From patchwork Fri Oct 13 21:28:15 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Richard Henderson X-Patchwork-Id: 1848666 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=linaro.org header.i=@linaro.org header.a=rsa-sha256 header.s=google header.b=abvPZfgu; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4S6fzd6hrNz1yqm for ; Sat, 14 Oct 2023 08:39:41 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qrPm1-0000V2-CP; Fri, 13 Oct 2023 17:32:45 -0400 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 1qrPlz-0000HF-4a for qemu-devel@nongnu.org; Fri, 13 Oct 2023 17:32:43 -0400 Received: from mail-pf1-x433.google.com ([2607:f8b0:4864:20::433]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1qrPlw-0001PH-6y for qemu-devel@nongnu.org; Fri, 13 Oct 2023 17:32:42 -0400 Received: by mail-pf1-x433.google.com with SMTP id d2e1a72fcca58-6b89ab5ddb7so250405b3a.0 for ; Fri, 13 Oct 2023 14:32:39 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; t=1697232758; x=1697837558; darn=nongnu.org; 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=za8xgsP66RnIsX8XSuhueYFTe+nOiG3yFbpulQj0Di4=; b=abvPZfgu+btOo0d4lVBofD7GLFQrFocrur3h71w97FRDzV6I9hO06hLWpBJmdkAMfY di1fCqY03jIKOGeW06K03b0hSKn8DaRuYkQGRtk/6HPAg5fx12CUKoXNeWkGa+V0uTsm HuGMRhvdFUlkGnI3ULWJAQu2LtIGlrCpqrdc6ml351g4Hi309UhKmCnHZQubO5wIj5vd Z+Tl10GNqYu7y39qdUnWZa3isddmmKaz+1BUYXv97NviUZ+iSq4Wdgbw1QXw9tzvelWr CVHx994eMY2azOWNWkrwQJrgqWKFQmkfzDiXL/7lHUpH3Tqe5HhGlUcfTec5syUw7dRm FMfw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1697232758; x=1697837558; 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=za8xgsP66RnIsX8XSuhueYFTe+nOiG3yFbpulQj0Di4=; b=T1w0bBeUcCr6pTeDiJCe5AFLtGt0NBUi1b2FaxWOQvc2VsSG+2SLd5Esdwn5PUqPZK Z4uh+TsakqiJd4sXSB4w0iGhc1DHQ/tvkmuU7VyzZPhRZ0HMgAsTUoqN2QSxNXKVBkNC PFs4/HRNpMFTj76c8s4z/EHR80qcj4H332gdzdrQ7+D8Vaa2xd6e3ip1KuH6QskINWBT zIFOgd/oXos/qmIXV4CXqzVZWSGSN1kA6qOysNtrDbUc5MpoMmjqXcU1Ai+gcd3ZUlnq h7SUImf7DqAPbtMWfLtdWZ7NCbak85ywiqbjGzDCttB/FEEWnQE8n8WZWiTd/jwbiPf2 nJlQ== X-Gm-Message-State: AOJu0YxbN6eSBw9nCpQnKxpKq8uOQ38k7Koyq1l8U5CM+eEfUdjhDxQf uxXOfjTTeNR2iLBp04NKsSA7ItC27FO5XXG8E8M= X-Google-Smtp-Source: AGHT+IF2rT3NRdlX74z5bYTU13XzzKdtCmx14Vg0hgtxGTtRy2Sa2aN8fBWdlfcOUVMmdpmJkC8/cQ== X-Received: by 2002:a05:6a00:198f:b0:68f:a661:d037 with SMTP id d15-20020a056a00198f00b0068fa661d037mr30815527pfl.3.1697232758407; Fri, 13 Oct 2023 14:32:38 -0700 (PDT) Received: from stoup.. ([71.212.149.95]) by smtp.gmail.com with ESMTPSA id j5-20020aa78d05000000b0068842ebfd10sm13977191pfe.160.2023.10.13.14.32.37 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 13 Oct 2023 14:32:37 -0700 (PDT) From: Richard Henderson To: qemu-devel@nongnu.org Cc: mark.cave-ayland@ilande.co.uk, atar4qemu@gmail.com Subject: [PATCH 54/85] target/sparc: Move EDGE* to decodetree Date: Fri, 13 Oct 2023 14:28:15 -0700 Message-Id: <20231013212846.165724-55-richard.henderson@linaro.org> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231013212846.165724-1-richard.henderson@linaro.org> References: <20231013212846.165724-1-richard.henderson@linaro.org> MIME-Version: 1.0 Received-SPF: pass client-ip=2607:f8b0:4864:20::433; envelope-from=richard.henderson@linaro.org; helo=mail-pf1-x433.