From patchwork Mon Jun 26 18:58:27 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Robin Dapp X-Patchwork-Id: 113089 Return-Path: Delivered-To: ouuuleilei@gmail.com Received: by 2002:a59:994d:0:b0:3d9:f83d:47d9 with SMTP id k13csp7690319vqr; Mon, 26 Jun 2023 12:00:11 -0700 (PDT) X-Google-Smtp-Source: ACHHUZ5+79VFihnemMaGRw1JWjvNMWpWn+WWstFCYQqqksWOjms7X883v5HTYV2YN26mBZpTzthj X-Received: by 2002:a17:907:a03:b0:982:2278:bcef with SMTP id bb3-20020a1709070a0300b009822278bcefmr28335735ejc.60.1687806011509; Mon, 26 Jun 2023 12:00:11 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1687806011; cv=none; d=google.com; s=arc-20160816; b=ZtoZMhVSNS+driGS/IzpBk63Rpbjho0Nq32InYojLYwb+PTg+E67TosZZl4vlke1E9 jN/jVCoR8nj/HSGKWI9Jk15pz2PJjhapfYvsmb9Qh4pVhuA5bqUc02MyZwzuAozILUPd GsbHVmRpk1xrLjFrnS4yQs191FGc1xDQRGuy3GoENQtytV8Py2hcKPqoWHmJMmyd8Zlm y0063R6A840ywu47bSVGgxACpYFEO7FuAMSTH/pzlGHFi603FKR4WLe3YiOHTSdEYnvI M8EoDY343HtxR0cK9C1IM/oZjIFo580Hx+qEK+5D+Ug+Mx2kA1S+RRX1B6Q0SKbGbrmN RT6g== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=sender:errors-to:reply-to:from:list-subscribe:list-help:list-post :list-archive:list-unsubscribe:list-id:precedence :content-transfer-encoding:to:subject:content-language:cc:user-agent :mime-version:date:message-id:dmarc-filter:delivered-to :dkim-signature:dkim-filter; bh=WogPw3SvqbnLWv85JbSDBOeODRbW84nFpnVnaSTZ/l4=; fh=0vXgoIsEThH6Qba7saR4QWpKnBkTEEL72tevRHKcYxQ=; b=wgIer+Sb6944vDu1wsggX6hvunhlzkDM1SEjSZyffPkv2W91/GX7vO2iuXo37sjoGx Nm3BdCuG6AxrOKfA5WgApbPoQVrJtYfex1yPQyj3bcEq4hohVA5Qx31dsl/OsRZ01GGL /FA5hJQzE4GGQxKNtnW8uMf3cEUibnhozOK8UqfNGtwrafgjjTmA31EtuP08+0ufA9ri jAYEmBC8ig8jDRb7Zg6IW3HRPkRveT9GMouAGe3WRReRYvm24Mjgj+F4EQbCQCCdz0cg jPXwBZYDcUOWSRgZrRLhgBzq+Wqa8MpFeBwE1qYYyZx9E8hkHeJmh+DrIOngGFk/DIvI kldg== ARC-Authentication-Results: i=1; mx.google.com; dkim=pass header.i=@gcc.gnu.org header.s=default header.b=JKAiofxL; spf=pass (google.com: domain of gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org designates 2620:52:3:1:0:246e:9693:128c as permitted sender) smtp.mailfrom="gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org"; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=gnu.org Received: from sourceware.org (server2.sourceware.org. [2620:52:3:1:0:246e:9693:128c]) by mx.google.com with ESMTPS id xa10-20020a170907b9ca00b009787b15aa51si3120732ejc.713.2023.06.26.12.00.11 for (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Mon, 26 Jun 2023 12:00:11 -0700 (PDT) Received-SPF: pass (google.com: domain of gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org designates 2620:52:3:1:0:246e:9693:128c as permitted sender) client-ip=2620:52:3:1:0:246e:9693:128c; Authentication-Results: mx.google.com; dkim=pass header.i=@gcc.gnu.org header.s=default header.b=JKAiofxL; spf=pass (google.com: domain of gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org designates 2620:52:3:1:0:246e:9693:128c as permitted sender) smtp.mailfrom="gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org"; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=gnu.org Received: from server2.sourceware.org (localhost [IPv6:::1]) by sourceware.org (Postfix) with ESMTP id 641563857353 for ; Mon, 26 Jun 2023 18:59:29 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 641563857353 