From nobody Mon Feb 9 22:08:09 2026 Delivered-To: importer@patchew.org Authentication-Results: mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass(p=none dis=none) header.from=redhat.com ARC-Seal: i=1; a=rsa-sha256; t=1662058037; cv=none; d=zohomail.com; s=zohoarc; b=XXyp9/wSN3BbGizMFPBHF8VgcgnLZLo9gWs69+vzDuMA2vw7gvmDPhrQHx5a4LkVp8fVVEpm5/S7dwc6LtNjlKGVhsAxo3yhhidoEJ/IW/w/oJA1gfbl3+ILqJC6hMDlhD47NKYhgus6zifQ4VRk5wZE3XVPVinKokb+HP8xh9U= ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=zohomail.com; s=zohoarc; t=1662058037; h=Content-Type:Content-Transfer-Encoding:Cc:Date:From:In-Reply-To:List-Subscribe:List-Post:List-Id:List-Archive:List-Help:List-Unsubscribe:MIME-Version:Message-ID:References:Sender:Subject:To; bh=MvcLLysMARUchQ4M6kaWE+2EUeOvL+cfwn/GHeZEhU8=; b=G+ZXr1v8Y29gu0ZsSrNpGHrDWcvvc8p+HNGbelfJc4Hm56bamId1jg3kmZAKdO4pcHjSV3f/vjB7sXrztQ/OybxnHuff2AE2Ia53HdVlHwqktA0aoVX7zrmvPjx/bbGWoOJv7AsDFsdfgR/GnjjC1yfhQuB8gyX7gTgb8o1QCxI= ARC-Authentication-Results: i=1; mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass header.from= (p=none dis=none) Return-Path: Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) by mx.zohomail.com with SMTPS id 1662058037930188.5108160403164; Thu, 1 Sep 2022 11:47:17 -0700 (PDT) Received: from localhost ([::1]:54392 helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1oTpDf-0004nZ-Ju for importer@patchew.org; Thu, 01 Sep 2022 14:47:15 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]:41788) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1oTosU-0008CF-0Q for qemu-devel@nongnu.org; Thu, 01 Sep 2022 14:25:22 -0400 Received: from us-smtp-delivery-124.mimecast.com ([170.10.133.124]:58031) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1oTosM-00006a-JH for qemu-devel@nongnu.org; Thu, 01 Sep 2022 14:25:21 -0400 Received: from mail-ed1-f70.google.com (mail-ed1-f70.google.com [209.85.208.70]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_128_GCM_SHA256) id us-mta-52-KEfzrnUwP1S8wibRogjA5A-1; Thu, 01 Sep 2022 14:25:12 -0400 Received: by mail-ed1-f70.google.com with SMTP id y14-20020a056402440e00b0044301c7ccd9so12215289eda.19 for ; Thu, 01 Sep 2022 11:25:12 -0700 (PDT) Received: from goa-sendmail ([2001:b07:6468:f312:9af8:e5f5:7516:fa89]) by smtp.gmail.com with ESMTPSA id d13-20020a50fb0d000000b00445bda73fbesm1796895edq.33.2022.09.01.11.25.09 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Thu, 01 Sep 2022 11:25:10 