Sort tunables list at the time it's generated. Note: adding new
tunables will cause other tunable IDs to change, but that was
the case before anyway. POSIX does not guarantee the order of "foo
in bar" AWK operators, so the order was indeterminate before anyway.
Even depending on the order to be the same across multiple calls,
such as in this script, is undefined, so sorting the list resolves
that also.
Note that sorting is not dependent on the user's locale.
The __stack_prot is used by Linux to make the stack executable if
a modules requires it. It is also marked as RELRO, which requires
to change the segment permission to RW to update it.
Also, there is no need to keep track of the flags: either the stack
will have the default permission of the ABI or should be change to
PROT_READ | PROT_WRITE | PROT_EXEC. The only additional flag,
PROT_GROWSDOWN or PROT_GROWSUP, is Linux only and can be deducted
from _STACK_GROWS_DOWN/_STACK_GROWS_UP.
Also, the check_consistency function was already removed some time
ago.
Checked on x86_64-linux-gnu and i686-linux-gnu.
Reviewed-by: Florian Weimer <fweimer@redhat.com>
These comments were written in 2003 (added in 2c008571c3), predating
the addition of getdelim(3)/getline(3) in POSIX.1-2008.
Reviewed-by: Sam James <sam@gentoo.org>
Reviewed-by: Florian Weimer <fweimer@redhat.com>
On i386, set the default minimum ISA level to 0, not 1 (baseline which
includes SSE2). There are no changes in config.h nor in config.make on
x86-64. This fixes BZ #31867.
Signed-off-by: H.J. Lu <hjl.tools@gmail.com>
Tested-by: Ian Jordan <immoloism@gmail.com>
Reviewed-by: Sam James <sam@gentoo.org>
Reviewed-by: Florian Weimer <fweimer@redhat.com>
In commit 46b5e98ef6 ("x86: Add seperate non-temporal tunable for
memset") a tunable threshold for enabling non-temporal memset was added,
but only for Intel hardware.
Since that commit, new benchmark results suggest that non-temporal
memset is beneficial on AMD, as well, so allow this tunable to be set
for AMD.
See:
https://docs.google.com/spreadsheets/d/1opzukzvum4n6-RUVHTGddV6RjAEil4P2uMjjQGLbLcU/edit?usp=sharing
which has been updated to include data using different stategies for
large memset on AMD Zen2, Zen3, and Zen4.
Signed-off-by: Joe Damato <jdamato@fastly.com>
Reviewed-by: Noah Goldstein <goldstein.w.n@gmail.com>
The manpage says that when the passed size is zero, they should set the
expected size and return 0. ERANGE shall be returned only when the non-zero
passed size is not large enough.
When no translator is set, __file_get_translator would return EINVAL
which is a confusing value. Better check for a passive translation
before getting the value.
The error message in xgetsockname was incorrectly referring to a
different function. This commit fixes that.
Suggested-by: Arjun Shankar <arjun@redhat.com>
Signed-off-by: Avinal Kumar <avinal.xlvii@gmail.com>
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
These are required by the upcoming POSIX standard and are available on
other platforms.
Link: https://austingroupbugs.net/view.php?id=339
Signed-off-by: Mohamed Akram <mohd.akram@outlook.com>
Reviewed-by: Arjun Shankar <arjun@redhat.com>
As of Linux kernel 6.9, some ioctls and a parameters structure have been
introduced which allow user programs to control whether a particular
epoll context will busy poll.
Update the headers to include these for the convenience of user apps.
The ioctls were added in Linux kernel 6.9 commit 18e2bf0edf4dd
("eventpoll: Add epoll ioctl for epoll_params") [1] to
include/uapi/linux/eventpoll.h.
[1]: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/diff/?h=v6.9&id=18e2bf0edf4dd
Signed-off-by: Joe Damato <jdamato@fastly.com>
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
difftime can signal an inexact conversion when converting to double,
so it should not be marked as pure or nothrow (BZ 31808).
