glibc/sysdeps/x86_64/multiarch/strcmp.c
Aurelien Jarno 4d64c64457 x86-64: Require BMI2 for AVX2 strcmp implementation
The AVX2 strcmp implementation uses the 'bzhi' instruction, which
belongs to the BMI2 CPU feature.

NB: It also uses the 'tzcnt' BMI1 instruction, but it is executed as BSF
as BSF if the CPU doesn't support TZCNT, and produces the same result
for non-zero input.

Partially fixes: b77b06e0e2 ("x86: Optimize strcmp-avx2.S")
Partially resolves: BZ #29611

Reviewed-by: Noah Goldstein  <goldstein.w.n@gmail.com>
2022-10-03 23:46:11 +02:00

86 lines
2.9 KiB
C

/* Multiple versions of strcmp.
All versions must be listed in ifunc-impl-list.c.
Copyright (C) 2017-2022 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
The GNU C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with the GNU C Library; if not, see
<https://www.gnu.org/licenses/>. */
/* Define multiple versions only for the definition in libc. */
#if IS_IN (libc)
# define strcmp __redirect_strcmp
# include <string.h>
# undef strcmp
# define SYMBOL_NAME strcmp
# include <init-arch.h>
extern __typeof (REDIRECT_NAME) OPTIMIZE (evex) attribute_hidden;
extern __typeof (REDIRECT_NAME) OPTIMIZE (avx2) attribute_hidden;
extern __typeof (REDIRECT_NAME) OPTIMIZE (avx2_rtm) attribute_hidden;
extern __typeof (REDIRECT_NAME) OPTIMIZE (sse42) attribute_hidden;
extern __typeof (REDIRECT_NAME)
OPTIMIZE (sse2_unaligned) attribute_hidden;
extern __typeof (REDIRECT_NAME) OPTIMIZE (sse2) attribute_hidden;
static inline void *
IFUNC_SELECTOR (void)
{
const struct cpu_features *cpu_features = __get_cpu_features ();
if (X86_ISA_CPU_FEATURE_USABLE_P (cpu_features, AVX2)
&& X86_ISA_CPU_FEATURE_USABLE_P (cpu_features, BMI2)
&& X86_ISA_CPU_FEATURES_ARCH_P (cpu_features,
AVX_Fast_Unaligned_Load, ))
{
if (X86_ISA_CPU_FEATURE_USABLE_P (cpu_features, AVX512VL)
&& X86_ISA_CPU_FEATURE_USABLE_P (cpu_features, AVX512BW))
return OPTIMIZE (evex);
if (CPU_FEATURE_USABLE_P (cpu_features, RTM))
return OPTIMIZE (avx2_rtm);
if (X86_ISA_CPU_FEATURES_ARCH_P (cpu_features,
Prefer_No_VZEROUPPER, !))
return OPTIMIZE (avx2);
}
if (X86_ISA_CPU_FEATURE_USABLE_P (cpu_features, SSE4_2)
/* Keep this as runtime check. Some ISA level >= 2 CPUs such as
Tremont, Silvermont, and more check this. */
&& !CPU_FEATURES_ARCH_P (cpu_features, Slow_SSE4_2))
return OPTIMIZE (sse42);
/* Keep this as runtime check. The standard SSE2 version has
meaningful optimizations around keeping all loads aligned in the
main loop which can benefit some ISA level >= 2 CPUs. */
if (CPU_FEATURES_ARCH_P (cpu_features, Fast_Unaligned_Load))
return OPTIMIZE (sse2_unaligned);
return OPTIMIZE (sse2);
}
libc_ifunc_redirected (__redirect_strcmp, strcmp, IFUNC_SELECTOR ());
# ifdef SHARED
__hidden_ver1 (strcmp, __GI_strcmp, __redirect_strcmp)
__attribute__ ((visibility ("hidden"))) __attribute_copy__ (strcmp);
# endif
#endif