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1662056714; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=MvcLLysMARUchQ4M6kaWE+2EUeOvL+cfwn/GHeZEhU8=; b=VZtcn4s8f/9R0ywK1y6gQYW2GzLF/hbyw77MHNMuOWE9WZhSu7B2/5tQ8KEeSYo0flseb4 wUeBcs87rLrOyxWumskLWV81wIwhWvu4/0oyv2TL9XJ4grrSxv7TLUP2Q3XV49pMS56OpP jrre4SYQC2cBZe3OUhlWy7pRDfad3v0= X-MC-Unique: KEfzrnUwP1S8wibRogjA5A-1 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; bh=MvcLLysMARUchQ4M6kaWE+2EUeOvL+cfwn/GHeZEhU8=; b=Gh4zx8yajOWPYIYWG2SY8erL1LRuEOg73FM38Ztt5QM5LULkAUvGkcYybkVAxw3N3w 7TQTCgN9NwI0SuWPplgmV0MF7jRBWUAvucAd4kD5jOzDSWbYn78tmGGMSCZzaPTtAHin Ods4iD+g9hX6xY8t/N6osLEGaiR20kUDNXhiFuyMVFLQci2b0d8mqtKp37WqqEHQV133 90ZVi2VHrWhkpvqE1fxj18k3rc0YIp4vDl8XscAxPy3N+QGKPPXaAMcEOUsz571YsYXM uBpe+eC0yCMgBDgz3SdaaZUnDtzwwssdsMhS7SGU2U8IxB1nmu4ZZ1C3z7c/lpWrKmA6 UuyA== X-Gm-Message-State: ACgBeo0b7E2MvnhfqhAta0YSBra3+LCtAgbzcIQNeU8Vnaqv/2RiPnKb wrSLmFP4DCZ1L0IuEIAXFuT2nmPL3hxjNmN56F+S5cWxuSrfbiQ8DkiJkZBqc5dpKMMfhwQzgLR oms/4OEAY03INzcNxUYPftzx0U1ql4Liy65z6poUNeQpW0EAXB8jh3uEaOGwd0n1mO/Q= X-Received: by 2002:a17:907:7394:b0:73a:d077:9ba1 with SMTP id er20-20020a170907739400b0073ad0779ba1mr23679218ejc.697.1662056711303; Thu, 01 Sep 2022 11:25:11 -0700 (PDT) X-Google-Smtp-Source: AA6agR5bVhbsVaV6jU7VtMvS85IL5XiHPCUAyXjKF/vbh9GhYMQgpHbJ7KYjbJ4V58H4n4qF9YCAQw== X-Received: by 2002:a17:907:7394:b0:73a:d077:9ba1 with SMTP id er20-20020a170907739400b0073ad0779ba1mr23679192ejc.697.1662056710948; Thu, 01 Sep 2022 11:25:10 -0700 (PDT) From: Paolo Bonzini To: qemu-devel@nongnu.org Cc: Paul Brook , Richard Henderson Subject: [PULL 20/39] target/i386: Rework sse_op_table6/7 Date: Thu, 1 Sep 2022 20:24:10 +0200 Message-Id: <20220901182429.93533-21-pbonzini@redhat.com> X-Mailer: git-send-email 2.37.2 In-Reply-To: <20220901182429.93533-1-pbonzini@redhat.com> References: <20220901182429.93533-1-pbonzini@redhat.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Received-SPF: pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) client-ip=209.51.188.17; envelope-from=qemu-devel-bounces+importer=patchew.org@nongnu.org; helo=lists.gnu.org; Received-SPF: pass client-ip=170.10.133.124; envelope-from=pbonzini@redhat.com; helo=us-smtp-delivery-124.mimecast.