From patchwork Tue Sep 19 08:16:11 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "juzhe.zhong@rivai.ai" X-Patchwork-Id: 141734 Return-Path: Delivered-To: ouuuleilei@gmail.com Received: by 2002:a05:612c:172:b0:3f2:4152:657d with SMTP id h50csp3215879vqi; Tue, 19 Sep 2023 01:21:11 -0700 (PDT) X-Google-Smtp-Source: AGHT+IG8d5ucYivoO0KZZgHKdFhojH7s/qsJWmr8Abkgy+lvfiWRqcFCmjWimcwIifpGD2H0ovy4 X-Received: by 2002:a17:906:113:b0:9aa:6fc3:bbff with SMTP id 19-20020a170906011300b009aa6fc3bbffmr11144126eje.18.1695111671146; Tue, 19 Sep 2023 01:21:11 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1695111671; cv=none; d=google.com; s=arc-20160816; b=l7l7Ohy1eDuBKacAUThFpaSzTBG41s3wZhLUzuMUcIXtD1itonpR1933lbjNR8Jo8f AtTpQwPzJCUSyqCuoxwUnwQHZdr4Og7hy2CxAwl8rX4kvlTNsgEXziw1icqfouLi46F3 kLyv/IaV1kUCu5F7Xsc0BcHS4ERH6yaMZc7jMShpDU1HoKMm9FDV1nhLKurtLyPtS9If ECmA8b6TUSHpjUnyBubDaCp0Z4UDPtXo/+YIi+bkgPHsWUH18fOR8aCuaLOksH1z0bi3 +bovn6yZ0w6DUOTXEm0BS0Oc9ggmVFs4ET1yCg75Dj5KYAyWscNtGRhb+WRksXgJk/2N j09A== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=sender:errors-to:list-subscribe:list-help:list-post:list-archive :list-unsubscribe:list-id:precedence:feedback-id :content-transfer-encoding:mime-version:message-id:date:subject:cc :to:from:dmarc-filter:delivered-to; bh=F4E5ZK2OyyaEuAmpmMJ1Eb+FS+QeiPHZNy0gqibL4Cc=; fh=12MRPJmZ1mgDpHqWoogMKqnaGRGM2b7lcuJroqfjJiw=; b=K/ogH83F1K0s2L7E6sBqdQjNDX3lbAYwYJEQ3J+Ch4F3ilsI+P95/LUio9IOJlGz0T toBzWiSH3YwVBahI1oq3rdS1rJ3nsj7IK2J7YtmQWecccXaKJdVr0cm/gRyHeA1EdRok 5A4wYTuxoXh1JmBw2cCYEkuhH1mvVGqb2lLYk/WCww8p/oizqby6N6wvNinrmuBkxbqF u6sFTOI5hxF8pJ0ArZf8JA7zhV6ic8hjvEHi+eyCpjIAJnp/MBGz2tydmrYHr0PfMJx0 Q9kqAoNi0Uhyxa+6ANzv0ot7umLLKIOKkPWz/w+ThJDpPFamshc30f5Eas0rsbYmkyqh hNXQ== ARC-Authentication-Results: i=1; mx.google.com; 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" Received: from server2.sourceware.org (server2.sourceware.org. [2620:52:3:1:0:246e:9693:128c]) by mx.google.com with ESMTPS id cw17-20020a170906c79100b0099367350a83si9988926ejb.898.2023.09.19.01.21.10 for (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 19 Sep 2023 01:21:11 -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; 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" Received: from server2.sourceware.org (localhost [IPv6:::1]) by sourceware.org (Postfix) with ESMTP id C2288385CC8A for ; Tue, 19 Sep 2023 08:19:22 +0000 (GMT) X-Original-To: gcc-patches@gcc.gnu.org Delivered-To: gcc-patches@gcc.gnu.org Received: from smtpbguseast2.qq.com (smtpbguseast2.qq.com [54.204.34.130]) by sourceware.org (Postfix) with ESMTPS id 1CB6A385696E for ; Tue, 19 Sep 2023 08:16:18 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org 1CB6A385696E Authentication-Results: sourceware.org; dmarc=none (p=none dis=none) header.from=rivai.ai Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=rivai.ai X-QQ-mid: bizesmtp74t1695111373tgoiywbn Received: from rios-cad121.hadoop.rioslab.org ( [58.60.1.9]) by bizesmtp.qq.com (ESMTP) with id ; Tue, 19 Sep 2023 16:16:12 +0800 (CST) X-QQ-SSF: 01400000000000G0V000000A0000000 X-QQ-FEAT: +ynUkgUhZJn90g5c68u3H2iXN4hb+pE9z9xLhvMMWhPzWGTcsUQVHupxZMy1d YOILKDqJ8l0eAfwtVZsKVb5pvSrzIAYyvlL28opm11Hx5icJmLzR/o/1kYccMwB9xm9NvD3 lyMYS1gmIXjEvWWvS0mJQ7WwPhouHemknSveY7k7XGOzP3xOn/07YXcEgBPbyy0BW86of6h sA8EIsEwm4Vhtgr6MZdKdiNClMzPdMwrBp8FdApFdm8qm8TCicaHb5KhXXfuqMg1+kN4/qF mgyUa1//gVp7ZDaRAoR0eoPjPHnMKbCmsL07b3rLNdGE6EUuDko6HagXmcoBj1FdEWE8C4V Xr77HbTMvzW3ZVyeK0amIvcArLjtFACprqZ0rfXVpbzrJ3LJFApqVW9w9atYA/YuvqssWjU Cd59p9dQJXw= X-QQ-GoodBg: 2 X-BIZMAIL-ID: 13454572190318052871 From: Juzhe-Zhong To: gcc-patches@gcc.gnu.org Cc: kito.cheng@gmail.com, kito.cheng@sifive.com, jeffreyalaw@gmail.com, rdapp.gcc@gmail.com, Juzhe-Zhong Subject: [Committed] RISC-V: Support integer FMA/FNMA VLS modes autovectorization Date: Tue, 19 Sep 2023 16:16:11 +0800 Message-Id: <20230919081611.2696019-1-juzhe.zhong@rivai.ai> X-Mailer: git-send-email 2.36.3 MIME-Version: 1.0 X-QQ-SENDSIZE: 520 Feedback-ID: bizesmtp:rivai.ai:qybglogicsvrgz:qybglogicsvrgz7a-one-0 X-Spam-Status: No, score=-11.7 required=5.0 tests=BAYES_00, GIT_PATCH_0, KAM_ASCII_DIVIDERS, KAM_DMARC_STATUS, KAM_SHORT, RCVD_IN_DNSWL_NONE, RCVD_IN_MSPIKE_H3, RCVD_IN_MSPIKE_WL, SPF_HELO_PASS, 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: gcc-patches@gcc.gnu.org X-Mailman-Version: 2.1.30 Precedence: list List-Id: Gcc-patches mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org Sender: "Gcc-patches" X-getmail-retrieved-from-mailbox: INBOX X-GMAIL-THRID: 1777453415487706182 X-GMAIL-MSGID: 1777453415487706182 Simpily extend the current VLA iterator and patterns. Regression passed with no difference. gcc/ChangeLog: * config/riscv/autovec.md: Add VLS modes. * config/riscv/vector.md: Ditto. gcc/testsuite/ChangeLog: * gcc.target/riscv/rvv/autovec/vls/def.h: Add VLS FMA/FNMA test. * gcc.target/riscv/rvv/autovec/vls/fma-1.c: New test. * gcc.target/riscv/rvv/autovec/vls/fma-2.c: New test. * gcc.target/riscv/rvv/autovec/vls/fma-3.c: New test. * gcc.target/riscv/rvv/autovec/vls/fma-4.c: New test. * gcc.target/riscv/rvv/autovec/vls/fnma-1.c: New test. * gcc.target/riscv/rvv/autovec/vls/fnma-2.c: New test. * gcc.target/riscv/rvv/autovec/vls/fnma-3.c: New test. * gcc.target/riscv/rvv/autovec/vls/fnma-4.c: New test. --- gcc/config/riscv/autovec.md | 24 +- gcc/config/riscv/vector.md | 296 +++++++++--------- .../gcc.target/riscv/rvv/autovec/vls/def.h | 18 ++ .../gcc.target/riscv/rvv/autovec/vls/fma-1.c | 45 +++ .../gcc.target/riscv/rvv/autovec/vls/fma-2.c | 43 +++ .../gcc.target/riscv/rvv/autovec/vls/fma-3.c | 41 +++ .../gcc.target/riscv/rvv/autovec/vls/fma-4.c | 39 +++ .../gcc.target/riscv/rvv/autovec/vls/fnma-1.c | 45 +++ .../gcc.target/riscv/rvv/autovec/vls/fnma-2.c | 43 +++ .../gcc.target/riscv/rvv/autovec/vls/fnma-3.c | 41 +++ .../gcc.target/riscv/rvv/autovec/vls/fnma-4.c | 39 +++ 11 files changed, 514 insertions(+), 160 deletions(-) create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-1.c create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-2.c create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-3.c create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-4.c create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-1.c create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-2.c create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-3.c create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-4.c diff --git a/gcc/config/riscv/autovec.md b/gcc/config/riscv/autovec.md index ac7599f3e0a..1aadb6eea1f 100644 --- a/gcc/config/riscv/autovec.md +++ b/gcc/config/riscv/autovec.md @@ -1079,12 +1079,12 @@ ;; ------------------------------------------------------------------------- (define_insn_and_split "fma4" - [(set (match_operand:VI 0 "register_operand") - (plus:VI - (mult:VI - (match_operand:VI 1 "register_operand") - (match_operand:VI 2 "register_operand")) - (match_operand:VI 3 "register_operand")))] + [(set (match_operand:V_VLSI 0 "register_operand") + (plus:V_VLSI + (mult:V_VLSI + (match_operand:V_VLSI 1 "register_operand") + (match_operand:V_VLSI 2 "register_operand")) + (match_operand:V_VLSI 3 "register_operand")))] "TARGET_VECTOR && can_create_pseudo_p ()" "#" "&& 1" @@ -1107,12 +1107,12 @@ ;; ------------------------------------------------------------------------- (define_insn_and_split "fnma4" - [(set (match_operand:VI 0 "register_operand") - (minus:VI - (match_operand:VI 3 "register_operand") - (mult:VI - (match_operand:VI 1 "register_operand") - (match_operand:VI 2 "register_operand"))))] + [(set (match_operand:V_VLSI 0 "register_operand") + (minus:V_VLSI + (match_operand:V_VLSI 3 "register_operand") + (mult:V_VLSI + (match_operand:V_VLSI 1 "register_operand") + (match_operand:V_VLSI 2 "register_operand"))))] "TARGET_VECTOR && can_create_pseudo_p ()" "#" "&& 1" diff --git a/gcc/config/riscv/vector.md b/gcc/config/riscv/vector.md index c7c6ec3d6f1..c5a1c9061c4 100644 --- a/gcc/config/riscv/vector.md +++ b/gcc/config/riscv/vector.md @@ -5144,8 +5144,8 @@ ;; ------------------------------------------------------------------------------- (define_expand "@pred_mul_plus" - [(set (match_operand:VI 0 "register_operand") - (if_then_else:VI + [(set (match_operand:V_VLSI 0 "register_operand") + (if_then_else:V_VLSI (unspec: [(match_operand: 1 "vector_mask_operand") (match_operand 6 "vector_length_operand") @@ -5154,20 +5154,20 @@ (match_operand 9 "const_int_operand") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (plus:VI - (mult:VI - (match_operand:VI 2 "register_operand") - (match_operand:VI 3 "register_operand")) - (match_operand:VI 4 "register_operand")) - (match_operand:VI 5 "vector_merge_operand")))] + (plus:V_VLSI + (mult:V_VLSI + (match_operand:V_VLSI 2 "register_operand") + (match_operand:V_VLSI 3 "register_operand")) + (match_operand:V_VLSI 4 "register_operand")) + (match_operand:V_VLSI 5 "vector_merge_operand")))] "TARGET_VECTOR" { riscv_vector::prepare_ternary_operands (operands); }) (define_insn "*pred_mul_plus_undef" - [(set (match_operand:VI 0 "register_operand" "=vd, vd,?&vd, vr, vr,?&vr") - (if_then_else:VI + [(set (match_operand:V_VLSI 0 "register_operand" "=vd, vd,?&vd, vr, vr,?&vr") + (if_then_else:V_VLSI (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm, vm,Wc1,Wc1, Wc1") (match_operand 6 "vector_length_operand" " rK, rK, rK, rK, rK, rK") @@ -5176,12 +5176,12 @@ (match_operand 9 "const_int_operand" " i, i, i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (plus:VI - (mult:VI - (match_operand:VI 3 "register_operand" " 0, vr, vr, 0, vr, vr") - (match_operand:VI 4 "register_operand" " vr, vr, vr, vr, vr, vr")) - (match_operand:VI 5 "register_operand" " vr, 0, vr, vr, 0, vr")) - (match_operand:VI 2 "vector_undef_operand")))] + (plus:V_VLSI + (mult:V_VLSI + (match_operand:V_VLSI 3 "register_operand" " 0, vr, vr, 0, vr, vr") + (match_operand:V_VLSI 4 "register_operand" " vr, vr, vr, vr, vr, vr")) + (match_operand:V_VLSI 5 "register_operand" " vr, 0, vr, vr, 0, vr")) + (match_operand:V_VLSI 2 "vector_undef_operand")))] "TARGET_VECTOR" "@ vmadd.vv\t%0,%4,%5%p1 @@ -5194,8 +5194,8 @@ (set_attr "mode" "")]) (define_insn "*pred_madd" - [(set (match_operand:VI 0 "register_operand" "=vd,?&vd, vr,?&vr") - (if_then_else:VI + [(set (match_operand:V_VLSI 0 "register_operand" "=vd,?&vd, vr,?&vr") + (if_then_else:V_VLSI (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm,Wc1, Wc1") (match_operand 5 "vector_length_operand" " rK, rK, rK, rK") @@ -5204,11 +5204,11 @@ (match_operand 8 "const_int_operand" " i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (plus:VI - (mult:VI - (match_operand:VI 2 "register_operand" " 0, vr, 0, vr") - (match_operand:VI 3 "register_operand" " vr, vr, vr, vr")) - (match_operand:VI 4 "register_operand" " vr, vr, vr, vr")) + (plus:V_VLSI + (mult:V_VLSI + (match_operand:V_VLSI 2 "register_operand" " 0, vr, 0, vr") + (match_operand:V_VLSI 3 "register_operand" " vr, vr, vr, vr")) + (match_operand:V_VLSI 4 "register_operand" " vr, vr, vr, vr")) (match_dup 2)))] "TARGET_VECTOR" "@ @@ -5225,8 +5225,8 @@ (set (attr "avl_type") (symbol_ref "INTVAL (operands[8])"))]) (define_insn "*pred_macc" - [(set (match_operand:VI 0 "register_operand" "=vd,?