From mboxrd@z Thu Jan 1 00:00:00 1970 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 8619AC6FA99 for ; Tue, 7 Mar 2023 21:45:21 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1pZf6x-0001UP-KC; Tue, 07 Mar 2023 16:44:43 -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 1pZf6u-0001Tr-Tg; Tue, 07 Mar 2023 16:44:41 -0500 Received: from mail-oa1-x2b.google.com ([2001:4860:4864:20::2b]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1pZf6s-0003wp-1F; Tue, 07 Mar 2023 16:44:40 -0500 Received: by mail-oa1-x2b.google.com with SMTP id 586e51a60fabf-17711f56136so1757779fac.12; Tue, 07 Mar 2023 13:44:37 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; t=1678225476; h=content-transfer-encoding:in-reply-to:from:references:cc:to :content-language:subject:user-agent:mime-version:date:message-id :from:to:cc:subject:date:message-id:reply-to; bh=6toeQQBQzW5G9zhhPcV/SA/n28d8pj5DO/ZttzbJMh0=; b=nc+g9Wt78urlI5IgVDvMHiPJ3xWjRPsxVbXoupBqHEz1qh47tMpX0SSjAMADnQdcqR xkzG9yrNe+TdPUB+pqxX8f7Wx61lQHd9we1bYBDcp0pwzrsehP9xQoXXS48IYiVkFq/w u6Na0+c4uVa/KgInMz8ND92Se7D6v8MiLo6MjJidBvZOJS8XrKvuQD6dHcZqFS9V1ZsI Z/qNDb4zC9ud/IrIdzAV7WP3iaDP8xiI3PjrzRmeikTPYApcdx9JkXPMt1TZiAXHBAnC HSal/eMNz4TUJvtXUBOuZfRh9VYgLyVKgfDL1TVCzn/teQW0cn0IbHfgctYlKaUttfGp vSzg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; t=1678225476; h=content-transfer-encoding:in-reply-to:from:references:cc:to :content-language:subject:user-agent:mime-version:date:message-id :x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=6toeQQBQzW5G9zhhPcV/SA/n28d8pj5DO/ZttzbJMh0=; b=nfUmB3P3Poqv9VXuvBCiHQlTp+wduubk9lBQ3FYOCWx9qBhRTQvMRRIyhq70rS2nsZ TL6vobNnahwGtl07jbFIIZBSJX9hY2VfBtEK6eNTt2EMuxRkOL1+6k+l0LV2ReiKhl0W y7hVzpcedyzQA5OUAsXRYJ6NEbbb72qLNF2MBuxgNGxhZMlABOwWDtfKsz1VuNno9Xx6 PZlpN6WkH3ijOX7EU5dXqaIa925lZjBgMmgQrR+1Ux3e5KxOURZ6/iVhdoGtQxvUbztU 6c+fJX8wv28N0EUIskwc7+n1SqOWGBUC//A5COULb+QU9A84Iv06J0G8Qv1Oxt11Zlod JjDA== X-Gm-Message-State: AO0yUKVTuwYNcynJkClv+gUrpiE+kr3Y5cEKwnJoesmUduuOV1iChTh8 8/1pVOjOA5IgY0ve2GAuoChyPeRs/bo= X-Google-Smtp-Source: AK7set+szvTsTDzskpzfN0tstsZLj8xpSixQ51fr2M62a1duSKz3oohtSEqpRdTtY7FLHVrQOxFaTQ== X-Received: by 2002:a05:6870:a546:b0:172:8ae4:af53 with SMTP id p6-20020a056870a54600b001728ae4af53mr10903574oal.21.1678225476262; Tue, 07 Mar 2023 13:44:36 -0800 (PST) Received: from [192.168.68.107] ([177.189.53.31]) by smtp.gmail.com with ESMTPSA id v3-20020a4aad83000000b00524faf3d2d7sm5367410oom.41.2023.03.07.13.44.33 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Tue, 07 Mar 2023 13:44:35 -0800 (PST) Message-ID: Date: Tue, 7 Mar 2023 18:44:31 -0300 MIME-Version: 1.0 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101 Thunderbird/102.8.0 Subject: Re: [PATCH v2 22/25] target/ppc: Avoid tcg_const_* in translate.c Content-Language: en-US To: Richard Henderson , qemu-devel@nongnu.org Cc: =?UTF-8?Q?C=c3=a9dric_Le_Goater?