From: Paolo Bonzini <pbonzini@redhat.com>
To: qemu-devel@nongnu.org
Cc: richard.henderson@linaro.org, paul@nowt.org
Subject: [PATCH v2 20/37] target/i386: reimplement 0x0f 0xd8-0xdf, 0xe8-0xef, 0xf8-0xff, add AVX
Date: Tue, 20 Sep 2022 19:24:50 +0200 [thread overview]
Message-ID: <20220920172507.95568-21-pbonzini@redhat.com> (raw)
In-Reply-To: <20220920172507.95568-1-pbonzini@redhat.com>
These are more simple integer instructions present in both MMX and SSE/AVX,
with no holes that were later occupied by newer instructions.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
target/i386/tcg/decode-new.c.inc | 28 ++++++++++++++++++++++++++++
target/i386/tcg/emit.c.inc | 32 ++++++++++++++++++++++++++++++++
target/i386/tcg/translate.c | 4 +++-
3 files changed, 63 insertions(+), 1 deletion(-)
diff --git a/target/i386/tcg/decode-new.c.inc b/target/i386/tcg/decode-new.c.inc
index 6635c41a2f..d207a1f0c1 100644
--- a/target/i386/tcg/decode-new.c.inc
+++ b/target/i386/tcg/decode-new.c.inc
@@ -273,6 +273,34 @@ static const X86OpEntry opcodes_0F[256] = {
[0x6d] = X86_OP_ENTRY3(PUNPCKHQDQ, V,x, H,x, W,x, vex4 p_66 avx2_256),
[0x6e] = X86_OP_ENTRY3(MOVD_to, V,x, None,None, E,y, vex5 mmx p_00_66), /* wrong dest Vy on SDM! */
[0x6f] = X86_OP_GROUP0(0F6F),
+
+ /* Incorrectly missing from 2-17 */
+ [0xd8] = X86_OP_ENTRY3(PSUBUSB, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xd9] = X86_OP_ENTRY3(PSUBUSW, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xda] = X86_OP_ENTRY3(PMINUB, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xdb] = X86_OP_ENTRY3(PAND, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xdc] = X86_OP_ENTRY3(PADDUSB, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xdd] = X86_OP_ENTRY3(PADDUSW, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xde] = X86_OP_ENTRY3(PMAXUB, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xdf] = X86_OP_ENTRY3(PANDN, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+
+ [0xe8] = X86_OP_ENTRY3(PSUBSB, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xe9] = X86_OP_ENTRY3(PSUBSW, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xea] = X86_OP_ENTRY3(PMINSW, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xeb] = X86_OP_ENTRY3(POR, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xec] = X86_OP_ENTRY3(PADDSB, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xed] = X86_OP_ENTRY3(PADDSW, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xee] = X86_OP_ENTRY3(PMAXSW, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xef] = X86_OP_ENTRY3(PXOR, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+
+ [0xf8] = X86_OP_ENTRY3(PSUBB, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xf9] = X86_OP_ENTRY3(PSUBW, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xfa] = X86_OP_ENTRY3(PSUBD, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xfb] = X86_OP_ENTRY3(PSUBQ, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xfc] = X86_OP_ENTRY3(PADDB, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xfd] = X86_OP_ENTRY3(PADDW, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ [0xfe] = X86_OP_ENTRY3(PADDD, V,x, H,x, W,x, vex4 mmx avx2_256 p_00_66),
+ /* 0xff = UD0 */
};
static void do_decode_0F(DisasContext *s, CPUX86State *env, X86OpEntry *entry, uint8_t *b)
diff --git a/target/i386/tcg/emit.c.inc b/target/i386/tcg/emit.c.inc
index 60b80e9d30..f1f7397869 100644
--- a/target/i386/tcg/emit.c.inc
+++ b/target/i386/tcg/emit.c.inc
@@ -328,9 +328,31 @@ static void gen_##uname(DisasContext *s, CPUX86State *env, X86DecodedInsn *decod
decode->op[2].offset, vec_len, vec_len); \
}
+BINARY_INT_GVEC(PADDB, tcg_gen_gvec_add, MO_8)
+BINARY_INT_GVEC(PADDW, tcg_gen_gvec_add, MO_16)