&vd, vr,?&vr") - (if_then_else:VI + [(set (match_operand:V_VLSI 0 "register_operand" "=vd,?&vd, vr,?&vr") + (if_then_else:V_VLSI (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm,Wc1, Wc1") (match_operand 5 "vector_length_operand" " rK, rK, rK, rK") @@ -5235,11 +5235,11 @@ (match_operand 8 "const_int_operand" " i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (plus:VI - (mult:VI - (match_operand:VI 2 "register_operand" " vr, vr, vr, vr") - (match_operand:VI 3 "register_operand" " vr, vr, vr, vr")) - (match_operand:VI 4 "register_operand" " 0, vr, 0, vr")) + (plus:V_VLSI + (mult:V_VLSI + (match_operand:V_VLSI 2 "register_operand" " vr, vr, vr, vr") + (match_operand:V_VLSI 3 "register_operand" " vr, vr, vr, vr")) + (match_operand:V_VLSI 4 "register_operand" " 0, vr, 0, vr")) (match_dup 4)))] "TARGET_VECTOR" "@ @@ -5256,8 +5256,8 @@ (set (attr "avl_type") (symbol_ref "INTVAL (operands[8])"))]) (define_expand "@pred_mul_plus_scalar" - [(set (match_operand:VI_QHS 0 "register_operand") - (if_then_else:VI_QHS + [(set (match_operand:V_VLSI_QHS 0 "register_operand") + (if_then_else:V_VLSI_QHS (unspec: [(match_operand: 1 "vector_mask_operand") (match_operand 6 "vector_length_operand") @@ -5266,19 +5266,19 @@ (match_operand 9 "const_int_operand") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (plus:VI_QHS - (mult:VI_QHS - (vec_duplicate:VI_QHS + (plus:V_VLSI_QHS + (mult:V_VLSI_QHS + (vec_duplicate:V_VLSI_QHS (match_operand: 2 "register_operand")) - (match_operand:VI_QHS 3 "register_operand")) - (match_operand:VI_QHS 4 "register_operand")) - (match_operand:VI_QHS 5 "register_operand")))] + (match_operand:V_VLSI_QHS 3 "register_operand")) + (match_operand:V_VLSI_QHS 4 "register_operand")) + (match_operand:V_VLSI_QHS 5 "register_operand")))] "TARGET_VECTOR" {}) (define_insn "*pred_madd_scalar" - [(set (match_operand:VI 0 "register_operand" "=vd,?&vd, vr,?&vr") - (if_then_else:VI + [(set (match_operand:V_VLSI 0 "register_operand" "=vd,?&vd, vr,?&vr") + (if_then_else:V_VLSI (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm,Wc1, Wc1") (match_operand 5 "vector_length_operand" " rK, rK, rK, rK") @@ -5287,12 +5287,12 @@ (match_operand 8 "const_int_operand" " i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (plus:VI - (mult:VI - (vec_duplicate:VI + (plus:V_VLSI + (mult:V_VLSI + (vec_duplicate:V_VLSI (match_operand: 2 "register_operand" " r, r, r, r")) - (match_operand:VI 3 "register_operand" " 0, vr, 0, vr")) - (match_operand:VI 4 "register_operand" " vr, vr, vr, vr")) + (match_operand:V_VLSI 3 "register_operand" " 0, vr, 0, vr")) + (match_operand:V_VLSI 4 "register_operand" " vr, vr, vr, vr")) (match_dup 3)))] "TARGET_VECTOR" "@ @@ -5309,8 +5309,8 @@ (set (attr "avl_type") (symbol_ref "INTVAL (operands[8])"))]) (define_insn "*pred_macc_scalar" - [(set (match_operand:VI 0 "register_operand" "=vd,?&vd, vr,?&vr") - (if_then_else:VI + [(set (match_operand:V_VLSI 0 "register_operand" "=vd,?&vd, vr,?&vr") + (if_then_else:V_VLSI (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm,Wc1, Wc1") (match_operand 5 "vector_length_operand" " rK, rK, rK, rK") @@ -5319,12 +5319,12 @@ (match_operand 8 "const_int_operand" " i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (plus:VI - (mult:VI - (vec_duplicate:VI + (plus:V_VLSI + (mult:V_VLSI + (vec_duplicate:V_VLSI (match_operand: 2 "register_operand" " r, r, r, r")) - (match_operand:VI 3 "register_operand" " vr, vr, vr, vr")) - (match_operand:VI 4 "register_operand" " 0, vr, 0, vr")) + (match_operand:V_VLSI 3 "register_operand" " vr, vr, vr, vr")) + (match_operand:V_VLSI 4 "register_operand" " 0, vr, 0, vr")) (match_dup 4)))] "TARGET_VECTOR" "@ @@ -5341,8 +5341,8 @@ (set (attr "avl_type") (symbol_ref "INTVAL (operands[8])"))]) (define_expand "@pred_mul_plus_scalar" - [(set (match_operand:VI_D 0 "register_operand") - (if_then_else:VI_D + [(set (match_operand:V_VLSI_D 0 "register_operand") + (if_then_else:V_VLSI_D (unspec: [(match_operand: 1 "vector_mask_operand") (match_operand 6 "vector_length_operand") @@ -5351,13 +5351,13 @@ (match_operand 9 "const_int_operand") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (plus:VI_D - (mult:VI_D - (vec_duplicate:VI_D + (plus:V_VLSI_D + (mult:V_VLSI_D + (vec_duplicate:V_VLSI_D (match_operand: 2 "reg_or_int_operand")) - (match_operand:VI_D 3 "register_operand")) - (match_operand:VI_D 4 "register_operand")) - (match_operand:VI_D 5 "register_operand")))] + (match_operand:V_VLSI_D 3 "register_operand")) + (match_operand:V_VLSI_D 4 "register_operand")) + (match_operand:V_VLSI_D 5 "register_operand")))] "TARGET_VECTOR" { if (riscv_vector::sew64_scalar_helper ( @@ -5375,8 +5375,8 @@ }) (define_insn "*pred_madd_extended_scalar" - [(set (match_operand:VI_D 0 "register_operand" "=vd,?&vd, vr,?&vr") - (if_then_else:VI_D + [(set (match_operand:V_VLSI_D 0 "register_operand" "=vd,?&vd, vr,?