diff options
| author | Amrita H S <amritahs@linux.ibm.com> | 2024-05-06 09:01:29 -0500 |
|---|---|---|
| committer | Peter Bergner <bergner@linux.ibm.com> | 2024-05-06 09:01:29 -0500 |
| commit | 23f0d81608d0ca6379894ef81670cf30af7fd081 (patch) | |
| tree | 2c6ec46bf4d7bfb00c46f215a5735cb49db65e31 /sysdeps/powerpc/powerpc64/multiarch/strncmp.c | |
| parent | ea73eb5f581ef5931fd67005aa0c526ba43366c9 (diff) | |
| download | glibc-23f0d81608d0ca6379894ef81670cf30af7fd081.zip glibc-23f0d81608d0ca6379894ef81670cf30af7fd081.tar.gz glibc-23f0d81608d0ca6379894ef81670cf30af7fd081.tar.bz2 | |
powerpc: Optimized strncmp for power10
This patch is based on __strcmp_power10.
Improvements from __strncmp_power9:
1. Uses new POWER10 instructions
- This code uses lxvp to decrease contention on load
by loading 32 bytes per instruction.
2. Performance implication
- This version has around 38% better performance on average.
- Minor performance regression is seen for few small sizes
and specific combination of alignments.
Signed-off-by: Amrita H S <amritahs@linux.ibm.com>
Reviewed-by: Peter Bergner <bergner@linux.ibm.com>
Diffstat (limited to 'sysdeps/powerpc/powerpc64/multiarch/strncmp.c')
| -rw-r--r-- | sysdeps/powerpc/powerpc64/multiarch/strncmp.c | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/sysdeps/powerpc/powerpc64/multiarch/strncmp.c b/sysdeps/powerpc/powerpc64/multiarch/strncmp.c index 6178f4a..a5ed67f 100644 --- a/sysdeps/powerpc/powerpc64/multiarch/strncmp.c +++ b/sysdeps/powerpc/powerpc64/multiarch/strncmp.c @@ -29,6 +29,7 @@ extern __typeof (strncmp) __strncmp_ppc attribute_hidden; extern __typeof (strncmp) __strncmp_power8 attribute_hidden; # ifdef __LITTLE_ENDIAN__ extern __typeof (strncmp) __strncmp_power9 attribute_hidden; +extern __typeof (strncmp) __strncmp_power10 attribute_hidden; # endif # undef strncmp @@ -36,6 +37,9 @@ extern __typeof (strncmp) __strncmp_power9 attribute_hidden; ifunc symbol properly. */ libc_ifunc_redirected (__redirect_strncmp, strncmp, # ifdef __LITTLE_ENDIAN__ + (hwcap2 & PPC_FEATURE2_ARCH_3_1 + && hwcap & PPC_FEATURE_HAS_VSX) + ? __strncmp_power10 : (hwcap2 & PPC_FEATURE2_ARCH_3_00 && hwcap & PPC_FEATURE_HAS_ALTIVEC) ? __strncmp_power9 : |