| Age | Commit message (Collapse) | Author |
|---|
|
Avoid calling select with empty sets which hangs the process
This makes uClibc-ng act like glibc and musl
Without this fix the test_poll of python3 testsuite hangs forever
Scenario of the issue:
If you call poll with only invalid file descriptors, like in python3
testsuite
(https://github.com/python/cpython/blob/master/Lib/test/test_poll.py#L83)
You will go through uClibc poll emulation code, which is based on
select syscall.
Your first call to select will fail, it will return -1 and errno will be
set to EBADF: https://github.com/wbx-github/uclibc-ng/blob/master/libc/sysdeps/linux/common/poll.c#L120
Then you will go through the for loop which tests individually each file descriptor by calling
select on each one: https://github.com/wbx-github/uclibc-ng/blob/master/libc/sysdeps/linux/common/poll.c#L163
each call will also return -1 with errno being equal to EBADF.
Therefore all pollfd will have the POLLNVAL flag in their respective revents field.
And, the most important, rset/wset/xset will stay empty.
Then the for loop ends, the "continue" makes the while loop run again.
The following select() is run again: https://github.com/wbx-github/uclibc-ng/blob/master/libc/sysdeps/linux/common/poll.c#L120
But this time the sets are empty.
If the poll was called with timeout set to -1, this select will hang forever because there is no timeout
and the sets are empty so no event will ever wake it up.
test program:
int main(void)
{
struct pollfd pfd;
int ret;
int pipe_fds[2];
pipe(pipe_fds);
close(pipe_fds[0]);
close(pipe_fds[1]);
pfd.fd = pipe_fds[0];
pfd.events = POLLIN | POLLOUT | POLLPRI;
pfd.revents = 0;
ret = poll(&pfd, 1, -1);
printf("ret: %d\n", ret);
if (ret < 0)
printf("error: %s", strerror(errno));
else {
puts("revents: ");
if (pfd.revents & POLLERR)
printf(" POLLERR");
if (pfd.revents & POLLHUP)
printf(" POLLHUP");
if (pfd.revents & POLLNVAL)
printf(" POLLNVAL");
puts("");
}
return 0;
}
This hangs on uClibc-ng aarch64 and Kalray's arch (kv3) but does the following on musl and glibc:
"
ret: 1
revents:
POLLNVAL
"
strace output of this program with uClibc *without* the patch applied:
pselect6(4, [3], [3], [3], NULL, NULL) = -1 EBADF (Bad file descriptor)
pselect6(4, [3], [3], [3], {tv_sec=0, tv_nsec=0}, NULL) = -1 EBADF (Bad file descriptor)
pselect6(0, 0x7ffffffb80, 0x7ffffffb68, 0x7ffffffb50, NULL, NULL
(never finishes)
strace output of this program with uClibc *with* the patch applied:
pselect6(4, [3], [3], [3], NULL, NULL) = -1 EBADF (Bad file descriptor)
pselect6(4, [3], [3], [3], {tv_sec=0, tv_nsec=0}, NULL) = -1 EBADF (Bad file descriptor)
write(1, "ret: 1\n", 7ret: 1
) = 7
write(1, "revents: \n", 10revents:
) = 10
write(1, " POLLNVAL\n", 10 POLLNVAL
) = 10
exit_group(0) = ?
+++ exited with 0 +++
|
|
Hi,
This diff fixes a typo in the PTRACE_EVENT_SECCOMP event code.
The typo itself was introduced in 2012 when syncing with glibc header
files and was itself fixed in 2013 in the glibc headers.
|
|
basically from or1k port of uClibc-ng, with fixes for structures in
pthreadtypes.h from 64 bit architectures.
18 testsuite failures counted.
|
|
|
|
|
|
|
|
The return type of syscall() is long so __syscall_error, which is jumped
to by syscall handlers to stash an error number into errno, must return
long too otherwhise it returs 4294967295L instead of -1L. For example,
syscall for x86_64 is defined in libc/sysdeps/linux/x86_64/syscall.S as
syscall:
movq %rdi, %rax /* Syscall number -> rax. */
movq %rsi, %rdi /* shift arg1 - arg5. */
movq %rdx, %rsi
movq %rcx, %rdx
movq %r8, %r10
movq %r9, %r8
movq 8(%rsp),%r9 /* arg6 is on the stack. */
syscall /* Do the system call. */
cmpq $-4095, %rax /* Check %rax for error. */
jae __syscall_error /* Branch forward if it failed. */
ret /* Return to caller. */
In libc/sysdeps/linux/x86_64/__syscall_error.c, __syscall_error is
defined as
int __syscall_error(void) attribute_hidden;
int __syscall_error(void)
{
register int err_no __asm__ ("%rcx");
__asm__ ("mov %rax, %rcx\n\t"
"neg %rcx");
__set_errno(err_no);
return -1;
}
So __syscall_error returns -1 as a 32-bit int in a 64-bit register, %rax
(0x00000000ffffffff, whose decimal value is decimal 4294967295) and a
test like this always returns false:
if (syscall(number, ...) == -1)
foo();
Fix the error by making __syscall_error return a long, like syscall().