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, DKIMWL_WL_HIGH=-0.001, 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, T_SCC_BODY_TEXT_LINE=-0.01 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+importer=patchew.org@nongnu.org Sender: "Qemu-devel" X-ZohoMail-DKIM: pass (identity @redhat.com) X-ZM-MESSAGEID: 1662058040165100001 Content-Type: text/plain; charset="utf-8" From: Paul Brook Add a flags field each row in sse_op_table6 and sse_op_table7. Initially this is only used as a replacement for the magic SSE41_SPECIAL pointer. The other flags are mostly relevant for the AVX implementation but can be applied to SSE as well. Signed-off-by: Paul Brook Message-Id: <20220424220204.2493824-6-paul@nowt.org> Reviewed-by: Richard Henderson Signed-off-by: Paolo Bonzini --- target/i386/tcg/translate.c | 230 ++++++++++++++++++++---------------- 1 file changed, 131 insertions(+), 99 deletions(-) diff --git a/target/i386/tcg/translate.c b/target/i386/tcg/translate.c index 7332bbcf44..b7321b7588 100644 --- a/target/i386/tcg/translate.c +++ b/target/i386/tcg/translate.c @@ -2976,7 +2976,6 @@ static const struct SSEOpHelper_table1 sse_op_table1[= 256] =3D { #undef SSE_SPECIAL =20 #define MMX_OP2(x) { gen_helper_ ## x ## _mmx, gen_helper_ ## x ## _xmm } -#define SSE_SPECIAL_FN ((void *)1) =20 static const SSEFunc_0_epp sse_op_table2[3 * 8][2] =3D { [0 + 2] =3D MMX_OP2(psrlw), @@ -3060,113 +3059,134 @@ static const SSEFunc_0_epp sse_op_table5[256] =3D= { [0xbf] =3D gen_helper_pavgb_mmx /* pavgusb */ }; =20 -struct SSEOpHelper_epp { +struct SSEOpHelper_table6 { SSEFunc_0_epp op[2]; uint32_t ext_mask; + int flags; }; =20 -struct SSEOpHelper_eppi { +struct SSEOpHelper_table7 { SSEFunc_0_eppi op[2]; uint32_t ext_mask; + int flags; }; =20 -#define SSSE3_OP(x) { MMX_OP2(x), CPUID_EXT_SSSE3 } -#define SSE41_OP(x) { { NULL, gen_helper_ ## x ## _xmm }, CPUID_EXT_SSE41 } -#define SSE42_OP(x) { { NULL, gen_helper_ ## x ## _xmm }, CPUID_EXT_SSE42 } -#define SSE41_SPECIAL { { NULL, SSE_SPECIAL_FN }, CPUID_EXT_SSE41 } -#define PCLMULQDQ_OP(x) { { NULL, gen_helper_ ## x ## _xmm }, \ - CPUID_EXT_PCLMULQDQ } -#define AESNI_OP(x) { { NULL, gen_helper_ ## x ## _xmm }, CPUID_EXT_AES } +#define gen_helper_special_xmm NULL =20 -static const struct SSEOpHelper_epp sse_op_table6[256] =3D { - [0x00] =3D SSSE3_OP(pshufb), - [0x01] =3D SSSE3_OP(phaddw), - [0x02] =3D SSSE3_OP(phaddd), - [0x03] =3D SSSE3_OP(phaddsw), - [0x04] =3D SSSE3_OP(pmaddubsw), - [0x05] =3D SSSE3_OP(phsubw), - [0x06] =3D SSSE3_OP(phsubd), - [0x07] =3D SSSE3_OP(phsubsw), - [0x08] =3D SSSE3_OP(psignb), - [0x09] =3D SSSE3_OP(psignw), - [0x0a] =3D