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org; s=default; t=1687805969; bh=WogPw3SvqbnLWv85JbSDBOeODRbW84nFpnVnaSTZ/l4=; h=Date:Cc:Subject:To:List-Id:List-Unsubscribe:List-Archive: List-Post:List-Help:List-Subscribe:From:Reply-To:From; b=JKAiofxLBPgiJiCBUcDoYYl5SfDqN74xK8ouhFz2HK25zHdTteLZl0hsZPuMud3XW hDx6kaglI3YP3Xqdi0eimalmzQxuyigTWErxkaOoNZW7b9TYN6qTfWK3Yob5VPO29Y GZ18k33gTLnTkn2uDpMTZEEcXu4NwydOOCcxeQNE= X-Original-To: gcc-patches@gcc.gnu.org Delivered-To: gcc-patches@gcc.gnu.org Received: from mail-wm1-x32f.google.com (mail-wm1-x32f.google.com [IPv6:2a00:1450:4864:20::32f]) by sourceware.org (Postfix) with ESMTPS id 1837B3858033 for ; Mon, 26 Jun 2023 18:58:31 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org 1837B3858033 Received: by mail-wm1-x32f.google.com with SMTP id 5b1f17b1804b1-3f9b4bf99c2so47150385e9.3 for ; Mon, 26 Jun 2023 11:58:31 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20221208; t=1687805909; x=1690397909; h=content-transfer-encoding:to:subject:from:content-language:cc :user-agent:mime-version:date:message-id:x-gm-message-state:from:to :cc:subject:date:message-id:reply-to; bh=WogPw3SvqbnLWv85JbSDBOeODRbW84nFpnVnaSTZ/l4=; b=L5UrmPZ255Se7XOktYwYILUg/xl+15r6pZ+9dLCZMd4u+JLHWeJNUPF9ymGpmTh4dh rPXR7EQpohKq5/AIoVltot0O6o/03SEd5rQRTxm++6O2Lc88IaC5aIt1RqqdDIpdCnsU ZmQw4iZGSlccpbCXlLI7yVdGD0p6SQ/WI+jh84CaeFnev6KKypMxSohvI/uiTMJgrVjj 28EAAub66H89dAEWh7KPpvCGNOmCSSrCb5Kt9sPYBFW5u1PbqPY3cK08Xqqbj9CvBpve vftOT/nibimv3VzT49Ioon3FZkxq4nbsTR4v6RzrxtnXqRSS4UA9In5uTnmNe0i+Oxaa f1/w== X-Gm-Message-State: AC+VfDyD7VnmhyS/LurlV5rWCQBblVStujw56ngaZelwk/vOc9USNdmN k/hTWB1DLtHWuoSivsS2SFMYN6UMqwA= X-Received: by 2002:adf:e549:0:b0:313:e9f6:3378 with SMTP id z9-20020adfe549000000b00313e9f63378mr4925330wrm.4.1687805908946; Mon, 26 Jun 2023 11:58:28 -0700 (PDT) Received: from [192.168.1.24] (ip-046-005-130-086.um12.pools.vodafone-ip.de. [46.5.130.86]) by smtp.gmail.com with ESMTPSA id z3-20020a5d4c83000000b0030497b3224bsm8132073wrs.64.2023.06.26.11.58.27 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Mon, 26 Jun 2023 11:58:28 -0700 (PDT) Message-ID: Date: Mon, 26 Jun 2023 20:58:27 +0200 MIME-Version: 1.0 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101 Thunderbird/102.12.0 Cc: rdapp.gcc@gmail.com Content-Language: en-US Subject: [PATCH] RISC-V: Add autovec FP int->float conversion. To: gcc-patches , palmer , Kito Cheng , "juzhe.zhong@rivai.ai" , jeffreyalaw X-Spam-Status: No, score=-8.9 required=5.0 tests=BAYES_00, DKIM_SIGNED, DKIM_VALID, DKIM_VALID_AU, DKIM_VALID_EF, FREEMAIL_FROM, GIT_PATCH_0, KAM_ASCII_DIVIDERS, 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: gcc-patches@gcc.gnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Gcc-patches mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-Patchwork-Original-From: Robin Dapp via Gcc-patches From: Robin Dapp Reply-To: Robin Dapp Errors-To: gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org Sender: "Gcc-patches" X-getmail-retrieved-from-mailbox: =?utf-8?q?INBOX?= X-GMAIL-THRID: =?utf-8?q?1769792876396396293?= X-GMAIL-MSGID: =?utf-8?q?1769792876396396293?