mirrors/dynarmic
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
dynarmic/xbyak/xbyak_mnemonic.h

2332 lines
247 KiB
C
Raw Normal View History

Squashed 'externals/xbyak/' changes from 88f2f771f..a1ac3750f a1ac3750f Merge branch 'dev' 5f4ba971f v6.68 cac2c175f update doc 1b08a8248 add test of prefetchiti bef70d9b1 add prefetchit{0,1} f66785876 add detection of prefetchiti 124617ac9 Merge branch 'dev' 1bce8be17 Merge commit 'fbb18f6' into dev bebfe64a3 [test] X32 is optional and remove CFLAGS d8c332cef Merge branch 'Tachi107-small-changes' into dev fbb18f69d ci: use containers 8a974696c test: simplify makefile 1efe9fe7c Merge branch 'dev' bb70083e6 v6.67 436e452d7 update doc 47ff6ef42 I get an error after all on GitHub action 445c0dcec add test of CMPccXADD c9347907d add CMPccXADD dc792cc56 add detection of cmpccxadd 62be84cc8 fix detection of boost b5ac7b0f0 Merge branch 'dev' 05dd400e0 recover a removed line 94eff6246 v6.66 ec3fadeba update doc 00bfaaa7f add test of rao-int 0f2f1aaa6 support rio-int 6c047f480 detect rao-int f07c5c255 fix args of EmptyAllocator::alloc in test 04d3eb5f5 stop if a test script causes an error 1c1d2366f disable boost sample cfb1127c3 tweak 5fcbeb7c4 [sample] change the way of detection of boost bafc1ee60 CXX uses g++ as default value d8cabc6cb remove warning of blace bc73a0816 remove -fno-operator-names option 6989aea94 use CXX instead of g++ a7c5a1bd7 use English 1bfbd8c4d Fix incorrect format strings 0ecef5c28 Merge branch 'dev' 7556c20ba update doc a15709271 v6.65 3b83aab3e add detect_x32 to TARGET 8c64bbbc3 use gcc instead of dpkg for portability 5e9a9b96f test_avx512.sh runs on x32 8ae01b0c2 disable some tests on x32 83b3da217 x32 does not check large disp 693ab8c9d sizeof(void*) = 4 on x32, so disable the test 348e3e548 Merge branch 'dev' 11b9c4dc0 v6.64 459636196 add T_M_K flag to vpmov* 1d3722928 add include path 5e27eddae move CrearError from LabelManager::reset() to CodeGenerator::reset() f8ea5c28d Merge branch 'dev' 20b2b1eae v6.63 3706869f8 desc. of setDefaultEncoding d6f2d7577 add test of setDefaultEncoding 3b0a19c41 vpmadd52{h,l}uq for avx-ifma 95752ebd7 add tAVX_IFMA cd36e31ea [sample] show AMX_FP16/AVX_VNNI_INT8/AVX_NE_CONVERT e5858af27 add setDefaultEncoding 2f7fb0220 modify gen.cpp for AVX-NE-CONVERT/AVX-VNNI-INT8/AMX-FP16 1c5cb7efa add AVX-NE-CONVERT instructions 564fe9acd add AVX-VNNI-INT8 instructions cd14d07b1 add AMX-FP16 instruction 7811f593c Merge pull request #161 from scribam/patch-1 2218f6c08 Update changelog.md 7bccdbbb4 Merge pull request #160 from herumi/dev 5fcf87596 compile nasm-2.15 instead of apt install e31961ea8 v6.62 ff4f9e65c update doc 9c8fb81db disable wrong detection of gcc -Warray-bounds 6b7519659 add serialize e16582696 Merge branch 'JonLiu1993-vcpkg-installition' into dev edbb410fb Add vcpkg installation instructions 055d31242 Merge branch 'dev' 21ab98441 Merge branch 'akodanka-enable_CIV' into dev ed4d598e3 Changes to compile xbyak project for openvino b652430c4 mingw uses __cpuidex 48457bfa0 Merge branch 'dev' 29cb524d1 v6.61.2 1a9a0b0e1 avoid including algorithm header in xbyak_util.h 6fadefd04 Merge branch 'dev' fc1c18a9d update doc f7cae7f11 v6.61.1 6f5ec5cf3 Merge branch 'doyaGu-master' into dev 4554d6bb9 Fix error related to XBYAK_NOEXCEPT git-subtree-dir: externals/xbyak git-subtree-split: a1ac3750f9a639b5a6c6d6c7da4259b8d6790989
2022-12-30 23:05:02 +00:00
const char *getVersionString() const { return "6.68"; }
void aadd(const Address& addr, const Reg32e &reg) { opModM(addr, reg, 0x0F, 0x38, 0x0FC); }
void aand(const Address& addr, const Reg32e &reg) { db(0x66); opModM(addr, reg, 0x0F, 0x38, 0x0FC); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void adc(const Operand& op, uint32_t imm) { opRM_I(op, imm, 0x10, 2); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
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); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void add(const Operand& op, uint32_t imm) { opRM_I(op, imm, 0x00, 0); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void add(const Operand& op1, const Operand& op2) { opRM_RM(op1, op2, 0x00); }
void addpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x58, 0x66, isXMM_XMMorMEM); }
void addps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x58, 0x100, isXMM_XMMorMEM); }
void addsd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x58, 0xF2, isXMM_XMMorMEM); }
void addss(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x58, 0xF3, isXMM_XMMorMEM); }
void addsubpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xD0, 0x66, isXMM_XMMorMEM); }
void addsubps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xD0, 0xF2, isXMM_XMMorMEM); }
void adox(const Reg32e& reg, const Operand& op) { opGen(reg, op, 0xF6, 0xF3, isREG32_REG32orMEM, NONE, 0x38); }
void aesdec(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xDE, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void aesdeclast(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xDF, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void aesenc(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xDC, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void aesenclast(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xDD, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void aesimc(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xDB, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void aeskeygenassist(const Xmm& xmm, const Operand& op, uint8_t imm) { opGen(xmm, op, 0xDF, 0x66, isXMM_XMMorMEM, imm, 0x3A); }
void and_(const Operand& op, uint32_t imm) { opRM_I(op, imm, 0x20, 4); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void and_(const Operand& op1, const Operand& op2) { opRM_RM(op1, op2, 0x20); }
void andn(const Reg32e& r1, const Reg32e& r2, const Operand& op) { opGpr(r1, r2, op, T_0F38, 0xf2, true); }
void andnpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x55, 0x66, isXMM_XMMorMEM); }
void andnps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x55, 0x100, isXMM_XMMorMEM); }
void andpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x54, 0x66, isXMM_XMMorMEM); }
void andps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x54, 0x100, isXMM_XMMorMEM); }
Squashed 'externals/xbyak/' changes from 88f2f771f..a1ac3750f a1ac3750f Merge branch 'dev' 5f4ba971f v6.68 cac2c175f update doc 1b08a8248 add test of prefetchiti bef70d9b1 add prefetchit{0,1} f66785876 add detection of prefetchiti 124617ac9 Merge branch 'dev' 1bce8be17 Merge commit 'fbb18f6' into dev bebfe64a3 [test] X32 is optional and remove CFLAGS d8c332cef Merge branch 'Tachi107-small-changes' into dev fbb18f69d ci: use containers 8a974696c test: simplify makefile 1efe9fe7c Merge branch 'dev' bb70083e6 v6.67 436e452d7 update doc 47ff6ef42 I get an error after all on GitHub action 445c0dcec add test of CMPccXADD c9347907d add CMPccXADD dc792cc56 add detection of cmpccxadd 62be84cc8 fix detection of boost b5ac7b0f0 Merge branch 'dev' 05dd400e0 recover a removed line 94eff6246 v6.66 ec3fadeba update doc 00bfaaa7f add test of rao-int 0f2f1aaa6 support rio-int 6c047f480 detect rao-int f07c5c255 fix args of EmptyAllocator::alloc in test 04d3eb5f5 stop if a test script causes an error 1c1d2366f disable boost sample cfb1127c3 tweak 5fcbeb7c4 [sample] change the way of detection of boost bafc1ee60 CXX uses g++ as default value d8cabc6cb remove warning of blace bc73a0816 remove -fno-operator-names option 6989aea94 use CXX instead of g++ a7c5a1bd7 use English 1bfbd8c4d Fix incorrect format strings 0ecef5c28 Merge branch 'dev' 7556c20ba update doc a15709271 v6.65 3b83aab3e add detect_x32 to TARGET 8c64bbbc3 use gcc instead of dpkg for portability 5e9a9b96f test_avx512.sh runs on x32 8ae01b0c2 disable some tests on x32 83b3da217 x32 does not check large disp 693ab8c9d sizeof(void*) = 4 on x32, so disable the test 348e3e548 Merge branch 'dev' 11b9c4dc0 v6.64 459636196 add T_M_K flag to vpmov* 1d3722928 add include path 5e27eddae move CrearError from LabelManager::reset() to CodeGenerator::reset() f8ea5c28d Merge branch 'dev' 20b2b1eae v6.63 3706869f8 desc. of setDefaultEncoding d6f2d7577 add test of setDefaultEncoding 3b0a19c41 vpmadd52{h,l}uq for avx-ifma 95752ebd7 add tAVX_IFMA cd36e31ea [sample] show AMX_FP16/AVX_VNNI_INT8/AVX_NE_CONVERT e5858af27 add setDefaultEncoding 2f7fb0220 modify gen.cpp for AVX-NE-CONVERT/AVX-VNNI-INT8/AMX-FP16 1c5cb7efa add AVX-NE-CONVERT instructions 564fe9acd add AVX-VNNI-INT8 instructions cd14d07b1 add AMX-FP16 instruction 7811f593c Merge pull request #161 from scribam/patch-1 2218f6c08 Update changelog.md 7bccdbbb4 Merge pull request #160 from herumi/dev 5fcf87596 compile nasm-2.15 instead of apt install e31961ea8 v6.62 ff4f9e65c update doc 9c8fb81db disable wrong detection of gcc -Warray-bounds 6b7519659 add serialize e16582696 Merge branch 'JonLiu1993-vcpkg-installition' into dev edbb410fb Add vcpkg installation instructions 055d31242 Merge branch 'dev' 21ab98441 Merge branch 'akodanka-enable_CIV' into dev ed4d598e3 Changes to compile xbyak project for openvino b652430c4 mingw uses __cpuidex 48457bfa0 Merge branch 'dev' 29cb524d1 v6.61.2 1a9a0b0e1 avoid including algorithm header in xbyak_util.h 6fadefd04 Merge branch 'dev' fc1c18a9d update doc f7cae7f11 v6.61.1 6f5ec5cf3 Merge branch 'doyaGu-master' into dev 4554d6bb9 Fix error related to XBYAK_NOEXCEPT git-subtree-dir: externals/xbyak git-subtree-split: a1ac3750f9a639b5a6c6d6c7da4259b8d6790989
2022-12-30 23:05:02 +00:00
void aor(const Address& addr, const Reg32e &reg) { db(0xF2); opModM(addr, reg, 0x0F, 0x38, 0x0FC); }
void axor(const Address& addr, const Reg32e &reg) { db(0xF3); opModM(addr, reg, 0x0F, 0x38, 0x0FC); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void bextr(const Reg32e& r1, const Operand& op, const Reg32e& r2) { opGpr(r1, op, r2, T_0F38, 0xf7, false); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void blendpd(const Xmm& xmm, const Operand& op, int imm) { opGen(xmm, op, 0x0D, 0x66, isXMM_XMMorMEM, static_cast<uint8_t>(imm), 0x3A); }
void blendps(const Xmm& xmm, const Operand& op, int imm) { opGen(xmm, op, 0x0C, 0x66, isXMM_XMMorMEM, static_cast<uint8_t>(imm), 0x3A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void blendvpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x15, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void blendvps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x14, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void blsi(const Reg32e& r, const Operand& op) { opGpr(Reg32e(3, r.getBit()), op, r, T_0F38, 0xf3, false); }
void blsmsk(const Reg32e& r, const Operand& op) { opGpr(Reg32e(2, r.getBit()), op, r, T_0F38, 0xf3, false); }
void blsr(const Reg32e& r, const Operand& op) { opGpr(Reg32e(1, r.getBit()), op, r, T_0F38, 0xf3, false); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
void bnd() { db(0xF2); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void bndcl(const BoundsReg& bnd, const Operand& op) { db(0xF3); opR_ModM(op, i32e, bnd.getIdx(), 0x0F, 0x1A, NONE, !op.isMEM()); }
void bndcn(const BoundsReg& bnd, const Operand& op) { db(0xF2); opR_ModM(op, i32e, bnd.getIdx(), 0x0F, 0x1B, NONE, !op.isMEM()); }
void bndcu(const BoundsReg& bnd, const Operand& op) { db(0xF2); opR_ModM(op, i32e, bnd.getIdx(), 0x0F, 0x1A, NONE, !op.isMEM()); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
void bndldx(const BoundsReg& bnd, const Address& addr) { opMIB(addr, bnd, 0x0F, 0x1A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void bndmk(const BoundsReg& bnd, const Address& addr) { db(0xF3); opModM(addr, bnd, 0x0F, 0x1B); }
void bndmov(const Address& addr, const BoundsReg& bnd) { db(0x66); opModM(addr, bnd, 0x0F, 0x1B); }
void bndmov(const BoundsReg& bnd, const Operand& op) { db(0x66); opModRM(bnd, op, op.isBNDREG(), op.isMEM(), 0x0F, 0x1A); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
void bndstx(const Address& addr, const BoundsReg& bnd) { opMIB(addr, bnd, 0x0F, 0x1B); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void bsf(const Reg&reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0xBC); }
void bsr(const Reg&reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0xBD); }
void bswap(const Reg32e& reg) { opModR(Reg32(1), reg, 0x0F); }
void bt(const Operand& op, const Reg& reg) { opModRM(reg, op, op.isREG(16|32|64) && op.getBit() == reg.getBit(), op.isMEM(), 0x0f, 0xA3); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void bt(const Operand& op, uint8_t imm) { opR_ModM(op, 16|32|64, 4, 0x0f, 0xba, NONE, false, 1); db(imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void btc(const Operand& op, const Reg& reg) { opModRM(reg, op, op.isREG(16|32|64) && op.getBit() == reg.getBit(), op.isMEM(), 0x0f, 0xBB); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void btc(const Operand& op, uint8_t imm) { opR_ModM(op, 16|32|64, 7, 0x0f, 0xba, NONE, false, 1); db(imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void btr(const Operand& op, const Reg& reg) { opModRM(reg, op, op.isREG(16|32|64) && op.getBit() == reg.getBit(), op.isMEM(), 0x0f, 0xB3); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void btr(const Operand& op, uint8_t imm) { opR_ModM(op, 16|32|64, 6, 0x0f, 0xba, NONE, false, 1); db(imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void bts(const Operand& op, const Reg& reg) { opModRM(reg, op, op.isREG(16|32|64) && op.getBit() == reg.getBit(), op.isMEM(), 0x0f, 0xAB); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void bts(const Operand& op, uint8_t imm) { opR_ModM(op, 16|32|64, 5, 0x0f, 0xba, NONE, false, 1); db(imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void bzhi(const Reg32e& r1, const Operand& op, const Reg32e& r2) { opGpr(r1, op, r2, T_0F38, 0xf5, false); }
void cbw() { db(0x66); db(0x98); }
void cdq() { db(0x99); }
void clc() { db(0xF8); }
void cld() { db(0xFC); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void cldemote(const Address& addr) { opMIB(addr, eax, 0x0F, 0x1C); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void clflush(const Address& addr) { opModM(addr, Reg32(7), 0x0F, 0xAE); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void clflushopt(const Address& addr) { db(0x66); opModM(addr, Reg32(7), 0x0F, 0xAE); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void cli() { db(0xFA); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void clwb(const Address& addr) { db(0x66); opMIB(addr, esi, 0x0F, 0xAE); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void clzero() { db(0x0F); db(0x01); db(0xFC); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void cmc() { db(0xF5); }
void cmova(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 7); }//-V524
void cmovae(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 3); }//-V524
void cmovb(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 2); }//-V524
void cmovbe(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 6); }//-V524
void cmovc(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 2); }//-V524
void cmove(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 4); }//-V524
void cmovg(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 15); }//-V524
void cmovge(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 13); }//-V524
void cmovl(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 12); }//-V524
void cmovle(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 14); }//-V524
void cmovna(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 6); }//-V524
void cmovnae(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 2); }//-V524
void cmovnb(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 3); }//-V524
void cmovnbe(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 7); }//-V524
void cmovnc(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 3); }//-V524
void cmovne(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 5); }//-V524
void cmovng(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 14); }//-V524
void cmovnge(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 12); }//-V524
void cmovnl(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 13); }//-V524
void cmovnle(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 15); }//-V524
void cmovno(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 1); }//-V524
void cmovnp(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 11); }//-V524
void cmovns(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 9); }//-V524
void cmovnz(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 5); }//-V524
void cmovo(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 0); }//-V524
void cmovp(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 10); }//-V524
void cmovpe(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 10); }//-V524
void cmovpo(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 11); }//-V524
void cmovs(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 8); }//-V524
void cmovz(const Reg& reg, const Operand& op) { opModRM(reg, op, op.isREG(16 | i32e), op.isMEM(), 0x0F, 0x40 | 4); }//-V524
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void cmp(const Operand& op, uint32_t imm) { opRM_I(op, imm, 0x38, 7); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void cmp(const Operand& op1, const Operand& op2) { opRM_RM(op1, op2, 0x38); }
void cmpeqpd(const Xmm& x, const Operand& op) { cmppd(x, op, 0); }
void cmpeqps(const Xmm& x, const Operand& op) { cmpps(x, op, 0); }
void cmpeqsd(const Xmm& x, const Operand& op) { cmpsd(x, op, 0); }
void cmpeqss(const Xmm& x, const Operand& op) { cmpss(x, op, 0); }
void cmplepd(const Xmm& x, const Operand& op) { cmppd(x, op, 2); }
void cmpleps(const Xmm& x, const Operand& op) { cmpps(x, op, 2); }
void cmplesd(const Xmm& x, const Operand& op) { cmpsd(x, op, 2); }
void cmpless(const Xmm& x, const Operand& op) { cmpss(x, op, 2); }
void cmpltpd(const Xmm& x, const Operand& op) { cmppd(x, op, 1); }
void cmpltps(const Xmm& x, const Operand& op) { cmpps(x, op, 1); }
void cmpltsd(const Xmm& x, const Operand& op) { cmpsd(x, op, 1); }
void cmpltss(const Xmm& x, const Operand& op) { cmpss(x, op, 1); }
void cmpneqpd(const Xmm& x, const Operand& op) { cmppd(x, op, 4); }
void cmpneqps(const Xmm& x, const Operand& op) { cmpps(x, op, 4); }
void cmpneqsd(const Xmm& x, const Operand& op) { cmpsd(x, op, 4); }
void cmpneqss(const Xmm& x, const Operand& op) { cmpss(x, op, 4); }
void cmpnlepd(const Xmm& x, const Operand& op) { cmppd(x, op, 6); }
void cmpnleps(const Xmm& x, const Operand& op) { cmpps(x, op, 6); }
void cmpnlesd(const Xmm& x, const Operand& op) { cmpsd(x, op, 6); }
void cmpnless(const Xmm& x, const Operand& op) { cmpss(x, op, 6); }
void cmpnltpd(const Xmm& x, const Operand& op) { cmppd(x, op, 5); }
void cmpnltps(const Xmm& x, const Operand& op) { cmpps(x, op, 5); }
void cmpnltsd(const Xmm& x, const Operand& op) { cmpsd(x, op, 5); }
void cmpnltss(const Xmm& x, const Operand& op) { cmpss(x, op, 5); }
void cmpordpd(const Xmm& x, const Operand& op) { cmppd(x, op, 7); }
void cmpordps(const Xmm& x, const Operand& op) { cmpps(x, op, 7); }
void cmpordsd(const Xmm& x, const Operand& op) { cmpsd(x, op, 7); }
void cmpordss(const Xmm& x, const Operand& op) { cmpss(x, op, 7); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void cmppd(const Xmm& xmm, const Operand& op, uint8_t imm8) { opGen(xmm, op, 0xC2, 0x66, isXMM_XMMorMEM, imm8); }
void cmpps(const Xmm& xmm, const Operand& op, uint8_t imm8) { opGen(xmm, op, 0xC2, 0x100, isXMM_XMMorMEM, imm8); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void cmpsb() { db(0xA6); }
void cmpsd() { db(0xA7); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void cmpsd(const Xmm& xmm, const Operand& op, uint8_t imm8) { opGen(xmm, op, 0xC2, 0xF2, isXMM_XMMorMEM, imm8); }
void cmpss(const Xmm& xmm, const Operand& op, uint8_t imm8) { opGen(xmm, op, 0xC2, 0xF3, isXMM_XMMorMEM, imm8); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void cmpsw() { db(0x66); db(0xA7); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void cmpunordpd(const Xmm& x, const Operand& op) { cmppd(x, op, 3); }
void cmpunordps(const Xmm& x, const Operand& op) { cmpps(x, op, 3); }
void cmpunordsd(const Xmm& x, const Operand& op) { cmpsd(x, op, 3); }
void cmpunordss(const Xmm& x, const Operand& op) { cmpss(x, op, 3); }
void cmpxchg(const Operand& op, const Reg& reg) { opModRM(reg, op, (op.isREG() && reg.isREG() && op.getBit() == reg.getBit()), op.isMEM(), 0x0F, 0xB0 | (reg.isBit(8) ? 0 : 1)); }
void cmpxchg8b(const Address& addr) { opModM(addr, Reg32(1), 0x0F, 0xC7); }
void comisd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x2F, 0x66, isXMM_XMMorMEM); }
void comiss(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x2F, 0x100, isXMM_XMMorMEM); }
void cpuid() { db(0x0F); db(0xA2); }
void crc32(const Reg32e& reg, const Operand& op) { if (reg.isBit(32) && op.isBit(16)) db(0x66); db(0xF2); opModRM(reg, op, op.isREG(), op.isMEM(), 0x0F, 0x38, 0xF0 | (op.isBit(8) ? 0 : 1)); }
void cvtdq2pd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xE6, 0xF3, isXMM_XMMorMEM); }
void cvtdq2ps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5B, 0x100, isXMM_XMMorMEM); }
void cvtpd2dq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xE6, 0xF2, isXMM_XMMorMEM); }
void cvtpd2pi(const Operand& reg, const Operand& op) { opGen(reg, op, 0x2D, 0x66, isMMX_XMMorMEM); }
void cvtpd2ps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5A, 0x66, isXMM_XMMorMEM); }
void cvtpi2pd(const Operand& reg, const Operand& op) { opGen(reg, op, 0x2A, 0x66, isXMM_MMXorMEM); }
void cvtpi2ps(const Operand& reg, const Operand& op) { opGen(reg, op, 0x2A, 0x100, isXMM_MMXorMEM); }
void cvtps2dq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5B, 0x66, isXMM_XMMorMEM); }
void cvtps2pd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5A, 0x100, isXMM_XMMorMEM); }
void cvtps2pi(const Operand& reg, const Operand& op) { opGen(reg, op, 0x2D, 0x100, isMMX_XMMorMEM); }
void cvtsd2si(const Operand& reg, const Operand& op) { opGen(reg, op, 0x2D, 0xF2, isREG32_XMMorMEM); }
void cvtsd2ss(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5A, 0xF2, isXMM_XMMorMEM); }
void cvtsi2sd(const Operand& reg, const Operand& op) { opGen(reg, op, 0x2A, 0xF2, isXMM_REG32orMEM); }
void cvtsi2ss(const Operand& reg, const Operand& op) { opGen(reg, op, 0x2A, 0xF3, isXMM_REG32orMEM); }
void cvtss2sd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5A, 0xF3, isXMM_XMMorMEM); }
void cvtss2si(const Operand& reg, const Operand& op) { opGen(reg, op, 0x2D, 0xF3, isREG32_XMMorMEM); }
void cvttpd2dq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xE6, 0x66, isXMM_XMMorMEM); }
void cvttpd2pi(const Operand& reg, const Operand& op) { opGen(reg, op, 0x2C, 0x66, isMMX_XMMorMEM); }
void cvttps2dq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5B, 0xF3, isXMM_XMMorMEM); }
void cvttps2pi(const Operand& reg, const Operand& op) { opGen(reg, op, 0x2C, 0x100, isMMX_XMMorMEM); }
void cvttsd2si(const Operand& reg, const Operand& op) { opGen(reg, op, 0x2C, 0xF2, isREG32_XMMorMEM); }
void cvttss2si(const Operand& reg, const Operand& op) { opGen(reg, op, 0x2C, 0xF3, isREG32_XMMorMEM); }
void cwd() { db(0x66); db(0x99); }
void cwde() { db(0x98); }
void dec(const Operand& op) { opIncDec(op, 0x48, 1); }
void div(const Operand& op) { opR_ModM(op, 0, 6, 0xF6); }
void divpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5E, 0x66, isXMM_XMMorMEM); }
void divps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5E, 0x100, isXMM_XMMorMEM); }
void divsd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5E, 0xF2, isXMM_XMMorMEM); }
void divss(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5E, 0xF3, isXMM_XMMorMEM); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void dppd(const Xmm& xmm, const Operand& op, int imm) { opGen(xmm, op, 0x41, 0x66, isXMM_XMMorMEM, static_cast<uint8_t>(imm), 0x3A); }
void dpps(const Xmm& xmm, const Operand& op, int imm) { opGen(xmm, op, 0x40, 0x66, isXMM_XMMorMEM, static_cast<uint8_t>(imm), 0x3A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void emms() { db(0x0F); db(0x77); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void endbr32() { db(0xF3); db(0x0F); db(0x1E); db(0xFB); }
void endbr64() { db(0xF3); db(0x0F); db(0x1E); db(0xFA); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void enter(uint16_t x, uint8_t y) { db(0xC8); dw(x); db(y); }
void extractps(const Operand& op, const Xmm& xmm, uint8_t imm) { opExt(op, xmm, 0x17, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void f2xm1() { db(0xD9); db(0xF0); }
void fabs() { db(0xD9); db(0xE1); }
void fadd(const Address& addr) { opFpuMem(addr, 0x00, 0xD8, 0xDC, 0, 0); }
void fadd(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xD8C0, 0xDCC0); }
void fadd(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xD8C0, 0xDCC0); }
void faddp() { db(0xDE); db(0xC1); }
void faddp(const Fpu& reg1) { opFpuFpu(reg1, st0, 0x0000, 0xDEC0); }
void faddp(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0x0000, 0xDEC0); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void fbld(const Address& addr) { opModM(addr, Reg32(4), 0xDF, 0x100); }
void fbstp(const Address& addr) { opModM(addr, Reg32(6), 0xDF, 0x100); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void fchs() { db(0xD9); db(0xE0); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void fclex() { db(0x9B); db(0xDB); db(0xE2); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void fcmovb(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xDAC0, 0x00C0); }
void fcmovb(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xDAC0, 0x00C0); }
void fcmovbe(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xDAD0, 0x00D0); }
void fcmovbe(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xDAD0, 0x00D0); }
void fcmove(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xDAC8, 0x00C8); }
void fcmove(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xDAC8, 0x00C8); }
void fcmovnb(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xDBC0, 0x00C0); }
void fcmovnb(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xDBC0, 0x00C0); }
void fcmovnbe(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xDBD0, 0x00D0); }
void fcmovnbe(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xDBD0, 0x00D0); }
void fcmovne(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xDBC8, 0x00C8); }
void fcmovne(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xDBC8, 0x00C8); }
void fcmovnu(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xDBD8, 0x00D8); }
void fcmovnu(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xDBD8, 0x00D8); }
void fcmovu(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xDAD8, 0x00D8); }
void fcmovu(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xDAD8, 0x00D8); }
void fcom() { db(0xD8); db(0xD1); }
void fcom(const Address& addr) { opFpuMem(addr, 0x00, 0xD8, 0xDC, 2, 0); }
void fcom(const Fpu& reg) { opFpu(reg, 0xD8, 0xD0); }
void fcomi(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xDBF0, 0x00F0); }
void fcomi(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xDBF0, 0x00F0); }
void fcomip(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xDFF0, 0x00F0); }
void fcomip(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xDFF0, 0x00F0); }
void fcomp() { db(0xD8); db(0xD9); }
void fcomp(const Address& addr) { opFpuMem(addr, 0x00, 0xD8, 0xDC, 3, 0); }
void fcomp(const Fpu& reg) { opFpu(reg, 0xD8, 0xD8); }
void fcompp() { db(0xDE); db(0xD9); }
void fcos() { db(0xD9); db(0xFF); }
void fdecstp() { db(0xD9); db(0xF6); }
void fdiv(const Address& addr) { opFpuMem(addr, 0x00, 0xD8, 0xDC, 6, 0); }
void fdiv(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xD8F0, 0xDCF8); }
void fdiv(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xD8F0, 0xDCF8); }
void fdivp() { db(0xDE); db(0xF9); }
void fdivp(const Fpu& reg1) { opFpuFpu(reg1, st0, 0x0000, 0xDEF8); }
void fdivp(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0x0000, 0xDEF8); }
void fdivr(const Address& addr) { opFpuMem(addr, 0x00, 0xD8, 0xDC, 7, 0); }
void fdivr(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xD8F8, 0xDCF0); }
void fdivr(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xD8F8, 0xDCF0); }
void fdivrp() { db(0xDE); db(0xF1); }
void fdivrp(const Fpu& reg1) { opFpuFpu(reg1, st0, 0x0000, 0xDEF0); }
void fdivrp(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0x0000, 0xDEF0); }
void ffree(const Fpu& reg) { opFpu(reg, 0xDD, 0xC0); }
void fiadd(const Address& addr) { opFpuMem(addr, 0xDE, 0xDA, 0x00, 0, 0); }
void ficom(const Address& addr) { opFpuMem(addr, 0xDE, 0xDA, 0x00, 2, 0); }
void ficomp(const Address& addr) { opFpuMem(addr, 0xDE, 0xDA, 0x00, 3, 0); }
void fidiv(const Address& addr) { opFpuMem(addr, 0xDE, 0xDA, 0x00, 6, 0); }
void fidivr(const Address& addr) { opFpuMem(addr, 0xDE, 0xDA, 0x00, 7, 0); }
void fild(const Address& addr) { opFpuMem(addr, 0xDF, 0xDB, 0xDF, 0, 5); }
void fimul(const Address& addr) { opFpuMem(addr, 0xDE, 0xDA, 0x00, 1, 0); }
void fincstp() { db(0xD9); db(0xF7); }
void finit() { db(0x9B); db(0xDB); db(0xE3); }
void fist(const Address& addr) { opFpuMem(addr, 0xDF, 0xDB, 0x00, 2, 0); }
void fistp(const Address& addr) { opFpuMem(addr, 0xDF, 0xDB, 0xDF, 3, 7); }
void fisttp(const Address& addr) { opFpuMem(addr, 0xDF, 0xDB, 0xDD, 1, 0); }
void fisub(const Address& addr) { opFpuMem(addr, 0xDE, 0xDA, 0x00, 4, 0); }
void fisubr(const Address& addr) { opFpuMem(addr, 0xDE, 0xDA, 0x00, 5, 0); }
void fld(const Address& addr) { opFpuMem(addr, 0x00, 0xD9, 0xDD, 0, 0); }
void fld(const Fpu& reg) { opFpu(reg, 0xD9, 0xC0); }
void fld1() { db(0xD9); db(0xE8); }
void fldcw(const Address& addr) { opModM(addr, Reg32(5), 0xD9, 0x100); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void fldenv(const Address& addr) { opModM(addr, Reg32(4), 0xD9, 0x100); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void fldl2e() { db(0xD9); db(0xEA); }
void fldl2t() { db(0xD9); db(0xE9); }
void fldlg2() { db(0xD9); db(0xEC); }
void fldln2() { db(0xD9); db(0xED); }
void fldpi() { db(0xD9); db(0xEB); }
void fldz() { db(0xD9); db(0xEE); }
void fmul(const Address& addr) { opFpuMem(addr, 0x00, 0xD8, 0xDC, 1, 0); }
void fmul(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xD8C8, 0xDCC8); }
void fmul(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xD8C8, 0xDCC8); }
void fmulp() { db(0xDE); db(0xC9); }
void fmulp(const Fpu& reg1) { opFpuFpu(reg1, st0, 0x0000, 0xDEC8); }
void fmulp(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0x0000, 0xDEC8); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void fnclex() { db(0xDB); db(0xE2); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void fninit() { db(0xDB); db(0xE3); }
void fnop() { db(0xD9); db(0xD0); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void fnsave(const Address& addr) { opModM(addr, Reg32(6), 0xDD, 0x100); }
void fnstcw(const Address& addr) { opModM(addr, Reg32(7), 0xD9, 0x100); }
void fnstenv(const Address& addr) { opModM(addr, Reg32(6), 0xD9, 0x100); }
void fnstsw(const Address& addr) { opModM(addr, Reg32(7), 0xDD, 0x100); }
void fnstsw(const Reg16& r) { if (r.getIdx() != Operand::AX) XBYAK_THROW(ERR_BAD_PARAMETER) db(0xDF); db(0xE0); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void fpatan() { db(0xD9); db(0xF3); }
void fprem() { db(0xD9); db(0xF8); }
void fprem1() { db(0xD9); db(0xF5); }
void fptan() { db(0xD9); db(0xF2); }
void frndint() { db(0xD9); db(0xFC); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void frstor(const Address& addr) { opModM(addr, Reg32(4), 0xDD, 0x100); }
void fsave(const Address& addr) { db(0x9B); opModM(addr, Reg32(6), 0xDD, 0x100); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void fscale() { db(0xD9); db(0xFD); }
void fsin() { db(0xD9); db(0xFE); }
void fsincos() { db(0xD9); db(0xFB); }
void fsqrt() { db(0xD9); db(0xFA); }
void fst(const Address& addr) { opFpuMem(addr, 0x00, 0xD9, 0xDD, 2, 0); }
void fst(const Fpu& reg) { opFpu(reg, 0xDD, 0xD0); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void fstcw(const Address& addr) { db(0x9B); opModM(addr, Reg32(7), 0xD9, 0x100); }
void fstenv(const Address& addr) { db(0x9B); opModM(addr, Reg32(6), 0xD9, 0x100); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void fstp(const Address& addr) { opFpuMem(addr, 0x00, 0xD9, 0xDD, 3, 0); }
void fstp(const Fpu& reg) { opFpu(reg, 0xDD, 0xD8); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void fstsw(const Address& addr) { db(0x9B); opModM(addr, Reg32(7), 0xDD, 0x100); }
void fstsw(const Reg16& r) { if (r.getIdx() != Operand::AX) XBYAK_THROW(ERR_BAD_PARAMETER) db(0x9B); db(0xDF); db(0xE0); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void fsub(const Address& addr) { opFpuMem(addr, 0x00, 0xD8, 0xDC, 4, 0); }
void fsub(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xD8E0, 0xDCE8); }
void fsub(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xD8E0, 0xDCE8); }
void fsubp() { db(0xDE); db(0xE9); }
void fsubp(const Fpu& reg1) { opFpuFpu(reg1, st0, 0x0000, 0xDEE8); }
void fsubp(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0x0000, 0xDEE8); }
void fsubr(const Address& addr) { opFpuMem(addr, 0x00, 0xD8, 0xDC, 5, 0); }
void fsubr(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xD8E8, 0xDCE0); }
void fsubr(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xD8E8, 0xDCE0); }
void fsubrp() { db(0xDE); db(0xE1); }
void fsubrp(const Fpu& reg1) { opFpuFpu(reg1, st0, 0x0000, 0xDEE0); }
void fsubrp(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0x0000, 0xDEE0); }
void ftst() { db(0xD9); db(0xE4); }
void fucom() { db(0xDD); db(0xE1); }
void fucom(const Fpu& reg) { opFpu(reg, 0xDD, 0xE0); }
void fucomi(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xDBE8, 0x00E8); }
void fucomi(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xDBE8, 0x00E8); }
void fucomip(const Fpu& reg1) { opFpuFpu(st0, reg1, 0xDFE8, 0x00E8); }
void fucomip(const Fpu& reg1, const Fpu& reg2) { opFpuFpu(reg1, reg2, 0xDFE8, 0x00E8); }
void fucomp() { db(0xDD); db(0xE9); }
void fucomp(const Fpu& reg) { opFpu(reg, 0xDD, 0xE8); }
void fucompp() { db(0xDA); db(0xE9); }
void fwait() { db(0x9B); }
void fxam() { db(0xD9); db(0xE5); }
void fxch() { db(0xD9); db(0xC9); }
void fxch(const Fpu& reg) { opFpu(reg, 0xD9, 0xC8); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void fxrstor(const Address& addr) { opModM(addr, Reg32(1), 0x0F, 0xAE); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void fxtract() { db(0xD9); db(0xF4); }
void fyl2x() { db(0xD9); db(0xF1); }
void fyl2xp1() { db(0xD9); db(0xF9); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void gf2p8affineinvqb(const Xmm& xmm, const Operand& op, int imm) { opGen(xmm, op, 0xCF, 0x66, isXMM_XMMorMEM, static_cast<uint8_t>(imm), 0x3A); }
void gf2p8affineqb(const Xmm& xmm, const Operand& op, int imm) { opGen(xmm, op, 0xCE, 0x66, isXMM_XMMorMEM, static_cast<uint8_t>(imm), 0x3A); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
void gf2p8mulb(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xCF, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void haddpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x7C, 0x66, isXMM_XMMorMEM); }
void haddps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x7C, 0xF2, isXMM_XMMorMEM); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void hlt() { db(0xF4); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void hsubpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x7D, 0x66, isXMM_XMMorMEM); }
void hsubps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x7D, 0xF2, isXMM_XMMorMEM); }
void idiv(const Operand& op) { opR_ModM(op, 0, 7, 0xF6); }
void imul(const Operand& op) { opR_ModM(op, 0, 5, 0xF6); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void in_(const Reg& a, const Reg& d) { opInOut(a, d, 0xEC); }
void in_(const Reg& a, uint8_t v) { opInOut(a, 0xE4, v); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void inc(const Operand& op) { opIncDec(op, 0x40, 0); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void insertps(const Xmm& xmm, const Operand& op, uint8_t imm) { opGen(xmm, op, 0x21, 0x66, isXMM_XMMorMEM, imm, 0x3A); }
void int3() { db(0xCC); }
void int_(uint8_t x) { db(0xCD); db(x); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void ja(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x77, 0x87, 0x0F); }//-V524
void ja(const char *label, LabelType type = T_AUTO) { ja(std::string(label), type); }//-V524
void ja(const void *addr) { opJmpAbs(addr, T_NEAR, 0x77, 0x87, 0x0F); }//-V524
void ja(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x77, 0x87, 0x0F); }//-V524
void jae(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x73, 0x83, 0x0F); }//-V524
void jae(const char *label, LabelType type = T_AUTO) { jae(std::string(label), type); }//-V524
void jae(const void *addr) { opJmpAbs(addr, T_NEAR, 0x73, 0x83, 0x0F); }//-V524