= , David Gibson , Greg Kurz , qemu-ppc@nongnu.org References: <20230307183503.2512684-1-richard.henderson@linaro.org> <20230307183503.2512684-23-richard.henderson@linaro.org> From: Daniel Henrique Barboza In-Reply-To: <20230307183503.2512684-23-richard.henderson@linaro.org> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit Received-SPF: pass client-ip=2001:4860:4864:20::2b; envelope-from=danielhb413@gmail.com; helo=mail-oa1-x2b.google.com X-Spam_score_int: -18 X-Spam_score: -1.9 X-Spam_bar: - X-Spam_report: (-1.9 / 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, FREEMAIL_ENVFROM_END_DIGIT=0.25, FREEMAIL_FROM=0.001, NICE_REPLY_A=-0.001, 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 On 3/7/23 15:35, Richard Henderson wrote: > All remaining uses are strictly read-only. > > Signed-off-by: Richard Henderson > --- Reviewed-by: Daniel Henrique Barboza > Cc: Daniel Henrique Barboza > Cc: Cédric Le Goater > Cc: David Gibson > Cc: Greg Kurz > Cc: qemu-ppc@nongnu.org > --- > target/ppc/translate.c | 142 +++++++++++++++++++++-------------------- > 1 file changed, 72 insertions(+), 70 deletions(-) > > diff --git a/target/ppc/translate.c b/target/ppc/translate.c > index 7ec940b7d8..9d05357d03 100644 > --- a/target/ppc/translate.c > +++ b/target/ppc/translate.c > @@ -263,8 +263,8 @@ static void gen_exception_err(DisasContext *ctx, uint32_t excp, uint32_t error) > * faulting instruction > */ > gen_update_nip(ctx, ctx->cia); > - t0 = tcg_const_i32(excp); > - t1 = tcg_const_i32(error); > + t0 = tcg_constant_i32(excp); > + t1 = tcg_constant_i32(error); > gen_helper_raise_exception_err(cpu_env, t0, t1); > ctx->base.is_jmp = DISAS_NORETURN; > } > @@ -278,7 +278,7 @@ static void gen_exception(DisasContext *ctx, uint32_t excp) > * faulting instruction > */ > gen_update_nip(ctx, ctx->cia); > - t0 = tcg_const_i32(excp); > + t0 = tcg_constant_i32(excp); > gen_helper_raise_exception(cpu_env, t0); > ctx->base.is_jmp = DISAS_NORETURN; > } > @@ -289,7 +289,7 @@ static void gen_exception_nip(DisasContext *ctx, uint32_t excp, > TCGv_i32 t0; > > gen_update_nip(ctx, nip); > - t0 = tcg_const_i32(excp); > + t0 = tcg_constant_i32(excp); > gen_helper_raise_exception(cpu_env, t0); > ctx->base.is_jmp = DISAS_NORETURN; > } > @@ -386,7 +386,7 @@ void spr_noaccess(DisasContext *ctx, int gprn, int sprn) > static void spr_load_dump_spr(int sprn) > { > #ifdef PPC_DUMP_SPR_ACCESSES > - TCGv_i32 t0 = tcg_const_i32(sprn); > + TCGv_i32 t0 = tcg_constant_i32(sprn); > gen_helper_load_dump_spr(cpu_env, t0); > #endif > } > @@ -400,7 +400,7 @@ void spr_read_generic(DisasContext *ctx, int gprn, int sprn) > static void spr_store_dump_spr(int sprn) > { > #ifdef PPC_DUMP_SPR_ACCESSES > - TCGv_i32 t0 = tcg_const_i32(sprn); > + TCGv_i32 t0 = tcg_constant_i32(sprn); > gen_helper_store_dump_spr(cpu_env, t0); > #endif > } > @@ -672,25 +672,25 @@ void spr_read_ibat_h(DisasContext *ctx, int gprn, int sprn) > > void