&vr") + (if_then_else:V_VLSI_D (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm,Wc1, Wc1") (match_operand 5 "vector_length_operand" " rK, rK, rK, rK") @@ -5385,13 +5385,13 @@ (match_operand 8 "const_int_operand" " i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (plus:VI_D - (mult:VI_D - (vec_duplicate:VI_D + (plus:V_VLSI_D + (mult:V_VLSI_D + (vec_duplicate:V_VLSI_D (sign_extend: (match_operand: 2 "register_operand" " r, r, r, r"))) - (match_operand:VI_D 3 "register_operand" " 0, vr, 0, vr")) - (match_operand:VI_D 4 "register_operand" " vr, vr, vr, vr")) + (match_operand:V_VLSI_D 3 "register_operand" " 0, vr, 0, vr")) + (match_operand:V_VLSI_D 4 "register_operand" " vr, vr, vr, vr")) (match_dup 3)))] "TARGET_VECTOR" "@ @@ -5408,8 +5408,8 @@ (set (attr "avl_type") (symbol_ref "INTVAL (operands[8])"))]) (define_insn "*pred_macc_extended_scalar" - [(set (match_operand:VI_D 0 "register_operand" "=vd,?&vd, vr,?&vr") - (if_then_else:VI_D + [(set (match_operand:V_VLSI_D 0 "register_operand" "=vd,?&vd, vr,?&vr") + (if_then_else:V_VLSI_D (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm,Wc1, Wc1") (match_operand 5 "vector_length_operand" " rK, rK, rK, rK") @@ -5418,13 +5418,13 @@ (match_operand 8 "const_int_operand" " i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (plus:VI_D - (mult:VI_D - (vec_duplicate:VI_D + (plus:V_VLSI_D + (mult:V_VLSI_D + (vec_duplicate:V_VLSI_D (sign_extend: (match_operand: 2 "register_operand" " r, r, r, r"))) - (match_operand:VI_D 3 "register_operand" " vr, vr, vr, vr")) - (match_operand:VI_D 4 "register_operand" " 0, vr, 0, vr")) + (match_operand:V_VLSI_D 3 "register_operand" " vr, vr, vr, vr")) + (match_operand:V_VLSI_D 4 "register_operand" " 0, vr, 0, vr")) (match_dup 4)))] "TARGET_VECTOR" "@ @@ -5441,8 +5441,8 @@ (set (attr "avl_type") (symbol_ref "INTVAL (operands[8])"))]) (define_expand "@pred_minus_mul" - [(set (match_operand:VI 0 "register_operand") - (if_then_else:VI + [(set (match_operand:V_VLSI 0 "register_operand") + (if_then_else:V_VLSI (unspec: [(match_operand: 1 "vector_mask_operand") (match_operand 6 "vector_length_operand") @@ -5451,20 +5451,20 @@ (match_operand 9 "const_int_operand") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (minus:VI - (match_operand:VI 4 "register_operand") - (mult:VI - (match_operand:VI 2 "register_operand") - (match_operand:VI 3 "register_operand"))) - (match_operand:VI 5 "vector_merge_operand")))] + (minus:V_VLSI + (match_operand:V_VLSI 4 "register_operand") + (mult:V_VLSI + (match_operand:V_VLSI 2 "register_operand") + (match_operand:V_VLSI 3 "register_operand"))) + (match_operand:V_VLSI 5 "vector_merge_operand")))] "TARGET_VECTOR" { riscv_vector::prepare_ternary_operands (operands); }) (define_insn "*pred_minus_mul_undef" - [(set (match_operand:VI 0 "register_operand" "=vd, vd,?&vd, vr, vr,?&vr") - (if_then_else:VI + [(set (match_operand:V_VLSI 0 "register_operand" "=vd, vd,?&vd, vr, vr,?&vr") + (if_then_else:V_VLSI (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm, vm,Wc1,Wc1, Wc1") (match_operand 6 "vector_length_operand" " rK, rK, rK, rK, rK, rK") @@ -5473,12 +5473,12 @@ (match_operand 9 "const_int_operand" " i, i, i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (minus:VI - (match_operand:VI 5 "register_operand" " vr, 0, vr, vr, 0, vr") - (mult:VI - (match_operand:VI 3 "register_operand" " 0, vr, vr, 0, vr, vr") - (match_operand:VI 4 "register_operand" " vr, vr, vr, vr, vr, vr"))) - (match_operand:VI 2 "vector_undef_operand")))] + (minus:V_VLSI + (match_operand:V_VLSI 5 "register_operand" " vr, 0, vr, vr, 0, vr") + (mult:V_VLSI + (match_operand:V_VLSI 3 "register_operand" " 0, vr, vr, 0, vr, vr") + (match_operand:V_VLSI 4 "register_operand" " vr, vr, vr, vr, vr, vr"))) + (match_operand:V_VLSI 2 "vector_undef_operand")))] "TARGET_VECTOR" "@ vnmsub.vv\t%0,%4,%5%p1 @@ -5491,8 +5491,8 @@ (set_attr "mode" "")]) (define_insn "*pred_nmsub" - [(set (match_operand:VI 0 "register_operand" "=vd,?&vd, vr,?&vr") - (if_then_else:VI + [(set (match_operand:V_VLSI 0 "register_operand" "=vd,?&vd, vr,?&vr") + (if_then_else:V_VLSI (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm,Wc1, Wc1") (match_operand 5 "vector_length_operand" " rK, rK, rK, rK") @@ -5501,11 +5501,11 @@ (match_operand 8 "const_int_operand" " i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (minus:VI - (match_operand:VI 4 "register_operand" " vr, vr, vr, vr") - (mult:VI - (match_operand:VI 2 "register_operand" " 0, vr, 0, vr") - (match_operand:VI 3 "register_operand" " vr, vr, vr, vr"))) + (minus:V_VLSI + (match_operand:V_VLSI 4 "register_operand" " vr, vr, vr, vr") + (mult:V_VLSI + (match_operand:V_VLSI 2 "register_operand" " 0, vr, 0, vr") + (match_operand:V_VLSI 3 "register_operand" " vr, vr, vr, vr"))) (match_dup 2)))] "TARGET_VECTOR" "@ @@ -5522,8 +5522,8 @@ (set (attr "avl_type") (symbol_ref "INTVAL (operands[8])"))]) (define_insn "*pred_nmsac" - [(set (match_operand:VI 0 "register_operand" "=vd,?&vd, vr,?&vr") - (if_then_else:VI + [(set (match_operand:V_VLSI 0 "register_operand" "=vd,?&vd, vr,?