void jae(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x73, 0x83, 0x0F); }//-V524
void jb(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x72, 0x82, 0x0F); }//-V524
void jb(const char *label, LabelType type = T_AUTO) { jb(std::string(label), type); }//-V524
void jb(const void *addr) { opJmpAbs(addr, T_NEAR, 0x72, 0x82, 0x0F); }//-V524
void jb(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x72, 0x82, 0x0F); }//-V524
void jbe(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x76, 0x86, 0x0F); }//-V524
void jbe(const char *label, LabelType type = T_AUTO) { jbe(std::string(label), type); }//-V524
void jbe(const void *addr) { opJmpAbs(addr, T_NEAR, 0x76, 0x86, 0x0F); }//-V524
void jbe(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x76, 0x86, 0x0F); }//-V524
void jc(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x72, 0x82, 0x0F); }//-V524
void jc(const char *label, LabelType type = T_AUTO) { jc(std::string(label), type); }//-V524
void jc(const void *addr) { opJmpAbs(addr, T_NEAR, 0x72, 0x82, 0x0F); }//-V524
void jc(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x72, 0x82, 0x0F); }//-V524
void je(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x74, 0x84, 0x0F); }//-V524
void je(const char *label, LabelType type = T_AUTO) { je(std::string(label), type); }//-V524
void je(const void *addr) { opJmpAbs(addr, T_NEAR, 0x74, 0x84, 0x0F); }//-V524
void je(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x74, 0x84, 0x0F); }//-V524
void jg(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x7F, 0x8F, 0x0F); }//-V524
void jg(const char *label, LabelType type = T_AUTO) { jg(std::string(label), type); }//-V524
void jg(const void *addr) { opJmpAbs(addr, T_NEAR, 0x7F, 0x8F, 0x0F); }//-V524
void jg(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x7F, 0x8F, 0x0F); }//-V524
void jge(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x7D, 0x8D, 0x0F); }//-V524
void jge(const char *label, LabelType type = T_AUTO) { jge(std::string(label), type); }//-V524
void jge(const void *addr) { opJmpAbs(addr, T_NEAR, 0x7D, 0x8D, 0x0F); }//-V524
void jge(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x7D, 0x8D, 0x0F); }//-V524
void jl(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x7C, 0x8C, 0x0F); }//-V524
void jl(const char *label, LabelType type = T_AUTO) { jl(std::string(label), type); }//-V524
void jl(const void *addr) { opJmpAbs(addr, T_NEAR, 0x7C, 0x8C, 0x0F); }//-V524
void jl(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x7C, 0x8C, 0x0F); }//-V524
void jle(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x7E, 0x8E, 0x0F); }//-V524
void jle(const char *label, LabelType type = T_AUTO) { jle(std::string(label), type); }//-V524
void jle(const void *addr) { opJmpAbs(addr, T_NEAR, 0x7E, 0x8E, 0x0F); }//-V524
void jle(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x7E, 0x8E, 0x0F); }//-V524
void jna(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x76, 0x86, 0x0F); }//-V524
void jna(const char *label, LabelType type = T_AUTO) { jna(std::string(label), type); }//-V524
void jna(const void *addr) { opJmpAbs(addr, T_NEAR, 0x76, 0x86, 0x0F); }//-V524
void jna(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x76, 0x86, 0x0F); }//-V524
void jnae(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x72, 0x82, 0x0F); }//-V524
void jnae(const char *label, LabelType type = T_AUTO) { jnae(std::string(label), type); }//-V524
void jnae(const void *addr) { opJmpAbs(addr, T_NEAR, 0x72, 0x82, 0x0F); }//-V524
void jnae(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x72, 0x82, 0x0F); }//-V524
void jnb(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x73, 0x83, 0x0F); }//-V524
void jnb(const char *label, LabelType type = T_AUTO) { jnb(std::string(label), type); }//-V524
void jnb(const void *addr) { opJmpAbs(addr, T_NEAR, 0x73, 0x83, 0x0F); }//-V524
void jnb(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x73, 0x83, 0x0F); }//-V524
void jnbe(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x77, 0x87, 0x0F); }//-V524
void jnbe(const char *label, LabelType type = T_AUTO) { jnbe(std::string(label), type); }//-V524
void jnbe(const void *addr) { opJmpAbs(addr, T_NEAR, 0x77, 0x87, 0x0F); }//-V524
void jnbe(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x77, 0x87, 0x0F); }//-V524
void jnc(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x73, 0x83, 0x0F); }//-V524
void jnc(const char *label, LabelType type = T_AUTO) { jnc(std::string(label), type); }//-V524
void jnc(const void *addr) { opJmpAbs(addr, T_NEAR, 0x73, 0x83, 0x0F); }//-V524
void jnc(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x73, 0x83, 0x0F); }//-V524
void jne(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x75, 0x85, 0x0F); }//-V524
void jne(const char *label, LabelType type = T_AUTO) { jne(std::string(label), type); }//-V524
void jne(const void *addr) { opJmpAbs(addr, T_NEAR, 0x75, 0x85, 0x0F); }//-V524
void jne(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x75, 0x85, 0x0F); }//-V524
void jng(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x7E, 0x8E, 0x0F); }//-V524
void jng(const char *label, LabelType type = T_AUTO) { jng(std::string(label), type); }//-V524
void jng(const void *addr) { opJmpAbs(addr, T_NEAR, 0x7E, 0x8E, 0x0F); }//-V524
void jng(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x7E, 0x8E, 0x0F); }//-V524
void jnge(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x7C, 0x8C, 0x0F); }//-V524
void jnge(const char *label, LabelType type = T_AUTO) { jnge(std::string(label), type); }//-V524
void jnge(const void *addr) { opJmpAbs(addr, T_NEAR, 0x7C, 0x8C, 0x0F); }//-V524
void jnge(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x7C, 0x8C, 0x0F); }//-V524
void jnl(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x7D, 0x8D, 0x0F); }//-V524
void jnl(const char *label, LabelType type = T_AUTO) { jnl(std::string(label), type); }//-V524
void jnl(const void *addr) { opJmpAbs(addr, T_NEAR, 0x7D, 0x8D, 0x0F); }//-V524
void jnl(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x7D, 0x8D, 0x0F); }//-V524
void jnle(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x7F, 0x8F, 0x0F); }//-V524
void jnle(const char *label, LabelType type = T_AUTO) { jnle(std::string(label), type); }//-V524
void jnle(const void *addr) { opJmpAbs(addr, T_NEAR, 0x7F, 0x8F, 0x0F); }//-V524
void jnle(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x7F, 0x8F, 0x0F); }//-V524
void jno(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x71, 0x81, 0x0F); }//-V524
void jno(const char *label, LabelType type = T_AUTO) { jno(std::string(label), type); }//-V524
void jno(const void *addr) { opJmpAbs(addr, T_NEAR, 0x71, 0x81, 0x0F); }//-V524
void jno(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x71, 0x81, 0x0F); }//-V524
void jnp(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x7B, 0x8B, 0x0F); }//-V524
void jnp(const char *label, LabelType type = T_AUTO) { jnp(std::string(label), type); }//-V524
void jnp(const void *addr) { opJmpAbs(addr, T_NEAR, 0x7B, 0x8B, 0x0F); }//-V524
void jnp(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x7B, 0x8B, 0x0F); }//-V524
void jns(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x79, 0x89, 0x0F); }//-V524
void jns(const char *label, LabelType type = T_AUTO) { jns(std::string(label), type); }//-V524
void jns(const void *addr) { opJmpAbs(addr, T_NEAR, 0x79, 0x89, 0x0F); }//-V524
void jns(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x79, 0x89, 0x0F); }//-V524
void jnz(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x75, 0x85, 0x0F); }//-V524
void jnz(const char *label, LabelType type = T_AUTO) { jnz(std::string(label), type); }//-V524
void jnz(const void *addr) { opJmpAbs(addr, T_NEAR, 0x75, 0x85, 0x0F); }//-V524
void jnz(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x75, 0x85, 0x0F); }//-V524
void jo(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x70, 0x80, 0x0F); }//-V524
void jo(const char *label, LabelType type = T_AUTO) { jo(std::string(label), type); }//-V524
void jo(const void *addr) { opJmpAbs(addr, T_NEAR, 0x70, 0x80, 0x0F); }//-V524
void jo(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x70, 0x80, 0x0F); }//-V524
void jp(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x7A, 0x8A, 0x0F); }//-V524
void jp(const char *label, LabelType type = T_AUTO) { jp(std::string(label), type); }//-V524
void jp(const void *addr) { opJmpAbs(addr, T_NEAR, 0x7A, 0x8A, 0x0F); }//-V524
void jp(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x7A, 0x8A, 0x0F); }//-V524
void jpe(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x7A, 0x8A, 0x0F); }//-V524
void jpe(const char *label, LabelType type = T_AUTO) { jpe(std::string(label), type); }//-V524
void jpe(const void *addr) { opJmpAbs(addr, T_NEAR, 0x7A, 0x8A, 0x0F); }//-V524
void jpe(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x7A, 0x8A, 0x0F); }//-V524
void jpo(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x7B, 0x8B, 0x0F); }//-V524
void jpo(const char *label, LabelType type = T_AUTO) { jpo(std::string(label), type); }//-V524
void jpo(const void *addr) { opJmpAbs(addr, T_NEAR, 0x7B, 0x8B, 0x0F); }//-V524
void jpo(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x7B, 0x8B, 0x0F); }//-V524
void js(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x78, 0x88, 0x0F); }//-V524
void js(const char *label, LabelType type = T_AUTO) { js(std::string(label), type); }//-V524
void js(const void *addr) { opJmpAbs(addr, T_NEAR, 0x78, 0x88, 0x0F); }//-V524
void js(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x78, 0x88, 0x0F); }//-V524
void jz(const Label& label, LabelType type = T_AUTO) { opJmp(label, type, 0x74, 0x84, 0x0F); }//-V524
void jz(const char *label, LabelType type = T_AUTO) { jz(std::string(label), type); }//-V524
void jz(const void *addr) { opJmpAbs(addr, T_NEAR, 0x74, 0x84, 0x0F); }//-V524
void jz(std::string label, LabelType type = T_AUTO) { opJmp(label, type, 0x74, 0x84, 0x0F); }//-V524
void lahf() { db(0x9F); }
void lddqu(const Xmm& xmm, const Address& addr) { db(0xF2); opModM(addr, xmm, 0x0F, 0xF0); }
void ldmxcsr(const Address& addr) { opModM(addr, Reg32(2), 0x0F, 0xAE); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void lea(const Reg& reg, const Address& addr) { if (!reg.isBit(16 | i32e)) XBYAK_THROW(ERR_BAD_SIZE_OF_REGISTER) opModM(addr, reg, 0x8D); }
void leave() { db(0xC9); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void lfence() { db(0x0F); db(0xAE); db(0xE8); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void lfs(const Reg& reg, const Address& addr) { opLoadSeg(addr, reg, 0x0F, 0xB4); }
void lgs(const Reg& reg, const Address& addr) { opLoadSeg(addr, reg, 0x0F, 0xB5); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void lock() { db(0xF0); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void lodsb() { db(0xAC); }
void lodsd() { db(0xAD); }
void lodsw() { db(0x66); db(0xAD); }
void loop(const Label& label) { opJmp(label, T_SHORT, 0xE2, 0, 0); }
void loop(const char *label) { loop(std::string(label)); }
void loop(std::string label) { opJmp(label, T_SHORT, 0xE2, 0, 0); }
void loope(const Label& label) { opJmp(label, T_SHORT, 0xE1, 0, 0); }
void loope(const char *label) { loope(std::string(label)); }
void loope(std::string label) { opJmp(label, T_SHORT, 0xE1, 0, 0); }
void loopne(const Label& label) { opJmp(label, T_SHORT, 0xE0, 0, 0); }
void loopne(const char *label) { loopne(std::string(label)); }
void loopne(std::string label) { opJmp(label, T_SHORT, 0xE0, 0, 0); }
void lss(const Reg& reg, const Address& addr) { opLoadSeg(addr, reg, 0x0F, 0xB2); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void lzcnt(const Reg&reg, const Operand& op) { opSp1(reg, op, 0xF3, 0x0F, 0xBD); }
void maskmovdqu(const Xmm& reg1, const Xmm& reg2) { db(0x66); opModR(reg1, reg2, 0x0F, 0xF7); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void maskmovq(const Mmx& reg1, const Mmx& reg2) { if (!reg1.isMMX() || !reg2.isMMX()) XBYAK_THROW(ERR_BAD_COMBINATION) opModR(reg1, reg2, 0x0F, 0xF7); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void maxpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5F, 0x66, isXMM_XMMorMEM); }
void maxps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5F, 0x100, isXMM_XMMorMEM); }
void maxsd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5F, 0xF2, isXMM_XMMorMEM); }
void maxss(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5F, 0xF3, isXMM_XMMorMEM); }
void mfence() { db(0x0F); db(0xAE); db(0xF0); }
void minpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5D, 0x66, isXMM_XMMorMEM); }
void minps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5D, 0x100, isXMM_XMMorMEM); }
void minsd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5D, 0xF2, isXMM_XMMorMEM); }
void minss(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5D, 0xF3, isXMM_XMMorMEM); }
void monitor() { db(0x0F); db(0x01); db(0xC8); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void monitorx() { db(0x0F); db(0x01); db(0xFA); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void movapd(const Address& addr, const Xmm& xmm) { db(0x66); opModM(addr, xmm, 0x0F, 0x29); }
void movapd(const Xmm& xmm, const Operand& op) { opMMX(xmm, op, 0x28, 0x66); }
void movaps(const Address& addr, const Xmm& xmm) { opModM(addr, xmm, 0x0F, 0x29); }
void movaps(const Xmm& xmm, const Operand& op) { opMMX(xmm, op, 0x28, 0x100); }
void movbe(const Address& addr, const Reg& reg) { opModM(addr, reg, 0x0F, 0x38, 0xF1); }
void movbe(const Reg& reg, const Address& addr) { opModM(addr, reg, 0x0F, 0x38, 0xF0); }
void movd(const Address& addr, const Mmx& mmx) { if (mmx.isXMM()) db(0x66); opModM(addr, mmx, 0x0F, 0x7E); }
void movd(const Mmx& mmx, const Address& addr) { if (mmx.isXMM()) db(0x66); opModM(addr, mmx, 0x0F, 0x6E); }
void movd(const Mmx& mmx, const Reg32& reg) { if (mmx.isXMM()) db(0x66); opModR(mmx, reg, 0x0F, 0x6E); }
void movd(const Reg32& reg, const Mmx& mmx) { if (mmx.isXMM()) db(0x66); opModR(mmx, reg, 0x0F, 0x7E); }
void movddup(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x12, 0xF2, isXMM_XMMorMEM, NONE, NONE); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void movdir64b(const Reg& reg, const Address& addr) { db(0x66); opModM(addr, reg.cvt32(), 0x0F, 0x38, 0xF8); }
void movdiri(const Address& addr, const Reg32e& reg) { opModM(addr, reg, 0x0F, 0x38, 0xF9); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void movdq2q(const Mmx& mmx, const Xmm& xmm) { db(0xF2); opModR(mmx, xmm, 0x0F, 0xD6); }
void movdqa(const Address& addr, const Xmm& xmm) { db(0x66); opModM(addr, xmm, 0x0F, 0x7F); }
void movdqa(const Xmm& xmm, const Operand& op) { opMMX(xmm, op, 0x6F, 0x66); }
void movdqu(const Address& addr, const Xmm& xmm) { db(0xF3); opModM(addr, xmm, 0x0F, 0x7F); }
void movdqu(const Xmm& xmm, const Operand& op) { opMMX(xmm, op, 0x6F, 0xF3); }
void movhlps(const Xmm& reg1, const Xmm& reg2) { opModR(reg1, reg2, 0x0F, 0x12); }
void movhpd(const Operand& op1, const Operand& op2) { opMovXMM(op1, op2, 0x16, 0x66); }
void movhps(const Operand& op1, const Operand& op2) { opMovXMM(op1, op2, 0x16, 0x100); }
void movlhps(const Xmm& reg1, const Xmm& reg2) { opModR(reg1, reg2, 0x0F, 0x16); }
void movlpd(const Operand& op1, const Operand& op2) { opMovXMM(op1, op2, 0x12, 0x66); }
void movlps(const Operand& op1, const Operand& op2) { opMovXMM(op1, op2, 0x12, 0x100); }
void movmskpd(const Reg32e& reg, const Xmm& xmm) { db(0x66); movmskps(reg, xmm); }
void movmskps(const Reg32e& reg, const Xmm& xmm) { opModR(reg, xmm, 0x0F, 0x50); }
void movntdq(const Address& addr, const Xmm& reg) { opModM(addr, Reg16(reg.getIdx()), 0x0F, 0xE7); }
void movntdqa(const Xmm& xmm, const Address& addr) { db(0x66); opModM(addr, xmm, 0x0F, 0x38, 0x2A); }
void movnti(const Address& addr, const Reg32e& reg) { opModM(addr, reg, 0x0F, 0xC3); }
void movntpd(const Address& addr, const Xmm& reg) { opModM(addr, Reg16(reg.getIdx()), 0x0F, 0x2B); }
void movntps(const Address& addr, const Xmm& xmm) { opModM(addr, Mmx(xmm.getIdx()), 0x0F, 0x2B); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void movntq(const Address& addr, const Mmx& mmx) { if (!mmx.isMMX()) XBYAK_THROW(ERR_BAD_COMBINATION) opModM(addr, mmx, 0x0F, 0xE7); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void movq(const Address& addr, const Mmx& mmx) { if (mmx.isXMM()) db(0x66); opModM(addr, mmx, 0x0F, mmx.isXMM() ? 0xD6 : 0x7F); }
void movq(const Mmx& mmx, const Operand& op) { if (mmx.isXMM()) db(0xF3); opModRM(mmx, op, (mmx.getKind() == op.getKind()), op.isMEM(), 0x0F, mmx.isXMM() ? 0x7E : 0x6F); }
void movq2dq(const Xmm& xmm, const Mmx& mmx) { db(0xF3); opModR(xmm, mmx, 0x0F, 0xD6); }
void movsb() { db(0xA4); }
void movsd() { db(0xA5); }
void movsd(const Address& addr, const Xmm& xmm) { db(0xF2); opModM(addr, xmm, 0x0F, 0x11); }
void movsd(const Xmm& xmm, const Operand& op) { opMMX(xmm, op, 0x10, 0xF2); }
void movshdup(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x16, 0xF3, isXMM_XMMorMEM, NONE, NONE); }
void movsldup(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x12, 0xF3, isXMM_XMMorMEM, NONE, NONE); }
void movss(const Address& addr, const Xmm& xmm) { db(0xF3); opModM(addr, xmm, 0x0F, 0x11); }
void movss(const Xmm& xmm, const Operand& op) { opMMX(xmm, op, 0x10, 0xF3); }
void movsw() { db(0x66); db(0xA5); }
void movsx(const Reg& reg, const Operand& op) { opMovxx(reg, op, 0xBE); }
void movupd(const Address& addr, const Xmm& xmm) { db(0x66); opModM(addr, xmm, 0x0F, 0x11); }
void movupd(const Xmm& xmm, const Operand& op) { opMMX(xmm, op, 0x10, 0x66); }
void movups(const Address& addr, const Xmm& xmm) { opModM(addr, xmm, 0x0F, 0x11); }
void movups(const Xmm& xmm, const Operand& op) { opMMX(xmm, op, 0x10, 0x100); }
void movzx(const Reg& reg, const Operand& op) { opMovxx(reg, op, 0xB6); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void mpsadbw(const Xmm& xmm, const Operand& op, int imm) { opGen(xmm, op, 0x42, 0x66, isXMM_XMMorMEM, static_cast<uint8_t>(imm), 0x3A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void mul(const Operand& op) { opR_ModM(op, 0, 4, 0xF6); }
void mulpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x59, 0x66, isXMM_XMMorMEM); }
void mulps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x59, 0x100, isXMM_XMMorMEM); }
void mulsd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x59, 0xF2, isXMM_XMMorMEM); }
void mulss(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x59, 0xF3, isXMM_XMMorMEM); }
void mulx(const Reg32e& r1, const Reg32e& r2, const Operand& op) { opGpr(r1, r2, op, T_F2 | T_0F38, 0xf6, true); }
void mwait() { db(0x0F); db(0x01); db(0xC9); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void mwaitx() { db(0x0F); db(0x01); db(0xFB); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void neg(const Operand& op) { opR_ModM(op, 0, 3, 0xF6); }
void not_(const Operand& op) { opR_ModM(op, 0, 2, 0xF6); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void or_(const Operand& op, uint32_t imm) { opRM_I(op, imm, 0x08, 1); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void or_(const Operand& op1, const Operand& op2) { opRM_RM(op1, op2, 0x08); }
void orpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x56, 0x66, isXMM_XMMorMEM); }
void orps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x56, 0x100, isXMM_XMMorMEM); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void out_(const Reg& d, const Reg& a) { opInOut(a, d, 0xEE); }
void out_(uint8_t v, const Reg& a) { opInOut(a, 0xE6, v); }
void outsb() { db(0x6E); }
void outsd() { db(0x6F); }
void outsw() { db(0x66); db(0x6F); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void pabsb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x1C, 0x66, NONE, 0x38); }
void pabsd(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x1E, 0x66, NONE, 0x38); }
void pabsw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x1D, 0x66, NONE, 0x38); }
void packssdw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x6B); }
void packsswb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x63); }
void packusdw(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x2B, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void packuswb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x67); }
void paddb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xFC); }
void paddd(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xFE); }
void paddq(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xD4); }
void paddsb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xEC); }
void paddsw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xED); }
void paddusb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xDC); }
void paddusw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xDD); }
void paddw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xFD); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void palignr(const Mmx& mmx, const Operand& op, int imm) { opMMX(mmx, op, 0x0f, 0x66, static_cast<uint8_t>(imm), 0x3a); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void pand(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xDB); }
void pandn(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xDF); }
void pause() { db(0xF3); db(0x90); }
void pavgb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xE0); }
void pavgw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xE3); }
void pblendvb(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x10, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void pblendw(const Xmm& xmm, const Operand& op, int imm) { opGen(xmm, op, 0x0E, 0x66, isXMM_XMMorMEM, static_cast<uint8_t>(imm), 0x3A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void pclmulhqhdq(const Xmm& xmm, const Operand& op) { pclmulqdq(xmm, op, 0x11); }
void pclmulhqlqdq(const Xmm& xmm, const Operand& op) { pclmulqdq(xmm, op, 0x01); }
void pclmullqhdq(const Xmm& xmm, const Operand& op) { pclmulqdq(xmm, op, 0x10); }
void pclmullqlqdq(const Xmm& xmm, const Operand& op) { pclmulqdq(xmm, op, 0x00); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void pclmulqdq(const Xmm& xmm, const Operand& op, int imm) { opGen(xmm, op, 0x44, 0x66, isXMM_XMMorMEM, static_cast<uint8_t>(imm), 0x3A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void pcmpeqb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x74); }
void pcmpeqd(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x76); }
void pcmpeqq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x29, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pcmpeqw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x75); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void pcmpestri(const Xmm& xmm, const Operand& op, uint8_t imm) { opGen(xmm, op, 0x61, 0x66, isXMM_XMMorMEM, imm, 0x3A); }
void pcmpestrm(const Xmm& xmm, const Operand& op, uint8_t imm) { opGen(xmm, op, 0x60, 0x66, isXMM_XMMorMEM, imm, 0x3A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void pcmpgtb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x64); }
void pcmpgtd(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x66); }
void pcmpgtq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x37, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pcmpgtw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x65); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void pcmpistri(const Xmm& xmm, const Operand& op, uint8_t imm) { opGen(xmm, op, 0x63, 0x66, isXMM_XMMorMEM, imm, 0x3A); }
void pcmpistrm(const Xmm& xmm, const Operand& op, uint8_t imm) { opGen(xmm, op, 0x62, 0x66, isXMM_XMMorMEM, imm, 0x3A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void pdep(const Reg32e& r1, const Reg32e& r2, const Operand& op) { opGpr(r1, r2, op, T_F2 | T_0F38, 0xf5, true); }
void pext(const Reg32e& r1, const Reg32e& r2, const Operand& op) { opGpr(r1, r2, op, T_F3 | T_0F38, 0xf5, true); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void pextrb(const Operand& op, const Xmm& xmm, uint8_t imm) { opExt(op, xmm, 0x14, imm); }
void pextrd(const Operand& op, const Xmm& xmm, uint8_t imm) { opExt(op, xmm, 0x16, imm); }
void pextrw(const Operand& op, const Mmx& xmm, uint8_t imm) { opExt(op, xmm, 0x15, imm, true); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void phaddd(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x02, 0x66, NONE, 0x38); }
void phaddsw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x03, 0x66, NONE, 0x38); }
void phaddw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x01, 0x66, NONE, 0x38); }
void phminposuw(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x41, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void phsubd(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x06, 0x66, NONE, 0x38); }
void phsubsw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x07, 0x66, NONE, 0x38); }
void phsubw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x05, 0x66, NONE, 0x38); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void pinsrb(const Xmm& xmm, const Operand& op, uint8_t imm) { opGen(xmm, op, 0x20, 0x66, isXMM_REG32orMEM, imm, 0x3A); }
void pinsrd(const Xmm& xmm, const Operand& op, uint8_t imm) { opGen(xmm, op, 0x22, 0x66, isXMM_REG32orMEM, imm, 0x3A); }
void pinsrw(const Mmx& mmx, const Operand& op, int imm) { if (!op.isREG(32) && !op.isMEM()) XBYAK_THROW(ERR_BAD_COMBINATION) opGen(mmx, op, 0xC4, mmx.isXMM() ? 0x66 : NONE, 0, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void pmaddubsw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x04, 0x66, NONE, 0x38); }
void pmaddwd(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xF5); }
void pmaxsb(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x3C, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmaxsd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x3D, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmaxsw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xEE); }
void pmaxub(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xDE); }
void pmaxud(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x3F, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmaxuw(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x3E, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pminsb(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x38, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pminsd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x39, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pminsw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xEA); }
void pminub(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xDA); }
void pminud(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x3B, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pminuw(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x3A, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmovmskb(const Reg32e& reg, const Mmx& mmx) { if (mmx.isXMM()) db(0x66); opModR(reg, mmx, 0x0F, 0xD7); }
void pmovsxbd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x21, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmovsxbq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x22, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmovsxbw(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x20, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmovsxdq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x25, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmovsxwd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x23, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmovsxwq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x24, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmovzxbd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x31, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmovzxbq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x32, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmovzxbw(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x30, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmovzxdq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x35, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmovzxwd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x33, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmovzxwq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x34, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmuldq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x28, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmulhrsw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x0B, 0x66, NONE, 0x38); }
void pmulhuw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xE4); }
void pmulhw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xE5); }
void pmulld(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x40, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void pmullw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xD5); }
void pmuludq(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xF4); }
void popcnt(const Reg&reg, const Operand& op) { opSp1(reg, op, 0xF3, 0x0F, 0xB8); }
void popf() { db(0x9D); }
void por(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xEB); }
Squashed 'externals/xbyak/' changes from 88f2f771f..a1ac3750f a1ac3750f Merge branch 'dev' 5f4ba971f v6.68 cac2c175f update doc 1b08a8248 add test of prefetchiti bef70d9b1 add prefetchit{0,1} f66785876 add detection of prefetchiti 124617ac9 Merge branch 'dev' 1bce8be17 Merge commit 'fbb18f6' into dev bebfe64a3 [test] X32 is optional and remove CFLAGS d8c332cef Merge branch 'Tachi107-small-changes' into dev fbb18f69d ci: use containers 8a974696c test: simplify makefile 1efe9fe7c Merge branch 'dev' bb70083e6 v6.67 436e452d7 update doc 47ff6ef42 I get an error after all on GitHub action 445c0dcec add test of CMPccXADD c9347907d add CMPccXADD dc792cc56 add detection of cmpccxadd 62be84cc8 fix detection of boost b5ac7b0f0 Merge branch 'dev' 05dd400e0 recover a removed line 94eff6246 v6.66 ec3fadeba update doc 00bfaaa7f add test of rao-int 0f2f1aaa6 support rio-int 6c047f480 detect rao-int f07c5c255 fix args of EmptyAllocator::alloc in test 04d3eb5f5 stop if a test script causes an error 1c1d2366f disable boost sample cfb1127c3 tweak 5fcbeb7c4 [sample] change the way of detection of boost bafc1ee60 CXX uses g++ as default value d8cabc6cb remove warning of blace bc73a0816 remove -fno-operator-names option 6989aea94 use CXX instead of g++ a7c5a1bd7 use English 1bfbd8c4d Fix incorrect format strings 0ecef5c28 Merge branch 'dev' 7556c20ba update doc a15709271 v6.65 3b83aab3e add detect_x32 to TARGET 8c64bbbc3 use gcc instead of dpkg for portability 5e9a9b96f test_avx512.sh runs on x32 8ae01b0c2 disable some tests on x32 83b3da217 x32 does not check large disp 693ab8c9d sizeof(void*) = 4 on x32, so disable the test 348e3e548 Merge branch 'dev' 11b9c4dc0 v6.64 459636196 add T_M_K flag to vpmov* 1d3722928 add include path 5e27eddae move CrearError from LabelManager::reset() to CodeGenerator::reset() f8ea5c28d Merge branch 'dev' 20b2b1eae v6.63 3706869f8 desc. of setDefaultEncoding d6f2d7577 add test of setDefaultEncoding 3b0a19c41 vpmadd52{h,l}uq for avx-ifma 95752ebd7 add tAVX_IFMA cd36e31ea [sample] show AMX_FP16/AVX_VNNI_INT8/AVX_NE_CONVERT e5858af27 add setDefaultEncoding 2f7fb0220 modify gen.cpp for AVX-NE-CONVERT/AVX-VNNI-INT8/AMX-FP16 1c5cb7efa add AVX-NE-CONVERT instructions 564fe9acd add AVX-VNNI-INT8 instructions cd14d07b1 add AMX-FP16 instruction 7811f593c Merge pull request #161 from scribam/patch-1 2218f6c08 Update changelog.md 7bccdbbb4 Merge pull request #160 from herumi/dev 5fcf87596 compile nasm-2.15 instead of apt install e31961ea8 v6.62 ff4f9e65c update doc 9c8fb81db disable wrong detection of gcc -Warray-bounds 6b7519659 add serialize e16582696 Merge branch 'JonLiu1993-vcpkg-installition' into dev edbb410fb Add vcpkg installation instructions 055d31242 Merge branch 'dev' 21ab98441 Merge branch 'akodanka-enable_CIV' into dev ed4d598e3 Changes to compile xbyak project for openvino b652430c4 mingw uses __cpuidex 48457bfa0 Merge branch 'dev' 29cb524d1 v6.61.2 1a9a0b0e1 avoid including algorithm header in xbyak_util.h 6fadefd04 Merge branch 'dev' fc1c18a9d update doc f7cae7f11 v6.61.1 6f5ec5cf3 Merge branch 'doyaGu-master' into dev 4554d6bb9 Fix error related to XBYAK_NOEXCEPT git-subtree-dir: externals/xbyak git-subtree-split: a1ac3750f9a639b5a6c6d6c7da4259b8d6790989
2022-12-30 23:05:02 +00:00
void prefetchit0(const Address& addr) { opModM(addr, Reg32(7), 0x0F, 0x18); }
void prefetchit1(const Address& addr) { opModM(addr, Reg32(6), 0x0F, 0x18); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void prefetchnta(const Address& addr) { opModM(addr, Reg32(0), 0x0F, 0x18); }
void prefetcht0(const Address& addr) { opModM(addr, Reg32(1), 0x0F, 0x18); }
void prefetcht1(const Address& addr) { opModM(addr, Reg32(2), 0x0F, 0x18); }
void prefetcht2(const Address& addr) { opModM(addr, Reg32(3), 0x0F, 0x18); }
void prefetchw(const Address& addr) { opModM(addr, Reg32(1), 0x0F, 0x0D); }
void prefetchwt1(const Address& addr) { opModM(addr, Reg32(2), 0x0F, 0x0D); }
void psadbw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xF6); }
void pshufb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x00, 0x66, NONE, 0x38); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void pshufd(const Mmx& mmx, const Operand& op, uint8_t imm8) { opMMX(mmx, op, 0x70, 0x66, imm8); }
void pshufhw(const Mmx& mmx, const Operand& op, uint8_t imm8) { opMMX(mmx, op, 0x70, 0xF3, imm8); }
void pshuflw(const Mmx& mmx, const Operand& op, uint8_t imm8) { opMMX(mmx, op, 0x70, 0xF2, imm8); }
void pshufw(const Mmx& mmx, const Operand& op, uint8_t imm8) { opMMX(mmx, op, 0x70, 0x00, imm8); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void psignb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x08, 0x66, NONE, 0x38); }
void psignd(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x0A, 0x66, NONE, 0x38); }
void psignw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x09, 0x66, NONE, 0x38); }
void pslld(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xF2); }
void pslld(const Mmx& mmx, int imm8) { opMMX_IMM(mmx, imm8, 0x72, 6); }
void pslldq(const Xmm& xmm, int imm8) { opMMX_IMM(xmm, imm8, 0x73, 7); }
void psllq(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xF3); }
void psllq(const Mmx& mmx, int imm8) { opMMX_IMM(mmx, imm8, 0x73, 6); }
void psllw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xF1); }
void psllw(const Mmx& mmx, int imm8) { opMMX_IMM(mmx, imm8, 0x71, 6); }
void psrad(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xE2); }
void psrad(const Mmx& mmx, int imm8) { opMMX_IMM(mmx, imm8, 0x72, 4); }
void psraw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xE1); }
void psraw(const Mmx& mmx, int imm8) { opMMX_IMM(mmx, imm8, 0x71, 4); }
void psrld(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xD2); }
void psrld(const Mmx& mmx, int imm8) { opMMX_IMM(mmx, imm8, 0x72, 2); }
void psrldq(const Xmm& xmm, int imm8) { opMMX_IMM(xmm, imm8, 0x73, 3); }
void psrlq(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xD3); }
void psrlq(const Mmx& mmx, int imm8) { opMMX_IMM(mmx, imm8, 0x73, 2); }
void psrlw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xD1); }
void psrlw(const Mmx& mmx, int imm8) { opMMX_IMM(mmx, imm8, 0x71, 2); }
void psubb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xF8); }
void psubd(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xFA); }
void psubq(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xFB); }
void psubsb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xE8); }
void psubsw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xE9); }
void psubusb(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xD8); }
void psubusw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xD9); }
void psubw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xF9); }
void ptest(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x17, 0x66, isXMM_XMMorMEM, NONE, 0x38); }
void punpckhbw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x68); }
void punpckhdq(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x6A); }
void punpckhqdq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x6D, 0x66, isXMM_XMMorMEM); }
void punpckhwd(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x69); }
void punpcklbw(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x60); }
void punpckldq(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x62); }
void punpcklqdq(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x6C, 0x66, isXMM_XMMorMEM); }
void punpcklwd(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0x61); }
void pushf() { db(0x9C); }
void pxor(const Mmx& mmx, const Operand& op) { opMMX(mmx, op, 0xEF); }
void rcl(const Operand& op, const Reg8& _cl) { opShift(op, _cl, 2); }
void rcl(const Operand& op, int imm) { opShift(op, imm, 2); }
