Add x86_rep_movsb_threshold and x86_rep_stosb_threshold to tunables
to update thresholds for "rep movsb" and "rep stosb" at run-time.
Note that the user specified threshold for "rep movsb" smaller than
the minimum threshold will be ignored.
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
strcmp-avx2.S: In avx2 strncmp function, strings are compared in
chunks of 4 vector size(i.e. 32x4=128 byte for avx2). After first 4
vector size comparison, code must check whether it already passed
the given offset. This patch implement avx2 offset check condition
for strncmp function, if both string compare same for first 4 vector
size.
The existing macros are fragile and expect local variables with a
certain name. Fix this by defining them as functions with default
implementation in a new header dl-runtime.h which arches can override
if need be.
This came up during ARC port review, hence the need for argument pltgot
in reloc_index() which is not needed by existing ports.
This patch potentially only affects hppa/x86 ports,
build tested for both those configs and a few more.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
The getcpu cache was removed from the kernel in Linux 2.6.24. glibc
support from the sched_getcpu implementation was removed in commit
dd26c44403 ("Consolidate sched_getcpu").
Since __x86_shared_non_temporal_threshold is defined as
long int __x86_shared_non_temporal_threshold;
and long int is 4 bytes for x32, use RDX_LP to compare against
__x86_shared_non_temporal_threshold in assembly code.
This patch removes the IEEE_DOUBLE_BIG_ENDIAN and
IEEE_DOUBLE_MIXED_ENDIAN macros from gmp-impl.h and gmp-mparam.h, and
the ieee_double_extract union from gmp-impl.h. The macros were used
only in defining the union, which was used nowhere in glibc. As GMP's
gmp-impl.h is over 5000 lines, the file in glibc is so far from the
GMP version that it doesn't seem to make sense to keep things there
that are not relevant in glibc. (I expect there is plenty more in the
header after this patch that is also not relevant in glibc and can be
cleaned up later.)
Tested with build-many-glibcs.py that installed stripped shared
libraries are unchanged by this patch.
The __sfp_handle_exceptions is not fully correct regarding raising
exceptions, since there is no direct way to raise only FP_EX_OVERFLOW
nor FP_EX_UNDERFLOW for SSE mode. Both libgcc and feraiseexcept rely
on x87 mode to accomplish it.
This reverts commit 460ee50de0.
Checked on x86_64.
The exported x86_64 fenv.h functions operate on both i387 and SSE (since
they should work on both float, double, and long double) while the
internal libc_fe* set either SSE (float, double, and float128) or
i387 (long double).
The libgcc __sfp_handle_exceptions (used on float128 implementation),
however, will set either SEE or i387 exception depending of the
exception to raise. This broke the internal assumption of float128
where only SSE operations will be used.
This patch reimplements the libgcc __sfp_handle_exceptions to use only
SSE operations and sets libgcc to use it instead of its own
implementation.
And I think we should fix libgcc in a similar manner, since checking on
config/i386/64/sfp-machine.h it already only supports SSE rounding mode
and x86_64 ABI also expectes float128 to use SSE registers [1]
(although it is not clear on how future implementation might implement
it).
Checked on x86_64-linux-gnu.
[1] https://github.com/hjl-tools/x86-psABI/wiki/X86-psABI
All cancellable syscalls are done by C implementations, so there is no
no need to use a specialized implementation to optimize register usage.
It fixes BZ #25765.
Checked on x86_64-linux-gnu.
The corner cases included were generated using exhaustive search
for all float/binary32 values on x86_64 (comparing to MPFR for
correct rounding to nearest).
For the j0/j1/y0 functions, only cases with ulp error <= 9 were
included.
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
With mathinline removal there is no need to keep building and testing
inline math tests.
The gen-libm-tests.py support to generate ULP_I_* is removed and all
libm-test-ulps files are updated to longer have the
i{float,double,ldouble} entries. The support for no-test-inline is
also removed from both gen-auto-libm-tests and the
auto-libm-test-out-* were regenerated.
Checked on x86_64-linux-gnu and i686-linux-gnu.
Exporting functions and relying on symbol interposition from libc.so
makes the choice of implementation dependent on DT_NEEDED order, which
is not what some compiler drivers expect.
