[clang] Bulk port 64-bit x86 builtins to TableGen (PR #121043)
Chandler Carruth via cfe-commits
cfe-commits at lists.llvm.org
Tue Dec 24 01:15:07 PST 2024
https://github.com/chandlerc created https://github.com/llvm/llvm-project/pull/121043
This PR follows https://github.com/llvm/llvm-project/pull/120831 (the PR contains both, only review the last
commit here as the other commit will be reviewed on the other PR).
Similar to that PR, this does a very mechanical port of X86 builtins to
TableGen. There is a *lot* of improvement available here to use TableGen
more effectively and collapse repeated structures. But those can now be
follow-up PRs that restructure *within* the `.td` file.
The current structure produces a file that exactly matches the original
X-macros except for the differences outlined in https://github.com/llvm/llvm-project/pull/120831:
- Horizontal whitespace
- `long long` types now use `long long` outside of OpenCL, but switch to
`long` in OpenCL (if relevant at all).
Otherwise, only the order of builtins change, and no tests regress.
>From 9d031c423a782c43f3fd716bd5db215436e36b95 Mon Sep 17 00:00:00 2001
From: Chandler Carruth <chandlerc at gmail.com>
Date: Sat, 21 Dec 2024 07:00:28 +0000
Subject: [PATCH 1/2] Mechanically port bulk of x86 builtins to TableGen
The goal is to make incremental (if small) progress towards fully
TableGen'ed builtins, and to unblock #120534 by gaining access to more
powerful TableGen-based representations.
The bulk `.td` file addition was generated with the help of a very rough
Python script. That script made no attempt to be robust or reusable, it
specifically handled only the cases in the X86 `.def` file.
Four entries from the `.def` file were not handled automatically as they
used `BUILTIN` rather than `TARGET_BUILTIN`. These were ported by hand
to an empty-feature `TargetBuiltin` entry, which seems like a better
match.
For all the automatically ported entries, the results were compared by
sorting and diffing the `.def` file and the generated `.inc` file. The
only differences were:
- Different horizontal whitespace
- Additional entries that had already been ported to the `.td` file.
- Systematically using `Oi` instead of `LLi` for the type `long long
int`. The `.def` file uses a mixture of `Oi` and `LLi`. I chose the
shorter encoding.
This gives me high confidence in the correctness of the change.
---
clang/include/clang/Basic/BuiltinsBase.td | 11 +-
clang/include/clang/Basic/BuiltinsX86.def | 2225 -------
clang/include/clang/Basic/BuiltinsX86.td | 5387 +++++++++++++++++
clang/include/clang/Basic/TargetBuiltins.h | 2 -
clang/lib/Basic/Targets/X86.cpp | 8 -
clang/utils/TableGen/ClangBuiltinsEmitter.cpp | 24 +
6 files changed, 5419 insertions(+), 2238 deletions(-)
delete mode 100644 clang/include/clang/Basic/BuiltinsX86.def
diff --git a/clang/include/clang/Basic/BuiltinsBase.td b/clang/include/clang/Basic/BuiltinsBase.td
index cff182f3f282cb..afed3c815d3290 100644
--- a/clang/include/clang/Basic/BuiltinsBase.td
+++ b/clang/include/clang/Basic/BuiltinsBase.td
@@ -95,9 +95,6 @@ class CustomEntry {
}
class AtomicBuiltin : Builtin;
-class TargetBuiltin : Builtin {
- string Features = "";
-}
class LibBuiltin<string header, string languages = "ALL_LANGUAGES"> : Builtin {
string Header = header;
@@ -122,6 +119,14 @@ class OCL_DSELangBuiltin : LangBuiltin<"OCL_DSE">;
class OCL_GASLangBuiltin : LangBuiltin<"OCL_GAS">;
class OCLLangBuiltin : LangBuiltin<"ALL_OCL_LANGUAGES">;
+class TargetBuiltin : Builtin {
+ string Features = "";
+}
+class TargetLibBuiltin : TargetBuiltin {
+ string Header;
+ string Languages = "ALL_LANGUAGES";
+}
+
class Template<list<string> substitutions,
list<string> affixes,
bit as_prefix = 0> {
diff --git a/clang/include/clang/Basic/BuiltinsX86.def b/clang/include/clang/Basic/BuiltinsX86.def
deleted file mode 100644
index 352b3a9ec594a7..00000000000000
--- a/clang/include/clang/Basic/BuiltinsX86.def
+++ /dev/null
@@ -1,2225 +0,0 @@
-//===--- BuiltinsX86.def - X86 Builtin function database --------*- C++ -*-===//
-//
-// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
-// See https://llvm.org/LICENSE.txt for license information.
-// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
-//
-//===----------------------------------------------------------------------===//
-//
-// This file defines the X86-specific builtin function database. Users of
-// this file must define the BUILTIN macro to make use of this information.
-//
-//===----------------------------------------------------------------------===//
-
-// The format of this database matches clang/Basic/Builtins.def.
-
-// FIXME: Ideally we would be able to pull this information from what
-// LLVM already knows about X86 builtins. We need to match the LLVM
-// definition anyway, since code generation will lower to the
-// intrinsic if one exists.
-
-#if defined(BUILTIN) && !defined(TARGET_BUILTIN)
-# define TARGET_BUILTIN(ID, TYPE, ATTRS, FEATURE) BUILTIN(ID, TYPE, ATTRS)
-#endif
-
-#if defined(BUILTIN) && !defined(TARGET_HEADER_BUILTIN)
-# define TARGET_HEADER_BUILTIN(ID, TYPE, ATTRS, HEADER, LANG, FEATURE) BUILTIN(ID, TYPE, ATTRS)
-#endif
-
-// MMX
-//
-// All MMX instructions will be generated via builtins. Any MMX vector
-// types (<1 x i64>, <2 x i32>, etc.) that aren't used by these builtins will be
-// expanded by the back-end.
-// FIXME: _mm_prefetch must be a built-in because it takes a compile-time constant
-// argument and our prior approach of using a #define to the current built-in
-// doesn't work in the presence of re-declaration of _mm_prefetch for windows.
-TARGET_BUILTIN(_mm_prefetch, "vcC*i", "nc", "mmx")
-
-// SSE intrinsics.
-
-TARGET_BUILTIN(__builtin_ia32_ldmxcsr, "vUi", "n", "sse")
-TARGET_HEADER_BUILTIN(_mm_setcsr, "vUi", "nh",XMMINTRIN_H, ALL_LANGUAGES, "sse")
-TARGET_BUILTIN(__builtin_ia32_stmxcsr, "Ui", "n", "sse")
-TARGET_HEADER_BUILTIN(_mm_getcsr, "Ui", "nh", XMMINTRIN_H, ALL_LANGUAGES, "sse")
-TARGET_BUILTIN(__builtin_ia32_cvtss2si, "iV4f", "ncV:128:", "sse")
-TARGET_BUILTIN(__builtin_ia32_cvttss2si, "iV4f", "ncV:128:", "sse")
-TARGET_BUILTIN(__builtin_ia32_movmskps, "iV4f", "nV:128:", "sse")
-TARGET_BUILTIN(__builtin_ia32_sfence, "v", "n", "sse")
-TARGET_HEADER_BUILTIN(_mm_sfence, "v", "nh", XMMINTRIN_H, ALL_LANGUAGES, "sse")
-TARGET_BUILTIN(__builtin_ia32_rcpps, "V4fV4f", "ncV:128:", "sse")
-TARGET_BUILTIN(__builtin_ia32_rcpss, "V4fV4f", "ncV:128:", "sse")
-TARGET_BUILTIN(__builtin_ia32_rsqrtps, "V4fV4f", "ncV:128:", "sse")
-TARGET_BUILTIN(__builtin_ia32_rsqrtss, "V4fV4f", "ncV:128:", "sse")
-TARGET_BUILTIN(__builtin_ia32_sqrtps, "V4fV4f", "ncV:128:", "sse")
-TARGET_BUILTIN(__builtin_ia32_sqrtss, "V4fV4f", "ncV:128:", "sse")
-TARGET_BUILTIN(__builtin_ia32_shufps, "V4fV4fV4fIi", "ncV:128:", "sse")
-
-TARGET_BUILTIN(__builtin_ia32_maskmovdqu, "vV16cV16cc*", "nV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_movmskpd, "iV2d", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_pmovmskb128, "iV16c", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_movnti, "vi*i", "n", "sse2")
-TARGET_BUILTIN(__builtin_ia32_pshufd, "V4iV4iIi", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_pshuflw, "V8sV8sIi", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_pshufhw, "V8sV8sIi", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psadbw128, "V2OiV16cV16c", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_sqrtpd, "V2dV2d", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_sqrtsd, "V2dV2d", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_shufpd, "V2dV2dV2dIi", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2dq, "V2OiV2d", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2ps, "V4fV2d", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_cvttpd2dq, "V4iV2d", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_cvtsd2si, "iV2d", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_cvttsd2si, "iV2d", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_cvtsd2ss, "V4fV4fV2d", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_cvtps2dq, "V4iV4f", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_cvttps2dq, "V4iV4f", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_clflush, "vvC*", "n", "sse2")
-TARGET_HEADER_BUILTIN(_mm_clflush, "vvC*", "nh", EMMINTRIN_H, ALL_LANGUAGES, "sse2")
-TARGET_BUILTIN(__builtin_ia32_lfence, "v", "n", "sse2")
-TARGET_HEADER_BUILTIN(_mm_lfence, "v", "nh", EMMINTRIN_H, ALL_LANGUAGES, "sse2")
-TARGET_BUILTIN(__builtin_ia32_mfence, "v", "n", "sse2")
-TARGET_HEADER_BUILTIN(_mm_mfence, "v", "nh", EMMINTRIN_H, ALL_LANGUAGES, "sse2")
-TARGET_BUILTIN(__builtin_ia32_pause, "v", "n", "")
-TARGET_HEADER_BUILTIN(_mm_pause, "v", "nh", EMMINTRIN_H, ALL_LANGUAGES, "")
-TARGET_BUILTIN(__builtin_ia32_pmuludq128, "V2OiV4iV4i", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psraw128, "V8sV8sV8s", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psrad128, "V4iV4iV4i", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psrlw128, "V8sV8sV8s", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psrld128, "V4iV4iV4i", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psrlq128, "V2OiV2OiV2Oi", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psllw128, "V8sV8sV8s", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_pslld128, "V4iV4iV4i", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psllq128, "V2OiV2OiV2Oi", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psllwi128, "V8sV8si", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_pslldi128, "V4iV4ii", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psllqi128, "V2OiV2Oii", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psrlwi128, "V8sV8si", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psrldi128, "V4iV4ii", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psrlqi128, "V2OiV2Oii", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psrawi128, "V8sV8si", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psradi128, "V4iV4ii", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_pmaddwd128, "V4iV8sV8s", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_pslldqi128_byteshift, "V2OiV2OiIi", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_psrldqi128_byteshift, "V2OiV2OiIi", "ncV:128:", "sse2")
-
-TARGET_BUILTIN(__builtin_ia32_monitor, "vvC*UiUi", "n", "sse3")
-TARGET_BUILTIN(__builtin_ia32_mwait, "vUiUi", "n", "sse3")
-TARGET_BUILTIN(__builtin_ia32_lddqu, "V16ccC*", "nV:128:", "sse3")
-
-TARGET_BUILTIN(__builtin_ia32_palignr128, "V16cV16cV16cIi", "ncV:128:", "ssse3")
-
-TARGET_BUILTIN(__builtin_ia32_insertps128, "V4fV4fV4fIc", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_pblendvb128, "V16cV16cV16cV16c", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_pblendw128, "V8sV8sV8sIi", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_blendpd, "V2dV2dV2dIi", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_blendps, "V4fV4fV4fIi", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_blendvpd, "V2dV2dV2dV2d", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_blendvps, "V4fV4fV4fV4f", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_packusdw128, "V8sV4iV4i", "ncV:128:", "sse4.1")
-
-TARGET_BUILTIN(__builtin_ia32_pmuldq128, "V2OiV4iV4i", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_roundps, "V4fV4fIi", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_roundss, "V4fV4fV4fIi", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_roundsd, "V2dV2dV2dIi", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_roundpd, "V2dV2dIi", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_dpps, "V4fV4fV4fIc", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_dppd, "V2dV2dV2dIc", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_ptestz128, "iV2OiV2Oi", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_ptestc128, "iV2OiV2Oi", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_ptestnzc128, "iV2OiV2Oi", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_mpsadbw128, "V16cV16cV16cIc", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_phminposuw128, "V8sV8s", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_vec_ext_v16qi, "cV16cIi", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_vec_set_v16qi, "V16cV16ccIi", "ncV:128:", "sse4.1")
-TARGET_BUILTIN(__builtin_ia32_vec_set_v4si, "V4iV4iiIi", "ncV:128:", "sse4.1")
-
-// SSE 4.2
-TARGET_BUILTIN(__builtin_ia32_pcmpistrm128, "V16cV16cV16cIc", "ncV:128:", "sse4.2")
-TARGET_BUILTIN(__builtin_ia32_pcmpistri128, "iV16cV16cIc", "ncV:128:", "sse4.2")
-TARGET_BUILTIN(__builtin_ia32_pcmpestrm128, "V16cV16ciV16ciIc", "ncV:128:", "sse4.2")
-TARGET_BUILTIN(__builtin_ia32_pcmpestri128, "iV16ciV16ciIc","ncV:128:", "sse4.2")
-
-TARGET_BUILTIN(__builtin_ia32_pcmpistria128, "iV16cV16cIc","ncV:128:", "sse4.2")
-TARGET_BUILTIN(__builtin_ia32_pcmpistric128, "iV16cV16cIc","ncV:128:", "sse4.2")
-TARGET_BUILTIN(__builtin_ia32_pcmpistrio128, "iV16cV16cIc","ncV:128:", "sse4.2")
-TARGET_BUILTIN(__builtin_ia32_pcmpistris128, "iV16cV16cIc","ncV:128:", "sse4.2")
-TARGET_BUILTIN(__builtin_ia32_pcmpistriz128, "iV16cV16cIc","ncV:128:", "sse4.2")
-TARGET_BUILTIN(__builtin_ia32_pcmpestria128, "iV16ciV16ciIc","ncV:128:", "sse4.2")
-TARGET_BUILTIN(__builtin_ia32_pcmpestric128, "iV16ciV16ciIc","ncV:128:", "sse4.2")
-TARGET_BUILTIN(__builtin_ia32_pcmpestrio128, "iV16ciV16ciIc","ncV:128:", "sse4.2")
-TARGET_BUILTIN(__builtin_ia32_pcmpestris128, "iV16ciV16ciIc","ncV:128:", "sse4.2")
-TARGET_BUILTIN(__builtin_ia32_pcmpestriz128, "iV16ciV16ciIc","ncV:128:", "sse4.2")
-
-TARGET_BUILTIN(__builtin_ia32_crc32qi, "UiUiUc", "nc", "crc32")
-TARGET_BUILTIN(__builtin_ia32_crc32hi, "UiUiUs", "nc", "crc32")
-TARGET_BUILTIN(__builtin_ia32_crc32si, "UiUiUi", "nc", "crc32")
-
-// SSE4a
-TARGET_BUILTIN(__builtin_ia32_extrqi, "V2OiV2OiIcIc", "ncV:128:", "sse4a")
-TARGET_BUILTIN(__builtin_ia32_extrq, "V2OiV2OiV16c", "ncV:128:", "sse4a")
-TARGET_BUILTIN(__builtin_ia32_insertqi, "V2OiV2OiV2OiIcIc", "ncV:128:", "sse4a")
-TARGET_BUILTIN(__builtin_ia32_insertq, "V2OiV2OiV2Oi", "ncV:128:", "sse4a")
-TARGET_BUILTIN(__builtin_ia32_movntsd, "vd*V2d", "nV:128:", "sse4a")
-TARGET_BUILTIN(__builtin_ia32_movntss, "vf*V4f", "nV:128:", "sse4a")
-
-// AES
-TARGET_BUILTIN(__builtin_ia32_aesenc128, "V2OiV2OiV2Oi", "ncV:128:", "aes")
-TARGET_BUILTIN(__builtin_ia32_aesenclast128, "V2OiV2OiV2Oi", "ncV:128:", "aes")
-TARGET_BUILTIN(__builtin_ia32_aesdec128, "V2OiV2OiV2Oi", "ncV:128:", "aes")
-TARGET_BUILTIN(__builtin_ia32_aesdeclast128, "V2OiV2OiV2Oi", "ncV:128:", "aes")
-TARGET_BUILTIN(__builtin_ia32_aesimc128, "V2OiV2Oi", "ncV:128:", "aes")
-TARGET_BUILTIN(__builtin_ia32_aeskeygenassist128, "V2OiV2OiIc", "ncV:128:", "aes")
-
-// VAES
-TARGET_BUILTIN(__builtin_ia32_aesenc256, "V4OiV4OiV4Oi", "ncV:256:", "vaes")
-TARGET_BUILTIN(__builtin_ia32_aesenc512, "V8OiV8OiV8Oi", "ncV:512:", "avx512f,evex512,vaes")
-TARGET_BUILTIN(__builtin_ia32_aesenclast256, "V4OiV4OiV4Oi", "ncV:256:", "vaes")
-TARGET_BUILTIN(__builtin_ia32_aesenclast512, "V8OiV8OiV8Oi", "ncV:512:", "avx512f,evex512,vaes")
-TARGET_BUILTIN(__builtin_ia32_aesdec256, "V4OiV4OiV4Oi", "ncV:256:", "vaes")
-TARGET_BUILTIN(__builtin_ia32_aesdec512, "V8OiV8OiV8Oi", "ncV:512:", "avx512f,evex512,vaes")
-TARGET_BUILTIN(__builtin_ia32_aesdeclast256, "V4OiV4OiV4Oi", "ncV:256:", "vaes")
-TARGET_BUILTIN(__builtin_ia32_aesdeclast512, "V8OiV8OiV8Oi", "ncV:512:", "avx512f,evex512,vaes")
-
-// GFNI
-TARGET_BUILTIN(__builtin_ia32_vgf2p8affineinvqb_v16qi, "V16cV16cV16cIc", "ncV:128:", "gfni")
-TARGET_BUILTIN(__builtin_ia32_vgf2p8affineinvqb_v32qi, "V32cV32cV32cIc", "ncV:256:", "avx,gfni")
-TARGET_BUILTIN(__builtin_ia32_vgf2p8affineinvqb_v64qi, "V64cV64cV64cIc", "ncV:512:", "avx512f,evex512,gfni")
-TARGET_BUILTIN(__builtin_ia32_vgf2p8affineqb_v16qi, "V16cV16cV16cIc", "ncV:128:", "gfni")
-TARGET_BUILTIN(__builtin_ia32_vgf2p8affineqb_v32qi, "V32cV32cV32cIc", "ncV:256:", "avx,gfni")
-TARGET_BUILTIN(__builtin_ia32_vgf2p8affineqb_v64qi, "V64cV64cV64cIc", "ncV:512:", "avx512f,evex512,gfni")
-TARGET_BUILTIN(__builtin_ia32_vgf2p8mulb_v16qi, "V16cV16cV16c", "ncV:128:", "gfni")
-TARGET_BUILTIN(__builtin_ia32_vgf2p8mulb_v32qi, "V32cV32cV32c", "ncV:256:", "avx,gfni")
-TARGET_BUILTIN(__builtin_ia32_vgf2p8mulb_v64qi, "V64cV64cV64c", "ncV:512:", "avx512f,evex512,gfni")
-
-// CLMUL
-TARGET_BUILTIN(__builtin_ia32_pclmulqdq128, "V2OiV2OiV2OiIc", "ncV:128:", "pclmul")
-
-// VPCLMULQDQ
-TARGET_BUILTIN(__builtin_ia32_pclmulqdq256, "V4OiV4OiV4OiIc", "ncV:256:", "vpclmulqdq")
-TARGET_BUILTIN(__builtin_ia32_pclmulqdq512, "V8OiV8OiV8OiIc", "ncV:512:", "avx512f,evex512,vpclmulqdq")
-
-// AVX
-TARGET_BUILTIN(__builtin_ia32_vpermilvarpd, "V2dV2dV2Oi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vpermilvarps, "V4fV4fV4i", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vpermilvarpd256, "V4dV4dV4Oi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vpermilvarps256, "V8fV8fV8i", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_blendpd256, "V4dV4dV4dIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_blendps256, "V8fV8fV8fIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_blendvpd256, "V4dV4dV4dV4d", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_blendvps256, "V8fV8fV8fV8f", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_shufpd256, "V4dV4dV4dIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_shufps256, "V8fV8fV8fIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_dpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmppd256, "V4dV4dV4dIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vextractf128_pd256, "V2dV4dIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vextractf128_ps256, "V4fV8fIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vextractf128_si256, "V4iV8iIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2ps256, "V4fV4d", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cvtps2dq256, "V8iV8f", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cvttpd2dq256, "V4iV4d", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2dq256, "V4iV4d", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cvttps2dq256, "V8iV8f", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vperm2f128_pd256, "V4dV4dV4dIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vperm2f128_ps256, "V8fV8fV8fIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vperm2f128_si256, "V8iV8iV8iIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vpermilpd, "V2dV2dIi", "ncV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vpermilps, "V4fV4fIi", "ncV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vpermilpd256, "V4dV4dIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vpermilps256, "V8fV8fIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vinsertf128_pd256, "V4dV4dV2dIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vinsertf128_ps256, "V8fV8fV4fIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vinsertf128_si256, "V8iV8iV4iIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_sqrtpd256, "V4dV4d", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_sqrtps256, "V8fV8f", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_rsqrtps256, "V8fV8f", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_rcpps256, "V8fV8f", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_roundpd256, "V4dV4dIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_roundps256, "V8fV8fIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vtestzpd, "iV2dV2d", "ncV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vtestcpd, "iV2dV2d", "ncV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vtestnzcpd, "iV2dV2d", "ncV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vtestzps, "iV4fV4f", "ncV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vtestcps, "iV4fV4f", "ncV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vtestnzcps, "iV4fV4f", "ncV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vtestzpd256, "iV4dV4d", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vtestcpd256, "iV4dV4d", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vtestnzcpd256, "iV4dV4d", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vtestzps256, "iV8fV8f", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vtestcps256, "iV8fV8f", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vtestnzcps256, "iV8fV8f", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_ptestz256, "iV4OiV4Oi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_ptestc256, "iV4OiV4Oi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_ptestnzc256, "iV4OiV4Oi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_movmskpd256, "iV4d", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_movmskps256, "iV8f", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vzeroall, "v", "n", "avx")
-TARGET_BUILTIN(__builtin_ia32_vzeroupper, "v", "n", "avx")
-TARGET_BUILTIN(__builtin_ia32_lddqu256, "V32ccC*", "nV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_maskloadpd, "V2dV2dC*V2Oi", "nV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_maskloadps, "V4fV4fC*V4i", "nV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_maskloadpd256, "V4dV4dC*V4Oi", "nV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_maskloadps256, "V8fV8fC*V8i", "nV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_maskstorepd, "vV2d*V2OiV2d", "nV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_maskstoreps, "vV4f*V4iV4f", "nV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_maskstorepd256, "vV4d*V4OiV4d", "nV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_maskstoreps256, "vV8f*V8iV8f", "nV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vec_ext_v32qi, "cV32cIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vec_ext_v16hi, "sV16sIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vec_ext_v8si, "iV8iIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vec_set_v32qi, "V32cV32ccIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vec_set_v16hi, "V16sV16ssIi", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_vec_set_v8si, "V8iV8iiIi", "ncV:256:", "avx")
-
-// AVX2
-TARGET_BUILTIN(__builtin_ia32_mpsadbw256, "V32cV32cV32cIc", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_packsswb256, "V32cV16sV16s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_packssdw256, "V16sV8iV8i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_packuswb256, "V32cV16sV16s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_packusdw256, "V16sV8iV8i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_palignr256, "V32cV32cV32cIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pavgb256, "V32cV32cV32c", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pavgw256, "V16sV16sV16s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pblendvb256, "V32cV32cV32cV32c", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pblendw256, "V16sV16sV16sIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_phaddw256, "V16sV16sV16s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_phaddd256, "V8iV8iV8i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_phaddsw256, "V16sV16sV16s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_phsubw256, "V16sV16sV16s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_phsubd256, "V8iV8iV8i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_phsubsw256, "V16sV16sV16s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pmaddubsw256, "V16sV32cV32c", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pmaddwd256, "V8iV16sV16s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pmovmskb256, "iV32c", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pmuldq256, "V4OiV8iV8i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pmulhrsw256, "V16sV16sV16s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pmulhuw256, "V16sV16sV16s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pmulhw256, "V16sV16sV16s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pmuludq256, "V4OiV8iV8i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psadbw256, "V4OiV32cV32c", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pshufb256, "V32cV32cV32c", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pshufd256, "V8iV8iIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pshuflw256, "V16sV16sIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pshufhw256, "V16sV16sIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psignb256, "V32cV32cV32c", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psignw256, "V16sV16sV16s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psignd256, "V8iV8iV8i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psllwi256, "V16sV16si", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psllw256, "V16sV16sV8s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pslldqi256_byteshift, "V4OiV4OiIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pslldi256, "V8iV8ii", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pslld256, "V8iV8iV4i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psllqi256, "V4OiV4Oii", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psllq256, "V4OiV4OiV2Oi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrawi256, "V16sV16si", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psraw256, "V16sV16sV8s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psradi256, "V8iV8ii", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrad256, "V8iV8iV4i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrldqi256_byteshift, "V4OiV4OiIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrlwi256, "V16sV16si", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrlw256, "V16sV16sV8s", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrldi256, "V8iV8ii", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrld256, "V8iV8iV4i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrlqi256, "V4OiV4Oii", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrlq256, "V4OiV4OiV2Oi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pblendd128, "V4iV4iV4iIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_pblendd256, "V8iV8iV8iIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_permvarsi256, "V8iV8iV8i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_permdf256, "V4dV4dIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_permvarsf256, "V8fV8fV8i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_permti256, "V4OiV4OiV4OiIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_permdi256, "V4OiV4OiIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_extract128i256, "V2OiV4OiIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_insert128i256, "V4OiV4OiV2OiIi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_maskloadd256, "V8iV8iC*V8i", "nV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_maskloadq256, "V4OiV4OiC*V4Oi", "nV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_maskloadd, "V4iV4iC*V4i", "nV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_maskloadq, "V2OiV2OiC*V2Oi", "nV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_maskstored256, "vV8i*V8iV8i", "nV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_maskstoreq256, "vV4Oi*V4OiV4Oi", "nV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_maskstored, "vV4i*V4iV4i", "nV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_maskstoreq, "vV2Oi*V2OiV2Oi", "nV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psllv8si, "V8iV8iV8i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psllv4si, "V4iV4iV4i", "ncV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psllv4di, "V4OiV4OiV4Oi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psllv2di, "V2OiV2OiV2Oi", "ncV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrav8si, "V8iV8iV8i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrav4si, "V4iV4iV4i", "ncV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrlv8si, "V8iV8iV8i", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrlv4si, "V4iV4iV4i", "ncV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrlv4di, "V4OiV4OiV4Oi", "ncV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_psrlv2di, "V2OiV2OiV2Oi", "ncV:128:", "avx2")
-
-// GATHER
-TARGET_BUILTIN(__builtin_ia32_gatherd_pd, "V2dV2ddC*V4iV2dIc", "nV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherd_pd256, "V4dV4ddC*V4iV4dIc", "nV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherq_pd, "V2dV2ddC*V2OiV2dIc", "nV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherq_pd256, "V4dV4ddC*V4OiV4dIc", "nV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherd_ps, "V4fV4ffC*V4iV4fIc", "nV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherd_ps256, "V8fV8ffC*V8iV8fIc", "nV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherq_ps, "V4fV4ffC*V2OiV4fIc", "nV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherq_ps256, "V4fV4ffC*V4OiV4fIc", "nV:256:", "avx2")
-
-TARGET_BUILTIN(__builtin_ia32_gatherd_q, "V2OiV2OiOiC*V4iV2OiIc", "nV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherd_q256, "V4OiV4OiOiC*V4iV4OiIc", "nV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherq_q, "V2OiV2OiOiC*V2OiV2OiIc", "nV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherq_q256, "V4OiV4OiOiC*V4OiV4OiIc", "nV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherd_d, "V4iV4iiC*V4iV4iIc", "nV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherd_d256, "V8iV8iiC*V8iV8iIc", "nV:256:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherq_d, "V4iV4iiC*V2OiV4iIc", "nV:128:", "avx2")
-TARGET_BUILTIN(__builtin_ia32_gatherq_d256, "V4iV4iiC*V4OiV4iIc", "nV:256:", "avx2")
-
-// F16C
-TARGET_BUILTIN(__builtin_ia32_vcvtps2ph, "V8sV4fIi", "ncV:128:", "f16c")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2ph256, "V8sV8fIi", "ncV:256:", "f16c")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2ps, "V4fV8s", "ncV:128:", "f16c")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2ps256, "V8fV8s", "ncV:256:", "f16c")
-
-// RDRAND
-TARGET_BUILTIN(__builtin_ia32_rdrand16_step, "UiUs*", "n", "rdrnd")
-TARGET_BUILTIN(__builtin_ia32_rdrand32_step, "UiUi*", "n", "rdrnd")
-
-// FXSR
-TARGET_BUILTIN(__builtin_ia32_fxrstor, "vv*", "n", "fxsr")
-TARGET_BUILTIN(__builtin_ia32_fxsave, "vv*", "n", "fxsr")
-
-// XSAVE
-TARGET_BUILTIN(__builtin_ia32_xsave, "vv*UOi", "n", "xsave")
-TARGET_BUILTIN(__builtin_ia32_xrstor, "vv*UOi", "n", "xsave")
-TARGET_BUILTIN(__builtin_ia32_xgetbv, "UOiUi", "n", "xsave")
-TARGET_HEADER_BUILTIN(_xgetbv, "UWiUi", "nh", IMMINTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_BUILTIN(__builtin_ia32_xsetbv, "vUiUOi", "n", "xsave")
-TARGET_HEADER_BUILTIN(_xsetbv, "vUiUWi", "nh", IMMINTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_BUILTIN(__builtin_ia32_xsaveopt, "vv*UOi", "n", "xsaveopt")
-TARGET_BUILTIN(__builtin_ia32_xrstors, "vv*UOi", "n", "xsaves")
-TARGET_BUILTIN(__builtin_ia32_xsavec, "vv*UOi", "n", "xsavec")
-TARGET_BUILTIN(__builtin_ia32_xsaves, "vv*UOi", "n", "xsaves")
-
-// SHSTK
-TARGET_BUILTIN(__builtin_ia32_incsspd, "vUi", "n", "shstk")
-TARGET_BUILTIN(__builtin_ia32_rdsspd, "UiUi", "n", "shstk")
-TARGET_BUILTIN(__builtin_ia32_saveprevssp, "v", "n", "shstk")
-TARGET_BUILTIN(__builtin_ia32_rstorssp, "vv*", "n", "shstk")
-TARGET_BUILTIN(__builtin_ia32_wrssd, "vUiv*", "n", "shstk")
-TARGET_BUILTIN(__builtin_ia32_wrussd, "vUiv*", "n", "shstk")
-TARGET_BUILTIN(__builtin_ia32_setssbsy, "v", "n", "shstk")
-TARGET_BUILTIN(__builtin_ia32_clrssbsy, "vv*", "n", "shstk")
-
-//CLFLUSHOPT
-TARGET_BUILTIN(__builtin_ia32_clflushopt, "vvC*", "n", "clflushopt")
-
-//CLWB
-TARGET_BUILTIN(__builtin_ia32_clwb, "vvC*", "n", "clwb")
-
-//WB[NO]INVD
-TARGET_BUILTIN(__builtin_ia32_wbinvd, "v", "n", "")
-TARGET_BUILTIN(__builtin_ia32_wbnoinvd, "v", "n", "wbnoinvd")
-
-// ADX
-TARGET_BUILTIN(__builtin_ia32_addcarryx_u32, "UcUcUiUiUi*", "nE", "")
-TARGET_BUILTIN(__builtin_ia32_subborrow_u32, "UcUcUiUiUi*", "nE", "")
-
-// RDSEED
-TARGET_BUILTIN(__builtin_ia32_rdseed16_step, "UiUs*", "n", "rdseed")
-TARGET_BUILTIN(__builtin_ia32_rdseed32_step, "UiUi*", "n", "rdseed")
-
-// LZCNT
-TARGET_BUILTIN(__builtin_ia32_lzcnt_u16, "UsUs", "ncE", "lzcnt")
-TARGET_BUILTIN(__builtin_ia32_lzcnt_u32, "UiUi", "ncE", "lzcnt")
-
-// BMI
-TARGET_BUILTIN(__builtin_ia32_bextr_u32, "UiUiUi", "ncE", "bmi")
-TARGET_BUILTIN(__builtin_ia32_tzcnt_u16, "UsUs", "ncE", "")
-TARGET_BUILTIN(__builtin_ia32_tzcnt_u32, "UiUi", "ncE", "")
-
-// BMI2
-TARGET_BUILTIN(__builtin_ia32_bzhi_si, "UiUiUi", "ncE", "bmi2")
-TARGET_BUILTIN(__builtin_ia32_pdep_si, "UiUiUi", "ncE", "bmi2")
-TARGET_BUILTIN(__builtin_ia32_pext_si, "UiUiUi", "ncE", "bmi2")
-
-// TBM
-TARGET_BUILTIN(__builtin_ia32_bextri_u32, "UiUiIUi", "ncE", "tbm")
-
-// LWP
-TARGET_BUILTIN(__builtin_ia32_llwpcb, "vv*", "n", "lwp")
-TARGET_BUILTIN(__builtin_ia32_slwpcb, "v*", "n", "lwp")
-TARGET_BUILTIN(__builtin_ia32_lwpins32, "UcUiUiIUi", "n", "lwp")
-TARGET_BUILTIN(__builtin_ia32_lwpval32, "vUiUiIUi", "n", "lwp")
-
-// SHA
-TARGET_BUILTIN(__builtin_ia32_sha1rnds4, "V4iV4iV4iIc", "ncV:128:", "sha")
-TARGET_BUILTIN(__builtin_ia32_sha1nexte, "V4iV4iV4i", "ncV:128:", "sha")
-TARGET_BUILTIN(__builtin_ia32_sha1msg1, "V4iV4iV4i", "ncV:128:", "sha")
-TARGET_BUILTIN(__builtin_ia32_sha1msg2, "V4iV4iV4i", "ncV:128:", "sha")
-TARGET_BUILTIN(__builtin_ia32_sha256rnds2, "V4iV4iV4iV4i", "ncV:128:", "sha")
-TARGET_BUILTIN(__builtin_ia32_sha256msg1, "V4iV4iV4i", "ncV:128:", "sha")
-TARGET_BUILTIN(__builtin_ia32_sha256msg2, "V4iV4iV4i", "ncV:128:", "sha")
-
-// FMA
-TARGET_BUILTIN(__builtin_ia32_vfmaddps, "V4fV4fV4fV4f", "ncV:128:", "fma|fma4")
-TARGET_BUILTIN(__builtin_ia32_vfmaddpd, "V2dV2dV2dV2d", "ncV:128:", "fma|fma4")
-TARGET_BUILTIN(__builtin_ia32_vfmaddss3, "V4fV4fV4fV4f", "ncV:128:", "fma")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsd3, "V2dV2dV2dV2d", "ncV:128:", "fma")
-TARGET_BUILTIN(__builtin_ia32_vfmaddss, "V4fV4fV4fV4f", "ncV:128:", "fma4")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsd, "V2dV2dV2dV2d", "ncV:128:", "fma4")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubps, "V4fV4fV4fV4f", "ncV:128:", "fma|fma4")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubpd, "V2dV2dV2dV2d", "ncV:128:", "fma|fma4")
-TARGET_BUILTIN(__builtin_ia32_vfmaddps256, "V8fV8fV8fV8f", "ncV:256:", "fma|fma4")
-TARGET_BUILTIN(__builtin_ia32_vfmaddpd256, "V4dV4dV4dV4d", "ncV:256:", "fma|fma4")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubps256, "V8fV8fV8fV8f", "ncV:256:", "fma|fma4")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubpd256, "V4dV4dV4dV4d", "ncV:256:", "fma|fma4")
-
-TARGET_BUILTIN(__builtin_ia32_vfmaddpd512_mask, "V8dV8dV8dV8dUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddpd512_maskz, "V8dV8dV8dV8dUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddpd512_mask3, "V8dV8dV8dV8dUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmsubpd512_mask3, "V8dV8dV8dV8dUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddps512_mask, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddps512_maskz, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddps512_mask3, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmsubps512_mask3, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubpd512_mask, "V8dV8dV8dV8dUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubpd512_maskz, "V8dV8dV8dV8dUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubpd512_mask3, "V8dV8dV8dV8dUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmsubaddpd512_mask3, "V8dV8dV8dV8dUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubps512_mask, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubps512_maskz, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubps512_mask3, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmsubaddps512_mask3, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512f,evex512")
-
-// XOP
-TARGET_BUILTIN(__builtin_ia32_vpmacssww, "V8sV8sV8sV8s", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpmacsww, "V8sV8sV8sV8s", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpmacsswd, "V4iV8sV8sV4i", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpmacswd, "V4iV8sV8sV4i", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpmacssdd, "V4iV4iV4iV4i", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpmacsdd, "V4iV4iV4iV4i", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpmacssdql, "V2OiV4iV4iV2Oi", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpmacsdql, "V2OiV4iV4iV2Oi", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpmacssdqh, "V2OiV4iV4iV2Oi", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpmacsdqh, "V2OiV4iV4iV2Oi", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpmadcsswd, "V4iV8sV8sV4i", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpmadcswd, "V4iV8sV8sV4i", "ncV:128:", "xop")
-
-TARGET_BUILTIN(__builtin_ia32_vphaddbw, "V8sV16c", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphaddbd, "V4iV16c", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphaddbq, "V2OiV16c", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphaddwd, "V4iV8s", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphaddwq, "V2OiV8s", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphadddq, "V2OiV4i", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphaddubw, "V8sV16c", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphaddubd, "V4iV16c", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphaddubq, "V2OiV16c", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphadduwd, "V4iV8s", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphadduwq, "V2OiV8s", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphaddudq, "V2OiV4i", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphsubbw, "V8sV16c", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphsubwd, "V4iV8s", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vphsubdq, "V2OiV4i", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpperm, "V16cV16cV16cV16c", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vprotb, "V16cV16cV16c", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vprotw, "V8sV8sV8s", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vprotd, "V4iV4iV4i", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vprotq, "V2OiV2OiV2Oi", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vprotbi, "V16cV16cIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vprotwi, "V8sV8sIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vprotdi, "V4iV4iIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vprotqi, "V2OiV2OiIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpshlb, "V16cV16cV16c", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpshlw, "V8sV8sV8s", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpshld, "V4iV4iV4i", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpshlq, "V2OiV2OiV2Oi", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpshab, "V16cV16cV16c", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpshaw, "V8sV8sV8s", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpshad, "V4iV4iV4i", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpshaq, "V2OiV2OiV2Oi", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpcomub, "V16cV16cV16cIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpcomuw, "V8sV8sV8sIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpcomud, "V4iV4iV4iIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpcomuq, "V2OiV2OiV2OiIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpcomb, "V16cV16cV16cIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpcomw, "V8sV8sV8sIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpcomd, "V4iV4iV4iIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpcomq, "V2OiV2OiV2OiIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpermil2pd, "V2dV2dV2dV2OiIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpermil2pd256, "V4dV4dV4dV4OiIc", "ncV:256:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpermil2ps, "V4fV4fV4fV4iIc", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vpermil2ps256, "V8fV8fV8fV8iIc", "ncV:256:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vfrczss, "V4fV4f", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vfrczsd, "V2dV2d", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vfrczps, "V4fV4f", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vfrczpd, "V2dV2d", "ncV:128:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vfrczps256, "V8fV8f", "ncV:256:", "xop")
-TARGET_BUILTIN(__builtin_ia32_vfrczpd256, "V4dV4d", "ncV:256:", "xop")
-
-TARGET_BUILTIN(__builtin_ia32_xbegin, "i", "n", "rtm")
-TARGET_BUILTIN(__builtin_ia32_xend, "v", "n", "rtm")
-TARGET_BUILTIN(__builtin_ia32_xabort, "vIc", "n", "rtm")
-TARGET_BUILTIN(__builtin_ia32_xtest, "i", "n", "rtm")
-
-BUILTIN(__builtin_ia32_rdpmc, "UOii", "")
-BUILTIN(__builtin_ia32_rdtsc, "UOi", "")
-BUILTIN(__rdtsc, "UOi", "")
-BUILTIN(__builtin_ia32_rdtscp, "UOiUi*", "")
-
-TARGET_BUILTIN(__builtin_ia32_rdpid, "Ui", "n", "rdpid")
-TARGET_BUILTIN(__builtin_ia32_rdpru, "ULLii", "n", "rdpru")
-
-// PKU
-TARGET_BUILTIN(__builtin_ia32_rdpkru, "Ui", "n", "pku")
-TARGET_BUILTIN(__builtin_ia32_wrpkru, "vUi", "n", "pku")
-
-// AVX-512
-TARGET_BUILTIN(__builtin_ia32_sqrtpd512, "V8dV8dIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_sqrtps512, "V16fV16fIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_rsqrt14sd_mask, "V2dV2dV2dV2dUc", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_rsqrt14ss_mask, "V4fV4fV4fV4fUc", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_rsqrt14pd512_mask, "V8dV8dV8dUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_rsqrt14ps512_mask, "V16fV16fV16fUs", "ncV:512:", "avx512f,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_rcp14sd_mask, "V2dV2dV2dV2dUc", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_rcp14ss_mask, "V4fV4fV4fV4fUc", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_rcp14pd512_mask, "V8dV8dV8dUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_rcp14ps512_mask, "V16fV16fV16fUs", "ncV:512:", "avx512f,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_cvttps2dq512_mask, "V16iV16fV16iUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvttps2udq512_mask, "V16iV16fV16iUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvttpd2dq512_mask, "V8iV8dV8iUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvttpd2udq512_mask, "V8iV8dV8iUcIi", "ncV:512:", "avx512f,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_cmpps512_mask, "UsV16fV16fIiUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cmpps256_mask, "UcV8fV8fIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cmpps128_mask, "UcV4fV4fIiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cmppd512_mask, "UcV8dV8dIiUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cmppd256_mask, "UcV4dV4dIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cmppd128_mask, "UcV2dV2dIiUc", "ncV:128:", "avx512vl")
