From patchwork Thu Oct 10 21:50:11 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: =?utf-8?q?Philippe_Mathieu-Daud=C3=A9?= X-Patchwork-Id: 13831227 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 55133D24455 for ; Thu, 10 Oct 2024 21:52:32 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1sz14t-0002mj-H6; Thu, 10 Oct 2024 17:52:11 -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 1sz14r-0002gH-OG for qemu-devel@nongnu.org; Thu, 10 Oct 2024 17:52:09 -0400 Received: from mail-pf1-x42b.google.com ([2607:f8b0:4864:20::42b]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1sz14o-0004pi-IZ for qemu-devel@nongnu.org; Thu, 10 Oct 2024 17:52:08 -0400 Received: by mail-pf1-x42b.google.com with SMTP id d2e1a72fcca58-71de9e1f374so1088943b3a.1 for ; Thu, 10 Oct 2024 14:52:06 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; t=1728597124; x=1729201924; 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=XgHZ7eZfaFz/ArZ5SBx3nJXEi+AyeO465ylD8WzrsqQ=; b=Dsxh7iSkwEdOP6Rmh1JtXi2aDMxytHhk73je/6d/wxuBBR0TAeMh/jI2BclWq5xPA7 JO/aZ4vO/2uL7mEtFe+PlykLFbMijh8N9HErFVG9kQZ04pZtDwQ08BuU1AVncPVyTCwA krs9Dqfua8m9F77oXZiuZLCDLumy0yelYx9Cg8qugNoxswGYDJYmNXTqBdFhN+fbWPfO 6JXa5bEvpkUNalLzsQH1SKZM57cQQkcXVd1ea12WQaIvk4zFq+dyB8u0btV8SwwWJDAB O3nYhZGPi/VOKL9tCF/aDcLFqNh82+qo0iOSiXOYQurn/4dMr6P7+nznGiVFsdLB3vvn zpmg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1728597124; x=1729201924; 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=XgHZ7eZfaFz/ArZ5SBx3nJXEi+AyeO465ylD8WzrsqQ=; b=qcAo+VadpI2hcwIwQd0840SwhuEwP6VOhywfKKxniCIdDebut+wy/s0iITgvnmsaCg cWAkK9I4Mf+piMP/0CN/4aPpw5fAL7+5Y/B7nBH1ODUe4XHyHXRNvJK8uXQgFmugD5tb OLR/aDe91YwPmjDjDcKIv7IVWO2sBuzValMITwYqwHQw8Y85UaVrTPMwDv0JfS96HFAk bAa3aS9g2/F87Ue3EyDC7OIZfiefOm6intvNv4zAkEM7TaKzlVHllilRYWrXcEi6vdoH 73zHCr7l1wmyeKkzWccNdMkC8wzPnr+eyS+d322iVd1j5MQsrNoVKIRWFyMAwTNez6Aa GNyw== X-Gm-Message-State: AOJu0YxUp9SsEwW3PhzMSxSxNadUNd5LZFQK1qy7i9HMfMgqTa7xGKOB FDOOzpe7YXuCthnfJe7/uiUvu+O7UT/vjTzgdC+Sx7ngEryyKOkZq8CD2+5JXU2VufDWr1TxBbB V X-Google-Smtp-Source: AGHT+IFkwLKke12ZpcNJsGZyk+pYWNk6HORZHOD6IQ8j8pfvUcP7igR3O5DmUjW19hjYrU+5DWWBHA== X-Received: by 2002:a05:6a00:1310:b0:71e:cc7:c511 with SMTP id d2e1a72fcca58-71e3807443cmr793114b3a.23.1728597124319; Thu, 10 Oct 2024 14:52:04 -0700 (PDT) Received: from localhost.localdomain ([45.176.89.169]) by smtp.gmail.com with ESMTPSA id d2e1a72fcca58-71e2ab0b7a1sm1491126b3a.197.2024.10.10.14.52.01 (version=TLS1_3 cipher=TLS_CHACHA20_POLY1305_SHA256 bits=256/256); Thu, 10 Oct 2024 14:52:03 -0700 (PDT) From: =?utf-8?q?Philippe_Mathieu-Daud=C3=A9?= To: qemu-devel@nongnu.org Cc: Huacai Chen , =?utf-8?q?Philippe_Mathieu-Daud?= =?utf-8?q?=C3=A9?= , Jiaxun Yang , =?utf-8?q?Herv=C3=A9_Poussineau?