This commit replaces one magic mechanism (symbol interposition) with
another one (preprocessor-/compiler-based redirection). This makes
the hand-over from the minimal malloc to the full malloc more
explicit.
Removing the ABI symbols is backwards-compatible because libc.so is
always in scope, and the dynamic loader will find the malloc-related
symbols there since commit f0b2132b35
("ld.so: Support moving versioned symbols between sonames
[BZ #24741]").
Reviewed-by: Carlos O'Donell <carlos@redhat.com>
With all Linux ABIs using the expected Linux kABI to indicate
syscalls errors, there is no need to replicate the INLINE_SYSCALL.
The generic Linux sysdep.h includes errno.h even for !__ASSEMBLER__,
which is ok now and it allows cleanup some archaic code that assume
otherwise.
Checked with a build against all affected ABIs.
This patch adds a new macro, libm_alias_finite, to define all _finite
symbol. It sets all _finite symbol as compat symbol based on its first
version (obtained from the definition at built generated first-versions.h).
The <fn>f128_finite symbols were introduced in GLIBC 2.26 and so need
special treatment in code that is shared between long double and float128.
It is done by adding a list, similar to internal symbol redifinition,
on sysdeps/ieee754/float128/float128_private.h.
Alpha also needs some tricky changes to ensure we still emit 2 compat
symbols for sqrt(f).
Passes buildmanyglibc.
Co-authored-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
Reviewed-by: Siddhesh Poyarekar <siddhesh@sourceware.org>
This patch replaces s_trunc.c in sysdeps/dbl-64 with the one in
sysdeps/dbl-64/wordsize-64 and removes the latter one.
The code is not changed except changes in code style.
Also adjusted the include path in x86_64 and sparc64 files.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
This patch replaces s_ceil.c in sysdeps/dbl-64 with the one in
sysdeps/dbl-64/wordsize-64 and removes the latter one.
The code is not changed except changes in code style.
Also adjusted the include path in x86_64 and sparc64 files.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
This patch replaces s_floor.c in sysdeps/dbl-64 with the one in
sysdeps/dbl-64/wordsize-64 and removes the latter one.
The code is not changed except changes in code style.
Also adjusted the include path in x86_64 and sparc64 files.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
This patch replaces s_rint.c in sysdeps/dbl-64 with the one in
sysdeps/dbl-64/wordsize-64 and removes the latter one.
The code is not changed except changes in code style.
Also adjusted the include path in x86_64 file.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
This patch replaces s_nearbyint.c in sysdeps/dbl-64 with the one in
sysdeps/dbl-64/wordsize-64 and removes the latter one.
The code is not changed except changes in code style.
Also adjusted the include path in x86_64 file.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
This commit adds missing skip_ifunc checks to aarch64, arm, i386,
sparc, and x86_64. A new test case ensures that IRELATIVE IFUNC
resolvers do not run in various diagnostic modes of the dynamic
loader.
Reviewed-By: Szabolcs Nagy <szabolcs.nagy@arm.com>
This patch new build tests to check for internal fields offsets for
internal pthread_rwlock_t definition. Althoug the '__data.__flags'
field layout should be preserved due static initializators, the patch
also adds tests for the futexes that may be used in a shared memory
(although using different libc version in such scenario is not really
supported).
Checked with a build against all affected ABIs.
Change-Id: Iccc103d557de13d17e4a3f59a0cad2f4a640c148
The offsets of pthread_mutex_t __data.__nusers, __data.__spins,
__data.elision, __data.list are not required to be constant over
the releases. Only the __data.__kind is used for static
initializers.
This patch also adds an additional size check for __data.__kind.
Checked with a build against affected ABIs.
Change-Id: I7a4e48cc91b4c4ada57e9a5d1b151fb702bfaa9f
Remove _finite tests and references from x86_64. Rather than calling
__exp_finite, use exp directly (since it's the same entry point).
x86_64 builds and passes testsuite.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
Use the generic C memset/memcpy/memmove in benchtests since comparing
against a slow byte-oriented implementation makes no sense.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
2019-08-29 Wilco Dijkstra <wdijkstr@arm.com>
* benchtests/bench-memcpy.c (simple_memcpy): Remove.
