specific data. + + When -lpthread is used, it provides initialized definitions for these + variables (in specific.c), which override these uninitialized definitions. + + If -lpthread is not used, these uninitialized variables default to zero, + which the __libc_tsd_* macros check for. */ + +void *(*__libc_internal_tsd_get) __P ((enum __libc_tsd_key_t)); +int (*__libc_internal_tsd_set) __P ((enum __libc_tsd_key_t, + __const void *)); diff --git a/libpthread/linuxthreads/oldsemaphore.c b/libpthread/linuxthreads/oldsemaphore.c new file mode 100644 index 000000000..a634bad8e --- /dev/null +++ b/libpthread/linuxthreads/oldsemaphore.c @@ -0,0 +1,235 @@ +/* + * This file contains the old semaphore code that we need to + * preserve for glibc-2.0 backwards compatibility. Port to glibc 2.1 + * done by Cristian Gafton. + */ + +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* Semaphores a la POSIX 1003.1b */ + +#include <errno.h> +#include "pthread.h" +#include "internals.h" +#include "spinlock.h" +#include "restart.h" +#include "queue.h" + +typedef struct { + long int sem_status; + int sem_spinlock; +} old_sem_t; + +/* Maximum value the semaphore can have. */ +#define SEM_VALUE_MAX ((int) ((~0u) >> 1)) + +static inline int sem_compare_and_swap(old_sem_t *sem, long oldval, long newval) +{ + return compare_and_swap(&sem->sem_status, oldval, newval, &sem->sem_spinlock); +} + +/* The state of a semaphore is represented by a long int encoding + either the semaphore count if >= 0 and no thread is waiting on it, + or the head of the list of threads waiting for the semaphore. + To distinguish the two cases, we encode the semaphore count N + as 2N+1, so that it has the lowest bit set. + + A sequence of sem_wait operations on a semaphore initialized to N + result in the following successive states: + 2N+1, 2N-1, ..., 3, 1, &first_waiting_thread, &second_waiting_thread, ... +*/ + +static void sem_restart_list(pthread_descr waiting); + +int __old_sem_init(old_sem_t *sem, int pshared, unsigned int value) +{ + if (value > SEM_VALUE_MAX) { + errno = EINVAL; + return -1; + } + if (pshared) { + errno = ENOSYS; + return -1; + } + sem->sem_spinlock = 0; + sem->sem_status = ((long)value << 1) + 1; + return 0; +} + +/* Function called by pthread_cancel to remove the thread from + waiting inside __old_sem_wait. Here we simply unconditionally + indicate that the thread is to be woken, by returning 1. */ + +static int old_sem_extricate_func(void *obj, pthread_descr th) +{ + return 1; +} + +int __old_sem_wait(old_sem_t * sem) +{ + long oldstatus, newstatus; + volatile pthread_descr self = thread_self(); + pthread_descr * th; + pthread_extricate_if extr; + + /* Set up extrication interface */ + extr.pu_object = 0; + extr.pu_extricate_func = old_sem_extricate_func; + + while (1) { + /* Register extrication interface */ + __pthread_set_own_extricate_if(self, &extr); + do { + oldstatus = sem->sem_status; + if ((oldstatus & 1) && (oldstatus != 1)) + newstatus = oldstatus - 2; + else { + newstatus = (long) self; + self->p_nextwaiting = (pthread_descr) oldstatus; + } + } + while (! sem_compare_and_swap(sem, oldstatus, newstatus)); + if (newstatus & 1) { + /* We got the semaphore. */ + __pthread_set_own_extricate_if(self, 0); + return 0; + } + /* Wait for sem_post or cancellation */ + suspend(self); + __pthread_set_own_extricate_if(self, 0); + + /* This is a cancellation point */ + if (self->p_canceled && self->p_cancelstate == PTHREAD_CANCEL_ENABLE) { + /* Remove ourselves from the waiting list if we're still on it */ + /* First check if we're at the head of the list. */ + do { + oldstatus = sem->sem_status; + if (oldstatus != (long) self) break; + newstatus = (long) self->p_nextwaiting; + } + while (! sem_compare_and_swap(sem, oldstatus, newstatus)); + /* Now, check if we're somewhere in the list. + There's a race condition with sem_post here, but it does not matter: + the net result is that at the time pthread_exit is called, + self is no longer reachable from sem->sem_status. */ + if (oldstatus != (long) self && (oldstatus & 1) == 0) { + for (th = &(((pthread_descr) oldstatus)->p_nextwaiting); + *th != NULL && *th != (pthread_descr) 1; + th = &((*th)->p_nextwaiting)) { + if (*th == self) { + *th = self->p_nextwaiting; + break; + } + } + } + pthread_exit(PTHREAD_CANCELED); + } + } +} + +int __old_sem_trywait(old_sem_t * sem) +{ + long oldstatus, newstatus; + + do { + oldstatus = sem->sem_status; + if ((oldstatus & 1) == 0 || (oldstatus == 1)) { + errno = EAGAIN; + return -1; + } + newstatus = oldstatus - 2; + } + while (! sem_compare_and_swap(sem, oldstatus, newstatus)); + return 0; +} + +int __old_sem_post(old_sem_t * sem) +{ + long oldstatus, newstatus; + + do { + oldstatus = sem->sem_status; + if ((oldstatus & 1) == 0) + newstatus = 3; + else { + if (oldstatus >= SEM_VALUE_MAX) { + /* Overflow */ + errno = ERANGE; + return -1; + } + newstatus = oldstatus + 2; + } + } + while (! sem_compare_and_swap(sem, oldstatus, newstatus)); + if ((oldstatus & 1) == 0) + sem_restart_list((pthread_descr) oldstatus); + return 0; +} + +int __old_sem_getvalue(old_sem_t * sem, int * sval) +{ + long status = sem->sem_status; + if (status & 1) + *sval = (int)((unsigned long) status >> 1); + else + *sval = 0; + return 0; +} + +int __old_sem_destroy(old_sem_t * sem) +{ + if ((sem->sem_status & 1) == 0) { + errno = EBUSY; + return -1; + } + return 0; +} + +/* Auxiliary function for restarting all threads on a waiting list, + in priority order. */ + +static void sem_restart_list(pthread_descr waiting) +{ + pthread_descr th, towake, *p; + + /* Sort list of waiting threads by decreasing priority (insertion sort) */ + towake = NULL; + while (waiting != (pthread_descr) 1) { + th = waiting; + waiting = waiting->p_nextwaiting; + p = &towake; + while (*p != NULL && th->p_priority < (*p)->p_priority) + p = &((*p)->p_nextwaiting); + th->p_nextwaiting = *p; + *p = th; + } + /* Wake up threads in priority order */ + while (towake != NULL) { + th = towake; + towake = towake->p_nextwaiting; + th->p_nextwaiting = NULL; + restart(th); + } +} + +#if defined PIC && DO_VERSIONING +symbol_version (__old_sem_init, sem_init, GLIBC_2.0); +symbol_version (__old_sem_wait, sem_wait, GLIBC_2.0); +symbol_version (__old_sem_trywait, sem_trywait, GLIBC_2.0); +symbol_version (__old_sem_post, sem_post, GLIBC_2.0); +symbol_version (__old_sem_getvalue, sem_getvalue, GLIBC_2.0); +symbol_version (__old_sem_destroy, sem_destroy, GLIBC_2.0); +#endif + diff --git a/libpthread/linuxthreads/pt-machine.c b/libpthread/linuxthreads/pt-machine.c new file mode 100644 index 000000000..438008d5d --- /dev/null +++ b/libpthread/linuxthreads/pt-machine.c @@ -0,0 +1,22 @@ +/* "Instantiation of machine-dependent pthreads inline functions. + Copyright (C) 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#define PT_EI + +#include <pt-machine.h> diff --git a/libpthread/linuxthreads/ptfork.c b/libpthread/linuxthreads/ptfork.c new file mode 100644 index 000000000..0c4e252d6 --- /dev/null +++ b/libpthread/linuxthreads/ptfork.c @@ -0,0 +1,107 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* mods for uClibc: removed strong alias and defined funcs properly */ + +/* The "atfork" stuff */ + +#include <errno.h> +#include <stddef.h> +#include <stdlib.h> +#include <unistd.h> +#include "pthread.h" +#include "internals.h" + +struct handler_list { + void (*handler)(void); + struct handler_list * next; +}; + +static pthread_mutex_t pthread_atfork_lock = PTHREAD_MUTEX_INITIALIZER; +static struct handler_list * pthread_atfork_prepare = NULL; +static struct handler_list * pthread_atfork_parent = NULL; +static struct handler_list * pthread_atfork_child = NULL; + +static void pthread_insert_list(struct handler_list ** list, + void (*handler)(void), + struct handler_list * newlist, + int at_end) +{ + if (handler == NULL) return; + if (at_end) { + while(*list != NULL) list = &((*list)->next); + } + newlist->handler = handler; + newlist->next = *list; + *list = newlist; +} + +struct handler_list_block { + struct handler_list prepare, parent, child; +}; + +int pthread_atfork(void (*prepare)(void), + void (*parent)(void), + void (*child)(void)) +{ + struct handler_list_block * block = + (struct handler_list_block *) malloc(sizeof(struct handler_list_block)); + if (block == NULL) return ENOMEM; + pthread_mutex_lock(&pthread_atfork_lock); + /* "prepare" handlers are called in LIFO */ + pthread_insert_list(&pthread_atfork_prepare, prepare, &block->prepare, 0); + /* "parent" handlers are called in FIFO */ + pthread_insert_list(&pthread_atfork_parent, parent, &block->parent, 1); + /* "child" handlers are called in FIFO */ + pthread_insert_list(&pthread_atfork_child, child, &block->child, 1); + pthread_mutex_unlock(&pthread_atfork_lock); + return 0; +} +//strong_alias (__pthread_atfork, pthread_atfork) + +static inline void pthread_call_handlers(struct handler_list * list) +{ + for (/*nothing*/; list != NULL; list = list->next) (list->handler)(); +} + +extern int __libc_fork(void); + +pid_t __fork(void) +{ + pid_t pid; + struct handler_list * prepare, * child, * parent; + + pthread_mutex_lock(&pthread_atfork_lock); + prepare = pthread_atfork_prepare; + child = pthread_atfork_child; + parent = pthread_atfork_parent; + pthread_mutex_unlock(&pthread_atfork_lock); + pthread_call_handlers(prepare); + pid = __libc_fork(); + if (pid == 0) { + __pthread_reset_main_thread(); + __fresetlockfiles(); + pthread_call_handlers(child); + } else { + pthread_call_handlers(parent); + } + return pid; +} +weak_alias (__fork, fork); + +pid_t __vfork(void) +{ + return __fork(); +} +weak_alias (__vfork, vfork); diff --git a/libpthread/linuxthreads/pthread.c b/libpthread/linuxthreads/pthread.c new file mode 100644 index 000000000..51d600a18 --- /dev/null +++ b/libpthread/linuxthreads/pthread.c @@ -0,0 +1,876 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* Thread creation, initialization, and basic low-level routines */ + +#define __FORCE_GLIBC +#include <features.h> +#include <errno.h> +#include <netdb.h> /* for h_errno */ +#include <stddef.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <unistd.h> +#include <fcntl.h> +#include <sys/wait.h> +#include <sys/resource.h> +#include "pthread.h" +#include "internals.h" +#include "spinlock.h" +#include "restart.h" +#include "debug.h" /* added to linuxthreads -StS */ + +/* mods for uClibc: getpwd and getpagesize are the syscalls */ +#define __getpid getpid +#define __getpagesize getpagesize +/* mods for uClibc: __libc_sigaction is not in any standard headers */ +#include <signal.h> +extern int __libc_sigaction (int sig, const struct sigaction *act, struct sigaction *oact); + + +/* These variables are used by the setup code. */ +extern int _errno; +extern int _h_errno; + + +/* Descriptor of the initial thread */ + +struct _pthread_descr_struct __pthread_initial_thread = { + &__pthread_initial_thread, /* pthread_descr p_nextlive */ + &__pthread_initial_thread, /* pthread_descr p_prevlive */ + NULL, /* pthread_descr p_nextwaiting */ + NULL, /* pthread_descr p_nextlock */ + PTHREAD_THREADS_MAX, /* pthread_t p_tid */ + 0, /* int p_pid */ + 0, /* int p_priority */ + &__pthread_handles[0].h_lock, /* struct _pthread_fastlock * p_lock */ + 0, /* int p_signal */ + NULL, /* sigjmp_buf * p_signal_buf */ + NULL, /* sigjmp_buf * p_cancel_buf */ + 0, /* char p_terminated */ + 0, /* char p_detached */ + 0, /* char p_exited */ + NULL, /* void * p_retval */ + 0, /* int p_retval */ + NULL, /* pthread_descr p_joining */ + NULL, /* struct _pthread_cleanup_buffer * p_cleanup */ + 0, /* char p_cancelstate */ + 0, /* char p_canceltype */ + 0, /* char p_canceled */ + &_errno, /* int *p_errnop */ + 0, /* int p_errno */ + &_h_errno, /* int *p_h_errnop */ + 0, /* int p_h_errno */ + NULL, /* char * p_in_sighandler */ + 0, /* char p_sigwaiting */ + PTHREAD_START_ARGS_INITIALIZER, /* struct pthread_start_args p_start_args */ + {NULL}, /* void ** p_specific[PTHREAD_KEY_1STLEVEL_SIZE] */ + {NULL}, /* void * p_libc_specific[_LIBC_TSD_KEY_N] */ + 0, /* int p_userstack */ + NULL, /* void * p_guardaddr */ + 0, /* size_t p_guardsize */ + &__pthread_initial_thread, /* pthread_descr p_self */ + 0, /* Always index 0 */ + 0, /* int p_report_events */ + {{{0, }}, 0, NULL}, /* td_eventbuf_t p_eventbuf */ + ATOMIC_INITIALIZER, /* struct pthread_atomic p_resume_count */ + 0, /* char p_woken_by_cancel */ + NULL, /* struct pthread_extricate_if *p_extricate */ + NULL, /* pthread_readlock_info *p_readlock_list; */ + NULL, /* pthread_readlock_info *p_readlock_free; */ + 0 /* int p_untracked_readlock_count; */ +}; + +/* Descriptor of the manager thread; none of this is used but the error + variables, the p_pid and p_priority fields, + and the address for identification. */ + +struct _pthread_descr_struct __pthread_manager_thread = { + NULL, /* pthread_descr p_nextlive */ + NULL, /* pthread_descr p_prevlive */ + NULL, /* pthread_descr p_nextwaiting */ + NULL, /* pthread_descr p_nextlock */ + 0, /* int p_tid */ + 0, /* int p_pid */ + 0, /* int p_priority */ + &__pthread_handles[1].h_lock, /* struct _pthread_fastlock * p_lock */ + 0, /* int p_signal */ + NULL, /* sigjmp_buf * p_signal_buf */ + NULL, /* sigjmp_buf * p_cancel_buf */ + 0, /* char p_terminated */ + 0, /* char p_detached */ + 0, /* char p_exited */ + NULL, /* void * p_retval */ + 0, /* int p_retval */ + NULL, /* pthread_descr p_joining */ + NULL, /* struct _pthread_cleanup_buffer * p_cleanup */ + 0, /* char p_cancelstate */ + 0, /* char p_canceltype */ + 0, /* char p_canceled */ + &__pthread_manager_thread.p_errno, /* int *p_errnop */ + 0, /* int p_errno */ + NULL, /* int *p_h_errnop */ + 0, /* int p_h_errno */ + NULL, /* char * p_in_sighandler */ + 0, /* char p_sigwaiting */ + PTHREAD_START_ARGS_INITIALIZER, /* struct pthread_start_args p_start_args */ + {NULL}, /* void ** p_specific[PTHREAD_KEY_1STLEVEL_SIZE] */ + {NULL}, /* void * p_libc_specific[_LIBC_TSD_KEY_N] */ + 0, /* int p_userstack */ + NULL, /* void * p_guardaddr */ + 0, /* size_t p_guardsize */ + &__pthread_manager_thread, /* pthread_descr p_self */ + 1, /* Always index 1 */ + 0, /* int p_report_events */ + {{{0, }}, 0, NULL}, /* td_eventbuf_t p_eventbuf */ + ATOMIC_INITIALIZER, /* struct pthread_atomic p_resume_count */ + 0, /* char p_woken_by_cancel */ + NULL, /* struct pthread_extricate_if *p_extricate */ + NULL, /* pthread_readlock_info *p_readlock_list; */ + NULL, /* pthread_readlock_info *p_readlock_free; */ + 0 /* int p_untracked_readlock_count; */ +}; + +/* Pointer to the main thread (the father of the thread manager thread) */ +/* Originally, this is the initial thread, but this changes after fork() */ + +pthread_descr __pthread_main_thread = &__pthread_initial_thread; + +/* Limit between the stack of the initial thread (above) and the + stacks of other threads (below). Aligned on a STACK_SIZE boundary. */ + +char *__pthread_initial_thread_bos = NULL; + +/* For non-MMU systems also remember to stack top of the initial thread. + * This is adapted when other stacks are malloc'ed since we don't know + * the bounds a-priori. -StS */ + +#ifndef __UCLIBC_HAS_MMU__ +char *__pthread_initial_thread_tos = NULL; +#endif /* __UCLIBC_HAS_MMU__ */ + +/* File descriptor for sending requests to the thread manager. */ +/* Initially -1, meaning that the thread manager is not running. */ + +int __pthread_manager_request = -1; + +/* Other end of the pipe for sending requests to the thread manager. */ + +int __pthread_manager_reader; + +/* Limits of the thread manager stack */ + +char *__pthread_manager_thread_bos = NULL; +char *__pthread_manager_thread_tos = NULL; + +/* For process-wide exit() */ + +int __pthread_exit_requested = 0; +int __pthread_exit_code = 0; + +/* Pointers that select new or old suspend/resume functions + based on availability of rt signals. */ + +void (*__pthread_restart)(pthread_descr) = __pthread_restart_old; +void (*__pthread_suspend)(pthread_descr) = __pthread_suspend_old; + +/* Communicate relevant LinuxThreads constants to gdb */ + +const int __pthread_threads_max = PTHREAD_THREADS_MAX; +const int __pthread_sizeof_handle = sizeof(struct pthread_handle_struct); +const int __pthread_offsetof_descr = offsetof(struct pthread_handle_struct, + h_descr); +const int __pthread_offsetof_pid = offsetof(struct _pthread_descr_struct, + p_pid); + +/* Forward declarations */ + +static void pthread_exit_process(int retcode, void *arg); +#ifndef __i386__ +static void pthread_handle_sigcancel(int sig); +static void pthread_handle_sigrestart(int sig); +#else +static void pthread_handle_sigcancel(int sig, struct sigcontext ctx); +static void pthread_handle_sigrestart(int sig, struct sigcontext ctx); +#endif +static void pthread_handle_sigdebug(int sig); + +/* Signal numbers used for the communication. + In these variables we keep track of the used variables. If the + platform does not support any real-time signals we will define the + values to some unreasonable value which will signal failing of all + the functions below. */ +#ifndef __SIGRTMIN +static int current_rtmin = -1; +static int current_rtmax = -1; +int __pthread_sig_restart = SIGUSR1; +int __pthread_sig_cancel = SIGUSR2; +int __pthread_sig_debug = 0; +#else +static int current_rtmin; +static int current_rtmax; + +#if __SIGRTMAX - __SIGRTMIN >= 3 +int __pthread_sig_restart = __SIGRTMIN; +int __pthread_sig_cancel = __SIGRTMIN + 1; +int __pthread_sig_debug = __SIGRTMIN + 2; +#else +int __pthread_sig_restart = SIGUSR1; +int __pthread_sig_cancel = SIGUSR2; +int __pthread_sig_debug = 0; +#endif + +static int rtsigs_initialized; + +#include "testrtsig.h" + +static void +init_rtsigs (void) +{ + if (!kernel_has_rtsig ()) + { + current_rtmin = -1; + current_rtmax = -1; +#if __SIGRTMAX - __SIGRTMIN >= 3 + __pthread_sig_restart = SIGUSR1; + __pthread_sig_cancel = SIGUSR2; + __pthread_sig_debug = 0; +#endif +PDEBUG("no rt-sigs, sig_restart=%d, sig_cancel=%d.