= Hi, this patch adds the autovec expander for vfcvt.f.x.v and tests for it. In addition, it modifies the zfhmin-1 test so it doesn't scan for "no vectorization" but rather check that we do not emit any (RTL) vector operations (other than float/float conversions) with a VNx..HFmode. Regards Robin gcc/ChangeLog: * config/riscv/autovec.md (2): New expander. gcc/testsuite/ChangeLog: * gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-run.c: Adjust. * gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-rv32gcv.c: Dito. * gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-rv64gcv.c: Dito. * gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-template.h: Dito. * gcc.target/riscv/rvv/autovec/zvfhmin-1.c: Add int/float conversions. * gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-run.c: New test. * gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-rv32gcv.c: New test. * gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-rv64gcv.c: New test. * gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-template.h: New test. * gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-zvfh-run.c: New test. --- gcc/config/riscv/autovec.md | 19 ++++ .../rvv/autovec/conversions/vfcvt-itof-run.c | 96 +++++++++++++++++++ .../autovec/conversions/vfcvt-itof-rv32gcv.c | 7 ++ .../autovec/conversions/vfcvt-itof-rv64gcv.c | 7 ++ .../autovec/conversions/vfcvt-itof-template.h | 20 ++++ .../autovec/conversions/vfcvt-itof-zvfh-run.c | 64 +++++++++++++ .../rvv/autovec/conversions/vfcvt_rtz-run.c | 44 +++++++++ .../autovec/conversions/vfcvt_rtz-rv32gcv.c | 5 +- .../autovec/conversions/vfcvt_rtz-rv64gcv.c | 5 +- .../autovec/conversions/vfcvt_rtz-template.h | 6 +- .../autovec/conversions/vfcvt_rtz-zvfh-run.c | 64 +++++++++++++ .../gcc.target/riscv/rvv/autovec/zvfhmin-1.c | 48 ++++++++-- 12 files changed, 374 insertions(+), 11 deletions(-) create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-run.c create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-rv32gcv.c create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-rv64gcv.c create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-template.h create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-zvfh-run.c create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-zvfh-run.c diff --git a/gcc/config/riscv/autovec.md b/gcc/config/riscv/autovec.md index 5cc48f966aa..c2913128362 100644 --- a/gcc/config/riscv/autovec.md +++ b/gcc/config/riscv/autovec.md @@ -573,6 +573,25 @@ (define_expand "2" DONE; }) +;; ------------------------------------------------------------------------- +;; ---- [FP<-INT] Conversions +;; ------------------------------------------------------------------------- +;; Includes: +;; - vfcvt.f.xu.v +;; - vfcvt.f.x.v +;; ------------------------------------------------------------------------- + +(define_expand "2" + [(set (match_operand:VF 0 "register_operand") + (any_float:VF + (match_operand: 1 "register_operand")))] + "TARGET_VECTOR" +{ + insn_code icode = code_for_pred (, mode); + riscv_vector::emit_vlmax_fp_insn (icode, riscv_vector::RVV_UNOP, operands); + DONE; +}) + ;; ========================================================================= ;; == Unary arithmetic ;; ========================================================================= diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-run.