(generic_memcpy): Include generic C memcpy.
* benchtests/bench-memmove.c (simple_memmove): Remove.
(generic_memmove): Include generic C memmove.
* benchtests/bench-memset.c (simple_memset): Remove.
(generic_memset): Include generic C memset.
* benchtests/bench-memset-large.c (simple_memset): Remove.
(generic_memset): Include generic C memset.
* benchtests/bench-memset-walk.c (simple_memset): Remove.
(generic_memset): Include generic C memset.
* string/memcpy.c (MEMCPY): Add defines to enable redirection.
* string/memset.c (MEMSET): Likewise.
* sysdeps/x86_64/memcopy.h: Remove empty file.
When compiled with -O3 and AVX, GCC 8 and 9 optimize some loops in
sysdeps/ieee754/dbl-64/branred.c with 256-bit vector instructions,
which leads to store forward stall:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90579
There is no easy fix in compiler. This patch limits vector width to
128 bits to work around this issue. It improves performance of sin
and cos by more than 40% on Skylake compiled with -O3 -march=skylake.
Tested with GCC 7/8/9 on x86-64.
[BZ #24603]
* sysdeps/x86_64/configure.ac: Check if -mprefer-vector-width=128
works.
* sysdeps/x86_64/configure: Regenerated.
* sysdeps/x86_64/fpu/Makefile (CFLAGS-branred.c): New. Set
to -mprefer-vector-width=128 if supported.
Since sysdeps/i386/dl-lookupcfg.h and sysdeps/x86_64/dl-lookupcfg.h are
identical, we can replace them with sysdeps/x86/dl-lookupcfg.h.
* sysdeps/i386/dl-lookupcfg.h: Moved to ...
* sysdeps/x86/dl-lookupcfg.h: Here.
* sysdeps/x86_64/dl-lookupcfg.h: Removed.
This patch rewrites wcscat using wcslen and wcscpy. This is similar to
the optimization done on strcat by 6e46de42fe.
The strcpy changes are mainly to add the internal alias to avoid PLT
calls.
Checked on x86_64-linux-gnu and a build against the affected
architectures.
* include/wchar.h (__wcscpy): New prototype.
* sysdeps/powerpc/powerpc32/power4/multiarch/wcscpy-ppc32.c
(__wcscpy): Route internal symbol to generic implementation.
* sysdeps/powerpc/powerpc32/power4/multiarch/wcscpy.c (wcscpy):
Add internal __wcscpy alias.
* sysdeps/powerpc/powerpc64/multiarch/wcscpy.c (wcscpy): Likewise.
* sysdeps/s390/wcscpy.c (wcscpy): Likewise.
* sysdeps/x86_64/multiarch/wcscpy.c (wcscpy): Likewise.
* wcsmbs/wcscpy.c (wcscpy): Add
* sysdeps/x86_64/multiarch/wcscpy-c.c (WCSCPY): Adjust macro to
use generic implementation.
* wcsmbs/wcscat.c (wcscat): Rewrite using wcslen and wcscpy.
This patch adds fall-through comments in some cases where -Wextra
produces implicit-fallthrough warnings.
The patch is non-exhaustive. Apart from architecture-specific code
for non-x86_64 architectures, it does not change sunrpc/xdr.c (legacy
code, probably should have such changes, but left to be dealt with
separately), or places that already had comments about the
fall-through but not matching the form expected by
-Wimplicit-fallthrough=3 (the default level with -Wextra; my
inclination is to adjust those comments to match rather than
downgrading to -Wimplicit-fallthrough=1 to allow any comment), or one
place where I thought the implicit fallthrough was not correct and so
should be handled separately as a bug fix. I think the key thing to
consider in review of this patch is whether the fall-through is indeed
intended and correct in each place where such a comment is added.
Tested for x86_64.
* elf/dl-exception.c (_dl_exception_create_format): Add
fall-through comments.
* elf/ldconfig.c (parse_conf_include): Likewise.
* elf/rtld.c (print_statistics): Likewise.
* locale/programs/charmap.c (parse_charmap): Likewise.
* misc/mntent_r.c (__getmntent_r): Likewise.
* posix/wordexp.c (parse_arith): Likewise.