SSSE3_OP(psignd), - [0x0b] =3D SSSE3_OP(pmulhrsw), - [0x10] =3D SSE41_OP(pblendvb), - [0x14] =3D SSE41_OP(blendvps), - [0x15] =3D SSE41_OP(blendvpd), - [0x17] =3D SSE41_OP(ptest), - [0x1c] =3D SSSE3_OP(pabsb), - [0x1d] =3D SSSE3_OP(pabsw), - [0x1e] =3D SSSE3_OP(pabsd), - [0x20] =3D SSE41_OP(pmovsxbw), - [0x21] =3D SSE41_OP(pmovsxbd), - [0x22] =3D SSE41_OP(pmovsxbq), - [0x23] =3D SSE41_OP(pmovsxwd), - [0x24] =3D SSE41_OP(pmovsxwq), - [0x25] =3D SSE41_OP(pmovsxdq), - [0x28] =3D SSE41_OP(pmuldq), - [0x29] =3D SSE41_OP(pcmpeqq), - [0x2a] =3D SSE41_SPECIAL, /* movntqda */ - [0x2b] =3D SSE41_OP(packusdw), - [0x30] =3D SSE41_OP(pmovzxbw), - [0x31] =3D SSE41_OP(pmovzxbd), - [0x32] =3D SSE41_OP(pmovzxbq), - [0x33] =3D SSE41_OP(pmovzxwd), - [0x34] =3D SSE41_OP(pmovzxwq), - [0x35] =3D SSE41_OP(pmovzxdq), - [0x37] =3D SSE42_OP(pcmpgtq), - [0x38] =3D SSE41_OP(pminsb), - [0x39] =3D SSE41_OP(pminsd), - [0x3a] =3D SSE41_OP(pminuw), - [0x3b] =3D SSE41_OP(pminud), - [0x3c] =3D SSE41_OP(pmaxsb), - [0x3d] =3D SSE41_OP(pmaxsd), - [0x3e] =3D SSE41_OP(pmaxuw), - [0x3f] =3D SSE41_OP(pmaxud), - [0x40] =3D SSE41_OP(pmulld), - [0x41] =3D SSE41_OP(phminposuw), - [0xdb] =3D AESNI_OP(aesimc), - [0xdc] =3D AESNI_OP(aesenc), - [0xdd] =3D AESNI_OP(aesenclast), - [0xde] =3D AESNI_OP(aesdec), - [0xdf] =3D AESNI_OP(aesdeclast), +#define OP(name, op, flags, ext, mmx_name) \ + {{mmx_name, gen_helper_ ## name ## _xmm}, CPUID_EXT_ ## ext, flags} +#define BINARY_OP_MMX(name, ext) \ + OP(name, op1, SSE_OPF_MMX, ext, gen_helper_ ## name ## _mmx) +#define BINARY_OP(name, ext, flags) \ + OP(name, op1, flags, ext, NULL) +#define UNARY_OP_MMX(name, ext) \ + OP(name, op1, SSE_OPF_MMX, ext, gen_helper_ ## name ## _mmx) +#define UNARY_OP(name, ext, flags) \ + OP(name, op1, flags, ext, NULL) +#define BLENDV_OP(name, ext, flags) OP(name, op1, 0, ext, NULL) +#define CMP_OP(name, ext) OP(name, op1, SSE_OPF_CMP, ext, NULL) +#define SPECIAL_OP(ext) OP(special, op1, SSE_OPF_SPECIAL, ext, NULL) + +/* prefix [66] 0f 38 */ +static const struct SSEOpHelper_table6 sse_op_table6[256] =3D { + [0x00] =3D BINARY_OP_MMX(pshufb, SSSE3), + [0x01] =3D BINARY_OP_MMX(phaddw, SSSE3), + [0x02] =3D BINARY_OP_MMX(phaddd, SSSE3), + [0x03] =3D BINARY_OP_MMX(phaddsw, SSSE3), + [0x04] =3D BINARY_OP_MMX(pmaddubsw, SSSE3), + [0x05] =3D BINARY_OP_MMX(phsubw, SSSE3), + [0x06] =3D BINARY_OP_MMX(phsubd, SSSE3), + [0x07] =3D BINARY_OP_MMX(phsubsw, SSSE3), + [0x08] =3D