void rcpps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x53, 0x100, isXMM_XMMorMEM); }
void rcpss(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x53, 0xF3, isXMM_XMMorMEM); }
void rcr(const Operand& op, const Reg8& _cl) { opShift(op, _cl, 3); }
void rcr(const Operand& op, int imm) { opShift(op, imm, 3); }
void rdmsr() { db(0x0F); db(0x32); }
void rdpmc() { db(0x0F); db(0x33); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void rdrand(const Reg& r) { if (r.isBit(8)) XBYAK_THROW(ERR_BAD_SIZE_OF_REGISTER) opModR(Reg(6, Operand::REG, r.getBit()), r, 0x0F, 0xC7); }
void rdseed(const Reg& r) { if (r.isBit(8)) XBYAK_THROW(ERR_BAD_SIZE_OF_REGISTER) opModR(Reg(7, Operand::REG, r.getBit()), r, 0x0F, 0xC7); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void rdtsc() { db(0x0F); db(0x31); }
void rdtscp() { db(0x0F); db(0x01); db(0xF9); }
void rep() { db(0xF3); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void repe() { db(0xF3); }
void repne() { db(0xF2); }
void repnz() { db(0xF2); }
void repz() { db(0xF3); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void ret(int imm = 0) { if (imm) { db(0xC2); dw(imm); } else { db(0xC3); } }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void retf(int imm = 0) { if (imm) { db(0xCA); dw(imm); } else { db(0xCB); } }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void rol(const Operand& op, const Reg8& _cl) { opShift(op, _cl, 0); }
void rol(const Operand& op, int imm) { opShift(op, imm, 0); }
void ror(const Operand& op, const Reg8& _cl) { opShift(op, _cl, 1); }
void ror(const Operand& op, int imm) { opShift(op, imm, 1); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void rorx(const Reg32e& r, const Operand& op, uint8_t imm) { opGpr(r, op, Reg32e(0, r.getBit()), T_0F3A | T_F2, 0xF0, false, imm); }
void roundpd(const Xmm& xmm, const Operand& op, uint8_t imm) { opGen(xmm, op, 0x09, 0x66, isXMM_XMMorMEM, imm, 0x3A); }
void roundps(const Xmm& xmm, const Operand& op, uint8_t imm) { opGen(xmm, op, 0x08, 0x66, isXMM_XMMorMEM, imm, 0x3A); }
void roundsd(const Xmm& xmm, const Operand& op, int imm) { opGen(xmm, op, 0x0B, 0x66, isXMM_XMMorMEM, static_cast<uint8_t>(imm), 0x3A); }
void roundss(const Xmm& xmm, const Operand& op, int imm) { opGen(xmm, op, 0x0A, 0x66, isXMM_XMMorMEM, static_cast<uint8_t>(imm), 0x3A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void rsqrtps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x52, 0x100, isXMM_XMMorMEM); }
void rsqrtss(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x52, 0xF3, isXMM_XMMorMEM); }
void sahf() { db(0x9E); }
void sal(const Operand& op, const Reg8& _cl) { opShift(op, _cl, 4); }
void sal(const Operand& op, int imm) { opShift(op, imm, 4); }
void sar(const Operand& op, const Reg8& _cl) { opShift(op, _cl, 7); }
void sar(const Operand& op, int imm) { opShift(op, imm, 7); }
void sarx(const Reg32e& r1, const Operand& op, const Reg32e& r2) { opGpr(r1, op, r2, T_F3 | T_0F38, 0xf7, false); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void sbb(const Operand& op, uint32_t imm) { opRM_I(op, imm, 0x18, 3); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void sbb(const Operand& op1, const Operand& op2) { opRM_RM(op1, op2, 0x18); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void scasb() { db(0xAE); }
void scasd() { db(0xAF); }
void scasw() { db(0x66); db(0xAF); }
Squashed 'externals/xbyak/' changes from 88f2f771f..a1ac3750f a1ac3750f Merge branch 'dev' 5f4ba971f v6.68 cac2c175f update doc 1b08a8248 add test of prefetchiti bef70d9b1 add prefetchit{0,1} f66785876 add detection of prefetchiti 124617ac9 Merge branch 'dev' 1bce8be17 Merge commit 'fbb18f6' into dev bebfe64a3 [test] X32 is optional and remove CFLAGS d8c332cef Merge branch 'Tachi107-small-changes' into dev fbb18f69d ci: use containers 8a974696c test: simplify makefile 1efe9fe7c Merge branch 'dev' bb70083e6 v6.67 436e452d7 update doc 47ff6ef42 I get an error after all on GitHub action 445c0dcec add test of CMPccXADD c9347907d add CMPccXADD dc792cc56 add detection of cmpccxadd 62be84cc8 fix detection of boost b5ac7b0f0 Merge branch 'dev' 05dd400e0 recover a removed line 94eff6246 v6.66 ec3fadeba update doc 00bfaaa7f add test of rao-int 0f2f1aaa6 support rio-int 6c047f480 detect rao-int f07c5c255 fix args of EmptyAllocator::alloc in test 04d3eb5f5 stop if a test script causes an error 1c1d2366f disable boost sample cfb1127c3 tweak 5fcbeb7c4 [sample] change the way of detection of boost bafc1ee60 CXX uses g++ as default value d8cabc6cb remove warning of blace bc73a0816 remove -fno-operator-names option 6989aea94 use CXX instead of g++ a7c5a1bd7 use English 1bfbd8c4d Fix incorrect format strings 0ecef5c28 Merge branch 'dev' 7556c20ba update doc a15709271 v6.65 3b83aab3e add detect_x32 to TARGET 8c64bbbc3 use gcc instead of dpkg for portability 5e9a9b96f test_avx512.sh runs on x32 8ae01b0c2 disable some tests on x32 83b3da217 x32 does not check large disp 693ab8c9d sizeof(void*) = 4 on x32, so disable the test 348e3e548 Merge branch 'dev' 11b9c4dc0 v6.64 459636196 add T_M_K flag to vpmov* 1d3722928 add include path 5e27eddae move CrearError from LabelManager::reset() to CodeGenerator::reset() f8ea5c28d Merge branch 'dev' 20b2b1eae v6.63 3706869f8 desc. of setDefaultEncoding d6f2d7577 add test of setDefaultEncoding 3b0a19c41 vpmadd52{h,l}uq for avx-ifma 95752ebd7 add tAVX_IFMA cd36e31ea [sample] show AMX_FP16/AVX_VNNI_INT8/AVX_NE_CONVERT e5858af27 add setDefaultEncoding 2f7fb0220 modify gen.cpp for AVX-NE-CONVERT/AVX-VNNI-INT8/AMX-FP16 1c5cb7efa add AVX-NE-CONVERT instructions 564fe9acd add AVX-VNNI-INT8 instructions cd14d07b1 add AMX-FP16 instruction 7811f593c Merge pull request #161 from scribam/patch-1 2218f6c08 Update changelog.md 7bccdbbb4 Merge pull request #160 from herumi/dev 5fcf87596 compile nasm-2.15 instead of apt install e31961ea8 v6.62 ff4f9e65c update doc 9c8fb81db disable wrong detection of gcc -Warray-bounds 6b7519659 add serialize e16582696 Merge branch 'JonLiu1993-vcpkg-installition' into dev edbb410fb Add vcpkg installation instructions 055d31242 Merge branch 'dev' 21ab98441 Merge branch 'akodanka-enable_CIV' into dev ed4d598e3 Changes to compile xbyak project for openvino b652430c4 mingw uses __cpuidex 48457bfa0 Merge branch 'dev' 29cb524d1 v6.61.2 1a9a0b0e1 avoid including algorithm header in xbyak_util.h 6fadefd04 Merge branch 'dev' fc1c18a9d update doc f7cae7f11 v6.61.1 6f5ec5cf3 Merge branch 'doyaGu-master' into dev 4554d6bb9 Fix error related to XBYAK_NOEXCEPT git-subtree-dir: externals/xbyak git-subtree-split: a1ac3750f9a639b5a6c6d6c7da4259b8d6790989
2022-12-30 23:05:02 +00:00
void serialize() { db(0x0F); db(0x01); db(0xE8); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void seta(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 7); }//-V524
void setae(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 3); }//-V524
void setb(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 2); }//-V524
void setbe(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 6); }//-V524
void setc(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 2); }//-V524
void sete(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 4); }//-V524
void setg(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 15); }//-V524
void setge(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 13); }//-V524
void setl(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 12); }//-V524
void setle(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 14); }//-V524
void setna(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 6); }//-V524
void setnae(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 2); }//-V524
void setnb(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 3); }//-V524
void setnbe(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 7); }//-V524
void setnc(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 3); }//-V524
void setne(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 5); }//-V524
void setng(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 14); }//-V524
void setnge(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 12); }//-V524
void setnl(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 13); }//-V524
void setnle(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 15); }//-V524
void setno(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 1); }//-V524
void setnp(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 11); }//-V524
void setns(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 9); }//-V524
void setnz(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 5); }//-V524
void seto(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 0); }//-V524
void setp(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 10); }//-V524
void setpe(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 10); }//-V524
void setpo(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 11); }//-V524
void sets(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 8); }//-V524
void setz(const Operand& op) { opR_ModM(op, 8, 0, 0x0F, 0x90 | 4); }//-V524
void sfence() { db(0x0F); db(0xAE); db(0xF8); }
void sha1msg1(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xC9, NONE, isXMM_XMMorMEM, NONE, 0x38); }
void sha1msg2(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xCA, NONE, isXMM_XMMorMEM, NONE, 0x38); }
void sha1nexte(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xC8, NONE, isXMM_XMMorMEM, NONE, 0x38); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void sha1rnds4(const Xmm& xmm, const Operand& op, uint8_t imm) { opGen(xmm, op, 0xCC, NONE, isXMM_XMMorMEM, imm, 0x3A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void sha256msg1(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xCC, NONE, isXMM_XMMorMEM, NONE, 0x38); }
void sha256msg2(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xCD, NONE, isXMM_XMMorMEM, NONE, 0x38); }
void sha256rnds2(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0xCB, NONE, isXMM_XMMorMEM, NONE, 0x38); }
void shl(const Operand& op, const Reg8& _cl) { opShift(op, _cl, 4); }
void shl(const Operand& op, int imm) { opShift(op, imm, 4); }
void shld(const Operand& op, const Reg& reg, const Reg8& _cl) { opShxd(op, reg, 0, 0xA4, &_cl); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void shld(const Operand& op, const Reg& reg, uint8_t imm) { opShxd(op, reg, imm, 0xA4); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void shlx(const Reg32e& r1, const Operand& op, const Reg32e& r2) { opGpr(r1, op, r2, T_66 | T_0F38, 0xf7, false); }
void shr(const Operand& op, const Reg8& _cl) { opShift(op, _cl, 5); }
void shr(const Operand& op, int imm) { opShift(op, imm, 5); }
void shrd(const Operand& op, const Reg& reg, const Reg8& _cl) { opShxd(op, reg, 0, 0xAC, &_cl); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void shrd(const Operand& op, const Reg& reg, uint8_t imm) { opShxd(op, reg, imm, 0xAC); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void shrx(const Reg32e& r1, const Operand& op, const Reg32e& r2) { opGpr(r1, op, r2, T_F2 | T_0F38, 0xf7, false); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void shufpd(const Xmm& xmm, const Operand& op, uint8_t imm8) { opGen(xmm, op, 0xC6, 0x66, isXMM_XMMorMEM, imm8); }
void shufps(const Xmm& xmm, const Operand& op, uint8_t imm8) { opGen(xmm, op, 0xC6, 0x100, isXMM_XMMorMEM, imm8); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void sqrtpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x51, 0x66, isXMM_XMMorMEM); }
void sqrtps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x51, 0x100, isXMM_XMMorMEM); }
void sqrtsd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x51, 0xF2, isXMM_XMMorMEM); }
void sqrtss(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x51, 0xF3, isXMM_XMMorMEM); }
void stac() { db(0x0F); db(0x01); db(0xCB); }
void stc() { db(0xF9); }
void std() { db(0xFD); }
void sti() { db(0xFB); }
void stmxcsr(const Address& addr) { opModM(addr, Reg32(3), 0x0F, 0xAE); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void stosb() { db(0xAA); }
void stosd() { db(0xAB); }
void stosw() { db(0x66); db(0xAB); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void sub(const Operand& op, uint32_t imm) { opRM_I(op, imm, 0x28, 5); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void sub(const Operand& op1, const Operand& op2) { opRM_RM(op1, op2, 0x28); }
void subpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5C, 0x66, isXMM_XMMorMEM); }
void subps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5C, 0x100, isXMM_XMMorMEM); }
void subsd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5C, 0xF2, isXMM_XMMorMEM); }
void subss(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x5C, 0xF3, isXMM_XMMorMEM); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void sysenter() { db(0x0F); db(0x34); }
void sysexit() { db(0x0F); db(0x35); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void tpause(const Reg32& r) { int idx = r.getIdx(); if (idx > 7) XBYAK_THROW(ERR_BAD_PARAMETER) db(0x66); db(0x0F); db(0xAE); setModRM(3, 6, idx); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void tzcnt(const Reg&reg, const Operand& op) { opSp1(reg, op, 0xF3, 0x0F, 0xBC); }
void ucomisd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x2E, 0x66, isXMM_XMMorMEM); }
void ucomiss(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x2E, 0x100, isXMM_XMMorMEM); }
void ud2() { db(0x0F); db(0x0B); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void umonitor(const Reg& r) { int idx = r.getIdx(); if (idx > 7) XBYAK_THROW(ERR_BAD_PARAMETER) int bit = r.getBit(); if (BIT != bit) { if ((BIT == 32 && bit == 16) || (BIT == 64 && bit == 32)) { db(0x67); } else { XBYAK_THROW(ERR_BAD_SIZE_OF_REGISTER) } } db(0xF3); db(0x0F); db(0xAE); setModRM(3, 6, idx); }
void umwait(const Reg32& r) { int idx = r.getIdx(); if (idx > 7) XBYAK_THROW(ERR_BAD_PARAMETER) db(0xF2); db(0x0F); db(0xAE); setModRM(3, 6, idx); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void unpckhpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x15, 0x66, isXMM_XMMorMEM); }
void unpckhps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x15, 0x100, isXMM_XMMorMEM); }
void unpcklpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x14, 0x66, isXMM_XMMorMEM); }
void unpcklps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x14, 0x100, isXMM_XMMorMEM); }
void vaddpd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_66 | T_EW1 | T_YMM | T_EVEX | T_ER_Z | T_B64, 0x58); }
void vaddps(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_EW0 | T_YMM | T_EVEX | T_ER_Z | T_B32, 0x58); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vaddsd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_F2 | T_EW1 | T_EVEX | T_ER_X | T_N8, 0x58); }
void vaddss(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_F3 | T_EW0 | T_EVEX | T_ER_X | T_N4, 0x58); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vaddsubpd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_66 | T_0F | T_YMM, 0xD0); }
void vaddsubps(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_F2 | T_0F | T_YMM, 0xD0); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
void vaesdec(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_66 | T_0F38 | T_YMM | T_EVEX, 0xDE); }
void vaesdeclast(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_66 | T_0F38 | T_YMM | T_EVEX, 0xDF); }
void vaesenc(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_66 | T_0F38 | T_YMM | T_EVEX, 0xDC); }
void vaesenclast(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_66 | T_0F38 | T_YMM | T_EVEX, 0xDD); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vaesimc(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F38 | T_W0, 0xDB); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vaeskeygenassist(const Xmm& xm, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F3A, 0xDF, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vandnpd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_66 | T_EW1 | T_YMM | T_EVEX | T_ER_Z | T_B64, 0x55); }
void vandnps(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_EW0 | T_YMM | T_EVEX | T_ER_Z | T_B32, 0x55); }
void vandpd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_66 | T_EW1 | T_YMM | T_EVEX | T_ER_Z | T_B64, 0x54); }
void vandps(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_EW0 | T_YMM | T_EVEX | T_ER_Z | T_B32, 0x54); }
Squashed 'externals/xbyak/' changes from 88f2f771f..a1ac3750f a1ac3750f Merge branch 'dev' 5f4ba971f v6.68 cac2c175f update doc 1b08a8248 add test of prefetchiti bef70d9b1 add prefetchit{0,1} f66785876 add detection of prefetchiti 124617ac9 Merge branch 'dev' 1bce8be17 Merge commit 'fbb18f6' into dev bebfe64a3 [test] X32 is optional and remove CFLAGS d8c332cef Merge branch 'Tachi107-small-changes' into dev fbb18f69d ci: use containers 8a974696c test: simplify makefile 1efe9fe7c Merge branch 'dev' bb70083e6 v6.67 436e452d7 update doc 47ff6ef42 I get an error after all on GitHub action 445c0dcec add test of CMPccXADD c9347907d add CMPccXADD dc792cc56 add detection of cmpccxadd 62be84cc8 fix detection of boost b5ac7b0f0 Merge branch 'dev' 05dd400e0 recover a removed line 94eff6246 v6.66 ec3fadeba update doc 00bfaaa7f add test of rao-int 0f2f1aaa6 support rio-int 6c047f480 detect rao-int f07c5c255 fix args of EmptyAllocator::alloc in test 04d3eb5f5 stop if a test script causes an error 1c1d2366f disable boost sample cfb1127c3 tweak 5fcbeb7c4 [sample] change the way of detection of boost bafc1ee60 CXX uses g++ as default value d8cabc6cb remove warning of blace bc73a0816 remove -fno-operator-names option 6989aea94 use CXX instead of g++ a7c5a1bd7 use English 1bfbd8c4d Fix incorrect format strings 0ecef5c28 Merge branch 'dev' 7556c20ba update doc a15709271 v6.65 3b83aab3e add detect_x32 to TARGET 8c64bbbc3 use gcc instead of dpkg for portability 5e9a9b96f test_avx512.sh runs on x32 8ae01b0c2 disable some tests on x32 83b3da217 x32 does not check large disp 693ab8c9d sizeof(void*) = 4 on x32, so disable the test 348e3e548 Merge branch 'dev' 11b9c4dc0 v6.64 459636196 add T_M_K flag to vpmov* 1d3722928 add include path 5e27eddae move CrearError from LabelManager::reset() to CodeGenerator::reset() f8ea5c28d Merge branch 'dev' 20b2b1eae v6.63 3706869f8 desc. of setDefaultEncoding d6f2d7577 add test of setDefaultEncoding 3b0a19c41 vpmadd52{h,l}uq for avx-ifma 95752ebd7 add tAVX_IFMA cd36e31ea [sample] show AMX_FP16/AVX_VNNI_INT8/AVX_NE_CONVERT e5858af27 add setDefaultEncoding 2f7fb0220 modify gen.cpp for AVX-NE-CONVERT/AVX-VNNI-INT8/AMX-FP16 1c5cb7efa add AVX-NE-CONVERT instructions 564fe9acd add AVX-VNNI-INT8 instructions cd14d07b1 add AMX-FP16 instruction 7811f593c Merge pull request #161 from scribam/patch-1 2218f6c08 Update changelog.md 7bccdbbb4 Merge pull request #160 from herumi/dev 5fcf87596 compile nasm-2.15 instead of apt install e31961ea8 v6.62 ff4f9e65c update doc 9c8fb81db disable wrong detection of gcc -Warray-bounds 6b7519659 add serialize e16582696 Merge branch 'JonLiu1993-vcpkg-installition' into dev edbb410fb Add vcpkg installation instructions 055d31242 Merge branch 'dev' 21ab98441 Merge branch 'akodanka-enable_CIV' into dev ed4d598e3 Changes to compile xbyak project for openvino b652430c4 mingw uses __cpuidex 48457bfa0 Merge branch 'dev' 29cb524d1 v6.61.2 1a9a0b0e1 avoid including algorithm header in xbyak_util.h 6fadefd04 Merge branch 'dev' fc1c18a9d update doc f7cae7f11 v6.61.1 6f5ec5cf3 Merge branch 'doyaGu-master' into dev 4554d6bb9 Fix error related to XBYAK_NOEXCEPT git-subtree-dir: externals/xbyak git-subtree-split: a1ac3750f9a639b5a6c6d6c7da4259b8d6790989
2022-12-30 23:05:02 +00:00
void vbcstnebf162ps(const Xmm& x, const Address& addr) { opVex(x, 0, addr, T_F3 | T_0F38 | T_W0 | T_YMM | T_B16, 0xB1); }
void vbcstnesh2ps(const Xmm& x, const Address& addr) { opVex(x, 0, addr, T_66 | T_0F38 | T_W0 | T_YMM | T_B16, 0xB1); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vblendpd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_W0 | T_YMM, 0x0D, imm); }
void vblendps(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_W0 | T_YMM, 0x0C, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vblendvpd(const Xmm& x1, const Xmm& x2, const Operand& op, const Xmm& x4) { opAVX_X_X_XM(x1, x2, op, T_0F3A | T_66 | T_YMM, 0x4B, x4.getIdx() << 4); }
void vblendvps(const Xmm& x1, const Xmm& x2, const Operand& op, const Xmm& x4) { opAVX_X_X_XM(x1, x2, op, T_0F3A | T_66 | T_YMM, 0x4A, x4.getIdx() << 4); }
void vbroadcastf128(const Ymm& y, const Address& addr) { opAVX_X_XM_IMM(y, addr, T_0F38 | T_66 | T_W0 | T_YMM, 0x1A); }
void vbroadcasti128(const Ymm& y, const Address& addr) { opAVX_X_XM_IMM(y, addr, T_0F38 | T_66 | T_W0 | T_YMM, 0x5A); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vbroadcastsd(const Ymm& y, const Operand& op) { if (!op.isMEM() && !(y.isYMM() && op.isXMM()) && !(y.isZMM() && op.isXMM())) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_XM_IMM(y, op, T_0F38 | T_66 | T_W0 | T_YMM | T_EVEX | T_EW1 | T_N8, 0x19); }
void vbroadcastss(const Xmm& x, const Operand& op) { if (!(op.isXMM() || op.isMEM())) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_XM_IMM(x, op, T_N4 | T_66 | T_0F38 | T_W0 | T_YMM | T_EVEX, 0x18); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcmpeq_ospd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 16); }
void vcmpeq_osps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 16); }
void vcmpeq_ossd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 16); }
void vcmpeq_osss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 16); }
void vcmpeq_uqpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 8); }
void vcmpeq_uqps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 8); }
void vcmpeq_uqsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 8); }
void vcmpeq_uqss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 8); }
void vcmpeq_uspd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 24); }
void vcmpeq_usps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 24); }
void vcmpeq_ussd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 24); }
void vcmpeq_usss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 24); }
void vcmpeqpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 0); }
void vcmpeqps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 0); }
void vcmpeqsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 0); }
void vcmpeqss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 0); }
void vcmpfalse_ospd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 27); }
void vcmpfalse_osps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 27); }
void vcmpfalse_ossd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 27); }
void vcmpfalse_osss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 27); }
void vcmpfalsepd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 11); }
void vcmpfalseps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 11); }
void vcmpfalsesd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 11); }
void vcmpfalsess(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 11); }
void vcmpge_oqpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 29); }
void vcmpge_oqps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 29); }
void vcmpge_oqsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 29); }
void vcmpge_oqss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 29); }
void vcmpgepd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 13); }
void vcmpgeps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 13); }
void vcmpgesd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 13); }
void vcmpgess(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 13); }
void vcmpgt_oqpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 30); }
void vcmpgt_oqps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 30); }
void vcmpgt_oqsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 30); }
void vcmpgt_oqss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 30); }
void vcmpgtpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 14); }
void vcmpgtps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 14); }
void vcmpgtsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 14); }
void vcmpgtss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 14); }
void vcmple_oqpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 18); }
void vcmple_oqps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 18); }
void vcmple_oqsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 18); }
void vcmple_oqss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 18); }
void vcmplepd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 2); }
void vcmpleps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 2); }
void vcmplesd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 2); }
void vcmpless(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 2); }
void vcmplt_oqpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 17); }
void vcmplt_oqps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 17); }
void vcmplt_oqsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 17); }
void vcmplt_oqss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 17); }
void vcmpltpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 1); }
void vcmpltps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 1); }
void vcmpltsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 1); }
void vcmpltss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 1); }
void vcmpneq_oqpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 12); }
void vcmpneq_oqps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 12); }
void vcmpneq_oqsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 12); }
void vcmpneq_oqss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 12); }
void vcmpneq_ospd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 28); }
void vcmpneq_osps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 28); }
void vcmpneq_ossd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 28); }
void vcmpneq_osss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 28); }
void vcmpneq_uspd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 20); }
void vcmpneq_usps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 20); }
void vcmpneq_ussd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 20); }
void vcmpneq_usss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 20); }
void vcmpneqpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 4); }
void vcmpneqps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 4); }
void vcmpneqsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 4); }
void vcmpneqss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 4); }
void vcmpnge_uqpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 25); }
void vcmpnge_uqps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 25); }
void vcmpnge_uqsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 25); }
void vcmpnge_uqss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 25); }
void vcmpngepd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 9); }
void vcmpngeps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 9); }
void vcmpngesd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 9); }
void vcmpngess(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 9); }
void vcmpngt_uqpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 26); }
void vcmpngt_uqps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 26); }
void vcmpngt_uqsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 26); }
void vcmpngt_uqss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 26); }
void vcmpngtpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 10); }
void vcmpngtps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 10); }
void vcmpngtsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 10); }
void vcmpngtss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 10); }
void vcmpnle_uqpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 22); }
void vcmpnle_uqps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 22); }
void vcmpnle_uqsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 22); }
void vcmpnle_uqss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 22); }
void vcmpnlepd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 6); }
void vcmpnleps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 6); }
void vcmpnlesd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 6); }
void vcmpnless(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 6); }
void vcmpnlt_uqpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 21); }
void vcmpnlt_uqps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 21); }
void vcmpnlt_uqsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 21); }
void vcmpnlt_uqss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 21); }
void vcmpnltpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 5); }
void vcmpnltps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 5); }
void vcmpnltsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 5); }
void vcmpnltss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 5); }
void vcmpord_spd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 23); }
void vcmpord_sps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 23); }
void vcmpord_ssd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 23); }
void vcmpord_sss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 23); }
void vcmpordpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 7); }
void vcmpordps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 7); }
void vcmpordsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 7); }
void vcmpordss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 7); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vcmppd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM, 0xC2, imm); }
void vcmpps(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_0F | T_YMM, 0xC2, imm); }
void vcmpsd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_F2 | T_0F, 0xC2, imm); }
void vcmpss(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_F3 | T_0F, 0xC2, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcmptrue_uspd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 31); }
void vcmptrue_usps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 31); }
void vcmptrue_ussd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 31); }
void vcmptrue_usss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 31); }
void vcmptruepd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 15); }
void vcmptrueps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 15); }
void vcmptruesd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 15); }
void vcmptruess(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 15); }
void vcmpunord_spd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 19); }
void vcmpunord_sps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 19); }
void vcmpunord_ssd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 19); }
void vcmpunord_sss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 19); }
void vcmpunordpd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmppd(x1, x2, op, 3); }
void vcmpunordps(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpps(x1, x2, op, 3); }
void vcmpunordsd(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpsd(x1, x2, op, 3); }
void vcmpunordss(const Xmm& x1, const Xmm& x2, const Operand& op) { vcmpss(x1, x2, op, 3); }
void vcomisd(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N8 | T_66 | T_0F | T_EW1 | T_EVEX | T_SAE_X, 0x2F); }
void vcomiss(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N4 | T_0F | T_EW0 | T_EVEX | T_SAE_X, 0x2F); }
void vcvtdq2pd(const Xmm& x, const Operand& op) { checkCvt1(x, op); opVex(x, 0, op, T_0F | T_F3 | T_YMM | T_EVEX | T_EW0 | T_B32 | T_N8 | T_N_VL, 0xE6); }
void vcvtdq2ps(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_0F | T_EW0 | T_YMM | T_EVEX | T_ER_Z | T_B32, 0x5B); }
Squashed 'externals/xbyak/' changes from 88f2f771f..a1ac3750f a1ac3750f Merge branch 'dev' 5f4ba971f v6.68 cac2c175f update doc 1b08a8248 add test of prefetchiti bef70d9b1 add prefetchit{0,1} f66785876 add detection of prefetchiti 124617ac9 Merge branch 'dev' 1bce8be17 Merge commit 'fbb18f6' into dev bebfe64a3 [test] X32 is optional and remove CFLAGS d8c332cef Merge branch 'Tachi107-small-changes' into dev fbb18f69d ci: use containers 8a974696c test: simplify makefile 1efe9fe7c Merge branch 'dev' bb70083e6 v6.67 436e452d7 update doc 47ff6ef42 I get an error after all on GitHub action 445c0dcec add test of CMPccXADD c9347907d add CMPccXADD dc792cc56 add detection of cmpccxadd 62be84cc8 fix detection of boost b5ac7b0f0 Merge branch 'dev' 05dd400e0 recover a removed line 94eff6246 v6.66 ec3fadeba update doc 00bfaaa7f add test of rao-int 0f2f1aaa6 support rio-int 6c047f480 detect rao-int f07c5c255 fix args of EmptyAllocator::alloc in test 04d3eb5f5 stop if a test script causes an error 1c1d2366f disable boost sample cfb1127c3 tweak 5fcbeb7c4 [sample] change the way of detection of boost bafc1ee60 CXX uses g++ as default value d8cabc6cb remove warning of blace bc73a0816 remove -fno-operator-names option 6989aea94 use CXX instead of g++ a7c5a1bd7 use English 1bfbd8c4d Fix incorrect format strings 0ecef5c28 Merge branch 'dev' 7556c20ba update doc a15709271 v6.65 3b83aab3e add detect_x32 to TARGET 8c64bbbc3 use gcc instead of dpkg for portability 5e9a9b96f test_avx512.sh runs on x32 8ae01b0c2 disable some tests on x32 83b3da217 x32 does not check large disp 693ab8c9d sizeof(void*) = 4 on x32, so disable the test 348e3e548 Merge branch 'dev' 11b9c4dc0 v6.64 459636196 add T_M_K flag to vpmov* 1d3722928 add include path 5e27eddae move CrearError from LabelManager::reset() to CodeGenerator::reset() f8ea5c28d Merge branch 'dev' 20b2b1eae v6.63 3706869f8 desc. of setDefaultEncoding d6f2d7577 add test of setDefaultEncoding 3b0a19c41 vpmadd52{h,l}uq for avx-ifma 95752ebd7 add tAVX_IFMA cd36e31ea [sample] show AMX_FP16/AVX_VNNI_INT8/AVX_NE_CONVERT e5858af27 add setDefaultEncoding 2f7fb0220 modify gen.cpp for AVX-NE-CONVERT/AVX-VNNI-INT8/AMX-FP16 1c5cb7efa add AVX-NE-CONVERT instructions 564fe9acd add AVX-VNNI-INT8 instructions cd14d07b1 add AMX-FP16 instruction 7811f593c Merge pull request #161 from scribam/patch-1 2218f6c08 Update changelog.md 7bccdbbb4 Merge pull request #160 from herumi/dev 5fcf87596 compile nasm-2.15 instead of apt install e31961ea8 v6.62 ff4f9e65c update doc 9c8fb81db disable wrong detection of gcc -Warray-bounds 6b7519659 add serialize e16582696 Merge branch 'JonLiu1993-vcpkg-installition' into dev edbb410fb Add vcpkg installation instructions 055d31242 Merge branch 'dev' 21ab98441 Merge branch 'akodanka-enable_CIV' into dev ed4d598e3 Changes to compile xbyak project for openvino b652430c4 mingw uses __cpuidex 48457bfa0 Merge branch 'dev' 29cb524d1 v6.61.2 1a9a0b0e1 avoid including algorithm header in xbyak_util.h 6fadefd04 Merge branch 'dev' fc1c18a9d update doc f7cae7f11 v6.61.1 6f5ec5cf3 Merge branch 'doyaGu-master' into dev 4554d6bb9 Fix error related to XBYAK_NOEXCEPT git-subtree-dir: externals/xbyak git-subtree-split: a1ac3750f9a639b5a6c6d6c7da4259b8d6790989
2022-12-30 23:05:02 +00:00
void vcvtneebf162ps(const Xmm& x, const Address& addr) { opVex(x, 0, addr, T_F3 | T_0F38 | T_W0 | T_YMM, 0xB0); }
void vcvtneeph2ps(const Xmm& x, const Address& addr) { opVex(x, 0, addr, T_66 | T_0F38 | T_W0 | T_YMM, 0xB0); }
void vcvtneobf162ps(const Xmm& x, const Address& addr) { opVex(x, 0, addr, T_F2 | T_0F38 | T_W0 | T_YMM, 0xB0); }
void vcvtneoph2ps(const Xmm& x, const Address& addr) { opVex(x, 0, addr, T_0F38 | T_W0 | T_YMM, 0xB0); }
void vcvtneps2bf16(const Xmm& x, const Operand& op, PreferredEncoding encoding = DefaultEncoding) { opCvt2(x, op, T_F3 | T_0F38 | T_EW0 | T_YMM | T_SAE_Z | T_B32 | orEvexIf(encoding), 0x72); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcvtpd2dq(const Xmm& x, const Operand& op) { opCvt2(x, op, T_0F | T_F2 | T_YMM | T_EVEX | T_EW1 | T_B64 | T_ER_Z, 0xE6); }
void vcvtpd2ps(const Xmm& x, const Operand& op) { opCvt2(x, op, T_0F | T_66 | T_YMM | T_EVEX | T_EW1 | T_B64 | T_ER_Z, 0x5A); }
void vcvtph2ps(const Xmm& x, const Operand& op) { checkCvt1(x, op); opVex(x, 0, op, T_0F38 | T_66 | T_W0 | T_EVEX | T_EW0 | T_N8 | T_N_VL | T_SAE_Y, 0x13); }
void vcvtps2dq(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F | T_EW0 | T_YMM | T_EVEX | T_ER_Z | T_B32, 0x5B); }
void vcvtps2pd(const Xmm& x, const Operand& op) { checkCvt1(x, op); opVex(x, 0, op, T_0F | T_YMM | T_EVEX | T_EW0 | T_B32 | T_N8 | T_N_VL | T_SAE_Y, 0x5A); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcvtps2ph(const Operand& op, const Xmm& x, uint8_t imm) { checkCvt1(x, op); opVex(x, 0, op, T_0F3A | T_66 | T_W0 | T_EVEX | T_EW0 | T_N8 | T_N_VL | T_SAE_Y | T_M_K, 0x1D, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcvtsd2si(const Reg32& r, const Operand& op) { opAVX_X_X_XM(Xmm(r.getIdx()), xm0, op, T_0F | T_F2 | T_W0 | T_EVEX | T_EW0 | T_N4 | T_ER_X, 0x2D); }