+BINARY_INT_GVEC(PADDD, tcg_gen_gvec_add, MO_32)
+BINARY_INT_GVEC(PADDSB, tcg_gen_gvec_ssadd, MO_8)
+BINARY_INT_GVEC(PADDSW, tcg_gen_gvec_ssadd, MO_16)
+BINARY_INT_GVEC(PADDUSB, tcg_gen_gvec_usadd, MO_8)
+BINARY_INT_GVEC(PADDUSW, tcg_gen_gvec_usadd, MO_16)
+BINARY_INT_GVEC(PAND, tcg_gen_gvec_and, MO_64)
BINARY_INT_GVEC(PCMPGTB, tcg_gen_gvec_cmp, TCG_COND_GT, MO_8)
BINARY_INT_GVEC(PCMPGTW, tcg_gen_gvec_cmp, TCG_COND_GT, MO_16)
BINARY_INT_GVEC(PCMPGTD, tcg_gen_gvec_cmp, TCG_COND_GT, MO_32)
+BINARY_INT_GVEC(PMAXSW, tcg_gen_gvec_smax, MO_16)
+BINARY_INT_GVEC(PMAXUB, tcg_gen_gvec_umax, MO_8)
+BINARY_INT_GVEC(PMINSW, tcg_gen_gvec_smin, MO_16)
+BINARY_INT_GVEC(PMINUB, tcg_gen_gvec_umin, MO_8)
+BINARY_INT_GVEC(POR, tcg_gen_gvec_or, MO_64)
+BINARY_INT_GVEC(PSUBB, tcg_gen_gvec_sub, MO_8)
+BINARY_INT_GVEC(PSUBW, tcg_gen_gvec_sub, MO_16)
+BINARY_INT_GVEC(PSUBD, tcg_gen_gvec_sub, MO_32)
+BINARY_INT_GVEC(PSUBQ, tcg_gen_gvec_sub, MO_64)
+BINARY_INT_GVEC(PSUBSB, tcg_gen_gvec_sssub, MO_8)
+BINARY_INT_GVEC(PSUBSW, tcg_gen_gvec_sssub, MO_16)
+BINARY_INT_GVEC(PSUBUSB, tcg_gen_gvec_ussub, MO_8)
+BINARY_INT_GVEC(PSUBUSW, tcg_gen_gvec_ussub, MO_16)
+BINARY_INT_GVEC(PXOR, tcg_gen_gvec_xor, MO_64)
/*
@@ -609,6 +631,16 @@ static void gen_MULX(DisasContext *s, CPUX86State *env, X86DecodedInsn *decode)
}
+static void gen_PANDN(DisasContext *s, CPUX86State *env, X86DecodedInsn *decode)
+{
+ int vec_len = vector_len(s, decode);
+
+ /* Careful, operand order is reversed! */
+ tcg_gen_gvec_andc(MO_64,
+ decode->op[0].offset, decode->op[2].offset,
+ decode->op[1].offset, vec_len, vec_len);
+}
+
static void gen_PDEP(DisasContext *s, CPUX86State *env, X86DecodedInsn *decode)
{
MemOp ot = decode->op[1].ot;
diff --git a/target/i386/tcg/translate.c b/target/i386/tcg/translate.c
index 1ffbf2f4f8..fd565650a9 100644
--- a/target/i386/tcg/translate.c
+++ b/target/i386/tcg/translate.c
@@ -4694,7 +4694,9 @@ static target_ulong disas_insn(DisasContext *s, CPUState *cpu)
#ifndef CONFIG_USER_ONLY
use_new &= b <= limit;
#endif
- if (use_new && (b >= 0x160 && b <= 0x16f)) {
+ if (use_new &&
+ ((b >= 0x160 && b <= 0x16f) ||
+ (b >= 0x1d8 && b <= 0x1ff && (b & 8)))) {
disas_insn_new(s, cpu, b + 0x100);
return s->pc;
}
--
2.37.2
next prev parent reply other threads:[~2022-09-20 22:26 UTC|newest]
Thread overview: 61+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-20 17:24 [PATCH v2 00/37] target/i386: new decoder + AVX implementation Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 01/37] target/i386: Define XMMReg and access macros, align ZMM registers Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 02/37] target/i386: make ldo/sto operations consistent with ldq Paolo Bonzini
2022-09-24 15:00 ` Richard Henderson
2022-09-20 17:24 ` [PATCH v2 03/37] target/i386: REPZ and REPNZ are mutually exclusive Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 04/37] target/i386: introduce insn_get_addr Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 05/37] target/i386: add core of new i386 decoder Paolo Bonzini
2022-09-24 15:09 ` Richard Henderson
2022-09-20 17:24 ` [PATCH v2 06/37] target/i386: add ALU load/writeback core Paolo Bonzini
2022-09-24 15:13 ` Richard Henderson
2022-09-20 17:24 ` [PATCH v2 07/37] target/i386: add CPUID[EAX=7, ECX=0].ECX to DisasContext Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 08/37] target/i386: add CPUID feature checks to new decoder Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 09/37] target/i386: add AVX_EN hflag Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 10/37] target/i386: validate VEX prefixes via the instructions' exception classes Paolo Bonzini
2022-09-24 20:19 ` Richard Henderson
2022-09-20 17:24 ` [PATCH v2 11/37] target/i386: validate SSE prefixes directly in the decoding table Paolo Bonzini