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-run.c new file mode 100644 index 00000000000..dfe73285b36 --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-run.c @@ -0,0 +1,96 @@ +/* { dg-do run { target { riscv_vector } } } */ +/* { dg-additional-options "-std=c99 -fno-vect-cost-model --param=riscv-autovec-preference=scalable" } */ + +#include "vfcvt-itof-template.h" + +#define RUN(TYPE1, TYPE2, NUM) \ + TYPE1 src##TYPE1##TYPE2##NUM[NUM]; \ + TYPE2 dst##TYPE1##TYPE2##NUM[NUM]; \ + for (int i = 0; i < NUM; i++) \ + { \ + src##TYPE1##TYPE2##NUM[i] = i * -3 - 88932; \ + } \ + vfcvt_##TYPE1##TYPE2 (dst##TYPE1##TYPE2##NUM, src##TYPE1##TYPE2##NUM, NUM); \ + for (int i = 0; i < NUM; i++) \ + if (dst##TYPE1##TYPE2##NUM[i] != (TYPE2) src##TYPE1##TYPE2##NUM[i]) \ + __builtin_abort (); + +#define RUN2(TYPE1, TYPE2, NUM) \ + TYPE1 src##TYPE1##TYPE2##NUM[NUM]; \ + TYPE2 dst##TYPE1##TYPE2##NUM[NUM]; \ + for (int i = 0; i < NUM; i++) \ + { \ + src##TYPE1##TYPE2##NUM[i] = i * 3 + 88932; \ + } \ + vfcvt_##TYPE1##TYPE2 (dst##TYPE1##TYPE2##NUM, src##TYPE1##TYPE2##NUM, NUM); \ + for (int i = 0; i < NUM; i++) \ + if (dst##TYPE1##TYPE2##NUM[i] != (TYPE2) src##TYPE1##TYPE2##NUM[i]) \ + __builtin_abort (); + +int +main () +{ + RUN (int32_t, float, 3) + RUN (int32_t, float, 4) + RUN (int32_t, float, 7) + RUN (int32_t, float, 99) + RUN (int32_t, float, 119) + RUN (int32_t, float, 128) + RUN (int32_t, float, 256) + RUN (int32_t, float, 279) + RUN (int32_t, float, 555) + RUN (int32_t, float, 1024) + RUN (int32_t, float, 1389) + RUN (int32_t, float, 2048) + RUN (int32_t, float, 3989) + RUN (int32_t, float, 4096) + RUN (int32_t, float, 5975) + + RUN2 (uint32_t, float, 3) + RUN2 (uint32_t, float, 4) + RUN2 (uint32_t, float, 7) + RUN2 (uint32_t, float, 99) + RUN2 (uint32_t, float, 119) + RUN2 (uint32_t, float, 128) + RUN2 (uint32_t, float, 256) + RUN2 (uint32_t, float, 279) + RUN2 (uint32_t, float, 555) + RUN2 (uint32_t, float, 1024) + RUN2 (uint32_t, float, 1389) + RUN2 (uint32_t, float, 2048) + RUN2 (uint32_t, float, 3989) + RUN2 (uint32_t, float, 4096) + RUN2 (uint32_t, float, 5975) + + RUN (int64_t, double, 3) + RUN (int64_t, double, 4) + RUN (int64_t, double, 7) + RUN (int64_t, double, 99) + RUN (int64_t, double, 119) + RUN (int64_t, double, 128) + RUN (int64_t, double, 256) + RUN (int64_t, double, 279) + RUN (int64_t, double, 555) + RUN (int64_t, double, 1024) + RUN (int64_t, double, 1389) + RUN (int64_t, double, 2048) + RUN (int64_t, double, 3989) + RUN (int64_t, double, 4096) + RUN (int64_t, double, 5975) + + RUN2 (uint64_t, double, 3) + RUN2 (uint64_t, double, 4) + RUN2 (uint64_t, double, 7) + RUN2 (uint64_t, double, 99) + RUN2 (uint64_t, double, 119) + RUN2 (uint64_t, double, 128) + RUN2 (uint64_t, double, 256) + RUN2 (uint64_t, double, 279) + RUN2 (uint64_t, double, 555) + RUN2 (uint64_t, double, 1024) + RUN2 (uint64_t, double, 1389) + RUN2 (uint64_t, double, 2048) + RUN2 (uint64_t, double, 3989) + RUN2 (uint64_t, double, 4096) + RUN2 (uint64_t, double, 5975) +} diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-rv32gcv.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-rv32gcv.c new file mode 100644 index 00000000000..dae14423fd3 --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-rv32gcv.c @@ -0,0 +1,7 @@ +/* { dg-do compile } */ +/* { dg-additional-options "-std=c99 -fno-vect-cost-model -march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable" } */ + +#include "vfcvt-itof-template.h" + +/* { dg-final { scan-assembler-times {\tvfcvt\.f\.x\.v} 3 } } */ +/* { dg-final { scan-assembler-times {\tvfcvt\.f\.xu\.v} 3 } } */ diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-rv64gcv.