Although we could do something more complicated, in which difftime is
plain on modern platforms but const and nothrow on obsolescent
platforms with 32-bit time_t, it hardly seems worth the trouble.
difftime is used so rarely that it's not worth taking pains to
optimize calls to it on obsolescent platforms.
Reviewed-by: DJ Delorie <dj@redhat.com>
The test aims to ensure that malloc uses the alternate path to
allocate memory when sbrk() or brk() fails.To achieve this,
the test first creates an obstruction at current program break,
tests that obstruction with a failing sbrk(), then checks if malloc
is still returning a valid ptr thus inferring that malloc() used
mmap() instead of brk() or sbrk() to allocate the memory.
Reviewed-by: Arjun Shankar <arjun@redhat.com>
Reviewed-by: Zack Weinberg <zack@owlfolio.org>
Left shift of ki is undefined when ki<0, copy the logic from exp,
which uses unsigned arithmetics, to fix it.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
Deallocate the memory for the FILE structure when seeking to the end fails
in append mode.
Fixes: ea33158c96 ("Fix offset caching for streams and use it for ftell (BZ #16680)")
Linux 6.9 adds the ELF note type NT_ARM_FPMR. Add this to glibc's
elf.h, along with the previously missed NT_ARM_SSVE, NT_ARM_ZA and
NT_ARM_ZT (added in older kernel versions).
Tested for x86_64.
This has been confirmed to work around some interposed mallocs. Here
is a discussion of the impact test ust/libc-wrapper/test_libc-wrapper
in lttng-tools:
New TLS usage in libgcc_s.so.1, compatibility impact
<https://inbox.sourceware.org/libc-alpha/8734v1ieke.fsf@oldenburg.str.redhat.com/>
Reportedly, this patch also papers over a similar issue when tcmalloc
2.9.1 is not compiled with -ftls-model=initial-exec. Of course the
goal really should be to compile mallocs with the initial-exec TLS
model, but this commit appears to be a useful interim workaround.
Fixes commit d2123d6827 ("elf: Fix slow
tls access after dlopen [BZ #19924]").
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
The tuning for non-temporal stores for memset vs memcpy is not always
the same. This includes both the exact value and whether non-temporal
stores are profitable at all for a given arch.
This patch add `x86_memset_non_temporal_threshold`. Currently we
disable non-temporal stores for non Intel vendors as the only
benchmarks showing its benefit have been on Intel hardware.
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
Previously we use `rep stosb` for all medium/large memsets. This is
notably worse than non-temporal stores for large (above a
few MBs) memsets.
See:
https://docs.google.com/spreadsheets/d/1opzukzvum4n6-RUVHTGddV6RjAEil4P2uMjjQGLbLcU/edit?usp=sharing
For data using different stategies for large memset on ICX and SKX.
Using non-temporal stores can be up to 3x faster on ICX and 2x faster
on SKX. Historically, these numbers would not have been so good
because of the zero-over-zero writeback optimization that `rep stosb`
is able to do. But, the zero-over-zero writeback optimization has been
removed as a potential side-channel attack, so there is no longer any
good reason to only rely on `rep stosb` for large memsets. On the flip
size, non-temporal writes can avoid data in their RFO requests saving
memory bandwidth.
All of the other changes to the file are to re-organize the
code-blocks to maintain "good" alignment given the new code added in
the `L(stosb_local)` case.
The results from running the GLIBC memset benchmarks on TGL-client for
N=20 runs:
Geometric Mean across the suite New / Old EXEX256: 0.979
Geometric Mean across the suite New / Old EXEX512: 0.979
Geometric Mean across the suite New / Old AVX2 : 0.986
Geometric Mean across the suite New / Old SSE2 : 0.979
Most of the cases are essentially unchanged, this is mostly to show
that adding the non-temporal case didn't add any regressions to the
other cases.