-
-TARGET_BUILTIN(__builtin_ia32_rndscaleps_mask, "V16fV16fIiV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_rndscalepd_mask, "V8dV8dIiV8dUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtps2dq512_mask, "V16iV16fV16iUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2dq512_mask, "V8iV8dV8iUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtps2udq512_mask, "V16iV16fV16iUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2udq512_mask, "V8iV8dV8iUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_minps512, "V16fV16fV16fIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_minpd512, "V8dV8dV8dIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_maxps512, "V16fV16fV16fIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_maxpd512, "V8dV8dV8dIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtdq2ps512_mask, "V16fV16iV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtudq2ps512_mask, "V16fV16iV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2ps512_mask, "V8fV8dV8fUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2ph512_mask, "V16sV16fIiV16sUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2ps512_mask, "V16fV16sV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmuldq512, "V8OiV16iV16i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmuludq512, "V8OiV16iV16i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_loaddqusi512_mask, "V16iiC*V16iUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_loaddqudi512_mask, "V8OiOiC*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_loadups512_mask, "V16ffC*V16fUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_loadaps512_mask, "V16fV16fC*V16fUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_loadupd512_mask, "V8ddC*V8dUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_loadapd512_mask, "V8dV8dC*V8dUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_storedqudi512_mask, "vOi*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_storedqusi512_mask, "vi*V16iUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_storeupd512_mask, "vd*V8dUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_storeapd512_mask, "vV8d*V8dUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_storeups512_mask, "vf*V16fUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_storeaps512_mask, "vV16f*V16fUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_alignq512, "V8OiV8OiV8OiIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_alignd512, "V16iV16iV16iIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_alignd128, "V4iV4iV4iIi", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_alignd256, "V8iV8iV8iIi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_alignq128, "V2OiV2OiV2OiIi", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_alignq256, "V4OiV4OiV4OiIi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_extractf64x4_mask, "V4dV8dIiV4dUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_extractf32x4_mask, "V4fV16fIiV4fUc", "ncV:512:", "avx512f,evex512")
-
-// AVX-VNNI and AVX512-VNNI
-TARGET_BUILTIN(__builtin_ia32_vpdpbusd128, "V4iV4iV4iV4i", "ncV:128:", "avx512vl,avx512vnni|avxvnni")
-TARGET_BUILTIN(__builtin_ia32_vpdpbusd256, "V8iV8iV8iV8i", "ncV:256:", "avx512vl,avx512vnni|avxvnni")
-TARGET_BUILTIN(__builtin_ia32_vpdpbusd512, "V16iV16iV16iV16i", "ncV:512:", "avx512vnni,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpdpbusds128, "V4iV4iV4iV4i", "ncV:128:", "avx512vl,avx512vnni|avxvnni")
-TARGET_BUILTIN(__builtin_ia32_vpdpbusds256, "V8iV8iV8iV8i", "ncV:256:", "avx512vl,avx512vnni|avxvnni")
-TARGET_BUILTIN(__builtin_ia32_vpdpbusds512, "V16iV16iV16iV16i", "ncV:512:", "avx512vnni,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpdpwssd128, "V4iV4iV4iV4i", "ncV:128:", "avx512vl,avx512vnni|avxvnni")
-TARGET_BUILTIN(__builtin_ia32_vpdpwssd256, "V8iV8iV8iV8i", "ncV:256:", "avx512vl,avx512vnni|avxvnni")
-TARGET_BUILTIN(__builtin_ia32_vpdpwssd512, "V16iV16iV16iV16i", "ncV:512:", "avx512vnni,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpdpwssds128, "V4iV4iV4iV4i", "ncV:128:", "avx512vl,avx512vnni|avxvnni")
-TARGET_BUILTIN(__builtin_ia32_vpdpwssds256, "V8iV8iV8iV8i", "ncV:256:", "avx512vl,avx512vnni|avxvnni")
-TARGET_BUILTIN(__builtin_ia32_vpdpwssds512, "V16iV16iV16iV16i", "ncV:512:", "avx512vnni,evex512")
-
-// AVX-VNNI-INT8
-TARGET_BUILTIN(__builtin_ia32_vpdpbssd128, "V4iV4iV4iV4i", "ncV:128:", "avxvnniint8|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpbssd256, "V8iV8iV8iV8i", "ncV:256:", "avxvnniint8|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpbssds128, "V4iV4iV4iV4i", "ncV:128:", "avxvnniint8|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpbssds256, "V8iV8iV8iV8i", "ncV:256:", "avxvnniint8|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpbsud128, "V4iV4iV4iV4i", "ncV:128:", "avxvnniint8|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpbsud256, "V8iV8iV8iV8i", "ncV:256:", "avxvnniint8|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpbsuds128, "V4iV4iV4iV4i", "ncV:128:", "avxvnniint8|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpbsuds256, "V8iV8iV8iV8i", "ncV:256:", "avxvnniint8|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpbuud128, "V4iV4iV4iV4i", "ncV:128:", "avxvnniint8|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpbuud256, "V8iV8iV8iV8i", "ncV:256:", "avxvnniint8|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpbuuds128, "V4iV4iV4iV4i", "ncV:128:", "avxvnniint8|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpbuuds256, "V8iV8iV8iV8i", "ncV:256:", "avxvnniint8|avx10.2-256")
-
-// MOVRS
-TARGET_BUILTIN(__builtin_ia32_prefetchrs, "vvC*", "nc", "movrs")
-
-TARGET_BUILTIN(__builtin_ia32_gather3div2df, "V2dV2dvC*V2OiUcIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3div2di, "V2OiV2OivC*V2OiUcIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3div4df, "V4dV4dvC*V4OiUcIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3div4di, "V4OiV4OivC*V4OiUcIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3div4sf, "V4fV4fvC*V2OiUcIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3div4si, "V4iV4ivC*V2OiUcIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3div8sf, "V4fV4fvC*V4OiUcIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3div8si, "V4iV4ivC*V4OiUcIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3siv2df, "V2dV2dvC*V4iUcIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3siv2di, "V2OiV2OivC*V4iUcIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3siv4df, "V4dV4dvC*V4iUcIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3siv4di, "V4OiV4OivC*V4iUcIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3siv4sf, "V4fV4fvC*V4iUcIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3siv4si, "V4iV4ivC*V4iUcIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3siv8sf, "V8fV8fvC*V8iUcIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gather3siv8si, "V8iV8ivC*V8iUcIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_gathersiv8df, "V8dV8dvC*V8iUcIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_gathersiv16sf, "V16fV16fvC*V16iUsIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_gatherdiv8df, "V8dV8dvC*V8OiUcIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_gatherdiv16sf, "V8fV8fvC*V8OiUcIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_gathersiv8di, "V8OiV8OivC*V8iUcIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_gathersiv16si, "V16iV16ivC*V16iUsIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_gatherdiv8di, "V8OiV8OivC*V8OiUcIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_gatherdiv16si, "V8iV8ivC*V8OiUcIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_scattersiv8df, "vv*UcV8iV8dIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_scattersiv16sf, "vv*UsV16iV16fIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_scatterdiv8df, "vv*UcV8OiV8dIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_scatterdiv16sf, "vv*UcV8OiV8fIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_scattersiv8di, "vv*UcV8iV8OiIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_scattersiv16si, "vv*UsV16iV16iIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_scatterdiv8di, "vv*UcV8OiV8OiIi", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_scatterdiv16si, "vv*UcV8OiV8iIi", "nV:512:", "avx512f,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_knotqi, "UcUc", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_knothi, "UsUs", "nc", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_knotsi, "UiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_knotdi, "UOiUOi", "nc", "avx512bw")
-
-TARGET_BUILTIN(__builtin_ia32_cmpb128_mask, "UsV16cV16cIiUs", "ncV:128:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_cmpd128_mask, "UcV4iV4iIiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cmpq128_mask, "UcV2OiV2OiIiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cmpw128_mask, "UcV8sV8sIiUc", "ncV:128:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_cmpb256_mask, "UiV32cV32cIiUi", "ncV:256:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_cmpd256_mask, "UcV8iV8iIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cmpq256_mask, "UcV4OiV4OiIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cmpw256_mask, "UsV16sV16sIiUs", "ncV:256:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_cmpb512_mask, "UOiV64cV64cIiUOi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_cmpd512_mask, "UsV16iV16iIiUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cmpq512_mask, "UcV8OiV8OiIiUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cmpw512_mask, "UiV32sV32sIiUi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_ucmpb128_mask, "UsV16cV16cIiUs", "ncV:128:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_ucmpd128_mask, "UcV4iV4iIiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_ucmpq128_mask, "UcV2OiV2OiIiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_ucmpw128_mask, "UcV8sV8sIiUc", "ncV:128:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_ucmpb256_mask, "UiV32cV32cIiUi", "ncV:256:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_ucmpd256_mask, "UcV8iV8iIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_ucmpq256_mask, "UcV4OiV4OiIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_ucmpw256_mask, "UsV16sV16sIiUs", "ncV:256:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_ucmpb512_mask, "UOiV64cV64cIiUOi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_ucmpd512_mask, "UsV16iV16iIiUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_ucmpq512_mask, "UcV8OiV8OiIiUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_ucmpw512_mask, "UiV32sV32sIiUi", "ncV:512:", "avx512bw,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_packssdw512, "V32sV16iV16i", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_packsswb512, "V64cV32sV32s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_packusdw512, "V32sV16iV16i", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_packuswb512, "V64cV32sV32s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_pavgb512, "V64cV64cV64c", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_pavgw512, "V32sV32sV32s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_pshufb512, "V64cV64cV64c", "ncV:512:", "avx512bw,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_vpconflictdi_128, "V2OiV2Oi", "ncV:128:", "avx512cd,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpconflictdi_256, "V4OiV4Oi", "ncV:256:", "avx512cd,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpconflictsi_128, "V4iV4i", "ncV:128:", "avx512cd,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpconflictsi_256, "V8iV8i", "ncV:256:", "avx512cd,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpconflictdi_512, "V8OiV8Oi", "ncV:512:", "avx512cd,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpconflictsi_512, "V16iV16i", "ncV:512:", "avx512cd,evex512")
-TARGET_BUILTIN(__builtin_ia32_vplzcntd_512, "V16iV16i", "ncV:512:", "avx512cd,evex512")
-TARGET_BUILTIN(__builtin_ia32_vplzcntq_512, "V8OiV8Oi", "ncV:512:", "avx512cd,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_vpshufbitqmb128_mask, "UsV16cV16cUs", "ncV:128:", "avx512vl,avx512bitalg")
-TARGET_BUILTIN(__builtin_ia32_vpshufbitqmb256_mask, "UiV32cV32cUi", "ncV:256:", "avx512vl,avx512bitalg")
-TARGET_BUILTIN(__builtin_ia32_vpshufbitqmb512_mask, "UOiV64cV64cUOi", "ncV:512:", "avx512bitalg,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_pmulhrsw512, "V32sV32sV32s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmulhuw512, "V32sV32sV32s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmulhw512, "V32sV32sV32s", "ncV:512:", "avx512bw,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_addpd512, "V8dV8dV8dIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_addps512, "V16fV16fV16fIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_divpd512, "V8dV8dV8dIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_divps512, "V16fV16fV16fIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_mulpd512, "V8dV8dV8dIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_mulps512, "V16fV16fV16fIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_subpd512, "V8dV8dV8dIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_subps512, "V16fV16fV16fIi", "ncV:512:", "avx512f,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_pmaddubsw512, "V32sV64cV64c", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmaddwd512, "V16iV32sV32s", "ncV:512:", "avx512bw,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_addss_round_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_divss_round_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_mulss_round_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_subss_round_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_maxss_round_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_minss_round_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_addsd_round_mask, "V2dV2dV2dV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_divsd_round_mask, "V2dV2dV2dV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_mulsd_round_mask, "V2dV2dV2dV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_subsd_round_mask, "V2dV2dV2dV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_maxsd_round_mask, "V2dV2dV2dV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_minsd_round_mask, "V2dV2dV2dV2dUcIi", "ncV:128:", "avx512f")
-
-TARGET_BUILTIN(__builtin_ia32_compressdf128_mask, "V2dV2dV2dUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_compressdf256_mask, "V4dV4dV4dUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_compressdi128_mask, "V2OiV2OiV2OiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_compressdi256_mask, "V4OiV4OiV4OiUc", "ncV:256:", "avx512vl")
-
-TARGET_BUILTIN(__builtin_ia32_compresshi128_mask, "V8sV8sV8sUc", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_compresshi256_mask, "V16sV16sV16sUs", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_compressqi128_mask, "V16cV16cV16cUs", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_compressqi256_mask, "V32cV32cV32cUi", "ncV:256:", "avx512vl,avx512vbmi2")
-
-TARGET_BUILTIN(__builtin_ia32_compresssf128_mask, "V4fV4fV4fUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_compresssf256_mask, "V8fV8fV8fUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_compresssi128_mask, "V4iV4iV4iUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_compresssi256_mask, "V8iV8iV8iUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_compressstoredf128_mask, "vV2d*V2dUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_compressstoredf256_mask, "vV4d*V4dUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_compressstoredi128_mask, "vV2Oi*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_compressstoredi256_mask, "vV4Oi*V4OiUc", "nV:256:", "avx512vl")
-
-TARGET_BUILTIN(__builtin_ia32_compressstorehi128_mask, "vV8s*V8sUc", "nV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_compressstorehi256_mask, "vV16s*V16sUs", "nV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_compressstoreqi128_mask, "vV16c*V16cUs", "nV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_compressstoreqi256_mask, "vV32c*V32cUi", "nV:256:", "avx512vl,avx512vbmi2")
-
-TARGET_BUILTIN(__builtin_ia32_compressstoresf128_mask, "vV4f*V4fUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_compressstoresf256_mask, "vV8f*V8fUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_compressstoresi128_mask, "vV4i*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_compressstoresi256_mask, "vV8i*V8iUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2dq128_mask, "V4iV2dV4iUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2ps_mask, "V4fV2dV4fUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2udq128_mask, "V4iV2dV4iUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2udq256_mask, "V4iV4dV4iUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtps2udq128_mask, "V4iV4fV4iUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtps2udq256_mask, "V8iV8fV8iUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvttpd2dq128_mask, "V4iV2dV4iUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvttpd2udq128_mask, "V4iV2dV4iUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvttpd2udq256_mask, "V4iV4dV4iUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvttps2udq128_mask, "V4iV4fV4iUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvttps2udq256_mask, "V8iV8fV8iUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expanddf128_mask, "V2dV2dV2dUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expanddf256_mask, "V4dV4dV4dUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expanddi128_mask, "V2OiV2OiV2OiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expanddi256_mask, "V4OiV4OiV4OiUc", "ncV:256:", "avx512vl")
-
-TARGET_BUILTIN(__builtin_ia32_expandhi128_mask, "V8sV8sV8sUc", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_expandhi256_mask, "V16sV16sV16sUs", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_expandqi128_mask, "V16cV16cV16cUs", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_expandqi256_mask, "V32cV32cV32cUi", "ncV:256:", "avx512vl,avx512vbmi2")
-
-TARGET_BUILTIN(__builtin_ia32_expandloaddf128_mask, "V2dV2dC*V2dUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expandloaddf256_mask, "V4dV4dC*V4dUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expandloaddi128_mask, "V4iV2OiC*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expandloaddi256_mask, "V4OiV4OiC*V4OiUc", "nV:256:", "avx512vl")
-
-TARGET_BUILTIN(__builtin_ia32_expandloadhi128_mask, "V8sV8sC*V8sUc", "nV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_expandloadhi256_mask, "V16sV16sC*V16sUs", "nV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_expandloadqi128_mask, "V16cV16cC*V16cUs", "nV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_expandloadqi256_mask, "V32cV32cC*V32cUi", "nV:256:", "avx512vl,avx512vbmi2")
-
-TARGET_BUILTIN(__builtin_ia32_expandloadsf128_mask, "V4fV4fC*V4fUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expandloadsf256_mask, "V8fV8fC*V8fUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expandloadsi128_mask, "V4iV4iC*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expandloadsi256_mask, "V8iV8iC*V8iUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expandsf128_mask, "V4fV4fV4fUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expandsf256_mask, "V8fV8fV8fUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expandsi128_mask, "V4iV4iV4iUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_expandsi256_mask, "V8iV8iV8iUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_getexppd128_mask, "V2dV2dV2dUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_getexppd256_mask, "V4dV4dV4dUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_getexpps128_mask, "V4fV4fV4fUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_getexpps256_mask, "V8fV8fV8fUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rndscalepd_128_mask, "V2dV2dIiV2dUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rndscalepd_256_mask, "V4dV4dIiV4dUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rndscaleps_128_mask, "V4fV4fIiV4fUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rndscaleps_256_mask, "V8fV8fIiV8fUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scalefpd128_mask, "V2dV2dV2dV2dUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scalefpd256_mask, "V4dV4dV4dV4dUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scalefps128_mask, "V4fV4fV4fV4fUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scalefps256_mask, "V8fV8fV8fV8fUc", "ncV:256:", "avx512vl")
-
-TARGET_BUILTIN(__builtin_ia32_scatterdiv2df, "vv*UcV2OiV2dIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scatterdiv2di, "vv*UcV2OiV2OiIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scatterdiv4df, "vv*UcV4OiV4dIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scatterdiv4di, "vv*UcV4OiV4OiIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scatterdiv4sf, "vv*UcV2OiV4fIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scatterdiv4si, "vv*UcV2OiV4iIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scatterdiv8sf, "vv*UcV4OiV4fIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scatterdiv8si, "vv*UcV4OiV4iIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scattersiv2df, "vv*UcV4iV2dIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scattersiv2di, "vv*UcV4iV2OiIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scattersiv4df, "vv*UcV4iV4dIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scattersiv4di, "vv*UcV4iV4OiIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scattersiv4sf, "vv*UcV4iV4fIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scattersiv4si, "vv*UcV4iV4iIi", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scattersiv8sf, "vv*UcV8iV8fIi", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scattersiv8si, "vv*UcV8iV8iIi", "nV:256:", "avx512vl")
-
-TARGET_BUILTIN(__builtin_ia32_vpermi2vard128, "V4iV4iV4iV4i", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpermi2vard256, "V8iV8iV8iV8i", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpermi2vard512, "V16iV16iV16iV16i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varpd128, "V2dV2dV2OiV2d", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varpd256, "V4dV4dV4OiV4d", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varpd512, "V8dV8dV8OiV8d", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varps128, "V4fV4fV4iV4f", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varps256, "V8fV8fV8iV8f", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varps512, "V16fV16fV16iV16f", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varq128, "V2OiV2OiV2OiV2Oi", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varq256, "V4OiV4OiV4OiV4Oi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varq512, "V8OiV8OiV8OiV8Oi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varqi128, "V16cV16cV16cV16c", "ncV:128:", "avx512vbmi,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varqi256, "V32cV32cV32cV32c", "ncV:256:", "avx512vbmi,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varqi512, "V64cV64cV64cV64c", "ncV:512:", "avx512vbmi,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varhi128, "V8sV8sV8sV8s", "ncV:128:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varhi256, "V16sV16sV16sV16s", "ncV:256:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_vpermi2varhi512, "V32sV32sV32sV32s", "ncV:512:", "avx512bw,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_vpshldd128, "V4iV4iV4iIi", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshldd256, "V8iV8iV8iIi", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshldd512, "V16iV16iV16iIi", "ncV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpshldq128, "V2OiV2OiV2OiIi", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshldq256, "V4OiV4OiV4OiIi", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshldq512, "V8OiV8OiV8OiIi", "ncV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpshldw128, "V8sV8sV8sIi", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshldw256, "V16sV16sV16sIi", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshldw512, "V32sV32sV32sIi", "ncV:512:", "avx512vbmi2,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_vpshldvd128, "V4iV4iV4iV4i", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshldvd256, "V8iV8iV8iV8i", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshldvd512, "V16iV16iV16iV16i", "ncV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpshldvq128, "V2OiV2OiV2OiV2Oi", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshldvq256, "V4OiV4OiV4OiV4Oi", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshldvq512, "V8OiV8OiV8OiV8Oi", "ncV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpshldvw128, "V8sV8sV8sV8s", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshldvw256, "V16sV16sV16sV16s", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshldvw512, "V32sV32sV32sV32s", "ncV:512:", "avx512vbmi2,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_vpshrdvd128, "V4iV4iV4iV4i", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshrdvd256, "V8iV8iV8iV8i", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshrdvd512, "V16iV16iV16iV16i", "ncV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpshrdvq128, "V2OiV2OiV2OiV2Oi", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshrdvq256, "V4OiV4OiV4OiV4Oi", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshrdvq512, "V8OiV8OiV8OiV8Oi", "ncV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpshrdvw128, "V8sV8sV8sV8s", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshrdvw256, "V16sV16sV16sV16s", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshrdvw512, "V32sV32sV32sV32s", "ncV:512:", "avx512vbmi2,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_vpshrdd128, "V4iV4iV4iIi", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshrdd256, "V8iV8iV8iIi", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshrdd512, "V16iV16iV16iIi", "ncV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpshrdq128, "V2OiV2OiV2OiIi", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshrdq256, "V4OiV4OiV4OiIi", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshrdq512, "V8OiV8OiV8OiIi", "ncV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpshrdw128, "V8sV8sV8sIi", "ncV:128:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshrdw256, "V16sV16sV16sIi", "ncV:256:", "avx512vl,avx512vbmi2")
-TARGET_BUILTIN(__builtin_ia32_vpshrdw512, "V32sV32sV32sIi", "ncV:512:", "avx512vbmi2,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_pmovswb512_mask, "V32cV32sV32cUi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovuswb512_mask, "V32cV32sV32cUi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovwb512_mask, "V32cV32sV32cUi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2qq128_mask, "V2OiV2dV2OiUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2qq256_mask, "V4OiV4dV4OiUc", "ncV:256:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2uqq128_mask, "V2OiV2dV2OiUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2uqq256_mask, "V4OiV4dV4OiUc", "ncV:256:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvtps2qq128_mask, "V2OiV4fV2OiUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvtps2qq256_mask, "V4OiV4fV4OiUc", "ncV:256:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvtps2uqq128_mask, "V2OiV4fV2OiUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvtps2uqq256_mask, "V4OiV4fV4OiUc", "ncV:256:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvtqq2ps128_mask, "V4fV2OiV4fUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvttpd2qq128_mask, "V2OiV2dV2OiUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvttpd2qq256_mask, "V4OiV4dV4OiUc", "ncV:256:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvttpd2uqq128_mask, "V2OiV2dV2OiUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvttpd2uqq256_mask, "V4OiV4dV4OiUc", "ncV:256:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvttps2qq128_mask, "V2OiV4fV2OiUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvttps2qq256_mask, "V4OiV4fV4OiUc", "ncV:256:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvttps2uqq128_mask, "V2OiV4fV2OiUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvttps2uqq256_mask, "V4OiV4fV4OiUc", "ncV:256:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_cvtuqq2ps128_mask, "V4fV2OiV4fUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_rangepd128_mask, "V2dV2dV2dIiV2dUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_rangepd256_mask, "V4dV4dV4dIiV4dUc", "ncV:256:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_rangeps128_mask, "V4fV4fV4fIiV4fUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_rangeps256_mask, "V8fV8fV8fIiV8fUc", "ncV:256:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_rangesd128_round_mask, "V2dV2dV2dV2dUcIiIi", "ncV:128:", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_rangess128_round_mask, "V4fV4fV4fV4fUcIiIi", "ncV:128:", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_reducepd128_mask, "V2dV2dIiV2dUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_reducepd256_mask, "V4dV4dIiV4dUc", "ncV:256:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_reduceps128_mask, "V4fV4fIiV4fUc", "ncV:128:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_reduceps256_mask, "V8fV8fIiV8fUc", "ncV:256:", "avx512vl,avx512dq")
-TARGET_BUILTIN(__builtin_ia32_reducesd_mask, "V2dV2dV2dV2dUcIiIi", "ncV:128:", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_reducess_mask, "V4fV4fV4fV4fUcIiIi", "ncV:128:", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_pmovswb128_mask, "V16cV8sV16cUc", "ncV:128:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_pmovswb256_mask, "V16cV16sV16cUs", "ncV:256:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_pmovuswb128_mask, "V16cV8sV16cUc", "ncV:128:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_pmovuswb256_mask, "V16cV16sV16cUs", "ncV:256:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_pmovwb128_mask, "V16cV8sV16cUc", "ncV:128:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2qq512_mask, "V8OiV8dV8OiUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtpd2uqq512_mask, "V8OiV8dV8OiUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtps2qq512_mask, "V8OiV8fV8OiUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtps2uqq512_mask, "V8OiV8fV8OiUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtqq2pd512_mask, "V8dV8OiV8dUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtqq2ps512_mask, "V8fV8OiV8fUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvttpd2qq512_mask, "V8OiV8dV8OiUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvttpd2uqq512_mask, "V8OiV8dV8OiUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvttps2qq512_mask, "V8OiV8fV8OiUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvttps2uqq512_mask, "V8OiV8fV8OiUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtuqq2pd512_mask, "V8dV8OiV8dUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtuqq2ps512_mask, "V8fV8OiV8fUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_rangepd512_mask, "V8dV8dV8dIiV8dUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_rangeps512_mask, "V16fV16fV16fIiV16fUsIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_reducepd512_mask, "V8dV8dIiV8dUcIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduceps512_mask, "V16fV16fIiV16fUsIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_prold512, "V16iV16iIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_prolq512, "V8OiV8OiIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_prold128, "V4iV4iIi", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prold256, "V8iV8iIi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prolq128, "V2OiV2OiIi", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prolq256, "V4OiV4OiIi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prolvd512, "V16iV16iV16i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_prolvq512, "V8OiV8OiV8Oi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_prord512, "V16iV16iIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_prorq512, "V8OiV8OiIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_prolvd128, "V4iV4iV4i", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prolvd256, "V8iV8iV8i", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prolvq128, "V2OiV2OiV2Oi", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prolvq256, "V4OiV4OiV4Oi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prord128, "V4iV4iIi", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prord256, "V8iV8iIi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prorq128, "V2OiV2OiIi", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prorq256, "V4OiV4OiIi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prorvd512, "V16iV16iV16i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_prorvq512, "V8OiV8OiV8Oi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_prorvd128, "V4iV4iV4i", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prorvd256, "V8iV8iV8i", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prorvq128, "V2OiV2OiV2Oi", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_prorvq256, "V4OiV4OiV4Oi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pshufhw512, "V32sV32sIi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_pshuflw512, "V32sV32sIi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_psllv32hi, "V32sV32sV32s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_psllw512, "V32sV32sV8s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_psllwi512, "V32sV32si", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_psllv16hi, "V16sV16sV16s", "ncV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_psllv8hi, "V8sV8sV8s", "ncV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pslldi512, "V16iV16ii", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psllqi512, "V8OiV8Oii", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrlv32hi, "V32sV32sV32s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrlv16hi, "V16sV16sV16s", "ncV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_psrlv8hi, "V8sV8sV8s", "ncV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_psrldi512, "V16iV16ii", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrlqi512, "V8OiV8Oii", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrav32hi, "V32sV32sV32s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrav16hi, "V16sV16sV16s", "ncV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_psrav8hi, "V8sV8sV8s", "ncV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_psravq128, "V2OiV2OiV2Oi", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_psravq256, "V4OiV4OiV4Oi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_psraw512, "V32sV32sV8s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrawi512, "V32sV32si", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrlw512, "V32sV32sV8s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrlwi512, "V32sV32si", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_pslldqi512_byteshift, "V8OiV8OiIi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrldqi512_byteshift, "V8OiV8OiIi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_movdqa32load128_mask, "V4iV4iC*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_movdqa32load256_mask, "V8iV8iC*V8iUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_movdqa32load512_mask, "V16iV16iC*V16iUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_movdqa32store512_mask, "vV16i*V16iUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_movdqa64load512_mask, "V8OiV8OiC*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_movdqa64store512_mask, "vV8Oi*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_movdqa32store128_mask, "vV4i*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_movdqa32store256_mask, "vV8i*V8iUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_movdqa64load128_mask, "V2OiV2OiC*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_movdqa64load256_mask, "V4OiV4OiC*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_movdqa64store128_mask, "vV2Oi*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_movdqa64store256_mask, "vV4Oi*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpmadd52huq512, "V8OiV8OiV8OiV8Oi", "ncV:512:", "avx512ifma,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpmadd52luq512, "V8OiV8OiV8OiV8Oi", "ncV:512:", "avx512ifma,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpmadd52huq128, "V2OiV2OiV2OiV2Oi", "ncV:128:", "avx512ifma,avx512vl|avxifma")
-TARGET_BUILTIN(__builtin_ia32_vpmadd52huq256, "V4OiV4OiV4OiV4Oi", "ncV:256:", "avx512ifma,avx512vl|avxifma")
-TARGET_BUILTIN(__builtin_ia32_vpmadd52luq128, "V2OiV2OiV2OiV2Oi", "ncV:128:", "avx512ifma,avx512vl|avxifma")
-TARGET_BUILTIN(__builtin_ia32_vpmadd52luq256, "V4OiV4OiV4OiV4Oi", "ncV:256:", "avx512ifma,avx512vl|avxifma")
