Leonardo Sandoval 1457016337 x86-64: Optimize strcmp/wcscmp and strncmp/wcsncmp with AVX2 ... Optimize x86-64 strcmp/wcscmp and strncmp/wcsncmp with AVX2. It uses vector comparison as much as possible. Peak performance observed on a SkyLake machine: 9x, 3x, 2.5x and 5.5x for strcmp, strncmp, wcscmp and wcsncmp, respectively. The larger the comparison length, the more benefit using avx2 functions, except on the strcmp, where peak is observed at length == 32 bytes. Select AVX2 strcmp/wcscmp on AVX2 machines where vzeroupper is preferred and AVX unaligned load is fast. NB: It uses TZCNT instead of BSF since TZCNT produces the same result as BSF for non-zero input. TZCNT is faster than BSF and is executed as BSF if machine doesn't support TZCNT. * sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Add strcmp-avx2, strncmp-avx2, wcscmp-avx2, wcscmp-sse2, wcsncmp-avx2 and wcsncmp-sse2. * sysdeps/x86_64/multiarch/ifunc-impl-list.c (__libc_ifunc_impl_list): Add tests for __strcmp_avx2, __strncmp_avx2, __wcscmp_avx2, __wcsncmp_avx2, __wcscmp_sse2 and __wcsncmp_sse2. * sysdeps/x86_64/multiarch/strcmp.c (OPTIMIZE (avx2)): (IFUNC_SELECTOR): Return OPTIMIZE (avx2) on AVX 2 machines if AVX unaligned load is fast and vzeroupper is preferred. * sysdeps/x86_64/multiarch/strncmp.c: Likewise. * sysdeps/x86_64/multiarch/strcmp-avx2.S: New file. * sysdeps/x86_64/multiarch/strncmp-avx2.S: Likewise. * sysdeps/x86_64/multiarch/wcscmp-avx2.S: Likewise. * sysdeps/x86_64/multiarch/wcscmp-sse2.S: Likewise. * sysdeps/x86_64/multiarch/wcscmp.c: Likewise. * sysdeps/x86_64/multiarch/wcsncmp-avx2.S: Likewise. * sysdeps/x86_64/multiarch/wcsncmp-sse2.c: Likewise. * sysdeps/x86_64/multiarch/wcsncmp.c: Likewise. * sysdeps/x86_64/wcscmp.S (__wcscmp): Add alias only if __wcscmp is undefined.		2018-06-01 16:32:43 -05:00
..
aarch64	Remove sysdeps/aarch64/soft-fp directory.	2018-05-22 17:23:34 +00:00
alpha	Remove sysdeps/alpha/soft-fp directory.	2018-05-23 17:29:20 +00:00
arm	Drop fpregset unused symbol exposition	2018-04-20 01:27:13 +02:00
generic	Remove sysdeps/generic/libcidn.abilist	2018-06-01 11:25:41 +02:00
gnu	Increase robustness of internal dlopen() by using RTLD_NOW [BZ #22766]	2018-04-26 10:41:43 -03:00
hppa	R_PARISC_TLS_DTPOFF32 reloc handling	2018-05-13 08:32:28 +09:30
htl	hurd: Avoid exposing all <sched.h> symbols from sys/types.h	2018-04-19 20:24:36 +02:00
hurd	hurd: Fix hurd installed headers test	2018-04-20 00:16:40 +02:00
i386	math: Update i686 ulps (--disable-multi-arch configuration)	2018-06-01 22:37:55 +02:00
ia64	elf: Unify symbol address run-time calculation [BZ #19818]	2018-04-04 23:09:37 +01:00
ieee754	Fix i686-linux-gnu build with GCC mainline.	2018-05-22 16:55:04 +00:00
init_array	sysdeps/init_array: Add PREINIT_FUNCTION to crti.S	2018-01-29 10:22:26 -08:00
m68k	Do not include math-barriers.h in math_private.h.	2018-05-11 15:11:38 +00:00
mach	Add narrowing divide functions.	2018-05-17 00:40:52 +00:00
microblaze	elf: Unify symbol address run-time calculation [BZ #19818]	2018-04-04 23:09:37 +01:00
mips	Update MIPS libm-test-ulps.	2018-05-16 15:35:26 +00:00
nios2	Make powerpc-nofpu __sqrtsf2, __sqrtdf2 compat symbols (bug 18473).	2018-06-01 17:25:12 +00:00
nptl	Fix comment typo	2018-05-08 14:59:13 +02:00
posix	Switch IDNA implementation to libidn2 [BZ #19728] [BZ #19729] [BZ #22247]	2018-05-23 15:27:24 +02:00
powerpc	Make powerpc-nofpu __sqrtsf2, __sqrtdf2 compat symbols (bug 18473).	2018-06-01 17:25:12 +00:00
pthread	hurd: fix sigevent's sigev_notify_attributes field type	2018-04-19 21:43:44 +02:00
riscv	elf: Unify symbol address run-time calculation [BZ #19818]	2018-04-04 23:09:37 +01:00
s390	nptl: Remove __ASSUME_PRIVATE_FUTEX	2018-05-17 04:25:10 -07:00
sh	Remove sysdeps/sh/soft-fp directory.	2018-05-23 20:05:31 +00:00
sparc	Remove sysdeps/sparc/sparc64/soft-fp directory.	2018-05-25 20:00:51 +00:00
unix	Remove sysdeps/powerpc/soft-fp directory.	2018-05-24 22:02:32 +00:00
wordsize-32	Use libc_hidden_* for strtoumax (bug 15105).	2018-02-28 14:16:21 +00:00
wordsize-64	Use libc_hidden_* for strtoumax (bug 15105).	2018-02-28 14:16:21 +00:00
x86	x86-64: Check Prefer_FSRM in ifunc-memmove.h	2018-05-21 16:54:59 -07:00
x86_64	x86-64: Optimize strcmp/wcscmp and strncmp/wcsncmp with AVX2	2018-06-01 16:32:43 -05:00