From patchwork Fri Sep 2 20:39:35 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Adhemerval Zanella Netto X-Patchwork-Id: 57298 Return-Path: X-Original-To: patchwork@sourceware.org Delivered-To: patchwork@sourceware.org Received: from server2.sourceware.org (localhost [IPv6:::1]) by sourceware.org (Postfix) with ESMTP id AE2E6388553A for ; Fri, 2 Sep 2022 20:41:44 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org AE2E6388553A DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=sourceware.org; s=default; t=1662151304; bh=yaswVVzodwomNg7OJoqsjfX9xQUObSy00raw0PfY2Ys=; h=To:Subject:Date:In-Reply-To:References:List-Id:List-Unsubscribe: List-Archive:List-Post:List-Help:List-Subscribe:From:Reply-To:Cc: From; b=TZPXWyc6Adf86Vip8OXOerVq6BQEZOZ5Jd9EVEyNtRBHyTxerZmU45vIMvi1XEH9z dR/mtCPCJ5P78lc1Sa596K3gTYO97KvY+f0YgImeRVkfFaCGdTCk/uLw3MCl1gtj5c HhAKoAPHZRw3SPsE+3BhpkCTKkulOZyouvnJWHxQ= X-Original-To: libc-alpha@sourceware.org Delivered-To: libc-alpha@sourceware.org Received: from mail-oa1-x31.google.com (mail-oa1-x31.google.com [IPv6:2001:4860:4864:20::31]) by sourceware.org (Postfix) with ESMTPS id 148B13857349 for ; Fri, 2 Sep 2022 20:40:06 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.1 sourceware.org 148B13857349 Received: by mail-oa1-x31.google.com with SMTP id 586e51a60fabf-11ee4649dfcso7673160fac.1 for ; Fri, 02 Sep 2022 13:40:06 -0700 (PDT) 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; bh=yaswVVzodwomNg7OJoqsjfX9xQUObSy00raw0PfY2Ys=; b=I6o5dG6UK3BDXN+v3FmePCn3bx2r58gOFlKhPXhU4dPmnHj7loySPNFLnonNSI9eiE 9aiq9/20Le1SOhAM8PCcIocY1BMORVLTu4wmzsai4DRQXkiKsDHNcRaS7PlV/87L2wZX wLueYsKr1qDGyOigoDynjHAY5fAD4RfGRK6kG3cK/o47LoYbbmjKWmswJ3IYmDw7HlqT w6JqhH2nH+J/lF0DN/NEquqeQ65sUvQH0rnBEt/4pF6sGNqjsgTnoVHqyiAaHnMJeN+D cMEd73K0zlrvzFnMgXI4aQuWyyGPCdzFbi4V8sVvelJ1/hYJH3vTR4XuryiiDS55Xl9f Odsg== X-Gm-Message-State: ACgBeo3bHOX+HVKJ4moHz8dgUmDbJUwf+yzB8DtW2mksEB5Lf0b4jjSo NIzhpsk628RmfLOvacyFZF0R2hHLe5IjrA== X-Google-Smtp-Source: AA6agR4cUM7K+ALBU6QEXWaXQcZI9xJ2duXPsrN0qTek2oWDLquIXcvTuBhlgbsrYbyoCkR7W99mwg== X-Received: by 2002:a05:6870:b00f:b0:11f:b1e:6142 with SMTP id y15-20020a056870b00f00b0011f0b1e6142mr3077773oae.234.1662151205632; Fri, 02 Sep 2022 13:40:05 -0700 (PDT) Received: from mandiga.. ([2804:1b3:a7c0:dfed:72d1:946c:4eb1:ce5f]) by smtp.gmail.com with ESMTPSA id i27-20020a4a6f5b000000b00425678b9c4bsm1107902oof.0.2022.09.02.13.40.04 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 02 Sep 2022 13:40:05 -0700 (PDT) To: libc-alpha@sourceware.org Subject: [PATCH 12/17] hppa: Add memcopy.h Date: Fri, 2 Sep 2022 17:39:35 -0300 Message-Id: <20220902203940.2385967-13-adhemerval.zanella@linaro.org> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20220902203940.2385967-1-adhemerval.zanella@linaro.org> References: <20220902203940.2385967-1-adhemerval.zanella@linaro.org> MIME-Version: 1.0 X-Spam-Status: No, score=-12.7 required=5.0 tests=BAYES_00, DKIM_SIGNED, DKIM_VALID, DKIM_VALID_AU, DKIM_VALID_EF, GIT_PATCH_0, KAM_SHORT, RCVD_IN_DNSWL_NONE, SPF_HELO_NONE, SPF_PASS, TXREP, T_SCC_BODY_TEXT_LINE 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: Adhemerval Zanella via Libc-alpha From: Adhemerval Zanella Netto Reply-To: Adhemerval Zanella Cc: Richard Henderson , Joseph Myers , caiyinyu Errors-To: libc-alpha-bounces+patchwork=sourceware.org@sourceware.org Sender: "Libc-alpha" From: Richard Henderson GCC's combine pass cannot merge (x >> c | y << (32 - c)) into a double-word shift unless (1) the subtract is in the same basic block and (2) the result of the subtract is used exactly once. Neither condition is true for any use of MERGE. By forcing the use of a double-word shift, we not only reduce contention on SAR, but also allow the setting of SAR to be hoisted outside of a loop. Checked on hppa-linux-gnu. --- sysdeps/hppa/memcopy.h | 42 ++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 42 insertions(+) create mode 100644 sysdeps/hppa/memcopy.h diff --git a/sysdeps/hppa/memcopy.h b/sysdeps/hppa/memcopy.h new file mode 100644 index 0000000000..288b5e9520 --- /dev/null +++ b/sysdeps/hppa/memcopy.h @@ -0,0 +1,42 @@ +/* Definitions for memory copy functions, PA-RISC version. + 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 + . */ + +#include + +/* Use a single double-word shift instead of two shifts and an ior. + If the uses of MERGE were close to the computation of shl/shr, + the compiler might have been able to create this itself. + But instead that computation is well separated. + + Using an inline function instead of a macro is the easiest way + to ensure that the types are correct. */ + +#undef MERGE + +static inline op_t +MERGE (op_t w0, int shl, op_t w1, int shr) +{ + _Static_assert (OPSIZ == 4 || OPSIZ == 8, "Invalid OPSIZE"); + + op_t res; + if (OPSIZ == 4) + asm ("shrpw %1,%2,%%sar,%0" : "=r"(res) : "r"(w0), "r"(w1), "q"(shr)); + else if (OPSIZ == 8) + asm ("shrpd %1,%2,%%sar,%0" : "=r"(res) : "r"(w0), "r"(w1), "q"(shr)); + return res; +}