The results on the memset-large benchmark suite on TGL-client for N=20
runs:
Geometric Mean across the suite New / Old EXEX256: 0.926
Geometric Mean across the suite New / Old EXEX512: 0.925
Geometric Mean across the suite New / Old AVX2 : 0.928
Geometric Mean across the suite New / Old SSE2 : 0.924
So roughly a 7.5% speedup. This is lower than what we see on servers
(likely because clients typically have faster single-core bandwidth so
saving bandwidth on RFOs is less impactful), but still advantageous.
Full test-suite passes on x86_64 w/ and w/o multiarch.
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
This new note type is defined at https://systemd.io/ELF_DLOPEN_METADATA/
and is used to list shared library dependencies loaded via dlopen().
Distro packagers can use this, via tools like those available at
https://github.com/systemd/package-notes to automatically generate
dependencies when building projects that make use of this
specification.
By defining the note id here we can use it in other projects as a
stable identifier, for example in 'readelf' to pretty-print its
content.
Signed-off-by: Luca Boccassi <bluca@debian.org>
Reviewed-by: Arjun Shankar <arjun@redhat.com>
Page was renamed some time ago, there's a redirect but better
to point to the right one
Signed-off-by: Luca Boccassi <bluca@debian.org>
Reviewed-by: Arjun Shankar <arjun@redhat.com>
A space is added before the left bracket of the x86_64 elf_machine_rela
function, in order to harmonize with the rest of the implementation of
the function and to make it easier to retrieve the function. The lines
where the function definition is located has been re-indented, as well
as its left curly bracket placed in the correct position.
Signed-off-by: Xin Wang <yw987194828@gmail.com>
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
Because difftime's behavior depends on the floating-point environment,
the function is pure, not const (BZ 31802).
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
pidfd_getpid.c has
/* Ignore invalid large values. */
if (INT_MULTIPLY_WRAPV (10, n, &n)
|| INT_ADD_WRAPV (n, *l++ - '0', &n))
return -1;
For GCC older than GCC 7, INT_ADD_WRAPV(a, b, r) is defined as
_GL_INT_OP_WRAPV (a, b, r, +, _GL_INT_ADD_RANGE_OVERFLOW)
and *l++ - '0' is evaluated twice. Fix BZ #31798 by moving "l++" out of
the if statement. Tested with GCC 6.4 and GCC 14.1.
Signed-off-by: H.J. Lu <hjl.tools@gmail.com>
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
This reverts commit 30a745450e.
On ppc64le, without <stdio.h>, vfscanf is used and with <stdio.h>
__isoc23_vfscanfieee128 is used. I am reverting this since it doesn't
work on all targets.
Add a test for fscanf of long double without including <stdio.h>.
Signed-off-by: H.J. Lu <hjl.tools@gmail.com>
Reviewed-by: Sunil K Pandey <skpgkp2@gmail.com>
The powerpc32 have an extra versionsort provided by LFS
versionsort64.o. It seems that 5226a81f55
used the wrong check to create the alias for the LFS to non-LFS version.
It should not matter for _DIRENT_MATCHES_DIRENT64 since both symbols
have the same implementation.
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
This patch updates the kernel version in the tests tst-mman-consts.py
and tst-mount-consts.py to 6.9. (There are no new constants covered
by these tests in 6.9 that need any other header changes;
tst-pidfd-consts.py was updated separately along with adding new
constants relevant to that test.)
Tested with build-many-glibcs.py.
The libc.a for alpha, s390, and sparcv9 does not provide
copysignf64x, copysignf128, frexpf64x, frexpf128, modff64x, and
modff128.
Checked with a static build for the affected ABIs.
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
Both the generic and POWER6 versions provide definitions of the
symbol, which are already provided by the ifunc resolver.
Checked on powerpc-linux-gnu-power4.
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
For powerpc64 the generic version provides a weak definition of
strchrnul, which are already provided by the ifunc resolver. The
powerpc32 version is slight different, where for static case there
is no iFUNC support.
The strncasecmp_l is provided ifunc resolver.
Checked on powerpc-linux-gnu-power4 and powerpc64-linux-gnu.
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
The generic version provides weak definitions of memchr/strlen,
which are already provided by the ifunc resolvers.
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