The problem can be circumvented by the caller by coercing the returned
value to int before comparing it to -1:
if ((int) syscall(number, ...) == -1)
foo();
The same problem probably occurs on other 64-bit systems but so far only
x86_64 was tested, so this change must be considered experimental.
Signed-off-by: Carlos Santos <unixmania@gmail.com>
|
|
(gdb) print offsetof(struct __jmp_buf_tag, __mask_was_saved)
$12 = (int *) 0x1f0
using https://stackoverflow.com/a/39663128/2171120
|
|
Update from linux/arch/sparc/include/uapi/asm/uctx.h
|
|
|
|
|
|
This symbol was causing a build failure with the new toolchain. It
looks like it has always been wrong.
The main issue was checking for PIC rather than __PIC__.
Remove all PSEUDO_* macros and ther SYSCALL_ERROR_NAME macro as they are
not needed by uclibc-ng, they are used in glibc for building up syscalls
there, but not here.
Fixes error:
/opt/shorne/software/or1k-linux/bin/../lib/gcc/or1k-linux/9.0.1/../../../../or1k-linux/bin/ld: libc/libc_so.a(or1k_clone.os): pc-relative relocation against dynamic symbol __syscall_error
/opt/shorne/software/or1k-linux/bin/../lib/gcc/or1k-linux/9.0.1/../../../../or1k-linux/bin/ld: final link failed: bad value
Signed-off-by: Stafford Horne <shorne@gmail.com>
|
|
This patch fixes the incorrect guard by __USE_MISC of struct winsize and
struct termio in powerpc termios header. Current states leads to build
failures if the program defines _XOPEN_SOURCE, but not _DEFAULT_SOURCE
or either _BSD_SOURCE or _SVID_SOURCE. Without any definition,
__USE_MISC will not be defined and neither the struct definitions.
This patch copies the default Linux ioctl-types.h by adjusting only the
character control field (c_cc) size in struct termio.
Signed-off-by: Vadim Kochan <vadim4j@gmail.com>
[Vadim: adopted for uclibc ]
|
|
We don't support shared libraries and thus _init/_fini. But loading
nommu binaries blows they aren't cleared, so do that.
Signed-off-by: Christoph Hellwig <hch@lst.de>
|
|
This avoids a nommu build failure.
Signed-off-by: Christoph Hellwig <hch@lst.de>
|
|
uClibc-ng don't build with gcc 9.1 [1] due to a new check that
"catch illegal asm constraint usage" [2].
gcc 9.1 print this error:
"invalid hard register usage between earlyclobber operand and input operand"
The asm constraint is present in uClibc since it support sparc (back in 2002)[3].
Note: There is no such constraint is Glibc counterpart code [4].
[1] https://gitlab.com/kubu93/toolchains-builder/-/jobs/205435757
[2] https://github.com/gcc-mirror/gcc/commit/b782636f28f5c378897c238081d28d7a4a6ca578
[3] https://cgit.uclibc-ng.org/cgi/cgit/uclibc-ng.git/commit/?id=3b6d086531102b6d09ce852feb1e370d5dca3ce9
[4]
+https://sourceware.org/git/?p=glibc.git;a=blob;f=sysdeps/unix/sysv/linux/sparc/sysdep.h;h=981b2a26b7a91093f821c97876
+e55bc4be2d9f8a;hb=HEAD
|
|
preadv/pwritev don't provide separate version for 64-bit wide off_t,
and default to 32-bit wide off_t, which results in a mismatch between
declaration and definition for user programs built with
-D_FILE_OFFSET_BITS=64.
Make offset argument of both functions __off64_t.
This fixes test misc/tst-preadvwritev on xtensa.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
|
|
Definitions of POLLWRNORM, POLLWRBAND and POLLREMOVE in xtensa linux
kernel are non-standard. Provide bits/poll.h with correct values for
these constants.
This fixes the following strace build errors:
In file included from xlat/pollflags.h:4:0,
from poll.c:34:
./static_assert.h:40:24: error: static assertion failed: "POLLWRBAND != 0x0100"
# define static_assert _Static_assert
^
xlat/pollflags.h:75:1: note: in expansion of macro ‘static_assert’
static_assert((POLLWRBAND) == (0x0100), "POLLWRBAND != 0x0100");
^~~~~~~~~~~~~
./static_assert.h:40:24: error: static assertion failed: "POLLREMOVE != 0x0800"
# define static_assert _Static_assert
^
xlat/pollflags.h:117:1: note: in expansion of macro ‘static_assert’
static_assert((POLLREMOVE) == (0x0800), "POLLREMOVE != 0x0800");
^~~~~~~~~~~~~
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
|
|
I've got several patches to fix ltp/openmp/uclibc-ng-test testcase fail on
c-sky.