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-rv64gcv.c new file mode 100644 index 00000000000..ccb2bb5544d --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-rv64gcv.c @@ -0,0 +1,7 @@ +/* { dg-do compile } */ +/* { dg-additional-options "-std=c99 -fno-vect-cost-model -march=rv64gcv_zvfh -mabi=lp64d --param=riscv-autovec-preference=scalable" } */ + +#include "vfcvt-itof-template.h" + +/* { dg-final { scan-assembler-times {\tvfcvt\.f\.x\.v} 3 } } */ +/* { dg-final { scan-assembler-times {\tvfcvt\.f\.xu\.v} 3 } } */ diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-template.h b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-template.h new file mode 100644 index 00000000000..63a5f81cdbc --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-template.h @@ -0,0 +1,20 @@ +#include + +#define TEST(TYPE1, TYPE2) \ + __attribute__ ((noipa)) \ + void vfcvt_##TYPE1##TYPE2 (TYPE2 *restrict dst, \ + TYPE1 *restrict a, int n) \ + { \ + for (int i = 0; i < n; i++) \ + dst[i] = (TYPE1) a[i]; \ + } + +#define TEST_ALL() \ + TEST (int32_t, float) \ + TEST (uint32_t, float) \ + TEST (int64_t, double) \ + TEST (uint64_t, double) \ + TEST (int16_t, _Float16) \ + TEST (uint16_t, _Float16) \ + +TEST_ALL () diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-zvfh-run.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-zvfh-run.c new file mode 100644 index 00000000000..e7a8f0afbc2 --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt-itof-zvfh-run.c @@ -0,0 +1,64 @@ +/* { dg-do run { target { riscv_vector && riscv_zvfh_hw } } } */ +/* { dg-additional-options "-std=c99 -march=rv64gcv_zvfh -mabi=lp64d -fno-vect-cost-model --param=riscv-autovec-preference=scalable" } */ + +#include "vfcvt-itof-template.h" + +#define RUN(TYPE1, TYPE2, NUM) \ + TYPE1 src##TYPE1##TYPE2##NUM[NUM]; \ + TYPE2 dst##TYPE1##TYPE2##NUM[NUM]; \ + for (int i = 0; i < NUM; i++) \ + { \ + src##TYPE1##TYPE2##NUM[i] = i * 3 - 8932; \ + } \ + vfcvt_##TYPE1##TYPE2 (dst##TYPE1##TYPE2##NUM, src##TYPE1##TYPE2##NUM, NUM); \ + for (int i = 0; i < NUM; i++) \ + if (dst##TYPE1##TYPE2##NUM[i] != (TYPE2) src##TYPE1##TYPE2##NUM[i]) \ + __builtin_abort (); + +#define RUN2(TYPE1, TYPE2, NUM) \ + TYPE1 src##TYPE1##TYPE2##NUM[NUM]; \ + TYPE2 dst##TYPE1##TYPE2##NUM[NUM]; \ + for (int i = 0; i < NUM; i++) \ + { \ + src##TYPE1##TYPE2##NUM[i] = i * 3 + 8932; \ + } \ + vfcvt_##TYPE1##TYPE2 (dst##TYPE1##TYPE2##NUM, src##TYPE1##TYPE2##NUM, NUM); \ + for (int i = 0; i < NUM; i++) \ + if (dst##TYPE1##TYPE2##NUM[i] != (TYPE2) src##TYPE1##TYPE2##NUM[i]) \ + __builtin_abort (); + +int +main () +{ + RUN (int16_t, _Float16, 3) + RUN (int16_t, _Float16, 4) + RUN (int16_t, _Float16, 7) + RUN (int16_t, _Float16, 99) + RUN (int16_t, _Float16, 119) + RUN (int16_t, _Float16, 128) + RUN (int16_t, _Float16, 256) + RUN (int16_t, _Float16, 279) + RUN (int16_t, _Float16, 555) + RUN (int16_t, _Float16, 1024) + RUN (int16_t, _Float16, 1389) + RUN (int16_t, _Float16, 2048) + RUN (int16_t, _Float16, 3989) + RUN (int16_t, _Float16, 4096) + RUN (int16_t, _Float16, 5975) + + RUN2 (uint16_t, _Float16, 3) + RUN2 (uint16_t, _Float16, 