= , Aleksandar Rikalo , Aurelien Jarno , Richard Henderson Subject: [PATCH v2 13/16] target/mips: Use tcg_constant_tl() instead of tcg_gen_movi_tl() Date: Thu, 10 Oct 2024 18:50:11 -0300 Message-ID: <20241010215015.44326-14-philmd@linaro.org> X-Mailer: git-send-email 2.45.2 In-Reply-To: <20241010215015.44326-1-philmd@linaro.org> References: <20241010215015.44326-1-philmd@linaro.org> MIME-Version: 1.0 Received-SPF: pass client-ip=2607:f8b0:4864:20::42b; envelope-from=philmd@linaro.org; helo=mail-pf1-x42b.google.com X-Spam_score_int: 12 X-Spam_score: 1.2 X-Spam_bar: + X-Spam_report: (1.2 / 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, RCVD_IN_SBL_CSS=3.335, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=no 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 Directly use tcg_constant_tl() for constant integer, this save a call to tcg_gen_movi_tl(), often saving a temp register. Most of the places found using the following Coccinelle spatch script: @@ identifier tmp; constant val; @@ * TCGv tmp = tcg_temp_new(); ... * tcg_gen_movi_tl(tmp, val); @@ identifier tmp; int val; @@ * TCGv tmp = tcg_temp_new(); ... * tcg_gen_movi_i64(tmp, val); Suggested-by: Richard Henderson Signed-off-by: Philippe Mathieu-Daudé Reviewed-by: Richard Henderson Message-Id: <20241004202621.4321-2-philmd@linaro.org> --- target/mips/tcg/translate.c | 49 ++++------ target/mips/tcg/nanomips_translate.c.inc | 110 +++++++++-------------- 2 files changed, 58 insertions(+), 101 deletions(-) diff --git a/target/mips/tcg/translate.c b/target/mips/tcg/translate.c index 8ff1d59eb64..254948583c8 100644 --- a/target/mips/tcg/translate.c +++ b/target/mips/tcg/translate.c @@ -2252,8 +2252,7 @@ static void gen_st_cond(DisasContext *ctx, int rt, int base, int offset, /* compare the address against that of the preceding LL */ gen_base_offset_addr(ctx, addr, base, offset); tcg_gen_brcond_tl(TCG_COND_EQ, addr, cpu_lladdr, l1); - tcg_gen_movi_tl(t0, 0); - gen_store_gpr(t0, rt); + gen_store_gpr(tcg_constant_tl(0), rt); tcg_gen_br(done); gen_set_label(l1); @@ -3059,8 +3058,7 @@ static void gen_r6_muldiv(DisasContext *ctx, int opc, int rd, int rs, int rt) tcg_gen_and_tl(t2, t2, t3); tcg_gen_setcondi_tl(TCG_COND_EQ, t3, t1, 0); tcg_gen_or_tl(t2, t2, t3); - tcg_gen_movi_tl(t3, 0); - tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, t3, t2, t1); + tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, tcg_constant_tl(0), t2, t1); tcg_gen_div_tl(cpu_gpr[rd], t0, t1); tcg_gen_ext32s_tl(cpu_gpr[rd], cpu_gpr[rd]); } @@ -3076,30 +3074,27 @@ static void gen_r6_muldiv(DisasContext *ctx, int opc, int rd, int rs, int rt) tcg_gen_and_tl(t2, t2, t3); tcg_gen_setcondi_tl(TCG_COND_EQ, t3, t1, 0); tcg_gen_or_tl(t2, t2, t3); - tcg_gen_movi_tl(t3, 0); - tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, t3, t2, t1); + tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, tcg_constant_tl(0), t2, t1); tcg_gen_rem_tl(cpu_gpr[rd], t0, t1); tcg_gen_ext32s_tl(cpu_gpr[rd], cpu_gpr[rd]); } break; case