diff options
| author | Leonardo Sandoval <leonardo.sandoval.gonzalez@linux.intel.com> | 2018-05-03 11:09:30 -0500 |
|---|---|---|
| committer | Leonardo Sandoval <leonardo.sandoval.gonzalez@linux.intel.com> | 2018-06-01 16:32:43 -0500 |
| commit | 1457016337072d1b6739f571846b619596990cb7 (patch) | |
| tree | 985983beef1070b9e01f6f31ad974f5038278799 /sysdeps/x86_64/multiarch/Makefile | |
| parent | e02c026f38505cd474ff1bdaa88fc671804f5805 (diff) | |
| download | glibc-1457016337072d1b6739f571846b619596990cb7.tar.gz glibc-1457016337072d1b6739f571846b619596990cb7.tar.xz glibc-1457016337072d1b6739f571846b619596990cb7.zip | |
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.
Diffstat (limited to 'sysdeps/x86_64/multiarch/Makefile')
| -rw-r--r-- | sysdeps/x86_64/multiarch/Makefile | 6 |
1 files changed, 4 insertions, 2 deletions
diff --git a/sysdeps/x86_64/multiarch/Makefile b/sysdeps/x86_64/multiarch/Makefile index 68257c4017..bb5e970735 100644 --- a/sysdeps/x86_64/multiarch/Makefile +++ b/sysdeps/x86_64/multiarch/Makefile @@ -6,8 +6,8 @@ ifeq ($(subdir),string) sysdep_routines += strncat-c stpncpy-c strncpy-c \ strcmp-sse2 strcmp-sse2-unaligned strcmp-ssse3 \ - strcmp-sse4_2 \ - strncmp-sse2 strncmp-ssse3 strncmp-sse4_2 \ + strcmp-sse4_2 strcmp-avx2 \ + strncmp-sse2 strncmp-ssse3 strncmp-sse4_2 strncmp-avx2 \ memchr-sse2 rawmemchr-sse2 memchr-avx2 rawmemchr-avx2 \ memrchr-sse2 memrchr-avx2 \ memcmp-sse2 \ @@ -51,6 +51,8 @@ ifeq ($(subdir),wcsmbs) sysdep_routines += wmemcmp-sse4 wmemcmp-ssse3 wmemcmp-c \ wmemcmp-avx2-movbe \ wmemchr-sse2 wmemchr-avx2 \ + wcscmp-sse2 wcscmp-avx2 \ + wcsncmp-sse2 wcsncmp-avx2 \ wcscpy-ssse3 wcscpy-c \ wcschr-sse2 wcschr-avx2 \ wcsrchr-sse2 wcsrchr-avx2 \ |