From patchwork Fri Nov 18 20:35:23 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Noah Goldstein X-Patchwork-Id: 1705800 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=sourceware.org (client-ip=8.43.85.97; helo=sourceware.org; envelope-from=libc-alpha-bounces+incoming=patchwork.ozlabs.org@sourceware.org; receiver=) Authentication-Results: legolas.ozlabs.org; dkim=pass (1024-bit key; secure) header.d=sourceware.org header.i=@sourceware.org header.a=rsa-sha256 header.s=default header.b=BY/thJl1; dkim-atps=neutral Received: from sourceware.org (server2.sourceware.org [8.43.85.97]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (P-384) server-digest SHA384) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4NDT8f5yDqz23mg for ; Sat, 19 Nov 2022 07:36:34 +1100 (AEDT) Received: from server2.sourceware.org (localhost [IPv6:::1]) by sourceware.org (Postfix) with ESMTP id 039603852237 for ; Fri, 18 Nov 2022 20:36:32 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 039603852237 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=sourceware.org; s=default; t=1668803792; bh=cWW+2DV3dUumSjG+qKhIcI3InpxQ3MSaryoBt3Sb29Q=; h=To:Cc:Subject:Date:In-Reply-To:References:List-Id: List-Unsubscribe:List-Archive:List-Post:List-Help:List-Subscribe: From:Reply-To:From; b=BY/thJl1Ygi8UCC+aQPN5vkSEXeMx2BhQPMTLgMdEbgmPmv/bbLqjvVnRnAxQ+4HF llL+kCL1l2PSnJAuiEnv2/xUsD3wraz1qRiyekUy8rLocs4Lz7Hup9Kv7024IcKdym HQUO73nhLSIqym6m0ocqWzNAiMw345tlH3CjYgbY= X-Original-To: libc-alpha@sourceware.org Delivered-To: libc-alpha@sourceware.org Received: from mail-pl1-x62e.google.com (mail-pl1-x62e.google.com [IPv6:2607:f8b0:4864:20::62e]) by sourceware.org (Postfix) with ESMTPS id D9A673853D45 for ; Fri, 18 Nov 2022 20:35:30 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.1 sourceware.org D9A673853D45 Received: by mail-pl1-x62e.google.com with SMTP id io19so5549096plb.8 for ; Fri, 18 Nov 2022 12:35:30 -0800 (PST) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=cWW+2DV3dUumSjG+qKhIcI3InpxQ3MSaryoBt3Sb29Q=; b=p2R6ZoKovpT0me/M7bo1VyYlOJv3b/F9QHsjU80G216oUMp8qa/7g95Ug4oNeFtAMj aMAPmBDHf07fF1m4GnSfXAhIm+zgxkV2qtNU/eChcF5OI18CvILUu5xnXldsCznyIisC g30DnrP6ExFKCkbXezHEoCaWqQJHGtWnzsB3VRQEkla5byH+73sSuSBaznNDl6O8pj5d GXXvdZByGQnwj+Js883f6WdHiNaThXb65iijSqU+gM/NTPUqBP8PpeLyuT+Tl6estJB+ HIqxxl9bJ+hn+G+OSmSt/oZhaQ4LsDqSoNS967EhBRPUCY28m29tUbZpHMPqSk/ppKR3 Zdmw== X-Gm-Message-State: ANoB5pn5dsIR06icgWM+LbuMcQE1AjkguozvHFfTUcoBb2nvi7UEA0Uu eIDmwLl0YvzlNyK6nmjdx+xND+r4C98= X-Google-Smtp-Source: AA0mqf7Dt9+efBGMBzwEzjhrDnJaEWiODMSz2f+4AWLUJKreYKZyQwGXeoxZCxMPap7+aLq1YzQ7sw== X-Received: by 2002:a17:902:6505:b0:172:7a15:9a61 with SMTP id b5-20020a170902650500b001727a159a61mr1119341plk.71.1668803729128; Fri, 18 Nov 2022 12:35:29 -0800 (PST) Received: from noahgold-desk.. ([192.55.60.47]) by smtp.gmail.com with ESMTPSA id n12-20020a17090a2fcc00b0020087d7e778sm5750113pjm.37.2022.11.18.12.35.27 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 18 Nov 2022 12:35:28 -0800 (PST) To: libc-alpha@sourceware.org Cc: goldstein.w.n@gmail.com, hjl.tools@gmail.com, andrey.kolesov@intel.com, carlos@systemhalted.org Subject: [PATCH v3 2/2] x86/fpu: Cleanup code in svml_{s|d}_wrapper_impl.h Date: Fri, 18 Nov 2022 12:35:23 -0800 Message-Id: <20221118203523.1142824-2-goldstein.w.n@gmail.