(parse_backtick): Likewise.
* resolv/ns_ttl.c (ns_parse_ttl): Likewise.
* sysdeps/x86/cpu-features.c (init_cpu_features): Likewise.
* sysdeps/x86_64/dl-machine.h (elf_machine_rela): Likewise.
This fixes the same bug in fnmatch that was fixed by commit 7e2f0d2d77 for
regexp matching. As a side effect it also removes the use of an unbound
VLA.
Since the size argument is unsigned. we should use unsigned Jcc
instructions, instead of signed, to check size.
Tested on x86-64 and x32, with and without --disable-multi-arch.
[BZ #24155]
CVE-2019-7309
* NEWS: Updated for CVE-2019-7309.
* sysdeps/x86_64/memcmp.S: Use RDX_LP for size. Clear the
upper 32 bits of RDX register for x32. Use unsigned Jcc
instructions, instead of signed.
* sysdeps/x86_64/x32/Makefile (tests): Add tst-size_t-memcmp-2.
* sysdeps/x86_64/x32/tst-size_t-memcmp-2.c: New test.
On x32, the size_t parameter may be passed in the lower 32 bits of a
64-bit register with the non-zero upper 32 bits. The string/memory
functions written in assembly can only use the lower 32 bits of a
64-bit register as length or must clear the upper 32 bits before using
the full 64-bit register for length.
This pach fixes strnlen/wcsnlen for x32. Tested on x86-64 and x32. On
x86-64, libc.so is the same with and withou the fix.
[BZ# 24097]
CVE-2019-6488
* sysdeps/x86_64/multiarch/strlen-avx2.S: Use RSI_LP for length.
Clear the upper 32 bits of RSI register.
* sysdeps/x86_64/strlen.S: Use RSI_LP for length.
* sysdeps/x86_64/x32/Makefile (tests): Add tst-size_t-strnlen
and tst-size_t-wcsnlen.
* sysdeps/x86_64/x32/tst-size_t-strnlen.c: New file.
* sysdeps/x86_64/x32/tst-size_t-wcsnlen.c: Likewise.
On x32, the size_t parameter may be passed in the lower 32 bits of a
64-bit register with the non-zero upper 32 bits. The string/memory
functions written in assembly can only use the lower 32 bits of a
64-bit register as length or must clear the upper 32 bits before using
the full 64-bit register for length.
This pach fixes strncpy for x32. Tested on x86-64 and x32. On x86-64,
libc.so is the same with and withou the fix.
[BZ# 24097]
CVE-2019-6488
* sysdeps/x86_64/multiarch/strcpy-avx2.S: Use RDX_LP for length.
* sysdeps/x86_64/multiarch/strcpy-sse2-unaligned.S: Likewise.
* sysdeps/x86_64/multiarch/strcpy-ssse3.S: Likewise.
* sysdeps/x86_64/x32/Makefile (tests): Add tst-size_t-strncpy.
* sysdeps/x86_64/x32/tst-size_t-strncpy.c: New file.
On x32, the size_t parameter may be passed in the lower 32 bits of a
64-bit register with the non-zero upper 32 bits. The string/memory
functions written in assembly can only use the lower 32 bits of a
64-bit register as length or must clear the upper 32 bits before using
the full 64-bit register for length.
This pach fixes the strncmp family for x32. Tested on x86-64 and x32.
On x86-64, libc.so is the same with and withou the fix.
[BZ# 24097]
CVE-2019-6488
* sysdeps/x86_64/multiarch/strcmp-avx2.S: Use RDX_LP for length.
* sysdeps/x86_64/multiarch/strcmp-sse42.S: Likewise.
* sysdeps/x86_64/strcmp.S: Likewise.
* sysdeps/x86_64/x32/Makefile (tests): Add tst-size_t-strncasecmp,
tst-size_t-strncmp and tst-size_t-wcsncmp.
* sysdeps/x86_64/x32/tst-size_t-strncasecmp.c: New file.
* sysdeps/x86_64/x32/tst-size_t-strncmp.c: Likewise.
* sysdeps/x86_64/x32/tst-size_t-wcsncmp.c: Likewise.