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+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Signed-off-by: Richard Henderson --- target/sparc/insns.decode | 22 ++- target/sparc/translate.c | 272 ++++++++++++++++---------------------- 2 files changed, 134 insertions(+), 160 deletions(-) diff --git a/target/sparc/insns.decode b/target/sparc/insns.decode index 185bcadab7..b4bbdd43ec 100644 --- a/target/sparc/insns.decode +++ b/target/sparc/insns.decode @@ -30,6 +30,9 @@ CALL 01 i:s30 @n_r_ri .. ..... ...... rs1:5 imm:1 rs2_or_imm:s13 &r_r_ri rd=0 @r_r_ri .. rd:5 ...... rs1:5 imm:1 rs2_or_imm:s13 &r_r_ri +&r_r_r rd rs1 rs2 +@r_r_r .. rd:5 ...... rs1:5 . ........ rs2:5 &r_r_r + Tcc_r 10 0 cond:4 111010 rs1:5 0 cc:1 0000000 rs2:5 Tcc_i 10 0 cond:4 111010 rs1:5 1 cc:1 0000 i:8 @@ -225,7 +228,24 @@ RESTORE 10 ..... 111101 ..... . ............. @r_r_ri DONE 10 00000 111110 00000 0 0000000000000 RETRY 10 00001 111110 00000 0 0000000000000 -NCP 10 ----- 110110 ----- --------- ----- # v8 CPop1 +{ + [ + EDGE8cc 10 ..... 110110 ..... 0 0000 0000 ..... @r_r_r + EDGE8N 10 ..... 110110 ..... 0 0000 0001 ..... @r_r_r + EDGE8Lcc 10 ..... 110110 ..... 0 0000 0010 ..... @r_r_r + EDGE8LN 10 ..... 110110 ..... 0 0000 0011 ..... @r_r_r + EDGE16cc 10 ..... 110110 ..... 0 0000 0100 ..... @r_r_r + EDGE16N 10 ..... 110110 ..... 0 0000 0101 ..... @r_r_r + EDGE16Lcc 10 ..... 110110 ..... 0 0000 0110 ..... @r_r_r + EDGE16LN 10 ..... 110110 ..... 0 0000 0111 ..... @r_r_r + EDGE32cc 10 ..... 110110 ..... 0 0000 1000 ..... @r_r_r + EDGE32N 10 ..... 110110 ..... 0 0000 1001 ..... @r_r_r + EDGE32Lcc 10 ..... 110110 ..... 0 0000 1010 ..... @r_r_r + EDGE32LN 10 ..... 110110 ..... 0 0000 1011 ..... @r_r_r + ] + NCP 10 ----- 110110 ----- --------- ----- # v8 CPop1 +} + NCP 10 ----- 110111 ----- --------- ----- # v8 CPop2 ## diff --git a/target/sparc/translate.c b/target/sparc/translate.c index 6598557bb5..56f32551af 100644 --- a/target/sparc/translate.c +++ b/target/sparc/translate.c @@ -2620,93 +2620,6 @@ static void gen_load_trap_state_at_tl(TCGv_ptr r_tsptr) } } -static void gen_edge(DisasContext *dc, TCGv dst, TCGv s1, TCGv s2, - int width, bool cc, bool left) -{ - TCGv lo1, lo2; - uint64_t amask, tabl, tabr; - int shift, imask, omask; - - if (cc) { - tcg_gen_mov_tl(cpu_cc_src, s1); - tcg_gen_mov_tl(cpu_cc_src2, s2); - tcg_gen_sub_tl(cpu_cc_dst, s1, s2); - tcg_gen_movi_i32(cpu_cc_op, CC_OP_SUB); - dc->cc_op = CC_OP_SUB; - } - - /* Theory of operation: there are two tables, left and right (not to - be confused with the left and right versions of the opcode). These - are indexed by the low 3 bits of the inputs. To make things "easy", - these tables are loaded into two constants, TABL and TABR below. - The operation index = (input & imask) << shift calculates the index - into the constant, while val = (table >> index) & omask calculates - the value we're looking for. */ - switch (width) { - case 8: - imask = 0x7; - shift = 3; - omask = 0xff; - if (left) { - tabl = 0x80c0e0f0f8fcfeffULL; - tabr = 0xff7f3f1f0f070301ULL; - } else { - tabl = 0x0103070f1f3f7fffULL; - tabr = 0xfffefcf8f0e0c080ULL; - } - break; - case 16: - imask = 0x6; - shift = 1; - omask = 0xf; - if (left) { - tabl = 0x8cef; - tabr = 0xf731; - } else { - tabl = 0x137f; - tabr = 0xfec8; - } - break; - case 32: - imask = 0x4; - shift = 0; - omask = 0x3; - if (left) { - tabl = (2 << 2) | 3; - tabr = (3 << 2) | 1; - } else { - tabl = (1 << 2) | 3; - tabr = (3 << 2) | 2; - } - break; - default: - abort(); - } - - lo1 = tcg_temp_new(); - lo2 = tcg_temp_new(); - tcg_gen_andi_tl(lo1, s1, imask); - tcg_gen_andi_tl(lo2, s2, imask); - tcg_gen_shli_tl(lo1, lo1, shift); - tcg_gen_shli_tl(lo2, lo2, shift); - - tcg_gen_shr_tl(lo1, tcg_constant_tl(tabl), lo1); - tcg_gen_shr_tl(lo2, tcg_constant_tl(tabr), lo2); - tcg_gen_andi_tl(lo1, lo1, omask); - tcg_gen_andi_tl(lo2, lo2, omask); - - amask = -8; - if (AM_CHECK(dc)) { - amask &= 0xffffffffULL; - } - tcg_gen_andi_tl(s1, s1, amask); - tcg_gen_andi_tl(s2, s2, amask); - - /* Compute dst = (s1 == s2 ? lo1 : lo1 & lo2). */ - tcg_gen_and_tl(lo2, lo2, lo1); - tcg_gen_movcond_tl(TCG_COND_EQ, dst, s1, s2, lo1, lo2); -} - static void gen_alignaddr(TCGv dst, TCGv s1, TCGv s2, bool left) { TCGv tmp = tcg_temp_new(); @@ -2771,6 +2684,8 @@ static int extract_qfpreg(DisasContext *dc, int x) #define avail_GL(C) ((C)->def->features & CPU_FEATURE_GL) #define avail_HYPV(C) ((C)->def->features & CPU_FEATURE_HYPV) #define avail_POWERDOWN(C) ((C)->def->features & CPU_FEATURE_POWERDOWN) +#define avail_VIS1(C) ((C)->def->features & CPU_FEATURE_VIS1) +#define avail_VIS2(C) ((C)->def->features & CPU_FEATURE_VIS2) /* Default case for non jump instructions. */ static bool advance_pc(DisasContext *dc) @@ -4219,6 +4134,116 @@ static bool trans_MULScc(DisasContext *dc, arg_r_r_ri *a) return do_cc_arith(dc, a, CC_OP_ADD, gen_op_mulscc, NULL); } +static bool gen_edge(DisasContext *dc, arg_r_r_r *a, + int width, bool cc, bool left) +{ + TCGv dst, s1, s2, lo1, lo2; + uint64_t amask, tabl, tabr; + int shift, imask, omask; + + dst = gen_dest_gpr(dc, a->rd); + s1 = gen_load_gpr(dc, a->rs1); + s2 = gen_load_gpr(dc, a->rs2); + + if (cc) { + tcg_gen_mov_tl(cpu_cc_src, s1); + tcg_gen_mov_tl(cpu_cc_src2, s2); + tcg_gen_sub_tl(cpu_cc_dst, s1, s2); + tcg_gen_movi_i32(cpu_cc_op, CC_OP_SUB); + dc->cc_op = CC_OP_SUB; + } + + /* + * Theory of operation: there are two tables, left and right (not to + * be confused with the left and right versions of the opcode). These + * are indexed by the low 3 bits of the inputs. To make things "easy", + * these tables are loaded into two constants, TABL and TABR below. + * The operation index = (input & imask) << shift calculates the index + * into the constant, while val = (table >> index) & omask calculates + * the value we're looking for. + */ + switch (width) { + case 8: + imask = 0x7; + shift = 3; + omask = 0xff; + if (left) { + tabl = 0x80c0e0f0f8fcfeffULL; + tabr = 0xff7f3f1f0f070301ULL; + } else { + tabl = 0x0103070f1f3f7fffULL; + tabr = 0xfffefcf8f0e0c080ULL; + } + break; + case 16: + imask = 0x6; + shift = 1; + omask = 0xf; + if (left) { + tabl = 0x8cef; + tabr = 0xf731; + } else { + tabl = 0x137f; + tabr = 0xfec8; + } + break; + case 32: + imask = 0x4; + shift = 0; + omask = 0x3; + if (left) { + tabl = (2 << 2) | 3; + tabr = (3 << 2) | 1; + } else { + tabl = (1 << 2) | 3; + tabr = (3 << 2) | 2; + } + break; + default: + abort(); + } + + lo1 = tcg_temp_new(); + lo2 = tcg_temp_new(); + tcg_gen_andi_tl(lo1, s1, imask); + tcg_gen_andi_tl(lo2, s2, imask); + tcg_gen_shli_tl(lo1, lo1, shift); + tcg_gen_shli_tl(lo2, lo2, shift); + + tcg_gen_shr_tl(lo1, tcg_constant_tl(tabl), lo1); + tcg_gen_shr_tl(lo2, tcg_constant_tl(tabr), lo2); + tcg_gen_andi_tl(lo1, lo1, omask); + tcg_gen_andi_tl(lo2, lo2, omask); + + amask = -8; + if (AM_CHECK(dc)) { + amask &= 0xffffffffULL; + } + tcg_gen_andi_tl(s1, s1, amask); + tcg_gen_andi_tl(s2, s2, amask); + + /* Compute dst = (s1 == s2 ? lo1 : lo1 & lo2). */ + tcg_gen_and_tl(lo2, lo2, lo1); + tcg_gen_movcond_tl(TCG_COND_EQ, dst, s1, s2, lo1, lo2); + + gen_store_gpr(dc, a->rd, dst); + return advance_pc(dc); +} + +TRANS(EDGE8cc, VIS1, gen_edge, a, 8, 1, 0) +TRANS(EDGE8Lcc, VIS1, gen_edge, a, 8, 1, 1) +TRANS(EDGE16cc, VIS1, gen_edge, a, 16, 1, 0) +TRANS(EDGE16Lcc, VIS1, gen_edge, a, 16, 1, 1) +TRANS(EDGE32cc, VIS1, gen_edge, a, 32, 1, 0) +TRANS(EDGE32Lcc, VIS1, gen_edge, a, 32, 1, 1) + +TRANS(EDGE8N, VIS2, gen_edge, a, 8, 0, 0) +TRANS(EDGE8LN, VIS2, gen_edge, a, 8, 0, 1) +TRANS(EDGE16N, VIS2, gen_edge, a, 16, 0, 0) +TRANS(EDGE16LN, VIS2, gen_edge, a, 16, 0, 1) +TRANS(EDGE32N, VIS2, gen_edge, a, 32, 0, 0) +TRANS(EDGE32LN, VIS2, gen_edge, a, 32, 0, 1) + static bool do_shift_r(DisasContext *dc, arg_shiftr *a, bool l, bool u) { TCGv dst, src1, src2; @@ -5097,89 +5122,18 @@ static void disas_sparc_legacy(DisasContext *dc, unsigned int insn) switch (opf) { case 0x000: /* VIS I edge8cc */ - CHECK_FPU_FEATURE(dc, VIS1); - cpu_src1 = gen_load_gpr(dc, rs1); - cpu_src2 = gen_load_gpr(dc, rs2); - gen_edge(dc, cpu_dst, cpu_src1, cpu_src2, 8, 1, 0); - gen_store_gpr(dc, rd, cpu_dst); - break; case 0x001: /* VIS II edge8n */ - CHECK_FPU_FEATURE(dc, VIS2); - cpu_src1 = gen_load_gpr(dc, rs1); - cpu_src2 = gen_load_gpr(dc, rs2); - gen_edge(dc, cpu_dst, cpu_src1, cpu_src2, 8, 0, 0); - gen_store_gpr(dc, rd, cpu_dst); - break; case 0x002: /* VIS I edge8lcc */ - CHECK_FPU_FEATURE(dc, VIS1); - cpu_src1 = gen_load_gpr(dc, rs1); - cpu_src2 = gen_load_gpr(dc, rs2); - gen_edge(dc, cpu_dst, cpu_src1, cpu_src2, 8, 1, 1); - gen_store_gpr(dc, rd, cpu_dst); - break; case 0x003: /* VIS II edge8ln */ - CHECK_FPU_FEATURE(dc, VIS2); - cpu_src1 = gen_load_gpr(dc, rs1); - cpu_src2 = gen_load_gpr(dc, rs2); - gen_edge(dc, cpu_dst, cpu_src1, cpu_src2, 8, 0, 1); - gen_store_gpr(dc, rd, cpu_dst); - break; case 0x004: /* VIS I edge16cc */ - CHECK_FPU_FEATURE(dc, VIS1); - cpu_src1 = gen_load_gpr(dc, rs1); - cpu_src2 = gen_load_gpr(dc, rs2); - gen_edge(dc, cpu_dst, cpu_src1, cpu_src2, 16, 1, 0); - gen_store_gpr(dc, rd, cpu_dst); - break; case 0x005: /* VIS II edge16n */ - CHECK_FPU_FEATURE(dc, VIS2); - cpu_src1 = gen_load_gpr(dc, rs1); - cpu_src2 = gen_load_gpr(dc, rs2); - gen_edge(dc, cpu_dst, cpu_src1, cpu_src2, 16, 0, 0); - gen_store_gpr(dc, rd, cpu_dst); - break; case 0x006: /* VIS I edge16lcc */ - CHECK_FPU_FEATURE(dc, VIS1); - cpu_src1 = gen_load_gpr(dc, rs1); - cpu_src2 = gen_load_gpr(dc, rs2); - gen_edge(dc, cpu_dst, cpu_src1, cpu_src2, 16, 1, 1); - gen_store_gpr(dc, rd, cpu_dst); - break; case 0x007: /* VIS II edge16ln */ - CHECK_FPU_FEATURE(dc, VIS2); - cpu_src1 = gen_load_gpr(dc, rs1); - cpu_src2 = gen_load_gpr(dc, rs2); - gen_edge(dc, cpu_dst, cpu_src1, cpu_src2, 16, 0, 1); - gen_store_gpr(dc, rd, cpu_dst); - break; case 0x008: /* VIS I edge32cc */ - CHECK_FPU_FEATURE(dc, VIS1); - cpu_src1 = gen_load_gpr(dc, rs1); - cpu_src2 = gen_load_gpr(dc, rs2); - gen_edge(dc, cpu_dst, cpu_src1, cpu_src2, 32, 1, 0); - gen_store_gpr(dc, rd, cpu_dst); - break; case 0x009: /* VIS II edge32n */ - CHECK_FPU_FEATURE(dc, VIS2); - cpu_src1 = gen_load_gpr(dc, rs1); - cpu_src2 = gen_load_gpr(dc, rs2); - gen_edge(dc, cpu_dst, cpu_src1, cpu_src2, 32, 0, 0); - gen_store_gpr(dc, rd, cpu_dst); - break; case 0x00a: /* VIS I edge32lcc */ - CHECK_FPU_FEATURE(dc, VIS1); - cpu_src1 = gen_load_gpr(dc, rs1); - cpu_src2 = gen_load_gpr(dc, rs2); - gen_edge(dc, cpu_dst, cpu_src1, cpu_src2, 32, 1, 1); - gen_store_gpr(dc, rd, cpu_dst); - break; case 0x00b: /* VIS II edge32ln */ - CHECK_FPU_FEATURE(dc, VIS2); - cpu_src1 = gen_load_gpr(dc, rs1); - cpu_src2 = gen_load_gpr(dc, rs2); - gen_edge(dc, cpu_dst, cpu_src1, cpu_src2, 32, 0, 1); - gen_store_gpr(dc, rd, cpu_dst); - break; + g_assert_not_reached(); /* in decodetree */ case 0x010: /* VIS I array8 */ CHECK_FPU_FEATURE(dc, VIS1); cpu_src1 = gen_load_gpr(dc, rs1);