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 X-Spam-Level: X-Spam-Status: No, score=-12.8 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_PATCH, MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_HELO_NONE,SPF_PASS,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 236AAC433E2 for ; Fri, 28 Aug 2020 14:46:06 +0000 (UTC) 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 mail.kernel.org (Postfix) with ESMTPS id C7B83207DF for ; Fri, 28 Aug 2020 14:46:05 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (2048-bit key) header.d=linaro.org header.i=@linaro.org header.b="nGd0FEkc" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org C7B83207DF Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=linaro.org Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Received: from localhost ([::1]:36804 helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1kBfdl-0002K6-2g for qemu-devel@archiver.kernel.org; Fri, 28 Aug 2020 10:46:05 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]:51736) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1kBfFE-0001eS-82 for qemu-devel@nongnu.org; Fri, 28 Aug 2020 10:20:44 -0400 Received: from mail-pf1-x436.google.com ([2607:f8b0:4864:20::436]:45328) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1kBfFB-0005KT-3c for qemu-devel@nongnu.org; Fri, 28 Aug 2020 10:20:43 -0400 Received: by mail-pf1-x436.google.com with SMTP id k15so723866pfc.12 for ; Fri, 28 Aug 2020 07:20:40 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=QQQ2WFgIL0ZEv9AAY/hbIBgOjAfsOiOu/hJOwBASNX8=; b=nGd0FEkcAJSFcRp4wIGs9vlxhy/de10JrKrdhB4CNCrWaiH2jjpmvEjvM0fxFYPGmc JkO8CndmQg3pxKlxhO61jiYP0iMWXzu1IvWxmoXbbn29vhDJpQONXEXIsxyARqeWMkv8 eO3GrzFMqzrmF6nQmi1wQYzexJ9QHMU5isMepTcuIawt8/QamWe00K+YbChphJMMOFmd ZdiaScBZlGcy5xZN6nuLaqFdKNivV8s2aWMio6sfDrOlPMJUV4HTD/lv2j9+nxIu1NbI 9oN/yY1i4onsxxCJRQ4UROn1t4TC9RThwNI7JZD7YPgKofkbqacL6LuJKpX7YEqmPySY aG4Q== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=QQQ2WFgIL0ZEv9AAY/hbIBgOjAfsOiOu/hJOwBASNX8=; b=J0CiWl9MHAfxBQ69hoI2UWn0btK6AyjtlBzbwm/AHASpT+FB6eYeWwklBRGNEs4X8Z jexEOJqICRosurpy8YG+DzVs8dFxQCSqhYA84QwNTFivoGz4PJJMt0PcAXQd5eHL9EVQ W+CGlsCPoszwdn1ygc+fVKTWnt2IKL1CtJb6EmVH7Mp6r5pmnQa/ReC843Mhg4RWbw89 Dpze1+9NUU7R8AyIY8SMzXpRGOPs9WzQdgJNfTuXv4F+S17X/m4t+YlI5zhpiw7TgnNM 6u9uc9Q4mFkzi/1My8FZU9IIElkv0ws8BGO51nElmo1itjWCU59MW4/7gYb48bOdJSJD 8GSA== X-Gm-Message-State: AOAM532u7U2PX/3R5iiKxhoIvPD2ZzFuWib5+3bT/jEodFupVYIfMxec Mp9pW5+EqhMwCHUV+VOMVyLw7InlwkfWhg== X-Google-Smtp-Source: ABdhPJx+X+ToUtq7tESVNC+FIsh1utQQ0dpZylUJKPEYb7LnbB2nCPzNMy5eB/LN1we6FYIpH0SfLg== X-Received: by 2002:a63:8543:: with SMTP id u64mr1427842pgd.350.1598624438542; Fri, 28 Aug 2020 07:20:38 -0700 (PDT) Received: from localhost.localdomain ([71.212.141.89]) by smtp.gmail.com with ESMTPSA id j3sm1403080pjw.23.2020.08.28.07.20.37 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 28 Aug 2020 07:20:37 -0700 (PDT) From: Richard Henderson To: qemu-devel@nongnu.org Subject: [PATCH v2 53/76] target/microblaze: Convert dec_load and dec_store to decodetree Date: Fri, 28 Aug 2020 07:19:06 -0700 Message-Id: <20200828141929.77854-54-richard.henderson@linaro.