Merge branch 'dev'
diff --git a/CMakeLists.txt b/CMakeLists.txt
index 238664f..26609ff 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -1,6 +1,6 @@
cmake_minimum_required(VERSION 2.6...3.0.2)
-project(xbyak LANGUAGES CXX VERSION 6.63)
+project(xbyak LANGUAGES CXX VERSION 6.64)
file(GLOB headers xbyak/*.h)
diff --git a/doc/changelog.md b/doc/changelog.md
index 484b6bd..67630f8 100644
--- a/doc/changelog.md
+++ b/doc/changelog.md
@@ -1,7 +1,8 @@
# History
-* 2022/Oct/06 vpmadd52{h,l}uq support AVX-IFMA
-* 2022/Oct/05 support amx_fp16/avx_vnni_int8/avx_ne_convert and add setDefaultEncoding()
+* 2022/Nov/04 ver 6.64 some vmov* support addressing with mask
+* 2022/Oct/06 ver 6.63 vpmadd52{h,l}uq support AVX-IFMA
+* 2022/Oct/05 ver 6.63 support amx_fp16/avx_vnni_int8/avx_ne_convert and add setDefaultEncoding()
* 2022/Aug/15 ver 6.62 add serialize instruction
* 2022/Aug/02 ver 6.61.1 noexcept is supported by Visual Studio 2015 or later
* 2022/Jul/29 ver 6.61 fix exception of movzx eax, ah in 64-bit mode
diff --git a/gen/gen_avx512.cpp b/gen/gen_avx512.cpp
index 41b28e9..8283a54 100644
--- a/gen/gen_avx512.cpp
+++ b/gen/gen_avx512.cpp
@@ -692,29 +692,29 @@
int type;
int mode;
} tbl[] = {
- { 0x32, "vpmovqb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N2 | T_N_VL, false },
- { 0x22, "vpmovsqb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N2 | T_N_VL, false },
- { 0x12, "vpmovusqb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N2 | T_N_VL, false },
+ { 0x32, "vpmovqb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N2 | T_N_VL | T_M_K, false },
+ { 0x22, "vpmovsqb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N2 | T_N_VL | T_M_K, false },
+ { 0x12, "vpmovusqb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N2 | T_N_VL | T_M_K, false },
- { 0x34, "vpmovqw", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N4 | T_N_VL, false },
- { 0x24, "vpmovsqw", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N4 | T_N_VL, false },
- { 0x14, "vpmovusqw", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N4 | T_N_VL, false },
+ { 0x34, "vpmovqw", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N4 | T_N_VL | T_M_K, false },
+ { 0x24, "vpmovsqw", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N4 | T_N_VL | T_M_K, false },
+ { 0x14, "vpmovusqw", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N4 | T_N_VL | T_M_K, false },
- { 0x35, "vpmovqd", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL, true },
- { 0x25, "vpmovsqd", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL, true },
- { 0x15, "vpmovusqd", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL, true },
+ { 0x35, "vpmovqd", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL | T_M_K, true },
+ { 0x25, "vpmovsqd", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL | T_M_K, true },
+ { 0x15, "vpmovusqd", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL | T_M_K, true },
- { 0x31, "vpmovdb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N4 | T_N_VL, false },
- { 0x21, "vpmovsdb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N4 | T_N_VL, false },
- { 0x11, "vpmovusdb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N4 | T_N_VL, false },
+ { 0x31, "vpmovdb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N4 | T_N_VL | T_M_K, false },
+ { 0x21, "vpmovsdb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N4 | T_N_VL | T_M_K, false },
+ { 0x11, "vpmovusdb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N4 | T_N_VL | T_M_K, false },
- { 0x33, "vpmovdw", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL, true },
- { 0x23, "vpmovsdw", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL, true },
- { 0x13, "vpmovusdw", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL, true },
+ { 0x33, "vpmovdw", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL | T_M_K, true },