-TARGET_BUILTIN(__builtin_ia32_vcomisd, "iV2dV2dIiIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vcomiss, "iV4fV4fIiIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_kunpckdi, "UOiUOiUOi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kunpcksi, "UiUiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_loaddquhi512_mask, "V32sV32sC*V32sUi", "nV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_loaddquqi512_mask, "V64cV64cC*V64cUOi", "nV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_fixupimmpd512_mask, "V8dV8dV8dV8OiIiUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_fixupimmpd512_maskz, "V8dV8dV8dV8OiIiUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_fixupimmps512_mask, "V16fV16fV16fV16iIiUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_fixupimmps512_maskz, "V16fV16fV16fV16iIiUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_fixupimmsd_mask, "V2dV2dV2dV2OiIiUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_fixupimmsd_maskz, "V2dV2dV2dV2OiIiUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_fixupimmss_mask, "V4fV4fV4fV4iIiUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_fixupimmss_maskz, "V4fV4fV4fV4iIiUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_getexpsd128_round_mask, "V2dV2dV2dV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_getexpss128_round_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_getmantsd_round_mask, "V2dV2dV2dIiV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_getmantss_round_mask, "V4fV4fV4fIiV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_loaddquhi128_mask, "V8sV8sC*V8sUc", "nV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddquhi256_mask, "V16sV16sC*V16sUs", "nV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddquqi128_mask, "V16cV16cC*V16cUs", "nV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddquqi256_mask, "V32cV32cC*V32cUi", "nV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fixupimmpd128_mask, "V2dV2dV2dV2OiIiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fixupimmpd128_maskz, "V2dV2dV2dV2OiIiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fixupimmpd256_mask, "V4dV4dV4dV4OiIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fixupimmpd256_maskz, "V4dV4dV4dV4OiIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fixupimmps128_mask, "V4fV4fV4fV4iIiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fixupimmps128_maskz, "V4fV4fV4fV4iIiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fixupimmps256_mask, "V8fV8fV8fV8iIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fixupimmps256_maskz, "V8fV8fV8fV8iIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadapd128_mask, "V2dV2dC*V2dUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadsd128_mask, "V2dV2dC*V2dUc", "nV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_loadapd256_mask, "V4dV4dC*V4dUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadaps128_mask, "V4fV4fC*V4fUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadss128_mask, "V4fV4fC*V4fUc", "nV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_loadaps256_mask, "V8fV8fC*V8fUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddqudi128_mask, "V2OiV2OiC*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddqudi256_mask, "V4OiV4OiC*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddqusi128_mask, "V4iV4iC*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loaddqusi256_mask, "V8iV8iC*V8iUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadupd128_mask, "V2dV2dC*V2dUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadupd256_mask, "V4dV4dC*V4dUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadups128_mask, "V4fV4fC*V4fUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_loadups256_mask, "V8fV8fC*V8fUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_storedquhi512_mask, "vV32s*V32sUi", "nV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_storedquqi512_mask, "vV64c*V64cUOi", "nV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_storedquhi128_mask, "vV8s*V8sUc", "nV:128:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_storedquhi256_mask, "vV16s*V16sUs", "nV:256:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_storedquqi128_mask, "vV16c*V16cUs", "nV:128:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_storedquqi256_mask, "vV32c*V32cUi", "nV:256:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_storeapd128_mask, "vV2d*V2dUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_storesd128_mask, "vV2d*V2dUc", "nV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_storeapd256_mask, "vV4d*V4dUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_storeaps128_mask, "vV4f*V4fUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_storess128_mask, "vV4f*V4fUc", "nV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_storeaps256_mask, "vV8f*V8fUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_storedqudi128_mask, "vV2Oi*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_storedqudi256_mask, "vV4Oi*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_storedqusi128_mask, "vV4i*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_storedqusi256_mask, "vV8i*V8iUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_storeupd128_mask, "vV2d*V2dUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_storeupd256_mask, "vV4d*V4dUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_storeups128_mask, "vV4f*V4fUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_storeups256_mask, "vV8f*V8fUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rcp14pd128_mask, "V2dV2dV2dUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rcp14pd256_mask, "V4dV4dV4dUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rcp14ps128_mask, "V4fV4fV4fUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rcp14ps256_mask, "V8fV8fV8fUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vplzcntd_128, "V4iV4i", "ncV:128:", "avx512cd,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vplzcntd_256, "V8iV8i", "ncV:256:", "avx512cd,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vplzcntq_128, "V2OiV2Oi", "ncV:128:", "avx512cd,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vplzcntq_256, "V4OiV4Oi", "ncV:256:", "avx512cd,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtsd2si32, "iV2dIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vcvtsd2usi32, "UiV2dIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vcvtss2si32, "iV4fIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vcvtss2usi32, "UiV4fIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vcvttsd2si32, "iV2dIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vcvttsd2usi32, "UiV2dIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vcvttss2si32, "iV4fIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vcvttss2usi32, "UiV4fIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vpermilpd512, "V8dV8dIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpermilps512, "V16fV16fIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpermilvarpd512, "V8dV8dV8Oi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpermilvarps512, "V16fV16fV16i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_rndscalesd_round_mask, "V2dV2dV2dV2dUcIiIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_rndscaless_round_mask, "V4fV4fV4fV4fUcIiIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_scalefpd512_mask, "V8dV8dV8dV8dUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_scalefps512_mask, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_scalefsd_round_mask, "V2dV2dV2dV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_scalefss_round_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_psradi512, "V16iV16ii", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psraqi512, "V8OiV8Oii", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psraq128, "V2OiV2OiV2Oi", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_psraq256, "V4OiV4OiV2Oi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_psraqi128, "V2OiV2Oii", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_psraqi256, "V4OiV4Oii", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pslld512, "V16iV16iV4i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psllq512, "V8OiV8OiV2Oi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psllv16si, "V16iV16iV16i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psllv8di, "V8OiV8OiV8Oi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrad512, "V16iV16iV4i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psraq512, "V8OiV8OiV2Oi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrav16si, "V16iV16iV16i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrav8di, "V8OiV8OiV8Oi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrld512, "V16iV16iV4i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrlq512, "V8OiV8OiV2Oi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrlv16si, "V16iV16iV16i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_psrlv8di, "V8OiV8OiV8Oi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pternlogd512_mask, "V16iV16iV16iV16iIiUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pternlogd512_maskz, "V16iV16iV16iV16iIiUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pternlogq512_mask, "V8OiV8OiV8OiV8OiIiUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pternlogq512_maskz, "V8OiV8OiV8OiV8OiIiUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pternlogd128_mask, "V4iV4iV4iV4iIiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pternlogd128_maskz, "V4iV4iV4iV4iIiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pternlogd256_mask, "V8iV8iV8iV8iIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pternlogd256_maskz, "V8iV8iV8iV8iIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pternlogq128_mask, "V2OiV2OiV2OiV2OiIiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pternlogq128_maskz, "V2OiV2OiV2OiV2OiIiUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pternlogq256_mask, "V4OiV4OiV4OiV4OiIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pternlogq256_maskz, "V4OiV4OiV4OiV4OiIiUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_shuf_f32x4, "V16fV16fV16fIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_shuf_f64x2, "V8dV8dV8dIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_shuf_i32x4, "V16iV16iV16iIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_shuf_i64x2, "V8OiV8OiV8OiIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_shufpd512, "V8dV8dV8dIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_shufps512, "V16fV16fV16fIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_shuf_f32x4_256, "V8fV8fV8fIi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_shuf_f64x2_256, "V4dV4dV4dIi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_shuf_i32x4_256, "V8iV8iV8iIi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_shuf_i64x2_256, "V4OiV4OiV4OiIi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_sqrtsd_round_mask, "V2dV2dV2dV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_sqrtss_round_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_rsqrt14pd128_mask, "V2dV2dV2dUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rsqrt14pd256_mask, "V4dV4dV4dUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rsqrt14ps128_mask, "V4fV4fV4fUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rsqrt14ps256_mask, "V8fV8fV8fUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtb2mask512, "UOiV64c", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtmask2b512, "V64cUOi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtmask2w512, "V32sUi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtd2mask512, "UsV16i", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtmask2d512, "V16iUs", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtmask2q512, "V8OiUc", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtq2mask512, "UcV8Oi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtb2mask128, "UsV16c", "ncV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtb2mask256, "UiV32c", "ncV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtmask2b128, "V16cUs", "ncV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtmask2b256, "V32cUi", "ncV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtmask2w128, "V8sUc", "ncV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtmask2w256, "V16sUs", "ncV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtd2mask128, "UcV4i", "ncV:128:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtd2mask256, "UcV8i", "ncV:256:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtmask2d128, "V4iUc", "ncV:128:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtmask2d256, "V8iUc", "ncV:256:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtmask2q128, "V2OiUc", "ncV:128:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtmask2q256, "V4OiUc", "ncV:256:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtq2mask128, "UcV2Oi", "ncV:128:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtq2mask256, "UcV4Oi", "ncV:256:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsdb512_mask, "V16cV16iV16cUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovsdb512mem_mask, "vV16c*V16iUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovswb512mem_mask, "vV32c*V32sUi", "nV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovsdw512_mask, "V16sV16iV16sUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovsdw512mem_mask, "vV16s*V16iUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovsqb512_mask, "V16cV8OiV16cUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovsqb512mem_mask, "vV16c*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovsqd512_mask, "V8iV8OiV8iUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovsqd512mem_mask, "vV8i*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovsqw512_mask, "V8sV8OiV8sUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovsqw512mem_mask, "vV8s*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovsdb128_mask, "V16cV4iV16cUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsdb128mem_mask, "vV16c*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovswb128mem_mask, "vV16c*V8sUc", "nV:128:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_pmovsdb256_mask, "V16cV8iV16cUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsdb256mem_mask, "vV16c*V8iUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovswb256mem_mask, "vV16c*V16sUs", "nV:256:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_pmovsdw128_mask, "V8sV4iV8sUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsdw128mem_mask, "vV8s*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsdw256_mask, "V8sV8iV8sUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsdw256mem_mask, "vV8s*V8iUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsqb128_mask, "V16cV2OiV16cUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsqb128mem_mask, "vV16c*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsqb256_mask, "V16cV4OiV16cUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsqb256mem_mask, "vV16c*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsqd128_mask, "V4iV2OiV4iUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsqd128mem_mask, "vV4i*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsqd256_mask, "V4iV4OiV4iUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsqd256mem_mask, "vV4i*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsqw128_mask, "V8sV2OiV8sUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsqw128mem_mask, "vV8s*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsqw256_mask, "V8sV4OiV8sUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovsqw256mem_mask, "vV8s*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusdb512_mask, "V16cV16iV16cUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovusdb512mem_mask, "vV16c*V16iUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovuswb512mem_mask, "vV32c*V32sUi", "nV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovusdw512_mask, "V16sV16iV16sUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovusdw512mem_mask, "vV16s*V16iUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovusqb512_mask, "V16cV8OiV16cUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovusqb512mem_mask, "vV16c*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovusqd512_mask, "V8iV8OiV8iUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovusqd512mem_mask, "vV8i*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovusqw512_mask, "V8sV8OiV8sUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovusqw512mem_mask, "vV8s*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovusdb128_mask, "V16cV4iV16cUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusdb128mem_mask, "vV16c*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovuswb128mem_mask, "vV16c*V8sUc", "nV:128:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_pmovusdb256_mask, "V16cV8iV16cUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusdb256mem_mask, "vV16c*V8iUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovuswb256mem_mask, "vV16c*V16sUs", "nV:256:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_pmovusdw128_mask, "V8sV4iV8sUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusdw128mem_mask, "vV8s*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusdw256_mask, "V8sV8iV8sUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusdw256mem_mask, "vV8s*V8iUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusqb128_mask, "V16cV2OiV16cUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusqb128mem_mask, "vV16c*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusqb256_mask, "V16cV4OiV16cUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusqb256mem_mask, "vV16c*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusqd128_mask, "V4iV2OiV4iUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusqd128mem_mask, "vV4i*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusqd256_mask, "V4iV4OiV4iUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusqd256mem_mask, "vV4i*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusqw128_mask, "V8sV2OiV8sUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusqw128mem_mask, "vV8s*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusqw256_mask, "V8sV4OiV8sUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovusqw256mem_mask, "vV8s*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovdb512_mask, "V16cV16iV16cUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovdb512mem_mask, "vV16c*V16iUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovwb512mem_mask, "vV32c*V32sUi", "nV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovdw512_mask, "V16sV16iV16sUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovdw512mem_mask, "vV16s*V16iUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovqb512_mask, "V16cV8OiV16cUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovqb512mem_mask, "vV16c*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovqd512_mask, "V8iV8OiV8iUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovqd512mem_mask, "vV8i*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovqw512_mask, "V8sV8OiV8sUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovqw512mem_mask, "vV8s*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_pmovdb128_mask, "V16cV4iV16cUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovwb128mem_mask, "vV16c*V8sUc", "nV:128:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_pmovdb128mem_mask, "vV16c*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovdb256_mask, "V16cV8iV16cUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovdb256mem_mask, "vV16c*V8iUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovwb256mem_mask, "vV16c*V16sUs", "nV:256:", "avx512vl,avx512bw")
-TARGET_BUILTIN(__builtin_ia32_pmovdw128_mask, "V8sV4iV8sUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovdw128mem_mask, "vV8s*V4iUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovdw256_mask, "V8sV8iV8sUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovdw256mem_mask, "vV8s*V8iUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovqb128_mask, "V16cV2OiV16cUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovqb128mem_mask, "vV16c*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovqb256_mask, "V16cV4OiV16cUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovqb256mem_mask, "vV16c*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovqd128_mask, "V4iV2OiV4iUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovqd128mem_mask, "vV4i*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovqd256mem_mask, "vV4i*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovqw128_mask, "V8sV2OiV8sUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovqw128mem_mask, "vV8s*V2OiUc", "nV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovqw256_mask, "V8sV4OiV8sUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_pmovqw256mem_mask, "vV8s*V4OiUc", "nV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_extractf32x8_mask, "V8fV16fIiV8fUc", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_extractf64x2_512_mask, "V2dV8dIiV2dUc", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_extracti32x8_mask, "V8iV16iIiV8iUc", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_extracti64x2_512_mask, "V2OiV8OiIiV2OiUc", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_extracti32x4_mask, "V4iV16iIiV4iUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_extracti64x4_mask, "V4OiV8OiIiV4OiUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_extractf64x2_256_mask, "V2dV4dIiV2dUc", "ncV:256:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_extracti64x2_256_mask, "V2OiV4OiIiV2OiUc", "ncV:256:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_extractf32x4_256_mask, "V4fV8fIiV4fUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_extracti32x4_256_mask, "V4iV8iIiV4iUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_insertf32x8, "V16fV16fV8fIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_insertf64x2_512, "V8dV8dV2dIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_inserti32x8, "V16iV16iV8iIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_inserti64x2_512, "V8OiV8OiV2OiIi", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_insertf64x4, "V8dV8dV4dIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_inserti64x4, "V8OiV8OiV4OiIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_insertf64x2_256, "V4dV4dV2dIi", "ncV:256:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_inserti64x2_256, "V4OiV4OiV2OiIi", "ncV:256:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_insertf32x4_256, "V8fV8fV4fIi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_inserti32x4_256, "V8iV8iV4iIi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_insertf32x4, "V16fV16fV4fIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_inserti32x4, "V16iV16iV4iIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_getmantpd128_mask, "V2dV2dIiV2dUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_getmantpd256_mask, "V4dV4dIiV4dUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_getmantps128_mask, "V4fV4fIiV4fUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_getmantps256_mask, "V8fV8fIiV8fUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_getmantpd512_mask, "V8dV8dIiV8dUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_getmantps512_mask, "V16fV16fIiV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_getexppd512_mask, "V8dV8dV8dUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_getexpps512_mask, "V16fV16fV16fUsIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddss3_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vfmaddss3_maskz, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vfmaddss3_mask3, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsd3_mask, "V2dV2dV2dV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsd3_maskz, "V2dV2dV2dV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsd3_mask3, "V2dV2dV2dV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vfmsubsd3_mask3, "V2dV2dV2dV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vfmsubss3_mask3, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_permdf512, "V8dV8dIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_permdi512, "V8OiV8OiIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_permvarhi512, "V32sV32sV32s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_permvardf512, "V8dV8dV8Oi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_permvardi512, "V8OiV8OiV8Oi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_permvarsf512, "V16fV16fV16i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_permvarsi512, "V16iV16iV16i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_permvarqi512, "V64cV64cV64c", "ncV:512:", "avx512vbmi,evex512")
-TARGET_BUILTIN(__builtin_ia32_permvarqi128, "V16cV16cV16c", "ncV:128:", "avx512vbmi,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_permvarqi256, "V32cV32cV32c", "ncV:256:", "avx512vbmi,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_permvarhi128, "V8sV8sV8s", "ncV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_permvarhi256, "V16sV16sV16s", "ncV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_permvardf256, "V4dV4dV4Oi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_permvardi256, "V4OiV4OiV4Oi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fpclasspd128_mask, "UcV2dIiUc", "ncV:128:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fpclasspd256_mask, "UcV4dIiUc", "ncV:256:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fpclassps128_mask, "UcV4fIiUc", "ncV:128:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fpclassps256_mask, "UcV8fIiUc", "ncV:256:", "avx512dq,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fpclassps512_mask, "UsV16fIiUs", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_fpclasspd512_mask, "UcV8dIiUc", "ncV:512:", "avx512dq,evex512")
-TARGET_BUILTIN(__builtin_ia32_fpclasssd_mask, "UcV2dIiUc", "ncV:128:", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_fpclassss_mask, "UcV4fIiUc", "ncV:128:", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_kaddqi, "UcUcUc", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_kaddhi, "UsUsUs", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_kaddsi, "UiUiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kadddi, "UOiUOiUOi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kandqi, "UcUcUc", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_kandhi, "UsUsUs", "nc", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_kandsi, "UiUiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kanddi, "UOiUOiUOi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kandnqi, "UcUcUc", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_kandnhi, "UsUsUs", "nc", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_kandnsi, "UiUiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kandndi, "UOiUOiUOi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_korqi, "UcUcUc", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_korhi, "UsUsUs", "nc", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_korsi, "UiUiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kordi, "UOiUOiUOi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kortestcqi, "iUcUc", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_kortestzqi, "iUcUc", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_kortestchi, "iUsUs", "nc", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_kortestzhi, "iUsUs", "nc", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_kortestcsi, "iUiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kortestzsi, "iUiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kortestcdi, "iUOiUOi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kortestzdi, "iUOiUOi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_ktestcqi, "iUcUc", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_ktestzqi, "iUcUc", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_ktestchi, "iUsUs", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_ktestzhi, "iUsUs", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_ktestcsi, "iUiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_ktestzsi, "iUiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_ktestcdi, "iUOiUOi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_ktestzdi, "iUOiUOi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kunpckhi, "UsUsUs", "nc", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_kxnorqi, "UcUcUc", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_kxnorhi, "UsUsUs", "nc", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_kxnorsi, "UiUiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kxnordi, "UOiUOiUOi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kxorqi, "UcUcUc", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_kxorhi, "UsUsUs", "nc", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_kxorsi, "UiUiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kxordi, "UOiUOiUOi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kshiftliqi, "UcUcIUi", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_kshiftlihi, "UsUsIUi", "nc", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_kshiftlisi, "UiUiIUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kshiftlidi, "UOiUOiIUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kshiftriqi, "UcUcIUi", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_kshiftrihi, "UsUsIUi", "nc", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_kshiftrisi, "UiUiIUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kshiftridi, "UOiUOiIUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kmovb, "UcUc", "nc", "avx512dq")
-TARGET_BUILTIN(__builtin_ia32_kmovw, "UsUs", "nc", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_kmovd, "UiUi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_kmovq, "UOiUOi", "nc", "avx512bw")
-TARGET_BUILTIN(__builtin_ia32_palignr512, "V64cV64cV64cIi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_dbpsadbw128, "V8sV16cV16cIi", "ncV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_dbpsadbw256, "V16sV32cV32cIi", "ncV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_dbpsadbw512, "V32sV64cV64cIi", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_psadbw512, "V8OiV64cV64c", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_compressdf512_mask, "V8dV8dV8dUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_compressdi512_mask, "V8OiV8OiV8OiUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_compresshi512_mask, "V32sV32sV32sUi", "ncV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_compressqi512_mask, "V64cV64cV64cUOi", "ncV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_compresssf512_mask, "V16fV16fV16fUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_compresssi512_mask, "V16iV16iV16iUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cmpsd_mask, "UcV2dV2dIiUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_cmpss_mask, "UcV4fV4fIiUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_pshufd512, "V16iV16iIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_expanddf512_mask, "V8dV8dV8dUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_expanddi512_mask, "V8OiV8OiV8OiUc", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_expandhi512_mask, "V32sV32sV32sUi", "ncV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_expandqi512_mask, "V64cV64cV64cUOi", "ncV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_expandloaddf512_mask, "V8dV8dC*V8dUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_expandloaddi512_mask, "V8OiV8OiC*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_expandloadhi512_mask, "V32sV32sC*V32sUi", "nV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_expandloadqi512_mask, "V64cV64cC*V64cUOi", "nV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_expandloadsf512_mask, "V16fV16fC*V16fUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_expandloadsi512_mask, "V16iV16iC*V16iUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_expandsf512_mask, "V16fV16fV16fUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_expandsi512_mask, "V16iV16iV16iUs", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtps2pd512_mask, "V8dV8fV8dUcIi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_compressstoredf512_mask, "vV8d*V8dUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_compressstoredi512_mask, "vV8Oi*V8OiUc", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_compressstorehi512_mask, "vV32s*V32sUi", "nV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_compressstoreqi512_mask, "vV64c*V64cUOi", "nV:512:", "avx512vbmi2,evex512")
-TARGET_BUILTIN(__builtin_ia32_compressstoresf512_mask, "vV16f*V16fUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_compressstoresi512_mask, "vV16i*V16iUs", "nV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2ps_mask, "V4fV8sV4fUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2ps256_mask, "V8fV8sV8fUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2ph_mask, "V8sV4fIiV8sUc", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2ph256_mask, "V8sV8fIiV8sUc", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtw2mask512, "UiV32s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtw2mask128, "UcV8s", "ncV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtw2mask256, "UsV16s", "ncV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtsd2ss_round_mask, "V4fV4fV2dV4fUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_cvtsi2ss32, "V4fV4fiIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_cvtss2sd_round_mask, "V2dV2dV4fV2dUcIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_cvtusi2ss32, "V4fV4fUiIi", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_vpmultishiftqb512, "V64cV64cV64c", "ncV:512:", "avx512vbmi,evex512")
-TARGET_BUILTIN(__builtin_ia32_vpmultishiftqb128, "V16cV16cV16c", "ncV:128:", "avx512vbmi,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vpmultishiftqb256, "V32cV32cV32c", "ncV:256:", "avx512vbmi,avx512vl")
-
-// bf16 intrinsics
-TARGET_BUILTIN(__builtin_ia32_cvtne2ps2bf16_128, "V8yV4fV4f", "ncV:128:", "avx512bf16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtne2ps2bf16_256, "V16yV8fV8f", "ncV:256:", "avx512bf16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtne2ps2bf16_512, "V32yV16fV16f", "ncV:512:", "avx512bf16,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtneps2bf16_128_mask, "V8yV4fV8yUc", "ncV:128:", "avx512bf16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtneps2bf16_256_mask, "V8yV8fV8yUc", "ncV:256:", "avx512bf16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cvtneps2bf16_512_mask, "V16yV16fV16yUs", "ncV:512:", "avx512bf16,evex512")
-TARGET_BUILTIN(__builtin_ia32_dpbf16ps_128, "V4fV4fV8yV8y", "ncV:128:", "avx512bf16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_dpbf16ps_256, "V8fV8fV16yV16y", "ncV:256:", "avx512bf16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_dpbf16ps_512, "V16fV16fV32yV32y", "ncV:512:", "avx512bf16,evex512")
-TARGET_BUILTIN(__builtin_ia32_cvtsbf162ss_32, "fy", "nc", "avx512bf16")
-
-TARGET_BUILTIN(__builtin_ia32_vp2intersect_q_512, "vV8OiV8OiUc*Uc*", "nV:512:", "avx512vp2intersect,evex512")
-TARGET_BUILTIN(__builtin_ia32_vp2intersect_q_256, "vV4OiV4OiUc*Uc*", "nV:256:", "avx512vp2intersect,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vp2intersect_q_128, "vV2OiV2OiUc*Uc*", "nV:128:", "avx512vp2intersect,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vp2intersect_d_512, "vV16iV16iUs*Us*", "nV:512:", "avx512vp2intersect,evex512")
-TARGET_BUILTIN(__builtin_ia32_vp2intersect_d_256, "vV8iV8iUc*Uc*", "nV:256:", "avx512vp2intersect,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vp2intersect_d_128, "vV4iV4iUc*Uc*", "nV:128:", "avx512vp2intersect,avx512vl")
-
-// AVX512 fp16 intrinsics
-TARGET_BUILTIN(__builtin_ia32_vcomish, "iV8xV8xIiIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_addph512, "V32xV32xV32xIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_subph512, "V32xV32xV32xIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_mulph512, "V32xV32xV32xIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_divph512, "V32xV32xV32xIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_maxph512, "V32xV32xV32xIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_minph512, "V32xV32xV32xIi", "ncV:512:", "avx512fp16,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_minph256, "V16xV16xV16x", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_minph128, "V8xV8xV8x", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_maxph256, "V16xV16xV16x", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_maxph128, "V8xV8xV8x", "ncV:128:", "avx512fp16,avx512vl")
-
-TARGET_BUILTIN(__builtin_ia32_addsh_round_mask, "V8xV8xV8xV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_divsh_round_mask, "V8xV8xV8xV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_mulsh_round_mask, "V8xV8xV8xV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_subsh_round_mask, "V8xV8xV8xV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_maxsh_round_mask, "V8xV8xV8xV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_minsh_round_mask, "V8xV8xV8xV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_cmpph512_mask, "UiV32xV32xIiUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_cmpph256_mask, "UsV16xV16xIiUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cmpph128_mask, "UcV8xV8xIiUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_cmpsh_mask, "UcV8xV8xIiUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_loadsh128_mask, "V8xV8xC*V8xUc", "nV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_storesh128_mask, "vV8x*V8xUc", "nV:128:", "avx512fp16")
-
-TARGET_BUILTIN(__builtin_ia32_rcpph128_mask, "V8xV8xV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rcpph256_mask, "V16xV16xV16xUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rcpph512_mask, "V32xV32xV32xUi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_rsqrtph128_mask, "V8xV8xV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rsqrtph256_mask, "V16xV16xV16xUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rsqrtph512_mask, "V32xV32xV32xUi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_getmantph128_mask, "V8xV8xIiV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_getmantph256_mask, "V16xV16xIiV16xUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_getmantph512_mask, "V32xV32xIiV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_getexpph128_mask, "V8xV8xV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_getexpph256_mask, "V16xV16xV16xUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_getexpph512_mask, "V32xV32xV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_scalefph128_mask, "V8xV8xV8xV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scalefph256_mask, "V16xV16xV16xV16xUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_scalefph512_mask, "V32xV32xV32xV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_rndscaleph_128_mask, "V8xV8xIiV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rndscaleph_256_mask, "V16xV16xIiV16xUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_rndscaleph_mask, "V32xV32xIiV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduceph128_mask, "V8xV8xIiV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_reduceph256_mask, "V16xV16xIiV16xUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_reduceph512_mask, "V32xV32xIiV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_rcpsh_mask, "V8xV8xV8xV8xUc", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_rsqrtsh_mask, "V8xV8xV8xV8xUc", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_getmantsh_round_mask, "V8xV8xV8xIiV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_getexpsh128_round_mask, "V8xV8xV8xV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_scalefsh_round_mask, "V8xV8xV8xV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_rndscalesh_round_mask, "V8xV8xV8xV8xUcIiIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_reducesh_mask, "V8xV8xV8xV8xUcIiIi", "ncV:128:", "avx512fp16")
-
-TARGET_BUILTIN(__builtin_ia32_sqrtph, "V8xV8x", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_sqrtph256, "V16xV16x", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_sqrtph512, "V32xV32xIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_sqrtsh_round_mask, "V8xV8xV8xV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_fpclassph128_mask, "UcV8xIiUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fpclassph256_mask, "UsV16xIiUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_fpclassph512_mask, "UiV32xIiUi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_fpclasssh_mask, "UcV8xIiUc", "ncV:128:", "avx512fp16")
-
-TARGET_BUILTIN(__builtin_ia32_vcvtpd2ph128_mask, "V8xV2dV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtpd2ph256_mask, "V8xV4dV8xUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtpd2ph512_mask, "V8xV8dV8xUcIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2pd128_mask, "V2dV8xV2dUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2pd256_mask, "V4dV8xV4dUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2pd512_mask, "V8dV8xV8dUcIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtsh2ss_round_mask, "V4fV4fV8xV4fUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vcvtss2sh_round_mask, "V8xV8xV4fV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vcvtsd2sh_round_mask, "V8xV8xV2dV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vcvtsh2sd_round_mask, "V2dV2dV8xV2dUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2w128_mask, "V8sV8xV8sUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2w256_mask, "V16sV16xV16sUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2w512_mask, "V32sV32xV32sUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2w128_mask, "V8sV8xV8sUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2w256_mask, "V16sV16xV16sUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2w512_mask, "V32sV32xV32sUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtw2ph128_mask, "V8xV8sV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtw2ph256_mask, "V16xV16sV16xUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtw2ph512_mask, "V32xV32sV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2uw128_mask, "V8UsV8xV8UsUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2uw256_mask, "V16UsV16xV16UsUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2uw512_mask, "V32UsV32xV32UsUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2uw128_mask, "V8UsV8xV8UsUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2uw256_mask, "V16UsV16xV16UsUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2uw512_mask, "V32UsV32xV32UsUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtuw2ph128_mask, "V8xV8UsV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtuw2ph256_mask, "V16xV16UsV16xUs", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtuw2ph512_mask, "V32xV32UsV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2dq128_mask, "V4iV8xV4iUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2dq256_mask, "V8iV8xV8iUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2dq512_mask, "V16iV16xV16iUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2udq128_mask, "V4UiV8xV4UiUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2udq256_mask, "V8UiV8xV8UiUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2udq512_mask, "V16UiV16xV16UiUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtdq2ph128_mask, "V8xV4iV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtdq2ph256_mask, "V8xV8iV8xUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtdq2ph512_mask, "V16xV16iV16xUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtudq2ph128_mask, "V8xV4UiV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtudq2ph256_mask, "V8xV8UiV8xUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtudq2ph512_mask, "V16xV16UiV16xUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2dq128_mask, "V4iV8xV4iUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2dq256_mask, "V8iV8xV8iUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2dq512_mask, "V16iV16xV16iUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2udq128_mask, "V4UiV8xV4UiUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2udq256_mask, "V8UiV8xV8UiUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2udq512_mask, "V16UiV16xV16UiUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtqq2ph128_mask, "V8xV2OiV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtqq2ph256_mask, "V8xV4OiV8xUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtqq2ph512_mask, "V8xV8OiV8xUcIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2qq128_mask, "V2OiV8xV2OiUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2qq256_mask, "V4OiV8xV4OiUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2qq512_mask, "V8OiV8xV8OiUcIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtuqq2ph128_mask, "V8xV2UOiV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtuqq2ph256_mask, "V8xV4UOiV8xUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtuqq2ph512_mask, "V8xV8UOiV8xUcIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2uqq128_mask, "V2UOiV8xV2UOiUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2uqq256_mask, "V4UOiV8xV4UOiUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2uqq512_mask, "V8UOiV8xV8UOiUcIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2qq128_mask, "V2OiV8xV2OiUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2qq256_mask, "V4OiV8xV4OiUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2qq512_mask, "V8OiV8xV8OiUcIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2uqq128_mask, "V2UOiV8xV2UOiUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2uqq256_mask, "V4UOiV8xV4UOiUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2uqq512_mask, "V8UOiV8xV8UOiUcIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtsh2si32, "iV8xIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vcvtsh2usi32, "UiV8xIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vcvtusi2sh, "V8xV8xUiIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vcvtsi2sh, "V8xV8xiIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vcvttsh2si32, "iV8xIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vcvttsh2usi32, "UiV8xIi", "ncV:128:", "avx512fp16")
-
-TARGET_BUILTIN(__builtin_ia32_vcvtph2psx128_mask, "V4fV8xV4fUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2psx256_mask, "V8fV8xV8fUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2psx512_mask, "V16fV16xV16fUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2phx128_mask, "V8xV4fV8xUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2phx256_mask, "V8xV8fV8xUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2phx512_mask, "V16xV16fV16xUsIi", "ncV:512:", "avx512fp16,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_vfmaddph, "V8xV8xV8xV8x", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfmaddph256, "V16xV16xV16xV16x", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfmaddph512_mask, "V32xV32xV32xV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddph512_mask3, "V32xV32xV32xV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddph512_maskz, "V32xV32xV32xV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubph, "V8xV8xV8xV8x", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubph256, "V16xV16xV16xV16x", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubph512_mask, "V32xV32xV32xV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubph512_maskz, "V32xV32xV32xV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubph512_mask3, "V32xV32xV32xV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_vfmsubaddph512_mask3, "V32xV32xV32xV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmsubph512_mask3, "V32xV32xV32xV32xUiIi", "ncV:512:", "avx512fp16,evex512")
-
-TARGET_BUILTIN(__builtin_ia32_vfmaddsh3_mask, "V8xV8xV8xV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsh3_maskz, "V8xV8xV8xV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsh3_mask3, "V8xV8xV8xV8xUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vfmsubsh3_mask3, "V8xV8xV8xV8xUcIi", "ncV:128:", "avx512fp16")