spr_write_ibatu(DisasContext *ctx, int sprn, int gprn) > { > - TCGv_i32 t0 = tcg_const_i32((sprn - SPR_IBAT0U) / 2); > + TCGv_i32 t0 = tcg_constant_i32((sprn - SPR_IBAT0U) / 2); > gen_helper_store_ibatu(cpu_env, t0, cpu_gpr[gprn]); > } > > void spr_write_ibatu_h(DisasContext *ctx, int sprn, int gprn) > { > - TCGv_i32 t0 = tcg_const_i32(((sprn - SPR_IBAT4U) / 2) + 4); > + TCGv_i32 t0 = tcg_constant_i32(((sprn - SPR_IBAT4U) / 2) + 4); > gen_helper_store_ibatu(cpu_env, t0, cpu_gpr[gprn]); > } > > void spr_write_ibatl(DisasContext *ctx, int sprn, int gprn) > { > - TCGv_i32 t0 = tcg_const_i32((sprn - SPR_IBAT0L) / 2); > + TCGv_i32 t0 = tcg_constant_i32((sprn - SPR_IBAT0L) / 2); > gen_helper_store_ibatl(cpu_env, t0, cpu_gpr[gprn]); > } > > void spr_write_ibatl_h(DisasContext *ctx, int sprn, int gprn) > { > - TCGv_i32 t0 = tcg_const_i32(((sprn - SPR_IBAT4L) / 2) + 4); > + TCGv_i32 t0 = tcg_constant_i32(((sprn - SPR_IBAT4L) / 2) + 4); > gen_helper_store_ibatl(cpu_env, t0, cpu_gpr[gprn]); > } > > @@ -712,25 +712,25 @@ void spr_read_dbat_h(DisasContext *ctx, int gprn, int sprn) > > void spr_write_dbatu(DisasContext *ctx, int sprn, int gprn) > { > - TCGv_i32 t0 = tcg_const_i32((sprn - SPR_DBAT0U) / 2); > + TCGv_i32 t0 = tcg_constant_i32((sprn - SPR_DBAT0U) / 2); > gen_helper_store_dbatu(cpu_env, t0, cpu_gpr[gprn]); > } > > void spr_write_dbatu_h(DisasContext *ctx, int sprn, int gprn) > { > - TCGv_i32 t0 = tcg_const_i32(((sprn - SPR_DBAT4U) / 2) + 4); > + TCGv_i32 t0 = tcg_constant_i32(((sprn - SPR_DBAT4U) / 2) + 4); > gen_helper_store_dbatu(cpu_env, t0, cpu_gpr[gprn]); > } > > void spr_write_dbatl(DisasContext *ctx, int sprn, int gprn) > { > - TCGv_i32 t0 = tcg_const_i32((sprn - SPR_DBAT0L) / 2); > + TCGv_i32 t0 = tcg_constant_i32((sprn - SPR_DBAT0L) / 2); > gen_helper_store_dbatl(cpu_env, t0, cpu_gpr[gprn]); > } > > void spr_write_dbatl_h(DisasContext *ctx, int sprn, int gprn) > { > - TCGv_i32 t0 = tcg_const_i32(((sprn - SPR_DBAT4L) / 2) + 4); > + TCGv_i32 t0 = tcg_constant_i32(((sprn - SPR_DBAT4L) / 2) + 4); > gen_helper_store_dbatl(cpu_env, t0, cpu_gpr[gprn]); > } > > @@ -1040,13 +1040,15 @@ void spr_write_booke206_mmucsr0(DisasContext *ctx, int sprn, int gprn) > > void spr_write_booke_pid(DisasContext *ctx, int sprn, int gprn) > { > - TCGv_i32 t0 = tcg_const_i32(sprn); > + TCGv_i32 t0 = tcg_constant_i32(sprn); > gen_helper_booke_setpid(cpu_env, t0, cpu_gpr[gprn]); > } > + > void spr_write_eplc(DisasContext *ctx, int sprn, int gprn) > { > gen_helper_booke_set_eplc(cpu_env, cpu_gpr[gprn]); > } > + > void spr_write_epsc(DisasContext *ctx, int sprn, int gprn) > { > gen_helper_booke_set_epsc(cpu_env, cpu_gpr[gprn]); > @@ -1080,9 +1082,9 @@ void spr_read_mas73(DisasContext *ctx, int gprn, int sprn) > static void gen_fscr_facility_check(DisasContext *ctx, int facility_sprn, > int bit, int sprn, int cause) > { > - TCGv_i32 t1 = tcg_const_i32(bit); > - TCGv_i32 