qiurui
/
glibc
mirror of git://sourceware.org/git/glibc.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

x86: Set Prefer_No_VZEROUPPER and add Prefer_AVX2_STRCMP 

1. Set Prefer_No_VZEROUPPER if RTM is usable to avoid RTM abort triggered
by VZEROUPPER inside a transactionally executing RTM region.
2. Since to compare 2 32-byte strings, 256-bit EVEX strcmp requires 2
loads, 3 VPCMPs and 2 KORDs while AVX2 strcmp requires 1 load, 2 VPCMPEQs,
1 VPMINU and 1 VPMOVMSKB, AVX2 strcmp is faster than EVEX strcmp.  Add
Prefer_AVX2_STRCMP to prefer AVX2 strcmp family functions.

(cherry picked from commit 1da50d4bda)
This commit is contained in:
H.J. Lu 2021-02-26 05:36:59 -08:00
parent 43c27a754b
commit b2e59f87ae
3 changed files with 23 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
sysdeps/x86/cpu-features.c
View File

@ -347,8 +347,24 @@ init_cpu_features (struct cpu_features *cpu_features)
cpu_features->feature[index_arch_Prefer_No_VZEROUPPER]
|= bit_arch_Prefer_No_VZEROUPPER;
else
cpu_features->feature[index_arch_Prefer_No_AVX512]
|= bit_arch_Prefer_No_AVX512;
{
cpu_features->feature[index_arch_Prefer_No_AVX512]
|= bit_arch_Prefer_No_AVX512;
/* Avoid RTM abort triggered by VZEROUPPER inside a
transactionally executing RTM region. */
if (CPU_FEATURES_CPU_P (cpu_features, RTM))
cpu_features->feature[index_arch_Prefer_No_VZEROUPPER]
|= bit_arch_Prefer_No_VZEROUPPER;
/* Since to compare 2 32-byte strings, 256-bit EVEX strcmp
requires 2 loads, 3 VPCMPs and 2 KORDs while AVX2 strcmp
requires 1 load, 2 VPCMPEQs, 1 VPMINU and 1 VPMOVMSKB,
AVX2 strcmp is faster than EVEX strcmp. */
if (CPU_FEATURES_ARCH_P (cpu_features, AVX2_Usable))
cpu_features->feature[index_arch_Prefer_AVX2_STRCMP]
|= bit_arch_Prefer_AVX2_STRCMP;
}
}
/* This spells out "AuthenticAMD". */
else if (ebx == 0x68747541 && ecx == 0x444d4163 && edx == 0x69746e65)

2
sysdeps/x86/cpu-features.h
View File

@ -41,6 +41,7 @@
#define bit_arch_MathVec_Prefer_No_AVX512 (1 << 21)
#define bit_arch_XSAVEC_Usable (1 << 22)
#define bit_arch_Prefer_FSRM (1 << 23)
#define bit_arch_Prefer_AVX2_STRCMP (1 << 24)
/* CPUID Feature flags. */
@ -266,6 +267,7 @@ extern const struct cpu_features *__get_cpu_features (void)
# define index_arch_MathVec_Prefer_No_AVX512 FEATURE_INDEX_1
# define index_arch_XSAVEC_Usable FEATURE_INDEX_1
# define index_arch_Prefer_FSRM FEATURE_INDEX_1
# define index_arch_Prefer_AVX2_STRCMP FEATURE_INDEX_1
#endif /* !__ASSEMBLER__ */

3
sysdeps/x86/cpu-tunables.c
View File

@ -282,6 +282,9 @@ TUNABLE_CALLBACK (set_hwcaps) (tunable_val_t *valp)
CHECK_GLIBC_IFUNC_ARCH_BOTH (n, cpu_features,
Fast_Copy_Backward, disable,
18);
CHECK_GLIBC_IFUNC_ARCH_NEED_ARCH_BOTH
(n, cpu_features, Prefer_AVX2_STRCMP, AVX2_Usable,
disable, 18);
}
break;
case 19: