From patchwork Fri Dec 23 18:00:07 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: =?utf-8?q?Christoph_M=C3=BCllner?= X-Patchwork-Id: 13081050 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 0675EC4332F for ; Fri, 23 Dec 2022 18:05:50 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1p8mLd-00086c-Lh; Fri, 23 Dec 2022 13:00:45 -0500 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1p8mLR-00080G-Qb for qemu-devel@nongnu.org; Fri, 23 Dec 2022 13:00:34 -0500 Received: from mail-ej1-x633.google.com ([2a00:1450:4864:20::633]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1p8mLP-0003LJ-QH for qemu-devel@nongnu.org; Fri, 23 Dec 2022 13:00:33 -0500 Received: by mail-ej1-x633.google.com with SMTP id u19so13510930ejm.8 for ; Fri, 23 Dec 2022 10:00:30 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=vrull.eu; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=ILaYvgfeEcTF/kKZX/Nqin+7BggI+czN6sJ0PIGCsv0=; b=THR44tTKsEmJtU3D1bbEv4iLKsGb45Q5sDC4BymxpXXsoHSY0YZ4tfAa5LT+IGZS/K 6SP0zGuNWQjFhYbInuy4Fg/FdZLZLbFHBC2A8auipMWusQwoTqlwnb92ZoBZvwTWCshQ bqNGoE/nD8bJ3njhxr9OLOZTcJgil9CpZMZN6DfHG/EWOlyv5VD7fCf8/+wCvGhGB81m lyDTim9u8HeEwRha/xhv9ASUQOD5WCdaIBt9ZpUGa6xIo1jqqUhjcUH3x/m55jeSm5qs RSFyjjmRmDK+aY0T2VB27JVIeZMAVFidilYhE6l6fHA3EZNmCXovtY/Y3OBAGaDMmFlW H5Pw== 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=ILaYvgfeEcTF/kKZX/Nqin+7BggI+czN6sJ0PIGCsv0=; b=YdwT01KfgJDYMedLNR2g12Fzhh1WWOsoXojoVDD8XRZs2npqSonMzbZaGP/2eHnmsk 902hHUooHZ8cTeWduXPt+XTgf7B0Wi/oOL75ugfxzrzP9+TEIAYVuxLLGhOE8uuCwg0b gntDR6I7ANlOflyPY47DNDxAMK8Mp9xARXGIJGmzTimzgwRHB0Fr+kCmbt9p3/lIW9f/ tbYWcFi2d0rviD9yd9+bmfwG/rGr+QI4TeGbR+4EbYPiUOOc66ieMsXr4DsjhmGmXLfV d0nWc9d81XQy+LS8qDlJCNNl9/pxngAeUYyG5bsFe3wgcJ0TheXKCWSQwcu+lzC9FpE3 fJJA== X-Gm-Message-State: AFqh2kp+qIKds1B3/5rM9U7iw2s0dMw1Mp+YR0t7zCPxLzH+mYvDmEC5 s7PovncCojVkeSK3a16MjTDzgw== X-Google-Smtp-Source: AMrXdXsmkjYL+7H/XijhAH8F0hVBu/1MgeTc/Htta1O1CWgI09F++Zzm0dW0jbnMC32G9SskuUjVCA== X-Received: by 2002:a17:906:b108:b0:843:a9fe:f115 with SMTP id u8-20020a170906b10800b00843a9fef115mr7008226ejy.32.1671818429849; Fri, 23 Dec 2022 10:00:29 -0800 (PST) Received: from beast.fritz.box (62-178-148-172.cable.dynamic.surfer.at. [62.178.148.172]) by smtp.gmail.com with ESMTPSA id p14-20020a056402154e00b0047d4685878esm1734937edx.51.2022.12.23.10.00.28 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 23 Dec 2022 10:00:29 -0800 (PST) From: Christoph Muellner To: qemu-riscv@nongnu.org, qemu-devel@nongnu.org, Alistair Francis , Bin Meng , Philipp Tomsich , =?utf-8?q?Heiko_St=C3=BCbner?= , Palmer Dabbelt , Richard Henderson , Nelson Chu , Kito Cheng , Cooper Qu , Lifang Xia , Yunhai Shang , Zhiwei Liu Cc: =?utf-8?q?Christoph_M=C3=BCllner?= Subject: [PATCH v2 07/15] RISC-V: Adding T-Head multiply-accumulate instructions Date: Fri, 23 Dec 2022 19:00:07 +0100 Message-Id: <20221223180016.2068508-8-christoph.muellner@vrull.eu> X-Mailer: git-send-email 2.38.1 In-Reply-To: <20221223180016.2068508-1-christoph.muellner@vrull.eu> References: <20221223180016.2068508-1-christoph.muellner@vrull.eu> MIME-Version: 1.0 Received-SPF: pass client-ip=2a00:1450:4864:20::633; envelope-from=christoph.muellner@vrull.eu; helo=mail-ej1-x633.google.com X-Spam_score_int: -20 X-Spam_score: -2.1 X-Spam_bar: -- X-Spam_report: (-2.1 / 5.0 requ) BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_NONE=-0.0001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org From: Christoph Müllner This patch adds support for the T-Head MAC instructions. The patch uses the T-Head specific decoder and translation. Changes in v2: - Add ISA_EXT_DATA_ENTRY() - Use single decoder for XThead extensions Co-developed-by: LIU Zhiwei Signed-off-by: Christoph Müllner Reviewed-by: Alistair Francis --- target/riscv/cpu.c | 2 + target/riscv/cpu.h | 1 + target/riscv/insn_trans/trans_xthead.c.inc | 83 ++++++++++++++++++++++ target/riscv/translate.c | 3 +- target/riscv/xthead.decode | 8 +++ 5 files changed, 96 insertions(+), 1 deletion(-) diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index 36a53784dd..88ad2138db 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -113,6 +113,7 @@ static const struct isa_ext_data isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(xtheadbs, true, PRIV_VERSION_1_11_0, ext_xtheadbs), ISA_EXT_DATA_ENTRY(xtheadcmo, true, PRIV_VERSION_1_11_0, ext_xtheadcmo), ISA_EXT_DATA_ENTRY(xtheadcondmov, true, PRIV_VERSION_1_11_0, ext_xtheadcondmov), + ISA_EXT_DATA_ENTRY(xtheadmac, true, PRIV_VERSION_1_11_0, ext_xtheadmac), ISA_EXT_DATA_ENTRY(xtheadsync, true, PRIV_VERSION_1_11_0, ext_xtheadsync), ISA_EXT_DATA_ENTRY(xventanacondops, true, PRIV_VERSION_1_12_0, ext_XVentanaCondOps), }; @@ -1071,6 +1072,7 @@ static Property riscv_cpu_extensions[] = { DEFINE_PROP_BOOL("xtheadbs", RISCVCPU, cfg.ext_xtheadbs, false), DEFINE_PROP_BOOL("xtheadcmo", RISCVCPU, cfg.ext_xtheadcmo, false), DEFINE_PROP_BOOL("xtheadcondmov", RISCVCPU, cfg.ext_xtheadcondmov, false), + DEFINE_PROP_BOOL("xtheadmac", RISCVCPU, cfg.ext_xtheadmac, false), DEFINE_PROP_BOOL("xtheadsync", RISCVCPU, cfg.ext_xtheadsync, false), DEFINE_PROP_BOOL("xventanacondops", RISCVCPU, cfg.ext_XVentanaCondOps, false), diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h index 01f035d8e9..92198be9d8 100644 --- a/target/riscv/cpu.h +++ b/target/riscv/cpu.h @@ -470,6 +470,7 @@ struct RISCVCPUConfig { bool ext_xtheadbs; bool ext_xtheadcmo; bool ext_xtheadcondmov; + bool ext_xtheadmac; bool ext_xtheadsync; bool ext_XVentanaCondOps; diff --git a/target/riscv/insn_trans/trans_xthead.c.inc b/target/riscv/insn_trans/trans_xthead.c.inc index bf549bbd74..109be58c9b 100644 --- a/target/riscv/insn_trans/trans_xthead.c.inc +++ b/target/riscv/insn_trans/trans_xthead.c.inc @@ -46,6 +46,12 @@ } \ } while (0) +#define REQUIRE_XTHEADMAC(ctx) do { \ + if (!ctx->cfg_ptr->ext_xtheadmac) { \ + return false; \ + } \ +} while (0) + #define REQUIRE_XTHEADSYNC(ctx) do { \ if (!ctx->cfg_ptr->ext_xtheadsync) { \ return false; \ @@ -307,6 +313,83 @@ static bool trans_th_mvnez(DisasContext *ctx, arg_th_mveqz *a) return gen_th_condmove(ctx, a, TCG_COND_NE); } +/* XTheadMac */ + +static bool gen_th_mac(DisasContext *ctx, arg_r *a, + void (*accumulate_func)(TCGv, TCGv, TCGv), + void (*extend_operand_func)(TCGv, TCGv)) +{ + TCGv dest = dest_gpr(ctx, a->rd); + TCGv src0 = get_gpr(ctx, a->rd, EXT_NONE); + TCGv src1 = get_gpr(ctx, a->rs1, EXT_NONE); + TCGv src2 = get_gpr(ctx, a->rs2, EXT_NONE); + TCGv tmp = tcg_temp_new(); + + if (extend_operand_func) { + TCGv tmp2 = tcg_temp_new(); + extend_operand_func(tmp, src1); + extend_operand_func(tmp2, src2); + tcg_gen_mul_tl(tmp, tmp, tmp2); + tcg_temp_free(tmp2); + } else { + tcg_gen_mul_tl(tmp, src1, src2); + } + + accumulate_func(dest, src0, tmp); + gen_set_gpr(ctx, a->rd, dest); + tcg_temp_free(tmp); + + return true; +} + +/* th.mula: "rd = rd + rs1 * rs2" */ +static bool trans_th_mula(DisasContext *ctx, arg_th_mula *a) +{ + REQUIRE_XTHEADMAC(ctx); + return gen_th_mac(ctx, a, tcg_gen_add_tl, NULL); +} + +/* th.mulah: "rd = sext.w(rd + sext.w(rs1[15:0]) * sext.w(rs2[15:0]))" */ +static bool trans_th_mulah(DisasContext *ctx, arg_th_mulah *a) +{ + REQUIRE_XTHEADMAC(ctx); + ctx->ol = MXL_RV32; + return gen_th_mac(ctx, a, tcg_gen_add_tl, tcg_gen_ext16s_tl); +} + +/* th.mulaw: "rd = sext.w(rd + rs1 * rs2)" */ +static bool trans_th_mulaw(DisasContext *ctx, arg_th_mulaw *a) +{ + REQUIRE_XTHEADMAC(ctx); + REQUIRE_64BIT(ctx); + ctx->ol = MXL_RV32; + return gen_th_mac(ctx, a, tcg_gen_add_tl, NULL); +} + +/* th.muls: "rd = rd - rs1 * rs2" */ +static bool trans_th_muls(DisasContext *ctx, arg_th_muls *a) +{ + REQUIRE_XTHEADMAC(ctx); + return gen_th_mac(ctx, a, tcg_gen_sub_tl, NULL); +} + +/* th.mulsh: "rd = sext.w(rd - sext.w(rs1[15:0]) * sext.w(rs2[15:0]))" */ +static bool trans_th_mulsh(DisasContext *ctx, arg_th_mulsh *a) +{ + REQUIRE_XTHEADMAC(ctx); + ctx->ol = MXL_RV32; + return gen_th_mac(ctx, a, tcg_gen_sub_tl, tcg_gen_ext16s_tl); +} + +/* th.mulsw: "rd = sext.w(rd - rs1 * rs2)" */ +static bool trans_th_mulsw(DisasContext *ctx, arg_th_mulsw *a) +{ + REQUIRE_XTHEADMAC(ctx); + REQUIRE_64BIT(ctx); + ctx->ol = MXL_RV32; + return gen_th_mac(ctx, a, tcg_gen_sub_tl, NULL); +} + /* XTheadSync */ static bool trans_th_sfence_vmas(DisasContext *ctx, arg_th_sfence_vmas *a) diff --git a/target/riscv/translate.c b/target/riscv/translate.c index f15883b16b..36f512baa8 100644 --- a/target/riscv/translate.c +++ b/target/riscv/translate.c @@ -129,7 +129,8 @@ static bool has_xthead_p(DisasContext *ctx __attribute__((__unused__))) { return ctx->cfg_ptr->ext_xtheadba || ctx->cfg_ptr->ext_xtheadbb || ctx->cfg_ptr->ext_xtheadbs || ctx->cfg_ptr->ext_xtheadcmo || - ctx->cfg_ptr->ext_xtheadcondmov || ctx->cfg_ptr->ext_xtheadsync; + ctx->cfg_ptr->ext_xtheadcondmov || ctx->cfg_ptr->ext_xtheadmac || + ctx->cfg_ptr->ext_xtheadsync; } #define MATERIALISE_EXT_PREDICATE(ext) \ diff --git a/target/riscv/xthead.decode b/target/riscv/xthead.decode index a8ebd8a18b..696de6cecf 100644 --- a/target/riscv/xthead.decode +++ b/target/riscv/xthead.decode @@ -88,6 +88,14 @@ th_l2cache_iall 0000000 10110 00000 000 00000 0001011 th_mveqz 0100000 ..... ..... 001 ..... 0001011 @r th_mvnez 0100001 ..... ..... 001 ..... 0001011 @r +# XTheadMac +th_mula 00100 00 ..... ..... 001 ..... 0001011 @r +th_mulah 00101 00 ..... ..... 001 ..... 0001011 @r +th_mulaw 00100 10 ..... ..... 001 ..... 0001011 @r +th_muls 00100 01 ..... ..... 001 ..... 0001011 @r +th_mulsh 00101 01 ..... ..... 001 ..... 0001011 @r +th_mulsw 00100 11 ..... ..... 001 ..... 0001011 @r + # XTheadSync th_sfence_vmas 0000010 ..... ..... 000 00000 0001011 @rs2_s th_sync 0000000 11000 00000 000 00000 0001011