void vcvtsd2ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_F2 | T_0F | T_EW1 | T_EVEX | T_ER_X, 0x5A); }
void vcvtsi2sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opCvt3(x1, x2, op, T_0F | T_F2 | T_EVEX, T_W1 | T_EW1 | T_ER_X | T_N8, T_W0 | T_EW0 | T_N4, 0x2A); }
void vcvtsi2ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opCvt3(x1, x2, op, T_0F | T_F3 | T_EVEX | T_ER_X, T_W1 | T_EW1 | T_N8, T_W0 | T_EW0 | T_N4, 0x2A); }
void vcvtss2sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_F3 | T_0F | T_EW0 | T_EVEX | T_SAE_X, 0x5A); }
void vcvtss2si(const Reg32& r, const Operand& op) { opAVX_X_X_XM(Xmm(r.getIdx()), xm0, op, T_0F | T_F3 | T_W0 | T_EVEX | T_EW0 | T_ER_X | T_N8, 0x2D); }
void vcvttpd2dq(const Xmm& x, const Operand& op) { opCvt2(x, op, T_66 | T_0F | T_YMM | T_EVEX |T_EW1 | T_B64 | T_ER_Z, 0xE6); }
void vcvttps2dq(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_F3 | T_0F | T_EW0 | T_YMM | T_EVEX | T_SAE_Z | T_B32, 0x5B); }
void vcvttsd2si(const Reg32& r, const Operand& op) { opAVX_X_X_XM(Xmm(r.getIdx()), xm0, op, T_0F | T_F2 | T_W0 | T_EVEX | T_EW0 | T_N4 | T_SAE_X, 0x2C); }
void vcvttss2si(const Reg32& r, const Operand& op) { opAVX_X_X_XM(Xmm(r.getIdx()), xm0, op, T_0F | T_F3 | T_W0 | T_EVEX | T_EW0 | T_SAE_X | T_N8, 0x2C); }
void vdivpd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_66 | T_EW1 | T_YMM | T_EVEX | T_ER_Z | T_B64, 0x5E); }
void vdivps(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_EW0 | T_YMM | T_EVEX | T_ER_Z | T_B32, 0x5E); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vdivsd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_F2 | T_EW1 | T_EVEX | T_ER_X | T_N8, 0x5E); }
void vdivss(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_F3 | T_EW0 | T_EVEX | T_ER_X | T_N4, 0x5E); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vdppd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_W0, 0x41, imm); }
void vdpps(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_W0 | T_YMM, 0x40, imm); }
void vextractf128(const Operand& op, const Ymm& y, uint8_t imm) { if (!(op.isXMEM() && y.isYMM())) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(y, 0, op, T_0F3A | T_66 | T_W0 | T_YMM, 0x19, imm); }
void vextracti128(const Operand& op, const Ymm& y, uint8_t imm) { if (!(op.isXMEM() && y.isYMM())) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(y, 0, op, T_0F3A | T_66 | T_W0 | T_YMM, 0x39, imm); }
void vextractps(const Operand& op, const Xmm& x, uint8_t imm) { if (!((op.isREG(32) || op.isMEM()) && x.isXMM())) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(x, 0, op, T_0F3A | T_66 | T_W0 | T_EVEX | T_N4, 0x17, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vfmadd132pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0x98); }
void vfmadd132ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x98); }
void vfmadd132sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_W1 | T_EW1 | T_EVEX | T_ER_X, 0x99); }
void vfmadd132ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_W0 | T_EW0 | T_EVEX | T_ER_X, 0x99); }
void vfmadd213pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0xA8); }
void vfmadd213ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0xA8); }
void vfmadd213sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_W1 | T_EW1 | T_EVEX | T_ER_X, 0xA9); }
void vfmadd213ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_W0 | T_EW0 | T_EVEX | T_ER_X, 0xA9); }
void vfmadd231pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0xB8); }
void vfmadd231ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0xB8); }
void vfmadd231sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_W1 | T_EW1 | T_EVEX | T_ER_X, 0xB9); }
void vfmadd231ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_W0 | T_EW0 | T_EVEX | T_ER_X, 0xB9); }
void vfmaddsub132pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0x96); }
void vfmaddsub132ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x96); }
void vfmaddsub213pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0xA6); }
void vfmaddsub213ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0xA6); }
void vfmaddsub231pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0xB6); }
void vfmaddsub231ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0xB6); }
void vfmsub132pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0x9A); }
void vfmsub132ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x9A); }
void vfmsub132sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_W1 | T_EW1 | T_EVEX | T_ER_X, 0x9B); }
void vfmsub132ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_W0 | T_EW0 | T_EVEX | T_ER_X, 0x9B); }
void vfmsub213pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0xAA); }
void vfmsub213ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0xAA); }
void vfmsub213sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_W1 | T_EW1 | T_EVEX | T_ER_X, 0xAB); }
void vfmsub213ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_W0 | T_EW0 | T_EVEX | T_ER_X, 0xAB); }
void vfmsub231pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0xBA); }
void vfmsub231ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0xBA); }
void vfmsub231sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_W1 | T_EW1 | T_EVEX | T_ER_X, 0xBB); }
void vfmsub231ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_W0 | T_EW0 | T_EVEX | T_ER_X, 0xBB); }
void vfmsubadd132pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0x97); }
void vfmsubadd132ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x97); }
void vfmsubadd213pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0xA7); }
void vfmsubadd213ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0xA7); }
void vfmsubadd231pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0xB7); }
void vfmsubadd231ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0xB7); }
void vfnmadd132pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0x9C); }
void vfnmadd132ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x9C); }
void vfnmadd132sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_W1 | T_EW1 | T_EVEX | T_ER_X, 0x9D); }
void vfnmadd132ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_W0 | T_EW0 | T_EVEX | T_ER_X, 0x9D); }
void vfnmadd213pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0xAC); }
void vfnmadd213ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0xAC); }
void vfnmadd213sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_W1 | T_EW1 | T_EVEX | T_ER_X, 0xAD); }
void vfnmadd213ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_W0 | T_EW0 | T_EVEX | T_ER_X, 0xAD); }
void vfnmadd231pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0xBC); }
void vfnmadd231ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0xBC); }
void vfnmadd231sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_W1 | T_EW1 | T_EVEX | T_ER_X, 0xBD); }
void vfnmadd231ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_W0 | T_EW0 | T_EVEX | T_ER_X, 0xBD); }
void vfnmsub132pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0x9E); }
void vfnmsub132ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x9E); }
void vfnmsub132sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_W1 | T_EW1 | T_EVEX | T_ER_X, 0x9F); }
void vfnmsub132ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_W0 | T_EW0 | T_EVEX | T_ER_X, 0x9F); }
void vfnmsub213pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0xAE); }
void vfnmsub213ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0xAE); }
void vfnmsub213sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_W1 | T_EW1 | T_EVEX | T_ER_X, 0xAF); }
void vfnmsub213ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_W0 | T_EW0 | T_EVEX | T_ER_X, 0xAF); }
void vfnmsub231pd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0xBE); }
void vfnmsub231ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0xBE); }
void vfnmsub231sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_W1 | T_EW1 | T_EVEX | T_ER_X, 0xBF); }
void vfnmsub231ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_W0 | T_EW0 | T_EVEX | T_ER_X, 0xBF); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vgatherdpd(const Xmm& x1, const Address& addr, const Xmm& x2) { opGather(x1, addr, x2, T_0F38 | T_66 | T_YMM | T_VSIB | T_W1, 0x92, 0); }
void vgatherdps(const Xmm& x1, const Address& addr, const Xmm& x2) { opGather(x1, addr, x2, T_0F38 | T_66 | T_YMM | T_VSIB | T_W0, 0x92, 1); }
void vgatherqpd(const Xmm& x1, const Address& addr, const Xmm& x2) { opGather(x1, addr, x2, T_0F38 | T_66 | T_YMM | T_VSIB | T_W1, 0x93, 1); }
void vgatherqps(const Xmm& x1, const Address& addr, const Xmm& x2) { opGather(x1, addr, x2, T_0F38 | T_66 | T_YMM | T_VSIB | T_W0, 0x93, 2); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vgf2p8affineinvqb(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_W1 | T_EW1 | T_YMM | T_EVEX | T_SAE_Z | T_B64, 0xCF, imm); }
void vgf2p8affineqb(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_W1 | T_EW1 | T_YMM | T_EVEX | T_SAE_Z | T_B64, 0xCE, imm); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
void vgf2p8mulb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_SAE_Z, 0xCF); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vhaddpd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_66 | T_0F | T_YMM, 0x7C); }
void vhaddps(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_F2 | T_0F | T_YMM, 0x7C); }
void vhsubpd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_66 | T_0F | T_YMM, 0x7D); }
void vhsubps(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_F2 | T_0F | T_YMM, 0x7D); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vinsertf128(const Ymm& y1, const Ymm& y2, const Operand& op, uint8_t imm) { if (!(y1.isYMM() && y2.isYMM() && op.isXMEM())) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(y1, &y2, op, T_0F3A | T_66 | T_W0 | T_YMM, 0x18, imm); }
void vinserti128(const Ymm& y1, const Ymm& y2, const Operand& op, uint8_t imm) { if (!(y1.isYMM() && y2.isYMM() && op.isXMEM())) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(y1, &y2, op, T_0F3A | T_66 | T_W0 | T_YMM, 0x38, imm); }
void vinsertps(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F3A | T_W0 | T_EW0 | T_EVEX, 0x21, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vlddqu(const Xmm& x, const Address& addr) { opAVX_X_X_XM(x, cvtIdx0(x), addr, T_0F | T_F2 | T_W0 | T_YMM, 0xF0); }
void vldmxcsr(const Address& addr) { opAVX_X_X_XM(xm2, xm0, addr, T_0F, 0xAE); }
void vmaskmovdqu(const Xmm& x1, const Xmm& x2) { opAVX_X_X_XM(x1, xm0, x2, T_0F | T_66, 0xF7); }
void vmaskmovpd(const Address& addr, const Xmm& x1, const Xmm& x2) { opAVX_X_X_XM(x2, x1, addr, T_0F38 | T_66 | T_W0 | T_YMM, 0x2F); }
void vmaskmovpd(const Xmm& x1, const Xmm& x2, const Address& addr) { opAVX_X_X_XM(x1, x2, addr, T_0F38 | T_66 | T_W0 | T_YMM, 0x2D); }
void vmaskmovps(const Address& addr, const Xmm& x1, const Xmm& x2) { opAVX_X_X_XM(x2, x1, addr, T_0F38 | T_66 | T_W0 | T_YMM, 0x2E); }
void vmaskmovps(const Xmm& x1, const Xmm& x2, const Address& addr) { opAVX_X_X_XM(x1, x2, addr, T_0F38 | T_66 | T_W0 | T_YMM, 0x2C); }
void vmaxpd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_66 | T_EW1 | T_YMM | T_EVEX | T_ER_Z | T_B64, 0x5F); }
void vmaxps(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_EW0 | T_YMM | T_EVEX | T_ER_Z | T_B32, 0x5F); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vmaxsd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_F2 | T_EW1 | T_EVEX | T_ER_X | T_N8, 0x5F); }
void vmaxss(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_F3 | T_EW0 | T_EVEX | T_ER_X | T_N4, 0x5F); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vminpd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_66 | T_EW1 | T_YMM | T_EVEX | T_ER_Z | T_B64, 0x5D); }
void vminps(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_EW0 | T_YMM | T_EVEX | T_ER_Z | T_B32, 0x5D); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vminsd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_F2 | T_EW1 | T_EVEX | T_ER_X | T_N8, 0x5D); }
void vminss(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_F3 | T_EW0 | T_EVEX | T_ER_X | T_N4, 0x5D); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vmovapd(const Address& addr, const Xmm& xmm) { opAVX_X_XM_IMM(xmm, addr, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX | T_M_K, 0x29); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovapd(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX, 0x28); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vmovaps(const Address& addr, const Xmm& xmm) { opAVX_X_XM_IMM(xmm, addr, T_0F | T_EW0 | T_YMM | T_EVEX | T_M_K, 0x29); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovaps(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_0F | T_EW0 | T_YMM | T_EVEX, 0x28); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vmovd(const Operand& op, const Xmm& x) { if (!op.isREG(32) && !op.isMEM()) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_X_XM(x, xm0, op, T_0F | T_66 | T_W0 | T_EVEX | T_N4, 0x7E); }
void vmovd(const Xmm& x, const Operand& op) { if (!op.isREG(32) && !op.isMEM()) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_X_XM(x, xm0, op, T_0F | T_66 | T_W0 | T_EVEX | T_N4, 0x6E); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovddup(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_DUP | T_F2 | T_0F | T_EW1 | T_YMM | T_EVEX | T_ER_X | T_ER_Y | T_ER_Z, 0x12); }
void vmovdqa(const Address& addr, const Xmm& xmm) { opAVX_X_XM_IMM(xmm, addr, T_66 | T_0F | T_YMM, 0x7F); }
void vmovdqa(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F | T_YMM, 0x6F); }
void vmovdqu(const Address& addr, const Xmm& xmm) { opAVX_X_XM_IMM(xmm, addr, T_F3 | T_0F | T_YMM, 0x7F); }
void vmovdqu(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_F3 | T_0F | T_YMM, 0x6F); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vmovhlps(const Xmm& x1, const Xmm& x2, const Operand& op = Operand()) { if (!op.isNone() && !op.isXMM()) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_X_XM(x1, x2, op, T_0F | T_EVEX | T_EW0, 0x12); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovhpd(const Address& addr, const Xmm& x) { opAVX_X_X_XM(x, xm0, addr, T_0F | T_66 | T_EVEX | T_EW1 | T_N8, 0x17); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vmovhpd(const Xmm& x, const Operand& op1, const Operand& op2 = Operand()) { if (!op2.isNone() && !op2.isMEM()) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_X_XM(x, op1, op2, T_0F | T_66 | T_EVEX | T_EW1 | T_N8, 0x16); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovhps(const Address& addr, const Xmm& x) { opAVX_X_X_XM(x, xm0, addr, T_0F | T_EVEX | T_EW0 | T_N8, 0x17); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vmovhps(const Xmm& x, const Operand& op1, const Operand& op2 = Operand()) { if (!op2.isNone() && !op2.isMEM()) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_X_XM(x, op1, op2, T_0F | T_EVEX | T_EW0 | T_N8, 0x16); }
void vmovlhps(const Xmm& x1, const Xmm& x2, const Operand& op = Operand()) { if (!op.isNone() && !op.isXMM()) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_X_XM(x1, x2, op, T_0F | T_EVEX | T_EW0, 0x16); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovlpd(const Address& addr, const Xmm& x) { opAVX_X_X_XM(x, xm0, addr, T_0F | T_66 | T_EVEX | T_EW1 | T_N8, 0x13); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vmovlpd(const Xmm& x, const Operand& op1, const Operand& op2 = Operand()) { if (!op2.isNone() && !op2.isMEM()) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_X_XM(x, op1, op2, T_0F | T_66 | T_EVEX | T_EW1 | T_N8, 0x12); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovlps(const Address& addr, const Xmm& x) { opAVX_X_X_XM(x, xm0, addr, T_0F | T_EVEX | T_EW0 | T_N8, 0x13); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vmovlps(const Xmm& x, const Operand& op1, const Operand& op2 = Operand()) { if (!op2.isNone() && !op2.isMEM()) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_X_XM(x, op1, op2, T_0F | T_EVEX | T_EW0 | T_N8, 0x12); }
void vmovmskpd(const Reg& r, const Xmm& x) { if (!r.isBit(i32e)) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_X_XM(x.isXMM() ? Xmm(r.getIdx()) : Ymm(r.getIdx()), cvtIdx0(x), x, T_0F | T_66 | T_W0 | T_YMM, 0x50); }
void vmovmskps(const Reg& r, const Xmm& x) { if (!r.isBit(i32e)) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_X_XM(x.isXMM() ? Xmm(r.getIdx()) : Ymm(r.getIdx()), cvtIdx0(x), x, T_0F | T_W0 | T_YMM, 0x50); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovntdq(const Address& addr, const Xmm& x) { opVex(x, 0, addr, T_0F | T_66 | T_YMM | T_EVEX | T_EW0, 0xE7); }
void vmovntdqa(const Xmm& x, const Address& addr) { opVex(x, 0, addr, T_0F38 | T_66 | T_YMM | T_EVEX | T_EW0, 0x2A); }
void vmovntpd(const Address& addr, const Xmm& x) { opVex(x, 0, addr, T_0F | T_66 | T_YMM | T_EVEX | T_EW1, 0x2B); }
void vmovntps(const Address& addr, const Xmm& x) { opVex(x, 0, addr, T_0F | T_YMM | T_EVEX | T_EW0, 0x2B); }
void vmovq(const Address& addr, const Xmm& x) { opAVX_X_X_XM(x, xm0, addr, T_0F | T_66 | T_EVEX | T_EW1 | T_N8, x.getIdx() < 16 ? 0xD6 : 0x7E); }
void vmovq(const Xmm& x, const Address& addr) { int type, code; if (x.getIdx() < 16) { type = T_0F | T_F3; code = 0x7E; } else { type = T_0F | T_66 | T_EVEX | T_EW1 | T_N8; code = 0x6E; } opAVX_X_X_XM(x, xm0, addr, type, code); }
void vmovq(const Xmm& x1, const Xmm& x2) { opAVX_X_X_XM(x1, xm0, x2, T_0F | T_F3 | T_EVEX | T_EW1 | T_N8, 0x7E); }
void vmovsd(const Address& addr, const Xmm& x) { opAVX_X_X_XM(x, xm0, addr, T_N8 | T_F2 | T_0F | T_EW1 | T_EVEX | T_M_K, 0x11); }
void vmovsd(const Xmm& x, const Address& addr) { opAVX_X_X_XM(x, xm0, addr, T_N8 | T_F2 | T_0F | T_EW1 | T_EVEX, 0x10); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vmovsd(const Xmm& x1, const Xmm& x2, const Operand& op = Operand()) { if (!op.isNone() && !op.isXMM()) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_X_XM(x1, x2, op, T_N8 | T_F2 | T_0F | T_EW1 | T_EVEX, 0x10); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovshdup(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_F3 | T_0F | T_EW0 | T_YMM | T_EVEX, 0x16); }
void vmovsldup(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_F3 | T_0F | T_EW0 | T_YMM | T_EVEX, 0x12); }
void vmovss(const Address& addr, const Xmm& x) { opAVX_X_X_XM(x, xm0, addr, T_N4 | T_F3 | T_0F | T_EW0 | T_EVEX | T_M_K, 0x11); }
void vmovss(const Xmm& x, const Address& addr) { opAVX_X_X_XM(x, xm0, addr, T_N4 | T_F3 | T_0F | T_EW0 | T_EVEX, 0x10); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vmovss(const Xmm& x1, const Xmm& x2, const Operand& op = Operand()) { if (!op.isNone() && !op.isXMM()) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_X_XM(x1, x2, op, T_N4 | T_F3 | T_0F | T_EW0 | T_EVEX, 0x10); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vmovupd(const Address& addr, const Xmm& xmm) { opAVX_X_XM_IMM(xmm, addr, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX | T_M_K, 0x11); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovupd(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX, 0x10); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vmovups(const Address& addr, const Xmm& xmm) { opAVX_X_XM_IMM(xmm, addr, T_0F | T_EW0 | T_YMM | T_EVEX | T_M_K, 0x11); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovups(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_0F | T_EW0 | T_YMM | T_EVEX, 0x10); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vmpsadbw(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_W0 | T_YMM, 0x42, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmulpd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_66 | T_EW1 | T_YMM | T_EVEX | T_ER_Z | T_B64, 0x59); }
void vmulps(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_EW0 | T_YMM | T_EVEX | T_ER_Z | T_B32, 0x59); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vmulsd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_F2 | T_EW1 | T_EVEX | T_ER_X | T_N8, 0x59); }
void vmulss(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_F3 | T_EW0 | T_EVEX | T_ER_X | T_N4, 0x59); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vorpd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_66 | T_EW1 | T_YMM | T_EVEX | T_ER_Z | T_B64, 0x56); }
void vorps(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_EW0 | T_YMM | T_EVEX | T_ER_Z | T_B32, 0x56); }
void vpabsb(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F38 | T_YMM | T_EVEX, 0x1C); }
void vpabsd(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x1E); }
void vpabsw(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F38 | T_YMM | T_EVEX, 0x1D); }
void vpackssdw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW0 | T_YMM | T_EVEX | T_B32, 0x6B); }
void vpacksswb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0x63); }
void vpackusdw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x2B); }
void vpackuswb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0x67); }
void vpaddb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xFC); }
void vpaddd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW0 | T_YMM | T_EVEX | T_B32, 0xFE); }
void vpaddq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX | T_B64, 0xD4); }
void vpaddsb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xEC); }
void vpaddsw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xED); }
void vpaddusb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xDC); }
void vpaddusw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xDD); }
void vpaddw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xFD); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpalignr(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_YMM | T_EVEX, 0x0F, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpand(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM, 0xDB); }
void vpandn(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM, 0xDF); }
void vpavgb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xE0); }
void vpavgw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xE3); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpblendd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_W0 | T_YMM, 0x02, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpblendvb(const Xmm& x1, const Xmm& x2, const Operand& op, const Xmm& x4) { opAVX_X_X_XM(x1, x2, op, T_0F3A | T_66 | T_YMM, 0x4C, x4.getIdx() << 4); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpblendw(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_W0 | T_YMM, 0x0E, imm); }
void vpbroadcastb(const Xmm& x, const Operand& op) { if (!(op.isXMM() || op.isMEM())) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_XM_IMM(x, op, T_N1 | T_66 | T_0F38 | T_W0 | T_YMM | T_EVEX, 0x78); }
void vpbroadcastd(const Xmm& x, const Operand& op) { if (!(op.isXMM() || op.isMEM())) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_XM_IMM(x, op, T_N4 | T_66 | T_0F38 | T_W0 | T_YMM | T_EVEX, 0x58); }
void vpbroadcastq(const Xmm& x, const Operand& op) { if (!(op.isXMM() || op.isMEM())) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_XM_IMM(x, op, T_N8 | T_66 | T_0F38 | T_W0 | T_EW1 | T_YMM | T_EVEX, 0x59); }
void vpbroadcastw(const Xmm& x, const Operand& op) { if (!(op.isXMM() || op.isMEM())) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_XM_IMM(x, op, T_N2 | T_66 | T_0F38 | T_W0 | T_YMM | T_EVEX, 0x79); }
void vpclmulqdq(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_W0 | T_YMM | T_EVEX, 0x44, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpcmpeqb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM, 0x74); }
void vpcmpeqd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM, 0x76); }
void vpcmpeqq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM, 0x29); }
void vpcmpeqw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM, 0x75); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpcmpestri(const Xmm& xm, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F3A, 0x61, imm); }
void vpcmpestrm(const Xmm& xm, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F3A, 0x60, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpcmpgtb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM, 0x64); }
void vpcmpgtd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM, 0x66); }
void vpcmpgtq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM, 0x37); }
void vpcmpgtw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM, 0x65); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpcmpistri(const Xmm& xm, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F3A, 0x63, imm); }
void vpcmpistrm(const Xmm& xm, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F3A, 0x62, imm); }
Squashed 'externals/xbyak/' changes from 88f2f771f..a1ac3750f a1ac3750f Merge branch 'dev' 5f4ba971f v6.68 cac2c175f update doc 1b08a8248 add test of prefetchiti bef70d9b1 add prefetchit{0,1} f66785876 add detection of prefetchiti 124617ac9 Merge branch 'dev' 1bce8be17 Merge commit 'fbb18f6' into dev bebfe64a3 [test] X32 is optional and remove CFLAGS d8c332cef Merge branch 'Tachi107-small-changes' into dev fbb18f69d ci: use containers 8a974696c test: simplify makefile 1efe9fe7c Merge branch 'dev' bb70083e6 v6.67 436e452d7 update doc 47ff6ef42 I get an error after all on GitHub action 445c0dcec add test of CMPccXADD c9347907d add CMPccXADD dc792cc56 add detection of cmpccxadd 62be84cc8 fix detection of boost b5ac7b0f0 Merge branch 'dev' 05dd400e0 recover a removed line 94eff6246 v6.66 ec3fadeba update doc 00bfaaa7f add test of rao-int 0f2f1aaa6 support rio-int 6c047f480 detect rao-int f07c5c255 fix args of EmptyAllocator::alloc in test 04d3eb5f5 stop if a test script causes an error 1c1d2366f disable boost sample cfb1127c3 tweak 5fcbeb7c4 [sample] change the way of detection of boost bafc1ee60 CXX uses g++ as default value d8cabc6cb remove warning of blace bc73a0816 remove -fno-operator-names option 6989aea94 use CXX instead of g++ a7c5a1bd7 use English 1bfbd8c4d Fix incorrect format strings 0ecef5c28 Merge branch 'dev' 7556c20ba update doc a15709271 v6.65 3b83aab3e add detect_x32 to TARGET 8c64bbbc3 use gcc instead of dpkg for portability 5e9a9b96f test_avx512.sh runs on x32 8ae01b0c2 disable some tests on x32 83b3da217 x32 does not check large disp 693ab8c9d sizeof(void*) = 4 on x32, so disable the test 348e3e548 Merge branch 'dev' 11b9c4dc0 v6.64 459636196 add T_M_K flag to vpmov* 1d3722928 add include path 5e27eddae move CrearError from LabelManager::reset() to CodeGenerator::reset() f8ea5c28d Merge branch 'dev' 20b2b1eae v6.63 3706869f8 desc. of setDefaultEncoding d6f2d7577 add test of setDefaultEncoding 3b0a19c41 vpmadd52{h,l}uq for avx-ifma 95752ebd7 add tAVX_IFMA cd36e31ea [sample] show AMX_FP16/AVX_VNNI_INT8/AVX_NE_CONVERT e5858af27 add setDefaultEncoding 2f7fb0220 modify gen.cpp for AVX-NE-CONVERT/AVX-VNNI-INT8/AMX-FP16 1c5cb7efa add AVX-NE-CONVERT instructions 564fe9acd add AVX-VNNI-INT8 instructions cd14d07b1 add AMX-FP16 instruction 7811f593c Merge pull request #161 from scribam/patch-1 2218f6c08 Update changelog.md 7bccdbbb4 Merge pull request #160 from herumi/dev 5fcf87596 compile nasm-2.15 instead of apt install e31961ea8 v6.62 ff4f9e65c update doc 9c8fb81db disable wrong detection of gcc -Warray-bounds 6b7519659 add serialize e16582696 Merge branch 'JonLiu1993-vcpkg-installition' into dev edbb410fb Add vcpkg installation instructions 055d31242 Merge branch 'dev' 21ab98441 Merge branch 'akodanka-enable_CIV' into dev ed4d598e3 Changes to compile xbyak project for openvino b652430c4 mingw uses __cpuidex 48457bfa0 Merge branch 'dev' 29cb524d1 v6.61.2 1a9a0b0e1 avoid including algorithm header in xbyak_util.h 6fadefd04 Merge branch 'dev' fc1c18a9d update doc f7cae7f11 v6.61.1 6f5ec5cf3 Merge branch 'doyaGu-master' into dev 4554d6bb9 Fix error related to XBYAK_NOEXCEPT git-subtree-dir: externals/xbyak git-subtree-split: a1ac3750f9a639b5a6c6d6c7da4259b8d6790989
2022-12-30 23:05:02 +00:00
void vpdpbssd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_F2 | T_0F38 | T_W0 | T_YMM, 0x50); }
void vpdpbssds(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_F2 | T_0F38 | T_W0 | T_YMM, 0x51); }
void vpdpbsud(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_F3 | T_0F38 | T_W0 | T_YMM, 0x50); }
void vpdpbsuds(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_F3 | T_0F38 | T_W0 | T_YMM, 0x51); }
void vpdpbusd(const Xmm& x1, const Xmm& x2, const Operand& op, PreferredEncoding encoding = DefaultEncoding) { opEncoding(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_SAE_Z | T_B32, 0x50, encoding); }
void vpdpbusds(const Xmm& x1, const Xmm& x2, const Operand& op, PreferredEncoding encoding = DefaultEncoding) { opEncoding(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_SAE_Z | T_B32, 0x51, encoding); }
void vpdpbuud(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_0F38 | T_W0 | T_YMM, 0x50); }
void vpdpbuuds(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_0F38 | T_W0 | T_YMM, 0x51); }
void vpdpwssd(const Xmm& x1, const Xmm& x2, const Operand& op, PreferredEncoding encoding = DefaultEncoding) { opEncoding(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_SAE_Z | T_B32, 0x52, encoding); }
void vpdpwssds(const Xmm& x1, const Xmm& x2, const Operand& op, PreferredEncoding encoding = DefaultEncoding) { opEncoding(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_SAE_Z | T_B32, 0x53, encoding); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vperm2f128(const Ymm& y1, const Ymm& y2, const Operand& op, uint8_t imm) { if (!(y1.isYMM() && y2.isYMM() && op.isYMEM())) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(y1, &y2, op, T_0F3A | T_66 | T_W0 | T_YMM, 0x06, imm); }
void vperm2i128(const Ymm& y1, const Ymm& y2, const Operand& op, uint8_t imm) { if (!(y1.isYMM() && y2.isYMM() && op.isYMEM())) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(y1, &y2, op, T_0F3A | T_66 | T_W0 | T_YMM, 0x46, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpermd(const Ymm& y1, const Ymm& y2, const Operand& op) { opAVX_X_X_XM(y1, y2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x36); }
void vpermilpd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW1 | T_YMM | T_EVEX | T_B64, 0x0D); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpermilpd(const Xmm& xm, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_EVEX | T_B64, 0x05, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpermilps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x0C); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpermilps(const Xmm& xm, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_EVEX | T_B32, 0x04, imm); }
void vpermpd(const Ymm& y, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(y, op, T_66 | T_0F3A | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0x01, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpermpd(const Ymm& y1, const Ymm& y2, const Operand& op) { opAVX_X_X_XM(y1, y2, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x16); }
void vpermps(const Ymm& y1, const Ymm& y2, const Operand& op) { opAVX_X_X_XM(y1, y2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x16); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpermq(const Ymm& y, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(y, op, T_66 | T_0F3A | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0x00, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpermq(const Ymm& y1, const Ymm& y2, const Operand& op) { opAVX_X_X_XM(y1, y2, op, T_66 | T_0F38 | T_W0 | T_EW1 | T_YMM | T_EVEX | T_B64, 0x36); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpextrb(const Operand& op, const Xmm& x, uint8_t imm) { if (!((op.isREG(8|16|i32e) || op.isMEM()) && x.isXMM())) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(x, 0, op, T_0F3A | T_66 | T_EVEX | T_N1, 0x14, imm); }
void vpextrd(const Operand& op, const Xmm& x, uint8_t imm) { if (!((op.isREG(32) || op.isMEM()) && x.isXMM())) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(x, 0, op, T_0F3A | T_66 | T_W0 | T_EVEX | T_EW0 | T_N4, 0x16, imm); }
void vpextrq(const Operand& op, const Xmm& x, uint8_t imm) { if (!((op.isREG(64) || op.isMEM()) && x.isXMM())) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(x, 0, op, T_0F3A | T_66 | T_W1 | T_EVEX | T_EW1 | T_N8, 0x16, imm); }
void vpextrw(const Operand& op, const Xmm& x, uint8_t imm) { if (!((op.isREG(16|i32e) || op.isMEM()) && x.isXMM())) XBYAK_THROW(ERR_BAD_COMBINATION) if (op.isREG() && x.getIdx() < 16) { opAVX_X_X_XM(Xmm(op.getIdx()), xm0, x, T_0F | T_66, 0xC5, imm); } else { opVex(x, 0, op, T_0F3A | T_66 | T_EVEX | T_N2, 0x15, imm); } }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vpgatherdd(const Xmm& x1, const Address& addr, const Xmm& x2) { opGather(x1, addr, x2, T_0F38 | T_66 | T_YMM | T_VSIB | T_W0, 0x90, 1); }
void vpgatherdq(const Xmm& x1, const Address& addr, const Xmm& x2) { opGather(x1, addr, x2, T_0F38 | T_66 | T_YMM | T_VSIB | T_W1, 0x90, 0); }
void vpgatherqd(const Xmm& x1, const Address& addr, const Xmm& x2) { opGather(x1, addr, x2, T_0F38 | T_66 | T_YMM | T_VSIB | T_W0, 0x91, 2); }
void vpgatherqq(const Xmm& x1, const Address& addr, const Xmm& x2) { opGather(x1, addr, x2, T_0F38 | T_66 | T_YMM | T_VSIB | T_W1, 0x91, 1); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vphaddd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM, 0x02); }
void vphaddsw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM, 0x03); }
void vphaddw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM, 0x01); }
void vphminposuw(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F38, 0x41); }
void vphsubd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM, 0x06); }
void vphsubsw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM, 0x07); }
void vphsubw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM, 0x05); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpinsrb(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { if (!(x1.isXMM() && x2.isXMM() && (op.isREG(32) || op.isMEM()))) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(x1, &x2, op, T_0F3A | T_66 | T_EVEX | T_N1, 0x20, imm); }
void vpinsrd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { if (!(x1.isXMM() && x2.isXMM() && (op.isREG(32) || op.isMEM()))) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(x1, &x2, op, T_0F3A | T_66 | T_W0 | T_EVEX | T_EW0 | T_N4, 0x22, imm); }
void vpinsrq(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { if (!(x1.isXMM() && x2.isXMM() && (op.isREG(64) || op.isMEM()))) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(x1, &x2, op, T_0F3A | T_66 | T_W1 | T_EVEX | T_EW1 | T_N8, 0x22, imm); }
void vpinsrw(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { if (!(x1.isXMM() && x2.isXMM() && (op.isREG(32) || op.isMEM()))) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(x1, &x2, op, T_0F | T_66 | T_EVEX | T_N2, 0xC4, imm); }
Squashed 'externals/xbyak/' changes from 88f2f771f..a1ac3750f a1ac3750f Merge branch 'dev' 5f4ba971f v6.68 cac2c175f update doc 1b08a8248 add test of prefetchiti bef70d9b1 add prefetchit{0,1} f66785876 add detection of prefetchiti 124617ac9 Merge branch 'dev' 1bce8be17 Merge commit 'fbb18f6' into dev bebfe64a3 [test] X32 is optional and remove CFLAGS d8c332cef Merge branch 'Tachi107-small-changes' into dev fbb18f69d ci: use containers 8a974696c test: simplify makefile 1efe9fe7c Merge branch 'dev' bb70083e6 v6.67 436e452d7 update doc 47ff6ef42 I get an error after all on GitHub action 445c0dcec add test of CMPccXADD c9347907d add CMPccXADD dc792cc56 add detection of cmpccxadd 62be84cc8 fix detection of boost b5ac7b0f0 Merge branch 'dev' 05dd400e0 recover a removed line 94eff6246 v6.66 ec3fadeba update doc 00bfaaa7f add test of rao-int 0f2f1aaa6 support rio-int 6c047f480 detect rao-int f07c5c255 fix args of EmptyAllocator::alloc in test 04d3eb5f5 stop if a test script causes an error 1c1d2366f disable boost sample cfb1127c3 tweak 5fcbeb7c4 [sample] change the way of detection of boost bafc1ee60 CXX uses g++ as default value d8cabc6cb remove warning of blace bc73a0816 remove -fno-operator-names option 6989aea94 use CXX instead of g++ a7c5a1bd7 use English 1bfbd8c4d Fix incorrect format strings 0ecef5c28 Merge branch 'dev' 7556c20ba update doc a15709271 v6.65 3b83aab3e add detect_x32 to TARGET 8c64bbbc3 use gcc instead of dpkg for portability 5e9a9b96f test_avx512.sh runs on x32 8ae01b0c2 disable some tests on x32 83b3da217 x32 does not check large disp 693ab8c9d sizeof(void*) = 4 on x32, so disable the test 348e3e548 Merge branch 'dev' 11b9c4dc0 v6.64 459636196 add T_M_K flag to vpmov* 1d3722928 add include path 5e27eddae move CrearError from LabelManager::reset() to CodeGenerator::reset() f8ea5c28d Merge branch 'dev' 20b2b1eae v6.63 3706869f8 desc. of setDefaultEncoding d6f2d7577 add test of setDefaultEncoding 3b0a19c41 vpmadd52{h,l}uq for avx-ifma 95752ebd7 add tAVX_IFMA cd36e31ea [sample] show AMX_FP16/AVX_VNNI_INT8/AVX_NE_CONVERT e5858af27 add setDefaultEncoding 2f7fb0220 modify gen.cpp for AVX-NE-CONVERT/AVX-VNNI-INT8/AMX-FP16 1c5cb7efa add AVX-NE-CONVERT instructions 564fe9acd add AVX-VNNI-INT8 instructions cd14d07b1 add AMX-FP16 instruction 7811f593c Merge pull request #161 from scribam/patch-1 2218f6c08 Update changelog.md 7bccdbbb4 Merge pull request #160 from herumi/dev 5fcf87596 compile nasm-2.15 instead of apt install e31961ea8 v6.62 ff4f9e65c update doc 9c8fb81db disable wrong detection of gcc -Warray-bounds 6b7519659 add serialize e16582696 Merge branch 'JonLiu1993-vcpkg-installition' into dev edbb410fb Add vcpkg installation instructions 055d31242 Merge branch 'dev' 21ab98441 Merge branch 'akodanka-enable_CIV' into dev ed4d598e3 Changes to compile xbyak project for openvino b652430c4 mingw uses __cpuidex 48457bfa0 Merge branch 'dev' 29cb524d1 v6.61.2 1a9a0b0e1 avoid including algorithm header in xbyak_util.h 6fadefd04 Merge branch 'dev' fc1c18a9d update doc f7cae7f11 v6.61.1 6f5ec5cf3 Merge branch 'doyaGu-master' into dev 4554d6bb9 Fix error related to XBYAK_NOEXCEPT git-subtree-dir: externals/xbyak git-subtree-split: a1ac3750f9a639b5a6c6d6c7da4259b8d6790989