\n", __pthread_sig_restart, __pthread_sig_cancel ); + __pthread_init_condvar(0); + } + else + { +#if __SIGRTMAX - __SIGRTMIN >= 3 + current_rtmin = __SIGRTMIN + 3; + __pthread_restart = __pthread_restart_new; + __pthread_suspend = __pthread_wait_for_restart_signal; + __pthread_init_condvar(1); +#else + current_rtmin = __SIGRTMIN; + __pthread_init_condvar(0); +#endif + + current_rtmax = __SIGRTMAX; +PDEBUG("have rt-sigs, rtmin = %d, rtmax = %d.\n", current_rtmin, current_rtmax); + } + + rtsigs_initialized = 1; +} +#endif + +/* Return number of available real-time signal with highest priority. */ +int +__libc_current_sigrtmin (void) +{ +#ifdef __SIGRTMIN + if (!rtsigs_initialized) + init_rtsigs (); +#endif + return current_rtmin; +} + +/* Return number of available real-time signal with lowest priority. */ +int +__libc_current_sigrtmax (void) +{ +#ifdef __SIGRTMIN + if (!rtsigs_initialized) + init_rtsigs (); +#endif + return current_rtmax; +} + +/* Allocate real-time signal with highest/lowest available + priority. Please note that we don't use a lock since we assume + this function to be called at program start. */ +int +__libc_allocate_rtsig (int high) +{ +#ifndef __SIGRTMIN + return -1; +#else + if (!rtsigs_initialized) + init_rtsigs (); + if (current_rtmin == -1 || current_rtmin > current_rtmax) + /* We don't have anymore signal available. */ + return -1; + + return high ? current_rtmin++ : current_rtmax--; +#endif +} + +/* Initialize the pthread library. + Initialization is split in two functions: + - a constructor function that blocks the __pthread_sig_restart signal + (must do this very early, since the program could capture the signal + mask with e.g. sigsetjmp before creating the first thread); + - a regular function called from pthread_create when needed. */ + +static void pthread_initialize(void) __attribute__((constructor)); + +static void pthread_initialize(void) +{ + struct sigaction sa; + sigset_t mask; + struct rlimit limit; + int max_stack; + + /* If already done (e.g. by a constructor called earlier!), bail out */ + if (__pthread_initial_thread_bos != NULL) return; +#ifdef TEST_FOR_COMPARE_AND_SWAP + /* Test if compare-and-swap is available */ + __pthread_has_cas = compare_and_swap_is_available(); +#endif + /* For the initial stack, reserve at least STACK_SIZE bytes of stack + below the current stack address, and align that on a + STACK_SIZE boundary. */ + __pthread_initial_thread_bos = + (char *)(((long)CURRENT_STACK_FRAME - 2 * STACK_SIZE) & ~(STACK_SIZE - 1)); + /* Update the descriptor for the initial thread. */ + __pthread_initial_thread.p_pid = __getpid(); + /* If we have special thread_self processing, initialize that for the + main thread now. */ +#ifdef INIT_THREAD_SELF + INIT_THREAD_SELF(&__pthread_initial_thread, 0); +#endif + /* The errno/h_errno variable of the main thread are the global ones. */ + __pthread_initial_thread.p_errnop = &_errno; + __pthread_initial_thread.p_h_errnop = &_h_errno; + /* Play with the stack size limit to make sure that no stack ever grows + beyond STACK_SIZE minus two pages (one page for the thread descriptor + immediately beyond, and one page to act as a guard page). */ + +#ifdef __UCLIBC_HAS_MMU__ + /* We cannot allocate a huge chunk of memory to mmap all thread stacks later + * on a non-MMU system. Thus, we don't need the rlimit either. -StS */ + getrlimit(RLIMIT_STACK, &limit); + max_stack = STACK_SIZE - 2 * __getpagesize(); + if (limit.rlim_cur > max_stack) { + limit.rlim_cur = max_stack; + setrlimit(RLIMIT_STACK, &limit); + } +#else + /* For non-MMU assume __pthread_initial_thread_tos at upper page boundary, and + * __pthread_initial_thread_bos at address 0. These bounds are refined as we + * malloc other stack frames such that they don't overlap. -StS + */ + __pthread_initial_thread_tos = + (char *)(((long)CURRENT_STACK_FRAME + __getpagesize()) & ~(__getpagesize() - 1)); + __pthread_initial_thread_bos = (char *) 1; /* set it non-zero so we know we have been here */ + PDEBUG("initial thread stack bounds: bos=%p, tos=%p\n", + __pthread_initial_thread_bos, __pthread_initial_thread_tos); +#endif /* __UCLIBC_HAS_MMU__ */ + +#ifdef __SIGRTMIN + /* Initialize real-time signals. */ + init_rtsigs (); +#endif + /* Setup signal handlers for the initial thread. + Since signal handlers are shared between threads, these settings + will be inherited by all other threads. */ +#ifndef __i386__ + sa.sa_handler = pthread_handle_sigrestart; +#else + sa.sa_handler = (__sighandler_t) pthread_handle_sigrestart; +#endif + sigemptyset(&sa.sa_mask); + sa.sa_flags = 0; + __libc_sigaction(__pthread_sig_restart, &sa, NULL); +#ifndef __i386__ + sa.sa_handler = pthread_handle_sigcancel; +#else + sa.sa_handler = (__sighandler_t) pthread_handle_sigcancel; +#endif + sa.sa_flags = 0; + __libc_sigaction(__pthread_sig_cancel, &sa, NULL); + if (__pthread_sig_debug > 0) { + sa.sa_handler = pthread_handle_sigdebug; + sigemptyset(&sa.sa_mask); + sa.sa_flags = 0; + __libc_sigaction(__pthread_sig_debug, &sa, NULL); + } + /* Initially, block __pthread_sig_restart. Will be unblocked on demand. */ + sigemptyset(&mask); + sigaddset(&mask, __pthread_sig_restart); +PDEBUG("block mask = %x\n", mask); + sigprocmask(SIG_BLOCK, &mask, NULL); + /* Register an exit function to kill all other threads. */ + /* Do it early so that user-registered atexit functions are called + before pthread_exit_process. */ + on_exit(pthread_exit_process, NULL); +} + +void __pthread_initialize(void) +{ + pthread_initialize(); +} + +int __pthread_initialize_manager(void) +{ + int manager_pipe[2]; + int pid; + struct pthread_request request; + + /* If basic initialization not done yet (e.g. we're called from a + constructor run before our constructor), do it now */ + if (__pthread_initial_thread_bos == NULL) pthread_initialize(); + /* Setup stack for thread manager */ + __pthread_manager_thread_bos = malloc(THREAD_MANAGER_STACK_SIZE); + if (__pthread_manager_thread_bos == NULL) return -1; + __pthread_manager_thread_tos = + __pthread_manager_thread_bos + THREAD_MANAGER_STACK_SIZE; + + /* On non-MMU systems we make sure that the initial thread bounds don't overlap + * with the manager stack frame */ + NOMMU_INITIAL_THREAD_BOUNDS(__pthread_manager_thread_tos,__pthread_manager_thread_bos); + PDEBUG("manager stack: size=%d, bos=%p, tos=%p\n", THREAD_MANAGER_STACK_SIZE, + __pthread_manager_thread_bos, __pthread_manager_thread_tos); + PDEBUG("initial stack: estimate bos=%p, tos=%p\n", + __pthread_initial_thread_bos, __pthread_initial_thread_tos); + + /* Setup pipe to communicate with thread manager */ + if (pipe(manager_pipe) == -1) { + free(__pthread_manager_thread_bos); + return -1; + } + /* Start the thread manager */ + pid = 0; + if (__pthread_initial_thread.p_report_events) + { + /* It's a bit more complicated. We have to report the creation of + the manager thread. */ + int idx = __td_eventword (TD_CREATE); + uint32_t mask = __td_eventmask (TD_CREATE); + + if ((mask & (__pthread_threads_events.event_bits[idx] + | __pthread_initial_thread.p_eventbuf.eventmask.event_bits[idx])) + != 0) + { + pid = clone(__pthread_manager_event, + (void **) __pthread_manager_thread_tos, + CLONE_VM | CLONE_FS | CLONE_FILES | CLONE_SIGHAND, + (void *)(long)manager_pipe[0]); + + if (pid != -1) + { + /* Now fill in the information about the new thread in + the newly created thread's data structure. We cannot let + the new thread do this since we don't know whether it was + already scheduled when we send the event. */ + __pthread_manager_thread.p_eventbuf.eventdata = + &__pthread_manager_thread; + __pthread_manager_thread.p_eventbuf.eventnum = TD_CREATE; + __pthread_last_event = &__pthread_manager_thread; + __pthread_manager_thread.p_tid = 2* PTHREAD_THREADS_MAX + 1; + __pthread_manager_thread.p_pid = pid; + + /* Now call the function which signals the event. */ + __linuxthreads_create_event (); + + /* Now restart the thread. */ + __pthread_unlock(__pthread_manager_thread.p_lock); + } + } + } + + if (pid == 0) { + pid = clone(__pthread_manager, (void **) __pthread_manager_thread_tos, + CLONE_VM | CLONE_FS | CLONE_FILES | CLONE_SIGHAND, + (void *)(long)manager_pipe[0]); + } + if (pid == -1) { + free(__pthread_manager_thread_bos); + __libc_close(manager_pipe[0]); + __libc_close(manager_pipe[1]); + return -1; + } + __pthread_manager_request = manager_pipe[1]; /* writing end */ + __pthread_manager_reader = manager_pipe[0]; /* reading end */ + __pthread_manager_thread.p_tid = 2* PTHREAD_THREADS_MAX + 1; + __pthread_manager_thread.p_pid = pid; + /* Make gdb aware of new thread manager */ + if (__pthread_threads_debug && __pthread_sig_debug > 0) + { + raise(__pthread_sig_debug); + /* We suspend ourself and gdb will wake us up when it is + ready to handle us. */ + __pthread_wait_for_restart_signal(thread_self()); + } + /* Synchronize debugging of the thread manager */ +PDEBUG("send REQ_DEBUG to manager thread\n"); + request.req_kind = REQ_DEBUG; + __libc_write(__pthread_manager_request, (char *) &request, sizeof(request)); + return 0; +} + +/* Thread creation */ + +int pthread_create(pthread_t *thread, const pthread_attr_t *attr, + void * (*start_routine)(void *), void *arg) +{ + pthread_descr self = thread_self(); + struct pthread_request request; + if (__pthread_manager_request < 0) { + if (__pthread_initialize_manager() < 0) return EAGAIN; + } + request.req_thread = self; + request.req_kind = REQ_CREATE; + request.req_args.create.attr = attr; + request.req_args.create.fn = start_routine; + request.req_args.create.arg = arg; + sigprocmask(SIG_SETMASK, (const sigset_t *) NULL, + &request.req_args.create.mask); + PDEBUG("write REQ_CREATE to manager thread\n"); + __libc_write(__pthread_manager_request, (char *) &request, sizeof(request)); +PDEBUG("before suspend(self)\n"); + suspend(self); +PDEBUG("after suspend(self)\n"); + if (THREAD_GETMEM(self, p_retcode) == 0) + *thread = (pthread_t) THREAD_GETMEM(self, p_retval); + return THREAD_GETMEM(self, p_retcode); +} + +/* Simple operations on thread identifiers */ + +pthread_t pthread_self(void) +{ + pthread_descr self = thread_self(); + return THREAD_GETMEM(self, p_tid); +} + +int pthread_equal(pthread_t thread1, pthread_t thread2) +{ + return thread1 == thread2; +} + +/* Helper function for thread_self in the case of user-provided stacks */ + +#ifndef THREAD_SELF + +pthread_descr __pthread_find_self() +{ + char * sp = CURRENT_STACK_FRAME; + pthread_handle h; + + /* __pthread_handles[0] is the initial thread, __pthread_handles[1] is + the manager threads handled specially in thread_self(), so start at 2 */ + h = __pthread_handles + 2; + while (! (sp <= (char *) h->h_descr && sp >= h->h_bottom)) h++; + +#ifdef DEBUG_PT + if (h->h_descr == NULL) { + printf("*** "__FUNCTION__" ERROR descriptor is NULL!!!!! ***\n\n"); + _exit(1); + } +#endif + + return h->h_descr; +} + +#endif + +/* Thread scheduling */ + +int pthread_setschedparam(pthread_t thread, int policy, + const struct sched_param *param) +{ + pthread_handle handle = thread_handle(thread); + pthread_descr th; + + __pthread_lock(&handle->h_lock, NULL); + if (invalid_handle(handle, thread)) { + __pthread_unlock(&handle->h_lock); + return ESRCH; + } + th = handle->h_descr; + if (sched_setscheduler(th->p_pid, policy, param) == -1) { + __pthread_unlock(&handle->h_lock); + return errno; + } + th->p_priority = policy == SCHED_OTHER ? 0 : param->sched_priority; + __pthread_unlock(&handle->h_lock); + if (__pthread_manager_request >= 0) + __pthread_manager_adjust_prio(th->p_priority); + return 0; +} + +int pthread_getschedparam(pthread_t thread, int *policy, + struct sched_param *param) +{ + pthread_handle handle = thread_handle(thread); + int pid, pol; + + __pthread_lock(&handle->h_lock, NULL); + if (invalid_handle(handle, thread)) { + __pthread_unlock(&handle->h_lock); + return ESRCH; + } + pid = handle->h_descr->p_pid; + __pthread_unlock(&handle->h_lock); + pol = sched_getscheduler(pid); + if (pol == -1) return errno; + if (sched_getparam(pid, param) == -1) return errno; + *policy = pol; + return 0; +} + +/* Process-wide exit() request */ + +static void pthread_exit_process(int retcode, void *arg) +{ + struct pthread_request request; + pthread_descr self = thread_self(); + + if (__pthread_manager_request >= 0) { + request.req_thread = self; + request.req_kind = REQ_PROCESS_EXIT; + request.req_args.exit.code = retcode; + __libc_write(__pthread_manager_request, + (char *) &request, sizeof(request)); + suspend(self); + /* Main thread should accumulate times for thread manager and its + children, so that timings for main thread account for all threads. */ + if (self == __pthread_main_thread) + waitpid(__pthread_manager_thread.p_pid, NULL, __WCLONE); + } +} + +/* The handler for the RESTART signal just records the signal received + in the thread descriptor, and optionally performs a siglongjmp + (for pthread_cond_timedwait). */ + +#ifndef __i386__ +static void pthread_handle_sigrestart(int sig) +{ + pthread_descr self = thread_self(); + PDEBUG("got called in non-i386 mode for %u\n", self); +#else +static void pthread_handle_sigrestart(int sig, struct sigcontext ctx) +{ + pthread_descr self; + asm volatile ("movw %w0,%%gs" : : "r" (ctx.gs)); + self = thread_self(); + PDEBUG("got called in i386-mode for %u\n", self); +#endif + THREAD_SETMEM(self, p_signal, sig); + if (THREAD_GETMEM(self, p_signal_jmp) != NULL) + siglongjmp(*THREAD_GETMEM(self, p_signal_jmp), 1); +} + +/* The handler for the CANCEL signal checks for cancellation + (in asynchronous mode), for process-wide exit and exec requests. + For the thread manager thread, redirect the signal to + __pthread_manager_sighandler. */ + +#ifndef __i386__ +static void pthread_handle_sigcancel(int sig) +{ + pthread_descr self = thread_self(); + sigjmp_buf * jmpbuf; +#else +static void pthread_handle_sigcancel(int sig, struct sigcontext ctx) +{ + pthread_descr self; + sigjmp_buf * jmpbuf; + asm volatile ("movw %w0,%%gs" : : "r" (ctx.gs)); + self = thread_self(); +#endif + + if (self == &__pthread_manager_thread) + { + __pthread_manager_sighandler(sig); + return; + } + if (__pthread_exit_requested) { + /* Main thread should accumulate times for thread manager and its + children, so that timings for main thread account for all threads. */ + if (self == __pthread_main_thread) + waitpid(__pthread_manager_thread.p_pid, NULL, __WCLONE); + _exit(__pthread_exit_code); + } + if (THREAD_GETMEM(self, p_canceled) + && THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE) { + if (THREAD_GETMEM(self, p_canceltype) == PTHREAD_CANCEL_ASYNCHRONOUS) + pthread_exit(PTHREAD_CANCELED); + jmpbuf = THREAD_GETMEM(self, p_cancel_jmp); + if (jmpbuf != NULL) { + THREAD_SETMEM(self, p_cancel_jmp, NULL); + siglongjmp(*jmpbuf, 1); + } + } +} + +/* Handler for the DEBUG signal. + The debugging strategy is as follows: + On reception of a REQ_DEBUG request (sent by new threads created to + the thread manager under debugging mode), the thread manager throws + __pthread_sig_debug to itself. The debugger (if active) intercepts + this signal, takes into account new threads and continue execution + of the thread manager by propagating the signal because it doesn't + know what it is specifically done for. In the current implementation, + the thread manager simply discards it. */ + +static void pthread_handle_sigdebug(int sig) +{ + /* Nothing */ +} + +/* Reset the state of the thread machinery after a fork(). + Close the pipe used for requests and set the main thread to the forked + thread. + Notice that we can't free the stack segments, as the forked thread + may hold pointers into them. */ + +void __pthread_reset_main_thread() +{ + pthread_descr self = thread_self(); + + if (__pthread_manager_request != -1) { + /* Free the thread manager stack */ + free(__pthread_manager_thread_bos); + __pthread_manager_thread_bos = __pthread_manager_thread_tos = NULL; + /* Close the two ends of the pipe */ + __libc_close(__pthread_manager_request); + __libc_close(__pthread_manager_reader); + __pthread_manager_request = __pthread_manager_reader = -1; + } + + /* Update the pid of the main thread */ + THREAD_SETMEM(self, p_pid, __getpid()); + /* Make the forked thread the main thread */ + __pthread_main_thread = self; + THREAD_SETMEM(self, p_nextlive, self); + THREAD_SETMEM(self, p_prevlive, self); + /* Now this thread modifies the global variables. */ + THREAD_SETMEM(self, p_errnop, &_errno); + THREAD_SETMEM(self, p_h_errnop, &_h_errno); +} + +/* Process-wide exec() request */ + +void __pthread_kill_other_threads_np(void) +{ + struct sigaction sa; + /* Terminate all other threads and thread manager */ + pthread_exit_process(0, NULL); + /* Make current thread the main thread in case the calling thread + changes its mind, does not exec(), and creates new threads instead. */ + __pthread_reset_main_thread(); + /* Reset the signal handlers behaviour for the signals the + implementation uses since this would be passed to the new + process. */ + sigemptyset(&sa.sa_mask); + sa.sa_flags = 0; + sa.sa_handler = SIG_DFL; + __libc_sigaction(__pthread_sig_restart, &sa, NULL); + __libc_sigaction(__pthread_sig_cancel, &sa, NULL); + if (__pthread_sig_debug > 0) + __libc_sigaction(__pthread_sig_debug, &sa, NULL); +} +weak_alias (__pthread_kill_other_threads_np, pthread_kill_other_threads_np) + +/* Concurrency symbol level. */ +static int current_level; + +int __pthread_setconcurrency(int level) +{ + /* We don't do anything unless we have found a useful interpretation. */ + current_level = level; + return 0; +} +weak_alias (__pthread_setconcurrency, pthread_setconcurrency) + +int __pthread_getconcurrency(void) +{ + return current_level; +} +weak_alias (__pthread_getconcurrency, pthread_getconcurrency) + +void __pthread_set_own_extricate_if(pthread_descr self, pthread_extricate_if *peif) +{ + __pthread_lock(self->p_lock, self); + THREAD_SETMEM(self, p_extricate, peif); + __pthread_unlock(self->p_lock); +} + +/* Primitives for controlling thread execution */ + +void __pthread_wait_for_restart_signal(pthread_descr self) +{ + sigset_t mask; + + sigprocmask(SIG_SETMASK, NULL, &mask); /* Get current signal mask */ + sigdelset(&mask, __pthread_sig_restart); /* Unblock the restart signal */ + do { + self->p_signal = 0; + PDEBUG("temporary block mask = %x\n", mask); + sigsuspend(&mask); /* Wait for signal */ + PDEBUG(" *** after sigsuspend *** \n"); + } while (self->p_signal !