t2 = tcg_const_i32(sprn); > - TCGv_i32 t3 = tcg_const_i32(cause); > + TCGv_i32 t1 = tcg_constant_i32(bit); > + TCGv_i32 t2 = tcg_constant_i32(sprn); > + TCGv_i32 t3 = tcg_constant_i32(cause); > > gen_helper_fscr_facility_check(cpu_env, t1, t2, t3); > } > @@ -1090,9 +1092,9 @@ static void gen_fscr_facility_check(DisasContext *ctx, int facility_sprn, > static void gen_msr_facility_check(DisasContext *ctx, int facility_sprn, > int bit, int sprn, int cause) > { > - TCGv_i32 t1 = tcg_const_i32(bit); > - TCGv_i32 t2 = tcg_const_i32(sprn); > - TCGv_i32 t3 = tcg_const_i32(cause); > + TCGv_i32 t1 = tcg_constant_i32(bit); > + TCGv_i32 t2 = tcg_constant_i32(sprn); > + TCGv_i32 t3 = tcg_constant_i32(cause); > > gen_helper_msr_facility_check(cpu_env, t1, t2, t3); > } > @@ -1388,7 +1390,7 @@ static inline void gen_op_cmp(TCGv arg0, TCGv arg1, int s, int crf) > > static inline void gen_op_cmpi(TCGv arg0, target_ulong arg1, int s, int crf) > { > - TCGv t0 = tcg_const_tl(arg1); > + TCGv t0 = tcg_constant_tl(arg1); > gen_op_cmp(arg0, t0, s, crf); > } > > @@ -1409,7 +1411,7 @@ static inline void gen_op_cmp32(TCGv arg0, TCGv arg1, int s, int crf) > > static inline void gen_op_cmpi32(TCGv arg0, target_ulong arg1, int s, int crf) > { > - TCGv t0 = tcg_const_tl(arg1); > + TCGv t0 = tcg_constant_tl(arg1); > gen_op_cmp32(arg0, t0, s, crf); > } > > @@ -1476,7 +1478,7 @@ static void gen_isel(DisasContext *ctx) > tcg_gen_extu_i32_tl(t0, cpu_crf[bi >> 2]); > tcg_gen_andi_tl(t0, t0, mask); > > - zr = tcg_const_tl(0); > + zr = tcg_constant_tl(0); > tcg_gen_movcond_tl(TCG_COND_NE, cpu_gpr[rD(ctx->opcode)], t0, zr, > rA(ctx->opcode) ? cpu_gpr[rA(ctx->opcode)] : zr, > cpu_gpr[rB(ctx->opcode)]); > @@ -1568,7 +1570,7 @@ static inline void gen_op_arith_add(DisasContext *ctx, TCGv ret, TCGv arg1, > tcg_gen_mov_tl(ca32, ca); > } > } else { > - TCGv zero = tcg_const_tl(0); > + TCGv zero = tcg_constant_tl(0); > if (add_ca) { > tcg_gen_add2_tl(t0, ca, arg1, zero, ca, zero); > tcg_gen_add2_tl(t0, ca, t0, ca, arg2, zero); > @@ -1609,7 +1611,7 @@ static void glue(gen_, name)(DisasContext *ctx) \ > add_ca, compute_ca, compute_ov) \ > static void glue(gen_, name)(DisasContext *ctx) \ > { \ > - TCGv t0 = tcg_const_tl(const_val); \ > + TCGv t0 = tcg_constant_tl(const_val); \ > gen_op_arith_add(ctx, cpu_gpr[rD(ctx->opcode)], \ > cpu_gpr[rA(ctx->opcode)], t0, \ > ca, glue(ca, 32), \ > @@ -1636,7 +1638,7 @@ GEN_INT_ARITH_ADD_CONST(addzeo, 0x16, 0, cpu_ca, 1, 1, 1) > /* addic addic.