- fix a ltp testcase.
- fix the problem that pthread creat will fail when libomp is linked before
libc, the variable pagesize is not init.
- fix tst-cancel4 and tst-cancel16. tst-cancelx4 and tst-cancelx16 still fail
with this patch applied, cleanup handler is not called for open/creat/fcntl,
seems some thing wrong with unwind, I haven't check the rootcause yet.
|
|
Fix issues with aarch64 and df with mismatching header between kernel
and libc.
|
|
On Linux/MIPS (O32 ABI) for system call we have two result registers - v0 and a3.
v0 contains actual syscall result on success or error number on fail, a3 set to 0/1
for indicating syscall success/fail. (if a3 == 1, v0 contains errno).
Now as we can see from definition of handle_sys (arch/mips/kernel/scall32-o32.S),
handler treats returned by syscall function (let's call "original") values in
range [-EMAXERRNO; 0[ as -errno, a3 is set to 1 and final returned (to userspace)
value is (-original).
INLINE_SYSCALL_NOERR_NCS defined in mips/bits/syscalls.h will handle
this behaviour.
Signed-off-by: Volodymyr Boyko <boyko.cxx@gmail.com>
|
|
* ldso/ldso/arm/aeabi_read_tp.S: Add Thumb version.
* ldso/ldso/arm/dl-startup.h: Do not force ARM encoding, adjust
for Thumb.
* ldso/ldso/arm/resolve.S: Force Thumb encoding on Thumb-only
processors.
* libc/sysdeps/linux/arm/crt1.S: Do not force ARM encoding, adjust
for Thumb.
Signed-off-by: Mickaël Guêné <mickael.guene@st.com>
Signed-off-by: Christophe Lyon <christophe.lyon@st.com>
|
|
Added in kernel in kernel 3.10 in
1ff3c9677bff ("timekeeping: Add CLOCK_TAI clockid")
NOTE: CLOCK_SGI_CYCLE was not added, as it has been lately removed.
Signed-off-by: Petr Vorel <petr.vorel@gmail.com>
|
|
|
|
Signed-off-by: Guo Ren <ren_guo@c-sky.com>
|
|
Call getpid() in INTERNAL_SYSCALL will break the argument regs,
because gcc couldn't save destoryed regs for system call asm.
Ref to glibc, we could just remove all the check code.
Signed-off-by: Guo Ren <ren_guo@c-sky.com>
|
|
Change clone.S so that in FDPIC case we take into account the fact
that we are given a function descriptor.
* libc/sysdeps/linux/arm/clone.S (__clone): Support __FDPIC__.
Signed-off-by: Mickaël Guêné <mickael.guene@st.com>
Signed-off-by: Christophe Lyon <christophe.lyon@st.com>
|
|
Implements __gnu_Unwind_Find_got(), which is called from libgcc while
unwinding.
* libc/sysdeps/linux/arm/Makefile.arch (CSRC): Add find._got.c.
* libc/sysdeps/linux/arm/find_got.c: New file.
Signed-off-by: Mickaël Guêné <mickael.guene@st.com>
Signed-off-by: Christophe Lyon <christophe.lyon@st.com>
|
|
* libc/sysdeps/linux/arm/crt1.S: Define _start for __FDPIC__.
Signed-off-by: Mickaël Guêné <mickael.guene@st.com>
Signed-off-by: Christophe Lyon <christophe.lyon@st.com>
|
|
Add FDPIC dynamic relocations support, similar to what other FDPIC
targets do.
Lazy binding is implemented in a folllow-up patch.
Disable the SEND* macros because they involve relocations to
access constant strings that are unsupported by the existing
arm version.
Define DL_START, START, ARCH_NEEDS_BOOTSTRAP_RELOCS,
DL_CHECK_LIB_TYPE similarly to what other FDPIC targets do.
Define raise() because _dl_find_hash references __aeabi_uidivmod,
which uses __aeabi_idiv0 which in turn references raise.
* include/elf.h (R_ARM_FUNCDESC): Define.
(R_ARM_FUNCDESC_VALUE): Define.
* ldso/include/dl-string.h (SEND_STDERR, SEND_ADDRESS_STDERR)
(SEND_NUMBER_STDERR): Define empty for __FDPIC__.
* ldso/ldso/arm/dl-inlines.h: New file.
* ldso/ldso/arm/dl-startup.h (PERFORM_BOOTSTRAP_RELOC): Fix type
of load_addr. Fix handling of R_ARM_RELATIVE, add support for
R_ARM_FUNCDESC_VALUE.
(DL_START, START): Define for __FDPIC__.
(raise): Define.