BINARY_OP_MMX(psignb, SSSE3), + [0x09] =3D BINARY_OP_MMX(psignw, SSSE3), + [0x0a] =3D BINARY_OP_MMX(psignd, SSSE3), + [0x0b] =3D BINARY_OP_MMX(pmulhrsw, SSSE3), + [0x10] =3D BLENDV_OP(pblendvb, SSE41, SSE_OPF_MMX), + [0x14] =3D BLENDV_OP(blendvps, SSE41, 0), + [0x15] =3D BLENDV_OP(blendvpd, SSE41, 0), + [0x17] =3D CMP_OP(ptest, SSE41), + [0x1c] =3D UNARY_OP_MMX(pabsb, SSSE3), + [0x1d] =3D UNARY_OP_MMX(pabsw, SSSE3), + [0x1e] =3D UNARY_OP_MMX(pabsd, SSSE3), + [0x20] =3D UNARY_OP(pmovsxbw, SSE41, SSE_OPF_MMX), + [0x21] =3D UNARY_OP(pmovsxbd, SSE41, SSE_OPF_MMX), + [0x22] =3D UNARY_OP(pmovsxbq, SSE41, SSE_OPF_MMX), + [0x23] =3D UNARY_OP(pmovsxwd, SSE41, SSE_OPF_MMX), + [0x24] =3D UNARY_OP(pmovsxwq, SSE41, SSE_OPF_MMX), + [0x25] =3D UNARY_OP(pmovsxdq, SSE41, SSE_OPF_MMX), + [0x28] =3D BINARY_OP(pmuldq, SSE41, SSE_OPF_MMX), + [0x29] =3D BINARY_OP(pcmpeqq, SSE41, SSE_OPF_MMX), + [0x2a] =3D SPECIAL_OP(SSE41), /* movntqda */ + [0x2b] =3D BINARY_OP(packusdw, SSE41, SSE_OPF_MMX), + [0x30] =3D UNARY_OP(pmovzxbw, SSE41, SSE_OPF_MMX), + [0x31] =3D UNARY_OP(pmovzxbd, SSE41, SSE_OPF_MMX), + [0x32] =3D UNARY_OP(pmovzxbq, SSE41, SSE_OPF_MMX), + [0x33] =3D UNARY_OP(pmovzxwd, SSE41, SSE_OPF_MMX), + [0x34] =3D UNARY_OP(pmovzxwq, SSE41, SSE_OPF_MMX), + [0x35] =3D UNARY_OP(pmovzxdq, SSE41, SSE_OPF_MMX), + [0x37] =3D BINARY_OP(pcmpgtq, SSE41, SSE_OPF_MMX), + [0x38] =3D BINARY_OP(pminsb, SSE41, SSE_OPF_MMX), + [0x39] =3D BINARY_OP(pminsd, SSE41, SSE_OPF_MMX), + [0x3a] =3D BINARY_OP(pminuw, SSE41, SSE_OPF_MMX), + [0x3b] =3D BINARY_OP(pminud, SSE41, SSE_OPF_MMX), + [0x3c] =3D BINARY_OP(pmaxsb, SSE41, SSE_OPF_MMX), + [0x3d] =3D BINARY_OP(pmaxsd, SSE41, SSE_OPF_MMX), + [0x3e] =3D BINARY_OP(pmaxuw, SSE41, SSE_OPF_MMX), + [0x3f] =3D BINARY_OP(pmaxud, SSE41, SSE_OPF_MMX), + [0x40] =3D BINARY_OP(pmulld, SSE41, SSE_OPF_MMX), + [0x41] =3D UNARY_OP(phminposuw, SSE41, 0), + [0xdb] =3D UNARY_OP(aesimc, AES, 0), + [0xdc] =3D BINARY_OP(aesenc, AES, 0), + [0xdd] =3D BINARY_OP(aesenclast, AES, 0), + [0xde] =3D BINARY_OP(aesdec, AES, 0), + [0xdf] =3D BINARY_OP(aesdeclast, AES, 0), }; =20 -static const struct SSEOpHelper_eppi sse_op_table7[256] =3D { - [0x08] =3D SSE41_OP(roundps), - [0x09] =3D SSE41_OP(roundpd), - [0x0a] =3D SSE41_OP(roundss), - [0x0b] =3D SSE41_OP(roundsd), - [0x0c] =3D SSE41_OP(blendps), - [0x0d] =3D SSE41_OP(blendpd), - [0x0e] =3D SSE41_OP(pblendw), - [0x0f] =3D SSSE3_OP(palignr), - [0x14] =3D SSE41_SPECIAL, /* pextrb */ - [0x15] =3D SSE41_SPECIAL, /* pextrw */ - [0x16] =3D SSE41_SPECIAL, /* pextrd/pextrq */ - [0x17] =3D SSE41_SPECIAL, /* extractps */ - [0x20] =3D SSE41_SPECIAL, /* pinsrb */ - [0x21] =3D SSE41_SPECIAL, /* insertps */ - [0x22] =3D SSE41_SPECIAL, /* pinsrd/pinsrq */ - [0x40] =3D SSE41_OP(dpps), - [0x41] =3D SSE41_OP(dppd), - [0x42] =3D SSE41_OP(mpsadbw), - [0x44] =3D PCLMULQDQ_OP(pclmulqdq), - [0x60] =3D SSE42_OP(pcmpestrm), - [0x61] =3D SSE42_OP(pcmpestri), - [0x62] =3D SSE42_OP(pcmpistrm), - [0x63] =3D SSE42_OP(pcmpistri), - [0xdf] =3D AESNI_OP(aeskeygenassist), +/* prefix [66] 0f 3a */ +static const struct SSEOpHelper_table7 sse_op_table7[256] =3D { + [0x08] =3D UNARY_OP(roundps, SSE41, 0), + [0x09] =3D UNARY_OP(roundpd, SSE41, 0), + [0x0a] =3D UNARY_OP(roundss, SSE41, SSE_OPF_SCALAR), + [0x0b] =3D UNARY_OP(roundsd, SSE41, SSE_OPF_SCALAR), + [0x0c] =3D BINARY_OP(blendps, SSE41, 0), + [0x0d] =3D BINARY_OP(blendpd, SSE41, 0), + [0x0e] =3D BINARY_OP(pblendw, SSE41, SSE_OPF_MMX), + [0x0f] =3D BINARY_OP_MMX(palignr, SSSE3), + [0x14] =3D SPECIAL_OP(SSE41), /* pextrb */ + [0x15] =3D SPECIAL_OP(SSE41), /* pextrw */ + [0x16] =3D SPECIAL_OP(SSE41), /* pextrd/pextrq */ + [0x17] =3D SPECIAL_OP(SSE41), /* extractps */ + [0x20] =3D SPECIAL_OP(SSE41), /* pinsrb */ + [0x21] =3D SPECIAL_OP(SSE41), /* insertps */ + [0x22] =3D SPECIAL_OP(SSE41), /* pinsrd/pinsrq */ + [0x40] =3D BINARY_OP(dpps, SSE41, 0), + [0x41] =3D BINARY_OP(dppd, SSE41, 0), + [0x42] =3D BINARY_OP(mpsadbw, SSE41, SSE_OPF_MMX), + [0x44] =3D BINARY_OP(pclmulqdq, PCLMULQDQ, 0), + [0x60] =3D CMP_OP(pcmpestrm, SSE42), + [0x61] =3D CMP_OP(pcmpestri, SSE42), + [0x62] =3D CMP_OP(pcmpistrm, SSE42), + [0x63] =3D CMP_OP(pcmpistri, SSE42), + [0xdf] =3D UNARY_OP(aeskeygenassist, AES, 0), }; =20 +#undef OP +#undef BINARY_OP_MMX +#undef BINARY_OP +#undef UNARY_OP_MMX +#undef UNARY_OP +#undef BLENDV_OP +#undef SPECIAL_OP + static void gen_sse(CPUX86State *env, DisasContext *s, int b, target_ulong pc_start) { int b1, op1_offset, op2_offset, is_xmm, val; int modrm, mod, rm, reg; int sse_op_flags; + const struct SSEOpHelper_table6 *op6; + const struct SSEOpHelper_table7 *op7; SSEFunc_0_epp sse_fn_epp; - SSEFunc_0_eppi sse_fn_eppi; SSEFunc_0_ppi sse_fn_ppi; SSEFunc_0_eppt sse_fn_eppt; MemOp ot; @@ -3821,12 +3841,13 @@ static void gen_sse(CPUX86State *env, DisasContext = *s, int b, mod =3D (modrm >> 6) & 3; =20 assert(b1 < 2); - sse_fn_epp =3D sse_op_table6[b].op[b1]; - if (!sse_fn_epp) { + op6 =3D &sse_op_table6[b]; + if (op6->ext_mask =3D=3D 0) { goto unknown_op; } - if (!