&vr") + (if_then_else:V_VLSI (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm,Wc1, Wc1") (match_operand 5 "vector_length_operand" " rK, rK, rK, rK") @@ -5532,11 +5532,11 @@ (match_operand 8 "const_int_operand" " i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (minus:VI - (match_operand:VI 4 "register_operand" " 0, vr, 0, vr") - (mult:VI - (match_operand:VI 2 "register_operand" " vr, vr, vr, vr") - (match_operand:VI 3 "register_operand" " vr, vr, vr, vr"))) + (minus:V_VLSI + (match_operand:V_VLSI 4 "register_operand" " 0, vr, 0, vr") + (mult:V_VLSI + (match_operand:V_VLSI 2 "register_operand" " vr, vr, vr, vr") + (match_operand:V_VLSI 3 "register_operand" " vr, vr, vr, vr"))) (match_dup 4)))] "TARGET_VECTOR" "@ @@ -5553,8 +5553,8 @@ (set (attr "avl_type") (symbol_ref "INTVAL (operands[8])"))]) (define_expand "@pred_minus_mul_scalar" - [(set (match_operand:VI_QHS 0 "register_operand") - (if_then_else:VI_QHS + [(set (match_operand:V_VLSI_QHS 0 "register_operand") + (if_then_else:V_VLSI_QHS (unspec: [(match_operand: 1 "vector_mask_operand") (match_operand 6 "vector_length_operand") @@ -5563,19 +5563,19 @@ (match_operand 9 "const_int_operand") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (minus:VI_QHS - (match_operand:VI_QHS 4 "register_operand") - (mult:VI_QHS - (vec_duplicate:VI_QHS + (minus:V_VLSI_QHS + (match_operand:V_VLSI_QHS 4 "register_operand") + (mult:V_VLSI_QHS + (vec_duplicate:V_VLSI_QHS (match_operand: 2 "register_operand")) - (match_operand:VI_QHS 3 "register_operand"))) - (match_operand:VI_QHS 5 "register_operand")))] + (match_operand:V_VLSI_QHS 3 "register_operand"))) + (match_operand:V_VLSI_QHS 5 "register_operand")))] "TARGET_VECTOR" {}) (define_insn "*pred_nmsub_scalar" - [(set (match_operand:VI 0 "register_operand" "=vd,?&vd, vr,?&vr") - (if_then_else:VI + [(set (match_operand:V_VLSI 0 "register_operand" "=vd,?&vd, vr,?&vr") + (if_then_else:V_VLSI (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm,Wc1, Wc1") (match_operand 5 "vector_length_operand" " rK, rK, rK, rK") @@ -5584,12 +5584,12 @@ (match_operand 8 "const_int_operand" " i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (minus:VI - (match_operand:VI 4 "register_operand" " vr, vr, vr, vr") - (mult:VI - (vec_duplicate:VI + (minus:V_VLSI + (match_operand:V_VLSI 4 "register_operand" " vr, vr, vr, vr") + (mult:V_VLSI + (vec_duplicate:V_VLSI (match_operand: 2 "register_operand" " r, r, r, r")) - (match_operand:VI 3 "register_operand" " 0, vr, 0, vr"))) + (match_operand:V_VLSI 3 "register_operand" " 0, vr, 0, vr"))) (match_dup 3)))] "TARGET_VECTOR" "@ @@ -5606,8 +5606,8 @@ (set (attr "avl_type") (symbol_ref "INTVAL (operands[8])"))]) (define_insn "*pred_nmsac_scalar" - [(set (match_operand:VI 0 "register_operand" "=vd,?&vd, vr,?&vr") - (if_then_else:VI + [(set (match_operand:V_VLSI 0 "register_operand" "=vd,?&vd, vr,?&vr") + (if_then_else:V_VLSI (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm,Wc1, Wc1") (match_operand 5 "vector_length_operand" " rK, rK, rK, rK") @@ -5616,12 +5616,12 @@ (match_operand 8 "const_int_operand" " i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (minus:VI - (match_operand:VI 4 "register_operand" " 0, vr, 0, vr") - (mult:VI - (vec_duplicate:VI + (minus:V_VLSI + (match_operand:V_VLSI 4 "register_operand" " 0, vr, 0, vr") + (mult:V_VLSI + (vec_duplicate:V_VLSI (match_operand: 2 "register_operand" " r, r, r, r")) - (match_operand:VI 3 "register_operand" " vr, vr, vr, vr"))) + (match_operand:V_VLSI 3 "register_operand" " vr, vr, vr, vr"))) (match_dup 4)))] "TARGET_VECTOR" "@ @@ -5638,8 +5638,8 @@ (set (attr "avl_type") (symbol_ref "INTVAL (operands[8])"))]) (define_expand "@pred_minus_mul_scalar" - [(set (match_operand:VI_D 0 "register_operand") - (if_then_else:VI_D + [(set (match_operand:V_VLSI_D 0 "register_operand") + (if_then_else:V_VLSI_D (unspec: [(match_operand: 1 "vector_mask_operand") (match_operand 6 "vector_length_operand") @@ -5648,13 +5648,13 @@ (match_operand 9 "const_int_operand") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (minus:VI_D - (match_operand:VI_D 4 "register_operand") - (mult:VI_D - (vec_duplicate:VI_D + (minus:V_VLSI_D + (match_operand:V_VLSI_D 4 "register_operand") + (mult:V_VLSI_D + (vec_duplicate:V_VLSI_D (match_operand: 2 "reg_or_int_operand")) - (match_operand:VI_D 3 "register_operand"))) - (match_operand:VI_D 5 "register_operand")))] + (match_operand:V_VLSI_D 3 "register_operand"))) + (match_operand:V_VLSI_D 5 "register_operand")))] "TARGET_VECTOR" { if (riscv_vector::sew64_scalar_helper ( @@ -5672,8 +5672,8 @@ }) (define_insn "*pred_nmsub_extended_scalar" - [(set (match_operand:VI_D 0 "register_operand" "=vd,?&vd, vr,?&vr") - (if_then_else:VI_D + [(set (match_operand:V_VLSI_D 0 "register_operand" "=vd,?&vd, vr,?&vr") + (if_then_else:V_VLSI_D (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm,Wc1, Wc1") (match_operand 5 "vector_length_operand" " rK, rK, rK, rK") @@ -5682,13 +5682,13 @@ (match_operand 8 "const_int_operand" " i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (minus:VI_D - (match_operand:VI_D 4 "register_operand" " vr, vr, vr, vr") - (mult:VI_D - (vec_duplicate:VI_D + (minus:V_VLSI_D + (match_operand:V_VLSI_D 4 "register_operand" " vr, vr, vr, vr") + (mult:V_VLSI_D + (vec_duplicate:V_VLSI_D (sign_extend: (match_operand: 2 "register_operand" " r, r, r, r"))) - (match_operand:VI_D 3 "register_operand" " 0, vr, 0, vr"))) + (match_operand:V_VLSI_D 3 "register_operand" " 0, vr, 0, vr"))) (match_dup 3)))] "TARGET_VECTOR" "@ @@ -5705,8 +5705,8 @@ (set (attr "avl_type") (symbol_ref "INTVAL (operands[8])"))]) (define_insn "*pred_nmsac_extended_scalar" - [(set (match_operand:VI_D 0 "register_operand" "=vd,?&vd, vr,?&vr") - (if_then_else:VI_D + [(set (match_operand:V_VLSI_D 0 "register_operand" "=vd,?&vd, vr,?