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20221118203523.1142824-1-goldstein.w.n@gmail.com> References: <20221118190835.1033248-1-goldstein.w.n@gmail.com> <20221118203523.1142824-1-goldstein.w.n@gmail.com> MIME-Version: 1.0 X-Spam-Status: No, score=-12.2 required=5.0 tests=BAYES_00, DKIM_SIGNED, DKIM_VALID, DKIM_VALID_AU, DKIM_VALID_EF, FREEMAIL_FROM, GIT_PATCH_0, KAM_SHORT, RCVD_IN_DNSWL_NONE, SPF_HELO_NONE, SPF_PASS, TXREP autolearn=ham autolearn_force=no version=3.4.6 X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on server2.sourceware.org X-BeenThere: libc-alpha@sourceware.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Libc-alpha mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-Patchwork-Original-From: Noah Goldstein via Libc-alpha From: Noah Goldstein Reply-To: Noah Goldstein Errors-To: libc-alpha-bounces+incoming=patchwork.ozlabs.org@sourceware.org Sender: "Libc-alpha" 1. Remove unnecessary spills. 2. Split the shared avx/avx512 wrappers to a new file. 3. Fix some small nit missed optimizations. All math and mathvec tests pass on x86. --- sysdeps/x86_64/fpu/svml_d_wrapper_impl.h | 253 ++---------------- sysdeps/x86_64/fpu/svml_s_wrapper_impl.h | 309 ++++------------------ sysdeps/x86_64/fpu/svml_sd_wrapper_impl.h | 190 +++++++++++++ 3 files changed, 266 insertions(+), 486 deletions(-) create mode 100644 sysdeps/x86_64/fpu/svml_sd_wrapper_impl.h diff --git a/sysdeps/x86_64/fpu/svml_d_wrapper_impl.h b/sysdeps/x86_64/fpu/svml_d_wrapper_impl.h index b03a2122b9..52407da8ed 100644 --- a/sysdeps/x86_64/fpu/svml_d_wrapper_impl.h +++ b/sysdeps/x86_64/fpu/svml_d_wrapper_impl.h @@ -18,39 +18,38 @@ /* SSE2 ISA version as wrapper to scalar. */ .macro WRAPPER_IMPL_SSE2 callee - subq $40, %rsp - cfi_adjust_cfa_offset (40) + subq $24, %rsp + cfi_adjust_cfa_offset (24) movaps %xmm0, (%rsp) call JUMPTARGET(\callee) - movsd %xmm0, 16(%rsp) + movsd %xmm0, (%rsp) movsd 8(%rsp), %xmm0 call JUMPTARGET(\callee) - movsd 16(%rsp), %xmm1 - movsd %xmm0, 24(%rsp) + movsd (%rsp), %xmm1 unpcklpd %xmm0, %xmm1 movaps %xmm1, %xmm0 - addq $40, %rsp - cfi_adjust_cfa_offset (-40) + addq $24, %rsp + cfi_adjust_cfa_offset (-24) ret .endm + /* 2 argument SSE2 ISA version as wrapper to scalar. */ .macro WRAPPER_IMPL_SSE2_ff callee - subq $56, %rsp - cfi_adjust_cfa_offset (56) + subq $40, %rsp + cfi_adjust_cfa_offset (40) movaps %xmm0, (%rsp) movaps %xmm1, 16(%rsp) call JUMPTARGET(\callee) - movsd %xmm0, 32(%rsp) + movsd %xmm0, (%rsp) movsd 8(%rsp), %xmm0 movsd 24(%rsp), %xmm1 call JUMPTARGET(\callee) - movsd 32(%rsp), %xmm1 - movsd %xmm0, 40(%rsp) + movsd (%rsp), %xmm1 unpcklpd %xmm0, %xmm1 movaps %xmm1, %xmm0 - addq $56, %rsp - cfi_adjust_cfa_offset (-56) + addq $40, %rsp + cfi_adjust_cfa_offset (-40) ret .endm @@ -62,229 +61,25 @@ pushq %rbx cfi_adjust_cfa_offset (8) cfi_rel_offset (%rbx, 0) + subq $24, %rsp + cfi_adjust_cfa_offset (24) + movaps %xmm0, (%rsp) movq %rdi, %rbp movq %rsi, %rbx - subq $40, %rsp - cfi_adjust_cfa_offset (40) - leaq 16(%rsp), %rsi - leaq 24(%rsp), %rdi - movaps %xmm0, (%rsp) call JUMPTARGET(\callee) - leaq 16(%rsp), %rsi - leaq 24(%rsp), %rdi - movsd 24(%rsp), %xmm0 - movapd (%rsp), %xmm1 - movsd %xmm0, 0(%rbp) - unpckhpd %xmm1, %xmm1 - movsd 16(%rsp), %xmm0 - movsd %xmm0, (%rbx) - movapd %xmm1, %xmm0 + movsd 8(%rsp), %xmm0 + leaq 8(%rbp), %rdi + leaq 8(%rbx), %rsi call JUMPTARGET(\callee) - movsd 24(%rsp), %xmm0 - movsd %xmm0, 8(%rbp) - movsd 16(%rsp), %xmm0 - movsd %xmm0, 8(%rbx) - addq $40, %rsp - cfi_adjust_cfa_offset (-40) - popq %rbx + addq $24, %rsp + cfi_adjust_cfa_offset (-24) + pop %rbx cfi_adjust_cfa_offset (-8) cfi_restore (%rbx) - popq %rbp - cfi_adjust_cfa_offset (-8) - cfi_restore (%rbp) - ret -.endm - -/* AVX/AVX2 ISA version as wrapper to SSE ISA version. */ -.macro WRAPPER_IMPL_AVX callee - pushq %rbp - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%rbp, 0) - movq %rsp, %rbp - cfi_def_cfa_register (%rbp) - andq $-32, %rsp - subq $32, %rsp - vextractf128 $1, %ymm0, (%rsp) - vzeroupper - call HIDDEN_JUMPTARGET(\callee) - vmovapd %xmm0, 16(%rsp) - vmovaps (%rsp), %xmm0 - call HIDDEN_JUMPTARGET(\callee) - vmovapd %xmm0, %xmm1 - vmovapd 16(%rsp), %xmm0 - vinsertf128 $1, %xmm1, %ymm0, %ymm0 - movq %rbp, %rsp - cfi_def_cfa_register (%rsp) - popq %rbp - cfi_adjust_cfa_offset (-8) - cfi_restore (%rbp) - ret -.endm - -/* 2 argument AVX/AVX2 ISA version as wrapper to SSE ISA version. */ -.macro WRAPPER_IMPL_AVX_ff callee - pushq %rbp - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%rbp, 0) - movq %rsp, %rbp - cfi_def_cfa_register (%rbp) - andq $-32, %rsp - subq $64, %rsp - vextractf128 $1, %ymm0, 16(%rsp) - vextractf128 $1, %ymm1, (%rsp) - vzeroupper - call HIDDEN_JUMPTARGET(\callee) - vmovaps %xmm0, 32(%rsp) - vmovaps 16(%rsp), %xmm0 - vmovaps (%rsp), %xmm1 - call HIDDEN_JUMPTARGET(\callee) - vmovaps %xmm0, %xmm1 - vmovaps 32(%rsp), %xmm0 - vinsertf128 $1, %xmm1, %ymm0, %ymm0 - movq %rbp, %rsp - cfi_def_cfa_register (%rsp) - popq %rbp + pop %rbp cfi_adjust_cfa_offset (-8) cfi_restore (%rbp) ret .endm -/* 3 argument AVX/AVX2 ISA version as wrapper to SSE ISA version. */ -.macro WRAPPER_IMPL_AVX_fFF callee - pushq %rbp - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%rbp, 0) - movq %rsp, %rbp - cfi_def_cfa_register (%rbp) - andq $-32, %rsp - pushq %r13 - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%r13, 0) - pushq %r14 - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%r14, 0) - subq $48, %rsp - movq %rsi, %r14 - movq %rdi, %r13 - vextractf128 $1, %ymm0, 32(%rsp) - vzeroupper - call HIDDEN_JUMPTARGET(\callee) - vmovaps 32(%rsp), %xmm0 - lea (%rsp), %rdi - lea 16(%rsp), %rsi - call HIDDEN_JUMPTARGET(\callee) - vmovapd (%rsp), %xmm0 - vmovapd 16(%rsp), %xmm1 - vmovapd %xmm0, 16(%r13) - vmovapd %xmm1, 16(%r14) - addq $48, %rsp - popq %r14 - cfi_adjust_cfa_offset (-8) - cfi_restore (%r14) - popq %r13 - cfi_adjust_cfa_offset (-8) - cfi_restore (%r13) - movq %rbp, %rsp - cfi_def_cfa_register (%rsp) - popq %rbp - cfi_adjust_cfa_offset (-8) - cfi_restore (%rbp) - ret -.endm - -/* AVX512 ISA version as wrapper to AVX2 ISA version. */ -.macro WRAPPER_IMPL_AVX512 callee - pushq %rbp - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%rbp, 0) - movq %rsp, %rbp - cfi_def_cfa_register (%rbp) - andq $-64, %rsp - subq $128, %rsp - vmovups %zmm0, (%rsp) - vmovupd (%rsp), %ymm0 - call HIDDEN_JUMPTARGET(\callee) - vmovupd %ymm0, 64(%rsp) - vmovupd 32(%rsp), %ymm0 - call HIDDEN_JUMPTARGET(\callee) - vmovupd %ymm0, 96(%rsp) - vmovups 64(%rsp), %zmm0 - movq %rbp, %rsp - cfi_def_cfa_register (%rsp) - popq %rbp - cfi_adjust_cfa_offset (-8) - cfi_restore (%rbp) - ret -.endm - -/* 2 argument AVX512 ISA version as wrapper to AVX2 ISA version. */ -.macro WRAPPER_IMPL_AVX512_ff callee - pushq %rbp - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%rbp, 0) - movq %rsp, %rbp - cfi_def_cfa_register (%rbp) - andq $-64, %rsp - subq $192, %rsp - vmovups %zmm0, (%rsp) - vmovups %zmm1, 64(%rsp) - vmovupd (%rsp), %ymm0 - vmovupd 64(%rsp), %ymm1 - call HIDDEN_JUMPTARGET(\callee) - vmovupd %ymm0, 128(%rsp) - vmovupd 32(%rsp), %ymm0 - vmovupd 96(%rsp), %ymm1 - call HIDDEN_JUMPTARGET(\callee) - vmovupd %ymm0, 160(%rsp) - vmovups 128(%rsp), %zmm0 - movq %rbp, %rsp - cfi_def_cfa_register (%rsp) - popq %rbp - cfi_adjust_cfa_offset (-8) - cfi_restore (%rbp) - ret -.endm - -/* 3 argument AVX512 ISA version as wrapper to AVX2 ISA version. */ -.macro WRAPPER_IMPL_AVX512_fFF callee - pushq %rbp - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%rbp, 0) - movq %rsp, %rbp - cfi_def_cfa_register (%rbp) - andq $-64, %rsp - pushq %r12 - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%r12, 0) - pushq %r13 - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%r13, 0) - subq $176, %rsp - movq %rsi, %r13 - vmovups %zmm0, (%rsp) - movq %rdi, %r12 - vmovupd (%rsp), %ymm0 - call HIDDEN_JUMPTARGET(\callee) - vmovupd 32(%rsp), %ymm0 - lea 64(%rsp), %rdi - lea 96(%rsp), %rsi - call HIDDEN_JUMPTARGET(\callee) - vmovupd 64(%rsp), %ymm0 - vmovupd 96(%rsp), %ymm1 - vmovupd %ymm0, 32(%r12) - vmovupd %ymm1, 32(%r13) - vzeroupper - addq $176, %rsp - popq %r13 - cfi_adjust_cfa_offset (-8) - cfi_restore (%r13) - popq %r12 - cfi_adjust_cfa_offset (-8) - cfi_restore (%r12) - movq %rbp, %rsp - cfi_def_cfa_register (%rsp) - popq %rbp - cfi_adjust_cfa_offset (-8) - cfi_restore (%rbp) - ret -.endm +#include "svml_sd_wrapper_impl.h" diff --git a/sysdeps/x86_64/fpu/svml_s_wrapper_impl.h b/sysdeps/x86_64/fpu/svml_s_wrapper_impl.h index cecf6c8384..d9266563ef 100644 --- a/sysdeps/x86_64/fpu/svml_s_wrapper_impl.h +++ b/sysdeps/x86_64/fpu/svml_s_wrapper_impl.h @@ -18,61 +18,66 @@ /* SSE2 ISA version as wrapper to scalar. */ .macro WRAPPER_IMPL_SSE2 callee - subq $40, %rsp - cfi_adjust_cfa_offset (40) + push %rbx + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbx, 0) + subq $16, %rsp + cfi_adjust_cfa_offset (16) movaps %xmm0, (%rsp) call JUMPTARGET(\callee) - movss %xmm0, 16(%rsp) + movss %xmm0, (%rsp) movss 4(%rsp), %xmm0 call JUMPTARGET(\callee) - movss %xmm0, 20(%rsp) + movss %xmm0, 4(%rsp) movss 8(%rsp), %xmm0 call JUMPTARGET(\callee) - movss %xmm0, 24(%rsp) + movd %xmm0, %ebx movss 12(%rsp), %xmm0 call JUMPTARGET(\callee) - movss 16(%rsp), %xmm3 - movss 20(%rsp), %xmm2 - movss 24(%rsp), %xmm1 - movss %xmm0, 28(%rsp) - unpcklps %xmm1, %xmm3 - unpcklps %xmm0, %xmm2 - unpcklps %xmm2, %xmm3 - movaps %xmm3, %xmm0 - addq $40, %rsp - cfi_adjust_cfa_offset (-40) + movd %ebx, %xmm1 + unpcklps %xmm0, %xmm1 + movsd (%rsp), %xmm0 + unpcklpd %xmm1, %xmm0 + addq $16, %rsp + cfi_adjust_cfa_offset (-16) + pop %rbx + cfi_adjust_cfa_offset (-8) + cfi_restore (%rbx) ret .endm /* 2 argument SSE2 ISA version as wrapper to scalar. */ .macro WRAPPER_IMPL_SSE2_ff callee - subq $56, %rsp - cfi_adjust_cfa_offset (56) + push %rbx + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbx, 0) + subq $32, %rsp + cfi_adjust_cfa_offset (40) movaps %xmm0, (%rsp) movaps %xmm1, 16(%rsp) call JUMPTARGET(\callee) - movss %xmm0, 32(%rsp) - movss 4(%rsp), %xmm0 movss 20(%rsp), %xmm1 + movss %xmm0, 0(%rsp) + movss 4(%rsp), %xmm0 call JUMPTARGET(\callee) - movss %xmm0, 36(%rsp) - movss 8(%rsp), %xmm0 movss 24(%rsp), %xmm1 + movss %xmm0, 4(%rsp) + movss 8(%rsp), %xmm0 call JUMPTARGET(\callee) - movss %xmm0, 40(%rsp) - movss 12(%rsp), %xmm0 movss 28(%rsp), %xmm1 + movd %xmm0, %ebx + movss 12(%rsp), %xmm0 call JUMPTARGET(\callee) - movss 32(%rsp), %xmm3 - movss 36(%rsp), %xmm2 - movss 40(%rsp), %xmm1 - movss %xmm0, 44(%rsp) - unpcklps %xmm1, %xmm3 - unpcklps %xmm0, %xmm2 - unpcklps %xmm2, %xmm3 - movaps %xmm3, %xmm0 - addq $56, %rsp - cfi_adjust_cfa_offset (-56) + /* merge 4x results into xmm0. */ + movd %ebx, %xmm1 + unpcklps %xmm0, %xmm1 + movsd (%rsp), %xmm0 + unpcklpd %xmm1, %xmm0 + addq $32, %rsp + cfi_adjust_cfa_offset (-32) + pop %rbx + cfi_adjust_cfa_offset (-8) + cfi_restore (%rbx) ret .endm @@ -86,48 +91,24 @@ cfi_rel_offset (%rbx, 0) movq %rdi, %rbp movq %rsi, %rbx - subq $40, %rsp - cfi_adjust_cfa_offset (40) - leaq 24(%rsp), %rsi - leaq 28(%rsp), %rdi + subq $24, %rsp + cfi_adjust_cfa_offset (24) movaps %xmm0, (%rsp) call JUMPTARGET(\callee) - leaq 24(%rsp), %rsi - leaq 28(%rsp), %rdi - movss 28(%rsp), %xmm0 - movss %xmm0, 0(%rbp) - movaps (%rsp), %xmm1 - movss 24(%rsp), %xmm0 - movss %xmm0, (%rbx) - movaps %xmm1, %xmm0 - shufps $85, %xmm1, %xmm0 + movss 4(%rsp), %xmm0 + leaq 4(%rbp), %rdi + leaq 4(%rbx), %rsi call JUMPTARGET(\callee) - movss 28(%rsp), %xmm0 - leaq 24(%rsp), %rsi - movss %xmm0, 4(%rbp) - leaq 28(%rsp), %rdi - movaps (%rsp), %xmm1 - movss 24(%rsp), %xmm0 - movss %xmm0, 4(%rbx) - movaps %xmm1, %xmm0 - unpckhps %xmm1, %xmm0 + movss 8(%rsp), %xmm0 + leaq 8(%rbp), %rdi + leaq 8(%rbx), %rsi call JUMPTARGET(\callee) - movaps (%rsp), %xmm1 - leaq 24(%rsp), %rsi - leaq 28(%rsp), %rdi - movss 28(%rsp), %xmm0 - shufps $255, %xmm1, %xmm1 - movss %xmm0, 8(%rbp) - movss 24(%rsp), %xmm0 - movss %xmm0, 8(%rbx) - movaps %xmm1, %xmm0 + movss 12(%rsp), %xmm0 + leaq 12(%rbp), %rdi + leaq 12(%rbx), %rsi call JUMPTARGET(\callee) - movss 28(%rsp), %xmm0 - movss %xmm0, 12(%rbp) - movss 24(%rsp), %xmm0 - movss %xmm0, 12(%rbx) - addq $40, %rsp - cfi_adjust_cfa_offset (-40) + addq $24, %rsp + cfi_adjust_cfa_offset (-24) popq %rbx cfi_adjust_cfa_offset (-8) cfi_restore (%rbx) @@ -137,190 +118,4 @@ ret .endm -/* AVX/AVX2 ISA version as wrapper to SSE ISA version. */ -.macro WRAPPER_IMPL_AVX callee - pushq %rbp - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%rbp, 0) - movq %rsp, %rbp - cfi_def_cfa_register (%rbp) - andq $-32, %rsp - subq $32, %rsp - vextractf128 $1, %ymm0, (%rsp) - vzeroupper - call HIDDEN_JUMPTARGET(\callee) - vmovaps %xmm0, 16(%rsp) - vmovaps (%rsp), %xmm0 - call HIDDEN_JUMPTARGET(\callee) - vmovaps %xmm0, %xmm1 - vmovaps 16(%rsp), %xmm0 - vinsertf128 $1, %xmm1, %ymm0, %ymm0 - movq %rbp, %rsp - cfi_def_cfa_register (%rsp) - popq %rbp - cfi_adjust_cfa_offset (-8) - cfi_restore (%rbp) - ret -.endm - -/* 2 argument AVX/AVX2 ISA version as wrapper to SSE ISA version. */ -.macro WRAPPER_IMPL_AVX_ff callee - pushq %rbp - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%rbp, 0) - movq %rsp, %rbp - cfi_def_cfa_register (%rbp) - andq $-32, %rsp - subq $64, %rsp - vextractf128 $1, %ymm0, 16(%rsp) - vextractf128 $1, %ymm1, (%rsp) - vzeroupper - call HIDDEN_JUMPTARGET(\callee) - vmovaps %xmm0, 32(%rsp) - vmovaps 16(%rsp), %xmm0 - vmovaps (%rsp), %xmm1 - call HIDDEN_JUMPTARGET(\callee) - vmovaps %xmm0, %xmm1 - vmovaps 32(%rsp), %xmm0 - vinsertf128 $1, %xmm1, %ymm0, %ymm0 - movq %rbp, %rsp - cfi_def_cfa_register (%rsp) - popq %rbp - cfi_adjust_cfa_offset (-8) - cfi_restore (%rbp) - ret -.endm - -/* 3 argument AVX/AVX2 ISA version as wrapper to SSE ISA version. */ -.macro WRAPPER_IMPL_AVX_fFF callee - pushq %rbp - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%rbp, 0) - movq %rsp, %rbp - cfi_def_cfa_register (%rbp) - andq $-32, %rsp - pushq %r13 - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%r13, 0) - pushq %r14 - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%r14, 0) - subq $48, %rsp - movq %rsi, %r14 - vmovaps %ymm0, (%rsp) - movq %rdi, %r13 - vmovaps 16(%rsp), %xmm1 - vmovaps %xmm1, 32(%rsp) - vzeroupper - vmovaps (%rsp), %xmm0 - call HIDDEN_JUMPTARGET(\callee) - vmovaps 32(%rsp), %xmm0 - lea (%rsp), %rdi - lea 16(%rsp), %rsi - call HIDDEN_JUMPTARGET(\callee) - vmovaps (%rsp), %xmm0 - vmovaps 16(%rsp), %xmm1 - vmovaps %xmm0, 16(%r13) - vmovaps %xmm1, 16(%r14) - addq $48, %rsp - popq %r14 - cfi_adjust_cfa_offset (-8) - cfi_restore (%r14) - popq %r13 - cfi_adjust_cfa_offset (-8) - cfi_restore (%r13) - movq %rbp, %rsp - cfi_def_cfa_register (%rsp) - popq %rbp - cfi_adjust_cfa_offset (-8) - cfi_restore (%rbp) - ret -.endm - -/* AVX512 ISA version as wrapper to AVX2 ISA version. */ -.macro WRAPPER_IMPL_AVX512 callee - pushq %rbp - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%rbp, 0) - movq %rsp, %rbp - cfi_def_cfa_register (%rbp) - andq $-64, %rsp - subq $128, %rsp - vmovups %zmm0, (%rsp) - vmovupd (%rsp), %ymm0 - call HIDDEN_JUMPTARGET(\callee) - vmovupd %ymm0, 64(%rsp) - vmovupd 32(%rsp), %ymm0 - call HIDDEN_JUMPTARGET(\callee) - vmovupd %ymm0, 96(%rsp) - vmovups 64(%rsp), %zmm0 - movq %rbp, %rsp - cfi_def_cfa_register (%rsp) - popq %rbp - cfi_adjust_cfa_offset (-8) - cfi_restore (%rbp) - ret -.endm - -/* 2 argument AVX512 