(s->cpuid_ext_features & sse_op_table6[b].ext_mask)) + if (!(s->cpuid_ext_features & op6->ext_mask)) { goto illegal_op; + } =20 if (b1) { op1_offset =3D ZMM_OFFSET(reg); @@ -3863,6 +3884,9 @@ static void gen_sse(CPUX86State *env, DisasContext *s= , int b, } } } else { + if ((op6->flags & SSE_OPF_MMX) =3D=3D 0) { + goto unknown_op; + } op1_offset =3D offsetof(CPUX86State,fpregs[reg].mmx); if (mod =3D=3D 3) { op2_offset =3D offsetof(CPUX86State,fpregs[rm].mmx); @@ -3872,13 +3896,13 @@ static void gen_sse(CPUX86State *env, DisasContext = *s, int b, gen_ldq_env_A0(s, op2_offset); } } - if (sse_fn_epp =3D=3D SSE_SPECIAL_FN) { - goto unknown_op; + if (!op6->op[b1]) { + goto illegal_op; } =20 tcg_gen_addi_ptr(s->ptr0, cpu_env, op1_offset); tcg_gen_addi_ptr(s->ptr1, cpu_env, op2_offset); - sse_fn_epp(cpu_env, s->ptr0, s->ptr1); + op6->op[b1](cpu_env, s->ptr0, s->ptr1); =20 if (b =3D=3D 0x17) { set_cc_op(s, CC_OP_EFLAGS); @@ -4249,16 +4273,21 @@ static void gen_sse(CPUX86State *env, DisasContext = *s, int b, mod =3D (modrm >> 6) & 3; =20 assert(b1 < 2); - sse_fn_eppi =3D sse_op_table7[b].op[b1]; - if (!sse_fn_eppi) { + op7 =3D &sse_op_table7[b]; + if (op7->ext_mask =3D=3D 0) { goto unknown_op; } - if (!(s->cpuid_ext_features & sse_op_table7[b].ext_mask)) + if (!(s->cpuid_ext_features & op7->ext_mask)) { goto illegal_op; + } =20 s->rip_offset =3D 1; =20 - if (sse_fn_eppi =3D=3D SSE_SPECIAL_FN) { + if (op7->flags & SSE_OPF_SPECIAL) { + /* None of the "special" ops are valid on mmx registers */ + if (b1 =3D=3D 0) { + goto illegal_op; + } ot =3D mo_64_32(s->dflag); rm =3D (modrm & 7) | REX_B(s); if (mod !=3D 3) @@ -4403,6 +4432,9 @@ static void gen_sse(CPUX86State *env, DisasContext *s= , int b, gen_ldo_env_A0(s, op2_offset); } } else { + if ((op7->flags & SSE_OPF_MMX) =3D=3D 0) { + goto illegal_op; + } op1_offset =3D offsetof(CPUX86State,fpregs[reg].mmx); if (mod =3D=3D 3) { op2_offset =3D offsetof(CPUX86State,fpregs[rm].mmx); @@ -4425,7 +4457,7 @@ static void gen_sse(CPUX86State *env, DisasContext *s= , int b, =20 tcg_gen_addi_ptr(s->ptr0, cpu_env, op1_offset); tcg_gen_addi_ptr(s->ptr1, cpu_env, op2_offset); - sse_fn_eppi(cpu_env, s->ptr0, s->ptr1, tcg_const_i32(val)); + op7->op[b1](cpu_env, s->ptr0, s->ptr1, tcg_const_i32(val)); break; =20 case 0x33a: --=20 2.37.2