2022-12-30 23:05:02 +00:00
void vpmadd52huq(const Xmm& x1, const Xmm& x2, const Operand& op, PreferredEncoding encoding = DefaultEncoding) { opEncoding(x1, x2, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_B64, 0xB5, encoding); }
void vpmadd52luq(const Xmm& x1, const Xmm& x2, const Operand& op, PreferredEncoding encoding = DefaultEncoding) { opEncoding(x1, x2, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_B64, 0xB4, encoding); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpmaddubsw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM | T_EVEX, 0x04); }
void vpmaddwd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xF5); }
void vpmaskmovd(const Address& addr, const Xmm& x1, const Xmm& x2) { opAVX_X_X_XM(x2, x1, addr, T_0F38 | T_66 | T_W0 | T_YMM, 0x8E); }
void vpmaskmovd(const Xmm& x1, const Xmm& x2, const Address& addr) { opAVX_X_X_XM(x1, x2, addr, T_0F38 | T_66 | T_W0 | T_YMM, 0x8C); }
void vpmaskmovq(const Address& addr, const Xmm& x1, const Xmm& x2) { opAVX_X_X_XM(x2, x1, addr, T_0F38 | T_66 | T_W1 | T_YMM, 0x8E); }
void vpmaskmovq(const Xmm& x1, const Xmm& x2, const Address& addr) { opAVX_X_X_XM(x1, x2, addr, T_0F38 | T_66 | T_W1 | T_YMM, 0x8C); }
void vpmaxsb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM | T_EVEX, 0x3C); }
void vpmaxsd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x3D); }
void vpmaxsw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xEE); }
void vpmaxub(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xDE); }
void vpmaxud(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x3F); }
void vpmaxuw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM | T_EVEX, 0x3E); }
void vpminsb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM | T_EVEX, 0x38); }
void vpminsd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x39); }
void vpminsw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xEA); }
void vpminub(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xDA); }
void vpminud(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x3B); }
void vpminuw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM | T_EVEX, 0x3A); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpmovmskb(const Reg32e& r, const Xmm& x) { if (!x.is(Operand::XMM | Operand::YMM)) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(x.isYMM() ? Ymm(r.getIdx()) : Xmm(r.getIdx()), 0, x, T_0F | T_66 | T_YMM, 0xD7); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpmovsxbd(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N4 | T_N_VL | T_66 | T_0F38 | T_YMM | T_EVEX, 0x21); }
void vpmovsxbq(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N2 | T_N_VL | T_66 | T_0F38 | T_YMM | T_EVEX, 0x22); }
void vpmovsxbw(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N8 | T_N_VL | T_66 | T_0F38 | T_YMM | T_EVEX, 0x20); }
void vpmovsxdq(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N8 | T_N_VL | T_66 | T_0F38 | T_EW0 | T_YMM | T_EVEX, 0x25); }
void vpmovsxwd(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N8 | T_N_VL | T_66 | T_0F38 | T_YMM | T_EVEX, 0x23); }
void vpmovsxwq(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N4 | T_N_VL | T_66 | T_0F38 | T_YMM | T_EVEX, 0x24); }
void vpmovzxbd(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N4 | T_N_VL | T_66 | T_0F38 | T_YMM | T_EVEX, 0x31); }
void vpmovzxbq(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N2 | T_N_VL | T_66 | T_0F38 | T_YMM | T_EVEX, 0x32); }
void vpmovzxbw(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N8 | T_N_VL | T_66 | T_0F38 | T_YMM | T_EVEX, 0x30); }
void vpmovzxdq(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N8 | T_N_VL | T_66 | T_0F38 | T_EW0 | T_YMM | T_EVEX, 0x35); }
void vpmovzxwd(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N8 | T_N_VL | T_66 | T_0F38 | T_YMM | T_EVEX, 0x33); }
void vpmovzxwq(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N4 | T_N_VL | T_66 | T_0F38 | T_YMM | T_EVEX, 0x34); }
void vpmuldq(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_EVEX | T_B64, 0x28); }
void vpmulhrsw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM | T_EVEX, 0x0B); }
void vpmulhuw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xE4); }
void vpmulhw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xE5); }
void vpmulld(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x40); }
void vpmullw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xD5); }
void vpmuludq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX | T_B64, 0xF4); }
void vpor(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM, 0xEB); }
void vpsadbw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xF6); }
void vpshufb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM | T_EVEX, 0x00); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpshufd(const Xmm& xm, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F | T_EW0 | T_YMM | T_EVEX | T_B32, 0x70, imm); }
void vpshufhw(const Xmm& xm, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(xm, op, T_F3 | T_0F | T_YMM | T_EVEX, 0x70, imm); }
void vpshuflw(const Xmm& xm, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(xm, op, T_F2 | T_0F | T_YMM | T_EVEX, 0x70, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsignb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM, 0x08); }
void vpsignd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM, 0x0A); }
void vpsignw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_YMM, 0x09); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpslld(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 6), x, op, T_66 | T_0F | T_EW0 | T_YMM | T_EVEX | T_B32 | T_MEM_EVEX, 0x72, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpslld(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N16 | T_66 | T_0F | T_EW0 | T_YMM | T_EVEX, 0xF2); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpslldq(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 7), x, op, T_66 | T_0F | T_YMM | T_EVEX | T_MEM_EVEX, 0x73, imm); }
void vpsllq(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 6), x, op, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX | T_B64 | T_MEM_EVEX, 0x73, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsllq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N16 | T_66 | T_0F | T_EW1 | T_YMM | T_EVEX, 0xF3); }
void vpsllvd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x47); }
void vpsllvq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0x47); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsllw(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 6), x, op, T_66 | T_0F | T_YMM | T_EVEX | T_MEM_EVEX, 0x71, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsllw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N16 | T_66 | T_0F | T_YMM | T_EVEX, 0xF1); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsrad(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 4), x, op, T_66 | T_0F | T_EW0 | T_YMM | T_EVEX | T_B32 | T_MEM_EVEX, 0x72, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsrad(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N16 | T_66 | T_0F | T_EW0 | T_YMM | T_EVEX, 0xE2); }
void vpsravd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x46); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsraw(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 4), x, op, T_66 | T_0F | T_YMM | T_EVEX | T_MEM_EVEX, 0x71, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsraw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N16 | T_66 | T_0F | T_YMM | T_EVEX, 0xE1); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsrld(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 2), x, op, T_66 | T_0F | T_EW0 | T_YMM | T_EVEX | T_B32 | T_MEM_EVEX, 0x72, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsrld(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N16 | T_66 | T_0F | T_EW0 | T_YMM | T_EVEX, 0xD2); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsrldq(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 3), x, op, T_66 | T_0F | T_YMM | T_EVEX | T_MEM_EVEX, 0x73, imm); }
void vpsrlq(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 2), x, op, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX | T_B64 | T_MEM_EVEX, 0x73, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsrlq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N16 | T_66 | T_0F | T_EW1 | T_YMM | T_EVEX, 0xD3); }
void vpsrlvd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W0 | T_EW0 | T_YMM | T_EVEX | T_B32, 0x45); }
void vpsrlvq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_W1 | T_EW1 | T_YMM | T_EVEX | T_B64, 0x45); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsrlw(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 2), x, op, T_66 | T_0F | T_YMM | T_EVEX | T_MEM_EVEX, 0x71, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsrlw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N16 | T_66 | T_0F | T_YMM | T_EVEX, 0xD1); }
void vpsubb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xF8); }
void vpsubd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW0 | T_YMM | T_EVEX | T_B32, 0xFA); }
void vpsubq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX | T_B64, 0xFB); }
void vpsubsb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xE8); }
void vpsubsw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xE9); }
void vpsubusb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xD8); }
void vpsubusw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xD9); }
void vpsubw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0xF9); }
void vptest(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F38 | T_YMM, 0x17); }
void vpunpckhbw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0x68); }
void vpunpckhdq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW0 | T_YMM | T_EVEX | T_B32, 0x6A); }
void vpunpckhqdq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX | T_B64, 0x6D); }
void vpunpckhwd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0x69); }
void vpunpcklbw(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0x60); }
void vpunpckldq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW0 | T_YMM | T_EVEX | T_B32, 0x62); }
void vpunpcklqdq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX | T_B64, 0x6C); }
void vpunpcklwd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM | T_EVEX, 0x61); }
void vpxor(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_YMM, 0xEF); }
void vrcpps(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_0F | T_YMM, 0x53); }
void vrcpss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_F3 | T_0F, 0x53); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vroundpd(const Xmm& xm, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F3A | T_YMM, 0x09, imm); }
void vroundps(const Xmm& xm, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F3A | T_YMM, 0x08, imm); }
void vroundsd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_W0, 0x0B, imm); }
void vroundss(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_W0, 0x0A, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vrsqrtps(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_0F | T_YMM, 0x52); }
void vrsqrtss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_F3 | T_0F, 0x52); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vshufpd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX | T_B64, 0xC6, imm); }
void vshufps(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_0F | T_EW0 | T_YMM | T_EVEX | T_B32, 0xC6, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vsqrtpd(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX | T_ER_Z | T_B64, 0x51); }
void vsqrtps(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_0F | T_EW0 | T_YMM | T_EVEX | T_ER_Z | T_B32, 0x51); }
void vsqrtsd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_F2 | T_0F | T_EW1 | T_EVEX | T_ER_X, 0x51); }
void vsqrtss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_F3 | T_0F | T_EW0 | T_EVEX | T_ER_X, 0x51); }
void vstmxcsr(const Address& addr) { opAVX_X_X_XM(xm3, xm0, addr, T_0F, 0xAE); }
void vsubpd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_66 | T_EW1 | T_YMM | T_EVEX | T_ER_Z | T_B64, 0x5C); }
void vsubps(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_EW0 | T_YMM | T_EVEX | T_ER_Z | T_B32, 0x5C); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vsubsd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_F2 | T_EW1 | T_EVEX | T_ER_X | T_N8, 0x5C); }
void vsubss(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_F3 | T_EW0 | T_EVEX | T_ER_X | T_N4, 0x5C); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vtestpd(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F38 | T_YMM, 0x0F); }
void vtestps(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_66 | T_0F38 | T_YMM, 0x0E); }
void vucomisd(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N8 | T_66 | T_0F | T_EW1 | T_EVEX | T_SAE_X, 0x2E); }
void vucomiss(const Xmm& xm, const Operand& op) { opAVX_X_XM_IMM(xm, op, T_N4 | T_0F | T_EW0 | T_EVEX | T_SAE_X, 0x2E); }
void vunpckhpd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX | T_B64, 0x15); }
void vunpckhps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_0F | T_EW0 | T_YMM | T_EVEX | T_B32, 0x15); }
void vunpcklpd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW1 | T_YMM | T_EVEX | T_B64, 0x14); }
void vunpcklps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_0F | T_EW0 | T_YMM | T_EVEX | T_B32, 0x14); }
void vxorpd(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_66 | T_EW1 | T_YMM | T_EVEX | T_ER_Z | T_B64, 0x57); }
void vxorps(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_0F | T_EW0 | T_YMM | T_EVEX | T_ER_Z | T_B32, 0x57); }
void vzeroall() { db(0xC5); db(0xFC); db(0x77); }
void vzeroupper() { db(0xC5); db(0xF8); db(0x77); }
void wait() { db(0x9B); }
void wbinvd() { db(0x0F); db(0x09); }
void wrmsr() { db(0x0F); db(0x30); }
void xadd(const Operand& op, const Reg& reg) { opModRM(reg, op, (op.isREG() && reg.isREG() && op.getBit() == reg.getBit()), op.isMEM(), 0x0F, 0xC0 | (reg.isBit(8) ? 0 : 1)); }
void xgetbv() { db(0x0F); db(0x01); db(0xD0); }
void xlatb() { db(0xD7); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void xor_(const Operand& op, uint32_t imm) { opRM_I(op, imm, 0x30, 6); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void xor_(const Operand& op1, const Operand& op2) { opRM_RM(op1, op2, 0x30); }
void xorpd(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x57, 0x66, isXMM_XMMorMEM); }
void xorps(const Xmm& xmm, const Operand& op) { opGen(xmm, op, 0x57, 0x100, isXMM_XMMorMEM); }
#ifdef XBYAK_ENABLE_OMITTED_OPERAND
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vblendpd(const Xmm& x, const Operand& op, uint8_t imm) { vblendpd(x, x, op, imm); }
void vblendps(const Xmm& x, const Operand& op, uint8_t imm) { vblendps(x, x, op, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vblendvpd(const Xmm& x1, const Operand& op, const Xmm& x4) { vblendvpd(x1, x1, op, x4); }
void vblendvps(const Xmm& x1, const Operand& op, const Xmm& x4) { vblendvps(x1, x1, op, x4); }
void vcmpeq_ospd(const Xmm& x, const Operand& op) { vcmpeq_ospd(x, x, op); }
void vcmpeq_osps(const Xmm& x, const Operand& op) { vcmpeq_osps(x, x, op); }
void vcmpeq_ossd(const Xmm& x, const Operand& op) { vcmpeq_ossd(x, x, op); }
void vcmpeq_osss(const Xmm& x, const Operand& op) { vcmpeq_osss(x, x, op); }
void vcmpeq_uqpd(const Xmm& x, const Operand& op) { vcmpeq_uqpd(x, x, op); }
void vcmpeq_uqps(const Xmm& x, const Operand& op) { vcmpeq_uqps(x, x, op); }
void vcmpeq_uqsd(const Xmm& x, const Operand& op) { vcmpeq_uqsd(x, x, op); }
void vcmpeq_uqss(const Xmm& x, const Operand& op) { vcmpeq_uqss(x, x, op); }
void vcmpeq_uspd(const Xmm& x, const Operand& op) { vcmpeq_uspd(x, x, op); }
void vcmpeq_usps(const Xmm& x, const Operand& op) { vcmpeq_usps(x, x, op); }
void vcmpeq_ussd(const Xmm& x, const Operand& op) { vcmpeq_ussd(x, x, op); }
void vcmpeq_usss(const Xmm& x, const Operand& op) { vcmpeq_usss(x, x, op); }
void vcmpeqpd(const Xmm& x, const Operand& op) { vcmpeqpd(x, x, op); }
void vcmpeqps(const Xmm& x, const Operand& op) { vcmpeqps(x, x, op); }
void vcmpeqsd(const Xmm& x, const Operand& op) { vcmpeqsd(x, x, op); }
void vcmpeqss(const Xmm& x, const Operand& op) { vcmpeqss(x, x, op); }
void vcmpfalse_ospd(const Xmm& x, const Operand& op) { vcmpfalse_ospd(x, x, op); }
void vcmpfalse_osps(const Xmm& x, const Operand& op) { vcmpfalse_osps(x, x, op); }
void vcmpfalse_ossd(const Xmm& x, const Operand& op) { vcmpfalse_ossd(x, x, op); }
void vcmpfalse_osss(const Xmm& x, const Operand& op) { vcmpfalse_osss(x, x, op); }
void vcmpfalsepd(const Xmm& x, const Operand& op) { vcmpfalsepd(x, x, op); }
void vcmpfalseps(const Xmm& x, const Operand& op) { vcmpfalseps(x, x, op); }
void vcmpfalsesd(const Xmm& x, const Operand& op) { vcmpfalsesd(x, x, op); }
void vcmpfalsess(const Xmm& x, const Operand& op) { vcmpfalsess(x, x, op); }
void vcmpge_oqpd(const Xmm& x, const Operand& op) { vcmpge_oqpd(x, x, op); }
void vcmpge_oqps(const Xmm& x, const Operand& op) { vcmpge_oqps(x, x, op); }
void vcmpge_oqsd(const Xmm& x, const Operand& op) { vcmpge_oqsd(x, x, op); }
void vcmpge_oqss(const Xmm& x, const Operand& op) { vcmpge_oqss(x, x, op); }
void vcmpgepd(const Xmm& x, const Operand& op) { vcmpgepd(x, x, op); }
void vcmpgeps(const Xmm& x, const Operand& op) { vcmpgeps(x, x, op); }
void vcmpgesd(const Xmm& x, const Operand& op) { vcmpgesd(x, x, op); }
void vcmpgess(const Xmm& x, const Operand& op) { vcmpgess(x, x, op); }
void vcmpgt_oqpd(const Xmm& x, const Operand& op) { vcmpgt_oqpd(x, x, op); }
void vcmpgt_oqps(const Xmm& x, const Operand& op) { vcmpgt_oqps(x, x, op); }
void vcmpgt_oqsd(const Xmm& x, const Operand& op) { vcmpgt_oqsd(x, x, op); }
void vcmpgt_oqss(const Xmm& x, const Operand& op) { vcmpgt_oqss(x, x, op); }
void vcmpgtpd(const Xmm& x, const Operand& op) { vcmpgtpd(x, x, op); }
void vcmpgtps(const Xmm& x, const Operand& op) { vcmpgtps(x, x, op); }
void vcmpgtsd(const Xmm& x, const Operand& op) { vcmpgtsd(x, x, op); }
void vcmpgtss(const Xmm& x, const Operand& op) { vcmpgtss(x, x, op); }
void vcmple_oqpd(const Xmm& x, const Operand& op) { vcmple_oqpd(x, x, op); }
void vcmple_oqps(const Xmm& x, const Operand& op) { vcmple_oqps(x, x, op); }
void vcmple_oqsd(const Xmm& x, const Operand& op) { vcmple_oqsd(x, x, op); }
void vcmple_oqss(const Xmm& x, const Operand& op) { vcmple_oqss(x, x, op); }
void vcmplepd(const Xmm& x, const Operand& op) { vcmplepd(x, x, op); }
void vcmpleps(const Xmm& x, const Operand& op) { vcmpleps(x, x, op); }
void vcmplesd(const Xmm& x, const Operand& op) { vcmplesd(x, x, op); }
void vcmpless(const Xmm& x, const Operand& op) { vcmpless(x, x, op); }
void vcmplt_oqpd(const Xmm& x, const Operand& op) { vcmplt_oqpd(x, x, op); }
void vcmplt_oqps(const Xmm& x, const Operand& op) { vcmplt_oqps(x, x, op); }
void vcmplt_oqsd(const Xmm& x, const Operand& op) { vcmplt_oqsd(x, x, op); }
void vcmplt_oqss(const Xmm& x, const Operand& op) { vcmplt_oqss(x, x, op); }
void vcmpltpd(const Xmm& x, const Operand& op) { vcmpltpd(x, x, op); }
void vcmpltps(const Xmm& x, const Operand& op) { vcmpltps(x, x, op); }
void vcmpltsd(const Xmm& x, const Operand& op) { vcmpltsd(x, x, op); }
void vcmpltss(const Xmm& x, const Operand& op) { vcmpltss(x, x, op); }
void vcmpneq_oqpd(const Xmm& x, const Operand& op) { vcmpneq_oqpd(x, x, op); }
void vcmpneq_oqps(const Xmm& x, const Operand& op) { vcmpneq_oqps(x, x, op); }
void vcmpneq_oqsd(const Xmm& x, const Operand& op) { vcmpneq_oqsd(x, x, op); }
void vcmpneq_oqss(const Xmm& x, const Operand& op) { vcmpneq_oqss(x, x, op); }
void vcmpneq_ospd(const Xmm& x, const Operand& op) { vcmpneq_ospd(x, x, op); }
void vcmpneq_osps(const Xmm& x, const Operand& op) { vcmpneq_osps(x, x, op); }
void vcmpneq_ossd(const Xmm& x, const Operand& op) { vcmpneq_ossd(x, x, op); }
void vcmpneq_osss(const Xmm& x, const Operand& op) { vcmpneq_osss(x, x, op); }
void vcmpneq_uspd(const Xmm& x, const Operand& op) { vcmpneq_uspd(x, x, op); }
void vcmpneq_usps(const Xmm& x, const Operand& op) { vcmpneq_usps(x, x, op); }
void vcmpneq_ussd(const Xmm& x, const Operand& op) { vcmpneq_ussd(x, x, op); }
void vcmpneq_usss(const Xmm& x, const Operand& op) { vcmpneq_usss(x, x, op); }
void vcmpneqpd(const Xmm& x, const Operand& op) { vcmpneqpd(x, x, op); }
void vcmpneqps(const Xmm& x, const Operand& op) { vcmpneqps(x, x, op); }
void vcmpneqsd(const Xmm& x, const Operand& op) { vcmpneqsd(x, x, op); }
void vcmpneqss(const Xmm& x, const Operand& op) { vcmpneqss(x, x, op); }
void vcmpnge_uqpd(const Xmm& x, const Operand& op) { vcmpnge_uqpd(x, x, op); }
void vcmpnge_uqps(const Xmm& x, const Operand& op) { vcmpnge_uqps(x, x, op); }
void vcmpnge_uqsd(const Xmm& x, const Operand& op) { vcmpnge_uqsd(x, x, op); }
void vcmpnge_uqss(const Xmm& x, const Operand& op) { vcmpnge_uqss(x, x, op); }
void vcmpngepd(const Xmm& x, const Operand& op) { vcmpngepd(x, x, op); }
void vcmpngeps(const Xmm& x, const Operand& op) { vcmpngeps(x, x, op); }
void vcmpngesd(const Xmm& x, const Operand& op) { vcmpngesd(x, x, op); }
void vcmpngess(const Xmm& x, const Operand& op) { vcmpngess(x, x, op); }
void vcmpngt_uqpd(const Xmm& x, const Operand& op) { vcmpngt_uqpd(x, x, op); }
void vcmpngt_uqps(const Xmm& x, const Operand& op) { vcmpngt_uqps(x, x, op); }
void vcmpngt_uqsd(const Xmm& x, const Operand& op) { vcmpngt_uqsd(x, x, op); }
void vcmpngt_uqss(const Xmm& x, const Operand& op) { vcmpngt_uqss(x, x, op); }
void vcmpngtpd(const Xmm& x, const Operand& op) { vcmpngtpd(x, x, op); }
void vcmpngtps(const Xmm& x, const Operand& op) { vcmpngtps(x, x, op); }
void vcmpngtsd(const Xmm& x, const Operand& op) { vcmpngtsd(x, x, op); }
void vcmpngtss(const Xmm& x, const Operand& op) { vcmpngtss(x, x, op); }
void vcmpnle_uqpd(const Xmm& x, const Operand& op) { vcmpnle_uqpd(x, x, op); }
void vcmpnle_uqps(const Xmm& x, const Operand& op) { vcmpnle_uqps(x, x, op); }
void vcmpnle_uqsd(const Xmm& x, const Operand& op) { vcmpnle_uqsd(x, x, op); }
void vcmpnle_uqss(const Xmm& x, const Operand& op) { vcmpnle_uqss(x, x, op); }
void vcmpnlepd(const Xmm& x, const Operand& op) { vcmpnlepd(x, x, op); }
void vcmpnleps(const Xmm& x, const Operand& op) { vcmpnleps(x, x, op); }
void vcmpnlesd(const Xmm& x, const Operand& op) { vcmpnlesd(x, x, op); }
void vcmpnless(const Xmm& x, const Operand& op) { vcmpnless(x, x, op); }
void vcmpnlt_uqpd(const Xmm& x, const Operand& op) { vcmpnlt_uqpd(x, x, op); }
void vcmpnlt_uqps(const Xmm& x, const Operand& op) { vcmpnlt_uqps(x, x, op); }
void vcmpnlt_uqsd(const Xmm& x, const Operand& op) { vcmpnlt_uqsd(x, x, op); }
void vcmpnlt_uqss(const Xmm& x, const Operand& op) { vcmpnlt_uqss(x, x, op); }
void vcmpnltpd(const Xmm& x, const Operand& op) { vcmpnltpd(x, x, op); }
void vcmpnltps(const Xmm& x, const Operand& op) { vcmpnltps(x, x, op); }
void vcmpnltsd(const Xmm& x, const Operand& op) { vcmpnltsd(x, x, op); }
void vcmpnltss(const Xmm& x, const Operand& op) { vcmpnltss(x, x, op); }
void vcmpord_spd(const Xmm& x, const Operand& op) { vcmpord_spd(x, x, op); }
void vcmpord_sps(const Xmm& x, const Operand& op) { vcmpord_sps(x, x, op); }
void vcmpord_ssd(const Xmm& x, const Operand& op) { vcmpord_ssd(x, x, op); }
void vcmpord_sss(const Xmm& x, const Operand& op) { vcmpord_sss(x, x, op); }
void vcmpordpd(const Xmm& x, const Operand& op) { vcmpordpd(x, x, op); }
void vcmpordps(const Xmm& x, const Operand& op) { vcmpordps(x, x, op); }
void vcmpordsd(const Xmm& x, const Operand& op) { vcmpordsd(x, x, op); }
void vcmpordss(const Xmm& x, const Operand& op) { vcmpordss(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vcmppd(const Xmm& x, const Operand& op, uint8_t imm) { vcmppd(x, x, op, imm); }
void vcmpps(const Xmm& x, const Operand& op, uint8_t imm) { vcmpps(x, x, op, imm); }
void vcmpsd(const Xmm& x, const Operand& op, uint8_t imm) { vcmpsd(x, x, op, imm); }
void vcmpss(const Xmm& x, const Operand& op, uint8_t imm) { vcmpss(x, x, op, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcmptrue_uspd(const Xmm& x, const Operand& op) { vcmptrue_uspd(x, x, op); }
void vcmptrue_usps(const Xmm& x, const Operand& op) { vcmptrue_usps(x, x, op); }
void vcmptrue_ussd(const Xmm& x, const Operand& op) { vcmptrue_ussd(x, x, op); }
void vcmptrue_usss(const Xmm& x, const Operand& op) { vcmptrue_usss(x, x, op); }
void vcmptruepd(const Xmm& x, const Operand& op) { vcmptruepd(x, x, op); }
void vcmptrueps(const Xmm& x, const Operand& op) { vcmptrueps(x, x, op); }
void vcmptruesd(const Xmm& x, const Operand& op) { vcmptruesd(x, x, op); }
void vcmptruess(const Xmm& x, const Operand& op) { vcmptruess(x, x, op); }
void vcmpunord_spd(const Xmm& x, const Operand& op) { vcmpunord_spd(x, x, op); }
void vcmpunord_sps(const Xmm& x, const Operand& op) { vcmpunord_sps(x, x, op); }
void vcmpunord_ssd(const Xmm& x, const Operand& op) { vcmpunord_ssd(x, x, op); }
void vcmpunord_sss(const Xmm& x, const Operand& op) { vcmpunord_sss(x, x, op); }
void vcmpunordpd(const Xmm& x, const Operand& op) { vcmpunordpd(x, x, op); }
void vcmpunordps(const Xmm& x, const Operand& op) { vcmpunordps(x, x, op); }
void vcmpunordsd(const Xmm& x, const Operand& op) { vcmpunordsd(x, x, op); }
void vcmpunordss(const Xmm& x, const Operand& op) { vcmpunordss(x, x, op); }
void vcvtsd2ss(const Xmm& x, const Operand& op) { vcvtsd2ss(x, x, op); }
void vcvtsi2sd(const Xmm& x, const Operand& op) { vcvtsi2sd(x, x, op); }
void vcvtsi2ss(const Xmm& x, const Operand& op) { vcvtsi2ss(x, x, op); }
void vcvtss2sd(const Xmm& x, const Operand& op) { vcvtss2sd(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vdppd(const Xmm& x, const Operand& op, uint8_t imm) { vdppd(x, x, op, imm); }
void vdpps(const Xmm& x, const Operand& op, uint8_t imm) { vdpps(x, x, op, imm); }
void vinsertps(const Xmm& x, const Operand& op, uint8_t imm) { vinsertps(x, x, op, imm); }
void vmpsadbw(const Xmm& x, const Operand& op, uint8_t imm) { vmpsadbw(x, x, op, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpackssdw(const Xmm& x, const Operand& op) { vpackssdw(x, x, op); }
void vpacksswb(const Xmm& x, const Operand& op) { vpacksswb(x, x, op); }
void vpackusdw(const Xmm& x, const Operand& op) { vpackusdw(x, x, op); }
void vpackuswb(const Xmm& x, const Operand& op) { vpackuswb(x, x, op); }
void vpaddb(const Xmm& x, const Operand& op) { vpaddb(x, x, op); }
void vpaddd(const Xmm& x, const Operand& op) { vpaddd(x, x, op); }
void vpaddq(const Xmm& x, const Operand& op) { vpaddq(x, x, op); }
void vpaddsb(const Xmm& x, const Operand& op) { vpaddsb(x, x, op); }
void vpaddsw(const Xmm& x, const Operand& op) { vpaddsw(x, x, op); }
void vpaddusb(const Xmm& x, const Operand& op) { vpaddusb(x, x, op); }
void vpaddusw(const Xmm& x, const Operand& op) { vpaddusw(x, x, op); }
void vpaddw(const Xmm& x, const Operand& op) { vpaddw(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpalignr(const Xmm& x, const Operand& op, uint8_t imm) { vpalignr(x, x, op, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpand(const Xmm& x, const Operand& op) { vpand(x, x, op); }
void vpandn(const Xmm& x, const Operand& op) { vpandn(x, x, op); }
void vpavgb(const Xmm& x, const Operand& op) { vpavgb(x, x, op); }
void vpavgw(const Xmm& x, const Operand& op) { vpavgw(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpblendd(const Xmm& x, const Operand& op, uint8_t imm) { vpblendd(x, x, op, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpblendvb(const Xmm& x1, const Operand& op, const Xmm& x4) { vpblendvb(x1, x1, op, x4); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpblendw(const Xmm& x, const Operand& op, uint8_t imm) { vpblendw(x, x, op, imm); }
void vpclmulqdq(const Xmm& x, const Operand& op, uint8_t imm) { vpclmulqdq(x, x, op, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpcmpeqb(const Xmm& x, const Operand& op) { vpcmpeqb(x, x, op); }
void vpcmpeqd(const Xmm& x, const Operand& op) { vpcmpeqd(x, x, op); }
void vpcmpeqq(const Xmm& x, const Operand& op) { vpcmpeqq(x, x, op); }
void vpcmpeqw(const Xmm& x, const Operand& op) { vpcmpeqw(x, x, op); }
void vpcmpgtb(const Xmm& x, const Operand& op) { vpcmpgtb(x, x, op); }
void vpcmpgtd(const Xmm& x, const Operand& op) { vpcmpgtd(x, x, op); }
void vpcmpgtq(const Xmm& x, const Operand& op) { vpcmpgtq(x, x, op); }
void vpcmpgtw(const Xmm& x, const Operand& op) { vpcmpgtw(x, x, op); }
void vphaddd(const Xmm& x, const Operand& op) { vphaddd(x, x, op); }
void vphaddsw(const Xmm& x, const Operand& op) { vphaddsw(x, x, op); }
void vphaddw(const Xmm& x, const Operand& op) { vphaddw(x, x, op); }
void vphsubd(const Xmm& x, const Operand& op) { vphsubd(x, x, op); }
void vphsubsw(const Xmm& x, const Operand& op) { vphsubsw(x, x, op); }
void vphsubw(const Xmm& x, const Operand& op) { vphsubw(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpinsrb(const Xmm& x, const Operand& op, uint8_t imm) { vpinsrb(x, x, op, imm); }
void vpinsrd(const Xmm& x, const Operand& op, uint8_t imm) { vpinsrd(x, x, op, imm); }
void vpinsrq(const Xmm& x, const Operand& op, uint8_t imm) { vpinsrq(x, x, op, imm); }
void vpinsrw(const Xmm& x, const Operand& op, uint8_t imm) { vpinsrw(x, x, op, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpmaddubsw(const Xmm& x, const Operand& op) { vpmaddubsw(x, x, op); }
void vpmaddwd(const Xmm& x, const Operand& op) { vpmaddwd(x, x, op); }
void vpmaxsb(const Xmm& x, const Operand& op) { vpmaxsb(x, x, op); }
void vpmaxsd(const Xmm& x, const Operand& op) { vpmaxsd(x, x, op); }
void vpmaxsw(const Xmm& x, const Operand& op) { vpmaxsw(x, x, op); }
void vpmaxub(const Xmm& x, const Operand& op) { vpmaxub(x, x, op); }
void vpmaxud(const Xmm& x, const Operand& op) { vpmaxud(x, x, op); }
void vpmaxuw(const Xmm& x, const Operand& op) { vpmaxuw(x, x, op); }
void vpminsb(const Xmm& x, const Operand& op) { vpminsb(x, x, op); }
void vpminsd(const Xmm& x, const Operand& op) { vpminsd(x, x, op); }
void vpminsw(const Xmm& x, const Operand& op) { vpminsw(x, x, op); }
void vpminub(const Xmm& x, const Operand& op) { vpminub(x, x, op); }
void vpminud(const Xmm& x, const Operand& op) { vpminud(x, x, op); }
void vpminuw(const Xmm& x, const Operand& op) { vpminuw(x, x, op); }
void vpmuldq(const Xmm& x, const Operand& op) { vpmuldq(x, x, op); }
void vpmulhrsw(const Xmm& x, const Operand& op) { vpmulhrsw(x, x, op); }
void vpmulhuw(const Xmm& x, const Operand& op) { vpmulhuw(x, x, op); }
void vpmulhw(const Xmm& x, const Operand& op) { vpmulhw(x, x, op); }
void vpmulld(const Xmm& x, const Operand& op) { vpmulld(x, x, op); }
void vpmullw(const Xmm& x, const Operand& op) { vpmullw(x, x, op); }
void vpmuludq(const Xmm& x, const Operand& op) { vpmuludq(x, x, op); }
void vpor(const Xmm& x, const Operand& op) { vpor(x, x, op); }
void vpsadbw(const Xmm& x, const Operand& op) { vpsadbw(x, x, op); }
void vpsignb(const Xmm& x, const Operand& op) { vpsignb(x, x, op); }
void vpsignd(const Xmm& x, const Operand& op) { vpsignd(x, x, op); }
void vpsignw(const Xmm& x, const Operand& op) { vpsignw(x, x, op); }
void vpslld(const Xmm& x, const Operand& op) { vpslld(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpslld(const Xmm& x, uint8_t imm) { vpslld(x, x, imm); }
void vpslldq(const Xmm& x, uint8_t imm) { vpslldq(x, x, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsllq(const Xmm& x, const Operand& op) { vpsllq(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsllq(const Xmm& x, uint8_t imm) { vpsllq(x, x, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsllw(const Xmm& x, const Operand& op) { vpsllw(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsllw(const Xmm& x, uint8_t imm) { vpsllw(x, x, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsrad(const Xmm& x, const Operand& op) { vpsrad(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsrad(const Xmm& x, uint8_t imm) { vpsrad(x, x, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsraw(const Xmm& x, const Operand& op) { vpsraw(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsraw(const Xmm& x, uint8_t imm) { vpsraw(x, x, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsrld(const Xmm& x, const Operand& op) { vpsrld(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsrld(const Xmm& x, uint8_t imm) { vpsrld(x, x, imm); }
void vpsrldq(const Xmm& x, uint8_t imm) { vpsrldq(x, x, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsrlq(const Xmm& x, const Operand& op) { vpsrlq(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsrlq(const Xmm& x, uint8_t imm) { vpsrlq(x, x, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsrlw(const Xmm& x, const Operand& op) { vpsrlw(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsrlw(const Xmm& x, uint8_t imm) { vpsrlw(x, x, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsubb(const Xmm& x, const Operand& op) { vpsubb(x, x, op); }
void vpsubd(const Xmm& x, const Operand& op) { vpsubd(x, x, op); }
void vpsubq(const Xmm& x, const Operand& op) { vpsubq(x, x, op); }
void vpsubsb(const Xmm& x, const Operand& op) { vpsubsb(x, x, op); }
void vpsubsw(const Xmm& x, const Operand& op) { vpsubsw(x, x, op); }
void vpsubusb(const Xmm& x, const Operand& op) { vpsubusb(x, x, op); }
void vpsubusw(const Xmm& x, const Operand& op) { vpsubusw(x, x, op); }
void vpsubw(const Xmm& x, const Operand& op) { vpsubw(x, x, op); }
void vpunpckhbw(const Xmm& x, const Operand& op) { vpunpckhbw(x, x, op); }
void vpunpckhdq(const Xmm& x, const Operand& op) { vpunpckhdq(x, x, op); }
void vpunpckhqdq(const Xmm& x, const Operand& op) { vpunpckhqdq(x, x, op); }
void vpunpckhwd(const Xmm& x, const Operand& op) { vpunpckhwd(x, x, op); }
void vpunpcklbw(const Xmm& x, const Operand& op) { vpunpcklbw(x, x, op); }
void vpunpckldq(const Xmm& x, const Operand& op) { vpunpckldq(x, x, op); }
void vpunpcklqdq(const Xmm& x, const Operand& op) { vpunpcklqdq(x, x, op); }
void vpunpcklwd(const Xmm& x, const Operand& op) { vpunpcklwd(x, x, op); }
void vpxor(const Xmm& x, const Operand& op) { vpxor(x, x, op); }
void vrcpss(const Xmm& x, const Operand& op) { vrcpss(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vroundsd(const Xmm& x, const Operand& op, uint8_t imm) { vroundsd(x, x, op, imm); }
void vroundss(const Xmm& x, const Operand& op, uint8_t imm) { vroundss(x, x, op, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vrsqrtss(const Xmm& x, const Operand& op) { vrsqrtss(x, x, op); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vshufpd(const Xmm& x, const Operand& op, uint8_t imm) { vshufpd(x, x, op, imm); }
void vshufps(const Xmm& x, const Operand& op, uint8_t imm) { vshufps(x, x, op, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vsqrtsd(const Xmm& x, const Operand& op) { vsqrtsd(x, x, op); }
void vsqrtss(const Xmm& x, const Operand& op) { vsqrtss(x, x, op); }
void vunpckhpd(const Xmm& x, const Operand& op) { vunpckhpd(x, x, op); }
void vunpckhps(const Xmm& x, const Operand& op) { vunpckhps(x, x, op); }
void vunpcklpd(const Xmm& x, const Operand& op) { vunpcklpd(x, x, op); }
void vunpcklps(const Xmm& x, const Operand& op) { vunpcklps(x, x, op); }
#endif
#ifdef XBYAK64
void jecxz(std::string label) { db(0x67); opJmp(label, T_SHORT, 0xe3, 0, 0); }
void jecxz(const Label& label) { db(0x67); opJmp(label, T_SHORT, 0xe3, 0, 0); }
void jrcxz(std::string label) { opJmp(label, T_SHORT, 0xe3, 0, 0); }
void jrcxz(const Label& label) { opJmp(label, T_SHORT, 0xe3, 0, 0); }
void cdqe() { db(0x48); db(0x98); }
void cqo() { db(0x48); db(0x99); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void cmpsq() { db(0x48); db(0xA7); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void popfq() { db(0x9D); }