-
-TARGET_BUILTIN(__builtin_ia32_vfmaddcph128_mask, "V4fV4fV4fV4fUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfmaddcph128_maskz, "V4fV4fV4fV4fUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfmaddcph256_mask, "V8fV8fV8fV8fUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfmaddcph256_maskz, "V8fV8fV8fV8fUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfmaddcph512_mask, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddcph512_maskz, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddcph512_mask3, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcph128_mask, "V4fV4fV4fV4fUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcph128_maskz, "V4fV4fV4fV4fUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcph256_mask, "V8fV8fV8fV8fUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcph256_maskz, "V8fV8fV8fV8fUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcph512_mask, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcph512_maskz, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcph512_mask3, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddcsh_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vfmaddcsh_maskz, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcsh_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcsh_maskz, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vfmaddcsh_round_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vfmaddcsh_round_mask3, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcsh_round_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcsh_round_mask3, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512fp16")
-
-TARGET_BUILTIN(__builtin_ia32_vfmulcsh_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vfcmulcsh_mask, "V4fV4fV4fV4fUcIi", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_vfmulcph128_mask, "V4fV4fV4fV4fUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfmulcph256_mask, "V8fV8fV8fV8fUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfmulcph512_mask, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_vfcmulcph128_mask, "V4fV4fV4fV4fUc", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfcmulcph256_mask, "V8fV8fV8fV8fUc", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_vfcmulcph512_mask, "V16fV16fV16fV16fUsIi", "ncV:512:", "avx512fp16,evex512")
-
-// generic select intrinsics
-TARGET_BUILTIN(__builtin_ia32_selectb_128, "V16cUsV16cV16c", "ncV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectb_256, "V32cUiV32cV32c", "ncV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectb_512, "V64cUOiV64cV64c", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_selectw_128, "V8sUcV8sV8s", "ncV:128:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectw_256, "V16sUsV16sV16s", "ncV:256:", "avx512bw,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectw_512, "V32sUiV32sV32s", "ncV:512:", "avx512bw,evex512")
-TARGET_BUILTIN(__builtin_ia32_selectd_128, "V4iUcV4iV4i", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectd_256, "V8iUcV8iV8i", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectd_512, "V16iUsV16iV16i", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_selectph_128, "V8xUcV8xV8x", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectph_256, "V16xUsV16xV16x", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectph_512, "V32xUiV32xV32x", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_selectpbf_128, "V8yUcV8yV8y", "ncV:128:", "avx512bf16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectpbf_256, "V16yUsV16yV16y", "ncV:256:", "avx512bf16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectpbf_512, "V32yUiV32yV32y", "ncV:512:", "avx512bf16,evex512")
-TARGET_BUILTIN(__builtin_ia32_selectq_128, "V2OiUcV2OiV2Oi", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectq_256, "V4OiUcV4OiV4Oi", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectq_512, "V8OiUcV8OiV8Oi", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_selectps_128, "V4fUcV4fV4f", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectps_256, "V8fUcV8fV8f", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectps_512, "V16fUsV16fV16f", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_selectpd_128, "V2dUcV2dV2d", "ncV:128:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectpd_256, "V4dUcV4dV4d", "ncV:256:", "avx512vl")
-TARGET_BUILTIN(__builtin_ia32_selectpd_512, "V8dUcV8dV8d", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_selectsh_128, "V8xUcV8xV8x", "ncV:128:", "avx512fp16")
-TARGET_BUILTIN(__builtin_ia32_selectsbf_128, "V8yUcV8yV8y", "ncV:128:", "avx512bf16")
-TARGET_BUILTIN(__builtin_ia32_selectss_128, "V4fUcV4fV4f", "ncV:128:", "avx512f")
-TARGET_BUILTIN(__builtin_ia32_selectsd_128, "V2dUcV2dV2d", "ncV:128:", "avx512f")
-
-// generic reduction intrinsics
-TARGET_BUILTIN(__builtin_ia32_reduce_fadd_pd512, "ddV8d", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduce_fadd_ps512, "ffV16f", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduce_fadd_ph512, "xxV32x", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduce_fadd_ph256, "xxV16x", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_reduce_fadd_ph128, "xxV8x", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmax_pd512, "dV8d", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmax_ps512, "fV16f", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmax_ph512, "xV32x", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmax_ph256, "xV16x", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmax_ph128, "xV8x", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmin_pd512, "dV8d", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmin_ps512, "fV16f", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmin_ph512, "xV32x", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmin_ph256, "xV16x", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmin_ph128, "xV8x", "ncV:128:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmul_pd512, "ddV8d", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmul_ps512, "ffV16f", "ncV:512:", "avx512f,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmul_ph512, "xxV32x", "ncV:512:", "avx512fp16,evex512")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmul_ph256, "xxV16x", "ncV:256:", "avx512fp16,avx512vl")
-TARGET_BUILTIN(__builtin_ia32_reduce_fmul_ph128, "xxV8x", "ncV:128:", "avx512fp16,avx512vl")
-
-// MONITORX/MWAITX
-TARGET_BUILTIN(__builtin_ia32_monitorx, "vvC*UiUi", "n", "mwaitx")
-TARGET_BUILTIN(__builtin_ia32_mwaitx, "vUiUiUi", "n", "mwaitx")
-
-// WAITPKG
-TARGET_BUILTIN(__builtin_ia32_umonitor, "vvC*", "n", "waitpkg")
-TARGET_BUILTIN(__builtin_ia32_umwait, "UcUiUiUi", "n", "waitpkg")
-TARGET_BUILTIN(__builtin_ia32_tpause, "UcUiUiUi", "n", "waitpkg")
-
-// CLZERO
-TARGET_BUILTIN(__builtin_ia32_clzero, "vv*", "n", "clzero")
-
-// CLDEMOTE
-TARGET_BUILTIN(__builtin_ia32_cldemote, "vvC*", "n", "cldemote")
-
-// Direct Move
-TARGET_BUILTIN(__builtin_ia32_directstore_u32, "vUi*Ui", "n", "movdiri")
-TARGET_BUILTIN(__builtin_ia32_movdir64b, "vv*vC*", "n", "movdir64b")
-
-// PTWRITE
-TARGET_BUILTIN(__builtin_ia32_ptwrite32, "vUi", "n", "ptwrite")
-
-// INVPCID
-TARGET_BUILTIN(__builtin_ia32_invpcid, "vUiv*", "nc", "invpcid")
-
-// ENQCMD
-TARGET_BUILTIN(__builtin_ia32_enqcmd, "Ucv*vC*", "n", "enqcmd")
-TARGET_BUILTIN(__builtin_ia32_enqcmds, "Ucv*vC*", "n", "enqcmd")
-
-// KEY LOCKER
-TARGET_BUILTIN(__builtin_ia32_loadiwkey, "vV2OiV2OiV2OiUi", "nV:128:", "kl")
-TARGET_BUILTIN(__builtin_ia32_encodekey128_u32, "UiUiV2Oiv*", "nV:128:", "kl")
-TARGET_BUILTIN(__builtin_ia32_encodekey256_u32, "UiUiV2OiV2Oiv*", "nV:128:", "kl")
-TARGET_BUILTIN(__builtin_ia32_aesenc128kl_u8, "UcV2Oi*V2OivC*", "nV:128:", "kl")
-TARGET_BUILTIN(__builtin_ia32_aesenc256kl_u8, "UcV2Oi*V2OivC*", "nV:128:", "kl")
-TARGET_BUILTIN(__builtin_ia32_aesdec128kl_u8, "UcV2Oi*V2OivC*", "nV:128:", "kl")
-TARGET_BUILTIN(__builtin_ia32_aesdec256kl_u8, "UcV2Oi*V2OivC*", "nV:128:", "kl")
-TARGET_BUILTIN(__builtin_ia32_aesencwide128kl_u8, "UcV2Oi*V2OiC*vC*", "nV:128:", "kl,widekl")
-TARGET_BUILTIN(__builtin_ia32_aesencwide256kl_u8, "UcV2Oi*V2OiC*vC*", "nV:128:", "kl,widekl")
-TARGET_BUILTIN(__builtin_ia32_aesdecwide128kl_u8, "UcV2Oi*V2OiC*vC*", "nV:128:", "kl,widekl")
-TARGET_BUILTIN(__builtin_ia32_aesdecwide256kl_u8, "UcV2Oi*V2OiC*vC*", "nV:128:", "kl,widekl")
-
-// SERIALIZE
-TARGET_BUILTIN(__builtin_ia32_serialize, "v", "n", "serialize")
-
-// TSXLDTRK
-TARGET_BUILTIN(__builtin_ia32_xsusldtrk, "v", "n", "tsxldtrk")
-TARGET_BUILTIN(__builtin_ia32_xresldtrk, "v", "n", "tsxldtrk")
-
-// RAO-INT
-TARGET_BUILTIN(__builtin_ia32_aadd32, "vv*Si", "n", "raoint")
-TARGET_BUILTIN(__builtin_ia32_aand32, "vv*Si", "n", "raoint")
-TARGET_BUILTIN(__builtin_ia32_aor32, "vv*Si", "n", "raoint")
-TARGET_BUILTIN(__builtin_ia32_axor32, "vv*Si", "n", "raoint")
-
-// MSVC
-TARGET_HEADER_BUILTIN(_BitScanForward, "UcUNi*UNi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(_BitScanReverse, "UcUNi*UNi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-
-TARGET_HEADER_BUILTIN(_ReadWriteBarrier, "v", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(_ReadBarrier, "v", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(_WriteBarrier, "v", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-
-TARGET_HEADER_BUILTIN(__cpuid, "vi*i", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(__cpuidex, "vi*ii", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-
-TARGET_HEADER_BUILTIN(__emul, "LLiii", "nch", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(__emulu, "ULLiUiUi", "nch", INTRIN_H, ALL_MS_LANGUAGES, "")
-
-TARGET_HEADER_BUILTIN(_AddressOfReturnAddress, "v*", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-
-TARGET_HEADER_BUILTIN(__stosb, "vUc*Ucz", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(__int2c, "v", "nhr", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(__ud2, "v", "nhr", INTRIN_H, ALL_MS_LANGUAGES, "")
-
-TARGET_HEADER_BUILTIN(__readfsbyte, "UcUNi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(__readfsword, "UsUNi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(__readfsdword, "UNiUNi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(__readfsqword, "ULLiUNi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-
-TARGET_HEADER_BUILTIN(__readgsbyte, "UcUNi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(__readgsword, "UsUNi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(__readgsdword, "UNiUNi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(__readgsqword, "ULLiUNi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-
-// AVX10.2 VNNI FP16
-TARGET_BUILTIN(__builtin_ia32_vdpphps128, "V4fV4fV8xV8x", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vdpphps256, "V8fV8fV16xV16x", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vdpphps512, "V16fV16fV32xV32x", "ncV:512:", "avx10.2-512")
-
-// AVX10.2 VNNI INT8
-TARGET_BUILTIN(__builtin_ia32_vpdpbssd512, "V16iV16iV16iV16i", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vpdpbssds512, "V16iV16iV16iV16i", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vpdpbsud512, "V16iV16iV16iV16i", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vpdpbsuds512, "V16iV16iV16iV16i", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vpdpbuud512, "V16iV16iV16iV16i", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vpdpbuuds512, "V16iV16iV16iV16i", "ncV:512:", "avx10.2-512")
-
-// AVX10.2 VNNI INT16
-TARGET_BUILTIN(__builtin_ia32_vpdpwsud512, "V16iV16iV16iV16i", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vpdpwsuds512, "V16iV16iV16iV16i", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vpdpwusd512, "V16iV16iV16iV16i", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vpdpwusds512, "V16iV16iV16iV16i", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vpdpwuud512, "V16iV16iV16iV16i", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vpdpwuuds512, "V16iV16iV16iV16i", "nV:512:", "avx10.2-512")
-
-// AVX10.2 VMPSADBW
-TARGET_BUILTIN(__builtin_ia32_mpsadbw512, "V32sV64cV64cIc", "ncV:512:", "avx10.2-512")
-
-// AVX10.2 YMM Rounding
-TARGET_BUILTIN(__builtin_ia32_vaddpd256_round, "V4dV4dV4dIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vaddph256_round, "V16xV16xV16xIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vaddps256_round, "V8fV8fV8fIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcmppd256_round_mask, "UcV4dV4dIiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcmpph256_round_mask, "UsV16xV16xIiUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcmpps256_round_mask, "UcV8fV8fIiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtdq2ph256_round_mask, "V8xV8iV8xUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtdq2ps256_round_mask, "V8fV8iV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtpd2dq256_round_mask, "V4iV4dV4iUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtpd2ph256_round_mask, "V8xV4dV8xUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtpd2ps256_round_mask, "V4fV4dV4fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtpd2qq256_round_mask, "V4LLiV4dV4LLiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtpd2udq256_round_mask, "V4UiV4dV4UiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtpd2uqq256_round_mask, "V4ULLiV4dV4ULLiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2dq256_round_mask, "V8iV8xV8iUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2pd256_round_mask, "V4dV8xV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2psx256_round_mask, "V8fV8xV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2qq256_round_mask, "V4LLiV8xV4LLiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2udq256_round_mask, "V8UiV8xV8UiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2uqq256_round_mask, "V4ULLiV8xV4ULLiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2uw256_round_mask, "V16UsV16xV16UsUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2w256_round_mask, "V16sV16xV16sUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2dq256_round_mask, "V8iV8fV8iUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2pd256_round_mask, "V4dV4fV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2phx256_round_mask, "V8xV8fV8xUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2qq256_round_mask, "V4LLiV4fV4LLiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2udq256_round_mask, "V8UiV8fV8UiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2uqq256_round_mask, "V4ULLiV4fV4ULLiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtqq2pd256_round_mask, "V4dV4LLiV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtqq2ph256_round_mask, "V8xV4LLiV8xUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtqq2ps256_round_mask, "V4fV4LLiV4fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2dq256_round_mask, "V4iV4dV4iUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2qq256_round_mask, "V4LLiV4dV4LLiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2udq256_round_mask, "V4UiV4dV4UiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2uqq256_round_mask, "V4ULLiV4dV4ULLiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2dq256_round_mask, "V8iV8xV8iUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2qq256_round_mask, "V4LLiV8xV4LLiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2udq256_round_mask, "V8UiV8xV8UiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2uqq256_round_mask, "V4ULLiV8xV4ULLiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2uw256_round_mask, "V16UsV16xV16UsUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2w256_round_mask, "V16sV16xV16sUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2dq256_round_mask, "V8iV8fV8iUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2qq256_round_mask, "V4LLiV4fV4LLiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2udq256_round_mask, "V8UiV8fV8UiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2uqq256_round_mask, "V4ULLiV4fV4ULLiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtudq2ph256_round_mask, "V8xV8UiV8xUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtudq2ps256_round_mask, "V8fV8UiV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtuqq2pd256_round_mask, "V4dV4ULLiV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtuqq2ph256_round_mask, "V8xV4ULLiV8xUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtuqq2ps256_round_mask, "V4fV4ULLiV4fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtuw2ph256_round_mask, "V16xV16UsV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtw2ph256_round_mask, "V16xV16sV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vdivpd256_round, "V4dV4dV4dIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vdivph256_round, "V16xV16xV16xIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vdivps256_round, "V8fV8fV8fIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcph256_round_mask, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcph256_round_maskz, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfcmaddcph256_round_mask3, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfcmulcph256_round_mask, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfixupimmpd256_round_mask, "V4dV4dV4dV4LLiIiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfixupimmpd256_round_maskz, "V4dV4dV4dV4LLiIiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfixupimmps256_round_mask, "V8fV8fV8fV8iIiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfixupimmps256_round_maskz, "V8fV8fV8fV8iIiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddpd256_round_mask, "V4dV4dV4dV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddpd256_round_maskz, "V4dV4dV4dV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddpd256_round_mask3, "V4dV4dV4dV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddph256_round_mask, "V16xV16xV16xV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddph256_round_maskz, "V16xV16xV16xV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddph256_round_mask3, "V16xV16xV16xV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddps256_round_mask, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddps256_round_maskz, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddps256_round_mask3, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddcph256_round_mask, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddcph256_round_maskz, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddcph256_round_mask3, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubpd256_round_mask, "V4dV4dV4dV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubpd256_round_maskz, "V4dV4dV4dV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubpd256_round_mask3, "V4dV4dV4dV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubph256_round_mask, "V16xV16xV16xV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubph256_round_maskz, "V16xV16xV16xV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubph256_round_mask3, "V16xV16xV16xV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubps256_round_mask, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubps256_round_maskz, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddsubps256_round_mask3, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmsubpd256_round_mask3, "V4dV4dV4dV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmsubph256_round_mask3, "V16xV16xV16xV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmsubps256_round_mask3, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmsubaddpd256_round_mask3, "V4dV4dV4dV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmsubaddph256_round_mask3, "V16xV16xV16xV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmsubaddps256_round_mask3, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmulcph256_round_mask, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vgetexppd256_round_mask, "V4dV4dV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vgetexpph256_round_mask, "V16xV16xV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vgetexpps256_round_mask, "V8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vgetmantpd256_round_mask, "V4dV4dIiV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vgetmantph256_round_mask, "V16xV16xIiV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vgetmantps256_round_mask, "V8fV8fIiV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vmaxpd256_round, "V4dV4dV4dIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vmaxph256_round, "V16xV16xV16xIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vmaxps256_round, "V8fV8fV8fIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminpd256_round, "V4dV4dV4dIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminph256_round, "V16xV16xV16xIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminps256_round, "V8fV8fV8fIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vmulpd256_round, "V4dV4dV4dIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vmulph256_round, "V16xV16xV16xIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vmulps256_round, "V8fV8fV8fIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vrangepd256_round_mask, "V4dV4dV4dIiV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vrangeps256_round_mask, "V8fV8fV8fIiV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vreducepd256_round_mask, "V4dV4dIiV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vreduceph256_round_mask, "V16xV16xIiV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vreduceps256_round_mask, "V8fV8fIiV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vrndscalepd256_round_mask, "V4dV4dIiV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vrndscaleph256_round_mask, "V16xV16xIiV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vrndscaleps256_round_mask, "V8fV8fIiV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vscalefpd256_round_mask, "V4dV4dV4dV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vscalefph256_round_mask, "V16xV16xV16xV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vscalefps256_round_mask, "V8fV8fV8fV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vsqrtpd256_round, "V4dV4dIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vsqrtph256_round, "V16xV16xIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vsqrtps256_round, "V8fV8fIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vsubpd256_round, "V4dV4dV4dIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vsubph256_round, "V16xV16xV16xIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vsubps256_round, "V8fV8fV8fIi", "nV:256:", "avx10.2-256")
-
-// AVX-VNNI-INT16
-TARGET_BUILTIN(__builtin_ia32_vpdpwsud128, "V4iV4iV4iV4i", "nV:128:", "avxvnniint16|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpwsud256, "V8iV8iV8iV8i", "nV:256:", "avxvnniint16|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpwsuds128, "V4iV4iV4iV4i", "nV:128:", "avxvnniint16|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpwsuds256, "V8iV8iV8iV8i", "nV:256:", "avxvnniint16|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpwusd128, "V4iV4iV4iV4i", "nV:128:", "avxvnniint16|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpwusd256, "V8iV8iV8iV8i", "nV:256:", "avxvnniint16|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpwusds128, "V4iV4iV4iV4i", "nV:128:", "avxvnniint16|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpwusds256, "V8iV8iV8iV8i", "nV:256:", "avxvnniint16|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpwuud128, "V4iV4iV4iV4i", "nV:128:", "avxvnniint16|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpwuud256, "V8iV8iV8iV8i", "nV:256:", "avxvnniint16|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpwuuds128, "V4iV4iV4iV4i", "nV:128:", "avxvnniint16|avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vpdpwuuds256, "V8iV8iV8iV8i", "nV:256:", "avxvnniint16|avx10.2-256")
-
-// AVX10.2 SATCVT-DS
-TARGET_BUILTIN(__builtin_ia32_vcvttsd2sis32, "iV2dIi", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttsd2usis32, "UiV2dIi", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttss2sis32, "iV4fIi", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttss2usis32, "UiV4fIi", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2dqs128_mask, "V4iV2dV4iUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2dqs256_round_mask, "V4iV4dV4iUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2dqs512_round_mask, "V8iV8dV8iUcIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2udqs128_mask, "V4iV2dV4iUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2udqs256_round_mask, "V4iV4dV4iUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2udqs512_round_mask, "V8iV8dV8iUcIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2qqs128_mask, "V2OiV2dV2OiUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2qqs256_round_mask, "V4OiV4dV4OiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2qqs512_round_mask, "V8OiV8dV8OiUcIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2uqqs128_mask, "V2OiV2dV2OiUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2uqqs256_round_mask, "V4OiV4dV4OiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttpd2uqqs512_round_mask, "V8OiV8dV8OiUcIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2dqs128_mask, "V4iV4fV4iUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2dqs256_round_mask, "V8iV8fV8iUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2dqs512_round_mask, "V16iV16fV16iUsIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2udqs128_mask, "V4iV4fV4iUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2udqs256_round_mask, "V8iV8fV8iUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2udqs512_round_mask, "V16iV16fV16iUsIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2qqs128_mask, "V2OiV4fV2OiUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2qqs256_round_mask, "V4OiV4fV4OiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2qqs512_round_mask, "V8OiV8fV8OiUcIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2uqqs128_mask, "V2OiV4fV2OiUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2uqqs256_round_mask, "V4OiV4fV4OiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2uqqs512_round_mask, "V8OiV8fV8OiUcIi", "nV:512:", "avx10.2-512")
-
-// AVX-NE-CONVERT
-TARGET_BUILTIN(__builtin_ia32_vbcstnebf162ps128, "V4fyC*", "nV:128:", "avxneconvert")
-TARGET_BUILTIN(__builtin_ia32_vbcstnebf162ps256, "V8fyC*", "nV:256:", "avxneconvert")
-TARGET_BUILTIN(__builtin_ia32_vbcstnesh2ps128, "V4fxC*", "nV:128:", "avxneconvert")
-TARGET_BUILTIN(__builtin_ia32_vbcstnesh2ps256, "V8fxC*", "nV:256:", "avxneconvert")
-TARGET_BUILTIN(__builtin_ia32_vcvtneebf162ps128, "V4fV8yC*", "nV:128:", "avxneconvert")
-TARGET_BUILTIN(__builtin_ia32_vcvtneebf162ps256, "V8fV16yC*", "nV:256:", "avxneconvert")
-TARGET_BUILTIN(__builtin_ia32_vcvtneeph2ps128, "V4fV8xC*", "nV:128:", "avxneconvert")
-TARGET_BUILTIN(__builtin_ia32_vcvtneeph2ps256, "V8fV16xC*", "nV:256:", "avxneconvert")
-TARGET_BUILTIN(__builtin_ia32_vcvtneobf162ps128, "V4fV8yC*", "nV:128:", "avxneconvert")
-TARGET_BUILTIN(__builtin_ia32_vcvtneobf162ps256, "V8fV16yC*", "nV:256:", "avxneconvert")
-TARGET_BUILTIN(__builtin_ia32_vcvtneoph2ps128, "V4fV8xC*", "nV:128:", "avxneconvert")
-TARGET_BUILTIN(__builtin_ia32_vcvtneoph2ps256, "V8fV16xC*", "nV:256:", "avxneconvert")
-TARGET_BUILTIN(__builtin_ia32_vcvtneps2bf16128, "V8yV4f", "nV:128:", "avx512bf16,avx512vl|avxneconvert")
-TARGET_BUILTIN(__builtin_ia32_vcvtneps2bf16256, "V8yV8f", "nV:256:", "avx512bf16,avx512vl|avxneconvert")
-
-// SHA512
-TARGET_BUILTIN(__builtin_ia32_vsha512msg1, "V4ULLiV4ULLiV2ULLi", "nV:256:", "sha512")
-TARGET_BUILTIN(__builtin_ia32_vsha512msg2, "V4ULLiV4ULLiV4ULLi", "nV:256:", "sha512")
-TARGET_BUILTIN(__builtin_ia32_vsha512rnds2, "V4ULLiV4ULLiV4ULLiV2ULLi", "nV:256:", "sha512")
-
-TARGET_HEADER_BUILTIN(_InterlockedAnd64, "WiWiD*Wi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(_InterlockedDecrement64, "WiWiD*", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(_InterlockedExchange64, "WiWiD*Wi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(_InterlockedExchangeAdd64, "WiWiD*Wi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(_InterlockedExchangeSub64, "WiWiD*Wi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(_InterlockedIncrement64, "WiWiD*", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(_InterlockedOr64, "WiWiD*Wi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-TARGET_HEADER_BUILTIN(_InterlockedXor64, "WiWiD*Wi", "nh", INTRIN_H, ALL_MS_LANGUAGES, "")
-
-// SM3
-TARGET_BUILTIN(__builtin_ia32_vsm3msg1, "V4UiV4UiV4UiV4Ui", "nV:128:", "sm3")
-TARGET_BUILTIN(__builtin_ia32_vsm3msg2, "V4UiV4UiV4UiV4Ui", "nV:128:", "sm3")
-TARGET_BUILTIN(__builtin_ia32_vsm3rnds2, "V4UiV4UiV4UiV4UiIUi", "nV:128:", "sm3")
-
-// SM4
-TARGET_BUILTIN(__builtin_ia32_vsm4key4128, "V4UiV4UiV4Ui", "nV:128:", "sm4")
-TARGET_BUILTIN(__builtin_ia32_vsm4key4256, "V8UiV8UiV8Ui", "nV:256:", "sm4")
-TARGET_BUILTIN(__builtin_ia32_vsm4rnds4128, "V4UiV4UiV4Ui", "nV:128:", "sm4")
-TARGET_BUILTIN(__builtin_ia32_vsm4rnds4256, "V8UiV8UiV8Ui", "nV:256:", "sm4")
-
-// SM4_EVEX
-TARGET_BUILTIN(__builtin_ia32_vsm4key4512, "V16UiV16UiV16Ui", "nV:512:", "avx10.2-512,sm4")
-TARGET_BUILTIN(__builtin_ia32_vsm4rnds4512, "V16UiV16UiV16Ui", "nV:512:", "avx10.2-512,sm4")
-
-// AVX10 MINMAX
-TARGET_BUILTIN(__builtin_ia32_vminmaxnepbf16128, "V8yV8yV8yIi", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminmaxnepbf16256, "V16yV16yV16yIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminmaxnepbf16512, "V32yV32yV32yIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vminmaxpd128_mask, "V2dV2dV2dIiV2dUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminmaxpd256_round_mask, "V4dV4dV4dIiV4dUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminmaxpd512_round_mask, "V8dV8dV8dIiV8dUcIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vminmaxph128_mask, "V8xV8xV8xIiV8xUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminmaxph256_round_mask, "V16xV16xV16xIiV16xUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminmaxph512_round_mask, "V32xV32xV32xIiV32xUiIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vminmaxps128_mask, "V4fV4fV4fIiV4fUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminmaxps256_round_mask, "V8fV8fV8fIiV8fUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminmaxps512_round_mask, "V16fV16fV16fIiV16fUsIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vminmaxsd_round_mask, "V2dV2dV2dIiV2dUcIi", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminmaxsh_round_mask, "V8xV8xV8xIiV8xUcIi", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminmaxss_round_mask, "V4fV4fV4fIiV4fUcIi", "nV:128:", "avx10.2-256")
-
-// AVX10.2 SATCVT
-TARGET_BUILTIN(__builtin_ia32_vcvtnebf162ibs128, "V8UsV8y", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtnebf162ibs256, "V16UsV16y", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtnebf162ibs512, "V32UsV32y", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtnebf162iubs128, "V8UsV8y", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtnebf162iubs256, "V16UsV16y", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtnebf162iubs512, "V32UsV32y", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2ibs128_mask, "V8UsV8xV8UsUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2ibs256_mask, "V16UsV16xV16UsUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2ibs512_mask, "V32UsV32xV32UsUiIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2iubs128_mask, "V8UsV8xV8UsUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2iubs256_mask, "V16UsV16xV16UsUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtph2iubs512_mask, "V32UsV32xV32UsUiIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2ibs128_mask, "V4UiV4fV4UiUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2ibs256_mask, "V8UiV8fV8UiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2ibs512_mask, "V16UiV16fV16UiUsIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2iubs128_mask, "V4UiV4fV4UiUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2iubs256_mask, "V8UiV8fV8UiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtps2iubs512_mask, "V16UiV16fV16UiUsIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvttnebf162ibs128, "V8UsV8y", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttnebf162ibs256, "V16UsV16y", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttnebf162ibs512, "V32UsV32y", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvttnebf162iubs128, "V8UsV8y", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttnebf162iubs256, "V16UsV16y", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttnebf162iubs512, "V32UsV32y", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2ibs128_mask, "V8UsV8xV8UsUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2ibs256_mask, "V16UsV16xV16UsUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2ibs512_mask, "V32UsV32xV32UsUiIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2iubs128_mask, "V8UsV8xV8UsUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2iubs256_mask, "V16UsV16xV16UsUsIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttph2iubs512_mask, "V32UsV32xV32UsUiIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2ibs128_mask, "V4UiV4fV4UiUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2ibs256_mask, "V8UiV8fV8UiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2ibs512_mask, "V16UiV16fV16UiUsIi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2iubs128_mask, "V4UiV4fV4UiUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2iubs256_mask, "V8UiV8fV8UiUcIi", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvttps2iubs512_mask, "V16UiV16fV16UiUsIi", "nV:512:", "avx10.2-512")
-
-// AVX10.2 CONVERT
-TARGET_BUILTIN(__builtin_ia32_vcvt2ps2phx128_mask, "V8xV4fV4fV8xUc", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvt2ps2phx256_mask, "V16xV8fV8fV16xUsIi", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvt2ps2phx512_mask, "V32xV16fV16fV32xUiIi", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtbiasph2bf8_128_mask, "V16cV16cV8xV16cUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtbiasph2bf8_256_mask, "V16cV32cV16xV16cUs", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtbiasph2bf8_512_mask, "V32cV64cV32xV32cUi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtbiasph2bf8s_128_mask, "V16cV16cV8xV16cUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtbiasph2bf8s_256_mask, "V16cV32cV16xV16cUs", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtbiasph2bf8s_512_mask, "V32cV64cV32xV32cUi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtbiasph2hf8_128_mask, "V16cV16cV8xV16cUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtbiasph2hf8_256_mask, "V16cV32cV16xV16cUs", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtbiasph2hf8_512_mask, "V32cV64cV32xV32cUi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtbiasph2hf8s_128_mask, "V16cV16cV8xV16cUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtbiasph2hf8s_256_mask, "V16cV32cV16xV16cUs", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtbiasph2hf8s_512_mask, "V32cV64cV32xV32cUi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtne2ph2bf8_128, "V16cV8xV8x", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtne2ph2bf8_256, "V32cV16xV16x", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtne2ph2bf8_512, "V64cV32xV32x", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtne2ph2bf8s_128, "V16cV8xV8x", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtne2ph2bf8s_256, "V32cV16xV16x", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtne2ph2bf8s_512, "V64cV32xV32x", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtne2ph2hf8_128, "V16cV8xV8x", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtne2ph2hf8_256, "V32cV16xV16x", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtne2ph2hf8_512, "V64cV32xV32x", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtne2ph2hf8s_128, "V16cV8xV8x", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtne2ph2hf8s_256, "V32cV16xV16x", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtne2ph2hf8s_512, "V64cV32xV32x", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvthf8_2ph128_mask, "V8xV16cV8xUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvthf8_2ph256_mask, "V16xV16cV16xUs", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvthf8_2ph512_mask, "V32xV32cV32xUi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtneph2bf8_128_mask, "V16cV8xV16cUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtneph2bf8_256_mask, "V16cV16xV16cUs", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtneph2bf8_512_mask, "V32cV32xV32cUi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtneph2bf8s_128_mask, "V16cV8xV16cUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtneph2bf8s_256_mask, "V16cV16xV16cUs", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtneph2bf8s_512_mask, "V32cV32xV32cUi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtneph2hf8_128_mask, "V16cV8xV16cUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtneph2hf8_256_mask, "V16cV16xV16cUs", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtneph2hf8_512_mask, "V32cV32xV32cUi", "nV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcvtneph2hf8s_128_mask, "V16cV8xV16cUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtneph2hf8s_256_mask, "V16cV16xV16cUs", "nV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcvtneph2hf8s_512_mask, "V32cV32xV32cUi", "nV:512:", "avx10.2-512")
-
-// AVX10.2 BF16
-TARGET_BUILTIN(__builtin_ia32_loadsbf16128_mask, "V8yV8yC*V8yUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_storesbf16128_mask, "vV8y*V8yUc", "nV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vaddnepbf16128, "V8yV8yV8y", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vaddnepbf16256, "V16yV16yV16y", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vaddnepbf16512, "V32yV32yV32y", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vdivnepbf16128, "V8yV8yV8y", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vdivnepbf16256, "V16yV16yV16y", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vdivnepbf16512, "V32yV32yV32y", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vmaxpbf16128, "V8yV8yV8y", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vmaxpbf16256, "V16yV16yV16y", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vmaxpbf16512, "V32yV32yV32y", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vminpbf16128, "V8yV8yV8y", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminpbf16256, "V16yV16yV16y", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vminpbf16512, "V32yV32yV32y", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vmulnepbf16128, "V8yV8yV8y", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vmulnepbf16256, "V16yV16yV16y", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vmulnepbf16512, "V32yV32yV32y", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vsubnepbf16128, "V8yV8yV8y", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vsubnepbf16256, "V16yV16yV16y", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vsubnepbf16512, "V32yV32yV32y", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcomsbf16eq, "iV8yV8y", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcomsbf16lt, "iV8yV8y", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcomsbf16neq, "iV8yV8y", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcomsbf16ge, "iV8yV8y", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcomsbf16gt, "iV8yV8y", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcomsbf16le, "iV8yV8y", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcmppbf16512_mask,"UiV32yV32yIiUi", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vcmppbf16256_mask,"UsV16yV16yIiUs", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vcmppbf16128_mask,"UcV8yV8yIiUc", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfpclasspbf16128_mask, "UcV8yIiUc", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfpclasspbf16256_mask, "UsV16yIiUs", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfpclasspbf16512_mask, "UiV32yIiUi", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vscalefpbf16128_mask, "V8yV8yV8yV8yUc", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vscalefpbf16256_mask, "V16yV16yV16yV16yUs", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vscalefpbf16512_mask, "V32yV32yV32yV32yUi", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vrcppbf16128_mask, "V8yV8yV8yUc", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vrcppbf16256_mask, "V16yV16yV16yUs", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vrcppbf16512_mask, "V32yV32yV32yUi", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vgetexppbf16128_mask, "V8yV8yV8yUc", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vgetexppbf16256_mask, "V16yV16yV16yUs", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vgetexppbf16512_mask, "V32yV32yV32yUi", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vrsqrtpbf16128_mask, "V8yV8yV8yUc", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vrsqrtpbf16256_mask, "V16yV16yV16yUs", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vrsqrtpbf16512_mask, "V32yV32yV32yUi", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vreducenepbf16128_mask, "V8yV8yIiV8yUc", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vreducenepbf16256_mask, "V16yV16yIiV16yUs", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vreducenepbf16512_mask, "V32yV32yIiV32yUi", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vrndscalenepbf16_128_mask, "V8yV8yIiV8yUc", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vrndscalenepbf16_256_mask, "V16yV16yIiV16yUs", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vrndscalenepbf16_mask, "V32yV32yIiV32yUi", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vgetmantpbf16128_mask, "V8yV8yIiV8yUc", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vgetmantpbf16256_mask, "V16yV16yIiV16yUs", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vgetmantpbf16512_mask, "V32yV32yIiV32yUi", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vsqrtnepbf16, "V8yV8y", "ncV:128:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vsqrtnepbf16256, "V16yV16y", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vsqrtnepbf16512, "V32yV32y", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddnepbh512, "V32yV32yV32yV32y", "ncV:512:", "avx10.2-512")
-TARGET_BUILTIN(__builtin_ia32_vfmaddnepbh256, "V16yV16yV16yV16y", "ncV:256:", "avx10.2-256")
-TARGET_BUILTIN(__builtin_ia32_vfmaddnepbh128, "V8yV8yV8yV8y", "ncV:128:", "avx10.2-256")
-
-#undef BUILTIN
-#undef TARGET_BUILTIN
-#undef TARGET_HEADER_BUILTIN