=__pthread_sig_restart ); +} + +/* The _old variants are for 2.0 and early 2.1 kernels which don't have RT signals. + On these kernels, we use SIGUSR1 and SIGUSR2 for restart and cancellation. + Since the restart signal does not queue, we use an atomic counter to create + queuing semantics. This is needed to resolve a rare race condition in + pthread_cond_timedwait_relative. */ + +void __pthread_restart_old(pthread_descr th) +{ + if (atomic_increment(&th->p_resume_count) == -1) + kill(th->p_pid, __pthread_sig_restart); +} + +void __pthread_suspend_old(pthread_descr self) +{ + if (atomic_decrement(&self->p_resume_count) <= 0) + __pthread_wait_for_restart_signal(self); +} + +void __pthread_restart_new(pthread_descr th) +{ + kill(th->p_pid, __pthread_sig_restart); +} + +/* There is no __pthread_suspend_new because it would just + be a wasteful wrapper for __pthread_wait_for_restart_signal */ + +/* Debugging aid */ + +#ifdef DEBUG_PT +#include <stdarg.h> + +void __pthread_message(char * fmt, ...) +{ + char buffer[1024]; + va_list args; + sprintf(buffer, "%05d : ", __getpid()); + va_start(args, fmt); + vsnprintf(buffer + 8, sizeof(buffer) - 8, fmt, args); + va_end(args); + __libc_write(2, buffer, strlen(buffer)); +} + +#endif + + +#ifndef PIC +/* We need a hook to force the cancelation wrappers to be linked in when + static libpthread is used. */ +extern const int __pthread_provide_wrappers; +static const int *const __pthread_require_wrappers = + &__pthread_provide_wrappers; +#endif diff --git a/libpthread/linuxthreads/ptlongjmp.c b/libpthread/linuxthreads/ptlongjmp.c new file mode 100644 index 000000000..c0ea8223a --- /dev/null +++ b/libpthread/linuxthreads/ptlongjmp.c @@ -0,0 +1,55 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1998 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* Redefine siglongjmp and longjmp so that they interact correctly + with cleanup handlers */ + +#include <setjmp.h> +#include "pthread.h" +#include "internals.h" + +/* These functions are not declared anywhere since they shouldn't be + used at another place but here. */ +extern void __libc_siglongjmp (sigjmp_buf env, int val) + __attribute__ ((noreturn)); +extern void __libc_longjmp (sigjmp_buf env, int val) + __attribute__ ((noreturn)); + + +static void pthread_cleanup_upto(__jmp_buf target) +{ + pthread_descr self = thread_self(); + struct _pthread_cleanup_buffer * c; + + for (c = THREAD_GETMEM(self, p_cleanup); + c != NULL && _JMPBUF_UNWINDS(target, c); + c = c->__prev) + c->__routine(c->__arg); + THREAD_SETMEM(self, p_cleanup, c); + if (THREAD_GETMEM(self, p_in_sighandler) + && _JMPBUF_UNWINDS(target, THREAD_GETMEM(self, p_in_sighandler))) + THREAD_SETMEM(self, p_in_sighandler, NULL); +} + +void siglongjmp(sigjmp_buf env, int val) +{ + pthread_cleanup_upto(env->__jmpbuf); + __libc_siglongjmp(env, val); +} + +void longjmp(jmp_buf env, int val) +{ + pthread_cleanup_upto(env->__jmpbuf); + __libc_siglongjmp(env, val); +} diff --git a/libpthread/linuxthreads/queue.h b/libpthread/linuxthreads/queue.h new file mode 100644 index 000000000..28bd75531 --- /dev/null +++ b/libpthread/linuxthreads/queue.h @@ -0,0 +1,61 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* Waiting queues */ + +/* Waiting queues are represented by lists of thread descriptors + linked through their p_nextwaiting field. The lists are kept + sorted by decreasing priority, and then decreasing waiting time. */ + +static inline void enqueue(pthread_descr * q, pthread_descr th) +{ + int prio = th->p_priority; + ASSERT(th->p_nextwaiting == NULL); + for (; *q != NULL; q = &((*q)->p_nextwaiting)) { + if (prio > (*q)->p_priority) { + th->p_nextwaiting = *q; + *q = th; + return; + } + } + *q = th; +} + +static inline pthread_descr dequeue(pthread_descr * q) +{ + pthread_descr th; + th = *q; + if (th != NULL) { + *q = th->p_nextwaiting; + th->p_nextwaiting = NULL; + } + return th; +} + +static inline int remove_from_queue(pthread_descr * q, pthread_descr th) +{ + for (; *q != NULL; q = &((*q)->p_nextwaiting)) { + if (*q == th) { + *q = th->p_nextwaiting; + th->p_nextwaiting = NULL; + return 1; + } + } + return 0; +} + +static inline int queue_is_empty(pthread_descr * q) +{ + return *q == NULL; +} diff --git a/libpthread/linuxthreads/restart.h b/libpthread/linuxthreads/restart.h new file mode 100644 index 000000000..702d7d15c --- /dev/null +++ b/libpthread/linuxthreads/restart.h @@ -0,0 +1,27 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +#include <signal.h> + +/* Primitives for controlling thread execution */ + +static inline void restart(pthread_descr th) +{ + __pthread_restart(th); /* see pthread.c */ +} + +static inline void suspend(pthread_descr self) +{ + __pthread_suspend(self); /* see pthread.c */ +} diff --git a/libpthread/linuxthreads/rwlock.c b/libpthread/linuxthreads/rwlock.c new file mode 100644 index 000000000..977fd88af --- /dev/null +++ b/libpthread/linuxthreads/rwlock.c @@ -0,0 +1,486 @@ +/* Read-write lock implementation. + Copyright (C) 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Xavier Leroy <Xavier.Leroy@inria.fr> + and Ulrich Drepper <drepper@cygnus.com>, 1998. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#include <errno.h> +#include <pthread.h> +#include <stdlib.h> +#include "internals.h" +#include "queue.h" +#include "spinlock.h" +#include "restart.h" + +/* + * Check whether the calling thread already owns one or more read locks on the + * specified lock. If so, return a pointer to the read lock info structure + * corresponding to that lock. + */ + +static pthread_readlock_info * +rwlock_is_in_list(pthread_descr self, pthread_rwlock_t *rwlock) +{ + pthread_readlock_info *info; + + for (info = self->p_readlock_list; info != NULL; info = info->pr_next) + { + if (info->pr_lock == rwlock) + return info; + } + + return NULL; +} + +/* + * Add a new lock to the thread's list of locks for which it has a read lock. + * A new info node must be allocated for this, which is taken from the thread's + * free list, or by calling malloc. If malloc fails, a null pointer is + * returned. Otherwise the lock info structure is initialized and pushed + * onto the thread's list. + */ + +static pthread_readlock_info * +rwlock_add_to_list(pthread_descr self, pthread_rwlock_t *rwlock) +{ + pthread_readlock_info *info = self->p_readlock_free; + + if (info != NULL) + self->p_readlock_free = info->pr_next; + else + info = malloc(sizeof *info); + + if (info == NULL) + return NULL; + + info->pr_lock_count = 1; + info->pr_lock = rwlock; + info->pr_next = self->p_readlock_list; + self->p_readlock_list = info; + + return info; +} + +/* + * If the thread owns a read lock over the given pthread_rwlock_t, + * and this read lock is tracked in the thread's lock list, + * this function returns a pointer to the info node in that list. + * It also decrements the lock count within that node, and if + * it reaches zero, it removes the node from the list. + * If nothing is found, it returns a null pointer. + */ + +static pthread_readlock_info * +rwlock_remove_from_list(pthread_descr self, pthread_rwlock_t *rwlock) +{ + pthread_readlock_info **pinfo; + + for (pinfo = &self->p_readlock_list; *pinfo != NULL; pinfo = &(*pinfo)->pr_next) + { + if ((*pinfo)->pr_lock == rwlock) + { + pthread_readlock_info *info = *pinfo; + if (--info->pr_lock_count == 0) + *pinfo = info->pr_next; + return info; + } + } + + return NULL; +} + +/* + * This function checks whether the conditions are right to place a read lock. + * It returns 1 if so, otherwise zero. The rwlock's internal lock must be + * locked upon entry. + */ + +static int +rwlock_can_rdlock(pthread_rwlock_t *rwlock, int have_lock_already) +{ + /* Can't readlock; it is write locked. */ + if (rwlock->__rw_writer != NULL) + return 0; + + /* Lock prefers readers; get it. */ + if (rwlock->__rw_kind == PTHREAD_RWLOCK_PREFER_READER_NP) + return 1; + + /* Lock prefers writers, but none are waiting. */ + if (queue_is_empty(&rwlock->__rw_write_waiting)) + return 1; + + /* Writers are waiting, but this thread already has a read lock */ + if (have_lock_already) + return 1; + + /* Writers are waiting, and this is a new lock */ + return 0; +} + +/* + * This function helps support brain-damaged recursive read locking + * semantics required by Unix 98, while maintaining write priority. + * This basically determines whether this thread already holds a read lock + * already. It returns 1 if so, otherwise it returns 0. + * + * If the thread has any ``untracked read locks'' then it just assumes + * that this lock is among them, just to be safe, and returns 1. + * + * Also, if it finds the thread's lock in the list, it sets the pointer + * referenced by pexisting to refer to the list entry. + * + * If the thread has no untracked locks, and the lock is not found + * in its list, then it is added to the list. If this fails, + * then *pout_of_mem is set to 1. + */ + +static int +rwlock_have_already(pthread_descr *pself, pthread_rwlock_t *rwlock, + pthread_readlock_info **pexisting, int *pout_of_mem) +{ + pthread_readlock_info *existing = NULL; + int out_of_mem = 0, have_lock_already = 0; + pthread_descr self = *pself; + + if (rwlock->__rw_kind == PTHREAD_RWLOCK_PREFER_WRITER_NP) + { + if (!self) + self = thread_self(); + + existing = rwlock_is_in_list(self, rwlock); + + if (existing != NULL || self->p_untracked_readlock_count > 0) + have_lock_already = 1; + else + { + existing = rwlock_add_to_list(self, rwlock); + if (existing == NULL) + out_of_mem = 1; + } + } + + *pout_of_mem = out_of_mem; + *pexisting = existing; + *pself = self; + + return have_lock_already; +} + +int +pthread_rwlock_init (pthread_rwlock_t *rwlock, + const pthread_rwlockattr_t *attr) +{ + __pthread_init_lock(&rwlock->__rw_lock); + rwlock->__rw_readers = 0; + rwlock->__rw_writer = NULL; + rwlock->__rw_read_waiting = NULL; + rwlock->__rw_write_waiting = NULL; + + if (attr == NULL) + { + rwlock->__rw_kind = PTHREAD_RWLOCK_DEFAULT_NP; + rwlock->__rw_pshared = PTHREAD_PROCESS_PRIVATE; + } + else + { + rwlock->__rw_kind = attr->__lockkind; + rwlock->__rw_pshared = attr->__pshared; + } + + return 0; +} + + +int +pthread_rwlock_destroy (pthread_rwlock_t *rwlock) +{ + int readers; + _pthread_descr writer; + + __pthread_lock (&rwlock->__rw_lock, NULL); + readers = rwlock->__rw_readers; + writer = rwlock->__rw_writer; + __pthread_unlock (&rwlock->__rw_lock); + + if (readers > 0 || writer != NULL) + return EBUSY; + + return 0; +} + +int +pthread_rwlock_rdlock (pthread_rwlock_t *rwlock) +{ + pthread_descr self = NULL; + pthread_readlock_info *existing; + int out_of_mem, have_lock_already; + + have_lock_already = rwlock_have_already(&self, rwlock, + &existing, &out_of_mem); + + for (;;) + { + if (self == NULL) + self = thread_self (); + + __pthread_lock (&rwlock->__rw_lock, self); + + if (rwlock_can_rdlock(rwlock, have_lock_already)) + break; + + enqueue (&rwlock->__rw_read_waiting, self); + __pthread_unlock (&rwlock->__rw_lock); + suspend (self); /* This is not a cancellation point */ + } + + ++rwlock->__rw_readers; + __pthread_unlock (&rwlock->__rw_lock); + + if (have_lock_already || out_of_mem) + { + if (existing != NULL) + existing->pr_lock_count++; + else + self->p_untracked_readlock_count++; + } + + return 0; +} + +int +pthread_rwlock_tryrdlock (pthread_rwlock_t *rwlock) +{ + pthread_descr self = thread_self(); + pthread_readlock_info *existing; + int out_of_mem, have_lock_already; + int retval = EBUSY; + + have_lock_already = rwlock_have_already(&self, rwlock, + &existing, &out_of_mem); + + __pthread_lock (&rwlock->__rw_lock, self); + + /* 0 is passed to here instead of have_lock_already. + This is to meet Single Unix Spec requirements: + if writers are waiting, pthread_rwlock_tryrdlock + does not acquire a read lock, even if the caller has + one or more read locks already. */ + + if (rwlock_can_rdlock(rwlock, 0)) + { + ++rwlock->__rw_readers; + retval = 0; + } + + __pthread_unlock (&rwlock->__rw_lock); + + if (retval == 0) + { + if (have_lock_already || out_of_mem) + { + if (existing != NULL) + existing->pr_lock_count++; + else + self->p_untracked_readlock_count++; + } + } + + return retval; +} + + +int +pthread_rwlock_wrlock (pthread_rwlock_t *rwlock) +{ + pthread_descr self = thread_self (); + + while(1) + { + __pthread_lock (&rwlock->__rw_lock, self); + if (rwlock->__rw_readers == 0 && rwlock->__rw_writer == NULL) + { + rwlock->__rw_writer = self; + __pthread_unlock (&rwlock->__rw_lock); + return 0; + } + + /* Suspend ourselves, then try again */ + enqueue (&rwlock->__rw_write_waiting, self); + __pthread_unlock (&rwlock->__rw_lock); + suspend (self); /* This is not a cancellation point */ + } +} + + +int +pthread_rwlock_trywrlock (pthread_rwlock_t *rwlock) +{ + int result = EBUSY; + + __pthread_lock (&rwlock->__rw_lock, NULL); + if (rwlock->__rw_readers == 0 && rwlock->__rw_writer == NULL) + { + rwlock->__rw_writer = thread_self (); + result = 0; + } + __pthread_unlock (&rwlock->__rw_lock); + + return result; +} + + +int +pthread_rwlock_unlock (pthread_rwlock_t *rwlock) +{ + pthread_descr torestart; + pthread_descr th; + + __pthread_lock (&rwlock->__rw_lock, NULL); + if (rwlock->__rw_writer != NULL) + { + /* Unlocking a write lock. */ + if (rwlock->__rw_writer != thread_self ()) + { + __pthread_unlock (&rwlock->__rw_lock); + return EPERM; + } + rwlock->__rw_writer = NULL; + + if (rwlock->__rw_kind == PTHREAD_RWLOCK_PREFER_READER_NP + || (th = dequeue (&rwlock->__rw_write_waiting)) == NULL) + { + /* Restart all waiting readers. */ + torestart = rwlock->__rw_read_waiting; + rwlock->__rw_read_waiting = NULL; + __pthread_unlock (&rwlock->__rw_lock); + while ((th = dequeue (&torestart)) != NULL) + restart (th); + } + else + { + /* Restart one waiting writer. */ + __pthread_unlock (&rwlock->__rw_lock); + restart (th); + } + } + else + { + /* Unlocking a read lock. */ + if (rwlock->__rw_readers == 0) + { + __pthread_unlock (&rwlock->__rw_lock); + return EPERM; + } + + --rwlock->__rw_readers; + if (rwlock->__rw_readers == 0) + /* Restart one waiting writer, if any. */ + th = dequeue (&rwlock->__rw_write_waiting); + else + th = NULL; + + __pthread_unlock (&rwlock->__rw_lock); + if (th != NULL) + restart (th); + + /* Recursive lock fixup */ + + if (rwlock->__rw_kind == PTHREAD_RWLOCK_PREFER_WRITER_NP) + { + pthread_descr self = thread_self(); + pthread_readlock_info *victim = rwlock_remove_from_list(self, rwlock); + + if (victim != NULL) + { + if (victim->pr_lock_count == 0) + { + victim->pr_next = self->p_readlock_free; + self->p_readlock_free = victim; + } + } + else + { + if (self->p_untracked_readlock_count > 0) + self->p_untracked_readlock_count--; + } + } + } + + return 0; +} + + + +int +pthread_rwlockattr_init (pthread_rwlockattr_t *attr) +{ + attr->__lockkind = 0; + attr->__pshared = 0; + + return 0; +} + + +int +pthread_rwlockattr_destroy (pthread_rwlockattr_t *attr) +{ + return 0; +} + + +int +pthread_rwlockattr_getpshared (const pthread_rwlockattr_t *attr, int *pshared) +{ + *pshared = attr->__pshared; + return 0; +} + + +int +pthread_rwlockattr_setpshared (pthread_rwlockattr_t *attr, int pshared) +{ + if (pshared != PTHREAD_PROCESS_PRIVATE && pshared != PTHREAD_PROCESS_SHARED) + return EINVAL; + + attr->__pshared = pshared; + + return 0; +} + + +int +pthread_rwlockattr_getkind_np (const pthread_rwlockattr_t *attr, int *pref) +{ + *pref = attr->__lockkind; + return 0; +} + + +int +pthread_rwlockattr_setkind_np (pthread_rwlockattr_t *attr, int pref) +{ + if (pref != PTHREAD_RWLOCK_PREFER_READER_NP + && pref != PTHREAD_RWLOCK_PREFER_WRITER_NP + && pref != PTHREAD_RWLOCK_DEFAULT_NP) + return EINVAL; + + attr->__lockkind = pref; + + return 0; +} diff --git a/libpthread/linuxthreads/semaphore.c b/libpthread/linuxthreads/semaphore.c new file mode 100644 index 000000000..0297b3a1e --- /dev/null +++ b/libpthread/linuxthreads/semaphore.c @@ -0,0 +1,209 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* Semaphores a la POSIX 1003.1b */ + +#include <errno.h> +#include "pthread.h" +#include "semaphore.h" +#include "internals.h" +#include "spinlock.h" +#include "restart.h" +#include "queue.h" + +int __new_sem_init(sem_t *sem, int pshared, unsigned int value) +{ + if (value > SEM_VALUE_MAX) { + errno = EINVAL; + return -1; + } + if (pshared) { + errno = ENOSYS; + return -1; + } + __pthread_init_lock((struct _pthread_fastlock *) &sem->__sem_lock); + sem->__sem_value = value; + sem->__sem_waiting = NULL; + return 0; +} + +/* Function called by pthread_cancel to remove the thread from + waiting inside __new_sem_wait. */ + +static int new_sem_extricate_func(void *obj, pthread_descr th) +{ + volatile pthread_descr self = thread_self(); + sem_t *sem = obj; + int did_remove = 0; + + __pthread_lock((struct _pthread_fastlock *) &sem->__sem_lock, self); + did_remove = remove_from_queue(&sem->__sem_waiting, th); + __pthread_unlock((struct _pthread_fastlock *) &sem->__sem_lock); + + return did_remove; +} + +int __new_sem_wait(sem_t * sem) +{ + volatile pthread_descr self = thread_self(); + pthread_extricate_if extr; + int already_canceled = 0; + + /* Set up extrication interface */ + extr.pu_object = sem; + extr.pu_extricate_func = new_sem_extricate_func; + + __pthread_lock((struct _pthread_fastlock *) &sem->__sem_lock, self); + if (sem->__sem_value > 0) { + sem->__sem_value--; + __pthread_unlock((struct _pthread_fastlock *) &sem->__sem_lock); + return 0; + } + /* Register extrication interface */ + __pthread_set_own_extricate_if(self, &extr); + /* Enqueue only if not already cancelled. */ + if (!