void pushfq() { db(0x9C); }
void lodsq() { db(0x48); db(0xAD); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void movsq() { db(0x48); db(0xA5); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void scasq() { db(0x48); db(0xAF); }
void stosq() { db(0x48); db(0xAB); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void syscall() { db(0x0F); db(0x05); }
void sysret() { db(0x0F); db(0x07); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void cmpxchg16b(const Address& addr) { opModM(addr, Reg64(1), 0x0F, 0xC7); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void fxrstor64(const Address& addr) { opModM(addr, Reg64(1), 0x0F, 0xAE); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void movq(const Reg64& reg, const Mmx& mmx) { if (mmx.isXMM()) db(0x66); opModR(mmx, reg, 0x0F, 0x7E); }
void movq(const Mmx& mmx, const Reg64& reg) { if (mmx.isXMM()) db(0x66); opModR(mmx, reg, 0x0F, 0x6E); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void movsxd(const Reg64& reg, const Operand& op) { if (!op.isBit(32)) XBYAK_THROW(ERR_BAD_COMBINATION) opModRM(reg, op, op.isREG(), op.isMEM(), 0x63); }
void pextrq(const Operand& op, const Xmm& xmm, uint8_t imm) { if (!op.isREG(64) && !op.isMEM()) XBYAK_THROW(ERR_BAD_COMBINATION) opGen(Reg64(xmm.getIdx()), op, 0x16, 0x66, 0, imm, 0x3A); }
void pinsrq(const Xmm& xmm, const Operand& op, uint8_t imm) { if (!op.isREG(64) && !op.isMEM()) XBYAK_THROW(ERR_BAD_COMBINATION) opGen(Reg64(xmm.getIdx()), op, 0x22, 0x66, 0, imm, 0x3A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcvtss2si(const Reg64& r, const Operand& op) { opAVX_X_X_XM(Xmm(r.getIdx()), xm0, op, T_0F | T_F3 | T_W1 | T_EVEX | T_EW1 | T_ER_X | T_N8, 0x2D); }
void vcvttss2si(const Reg64& r, const Operand& op) { opAVX_X_X_XM(Xmm(r.getIdx()), xm0, op, T_0F | T_F3 | T_W1 | T_EVEX | T_EW1 | T_SAE_X | T_N8, 0x2C); }
void vcvtsd2si(const Reg64& r, const Operand& op) { opAVX_X_X_XM(Xmm(r.getIdx()), xm0, op, T_0F | T_F2 | T_W1 | T_EVEX | T_EW1 | T_N4 | T_ER_X, 0x2D); }
void vcvttsd2si(const Reg64& r, const Operand& op) { opAVX_X_X_XM(Xmm(r.getIdx()), xm0, op, T_0F | T_F2 | T_W1 | T_EVEX | T_EW1 | T_N4 | T_SAE_X, 0x2C); }
void vmovq(const Xmm& x, const Reg64& r) { opAVX_X_X_XM(x, xm0, Xmm(r.getIdx()), T_66 | T_0F | T_W1 | T_EVEX | T_EW1, 0x6E); }
void vmovq(const Reg64& r, const Xmm& x) { opAVX_X_X_XM(x, xm0, Xmm(r.getIdx()), T_66 | T_0F | T_W1 | T_EVEX | T_EW1, 0x7E); }
Squashed 'externals/xbyak/' changes from 88f2f771f..a1ac3750f a1ac3750f Merge branch 'dev' 5f4ba971f v6.68 cac2c175f update doc 1b08a8248 add test of prefetchiti bef70d9b1 add prefetchit{0,1} f66785876 add detection of prefetchiti 124617ac9 Merge branch 'dev' 1bce8be17 Merge commit 'fbb18f6' into dev bebfe64a3 [test] X32 is optional and remove CFLAGS d8c332cef Merge branch 'Tachi107-small-changes' into dev fbb18f69d ci: use containers 8a974696c test: simplify makefile 1efe9fe7c Merge branch 'dev' bb70083e6 v6.67 436e452d7 update doc 47ff6ef42 I get an error after all on GitHub action 445c0dcec add test of CMPccXADD c9347907d add CMPccXADD dc792cc56 add detection of cmpccxadd 62be84cc8 fix detection of boost b5ac7b0f0 Merge branch 'dev' 05dd400e0 recover a removed line 94eff6246 v6.66 ec3fadeba update doc 00bfaaa7f add test of rao-int 0f2f1aaa6 support rio-int 6c047f480 detect rao-int f07c5c255 fix args of EmptyAllocator::alloc in test 04d3eb5f5 stop if a test script causes an error 1c1d2366f disable boost sample cfb1127c3 tweak 5fcbeb7c4 [sample] change the way of detection of boost bafc1ee60 CXX uses g++ as default value d8cabc6cb remove warning of blace bc73a0816 remove -fno-operator-names option 6989aea94 use CXX instead of g++ a7c5a1bd7 use English 1bfbd8c4d Fix incorrect format strings 0ecef5c28 Merge branch 'dev' 7556c20ba update doc a15709271 v6.65 3b83aab3e add detect_x32 to TARGET 8c64bbbc3 use gcc instead of dpkg for portability 5e9a9b96f test_avx512.sh runs on x32 8ae01b0c2 disable some tests on x32 83b3da217 x32 does not check large disp 693ab8c9d sizeof(void*) = 4 on x32, so disable the test 348e3e548 Merge branch 'dev' 11b9c4dc0 v6.64 459636196 add T_M_K flag to vpmov* 1d3722928 add include path 5e27eddae move CrearError from LabelManager::reset() to CodeGenerator::reset() f8ea5c28d Merge branch 'dev' 20b2b1eae v6.63 3706869f8 desc. of setDefaultEncoding d6f2d7577 add test of setDefaultEncoding 3b0a19c41 vpmadd52{h,l}uq for avx-ifma 95752ebd7 add tAVX_IFMA cd36e31ea [sample] show AMX_FP16/AVX_VNNI_INT8/AVX_NE_CONVERT e5858af27 add setDefaultEncoding 2f7fb0220 modify gen.cpp for AVX-NE-CONVERT/AVX-VNNI-INT8/AMX-FP16 1c5cb7efa add AVX-NE-CONVERT instructions 564fe9acd add AVX-VNNI-INT8 instructions cd14d07b1 add AMX-FP16 instruction 7811f593c Merge pull request #161 from scribam/patch-1 2218f6c08 Update changelog.md 7bccdbbb4 Merge pull request #160 from herumi/dev 5fcf87596 compile nasm-2.15 instead of apt install e31961ea8 v6.62 ff4f9e65c update doc 9c8fb81db disable wrong detection of gcc -Warray-bounds 6b7519659 add serialize e16582696 Merge branch 'JonLiu1993-vcpkg-installition' into dev edbb410fb Add vcpkg installation instructions 055d31242 Merge branch 'dev' 21ab98441 Merge branch 'akodanka-enable_CIV' into dev ed4d598e3 Changes to compile xbyak project for openvino b652430c4 mingw uses __cpuidex 48457bfa0 Merge branch 'dev' 29cb524d1 v6.61.2 1a9a0b0e1 avoid including algorithm header in xbyak_util.h 6fadefd04 Merge branch 'dev' fc1c18a9d update doc f7cae7f11 v6.61.1 6f5ec5cf3 Merge branch 'doyaGu-master' into dev 4554d6bb9 Fix error related to XBYAK_NOEXCEPT git-subtree-dir: externals/xbyak git-subtree-split: a1ac3750f9a639b5a6c6d6c7da4259b8d6790989
2022-12-30 23:05:02 +00:00
void cmpbexadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xE6, false); }
void cmpbxadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xE2, false); }
void cmplexadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xEE, false); }
void cmplxadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xEC, false); }
void cmpnbexadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xE7, false); }
void cmpnbxadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xE3, false); }
void cmpnlexadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xEF, false); }
void cmpnlxadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xED, false); }
void cmpnoxadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xE1, false); }
void cmpnpxadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xEB, false); }
void cmpnsxadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xE9, false); }
void cmpnzxadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xE5, false); }
void cmpoxadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xE0, false); }
void cmppxadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xEA, false); }
void cmpsxadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xE8, false); }
void cmpzxadd(const Address& addr, const Reg32e& r1, const Reg32e& r2) { opGpr(r1, addr, r2, T_66 | T_0F38, 0xE4, false); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void ldtilecfg(const Address& addr) { opVex(tmm0, &tmm0, addr, T_0F38 | T_W0, 0x49); }
void sttilecfg(const Address& addr) { opVex(tmm0, &tmm0, addr, T_66 | T_0F38 | T_W0, 0x49); }
void tileloadd(const Tmm& tm, const Address& addr) { opAMX(tm, addr, T_F2 | T_0F38 | T_W0, 0x4b); }
void tileloaddt1(const Tmm& tm, const Address& addr) { opAMX(tm, addr, T_66 | T_0F38 | T_W0, 0x4b); }
void tilerelease() { db(0xc4); db(0xe2); db(0x78); db(0x49); db(0xc0); }
void tilestored(const Address& addr, const Tmm& tm) { opVex(tm, &tmm0, addr, T_F3 | T_0F38 | T_W0, 0x4b); }
void tilezero(const Tmm& Tmm) { opVex(Tmm, &tmm0, tmm0, T_F2 | T_0F38 | T_W0, 0x49); }
void tdpbssd(const Tmm& x1, const Tmm& x2, const Tmm& x3) { opVex(x1, &x3, x2, T_F2 | T_0F38 | T_W0, 0x5e); }
void tdpbsud(const Tmm& x1, const Tmm& x2, const Tmm& x3) { opVex(x1, &x3, x2, T_F3 | T_0F38 | T_W0, 0x5e); }
void tdpbusd(const Tmm& x1, const Tmm& x2, const Tmm& x3) { opVex(x1, &x3, x2, T_66 | T_0F38 | T_W0, 0x5e); }
void tdpbuud(const Tmm& x1, const Tmm& x2, const Tmm& x3) { opVex(x1, &x3, x2, T_0F38 | T_W0, 0x5e); }
Squashed 'externals/xbyak/' changes from 88f2f771f..a1ac3750f a1ac3750f Merge branch 'dev' 5f4ba971f v6.68 cac2c175f update doc 1b08a8248 add test of prefetchiti bef70d9b1 add prefetchit{0,1} f66785876 add detection of prefetchiti 124617ac9 Merge branch 'dev' 1bce8be17 Merge commit 'fbb18f6' into dev bebfe64a3 [test] X32 is optional and remove CFLAGS d8c332cef Merge branch 'Tachi107-small-changes' into dev fbb18f69d ci: use containers 8a974696c test: simplify makefile 1efe9fe7c Merge branch 'dev' bb70083e6 v6.67 436e452d7 update doc 47ff6ef42 I get an error after all on GitHub action 445c0dcec add test of CMPccXADD c9347907d add CMPccXADD dc792cc56 add detection of cmpccxadd 62be84cc8 fix detection of boost b5ac7b0f0 Merge branch 'dev' 05dd400e0 recover a removed line 94eff6246 v6.66 ec3fadeba update doc 00bfaaa7f add test of rao-int 0f2f1aaa6 support rio-int 6c047f480 detect rao-int f07c5c255 fix args of EmptyAllocator::alloc in test 04d3eb5f5 stop if a test script causes an error 1c1d2366f disable boost sample cfb1127c3 tweak 5fcbeb7c4 [sample] change the way of detection of boost bafc1ee60 CXX uses g++ as default value d8cabc6cb remove warning of blace bc73a0816 remove -fno-operator-names option 6989aea94 use CXX instead of g++ a7c5a1bd7 use English 1bfbd8c4d Fix incorrect format strings 0ecef5c28 Merge branch 'dev' 7556c20ba update doc a15709271 v6.65 3b83aab3e add detect_x32 to TARGET 8c64bbbc3 use gcc instead of dpkg for portability 5e9a9b96f test_avx512.sh runs on x32 8ae01b0c2 disable some tests on x32 83b3da217 x32 does not check large disp 693ab8c9d sizeof(void*) = 4 on x32, so disable the test 348e3e548 Merge branch 'dev' 11b9c4dc0 v6.64 459636196 add T_M_K flag to vpmov* 1d3722928 add include path 5e27eddae move CrearError from LabelManager::reset() to CodeGenerator::reset() f8ea5c28d Merge branch 'dev' 20b2b1eae v6.63 3706869f8 desc. of setDefaultEncoding d6f2d7577 add test of setDefaultEncoding 3b0a19c41 vpmadd52{h,l}uq for avx-ifma 95752ebd7 add tAVX_IFMA cd36e31ea [sample] show AMX_FP16/AVX_VNNI_INT8/AVX_NE_CONVERT e5858af27 add setDefaultEncoding 2f7fb0220 modify gen.cpp for AVX-NE-CONVERT/AVX-VNNI-INT8/AMX-FP16 1c5cb7efa add AVX-NE-CONVERT instructions 564fe9acd add AVX-VNNI-INT8 instructions cd14d07b1 add AMX-FP16 instruction 7811f593c Merge pull request #161 from scribam/patch-1 2218f6c08 Update changelog.md 7bccdbbb4 Merge pull request #160 from herumi/dev 5fcf87596 compile nasm-2.15 instead of apt install e31961ea8 v6.62 ff4f9e65c update doc 9c8fb81db disable wrong detection of gcc -Warray-bounds 6b7519659 add serialize e16582696 Merge branch 'JonLiu1993-vcpkg-installition' into dev edbb410fb Add vcpkg installation instructions 055d31242 Merge branch 'dev' 21ab98441 Merge branch 'akodanka-enable_CIV' into dev ed4d598e3 Changes to compile xbyak project for openvino b652430c4 mingw uses __cpuidex 48457bfa0 Merge branch 'dev' 29cb524d1 v6.61.2 1a9a0b0e1 avoid including algorithm header in xbyak_util.h 6fadefd04 Merge branch 'dev' fc1c18a9d update doc f7cae7f11 v6.61.1 6f5ec5cf3 Merge branch 'doyaGu-master' into dev 4554d6bb9 Fix error related to XBYAK_NOEXCEPT git-subtree-dir: externals/xbyak git-subtree-split: a1ac3750f9a639b5a6c6d6c7da4259b8d6790989
2022-12-30 23:05:02 +00:00
void tdpfp16ps(const Tmm &x1, const Tmm &x2, const Tmm &x3) { opVex(x1, &x3, x2, T_F2 | T_0F38 | T_W0, 0x5c); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void tdpbf16ps(const Tmm& x1, const Tmm& x2, const Tmm& x3) { opVex(x1, &x3, x2, T_F3 | T_0F38 | T_W0, 0x5c); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
#else
void jcxz(std::string label) { db(0x67); opJmp(label, T_SHORT, 0xe3, 0, 0); }
void jcxz(const Label& label) { db(0x67); opJmp(label, T_SHORT, 0xe3, 0, 0); }
void jecxz(std::string label) { opJmp(label, T_SHORT, 0xe3, 0, 0); }
void jecxz(const Label& label) { opJmp(label, T_SHORT, 0xe3, 0, 0); }
void aaa() { db(0x37); }
void aad() { db(0xD5); db(0x0A); }
void aam() { db(0xD4); db(0x0A); }
void aas() { db(0x3F); }
void daa() { db(0x27); }
void das() { db(0x2F); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void into() { db(0xCE); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void popad() { db(0x61); }
void popfd() { db(0x9D); }
void pusha() { db(0x60); }
void pushad() { db(0x60); }
void pushfd() { db(0x9C); }
void popa() { db(0x61); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void lds(const Reg& reg, const Address& addr) { opLoadSeg(addr, reg, 0xC5, 0x100); }
void les(const Reg& reg, const Address& addr) { opLoadSeg(addr, reg, 0xC4, 0x100); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
#endif
#ifndef XBYAK_NO_OP_NAMES
void and(const Operand& op1, const Operand& op2) { and_(op1, op2); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void and(const Operand& op, uint32_t imm) { and_(op, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void or(const Operand& op1, const Operand& op2) { or_(op1, op2); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void or(const Operand& op, uint32_t imm) { or_(op, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void xor(const Operand& op1, const Operand& op2) { xor_(op1, op2); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void xor(const Operand& op, uint32_t imm) { xor_(op, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void not(const Operand& op) { not_(op); }
#endif
#ifndef XBYAK_DISABLE_AVX512
void kaddb(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_66 | T_W0, 0x4A); }
void kaddd(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_66 | T_W1, 0x4A); }
void kaddq(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W1, 0x4A); }
void kaddw(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W0, 0x4A); }
void kandb(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_66 | T_W0, 0x41); }
void kandd(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_66 | T_W1, 0x41); }
void kandnb(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_66 | T_W0, 0x42); }
void kandnd(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_66 | T_W1, 0x42); }
void kandnq(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W1, 0x42); }
void kandnw(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W0, 0x42); }
void kandq(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W1, 0x41); }
void kandw(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W0, 0x41); }
void kmovb(const Address& addr, const Opmask& k) { opVex(k, 0, addr, T_L0 | T_0F | T_66 | T_W0, 0x91); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void kmovb(const Opmask& k, const Operand& op) { if (!op.isMEM() && !op.isOPMASK()) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(k, 0, op, T_L0 | T_0F | T_66 | T_W0, 0x90); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void kmovb(const Opmask& k, const Reg32& r) { opVex(k, 0, r, T_L0 | T_0F | T_66 | T_W0, 0x92); }
void kmovb(const Reg32& r, const Opmask& k) { opVex(r, 0, k, T_L0 | T_0F | T_66 | T_W0, 0x93); }
void kmovd(const Address& addr, const Opmask& k) { opVex(k, 0, addr, T_L0 | T_0F | T_66 | T_W1, 0x91); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void kmovd(const Opmask& k, const Operand& op) { if (!op.isMEM() && !op.isOPMASK()) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(k, 0, op, T_L0 | T_0F | T_66 | T_W1, 0x90); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void kmovd(const Opmask& k, const Reg32& r) { opVex(k, 0, r, T_L0 | T_0F | T_F2 | T_W0, 0x92); }
void kmovd(const Reg32& r, const Opmask& k) { opVex(r, 0, k, T_L0 | T_0F | T_F2 | T_W0, 0x93); }
void kmovq(const Address& addr, const Opmask& k) { opVex(k, 0, addr, T_L0 | T_0F | T_W1, 0x91); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void kmovq(const Opmask& k, const Operand& op) { if (!op.isMEM() && !op.isOPMASK()) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(k, 0, op, T_L0 | T_0F | T_W1, 0x90); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void kmovw(const Address& addr, const Opmask& k) { opVex(k, 0, addr, T_L0 | T_0F | T_W0, 0x91); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void kmovw(const Opmask& k, const Operand& op) { if (!op.isMEM() && !op.isOPMASK()) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(k, 0, op, T_L0 | T_0F | T_W0, 0x90); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void kmovw(const Opmask& k, const Reg32& r) { opVex(k, 0, r, T_L0 | T_0F | T_W0, 0x92); }
void kmovw(const Reg32& r, const Opmask& k) { opVex(r, 0, k, T_L0 | T_0F | T_W0, 0x93); }
void knotb(const Opmask& r1, const Opmask& r2) { opVex(r1, 0, r2, T_0F | T_66 | T_W0, 0x44); }
void knotd(const Opmask& r1, const Opmask& r2) { opVex(r1, 0, r2, T_0F | T_66 | T_W1, 0x44); }
void knotq(const Opmask& r1, const Opmask& r2) { opVex(r1, 0, r2, T_0F | T_W1, 0x44); }
void knotw(const Opmask& r1, const Opmask& r2) { opVex(r1, 0, r2, T_0F | T_W0, 0x44); }
void korb(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_66 | T_W0, 0x45); }
void kord(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_66 | T_W1, 0x45); }
void korq(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W1, 0x45); }
void kortestb(const Opmask& r1, const Opmask& r2) { opVex(r1, 0, r2, T_0F | T_66 | T_W0, 0x98); }
void kortestd(const Opmask& r1, const Opmask& r2) { opVex(r1, 0, r2, T_0F | T_66 | T_W1, 0x98); }
void kortestq(const Opmask& r1, const Opmask& r2) { opVex(r1, 0, r2, T_0F | T_W1, 0x98); }
void kortestw(const Opmask& r1, const Opmask& r2) { opVex(r1, 0, r2, T_0F | T_W0, 0x98); }
void korw(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W0, 0x45); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void kshiftlb(const Opmask& r1, const Opmask& r2, uint8_t imm) { opVex(r1, 0, r2, T_66 | T_0F3A | T_W0, 0x32, imm); }
void kshiftld(const Opmask& r1, const Opmask& r2, uint8_t imm) { opVex(r1, 0, r2, T_66 | T_0F3A | T_W0, 0x33, imm); }
void kshiftlq(const Opmask& r1, const Opmask& r2, uint8_t imm) { opVex(r1, 0, r2, T_66 | T_0F3A | T_W1, 0x33, imm); }
void kshiftlw(const Opmask& r1, const Opmask& r2, uint8_t imm) { opVex(r1, 0, r2, T_66 | T_0F3A | T_W1, 0x32, imm); }
void kshiftrb(const Opmask& r1, const Opmask& r2, uint8_t imm) { opVex(r1, 0, r2, T_66 | T_0F3A | T_W0, 0x30, imm); }
void kshiftrd(const Opmask& r1, const Opmask& r2, uint8_t imm) { opVex(r1, 0, r2, T_66 | T_0F3A | T_W0, 0x31, imm); }
void kshiftrq(const Opmask& r1, const Opmask& r2, uint8_t imm) { opVex(r1, 0, r2, T_66 | T_0F3A | T_W1, 0x31, imm); }
void kshiftrw(const Opmask& r1, const Opmask& r2, uint8_t imm) { opVex(r1, 0, r2, T_66 | T_0F3A | T_W1, 0x30, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void ktestb(const Opmask& r1, const Opmask& r2) { opVex(r1, 0, r2, T_0F | T_66 | T_W0, 0x99); }
void ktestd(const Opmask& r1, const Opmask& r2) { opVex(r1, 0, r2, T_0F | T_66 | T_W1, 0x99); }
void ktestq(const Opmask& r1, const Opmask& r2) { opVex(r1, 0, r2, T_0F | T_W1, 0x99); }
void ktestw(const Opmask& r1, const Opmask& r2) { opVex(r1, 0, r2, T_0F | T_W0, 0x99); }
void kunpckbw(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_66 | T_W0, 0x4B); }
void kunpckdq(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W1, 0x4B); }
void kunpckwd(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W0, 0x4B); }
void kxnorb(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_66 | T_W0, 0x46); }
void kxnord(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_66 | T_W1, 0x46); }
void kxnorq(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W1, 0x46); }
void kxnorw(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W0, 0x46); }
void kxorb(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_66 | T_W0, 0x47); }
void kxord(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_66 | T_W1, 0x47); }
void kxorq(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W1, 0x47); }
void kxorw(const Opmask& r1, const Opmask& r2, const Opmask& r3) { opVex(r1, &r2, r3, T_L1 | T_0F | T_W0, 0x47); }
void v4fmaddps(const Zmm& z1, const Zmm& z2, const Address& addr) { opAVX_X_X_XM(z1, z2, addr, T_0F38 | T_F2 | T_EW0 | T_YMM | T_MUST_EVEX | T_N16, 0x9A); }
void v4fmaddss(const Xmm& x1, const Xmm& x2, const Address& addr) { opAVX_X_X_XM(x1, x2, addr, T_0F38 | T_F2 | T_EW0 | T_MUST_EVEX | T_N16, 0x9B); }
void v4fnmaddps(const Zmm& z1, const Zmm& z2, const Address& addr) { opAVX_X_X_XM(z1, z2, addr, T_0F38 | T_F2 | T_EW0 | T_YMM | T_MUST_EVEX | T_N16, 0xAA); }
void v4fnmaddss(const Xmm& x1, const Xmm& x2, const Address& addr) { opAVX_X_X_XM(x1, x2, addr, T_0F38 | T_F2 | T_EW0 | T_MUST_EVEX | T_N16, 0xAB); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vaddph(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_MAP5 | T_EW0 | T_YMM | T_MUST_EVEX | T_ER_Z | T_B16, 0x58); }
void vaddsh(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_MAP5 | T_F3 | T_EW0 | T_MUST_EVEX | T_ER_X | T_N2, 0x58); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void valignd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX, 0x03, imm); }
void valignq(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX, 0x03, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vblendmpd(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, 0x65); }
void vblendmps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x65); }
void vbroadcastf32x2(const Ymm& y, const Operand& op) { opAVX_X_XM_IMM(y, op, T_66 | T_0F38 | T_YMM | T_MUST_EVEX | T_EW0 | T_N8, 0x19); }
void vbroadcastf32x4(const Ymm& y, const Address& addr) { opAVX_X_XM_IMM(y, addr, T_66 | T_0F38 | T_YMM | T_MUST_EVEX | T_EW0 | T_N16, 0x1A); }
void vbroadcastf32x8(const Zmm& y, const Address& addr) { opAVX_X_XM_IMM(y, addr, T_66 | T_0F38 | T_YMM | T_MUST_EVEX | T_EW0 | T_N32, 0x1B); }
void vbroadcastf64x2(const Ymm& y, const Address& addr) { opAVX_X_XM_IMM(y, addr, T_66 | T_0F38 | T_YMM | T_MUST_EVEX | T_EW1 | T_N16, 0x1A); }
void vbroadcastf64x4(const Zmm& y, const Address& addr) { opAVX_X_XM_IMM(y, addr, T_66 | T_0F38 | T_YMM | T_MUST_EVEX | T_EW1 | T_N32, 0x1B); }
void vbroadcasti32x2(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_YMM | T_MUST_EVEX | T_EW0 | T_N8, 0x59); }
void vbroadcasti32x4(const Ymm& y, const Operand& op) { opAVX_X_XM_IMM(y, op, T_66 | T_0F38 | T_YMM | T_MUST_EVEX | T_EW0 | T_N16, 0x5A); }
void vbroadcasti32x8(const Zmm& z, const Operand& op) { opAVX_X_XM_IMM(z, op, T_66 | T_0F38 | T_YMM | T_MUST_EVEX | T_EW0 | T_N32, 0x5B); }
void vbroadcasti64x2(const Ymm& y, const Operand& op) { opAVX_X_XM_IMM(y, op, T_66 | T_0F38 | T_YMM | T_MUST_EVEX | T_EW1 | T_N16, 0x5A); }
void vbroadcasti64x4(const Zmm& z, const Operand& op) { opAVX_X_XM_IMM(z, op, T_66 | T_0F38 | T_YMM | T_MUST_EVEX | T_EW1 | T_N32, 0x5B); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcmpeq_ospd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 16); }
void vcmpeq_osps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 16); }
void vcmpeq_ossd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 16); }
void vcmpeq_osss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 16); }
void vcmpeq_uqpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 8); }
void vcmpeq_uqps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 8); }
void vcmpeq_uqsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 8); }
void vcmpeq_uqss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 8); }
void vcmpeq_uspd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 24); }
void vcmpeq_usps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 24); }
void vcmpeq_ussd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 24); }
void vcmpeq_usss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 24); }
void vcmpeqpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 0); }
void vcmpeqps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 0); }
void vcmpeqsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 0); }
void vcmpeqss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 0); }
void vcmpfalse_ospd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 27); }
void vcmpfalse_osps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 27); }
void vcmpfalse_ossd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 27); }
void vcmpfalse_osss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 27); }
void vcmpfalsepd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 11); }
void vcmpfalseps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 11); }
void vcmpfalsesd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 11); }
void vcmpfalsess(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 11); }
void vcmpge_oqpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 29); }
void vcmpge_oqps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 29); }
void vcmpge_oqsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 29); }
void vcmpge_oqss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 29); }
void vcmpgepd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 13); }
void vcmpgeps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 13); }
void vcmpgesd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 13); }
void vcmpgess(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 13); }
void vcmpgt_oqpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 30); }
void vcmpgt_oqps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 30); }
void vcmpgt_oqsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 30); }
void vcmpgt_oqss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 30); }
void vcmpgtpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 14); }
void vcmpgtps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 14); }
void vcmpgtsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 14); }
void vcmpgtss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 14); }
void vcmple_oqpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 18); }
void vcmple_oqps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 18); }
void vcmple_oqsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 18); }
void vcmple_oqss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 18); }
void vcmplepd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 2); }
void vcmpleps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 2); }
void vcmplesd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 2); }
void vcmpless(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 2); }
void vcmplt_oqpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 17); }
void vcmplt_oqps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 17); }
void vcmplt_oqsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 17); }
void vcmplt_oqss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 17); }
void vcmpltpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 1); }
void vcmpltps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 1); }
void vcmpltsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 1); }
void vcmpltss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 1); }
void vcmpneq_oqpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 12); }
void vcmpneq_oqps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 12); }
void vcmpneq_oqsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 12); }
void vcmpneq_oqss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 12); }
void vcmpneq_ospd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 28); }
void vcmpneq_osps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 28); }
void vcmpneq_ossd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 28); }
void vcmpneq_osss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 28); }
void vcmpneq_uspd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 20); }
void vcmpneq_usps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 20); }
void vcmpneq_ussd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 20); }
void vcmpneq_usss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 20); }
void vcmpneqpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 4); }
void vcmpneqps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 4); }
void vcmpneqsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 4); }
void vcmpneqss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 4); }
void vcmpnge_uqpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 25); }
void vcmpnge_uqps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 25); }
void vcmpnge_uqsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 25); }
void vcmpnge_uqss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 25); }
void vcmpngepd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 9); }
void vcmpngeps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 9); }
void vcmpngesd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 9); }
void vcmpngess(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 9); }
void vcmpngt_uqpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 26); }
void vcmpngt_uqps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 26); }
void vcmpngt_uqsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 26); }
void vcmpngt_uqss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 26); }
void vcmpngtpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 10); }
void vcmpngtps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 10); }
void vcmpngtsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 10); }
void vcmpngtss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 10); }
void vcmpnle_uqpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 22); }
void vcmpnle_uqps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 22); }
void vcmpnle_uqsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 22); }
void vcmpnle_uqss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 22); }
void vcmpnlepd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 6); }
void vcmpnleps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 6); }
void vcmpnlesd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 6); }
void vcmpnless(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 6); }
void vcmpnlt_uqpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 21); }
void vcmpnlt_uqps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 21); }
void vcmpnlt_uqsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 21); }
void vcmpnlt_uqss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 21); }
void vcmpnltpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 5); }
void vcmpnltps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 5); }
void vcmpnltsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 5); }
void vcmpnltss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 5); }
void vcmpord_spd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 23); }
void vcmpord_sps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 23); }
void vcmpord_ssd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 23); }
void vcmpord_sss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 23); }
void vcmpordpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 7); }
void vcmpordps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 7); }
void vcmpordsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 7); }
void vcmpordss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 7); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vcmppd(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_66 | T_0F | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B64, 0xC2, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcmpph(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_0F3A | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B16, 0xC2, imm); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vcmpps(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_0F | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B32, 0xC2, imm); }
void vcmpsd(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_N8 | T_F2 | T_0F | T_EW1 | T_SAE_Z | T_MUST_EVEX, 0xC2, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcmpsh(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_N2 | T_F3 | T_0F3A | T_EW0 | T_SAE_X | T_MUST_EVEX, 0xC2, imm); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vcmpss(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_N4 | T_F3 | T_0F | T_EW0 | T_SAE_Z | T_MUST_EVEX, 0xC2, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcmptrue_uspd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 31); }
void vcmptrue_usps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 31); }
void vcmptrue_ussd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 31); }
void vcmptrue_usss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 31); }
void vcmptruepd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 15); }
void vcmptrueps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 15); }
void vcmptruesd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 15); }
void vcmptruess(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 15); }
void vcmpunord_spd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 19); }
void vcmpunord_sps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 19); }
void vcmpunord_ssd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 19); }
void vcmpunord_sss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 19); }
void vcmpunordpd(const Opmask& k, const Xmm& x, const Operand& op) { vcmppd(k, x, op, 3); }
void vcmpunordps(const Opmask& k, const Xmm& x, const Operand& op) { vcmpps(k, x, op, 3); }
void vcmpunordsd(const Opmask& k, const Xmm& x, const Operand& op) { vcmpsd(k, x, op, 3); }
void vcmpunordss(const Opmask& k, const Xmm& x, const Operand& op) { vcmpss(k, x, op, 3); }
void vcomish(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_MAP5 | T_MUST_EVEX | T_EW0 | T_SAE_X | T_N2, 0x2F); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
void vcompressb(const Operand& op, const Xmm& x) { opAVX_X_XM_IMM(x, op, T_N1 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x63); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcompresspd(const Operand& op, const Xmm& x) { opAVX_X_XM_IMM(x, op, T_N8 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX, 0x8A); }
void vcompressps(const Operand& op, const Xmm& x) { opAVX_X_XM_IMM(x, op, T_N4 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x8A); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
void vcompressw(const Operand& op, const Xmm& x) { opAVX_X_XM_IMM(x, op, T_N2 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX, 0x63); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcvtdq2ph(const Xmm& x, const Operand& op) { checkCvt4(x, op); opCvt(x, op, T_N16 | T_N_VL | T_MAP5 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B32, 0x5B); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vcvtne2ps2bf16(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_F2 | T_0F38 | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B32, 0x72); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcvtpd2ph(const Xmm& x, const Operand& op) { opCvt5(x, op, T_N16 | T_N_VL | T_66 | T_MAP5 | T_EW1 | T_ER_Z | T_MUST_EVEX | T_B64, 0x5A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcvtpd2qq(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F | T_EW1 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B64, 0x7B); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcvtpd2udq(const Xmm& x, const Operand& op) { opCvt2(x, op, T_0F | T_EW1 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B64, 0x79); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcvtpd2uqq(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F | T_EW1 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B64, 0x79); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcvtph2dq(const Xmm& x, const Operand& op) { checkCvt1(x, op); opVex(x, 0, op, T_N8 | T_N_VL | T_66 | T_MAP5 | T_EW0 | T_YMM | T_ER_Y | T_MUST_EVEX | T_B16, 0x5B); }
void vcvtph2pd(const Xmm& x, const Operand& op) { if (!op.isXMM() && !op.isMEM()) XBYAK_THROW(ERR_BAD_MEM_SIZE) opVex(x, 0, op, T_N4 | T_N_VL | T_MAP5 | T_EW0 | T_YMM | T_SAE_X | T_MUST_EVEX | T_B16, 0x5A); }
void vcvtph2psx(const Xmm& x, const Operand& op) { checkCvt1(x, op); opVex(x, 0, op, T_N8 | T_N_VL | T_66 | T_MAP6 | T_EW0 | T_YMM | T_SAE_Y | T_MUST_EVEX | T_B16, 0x13); }
void vcvtph2qq(const Xmm& x, const Operand& op) { if (!op.isXMM() && !op.isMEM()) XBYAK_THROW(ERR_BAD_MEM_SIZE) opVex(x, 0, op, T_N4 | T_N_VL | T_66 | T_MAP5 | T_EW0 | T_YMM | T_ER_X | T_MUST_EVEX | T_B16, 0x7B); }
void vcvtph2udq(const Xmm& x, const Operand& op) { checkCvt1(x, op); opVex(x, 0, op, T_N8 | T_N_VL | T_MAP5 | T_EW0 | T_YMM | T_ER_Y | T_MUST_EVEX | T_B16, 0x79); }
void vcvtph2uqq(const Xmm& x, const Operand& op) { if (!op.isXMM() && !op.isMEM()) XBYAK_THROW(ERR_BAD_MEM_SIZE) opVex(x, 0, op, T_N4 | T_N_VL | T_66 | T_MAP5 | T_EW0 | T_YMM | T_ER_X | T_MUST_EVEX | T_B16, 0x79); }