*/ > static inline void gen_op_addic(DisasContext *ctx, bool compute_rc0) > { > - TCGv c = tcg_const_tl(SIMM(ctx->opcode)); > + TCGv c = tcg_constant_tl(SIMM(ctx->opcode)); > gen_op_arith_add(ctx, cpu_gpr[rD(ctx->opcode)], cpu_gpr[rA(ctx->opcode)], > c, cpu_ca, cpu_ca32, 0, 1, 0, compute_rc0); > } > @@ -1709,7 +1711,7 @@ GEN_INT_ARITH_DIVW(divwo, 0x1F, 1, 1); > #define GEN_DIVE(name, hlpr, compute_ov) \ > static void gen_##name(DisasContext *ctx) \ > { \ > - TCGv_i32 t0 = tcg_const_i32(compute_ov); \ > + TCGv_i32 t0 = tcg_constant_i32(compute_ov); \ > gen_helper_##hlpr(cpu_gpr[rD(ctx->opcode)], cpu_env, \ > cpu_gpr[rA(ctx->opcode)], cpu_gpr[rB(ctx->opcode)], t0); \ > if (unlikely(Rc(ctx->opcode) != 0)) { \ > @@ -1802,8 +1804,8 @@ static inline void gen_op_arith_modw(DisasContext *ctx, TCGv ret, TCGv arg1, > tcg_gen_rem_i32(t3, t0, t1); > tcg_gen_ext_i32_tl(ret, t3); > } else { > - TCGv_i32 t2 = tcg_const_i32(1); > - TCGv_i32 t3 = tcg_const_i32(0); > + TCGv_i32 t2 = tcg_constant_i32(1); > + TCGv_i32 t3 = tcg_constant_i32(0); > tcg_gen_movcond_i32(TCG_COND_EQ, t1, t1, t3, t2, t1); > tcg_gen_remu_i32(t3, t0, t1); > tcg_gen_extu_i32_tl(ret, t3); > @@ -1842,8 +1844,8 @@ static inline void gen_op_arith_modd(DisasContext *ctx, TCGv ret, TCGv arg1, > tcg_gen_movcond_i64(TCG_COND_NE, t1, t2, t3, t2, t1); > tcg_gen_rem_i64(ret, t0, t1); > } else { > - TCGv_i64 t2 = tcg_const_i64(1); > - TCGv_i64 t3 = tcg_const_i64(0); > + TCGv_i64 t2 = tcg_constant_i64(1); > + TCGv_i64 t3 = tcg_constant_i64(0); > tcg_gen_movcond_i64(TCG_COND_EQ, t1, t1, t3, t2, t1); > tcg_gen_remu_i64(ret, t0, t1); > } > @@ -2038,7 +2040,7 @@ static inline void gen_op_arith_subf(DisasContext *ctx, TCGv ret, TCGv arg1, > } else if (add_ca) { > TCGv zero, inv1 = tcg_temp_new(); > tcg_gen_not_tl(inv1, arg1); > - zero = tcg_const_tl(0); > + zero = tcg_constant_tl(0); > tcg_gen_add2_tl(t0, cpu_ca, arg2, zero, cpu_ca, zero); > tcg_gen_add2_tl(t0, cpu_ca, t0, cpu_ca, inv1, zero); > gen_op_arith_compute_ca32(ctx, t0, inv1, arg2, cpu_ca32, 0); > @@ -2083,7 +2085,7 @@ static void glue(gen_, name)(DisasContext *ctx) \ > add_ca, compute_ca, compute_ov) \ > static void glue(gen_, name)(DisasContext *ctx) \ > { \ > - TCGv t0 = tcg_const_tl(const_val); \ > + TCGv t0 = tcg_constant_tl(const_val); \ > gen_op_arith_subf(ctx, cpu_gpr[rD(ctx->opcode)], \ > cpu_gpr[rA(ctx->opcode)], t0, \ > add_ca, compute_ca, compute_ov, Rc(ctx->opcode)); \ > @@ -2107,7 +2109,7 @@ GEN_INT_ARITH_SUBF_CONST(subfzeo, 0x16, 0, 1, 1, 1) > /* subfic */ > static void gen_subfic(DisasContext *ctx) > { > - TCGv c = tcg_const_tl(SIMM(ctx->opcode)); > + TCGv c = tcg_constant_tl(SIMM(ctx->opcode)); > gen_op_arith_subf(ctx, cpu_gpr[rD(ctx->opcode)], cpu_gpr[rA(ctx->opcode)], > c, 0, 1, 0, 0); > } > @@ -2115,7 +2117,7 @@ static void gen_subfic(DisasContext *ctx) > /* neg neg. nego nego. */ > static inline void gen_op_arith_neg(DisasContext *ctx, bool compute_ov) > { > - TCGv zero = tcg_const_tl(0); > + TCGv zero = tcg_constant_tl(0); > gen_op_arith_subf(ctx, cpu_gpr[rD(ctx->opcode)], cpu_gpr[rA(ctx->opcode)], > zero, 0, 0, compute_ov, Rc(ctx->opcode)); > } > @@ -2214,7 +2216,7 @@ GEN_LOGICAL2(nor, tcg_gen_nor_tl, 0x03, PPC_INTEGER); > #if defined(TARGET_PPC64) && !defined(CONFIG_USER_ONLY) > static void gen_pause(DisasContext *ctx) > { > - TCGv_i32 t0 = tcg_const_i32(0); > + TCGv_i32 t0 = tcg_constant_i32(0); > tcg_gen_st_i32(t0, cpu_env, > -offsetof(PowerPCCPU, env) + offsetof(CPUState, halted)); > > @@ -3256,7 +3258,7 @@ static void gen_lmw(DisasContext *ctx) > } > gen_set_access_type(ctx, ACCESS_INT); > t0 = tcg_temp_new(); > - t1 = tcg_const_i32(rD(ctx->opcode)); > + t1 = tcg_constant_i32(rD(ctx->opcode)); > gen_addr_imm_index(ctx, t0, 0); > gen_helper_lmw(cpu_env, t0, t1); > } > @@ -3273,7 +3275,7 @@ static void gen_stmw(DisasContext *ctx) > } > gen_set_access_type(ctx, ACCESS_INT); > t0 = tcg_temp_new(); > - t1 = tcg_const_i32(rS(ctx->opcode)); > + t1 = tcg_constant_i32(rS(ctx->opcode)); > gen_addr_imm_index(ctx, t0, 0); > gen_helper_stmw(cpu_env, t0, t1); > } > @@ -3311,8 +3313,8 @@ static void gen_lswi(DisasContext *ctx) > gen_set_access_type(ctx, ACCESS_INT); > t0 = tcg_temp_new(); > gen_addr_register(ctx, t0); > - t1 = tcg_const_i32(nb); > - t2 = tcg_const_i32(start); > + t1 = tcg_constant_i32(nb); > + t2 = tcg_constant_i32(start); > gen_helper_lsw(cpu_env, t0, t1, t2); > } > > @@ -3329,9 +3331,9 @@ static void gen_lswx(DisasContext *ctx) > gen_set_access_type(ctx, ACCESS_INT); > t0 = tcg_temp_new(); > gen_addr_reg_index(ctx, t0); > - t1 = tcg_const_i32(rD(ctx->opcode)); > - t2 = tcg_const_i32(rA(ctx->opcode)); > - t3 = tcg_const_i32(rB(ctx->opcode)); > + t1 = tcg_constant_i32(rD(ctx->opcode)); > + t2 = tcg_constant_i32(rA(ctx->opcode)); > + t3 = tcg_constant_i32(rB(ctx->opcode)); > gen_helper_lswx(cpu_env, t0, t1, t2, t3); > } > > @@ -3352,8 +3354,8 @@ static void gen_stswi(DisasContext *ctx) > if (nb == 0) { > nb = 32; > } > - t1 = tcg_const_i32(nb); > - t2 = tcg_const_i32(rS(ctx->opcode)); > + t1 = tcg_constant_i32(nb); > + t2 = tcg_constant_i32(rS(ctx->opcode)); > gen_helper_stsw(cpu_env, t0, t1, t2); > } > > @@ -3373,7 +3375,7 @@ static void gen_stswx(DisasContext *ctx) > t1 = tcg_temp_new_i32(); > tcg_gen_trunc_tl_i32(t1, cpu_xer); > tcg_gen_andi_i32(t1, t1, 0x7F); > - t2 = tcg_const_i32(rS(ctx->opcode)); > + t2 = tcg_constant_i32(rS(ctx->opcode)); > gen_helper_stsw(cpu_env, t0, t1, t2); > } > > @@ -3943,7 +3945,7 @@ static void gen_wait(DisasContext *ctx) > * to occur. > */ > if (wc == 0) { > - TCGv_i32 t0 = tcg_const_i32(1); > + TCGv_i32 t0 = tcg_constant_i32(1); > tcg_gen_st_i32(t0, cpu_env, > -offsetof(PowerPCCPU, env) + offsetof(CPUState, halted)); > /* Stop translation, as the CPU is supposed to sleep from now */ > @@ -3989,7 +3991,7 @@ static void gen_doze(DisasContext *ctx) > TCGv_i32 t; > > CHK_HV(ctx); > - t = tcg_const_i32(PPC_PM_DOZE); > + t = tcg_constant_i32(PPC_PM_DOZE); > gen_helper_pminsn(cpu_env, t); > /* Stop translation, as the CPU is supposed to sleep from now */ > gen_exception_nip(ctx, EXCP_HLT, ctx->base.pc_next); > @@ -4004,7 +4006,7 @@ static void gen_nap(DisasContext *ctx) > TCGv_i32 t; > > CHK_HV(ctx); > - t = tcg_const_i32(PPC_PM_NAP); > + t = tcg_constant_i32(PPC_PM_NAP); > gen_helper_pminsn(cpu_env, t); > /* Stop translation, as the CPU is supposed to sleep from now */ > gen_exception_nip(ctx, EXCP_HLT, ctx->base.pc_next); > @@ -4019,7 +4021,7 @@ static void gen_stop(DisasContext *ctx) > TCGv_i32 t; > > CHK_HV(ctx); > - t = tcg_const_i32(PPC_PM_STOP); > + t = tcg_constant_i32(PPC_PM_STOP); > gen_helper_pminsn(cpu_env, t); > /* Stop translation, as the CPU is supposed to sleep from now */ > gen_exception_nip(ctx, EXCP_HLT, ctx->base.pc_next); > @@ -4034,7 +4036,7 @@ static void gen_sleep(DisasContext *ctx) > TCGv_i32 t; > > CHK_HV(ctx); > - t = tcg_const_i32(PPC_PM_SLEEP); > + t = tcg_constant_i32(PPC_PM_SLEEP); > gen_helper_pminsn(cpu_env, t); > /* Stop translation, as the CPU is supposed to sleep from now */ > gen_exception_nip(ctx, EXCP_HLT, ctx->base.pc_next); > @@ -4049,7 +4051,7 @@ static void gen_rvwinkle(DisasContext *ctx) > TCGv_i32 t; > > CHK_HV(ctx); > - t = tcg_const_i32(PPC_PM_RVWINKLE); > + t = tcg_constant_i32(PPC_PM_RVWINKLE); > gen_helper_pminsn(cpu_env, t); > /* Stop translation, as the CPU is supposed to sleep from now */ > gen_exception_nip(ctx, EXCP_HLT, ctx->base.pc_next); > @@ -4506,7 +4508,7 @@ static void gen_tw(DisasContext *ctx) > if (check_unconditional_trap(ctx)) { > return; > } > - t0 = tcg_const_i32(TO(ctx->opcode)); > + t0 = tcg_constant_i32(TO(ctx->opcode)); > gen_helper_tw(cpu_env, cpu_gpr[rA(ctx->opcode)], cpu_gpr[rB(ctx->opcode)], > t0); > } > @@ -4520,8 +4522,8 @@ static void gen_twi(DisasContext *ctx) > if (check_unconditional_trap(ctx)) { > return; > } > - t0 = tcg_const_tl(SIMM(ctx->opcode)); > - t1 = tcg_const_i32(TO(ctx->opcode)); > + t0 = tcg_constant_tl(SIMM(ctx->opcode)); > + t1 = tcg_constant_i32(TO(ctx->opcode)); > gen_helper_tw(cpu_env, cpu_gpr[rA(ctx->opcode)], t0, t1); > } > > @@ -4534,7 +4536,7 @@ static void gen_td(DisasContext *ctx) > if (check_unconditional_trap(ctx)) { > return; > } > - t0 = tcg_const_i32(TO(ctx->opcode)); > + t0 = tcg_constant_i32(TO(ctx->opcode)); > gen_helper_td(cpu_env, cpu_gpr[rA(ctx->opcode)], cpu_gpr[rB(ctx->opcode)], > t0); > } > @@ -4548,8 +4550,8 @@ static void gen_tdi(DisasContext *ctx) > if (check_unconditional_trap(ctx)) { > return; > } > - t0 = tcg_const_tl(SIMM(ctx->opcode)); > - t1 = tcg_const_i32(TO(ctx->opcode)); > + t0 = tcg_constant_tl(SIMM(ctx->opcode)); > + t1 = tcg_constant_i32(TO(ctx->opcode)); > gen_helper_td(cpu_env, cpu_gpr[rA(ctx->opcode)], t0, t1); > } > #endif > @@ -5026,7 +5028,7 @@ static void gen_dcbz(DisasContext *ctx) > > gen_set_access_type(ctx, ACCESS_CACHE); > tcgv_addr = tcg_temp_new(); > - tcgv_op = tcg_const_i32(ctx->opcode & 0x03FF000); > + tcgv_op = tcg_constant_i32(ctx->opcode & 0x03FF000); > gen_addr_reg_index(ctx, tcgv_addr); > gen_helper_dcbz(cpu_env, tcgv_addr, tcgv_op); > } > @@ -5039,7 +5041,7 @@ static void gen_dcbzep(DisasContext *ctx) > > gen_set_access_type(ctx, ACCESS_CACHE); > tcgv_addr = tcg_temp_new(); > - tcgv_op = tcg_const_i32(ctx->opcode & 0x03FF000); > + tcgv_op = tcg_constant_i32(ctx->opcode & 0x03FF000); > gen_addr_reg_index(ctx, tcgv_addr); > gen_helper_dcbzep(cpu_env, tcgv_addr, tcgv_op); > } > @@ -5114,7 +5116,7 @@ static void gen_mfsr(DisasContext *ctx) > TCGv t0; > > CHK_SV(ctx); > - t0 = tcg_const_tl(SR(ctx->opcode)); > + t0 = tcg_constant_tl(SR(ctx->opcode)); > gen_helper_load_sr(cpu_gpr[rD(ctx->opcode)], cpu_env, t0); > #endif /* defined(CONFIG_USER_ONLY) */ > } > @@ -5143,7 +5145,7 @@ static void gen_mtsr(DisasContext *ctx) > TCGv t0; > > CHK_SV(ctx); > - t0 = tcg_const_tl(SR(ctx->opcode)); > + t0 = tcg_constant_tl(SR(ctx->opcode)); > gen_helper_store_sr(cpu_env, t0, cpu_gpr[rS(ctx->opcode)]); > #endif /* defined(CONFIG_USER_ONLY) */ > } > @@ -5175,7 +5177,7 @@ static void gen_mfsr_64b(DisasContext *ctx) > TCGv t0; > > CHK_SV(ctx); > - t0 = tcg_const_tl(SR(ctx->opcode)); > + t0 = tcg_constant_tl(SR(ctx->opcode)); > gen_helper_load_sr(cpu_gpr[rD(ctx->opcode)], cpu_env, t0); > #endif /* defined(CONFIG_USER_ONLY) */ > } > @@ -5204,7 +5206,7 @@ static void gen_mtsr_64b(DisasContext *ctx) > TCGv t0; > > CHK_SV(ctx); > - t0 = tcg_const_tl(SR(ctx->opcode)); > + t0 = tcg_constant_tl(SR(ctx->opcode)); > gen_helper_store_sr(cpu_env, t0, cpu_gpr[rS(ctx->opcode)]); > #endif /* defined(CONFIG_USER_ONLY) */ > } > @@ -5558,7 +5560,7 @@ static void gen_mfdcr(DisasContext *ctx) > TCGv dcrn; > > CHK_SV(ctx); > - dcrn = tcg_const_tl(SPR(ctx->opcode)); > + dcrn = tcg_constant_tl(SPR(ctx->opcode)); > gen_helper_load_dcr(cpu_gpr[rD(ctx->opcode)], cpu_env, dcrn); > #endif /* defined(CONFIG_USER_ONLY) */ > } > @@ -5572,7 +5574,7 @@ static void gen_mtdcr(DisasContext *ctx) > TCGv dcrn; > > CHK_SV(ctx); > - dcrn = tcg_const_tl(SPR(ctx->opcode)); > + dcrn = tcg_constant_tl(SPR(ctx->opcode)); > gen_helper_store_dcr(cpu_env, dcrn, cpu_gpr[rS(ctx->opcode)]); > #endif /* defined(CONFIG_USER_ONLY) */ > } > @@ -5793,7 +5795,7 @@ static void gen_tlbre_440(DisasContext *ctx) > case 1: > case 2: > { > - TCGv_i32 t0 = tcg_const_i32(rB(ctx->opcode)); > + TCGv_i32 t0 = tcg_constant_i32(rB(ctx->opcode)); > gen_helper_440_tlbre(cpu_gpr[rD(ctx->opcode)], cpu_env, > t0, cpu_gpr[rA(ctx->opcode)]); > } > @@ -5839,7 +5841,7 @@ static void gen_tlbwe_440(DisasContext *ctx) > case 1: > case 2: > { > - TCGv_i32 t0 = tcg_const_i32(rB(ctx->opcode)); > + TCGv_i32 t0 = tcg_constant_i32(rB(ctx->opcode)); > gen_helper_440_tlbwe(cpu_env, t0, cpu_gpr[rA(ctx->opcode)], > cpu_gpr[rS(ctx->opcode)]); > } > @@ -5981,7 +5983,7 @@ static void gen_wrteei(DisasContext *ctx) > /* dlmzb */ > static void gen_dlmzb(DisasContext *ctx) > { > - TCGv_i32 t0 = tcg_const_i32(Rc(ctx->opcode)); > + TCGv_i32 t0 = tcg_constant_i32(Rc(ctx->opcode)); > gen_helper_dlmzb(cpu_gpr[rA(ctx->opcode)], cpu_env, > cpu_gpr[rS(ctx->opcode)], cpu_gpr[rB(ctx->opcode)], t0); > }