&vr") + (if_then_else:V_VLSI_D (unspec: [(match_operand: 1 "vector_mask_operand" " vm, vm,Wc1, Wc1") (match_operand 5 "vector_length_operand" " rK, rK, rK, rK") @@ -5715,13 +5715,13 @@ (match_operand 8 "const_int_operand" " i, i, i, i") (reg:SI VL_REGNUM) (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE) - (minus:VI_D - (match_operand:VI_D 4 "register_operand" " 0, vr, 0, vr") - (mult:VI_D - (vec_duplicate:VI_D + (minus:V_VLSI_D + (match_operand:V_VLSI_D 4 "register_operand" " 0, vr, 0, vr") + (mult:V_VLSI_D + (vec_duplicate:V_VLSI_D (sign_extend: (match_operand: 2 "register_operand" " r, r, r, r"))) - (match_operand:VI_D 3 "register_operand" " vr, vr, vr, vr"))) + (match_operand:V_VLSI_D 3 "register_operand" " vr, vr, vr, vr"))) (match_dup 4)))] "TARGET_VECTOR" "@ diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h index 4d32de5a61f..7528ee73d9e 100644 --- a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h @@ -457,3 +457,21 @@ typedef double v512df __attribute__ ((vector_size (4096))); \ return v; \ } + +#define DEF_FMA_VV(PREFIX, NUM, TYPE) \ + void __attribute__ ((noinline, noclone)) \ + PREFIX##_##TYPE##NUM (TYPE *restrict a, TYPE *restrict b, TYPE *restrict c, \ + TYPE *restrict d) \ + { \ + for (int i = 0; i < NUM; ++i) \ + a[i] = b[i] * c[i] + d[i]; \ + } + +#define DEF_FNMA_VV(PREFIX, NUM, TYPE) \ + void __attribute__ ((noinline, noclone)) \ + PREFIX##_##TYPE##NUM (TYPE *restrict a, TYPE *restrict b, TYPE *restrict c, \ + TYPE *restrict d) \ + { \ + for (int i = 0; i < NUM; ++i) \ + a[i] = d[i] - b[i] * c[i]; \ + } diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-1.c new file mode 100644 index 00000000000..7f9073a4dc1 --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-1.c @@ -0,0 +1,45 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 --param=riscv-autovec-lmul=m8 -fdump-tree-optimized" } */ + +#include "def.h" + +DEF_FMA_VV (fma, 2, int8_t) +DEF_FMA_VV (fma, 4, int8_t) +DEF_FMA_VV (fma, 8, int8_t) +DEF_FMA_VV (fma, 16, int8_t) +DEF_FMA_VV (fma, 32, int8_t) +DEF_FMA_VV (fma, 64, int8_t) +DEF_FMA_VV (fma, 128, int8_t) +DEF_FMA_VV (fma, 256, int8_t) +DEF_FMA_VV (fma, 512, int8_t) +DEF_FMA_VV (fma, 1024, int8_t) +DEF_FMA_VV (fma, 2048, int8_t) +DEF_FMA_VV (fma, 4096, int8_t) + +DEF_FMA_VV (fma, 2, uint8_t) +DEF_FMA_VV (fma, 4, uint8_t) +DEF_FMA_VV (fma, 8, uint8_t) +DEF_FMA_VV (fma, 16, uint8_t) +DEF_FMA_VV (fma, 32, uint8_t) +DEF_FMA_VV (fma, 64, uint8_t) +DEF_FMA_VV (fma, 128, uint8_t) +DEF_FMA_VV (fma, 256, uint8_t) +DEF_FMA_VV (fma, 512, uint8_t) +DEF_FMA_VV (fma, 1024, uint8_t) +DEF_FMA_VV (fma, 2048, uint8_t) +DEF_FMA_VV (fma, 4096, uint8_t) + +/* { dg-final { scan-assembler-times {vma[c-d][c-d]\.vv} 24 } } */ +/* { dg-final { scan-assembler-not {csrr} } } */ +/* { dg-final { scan-tree-dump-not "1,1" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2,2" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4,4" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "16,16" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "32,32" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "64,64" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "128,128" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "256,256" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "512,512" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "1024,1024" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2048,2048" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4096,4096" "optimized" } } */ diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-2.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-2.c new file mode 100644 index 00000000000..ddc4b552fc7 --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-2.c @@ -0,0 +1,43 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 --param=riscv-autovec-lmul=m8 -fdump-tree-optimized" } */ + +#include "def.h" + +DEF_FMA_VV (fma, 2, int16_t) +DEF_FMA_VV (fma, 4, int16_t) +DEF_FMA_VV (fma, 8, int16_t) +DEF_FMA_VV (fma, 16, int16_t) +DEF_FMA_VV (fma, 32, int16_t) +DEF_FMA_VV (fma, 64, int16_t) +DEF_FMA_VV (fma, 128, int16_t) +DEF_FMA_VV (fma, 256, int16_t) +DEF_FMA_VV (fma, 512, int16_t) +DEF_FMA_VV (fma, 1024, int16_t) +DEF_FMA_VV (fma, 2048, int16_t) + +DEF_FMA_VV (fma, 2, uint16_t) +DEF_FMA_VV (fma, 4, uint16_t) +DEF_FMA_VV (fma, 8, uint16_t) +DEF_FMA_VV (fma, 16, uint16_t) +DEF_FMA_VV (fma, 32, uint16_t) +DEF_FMA_VV (fma, 64, uint16_t) +DEF_FMA_VV (fma, 128, uint16_t) +DEF_FMA_VV (fma, 256, uint16_t) +DEF_FMA_VV (fma, 512, uint16_t) +DEF_FMA_VV (fma, 1024, uint16_t) +DEF_FMA_VV (fma, 2048, uint16_t) + +/* { dg-final { scan-assembler-times {vma[c-d][c-d]\.vv} 22 } } */ +/* { dg-final { scan-assembler-not {csrr} } } */ +/* { dg-final { scan-tree-dump-not "1,1" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2,2" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4,4" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "16,16" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "32,32" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "64,64" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "128,128" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "256,256" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "512,512" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "1024,1024" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2048,2048" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4096,4096" "optimized" } } */ diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-3.