2022-09-24 20:23 ` Richard Henderson
2022-09-20 17:24 ` [PATCH v2 12/37] target/i386: move scalar 0F 38 and 0F 3A instruction to new decoder Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 13/37] target/i386: Prepare ops_sse_header.h for 256 bit AVX Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 14/37] target/i386: extend helpers to support VEX.V 3- and 4- operand encodings Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 15/37] target/i386: support operand merging in binary scalar helpers Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 16/37] target/i386: provide 3-operand versions of unary " Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 17/37] target/i386: implement additional AVX comparison operators Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 18/37] target/i386: Introduce 256-bit vector helpers Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 19/37] target/i386: reimplement 0x0f 0x60-0x6f, add AVX Paolo Bonzini
2022-09-24 20:31 ` Richard Henderson
2022-09-20 17:24 ` Paolo Bonzini [this message]
2022-09-24 20:32 ` [PATCH v2 20/37] target/i386: reimplement 0x0f 0xd8-0xdf, 0xe8-0xef, 0xf8-0xff, " Richard Henderson
2022-09-20 17:24 ` [PATCH v2 21/37] target/i386: reimplement 0x0f 0x50-0x5f, " Paolo Bonzini
2022-09-24 20:37 ` Richard Henderson
2022-09-20 17:24 ` [PATCH v2 22/37] target/i386: reimplement 0x0f 0x78-0x7f, " Paolo Bonzini
2022-09-24 20:43 ` Richard Henderson
2022-09-26 7:24 ` Paolo Bonzini
2022-09-26 7:46 ` Richard Henderson
2022-09-20 17:24 ` [PATCH v2 23/37] target/i386: reimplement 0x0f 0x70-0x77, " Paolo Bonzini
2022-09-24 20:53 ` Richard Henderson
2022-09-20 17:24 ` [PATCH v2 24/37] target/i386: reimplement 0x0f 0xd0-0xd7, 0xe0-0xe7, 0xf0-0xf7, " Paolo Bonzini
2022-09-24 20:56 ` Richard Henderson
2022-09-26 7:56 ` Paolo Bonzini
2022-09-20 17:24 ` [PATCH v2 25/37] target/i386: clarify (un)signedness of immediates from 0F3Ah opcodes Paolo Bonzini
2022-09-24 20:57 ` Richard Henderson
2022-09-20 17:24 ` [PATCH v2 26/37] target/i386: reimplement 0x0f 0x3a, add AVX Paolo Bonzini
2022-09-24 21:02 ` Richard Henderson
2022-09-20 17:24 ` [PATCH v2 27/37] target/i386: Use tcg gvec ops for pmovmskb Paolo Bonzini
2022-09-24 21:08 ` Richard Henderson
2022-09-20 17:24 ` [PATCH v2 28/37] target/i386: reimplement 0x0f 0x38, add AVX Paolo Bonzini
2022-09-24 21:14 ` Richard Henderson
2022-09-20 17:24 ` [PATCH v2 29/37] target/i386: reimplement 0x0f 0xc2, 0xc4-0xc6, " Paolo Bonzini
2022-09-20 17:25 ` [PATCH v2 30/37] target/i386: reimplement 0x0f 0x10-0x17, " Paolo Bonzini
2022-09-24 21:16 ` Richard Henderson
2022-09-20 17:25 ` [PATCH v2 31/37] target/i386: reimplement 0x0f 0x28-0x2f, " Paolo Bonzini
2022-09-24 21:18 ` Richard Henderson
2022-09-20 17:25 ` [PATCH v2 32/37] target/i386: implement XSAVE and XRSTOR of AVX registers Paolo Bonzini
2022-09-24 21:19 ` Richard Henderson
2022-09-20 17:25 ` [PATCH v2 33/37] target/i386: implement VLDMXCSR/VSTMXCSR Paolo Bonzini
2022-09-24 21:20 ` Richard Henderson
2022-09-20 17:25 ` [PATCH v2 34/37] target/i386: Enable AVX cpuid bits when using TCG Paolo Bonzini
2022-09-20 17:25 ` [PATCH v2 35/37] tests/tcg: extend SSE tests to AVX Paolo Bonzini
2022-09-20 17:25 ` [PATCH v2 36/37] target/i386: move 3DNow to the new decoder Paolo Bonzini
2022-09-24 21:24 ` Richard Henderson
2022-09-20 17:25 ` [PATCH v2 37/37] target/i386: remove old SSE decoder Paolo Bonzini
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20220920172507.95568-21-pbonzini@redhat.com \
--to=pbonzini@redhat.com \
--cc=paul@nowt.org \
--cc=qemu-devel@nongnu.org \
--cc=richard.henderson@linaro.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).