void vcvtph2uw(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_MAP5 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0x7D); }
void vcvtph2w(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_MAP5 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0x7D); }
void vcvtps2phx(const Xmm& x, const Operand& op) { checkCvt4(x, op); opCvt(x, op, T_N16 | T_N_VL | T_66 | T_MAP5 | T_EW0 | T_ER_Z | T_MUST_EVEX | T_B32, 0x1D); }
void vcvtps2qq(const Xmm& x, const Operand& op) { checkCvt1(x, op); opVex(x, 0, op, T_N8 | T_N_VL | T_66 | T_0F | T_EW0 | T_YMM | T_ER_Y | T_MUST_EVEX | T_B32, 0x7B); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcvtps2udq(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_0F | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B32, 0x79); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcvtps2uqq(const Xmm& x, const Operand& op) { checkCvt1(x, op); opVex(x, 0, op, T_N8 | T_N_VL | T_66 | T_0F | T_EW0 | T_YMM | T_ER_Y | T_MUST_EVEX | T_B32, 0x79); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcvtqq2pd(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_F3 | T_0F | T_EW1 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B64, 0xE6); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcvtqq2ph(const Xmm& x, const Operand& op) { opCvt5(x, op, T_N16 | T_N_VL | T_MAP5 | T_EW1 | T_ER_Z | T_MUST_EVEX | T_B64, 0x5B); }
void vcvtqq2ps(const Xmm& x, const Operand& op) { opCvt2(x, op, T_0F | T_EW1 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B64, 0x5B); }
void vcvtsd2sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_F2 | T_MAP5 | T_EW1 | T_ER_X | T_MUST_EVEX, 0x5A); }
void vcvtsd2usi(const Reg32e& r, const Operand& op) { int type = (T_N8 | T_F2 | T_0F | T_ER_X | T_MUST_EVEX) | (r.isREG(64) ? T_EW1 : T_EW0); opVex(r, &xm0, op, type, 0x79); }
void vcvtsh2sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_F3 | T_MAP5 | T_EW0 | T_SAE_X | T_MUST_EVEX, 0x5A); }
void vcvtsh2si(const Reg32e& r, const Operand& op) { int type = (T_N2 | T_F3 | T_MAP5 | T_ER_X | T_MUST_EVEX) | (r.isREG(64) ? T_EW1 : T_EW0); opVex(r, &xm0, op, type, 0x2D); }
void vcvtsh2ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_MAP6 | T_EW0 | T_SAE_X | T_MUST_EVEX, 0x13); }
void vcvtsh2usi(const Reg32e& r, const Operand& op) { int type = (T_N2 | T_F3 | T_MAP5 | T_ER_X | T_MUST_EVEX) | (r.isREG(64) ? T_EW1 : T_EW0); opVex(r, &xm0, op, type, 0x79); }
void vcvtsi2sh(const Xmm& x1, const Xmm& x2, const Operand& op) { if (!(x1.isXMM() && x2.isXMM() && op.isBit(32|64))) XBYAK_THROW(ERR_BAD_COMBINATION) int type = (T_F3 | T_MAP5 | T_ER_R | T_MUST_EVEX | T_M_K) | (op.isBit(32) ? (T_EW0 | T_N4) : (T_EW1 | T_N8)); opVex(x1, &x2, op, type, 0x2A); }
void vcvtss2sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_MAP5 | T_EW0 | T_ER_X | T_MUST_EVEX, 0x1D); }
void vcvtss2usi(const Reg32e& r, const Operand& op) { int type = (T_N4 | T_F3 | T_0F | T_ER_X | T_MUST_EVEX) | (r.isREG(64) ? T_EW1 : T_EW0); opVex(r, &xm0, op, type, 0x79); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcvttpd2qq(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B64, 0x7A); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcvttpd2udq(const Xmm& x, const Operand& op) { opCvt2(x, op, T_0F | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B64, 0x78); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcvttpd2uqq(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B64, 0x78); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcvttph2dq(const Xmm& x, const Operand& op) { checkCvt1(x, op); opVex(x, 0, op, T_N8 | T_N_VL | T_F3 | T_MAP5 | T_EW0 | T_YMM | T_SAE_Y | T_MUST_EVEX | T_B16, 0x5B); }
void vcvttph2qq(const Xmm& x, const Operand& op) { if (!op.isXMM() && !op.isMEM()) XBYAK_THROW(ERR_BAD_MEM_SIZE) opVex(x, 0, op, T_N4 | T_N_VL | T_66 | T_MAP5 | T_EW0 | T_YMM | T_SAE_X | T_MUST_EVEX | T_B16, 0x7A); }
void vcvttph2udq(const Xmm& x, const Operand& op) { checkCvt1(x, op); opVex(x, 0, op, T_N8 | T_N_VL | T_MAP5 | T_EW0 | T_YMM | T_SAE_Y | T_MUST_EVEX | T_B16, 0x78); }
void vcvttph2uqq(const Xmm& x, const Operand& op) { if (!op.isXMM() && !op.isMEM()) XBYAK_THROW(ERR_BAD_MEM_SIZE) opVex(x, 0, op, T_N4 | T_N_VL | T_66 | T_MAP5 | T_EW0 | T_YMM | T_SAE_X | T_MUST_EVEX | T_B16, 0x78); }
void vcvttph2uw(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_MAP5 | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B16, 0x7C); }
void vcvttph2w(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_MAP5 | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B16, 0x7C); }
void vcvttps2qq(const Xmm& x, const Operand& op) { checkCvt1(x, op); opVex(x, 0, op, T_N8 | T_N_VL | T_66 | T_0F | T_EW0 | T_YMM | T_SAE_Y | T_MUST_EVEX | T_B32, 0x7A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcvttps2udq(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_0F | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B32, 0x78); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcvttps2uqq(const Xmm& x, const Operand& op) { checkCvt1(x, op); opVex(x, 0, op, T_N8 | T_N_VL | T_66 | T_0F | T_EW0 | T_YMM | T_SAE_Y | T_MUST_EVEX | T_B32, 0x78); }
void vcvttsd2usi(const Reg32e& r, const Operand& op) { int type = (T_N8 | T_F2 | T_0F | T_SAE_X | T_MUST_EVEX) | (r.isREG(64) ? T_EW1 : T_EW0); opVex(r, &xm0, op, type, 0x78); }
void vcvttsh2si(const Reg32e& r, const Operand& op) { int type = (T_N2 | T_F3 | T_MAP5 | T_EW0 | T_SAE_X | T_MUST_EVEX) | (r.isREG(64) ? T_EW1 : T_EW0); opVex(r, &xm0, op, type, 0x2C); }
void vcvttsh2usi(const Reg32e& r, const Operand& op) { int type = (T_N2 | T_F3 | T_MAP5 | T_EW0 | T_SAE_X | T_MUST_EVEX) | (r.isREG(64) ? T_EW1 : T_EW0); opVex(r, &xm0, op, type, 0x78); }
void vcvttss2usi(const Reg32e& r, const Operand& op) { int type = (T_N4 | T_F3 | T_0F | T_SAE_X | T_MUST_EVEX) | (r.isREG(64) ? T_EW1 : T_EW0); opVex(r, &xm0, op, type, 0x78); }
void vcvtudq2pd(const Xmm& x, const Operand& op) { checkCvt1(x, op); opVex(x, 0, op, T_N8 | T_N_VL | T_F3 | T_0F | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x7A); }
void vcvtudq2ph(const Xmm& x, const Operand& op) { checkCvt4(x, op); opCvt(x, op, T_N16 | T_N_VL | T_F2 | T_MAP5 | T_EW0 | T_ER_Z | T_MUST_EVEX | T_B32, 0x7A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcvtudq2ps(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_F2 | T_0F | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B32, 0x7A); }
void vcvtuqq2pd(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_F3 | T_0F | T_EW1 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B64, 0x7A); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcvtuqq2ph(const Xmm& x, const Operand& op) { opCvt5(x, op, T_N16 | T_N_VL | T_F2 | T_MAP5 | T_EW1 | T_ER_Z | T_MUST_EVEX | T_B64, 0x7A); }
void vcvtuqq2ps(const Xmm& x, const Operand& op) { opCvt2(x, op, T_F2 | T_0F | T_EW1 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B64, 0x7A); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcvtusi2sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opCvt3(x1, x2, op, T_F2 | T_0F | T_MUST_EVEX, T_W1 | T_EW1 | T_ER_X | T_N8, T_W0 | T_EW0 | T_N4, 0x7B); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcvtusi2sh(const Xmm& x1, const Xmm& x2, const Operand& op) { if (!(x1.isXMM() && x2.isXMM() && op.isBit(32|64))) XBYAK_THROW(ERR_BAD_COMBINATION) int type = (T_F3 | T_MAP5 | T_ER_R | T_MUST_EVEX | T_M_K) | (op.isBit(32) ? (T_EW0 | T_N4) : (T_EW1 | T_N8)); opVex(x1, &x2, op, type, 0x7B); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vcvtusi2ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opCvt3(x1, x2, op, T_F3 | T_0F | T_MUST_EVEX | T_ER_X, T_W1 | T_EW1 | T_N8, T_W0 | T_EW0 | T_N4, 0x7B); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vcvtuw2ph(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_F2 | T_MAP5 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0x7D); }
void vcvtw2ph(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_F3 | T_MAP5 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0x7D); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vdbpsadbw(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX, 0x42, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vdivph(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_MAP5 | T_EW0 | T_YMM | T_MUST_EVEX | T_ER_Z | T_B16, 0x5E); }
void vdivsh(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_MAP5 | T_F3 | T_EW0 | T_MUST_EVEX | T_ER_X | T_N2, 0x5E); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vdpbf16ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_F3 | T_0F38 | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B32, 0x52); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vexp2pd(const Zmm& z, const Operand& op) { opAVX_X_XM_IMM(z, op, T_66 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW1 | T_B64 | T_SAE_Z, 0xC8); }
void vexp2ps(const Zmm& z, const Operand& op) { opAVX_X_XM_IMM(z, op, T_66 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_B32 | T_SAE_Z, 0xC8); }
void vexpandpd(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_N8 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX, 0x88); }
void vexpandps(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_N4 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x88); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vextractf32x4(const Operand& op, const Ymm& r, uint8_t imm) { if (!op.is(Operand::MEM | Operand::XMM)) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r, 0, op, T_N16 | T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX, 0x19, imm); }
void vextractf32x8(const Operand& op, const Zmm& r, uint8_t imm) { if (!op.is(Operand::MEM | Operand::YMM)) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r, 0, op, T_N32 | T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX, 0x1B, imm); }
void vextractf64x2(const Operand& op, const Ymm& r, uint8_t imm) { if (!op.is(Operand::MEM | Operand::XMM)) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r, 0, op, T_N16 | T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX, 0x19, imm); }
void vextractf64x4(const Operand& op, const Zmm& r, uint8_t imm) { if (!op.is(Operand::MEM | Operand::YMM)) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r, 0, op, T_N32 | T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX, 0x1B, imm); }
void vextracti32x4(const Operand& op, const Ymm& r, uint8_t imm) { if (!op.is(Operand::MEM | Operand::XMM)) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r, 0, op, T_N16 | T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX, 0x39, imm); }
void vextracti32x8(const Operand& op, const Zmm& r, uint8_t imm) { if (!op.is(Operand::MEM | Operand::YMM)) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r, 0, op, T_N32 | T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX, 0x3B, imm); }
void vextracti64x2(const Operand& op, const Ymm& r, uint8_t imm) { if (!op.is(Operand::MEM | Operand::XMM)) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r, 0, op, T_N16 | T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX, 0x39, imm); }
void vextracti64x4(const Operand& op, const Zmm& r, uint8_t imm) { if (!op.is(Operand::MEM | Operand::YMM)) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r, 0, op, T_N32 | T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX, 0x3B, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vfcmaddcph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_F2 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B32, 0x56); }
void vfcmulcph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_F2 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B32, 0xD6); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vfixupimmpd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B64, 0x54, imm); }
void vfixupimmps(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B32, 0x54, imm); }
void vfixupimmsd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F3A | T_EW1 | T_SAE_Z | T_MUST_EVEX, 0x55, imm); }
void vfixupimmss(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F3A | T_EW0 | T_SAE_Z | T_MUST_EVEX, 0x55, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vfmadd132ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0x98); }
void vfmadd132sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_ER_X | T_MUST_EVEX, 0x99); }
void vfmadd213ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0xA8); }
void vfmadd213sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_ER_X | T_MUST_EVEX, 0xA9); }
void vfmadd231ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0xB8); }
void vfmadd231sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_ER_X | T_MUST_EVEX, 0xB9); }
void vfmaddcph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_F3 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B32, 0x56); }
void vfmaddsub132ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0x96); }
void vfmaddsub213ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0xA6); }
void vfmaddsub231ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0xB6); }
void vfmsub132ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0x9A); }
void vfmsub132sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_ER_X | T_MUST_EVEX, 0x9B); }
void vfmsub213ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0xAA); }
void vfmsub213sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_ER_X | T_MUST_EVEX, 0xAB); }
void vfmsub231ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0xBA); }
void vfmsub231sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_ER_X | T_MUST_EVEX, 0xBB); }
void vfmsubadd132ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0x97); }
void vfmsubadd213ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0xA7); }
void vfmsubadd231ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0xB7); }
void vfmulcph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_F3 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B32, 0xD6); }
void vfnmadd132ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0x9C); }
void vfnmadd132sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_ER_X | T_MUST_EVEX, 0x9D); }
void vfnmadd213ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0xAC); }
void vfnmadd213sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_ER_X | T_MUST_EVEX, 0xAD); }
void vfnmadd231ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0xBC); }
void vfnmadd231sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_ER_X | T_MUST_EVEX, 0xBD); }
void vfnmsub132ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0x9E); }
void vfnmsub132sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_ER_X | T_MUST_EVEX, 0x9F); }
void vfnmsub213ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0xAE); }
void vfnmsub213sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_ER_X | T_MUST_EVEX, 0xAF); }
void vfnmsub231ph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0xBE); }
void vfnmsub231sh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_ER_X | T_MUST_EVEX, 0xBF); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vfpclasspd(const Opmask& k, const Operand& op, uint8_t imm) { if (!op.isBit(128|256|512)) XBYAK_THROW(ERR_BAD_MEM_SIZE) opVex(k.changeBit(op.getBit()), 0, op, T_66 | T_0F3A | T_MUST_EVEX | T_YMM | T_EW1 | T_B64, 0x66, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vfpclassph(const Opmask& k, const Operand& op, uint8_t imm) { if (!op.isBit(128|256|512)) XBYAK_THROW(ERR_BAD_MEM_SIZE) opVex(k.changeBit(op.getBit()), 0, op, T_0F3A | T_MUST_EVEX | T_YMM | T_EW0 | T_B16, 0x66, imm); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vfpclassps(const Opmask& k, const Operand& op, uint8_t imm) { if (!op.isBit(128|256|512)) XBYAK_THROW(ERR_BAD_MEM_SIZE) opVex(k.changeBit(op.getBit()), 0, op, T_66 | T_0F3A | T_MUST_EVEX | T_YMM | T_EW0 | T_B32, 0x66, imm); }
void vfpclasssd(const Opmask& k, const Operand& op, uint8_t imm) { if (!op.isXMEM()) XBYAK_THROW(ERR_BAD_MEM_SIZE) opVex(k, 0, op, T_66 | T_0F3A | T_MUST_EVEX | T_EW1 | T_N8, 0x67, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vfpclasssh(const Opmask& k, const Operand& op, uint8_t imm) { if (!op.isXMEM()) XBYAK_THROW(ERR_BAD_MEM_SIZE) opVex(k, 0, op, T_0F3A | T_MUST_EVEX | T_EW0 | T_N2, 0x67, imm); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vfpclassss(const Opmask& k, const Operand& op, uint8_t imm) { if (!op.isXMEM()) XBYAK_THROW(ERR_BAD_MEM_SIZE) opVex(k, 0, op, T_66 | T_0F3A | T_MUST_EVEX | T_EW0 | T_N4, 0x67, imm); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vgatherdpd(const Xmm& x, const Address& addr) { opGather2(x, addr, T_N8 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_VSIB, 0x92, 1); }
void vgatherdps(const Xmm& x, const Address& addr) { opGather2(x, addr, T_N4 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_VSIB, 0x92, 0); }
void vgatherpf0dpd(const Address& addr) { opGatherFetch(addr, zm1, T_N8 | T_66 | T_0F38 | T_EW1 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC6, Operand::YMM); }
void vgatherpf0dps(const Address& addr) { opGatherFetch(addr, zm1, T_N4 | T_66 | T_0F38 | T_EW0 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC6, Operand::ZMM); }
void vgatherpf0qpd(const Address& addr) { opGatherFetch(addr, zm1, T_N8 | T_66 | T_0F38 | T_EW1 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC7, Operand::ZMM); }
void vgatherpf0qps(const Address& addr) { opGatherFetch(addr, zm1, T_N4 | T_66 | T_0F38 | T_EW0 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC7, Operand::ZMM); }
void vgatherpf1dpd(const Address& addr) { opGatherFetch(addr, zm2, T_N8 | T_66 | T_0F38 | T_EW1 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC6, Operand::YMM); }
void vgatherpf1dps(const Address& addr) { opGatherFetch(addr, zm2, T_N4 | T_66 | T_0F38 | T_EW0 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC6, Operand::ZMM); }
void vgatherpf1qpd(const Address& addr) { opGatherFetch(addr, zm2, T_N8 | T_66 | T_0F38 | T_EW1 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC7, Operand::ZMM); }
void vgatherpf1qps(const Address& addr) { opGatherFetch(addr, zm2, T_N4 | T_66 | T_0F38 | T_EW0 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC7, Operand::ZMM); }
void vgatherqpd(const Xmm& x, const Address& addr) { opGather2(x, addr, T_N8 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_VSIB, 0x93, 0); }
void vgatherqps(const Xmm& x, const Address& addr) { opGather2(x, addr, T_N4 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_VSIB, 0x93, 2); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vgetexppd(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B64, 0x42); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vgetexpph(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B16, 0x42); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vgetexpps(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 | T_B32, 0x42); }
void vgetexpsd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_EW1 | T_SAE_X | T_MUST_EVEX, 0x43); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vgetexpsh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_SAE_X | T_MUST_EVEX, 0x43); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vgetexpss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_EW0 | T_SAE_X | T_MUST_EVEX, 0x43); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vgetmantpd(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(x, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B64, 0x26, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vgetmantph(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(x, op, T_0F3A | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B16, 0x26, imm); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vgetmantps(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(x, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B32, 0x26, imm); }
void vgetmantsd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F3A | T_EW1 | T_SAE_X | T_MUST_EVEX, 0x27, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vgetmantsh(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_0F3A | T_EW0 | T_SAE_X | T_MUST_EVEX, 0x27, imm); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vgetmantss(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F3A | T_EW0 | T_SAE_X | T_MUST_EVEX, 0x27, imm); }
void vinsertf32x4(const Ymm& r1, const Ymm& r2, const Operand& op, uint8_t imm) {if (!(r1.getKind() == r2.getKind() && op.is(Operand::MEM | Operand::XMM))) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r1, &r2, op, T_N16 | T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX, 0x18, imm); }
void vinsertf32x8(const Zmm& r1, const Zmm& r2, const Operand& op, uint8_t imm) {if (!op.is(Operand::MEM | Operand::YMM)) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r1, &r2, op, T_N32 | T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX, 0x1A, imm); }
void vinsertf64x2(const Ymm& r1, const Ymm& r2, const Operand& op, uint8_t imm) {if (!(r1.getKind() == r2.getKind() && op.is(Operand::MEM | Operand::XMM))) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r1, &r2, op, T_N16 | T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX, 0x18, imm); }
void vinsertf64x4(const Zmm& r1, const Zmm& r2, const Operand& op, uint8_t imm) {if (!op.is(Operand::MEM | Operand::YMM)) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r1, &r2, op, T_N32 | T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX, 0x1A, imm); }
void vinserti32x4(const Ymm& r1, const Ymm& r2, const Operand& op, uint8_t imm) {if (!(r1.getKind() == r2.getKind() && op.is(Operand::MEM | Operand::XMM))) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r1, &r2, op, T_N16 | T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX, 0x38, imm); }
void vinserti32x8(const Zmm& r1, const Zmm& r2, const Operand& op, uint8_t imm) {if (!op.is(Operand::MEM | Operand::YMM)) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r1, &r2, op, T_N32 | T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX, 0x3A, imm); }
void vinserti64x2(const Ymm& r1, const Ymm& r2, const Operand& op, uint8_t imm) {if (!(r1.getKind() == r2.getKind() && op.is(Operand::MEM | Operand::XMM))) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r1, &r2, op, T_N16 | T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX, 0x38, imm); }
void vinserti64x4(const Zmm& r1, const Zmm& r2, const Operand& op, uint8_t imm) {if (!op.is(Operand::MEM | Operand::YMM)) XBYAK_THROW(ERR_BAD_COMBINATION) opVex(r1, &r2, op, T_N32 | T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX, 0x3A, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vmaxph(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_MAP5 | T_EW0 | T_YMM | T_MUST_EVEX | T_ER_Z | T_B16, 0x5F); }
void vmaxsh(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_MAP5 | T_F3 | T_EW0 | T_MUST_EVEX | T_ER_X | T_N2, 0x5F); }
void vminph(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_MAP5 | T_EW0 | T_YMM | T_MUST_EVEX | T_ER_Z | T_B16, 0x5D); }
void vminsh(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_MAP5 | T_F3 | T_EW0 | T_MUST_EVEX | T_ER_X | T_N2, 0x5D); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vmovdqa32(const Address& addr, const Xmm& x) { opAVX_X_XM_IMM(x, addr, T_66 | T_0F | T_EW0 | T_YMM | T_ER_X | T_ER_Y | T_ER_Z | T_MUST_EVEX | T_M_K, 0x7F); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovdqa32(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F | T_EW0 | T_YMM | T_ER_X | T_ER_Y | T_ER_Z | T_MUST_EVEX, 0x6F); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vmovdqa64(const Address& addr, const Xmm& x) { opAVX_X_XM_IMM(x, addr, T_66 | T_0F | T_EW1 | T_YMM | T_ER_X | T_ER_Y | T_ER_Z | T_MUST_EVEX | T_M_K, 0x7F); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovdqa64(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F | T_EW1 | T_YMM | T_ER_X | T_ER_Y | T_ER_Z | T_MUST_EVEX, 0x6F); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vmovdqu16(const Address& addr, const Xmm& x) { opAVX_X_XM_IMM(x, addr, T_F2 | T_0F | T_EW1 | T_YMM | T_ER_X | T_ER_Y | T_ER_Z | T_MUST_EVEX | T_M_K, 0x7F); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovdqu16(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_F2 | T_0F | T_EW1 | T_YMM | T_ER_X | T_ER_Y | T_ER_Z | T_MUST_EVEX, 0x6F); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vmovdqu32(const Address& addr, const Xmm& x) { opAVX_X_XM_IMM(x, addr, T_F3 | T_0F | T_EW0 | T_YMM | T_ER_X | T_ER_Y | T_ER_Z | T_MUST_EVEX | T_M_K, 0x7F); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovdqu32(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_F3 | T_0F | T_EW0 | T_YMM | T_ER_X | T_ER_Y | T_ER_Z | T_MUST_EVEX, 0x6F); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vmovdqu64(const Address& addr, const Xmm& x) { opAVX_X_XM_IMM(x, addr, T_F3 | T_0F | T_EW1 | T_YMM | T_ER_X | T_ER_Y | T_ER_Z | T_MUST_EVEX | T_M_K, 0x7F); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovdqu64(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_F3 | T_0F | T_EW1 | T_YMM | T_ER_X | T_ER_Y | T_ER_Z | T_MUST_EVEX, 0x6F); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vmovdqu8(const Address& addr, const Xmm& x) { opAVX_X_XM_IMM(x, addr, T_F2 | T_0F | T_EW0 | T_YMM | T_ER_X | T_ER_Y | T_ER_Z | T_MUST_EVEX | T_M_K, 0x7F); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vmovdqu8(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_F2 | T_0F | T_EW0 | T_YMM | T_ER_X | T_ER_Y | T_ER_Z | T_MUST_EVEX, 0x6F); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vmovsh(const Address& addr, const Xmm& x) { opAVX_X_XM_IMM(x, addr, T_N2 | T_F3 | T_MAP5 | T_EW0 | T_MUST_EVEX | T_M_K, 0x11); }
void vmovsh(const Xmm& x, const Address& addr) { opAVX_X_X_XM(x, xm0, addr, T_N2 | T_F3 | T_MAP5 | T_EW0 | T_MUST_EVEX, 0x10); }
void vmovsh(const Xmm& x1, const Xmm& x2, const Xmm& x3) { opAVX_X_X_XM(x1, x2, x3, T_N2 | T_F3 | T_MAP5 | T_EW0 | T_MUST_EVEX, 0x10); }
void vmovw(const Address& addr, const Xmm& x) { opAVX_X_XM_IMM(x, addr, T_N2 | T_66 | T_MAP5 | T_MUST_EVEX, 0x7E); }
void vmovw(const Reg32e& r, const Xmm& x) { opAVX_X_X_XM(x, xm0, r, T_N2 | T_66 | T_MAP5 | T_MUST_EVEX, 0x7E); }
void vmovw(const Xmm& x, const Operand& op) { if (!op.isREG(32|64) && !op.isMEM()) XBYAK_THROW(ERR_BAD_COMBINATION) opAVX_X_X_XM(x, xm0, op, T_N2 | T_66 | T_MAP5 | T_MUST_EVEX, 0x6E); }
void vmulph(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_MAP5 | T_EW0 | T_YMM | T_MUST_EVEX | T_ER_Z | T_B16, 0x59); }
void vmulsh(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_MAP5 | T_F3 | T_EW0 | T_MUST_EVEX | T_ER_X | T_N2, 0x59); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vp2intersectd(const Opmask& k, const Xmm& x, const Operand& op) { if (k.getOpmaskIdx() != 0) XBYAK_THROW(ERR_OPMASK_IS_ALREADY_SET) opAVX_K_X_XM(k, x, op, T_F2 | T_0F38 | T_YMM | T_EVEX | T_EW0 | T_B32, 0x68); }
void vp2intersectq(const Opmask& k, const Xmm& x, const Operand& op) { if (k.getOpmaskIdx() != 0) XBYAK_THROW(ERR_OPMASK_IS_ALREADY_SET) opAVX_K_X_XM(k, x, op, T_F2 | T_0F38 | T_YMM | T_EVEX | T_EW1 | T_B64, 0x68); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vp4dpwssd(const Zmm& z1, const Zmm& z2, const Address& addr) { opAVX_X_X_XM(z1, z2, addr, T_0F38 | T_F2 | T_EW0 | T_YMM | T_MUST_EVEX | T_N16, 0x52); }
void vp4dpwssds(const Zmm& z1, const Zmm& z2, const Address& addr) { opAVX_X_X_XM(z1, z2, addr, T_0F38 | T_F2 | T_EW0 | T_YMM | T_MUST_EVEX | T_N16, 0x53); }
void vpabsq(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_MUST_EVEX | T_EW1 | T_B64 | T_YMM, 0x1F); }
void vpandd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0xDB); }
void vpandnd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0xDF); }
void vpandnq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0xDF); }
void vpandq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0xDB); }
void vpblendmb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x66); }
void vpblendmd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x64); }
void vpblendmq(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, 0x64); }
void vpblendmw(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, 0x66); }
void vpbroadcastb(const Xmm& x, const Reg8& r) { opVex(x, 0, r, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x7A); }
void vpbroadcastd(const Xmm& x, const Reg32& r) { opVex(x, 0, r, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x7C); }
void vpbroadcastmb2q(const Xmm& x, const Opmask& k) { opVex(x, 0, k, T_F3 | T_0F38 | T_YMM | T_MUST_EVEX | T_EW1, 0x2A); }
void vpbroadcastmw2d(const Xmm& x, const Opmask& k) { opVex(x, 0, k, T_F3 | T_0F38 | T_YMM | T_MUST_EVEX | T_EW0, 0x3A); }
void vpbroadcastw(const Xmm& x, const Reg16& r) { opVex(x, 0, r, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x7B); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpcmpb(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX, 0x3F, imm); }
void vpcmpd(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x1F, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpcmpeqb(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_66 | T_0F | T_YMM | T_MUST_EVEX, 0x74); }
void vpcmpeqd(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_66 | T_0F | T_YMM | T_MUST_EVEX | T_B32, 0x76); }
void vpcmpeqq(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x29); }
void vpcmpeqw(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_66 | T_0F | T_YMM | T_MUST_EVEX, 0x75); }
void vpcmpgtb(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_66 | T_0F | T_YMM | T_MUST_EVEX, 0x64); }
void vpcmpgtd(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_66 | T_0F | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x66); }
void vpcmpgtq(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x37); }
void vpcmpgtw(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_66 | T_0F | T_YMM | T_MUST_EVEX, 0x65); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpcmpq(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x1F, imm); }
void vpcmpub(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX, 0x3E, imm); }
void vpcmpud(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x1E, imm); }
void vpcmpuq(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x1E, imm); }
void vpcmpuw(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX, 0x3E, imm); }
void vpcmpw(const Opmask& k, const Xmm& x, const Operand& op, uint8_t imm) { opAVX_K_X_XM(k, x, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX, 0x3F, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpcompressd(const Operand& op, const Xmm& x) { opAVX_X_XM_IMM(x, op, T_N4 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x8B); }
void vpcompressq(const Operand& op, const Xmm& x) { opAVX_X_XM_IMM(x, op, T_N8 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX, 0x8B); }
void vpconflictd(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0xC4); }
void vpconflictq(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0xC4); }
void vpermb(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x8D); }
void vpermi2b(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x75); }
void vpermi2d(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x76); }
void vpermi2pd(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, 0x77); }
void vpermi2ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x77); }
void vpermi2q(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, 0x76); }
void vpermi2w(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, 0x75); }
void vpermt2b(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x7D); }
void vpermt2d(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x7E); }
void vpermt2pd(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, 0x7F); }
void vpermt2ps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x7F); }
void vpermt2q(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, 0x7E); }
void vpermt2w(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, 0x7D); }
void vpermw(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, 0x8D); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
void vpexpandb(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_N1 | T_66 | T_0F38 | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX, 0x62); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpexpandd(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_N4 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x89); }
void vpexpandq(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_N8 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX, 0x89); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
void vpexpandw(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_N2 | T_66 | T_0F38 | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX, 0x62); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vpgatherdd(const Xmm& x, const Address& addr) { opGather2(x, addr, T_N4 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_VSIB, 0x90, 0); }
void vpgatherdq(const Xmm& x, const Address& addr) { opGather2(x, addr, T_N8 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_VSIB, 0x90, 1); }
void vpgatherqd(const Xmm& x, const Address& addr) { opGather2(x, addr, T_N4 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_VSIB, 0x91, 2); }
void vpgatherqq(const Xmm& x, const Address& addr) { opGather2(x, addr, T_N8 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_VSIB, 0x91, 0); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vplzcntd(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x44); }
void vplzcntq(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x44); }
void vpmaxsq(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, 0x3D); }
void vpmaxuq(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, 0x3F); }
void vpminsq(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, 0x39); }
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); }
Squashed 'externals/xbyak/' changes from 88f2f771f..a1ac3750f a1ac3750f Merge branch 'dev' 5f4ba971f v6.68 cac2c175f update doc 1b08a8248 add test of prefetchiti bef70d9b1 add prefetchit{0,1} f66785876 add detection of prefetchiti 124617ac9 Merge branch 'dev' 1bce8be17 Merge commit 'fbb18f6' into dev bebfe64a3 [test] X32 is optional and remove CFLAGS d8c332cef Merge branch 'Tachi107-small-changes' into dev fbb18f69d ci: use containers 8a974696c test: simplify makefile 1efe9fe7c Merge branch 'dev' bb70083e6 v6.67 436e452d7 update doc 47ff6ef42 I get an error after all on GitHub action 445c0dcec add test of CMPccXADD c9347907d add CMPccXADD dc792cc56 add detection of cmpccxadd 62be84cc8 fix detection of boost b5ac7b0f0 Merge branch 'dev' 05dd400e0 recover a removed line 94eff6246 v6.66 ec3fadeba update doc 00bfaaa7f add test of rao-int 0f2f1aaa6 support rio-int 6c047f480 detect rao-int f07c5c255 fix args of EmptyAllocator::alloc in test 04d3eb5f5 stop if a test script causes an error 1c1d2366f disable boost sample cfb1127c3 tweak 5fcbeb7c4 [sample] change the way of detection of boost bafc1ee60 CXX uses g++ as default value d8cabc6cb remove warning of blace bc73a0816 remove -fno-operator-names option 6989aea94 use CXX instead of g++ a7c5a1bd7 use English 1bfbd8c4d Fix incorrect format strings 0ecef5c28 Merge branch 'dev' 7556c20ba update doc a15709271 v6.65 3b83aab3e add detect_x32 to TARGET 8c64bbbc3 use gcc instead of dpkg for portability 5e9a9b96f test_avx512.sh runs on x32 8ae01b0c2 disable some tests on x32 83b3da217 x32 does not check large disp 693ab8c9d sizeof(void*) = 4 on x32, so disable the test 348e3e548 Merge branch 'dev' 11b9c4dc0 v6.64 459636196 add T_M_K flag to vpmov* 1d3722928 add include path 5e27eddae move CrearError from LabelManager::reset() to CodeGenerator::reset() f8ea5c28d Merge branch 'dev' 20b2b1eae v6.63 3706869f8 desc. of setDefaultEncoding d6f2d7577 add test of setDefaultEncoding 3b0a19c41 vpmadd52{h,l}uq for avx-ifma 95752ebd7 add tAVX_IFMA cd36e31ea [sample] show AMX_FP16/AVX_VNNI_INT8/AVX_NE_CONVERT e5858af27 add setDefaultEncoding 2f7fb0220 modify gen.cpp for AVX-NE-CONVERT/AVX-VNNI-INT8/AMX-FP16 1c5cb7efa add AVX-NE-CONVERT instructions 564fe9acd add AVX-VNNI-INT8 instructions cd14d07b1 add AMX-FP16 instruction 7811f593c Merge pull request #161 from scribam/patch-1 2218f6c08 Update changelog.md 7bccdbbb4 Merge pull request #160 from herumi/dev 5fcf87596 compile nasm-2.15 instead of apt install e31961ea8 v6.62 ff4f9e65c update doc 9c8fb81db disable wrong detection of gcc -Warray-bounds 6b7519659 add serialize e16582696 Merge branch 'JonLiu1993-vcpkg-installition' into dev edbb410fb Add vcpkg installation instructions 055d31242 Merge branch 'dev' 21ab98441 Merge branch 'akodanka-enable_CIV' into dev ed4d598e3 Changes to compile xbyak project for openvino b652430c4 mingw uses __cpuidex 48457bfa0 Merge branch 'dev' 29cb524d1 v6.61.2 1a9a0b0e1 avoid including algorithm header in xbyak_util.h 6fadefd04 Merge branch 'dev' fc1c18a9d update doc f7cae7f11 v6.61.1 6f5ec5cf3 Merge branch 'doyaGu-master' into dev 4554d6bb9 Fix error related to XBYAK_NOEXCEPT git-subtree-dir: externals/xbyak git-subtree-split: a1ac3750f9a639b5a6c6d6c7da4259b8d6790989