diff --git a/clang/include/clang/Basic/BuiltinsX86.td b/clang/include/clang/Basic/BuiltinsX86.td
index cf8d2771310e3c..be2802f3908ff3 100644
--- a/clang/include/clang/Basic/BuiltinsX86.td
+++ b/clang/include/clang/Basic/BuiltinsX86.td
@@ -17,6 +17,21 @@ class X86Builtin<string prototype> : TargetBuiltin {
let Prototype = prototype;
}
+class X86NoPrefixBuiltin<string prototype> : TargetBuiltin {
+ let Spellings = [NAME];
+ let Prototype = prototype;
+}
+
+class X86LibBuiltin<string prototype> : TargetLibBuiltin {
+ let Spellings = [NAME];
+ let Prototype = prototype;
+}
+
+def rdpmc : X86Builtin<"unsigned long long int(int)">;
+def rdtsc : X86Builtin<"unsigned long long int()">;
+def __rdtsc : X86NoPrefixBuiltin<"unsigned long long int()">;
+def rdtscp : X86Builtin<"unsigned long long int(unsigned int*)">;
+
// Undefined Values
def undef128 : X86Builtin<"_Vector<2, double>()"> {
let Attributes = [Const, NoThrow, RequiredVectorWidth<128>];
@@ -135,3 +150,5375 @@ let Attributes = [Const, NoThrow, RequiredVectorWidth<256>], Features = "avx" in
def Op#ps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>)">;
}
}
+
+
+// Mechanically ported builtins from the original `.def` file.
+//
+// TODO: Build structured ways of synthesizing relevant groups and improve the
+// organization of the builtins below this point (and move them above it). The
+// current formulation is based on what was easiest to recognize from the
+// pre-TableGen version.
+
+let Features = "mmx", Attributes = [NoThrow, Const] in {
+ def _mm_prefetch : X86NoPrefixBuiltin<"void(char const *, int)">;
+}
+
+let Features = "sse", Attributes = [NoThrow] in {
+ def ldmxcsr : X86Builtin<"void(unsigned int)">;
+}
+
+let Features = "sse", Header = "xmmintrin.h", Attributes = [NoThrow, RequireDeclaration] in {
+ def _mm_setcsr : X86LibBuiltin<"void(unsigned int)">;
+}
+
+let Features = "sse", Attributes = [NoThrow] in {
+ def stmxcsr : X86Builtin<"unsigned int()">;
+}
+
+let Features = "sse", Header = "xmmintrin.h", Attributes = [NoThrow, RequireDeclaration] in {
+ def _mm_getcsr : X86LibBuiltin<"unsigned int()">;
+}
+
+let Features = "sse", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtss2si : X86Builtin<"int(_Vector<4, float>)">;
+ def cvttss2si : X86Builtin<"int(_Vector<4, float>)">;
+}
+
+let Features = "sse", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def movmskps : X86Builtin<"int(_Vector<4, float>)">;
+}
+
+let Features = "sse", Attributes = [NoThrow] in {
+ def sfence : X86Builtin<"void()">;
+}
+
+let Features = "sse", Header = "xmmintrin.h", Attributes = [NoThrow, RequireDeclaration] in {
+ def _mm_sfence : X86LibBuiltin<"void()">;
+}
+
+let Features = "sse", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rcpps : X86Builtin<"_Vector<4, float>(_Vector<4, float>)">;
+ def rcpss : X86Builtin<"_Vector<4, float>(_Vector<4, float>)">;
+ def rsqrtps : X86Builtin<"_Vector<4, float>(_Vector<4, float>)">;
+ def rsqrtss : X86Builtin<"_Vector<4, float>(_Vector<4, float>)">;
+ def sqrtps : X86Builtin<"_Vector<4, float>(_Vector<4, float>)">;
+ def sqrtss : X86Builtin<"_Vector<4, float>(_Vector<4, float>)">;
+ def shufps : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Constant int)">;
+}
+
+let Features = "sse2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def maskmovdqu : X86Builtin<"void(_Vector<16, char>, _Vector<16, char>, char *)">;
+}
+
+let Features = "sse2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def movmskpd : X86Builtin<"int(_Vector<2, double>)">;
+ def pmovmskb128 : X86Builtin<"int(_Vector<16, char>)">;
+}
+
+let Features = "sse2", Attributes = [NoThrow] in {
+ def movnti : X86Builtin<"void(int *, int)">;
+}
+
+let Features = "sse2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pshufd : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Constant int)">;
+ def pshuflw : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Constant int)">;
+ def pshufhw : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Constant int)">;
+ def psadbw128 : X86Builtin<"_Vector<2, long long int>(_Vector<16, char>, _Vector<16, char>)">;
+ def sqrtpd : X86Builtin<"_Vector<2, double>(_Vector<2, double>)">;
+ def sqrtsd : X86Builtin<"_Vector<2, double>(_Vector<2, double>)">;
+ def shufpd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Constant int)">;
+ def cvtpd2dq : X86Builtin<"_Vector<2, long long int>(_Vector<2, double>)">;
+ def cvtpd2ps : X86Builtin<"_Vector<4, float>(_Vector<2, double>)">;
+ def cvttpd2dq : X86Builtin<"_Vector<4, int>(_Vector<2, double>)">;
+ def cvtsd2si : X86Builtin<"int(_Vector<2, double>)">;
+ def cvttsd2si : X86Builtin<"int(_Vector<2, double>)">;
+ def cvtsd2ss : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<2, double>)">;
+ def cvtps2dq : X86Builtin<"_Vector<4, int>(_Vector<4, float>)">;
+ def cvttps2dq : X86Builtin<"_Vector<4, int>(_Vector<4, float>)">;
+}
+
+let Features = "sse2", Attributes = [NoThrow] in {
+ def clflush : X86Builtin<"void(void const *)">;
+}
+
+let Features = "sse2", Header = "emmintrin.h", Attributes = [NoThrow, RequireDeclaration] in {
+ def _mm_clflush : X86LibBuiltin<"void(void const *)">;
+}
+
+let Features = "sse2", Attributes = [NoThrow] in {
+ def lfence : X86Builtin<"void()">;
+}
+
+let Features = "sse2", Header = "emmintrin.h", Attributes = [NoThrow, RequireDeclaration] in {
+ def _mm_lfence : X86LibBuiltin<"void()">;
+}
+
+let Features = "sse2", Attributes = [NoThrow] in {
+ def mfence : X86Builtin<"void()">;
+}
+
+let Features = "sse2", Header = "emmintrin.h", Attributes = [NoThrow, RequireDeclaration] in {
+ def _mm_mfence : X86LibBuiltin<"void()">;
+}
+
+let Attributes = [NoThrow] in {
+ def pause : X86Builtin<"void()">;
+}
+
+let Header = "emmintrin.h", Attributes = [NoThrow, RequireDeclaration] in {
+ def _mm_pause : X86LibBuiltin<"void()">;
+}
+
+let Features = "sse2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmuludq128 : X86Builtin<"_Vector<2, long long int>(_Vector<4, int>, _Vector<4, int>)">;
+ def psraw128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>)">;
+ def psrad128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+ def psrlw128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>)">;
+ def psrld128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+ def psrlq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+ def psllw128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>)">;
+ def pslld128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+ def psllq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+ def psllwi128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>, int)">;
+ def pslldi128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, int)">;
+ def psllqi128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, int)">;
+ def psrlwi128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>, int)">;
+ def psrldi128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, int)">;
+ def psrlqi128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, int)">;
+ def psrawi128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>, int)">;
+ def psradi128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, int)">;
+ def pmaddwd128 : X86Builtin<"_Vector<4, int>(_Vector<8, short>, _Vector<8, short>)">;
+ def pslldqi128_byteshift : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Constant int)">;
+ def psrldqi128_byteshift : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Constant int)">;
+}
+
+let Features = "sse3", Attributes = [NoThrow] in {
+ def monitor : X86Builtin<"void(void const *, unsigned int, unsigned int)">;
+ def mwait : X86Builtin<"void(unsigned int, unsigned int)">;
+}
+
+let Features = "sse3", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def lddqu : X86Builtin<"_Vector<16, char>(char const *)">;
+}
+
+let Features = "ssse3", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def palignr128 : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>, _Constant int)">;
+}
+
+let Features = "sse4.1", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def insertps128 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Constant char)">;
+ def pblendvb128 : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>, _Vector<16, char>)">;
+ def pblendw128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>, _Constant int)">;
+ def blendpd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Constant int)">;
+ def blendps : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Constant int)">;
+ def blendvpd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>)">;
+ def blendvps : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>)">;
+ def packusdw128 : X86Builtin<"_Vector<8, short>(_Vector<4, int>, _Vector<4, int>)">;
+ def pmuldq128 : X86Builtin<"_Vector<2, long long int>(_Vector<4, int>, _Vector<4, int>)">;
+ def roundps : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Constant int)">;
+ def roundss : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Constant int)">;
+ def roundsd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Constant int)">;
+ def roundpd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Constant int)">;
+ def dpps : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Constant char)">;
+ def dppd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Constant char)">;
+ def ptestz128 : X86Builtin<"int(_Vector<2, long long int>, _Vector<2, long long int>)">;
+ def ptestc128 : X86Builtin<"int(_Vector<2, long long int>, _Vector<2, long long int>)">;
+ def ptestnzc128 : X86Builtin<"int(_Vector<2, long long int>, _Vector<2, long long int>)">;
+ def mpsadbw128 : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>, _Constant char)">;
+ def phminposuw128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>)">;
+ def vec_ext_v16qi : X86Builtin<"char(_Vector<16, char>, _Constant int)">;
+ def vec_set_v16qi : X86Builtin<"_Vector<16, char>(_Vector<16, char>, char, _Constant int)">;
+ def vec_set_v4si : X86Builtin<"_Vector<4, int>(_Vector<4, int>, int, _Constant int)">;
+}
+
+let Features = "sse4.2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pcmpistrm128 : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>, _Constant char)">;
+ def pcmpistri128 : X86Builtin<"int(_Vector<16, char>, _Vector<16, char>, _Constant char)">;
+ def pcmpestrm128 : X86Builtin<"_Vector<16, char>(_Vector<16, char>, int, _Vector<16, char>, int, _Constant char)">;
+ def pcmpestri128 : X86Builtin<"int(_Vector<16, char>, int, _Vector<16, char>, int, _Constant char)">;
+ def pcmpistria128 : X86Builtin<"int(_Vector<16, char>, _Vector<16, char>, _Constant char)">;
+ def pcmpistric128 : X86Builtin<"int(_Vector<16, char>, _Vector<16, char>, _Constant char)">;
+ def pcmpistrio128 : X86Builtin<"int(_Vector<16, char>, _Vector<16, char>, _Constant char)">;
+ def pcmpistris128 : X86Builtin<"int(_Vector<16, char>, _Vector<16, char>, _Constant char)">;
+ def pcmpistriz128 : X86Builtin<"int(_Vector<16, char>, _Vector<16, char>, _Constant char)">;
+ def pcmpestria128 : X86Builtin<"int(_Vector<16, char>, int, _Vector<16, char>, int, _Constant char)">;
+ def pcmpestric128 : X86Builtin<"int(_Vector<16, char>, int, _Vector<16, char>, int, _Constant char)">;
+ def pcmpestrio128 : X86Builtin<"int(_Vector<16, char>, int, _Vector<16, char>, int, _Constant char)">;
+ def pcmpestris128 : X86Builtin<"int(_Vector<16, char>, int, _Vector<16, char>, int, _Constant char)">;
+ def pcmpestriz128 : X86Builtin<"int(_Vector<16, char>, int, _Vector<16, char>, int, _Constant char)">;
+}
+
+let Features = "crc32", Attributes = [NoThrow, Const] in {
+ def crc32qi : X86Builtin<"unsigned int(unsigned int, unsigned char)">;
+ def crc32hi : X86Builtin<"unsigned int(unsigned int, unsigned short)">;
+ def crc32si : X86Builtin<"unsigned int(unsigned int, unsigned int)">;
+}
+
+let Features = "sse4a", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def extrqi : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Constant char, _Constant char)">;
+ def extrq : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<16, char>)">;
+ def insertqi : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Constant char, _Constant char)">;
+ def insertq : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "sse4a", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def movntsd : X86Builtin<"void(double *, _Vector<2, double>)">;
+ def movntss : X86Builtin<"void(float *, _Vector<4, float>)">;
+}
+
+let Features = "aes", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def aesenc128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+ def aesenclast128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+ def aesdec128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+ def aesdeclast128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+ def aesimc128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>)">;
+ def aeskeygenassist128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Constant char)">;
+}
+
+let Features = "vaes", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def aesenc256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512f,evex512,vaes", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def aesenc512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>)">;
+}
+
+let Features = "vaes", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def aesenclast256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512f,evex512,vaes", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def aesenclast512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>)">;
+}
+
+let Features = "vaes", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def aesdec256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512f,evex512,vaes", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def aesdec512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>)">;
+}
+
+let Features = "vaes", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def aesdeclast256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512f,evex512,vaes", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def aesdeclast512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>)">;
+}
+
+let Features = "gfni", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vgf2p8affineinvqb_v16qi : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>, _Constant char)">;
+}
+
+let Features = "avx,gfni", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vgf2p8affineinvqb_v32qi : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>, _Constant char)">;
+}
+
+let Features = "avx512f,evex512,gfni", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vgf2p8affineinvqb_v64qi : X86Builtin<"_Vector<64, char>(_Vector<64, char>, _Vector<64, char>, _Constant char)">;
+}
+
+let Features = "gfni", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vgf2p8affineqb_v16qi : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>, _Constant char)">;
+}
+
+let Features = "avx,gfni", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vgf2p8affineqb_v32qi : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>, _Constant char)">;
+}
+
+let Features = "avx512f,evex512,gfni", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vgf2p8affineqb_v64qi : X86Builtin<"_Vector<64, char>(_Vector<64, char>, _Vector<64, char>, _Constant char)">;
+}
+
+let Features = "gfni", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vgf2p8mulb_v16qi : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>)">;
+}
+
+let Features = "avx,gfni", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vgf2p8mulb_v32qi : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>)">;
+}
+
+let Features = "avx512f,evex512,gfni", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vgf2p8mulb_v64qi : X86Builtin<"_Vector<64, char>(_Vector<64, char>, _Vector<64, char>)">;
+}
+
+let Features = "pclmul", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pclmulqdq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Constant char)">;
+}
+
+let Features = "vpclmulqdq", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pclmulqdq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, _Constant char)">;
+}
+
+let Features = "avx512f,evex512,vpclmulqdq", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pclmulqdq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, _Constant char)">;
+}
+
+let Features = "avx", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpermilvarpd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, long long int>)">;
+ def vpermilvarps : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, int>)">;
+ def vpermilvarpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, long long int>)">;
+ def vpermilvarps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, int>)">;
+ def blendpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant int)">;
+ def blendps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant int)">;
+ def blendvpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, double>)">;
+ def blendvps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>)">;
+ def shufpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant int)">;
+ def shufps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant int)">;
+ def dpps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant char)">;
+ def cmppd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant char)">;
+ def cmpps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant char)">;
+ def vextractf128_pd256 : X86Builtin<"_Vector<2, double>(_Vector<4, double>, _Constant int)">;
+ def vextractf128_ps256 : X86Builtin<"_Vector<4, float>(_Vector<8, float>, _Constant int)">;
+ def vextractf128_si256 : X86Builtin<"_Vector<4, int>(_Vector<8, int>, _Constant int)">;
+ def cvtpd2ps256 : X86Builtin<"_Vector<4, float>(_Vector<4, double>)">;
+ def cvtps2dq256 : X86Builtin<"_Vector<8, int>(_Vector<8, float>)">;
+ def cvttpd2dq256 : X86Builtin<"_Vector<4, int>(_Vector<4, double>)">;
+ def cvtpd2dq256 : X86Builtin<"_Vector<4, int>(_Vector<4, double>)">;
+ def cvttps2dq256 : X86Builtin<"_Vector<8, int>(_Vector<8, float>)">;
+ def vperm2f128_pd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant int)">;
+ def vperm2f128_ps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant int)">;
+ def vperm2f128_si256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Constant int)">;
+}
+
+let Features = "avx", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpermilpd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Constant int)">;
+ def vpermilps : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Constant int)">;
+}
+
+let Features = "avx", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpermilpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Constant int)">;
+ def vpermilps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Constant int)">;
+ def vinsertf128_pd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<2, double>, _Constant int)">;
+ def vinsertf128_ps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<4, float>, _Constant int)">;
+ def vinsertf128_si256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<4, int>, _Constant int)">;
+ def sqrtpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>)">;
+ def sqrtps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>)">;
+ def rsqrtps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>)">;
+ def rcpps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>)">;
+ def roundpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Constant int)">;
+ def roundps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Constant int)">;
+}
+
+let Features = "avx", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vtestzpd : X86Builtin<"int(_Vector<2, double>, _Vector<2, double>)">;
+ def vtestcpd : X86Builtin<"int(_Vector<2, double>, _Vector<2, double>)">;
+ def vtestnzcpd : X86Builtin<"int(_Vector<2, double>, _Vector<2, double>)">;
+ def vtestzps : X86Builtin<"int(_Vector<4, float>, _Vector<4, float>)">;
+ def vtestcps : X86Builtin<"int(_Vector<4, float>, _Vector<4, float>)">;
+ def vtestnzcps : X86Builtin<"int(_Vector<4, float>, _Vector<4, float>)">;
+}
+
+let Features = "avx", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vtestzpd256 : X86Builtin<"int(_Vector<4, double>, _Vector<4, double>)">;
+ def vtestcpd256 : X86Builtin<"int(_Vector<4, double>, _Vector<4, double>)">;
+ def vtestnzcpd256 : X86Builtin<"int(_Vector<4, double>, _Vector<4, double>)">;
+ def vtestzps256 : X86Builtin<"int(_Vector<8, float>, _Vector<8, float>)">;
+ def vtestcps256 : X86Builtin<"int(_Vector<8, float>, _Vector<8, float>)">;
+ def vtestnzcps256 : X86Builtin<"int(_Vector<8, float>, _Vector<8, float>)">;
+ def ptestz256 : X86Builtin<"int(_Vector<4, long long int>, _Vector<4, long long int>)">;
+ def ptestc256 : X86Builtin<"int(_Vector<4, long long int>, _Vector<4, long long int>)">;
+ def ptestnzc256 : X86Builtin<"int(_Vector<4, long long int>, _Vector<4, long long int>)">;
+ def movmskpd256 : X86Builtin<"int(_Vector<4, double>)">;
+ def movmskps256 : X86Builtin<"int(_Vector<8, float>)">;
+}
+
+let Features = "avx", Attributes = [NoThrow] in {
+ def vzeroall : X86Builtin<"void()">;
+ def vzeroupper : X86Builtin<"void()">;
+}
+
+let Features = "avx", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def lddqu256 : X86Builtin<"_Vector<32, char>(char const *)">;
+}
+
+let Features = "avx", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def maskloadpd : X86Builtin<"_Vector<2, double>(_Vector<2, double const *>, _Vector<2, long long int>)">;
+ def maskloadps : X86Builtin<"_Vector<4, float>(_Vector<4, float const *>, _Vector<4, int>)">;
+}
+
+let Features = "avx", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def maskloadpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double const *>, _Vector<4, long long int>)">;
+ def maskloadps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float const *>, _Vector<8, int>)">;
+}
+
+let Features = "avx", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def maskstorepd : X86Builtin<"void(_Vector<2, double *>, _Vector<2, long long int>, _Vector<2, double>)">;
+ def maskstoreps : X86Builtin<"void(_Vector<4, float *>, _Vector<4, int>, _Vector<4, float>)">;
+}
+
+let Features = "avx", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def maskstorepd256 : X86Builtin<"void(_Vector<4, double *>, _Vector<4, long long int>, _Vector<4, double>)">;
+ def maskstoreps256 : X86Builtin<"void(_Vector<8, float *>, _Vector<8, int>, _Vector<8, float>)">;
+}
+
+let Features = "avx", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vec_ext_v32qi : X86Builtin<"char(_Vector<32, char>, _Constant int)">;
+ def vec_ext_v16hi : X86Builtin<"short(_Vector<16, short>, _Constant int)">;
+ def vec_ext_v8si : X86Builtin<"int(_Vector<8, int>, _Constant int)">;
+ def vec_set_v32qi : X86Builtin<"_Vector<32, char>(_Vector<32, char>, char, _Constant int)">;
+ def vec_set_v16hi : X86Builtin<"_Vector<16, short>(_Vector<16, short>, short, _Constant int)">;
+ def vec_set_v8si : X86Builtin<"_Vector<8, int>(_Vector<8, int>, int, _Constant int)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def mpsadbw256 : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>, _Constant char)">;
+ def packsswb256 : X86Builtin<"_Vector<32, char>(_Vector<16, short>, _Vector<16, short>)">;
+ def packssdw256 : X86Builtin<"_Vector<16, short>(_Vector<8, int>, _Vector<8, int>)">;
+ def packuswb256 : X86Builtin<"_Vector<32, char>(_Vector<16, short>, _Vector<16, short>)">;
+ def packusdw256 : X86Builtin<"_Vector<16, short>(_Vector<8, int>, _Vector<8, int>)">;
+ def palignr256 : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>, _Constant int)">;
+ def pavgb256 : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>)">;
+ def pavgw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">;
+ def pblendvb256 : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>, _Vector<32, char>)">;
+ def pblendw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>, _Constant int)">;
+ def phaddw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">;
+ def phaddd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>)">;
+ def phaddsw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">;
+ def phsubw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">;
+ def phsubd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>)">;
+ def phsubsw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">;
+ def pmaddubsw256 : X86Builtin<"_Vector<16, short>(_Vector<32, char>, _Vector<32, char>)">;
+ def pmaddwd256 : X86Builtin<"_Vector<8, int>(_Vector<16, short>, _Vector<16, short>)">;
+ def pmovmskb256 : X86Builtin<"int(_Vector<32, char>)">;
+ def pmuldq256 : X86Builtin<"_Vector<4, long long int>(_Vector<8, int>, _Vector<8, int>)">;
+ def pmulhrsw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">;
+ def pmulhuw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">;
+ def pmulhw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">;
+ def pmuludq256 : X86Builtin<"_Vector<4, long long int>(_Vector<8, int>, _Vector<8, int>)">;
+ def psadbw256 : X86Builtin<"_Vector<4, long long int>(_Vector<32, char>, _Vector<32, char>)">;
+ def pshufb256 : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>)">;
+ def pshufd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Constant int)">;
+ def pshuflw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Constant int)">;
+ def pshufhw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Constant int)">;
+ def psignb256 : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>)">;
+ def psignw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">;
+ def psignd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>)">;
+ def psllwi256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, int)">;
+ def psllw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<8, short>)">;
+ def pslldqi256_byteshift : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Constant int)">;
+ def pslldi256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, int)">;
+ def pslld256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<4, int>)">;
+ def psllqi256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, int)">;
+ def psllq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<2, long long int>)">;
+ def psrawi256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, int)">;
+ def psraw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<8, short>)">;
+ def psradi256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, int)">;
+ def psrad256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<4, int>)">;
+ def psrldqi256_byteshift : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Constant int)">;
+ def psrlwi256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, int)">;
+ def psrlw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<8, short>)">;
+ def psrldi256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, int)">;
+ def psrld256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<4, int>)">;
+ def psrlqi256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, int)">;
+ def psrlq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<2, long long int>)">;
+ def pblendd128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Constant int)">;
+ def pblendd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Constant int)">;
+ def permvarsi256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>)">;
+ def permdf256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Constant int)">;
+ def permvarsf256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, int>)">;
+ def permti256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, _Constant int)">;
+ def permdi256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Constant int)">;
+ def extract128i256 : X86Builtin<"_Vector<2, long long int>(_Vector<4, long long int>, _Constant int)">;
+ def insert128i256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<2, long long int>, _Constant int)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def maskloadd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int const *>, _Vector<8, int>)">;
+ def maskloadq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int const *>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def maskloadd : X86Builtin<"_Vector<4, int>(_Vector<4, int const *>, _Vector<4, int>)">;
+ def maskloadq : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int const *>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def maskstored256 : X86Builtin<"void(_Vector<8, int *>, _Vector<8, int>, _Vector<8, int>)">;
+ def maskstoreq256 : X86Builtin<"void(_Vector<4, long long int *>, _Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def maskstored : X86Builtin<"void(_Vector<4, int *>, _Vector<4, int>, _Vector<4, int>)">;
+ def maskstoreq : X86Builtin<"void(_Vector<2, long long int *>, _Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def psllv8si : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def psllv4si : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def psllv4di : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def psllv2di : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def psrav8si : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def psrav4si : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def psrlv8si : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def psrlv4si : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def psrlv4di : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def psrlv2di : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def gatherd_pd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, double const *, _Vector<4, int>, _Vector<2, double>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def gatherd_pd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, double const *, _Vector<4, int>, _Vector<4, double>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def gatherq_pd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, double const *, _Vector<2, long long int>, _Vector<2, double>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def gatherq_pd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, double const *, _Vector<4, long long int>, _Vector<4, double>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def gatherd_ps : X86Builtin<"_Vector<4, float>(_Vector<4, float>, float const *, _Vector<4, int>, _Vector<4, float>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def gatherd_ps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, float const *, _Vector<8, int>, _Vector<8, float>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def gatherq_ps : X86Builtin<"_Vector<4, float>(_Vector<4, float>, float const *, _Vector<2, long long int>, _Vector<4, float>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def gatherq_ps256 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, float const *, _Vector<4, long long int>, _Vector<4, float>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def gatherd_q : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, long long int const *, _Vector<4, int>, _Vector<2, long long int>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def gatherd_q256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, long long int const *, _Vector<4, int>, _Vector<4, long long int>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def gatherq_q : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, long long int const *, _Vector<2, long long int>, _Vector<2, long long int>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def gatherq_q256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, long long int const *, _Vector<4, long long int>, _Vector<4, long long int>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def gatherd_d : X86Builtin<"_Vector<4, int>(_Vector<4, int>, int const *, _Vector<4, int>, _Vector<4, int>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def gatherd_d256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, int const *, _Vector<8, int>, _Vector<8, int>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def gatherq_d : X86Builtin<"_Vector<4, int>(_Vector<4, int>, int const *, _Vector<2, long long int>, _Vector<4, int>, _Constant char)">;
+}
+
+let Features = "avx2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def gatherq_d256 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, int const *, _Vector<4, long long int>, _Vector<4, int>, _Constant char)">;
+}
+
+let Features = "f16c", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtps2ph : X86Builtin<"_Vector<8, short>(_Vector<4, float>, _Constant int)">;
+}
+
+let Features = "f16c", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtps2ph256 : X86Builtin<"_Vector<8, short>(_Vector<8, float>, _Constant int)">;
+}
+
+let Features = "f16c", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtph2ps : X86Builtin<"_Vector<4, float>(_Vector<8, short>)">;
+}
+
+let Features = "f16c", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtph2ps256 : X86Builtin<"_Vector<8, float>(_Vector<8, short>)">;
+}
+
+let Features = "rdrnd", Attributes = [NoThrow] in {
+ def rdrand16_step : X86Builtin<"unsigned int(unsigned short *)">;
+ def rdrand32_step : X86Builtin<"unsigned int(unsigned int *)">;
+}
+
+let Features = "fxsr", Attributes = [NoThrow] in {
+ def fxrstor : X86Builtin<"void(void *)">;
+ def fxsave : X86Builtin<"void(void *)">;
+}
+
+let Features = "xsave", Attributes = [NoThrow] in {
+ def xsave : X86Builtin<"void(void *, unsigned long long int)">;
+ def xrstor : X86Builtin<"void(void *, unsigned long long int)">;
+ def xgetbv : X86Builtin<"unsigned long long int(unsigned int)">;
+}
+
+let Header = "immintrin.h", Languages = "ALL_MS_LANGUAGES", Attributes = [NoThrow, RequireDeclaration] in {
+ def _xgetbv : X86LibBuiltin<"uint64_t(unsigned int)">;
+}
+
+let Features = "xsave", Attributes = [NoThrow] in {
+ def xsetbv : X86Builtin<"void(unsigned int, unsigned long long int)">;
+}
+
+let Header = "immintrin.h", Languages = "ALL_MS_LANGUAGES", Attributes = [NoThrow, RequireDeclaration] in {
+ def _xsetbv : X86LibBuiltin<"void(unsigned int, uint64_t)">;
+}
+
+let Features = "xsaveopt", Attributes = [NoThrow] in {
+ def xsaveopt : X86Builtin<"void(void *, unsigned long long int)">;
+}
+
+let Features = "xsaves", Attributes = [NoThrow] in {
+ def xrstors : X86Builtin<"void(void *, unsigned long long int)">;
+}
+
+let Features = "xsavec", Attributes = [NoThrow] in {
+ def xsavec : X86Builtin<"void(void *, unsigned long long int)">;
+}
+
+let Features = "xsaves", Attributes = [NoThrow] in {
+ def xsaves : X86Builtin<"void(void *, unsigned long long int)">;
+}
+
+let Features = "shstk", Attributes = [NoThrow] in {
+ def incsspd : X86Builtin<"void(unsigned int)">;
+ def rdsspd : X86Builtin<"unsigned int(unsigned int)">;
+ def saveprevssp : X86Builtin<"void()">;
+ def rstorssp : X86Builtin<"void(void *)">;
+ def wrssd : X86Builtin<"void(unsigned int, void *)">;
+ def wrussd : X86Builtin<"void(unsigned int, void *)">;
+ def setssbsy : X86Builtin<"void()">;
+ def clrssbsy : X86Builtin<"void(void *)">;
+}
+
+let Features = "clflushopt", Attributes = [NoThrow] in {
+ def clflushopt : X86Builtin<"void(void const *)">;
+}
+
+let Features = "clwb", Attributes = [NoThrow] in {
+ def clwb : X86Builtin<"void(void const *)">;
+}
+
+let Attributes = [NoThrow] in {
+ def wbinvd : X86Builtin<"void()">;
+}
+
+let Features = "wbnoinvd", Attributes = [NoThrow] in {
+ def wbnoinvd : X86Builtin<"void()">;
+}
+
+let Attributes = [NoThrow, Constexpr] in {
+ def addcarryx_u32 : X86Builtin<"unsigned char(unsigned char, unsigned int, unsigned int, unsigned int *)">;
+ def subborrow_u32 : X86Builtin<"unsigned char(unsigned char, unsigned int, unsigned int, unsigned int *)">;
+}
+
+let Features = "rdseed", Attributes = [NoThrow] in {
+ def rdseed16_step : X86Builtin<"unsigned int(unsigned short *)">;
+ def rdseed32_step : X86Builtin<"unsigned int(unsigned int *)">;
+}
+
+let Features = "lzcnt", Attributes = [NoThrow, Const, Constexpr] in {
+ def lzcnt_u16 : X86Builtin<"unsigned short(unsigned short)">;
+ def lzcnt_u32 : X86Builtin<"unsigned int(unsigned int)">;
+}
+
+let Features = "bmi", Attributes = [NoThrow, Const, Constexpr] in {
+ def bextr_u32 : X86Builtin<"unsigned int(unsigned int, unsigned int)">;
+}
+
+let Attributes = [NoThrow, Const, Constexpr] in {
+ def tzcnt_u16 : X86Builtin<"unsigned short(unsigned short)">;
+ def tzcnt_u32 : X86Builtin<"unsigned int(unsigned int)">;
+}
+
+let Features = "bmi2", Attributes = [NoThrow, Const, Constexpr] in {
+ def bzhi_si : X86Builtin<"unsigned int(unsigned int, unsigned int)">;
+ def pdep_si : X86Builtin<"unsigned int(unsigned int, unsigned int)">;
+ def pext_si : X86Builtin<"unsigned int(unsigned int, unsigned int)">;
+}
+
+let Features = "tbm", Attributes = [NoThrow, Const, Constexpr] in {
+ def bextri_u32 : X86Builtin<"unsigned int(unsigned int, _Constant unsigned int)">;
+}
+
+let Features = "lwp", Attributes = [NoThrow] in {
+ def llwpcb : X86Builtin<"void(void *)">;
+ def slwpcb : X86Builtin<"void *()">;
+ def lwpins32 : X86Builtin<"unsigned char(unsigned int, unsigned int, _Constant unsigned int)">;
+ def lwpval32 : X86Builtin<"void(unsigned int, unsigned int, _Constant unsigned int)">;
+}
+
+let Features = "sha", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def sha1rnds4 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Constant char)">;
+ def sha1nexte : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+ def sha1msg1 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+ def sha1msg2 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+ def sha256rnds2 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+ def sha256msg1 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+ def sha256msg2 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "fma|fma4", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfmaddps : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>)">;
+ def vfmaddpd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>)">;
+}
+
+let Features = "fma", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfmaddss3 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>)">;
+ def vfmaddsd3 : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>)">;
+}
+
+let Features = "fma4", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfmaddss : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>)">;
+ def vfmaddsd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>)">;
+}
+
+let Features = "fma|fma4", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfmaddsubps : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>)">;
+ def vfmaddsubpd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>)">;
+}
+
+let Features = "fma|fma4", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vfmaddps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>)">;
+ def vfmaddpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, double>)">;
+ def vfmaddsubps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>)">;
+ def vfmaddsubpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, double>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vfmaddpd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Vector<8, double>, unsigned char, _Constant int)">;
+ def vfmaddpd512_maskz : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Vector<8, double>, unsigned char, _Constant int)">;
+ def vfmaddpd512_mask3 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Vector<8, double>, unsigned char, _Constant int)">;
+ def vfmsubpd512_mask3 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Vector<8, double>, unsigned char, _Constant int)">;
+ def vfmaddps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def vfmaddps512_maskz : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def vfmaddps512_mask3 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def vfmsubps512_mask3 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def vfmaddsubpd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Vector<8, double>, unsigned char, _Constant int)">;
+ def vfmaddsubpd512_maskz : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Vector<8, double>, unsigned char, _Constant int)">;
+ def vfmaddsubpd512_mask3 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Vector<8, double>, unsigned char, _Constant int)">;
+ def vfmsubaddpd512_mask3 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Vector<8, double>, unsigned char, _Constant int)">;
+ def vfmaddsubps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def vfmaddsubps512_maskz : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def vfmaddsubps512_mask3 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def vfmsubaddps512_mask3 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+}
+
+let Features = "xop", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpmacssww : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>, _Vector<8, short>)">;
+ def vpmacsww : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>, _Vector<8, short>)">;
+ def vpmacsswd : X86Builtin<"_Vector<4, int>(_Vector<8, short>, _Vector<8, short>, _Vector<4, int>)">;
+ def vpmacswd : X86Builtin<"_Vector<4, int>(_Vector<8, short>, _Vector<8, short>, _Vector<4, int>)">;
+ def vpmacssdd : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+ def vpmacsdd : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+ def vpmacssdql : X86Builtin<"_Vector<2, long long int>(_Vector<4, int>, _Vector<4, int>, _Vector<2, long long int>)">;
+ def vpmacsdql : X86Builtin<"_Vector<2, long long int>(_Vector<4, int>, _Vector<4, int>, _Vector<2, long long int>)">;
+ def vpmacssdqh : X86Builtin<"_Vector<2, long long int>(_Vector<4, int>, _Vector<4, int>, _Vector<2, long long int>)">;
+ def vpmacsdqh : X86Builtin<"_Vector<2, long long int>(_Vector<4, int>, _Vector<4, int>, _Vector<2, long long int>)">;
+ def vpmadcsswd : X86Builtin<"_Vector<4, int>(_Vector<8, short>, _Vector<8, short>, _Vector<4, int>)">;
+ def vpmadcswd : X86Builtin<"_Vector<4, int>(_Vector<8, short>, _Vector<8, short>, _Vector<4, int>)">;
+ def vphaddbw : X86Builtin<"_Vector<8, short>(_Vector<16, char>)">;
+ def vphaddbd : X86Builtin<"_Vector<4, int>(_Vector<16, char>)">;
+ def vphaddbq : X86Builtin<"_Vector<2, long long int>(_Vector<16, char>)">;
+ def vphaddwd : X86Builtin<"_Vector<4, int>(_Vector<8, short>)">;
+ def vphaddwq : X86Builtin<"_Vector<2, long long int>(_Vector<8, short>)">;
+ def vphadddq : X86Builtin<"_Vector<2, long long int>(_Vector<4, int>)">;
+ def vphaddubw : X86Builtin<"_Vector<8, short>(_Vector<16, char>)">;
+ def vphaddubd : X86Builtin<"_Vector<4, int>(_Vector<16, char>)">;