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-3.c new file mode 100644 index 00000000000..3bb52ae5974 --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-3.c @@ -0,0 +1,41 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 --param=riscv-autovec-lmul=m8 -fdump-tree-optimized" } */ + +#include "def.h" + +DEF_FMA_VV (fma, 2, int32_t) +DEF_FMA_VV (fma, 4, int32_t) +DEF_FMA_VV (fma, 8, int32_t) +DEF_FMA_VV (fma, 16, int32_t) +DEF_FMA_VV (fma, 32, int32_t) +DEF_FMA_VV (fma, 64, int32_t) +DEF_FMA_VV (fma, 128, int32_t) +DEF_FMA_VV (fma, 256, int32_t) +DEF_FMA_VV (fma, 512, int32_t) +DEF_FMA_VV (fma, 1024, int32_t) + +DEF_FMA_VV (fma, 2, uint32_t) +DEF_FMA_VV (fma, 4, uint32_t) +DEF_FMA_VV (fma, 8, uint32_t) +DEF_FMA_VV (fma, 16, uint32_t) +DEF_FMA_VV (fma, 32, uint32_t) +DEF_FMA_VV (fma, 64, uint32_t) +DEF_FMA_VV (fma, 128, uint32_t) +DEF_FMA_VV (fma, 256, uint32_t) +DEF_FMA_VV (fma, 512, uint32_t) +DEF_FMA_VV (fma, 1024, uint32_t) + +/* { dg-final { scan-assembler-times {vma[c-d][c-d]\.vv} 20 } } */ +/* { dg-final { scan-assembler-not {csrr} } } */ +/* { dg-final { scan-tree-dump-not "1,1" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2,2" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4,4" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "16,16" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "32,32" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "64,64" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "128,128" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "256,256" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "512,512" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "1024,1024" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2048,2048" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4096,4096" "optimized" } } */ diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-4.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-4.c new file mode 100644 index 00000000000..903a4f723e6 --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fma-4.c @@ -0,0 +1,39 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 --param=riscv-autovec-lmul=m8 -fdump-tree-optimized" } */ + +#include "def.h" + +DEF_FMA_VV (fma, 2, int64_t) +DEF_FMA_VV (fma, 4, int64_t) +DEF_FMA_VV (fma, 8, int64_t) +DEF_FMA_VV (fma, 16, int64_t) +DEF_FMA_VV (fma, 32, int64_t) +DEF_FMA_VV (fma, 64, int64_t) +DEF_FMA_VV (fma, 128, int64_t) +DEF_FMA_VV (fma, 256, int64_t) +DEF_FMA_VV (fma, 512, int64_t) + +DEF_FMA_VV (fma, 2, uint64_t) +DEF_FMA_VV (fma, 4, uint64_t) +DEF_FMA_VV (fma, 8, uint64_t) +DEF_FMA_VV (fma, 16, uint64_t) +DEF_FMA_VV (fma, 32, uint64_t) +DEF_FMA_VV (fma, 64, uint64_t) +DEF_FMA_VV (fma, 128, uint64_t) +DEF_FMA_VV (fma, 256, uint64_t) +DEF_FMA_VV (fma, 512, uint64_t) + +/* { dg-final { scan-assembler-times {vma[c-d][c-d]\.vv} 18 } } */ +/* { dg-final { scan-assembler-not {csrr} } } */ +/* { dg-final { scan-tree-dump-not "1,1" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2,2" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4,4" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "16,16" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "32,32" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "64,64" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "128,128" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "256,256" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "512,512" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "1024,1024" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2048,2048" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4096,4096" "optimized" } } */ diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-1.c new file mode 100644 index 00000000000..418c767a7bf --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-1.c @@ -0,0 +1,45 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 --param=riscv-autovec-lmul=m8 -fdump-tree-optimized" } */ + +#include "def.h" + +DEF_FNMA_VV (fnma, 2, int8_t) +DEF_FNMA_VV (fnma, 4, int8_t) +DEF_FNMA_VV (fnma, 8, int8_t) +DEF_FNMA_VV (fnma, 16, int8_t) +DEF_FNMA_VV (fnma, 32, int8_t) +DEF_FNMA_VV (fnma, 64, int8_t) +DEF_FNMA_VV (fnma, 128, int8_t) +DEF_FNMA_VV (fnma, 256, int8_t) +DEF_FNMA_VV (fnma, 512, int8_t) +DEF_FNMA_VV (fnma, 1024, int8_t) +DEF_FNMA_VV (fnma, 2048, int8_t) +DEF_FNMA_VV (fnma, 4096, int8_t) + +DEF_FNMA_VV (fnma, 2, uint8_t) +DEF_FNMA_VV (fnma, 4, uint8_t) +DEF_FNMA_VV (fnma, 8, uint8_t) +DEF_FNMA_VV (fnma, 16, uint8_t) +DEF_FNMA_VV (fnma, 32, uint8_t) +DEF_FNMA_VV (fnma, 64, uint8_t) +DEF_FNMA_VV (fnma, 128, uint8_t) +DEF_FNMA_VV (fnma, 256, uint8_t) +DEF_FNMA_VV (fnma, 512, uint8_t) +DEF_FNMA_VV (fnma, 1024, uint8_t) +DEF_FNMA_VV (fnma, 2048, uint8_t) +DEF_FNMA_VV (fnma, 4096, uint8_t) + +/* { dg-final { scan-assembler-times {vnms[a-u][b-c]\.vv} 24 } } */ +/* { dg-final { scan-assembler-not {csrr} } } */ +/* { dg-final { scan-tree-dump-not "1,1" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2,2" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4,4" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "16,16" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "32,32" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "64,64" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "128,128" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "256,256" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "512,512" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "1024,1024" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2048,2048" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4096,4096" "optimized" } } */ diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-2.