* ldso/ldso/arm/dl-sysdep.h (ARCH_NEEDS_BOOTSTRAP_RELOCS): Define.
(DL_CHECK_LIB_TYPE): Define.
(elf_machine_type_class): Take into account FDPIC related
relocations.
(elf_machine_load_address): Support __FDPIC__.
(elf_machine_relative): Likewise.
* ldso/ldso/arm/elfinterp.c (_dl_linux_resolver): Dummy support
for __FDPIC__, implemented in a later patch.
(_dl_do_reloc): Fix reloc_adr computation for __FDPIC__, fix
handling of local symbols. Fix handling of R_ARM_RELATIVE, add
support for R_ARM_FUNCDESC_VALUE, R_ARM_FUNCDESC.
* ldso/ldso/arm/resolve.S: Make _dl_linux_resolve hidden.
* ldso/ldso/fdpic/dl-inlines.h (htab_delete): Declare.
* libc/sysdeps/linux/arm/bits/elf-fdpic.h: New file, similar to bfin's.
* libc/sysdeps/linux/arm/crtreloc.c: Likewise.
* libc/sysdeps/linux/arm/find_exidx.c (__dl_addr_in_loadaddr) Define.
(find_exidx_callback): Support __FDPIC__.
Signed-off-by: Mickaël Guêné <mickael.guene@st.com>
Signed-off-by: Christophe Lyon <christophe.lyon@st.com>
|
|
Only a simple hello world is tested in qemu system emulation.
|
|
In ltp testcase sendfile08.c, it use offset=NULL to test the api.
PATCH V2:
fixup the stupid missing check in the end. Sorry for lose test.
See "man sendfile" and it really support offset is NULL.
Signed-off-by: Guo Ren <ren_guo@c-sky.com>
|
|
This reverts commit b00fd230ed0b49b9f23d829ad5d09859f34bb754.
|
|
Ref the implement from the glibc and high=0 seems so bad.
Signed-off-by: Guo Ren <ren_guo@c-sky.com>
|
|
In ltp testcase sendfile08.c, it use offset=NULL to test the api.
See "man sendfile" and it really support offset is NULL.
Signed-off-by: Guo Ren <ren_guo@c-sky.com>
|
|
Those definitions exist on all other architectures, but were not
present in or1k specific headers when or1k support was merged.
On the kernel side, their support is completely architecture
independent, so we just need those definitions to make those fcntl()
calls available on or1k.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
|
|
Otherwise it breaks mips64 n64.
Should be used for aarch64/tilegx only.
|
|
Sync clone()/vfork() with GNU C Library.
Allow CFI and fix strsignal() / tcsetattr().
|
|
Signed-off-by: Guo Ren <ren_guo@c-sky.com>
|
|
This adds basic support for tile architecture.
Only static binaries, no ld.so or threading support.
Tested with qemu-tilegx only.
|
|
From glibc commit 9aa5c222b9e0409143410a02b6364a3b25dbf028
[BZ #22919]
|
|
|
|
Newer iproute2 package make use of some of defines.
Sync missing defines with GNU C library.
|
|
... so don't declare it. Otherwise, eudev finds it declared and tries
to link with it - and fails.
Signed-off-by: Alexey Neyman <stilor@att.net>
|
|
|
|
Signed-off-by: Guo Ren <ren_guo@c-sky.com>
|
|
Sync with GNU C library and consolidate duplicate non
architecture specific defines.
MAP_UNINITIALIZED is only defined to 0x4000000 and used by
the Linux kernel when CONFIG_MMAP_ALLOW_UNINITIALIZED is enabled.
CONFIG_MMAP_ALLOW_UNINITIALIZED is only available for nommu.
See Documentation/nommu-mmap.txt.
|
|
Fixes a buildroot build error with ffmpeg
In file included from /home/bernd/buildroot/output/host/aarch64-buildroot-linux-uclibc/sysroot/usr/include/signal.h:329:0,
from fftools/ffmpeg.h:26,
from fftools/ffmpeg_opt.c:23:
/home/bernd/buildroot/output/host/aarch64-buildroot-linux-uclibc/sysroot/usr/include/sys/ucontext.h:52:16:
error: field 'uc_mcontext' has incomplete type
mcontext_t uc_mcontext;
^~~~~~~~~~~
using this defconfig:
BR2_aarch64=y
BR2_PACKAGE_FFMPEG=y
sys/ucontext.h for other archs already include bits/sigcontext.h,
on aarch64 this is needed as well.
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
|
|
The fallback code is not used as all supported kernels have the
syscall. Check if a absolute path is returned from syscall.
|
|
It seems there is no real dependency to NPTL for these clock_*
functions when UCLIBC_ADVANCED_REALTIME is enabled.
No regressions found.
Reported-by: Baruch Siach <baruch@tkos.co.il>
|