+ { 0x23, "vpmovsdw", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL | T_M_K, true },
+ { 0x13, "vpmovusdw", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL | T_M_K, true },
- { 0x30, "vpmovwb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL, true },
- { 0x20, "vpmovswb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL, true },
- { 0x10, "vpmovuswb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL, true },
+ { 0x30, "vpmovwb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL | T_M_K, true },
+ { 0x20, "vpmovswb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL | T_M_K, true },
+ { 0x10, "vpmovuswb", T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_N8 | T_N_VL | T_M_K, true },
};
for (size_t i = 0; i < NUM_OF_ARRAY(tbl); i++) {
const Tbl& p = tbl[i];
diff --git a/meson.build b/meson.build
index cb06c03..d9e89ac 100644
--- a/meson.build
+++ b/meson.build
@@ -5,7 +5,7 @@
project(
'xbyak',
'cpp',
- version: '6.63',
+ version: '6.64',
license: 'BSD-3-Clause',
default_options: 'b_ndebug=if-release'
)
diff --git a/readme.md b/readme.md
index 20f08e8..8f72853 100644
--- a/readme.md
+++ b/readme.md
@@ -1,5 +1,5 @@
-# Xbyak 6.63 [![Badge Build]][Build Status]
+# Xbyak 6.64 [![Badge Build]][Build Status]
*A C++ JIT assembler for x86 (IA32), x64 (AMD64, x86-64)*
diff --git a/readme.txt b/readme.txt
index bf7875e..2306bf0 100644
--- a/readme.txt
+++ b/readme.txt
@@ -1,5 +1,5 @@
- C++用x86(IA-32), x64(AMD64, x86-64) JITアセンブラ Xbyak 6.63
+ C++用x86(IA-32), x64(AMD64, x86-64) JITアセンブラ Xbyak 6.64
-----------------------------------------------------------------------------
◎概要
@@ -402,6 +402,7 @@
-----------------------------------------------------------------------------
◎履歴
+2022/11/04 ver 6.64 vmov*命令をmaskつきアドレッシング対応修正
2022/10/06 ver 6.63 AVX-IFMA用のvpmadd52{h,l}uq対応
2022/10/05 amx_fp16/avx_vnni_int8/avx_ne_convertt対応とsetDefaultEncoding()追加
2022/09/15 ver 6.62 serialize追加
diff --git a/test/Makefile.win b/test/Makefile.win
index 4025ae2..96105b3 100644
--- a/test/Makefile.win
+++ b/test/Makefile.win
@@ -1,4 +1,4 @@
-OPT=/EHsc -I../xbyak /W4 -D_CRT_SECURE_NO_WARNINGS
+OPT=/EHsc -I../xbyak /W4 -D_CRT_SECURE_NO_WARNINGS -I ../
../xbyak/xbyak_mnemonic.h: ../gen/gen_code.exe ../gen/gen_avx512.exe
../gen/gen_code.exe > $@
../gen/gen_avx512.exe >> $@
diff --git a/test/make_512.cpp b/test/make_512.cpp
index 83994ab..39bfa99 100644
--- a/test/make_512.cpp
+++ b/test/make_512.cpp
@@ -1807,44 +1807,44 @@
put("vpmovd2m", K, _XMM | _YMM | _ZMM);
put("vpmovq2m", K, _XMM | _YMM | _ZMM);
- put("vpmovqb", XMM_KZ | _MEM, _XMM | _YMM | _ZMM);
- put("vpmovsqb", XMM_KZ | _MEM, _XMM | _YMM | _ZMM);
- put("vpmovusqb", XMM_KZ | _MEM, _XMM | _YMM | _ZMM);
+ put("vpmovqb", XMM_KZ | _MEM | MEM_K, _XMM | _YMM | _ZMM);
+ put("vpmovsqb", XMM_KZ | _MEM | MEM_K, _XMM | _YMM | _ZMM);
+ put("vpmovusqb", XMM_KZ | _MEM | MEM_K, _XMM | _YMM | _ZMM);
- put("vpmovqw", XMM_KZ | _MEM, _XMM | _YMM | _ZMM);
- put("vpmovsqw", XMM_KZ | _MEM, _XMM | _YMM | _ZMM);
- put("vpmovusqw", XMM_KZ | _MEM, _XMM | _YMM | _ZMM);
+ put("vpmovqw", XMM_KZ | _MEM | MEM_K, _XMM | _YMM | _ZMM);
+ put("vpmovsqw", XMM_KZ | _MEM | MEM_K, _XMM | _YMM | _ZMM);
+ put("vpmovusqw", XMM_KZ | _MEM | MEM_K, _XMM | _YMM | _ZMM);
- put("vpmovqd", XMM_KZ | _MEM, _XMM | _YMM);
- put("vpmovqd", YMM_KZ | _MEM, _ZMM);
+ put("vpmovqd", XMM_KZ | _MEM | MEM_K, _XMM | _YMM);
+ put("vpmovqd", YMM_KZ | _MEM | MEM_K, _ZMM);
- put("vpmovsqd", XMM_KZ | _MEM, _XMM | _YMM);
- put("vpmovsqd", YMM_KZ | _MEM, _ZMM);
+ put("vpmovsqd", XMM_KZ | _MEM | MEM_K, _XMM | _YMM);
+ put("vpmovsqd", YMM_KZ | _MEM | MEM_K, _ZMM);
- put("vpmovusqd", XMM_KZ | _MEM, _XMM | _YMM);
- put("vpmovusqd", YMM_KZ | _MEM, _ZMM);
+ put("vpmovusqd", XMM_KZ | _MEM | MEM_K, _XMM | _YMM);
+ put("vpmovusqd", YMM_KZ | _MEM | MEM_K, _ZMM);
- put("vpmovdb", XMM_KZ | _MEM, _XMM | _YMM | _ZMM);
- put("vpmovsdb", XMM_KZ | _MEM, _XMM | _YMM | _ZMM);
- put("vpmovusdb", XMM_KZ | _MEM, _XMM | _YMM | _ZMM);