4) + RUN2 (uint16_t, _Float16, 7) + RUN2 (uint16_t, _Float16, 99) + RUN2 (uint16_t, _Float16, 119) + RUN2 (uint16_t, _Float16, 128) + RUN2 (uint16_t, _Float16, 256) + RUN2 (uint16_t, _Float16, 279) + RUN2 (uint16_t, _Float16, 555) + RUN2 (uint16_t, _Float16, 1024) + RUN2 (uint16_t, _Float16, 1389) + RUN2 (uint16_t, _Float16, 2048) + RUN2 (uint16_t, _Float16, 3989) + RUN2 (uint16_t, _Float16, 4096) + RUN2 (uint16_t, _Float16, 5975) +} diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-run.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-run.c index 05f8d911ad7..e4dca6014a3 100644 --- a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-run.c +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-run.c @@ -4,6 +4,18 @@ #include "vfcvt_rtz-template.h" #define RUN(TYPE1, TYPE2, NUM) \ + TYPE1 src##TYPE1##TYPE2##NUM[NUM]; \ + TYPE2 dst##TYPE1##TYPE2##NUM[NUM]; \ + for (int i = 0; i < NUM; i++) \ + { \ + src##TYPE1##TYPE2##NUM[i] = i * -3.1315926 - 88932.947289; \ + } \ + vfcvt_##TYPE1##TYPE2 (dst##TYPE1##TYPE2##NUM, src##TYPE1##TYPE2##NUM, NUM); \ + for (int i = 0; i < NUM; i++) \ + if (dst##TYPE1##TYPE2##NUM[i] != (TYPE2) src##TYPE1##TYPE2##NUM[i]) \ + __builtin_abort (); + +#define RUN2(TYPE1, TYPE2, NUM) \ TYPE1 src##TYPE1##TYPE2##NUM[NUM]; \ TYPE2 dst##TYPE1##TYPE2##NUM[NUM]; \ for (int i = 0; i < NUM; i++) \ @@ -34,6 +46,22 @@ main () RUN (float, int32_t, 4096) RUN (float, int32_t, 5975) + RUN2 (float, uint32_t, 3) + RUN2 (float, uint32_t, 4) + RUN2 (float, uint32_t, 7) + RUN2 (float, uint32_t, 99) + RUN2 (float, uint32_t, 119) + RUN2 (float, uint32_t, 128) + RUN2 (float, uint32_t, 256) + RUN2 (float, uint32_t, 279) + RUN2 (float, uint32_t, 555) + RUN2 (float, uint32_t, 1024) + RUN2 (float, uint32_t, 1389) + RUN2 (float, uint32_t, 2048) + RUN2 (float, uint32_t, 3989) + RUN2 (float, uint32_t, 4096) + RUN2 (float, uint32_t, 5975) + RUN (double, int64_t, 3) RUN (double, int64_t, 4) RUN (double, int64_t, 7) @@ -49,4 +77,20 @@ main () RUN (double, int64_t, 3989) RUN (double, int64_t, 4096) RUN (double, int64_t, 5975) + + RUN2 (double, uint64_t, 3) + RUN2 (double, uint64_t, 4) + RUN2 (double, uint64_t, 7) + RUN2 (double, uint64_t, 99) + RUN2 (double, uint64_t, 119) + RUN2 (double, uint64_t, 128) + RUN2 (double, uint64_t, 256) + RUN2 (double, uint64_t, 279) + RUN2 (double, uint64_t, 555) + RUN2 (double, uint64_t, 1024) + RUN2 (double, uint64_t, 1389) + RUN2 (double, uint64_t, 2048) + RUN2 (double, uint64_t, 3989) + RUN2 (double, uint64_t, 4096) + RUN2 (double, uint64_t, 5975) } diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-rv32gcv.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-rv32gcv.c index 2f84631775f..0a79adf3510 100644 --- a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-rv32gcv.c +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-rv32gcv.c @@ -1,6 +1,7 @@ /* { dg-do compile } */ -/* { dg-additional-options "-std=c99 -fno-vect-cost-model -march=rv32gcv -mabi=ilp32d --param=riscv-autovec-preference=scalable" } */ +/* { dg-additional-options "-std=c99 -fno-vect-cost-model -march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable" } */ #include "vfcvt_rtz-template.h" -/* { dg-final { scan-assembler-times {\tvfcvt\.rtz} 2 } } */ +/* { dg-final { scan-assembler-times {\tvfcvt\.rtz\.x\.f\.v} 3 } } */ +/* { dg-final { scan-assembler-times {\tvfcvt\.rtz\.xu\.f\.v} 3 } } */ diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-rv64gcv.