From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: by sourceware.org (Postfix, from userid 1778) id A6DE13858D1E; Mon, 3 Oct 2022 21:55:54 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org A6DE13858D1E DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=sourceware.org; s=default; t=1664834154; bh=JrOeyp772GQSw3M1WAw+Z2Y+LGr1hTE4XAqOwdI5T5g=; h=From:To:Subject:Date:From; b=vmH/QV0Lhpu6H9BF+EEU52IepZntklVaIlDGtg8UGrH1zCII6oZIo2vIRyrCgF1S0 eGhXM+5qwXpVwP+R6m190dlsxMdo1R6n4lo22YnM7krjY6SP7nj/RiS3mSMqBYo2+M CdJts9GmPO/B276AkFcm2/CClGRJODfiJdS7JqJ8= Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit From: Aurelien Jarno To: glibc-cvs@sourceware.org Subject: [glibc/release/2.35/master] x86-64: Require BMI2 for AVX2 strncmp implementation X-Act-Checkin: glibc X-Git-Author: Aurelien Jarno X-Git-Refname: refs/heads/release/2.35/master X-Git-Oldrev: 6260de8867d9634e5b9b430ce321da6a3256e9a5 X-Git-Newrev: 7e1326c43a7ab2468946bd711983907d79ceca3b Message-Id: <20221003215554.A6DE13858D1E@sourceware.org> Date: Mon, 3 Oct 2022 21:55:54 +0000 (GMT) List-Id: https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=7e1326c43a7ab2468946bd711983907d79ceca3b commit 7e1326c43a7ab2468946bd711983907d79ceca3b Author: Aurelien Jarno Date: Mon Oct 3 23:16:46 2022 +0200 x86-64: Require BMI2 for AVX2 strncmp implementation The AVX2 strncmp implementations 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: b77b06e0e296 ("x86: Optimize strcmp-avx2.S") Partially resolves: BZ #29611 Reviewed-by: Noah Goldstein (cherry picked from commit fc7de1d9b99ae1676bc626ddca422d7abee0eb48) Diff: --- sysdeps/x86_64/multiarch/ifunc-impl-list.c | 7 +++++-- sysdeps/x86_64/multiarch/strncmp.c | 4 ++-- 2 files changed, 7 insertions(+), 4 deletions(-) diff --git a/sysdeps/x86_64/multiarch/ifunc-impl-list.c b/sysdeps/x86_64/multiarch/ifunc-impl-list.c index 136e3b8103..b27ae975e4 100644 --- a/sysdeps/x86_64/multiarch/ifunc-impl-list.c +++ b/sysdeps/x86_64/multiarch/ifunc-impl-list.c @@ -1040,15 +1040,18 @@ __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array, /* Support sysdeps/x86_64/multiarch/strncmp.c. */ IFUNC_IMPL (i, name, strncmp, IFUNC_IMPL_ADD (array, i, strncmp, - CPU_FEATURE_USABLE (AVX2), + (CPU_FEATURE_USABLE (AVX2) + && CPU_FEATURE_USABLE (BMI2)), __strncmp_avx2) IFUNC_IMPL_ADD (array, i, strncmp, (CPU_FEATURE_USABLE (AVX2) + && CPU_FEATURE_USABLE (BMI2) && CPU_FEATURE_USABLE (RTM)), __strncmp_avx2_rtm) IFUNC_IMPL_ADD (array, i, strncmp, (CPU_FEATURE_USABLE (AVX512VL) - && CPU_FEATURE_USABLE (AVX512BW)), + && CPU_FEATURE_USABLE (AVX512BW) + && CPU_FEATURE_USABLE (BMI2)), __strncmp_evex) IFUNC_IMPL_ADD (array, i, strncmp, CPU_FEATURE_USABLE (SSE4_2), __strncmp_sse42) diff --git a/sysdeps/x86_64/multiarch/strncmp.c b/sysdeps/x86_64/multiarch/strncmp.c index fca74199d8..4241da5bdf 100644 --- a/sysdeps/x86_64/multiarch/strncmp.c +++ b/sysdeps/x86_64/multiarch/strncmp.c @@ -39,11 +39,11 @@ IFUNC_SELECTOR (void) const struct cpu_features* cpu_features = __get_cpu_features (); if (CPU_FEATURE_USABLE_P (cpu_features, AVX2) + && CPU_FEATURE_USABLE_P (cpu_features, BMI2) && CPU_FEATURES_ARCH_P (cpu_features, AVX_Fast_Unaligned_Load)) { if (CPU_FEATURE_USABLE_P (cpu_features, AVX512VL) - && CPU_FEATURE_USABLE_P (cpu_features, AVX512BW) - && CPU_FEATURE_USABLE_P (cpu_features, BMI2)) + && CPU_FEATURE_USABLE_P (cpu_features, AVX512BW)) return OPTIMIZE (evex); if (CPU_FEATURE_USABLE_P (cpu_features, RTM))