+ def vphaddubq : X86Builtin<"_Vector<2, long long int>(_Vector<16, char>)">;
+ def vphadduwd : X86Builtin<"_Vector<4, int>(_Vector<8, short>)">;
+ def vphadduwq : X86Builtin<"_Vector<2, long long int>(_Vector<8, short>)">;
+ def vphaddudq : X86Builtin<"_Vector<2, long long int>(_Vector<4, int>)">;
+ def vphsubbw : X86Builtin<"_Vector<8, short>(_Vector<16, char>)">;
+ def vphsubwd : X86Builtin<"_Vector<4, int>(_Vector<8, short>)">;
+ def vphsubdq : X86Builtin<"_Vector<2, long long int>(_Vector<4, int>)">;
+ def vpperm : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>, _Vector<16, char>)">;
+ def vprotb : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>)">;
+ def vprotw : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>)">;
+ def vprotd : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+ def vprotq : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+ def vprotbi : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Constant char)">;
+ def vprotwi : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Constant char)">;
+ def vprotdi : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Constant char)">;
+ def vprotqi : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Constant char)">;
+ def vpshlb : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>)">;
+ def vpshlw : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>)">;
+ def vpshld : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+ def vpshlq : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+ def vpshab : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>)">;
+ def vpshaw : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>)">;
+ def vpshad : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+ def vpshaq : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+ def vpcomub : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>, _Constant char)">;
+ def vpcomuw : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>, _Constant char)">;
+ def vpcomud : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Constant char)">;
+ def vpcomuq : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Constant char)">;
+ def vpcomb : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>, _Constant char)">;
+ def vpcomw : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>, _Constant char)">;
+ def vpcomd : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Constant char)">;
+ def vpcomq : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Constant char)">;
+ def vpermil2pd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, long long int>, _Constant char)">;
+}
+
+let Features = "xop", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpermil2pd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, long long int>, _Constant char)">;
+}
+
+let Features = "xop", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpermil2ps : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, int>, _Constant char)">;
+}
+
+let Features = "xop", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpermil2ps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, int>, _Constant char)">;
+}
+
+let Features = "xop", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfrczss : X86Builtin<"_Vector<4, float>(_Vector<4, float>)">;
+ def vfrczsd : X86Builtin<"_Vector<2, double>(_Vector<2, double>)">;
+ def vfrczps : X86Builtin<"_Vector<4, float>(_Vector<4, float>)">;
+ def vfrczpd : X86Builtin<"_Vector<2, double>(_Vector<2, double>)">;
+}
+
+let Features = "xop", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vfrczps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>)">;
+ def vfrczpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>)">;
+}
+
+let Features = "rtm", Attributes = [NoThrow] in {
+ def xbegin : X86Builtin<"int()">;
+ def xend : X86Builtin<"void()">;
+ def xabort : X86Builtin<"void(_Constant char)">;
+ def xtest : X86Builtin<"int()">;
+}
+
+let Features = "rdpid", Attributes = [NoThrow] in {
+ def rdpid : X86Builtin<"unsigned int()">;
+}
+
+let Features = "rdpru", Attributes = [NoThrow] in {
+ def rdpru : X86Builtin<"unsigned long long int(int)">;
+}
+
+let Features = "pku", Attributes = [NoThrow] in {
+ def rdpkru : X86Builtin<"unsigned int()">;
+ def wrpkru : X86Builtin<"void(unsigned int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def sqrtpd512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Constant int)">;
+ def sqrtps512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Constant int)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rsqrt14sd_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char)">;
+ def rsqrt14ss_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def rsqrt14pd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, unsigned char)">;
+ def rsqrt14ps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, unsigned short)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rcp14sd_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char)">;
+ def rcp14ss_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def rcp14pd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, unsigned char)">;
+ def rcp14ps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, unsigned short)">;
+ def cvttps2dq512_mask : X86Builtin<"_Vector<16, int>(_Vector<16, float>, _Vector<16, int>, unsigned short, _Constant int)">;
+ def cvttps2udq512_mask : X86Builtin<"_Vector<16, int>(_Vector<16, float>, _Vector<16, int>, unsigned short, _Constant int)">;
+ def cvttpd2dq512_mask : X86Builtin<"_Vector<8, int>(_Vector<8, double>, _Vector<8, int>, unsigned char, _Constant int)">;
+ def cvttpd2udq512_mask : X86Builtin<"_Vector<8, int>(_Vector<8, double>, _Vector<8, int>, unsigned char, _Constant int)">;
+ def cmpps512_mask : X86Builtin<"unsigned short(_Vector<16, float>, _Vector<16, float>, _Constant int, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cmpps256_mask : X86Builtin<"unsigned char(_Vector<8, float>, _Vector<8, float>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cmpps128_mask : X86Builtin<"unsigned char(_Vector<4, float>, _Vector<4, float>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def cmppd512_mask : X86Builtin<"unsigned char(_Vector<8, double>, _Vector<8, double>, _Constant int, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cmppd256_mask : X86Builtin<"unsigned char(_Vector<4, double>, _Vector<4, double>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cmppd128_mask : X86Builtin<"unsigned char(_Vector<2, double>, _Vector<2, double>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def rndscaleps_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Constant int, _Vector<16, float>, unsigned short, _Constant int)">;
+ def rndscalepd_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Constant int, _Vector<8, double>, unsigned char, _Constant int)">;
+ def cvtps2dq512_mask : X86Builtin<"_Vector<16, int>(_Vector<16, float>, _Vector<16, int>, unsigned short, _Constant int)">;
+ def cvtpd2dq512_mask : X86Builtin<"_Vector<8, int>(_Vector<8, double>, _Vector<8, int>, unsigned char, _Constant int)">;
+ def cvtps2udq512_mask : X86Builtin<"_Vector<16, int>(_Vector<16, float>, _Vector<16, int>, unsigned short, _Constant int)">;
+ def cvtpd2udq512_mask : X86Builtin<"_Vector<8, int>(_Vector<8, double>, _Vector<8, int>, unsigned char, _Constant int)">;
+ def minps512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Constant int)">;
+ def minpd512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Constant int)">;
+ def maxps512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Constant int)">;
+ def maxpd512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Constant int)">;
+ def cvtdq2ps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, int>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def cvtudq2ps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, int>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def cvtpd2ps512_mask : X86Builtin<"_Vector<8, float>(_Vector<8, double>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vcvtps2ph512_mask : X86Builtin<"_Vector<16, short>(_Vector<16, float>, _Constant int, _Vector<16, short>, unsigned short)">;
+ def vcvtph2ps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, short>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def pmuldq512 : X86Builtin<"_Vector<8, long long int>(_Vector<16, int>, _Vector<16, int>)">;
+ def pmuludq512 : X86Builtin<"_Vector<8, long long int>(_Vector<16, int>, _Vector<16, int>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def loaddqusi512_mask : X86Builtin<"_Vector<16, int>(int const *, _Vector<16, int>, unsigned short)">;
+ def loaddqudi512_mask : X86Builtin<"_Vector<8, long long int>(long long int const *, _Vector<8, long long int>, unsigned char)">;
+ def loadups512_mask : X86Builtin<"_Vector<16, float>(float const *, _Vector<16, float>, unsigned short)">;
+ def loadaps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float const *>, _Vector<16, float>, unsigned short)">;
+ def loadupd512_mask : X86Builtin<"_Vector<8, double>(double const *, _Vector<8, double>, unsigned char)">;
+ def loadapd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double const *>, _Vector<8, double>, unsigned char)">;
+ def storedqudi512_mask : X86Builtin<"void(long long int *, _Vector<8, long long int>, unsigned char)">;
+ def storedqusi512_mask : X86Builtin<"void(int *, _Vector<16, int>, unsigned short)">;
+ def storeupd512_mask : X86Builtin<"void(double *, _Vector<8, double>, unsigned char)">;
+ def storeapd512_mask : X86Builtin<"void(_Vector<8, double *>, _Vector<8, double>, unsigned char)">;
+ def storeups512_mask : X86Builtin<"void(float *, _Vector<16, float>, unsigned short)">;
+ def storeaps512_mask : X86Builtin<"void(_Vector<16, float *>, _Vector<16, float>, unsigned short)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def alignq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, _Constant int)">;
+ def alignd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def alignd128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def alignd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def alignq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def alignq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, _Constant int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def extractf64x4_mask : X86Builtin<"_Vector<4, double>(_Vector<8, double>, _Constant int, _Vector<4, double>, unsigned char)">;
+ def extractf32x4_mask : X86Builtin<"_Vector<4, float>(_Vector<16, float>, _Constant int, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512vnni|avxvnni", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpdpbusd128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avx512vl,avx512vnni|avxvnni", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpdpbusd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx512vnni,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpdpbusd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+}
+
+let Features = "avx512vl,avx512vnni|avxvnni", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpdpbusds128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avx512vl,avx512vnni|avxvnni", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpdpbusds256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx512vnni,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpdpbusds512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+}
+
+let Features = "avx512vl,avx512vnni|avxvnni", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpdpwssd128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avx512vl,avx512vnni|avxvnni", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpdpwssd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx512vnni,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpdpwssd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+}
+
+let Features = "avx512vl,avx512vnni|avxvnni", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpdpwssds128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avx512vl,avx512vnni|avxvnni", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpdpwssds256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx512vnni,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpdpwssds512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+}
+
+let Features = "avxvnniint8|avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpdpbssd128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avxvnniint8|avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpdpbssd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avxvnniint8|avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpdpbssds128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avxvnniint8|avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpdpbssds256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avxvnniint8|avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpdpbsud128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avxvnniint8|avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpdpbsud256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avxvnniint8|avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpdpbsuds128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avxvnniint8|avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpdpbsuds256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avxvnniint8|avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpdpbuud128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avxvnniint8|avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpdpbuud256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avxvnniint8|avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpdpbuuds128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avxvnniint8|avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpdpbuuds256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "movrs", Attributes = [NoThrow, Const] in {
+ def prefetchrs : X86Builtin<"void(void const *)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def gather3div2df : X86Builtin<"_Vector<2, double>(_Vector<2, double>, void const *, _Vector<2, long long int>, unsigned char, _Constant int)">;
+ def gather3div2di : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, void const *, _Vector<2, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def gather3div4df : X86Builtin<"_Vector<4, double>(_Vector<4, double>, void const *, _Vector<4, long long int>, unsigned char, _Constant int)">;
+ def gather3div4di : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, void const *, _Vector<4, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def gather3div4sf : X86Builtin<"_Vector<4, float>(_Vector<4, float>, void const *, _Vector<2, long long int>, unsigned char, _Constant int)">;
+ def gather3div4si : X86Builtin<"_Vector<4, int>(_Vector<4, int>, void const *, _Vector<2, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def gather3div8sf : X86Builtin<"_Vector<4, float>(_Vector<4, float>, void const *, _Vector<4, long long int>, unsigned char, _Constant int)">;
+ def gather3div8si : X86Builtin<"_Vector<4, int>(_Vector<4, int>, void const *, _Vector<4, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def gather3siv2df : X86Builtin<"_Vector<2, double>(_Vector<2, double>, void const *, _Vector<4, int>, unsigned char, _Constant int)">;
+ def gather3siv2di : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, void const *, _Vector<4, int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def gather3siv4df : X86Builtin<"_Vector<4, double>(_Vector<4, double>, void const *, _Vector<4, int>, unsigned char, _Constant int)">;
+ def gather3siv4di : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, void const *, _Vector<4, int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def gather3siv4sf : X86Builtin<"_Vector<4, float>(_Vector<4, float>, void const *, _Vector<4, int>, unsigned char, _Constant int)">;
+ def gather3siv4si : X86Builtin<"_Vector<4, int>(_Vector<4, int>, void const *, _Vector<4, int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def gather3siv8sf : X86Builtin<"_Vector<8, float>(_Vector<8, float>, void const *, _Vector<8, int>, unsigned char, _Constant int)">;
+ def gather3siv8si : X86Builtin<"_Vector<8, int>(_Vector<8, int>, void const *, _Vector<8, int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def gathersiv8df : X86Builtin<"_Vector<8, double>(_Vector<8, double>, void const *, _Vector<8, int>, unsigned char, _Constant int)">;
+ def gathersiv16sf : X86Builtin<"_Vector<16, float>(_Vector<16, float>, void const *, _Vector<16, int>, unsigned short, _Constant int)">;
+ def gatherdiv8df : X86Builtin<"_Vector<8, double>(_Vector<8, double>, void const *, _Vector<8, long long int>, unsigned char, _Constant int)">;
+ def gatherdiv16sf : X86Builtin<"_Vector<8, float>(_Vector<8, float>, void const *, _Vector<8, long long int>, unsigned char, _Constant int)">;
+ def gathersiv8di : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, void const *, _Vector<8, int>, unsigned char, _Constant int)">;
+ def gathersiv16si : X86Builtin<"_Vector<16, int>(_Vector<16, int>, void const *, _Vector<16, int>, unsigned short, _Constant int)">;
+ def gatherdiv8di : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, void const *, _Vector<8, long long int>, unsigned char, _Constant int)">;
+ def gatherdiv16si : X86Builtin<"_Vector<8, int>(_Vector<8, int>, void const *, _Vector<8, long long int>, unsigned char, _Constant int)">;
+ def scattersiv8df : X86Builtin<"void(void *, unsigned char, _Vector<8, int>, _Vector<8, double>, _Constant int)">;
+ def scattersiv16sf : X86Builtin<"void(void *, unsigned short, _Vector<16, int>, _Vector<16, float>, _Constant int)">;
+ def scatterdiv8df : X86Builtin<"void(void *, unsigned char, _Vector<8, long long int>, _Vector<8, double>, _Constant int)">;
+ def scatterdiv16sf : X86Builtin<"void(void *, unsigned char, _Vector<8, long long int>, _Vector<8, float>, _Constant int)">;
+ def scattersiv8di : X86Builtin<"void(void *, unsigned char, _Vector<8, int>, _Vector<8, long long int>, _Constant int)">;
+ def scattersiv16si : X86Builtin<"void(void *, unsigned short, _Vector<16, int>, _Vector<16, int>, _Constant int)">;
+ def scatterdiv8di : X86Builtin<"void(void *, unsigned char, _Vector<8, long long int>, _Vector<8, long long int>, _Constant int)">;
+ def scatterdiv16si : X86Builtin<"void(void *, unsigned char, _Vector<8, long long int>, _Vector<8, int>, _Constant int)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const] in {
+ def knotqi : X86Builtin<"unsigned char(unsigned char)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const] in {
+ def knothi : X86Builtin<"unsigned short(unsigned short)">;
+}
+
+let Features = "avx512bw", Attributes = [NoThrow, Const] in {
+ def knotsi : X86Builtin<"unsigned int(unsigned int)">;
+ def knotdi : X86Builtin<"unsigned long long int(unsigned long long int)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cmpb128_mask : X86Builtin<"unsigned short(_Vector<16, char>, _Vector<16, char>, _Constant int, unsigned short)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cmpd128_mask : X86Builtin<"unsigned char(_Vector<4, int>, _Vector<4, int>, _Constant int, unsigned char)">;
+ def cmpq128_mask : X86Builtin<"unsigned char(_Vector<2, long long int>, _Vector<2, long long int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cmpw128_mask : X86Builtin<"unsigned char(_Vector<8, short>, _Vector<8, short>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cmpb256_mask : X86Builtin<"unsigned int(_Vector<32, char>, _Vector<32, char>, _Constant int, unsigned int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cmpd256_mask : X86Builtin<"unsigned char(_Vector<8, int>, _Vector<8, int>, _Constant int, unsigned char)">;
+ def cmpq256_mask : X86Builtin<"unsigned char(_Vector<4, long long int>, _Vector<4, long long int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cmpw256_mask : X86Builtin<"unsigned short(_Vector<16, short>, _Vector<16, short>, _Constant int, unsigned short)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def cmpb512_mask : X86Builtin<"unsigned long long int(_Vector<64, char>, _Vector<64, char>, _Constant int, unsigned long long int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def cmpd512_mask : X86Builtin<"unsigned short(_Vector<16, int>, _Vector<16, int>, _Constant int, unsigned short)">;
+ def cmpq512_mask : X86Builtin<"unsigned char(_Vector<8, long long int>, _Vector<8, long long int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def cmpw512_mask : X86Builtin<"unsigned int(_Vector<32, short>, _Vector<32, short>, _Constant int, unsigned int)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def ucmpb128_mask : X86Builtin<"unsigned short(_Vector<16, char>, _Vector<16, char>, _Constant int, unsigned short)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def ucmpd128_mask : X86Builtin<"unsigned char(_Vector<4, int>, _Vector<4, int>, _Constant int, unsigned char)">;
+ def ucmpq128_mask : X86Builtin<"unsigned char(_Vector<2, long long int>, _Vector<2, long long int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def ucmpw128_mask : X86Builtin<"unsigned char(_Vector<8, short>, _Vector<8, short>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def ucmpb256_mask : X86Builtin<"unsigned int(_Vector<32, char>, _Vector<32, char>, _Constant int, unsigned int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def ucmpd256_mask : X86Builtin<"unsigned char(_Vector<8, int>, _Vector<8, int>, _Constant int, unsigned char)">;
+ def ucmpq256_mask : X86Builtin<"unsigned char(_Vector<4, long long int>, _Vector<4, long long int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def ucmpw256_mask : X86Builtin<"unsigned short(_Vector<16, short>, _Vector<16, short>, _Constant int, unsigned short)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def ucmpb512_mask : X86Builtin<"unsigned long long int(_Vector<64, char>, _Vector<64, char>, _Constant int, unsigned long long int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def ucmpd512_mask : X86Builtin<"unsigned short(_Vector<16, int>, _Vector<16, int>, _Constant int, unsigned short)">;
+ def ucmpq512_mask : X86Builtin<"unsigned char(_Vector<8, long long int>, _Vector<8, long long int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def ucmpw512_mask : X86Builtin<"unsigned int(_Vector<32, short>, _Vector<32, short>, _Constant int, unsigned int)">;
+ def packssdw512 : X86Builtin<"_Vector<32, short>(_Vector<16, int>, _Vector<16, int>)">;
+ def packsswb512 : X86Builtin<"_Vector<64, char>(_Vector<32, short>, _Vector<32, short>)">;
+ def packusdw512 : X86Builtin<"_Vector<32, short>(_Vector<16, int>, _Vector<16, int>)">;
+ def packuswb512 : X86Builtin<"_Vector<64, char>(_Vector<32, short>, _Vector<32, short>)">;
+ def pavgb512 : X86Builtin<"_Vector<64, char>(_Vector<64, char>, _Vector<64, char>)">;
+ def pavgw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>)">;
+ def pshufb512 : X86Builtin<"_Vector<64, char>(_Vector<64, char>, _Vector<64, char>)">;
+}
+
+let Features = "avx512cd,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpconflictdi_128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>)">;
+}
+
+let Features = "avx512cd,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpconflictdi_256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>)">;
+}
+
+let Features = "avx512cd,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpconflictsi_128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>)">;
+}
+
+let Features = "avx512cd,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpconflictsi_256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>)">;
+}
+
+let Features = "avx512cd,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpconflictdi_512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>)">;
+ def vpconflictsi_512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>)">;
+ def vplzcntd_512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>)">;
+ def vplzcntq_512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>)">;
+}
+
+let Features = "avx512vl,avx512bitalg", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpshufbitqmb128_mask : X86Builtin<"unsigned short(_Vector<16, char>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx512vl,avx512bitalg", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpshufbitqmb256_mask : X86Builtin<"unsigned int(_Vector<32, char>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx512bitalg,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpshufbitqmb512_mask : X86Builtin<"unsigned long long int(_Vector<64, char>, _Vector<64, char>, unsigned long long int)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmulhrsw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>)">;
+ def pmulhuw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>)">;
+ def pmulhw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def addpd512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Constant int)">;
+ def addps512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Constant int)">;
+ def divpd512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Constant int)">;
+ def divps512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Constant int)">;
+ def mulpd512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Constant int)">;
+ def mulps512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Constant int)">;
+ def subpd512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Constant int)">;
+ def subps512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Constant int)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmaddubsw512 : X86Builtin<"_Vector<32, short>(_Vector<64, char>, _Vector<64, char>)">;
+ def pmaddwd512 : X86Builtin<"_Vector<16, int>(_Vector<32, short>, _Vector<32, short>)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def addss_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def divss_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def mulss_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def subss_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def maxss_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def minss_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def addsd_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int)">;
+ def divsd_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int)">;
+ def mulsd_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int)">;
+ def subsd_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int)">;
+ def maxsd_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int)">;
+ def minsd_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def compressdf128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def compressdf256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def compressdi128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def compressdi256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def compresshi128_mask : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def compresshi256_mask : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def compressqi128_mask : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def compressqi256_mask : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def compresssf128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def compresssf256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def compresssi128_mask : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def compresssi256_mask : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def compressstoredf128_mask : X86Builtin<"void(_Vector<2, double *>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def compressstoredf256_mask : X86Builtin<"void(_Vector<4, double *>, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def compressstoredi128_mask : X86Builtin<"void(_Vector<2, long long int *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def compressstoredi256_mask : X86Builtin<"void(_Vector<4, long long int *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def compressstorehi128_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def compressstorehi256_mask : X86Builtin<"void(_Vector<16, short *>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def compressstoreqi128_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def compressstoreqi256_mask : X86Builtin<"void(_Vector<32, char *>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def compressstoresf128_mask : X86Builtin<"void(_Vector<4, float *>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def compressstoresf256_mask : X86Builtin<"void(_Vector<8, float *>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def compressstoresi128_mask : X86Builtin<"void(_Vector<4, int *>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def compressstoresi256_mask : X86Builtin<"void(_Vector<8, int *>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtpd2dq128_mask : X86Builtin<"_Vector<4, int>(_Vector<2, double>, _Vector<4, int>, unsigned char)">;
+ def cvtpd2ps_mask : X86Builtin<"_Vector<4, float>(_Vector<2, double>, _Vector<4, float>, unsigned char)">;
+ def cvtpd2udq128_mask : X86Builtin<"_Vector<4, int>(_Vector<2, double>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtpd2udq256_mask : X86Builtin<"_Vector<4, int>(_Vector<4, double>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtps2udq128_mask : X86Builtin<"_Vector<4, int>(_Vector<4, float>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtps2udq256_mask : X86Builtin<"_Vector<8, int>(_Vector<8, float>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvttpd2dq128_mask : X86Builtin<"_Vector<4, int>(_Vector<2, double>, _Vector<4, int>, unsigned char)">;
+ def cvttpd2udq128_mask : X86Builtin<"_Vector<4, int>(_Vector<2, double>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvttpd2udq256_mask : X86Builtin<"_Vector<4, int>(_Vector<4, double>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvttps2udq128_mask : X86Builtin<"_Vector<4, int>(_Vector<4, float>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvttps2udq256_mask : X86Builtin<"_Vector<8, int>(_Vector<8, float>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def expanddf128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def expanddf256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def expanddi128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def expanddi256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def expandhi128_mask : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def expandhi256_mask : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def expandqi128_mask : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def expandqi256_mask : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def expandloaddf128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double const *>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def expandloaddf256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double const *>, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def expandloaddi128_mask : X86Builtin<"_Vector<4, int>(_Vector<2, long long int const *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def expandloaddi256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int const *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def expandloadhi128_mask : X86Builtin<"_Vector<8, short>(_Vector<8, short const *>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def expandloadhi256_mask : X86Builtin<"_Vector<16, short>(_Vector<16, short const *>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def expandloadqi128_mask : X86Builtin<"_Vector<16, char>(_Vector<16, char const *>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def expandloadqi256_mask : X86Builtin<"_Vector<32, char>(_Vector<32, char const *>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def expandloadsf128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float const *>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def expandloadsf256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float const *>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def expandloadsi128_mask : X86Builtin<"_Vector<4, int>(_Vector<4, int const *>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def expandloadsi256_mask : X86Builtin<"_Vector<8, int>(_Vector<8, int const *>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def expandsf128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def expandsf256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def expandsi128_mask : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def expandsi256_mask : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def getexppd128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def getexppd256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def getexpps128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def getexpps256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rndscalepd_128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Constant int, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def rndscalepd_256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Constant int, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rndscaleps_128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Constant int, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def rndscaleps_256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Constant int, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def scalefpd128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def scalefpd256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def scalefps128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def scalefps256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def scatterdiv2df : X86Builtin<"void(void *, unsigned char, _Vector<2, long long int>, _Vector<2, double>, _Constant int)">;
+ def scatterdiv2di : X86Builtin<"void(void *, unsigned char, _Vector<2, long long int>, _Vector<2, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def scatterdiv4df : X86Builtin<"void(void *, unsigned char, _Vector<4, long long int>, _Vector<4, double>, _Constant int)">;
+ def scatterdiv4di : X86Builtin<"void(void *, unsigned char, _Vector<4, long long int>, _Vector<4, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def scatterdiv4sf : X86Builtin<"void(void *, unsigned char, _Vector<2, long long int>, _Vector<4, float>, _Constant int)">;
+ def scatterdiv4si : X86Builtin<"void(void *, unsigned char, _Vector<2, long long int>, _Vector<4, int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def scatterdiv8sf : X86Builtin<"void(void *, unsigned char, _Vector<4, long long int>, _Vector<4, float>, _Constant int)">;
+ def scatterdiv8si : X86Builtin<"void(void *, unsigned char, _Vector<4, long long int>, _Vector<4, int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def scattersiv2df : X86Builtin<"void(void *, unsigned char, _Vector<4, int>, _Vector<2, double>, _Constant int)">;
+ def scattersiv2di : X86Builtin<"void(void *, unsigned char, _Vector<4, int>, _Vector<2, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def scattersiv4df : X86Builtin<"void(void *, unsigned char, _Vector<4, int>, _Vector<4, double>, _Constant int)">;
+ def scattersiv4di : X86Builtin<"void(void *, unsigned char, _Vector<4, int>, _Vector<4, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def scattersiv4sf : X86Builtin<"void(void *, unsigned char, _Vector<4, int>, _Vector<4, float>, _Constant int)">;
+ def scattersiv4si : X86Builtin<"void(void *, unsigned char, _Vector<4, int>, _Vector<4, int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def scattersiv8sf : X86Builtin<"void(void *, unsigned char, _Vector<8, int>, _Vector<8, float>, _Constant int)">;
+ def scattersiv8si : X86Builtin<"void(void *, unsigned char, _Vector<8, int>, _Vector<8, int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpermi2vard128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpermi2vard256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpermi2vard512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpermi2varpd128 : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, long long int>, _Vector<2, double>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpermi2varpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, long long int>, _Vector<4, double>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpermi2varpd512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, long long int>, _Vector<8, double>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpermi2varps128 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, int>, _Vector<4, float>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpermi2varps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, int>, _Vector<8, float>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpermi2varps512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, int>, _Vector<16, float>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpermi2varq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpermi2varq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpermi2varq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, _Vector<8, long long int>)">;
+}
+
+let Features = "avx512vbmi,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpermi2varqi128 : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>, _Vector<16, char>)">;
+}
+
+let Features = "avx512vbmi,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpermi2varqi256 : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>, _Vector<32, char>)">;
+}
+
+let Features = "avx512vbmi,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpermi2varqi512 : X86Builtin<"_Vector<64, char>(_Vector<64, char>, _Vector<64, char>, _Vector<64, char>)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpermi2varhi128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>, _Vector<8, short>)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpermi2varhi256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>, _Vector<16, short>)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpermi2varhi512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>, _Vector<32, short>)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpshldd128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Constant int)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpshldd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Constant int)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpshldd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Constant int)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpshldq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpshldq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpshldq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpshldw128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>, _Constant int)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpshldw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>, _Constant int)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpshldw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>, _Constant int)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpshldvd128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpshldvd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpshldvd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpshldvq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpshldvq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpshldvq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, _Vector<8, long long int>)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpshldvw128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>, _Vector<8, short>)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpshldvw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>, _Vector<16, short>)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpshldvw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>, _Vector<32, short>)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpshrdvd128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpshrdvd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpshrdvd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpshrdvq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpshrdvq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpshrdvq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, _Vector<8, long long int>)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpshrdvw128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>, _Vector<8, short>)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpshrdvw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>, _Vector<16, short>)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpshrdvw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>, _Vector<32, short>)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpshrdd128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Constant int)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpshrdd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Constant int)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpshrdd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Constant int)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpshrdq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpshrdq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpshrdq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpshrdw128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>, _Constant int)">;
+}
+
+let Features = "avx512vl,avx512vbmi2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpshrdw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>, _Constant int)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpshrdw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>, _Constant int)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovswb512_mask : X86Builtin<"_Vector<32, char>(_Vector<32, short>, _Vector<32, char>, unsigned int)">;
+ def pmovuswb512_mask : X86Builtin<"_Vector<32, char>(_Vector<32, short>, _Vector<32, char>, unsigned int)">;