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-rv64gcv.c index 40e3e7a450d..e74984798e6 100644 --- a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-rv64gcv.c +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-rv64gcv.c @@ -1,6 +1,7 @@ /* { dg-do compile } */ -/* { dg-additional-options "-std=c99 -fno-vect-cost-model -march=rv64gcv -mabi=lp64d --param=riscv-autovec-preference=scalable" } */ +/* { dg-additional-options "-std=c99 -fno-vect-cost-model -march=rv64gcv_zvfh -mabi=lp64d --param=riscv-autovec-preference=scalable" } */ #include "vfcvt_rtz-template.h" -/* { dg-final { scan-assembler-times {\tvfcvt\.rtz} 2 } } */ +/* { dg-final { scan-assembler-times {\tvfcvt\.rtz\.x\.f\.v} 3 } } */ +/* { dg-final { scan-assembler-times {\tvfcvt\.rtz\.xu\.f\.v} 3 } } */ diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-template.h b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-template.h index 73bc1ad5591..17f50b95f6c 100644 --- a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-template.h +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-template.h @@ -10,6 +10,10 @@ #define TEST_ALL() \ TEST (float, int32_t) \ - TEST (double, int64_t) + TEST (float, uint32_t) \ + TEST (double, int64_t) \ + TEST (double, uint64_t) \ + TEST (_Float16, int16_t) \ + TEST (_Float16, uint16_t) \ TEST_ALL () diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-zvfh-run.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-zvfh-run.c new file mode 100644 index 00000000000..e75aee7f36f --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/conversions/vfcvt_rtz-zvfh-run.c @@ -0,0 +1,64 @@ +/* { dg-do run { target { riscv_vector && riscv_zvfh_hw } } } */ +/* { dg-additional-options "-std=c99 -march=rv64gcv_zvfh -mabi=lp64d -fno-vect-cost-model --param=riscv-autovec-preference=scalable" } */ + +#include "vfcvt_rtz-template.h" + +#define RUN(TYPE1, TYPE2, NUM) \ + TYPE1 src##TYPE1##TYPE2##NUM[NUM]; \ + TYPE2 dst##TYPE1##TYPE2##NUM[NUM]; \ + for (int i = 0; i < NUM; i++) \ + { \ + src##TYPE1##TYPE2##NUM[i] = i * -3.1315926 - 8932.947289; \ + } \ + vfcvt_##TYPE1##TYPE2 (dst##TYPE1##TYPE2##NUM, src##TYPE1##TYPE2##NUM, NUM); \ + for (int i = 0; i < NUM; i++) \ + if (dst##TYPE1##TYPE2##NUM[i] != (TYPE2) src##TYPE1##TYPE2##NUM[i]) \ + __builtin_abort (); + +#define RUN2(TYPE1, TYPE2, NUM) \ + TYPE1 src##TYPE1##TYPE2##NUM[NUM]; \ + TYPE2 dst##TYPE1##TYPE2##NUM[NUM]; \ + for (int i = 0; i < NUM; i++) \ + { \ + src##TYPE1##TYPE2##NUM[i] = i * 3.1315926 + 8932.947289; \ + } \ + vfcvt_##TYPE1##TYPE2 (dst##TYPE1##TYPE2##NUM, src##TYPE1##TYPE2##NUM, NUM); \ + for (int i = 0; i < NUM; i++) \ + if (dst##TYPE1##TYPE2##NUM[i] != (TYPE2) src##TYPE1##TYPE2##NUM[i]) \ + __builtin_abort (); + +int +main () +{ + RUN (_Float16, int16_t, 3) + RUN (_Float16, int16_t, 4) + RUN (_Float16, int16_t, 7) + RUN (_Float16, int16_t, 99) + RUN (_Float16, int16_t, 119) + RUN (_Float16, int16_t, 128) + RUN (_Float16, int16_t, 256) + RUN (_Float16, int16_t, 279) + RUN (_Float16, int16_t, 555) + RUN (_Float16, int16_t, 1024) + RUN (_Float16, int16_t, 1389) + RUN (_Float16, int16_t, 2048) + RUN (_Float16, int16_t, 3989) + RUN (_Float16, int16_t, 