From patchwork Fri Apr 21 07:34:12 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Hau Hsu X-Patchwork-Id: 68105 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 4B4D33857720 for ; Fri, 21 Apr 2023 07:34:44 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 4B4D33857720 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=sourceware.org; s=default; t=1682062484; bh=iXYYco1kbXOODD2dFTtsWq6yKtRa6g72VD/nuRu/hoo=; h=To:Cc:Subject:Date:List-Id:List-Unsubscribe:List-Archive: List-Post:List-Help:List-Subscribe:From:Reply-To:From; b=tkSWML9zvil+WcSIoKQqtcHV408Vkk04GRXGOt+hX3qu3fNIzT5M368nzlAwZaZ8Q IU++qVtkBdCtdu4AnF+5+4vTO/QeSZg1lIFnY1i7CpsGeik/S0yefRZuEy6nkQg+Wx P++Yy05UU+QKximm04Cw4DBbsGuUM4mAXAppE6ME= X-Original-To: libc-alpha@sourceware.org Delivered-To: libc-alpha@sourceware.org Received: from hhsu-imac.local (1-169-217-217.dynamic-ip.hinet.net [1.169.217.217]) by sourceware.org (Postfix) with ESMTP id E602A3858C83 for ; Fri, 21 Apr 2023 07:34:21 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org E602A3858C83 Received: by hhsu-imac.local (Postfix, from userid 501) id 419D552C5C9F; Fri, 21 Apr 2023 15:34:15 +0800 (CST) To: libc-alpha@sourceware.org, hongrong.hsu@sifive.com, jerry.shih@sifive.com, nick.knight@sifive.com, kito.cheng@sifive.com Cc: greentime.hu@sifive.com, alice.chan@sifive.com, andrew@sifive.com, vincent.chen@sifive.com, hau.hsu@sifive.com, Yun Hsiang Subject: [PATCH v2 5/5] riscv: add vectorized __memcmpeq Date: Fri, 21 Apr 2023 15:34:12 +0800 Message-Id: <20230421073412.14355-1-hau.hsu@sifive.com> X-Mailer: git-send-email 2.34.1 MIME-Version: 1.0 X-Spam-Status: No, score=-6.2 required=5.0 tests=BAYES_00, GIT_PATCH_0, HEADER_FROM_DIFFERENT_DOMAINS, KAM_DMARC_QUARANTINE, KAM_DMARC_STATUS, KAM_LAZY_DOMAIN_SECURITY, KAM_SHORT, KHOP_HELO_FCRDNS, NO_DNS_FOR_FROM, RCVD_IN_PBL, RCVD_IN_SORBS_DUL, RDNS_DYNAMIC, SPF_HELO_NONE, SPF_NONE, 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: Hau Hsu via Libc-alpha From: Hau Hsu Reply-To: Hau Hsu Errors-To: libc-alpha-bounces+patchwork=sourceware.org@sourceware.org Sender: "Libc-alpha" From: Yun Hsiang This patch proposes implementations of __memcmpeq that leverage the RISC-V V extension (RVV), version 1.0. These routines assumes VLEN is at least 32 bits, as is required by all currently defined vector extensions, and they support arbitrarily large VLEN. All implementations work for both RV32 and RV64 platforms, and make no assumptions about page size. --- sysdeps/riscv/rvv/memcmp.S | 4 --- sysdeps/riscv/rvv/memcmpeq.S | 69 ++++++++++++++++++++++++++++++++++++ 2 files changed, 69 insertions(+), 4 deletions(-) create mode 100644 sysdeps/riscv/rvv/memcmpeq.S diff --git a/sysdeps/riscv/rvv/memcmp.S b/sysdeps/riscv/rvv/memcmp.S index b156ec524c..74d8361293 100644 --- a/sysdeps/riscv/rvv/memcmp.S +++ b/sysdeps/riscv/rvv/memcmp.S @@ -69,7 +69,3 @@ L(found): END(memcmp) libc_hidden_builtin_def (memcmp) -weak_alias (memcmp,bcmp) -strong_alias (memcmp, __memcmpeq) -libc_hidden_def (__memcmpeq) - diff --git a/sysdeps/riscv/rvv/memcmpeq.S b/sysdeps/riscv/rvv/memcmpeq.S new file mode 100644 index 0000000000..302bca6992 --- /dev/null +++ b/sysdeps/riscv/rvv/memcmpeq.S @@ -0,0 +1,69 @@ +/* RVV versions memcmp. RISC-V version. + Copyright (C) 2023 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Jerry Shih , + Yun Hsiang . + + 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 +#include + + +#define iResult a0 + +#define pSrc1 a0 +#define pSrc2 a1 +#define iNum a2 + +#define iVL a3 +#define iTemp a4 + +#define ELEM_LMUL_SETTING m1 +#define vData1 v0 +#define vData2 v8 +#define vMask v16 + +ENTRY(__memcmpeq) + +L(loop): + vsetvli iVL, iNum, e8, ELEM_LMUL_SETTING, ta, ma + + vle8.v vData1, (pSrc1) + vle8.v vData2, (pSrc2) + + vmsne.vv vMask, vData1, vData2 + sub iNum, iNum, iVL + vfirst.m iTemp, vMask + + // Skip the loop if we find the different value between pSrc1 and pSrc2. + bgez iTemp, L(found) + + add pSrc1, pSrc1, iVL + add pSrc2, pSrc2, iVL + + bnez iNum, L(loop) + + li iResult, 0 + ret + +L(found): + mv iResult, iVL + ret + +END(__memcmpeq) + +weak_alias (__memcmpeq, bcmp) +libc_hidden_def (__memcmpeq)