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200828141929.77854-1-richard.henderson@linaro.org> References: <20200828141929.77854-1-richard.henderson@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Received-SPF: pass client-ip=2607:f8b0:4864:20::436; envelope-from=richard.henderson@linaro.org; helo=mail-pf1-x436.google.com X-detected-operating-system: by eggs.gnu.org: No matching host in p0f cache. That's all we know. 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: 2.1.23 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: edgar.iglesias@xilinx.com Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: "Qemu-devel" Signed-off-by: Richard Henderson --- target/microblaze/insns.decode | 32 ++ target/microblaze/translate.c | 723 +++++++++++++++++++-------------- 2 files changed, 456 insertions(+), 299 deletions(-) diff --git a/target/microblaze/insns.decode b/target/microblaze/insns.decode index 87e8f5679b..47b92b9cbc 100644 --- a/target/microblaze/insns.decode +++ b/target/microblaze/insns.decode @@ -100,6 +100,22 @@ idivu 010010 ..... ..... ..... 000 0000 0010 @typea imm 101100 00000 00000 imm:16 +lbu 110000 ..... ..... ..... 0000 000 0000 @typea +lbur 110000 ..... ..... ..... 0100 000 0000 @typea +lbuea 110000 ..... ..... ..... 0001 000 0000 @typea +lbui 111000 ..... ..... ................ @typeb + +lhu 110001 ..... ..... ..... 0000 000 0000 @typea +lhur 110001 ..... ..... ..... 0100 000 0000 @typea +lhuea 110001 ..... ..... ..... 0001 000 0000 @typea +lhui 111001 ..... ..... ................ @typeb + +lw 110010 ..... ..... ..... 0000 000 0000 @typea +lwr 110010 ..... ..... ..... 0100 000 0000 @typea +lwea 110010 ..... ..... ..... 0001 000 0000 @typea +lwx 110010 ..... ..... ..... 1000 000 0000 @typea +lwi 111010 ..... ..... ................ @typeb + mul 010000 ..... ..... ..... 000 0000 0000 @typea mulh 010000 ..... ..... ..... 000 0000 0001 @typea mulhu 010000 ..... ..... ..... 000 0000 0011 @typea @@ -123,6 +139,22 @@ rsubic 001011 ..... ..... ................ @typeb rsubik 001101 ..... ..... ................ @typeb rsubikc 001111 ..... ..... ................ @typeb +sb 110100 ..... ..... ..... 0000 000 0000 @typea +sbr 110100 ..... ..... ..... 0100 000 0000 @typea +sbea 110100 ..... ..... ..... 0001 000 0000 @typea +sbi 111100 ..... ..... ................ @typeb + +sh 110101 ..... ..... ..... 0000 000 0000 @typea +shr 110101 ..... ..... ..... 0100 000 0000 @typea +shea 110101 ..... ..... ..... 0001 000 0000 @typea +shi 111101 ..... ..... ................ @typeb + +sw 110110 ..... ..... ..... 0000 000 0000 @typea +swr 110110 ..... ..... ..... 0100 000 0000 @typea +swea 110110 ..... ..... ..... 0001 000 0000 @typea +swx 110110 ..... ..... ..... 1000 000 0000 @typea +swi 111110 ..... ..... ................ @typeb + sext8 100100 ..... ..... 00000 000 0110 0000 @typea0 sext16 100100 ..... ..... 