On x32, the size_t parameter may be passed in the lower 32 bits of a
64-bit register with the non-zero upper 32 bits. The string/memory
functions written in assembly can only use the lower 32 bits of a
64-bit register as length or must clear the upper 32 bits before using
the full 64-bit register for length.
This pach fixes memset/wmemset for x32. Tested on x86-64 and x32. On
x86-64, libc.so is the same with and withou the fix.
[BZ# 24097]
CVE-2019-6488
* sysdeps/x86_64/multiarch/memset-avx512-no-vzeroupper.S: Use
RDX_LP for length. Clear the upper 32 bits of RDX register.
* sysdeps/x86_64/multiarch/memset-vec-unaligned-erms.S: Likewise.
* sysdeps/x86_64/x32/Makefile (tests): Add tst-size_t-wmemset.
* sysdeps/x86_64/x32/tst-size_t-memset.c: New file.
* sysdeps/x86_64/x32/tst-size_t-wmemset.c: Likewise.
On x32, the size_t parameter may be passed in the lower 32 bits of a
64-bit register with the non-zero upper 32 bits. The string/memory
functions written in assembly can only use the lower 32 bits of a
64-bit register as length or must clear the upper 32 bits before using
the full 64-bit register for length.
This pach fixes memrchr for x32. Tested on x86-64 and x32. On x86-64,
libc.so is the same with and withou the fix.
[BZ# 24097]
CVE-2019-6488
* sysdeps/x86_64/memrchr.S: Use RDX_LP for length.
* sysdeps/x86_64/multiarch/memrchr-avx2.S: Likewise.
* sysdeps/x86_64/x32/Makefile (tests): Add tst-size_t-memrchr.
* sysdeps/x86_64/x32/tst-size_t-memrchr.c: New file.
On x32, the size_t parameter may be passed in the lower 32 bits of a
64-bit register with the non-zero upper 32 bits. The string/memory
functions written in assembly can only use the lower 32 bits of a
64-bit register as length or must clear the upper 32 bits before using
the full 64-bit register for length.
This pach fixes memcpy for x32. Tested on x86-64 and x32. On x86-64,
libc.so is the same with and withou the fix.
[BZ# 24097]
CVE-2019-6488
* sysdeps/x86_64/multiarch/memcpy-ssse3-back.S: Use RDX_LP for
length. Clear the upper 32 bits of RDX register.
* sysdeps/x86_64/multiarch/memcpy-ssse3.S: Likewise.
* sysdeps/x86_64/multiarch/memmove-avx512-no-vzeroupper.S:
Likewise.
* sysdeps/x86_64/multiarch/memmove-vec-unaligned-erms.S:
Likewise.
* sysdeps/x86_64/x32/Makefile (tests): Add tst-size_t-memcpy.
tst-size_t-wmemchr.
* sysdeps/x86_64/x32/tst-size_t-memcpy.c: New file.
On x32, the size_t parameter may be passed in the lower 32 bits of a
64-bit register with the non-zero upper 32 bits. The string/memory
functions written in assembly can only use the lower 32 bits of a
64-bit register as length or must clear the upper 32 bits before using
the full 64-bit register for length.
This pach fixes memcmp/wmemcmp for x32. Tested on x86-64 and x32. On
x86-64, libc.so is the same with and withou the fix.
[BZ# 24097]
CVE-2019-6488
* sysdeps/x86_64/multiarch/memcmp-avx2-movbe.S: Use RDX_LP for
length. Clear the upper 32 bits of RDX register.
* sysdeps/x86_64/multiarch/memcmp-sse4.S: Likewise.
* sysdeps/x86_64/multiarch/memcmp-ssse3.S: Likewise.
* sysdeps/x86_64/x32/Makefile (tests): Add tst-size_t-memcmp and
tst-size_t-wmemcmp.
* sysdeps/x86_64/x32/tst-size_t-memcmp.c: New file.
* sysdeps/x86_64/x32/tst-size_t-wmemcmp.c: Likewise.
On x32, the size_t parameter may be passed in the lower 32 bits of a
64-bit register with the non-zero upper 32 bits. The string/memory
functions written in assembly can only use the lower 32 bits of a
64-bit register as length or must clear the upper 32 bits before using
the full 64-bit register for length.