(THREAD_GETMEM(self, p_canceled) + && THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE)) + enqueue(&sem->__sem_waiting, self); + else + already_canceled = 1; + __pthread_unlock((struct _pthread_fastlock *) &sem->__sem_lock); + + if (already_canceled) { + __pthread_set_own_extricate_if(self, 0); + pthread_exit(PTHREAD_CANCELED); + } + + /* Wait for sem_post or cancellation, or fall through if already canceled */ + suspend(self); + __pthread_set_own_extricate_if(self, 0); + + /* Terminate only if the wakeup came from cancellation. */ + /* Otherwise ignore cancellation because we got the semaphore. */ + + if (THREAD_GETMEM(self, p_woken_by_cancel) + && THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE) { + THREAD_SETMEM(self, p_woken_by_cancel, 0); + pthread_exit(PTHREAD_CANCELED); + } + /* We got the semaphore */ + return 0; +} + +int __new_sem_trywait(sem_t * sem) +{ + int retval; + + __pthread_lock((struct _pthread_fastlock *) &sem->__sem_lock, NULL); + if (sem->__sem_value == 0) { + errno = EAGAIN; + retval = -1; + } else { + sem->__sem_value--; + retval = 0; + } + __pthread_unlock((struct _pthread_fastlock *) &sem->__sem_lock); + return retval; +} + +int __new_sem_post(sem_t * sem) +{ + pthread_descr self = thread_self(); + pthread_descr th; + struct pthread_request request; + + if (THREAD_GETMEM(self, p_in_sighandler) == NULL) { + __pthread_lock((struct _pthread_fastlock *) &sem->__sem_lock, self); + if (sem->__sem_waiting == NULL) { + if (sem->__sem_value >= SEM_VALUE_MAX) { + /* Overflow */ + errno = ERANGE; + __pthread_unlock((struct _pthread_fastlock *) &sem->__sem_lock); + return -1; + } + sem->__sem_value++; + __pthread_unlock((struct _pthread_fastlock *) &sem->__sem_lock); + } else { + th = dequeue(&sem->__sem_waiting); + __pthread_unlock((struct _pthread_fastlock *) &sem->__sem_lock); + restart(th); + } + } else { + /* If we're in signal handler, delegate post operation to + the thread manager. */ + if (__pthread_manager_request < 0) { + if (__pthread_initialize_manager() < 0) { + errno = EAGAIN; + return -1; + } + } + request.req_kind = REQ_POST; + request.req_args.post = sem; + __libc_write(__pthread_manager_request, + (char *) &request, sizeof(request)); + } + return 0; +} + +int __new_sem_getvalue(sem_t * sem, int * sval) +{ + *sval = sem->__sem_value; + return 0; +} + +int __new_sem_destroy(sem_t * sem) +{ + if (sem->__sem_waiting != NULL) { + __set_errno (EBUSY); + return -1; + } + return 0; +} + +sem_t *sem_open(const char *name, int oflag, ...) +{ + __set_errno (ENOSYS); + return SEM_FAILED; +} + +int sem_close(sem_t *sem) +{ + __set_errno (ENOSYS); + return -1; +} + +int sem_unlink(const char *name) +{ + __set_errno (ENOSYS); + return -1; +} + +#if defined PIC && DO_VERSIONING +default_symbol_version (__new_sem_init, sem_init, GLIBC_2.1); +default_symbol_version (__new_sem_wait, sem_wait, GLIBC_2.1); +default_symbol_version (__new_sem_trywait, sem_trywait, GLIBC_2.1); +default_symbol_version (__new_sem_post, sem_post, GLIBC_2.1); +default_symbol_version (__new_sem_getvalue, sem_getvalue, GLIBC_2.1); +default_symbol_version (__new_sem_destroy, sem_destroy, GLIBC_2.1); +#else +# ifdef weak_alias +weak_alias (__new_sem_init, sem_init) +weak_alias (__new_sem_wait, sem_wait) +weak_alias (__new_sem_trywait, sem_trywait) +weak_alias (__new_sem_post, sem_post) +weak_alias (__new_sem_getvalue, sem_getvalue) +weak_alias (__new_sem_destroy, sem_destroy) +# endif +#endif + diff --git a/libpthread/linuxthreads/semaphore.h b/libpthread/linuxthreads/semaphore.h new file mode 100644 index 000000000..84742233b --- /dev/null +++ b/libpthread/linuxthreads/semaphore.h @@ -0,0 +1,80 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +#ifndef _SEMAPHORE_H +#define _SEMAPHORE_H 1 + +#include <features.h> +#include <sys/types.h> + +#ifndef _PTHREAD_DESCR_DEFINED +/* Thread descriptors. Needed for `sem_t' definition. */ +typedef struct _pthread_descr_struct *_pthread_descr; +# define _PTHREAD_DESCR_DEFINED +#endif + +/* System specific semaphore definition. */ +typedef struct +{ + struct + { + long int status; + int spinlock; + } __sem_lock; + int __sem_value; + _pthread_descr __sem_waiting; +} sem_t; + + + +/* Value returned if `sem_open' failed. */ +#define SEM_FAILED ((sem_t *) 0) + +/* Maximum value the semaphore can have. */ +#define SEM_VALUE_MAX ((int) ((~0u) >> 1)) + + +__BEGIN_DECLS + +/* Initialize semaphore object SEM to VALUE. If PSHARED then share it + with other processes. */ +extern int sem_init __P ((sem_t *__sem, int __pshared, unsigned int __value)); + +/* Free resources associated with semaphore object SEM. */ +extern int sem_destroy __P ((sem_t *__sem)); + +/* Open a named semaphore NAME with open flaot OFLAG. */ +extern sem_t *sem_open __P ((__const char *__name, int __oflag, ...)); + +/* Close descriptor for named semaphore SEM. */ +extern int sem_close __P ((sem_t *__sem)); + +/* Remove named semaphore NAME. */ +extern int sem_unlink __P ((__const char *__name)); + +/* Wait for SEM being posted. */ +extern int sem_wait __P ((sem_t *__sem)); + +/* Test whether SEM is posted. */ +extern int sem_trywait __P ((sem_t *__sem)); + +/* Post SEM. */ +extern int sem_post __P ((sem_t *__sem)); + +/* Get current value of SEM and store it in *SVAL. */ +extern int sem_getvalue __P ((sem_t *__sem, int *__sval)); + +__END_DECLS + +#endif /* semaphore.h */ diff --git a/libpthread/linuxthreads/signals.c b/libpthread/linuxthreads/signals.c new file mode 100644 index 000000000..de156e15a --- /dev/null +++ b/libpthread/linuxthreads/signals.c @@ -0,0 +1,239 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* Handling of signals */ + +#include <errno.h> +#include <signal.h> +#include <stdio.h> +#include "pthread.h" +#include "internals.h" +#include "spinlock.h" +#include <ucontext.h> +#include <sigcontextinfo.h> + +/* mods for uClibc: __libc_sigaction is not in any standard headers */ +extern int __libc_sigaction (int sig, const struct sigaction *act, struct sigaction *oact); + +int pthread_sigmask(int how, const sigset_t * newmask, sigset_t * oldmask) +{ + sigset_t mask; + + if (newmask != NULL) { + mask = *newmask; + /* Don't allow __pthread_sig_restart to be unmasked. + Don't allow __pthread_sig_cancel to be masked. */ + switch(how) { + case SIG_SETMASK: + sigaddset(&mask, __pthread_sig_restart); + sigdelset(&mask, __pthread_sig_cancel); + break; + case SIG_BLOCK: + sigdelset(&mask, __pthread_sig_cancel); + break; + case SIG_UNBLOCK: + sigdelset(&mask, __pthread_sig_restart); + break; + } + newmask = &mask; + } + if (sigprocmask(how, newmask, oldmask) == -1) + return errno; + else + return 0; +} + +int pthread_kill(pthread_t thread, int signo) +{ + pthread_handle handle = thread_handle(thread); + int pid; + + __pthread_lock(&handle->h_lock, NULL); + if (invalid_handle(handle, thread)) { + __pthread_unlock(&handle->h_lock); + return ESRCH; + } + pid = handle->h_descr->p_pid; + __pthread_unlock(&handle->h_lock); + if (kill(pid, signo) == -1) + return errno; + else + return 0; +} + +/* User-provided signal handlers */ +typedef void (*arch_sighandler_t) __PMT ((int, SIGCONTEXT)); +static union +{ + arch_sighandler_t old; + void (*rt) (int, struct siginfo *, struct ucontext *); +} sighandler[NSIG]; + +/* The wrapper around user-provided signal handlers */ +static void pthread_sighandler(int signo, SIGCONTEXT ctx) +{ + pthread_descr self = thread_self(); + char * in_sighandler; + /* If we're in a sigwait operation, just record the signal received + and return without calling the user's handler */ + if (THREAD_GETMEM(self, p_sigwaiting)) { + THREAD_SETMEM(self, p_sigwaiting, 0); + THREAD_SETMEM(self, p_signal, signo); + return; + } + /* Record that we're in a signal handler and call the user's + handler function */ + in_sighandler = THREAD_GETMEM(self, p_in_sighandler); + if (in_sighandler == NULL) + THREAD_SETMEM(self, p_in_sighandler, CURRENT_STACK_FRAME); + sighandler[signo].old(signo, SIGCONTEXT_EXTRA_ARGS ctx); + if (in_sighandler == NULL) + THREAD_SETMEM(self, p_in_sighandler, NULL); +} + +/* The same, this time for real-time signals. */ +static void pthread_sighandler_rt(int signo, struct siginfo *si, + struct ucontext *uc) +{ + pthread_descr self = thread_self(); + char * in_sighandler; + /* If we're in a sigwait operation, just record the signal received + and return without calling the user's handler */ + if (THREAD_GETMEM(self, p_sigwaiting)) { + THREAD_SETMEM(self, p_sigwaiting, 0); + THREAD_SETMEM(self, p_signal, signo); + return; + } + /* Record that we're in a signal handler and call the user's + handler function */ + in_sighandler = THREAD_GETMEM(self, p_in_sighandler); + if (in_sighandler == NULL) + THREAD_SETMEM(self, p_in_sighandler, CURRENT_STACK_FRAME); + sighandler[signo].rt(signo, si, uc); + if (in_sighandler == NULL) + THREAD_SETMEM(self, p_in_sighandler, NULL); +} + +/* The wrapper around sigaction. Install our own signal handler + around the signal. */ +int sigaction(int sig, const struct sigaction * act, + struct sigaction * oact) +{ + struct sigaction newact; + struct sigaction *newactp; + +printf(__FUNCTION__": pthreads wrapper!\n"); + if (sig == __pthread_sig_restart || + sig == __pthread_sig_cancel || + (sig == __pthread_sig_debug && __pthread_sig_debug > 0)) + return EINVAL; + if (act) + { + newact = *act; + if (act->sa_handler != SIG_IGN && act->sa_handler != SIG_DFL + && sig > 0 && sig < NSIG) + { + if (act->sa_flags & SA_SIGINFO) + newact.sa_handler = (__sighandler_t) pthread_sighandler_rt; + else + newact.sa_handler = (__sighandler_t) pthread_sighandler; + } + newactp = &newact; + } + else + newactp = NULL; + if (__libc_sigaction(sig, newactp, oact) == -1) + return -1; +printf(__FUNCTION__": signahdler installed, __sigaction successful\n"); + if (sig > 0 && sig < NSIG) + { + if (oact != NULL) + oact->sa_handler = (__sighandler_t) sighandler[sig].old; + if (act) + /* For the assignment is does not matter whether it's a normal + or real-time signal. */ + sighandler[sig].old = (arch_sighandler_t) act->sa_handler; + } + return 0; +} + +/* A signal handler that does nothing */ +static void pthread_null_sighandler(int sig) { } + +/* sigwait -- synchronously wait for a signal */ +int sigwait(const sigset_t * set, int * sig) +{ + volatile pthread_descr self = thread_self(); + sigset_t mask; + int s; + sigjmp_buf jmpbuf; + struct sigaction sa; + + /* Get ready to block all signals except those in set + and the cancellation signal. + Also check that handlers are installed on all signals in set, + and if not, install our dummy handler. This is conformant to + POSIX: "The effect of sigwait() on the signal actions for the + signals in set is unspecified." */ + sigfillset(&mask); + sigdelset(&mask, __pthread_sig_cancel); + for (s = 1; s <= NSIG; s++) { + if (sigismember(set, s) && + s != __pthread_sig_restart && + s != __pthread_sig_cancel && + s != __pthread_sig_debug) { + sigdelset(&mask, s); + if (sighandler[s].old == NULL || + sighandler[s].old == (arch_sighandler_t) SIG_DFL || + sighandler[s].old == (arch_sighandler_t) SIG_IGN) { + sa.sa_handler = pthread_null_sighandler; + sigemptyset(&sa.sa_mask); + sa.sa_flags = 0; + sigaction(s, &sa, NULL); + } + } + } + /* Test for cancellation */ + if (sigsetjmp(jmpbuf, 1) == 0) { + THREAD_SETMEM(self, p_cancel_jmp, &jmpbuf); + if (! (THREAD_GETMEM(self, p_canceled) + && THREAD_GETMEM(self, p_cancelstate) == PTHREAD_CANCEL_ENABLE)) { + /* Reset the signal count */ + THREAD_SETMEM(self, p_signal, 0); + /* Say we're in sigwait */ + THREAD_SETMEM(self, p_sigwaiting, 1); + /* Unblock the signals and wait for them */ + sigsuspend(&mask); + } + } + THREAD_SETMEM(self, p_cancel_jmp, NULL); + /* The signals are now reblocked. Check for cancellation */ + pthread_testcancel(); + /* We should have self->p_signal != 0 and equal to the signal received */ + *sig = THREAD_GETMEM(self, p_signal); + return 0; +} + +/* Redefine raise() to send signal to calling thread only, + as per POSIX 1003.1c */ +int raise (int sig) +{ + int retcode = pthread_kill(pthread_self(), sig); + if (retcode == 0) + return 0; + else { + errno = retcode; + return -1; + } +} diff --git a/libpthread/linuxthreads/specific.c b/libpthread/linuxthreads/specific.c new file mode 100644 index 000000000..14c4b29d1 --- /dev/null +++ b/libpthread/linuxthreads/specific.c @@ -0,0 +1,179 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* mods for uClibc: removed strong_alias'es */ + +/* Thread-specific data */ + +#include <errno.h> +#include <stddef.h> +#include <stdlib.h> +#include "pthread.h" +#include "internals.h" + +/* Table of keys. */ + +static struct pthread_key_struct pthread_keys[PTHREAD_KEYS_MAX] = + { { 0, NULL } }; + +/* Mutex to protect access to pthread_keys */ + +static pthread_mutex_t pthread_keys_mutex = PTHREAD_MUTEX_INITIALIZER; + +/* Create a new key */ + +int pthread_key_create(pthread_key_t * key, destr_function destr) +{ + int i; + + pthread_mutex_lock(&pthread_keys_mutex); + for (i = 0; i < PTHREAD_KEYS_MAX; i++) { + if (! pthread_keys[i].in_use) { + /* Mark key in use */ + pthread_keys[i].in_use = 1; + pthread_keys[i].destr = destr; + pthread_mutex_unlock(&pthread_keys_mutex); + *key = i; + return 0; + } + } + pthread_mutex_unlock(&pthread_keys_mutex); + return EAGAIN; +} +//strong_alias (__pthread_key_create, pthread_key_create) + +/* Delete a key */ + +int pthread_key_delete(pthread_key_t key) +{ + pthread_descr self = thread_self(); + pthread_descr th; + unsigned int idx1st, idx2nd; + + pthread_mutex_lock(&pthread_keys_mutex); + if (key >= PTHREAD_KEYS_MAX || !pthread_keys[key].in_use) { + pthread_mutex_unlock(&pthread_keys_mutex); + return EINVAL; + } + pthread_keys[key].in_use = 0; + pthread_keys[key].destr = NULL; + /* Set the value of the key to NULL in all running threads, so + that if the key is reallocated later by pthread_key_create, its + associated values will be NULL in all threads. */ + idx1st = key / PTHREAD_KEY_2NDLEVEL_SIZE; + idx2nd = key % PTHREAD_KEY_2NDLEVEL_SIZE; + th = self; + do { + /* If the thread already is terminated don't modify the memory. */ + if (!th->p_terminated && th->p_specific[idx1st] != NULL) + th->p_specific[idx1st][idx2nd] = NULL; + th = th->p_nextlive; + } while (th != self); + pthread_mutex_unlock(&pthread_keys_mutex); + return 0; +} + +/* Set the value of a key */ + +int pthread_setspecific(pthread_key_t key, const void * pointer) +{ + pthread_descr self = thread_self(); + unsigned int idx1st, idx2nd; + + if (key >= PTHREAD_KEYS_MAX || !pthread_keys[key].in_use) + return EINVAL; + idx1st = key / PTHREAD_KEY_2NDLEVEL_SIZE; + idx2nd = key % PTHREAD_KEY_2NDLEVEL_SIZE; + if (THREAD_GETMEM_NC(self, p_specific[idx1st]) == NULL) { + void *newp = calloc(PTHREAD_KEY_2NDLEVEL_SIZE, sizeof (void *)); + if (newp == NULL) + return ENOMEM; + THREAD_SETMEM_NC(self, p_specific[idx1st], newp); + } + THREAD_GETMEM_NC(self, p_specific[idx1st])[idx2nd] = (void *) pointer; + return 0; +} +//strong_alias (__pthread_setspecific, pthread_setspecific) + +/* Get the value of a key */ + +void * pthread_getspecific(pthread_key_t key) +{ + pthread_descr self = thread_self(); + unsigned int idx1st, idx2nd; + + if (key >= PTHREAD_KEYS_MAX) + return NULL; + idx1st = key / PTHREAD_KEY_2NDLEVEL_SIZE; + idx2nd = key % PTHREAD_KEY_2NDLEVEL_SIZE; + if (THREAD_GETMEM_NC(self, p_specific[idx1st]) == NULL + || !pthread_keys[key].in_use) + return NULL; + return THREAD_GETMEM_NC(self, p_specific[idx1st])[idx2nd]; +} +//strong_alias (__pthread_getspecific, pthread_getspecific) + +/* Call the destruction routines on all keys */ + +void __pthread_destroy_specifics() +{ + pthread_descr self = thread_self(); + int i, j, round, found_nonzero; + destr_function destr; + void * data; + + for (round = 0, found_nonzero = 1; + found_nonzero && round < PTHREAD_DESTRUCTOR_ITERATIONS; + round++) { + found_nonzero = 0; + for (i = 0; i < PTHREAD_KEY_1STLEVEL_SIZE; i++) + if (THREAD_GETMEM_NC(self, p_specific[i]) != NULL) + for (j = 0; j < PTHREAD_KEY_2NDLEVEL_SIZE; j++) { + destr = pthread_keys[i * PTHREAD_KEY_2NDLEVEL_SIZE + j].destr; + data = THREAD_GETMEM_NC(self, p_specific[i])[j]; + if (destr != NULL && data != NULL) { + THREAD_GETMEM_NC(self, p_specific[i])[j] = NULL; + destr(data); + found_nonzero = 1; + } + } + } + for (i = 0; i < PTHREAD_KEY_1STLEVEL_SIZE; i++) { + if (THREAD_GETMEM_NC(self, p_specific[i]) != NULL) + free(THREAD_GETMEM_NC(self, p_specific[i])); + } +} + +/* Thread-specific data for libc. */ + +static int +libc_internal_tsd_set(enum __libc_tsd_key_t key, const void * pointer) +{ + pthread_descr self = thread_self(); + + THREAD_SETMEM_NC(self, p_libc_specific[key], (void *) pointer); + return 0; +} +int (*__libc_internal_tsd_set)(enum __libc_tsd_key_t key, const void * pointer) + = libc_internal_tsd_set; + +static void * +libc_internal_tsd_get(enum __libc_tsd_key_t key) +{ + pthread_descr self = thread_self(); + + return THREAD_GETMEM_NC(self, p_libc_specific[key]); +} +void * (*__libc_internal_tsd_get)(enum __libc_tsd_key_t key) + = libc_internal_tsd_get; diff --git a/libpthread/linuxthreads/spinlock.c b/libpthread/linuxthreads/spinlock.c new file mode 100644 index 000000000..b1a99d975 --- /dev/null +++ b/libpthread/linuxthreads/spinlock.c @@ -0,0 +1,195 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1998 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +/* Internal locks */ + +#include <errno.h> +#include <sched.h> +#include <time.h> +#include "pthread.h" +#include "internals.h" +#include "spinlock.h" +#include "restart.h" + +/* The status field of a fastlock has the following meaning: + 0: fastlock is free + 1: fastlock is taken, no thread is waiting on it + ADDR: fastlock is taken, ADDR is address of thread descriptor for + first waiting thread, other waiting threads are linked via + their p_nextlock field. + The waiting list is not sorted by priority order. + Actually, we always insert at top of list (sole insertion mode + that can be performed without locking). + For __pthread_unlock, we perform a linear search in the list + to find the highest-priority, oldest waiting thread. + This is safe because there are no concurrent __pthread_unlock + operations -- only the thread that locked the mutex can unlock it. */ + +void internal_function __pthread_lock(struct _pthread_fastlock * lock, + pthread_descr self) +{ + long oldstatus, newstatus; + int spurious_wakeup_count = 0; + + do { + oldstatus = lock->__status; + if (oldstatus == 0) { + newstatus = 1; + } else { + if (self == NULL) + self = thread_self(); + newstatus = (long) self; + } + if (self != NULL) { + ASSERT(self->p_nextlock == NULL); + THREAD_SETMEM(self, p_nextlock, (pthread_descr) oldstatus); + } + } while(! compare_and_swap(&lock->__status, oldstatus, newstatus, + &lock->__spinlock)); + + /* Suspend with guard against spurious wakeup. + This can happen in pthread_cond_timedwait_relative, when the thread + wakes up due to timeout and is still on the condvar queue, and then + locks the queue to remove itself. At that point it may still be on the + queue, and may be resumed by a condition signal. */ + + if (oldstatus != 0) { + for (;;) { + suspend(self); + if (self->p_nextlock != NULL) { + /* Count resumes that don't belong to us. */ + spurious_wakeup_count++; + continue; + } + break; + } + } + + /* Put back any resumes we caught that don't belong to us. */ + while (spurious_wakeup_count--) + restart(self); +} + +void internal_function __pthread_unlock(struct _pthread_fastlock * lock) +{ + long oldstatus; + pthread_descr thr, * ptr, * maxptr; + int maxprio; + +again: + oldstatus = lock->__status; + if (oldstatus == 0 || oldstatus == 1) { + /* No threads are waiting for this lock. Please note that we also + enter this case if the lock is not taken at all. If this wouldn't + be done here we would crash further down. */ + if (! compare_and_swap(&lock->__status, oldstatus, 0, &lock->__spinlock)) + goto again; + return; + } + /* Find thread in waiting queue with maximal priority */ + ptr = (pthread_descr *) &lock->__status; + thr = (pthread_descr) oldstatus; + maxprio = 0; + maxptr = ptr; + while (thr != (pthread_descr) 1) { + if (thr->p_priority >= maxprio) { + maxptr = ptr; + maxprio = thr->p_priority; + } + ptr = &(thr->p_nextlock); + thr = *ptr; + } + /* Remove max prio thread from waiting list. */ + if (maxptr == (pthread_descr *) &lock->__status) { + /* If max prio thread is at head, remove it with compare-and-swap + to guard against concurrent lock operation */ + thr = (pthread_descr) oldstatus; + if (! compare_and_swap(&lock->__status, + oldstatus, (long)(thr->p_nextlock), + &lock->__spinlock)) + goto again; + } else { + /* No risk of concurrent access, remove max prio thread normally */ + thr = *maxptr; + *maxptr = thr->p_nextlock; + } + /* Wake up the selected waiting thread */ + thr->p_nextlock = NULL; + restart(thr); +} + +/* Compare-and-swap emulation with a spinlock */ + +#ifdef TEST_FOR_COMPARE_AND_SWAP +int __pthread_has_cas = 0; +#endif + +#if !defined HAS_COMPARE_AND_SWAP || defined TEST_FOR_COMPARE_AND_SWAP + +static void __pthread_acquire(int * spinlock); + +int __pthread_compare_and_swap(long * ptr, long oldval, long newval, + int * spinlock) +{ + int res; + if (testandset(spinlock)) __pthread_acquire(spinlock); + if (*ptr == oldval) { + *ptr = newval; res = 1; + } else { + res = 0; + } + *spinlock = 0; + return res; +} + +/* This function is called if the inlined test-and-set + in __pthread_compare_and_swap() failed */ + +/* The retry strategy is as follows: + - We test and set the spinlock MAX_SPIN_COUNT times, calling + sched_yield() each time. This gives ample opportunity for other + threads with priority >= our priority to make progress and + release the spinlock. + - If a thread with priority < our priority owns the spinlock, + calling sched_yield() repeatedly is useless, since we're preventing + the owning thread from making progress and releasing the spinlock. + So, after MAX_SPIN_LOCK attemps, we suspend the calling thread + using nanosleep(). This again should give time to the owning thread + for releasing the spinlock. + Notice that the nanosleep() interval must not be too small, + since the kernel does busy-waiting for short intervals in a realtime + process (!). The smallest duration that guarantees thread + suspension is currently 2ms. + - When nanosleep() returns, we try again, doing MAX_SPIN_COUNT + sched_yield(), then sleeping again if needed. */ + +static void __pthread_acquire(int * spinlock) +{ + int cnt = 0; + struct timespec tm; + + while (testandset(spinlock)) { + if (cnt < MAX_SPIN_COUNT) { + sched_yield(); + cnt++; + } else { + tm.tv_sec = 0; + tm.tv_nsec = SPIN_SLEEP_DURATION; + nanosleep(&tm, NULL); + cnt = 0; + } + } +} + +#endif diff --git a/libpthread/linuxthreads/spinlock.h b/libpthread/linuxthreads/spinlock.h new file mode 100644 index 000000000..aae18a27b --- /dev/null +++ b/libpthread/linuxthreads/spinlock.h @@ -0,0 +1,102 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1998 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +#if defined(TEST_FOR_COMPARE_AND_SWAP) + +extern int __pthread_has_cas; +extern int __pthread_compare_and_swap(long * ptr, long oldval, long newval, + int * spinlock); + +static inline int compare_and_swap(long * ptr, long oldval, long newval, + int * spinlock) +{ + if (__builtin_expect (__pthread_has_cas, 1)) + return __compare_and_swap(ptr, oldval, newval); + else + return __pthread_compare_and_swap(ptr, oldval, newval, spinlock); +} + +#elif defined(HAS_COMPARE_AND_SWAP) + +static inline int compare_and_swap(long * ptr, long oldval, long newval, + int * spinlock) +{ + return __compare_and_swap(ptr, oldval, newval); +} + +#else + +extern int __pthread_compare_and_swap(long * ptr, long oldval, long newval, + int * spinlock); + +static inline int compare_and_swap(long * ptr, long oldval, long newval, + int * spinlock) +{ + return __pthread_compare_and_swap(ptr, oldval, newval, spinlock); +} + +#endif + +/* Internal locks */ + +extern void internal_function __pthread_lock(struct _pthread_fastlock * lock, + pthread_descr self); +extern void internal_function __pthread_unlock(struct _pthread_fastlock *lock); + +static inline void __pthread_init_lock(struct _pthread_fastlock * lock) +{ + lock->__status = 0; + lock->__spinlock = 0; +} + +static inline int __pthread_trylock (struct _pthread_fastlock * lock) +{ + long oldstatus; + + do { + oldstatus = lock->__status; + if (oldstatus != 0) return EBUSY; + } while(! compare_and_swap(&lock->__status, 0, 1, &lock->__spinlock)); + return 0; +} + +#define LOCK_INITIALIZER {0, 0} + +/* Operations on pthread_atomic, which is defined in internals.h */ + +static inline long atomic_increment(struct pthread_atomic *pa) +{ + long oldval; + + do { + oldval = pa->p_count; + } while (!compare_and_swap(&pa->p_count, oldval, oldval + 1, &pa->p_spinlock)); + + return oldval; +} + + +static inline long atomic_decrement(struct pthread_atomic *pa) +{ + long oldval; + + do { + oldval = pa->p_count; + } while (!compare_and_swap(&pa->p_count, oldval, oldval - 1, &pa->p_spinlock)); + + return oldval; +} + +#define ATOMIC_INITIALIZER { 0, 0 } + diff --git a/libpthread/linuxthreads/sysdeps/alpha/pt-machine.h b/libpthread/linuxthreads/sysdeps/alpha/pt-machine.h new file mode 100644 index 000000000..e59c6906c --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/alpha/pt-machine.h @@ -0,0 +1,108 @@ +/* Machine-dependent pthreads configuration and inline functions. + Alpha version. + Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Richard Henderson <rth@tamu.edu>. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifndef PT_EI +# define PT_EI extern inline +#endif + +#include <asm/pal.h> + + +/* Get some notion of the current stack. Need not be exactly the top + of the stack, just something somewhere in the current frame. */ +#define CURRENT_STACK_FRAME stack_pointer +register char *stack_pointer __asm__("$30"); + + +/* Spinlock implementation; required. */ +PT_EI long int +testandset (int *spinlock) +{ + long int ret, temp; + + __asm__ __volatile__( + "/* Inline spinlock test & set */\n" + "1:\t" + "ldl_l %0,%3\n\t" + "bne %0,2f\n\t" + "or $31,1,%1\n\t" + "stl_c %1,%2\n\t" + "beq %1,1b\n" + "2:\tmb\n" + "/* End spinlock test & set */" + : "=&r"(ret), "=&r"(temp), "=m"(*spinlock) + : "m"(*spinlock) + : "memory"); + + return ret; +} + +/* Spinlock release; default is just set to zero. */ +#define RELEASE(spinlock) \ + __asm__ __volatile__("mb" : : : "memory"); \ + *spinlock = 0 + + +/* Begin allocating thread stacks at this address. Default is to allocate + them just below the initial program stack. */ +#define THREAD_STACK_START_ADDRESS 0x40000000000 + + +/* Return the thread descriptor for the current thread. */ +#define THREAD_SELF \ +({ \ + register pthread_descr __self __asm__("$0"); \ + __asm__ ("call_pal %1" : "=r"(__self) : "i"(PAL_rduniq) : "$0"); \ + __self; \ +}) + +/* Initialize the thread-unique value. */ +#define INIT_THREAD_SELF(descr, nr) \ +{ \ + register pthread_descr __self __asm__("$16") = (descr); \ + __asm__ __volatile__ ("call_pal %1" : : "r"(__self), "i"(PAL_wruniq)); \ +} + + +/* Compare-and-swap for semaphores. */ + +#define HAS_COMPARE_AND_SWAP +PT_EI int +__compare_and_swap (long int *p, long int oldval, long int newval) +{ + long int ret; + + __asm__ __volatile__ ( + "/* Inline compare & swap */\n" + "1:\t" + "ldq_l %0,%4\n\t" + "cmpeq %0,%2,%0\n\t" + "beq %0,2f\n\t" + "mov %3,%0\n\t" + "stq_c %0,%1\n\t" + "beq %0,1b\n\t" + "2:\tmb\n" + "/* End compare & swap */" + : "=&r"(ret), "=m"(*p) + : "r"(oldval), "r"(newval), "m"(*p)); + + return ret; +} diff --git a/libpthread/linuxthreads/sysdeps/alpha/sigcontextinfo.h b/libpthread/linuxthreads/sysdeps/alpha/sigcontextinfo.h new file mode 100644 index 000000000..eb6f4f075 --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/alpha/sigcontextinfo.h @@ -0,0 +1,25 @@ +/* Copyright (C) 1999 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, write to the Free + Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA + 02111-1307 USA. */ + +#define SIGCONTEXT struct sigcontext +#define SIGCONTEXT_EXTRA_ARGS +#define GET_PC(ctx) ((void *) (ctx).sc_pc) +#define GET_FRAME(ctx) ((void *) (ctx).sc_regs[15]) +#define GET_STACK(ctx) ((void *) (ctx).sc_regs[30]) +#define CALL_SIGHANDLER(handler, signo, ctx) \ + (handler)((signo), SIGCONTEXT_EXTRA_ARGS (ctx)) diff --git a/libpthread/linuxthreads/sysdeps/arm/bits/armsigctx.h b/libpthread/linuxthreads/sysdeps/arm/bits/armsigctx.h new file mode 100644 index 000000000..4530cdbda --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/arm/bits/armsigctx.h @@ -0,0 +1,73 @@ +/* Definition of `struct sigcontext' for Linux/ARM + Copyright (C) 1996, 1997, 1998, 1999, 2000 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, write to the Free + Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA + 02111-1307 USA. */ + +/* The format of struct sigcontext changed between 2.0 and 2.1 kernels. + Fortunately 2.0 puts a magic number in the first word and this is not + a legal value for `trap_no', so we can tell them apart. */ + +/* Early 2.2 and 2.3 kernels do not have the `fault_address' member in + the sigcontext structure. Unfortunately there is no reliable way + to test for its presence and this word will contain garbage for too-old + kernels. Versions 2.2.14 and 2.3.35 (plus later versions) are known to + include this element. */ + +#ifndef __ARMSIGCTX_H +#define __ARMSIGCTX_H 1 + +#include <asm/ptrace.h> + +union k_sigcontext + { + struct + { + unsigned long int trap_no; + unsigned long int error_code; + unsigned long int oldmask; + unsigned long int arm_r0; + unsigned long int arm_r1; + unsigned long int arm_r2; + unsigned long int arm_r3; + unsigned long int arm_r4; + unsigned long int arm_r5; + unsigned long int arm_r6; + unsigned long int arm_r7; + unsigned long int arm_r8; + unsigned long int arm_r9; + unsigned long int arm_r10; + unsigned long int arm_fp; + unsigned long int arm_ip; + unsigned long int arm_sp; + unsigned long int arm_lr; + unsigned long int arm_pc; + unsigned long int arm_cpsr; + unsigned long fault_address; + } v21; + struct + { + unsigned long int magic; + struct pt_regs reg; + unsigned long int trap_no; + unsigned long int error_code; + unsigned long int oldmask; + } v20; +}; + +#define SIGCONTEXT_2_0_MAGIC 0x4B534154 + +#endif /* bits/armsigctx.h */ diff --git a/libpthread/linuxthreads/sysdeps/arm/pt-machine.h b/libpthread/linuxthreads/sysdeps/arm/pt-machine.h new file mode 100644 index 000000000..d4dc4c4ed --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/arm/pt-machine.h @@ -0,0 +1,48 @@ +/* Machine-dependent pthreads configuration and inline functions. + ARM version. + Copyright (C) 1997, 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Philip Blundell <philb@gnu.org>. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifndef PT_EI +# define PT_EI extern inline +#endif + + +/* This will not work on ARM1 or ARM2 because SWP is lacking on those + machines. Unfortunately we have no way to detect this at compile + time; let's hope nobody tries to use one. */ + +/* Spinlock implementation; required. */ +PT_EI int +testandset (int *spinlock) +{ + register unsigned int ret; + + __asm__ __volatile__("swp %0, %1, [%2]" + : "=r"(ret) + : "0"(1), "r"(spinlock)); + + return ret; +} + + +/* Get some notion of the current stack. Need not be exactly the top + of the stack, just something somewhere in the current frame. */ +#define CURRENT_STACK_FRAME stack_pointer +register char * stack_pointer __asm__ ("sp"); diff --git a/libpthread/linuxthreads/sysdeps/i386/i686/pt-machine.h b/libpthread/linuxthreads/sysdeps/i386/i686/pt-machine.h new file mode 100644 index 000000000..8d9ea709b --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/i386/i686/pt-machine.h @@ -0,0 +1,67 @@ +/* Machine-dependent pthreads configuration and inline functions. + i686 version. + Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Richard Henderson <rth@tamu.edu>. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifndef PT_EI +# define PT_EI extern inline +#endif + + +/* Get some notion of the current stack. Need not be exactly the top + of the stack, just something somewhere in the current frame. */ +#define CURRENT_STACK_FRAME stack_pointer +register char * stack_pointer __asm__ ("%esp"); + + +/* Spinlock implementation; required. */ +PT_EI int +testandset (int *spinlock) +{ + int ret; + + __asm__ __volatile__ ( + "xchgl %0, %1" + : "=r"(ret), "=m"(*spinlock) + : "0"(1), "m"(*spinlock) + : "memory"); + + return ret; +} + + +/* Compare-and-swap for semaphores. It's always available on i686. */ +#define HAS_COMPARE_AND_SWAP + +PT_EI int +__compare_and_swap (long int *p, long int oldval, long int newval) +{ + char ret; + long int readval; + + __asm__ __volatile__ ("lock; cmpxchgl %3, %1; sete %0" + : "=q" (ret), "=m" (*p), "=a" (readval) + : "r" (newval), "m" (*p), "a" (oldval) + : "memory"); + return ret; +} + + +/* Use the LDT implementation only if the kernel is fixed. */ +//#include "../useldt.h" diff --git a/libpthread/linuxthreads/sysdeps/i386/pt-machine.h b/libpthread/linuxthreads/sysdeps/i386/pt-machine.h new file mode 100644 index 000000000..f542bb2d2 --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/i386/pt-machine.h @@ -0,0 +1,99 @@ +/* Machine-dependent pthreads configuration and inline functions. + i386 version. + Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Richard Henderson <rth@tamu.edu>. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifndef PT_EI +# define PT_EI extern inline +#endif + +/* Get some notion of the current stack. Need not be exactly the top + of the stack, just something somewhere in the current frame. */ +#define CURRENT_STACK_FRAME stack_pointer +register char * stack_pointer __asm__ ("%esp"); + + +/* Spinlock implementation; required. */ +PT_EI int +testandset (int *spinlock) +{ + int ret; + + __asm__ __volatile__( + "xchgl %0, %1" + : "=r"(ret), "=m"(*spinlock) + : "0"(1), "m"(*spinlock) + : "memory"); + + return ret; +} + + +/* Compare-and-swap for semaphores. + Available on the 486 and above, but not on the 386. + We test dynamically whether it's available or not. */ + +#define HAS_COMPARE_AND_SWAP +#define TEST_FOR_COMPARE_AND_SWAP + +PT_EI int +__compare_and_swap (long int *p, long int oldval, long int newval) +{ + char ret; + long int readval; + + __asm__ __volatile__ ("lock; cmpxchgl %3, %1; sete %0" + : "=q" (ret), "=m" (*p), "=a" (readval) + : "r" (newval), "m" (*p), "a" (oldval) + : "memory"); + return ret; +} + + +PT_EI int +get_eflags (void) +{ + int res; + __asm__ __volatile__ ("pushfl; popl %0" : "=r" (res) : ); + return res; +} + + +PT_EI void +set_eflags (int newflags) +{ + __asm__ __volatile__ ("pushl %0; popfl" : : "r" (newflags) : "cc"); +} + + +PT_EI int +compare_and_swap_is_available (void) +{ + int oldflags = get_eflags (); + int changed; + /* Flip AC bit in EFLAGS. */ + set_eflags (oldflags ^ 0x40000); + /* See if bit changed. */ + changed = (get_eflags () ^ oldflags) & 0x40000; + /* Restore EFLAGS. */ + set_eflags (oldflags); + /* If the AC flag did not change, it's a 386 and it lacks cmpxchg. + Otherwise, it's a 486 or above and it has cmpxchg. */ + return changed != 0; +} diff --git a/libpthread/linuxthreads/sysdeps/i386/sigcontextinfo.h b/libpthread/linuxthreads/sysdeps/i386/sigcontextinfo.h new file mode 100644 index 000000000..42c18b22b --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/i386/sigcontextinfo.h @@ -0,0 +1,24 @@ +/* Copyright (C) 1998, 1999 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#define SIGCONTEXT struct sigcontext +#define SIGCONTEXT_EXTRA_ARGS +#define GET_PC(ctx) ((void *) ctx.eip) +#define GET_FRAME(ctx) ((void *) ctx.ebp) +#define GET_STACK(ctx) ((void *) ctx.esp_at_signal) diff --git a/libpthread/linuxthreads/sysdeps/i386/useldt.h b/libpthread/linuxthreads/sysdeps/i386/useldt.h new file mode 100644 index 000000000..1a789e2e0 --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/i386/useldt.h @@ -0,0 +1,170 @@ +/* Special definitions for ix86 machine using segment register based + thread descriptor. + Copyright (C) 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Ulrich Drepper <drepper@cygnus.com>. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#include <stddef.h> /* For offsetof. */ + + +/* We don't want to include the kernel header. So duplicate the + information. */ + +/* Structure passed on `modify_ldt' call. */ +struct modify_ldt_ldt_s +{ + unsigned int entry_number; + unsigned long int base_addr; + unsigned int limit; + unsigned int seg_32bit:1; + unsigned int contents:2; + unsigned int read_exec_only:1; + unsigned int limit_in_pages:1; + unsigned int seg_not_present:1; + unsigned int useable:1; + unsigned int empty:25; +}; + +/* System call to set LDT entry. */ +extern int __modify_ldt (int, struct modify_ldt_ldt_s *, size_t); + + +/* Return the thread descriptor for the current thread. + + The contained asm must *not* be marked volatile since otherwise + assignments like + pthread_descr self = thread_self(); + do not get optimized away. */ +#define THREAD_SELF \ +({ \ + register pthread_descr __self; \ + __asm__ ("movl %%gs:%c1,%0" : "=r" (__self) \ + : "i" (offsetof (struct _pthread_descr_struct, p_self))); \ + __self; \ +}) + +/* Initialize the thread-unique value. */ +#define INIT_THREAD_SELF(descr, nr) \ +{ \ + struct modify_ldt_ldt_s ldt_entry = \ + { nr, (unsigned long int) descr, sizeof (*descr), 1, 0, 0, 0, 0, 1, 0 }; \ + if (__modify_ldt (1, &ldt_entry, sizeof (ldt_entry)) != 0) \ + abort (); \ + __asm__ __volatile__ ("movw %w0, %%gs" : : "r" (nr * 8 + 7)); \ +} + +/* Free resources associated with thread descriptor. */ +#define FREE_THREAD_SELF(descr, nr) \ +{ \ + struct modify_ldt_ldt_s ldt_entry = \ + { nr, 0, 0, 0, 0, 1, 0, 1, 0, 0 }; \ + __asm__ __volatile__ ("movw %w0,%%gs" : : "r" (0)); \ + __modify_ldt (1, &ldt_entry, sizeof (ldt_entry)); \ +} + +/* Read member of the thread descriptor directly. */ +#define THREAD_GETMEM(descr, member) \ +({ \ + __typeof__ (descr->member) __value; \ + if (sizeof (__value) == 1) \ + __asm__ __volatile__ ("movb %%gs:%P2,%b0" \ + : "=r" (__value) \ + : "0" (0), \ + "i" (offsetof (struct _pthread_descr_struct, \ + member))); \ + else \ + { \ + if (sizeof (__value) != 4) \ + /* There should not be any value with a size other than 1 or 4. */ \ + abort (); \ + \ + __asm__ __volatile__ ("movl %%gs:%P1,%0" \ + : "=r" (__value) \ + : "i" (offsetof (struct _pthread_descr_struct, \ + member))); \ + } \ + __value; \ +}) + +/* Same as THREAD_GETMEM, but the member offset can be non-constant. */ +#define THREAD_GETMEM_NC(descr, member) \ +({ \ + __typeof__ (descr->member) __value; \ + if (sizeof (__value) == 1) \ + __asm__ __volatile__ ("movb %%gs:(%2),%b0" \ + : "=r" (__value) \ + : "0" (0), \ + "r" (offsetof (struct _pthread_descr_struct, \ + member))); \ + else \ + { \ + if (sizeof (__value) != 4) \ + /* There should not be any value with a size other than 1 or 4. */ \ + abort (); \ + \ + __asm__ __volatile__ ("movl %%gs:(%1),%0" \ + : "=r" (__value) \ + : "r" (offsetof (struct _pthread_descr_struct, \ + member))); \ + } \ + __value; \ +}) + +/* Same as THREAD_SETMEM, but the member offset can be non-constant. */ +#define THREAD_SETMEM(descr, member, value) \ +({ \ + __typeof__ (descr->member) __value = (value); \ + if (sizeof (__value) == 1) \ + __asm__ __volatile__ ("movb %0,%%gs:%P1" : \ + : "r" (__value), \ + "i" (offsetof (struct _pthread_descr_struct, \ + member))); \ + else \ + { \ + if (sizeof (__value) != 4) \ + /* There should not be any value with a size other than 1 or 4. */ \ + abort (); \ + \ + __asm__ __volatile__ ("movl %0,%%gs:%P1" : \ + : "r" (__value), \ + "i" (offsetof (struct _pthread_descr_struct, \ + member))); \ + } \ +}) + +/* Set member of the thread descriptor directly. */ +#define THREAD_SETMEM_NC(descr, member, value) \ +({ \ + __typeof__ (descr->member) __value = (value); \ + if (sizeof (__value) == 1) \ + __asm__ __volatile__ ("movb %0,%%gs:(%1)" : \ + : "r" (__value), \ + "r" (offsetof (struct _pthread_descr_struct, \ + member))); \ + else \ + { \ + if (sizeof (__value) != 4) \ + /* There should not be any value with a size other than 1 or 4. */ \ + abort (); \ + \ + __asm__ __volatile__ ("movl %0,%%gs:(%1)" : \ + : "r" (__value), \ + "r" (offsetof (struct _pthread_descr_struct, \ + member))); \ + } \ +}) diff --git a/libpthread/linuxthreads/sysdeps/m68k/pt-machine.h b/libpthread/linuxthreads/sysdeps/m68k/pt-machine.h new file mode 100644 index 000000000..38ea68114 --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/m68k/pt-machine.h @@ -0,0 +1,62 @@ +/* Machine-dependent pthreads configuration and inline functions. + m68k version. + Copyright (C) 1996, 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Richard Henderson <rth@tamu.edu>. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If + not, write to the Free Software Foundation, Inc., + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef PT_EI +# define PT_EI extern inline +#endif + + +/* Spinlock implementation; required. */ +PT_EI int +testandset (int *spinlock) +{ + char ret; + + __asm__ __volatile__("tas %1; sne %0" + : "=dm"(ret), "=m"(*spinlock) + : "m"(*spinlock) + : "cc"); + + return ret; +} + + +/* Get some notion of the current stack. Need not be exactly the top + of the stack, just something somewhere in the current frame. */ +#define CURRENT_STACK_FRAME stack_pointer +register char * stack_pointer __asm__ ("%sp"); + + +/* Compare-and-swap for semaphores. */ + +#define HAS_COMPARE_AND_SWAP +PT_EI int +__compare_and_swap (long int *p, long int oldval, long int newval) +{ + char ret; + long int readval; + + __asm__ __volatile__ ("casl %2, %3, %1; seq %0" + : "=dm" (ret), "=m" (*p), "=d" (readval) + : "d" (newval), "m" (*p), "2" (oldval)); + + return ret; +} diff --git a/libpthread/linuxthreads/sysdeps/m68k/sigcontextinfo.h b/libpthread/linuxthreads/sysdeps/m68k/sigcontextinfo.h new file mode 100644 index 000000000..b7e08cfc9 --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/m68k/sigcontextinfo.h @@ -0,0 +1,26 @@ +/* Copyright (C) 1998, 1999, 2001 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Andreas Schwab <schwab@issan.informatik.uni-dortmund.de>, 1998. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, write to the Free + Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA + 02111-1307 USA. */ + +#define SIGCONTEXT int _code, struct sigcontext * +#define SIGCONTEXT_EXTRA_ARGS _code, +#define GET_PC(ctx) ((void *) (ctx)->sc_pc) +#define GET_FRAME(ctx) ((void *) __builtin_frame_address (1)) +#define GET_STACK(ctx) ((void *) (ctx)->sc_usp) +#define CALL_SIGHANDLER(handler, signo, ctx) \ + (handler)((signo), SIGCONTEXT_EXTRA_ARGS (ctx)) diff --git a/libpthread/linuxthreads/sysdeps/mips/pt-machine.h b/libpthread/linuxthreads/sysdeps/mips/pt-machine.h new file mode 100644 index 000000000..527392308 --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/mips/pt-machine.h @@ -0,0 +1,90 @@ +/* Machine-dependent pthreads configuration and inline functions. + + Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Ralf Baechle <ralf@gnu.ai.mit.edu>. + Based on the Alpha version by Richard Henderson <rth@tamu.edu>. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If + not, write to the Free Software Foundation, Inc., + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + + TODO: This version makes use of MIPS ISA 2 features. It won't + work on ISA 1. These machines will have to take the overhead of + a sysmips(MIPS_ATOMIC_SET, ...) syscall which isn't implemented + yet correctly. There is however a better solution for R3000 + uniprocessor machines possible. */ + +#ifndef PT_EI +# define PT_EI extern inline +#endif + + +/* Spinlock implementation; required. */ +PT_EI long int +testandset (int *spinlock) +{ + long int ret, temp; + + __asm__ __volatile__( + "# Inline spinlock test & set\n\t" + ".set\tmips2\n" + "1:\tll\t%0,%3\n\t" + "bnez\t%0,2f\n\t" + ".set\tnoreorder\n\t" + "li\t%1,1\n\t" + ".set\treorder\n\t" + "sc\t%1,%2\n\t" + "beqz\t%1,1b\n" + "2:\t.set\tmips0\n\t" + "/* End spinlock test & set */" + : "=&r"(ret), "=&r" (temp), "=m"(*spinlock) + : "m"(*spinlock) + : "memory"); + + return ret; +} + + +/* Get some notion of the current stack. Need not be exactly the top + of the stack, just something somewhere in the current frame. */ +#define CURRENT_STACK_FRAME stack_pointer +register char * stack_pointer __asm__ ("$29"); + + +/* Compare-and-swap for semaphores. */ + +#define HAS_COMPARE_AND_SWAP +PT_EI int +__compare_and_swap (long int *p, long int oldval, long int newval) +{ + long ret; + + __asm__ __volatile__ ( + "/* Inline compare & swap */\n\t" + ".set\tmips2\n" + "1:\tll\t%0,%4\n\t" + ".set\tnoreorder\n\t" + "bne\t%0,%2,2f\n\t" + "move\t%0,%3\n\t" + ".set\treorder\n\t" + "sc\t%0,%1\n\t" + "beqz\t%0,1b\n" + "2:\t.set\tmips0\n\t" + "/* End compare & swap */" + : "=&r"(ret), "=m"(*p) + : "r"(oldval), "r"(newval), "m"(*p)); + + return ret; +} diff --git a/libpthread/linuxthreads/sysdeps/mips/sigcontextinfo.h b/libpthread/linuxthreads/sysdeps/mips/sigcontextinfo.h new file mode 100644 index 000000000..a51c6f043 --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/mips/sigcontextinfo.h @@ -0,0 +1,27 @@ +/* Copyright (C) 2000, 2001 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Andreas Jaeger <aj@suse.de>, 2000. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, write to the Free + Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA + 02111-1307 USA. */ + + +#define SIGCONTEXT unsigned long _code, struct sigcontext * +#define SIGCONTEXT_EXTRA_ARGS _code, +#define GET_PC(ctx) ((void *) ctx->sc_pc) +#define GET_FRAME(ctx) ((void *) ctx->sc_regs[30]) +#define GET_STACK(ctx) ((void *) ctx->sc_regs[29]) +#define CALL_SIGHANDLER(handler, signo, ctx) \ + (handler)((signo), SIGCONTEXT_EXTRA_ARGS (ctx)) diff --git a/libpthread/linuxthreads/sysdeps/powerpc/pt-machine.h b/libpthread/linuxthreads/sysdeps/powerpc/pt-machine.h new file mode 100644 index 000000000..578369a7f --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/powerpc/pt-machine.h @@ -0,0 +1,69 @@ +/* Machine-dependent pthreads configuration and inline functions. + powerpc version. + Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If + not, write to the Free Software Foundation, Inc., + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* These routines are from Appendix G of the 'PowerPC 601 RISC Microprocessor + User's Manual', by IBM and Motorola. */ + +#ifndef PT_EI +# define PT_EI extern inline +#endif + +/* For multiprocessor systems, we want to ensure all memory accesses + are completed before we reset a lock. */ +#if 0 +/* on non multiprocessor systems, you can just: */ +#define sync() /* nothing */ +#else +#define sync() __asm__ __volatile__ ("sync") +#endif + +/* Get some notion of the current stack. Need not be exactly the top + of the stack, just something somewhere in the current frame. */ +#define CURRENT_STACK_FRAME stack_pointer +register char * stack_pointer __asm__ ("r1"); + +/* Compare-and-swap for semaphores. */ +/* note that test-and-set(x) is the same as compare-and-swap(x, 0, 1) */ + +#define HAS_COMPARE_AND_SWAP +#if BROKEN_PPC_ASM_CR0 +static +#else +PT_EI +#endif +int +__compare_and_swap (long int *p, long int oldval, long int newval) +{ + int ret; + + sync(); + __asm__ __volatile__( + "0: lwarx %0,0,%1 ;" + " xor. %0,%3,%0;" + " bne 1f;" + " stwcx. %2,0,%1;" + " bne- 0b;" + "1: " + : "=&r"(ret) + : "r"(p), "r"(newval), "r"(oldval) + : "cr0", "memory"); + sync(); + return ret == 0; +} diff --git a/libpthread/linuxthreads/sysdeps/powerpc/sigcontextinfo.h b/libpthread/linuxthreads/sysdeps/powerpc/sigcontextinfo.h new file mode 100644 index 000000000..138a15cfa --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/powerpc/sigcontextinfo.h @@ -0,0 +1,27 @@ +/* Copyright (C) 1998, 1999, 2001 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, write to the Free + Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA + 02111-1307 USA. */ + +#include <signal.h> + +#define SIGCONTEXT struct sigcontext * +#define SIGCONTEXT_EXTRA_ARGS +#define GET_PC(ctx) ((void *)((ctx)->regs->nip)) +#define GET_FRAME(ctx) (*(void **)((ctx)->regs->gpr[1])) +#define GET_STACK(ctx) ((void *)((ctx)->regs->gpr[1])) +#define CALL_SIGHANDLER(handler, signo, ctx) \ + (handler)((signo), SIGCONTEXT_EXTRA_ARGS (ctx)) diff --git a/libpthread/linuxthreads/sysdeps/pthread/bits/libc-lock.h b/libpthread/linuxthreads/sysdeps/pthread/bits/libc-lock.h new file mode 100644 index 000000000..a14cea1aa --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/pthread/bits/libc-lock.h @@ -0,0 +1,214 @@ +/* libc-internal interface for mutex locks. LinuxThreads version. + Copyright (C) 1996, 1997, 1998, 1999 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifndef _BITS_LIBC_LOCK_H +#define _BITS_LIBC_LOCK_H 1 + +#include <pthread.h> + +/* Mutex type. */ +#ifdef _LIBC +typedef pthread_mutex_t __libc_lock_t; +#else +typedef struct __libc_lock_opaque__ __libc_lock_t; +#endif + +/* Type