ISA version as wrapper to AVX2 ISA version. */ -.macro WRAPPER_IMPL_AVX512_ff callee - pushq %rbp - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%rbp, 0) - movq %rsp, %rbp - cfi_def_cfa_register (%rbp) - andq $-64, %rsp - subq $192, %rsp - vmovups %zmm0, (%rsp) - vmovups %zmm1, 64(%rsp) - vmovups (%rsp), %ymm0 - vmovups 64(%rsp), %ymm1 - call HIDDEN_JUMPTARGET(\callee) - vmovups %ymm0, 128(%rsp) - vmovups 32(%rsp), %ymm0 - vmovups 96(%rsp), %ymm1 - call HIDDEN_JUMPTARGET(\callee) - vmovups %ymm0, 160(%rsp) - vmovups 128(%rsp), %zmm0 - movq %rbp, %rsp - cfi_def_cfa_register (%rsp) - popq %rbp - cfi_adjust_cfa_offset (-8) - cfi_restore (%rbp) - ret -.endm - -/* 3 argument AVX512 ISA version as wrapper to AVX2 ISA version. */ -.macro WRAPPER_IMPL_AVX512_fFF callee - pushq %rbp - cfi_adjust_cfa_offset (8) - cfi_rel_offset (%rbp, 0) - movq %rsp, %rbp - cfi_def_cfa_register (%rbp) - andq $-64, %rsp - pushq %r12 - pushq %r13 - subq $176, %rsp - movq %rsi, %r13 - vmovaps %zmm0, (%rsp) - movq %rdi, %r12 - vmovaps (%rsp), %ymm0 - call HIDDEN_JUMPTARGET(\callee) - vmovaps 32(%rsp), %ymm0 - lea 64(%rsp), %rdi - lea 96(%rsp), %rsi - call HIDDEN_JUMPTARGET(\callee) - vmovaps 64(%rsp), %ymm0 - vmovaps 96(%rsp), %ymm1 - vmovaps %ymm0, 32(%r12) - vmovaps %ymm1, 32(%r13) - addq $176, %rsp - popq %r13 - popq %r12 - movq %rbp, %rsp - cfi_def_cfa_register (%rsp) - popq %rbp - cfi_adjust_cfa_offset (-8) - cfi_restore (%rbp) - ret -.endm +#include "svml_sd_wrapper_impl.h" diff --git a/sysdeps/x86_64/fpu/svml_sd_wrapper_impl.h b/sysdeps/x86_64/fpu/svml_sd_wrapper_impl.h new file mode 100644 index 0000000000..bd934ad578 --- /dev/null +++ b/sysdeps/x86_64/fpu/svml_sd_wrapper_impl.h @@ -0,0 +1,190 @@ +/* Common float/double wrapper implementations of vector math + functions. + Copyright (C) 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 + . */ + +/* AVX/AVX2 ISA version as wrapper to SSE ISA version. */ +.macro WRAPPER_IMPL_AVX callee + pushq %rbp + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbp, 0) + movq %rsp, %rbp + cfi_def_cfa_register (%rbp) + andq $-32, %rsp + subq $32, %rsp + vmovaps %ymm0, (%rsp) + vzeroupper + call HIDDEN_JUMPTARGET(\callee) + vmovaps %xmm0, (%rsp) + vmovaps 16(%rsp), %xmm0 + call HIDDEN_JUMPTARGET(\callee) + /* combine xmm0 (return of second call) with result of first + call (saved on stack). Might be worth exploring logic that + uses `vpblend` and reads in ymm1 using -16(rsp). */ + vmovaps (%rsp), %xmm1 + vinsertf128 $1, %xmm0, %ymm1, %ymm0 + movq %rbp, %rsp + cfi_def_cfa_register (%rsp) + popq %rbp + cfi_adjust_cfa_offset (-8) + cfi_restore (%rbp) + ret +.endm + +/* 2 argument AVX/AVX2 ISA version as wrapper to SSE ISA version. */ +.macro WRAPPER_IMPL_AVX_ff callee + pushq %rbp + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbp, 0) + movq %rsp, %rbp + cfi_def_cfa_register (%rbp) + andq $-32, %rsp + subq $64, %rsp + vmovaps %ymm0, (%rsp) + vmovaps %ymm1, 32(%rsp) + vzeroupper + call HIDDEN_JUMPTARGET(\callee) + vmovaps 48(%rsp), %xmm1 + vmovaps %xmm0, (%rsp) + vmovaps 16(%rsp), %xmm0 + call HIDDEN_JUMPTARGET(\callee) + /* combine