R6_OPC_DIVU: { - TCGv t2 = tcg_constant_tl(0); - TCGv t3 = tcg_constant_tl(1); tcg_gen_ext32u_tl(t0, t0); tcg_gen_ext32u_tl(t1, t1); - tcg_gen_movcond_tl(TCG_COND_EQ, t1, t1, t2, t3, t1); + tcg_gen_movcond_tl(TCG_COND_EQ, t1, t1, + tcg_constant_tl(0), tcg_constant_tl(1), t1); tcg_gen_divu_tl(cpu_gpr[rd], t0, t1); tcg_gen_ext32s_tl(cpu_gpr[rd], cpu_gpr[rd]); } break; case R6_OPC_MODU: { - TCGv t2 = tcg_constant_tl(0); - TCGv t3 = tcg_constant_tl(1); tcg_gen_ext32u_tl(t0, t0); tcg_gen_ext32u_tl(t1, t1); - tcg_gen_movcond_tl(TCG_COND_EQ, t1, t1, t2, t3, t1); + tcg_gen_movcond_tl(TCG_COND_EQ, t1, t1, + tcg_constant_tl(0), tcg_constant_tl(1), t1); tcg_gen_remu_tl(cpu_gpr[rd], t0, t1); tcg_gen_ext32s_tl(cpu_gpr[rd], cpu_gpr[rd]); } @@ -3154,8 +3149,7 @@ static void gen_r6_muldiv(DisasContext *ctx, int opc, int rd, int rs, int rt) tcg_gen_and_tl(t2, t2, t3); tcg_gen_setcondi_tl(TCG_COND_EQ, t3, t1, 0); tcg_gen_or_tl(t2, t2, t3); - tcg_gen_movi_tl(t3, 0); - tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, t3, t2, t1); + tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, tcg_constant_tl(0), t2, t1); tcg_gen_div_tl(cpu_gpr[rd], t0, t1); } break; @@ -3168,24 +3162,21 @@ static void gen_r6_muldiv(DisasContext *ctx, int opc, int rd, int rs, int rt) tcg_gen_and_tl(t2, t2, t3); tcg_gen_setcondi_tl(TCG_COND_EQ, t3, t1, 0); tcg_gen_or_tl(t2, t2, t3); - tcg_gen_movi_tl(t3, 0); - tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, t3, t2, t1); + tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, tcg_constant_tl(0), t2, t1); tcg_gen_rem_tl(cpu_gpr[rd], t0, t1); } break; case R6_OPC_DDIVU: { - TCGv t2 = tcg_constant_tl(0); - TCGv t3 = tcg_constant_tl(1); - tcg_gen_movcond_tl(TCG_COND_EQ, t1, t1, t2, t3, t1); + tcg_gen_movcond_tl(TCG_COND_EQ, t1, t1, + tcg_constant_tl(0), tcg_constant_tl(1), t1); tcg_gen_divu_i64(cpu_gpr[rd], t0, t1); } break; case R6_OPC_DMODU: { - TCGv t2 = tcg_constant_tl(0); - TCGv t3 = tcg_constant_tl(1); - tcg_gen_movcond_tl(TCG_COND_EQ, t1, t1, t2, t3, t1); + tcg_gen_movcond_tl(TCG_COND_EQ, t1, t1, + tcg_constant_tl(0), tcg_constant_tl(1), t1); tcg_gen_remu_i64(cpu_gpr[rd], t0, t1); } break; @@ -3238,8 +3229,7 @@ static void gen_div1_tx79(DisasContext *ctx, uint32_t opc, int rs, int rt) tcg_gen_and_tl(t2, t2, t3); tcg_gen_setcondi_tl(TCG_COND_EQ, t3, t1, 0); tcg_gen_or_tl(t2, t2, t3); - tcg_gen_movi_tl(t3, 0); - tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, t3, t2, t1); + tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, tcg_constant_tl(0), t2, t1); tcg_gen_div_tl(cpu_LO[1], t0, t1); tcg_gen_rem_tl(cpu_HI[1], t0, t1); tcg_gen_ext32s_tl(cpu_LO[1], cpu_LO[1]); @@ -3294,8 +3284,7 @@ static void gen_muldiv(DisasContext *ctx, uint32_t opc, tcg_gen_and_tl(t2, t2, t3); tcg_gen_setcondi_tl(TCG_COND_EQ, t3, t1, 0); tcg_gen_or_tl(t2, t2, t3); - tcg_gen_movi_tl(t3, 0); - tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, t3, t2, t1); + tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, tcg_constant_tl(0), t2, t1); tcg_gen_div_tl(cpu_LO[acc], t0, t1); tcg_gen_rem_tl(cpu_HI[acc], t0, t1); tcg_gen_ext32s_tl(cpu_LO[acc], cpu_LO[acc]); @@ -3347,17 +3336,15 @@ static void gen_muldiv(DisasContext *ctx, uint32_t opc, tcg_gen_and_tl(t2, t2, t3); tcg_gen_setcondi_tl(TCG_COND_EQ, t3, t1, 0); tcg_gen_or_tl(t2, t2, t3); - tcg_gen_movi_tl(t3, 0); - tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, t3, t2, t1); + tcg_gen_movcond_tl(TCG_COND_NE, t1, t2, tcg_constant_tl(0), t2, t1); tcg_gen_div_tl(cpu_LO[acc], t0, t1); tcg_gen_rem_tl(cpu_HI[acc], t0, t1); } break; case OPC_DDIVU: { - TCGv t2 = tcg_constant_tl(0); - TCGv t3 = tcg_constant_tl(1); - tcg_gen_movcond_tl(TCG_COND_EQ, t1, t1, t2, t3, t1); + tcg_gen_movcond_tl(TCG_COND_EQ, t1, t1, + tcg_constant_tl(0), tcg_constant_tl(1), t1); tcg_gen_divu_i64(cpu_LO[acc], t0, t1); tcg_gen_remu_i64(cpu_HI[acc], t0, t1); } diff --git a/target/mips/tcg/nanomips_translate.c.inc b/target/mips/tcg/nanomips_translate.c.inc index d6149ea26e4..6b7abf1f8f8 100644 --- a/target/mips/tcg/nanomips_translate.c.inc +++ b/target/mips/tcg/nanomips_translate.c.inc @@ -1052,8 +1052,7 @@ static void gen_scwp(DisasContext *ctx, uint32_t base, int16_t offset, tcg_gen_movi_tl(cpu_gpr[reg1], 0); } gen_set_label(lab_done); - tcg_gen_movi_tl(lladdr, -1); - tcg_gen_st_tl(lladdr, tcg_env, offsetof(CPUMIPSState, lladdr)); + tcg_gen_st_tl(tcg_constant_tl(-1), tcg_env, offsetof(CPUMIPSState, lladdr)); } static void gen_adjust_sp(DisasContext *ctx, int u) @@ -1543,7 +1542,6 @@ static void gen_pool32axf_1_nanomips_insn(DisasContext *ctx, uint32_t opc, { int16_t imm; TCGv t0 = tcg_temp_new(); - TCGv t1 = tcg_temp_new(); TCGv v0_t = tcg_temp_new(); gen_load_gpr(v0_t, v1); @@ -1570,12 +1568,10 @@ static void gen_pool32axf_1_nanomips_insn(DisasContext *ctx, uint32_t opc, check_dsp(ctx); switch (extract32(ctx->opcode, 12, 2)) { case NM_MTHLIP: - tcg_gen_movi_tl(t0, v2 >> 3); - gen_helper_mthlip(t0, v0_t, tcg_env); + gen_helper_mthlip(tcg_constant_tl(v2 >> 3), v0_t, tcg_env); break; case NM_SHILOV: - tcg_gen_movi_tl(t0, v2 >> 3); - gen_helper_shilo(t0, v0_t, tcg_env); + gen_helper_shilo(tcg_constant_tl(v2 >> 3), v0_t, tcg_env); break; default: gen_reserved_instruction(ctx); @@ -1587,39 +1583,34 @@ static void gen_pool32axf_1_nanomips_insn(DisasContext *ctx, uint32_t opc, imm = extract32(ctx->opcode, 14, 7); switch (extract32(ctx->opcode, 12, 2)) { case NM_RDDSP: - tcg_gen_movi_tl(t0, imm); - gen_helper_rddsp(t0, t0, tcg_env); + gen_helper_rddsp(t0, tcg_constant_tl(imm), tcg_env); gen_store_gpr(t0, ret); break; case NM_WRDSP: gen_load_gpr(t0, ret); - tcg_gen_movi_tl(t1, imm); - gen_helper_wrdsp(t0, t1, tcg_env); + gen_helper_wrdsp(t0, tcg_constant_tl(imm), tcg_env); break; case NM_EXTP: - tcg_gen_movi_tl(t0, v2 >> 3); - tcg_gen_movi_tl(t1, v1); - gen_helper_extp(t0, t0, t1, tcg_env); + gen_helper_extp(t0, tcg_constant_tl(v2 >> 3), + tcg_constant_tl(v1), tcg_env); gen_store_gpr(t0, ret); break; case NM_EXTPDP: - tcg_gen_movi_tl(t0, v2 >> 3); - tcg_gen_movi_tl(t1, v1); - gen_helper_extpdp(t0, t0, t1, tcg_env); + gen_helper_extpdp(t0, tcg_constant_tl(v2 >> 3), + tcg_constant_tl(v1), tcg_env); gen_store_gpr(t0, ret); break; } break; case NM_POOL32AXF_1_4: check_dsp(ctx); - tcg_gen_movi_tl(t0, v2 >> 2); switch (extract32(ctx->opcode, 12, 1)) { case NM_SHLL_QB: - gen_helper_shll_qb(t0, t0, v0_t, tcg_env); + gen_helper_shll_qb(t0, tcg_constant_tl(v2 >> 2), v0_t, tcg_env); gen_store_gpr(t0, ret); break; case NM_SHRL_QB: - gen_helper_shrl_qb(t0, t0, v0_t); + gen_helper_shrl_qb(t0, tcg_constant_tl(v2 >> 2), v0_t); gen_store_gpr(t0, ret); break; } @@ -1630,23 +1621,25 @@ static void gen_pool32axf_1_nanomips_insn(DisasContext *ctx, uint32_t opc, break; case NM_POOL32AXF_1_7: check_dsp(ctx); - tcg_gen_movi_tl(t0, v2 >> 3); - tcg_gen_movi_tl(t1, v1); switch (extract32(ctx->opcode, 12, 2)) { case NM_EXTR_W: - gen_helper_extr_w(t0, t0, t1, tcg_env); + gen_helper_extr_w(t0, tcg_constant_tl(v2 >> 3), + tcg_constant_tl(v1), tcg_env); gen_store_gpr(t0, ret); break; case NM_EXTR_R_W: - gen_helper_extr_r_w(t0, t0, t1, tcg_env); + gen_helper_extr_r_w(t0, tcg_constant_tl(v2 >> 3), + tcg_constant_tl(v1), tcg_env); gen_store_gpr(t0, ret); break; case NM_EXTR_RS_W: - gen_helper_extr_rs_w(t0, t0, t1, tcg_env); + gen_helper_extr_rs_w(t0, tcg_constant_tl(v2 >> 3), + tcg_constant_tl(v1), tcg_env); gen_store_gpr(t0, ret); break; case NM_EXTR_S_H: - gen_helper_extr_s_h(t0, t0, t1, tcg_env); + gen_helper_extr_s_h(t0, tcg_constant_tl(v2 >> 3), + tcg_constant_tl(v1), tcg_env); gen_store_gpr(t0, ret); break; } @@ -1848,8 +1841,7 @@ static void gen_pool32axf_2_nanomips_insn(DisasContext *ctx, uint32_t opc, case NM_EXTRV_W: check_dsp(ctx); gen_load_gpr(v1_t, rs); - tcg_gen_movi_tl(t0, rd >> 3); - gen_helper_extr_w(t0, t0, v1_t, tcg_env); + gen_helper_extr_w(t0, tcg_constant_tl(rd >> 3), v1_t, tcg_env); gen_store_gpr(t0, ret); break; } @@ -1903,8 +1895,7 @@ static void gen_pool32axf_2_nanomips_insn(DisasContext *ctx, uint32_t opc, break; case NM_EXTRV_R_W: check_dsp(ctx); - tcg_gen_movi_tl(t0, rd >> 3); - gen_helper_extr_r_w(t0, t0, v1_t, tcg_env); + gen_helper_extr_r_w(t0, tcg_constant_tl(rd >> 3), v1_t, tcg_env); gen_store_gpr(t0, ret); break; default: @@ -1923,8 +1914,7 @@ static void gen_pool32axf_2_nanomips_insn(DisasContext *ctx, uint32_t opc, break; case NM_EXTPV: check_dsp(ctx); - tcg_gen_movi_tl(t0, rd >> 3); - gen_helper_extp(t0, t0, v1_t, tcg_env); + gen_helper_extp(t0, tcg_constant_tl(rd >> 3), v1_t, tcg_env); gen_store_gpr(t0, ret); break; case NM_MSUB: @@ -1947,8 +1937,7 @@ static void gen_pool32axf_2_nanomips_insn(DisasContext *ctx, uint32_t opc, break; case NM_EXTRV_RS_W: check_dsp(ctx); - tcg_gen_movi_tl(t0, rd >> 3); - gen_helper_extr_rs_w(t0, t0, v1_t, tcg_env); + gen_helper_extr_rs_w(t0, tcg_constant_tl(rd >> 3), v1_t, tcg_env); gen_store_gpr(t0, ret); break; } @@ -1964,8 +1953,7 @@ static void gen_pool32axf_2_nanomips_insn(DisasContext *ctx, uint32_t opc, break; case NM_EXTPDPV: check_dsp(ctx); - tcg_gen_movi_tl(t0, rd >> 3); - gen_helper_extpdp(t0, t0, v1_t, tcg_env); + gen_helper_extpdp(t0, tcg_constant_tl(rd >> 3), v1_t, tcg_env); gen_store_gpr(t0, ret); break; case NM_MSUBU: @@ -1990,8 +1978,7 @@ static void gen_pool32axf_2_nanomips_insn(DisasContext *ctx, uint32_t opc, break; case NM_EXTRV_S_H: check_dsp(ctx); - tcg_gen_movi_tl(t0, rd >> 3); - gen_helper_extr_s_h(t0, t0, v1_t, tcg_env); + gen_helper_extr_s_h(t0, tcg_constant_tl(rd >> 3), v1_t, tcg_env); gen_store_gpr(t0, ret); break; } @@ -2149,24 +2136,22 @@ static void gen_pool32axf_7_nanomips_insn(DisasContext *ctx, uint32_t opc, switch (opc) { case NM_SHRA_R_QB: check_dsp_r2(ctx); - tcg_gen_movi_tl(t0, rd >> 2); switch (extract32(ctx->opcode, 12, 1)) { case 0: /* NM_SHRA_QB */ - gen_helper_shra_qb(t0, t0, rs_t); + gen_helper_shra_qb(t0, tcg_constant_tl(rd >> 2), rs_t); gen_store_gpr(t0, rt); break; case 1: /* NM_SHRA_R_QB */ - gen_helper_shra_r_qb(t0, t0, rs_t); + gen_helper_shra_r_qb(t0, tcg_constant_tl(rd >> 2), rs_t); gen_store_gpr(t0, rt); break; } break; case NM_SHRL_PH: check_dsp_r2(ctx); - tcg_gen_movi_tl(t0, rd >> 1); - gen_helper_shrl_ph(t0, t0, rs_t); + gen_helper_shrl_ph(t0, tcg_constant_tl(rd >> 1), rs_t); gen_store_gpr(t0, rt); break; case NM_REPL_QB: @@ -2180,8 +2165,7 @@ static void gen_pool32axf_7_nanomips_insn(DisasContext *ctx, uint32_t opc, (uint32_t)imm << 8 | (uint32_t)imm; result = (int32_t)result; - tcg_gen_movi_tl(t0, result); - gen_store_gpr(t0, rt); + gen_store_gpr(tcg_constant_tl(result), rt); } break; default: @@ -2302,10 +2286,9 @@ static void gen_compute_imm_branch(DisasContext *ctx, uint32_t opc, { TCGCond cond = TCG_COND_ALWAYS; TCGv t0 = tcg_temp_new(); - TCGv t1 = tcg_temp_new(); + TCGv timm = tcg_constant_tl(imm); gen_load_gpr(t0, rt); - tcg_gen_movi_tl(t1, imm); ctx->btarget = addr_add(ctx, ctx->base.pc_next + 4, offset); /* Load needed operands and calculate btarget */ @@ -2334,7 +2317,7 @@ static void gen_compute_imm_branch(DisasContext *ctx, uint32_t opc, } else { tcg_gen_shri_tl(t0, t0, imm); tcg_gen_andi_tl(t0, t0, 1); - tcg_gen_movi_tl(t1, 0); + timm = tcg_constant_tl(0); if (opc == NM_BBEQZC) { cond = TCG_COND_EQ; } else { @@ -2389,7 +2372,7 @@ static void gen_compute_imm_branch(DisasContext *ctx, uint32_t opc, /* Conditional compact branch */ TCGLabel *fs = gen_new_label(); - tcg_gen_brcond_tl(tcg_invert_cond(cond), t0, t1, fs); + tcg_gen_brcond_tl(tcg_invert_cond(cond), t0, timm, fs); gen_goto_tb(ctx, 1, ctx->btarget); gen_set_label(fs); @@ -2403,7 +2386,6 @@ static void gen_compute_nanomips_pbalrsc_branch(DisasContext *ctx, int rs, int rt) { TCGv t0 = tcg_temp_new(); - TCGv t1 = tcg_temp_new(); /* load rs */ gen_load_gpr(t0, rs); @@ -2415,8 +2397,7 @@ static void gen_compute_nanomips_pbalrsc_branch(DisasContext *ctx, int rs, /* calculate btarget */ tcg_gen_shli_tl(t0, t0, 1); - tcg_gen_movi_tl(t1, ctx->base.pc_next + 4); - gen_op_addr_add(ctx, btarget, t1, t0); + gen_op_addr_add(ctx, btarget, tcg_constant_tl(ctx->base.pc_next + 4), t0); /* branch completion */ clear_branch_hflags(ctx); @@ -3443,13 +3424,10 @@ static void gen_pool32a5_nanomips_insn(DisasContext *ctx, int opc, case NM_SHILO: check_dsp(ctx); { - TCGv tv0 = tcg_temp_new(); - TCGv tv1 = tcg_temp_new(); int16_t imm = extract32(ctx->opcode, 16, 7); - tcg_gen_movi_tl(tv0, rd >> 3); - tcg_gen_movi_tl(tv1, imm); - gen_helper_shilo(tv0, tv1, tcg_env); + gen_helper_shilo(tcg_constant_tl(rd >> 3), + tcg_constant_tl(imm), tcg_env); } break; case NM_MULEQ_S_W_PHL: @@ -3504,8 +3482,7 @@ static void gen_pool32a5_nanomips_insn(DisasContext *ctx, int opc, break; case NM_SHRA_R_W: check_dsp(ctx); - tcg_gen_movi_tl(t0, rd); - gen_helper_shra_r_w(v1_t, t0, v1_t); + gen_helper_shra_r_w(v1_t, tcg_constant_tl(rd), v1_t); gen_store_gpr(v1_t, rt); break; case NM_SHRA_R_PH: @@ -3545,8 +3522,7 @@ static void gen_pool32a5_nanomips_insn(DisasContext *ctx, int opc, break; case NM_SHLL_S_W: check_dsp(ctx); - tcg_gen_movi_tl(t0, rd); - gen_helper_shll_s_w(v1_t, t0, v1_t, tcg_env); + gen_helper_shll_s_w(v1_t, tcg_constant_tl(rd), v1_t, tcg_env); gen_store_gpr(v1_t, rt); break; case NM_REPL_PH: @@ -3727,31 +3703,25 @@ static int decode_nanomips_32_48_opc(CPUMIPSState *env, DisasContext *ctx) case NM_LWPC48: check_nms(ctx); if (rt != 0) { - TCGv t0; - t0 = tcg_temp_new(); - target_long addr = addr_add(ctx, ctx->base.pc_next + 6, addr_off); - tcg_gen_movi_tl(t0, addr); - tcg_gen_qemu_ld_tl(cpu_gpr[rt], t0, ctx->mem_idx, + tcg_gen_qemu_ld_tl(cpu_gpr[rt], tcg_constant_tl(addr), ctx->mem_idx, mo_endian(ctx) | MO_SL | ctx->default_tcg_memop_mask); } break; case NM_SWPC48: check_nms(ctx); { - TCGv t0, t1; - t0 = tcg_temp_new(); + TCGv t1; t1 = tcg_temp_new(); target_long addr = addr_add(ctx, ctx->base.pc_next + 6, addr_off); - tcg_gen_movi_tl(t0, addr); gen_load_gpr(t1, rt); - tcg_gen_qemu_st_tl(t1, t0, ctx->mem_idx, + tcg_gen_qemu_st_tl(t1, tcg_constant_tl(addr), ctx->mem_idx, mo_endian(ctx) | MO_UL | ctx->default_tcg_memop_mask); } break;