2022-12-30 23:05:02 +00:00
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); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
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); }
Squashed 'externals/xbyak/' changes from 88f2f771f..a1ac3750f a1ac3750f Merge branch 'dev' 5f4ba971f v6.68 cac2c175f update doc 1b08a8248 add test of prefetchiti bef70d9b1 add prefetchit{0,1} f66785876 add detection of prefetchiti 124617ac9 Merge branch 'dev' 1bce8be17 Merge commit 'fbb18f6' into dev bebfe64a3 [test] X32 is optional and remove CFLAGS d8c332cef Merge branch 'Tachi107-small-changes' into dev fbb18f69d ci: use containers 8a974696c test: simplify makefile 1efe9fe7c Merge branch 'dev' bb70083e6 v6.67 436e452d7 update doc 47ff6ef42 I get an error after all on GitHub action 445c0dcec add test of CMPccXADD c9347907d add CMPccXADD dc792cc56 add detection of cmpccxadd 62be84cc8 fix detection of boost b5ac7b0f0 Merge branch 'dev' 05dd400e0 recover a removed line 94eff6246 v6.66 ec3fadeba update doc 00bfaaa7f add test of rao-int 0f2f1aaa6 support rio-int 6c047f480 detect rao-int f07c5c255 fix args of EmptyAllocator::alloc in test 04d3eb5f5 stop if a test script causes an error 1c1d2366f disable boost sample cfb1127c3 tweak 5fcbeb7c4 [sample] change the way of detection of boost bafc1ee60 CXX uses g++ as default value d8cabc6cb remove warning of blace bc73a0816 remove -fno-operator-names option 6989aea94 use CXX instead of g++ a7c5a1bd7 use English 1bfbd8c4d Fix incorrect format strings 0ecef5c28 Merge branch 'dev' 7556c20ba update doc a15709271 v6.65 3b83aab3e add detect_x32 to TARGET 8c64bbbc3 use gcc instead of dpkg for portability 5e9a9b96f test_avx512.sh runs on x32 8ae01b0c2 disable some tests on x32 83b3da217 x32 does not check large disp 693ab8c9d sizeof(void*) = 4 on x32, so disable the test 348e3e548 Merge branch 'dev' 11b9c4dc0 v6.64 459636196 add T_M_K flag to vpmov* 1d3722928 add include path 5e27eddae move CrearError from LabelManager::reset() to CodeGenerator::reset() f8ea5c28d Merge branch 'dev' 20b2b1eae v6.63 3706869f8 desc. of setDefaultEncoding d6f2d7577 add test of setDefaultEncoding 3b0a19c41 vpmadd52{h,l}uq for avx-ifma 95752ebd7 add tAVX_IFMA cd36e31ea [sample] show AMX_FP16/AVX_VNNI_INT8/AVX_NE_CONVERT e5858af27 add setDefaultEncoding 2f7fb0220 modify gen.cpp for AVX-NE-CONVERT/AVX-VNNI-INT8/AMX-FP16 1c5cb7efa add AVX-NE-CONVERT instructions 564fe9acd add AVX-VNNI-INT8 instructions cd14d07b1 add AMX-FP16 instruction 7811f593c Merge pull request #161 from scribam/patch-1 2218f6c08 Update changelog.md 7bccdbbb4 Merge pull request #160 from herumi/dev 5fcf87596 compile nasm-2.15 instead of apt install e31961ea8 v6.62 ff4f9e65c update doc 9c8fb81db disable wrong detection of gcc -Warray-bounds 6b7519659 add serialize e16582696 Merge branch 'JonLiu1993-vcpkg-installition' into dev edbb410fb Add vcpkg installation instructions 055d31242 Merge branch 'dev' 21ab98441 Merge branch 'akodanka-enable_CIV' into dev ed4d598e3 Changes to compile xbyak project for openvino b652430c4 mingw uses __cpuidex 48457bfa0 Merge branch 'dev' 29cb524d1 v6.61.2 1a9a0b0e1 avoid including algorithm header in xbyak_util.h 6fadefd04 Merge branch 'dev' fc1c18a9d update doc f7cae7f11 v6.61.1 6f5ec5cf3 Merge branch 'doyaGu-master' into dev 4554d6bb9 Fix error related to XBYAK_NOEXCEPT git-subtree-dir: externals/xbyak git-subtree-split: a1ac3750f9a639b5a6c6d6c7da4259b8d6790989
2022-12-30 23:05:02 +00:00
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); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpmovw2m(const Opmask& k, const Xmm& x) { opVex(k, 0, x, T_F3 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW1, 0x29); }
Squashed 'externals/xbyak/' changes from 88f2f771f..a1ac3750f a1ac3750f Merge branch 'dev' 5f4ba971f v6.68 cac2c175f update doc 1b08a8248 add test of prefetchiti bef70d9b1 add prefetchit{0,1} f66785876 add detection of prefetchiti 124617ac9 Merge branch 'dev' 1bce8be17 Merge commit 'fbb18f6' into dev bebfe64a3 [test] X32 is optional and remove CFLAGS d8c332cef Merge branch 'Tachi107-small-changes' into dev fbb18f69d ci: use containers 8a974696c test: simplify makefile 1efe9fe7c Merge branch 'dev' bb70083e6 v6.67 436e452d7 update doc 47ff6ef42 I get an error after all on GitHub action 445c0dcec add test of CMPccXADD c9347907d add CMPccXADD dc792cc56 add detection of cmpccxadd 62be84cc8 fix detection of boost b5ac7b0f0 Merge branch 'dev' 05dd400e0 recover a removed line 94eff6246 v6.66 ec3fadeba update doc 00bfaaa7f add test of rao-int 0f2f1aaa6 support rio-int 6c047f480 detect rao-int f07c5c255 fix args of EmptyAllocator::alloc in test 04d3eb5f5 stop if a test script causes an error 1c1d2366f disable boost sample cfb1127c3 tweak 5fcbeb7c4 [sample] change the way of detection of boost bafc1ee60 CXX uses g++ as default value d8cabc6cb remove warning of blace bc73a0816 remove -fno-operator-names option 6989aea94 use CXX instead of g++ a7c5a1bd7 use English 1bfbd8c4d Fix incorrect format strings 0ecef5c28 Merge branch 'dev' 7556c20ba update doc a15709271 v6.65 3b83aab3e add detect_x32 to TARGET 8c64bbbc3 use gcc instead of dpkg for portability 5e9a9b96f test_avx512.sh runs on x32 8ae01b0c2 disable some tests on x32 83b3da217 x32 does not check large disp 693ab8c9d sizeof(void*) = 4 on x32, so disable the test 348e3e548 Merge branch 'dev' 11b9c4dc0 v6.64 459636196 add T_M_K flag to vpmov* 1d3722928 add include path 5e27eddae move CrearError from LabelManager::reset() to CodeGenerator::reset() f8ea5c28d Merge branch 'dev' 20b2b1eae v6.63 3706869f8 desc. of setDefaultEncoding d6f2d7577 add test of setDefaultEncoding 3b0a19c41 vpmadd52{h,l}uq for avx-ifma 95752ebd7 add tAVX_IFMA cd36e31ea [sample] show AMX_FP16/AVX_VNNI_INT8/AVX_NE_CONVERT e5858af27 add setDefaultEncoding 2f7fb0220 modify gen.cpp for AVX-NE-CONVERT/AVX-VNNI-INT8/AMX-FP16 1c5cb7efa add AVX-NE-CONVERT instructions 564fe9acd add AVX-VNNI-INT8 instructions cd14d07b1 add AMX-FP16 instruction 7811f593c Merge pull request #161 from scribam/patch-1 2218f6c08 Update changelog.md 7bccdbbb4 Merge pull request #160 from herumi/dev 5fcf87596 compile nasm-2.15 instead of apt install e31961ea8 v6.62 ff4f9e65c update doc 9c8fb81db disable wrong detection of gcc -Warray-bounds 6b7519659 add serialize e16582696 Merge branch 'JonLiu1993-vcpkg-installition' into dev edbb410fb Add vcpkg installation instructions 055d31242 Merge branch 'dev' 21ab98441 Merge branch 'akodanka-enable_CIV' into dev ed4d598e3 Changes to compile xbyak project for openvino b652430c4 mingw uses __cpuidex 48457bfa0 Merge branch 'dev' 29cb524d1 v6.61.2 1a9a0b0e1 avoid including algorithm header in xbyak_util.h 6fadefd04 Merge branch 'dev' fc1c18a9d update doc f7cae7f11 v6.61.1 6f5ec5cf3 Merge branch 'doyaGu-master' into dev 4554d6bb9 Fix error related to XBYAK_NOEXCEPT git-subtree-dir: externals/xbyak git-subtree-split: a1ac3750f9a639b5a6c6d6c7da4259b8d6790989
2022-12-30 23:05:02 +00:00
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); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
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); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
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); }
void vpopcntd(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 | T_B32, 0x55); }
void vpopcntq(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B64, 0x55); }
void vpopcntw(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX, 0x54); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpord(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0xEB); }
void vporq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0xEB); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vprold(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 1), x, op, T_66 | T_0F | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x72, imm); }
void vprolq(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 1), x, op, T_66 | T_0F | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x72, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vprolvd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x15); }
void vprolvq(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, 0x15); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vprord(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 0), x, op, T_66 | T_0F | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x72, imm); }
void vprorq(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 0), x, op, T_66 | T_0F | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x72, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vprorvd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x14); }
void vprorvq(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, 0x14); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vpscatterdd(const Address& addr, const Xmm& x) { opGather2(x, addr, T_N4 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K | T_VSIB, 0xA0, 0); }
void vpscatterdq(const Address& addr, const Xmm& x) { opGather2(x, addr, T_N8 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_M_K | T_VSIB, 0xA0, 1); }
void vpscatterqd(const Address& addr, const Xmm& x) { opGather2(x, addr, T_N4 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K | T_VSIB, 0xA1, 2); }
void vpscatterqq(const Address& addr, const Xmm& x) { opGather2(x, addr, T_N8 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_M_K | T_VSIB, 0xA1, 0); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpshldd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B32, 0x71, imm); }
void vpshldq(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B64, 0x71, imm); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
void vpshldvd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B32, 0x71); }
void vpshldvq(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_SAE_Z | T_MUST_EVEX | T_B64, 0x71); }
void vpshldvw(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_SAE_Z | T_MUST_EVEX, 0x70); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpshldw(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX, 0x70, imm); }
void vpshrdd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B32, 0x73, imm); }
void vpshrdq(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B64, 0x73, imm); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
void vpshrdvd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B32, 0x73); }
void vpshrdvq(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_SAE_Z | T_MUST_EVEX | T_B64, 0x73); }
void vpshrdvw(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_SAE_Z | T_MUST_EVEX, 0x72); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpshrdw(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX, 0x72, imm); }
Squashed 'externals/xbyak/' changes from d512551e..2794cde7 2794cde7 add xword, yword, etc. in Xbyak::util fb9c04e4 fix document for vfpclassps a51be78b fix test dependency 04fdfb1e update version e6354f8b add vgf2p8mulb 09a12642 add gf2p8affineqb d171ba0e add gf2p8affineinvqb 457f4fd0 add vpshufbitqmb 5af0ba39 add vpexpand{b,w} e450f965 vpopcnt{d,q} supports ptr_b 48499eb1 add vpdpbusd(s), vpdpwssd(s) 9c745109 add vpdpbusd, vpdpbusds 0e1a11b4 add vpopcnt{b,w,d,q} 9acfc132 add vpshrd(v){w,d,q} ac8de850 add vpshld(v){w,d,q} f181c259 add vcompressb, vcompressw 5a402477 vpclmulqdq supports AVX-512 9e16b40b vaes* supports AVX-512 7fde08e0 add flags for intel's manual 319433-030.pdf c5da3778 add test of v4fmaddps, vp4dpwssd, etc. e4fc9d8a fix mpx encoding d0b2fb62 add bnd(0xf2) prefix for MPX f12b5678 use db for array cd74ab44 remove bat file git-subtree-dir: externals/xbyak git-subtree-split: 2794cde79eb71e86490061cac9622ad0067b8d15
2020-04-22 20:45:52 +01:00
void vpshufbitqmb(const Opmask& k, const Xmm& x, const Operand& op) { opVex(k, &x, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x8F); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsllvw(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, 0x12); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpsraq(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_X_XM(Xmm(x.getKind(), 4), x, op, T_66 | T_0F | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x72, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vpsraq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N16 | T_66 | T_0F | T_EW1 | T_YMM | T_MUST_EVEX, 0xE2); }
void vpsravq(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, 0x46); }
void vpsravw(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, 0x11); }
void vpsrlvw(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, 0x10); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vpternlogd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x25, imm); }
void vpternlogq(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x25, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vptestmb(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x26); }
void vptestmd(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x27); }
void vptestmq(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x27); }
void vptestmw(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX, 0x26); }
void vptestnmb(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x26); }
void vptestnmd(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_F3 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x27); }
void vptestnmq(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_F3 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x27); }
void vptestnmw(const Opmask& k, const Xmm& x, const Operand& op) { opAVX_K_X_XM(k, x, op, T_F3 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX, 0x26); }
void vpxord(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0xEF); }
void vpxorq(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0xEF); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vrangepd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B64, 0x50, imm); }
void vrangeps(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B32, 0x50, imm); }
void vrangesd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F3A | T_EW1 | T_SAE_X | T_MUST_EVEX, 0x51, imm); }
void vrangess(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F3A | T_EW0 | T_SAE_X | T_MUST_EVEX, 0x51, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vrcp14pd(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x4C); }
void vrcp14ps(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x4C); }
void vrcp14sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_EW1 | T_MUST_EVEX, 0x4D); }
void vrcp14ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_EW0 | T_MUST_EVEX, 0x4D); }
void vrcp28pd(const Zmm& z, const Operand& op) { opAVX_X_XM_IMM(z, op, T_66 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW1 | T_B64 | T_SAE_Z, 0xCA); }
void vrcp28ps(const Zmm& z, const Operand& op) { opAVX_X_XM_IMM(z, op, T_66 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_B32 | T_SAE_Z, 0xCA); }
void vrcp28sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_EW1 | T_SAE_X | T_MUST_EVEX, 0xCB); }
void vrcp28ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_EW0 | T_SAE_X | T_MUST_EVEX, 0xCB); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vrcpph(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_MUST_EVEX | T_B16, 0x4C); }
void vrcpsh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_MUST_EVEX, 0x4D); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vreducepd(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(x, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B64, 0x56, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vreduceph(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(x, op, T_0F3A | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B16, 0x56, imm); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vreduceps(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(x, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B32, 0x56, imm); }
void vreducesd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F3A | T_EW1 | T_SAE_X | T_MUST_EVEX, 0x57, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vreducesh(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_0F3A | T_EW0 | T_SAE_X | T_MUST_EVEX, 0x57, imm); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vreducess(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F3A | T_EW0 | T_SAE_X | T_MUST_EVEX, 0x57, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vrndscalepd(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(x, op, T_66 | T_0F3A | T_EW1 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B64, 0x09, imm); }
void vrndscaleph(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(x, op, T_0F3A | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B16, 0x08, imm); }
void vrndscaleps(const Xmm& x, const Operand& op, uint8_t imm) { opAVX_X_XM_IMM(x, op, T_66 | T_0F3A | T_EW0 | T_YMM | T_SAE_Z | T_MUST_EVEX | T_B32, 0x08, imm); }
void vrndscalesd(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F3A | T_EW1 | T_SAE_X | T_MUST_EVEX, 0x0B, imm); }
void vrndscalesh(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_0F3A | T_EW0 | T_SAE_X | T_MUST_EVEX, 0x0A, imm); }
void vrndscaless(const Xmm& x1, const Xmm& x2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F3A | T_EW0 | T_SAE_X | T_MUST_EVEX, 0x0A, imm); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vrsqrt14pd(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_B64, 0x4E); }
void vrsqrt14ps(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_B32, 0x4E); }
void vrsqrt14sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX, 0x4F); }
void vrsqrt14ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX, 0x4F); }
void vrsqrt28pd(const Zmm& z, const Operand& op) { opAVX_X_XM_IMM(z, op, T_66 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW1 | T_B64 | T_SAE_Z, 0xCC); }
void vrsqrt28ps(const Zmm& z, const Operand& op) { opAVX_X_XM_IMM(z, op, T_66 | T_0F38 | T_MUST_EVEX | T_YMM | T_EW0 | T_B32 | T_SAE_Z, 0xCC); }
void vrsqrt28sd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_EW1 | T_SAE_X | T_MUST_EVEX, 0xCD); }
void vrsqrt28ss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_EW0 | T_SAE_X | T_MUST_EVEX, 0xCD); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vrsqrtph(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_MUST_EVEX | T_B16, 0x4E); }
void vrsqrtsh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_MUST_EVEX, 0x4F); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vscalefpd(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_ER_Z | T_MUST_EVEX | T_B64, 0x2C); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vscalefph(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_MAP6 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0x2C); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vscalefps(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_66 | T_0F38 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B32, 0x2C); }
void vscalefsd(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N8 | T_66 | T_0F38 | T_EW1 | T_ER_X | T_MUST_EVEX, 0x2D); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vscalefsh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_66 | T_MAP6 | T_EW0 | T_ER_X | T_MUST_EVEX, 0x2D); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
void vscalefss(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N4 | T_66 | T_0F38 | T_EW0 | T_ER_X | T_MUST_EVEX, 0x2D); }
Squashed 'externals/xbyak/' changes from 2794cde7..671fc805 671fc805 update test/cybozu 8ca86231 remove mutable in Address 8b93498f add cmpsb/scasb/... 7eb62750 avoid core_sharing_data_cache = 0 for some cloud envrionment 85767e95 support mingw64 59573e6e add PROTECT_RE mode for protect() 71b75f65 fix push(qword[mem]) 811f4959 Merge branch 'rsdubtso-master' 8e3cb711 Account for potentially zero 0xb leaf when parsing cache/topology via cpuid a816249f update version fe083912 fix to avoid zero division for some virtual machine f0a8f7fa update version cac09b7a Merge pull request #62 from mgouicem/master 1f96b5e0 Fixes an error raised by clang < 3.9 c0f885ac Merge pull request #61 from mgouicem/master bfe2d201 Change default value for n_cores in setCacheHierarchy. fd587b55 change format and add getter for data_cache_size 80b3c7b9 remove macro 88189609 Merge branch 'mgouicem-master' e6b79723 Adding queries to get the cpu topology on Intel architectures. 221384f0 vmov* supports [mem]|k|z c04141ef define XBYAK_NO_OP_NAMES for test af7f05ee add const for Label git-subtree-dir: externals/xbyak git-subtree-split: 671fc805d09d075f48d4625f183ef2e1ef725106
2020-04-22 20:51:10 +01:00
void vscatterdpd(const Address& addr, const Xmm& x) { opGather2(x, addr, T_N8 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_M_K | T_VSIB, 0xA2, 1); }
void vscatterdps(const Address& addr, const Xmm& x) { opGather2(x, addr, T_N4 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K | T_VSIB, 0xA2, 0); }
void vscatterpf0dpd(const Address& addr) { opGatherFetch(addr, zm5, T_N8 | T_66 | T_0F38 | T_EW1 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC6, Operand::YMM); }
void vscatterpf0dps(const Address& addr) { opGatherFetch(addr, zm5, T_N4 | T_66 | T_0F38 | T_EW0 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC6, Operand::ZMM); }
void vscatterpf0qpd(const Address& addr) { opGatherFetch(addr, zm5, T_N8 | T_66 | T_0F38 | T_EW1 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC7, Operand::ZMM); }
void vscatterpf0qps(const Address& addr) { opGatherFetch(addr, zm5, T_N4 | T_66 | T_0F38 | T_EW0 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC7, Operand::ZMM); }
void vscatterpf1dpd(const Address& addr) { opGatherFetch(addr, zm6, T_N8 | T_66 | T_0F38 | T_EW1 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC6, Operand::YMM); }
void vscatterpf1dps(const Address& addr) { opGatherFetch(addr, zm6, T_N4 | T_66 | T_0F38 | T_EW0 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC6, Operand::ZMM); }
void vscatterpf1qpd(const Address& addr) { opGatherFetch(addr, zm6, T_N8 | T_66 | T_0F38 | T_EW1 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC7, Operand::ZMM); }
void vscatterpf1qps(const Address& addr) { opGatherFetch(addr, zm6, T_N4 | T_66 | T_0F38 | T_EW0 | T_MUST_EVEX | T_M_K | T_VSIB, 0xC7, Operand::ZMM); }
void vscatterqpd(const Address& addr, const Xmm& x) { opGather2(x, addr, T_N8 | T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX | T_M_K | T_VSIB, 0xA3, 0); }
void vscatterqps(const Address& addr, const Xmm& x) { opGather2(x, addr, T_N4 | T_66 | T_0F38 | T_EW0 | T_YMM | T_MUST_EVEX | T_M_K | T_VSIB, 0xA3, 2); }
Squashed 'externals/xbyak/' changes from 73ac5866..0140eeff 0140eeff Merge branch 'dev' 1efe14b2 change the original behavior of SetError 83c89c7a rename and fix indent 8be7ca93 Merge branch 'sbogusev-master' into dev 070b4c09 make l_err() inline with block scope static TLS l_error 9a4e6579 v5.97 d0ced1bc XBYAK_ONLY_CLASS_CPU is for only util::Cpu bb967ae7 replace uint32 with uint32_t etc. c306b8e5 update to v5.95 605e4224 use noexcept if C++11 or later 7a17c2c8 remove warning 5dfa4462 use constexpr if c++14 or later 18c9caaa Merge branch 'densamoilov-fix-mov-interface' into dev 3966ba9d fix mov interface be492be1 change the behavior of push((byte|word), imm) to cast imm to int8_t/int16_t d9696b54 Merge pull request #102 from igorsafo/master ea73267f Cpu: make getNumCores constant ff0b10e9 Merge pull request #101 from densamoilov/use-thread_local-when-supported 0c4eafc3 use thread_local for XBYAK_TLS when supported c1aea35e CodeGenerator::reset() calls ClearError() b4df97b1 Merge branch 'cursey-no-winsock2-header' 6a47bb0e v5.94 9a1749e6 define WIN32_LEAN_AND_MEAN for including winsock2.h after xbyak.h 42dddb74 Remove #include <winsock2.h> 615b85fa update doc 9cd796a9 rename XBYAK_NOEXCEPTION to XBYAK_NO_EXCEPTION 7cdf227f use static to avoid multiple instance 38a28dec test_nm.bat supports noexcept 0fdffc6b XBYAK_NOEXCEPTION for -fno-exceptions eda6e2a3 v5.92 5c26c8bb mov(rax, imm64) on 32-bit env with XBYAK64 6208e3ae throw exception if not supported amx sibmem 2 c6737d14 mov amx insts from avx512 34ea5c16 throw exception if not supported amx sibmem 6f93fe35 fix test of sizeof(Operand) 5b89c3b2 remove T_TMM 5ce32858 gen_amx.cpp is merged into gen_avx512.cpp fe4f965f remove my alias for tmm registers 92f904d8 bit_ contains 8192 98b51da9 extend mnemonics with Intel(R) AMX ISA 8d1b4c9e add generation of Intel(R) AMX ISA mnemonics 8ded45d1 add support of Intel(R) AMX ISA b23c4b02 v5.912 ffe32a60 Merge branch 'rsdubtso-master' e7b7fd2f use MAP_JIT on macOS regardless of Xcode version 82b70e66 v5.911 ; XBYAK_USE_MMAP_ALLOCATOR is defined 2f6d9e34 fix test for mac a7d10a1e add link to GitHub Sponsor 96076265 accept k0 mask register (it means no mask) 7e3167e4 kmov{b,w,d,q} throws for unsupported reg f487d7b7 Merge pull request #91 from marcelotrevisani/patch-1 dc9e6a79 Possibility to specify a different PREFIX 5fc69fc8 remove warning of test e69e0b42 fix typo of type of Zmi 34f797e8 perf does not recognize too short function name 6cc0f4df Consider max defined as a macro on Windows 5722393d fix for zeroed-out 0xb leaf 6a4459a8 Merge branch 'tyfkda-feature/fix-segfault-in-calc' 47922ed9 Fix segmentation fault in calc sample 8f696e93 add test_avx512 to bat 00114d79 add .travis.yml a29fa27b refactor test 508b543c fix error of vfpclasspd 0d54f1b1 fix for windows 4da8fd4e add setDefaultJmpNEAR da7f7317 revert to the behavior before v5.84 if -fno-operator-names is defined 7dac9f61 update to v5.85 fe639332 enable MAP_JIT only if mojave or later 4443d791 specify MAP_JIT mmap flag on macOS 20ee4c2d update doc ca0e8395 [changed] XBYAK_NO_OP_NAMES is defined f32836da remove exit(1) a1e9adf2 v5.82 08b8b1ba Support AMD Zen New Instructions. 2501ba9a remove *.user and *.vcproj 5c2ea988 Merge branch 'jrmwng-feature/upgrade-to-vs2017/jrmwng' 35847f7a Merge branch 'feature/upgrade-to-vs2017/jrmwng' of https://github.com/jrmwng/xbyak into jrmwng-feature/upgrade-to-vs2017/jrmwng ef267775 address "warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/SAFESEH' specification" 4a6c59bb address a conflict of sharing intermediate directory by different projects 9577cbf3 inherit "some output locations" from parent or project defaults 6c5f7186 upgrade projects from VS2018 to VS2017 4ca0434b v5.81 72b4e95d add lds/lss/les/lfs/lgs cc8f037c fix ; move ERR_INTERNAL to the end 9e9ec1c3 add repe, repne, repne, prez eea0edc3 add some fpu mnemonics 06235fa6 add loop/loope/loopne 7fc0c2bb add enter/leave 9fa2ef3c add in_, out_ df208648 add lods{b,w,d,q}, outs{b,w,d} 4672d2cb add int3, int_, into 431977cb add pushfq, popfq 81c4749f syscall, sysenter, sysexit, sysret 1f1b53c4 add clflushopt, fldenv, fnstw b765db33 Profiler uses append mode 44dc3546 add Profiler class 42949334 update version to v5.802 91cb919b Merge branch 'vpirogov-master' a6452f82 fixed avx512_bf16 detection f41da5aa tweak ; vcvtneps2bf16 calls opCvt2 b12460ba [sample] fix typo of quantize.cpp b22f5881 add set_opt.bat for test on Windows f402faad add vp2intersectd/vp2intersectq 4cfd5208 add avx512_bf16 4033564c fix vcmppd/vcmpps for ptr_b git-subtree-dir: externals/xbyak git-subtree-split: 0140eeff1fffcf5069dea3abb57095695320971c
2020-09-19 16:27:42 +01:00
void vshuff32x4(const Ymm& y1, const Ymm& y2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(y1, y2, op, T_66 | T_0F3A | T_YMM | T_MUST_EVEX | T_EW0 | T_B32, 0x23, imm); }
void vshuff64x2(const Ymm& y1, const Ymm& y2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(y1, y2, op, T_66 | T_0F3A | T_YMM | T_MUST_EVEX | T_EW1 | T_B64, 0x23, imm); }
void vshufi32x4(const Ymm& y1, const Ymm& y2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(y1, y2, op, T_66 | T_0F3A | T_YMM | T_MUST_EVEX | T_EW0 | T_B32, 0x43, imm); }
void vshufi64x2(const Ymm& y1, const Ymm& y2, const Operand& op, uint8_t imm) { opAVX_X_X_XM(y1, y2, op, T_66 | T_0F3A | T_YMM | T_MUST_EVEX | T_EW1 | T_B64, 0x43, imm); }
Squashed 'externals/xbyak/' changes from 590c10e37..9357732aa 9357732aa v6.06 b161a3eb7 update changelog e5453b064 Merge branch 'dev' 2a265d9d9 memfd: keep file descriptor open during allocation lifetime 31ff018ed replace unsigned int with uint32_t in xbyak_util.h e427b2231 move Type into Cpu 8cf41063b refactoring Cpu 66d62968d add -cpuid to test_util cf7cb744c add comments 028112949 add detection of clzero c88007b03 update doc 4cb2e77a9 v6.052 015c27cf6 Merge branch 'dev' d808f9ecb add test of Cpu::has b48a7bb1b add operator== to Type 1a90f456c v6.051 96cb1d660 update changelog 1f5a77f56 Merge branch 'dev' 87e14f02a add test with noexception 30144f809 fix error when XBYAK_NO_EXCEPTION is defined ec15751df Merge branch 'dev' 4831b3fb3 v6.05 9ddf251f3 remove debug code 905b31bab fix typo 72d1ac118 add movdir64b a6665996b add detection of movdir64b 92ddc6dfd remove Type::operator<<() because it's ambiguous 9cdd40f5e add detection of movdiri 379f8bf37 add movdiri 0ad6db138 fix cldemote test for 32-bit 84ab46bb3 add cldemote to Cpu a84ddc12d support cldemote 3a6cc626e add clwb 38c40c02e detection of clflushopt c061ac839 avoid unnecessary replacement in readme.md 095ebbff1 extend Cpu::Type to 128 bit 3ea8e45d3 Merge branch 'dev' f7bfc2634 v6.041 ab4e52d67 update doc 2bef54399 include intrin.h on mingw 621ed0775 Merge branch 'dev' 4f5893e1e memfd_create: disable for ANDROID_API < 30 684d69db0 Merge branch 'dev' ea88b6d85 disable warning on mingw 7e8923fa5 add colon 9914216bd Merge branch 'dev' b335602ff [skip ci] tweak doc 25fbb3996 tweak doc 122054dc4 Formatted Supported OS Section d4fa7e46b Reworked Authors Section 9e4b64a74 Adjusted Spacing ff9fdb45f Use Github Sponsor Widget 893b31d46 Added Quicklinks 2d2adf78b Added License Badge 335b5941b Formatted Header 70603addf Moved Install / Usage Into Dedicated Files 1659d28fe Moved History Into Dedicated File fb953284b Merge branch 'dev' dbd96b277 add news 898c354e6 v6.04 f8e2ad1e9 add waitpkg detection to Cpu a220fd69a add umwait 64ec053e6 add umonitor 764d54f6f add tpause c68646e8d remove warning of vc 4e8214ca2 Merge branch 'dev' 360f4b673 test badSSE only for 32-bit mode a9fddc454 fix typo of readme 250b5dc39 fix typo of readme c46e92756 Merge branch 'dev' 31b7cd350 v6.03 a2f9ed085 rename isValidXMm to isValidSSE 2bc8fcbd3 add tests to badSSE 6de93fb88 add baseSSE test e2eda384c update doc 71a7b1773 fix condition to throw error for SSE instructions when using XMM16-XMM31 615b665cc sample/memfd shows /proc/self/maps 2861517f2 add memfd sample 507b0285e apt update at first 452c07f77 typedef for no-MmapAllocator 8af6e2026 minimize diff 0af3b5d07 Allocator: take optional name parameter and use it with memfd a67e24505 Merge branch 'dev' b3892c15e Merge pull request #138 from Tachi107/make-cxx-cpp-ld-flags b4eddaced build(make): honour CXXFLAGS, CPPFLAGS and LDFLAGS fbe60e590 test generates only a.asm 0a4a7571d Merge branch 'patch-1' of https://github.com/Tachi107/xbyak into Tachi107-patch-1 55b9f131b build(make): fix clean target in test/Makefile 7aef3ff54 Merge branch 'dev' 82e0deb8a v6.02 4d9906a94 fix condition to throw error for invalid displacements c79311a51 fix test_util.cpp 9b2c175b3 Merge branch 'dev' 06d797e33 Allow parallel feature checks 2a85bba3f Merge branch 'dev' 8d5af80a6 v6.01 df39606fe update doc 08f11817c supprt retf 1abfc3465 support call(mem, T_FAR) fb158f901 support jmp(mem, T_FAR) 9be47ceb2 Merge branch 'lioncash-hlt' into dev 3162eb16f add test of hlt bb55725a9 xbyak: Add hlt opcode function 47cf85fdb fix conflict b29e471ea build(meson): fix CMake Config file include dir 2cc21925f delete the sentences translated into Japanese from COPYRIGHT dca3930de remove tabs cecd204a5 Merge pull request #129 from Tachi107/meson-cmake-config 345de8a54 build(meson): generate CMake package config files e831805cc revert change of the type of Pack::util::operator[] fc9d953e5 util::Pack has not pointer but instance of Reg64 0868c54a9 use _WIN32 instead of _MSC_VER for mingw64 894a1b14a use original uint8_t dfc079ca1 add cstrs of Xbyak::util::Pack for 11 or 12 args 06e8f531f Merge branch 'Tachi107-meson' into dev ef90b6bd2 Makefile updates the version of meson.build 0000938f5 build: add Meson support This allows Meson users to use xbyak more easly, while it also provides a pkg-config file generator, useful to Linux distributions 757e4063f v6.00 b3489f548 add vcvtusi2sh 16d9898ab add vcvtsi2sh cfc03cb8f unify T_66, T_F3, T_F2 flags 02fa7057d add vcvttph2qq 4e72a9dc4 add vcvttsh2usi 05d08e05b add vcvttsh2si 621e6548c add vcvtsh2usi 34abda5c5 extend vcvtps2ph facd622b2 add vcvtw2ph 270af1cb2 add vcvtuw2ph 17dc697cd add vcvttph2w 62f022aea vcvttph2uw 836346bfd add vcvtph2w 300edb37b add vcvtph2uw 737904b5a add vcvtuqq2ph 52b2ebf18 add vcvtqq2ph 6761f1e05 add vcvtpd2ph 03f95b7e3 add vcvtudq2ph 64430b2bb add vcvtps2phx 88e426aa4 add vcvtdq2ph bf28a94a5 add vcvttph2uqq 4c4e665d3 add vcvtph2uqq 2b0099b63 add vcvtph2qq 3ff69a474 add vcvtph2pd 678b52956 add vcvttph2udq 72a5717e4 add vcvttph2dq 524d52bf6 add vcvtph2udq d1cf4db97 add vcvtph2psx fccd2c49b add vcvtph2dq 6530f4099 refactor gen-cvt c51e16156 tweak ed9f6c72a add vcvtsh2si ac1407bd1 add vcvtss2sh 2958a19bb add vcvtsh2ss 10d683303 add vcvtsh2sd 3920c950c add vcvtsd2sh 9a1b73932 add vmovw d86e4882f vmovsh xmm, addr 61f85a204 add vmovsh 847166cef add tests of vgetmant{ph,sh} c11a21cf4 add vgetmantph 55ab361f0 add tests of vgetexp{ph,sh} 8653f4152 add vgetexpsh 697eeb627 add vgetexpph ee920a991 add vfpclasssh 7aed436e6 add vfpclassph 56dca14e4 v5.997 b3b1e4e3c fix vrndscale* to support {sae} 681077eb1 add vrndscalesh 2f14eae85 add vrndscaleph f96870a44 add vreducesh 043c94ae9 add vreduceph 37bf3bb49 add vscalefsh c16f91c59 add vscalefph 61ad45935 add vsqrtsh e00b508d5 add vsqrtph 52765d54f add vrsqrtsh 2ec6a7ab4 add vrsqrtph b1ff7891d add vrcpph, vrcpsh 4f543ca0e add vucomish 412b95f02 add vf{,c}mulcph 9fc53baed add vf{,c}maddcph 0098ce98d add vf{,n}m{sub,add}sh 849dffb10 add vfmadd, vfnmadd, vfnmsub for avx512-fp16 eadb93d0b add vfmsubaddcph 9b8802cba add vfmaddsub{132,213,231}ph ab9481b4c add vcomish cd036ea17 fix vcmpsh b494b321a vcmpsh supports ptr_b eb2d63c83 vcmpph uses T_B16 a480b3dd9 v5.996 16d18b1d4 fix v{add,sub,mul,...}{sd,ss} to support T_rd_sae etc. 66c6ca1ab support vaddsh with T_rd_sae 6333ec099 add m16bcst 84053c8e7 add vcmpph 5df23d263 move FP16 to AVX-512 d5c7336f8 fix disp scaling of v{add,sub,mul,div,max,min}sh ebf29542d add vaddsh test f29689d02 add test of vaddph 2c4b6ac16 add v{add,sub,mul,div,min,max}{ph,sh} edf3c2f6f remove mask for mmm 2c561aad6 add T_MAP5 and T_MAP6 acd360c38 add Cpu::tAVX512_FP16 1554f479c remove unused flags(tSSE4a and tSSE5) of Cpu c313a8758 update CMakeLists.txt version 740b39e24 Merge pull request #122 from abouvier/cmake 10a1e5759 v5.995 0a557a099 rename XBYAK_MEMFD_CREATE to XBYAK_USE_MEMFD 0ad9b9bbe Merge branch 'captain5050-master' into dev c7bb66383 fix for mac and enable it if XBYAK_MEMFD_CREATE is defined c1e6569f1 fix cmake config files ccd4130ec Name mmap pages on Linux with memfd_create 2fb843c32 v5.994 413a66b44 add alias of vcmpXX{ps,pd,ss,sd} for mask register 6f4f76890 v5.993 19043cb9a add test for gather/scatter b5acb1d1c gather test does not generate bad combination of regs 67ec1674a check restriction of gather/scatter regs ea9814f4f check bad reg combination of gather a34850b2d add endbr32 and endbr64 git-subtree-dir: externals/xbyak git-subtree-split: 9357732aa2aa3cf97809027596dfa5c61d1515b2
2022-06-05 13:31:49 +01:00
void vsqrtph(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_MAP5 | T_EW0 | T_YMM | T_ER_Z | T_MUST_EVEX | T_B16, 0x51); }
void vsqrtsh(const Xmm& x1, const Xmm& x2, const Operand& op) { opAVX_X_X_XM(x1, x2, op, T_N2 | T_F3 | T_MAP5 | T_EW0 | T_ER_X | T_MUST_EVEX, 0x51); }
void vsubph(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_MAP5 | T_EW0 | T_YMM | T_MUST_EVEX | T_ER_Z | T_B16, 0x5C); }
void vsubsh(const Xmm& xmm, const Operand& op1, const Operand& op2 = Operand()) { opAVX_X_X_XM(xmm, op1, op2, T_MAP5 | T_F3 | T_EW0 | T_MUST_EVEX | T_ER_X | T_N2, 0x5C); }
void vucomish(const Xmm& x, const Operand& op) { opAVX_X_XM_IMM(x, op, T_MAP5 | T_MUST_EVEX | T_EW0 | T_SAE_X | T_N2, 0x2E); }
Squashed 'externals/xbyak/' content from commit d512551e git-subtree-dir: externals/xbyak git-subtree-split: d512551e914737300ba35f3c049d1b40effbe76d
2020-04-22 20:25:57 +01:00
#ifdef XBYAK64
void kmovq(const Opmask& k, const Reg64& r) { opVex(k, 0, r, T_L0 | T_0F | T_F2 | T_W1, 0x92); }
void kmovq(const Reg64& r, const Opmask& k) { opVex(r, 0, k, T_L0 | T_0F | T_F2 | T_W1, 0x93); }
void vpbroadcastq(const Xmm& x, const Reg64& r) { opVex(x, 0, r, T_66 | T_0F38 | T_EW1 | T_YMM | T_MUST_EVEX, 0x7C); }
#endif
#endif
Reference in a new issue Copy permalink