xmm0 (return of second call) with result of first + call (saved on stack). Might be worth exploring logic that + uses `vpblend` and reads in ymm1 using -16(rsp). */ + vmovaps (%rsp), %xmm1 + vinsertf128 $1, %xmm0, %ymm1, %ymm0 + movq %rbp, %rsp + cfi_def_cfa_register (%rsp) + popq %rbp + cfi_adjust_cfa_offset (-8) + cfi_restore (%rbp) + ret +.endm + +/* 3 argument AVX/AVX2 ISA version as wrapper to SSE ISA version. */ +.macro WRAPPER_IMPL_AVX_fFF callee + pushq %rbp + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbp, 0) + movq %rsp, %rbp + andq $-32, %rsp + subq $32, %rsp + vmovaps %ymm0, (%rsp) + pushq %rbx + pushq %r14 + movq %rdi, %rbx + movq %rsi, %r14 + vzeroupper + call HIDDEN_JUMPTARGET(\callee) + vmovaps 32(%rsp), %xmm0 + leaq 16(%rbx), %rdi + leaq 16(%r14), %rsi + call HIDDEN_JUMPTARGET(\callee) + popq %r14 + popq %rbx + movq %rbp, %rsp + cfi_def_cfa_register (%rsp) + popq %rbp + cfi_adjust_cfa_offset (-8) + cfi_restore (%rbp) + ret +.endm + +/* AVX512 ISA version as wrapper to AVX2 ISA version. */ +.macro WRAPPER_IMPL_AVX512 callee + pushq %rbp + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbp, 0) + movq %rsp, %rbp + cfi_def_cfa_register (%rbp) + andq $-64, %rsp + subq $64, %rsp + vmovups %zmm0, (%rsp) + call HIDDEN_JUMPTARGET(\callee) + vmovupd %ymm0, (%rsp) + vmovupd 32(%rsp), %ymm0 + call HIDDEN_JUMPTARGET(\callee) + /* combine ymm0 (return of second call) with result of first + call (saved on stack). */ + vmovaps (%rsp), %ymm1 + vinserti64x4 $0x1, %ymm0, %zmm1, %zmm0 + movq %rbp, %rsp + cfi_def_cfa_register (%rsp) + popq %rbp + cfi_adjust_cfa_offset (-8) + cfi_restore (%rbp) + ret +.endm + +/* 2 argument AVX512 ISA version as wrapper to AVX2 ISA version. */ +.macro WRAPPER_IMPL_AVX512_ff callee + pushq %rbp + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbp, 0) + movq %rsp, %rbp + cfi_def_cfa_register (%rbp) + andq $-64, %rsp + addq $-128, %rsp + vmovups %zmm0, (%rsp) + vmovups %zmm1, 64(%rsp) + /* ymm0 and ymm1 are already set. */ + call HIDDEN_JUMPTARGET(\callee) + vmovups 96(%rsp), %ymm1 + vmovaps %ymm0, (%rsp) + vmovups 32(%rsp), %ymm0 + call HIDDEN_JUMPTARGET(\callee) + /* combine ymm0 (return of second call) with result of first + call (saved on stack). */ + vmovaps (%rsp), %ymm1 + vinserti64x4 $0x1, %ymm0, %zmm1, %zmm0 + movq %rbp, %rsp + cfi_def_cfa_register (%rsp) + popq %rbp + cfi_adjust_cfa_offset (-8) + cfi_restore (%rbp) + ret +.endm + +/* 3 argument AVX512 ISA version as wrapper to AVX2 ISA version. */ +.macro WRAPPER_IMPL_AVX512_fFF callee + pushq %rbp + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbp, 0) + movq %rsp, %rbp + cfi_def_cfa_register (%rbp) + andq $-64, %rsp + subq $64, %rsp + vmovaps %zmm0, (%rsp) + pushq %rbx + pushq %r14 + movq %rdi, %rbx + movq %rsi, %r14 + /* ymm0 is already set. */ + call HIDDEN_JUMPTARGET(\callee) + vmovaps 48(%rsp), %ymm0 + leaq 32(%rbx), %rdi + leaq 32(%r14), %rsi + call HIDDEN_JUMPTARGET(\callee) + popq %r14 + popq %rbx + movq %rbp, %rsp + cfi_def_cfa_register (%rsp) + popq %rbp + cfi_adjust_cfa_offset (-8) + cfi_restore (%rbp) + ret +.endm