+ put("vpmovdb", XMM_KZ | _MEM | MEM_K, _XMM | _YMM | _ZMM);
+ put("vpmovsdb", XMM_KZ | _MEM | MEM_K, _XMM | _YMM | _ZMM);
+ put("vpmovusdb", XMM_KZ | _MEM | MEM_K, _XMM | _YMM | _ZMM);
- put("vpmovdw", XMM_KZ | _MEM, _XMM | _YMM);
- put("vpmovdw", YMM_KZ | _MEM, _ZMM);
+ put("vpmovdw", XMM_KZ | _MEM | MEM_K, _XMM | _YMM);
+ put("vpmovdw", YMM_KZ | _MEM | MEM_K, _ZMM);
- put("vpmovsdw", XMM_KZ | _MEM, _XMM | _YMM);
- put("vpmovsdw", YMM_KZ | _MEM, _ZMM);
+ put("vpmovsdw", XMM_KZ | _MEM | MEM_K, _XMM | _YMM);
+ put("vpmovsdw", YMM_KZ | _MEM | MEM_K, _ZMM);
- put("vpmovusdw", XMM_KZ | _MEM, _XMM | _YMM);
- put("vpmovusdw", YMM_KZ | _MEM, _ZMM);
+ put("vpmovusdw", XMM_KZ | _MEM | MEM_K, _XMM | _YMM);
+ put("vpmovusdw", YMM_KZ | _MEM | MEM_K, _ZMM);
- put("vpmovwb", XMM_KZ | _MEM, _XMM | _YMM);
- put("vpmovwb", YMM_KZ | _MEM, _ZMM);
+ put("vpmovwb", XMM_KZ | _MEM | MEM_K, _XMM | _YMM);
+ put("vpmovwb", YMM_KZ | _MEM | MEM_K, _ZMM);
- put("vpmovswb", XMM_KZ | _MEM, _XMM | _YMM);
- put("vpmovswb", YMM_KZ | _MEM, _ZMM);
+ put("vpmovswb", XMM_KZ | _MEM | MEM_K, _XMM | _YMM);
+ put("vpmovswb", YMM_KZ | _MEM | MEM_K, _ZMM);
- put("vpmovuswb", XMM_KZ | _MEM, _XMM | _YMM);
- put("vpmovuswb", YMM_KZ | _MEM, _ZMM);
+ put("vpmovuswb", XMM_KZ | _MEM | MEM_K, _XMM | _YMM);
+ put("vpmovuswb", YMM_KZ | _MEM | MEM_K, _ZMM);
}
void putRot()
{
diff --git a/xbyak/xbyak.h b/xbyak/xbyak.h
index 16a0ed8..99c4ec4 100644
--- a/xbyak/xbyak.h
+++ b/xbyak/xbyak.h
@@ -155,7 +155,7 @@
enum {
DEFAULT_MAX_CODE_SIZE = 4096,
- VERSION = 0x6630 /* 0xABCD = A.BC(.D) */
+ VERSION = 0x6640 /* 0xABCD = A.BC(.D) */
};
#ifndef MIE_INTEGER_TYPE_DEFINED
@@ -1485,7 +1485,6 @@
clabelDefList_.clear();
clabelUndefList_.clear();
resetLabelPtrList();
- ClearError();
}
void enterLocal()
{
@@ -2804,6 +2803,7 @@
}
void reset()
{
+ ClearError();
resetSize();
labelMgr_.reset();
labelMgr_.set(this);
diff --git a/xbyak/xbyak_mnemonic.h b/xbyak/xbyak_mnemonic.h
index bf2ab70..e626915 100644
--- a/xbyak/xbyak_mnemonic.h
+++ b/xbyak/xbyak_mnemonic.h
@@ -1,4 +1,4 @@
-const char *getVersionString() const { return "6.63"; }
+const char *getVersionString() const { return "6.64"; }
void adc(const Operand& op, uint32_t imm) { opRM_I(op, imm, 0x10, 2); }
void adc(const Operand& op1, const Operand& op2) { opRM_RM(op1, op2, 0x10); }
void adcx(const Reg32e& reg, const Operand& op) { opGen(reg, op, 0xF6, 0x66, isREG32_REG32orMEM, NONE, 0x38); }
@@ -2163,30 +2163,30 @@
void vpminuq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x3B); }
void vpmovb2m(const Opmask& k, const Xmm& x) { opVex(k, 0, x, T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0, 0x29); }
void vpmovd2m(const Opmask& k, const Xmm& x) { opVex(k, 0, x, T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0, 0x39); }
-void vpmovdb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N4 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x31, false); }
-void vpmovdw(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x33, true); }
+void vpmovdb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N4 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x31, false); }
+void vpmovdw(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x33, true); }
void vpmovm2b(const Xmm& x, const Opmask& k) { opVex(x, 0, k, T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0, 0x28); }
void vpmovm2d(const Xmm& x, const Opmask& k) { opVex(x, 0, k, T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0, 0x38); }
void vpmovm2q(const Xmm& x, const Opmask& k) { opVex(x, 0, k, T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW1, 0x38); }
void vpmovm2w(const Xmm& x, const Opmask& k) { opVex(x, 0, k, T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW1, 0x28); }
void vpmovq2m(const Opmask& k, const Xmm& x) { opVex(k, 0, x, T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW1, 0x39); }