for key to thread-specific data. */ +typedef pthread_key_t __libc_key_t; + +/* Define a lock variable NAME with storage class CLASS. The lock must be + initialized with __libc_lock_init before it can be used (or define it + with __libc_lock_define_initialized, below). Use `extern' for CLASS to + declare a lock defined in another module. In public structure + definitions you must use a pointer to the lock structure (i.e., NAME + begins with a `*'), because its storage size will not be known outside + of libc. */ +#define __libc_lock_define(CLASS,NAME) \ + CLASS __libc_lock_t NAME; + +/* Define an initialized lock variable NAME with storage class CLASS. + + For the C library we take a deeper look at the initializer. For this + implementation all fields are initialized to zero. Therefore we + don't initialize the variable which allows putting it into the BSS + section. */ +#define __libc_lock_define_initialized(CLASS,NAME) \ + CLASS __libc_lock_t NAME; + +/* Define an initialized recursive lock variable NAME with storage + class CLASS. */ +#define __libc_lock_define_initialized_recursive(CLASS,NAME) \ + CLASS __libc_lock_t NAME = PTHREAD_RECURSIVE_MUTEX_INITIALIZER_NP; + +/* Initialize the named lock variable, leaving it in a consistent, unlocked + state. */ +#define __libc_lock_init(NAME) \ + (__pthread_mutex_init != NULL ? __pthread_mutex_init (&(NAME), NULL) : 0); + +/* Same as last but this time we initialize a recursive mutex. */ +#define __libc_lock_init_recursive(NAME) \ + do { \ + if (__pthread_mutex_init != NULL) \ + { \ + pthread_mutexattr_t __attr; \ + __pthread_mutexattr_init (&__attr); \ + __pthread_mutexattr_settype (&__attr, PTHREAD_MUTEX_RECURSIVE_NP); \ + __pthread_mutex_init (&(NAME), &__attr); \ + __pthread_mutexattr_destroy (&__attr); \ + } \ + } while (0); + +/* Finalize the named lock variable, which must be locked. It cannot be + used again until __libc_lock_init is called again on it. This must be + called on a lock variable before the containing storage is reused. */ +#define __libc_lock_fini(NAME) \ + (__pthread_mutex_destroy != NULL ? __pthread_mutex_destroy (&(NAME)) : 0); + +/* Finalize recursive named lock. */ +#define __libc_lock_fini_recursive(NAME) __libc_lock_fini (NAME) + +/* Lock the named lock variable. */ +#define __libc_lock_lock(NAME) \ + (__pthread_mutex_lock != NULL ? __pthread_mutex_lock (&(NAME)) : 0); + +/* Lock the recursive named lock variable. */ +#define __libc_lock_lock_recursive(NAME) __libc_lock_lock (NAME) + +/* Try to lock the named lock variable. */ +#define __libc_lock_trylock(NAME) \ + (__pthread_mutex_trylock != NULL ? __pthread_mutex_trylock (&(NAME)) : 0) + +/* Try to lock the recursive named lock variable. */ +#define __libc_lock_trylock_recursive(NAME) __libc_lock_trylock (NAME) + +/* Unlock the named lock variable. */ +#define __libc_lock_unlock(NAME) \ + (__pthread_mutex_unlock != NULL ? __pthread_mutex_unlock (&(NAME)) : 0); + +/* Unlock the recursive named lock variable. */ +#define __libc_lock_unlock_recursive(NAME) __libc_lock_unlock (NAME) + + +/* Define once control variable. */ +#if PTHREAD_ONCE_INIT == 0 +/* Special case for static variables where we can avoid the initialization + if it is zero. */ +# define __libc_once_define(CLASS, NAME) \ + CLASS pthread_once_t NAME +#else +# define __libc_once_define(CLASS, NAME) \ + CLASS pthread_once_t NAME = PTHREAD_ONCE_INIT +#endif + +/* Call handler iff the first call. */ +#define __libc_once(ONCE_CONTROL, INIT_FUNCTION) \ + do { \ + if (__pthread_once != NULL) \ + __pthread_once (&(ONCE_CONTROL), (INIT_FUNCTION)); \ + else if ((ONCE_CONTROL) == 0) { \ + INIT_FUNCTION (); \ + (ONCE_CONTROL) = 1; \ + } \ + } while (0) + + +/* Start critical region with cleanup. */ +#define __libc_cleanup_region_start(FCT, ARG) \ + { struct _pthread_cleanup_buffer _buffer; \ + int _avail = _pthread_cleanup_push_defer != NULL; \ + if (_avail) { \ + _pthread_cleanup_push_defer (&_buffer, (FCT), (ARG)); \ + } + +/* End critical region with cleanup. */ +#define __libc_cleanup_region_end(DOIT) \ + if (_avail) { \ + _pthread_cleanup_pop_restore (&_buffer, (DOIT)); \ + } \ + } + +/* Sometimes we have to exit the block in the middle. */ +#define __libc_cleanup_end(DOIT) \ + if (_avail) { \ + _pthread_cleanup_pop_restore (&_buffer, (DOIT)); \ + } + +/* Create thread-specific key. */ +#define __libc_key_create(KEY, DESTRUCTOR) \ + (__pthread_key_create != NULL ? __pthread_key_create (KEY, DESTRUCTOR) : 1) + +/* Get thread-specific data. */ +#define __libc_getspecific(KEY) \ + (__pthread_getspecific != NULL ? __pthread_getspecific (KEY) : NULL) + +/* Set thread-specific data. */ +#define __libc_setspecific(KEY, VALUE) \ + (__pthread_setspecific != NULL ? __pthread_setspecific (KEY, VALUE) : 0) + + +/* Register handlers to execute before and after `fork'. */ +#define __libc_atfork(PREPARE, PARENT, CHILD) \ + (__pthread_atfork != NULL ? __pthread_atfork (PREPARE, PARENT, CHILD) : 0) + + +/* Make the pthread functions weak so that we can elide them from + single-threaded processes. */ +#ifndef __NO_WEAK_PTHREAD_ALIASES +# ifdef weak_extern +weak_extern (__pthread_mutex_init) +weak_extern (__pthread_mutex_destroy) +weak_extern (__pthread_mutex_lock) +weak_extern (__pthread_mutex_trylock) +weak_extern (__pthread_mutex_unlock) +weak_extern (__pthread_mutexattr_init) +weak_extern (__pthread_mutexattr_destroy) +weak_extern (__pthread_mutexattr_settype) +weak_extern (__pthread_key_create) +weak_extern (__pthread_setspecific) +weak_extern (__pthread_getspecific) +weak_extern (__pthread_once) +weak_extern (__pthread_initialize) +weak_extern (__pthread_atfork) +weak_extern (_pthread_cleanup_push_defer) +weak_extern (_pthread_cleanup_pop_restore) +# else +# pragma weak __pthread_mutex_init +# pragma weak __pthread_mutex_destroy +# pragma weak __pthread_mutex_lock +# pragma weak __pthread_mutex_trylock +# pragma weak __pthread_mutex_unlock +# pragma weak __pthread_mutexattr_init +# pragma weak __pthread_mutexattr_destroy +# pragma weak __pthread_mutexattr_settype +# pragma weak __pthread_key_create +# pragma weak __pthread_setspecific +# pragma weak __pthread_getspecific +# pragma weak __pthread_once +# pragma weak __pthread_initialize +# pragma weak __pthread_atfork +# pragma weak _pthread_cleanup_push_defer +# pragma weak _pthread_cleanup_pop_restore +# endif +#endif + +/* We need portable names for some functions. E.g., when they are + used as argument to __libc_cleanup_region_start. */ +#define __libc_mutex_unlock __pthread_mutex_unlock + +#endif /* bits/libc-lock.h */ diff --git a/libpthread/linuxthreads/sysdeps/pthread/bits/libc-tsd.h b/libpthread/linuxthreads/sysdeps/pthread/bits/libc-tsd.h new file mode 100644 index 000000000..e38cdf550 --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/pthread/bits/libc-tsd.h @@ -0,0 +1,44 @@ +/* libc-internal interface for thread-specific data. LinuxThreads version. + Copyright (C) 1997, 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifndef _BITS_LIBC_TSD_H +#define _BITS_LIBC_TSD_H 1 + +#include <features.h> + +/* Fast thread-specific data internal to libc. */ +enum __libc_tsd_key_t { _LIBC_TSD_KEY_MALLOC = 0, + _LIBC_TSD_KEY_DL_ERROR, + _LIBC_TSD_KEY_N }; + +extern void *(*__libc_internal_tsd_get) __P ((enum __libc_tsd_key_t)); +extern int (*__libc_internal_tsd_set) __P ((enum __libc_tsd_key_t, + __const void *)); + +#define __libc_tsd_define(CLASS, KEY) CLASS void *__libc_tsd_##KEY##_data; +#define __libc_tsd_get(KEY) \ + (__libc_internal_tsd_get != NULL \ + ? __libc_internal_tsd_get (_LIBC_TSD_KEY_##KEY) \ + : __libc_tsd_##KEY##_data) +#define __libc_tsd_set(KEY, VALUE) \ + (__libc_internal_tsd_set != NULL \ + ? __libc_internal_tsd_set (_LIBC_TSD_KEY_##KEY, (VALUE)) \ + : ((__libc_tsd_##KEY##_data = (VALUE)), 0)) + +#endif /* bits/libc-tsd.h */ diff --git a/libpthread/linuxthreads/sysdeps/pthread/bits/pthreadtypes.h b/libpthread/linuxthreads/sysdeps/pthread/bits/pthreadtypes.h new file mode 100644 index 000000000..db4c3790c --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/pthread/bits/pthreadtypes.h @@ -0,0 +1,122 @@ +/* Linuxthreads - a simple clone()-based implementation of Posix */ +/* threads for Linux. */ +/* Copyright (C) 1996 Xavier Leroy (Xavier.Leroy@inria.fr) */ +/* */ +/* This program is free software; you can redistribute it and/or */ +/* modify it under the terms of the GNU Library General Public License */ +/* as published by the Free Software Foundation; either version 2 */ +/* of the License, or (at your option) any later version. */ +/* */ +/* This program is distributed in the hope that it will be useful, */ +/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ +/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ +/* GNU Library General Public License for more details. */ + +#if !defined _BITS_TYPES_H && !defined _PTHREAD_H +# error "Never include <bits/pthreadtypes.h> directly; use <sys/types.h> instead." +#endif + +#ifndef _BITS_PTHREADTYPES_H +#define _BITS_PTHREADTYPES_H 1 + +#define __need_schedparam +#include <bits/sched.h> + +/* Fast locks (not abstract because mutexes and conditions aren't abstract). */ +struct _pthread_fastlock +{ + long int __status; /* "Free" or "taken" or head of waiting list */ + int __spinlock; /* For compare-and-swap emulation */ +}; + +#ifndef _PTHREAD_DESCR_DEFINED +/* Thread descriptors */ +typedef struct _pthread_descr_struct *_pthread_descr; +# define _PTHREAD_DESCR_DEFINED +#endif + + +/* Attributes for threads. */ +typedef struct +{ + int __detachstate; + int __schedpolicy; + struct __sched_param __schedparam; + int __inheritsched; + int __scope; + size_t __guardsize; + int __stackaddr_set; + void *__stackaddr; + size_t __stacksize; +} pthread_attr_t; + + +/* Conditions (not abstract because of PTHREAD_COND_INITIALIZER */ +typedef struct +{ + struct _pthread_fastlock __c_lock; /* Protect against concurrent access */ + _pthread_descr __c_waiting; /* Threads waiting on this condition */ +} pthread_cond_t; + + +/* Attribute for conditionally variables. */ +typedef struct +{ + int __dummy; +} pthread_condattr_t; + +/* Keys for thread-specific data */ +typedef unsigned int pthread_key_t; + + +/* Mutexes (not abstract because of PTHREAD_MUTEX_INITIALIZER). */ +/* (The layout is unnatural to maintain binary compatibility + with earlier releases of LinuxThreads.) */ +typedef struct +{ + int __m_reserved; /* Reserved for future use */ + int __m_count; /* Depth of recursive locking */ + _pthread_descr __m_owner; /* Owner thread (if recursive or errcheck) */ + int __m_kind; /* Mutex kind: fast, recursive or errcheck */ + struct _pthread_fastlock __m_lock; /* Underlying fast lock */ +} pthread_mutex_t; + + +/* Attribute for mutex. */ +typedef struct +{ + int __mutexkind; +} pthread_mutexattr_t; + + +/* Once-only execution */ +typedef int pthread_once_t; + + +#ifdef __USE_UNIX98 +/* Read-write locks. */ +typedef struct _pthread_rwlock_t +{ + struct _pthread_fastlock __rw_lock; /* Lock to guarantee mutual exclusion */ + int __rw_readers; /* Number of readers */ + _pthread_descr __rw_writer; /* Identity of writer, or NULL if none */ + _pthread_descr __rw_read_waiting; /* Threads waiting for reading */ + _pthread_descr __rw_write_waiting; /* Threads waiting for writing */ + int __rw_kind; /* Reader/Writer preference selection */ + int __rw_pshared; /* Shared between processes or not */ +} pthread_rwlock_t; + + +/* Attribute for read-write locks. */ +typedef struct +{ + int __lockkind; + int __pshared; +} pthread_rwlockattr_t; +#endif + + +/* Thread identifiers */ +typedef unsigned long int pthread_t; + +#endif /* bits/pthreadtypes.h */ diff --git a/libpthread/linuxthreads/sysdeps/pthread/bits/stdio-lock.h b/libpthread/linuxthreads/sysdeps/pthread/bits/stdio-lock.h new file mode 100644 index 000000000..edc69f6cf --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/pthread/bits/stdio-lock.h @@ -0,0 +1,39 @@ +/* Thread package specific definitions of stream lock type. + Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#include <pthread.h> + +typedef pthread_mutex_t _IO_lock_t; + +/* We need recursive (counting) mutexes. */ +#define _IO_lock_initializer PTHREAD_RECURSIVE_MUTEX_INITIALIZER_NP + + +#define _IO_cleanup_region_start(_fct, _fp) \ + __libc_cleanup_region_start (_fct, _fp) +#define _IO_cleanup_region_end(_doit) \ + __libc_cleanup_region_end (_doit) +#define _IO_lock_init(_name) \ + __libc_lock_init_recursive (_name) +#define _IO_lock_fini(_name) \ + __libc_lock_fini_recursive (_name) +#define _IO_lock_lock(_name) \ + __libc_lock_lock (_name) +#define _IO_lock_unlock(_name) \ + __libc_lock_unlock (_name) diff --git a/libpthread/linuxthreads/sysdeps/sparc/sparc32/pt-machine.h b/libpthread/linuxthreads/sysdeps/sparc/sparc32/pt-machine.h new file mode 100644 index 000000000..69af8579a --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/sparc/sparc32/pt-machine.h @@ -0,0 +1,66 @@ +/* Machine-dependent pthreads configuration and inline functions. + sparc version. + Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Richard Henderson <rth@tamu.edu>. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifndef PT_EI +# define PT_EI extern inline +#endif + +/* Spinlock implementation; required. */ +PT_EI int +testandset (int *spinlock) +{ + int ret; + + __asm__ __volatile__("ldstub %1,%0" + : "=r"(ret), "=m"(*spinlock) + : "m"(*spinlock)); + + return ret; +} + + +/* Spinlock release; default is just set to zero. */ +#define RELEASE(spinlock) \ + __asm__ __volatile__("stbar; stb %1,%0" : "=m"(*(spinlock)) : "r"(0)); + + +/* Get some notion of the current stack. Need not be exactly the top + of the stack, just something somewhere in the current frame. */ +#define CURRENT_STACK_FRAME stack_pointer +register char * stack_pointer __asm__("%sp"); + + +/* Registers %g6 and %g7 are reserved by the ABI for "system use". It + happens that Solaris uses %g6 for the thread pointer -- we do the same. */ +struct _pthread_descr_struct; +register struct _pthread_descr_struct *__thread_self __asm__("%g6"); + +/* Return the thread descriptor for the current thread. */ +#define THREAD_SELF __thread_self + +/* Initialize the thread-unique value. */ +#define INIT_THREAD_SELF(descr, nr) (__thread_self = (descr)) + +/* Access to data in the thread descriptor is easy. */ +#define THREAD_GETMEM(descr, member) __thread_self->member +#define THREAD_GETMEM_NC(descr, member) __thread_self->member +#define THREAD_SETMEM(descr, member, value) __thread_self->member = (value) +#define THREAD_SETMEM_NC(descr, member, value) __thread_self->member = (value) diff --git a/libpthread/linuxthreads/sysdeps/sparc/sparc32/sigcontextinfo.h b/libpthread/linuxthreads/sysdeps/sparc/sparc32/sigcontextinfo.h new file mode 100644 index 000000000..2c2770d07 --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/sparc/sparc32/sigcontextinfo.h @@ -0,0 +1,29 @@ +/* Copyright (C) 1999, 2000 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Jakub Jelinek <jakub@redhat.com>, 1999. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, write to the Free + Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA + 02111-1307 USA. */ + +#define SIGCONTEXT struct sigcontext * +#define SIGCONTEXT_EXTRA_ARGS +#define GET_PC(__ctx) ((void *) ((__ctx)->si_regs.pc)) +#define ADVANCE_STACK_FRAME(__next) \ + ((void *) (((unsigned *)(__next))+14)) + +#define GET_STACK(__ctx) ((void *) (__ctx)->si_regs.u_regs[14]) +#define GET_FRAME(__ctx) ADVANCE_STACK_FRAME (GET_STACK(__ctx)) +#define CALL_SIGHANDLER(handler, signo, ctx) \ + (handler)((signo), SIGCONTEXT_EXTRA_ARGS (ctx)) diff --git a/libpthread/linuxthreads/sysdeps/sparc/sparc64/pt-machine.h b/libpthread/linuxthreads/sysdeps/sparc/sparc64/pt-machine.h new file mode 100644 index 000000000..5560003fc --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/sparc/sparc64/pt-machine.h @@ -0,0 +1,77 @@ +/* Machine-dependent pthreads configuration and inline functions. + Sparc v9 version. + Copyright (C) 1997, 1998, 1999 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Richard Henderson <rth@tamu.edu>. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If + not, write to the Free Software Foundation, Inc., + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef PT_EI +# define PT_EI extern inline +#endif + + +/* Spinlock implementation; required. */ +PT_EI int +testandset (int *spinlock) +{ + int ret; + + __asm__ __volatile__("ldstub %1,%0" + : "=r"(ret), "=m"(*spinlock) : "m"(*spinlock)); + + return ret; +} + + +/* Get some notion of the current stack. Need not be exactly the top + of the stack, just something somewhere in the current frame. */ +#define CURRENT_STACK_FRAME stack_pointer +register char *stack_pointer __asm__ ("%sp"); + + +/* Registers %g6 and %g7 are reserved by the ABI for "system use". It + happens that Solaris uses %g6 for the thread pointer -- we do the same. */ +struct _pthread_descr_struct; +register struct _pthread_descr_struct *__thread_self __asm__("%g6"); + +/* Return the thread descriptor for the current thread. */ +#define THREAD_SELF __thread_self + +/* Initialize the thread-unique value. */ +#define INIT_THREAD_SELF(descr, nr) (__thread_self = (descr)) + + +/* Compare-and-swap for semaphores. */ + +#define HAS_COMPARE_AND_SWAP +PT_EI int +__compare_and_swap (long int *p, long int oldval, long int newval) +{ + long int readval; + + __asm__ __volatile__ ("casx [%4], %2, %0" + : "=r"(readval), "=m"(*p) + : "r"(oldval), "m"(*p), "r"(p), "0"(newval)); + + return readval == oldval; +} + +/* Access to data in the thread descriptor is easy. */ +#define THREAD_GETMEM(descr, member) __thread_self->member +#define THREAD_GETMEM_NC(descr, member) __thread_self->member +#define THREAD_SETMEM(descr, member, value) __thread_self->member = (value) +#define THREAD_SETMEM_NC(descr, member, value) __thread_self->member = (value) diff --git a/libpthread/linuxthreads/sysdeps/sparc/sparc64/sigcontextinfo.h b/libpthread/linuxthreads/sysdeps/sparc/sparc64/sigcontextinfo.h new file mode 100644 index 000000000..47e0d27f8 --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/sparc/sparc64/sigcontextinfo.h @@ -0,0 +1,32 @@ +/* Copyright (C) 1999 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Jakub Jelinek <jj@ultra.linux.cz>, 1999. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, write to the Free + Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA + 02111-1307 USA. */ + +#ifndef STACK_BIAS +#define STACK_BIAS 2047 +#endif +#define SIGCONTEXT struct sigcontext * +#define SIGCONTEXT_EXTRA_ARGS +#define GET_PC(__ctx) ((void *) ((__ctx)->sigc_regs.tpc)) +#define ADVANCE_STACK_FRAME(__next) \ + ((void *) (((unsigned long *) (((unsigned long int) (__next)) \ + + STACK_BIAS))+14)) +#define GET_STACK(__ctx) ((void *) ((__ctx)->sigc_regs.u_regs[14])) +#define GET_FRAME(__ctx) ADVANCE_STACK_FRAME (GET_STACK (__ctx)) +#define CALL_SIGHANDLER(handler, signo, ctx) \ + (handler)((signo), SIGCONTEXT_EXTRA_ARGS (ctx)) diff --git a/libpthread/linuxthreads/sysdeps/unix/sysv/linux/bits/local_lim.h b/libpthread/linuxthreads/sysdeps/unix/sysv/linux/bits/local_lim.h new file mode 100644 index 000000000..545a90b2e --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/unix/sysv/linux/bits/local_lim.h @@ -0,0 +1,55 @@ +/* Minimum guaranteed maximum values for system limits. Linux version. + Copyright (C) 1993, 94, 95, 96, 97, 98 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +/* The kernel header pollutes the namespace with the NR_OPEN symbol. + Remove this after including the header if necessary. */ +#ifndef NR_OPEN +# define __undef_NR_OPEN +#endif + +/* The kernel sources contain a file with all the needed information. */ +#include <linux/limits.h> + +/* Have to remove NR_OPEN? */ +#ifdef __undef_NR_OPEN +# undef NR_OPEN +# undef __undef_NR_OPEN +#endif + +/* The number of data keys per process. */ +#define _POSIX_THREAD_KEYS_MAX 128 +/* This is the value this implementation supports. */ +#define PTHREAD_KEYS_MAX 1024 + +/* Controlling the iterations of destructors for thread-specific data. */ +#define _POSIX_THREAD_DESTRUCTOR_ITERATIONS 4 +/* Number of iterations this implementation does. */ +#define PTHREAD_DESTRUCTOR_ITERATIONS _POSIX_THREAD_DESTRUCTOR_ITERATIONS + +/* The number of threads per process. */ +#define _POSIX_THREAD_THREADS_MAX 64 +/* This is the value this implementation supports. */ +#define PTHREAD_THREADS_MAX 1024 + +/* Maximum amount by which a process can descrease its asynchronous I/O + priority level. */ +#define AIO_PRIO_DELTA_MAX 20 + +/* Minimum size for a thread. We are free to choose a reasonable value. */ +#define PTHREAD_STACK_MIN 16384 diff --git a/libpthread/linuxthreads/sysdeps/unix/sysv/linux/bits/posix_opt.h b/libpthread/linuxthreads/sysdeps/unix/sysv/linux/bits/posix_opt.h new file mode 100644 index 000000000..15683b77e --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/unix/sysv/linux/bits/posix_opt.h @@ -0,0 +1,110 @@ +/* Define POSIX options for Linux. + Copyright (C) 1996, 1997, 1998, 1999 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifndef _POSIX_OPT_H +#define _POSIX_OPT_H 1 + +/* Job control is supported. */ +#define _POSIX_JOB_CONTROL 1 + +/* Processes have a saved set-user-ID and a saved set-group-ID. */ +#define _POSIX_SAVED_IDS 1 + +/* Priority scheduling is supported. */ +#define _POSIX_PRIORITY_SCHEDULING 1 + +/* Synchronizing file data is supported. */ +#define _POSIX_SYNCHRONIZED_IO 1 + +/* The fsync function is present. */ +#define _POSIX_FSYNC 1 + +/* Mapping of files to memory is supported. */ +#define _POSIX_MAPPED_FILES 1 + +/* Locking of all memory is supported. */ +#define _POSIX_MEMLOCK 1 + +/* Locking of ranges of memory is supported. */ +#define _POSIX_MEMLOCK_RANGE 1 + +/* Setting of memory protections is supported. */ +#define _POSIX_MEMORY_PROTECTION 1 + +/* Implementation supports `poll' function. */ +#define _POSIX_POLL 1 + +/* Implementation supports `select' and `pselect' functions. */ +#define _POSIX_SELECT 1 + +/* Only root can change owner of file. */ +#define _POSIX_CHOWN_RESTRICTED 1 + +/* `c_cc' member of 'struct termios' structure can be disabled by + using the value _POSIX_VDISABLE. */ +#define _POSIX_VDISABLE '\0' + +/* Filenames are not silently truncated. */ +#define _POSIX_NO_TRUNC 1 + +/* X/Open realtime support is available. */ +#define _XOPEN_REALTIME 1 + +/* X/Open realtime thread support is available. */ +#define _XOPEN_REALTIME_THREADS 1 + +/* XPG4.2 shared memory is supported. */ +#define _XOPEN_SHM 1 + +/* Tell we have POSIX threads. */ +#define _POSIX_THREADS 1 + +/* We have the reentrant functions described in POSIX. */ +#define _POSIX_REENTRANT_FUNCTIONS 1 +#define _POSIX_THREAD_SAFE_FUNCTIONS 1 + +/* We provide priority scheduling for threads. */ +#define _POSIX_THREAD_PRIORITY_SCHEDULING 1 + +/* We support user-defined stack sizes. */ +#define _POSIX_THREAD_ATTR_STACKSIZE 1 + +/* We support user-defined stacks. */ +#define _POSIX_THREAD_ATTR_STACKADDR 1 + +/* We support POSIX.1b semaphores, but only the non-shared form for now. */ +/*#define _POSIX_SEMAPHORES 1 XXX We are not quite there now. */ + +/* Real-time signals are supported. */ +#define _POSIX_REALTIME_SIGNALS 1 + +/* We support asynchronous I/O. */ +#define _POSIX_ASYNCHRONOUS_IO 1 +/* Alternative name for Unix98. */ +#define _LFS_ASYNCHRONOUS_IO 1 + +/* The LFS support in asynchronous I/O is also available. */ +#define _LFS64_ASYNCHRONOUS_IO 1 + +/* The rest of the LFS is also available. */ +#define _LFS_LARGEFILE 1 +#define _LFS64_LARGEFILE 1 +#define _LFS64_STDIO 1 + +#endif /* posix_opt.h */ diff --git a/libpthread/linuxthreads/sysdeps/unix/sysv/linux/bits/sigthread.h b/libpthread/linuxthreads/sysdeps/unix/sysv/linux/bits/sigthread.h new file mode 100644 index 000000000..c9b1dcf1b --- /dev/null +++ b/libpthread/linuxthreads/sysdeps/unix/sysv/linux/bits/sigthread.h @@ -0,0 +1,37 @@ +/* Signal handling function for threaded programs. + Copyright (C) 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifndef _BITS_SIGTHREAD_H +#define _BITS_SIGTHREAD_H 1 + +#if !defined _SIGNAL_H && !defined _PTHREAD_H +# error "Never include this file directly. Use <pthread.h> instead" +#endif + +/* Functions for handling signals. */ + +/* Modify the signal mask for the calling thread. The arguments have + the same meaning as for sigprocmask(2). */ +extern int pthread_sigmask __P ((int __how, __const __sigset_t *__newmask, + __sigset_t *__oldmask)); + +/* Send signal SIGNO to the given thread. */ +extern int pthread_kill __P ((pthread_t __thread, int __signo)); + +#endif /* bits/sigthread.h */ diff --git a/libpthread/linuxthreads/testrtsig.h b/libpthread/linuxthreads/testrtsig.h new file mode 100644 index 000000000..fb8b011f0 --- /dev/null +++ b/libpthread/linuxthreads/testrtsig.h @@ -0,0 +1,40 @@ +/* Test whether RT signals are really available. + Copyright (C) 1997 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Ulrich Drepper <drepper@cygnus.com>, 1997. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#include <limits.h> +#include <string.h> +#include <sys/utsname.h> + +static int +kernel_has_rtsig (void) +{ + return 0; /* hacked to test old uClibc that doesn't know about RT signals. 0.9.9 should work with RT signals. Make this proper in the end! */ +#ifdef RTSIG_MAX + return 1; +#else + return 0; +#endif + +/* + struct utsname name; + + return uname (&name) == 0 && __strverscmp (name.release, "2.1.70") >= 0; +*/ +} diff --git a/libpthread/linuxthreads/weaks.c b/libpthread/linuxthreads/weaks.c new file mode 100644 index 000000000..5281a2a5b --- /dev/null +++ b/libpthread/linuxthreads/weaks.c @@ -0,0 +1,120 @@ +/* The weak pthread functions for Linux. + Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#include <errno.h> +#include <limits.h> +#include <stdlib.h> + +extern int __pthread_return_0 __P ((void)); +extern int __pthread_return_1 __P ((void)); +extern void __pthread_return_void __P ((void)); + +/* NOTE: these require a strong alias in attr.c. I don't think we need this + * versioning stuff with uClibc. -StS + */ + +/* Those are pthread functions which return 0 if successful. */ +//#if defined HAVE_ELF && defined PIC && defined DO_VERSIONING +//weak_alias (__pthread_return_0, __libc_pthread_attr_init_2_0) +//symbol_version (__libc_pthread_attr_init_2_0, pthread_attr_init, GLIBC_2.0); +//weak_alias (__pthread_return_0, __libc_pthread_attr_init_2_1) +//default_symbol_version (__libc_pthread_attr_init_2_1, pthread_attr_init, +// GLIBC_2.1); +//#else +weak_alias (__pthread_return_0, pthread_attr_init) +//#endif +weak_alias (__pthread_return_0, pthread_attr_destroy) +weak_alias (__pthread_return_0, pthread_attr_setdetachstate) +weak_alias (__pthread_return_0, pthread_attr_getdetachstate) +weak_alias (__pthread_return_0, pthread_attr_setschedparam) +weak_alias (__pthread_return_0, pthread_attr_getschedparam) +weak_alias (__pthread_return_0, pthread_attr_setschedpolicy) +weak_alias (__pthread_return_0, pthread_attr_getschedpolicy) +weak_alias (__pthread_return_0, pthread_attr_setinheritsched) +weak_alias (__pthread_return_0, pthread_attr_getinheritsched) +weak_alias (__pthread_return_0, pthread_attr_setscope) +weak_alias (__pthread_return_0, pthread_attr_getscope) +weak_alias (__pthread_return_0, pthread_attr_setstackaddr) +weak_alias (__pthread_return_0, pthread_attr_getstackaddr) +weak_alias (__pthread_return_0, pthread_attr_setstacksize) +weak_alias (__pthread_return_0, pthread_attr_getstacksize) +weak_alias (__pthread_return_0, pthread_mutex_init) +weak_alias (__pthread_return_0, pthread_mutex_destroy) +weak_alias (__pthread_return_0, pthread_mutex_lock) +weak_alias (__pthread_return_0, pthread_mutex_trylock) +weak_alias (__pthread_return_0, pthread_mutex_unlock) +weak_alias (__pthread_return_0, pthread_mutexattr_init) +weak_alias (__pthread_return_0, pthread_mutexattr_destroy) +weak_alias (__pthread_return_0, pthread_mutexattr_settype) +weak_alias (__pthread_return_0, pthread_mutexattr_gettype) +weak_alias (__pthread_return_0, pthread_condattr_init) +weak_alias (__pthread_return_0, pthread_condattr_destroy) +weak_alias (__pthread_return_0, pthread_setschedparam) +weak_alias (__pthread_return_0, pthread_getschedparam) +weak_alias (__pthread_return_0, pthread_getcancelstate) +weak_alias (__pthread_return_0, pthread_setcancelstate) +weak_alias (__pthread_return_0, pthread_setcanceltype) +weak_alias (__pthread_return_0, pthread_setconcurrency) +weak_alias (__pthread_return_0, pthread_getconcurrency) +weak_alias (__pthread_return_0, pthread_self) +weak_alias (__pthread_return_0, pthread_cond_init) +weak_alias (__pthread_return_0, pthread_cond_destroy) +weak_alias (__pthread_return_0, pthread_cond_wait) +weak_alias (__pthread_return_0, pthread_cond_timedwait) +weak_alias (__pthread_return_0, pthread_cond_signal) +weak_alias (__pthread_return_0, pthread_cond_broadcast) +weak_alias (__pthread_return_0, pthread_rwlock_init) +weak_alias (__pthread_return_0, pthread_rwlock_destroy) +weak_alias (__pthread_return_0, pthread_rwlock_rdlock) +weak_alias (__pthread_return_0, pthread_rwlock_wrlock) +weak_alias (__pthread_return_0, pthread_rwlock_tryrdlock) +weak_alias (__pthread_return_0, pthread_rwlock_trywrlock) +weak_alias (__pthread_return_0, pthread_rwlock_unlock) +weak_alias (__pthread_return_0, pthread_rwlockattr_init) +weak_alias (__pthread_return_0, pthread_rwlockattr_destroy) +weak_alias (__pthread_return_0, pthread_rwlockattr_setpshared) +weak_alias (__pthread_return_0, pthread_rwlockattr_getpshared) + + +/* Those are pthread functions which return 1 if successful. */ +weak_alias (__pthread_return_1, pthread_equal) + +/* pthread_exit () is a special case. */ +void weak_function +pthread_exit (void *retval) +{ + exit (EXIT_SUCCESS); +} + +int +__pthread_return_0 (void) +{ + return 0; +} + +int +__pthread_return_1 (void) +{ + return 1; +} + +void +__pthread_return_void (void) +{ +} diff --git a/libpthread/linuxthreads/wrapsyscall.c b/libpthread/linuxthreads/wrapsyscall.c new file mode 100644 index 000000000..3b9ada160 --- /dev/null +++ b/libpthread/linuxthreads/wrapsyscall.c @@ -0,0 +1,178 @@ +/* Wrapper arpund system calls to provide cancelation points. + Copyright (C) 1996, 1997, 1998 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Ulrich Drepper <drepper@cygnus.com>, 1996. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#define __FORCE_GLIBC +#include <features.h> +#include <fcntl.h> +#include <sys/mman.h> +#include <pthread.h> +#include <unistd.h> +#include <stdarg.h> +#include <stddef.h> +#include <stdlib.h> +#include <termios.h> +#include <sys/resource.h> +#include <sys/wait.h> +#include <sys/socket.h> + + +#ifndef PIC +/* We need a hook to force this file to be linked in when static + libpthread is used. */ +const int __pthread_provide_wrappers = 0; +#endif + + +#define CANCELABLE_SYSCALL(res_type, name, param_list, params) \ +res_type __libc_##name param_list; \ +res_type \ +name param_list \ +{ \ + res_type result; \ + int oldtype; \ + pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, &oldtype); \ + result = __libc_##name params; \ + pthread_setcanceltype (oldtype, NULL); \ + return result; \ +} + +#define CANCELABLE_SYSCALL_VA(res_type, name, param_list, params, last_arg) \ +res_type __libc_##name param_list; \ +res_type \ +name param_list \ +{ \ + res_type result; \ + int oldtype; \ + va_list ap; \ + pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, &oldtype); \ + va_start (ap, last_arg); \ + result = __libc_##name params; \ + va_end (ap); \ + pthread_setcanceltype (oldtype, NULL); \ + return result; \ +} + + +/* close(2). */ +CANCELABLE_SYSCALL (int, close, (int fd), (fd)) + + +/* fcntl(2). */ +CANCELABLE_SYSCALL_VA (int, fcntl, (int fd, int cmd, ...), + (fd, cmd, va_arg (ap, long int)), cmd) + + +/* fsync(2). */ +CANCELABLE_SYSCALL (int, fsync, (int fd), (fd)) + + +/* lseek(2). */ +CANCELABLE_SYSCALL (off_t, lseek, (int fd, off_t offset, int whence), + (fd, offset, whence)) + + +/* msync(2). */ +/* This syscall not implemented in uClibc +CANCELABLE_SYSCALL (int, msync, (__ptr_t addr, size_t length, int flags), + (addr, length, flags)) +*/ + +/* nanosleep(2). */ +CANCELABLE_SYSCALL (int, nanosleep, (const struct timespec *requested_time, + struct timespec *remaining), + (requested_time, remaining)) + + +/* open(2). */ +CANCELABLE_SYSCALL_VA (int, open, (const char *pathname, int flags, ...), + (pathname, flags, va_arg (ap, mode_t)), flags) + + +/* pause(2). */ +CANCELABLE_SYSCALL (int, pause, (void), ()) + + +/* read(2). */ +CANCELABLE_SYSCALL (ssize_t, read, (int fd, void *buf, size_t count), + (fd, buf, count)) + + +/* system(3). */ +CANCELABLE_SYSCALL (int, system, (const char *line), (line)) + +/* tcdrain(2). */ +CANCELABLE_SYSCALL (int, tcdrain, (int fd), (fd)) + +/* wait(2). */ +CANCELABLE_SYSCALL (__pid_t, wait, (__WAIT_STATUS_DEFN stat_loc), (stat_loc)) + + +/* waitpid(2). */ +CANCELABLE_SYSCALL (__pid_t, waitpid, (__pid_t pid, int *stat_loc, + int options), + (pid, stat_loc, options)) + + +/* write(2). */ +CANCELABLE_SYSCALL (ssize_t, write, (int fd, const void *buf, size_t n), + (fd, buf, n)) + + +/* The following system calls are thread cancellation points specified + in XNS. */ + +/* accept(2). */ +CANCELABLE_SYSCALL (int, accept, (int fd, __SOCKADDR_ARG addr, + socklen_t *addr_len), + (fd, addr, addr_len)) + +/* connect(2). */ +CANCELABLE_SYSCALL (int, connect, (int fd, __CONST_SOCKADDR_ARG addr, + socklen_t len), + (fd, addr, len)) + +/* recv(2). */ +CANCELABLE_SYSCALL (int, recv, (int fd, __ptr_t buf, size_t n, int flags), + (fd, buf, n, flags)) + +/* recvfrom(2). */ +CANCELABLE_SYSCALL (int, recvfrom, (int fd, __ptr_t buf, size_t n, int flags, + __SOCKADDR_ARG addr, socklen_t *addr_len), + (fd, buf, n, flags, addr, addr_len)) + +/* recvmsg(2). */ +CANCELABLE_SYSCALL (int, recvmsg, (int fd, struct msghdr *message, int flags), + (fd, message, flags)) + +/* send(2). */ +CANCELABLE_SYSCALL (int, send, (int fd, const __ptr_t buf, size_t n, + int flags), + (fd, buf, n, flags)) + +/* sendmsg(2). */ +CANCELABLE_SYSCALL (int, sendmsg, (int fd, const struct msghdr *message, + int flags), + (fd, message, flags)) + +/* sendto(2). */ +CANCELABLE_SYSCALL (int, sendto, (int fd, const __ptr_t buf, size_t n, + int flags, __CONST_SOCKADDR_ARG addr, + socklen_t addr_len), + (fd, buf, n, flags, addr, addr_len)) |