This pach fixes memchr/wmemchr for x32. Tested on x86-64 and x32. On
x86-64, libc.so is the same with and withou the fix.
[BZ# 24097]
CVE-2019-6488
* sysdeps/x86_64/memchr.S: Use RDX_LP for length. Clear the
upper 32 bits of RDX register.
* sysdeps/x86_64/multiarch/memchr-avx2.S: Likewise.
* sysdeps/x86_64/x32/Makefile (tests): Add tst-size_t-memchr and
tst-size_t-wmemchr.
* sysdeps/x86_64/x32/test-size_t.h: New file.
* sysdeps/x86_64/x32/tst-size_t-memchr.c: Likewise.
* sysdeps/x86_64/x32/tst-size_t-wmemchr.c: Likewise.
Optimize x86-64 strcat/strncat, strcpy/strncpy and stpcpy/stpncpy with AVX2.
It uses vector comparison as much as possible. In general, the larger the
source string, the greater performance gain observed, reaching speedups of
1.6x compared to SSE2 unaligned routines. Select AVX2 strcat/strncat,
strcpy/strncpy and stpcpy/stpncpy on AVX2 machines where vzeroupper is
preferred and AVX unaligned load is fast.
* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add
strcat-avx2, strncat-avx2, strcpy-avx2, strncpy-avx2,
stpcpy-avx2 and stpncpy-avx2.
* sysdeps/x86_64/multiarch/ifunc-impl-list.c:
(__libc_ifunc_impl_list): Add tests for __strcat_avx2,
__strncat_avx2, __strcpy_avx2, __strncpy_avx2, __stpcpy_avx2
and __stpncpy_avx2.
* sysdeps/x86_64/multiarch/{ifunc-unaligned-ssse3.h =>
ifunc-strcpy.h}: rename header for a more generic name.
* sysdeps/x86_64/multiarch/ifunc-strcpy.h:
(IFUNC_SELECTOR): Return OPTIMIZE (avx2) on AVX 2 machines if
AVX unaligned load is fast and vzeroupper is preferred.
* sysdeps/x86_64/multiarch/stpcpy-avx2.S: New file
* sysdeps/x86_64/multiarch/stpncpy-avx2.S: Likewise
* sysdeps/x86_64/multiarch/strcat-avx2.S: Likewise
* sysdeps/x86_64/multiarch/strcpy-avx2.S: Likewise
* sysdeps/x86_64/multiarch/strncat-avx2.S: Likewise
* sysdeps/x86_64/multiarch/strncpy-avx2.S: Likewise
The x86 defines optimized THREAD_ATOMIC_* macros where reference always
the current thread instead of the one indicated by input 'descr' argument.
It work as long the input is the self thread pointer, however it generates
wrong code if the semantic is to set a bit atomicialy from another thread.
This is not an issue for current GLIBC usage, however the new cancellation
code expects that some synchronization code to atomically set bits from
different threads.
The generic code generates an additional load to reference to TLS segment,
for instance the code:
THREAD_ATOMIC_BIT_SET (THREAD_SELF, cancelhandling, CANCELED_BIT);
Compiles to:
lock;orl $4, %fs:776
Where with patch changes it now compiles to:
mov %fs:16,%rax
lock;orl $4, 776(%rax)
If some usage indeed proves to be a hotspot we can add an extra macro
with a more descriptive name (THREAD_ATOMIC_BIT_SET_SELF for instance)
where x86_64 might optimize it.
Checked on x86_64-linux-gnu.
* sysdeps/x86_64/nptl/tls.h (THREAD_ATOMIC_CMPXCHG_VAL,
THREAD_ATOMIC_AND, THREAD_ATOMIC_BIT_SET): Remove macros.
The current s_sincosf.c is faster than s_sincosf-sse2.S. On Broadwell
with FMA disabled, bench-sincosf shows:
Before After Improvement
max 154.032 114.517 34%
min 6.25 5.609 11%
mean 14.8728 12.8589 15%
* sysdeps/x86_64/fpu/s_sincosf.S: Removed.
* sysdeps/x86_64/fpu/multiarch/s_sincosf-sse2.S: Likewise.
* sysdeps/x86_64/fpu/multiarch/s_sincosf-sse2.c: New file.