-void vpmovqb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N2 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x32, false); }
-void vpmovqd(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x35, true); }
-void vpmovqw(const Operand& op, const Xmm& x) { opVmov(op, x, T_N4 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x34, false); }
-void vpmovsdb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N4 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x21, false); }
-void vpmovsdw(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x23, true); }
-void vpmovsqb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N2 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x22, false); }
-void vpmovsqd(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x25, true); }
-void vpmovsqw(const Operand& op, const Xmm& x) { opVmov(op, x, T_N4 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x24, false); }
-void vpmovswb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x20, true); }
-void vpmovusdb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N4 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x11, false); }
-void vpmovusdw(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x13, true); }
-void vpmovusqb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N2 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x12, false); }
-void vpmovusqd(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x15, true); }
-void vpmovusqw(const Operand& op, const Xmm& x) { opVmov(op, x, T_N4 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x14, false); }
-void vpmovuswb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x10, true); }
+void vpmovqb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N2 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x32, false); }
+void vpmovqd(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x35, true); }
+void vpmovqw(const Operand& op, const Xmm& x) { opVmov(op, x, T_N4 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x34, false); }
+void vpmovsdb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N4 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x21, false); }
+void vpmovsdw(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x23, true); }
+void vpmovsqb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N2 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x22, false); }
+void vpmovsqd(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x25, true); }
+void vpmovsqw(const Operand& op, const Xmm& x) { opVmov(op, x, T_N4 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x24, false); }
+void vpmovswb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x20, true); }
+void vpmovusdb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N4 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x11, false); }
+void vpmovusdw(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x13, true); }
+void vpmovusqb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N2 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x12, false); }
+void vpmovusqd(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x15, true); }
+void vpmovusqw(const Operand& op, const Xmm& x) { opVmov(op, x, T_N4 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x14, false); }
+void vpmovuswb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x10, true); }
void vpmovw2m(const Opmask& k, const Xmm& x) { opVex(k, 0, x, T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW1, 0x29); }
-void vpmovwb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x30, true); }
+void vpmovwb(const Operand& op, const Xmm& x) { opVmov(op, x, T_N8 | T_N_VL | T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K, 0x30, true); }
void vpmullq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x40); }
void vpmultishiftqb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x83); }
void vpopcntb(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX, 0x54); }