00000 000 0110 0001 @typea0 diff --git a/target/microblaze/translate.c b/target/microblaze/translate.c index a55e110171..d2baa7db0e 100644 --- a/target/microblaze/translate.c +++ b/target/microblaze/translate.c @@ -105,6 +105,17 @@ static inline void t_sync_flags(DisasContext *dc) } } +static inline void sync_jmpstate(DisasContext *dc) +{ + if (dc->jmp == JMP_DIRECT || dc->jmp == JMP_DIRECT_CC) { + if (dc->jmp == JMP_DIRECT) { + tcg_gen_movi_i32(cpu_btaken, 1); + } + dc->jmp = JMP_INDIRECT; + tcg_gen_movi_i32(cpu_btarget, dc->jmp_pc); + } +} + static void gen_raise_exception(DisasContext *dc, uint32_t index) { TCGv_i32 tmp = tcg_const_i32(index); @@ -668,6 +679,419 @@ static bool trans_wdic(DisasContext *dc, arg_wdic *a) DO_TYPEA(xor, false, tcg_gen_xor_i32) DO_TYPEBI(xori, false, tcg_gen_xori_i32) +static TCGv compute_ldst_addr_typea(DisasContext *dc, int ra, int rb) +{ + TCGv ret = tcg_temp_new(); + + /* If any of the regs is r0, set t to the value of the other reg. */ + if (ra && rb) { + TCGv_i32 tmp = tcg_temp_new_i32(); + tcg_gen_add_i32(tmp, cpu_R[ra], cpu_R[rb]); + tcg_gen_extu_i32_tl(ret, tmp); + tcg_temp_free_i32(tmp); + } else if (ra) { + tcg_gen_extu_i32_tl(ret, cpu_R[ra]); + } else if (rb) { + tcg_gen_extu_i32_tl(ret, cpu_R[rb]); + } else { + tcg_gen_movi_tl(ret, 0); + } + + if ((ra == 1 || rb == 1) && dc->cpu->cfg.stackprot) { + gen_helper_stackprot(cpu_env, ret); + } + return ret; +} + +static TCGv compute_ldst_addr_typeb(DisasContext *dc, int ra, int imm) +{ + TCGv ret = tcg_temp_new(); + + /* If any of the regs is r0, set t to the value of the other reg. */ + if (ra) { + TCGv_i32 tmp = tcg_temp_new_i32(); + tcg_gen_addi_i32(tmp, cpu_R[ra], imm); + tcg_gen_extu_i32_tl(ret, tmp); + tcg_temp_free_i32(tmp); + } else { + tcg_gen_movi_tl(ret, (uint32_t)imm); + } + + if (ra == 1 && dc->cpu->cfg.stackprot) { + gen_helper_stackprot(cpu_env, ret); + } + return ret; +} + +static TCGv compute_ldst_addr_ea(DisasContext *dc, int ra, int rb) +{ + int addr_size = dc->cpu->cfg.addr_size; + TCGv ret = tcg_temp_new(); + + if (addr_size == 32 || ra == 0) { + if (rb) { + tcg_gen_extu_i32_tl(ret, cpu_R[rb]); + } else { + tcg_gen_movi_tl(ret, 0); + } + } else { + if (rb) { + tcg_gen_concat_i32_i64(ret, cpu_R[rb], cpu_R[ra]); + } else { + tcg_gen_extu_i32_tl(ret, cpu_R[ra]); + tcg_gen_shli_tl(ret, ret, 32); + } + if (addr_size < 64) { + /* Mask off out of range bits. */ + tcg_gen_andi_i64(ret, ret, MAKE_64BIT_MASK(0, addr_size)); + } + } + return ret; +} + +static bool do_load(DisasContext *dc, int rd, TCGv addr, MemOp mop, + int mem_index, bool rev) +{ + TCGv_i32 v; + MemOp size = mop & MO_SIZE; + + /* + * When doing reverse accesses we need to do two things. + * + * 1. Reverse the address wrt endianness. + * 2. Byteswap the data lanes on the way back into the CPU core. + */ + if (rev) { + if (size > MO_8) { + mop ^= MO_BSWAP; + } + if (size < MO_32) { + tcg_gen_xori_tl(addr, addr, 3 - size); + } + } + + t_sync_flags(dc); + sync_jmpstate(dc); + + /* + * Microblaze gives MMU faults priority over faults due to + * unaligned addresses. That's why we speculatively do the load + * into v. If the load succeeds, we verify alignment of the + * address and if that succeeds we write into the destination reg. + */ + v = tcg_temp_new_i32(); + tcg_gen_qemu_ld_i32(v, addr, mem_index, mop); + + /* TODO: Convert to CPUClass::do_unaligned_access. */ + if (dc->cpu->cfg.unaligned_exceptions && size > MO_8) { + TCGv_i32 t0 = tcg_const_i32(0); + TCGv_i32 treg = tcg_const_i32(rd); + TCGv_i32 tsize = tcg_const_i32((1 << size) - 1); + + tcg_gen_movi_i32(cpu_pc, dc->base.pc_next); + gen_helper_memalign(cpu_env, addr, treg, t0, tsize); + + tcg_temp_free_i32(t0); + tcg_temp_free_i32(treg); + tcg_temp_free_i32(tsize); + } + + if (rd) { + tcg_gen_mov_i32(cpu_R[rd], v); + } + + tcg_temp_free_i32(v); + tcg_temp_free(addr); + return true; +} + +static bool trans_lbu(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + return do_load(dc, arg->rd, addr, MO_UB, dc->mem_index, false); +} + +static bool trans_lbur(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + return do_load(dc, arg->rd, addr, MO_UB, dc->mem_index, true); +} + +static bool trans_lbuea(DisasContext *dc, arg_typea *arg) +{ + if (trap_userspace(dc, true)) { + return true; + } + TCGv addr = compute_ldst_addr_ea(dc, arg->ra, arg->rb); + return do_load(dc, arg->rd, addr, MO_UB, MMU_NOMMU_IDX, false); +} + +static bool trans_lbui(DisasContext *dc, arg_typeb *arg) +{ + TCGv addr = compute_ldst_addr_typeb(dc, arg->ra, arg->imm); + return do_load(dc, arg->rd, addr, MO_UB, dc->mem_index, false); +} + +static bool trans_lhu(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + return do_load(dc, arg->rd, addr, MO_TEUW, dc->mem_index, false); +} + +static bool trans_lhur(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + return do_load(dc, arg->rd, addr, MO_TEUW, dc->mem_index, true); +} + +static bool trans_lhuea(DisasContext *dc, arg_typea *arg) +{ + if (trap_userspace(dc, true)) { + return true; + } + TCGv addr = compute_ldst_addr_ea(dc, arg->ra, arg->rb); + return do_load(dc, arg->rd, addr, MO_TEUW, MMU_NOMMU_IDX, false); +} + +static bool trans_lhui(DisasContext *dc, arg_typeb *arg) +{ + TCGv addr = compute_ldst_addr_typeb(dc, arg->ra, arg->imm); + return do_load(dc, arg->rd, addr, MO_TEUW, dc->mem_index, false); +} + +static bool trans_lw(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + return do_load(dc, arg->rd, addr, MO_TEUL, dc->mem_index, false); +} + +static bool trans_lwr(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + return do_load(dc, arg->rd, addr, MO_TEUL, dc->mem_index, true); +} + +static bool trans_lwea(DisasContext *dc, arg_typea *arg) +{ + if (trap_userspace(dc, true)) { + return true; + } + TCGv addr = compute_ldst_addr_ea(dc, arg->ra, arg->rb); + return do_load(dc, arg->rd, addr, MO_TEUL, MMU_NOMMU_IDX, false); +} + +static bool trans_lwi(DisasContext *dc, arg_typeb *arg) +{ + TCGv addr = compute_ldst_addr_typeb(dc, arg->ra, arg->imm); + return do_load(dc, arg->rd, addr, MO_TEUL, dc->mem_index, false); +} + +static bool trans_lwx(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + + /* lwx does not throw unaligned access errors, so force alignment */ + tcg_gen_andi_tl(addr, addr, ~3); + + t_sync_flags(dc); + sync_jmpstate(dc); + + tcg_gen_qemu_ld_i32(cpu_res_val, addr, dc->mem_index, MO_TEUL); + tcg_gen_mov_tl(cpu_res_addr, addr); + tcg_temp_free(addr); + + if (arg->rd) { + tcg_gen_mov_i32(cpu_R[arg->rd], cpu_res_val); + } + + /* No support for AXI exclusive so always clear C */ + tcg_gen_movi_i32(cpu_msr_c, 0); + return true; +} + +static bool do_store(DisasContext *dc, int rd, TCGv addr, MemOp mop, + int mem_index, bool rev) +{ + MemOp size = mop & MO_SIZE; + + /* + * When doing reverse accesses we need to do two things. + * + * 1. Reverse the address wrt endianness. + * 2. Byteswap the data lanes on the way back into the CPU core. + */ + if (rev) { + if (size > MO_8) { + mop ^= MO_BSWAP; + } + if (size < MO_32) { + tcg_gen_xori_tl(addr, addr, 3 - size); + } + } + + t_sync_flags(dc); + sync_jmpstate(dc); + + tcg_gen_qemu_st_i32(reg_for_read(dc, rd), addr, mem_index, mop); + + /* TODO: Convert to CPUClass::do_unaligned_access. */ + if (dc->cpu->cfg.unaligned_exceptions && size > MO_8) { + TCGv_i32 t1 = tcg_const_i32(1); + TCGv_i32 treg = tcg_const_i32(rd); + TCGv_i32 tsize = tcg_const_i32((1 << size) - 1); + + tcg_gen_movi_i32(cpu_pc, dc->base.pc_next); + /* FIXME: if the alignment is wrong, we should restore the value + * in memory. One possible way to achieve this is to probe + * the MMU prior to the memaccess, thay way we could put + * the alignment checks in between the probe and the mem + * access. + */ + gen_helper_memalign(cpu_env, addr, treg, t1, tsize); + + tcg_temp_free_i32(t1); + tcg_temp_free_i32(treg); + tcg_temp_free_i32(tsize); + } + + tcg_temp_free(addr); + return true; +} + +static bool trans_sb(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + return do_store(dc, arg->rd, addr, MO_UB, dc->mem_index, false); +} + +static bool trans_sbr(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + return do_store(dc, arg->rd, addr, MO_UB, dc->mem_index, true); +} + +static bool trans_sbea(DisasContext *dc, arg_typea *arg) +{ + if (trap_userspace(dc, true)) { + return true; + } + TCGv addr = compute_ldst_addr_ea(dc, arg->ra, arg->rb); + return do_store(dc, arg->rd, addr, MO_UB, MMU_NOMMU_IDX, false); +} + +static bool trans_sbi(DisasContext *dc, arg_typeb *arg) +{ + TCGv addr = compute_ldst_addr_typeb(dc, arg->ra, arg->imm); + return do_store(dc, arg->rd, addr, MO_UB, dc->mem_index, false); +} + +static bool trans_sh(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + return do_store(dc, arg->rd, addr, MO_TEUW, dc->mem_index, false); +} + +static bool trans_shr(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + return do_store(dc, arg->rd, addr, MO_TEUW, dc->mem_index, true); +} + +static bool trans_shea(DisasContext *dc, arg_typea *arg) +{ + if (trap_userspace(dc, true)) { + return true; + } + TCGv addr = compute_ldst_addr_ea(dc, arg->ra, arg->rb); + return do_store(dc, arg->rd, addr, MO_TEUW, MMU_NOMMU_IDX, false); +} + +static bool trans_shi(DisasContext *dc, arg_typeb *arg) +{ + TCGv addr = compute_ldst_addr_typeb(dc, arg->ra, arg->imm); + return do_store(dc, arg->rd, addr, MO_TEUW, dc->mem_index, false); +} + +static bool trans_sw(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + return do_store(dc, arg->rd, addr, MO_TEUL, dc->mem_index, false); +} + +static bool trans_swr(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + return do_store(dc, arg->rd, addr, MO_TEUL, dc->mem_index, true); +} + +static bool trans_swea(DisasContext *dc, arg_typea *arg) +{ + if (trap_userspace(dc, true)) { + return true; + } + TCGv addr = compute_ldst_addr_ea(dc, arg->ra, arg->rb); + return do_store(dc, arg->rd, addr, MO_TEUL, MMU_NOMMU_IDX, false); +} + +static bool trans_swi(DisasContext *dc, arg_typeb *arg) +{ + TCGv addr = compute_ldst_addr_typeb(dc, arg->ra, arg->imm); + return do_store(dc, arg->rd, addr, MO_TEUL, dc->mem_index, false); +} + +static bool trans_swx(DisasContext *dc, arg_typea *arg) +{ + TCGv addr = compute_ldst_addr_typea(dc, arg->ra, arg->rb); + TCGLabel *swx_done = gen_new_label(); + TCGLabel *swx_fail = gen_new_label(); + TCGv_i32 tval; + + t_sync_flags(dc); + sync_jmpstate(dc); + + /* swx does not throw unaligned access errors, so force alignment */ + tcg_gen_andi_tl(addr, addr, ~3); + + /* + * Compare the address vs the one we used during lwx. + * On mismatch, the operation fails. On match, addr dies at the + * branch, but we know we can use the equal version in the global. + * In either case, addr is no longer needed. + */ + tcg_gen_brcond_tl(TCG_COND_NE, cpu_res_addr, addr, swx_fail); + tcg_temp_free(addr); + + /* + * Compare the value loaded during lwx with current contents of + * the reserved location. + */ + tval = tcg_temp_new_i32(); + + tcg_gen_atomic_cmpxchg_i32(tval, cpu_res_addr, cpu_res_val, + reg_for_write(dc, arg->rd), + dc->mem_index, MO_TEUL); + + tcg_gen_brcond_i32(TCG_COND_NE, cpu_res_val, tval, swx_fail); + tcg_temp_free_i32(tval); + + /* Success */ + tcg_gen_movi_i32(cpu_msr_c, 0); + tcg_gen_br(swx_done); + + /* Failure */ + gen_set_label(swx_fail); + tcg_gen_movi_i32(cpu_msr_c, 1); + + gen_set_label(swx_done); + + /* + * Prevent the saved address from working again without another ldx. + * Akin to the pseudocode setting reservation = 0. + */ + tcg_gen_movi_tl(cpu_res_addr, -1); + return true; +} + static bool trans_zero(DisasContext *dc, arg_zero *arg) { /* If opcode_0_illegal, trap. */ @@ -887,303 +1311,6 @@ static void dec_msr(DisasContext *dc) } } -static inline void sync_jmpstate(DisasContext *dc) -{ - if (dc->jmp == JMP_DIRECT || dc->jmp == JMP_DIRECT_CC) { - if (dc->jmp == JMP_DIRECT) { - tcg_gen_movi_i32(cpu_btaken, 1); - } - dc->jmp = JMP_INDIRECT; - tcg_gen_movi_i32(cpu_btarget, dc->jmp_pc); - } -} - -static inline void compute_ldst_addr(DisasContext *dc, bool ea, TCGv t) -{ - /* Should be set to true if r1 is used by loadstores. */ - bool stackprot = false; - TCGv_i32 t32; - - /* All load/stores use ra. */ - if (dc->ra == 1 && dc->cpu->cfg.stackprot) { - stackprot = true; - } - - /* Treat the common cases first. */ - if (!dc->type_b) { - if (ea) { - int addr_size = dc->cpu->cfg.addr_size; - - if (addr_size == 32) { - tcg_gen_extu_i32_tl(t, cpu_R[dc->rb]); - return; - } - - tcg_gen_concat_i32_i64(t, cpu_R[dc->rb], cpu_R[dc->ra]); - if (addr_size < 64) { - /* Mask off out of range bits. */ - tcg_gen_andi_i64(t, t, MAKE_64BIT_MASK(0, addr_size)); - } - return; - } - - /* If any of the regs is r0, set t to the value of the other reg. */ - if (dc->ra == 0) { - tcg_gen_extu_i32_tl(t, cpu_R[dc->rb]); - return; - } else if (dc->rb == 0) { - tcg_gen_extu_i32_tl(t, cpu_R[dc->ra]); - return; - } - - if (dc->rb == 1 && dc->cpu->cfg.stackprot) { - stackprot = true; - } - - t32 = tcg_temp_new_i32(); - tcg_gen_add_i32(t32, cpu_R[dc->ra], cpu_R[dc->rb]); - tcg_gen_extu_i32_tl(t, t32); - tcg_temp_free_i32(t32); - - if (stackprot) { - gen_helper_stackprot(cpu_env, t); - } - return; - } - /* Immediate. */ - t32 = tcg_temp_new_i32(); - tcg_gen_addi_i32(t32, cpu_R[dc->ra], dec_alu_typeb_imm(dc)); - tcg_gen_extu_i32_tl(t, t32); - tcg_temp_free_i32(t32); - - if (stackprot) { - gen_helper_stackprot(cpu_env, t); - } - return; -} - -static void dec_load(DisasContext *dc) -{ - TCGv_i32 v; - TCGv addr; - unsigned int size; - bool rev = false, ex = false, ea = false; - int mem_index = dc->mem_index; - MemOp mop; - - mop = dc->opcode & 3; - size = 1 << mop; - if (!dc->type_b) { - ea = extract32(dc->ir, 7, 1); - rev = extract32(dc->ir, 9, 1); - ex = extract32(dc->ir, 10, 1); - } - mop |= MO_TE; - if (rev) { - mop ^= MO_BSWAP; - } - - if (trap_illegal(dc, size > 4)) { - return; - } - - if (trap_userspace(dc, ea)) { - return; - } - - t_sync_flags(dc); - addr = tcg_temp_new(); - compute_ldst_addr(dc, ea, addr); - /* Extended addressing bypasses the MMU. */ - mem_index = ea ? MMU_NOMMU_IDX : mem_index; - - /* - * When doing reverse accesses we need to do two things. - * - * 1. Reverse the address wrt endianness. - * 2. Byteswap the data lanes on the way back into the CPU core. - */ - if (rev && size != 4) { - /* Endian reverse the address. t is addr. */ - switch (size) { - case 1: - { - tcg_gen_xori_tl(addr, addr, 3); - break; - } - - case 2: - /* 00 -> 10 - 10 -> 00. */ - tcg_gen_xori_tl(addr, addr, 2); - break; - default: - cpu_abort(CPU(dc->cpu), "Invalid reverse size\n"); - break; - } - } - - /* lwx does not throw unaligned access errors, so force alignment */ - if (ex) { - tcg_gen_andi_tl(addr, addr, ~3); - } - - /* If we get a fault on a dslot, the jmpstate better be in sync. */ - sync_jmpstate(dc); - - /* Verify alignment if needed. */ - /* - * Microblaze gives MMU faults priority over faults due to - * unaligned addresses. That's why we speculatively do the load - * into v. If the load succeeds, we verify alignment of the - * address and if that succeeds we write into the destination reg. - */ - v = tcg_temp_new_i32(); - tcg_gen_qemu_ld_i32(v, addr, mem_index, mop); - - if (dc->cpu->cfg.unaligned_exceptions && size > 1) { - TCGv_i32 t0 = tcg_const_i32(0); - TCGv_i32 treg = tcg_const_i32(dc->rd); - TCGv_i32 tsize = tcg_const_i32(size - 1); - - tcg_gen_movi_i32(cpu_pc, dc->base.pc_next); - gen_helper_memalign(cpu_env, addr, treg, t0, tsize); - - tcg_temp_free_i32(t0); - tcg_temp_free_i32(treg); - tcg_temp_free_i32(tsize); - } - - if (ex) { - tcg_gen_mov_tl(cpu_res_addr, addr); - tcg_gen_mov_i32(cpu_res_val, v); - } - if (dc->rd) { - tcg_gen_mov_i32(cpu_R[dc->rd], v); - } - tcg_temp_free_i32(v); - - if (ex) { /* lwx */ - /* no support for AXI exclusive so always clear C */ - tcg_gen_movi_i32(cpu_msr_c, 0); - } - - tcg_temp_free(addr); -} - -static void dec_store(DisasContext *dc) -{ - TCGv addr; - TCGLabel *swx_skip = NULL; - unsigned int size; - bool rev = false, ex = false, ea = false; - int mem_index = dc->mem_index; - MemOp mop; - - mop = dc->opcode & 3; - size = 1 << mop; - if (!dc->type_b) { - ea = extract32(dc->ir, 7, 1); - rev = extract32(dc->ir, 9, 1); - ex = extract32(dc->ir, 10, 1); - } - mop |= MO_TE; - if (rev) { - mop ^= MO_BSWAP; - } - - if (trap_illegal(dc, size > 4)) { - return; - } - - trap_userspace(dc, ea); - - t_sync_flags(dc); - /* If we get a fault on a dslot, the jmpstate better be in sync. */ - sync_jmpstate(dc); - /* SWX needs a temp_local. */ - addr = ex ? tcg_temp_local_new() : tcg_temp_new(); - compute_ldst_addr(dc, ea, addr); - /* Extended addressing bypasses the MMU. */ - mem_index = ea ? MMU_NOMMU_IDX : mem_index; - - if (ex) { /* swx */ - TCGv_i32 tval; - - /* swx does not throw unaligned access errors, so force alignment */ - tcg_gen_andi_tl(addr, addr, ~3); - - tcg_gen_movi_i32(cpu_msr_c, 1); - swx_skip = gen_new_label(); - tcg_gen_brcond_tl(TCG_COND_NE, cpu_res_addr, addr, swx_skip); - - /* - * Compare the value loaded at lwx with current contents of - * the reserved location. - */ - tval = tcg_temp_new_i32(); - - tcg_gen_atomic_cmpxchg_i32(tval, addr, cpu_res_val, - cpu_R[dc->rd], mem_index, - mop); - - tcg_gen_brcond_i32(TCG_COND_NE, cpu_res_val, tval, swx_skip); - tcg_gen_movi_i32(cpu_msr_c, 0); - tcg_temp_free_i32(tval); - } - - if (rev && size != 4) { - /* Endian reverse the address. t is addr. */ - switch (size) { - case 1: - { - tcg_gen_xori_tl(addr, addr, 3); - break; - } - - case 2: - /* 00 -> 10 - 10 -> 00. */ - /* Force addr into the temp. */ - tcg_gen_xori_tl(addr, addr, 2); - break; - default: - cpu_abort(CPU(dc->cpu), "Invalid reverse size\n"); - break; - } - } - - if (!ex) { - tcg_gen_qemu_st_i32(cpu_R[dc->rd], addr, mem_index, mop); - } - - /* Verify alignment if needed. */ - if (dc->cpu->cfg.unaligned_exceptions && size > 1) { - TCGv_i32 t1 = tcg_const_i32(1); - TCGv_i32 treg = tcg_const_i32(dc->rd); - TCGv_i32 tsize = tcg_const_i32(size - 1); - - tcg_gen_movi_i32(cpu_pc, dc->base.pc_next); - /* FIXME: if the alignment is wrong, we should restore the value - * in memory. One possible way to achieve this is to probe - * the MMU prior to the memaccess, thay way we could put - * the alignment checks in between the probe and the mem - * access. - */ - gen_helper_memalign(cpu_env, addr, treg, t1, tsize); - - tcg_temp_free_i32(t1); - tcg_temp_free_i32(treg); - tcg_temp_free_i32(tsize); - } - - if (ex) { - gen_set_label(swx_skip); - } - - tcg_temp_free(addr); -} - static inline void eval_cc(DisasContext *dc, unsigned int cc, TCGv_i32 d, TCGv_i32 a) { @@ -1491,8 +1618,6 @@ static struct decoder_info { }; void (*dec)(DisasContext *dc); } decinfo[] = { - {DEC_LD, dec_load}, - {DEC_ST, dec_store}, {DEC_BR, dec_br}, {DEC_BCC, dec_bcc}, {DEC_RTS, dec_rts}, -- 2.25.1