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-2.c new file mode 100644 index 00000000000..c1b629a7b22 --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-2.c @@ -0,0 +1,43 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 --param=riscv-autovec-lmul=m8 -fdump-tree-optimized" } */ + +#include "def.h" + +DEF_FNMA_VV (fnma, 2, int16_t) +DEF_FNMA_VV (fnma, 4, int16_t) +DEF_FNMA_VV (fnma, 8, int16_t) +DEF_FNMA_VV (fnma, 16, int16_t) +DEF_FNMA_VV (fnma, 32, int16_t) +DEF_FNMA_VV (fnma, 64, int16_t) +DEF_FNMA_VV (fnma, 128, int16_t) +DEF_FNMA_VV (fnma, 256, int16_t) +DEF_FNMA_VV (fnma, 512, int16_t) +DEF_FNMA_VV (fnma, 1024, int16_t) +DEF_FNMA_VV (fnma, 2048, int16_t) + +DEF_FNMA_VV (fnma, 2, uint16_t) +DEF_FNMA_VV (fnma, 4, uint16_t) +DEF_FNMA_VV (fnma, 8, uint16_t) +DEF_FNMA_VV (fnma, 16, uint16_t) +DEF_FNMA_VV (fnma, 32, uint16_t) +DEF_FNMA_VV (fnma, 64, uint16_t) +DEF_FNMA_VV (fnma, 128, uint16_t) +DEF_FNMA_VV (fnma, 256, uint16_t) +DEF_FNMA_VV (fnma, 512, uint16_t) +DEF_FNMA_VV (fnma, 1024, uint16_t) +DEF_FNMA_VV (fnma, 2048, uint16_t) + +/* { dg-final { scan-assembler-times {vnms[a-u][b-c]\.vv} 22 } } */ +/* { dg-final { scan-assembler-not {csrr} } } */ +/* { dg-final { scan-tree-dump-not "1,1" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2,2" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4,4" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "16,16" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "32,32" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "64,64" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "128,128" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "256,256" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "512,512" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "1024,1024" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2048,2048" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4096,4096" "optimized" } } */ diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-3.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-3.c new file mode 100644 index 00000000000..bab693eda0b --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-3.c @@ -0,0 +1,41 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 --param=riscv-autovec-lmul=m8 -fdump-tree-optimized" } */ + +#include "def.h" + +DEF_FNMA_VV (fnma, 2, int32_t) +DEF_FNMA_VV (fnma, 4, int32_t) +DEF_FNMA_VV (fnma, 8, int32_t) +DEF_FNMA_VV (fnma, 16, int32_t) +DEF_FNMA_VV (fnma, 32, int32_t) +DEF_FNMA_VV (fnma, 64, int32_t) +DEF_FNMA_VV (fnma, 128, int32_t) +DEF_FNMA_VV (fnma, 256, int32_t) +DEF_FNMA_VV (fnma, 512, int32_t) +DEF_FNMA_VV (fnma, 1024, int32_t) + +DEF_FNMA_VV (fnma, 2, uint32_t) +DEF_FNMA_VV (fnma, 4, uint32_t) +DEF_FNMA_VV (fnma, 8, uint32_t) +DEF_FNMA_VV (fnma, 16, uint32_t) +DEF_FNMA_VV (fnma, 32, uint32_t) +DEF_FNMA_VV (fnma, 64, uint32_t) +DEF_FNMA_VV (fnma, 128, uint32_t) +DEF_FNMA_VV (fnma, 256, uint32_t) +DEF_FNMA_VV (fnma, 512, uint32_t) +DEF_FNMA_VV (fnma, 1024, uint32_t) + +/* { dg-final { scan-assembler-times {vnms[a-u][b-c]\.vv} 20 } } */ +/* { dg-final { scan-assembler-not {csrr} } } */ +/* { dg-final { scan-tree-dump-not "1,1" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2,2" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4,4" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "16,16" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "32,32" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "64,64" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "128,128" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "256,256" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "512,512" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "1024,1024" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2048,2048" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4096,4096" "optimized" } } */ diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-4.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-4.c new file mode 100644 index 00000000000..f0a7c5dee30 --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/fnma-4.c @@ -0,0 +1,39 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 --param=riscv-autovec-lmul=m8 -fdump-tree-optimized" } */ + +#include "def.h" + +DEF_FNMA_VV (fnma, 2, int64_t) +DEF_FNMA_VV (fnma, 4, int64_t) +DEF_FNMA_VV (fnma, 8, int64_t) +DEF_FNMA_VV (fnma, 16, int64_t) +DEF_FNMA_VV (fnma, 32, int64_t) +DEF_FNMA_VV (fnma, 64, int64_t) +DEF_FNMA_VV (fnma, 128, int64_t) +DEF_FNMA_VV (fnma, 256, int64_t) +DEF_FNMA_VV (fnma, 512, int64_t) + +DEF_FNMA_VV (fnma, 2, uint64_t) +DEF_FNMA_VV (fnma, 4, uint64_t) +DEF_FNMA_VV (fnma, 8, uint64_t) +DEF_FNMA_VV (fnma, 16, uint64_t) +DEF_FNMA_VV (fnma, 32, uint64_t) +DEF_FNMA_VV (fnma, 64, uint64_t) +DEF_FNMA_VV (fnma, 128, uint64_t) +DEF_FNMA_VV (fnma, 256, uint64_t) +DEF_FNMA_VV (fnma, 512, uint64_t) + +/* { dg-final { scan-assembler-times {vnms[a-u][b-c]\.vv} 18 } } */ +/* { dg-final { scan-assembler-not {csrr} } } */ +/* { dg-final { scan-tree-dump-not "1,1" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2,2" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4,4" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "16,16" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "32,32" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "64,64" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "128,128" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "256,256" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "512,512" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "1024,1024" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "2048,2048" "optimized" } } */ +/* { dg-final { scan-tree-dump-not "4096,4096" "optimized" } } */