4096) + RUN (_Float16, int16_t, 5975) + + RUN2 (_Float16, uint16_t, 3) + RUN2 (_Float16, uint16_t, 4) + RUN2 (_Float16, uint16_t, 7) + RUN2 (_Float16, uint16_t, 99) + RUN2 (_Float16, uint16_t, 119) + RUN2 (_Float16, uint16_t, 128) + RUN2 (_Float16, uint16_t, 256) + RUN2 (_Float16, uint16_t, 279) + RUN2 (_Float16, uint16_t, 555) + RUN2 (_Float16, uint16_t, 1024) + RUN2 (_Float16, uint16_t, 1389) + RUN2 (_Float16, uint16_t, 2048) + RUN2 (_Float16, uint16_t, 3989) + RUN2 (_Float16, uint16_t, 4096) + RUN2 (_Float16, uint16_t, 5975) +} diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/zvfhmin-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/zvfhmin-1.c index 109fcbcaf00..1c417902e24 100644 --- a/gcc/testsuite/gcc.target/riscv/rvv/autovec/zvfhmin-1.c +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/zvfhmin-1.c @@ -1,5 +1,7 @@ /* { dg-do compile } */ -/* { dg-options "-march=rv32gcv_zvfhmin -mabi=ilp32d --param riscv-autovec-preference=scalable -fdump-tree-vect-details" } */ +/* { dg-options "-march=rv32gcv_zvfhmin -mabi=ilp32d --param riscv-autovec-preference=scalable -ffast-math -fdump-rtl-final" } */ + +#include void f0 (_Float16 * __restrict a, _Float16 * __restrict b, int n) { @@ -40,10 +42,44 @@ void f6 (_Float16 * __restrict a, _Float16 * __restrict b, int n) void f7 (_Float16 * __restrict a, _Float16 * __restrict b, int n) { for (int i = 0; i < n; i++) - a[i] = __builtin_sqrtf (b[i]); + a[i] = __builtin_sqrtf16 (b[i]); +} + +void f8 (_Float16 * __restrict a, int16_t * __restrict b, int n) +{ + for (int i = 0; i < n; i++) + a[i] = (_Float16) (b[i]); +} + +void f9 (_Float16 * __restrict a, uint16_t * __restrict b, int n) +{ + for (int i = 0; i < n; i++) + a[i] = (_Float16) (b[i]); +} + +void f10 (int16_t * __restrict a, _Float16 * __restrict b, int n) +{ + for (int i = 0; i < n; i++) + a[i] = (int16_t) (b[i]); +} + +void f11 (uint16_t * __restrict a, _Float16 * __restrict b, int n) +{ + for (int i = 0; i < n; i++) + a[i] = (uint16_t) (b[i]); } -/* We can't enable FP16 NEG/PLUS/MINUS/MULT/DIV/ABS/SQRTF auto-vectorization - when -march="*zvfhmin*" because the min variant of the extension only - provides loads, stores and conversions. */ -/* { dg-final { scan-tree-dump-times "vectorized 1 loops in function" 0 "vect" } } */ +/* We can't enable FP16 NEG/PLUS/MINUS/MULT/DIV/ABS/SQRTF as well as int/float + conversion auto-vectorization when -march="*zvfhmin*" because the min + variant of the extension only provides loads, stores and conversions. + As we might still vectorize after promotion to float, we need to make + sure that no vector operations with an HFmode are being generated. */ +/* { dg-final { scan-rtl-dump-not "plus:VNx\[0-9\]+HF" "final" } } */ +/* { dg-final { scan-rtl-dump-not "minus:VNx\[0-9\]+HF" "final" } } */ +/* { dg-final { scan-rtl-dump-not "mult:VNx\[0-9\]+HF" "final" } } */ +/* { dg-final { scan-rtl-dump-not "div:VNx\[0-9\]+HF" "final" } } */ +/* { dg-final { scan-rtl-dump-not "neg:VNx\[0-9\]+HF" "final" } } */ +/* { dg-final { scan-rtl-dump-not "abs:VNx\[0-9\]+HF" "final" } } */ +/* { dg-final { scan-rtl-dump-not "sqrt:VNx\[0-9\]+HF" "final" } } */ +/* { dg-final { scan-rtl-dump-not "float:VNx\[0-9\]+HF" "final" } } */ +/* { dg-final { scan-rtl-dump-not "fix:VNx\[0-9\]+HI\s*.+reg:VNx\[0-9\]+HF" "final" } } */