From patchwork Sat Jan 16 11:48:33 2016 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Andrew Senkevich X-Patchwork-Id: 10403 Received: (qmail 106177 invoked by alias); 16 Jan 2016 11:49:08 -0000 Mailing-List: contact libc-alpha-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Unsubscribe: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: libc-alpha-owner@sourceware.org Delivered-To: mailing list libc-alpha@sourceware.org Received: (qmail 106162 invoked by uid 89); 16 Jan 2016 11:49:07 -0000 Authentication-Results: sourceware.org; auth=none X-Virus-Found: No X-Spam-SWARE-Status: No, score=-0.1 required=5.0 tests=AWL, BAYES_00, FREEMAIL_FROM, RCVD_IN_DNSWL_LOW, SPF_PASS autolearn=ham version=3.3.2 spammy=339, 33, 9, family, 2016-01-16 X-HELO: mail-lb0-f173.google.com X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-type:content-transfer-encoding; bh=O+KwyP4sd5sC4clyTvkiMVqPWLcrb84EoaLrIRprfRU=; b=Zad1F5g9UrUjqWkOKgCHi3/hpIGnqN+kcCNe2MWtXqsTCQ0owntzzUVMIBgngw4Uul z6J+qeWiXQ3vIHEuMCQOoyeIbFavUgHe8H8eWvuTgW39hbB9z1O5tQJa5gsGPdmuZnDY iCAmhX2bxrfjvRBmtLCFZY/7gOmgJboCIx/PLShHI25EaaitiUmePEENvxexEfGvnSCa IkVwvjIv5RurwNdbxz9Wb/8eJv9nc05EuB99e4CaQ7xJlKaHfnX2ULRjQ74qLbGM2z7b P7npJ3SeS2pmotCcctYe/aJa7dTnrvz69M7tHiHRGKrJ0EiRltxr5UNB3xREcm9dwl19 nA/g== X-Gm-Message-State: ALoCoQl6hn/ATKAFXNQWZyWaClNA2r1wH4whCN17NbwocmMX4Z4Wk2dLOJxvieOb6i90ZIB0BSFvINx3Vld/ogfJa0Sg+TxftA== X-Received: by 10.112.180.138 with SMTP id do10mr5066774lbc.127.1452944943159; Sat, 16 Jan 2016 03:49:03 -0800 (PST) MIME-Version: 1.0 In-Reply-To: <569971BB.5050702@redhat.com> References: <569971BB.5050702@redhat.com> From: Andrew Senkevich Date: Sat, 16 Jan 2016 14:48:33 +0300 Message-ID: Subject: Re: [PATCH] x86_64: memcpy/memmove family optimized with AVX512 To: Florian Weimer Cc: "H.J. Lu" , libc-alpha 2016-01-16 1:24 GMT+03:00 Florian Weimer : > On 01/15/2016 10:47 PM, Andrew Senkevich wrote: >> diff --git a/sysdeps/x86_64/multiarch/memcpy_chk.S b/sysdeps/x86_64/multiarch/memcpy_chk.S >> index 6476c62..a0d56d4 100644 >> --- a/sysdeps/x86_64/multiarch/memcpy_chk.S >> +++ b/sysdeps/x86_64/multiarch/memcpy_chk.S >> @@ -30,7 +30,15 @@ >> ENTRY(__memcpy_chk) >> .type __memcpy_chk, @gnu_indirect_function >> LOAD_RTLD_GLOBAL_RO_RDX >> - leaq __memcpy_chk_sse2(%rip), %rax >> +#ifdef HAVE_AVX512_ASM_SUPPORT >> + HAS_ARCH_FEATURE (AVX512F_Usable) >> + jz 1f >> +# HAS_ARCH_FEATURE (Prefer_No_VZEROUPPER) >> +# jz 1f > > The two lines starting with “#” look somewhat out of place. Is this > really intended? Florian, thank you, I have fixed it: --- WBR, Andrew diff --git a/sysdeps/x86_64/multiarch/memcpy_chk.S b/sysdeps/x86_64/multiarch/memcpy_chk.S index a0d56d4..648217e 100644 --- a/sysdeps/x86_64/multiarch/memcpy_chk.S +++ b/sysdeps/x86_64/multiarch/memcpy_chk.S @@ -33,9 +33,9 @@ ENTRY(__memcpy_chk) #ifdef HAVE_AVX512_ASM_SUPPORT HAS_ARCH_FEATURE (AVX512F_Usable) jz 1f -# HAS_ARCH_FEATURE (Prefer_No_VZEROUPPER) -# jz 1f - leaq __memcpy_avx512_no_vzeroupper(%rip), %rax + HAS_ARCH_FEATURE (Prefer_No_VZEROUPPER) + jz 1f + leaq __memcpy_chk_avx512_no_vzeroupper(%rip), %rax ret #endif 1: leaq __memcpy_chk_sse2(%rip), %rax