From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: by sourceware.org (Postfix, from userid 1778) id 53C6B385AC3C; Mon, 3 Oct 2022 22:05:01 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 53C6B385AC3C DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=sourceware.org; s=default; t=1664834701; bh=+wVBvgX+C4zb4hyFlxxdHHWf+uJkB1aSJDLzJ6Qriqs=; h=From:To:Subject:Date:From; b=CyierMp9vNGG99OnT4fT0J1ZZC27mmA1aR7Xi5sHFW2lznn91B04I8qWTs0ar0s45 w6rT2MHJYBm6rN5iPwzUP0NAgWPZCm4GdltusEub4suclGRMIPCNkG89jR4DG51dWa pMm0bT9oatkaWPnmFLG3ATkkmPlj5cgG+8B0VP24= 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/2.33/master] x86-64: Require BMI2 for AVX2 strncmp implementation X-Act-Checkin: glibc X-Git-Author: Aurelien Jarno X-Git-Refname: refs/heads/release/2.33/master X-Git-Oldrev: 26d4359b448a39054e627ad072a7480dd24f0313 X-Git-Newrev: 050d652900d5f88447af335b20b621344f16e192 Message-Id: <20221003220501.53C6B385AC3C@sourceware.org> Date: Mon, 3 Oct 2022 22:05:01 +0000 (GMT) List-Id: https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=050d652900d5f88447af335b20b621344f16e192 commit 050d652900d5f88447af335b20b621344f16e192 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 70931f1598..34d5f6efe5 100644 --- a/sysdeps/x86_64/multiarch/ifunc-impl-list.c +++ b/sysdeps/x86_64/multiarch/ifunc-impl-list.c @@ -1022,15 +1022,18 @@ __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 f94a421784..7632d7b2dd 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))