+ def pmovwb512_mask : X86Builtin<"_Vector<32, char>(_Vector<32, short>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtpd2qq128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<2, double>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtpd2qq256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, double>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtpd2uqq128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<2, double>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtpd2uqq256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, double>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtps2qq128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<4, float>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtps2qq256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, float>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtps2uqq128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<4, float>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtps2uqq256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, float>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtqq2ps128_mask : X86Builtin<"_Vector<4, float>(_Vector<2, long long int>, _Vector<4, float>, unsigned char)">;
+ def cvttpd2qq128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<2, double>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvttpd2qq256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, double>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvttpd2uqq128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<2, double>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvttpd2uqq256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, double>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvttps2qq128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<4, float>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvttps2qq256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, float>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvttps2uqq128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<4, float>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvttps2uqq256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, float>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtuqq2ps128_mask : X86Builtin<"_Vector<4, float>(_Vector<2, long long int>, _Vector<4, float>, unsigned char)">;
+ def rangepd128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Constant int, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def rangepd256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant int, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rangeps128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Constant int, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def rangeps256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant int, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rangesd128_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int, _Constant int)">;
+ def rangess128_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int, _Constant int)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def reducepd128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Constant int, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def reducepd256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Constant int, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def reduceps128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Constant int, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def reduceps256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Constant int, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def reducesd_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int, _Constant int)">;
+ def reducess_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int, _Constant int)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovswb128_mask : X86Builtin<"_Vector<16, char>(_Vector<8, short>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovswb256_mask : X86Builtin<"_Vector<16, char>(_Vector<16, short>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovuswb128_mask : X86Builtin<"_Vector<16, char>(_Vector<8, short>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovuswb256_mask : X86Builtin<"_Vector<16, char>(_Vector<16, short>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovwb128_mask : X86Builtin<"_Vector<16, char>(_Vector<8, short>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512dq,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def cvtpd2qq512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, double>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+ def cvtpd2uqq512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, double>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+ def cvtps2qq512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, float>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+ def cvtps2uqq512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, float>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+ def cvtqq2pd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, long long int>, _Vector<8, double>, unsigned char, _Constant int)">;
+ def cvtqq2ps512_mask : X86Builtin<"_Vector<8, float>(_Vector<8, long long int>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def cvttpd2qq512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, double>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+ def cvttpd2uqq512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, double>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+ def cvttps2qq512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, float>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+ def cvttps2uqq512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, float>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+ def cvtuqq2pd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, long long int>, _Vector<8, double>, unsigned char, _Constant int)">;
+ def cvtuqq2ps512_mask : X86Builtin<"_Vector<8, float>(_Vector<8, long long int>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def rangepd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Constant int, _Vector<8, double>, unsigned char, _Constant int)">;
+ def rangeps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Constant int, _Vector<16, float>, unsigned short, _Constant int)">;
+ def reducepd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Constant int, _Vector<8, double>, unsigned char, _Constant int)">;
+ def reduceps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Constant int, _Vector<16, float>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def prold512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Constant int)">;
+ def prolq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def prold128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def prold256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def prolq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def prolq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Constant int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def prolvd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>)">;
+ def prolvq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>)">;
+ def prord512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Constant int)">;
+ def prorq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def prolvd128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def prolvd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def prolvq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def prolvq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def prord128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def prord256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def prorq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def prorq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Constant int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def prorvd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>)">;
+ def prorvq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def prorvd128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def prorvd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def prorvq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def prorvq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pshufhw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Constant int)">;
+ def pshuflw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Constant int)">;
+ def psllv32hi : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>)">;
+ def psllw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<8, short>)">;
+ def psllwi512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, int)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def psllv16hi : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def psllv8hi : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pslldi512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, int)">;
+ def psllqi512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, int)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def psrlv32hi : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def psrlv16hi : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def psrlv8hi : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def psrldi512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, int)">;
+ def psrlqi512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, int)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def psrav32hi : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def psrav16hi : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def psrav8hi : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def psravq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def psravq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def psraw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<8, short>)">;
+ def psrawi512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, int)">;
+ def psrlw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<8, short>)">;
+ def psrlwi512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, int)">;
+ def pslldqi512_byteshift : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Constant int)">;
+ def psrldqi512_byteshift : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def movdqa32load128_mask : X86Builtin<"_Vector<4, int>(_Vector<4, int const *>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def movdqa32load256_mask : X86Builtin<"_Vector<8, int>(_Vector<8, int const *>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def movdqa32load512_mask : X86Builtin<"_Vector<16, int>(_Vector<16, int const *>, _Vector<16, int>, unsigned short)">;
+ def movdqa32store512_mask : X86Builtin<"void(_Vector<16, int *>, _Vector<16, int>, unsigned short)">;
+ def movdqa64load512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int const *>, _Vector<8, long long int>, unsigned char)">;
+ def movdqa64store512_mask : X86Builtin<"void(_Vector<8, long long int *>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def movdqa32store128_mask : X86Builtin<"void(_Vector<4, int *>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def movdqa32store256_mask : X86Builtin<"void(_Vector<8, int *>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def movdqa64load128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int const *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def movdqa64load256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int const *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def movdqa64store128_mask : X86Builtin<"void(_Vector<2, long long int *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def movdqa64store256_mask : X86Builtin<"void(_Vector<4, long long int *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512ifma,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpmadd52huq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, _Vector<8, long long int>)">;
+ def vpmadd52luq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, _Vector<8, long long int>)">;
+}
+
+let Features = "avx512ifma,avx512vl|avxifma", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpmadd52huq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx512ifma,avx512vl|avxifma", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpmadd52huq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512ifma,avx512vl|avxifma", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpmadd52luq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx512ifma,avx512vl|avxifma", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpmadd52luq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcomisd : X86Builtin<"int(_Vector<2, double>, _Vector<2, double>, _Constant int, _Constant int)">;
+ def vcomiss : X86Builtin<"int(_Vector<4, float>, _Vector<4, float>, _Constant int, _Constant int)">;
+}
+
+let Features = "avx512bw", Attributes = [NoThrow, Const] in {
+ def kunpckdi : X86Builtin<"unsigned long long int(unsigned long long int, unsigned long long int)">;
+ def kunpcksi : X86Builtin<"unsigned int(unsigned int, unsigned int)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def loaddquhi512_mask : X86Builtin<"_Vector<32, short>(_Vector<32, short const *>, _Vector<32, short>, unsigned int)">;
+ def loaddquqi512_mask : X86Builtin<"_Vector<64, char>(_Vector<64, char const *>, _Vector<64, char>, unsigned long long int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def fixupimmpd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Vector<8, long long int>, _Constant int, unsigned char, _Constant int)">;
+ def fixupimmpd512_maskz : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Vector<8, long long int>, _Constant int, unsigned char, _Constant int)">;
+ def fixupimmps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, int>, _Constant int, unsigned short, _Constant int)">;
+ def fixupimmps512_maskz : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, int>, _Constant int, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def fixupimmsd_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, long long int>, _Constant int, unsigned char, _Constant int)">;
+ def fixupimmsd_maskz : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, long long int>, _Constant int, unsigned char, _Constant int)">;
+ def fixupimmss_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, int>, _Constant int, unsigned char, _Constant int)">;
+ def fixupimmss_maskz : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, int>, _Constant int, unsigned char, _Constant int)">;
+ def getexpsd128_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int)">;
+ def getexpss128_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def getmantsd_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Constant int, _Vector<2, double>, unsigned char, _Constant int)">;
+ def getmantss_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Constant int, _Vector<4, float>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def loaddquhi128_mask : X86Builtin<"_Vector<8, short>(_Vector<8, short const *>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def loaddquhi256_mask : X86Builtin<"_Vector<16, short>(_Vector<16, short const *>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def loaddquqi128_mask : X86Builtin<"_Vector<16, char>(_Vector<16, char const *>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def loaddquqi256_mask : X86Builtin<"_Vector<32, char>(_Vector<32, char const *>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def fixupimmpd128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, long long int>, _Constant int, unsigned char)">;
+ def fixupimmpd128_maskz : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, long long int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def fixupimmpd256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, long long int>, _Constant int, unsigned char)">;
+ def fixupimmpd256_maskz : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, long long int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def fixupimmps128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, int>, _Constant int, unsigned char)">;
+ def fixupimmps128_maskz : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def fixupimmps256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, int>, _Constant int, unsigned char)">;
+ def fixupimmps256_maskz : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def loadapd128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double const *>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def loadsd128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double const *>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def loadapd256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double const *>, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def loadaps128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float const *>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def loadss128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float const *>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def loadaps256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float const *>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def loaddqudi128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int const *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def loaddqudi256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int const *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def loaddqusi128_mask : X86Builtin<"_Vector<4, int>(_Vector<4, int const *>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def loaddqusi256_mask : X86Builtin<"_Vector<8, int>(_Vector<8, int const *>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def loadupd128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double const *>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def loadupd256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double const *>, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def loadups128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float const *>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def loadups256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float const *>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def storedquhi512_mask : X86Builtin<"void(_Vector<32, short *>, _Vector<32, short>, unsigned int)">;
+ def storedquqi512_mask : X86Builtin<"void(_Vector<64, char *>, _Vector<64, char>, unsigned long long int)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def storedquhi128_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def storedquhi256_mask : X86Builtin<"void(_Vector<16, short *>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def storedquqi128_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def storedquqi256_mask : X86Builtin<"void(_Vector<32, char *>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def storeapd128_mask : X86Builtin<"void(_Vector<2, double *>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def storesd128_mask : X86Builtin<"void(_Vector<2, double *>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def storeapd256_mask : X86Builtin<"void(_Vector<4, double *>, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def storeaps128_mask : X86Builtin<"void(_Vector<4, float *>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def storess128_mask : X86Builtin<"void(_Vector<4, float *>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def storeaps256_mask : X86Builtin<"void(_Vector<8, float *>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def storedqudi128_mask : X86Builtin<"void(_Vector<2, long long int *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def storedqudi256_mask : X86Builtin<"void(_Vector<4, long long int *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def storedqusi128_mask : X86Builtin<"void(_Vector<4, int *>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def storedqusi256_mask : X86Builtin<"void(_Vector<8, int *>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def storeupd128_mask : X86Builtin<"void(_Vector<2, double *>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def storeupd256_mask : X86Builtin<"void(_Vector<4, double *>, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def storeups128_mask : X86Builtin<"void(_Vector<4, float *>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def storeups256_mask : X86Builtin<"void(_Vector<8, float *>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rcp14pd128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def rcp14pd256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rcp14ps128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def rcp14ps256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512cd,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vplzcntd_128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>)">;
+}
+
+let Features = "avx512cd,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vplzcntd_256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>)">;
+}
+
+let Features = "avx512cd,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vplzcntq_128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>)">;
+}
+
+let Features = "avx512cd,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vplzcntq_256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtsd2si32 : X86Builtin<"int(_Vector<2, double>, _Constant int)">;
+ def vcvtsd2usi32 : X86Builtin<"unsigned int(_Vector<2, double>, _Constant int)">;
+ def vcvtss2si32 : X86Builtin<"int(_Vector<4, float>, _Constant int)">;
+ def vcvtss2usi32 : X86Builtin<"unsigned int(_Vector<4, float>, _Constant int)">;
+ def vcvttsd2si32 : X86Builtin<"int(_Vector<2, double>, _Constant int)">;
+ def vcvttsd2usi32 : X86Builtin<"unsigned int(_Vector<2, double>, _Constant int)">;
+ def vcvttss2si32 : X86Builtin<"int(_Vector<4, float>, _Constant int)">;
+ def vcvttss2usi32 : X86Builtin<"unsigned int(_Vector<4, float>, _Constant int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpermilpd512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Constant int)">;
+ def vpermilps512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Constant int)">;
+ def vpermilvarpd512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, long long int>)">;
+ def vpermilvarps512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, int>)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rndscalesd_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int, _Constant int)">;
+ def rndscaless_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int, _Constant int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def scalefpd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Vector<8, double>, unsigned char, _Constant int)">;
+ def scalefps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def scalefsd_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int)">;
+ def scalefss_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def psradi512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, int)">;
+ def psraqi512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def psraq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def psraq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def psraqi128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def psraqi256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pslld512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<4, int>)">;
+ def psllq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<2, long long int>)">;
+ def psllv16si : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>)">;
+ def psllv8di : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>)">;
+ def psrad512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<4, int>)">;
+ def psraq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<2, long long int>)">;
+ def psrav16si : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>)">;
+ def psrav8di : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>)">;
+ def psrld512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<4, int>)">;
+ def psrlq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<2, long long int>)">;
+ def psrlv16si : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>)">;
+ def psrlv8di : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>)">;
+ def pternlogd512_mask : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>, _Constant int, unsigned short)">;
+ def pternlogd512_maskz : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>, _Constant int, unsigned short)">;
+ def pternlogq512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, _Vector<8, long long int>, _Constant int, unsigned char)">;
+ def pternlogq512_maskz : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, _Vector<8, long long int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pternlogd128_mask : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>, _Constant int, unsigned char)">;
+ def pternlogd128_maskz : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pternlogd256_mask : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>, _Constant int, unsigned char)">;
+ def pternlogd256_maskz : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pternlogq128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Vector<2, long long int>, _Constant int, unsigned char)">;
+ def pternlogq128_maskz : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, _Vector<2, long long int>, _Vector<2, long long int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pternlogq256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, _Vector<4, long long int>, _Constant int, unsigned char)">;
+ def pternlogq256_maskz : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, _Vector<4, long long int>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def shuf_f32x4 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Constant int)">;
+ def shuf_f64x2 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Constant int)">;
+ def shuf_i32x4 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Constant int)">;
+ def shuf_i64x2 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, _Constant int)">;
+ def shufpd512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Constant int)">;
+ def shufps512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def shuf_f32x4_256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant int)">;
+ def shuf_f64x2_256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant int)">;
+ def shuf_i32x4_256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Constant int)">;
+ def shuf_i64x2_256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>, _Constant int)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def sqrtsd_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int)">;
+ def sqrtss_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rsqrt14pd128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def rsqrt14pd256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rsqrt14ps128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def rsqrt14ps256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def cvtb2mask512 : X86Builtin<"unsigned long long int(_Vector<64, char>)">;
+ def cvtmask2b512 : X86Builtin<"_Vector<64, char>(unsigned long long int)">;
+ def cvtmask2w512 : X86Builtin<"_Vector<32, short>(unsigned int)">;
+}
+
+let Features = "avx512dq,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def cvtd2mask512 : X86Builtin<"unsigned short(_Vector<16, int>)">;
+ def cvtmask2d512 : X86Builtin<"_Vector<16, int>(unsigned short)">;
+ def cvtmask2q512 : X86Builtin<"_Vector<8, long long int>(unsigned char)">;
+ def cvtq2mask512 : X86Builtin<"unsigned char(_Vector<8, long long int>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtb2mask128 : X86Builtin<"unsigned short(_Vector<16, char>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtb2mask256 : X86Builtin<"unsigned int(_Vector<32, char>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtmask2b128 : X86Builtin<"_Vector<16, char>(unsigned short)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtmask2b256 : X86Builtin<"_Vector<32, char>(unsigned int)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtmask2w128 : X86Builtin<"_Vector<8, short>(unsigned char)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtmask2w256 : X86Builtin<"_Vector<16, short>(unsigned short)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtd2mask128 : X86Builtin<"unsigned char(_Vector<4, int>)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtd2mask256 : X86Builtin<"unsigned char(_Vector<8, int>)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtmask2d128 : X86Builtin<"_Vector<4, int>(unsigned char)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtmask2d256 : X86Builtin<"_Vector<8, int>(unsigned char)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtmask2q128 : X86Builtin<"_Vector<2, long long int>(unsigned char)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtmask2q256 : X86Builtin<"_Vector<4, long long int>(unsigned char)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtq2mask128 : X86Builtin<"unsigned char(_Vector<2, long long int>)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtq2mask256 : X86Builtin<"unsigned char(_Vector<4, long long int>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovsdb512_mask : X86Builtin<"_Vector<16, char>(_Vector<16, int>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovsdb512mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<16, int>, unsigned short)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovswb512mem_mask : X86Builtin<"void(_Vector<32, char *>, _Vector<32, short>, unsigned int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovsdw512_mask : X86Builtin<"_Vector<16, short>(_Vector<16, int>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovsdw512mem_mask : X86Builtin<"void(_Vector<16, short *>, _Vector<16, int>, unsigned short)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovsqb512_mask : X86Builtin<"_Vector<16, char>(_Vector<8, long long int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovsqb512mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovsqd512_mask : X86Builtin<"_Vector<8, int>(_Vector<8, long long int>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovsqd512mem_mask : X86Builtin<"void(_Vector<8, int *>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovsqw512_mask : X86Builtin<"_Vector<8, short>(_Vector<8, long long int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovsqw512mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovsdb128_mask : X86Builtin<"_Vector<16, char>(_Vector<4, int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovsdb128mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovswb128mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovsdb256_mask : X86Builtin<"_Vector<16, char>(_Vector<8, int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovsdb256mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovswb256mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovsdw128_mask : X86Builtin<"_Vector<8, short>(_Vector<4, int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovsdw128mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovsdw256_mask : X86Builtin<"_Vector<8, short>(_Vector<8, int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovsdw256mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovsqb128_mask : X86Builtin<"_Vector<16, char>(_Vector<2, long long int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovsqb128mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovsqb256_mask : X86Builtin<"_Vector<16, char>(_Vector<4, long long int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovsqb256mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovsqd128_mask : X86Builtin<"_Vector<4, int>(_Vector<2, long long int>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovsqd128mem_mask : X86Builtin<"void(_Vector<4, int *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovsqd256_mask : X86Builtin<"_Vector<4, int>(_Vector<4, long long int>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovsqd256mem_mask : X86Builtin<"void(_Vector<4, int *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovsqw128_mask : X86Builtin<"_Vector<8, short>(_Vector<2, long long int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovsqw128mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovsqw256_mask : X86Builtin<"_Vector<8, short>(_Vector<4, long long int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovsqw256mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovusdb512_mask : X86Builtin<"_Vector<16, char>(_Vector<16, int>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovusdb512mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<16, int>, unsigned short)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovuswb512mem_mask : X86Builtin<"void(_Vector<32, char *>, _Vector<32, short>, unsigned int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovusdw512_mask : X86Builtin<"_Vector<16, short>(_Vector<16, int>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovusdw512mem_mask : X86Builtin<"void(_Vector<16, short *>, _Vector<16, int>, unsigned short)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovusqb512_mask : X86Builtin<"_Vector<16, char>(_Vector<8, long long int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovusqb512mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovusqd512_mask : X86Builtin<"_Vector<8, int>(_Vector<8, long long int>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovusqd512mem_mask : X86Builtin<"void(_Vector<8, int *>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovusqw512_mask : X86Builtin<"_Vector<8, short>(_Vector<8, long long int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovusqw512mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovusdb128_mask : X86Builtin<"_Vector<16, char>(_Vector<4, int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovusdb128mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovuswb128mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovusdb256_mask : X86Builtin<"_Vector<16, char>(_Vector<8, int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovusdb256mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovuswb256mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovusdw128_mask : X86Builtin<"_Vector<8, short>(_Vector<4, int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovusdw128mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovusdw256_mask : X86Builtin<"_Vector<8, short>(_Vector<8, int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovusdw256mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovusqb128_mask : X86Builtin<"_Vector<16, char>(_Vector<2, long long int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovusqb128mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovusqb256_mask : X86Builtin<"_Vector<16, char>(_Vector<4, long long int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovusqb256mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovusqd128_mask : X86Builtin<"_Vector<4, int>(_Vector<2, long long int>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovusqd128mem_mask : X86Builtin<"void(_Vector<4, int *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovusqd256_mask : X86Builtin<"_Vector<4, int>(_Vector<4, long long int>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovusqd256mem_mask : X86Builtin<"void(_Vector<4, int *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovusqw128_mask : X86Builtin<"_Vector<8, short>(_Vector<2, long long int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovusqw128mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovusqw256_mask : X86Builtin<"_Vector<8, short>(_Vector<4, long long int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovusqw256mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovdb512_mask : X86Builtin<"_Vector<16, char>(_Vector<16, int>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovdb512mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<16, int>, unsigned short)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovwb512mem_mask : X86Builtin<"void(_Vector<32, char *>, _Vector<32, short>, unsigned int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovdw512_mask : X86Builtin<"_Vector<16, short>(_Vector<16, int>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovdw512mem_mask : X86Builtin<"void(_Vector<16, short *>, _Vector<16, int>, unsigned short)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovqb512_mask : X86Builtin<"_Vector<16, char>(_Vector<8, long long int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovqb512mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovqd512_mask : X86Builtin<"_Vector<8, int>(_Vector<8, long long int>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovqd512mem_mask : X86Builtin<"void(_Vector<8, int *>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pmovqw512_mask : X86Builtin<"_Vector<8, short>(_Vector<8, long long int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def pmovqw512mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovdb128_mask : X86Builtin<"_Vector<16, char>(_Vector<4, int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovwb128mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovdb128mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovdb256_mask : X86Builtin<"_Vector<16, char>(_Vector<8, int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovdb256mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl,avx512bw", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovwb256mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovdw128_mask : X86Builtin<"_Vector<8, short>(_Vector<4, int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovdw128mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovdw256_mask : X86Builtin<"_Vector<8, short>(_Vector<8, int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovdw256mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovqb128_mask : X86Builtin<"_Vector<16, char>(_Vector<2, long long int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovqb128mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovqb256_mask : X86Builtin<"_Vector<16, char>(_Vector<4, long long int>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovqb256mem_mask : X86Builtin<"void(_Vector<16, char *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovqd128_mask : X86Builtin<"_Vector<4, int>(_Vector<2, long long int>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovqd128mem_mask : X86Builtin<"void(_Vector<4, int *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovqd256mem_mask : X86Builtin<"void(_Vector<4, int *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def pmovqw128_mask : X86Builtin<"_Vector<8, short>(_Vector<2, long long int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def pmovqw128mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def pmovqw256_mask : X86Builtin<"_Vector<8, short>(_Vector<4, long long int>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def pmovqw256mem_mask : X86Builtin<"void(_Vector<8, short *>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512dq,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def extractf32x8_mask : X86Builtin<"_Vector<8, float>(_Vector<16, float>, _Constant int, _Vector<8, float>, unsigned char)">;
+ def extractf64x2_512_mask : X86Builtin<"_Vector<2, double>(_Vector<8, double>, _Constant int, _Vector<2, double>, unsigned char)">;
+ def extracti32x8_mask : X86Builtin<"_Vector<8, int>(_Vector<16, int>, _Constant int, _Vector<8, int>, unsigned char)">;
+ def extracti64x2_512_mask : X86Builtin<"_Vector<2, long long int>(_Vector<8, long long int>, _Constant int, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def extracti32x4_mask : X86Builtin<"_Vector<4, int>(_Vector<16, int>, _Constant int, _Vector<4, int>, unsigned char)">;
+ def extracti64x4_mask : X86Builtin<"_Vector<4, long long int>(_Vector<8, long long int>, _Constant int, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def extractf64x2_256_mask : X86Builtin<"_Vector<2, double>(_Vector<4, double>, _Constant int, _Vector<2, double>, unsigned char)">;
+ def extracti64x2_256_mask : X86Builtin<"_Vector<2, long long int>(_Vector<4, long long int>, _Constant int, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def extractf32x4_256_mask : X86Builtin<"_Vector<4, float>(_Vector<8, float>, _Constant int, _Vector<4, float>, unsigned char)">;
+ def extracti32x4_256_mask : X86Builtin<"_Vector<4, int>(_Vector<8, int>, _Constant int, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512dq,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def insertf32x8 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<8, float>, _Constant int)">;
+ def insertf64x2_512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<2, double>, _Constant int)">;
+ def inserti32x8 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<8, int>, _Constant int)">;
+ def inserti64x2_512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<2, long long int>, _Constant int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def insertf64x4 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<4, double>, _Constant int)">;
+ def inserti64x4 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<4, long long int>, _Constant int)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def insertf64x2_256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<2, double>, _Constant int)">;
+ def inserti64x2_256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<2, long long int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def insertf32x4_256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<4, float>, _Constant int)">;
+ def inserti32x4_256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<4, int>, _Constant int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def insertf32x4 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<4, float>, _Constant int)">;
+ def inserti32x4 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<4, int>, _Constant int)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def getmantpd128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Constant int, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def getmantpd256_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Constant int, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def getmantps128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Constant int, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def getmantps256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Constant int, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def getmantpd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Constant int, _Vector<8, double>, unsigned char, _Constant int)">;
+ def getmantps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Constant int, _Vector<16, float>, unsigned short, _Constant int)">;
+ def getexppd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, unsigned char, _Constant int)">;
+ def getexpps512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfmaddss3_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vfmaddss3_maskz : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vfmaddss3_mask3 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vfmaddsd3_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int)">;
+ def vfmaddsd3_maskz : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int)">;
+ def vfmaddsd3_mask3 : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int)">;
+ def vfmsubsd3_mask3 : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Vector<2, double>, unsigned char, _Constant int)">;
+ def vfmsubss3_mask3 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def permdf512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Constant int)">;
+ def permdi512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Constant int)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def permvarhi512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def permvardf512 : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, long long int>)">;
+ def permvardi512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>)">;
+ def permvarsf512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, int>)">;
+ def permvarsi512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>)">;
+}
+
+let Features = "avx512vbmi,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def permvarqi512 : X86Builtin<"_Vector<64, char>(_Vector<64, char>, _Vector<64, char>)">;
+}
+
+let Features = "avx512vbmi,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def permvarqi128 : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>)">;
+}
+
+let Features = "avx512vbmi,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def permvarqi256 : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def permvarhi128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def permvarhi256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def permvardf256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, long long int>)">;
+ def permvardi256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def fpclasspd128_mask : X86Builtin<"unsigned char(_Vector<2, double>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def fpclasspd256_mask : X86Builtin<"unsigned char(_Vector<4, double>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def fpclassps128_mask : X86Builtin<"unsigned char(_Vector<4, float>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512dq,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def fpclassps256_mask : X86Builtin<"unsigned char(_Vector<8, float>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512dq,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def fpclassps512_mask : X86Builtin<"unsigned short(_Vector<16, float>, _Constant int, unsigned short)">;
+ def fpclasspd512_mask : X86Builtin<"unsigned char(_Vector<8, double>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def fpclasssd_mask : X86Builtin<"unsigned char(_Vector<2, double>, _Constant int, unsigned char)">;
+ def fpclassss_mask : X86Builtin<"unsigned char(_Vector<4, float>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const] in {
+ def kaddqi : X86Builtin<"unsigned char(unsigned char, unsigned char)">;
+ def kaddhi : X86Builtin<"unsigned short(unsigned short, unsigned short)">;
+}
+
+let Features = "avx512bw", Attributes = [NoThrow, Const] in {
+ def kaddsi : X86Builtin<"unsigned int(unsigned int, unsigned int)">;
+ def kadddi : X86Builtin<"unsigned long long int(unsigned long long int, unsigned long long int)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const] in {
+ def kandqi : X86Builtin<"unsigned char(unsigned char, unsigned char)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const] in {
+ def kandhi : X86Builtin<"unsigned short(unsigned short, unsigned short)">;
+}
+
+let Features = "avx512bw", Attributes = [NoThrow, Const] in {
+ def kandsi : X86Builtin<"unsigned int(unsigned int, unsigned int)">;
+ def kanddi : X86Builtin<"unsigned long long int(unsigned long long int, unsigned long long int)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const] in {
+ def kandnqi : X86Builtin<"unsigned char(unsigned char, unsigned char)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const] in {
+ def kandnhi : X86Builtin<"unsigned short(unsigned short, unsigned short)">;
+}
+
+let Features = "avx512bw", Attributes = [NoThrow, Const] in {
+ def kandnsi : X86Builtin<"unsigned int(unsigned int, unsigned int)">;
+ def kandndi : X86Builtin<"unsigned long long int(unsigned long long int, unsigned long long int)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const] in {
+ def korqi : X86Builtin<"unsigned char(unsigned char, unsigned char)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const] in {
+ def korhi : X86Builtin<"unsigned short(unsigned short, unsigned short)">;
+}
+
+let Features = "avx512bw", Attributes = [NoThrow, Const] in {
+ def korsi : X86Builtin<"unsigned int(unsigned int, unsigned int)">;
+ def kordi : X86Builtin<"unsigned long long int(unsigned long long int, unsigned long long int)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const] in {
+ def kortestcqi : X86Builtin<"int(unsigned char, unsigned char)">;
+ def kortestzqi : X86Builtin<"int(unsigned char, unsigned char)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const] in {
+ def kortestchi : X86Builtin<"int(unsigned short, unsigned short)">;
+ def kortestzhi : X86Builtin<"int(unsigned short, unsigned short)">;
+}
+
+let Features = "avx512bw", Attributes = [NoThrow, Const] in {
+ def kortestcsi : X86Builtin<"int(unsigned int, unsigned int)">;
+ def kortestzsi : X86Builtin<"int(unsigned int, unsigned int)">;
+ def kortestcdi : X86Builtin<"int(unsigned long long int, unsigned long long int)">;
+ def kortestzdi : X86Builtin<"int(unsigned long long int, unsigned long long int)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const] in {
+ def ktestcqi : X86Builtin<"int(unsigned char, unsigned char)">;
+ def ktestzqi : X86Builtin<"int(unsigned char, unsigned char)">;
+ def ktestchi : X86Builtin<"int(unsigned short, unsigned short)">;
+ def ktestzhi : X86Builtin<"int(unsigned short, unsigned short)">;
+}
+
+let Features = "avx512bw", Attributes = [NoThrow, Const] in {
+ def ktestcsi : X86Builtin<"int(unsigned int, unsigned int)">;
+ def ktestzsi : X86Builtin<"int(unsigned int, unsigned int)">;
+ def ktestcdi : X86Builtin<"int(unsigned long long int, unsigned long long int)">;
+ def ktestzdi : X86Builtin<"int(unsigned long long int, unsigned long long int)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const] in {
+ def kunpckhi : X86Builtin<"unsigned short(unsigned short, unsigned short)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const] in {
+ def kxnorqi : X86Builtin<"unsigned char(unsigned char, unsigned char)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const] in {
+ def kxnorhi : X86Builtin<"unsigned short(unsigned short, unsigned short)">;
+}
+
+let Features = "avx512bw", Attributes = [NoThrow, Const] in {
+ def kxnorsi : X86Builtin<"unsigned int(unsigned int, unsigned int)">;
+ def kxnordi : X86Builtin<"unsigned long long int(unsigned long long int, unsigned long long int)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const] in {
+ def kxorqi : X86Builtin<"unsigned char(unsigned char, unsigned char)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const] in {
+ def kxorhi : X86Builtin<"unsigned short(unsigned short, unsigned short)">;
+}
+
+let Features = "avx512bw", Attributes = [NoThrow, Const] in {
+ def kxorsi : X86Builtin<"unsigned int(unsigned int, unsigned int)">;
+ def kxordi : X86Builtin<"unsigned long long int(unsigned long long int, unsigned long long int)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const] in {
+ def kshiftliqi : X86Builtin<"unsigned char(unsigned char, _Constant unsigned int)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const] in {
+ def kshiftlihi : X86Builtin<"unsigned short(unsigned short, _Constant unsigned int)">;
+}
+
+let Features = "avx512bw", Attributes = [NoThrow, Const] in {
+ def kshiftlisi : X86Builtin<"unsigned int(unsigned int, _Constant unsigned int)">;
+ def kshiftlidi : X86Builtin<"unsigned long long int(unsigned long long int, _Constant unsigned int)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const] in {
+ def kshiftriqi : X86Builtin<"unsigned char(unsigned char, _Constant unsigned int)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const] in {
+ def kshiftrihi : X86Builtin<"unsigned short(unsigned short, _Constant unsigned int)">;
+}
+
+let Features = "avx512bw", Attributes = [NoThrow, Const] in {
+ def kshiftrisi : X86Builtin<"unsigned int(unsigned int, _Constant unsigned int)">;
+ def kshiftridi : X86Builtin<"unsigned long long int(unsigned long long int, _Constant unsigned int)">;
+}
+
+let Features = "avx512dq", Attributes = [NoThrow, Const] in {
+ def kmovb : X86Builtin<"unsigned char(unsigned char)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const] in {
+ def kmovw : X86Builtin<"unsigned short(unsigned short)">;
+}
+
+let Features = "avx512bw", Attributes = [NoThrow, Const] in {
+ def kmovd : X86Builtin<"unsigned int(unsigned int)">;
+ def kmovq : X86Builtin<"unsigned long long int(unsigned long long int)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def palignr512 : X86Builtin<"_Vector<64, char>(_Vector<64, char>, _Vector<64, char>, _Constant int)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def dbpsadbw128 : X86Builtin<"_Vector<8, short>(_Vector<16, char>, _Vector<16, char>, _Constant int)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def dbpsadbw256 : X86Builtin<"_Vector<16, short>(_Vector<32, char>, _Vector<32, char>, _Constant int)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def dbpsadbw512 : X86Builtin<"_Vector<32, short>(_Vector<64, char>, _Vector<64, char>, _Constant int)">;
+ def psadbw512 : X86Builtin<"_Vector<8, long long int>(_Vector<64, char>, _Vector<64, char>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def compressdf512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, unsigned char)">;
+ def compressdi512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def compresshi512_mask : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>, unsigned int)">;
+ def compressqi512_mask : X86Builtin<"_Vector<64, char>(_Vector<64, char>, _Vector<64, char>, unsigned long long int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def compresssf512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, unsigned short)">;
+ def compresssi512_mask : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, unsigned short)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cmpsd_mask : X86Builtin<"unsigned char(_Vector<2, double>, _Vector<2, double>, _Constant int, unsigned char, _Constant int)">;
+ def cmpss_mask : X86Builtin<"unsigned char(_Vector<4, float>, _Vector<4, float>, _Constant int, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def pshufd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Constant int)">;
+ def expanddf512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, unsigned char)">;
+ def expanddi512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def expandhi512_mask : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>, unsigned int)">;
+ def expandqi512_mask : X86Builtin<"_Vector<64, char>(_Vector<64, char>, _Vector<64, char>, unsigned long long int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def expandloaddf512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double const *>, _Vector<8, double>, unsigned char)">;
+ def expandloaddi512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int const *>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def expandloadhi512_mask : X86Builtin<"_Vector<32, short>(_Vector<32, short const *>, _Vector<32, short>, unsigned int)">;
+ def expandloadqi512_mask : X86Builtin<"_Vector<64, char>(_Vector<64, char const *>, _Vector<64, char>, unsigned long long int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def expandloadsf512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float const *>, _Vector<16, float>, unsigned short)">;
+ def expandloadsi512_mask : X86Builtin<"_Vector<16, int>(_Vector<16, int const *>, _Vector<16, int>, unsigned short)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def expandsf512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, unsigned short)">;
+ def expandsi512_mask : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, unsigned short)">;
+ def cvtps2pd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, float>, _Vector<8, double>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def compressstoredf512_mask : X86Builtin<"void(_Vector<8, double *>, _Vector<8, double>, unsigned char)">;
+ def compressstoredi512_mask : X86Builtin<"void(_Vector<8, long long int *>, _Vector<8, long long int>, unsigned char)">;
+}
+
+let Features = "avx512vbmi2,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def compressstorehi512_mask : X86Builtin<"void(_Vector<32, short *>, _Vector<32, short>, unsigned int)">;
+ def compressstoreqi512_mask : X86Builtin<"void(_Vector<64, char *>, _Vector<64, char>, unsigned long long int)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def compressstoresf512_mask : X86Builtin<"void(_Vector<16, float *>, _Vector<16, float>, unsigned short)">;
+ def compressstoresi512_mask : X86Builtin<"void(_Vector<16, int *>, _Vector<16, int>, unsigned short)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtph2ps_mask : X86Builtin<"_Vector<4, float>(_Vector<8, short>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtph2ps256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, short>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtps2ph_mask : X86Builtin<"_Vector<8, short>(_Vector<4, float>, _Constant int, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtps2ph256_mask : X86Builtin<"_Vector<8, short>(_Vector<8, float>, _Constant int, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def cvtw2mask512 : X86Builtin<"unsigned int(_Vector<32, short>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtw2mask128 : X86Builtin<"unsigned char(_Vector<8, short>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtw2mask256 : X86Builtin<"unsigned short(_Vector<16, short>)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtsd2ss_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<2, double>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def cvtsi2ss32 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, int, _Constant int)">;
+ def cvtss2sd_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<4, float>, _Vector<2, double>, unsigned char, _Constant int)">;
+ def cvtusi2ss32 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512vbmi,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vpmultishiftqb512 : X86Builtin<"_Vector<64, char>(_Vector<64, char>, _Vector<64, char>)">;
+}
+
+let Features = "avx512vbmi,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vpmultishiftqb128 : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>)">;
+}
+
+let Features = "avx512vbmi,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vpmultishiftqb256 : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>)">;
+}
+
+let Features = "avx512bf16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtne2ps2bf16_128 : X86Builtin<"_Vector<8, __bf16>(_Vector<4, float>, _Vector<4, float>)">;
+}
+
+let Features = "avx512bf16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtne2ps2bf16_256 : X86Builtin<"_Vector<16, __bf16>(_Vector<8, float>, _Vector<8, float>)">;
+}
+
+let Features = "avx512bf16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def cvtne2ps2bf16_512 : X86Builtin<"_Vector<32, __bf16>(_Vector<16, float>, _Vector<16, float>)">;
+}
+
+let Features = "avx512bf16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtneps2bf16_128_mask : X86Builtin<"_Vector<8, __bf16>(_Vector<4, float>, _Vector<8, __bf16>, unsigned char)">;
+}
+
+let Features = "avx512bf16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cvtneps2bf16_256_mask : X86Builtin<"_Vector<8, __bf16>(_Vector<8, float>, _Vector<8, __bf16>, unsigned char)">;
+}
+
+let Features = "avx512bf16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def cvtneps2bf16_512_mask : X86Builtin<"_Vector<16, __bf16>(_Vector<16, float>, _Vector<16, __bf16>, unsigned short)">;
+}
+
+let Features = "avx512bf16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def dpbf16ps_128 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<8, __bf16>, _Vector<8, __bf16>)">;
+}
+
+let Features = "avx512bf16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def dpbf16ps_256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<16, __bf16>, _Vector<16, __bf16>)">;
+}
+
+let Features = "avx512bf16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def dpbf16ps_512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<32, __bf16>, _Vector<32, __bf16>)">;
+}
+
+let Features = "avx512bf16", Attributes = [NoThrow, Const] in {
+ def cvtsbf162ss_32 : X86Builtin<"float(__bf16)">;
+}
+
+let Features = "avx512vp2intersect,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vp2intersect_q_512 : X86Builtin<"void(_Vector<8, long long int>, _Vector<8, long long int>, unsigned char *, unsigned char *)">;
+}
+
+let Features = "avx512vp2intersect,avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vp2intersect_q_256 : X86Builtin<"void(_Vector<4, long long int>, _Vector<4, long long int>, unsigned char *, unsigned char *)">;
+}
+
+let Features = "avx512vp2intersect,avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vp2intersect_q_128 : X86Builtin<"void(_Vector<2, long long int>, _Vector<2, long long int>, unsigned char *, unsigned char *)">;
+}
+
+let Features = "avx512vp2intersect,evex512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vp2intersect_d_512 : X86Builtin<"void(_Vector<16, int>, _Vector<16, int>, unsigned short *, unsigned short *)">;
+}
+
+let Features = "avx512vp2intersect,avx512vl", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vp2intersect_d_256 : X86Builtin<"void(_Vector<8, int>, _Vector<8, int>, unsigned char *, unsigned char *)">;
+}
+
+let Features = "avx512vp2intersect,avx512vl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vp2intersect_d_128 : X86Builtin<"void(_Vector<4, int>, _Vector<4, int>, unsigned char *, unsigned char *)">;
+}
+
+let Features = "avx512fp16", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcomish : X86Builtin<"int(_Vector<8, _Float16>, _Vector<8, _Float16>, _Constant int, _Constant int)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def addph512 : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Constant int)">;
+ def subph512 : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Constant int)">;
+ def mulph512 : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Constant int)">;
+ def divph512 : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Constant int)">;
+ def maxph512 : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Constant int)">;
+ def minph512 : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def minph256 : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def minph128 : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def maxph256 : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def maxph128 : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>)">;
+}
+
+let Features = "avx512fp16", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def addsh_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def divsh_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def mulsh_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def subsh_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def maxsh_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def minsh_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def cmpph512_mask : X86Builtin<"unsigned int(_Vector<32, _Float16>, _Vector<32, _Float16>, _Constant int, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def cmpph256_mask : X86Builtin<"unsigned short(_Vector<16, _Float16>, _Vector<16, _Float16>, _Constant int, unsigned short)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cmpph128_mask : X86Builtin<"unsigned char(_Vector<8, _Float16>, _Vector<8, _Float16>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512fp16", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cmpsh_mask : X86Builtin<"unsigned char(_Vector<8, _Float16>, _Vector<8, _Float16>, _Constant int, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def loadsh128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16 const *>, _Vector<8, _Float16>, unsigned char)">;
+ def storesh128_mask : X86Builtin<"void(_Vector<8, _Float16 *>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rcpph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def rcpph256_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def rcpph512_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, unsigned int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rsqrtph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def rsqrtph256_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def rsqrtph512_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, unsigned int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def getmantph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Constant int, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def getmantph256_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Constant int, _Vector<16, _Float16>, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def getmantph512_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Constant int, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def getexpph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def getexpph256_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def getexpph512_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def scalefph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def scalefph256_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def scalefph512_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rndscaleph_128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Constant int, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def rndscaleph_256_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Constant int, _Vector<16, _Float16>, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def rndscaleph_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Constant int, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def reduceph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Constant int, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def reduceph256_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Constant int, _Vector<16, _Float16>, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def reduceph512_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Constant int, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def rcpsh_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char)">;
+ def rsqrtsh_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char)">;
+ def getmantsh_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Constant int, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def getexpsh128_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def scalefsh_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def rndscalesh_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int, _Constant int)">;
+ def reducesh_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def sqrtph : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def sqrtph256 : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def sqrtph512 : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Constant int)">;
+}
+
+let Features = "avx512fp16", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def sqrtsh_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def fpclassph128_mask : X86Builtin<"unsigned char(_Vector<8, _Float16>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def fpclassph256_mask : X86Builtin<"unsigned short(_Vector<16, _Float16>, _Constant int, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def fpclassph512_mask : X86Builtin<"unsigned int(_Vector<32, _Float16>, _Constant int, unsigned int)">;
+}
+
+let Features = "avx512fp16", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def fpclasssh_mask : X86Builtin<"unsigned char(_Vector<8, _Float16>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtpd2ph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<2, double>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtpd2ph256_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<4, double>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtpd2ph512_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, double>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtph2pd128_mask : X86Builtin<"_Vector<2, double>(_Vector<8, _Float16>, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtph2pd256_mask : X86Builtin<"_Vector<4, double>(_Vector<8, _Float16>, _Vector<4, double>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtph2pd512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, _Float16>, _Vector<8, double>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtsh2ss_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<8, _Float16>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vcvtss2sh_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<4, float>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def vcvtsd2sh_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<2, double>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def vcvtsh2sd_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<8, _Float16>, _Vector<2, double>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtph2w128_mask : X86Builtin<"_Vector<8, short>(_Vector<8, _Float16>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtph2w256_mask : X86Builtin<"_Vector<16, short>(_Vector<16, _Float16>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtph2w512_mask : X86Builtin<"_Vector<32, short>(_Vector<32, _Float16>, _Vector<32, short>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvttph2w128_mask : X86Builtin<"_Vector<8, short>(_Vector<8, _Float16>, _Vector<8, short>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvttph2w256_mask : X86Builtin<"_Vector<16, short>(_Vector<16, _Float16>, _Vector<16, short>, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvttph2w512_mask : X86Builtin<"_Vector<32, short>(_Vector<32, _Float16>, _Vector<32, short>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtw2ph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, short>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtw2ph256_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, short>, _Vector<16, _Float16>, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtw2ph512_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<32, short>, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtph2uw128_mask : X86Builtin<"_Vector<8, unsigned short>(_Vector<8, _Float16>, _Vector<8, unsigned short>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtph2uw256_mask : X86Builtin<"_Vector<16, unsigned short>(_Vector<16, _Float16>, _Vector<16, unsigned short>, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtph2uw512_mask : X86Builtin<"_Vector<32, unsigned short>(_Vector<32, _Float16>, _Vector<32, unsigned short>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvttph2uw128_mask : X86Builtin<"_Vector<8, unsigned short>(_Vector<8, _Float16>, _Vector<8, unsigned short>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvttph2uw256_mask : X86Builtin<"_Vector<16, unsigned short>(_Vector<16, _Float16>, _Vector<16, unsigned short>, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvttph2uw512_mask : X86Builtin<"_Vector<32, unsigned short>(_Vector<32, _Float16>, _Vector<32, unsigned short>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtuw2ph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, unsigned short>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtuw2ph256_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, unsigned short>, _Vector<16, _Float16>, unsigned short)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtuw2ph512_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<32, unsigned short>, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtph2dq128_mask : X86Builtin<"_Vector<4, int>(_Vector<8, _Float16>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtph2dq256_mask : X86Builtin<"_Vector<8, int>(_Vector<8, _Float16>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtph2dq512_mask : X86Builtin<"_Vector<16, int>(_Vector<16, _Float16>, _Vector<16, int>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtph2udq128_mask : X86Builtin<"_Vector<4, unsigned int>(_Vector<8, _Float16>, _Vector<4, unsigned int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtph2udq256_mask : X86Builtin<"_Vector<8, unsigned int>(_Vector<8, _Float16>, _Vector<8, unsigned int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtph2udq512_mask : X86Builtin<"_Vector<16, unsigned int>(_Vector<16, _Float16>, _Vector<16, unsigned int>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtdq2ph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<4, int>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtdq2ph256_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, int>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtdq2ph512_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, int>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtudq2ph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<4, unsigned int>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtudq2ph256_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, unsigned int>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtudq2ph512_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, unsigned int>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvttph2dq128_mask : X86Builtin<"_Vector<4, int>(_Vector<8, _Float16>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvttph2dq256_mask : X86Builtin<"_Vector<8, int>(_Vector<8, _Float16>, _Vector<8, int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvttph2dq512_mask : X86Builtin<"_Vector<16, int>(_Vector<16, _Float16>, _Vector<16, int>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvttph2udq128_mask : X86Builtin<"_Vector<4, unsigned int>(_Vector<8, _Float16>, _Vector<4, unsigned int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvttph2udq256_mask : X86Builtin<"_Vector<8, unsigned int>(_Vector<8, _Float16>, _Vector<8, unsigned int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvttph2udq512_mask : X86Builtin<"_Vector<16, unsigned int>(_Vector<16, _Float16>, _Vector<16, unsigned int>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtqq2ph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<2, long long int>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtqq2ph256_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<4, long long int>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtqq2ph512_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, long long int>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtph2qq128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<8, _Float16>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtph2qq256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<8, _Float16>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtph2qq512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, _Float16>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtuqq2ph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<2, unsigned long long int>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtuqq2ph256_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<4, unsigned long long int>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtuqq2ph512_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, unsigned long long int>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtph2uqq128_mask : X86Builtin<"_Vector<2, unsigned long long int>(_Vector<8, _Float16>, _Vector<2, unsigned long long int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtph2uqq256_mask : X86Builtin<"_Vector<4, unsigned long long int>(_Vector<8, _Float16>, _Vector<4, unsigned long long int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtph2uqq512_mask : X86Builtin<"_Vector<8, unsigned long long int>(_Vector<8, _Float16>, _Vector<8, unsigned long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvttph2qq128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<8, _Float16>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvttph2qq256_mask : X86Builtin<"_Vector<4, long long int>(_Vector<8, _Float16>, _Vector<4, long long int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvttph2qq512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, _Float16>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvttph2uqq128_mask : X86Builtin<"_Vector<2, unsigned long long int>(_Vector<8, _Float16>, _Vector<2, unsigned long long int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvttph2uqq256_mask : X86Builtin<"_Vector<4, unsigned long long int>(_Vector<8, _Float16>, _Vector<4, unsigned long long int>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvttph2uqq512_mask : X86Builtin<"_Vector<8, unsigned long long int>(_Vector<8, _Float16>, _Vector<8, unsigned long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtsh2si32 : X86Builtin<"int(_Vector<8, _Float16>, _Constant int)">;
+ def vcvtsh2usi32 : X86Builtin<"unsigned int(_Vector<8, _Float16>, _Constant int)">;
+ def vcvtusi2sh : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, unsigned int, _Constant int)">;
+ def vcvtsi2sh : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, int, _Constant int)">;
+ def vcvttsh2si32 : X86Builtin<"int(_Vector<8, _Float16>, _Constant int)">;
+ def vcvttsh2usi32 : X86Builtin<"unsigned int(_Vector<8, _Float16>, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtph2psx128_mask : X86Builtin<"_Vector<4, float>(_Vector<8, _Float16>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtph2psx256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, _Float16>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtph2psx512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, _Float16>, _Vector<16, float>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtps2phx128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<4, float>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvtps2phx256_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, float>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvtps2phx512_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, float>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfmaddph : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vfmaddph256 : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Vector<16, _Float16>)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vfmaddph512_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+ def vfmaddph512_mask3 : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+ def vfmaddph512_maskz : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfmaddsubph : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vfmaddsubph256 : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Vector<16, _Float16>)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vfmaddsubph512_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+ def vfmaddsubph512_maskz : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+ def vfmaddsubph512_mask3 : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+ def vfmsubaddph512_mask3 : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+ def vfmsubph512_mask3 : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx512fp16", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfmaddsh3_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def vfmaddsh3_maskz : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def vfmaddsh3_mask3 : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def vfmsubsh3_mask3 : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfmaddcph128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char)">;
+ def vfmaddcph128_maskz : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vfmaddcph256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char)">;
+ def vfmaddcph256_maskz : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vfmaddcph512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def vfmaddcph512_maskz : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def vfmaddcph512_mask3 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfcmaddcph128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char)">;
+ def vfcmaddcph128_maskz : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vfcmaddcph256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char)">;
+ def vfcmaddcph256_maskz : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vfcmaddcph512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def vfcmaddcph512_maskz : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+ def vfcmaddcph512_mask3 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512fp16", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfmaddcsh_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vfmaddcsh_maskz : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vfcmaddcsh_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vfcmaddcsh_maskz : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vfmaddcsh_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vfmaddcsh_round_mask3 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vfcmaddcsh_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vfcmaddcsh_round_mask3 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vfmulcsh_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vfcmulcsh_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfmulcph128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vfmulcph256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vfmulcph512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfcmulcph128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vfcmulcph256_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vfcmulcph512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def selectb_128 : X86Builtin<"_Vector<16, char>(unsigned short, _Vector<16, char>, _Vector<16, char>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def selectb_256 : X86Builtin<"_Vector<32, char>(unsigned int, _Vector<32, char>, _Vector<32, char>)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def selectb_512 : X86Builtin<"_Vector<64, char>(unsigned long long int, _Vector<64, char>, _Vector<64, char>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def selectw_128 : X86Builtin<"_Vector<8, short>(unsigned char, _Vector<8, short>, _Vector<8, short>)">;
+}
+
+let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def selectw_256 : X86Builtin<"_Vector<16, short>(unsigned short, _Vector<16, short>, _Vector<16, short>)">;
+}
+
+let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def selectw_512 : X86Builtin<"_Vector<32, short>(unsigned int, _Vector<32, short>, _Vector<32, short>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def selectd_128 : X86Builtin<"_Vector<4, int>(unsigned char, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def selectd_256 : X86Builtin<"_Vector<8, int>(unsigned char, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def selectd_512 : X86Builtin<"_Vector<16, int>(unsigned short, _Vector<16, int>, _Vector<16, int>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def selectph_128 : X86Builtin<"_Vector<8, _Float16>(unsigned char, _Vector<8, _Float16>, _Vector<8, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def selectph_256 : X86Builtin<"_Vector<16, _Float16>(unsigned short, _Vector<16, _Float16>, _Vector<16, _Float16>)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def selectph_512 : X86Builtin<"_Vector<32, _Float16>(unsigned int, _Vector<32, _Float16>, _Vector<32, _Float16>)">;
+}
+
+let Features = "avx512bf16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def selectpbf_128 : X86Builtin<"_Vector<8, __bf16>(unsigned char, _Vector<8, __bf16>, _Vector<8, __bf16>)">;
+}
+
+let Features = "avx512bf16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def selectpbf_256 : X86Builtin<"_Vector<16, __bf16>(unsigned short, _Vector<16, __bf16>, _Vector<16, __bf16>)">;
+}
+
+let Features = "avx512bf16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def selectpbf_512 : X86Builtin<"_Vector<32, __bf16>(unsigned int, _Vector<32, __bf16>, _Vector<32, __bf16>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def selectq_128 : X86Builtin<"_Vector<2, long long int>(unsigned char, _Vector<2, long long int>, _Vector<2, long long int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def selectq_256 : X86Builtin<"_Vector<4, long long int>(unsigned char, _Vector<4, long long int>, _Vector<4, long long int>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def selectq_512 : X86Builtin<"_Vector<8, long long int>(unsigned char, _Vector<8, long long int>, _Vector<8, long long int>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def selectps_128 : X86Builtin<"_Vector<4, float>(unsigned char, _Vector<4, float>, _Vector<4, float>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def selectps_256 : X86Builtin<"_Vector<8, float>(unsigned char, _Vector<8, float>, _Vector<8, float>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def selectps_512 : X86Builtin<"_Vector<16, float>(unsigned short, _Vector<16, float>, _Vector<16, float>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def selectpd_128 : X86Builtin<"_Vector<2, double>(unsigned char, _Vector<2, double>, _Vector<2, double>)">;
+}
+
+let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def selectpd_256 : X86Builtin<"_Vector<4, double>(unsigned char, _Vector<4, double>, _Vector<4, double>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def selectpd_512 : X86Builtin<"_Vector<8, double>(unsigned char, _Vector<8, double>, _Vector<8, double>)">;
+}
+
+let Features = "avx512fp16", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def selectsh_128 : X86Builtin<"_Vector<8, _Float16>(unsigned char, _Vector<8, _Float16>, _Vector<8, _Float16>)">;
+}
+
+let Features = "avx512bf16", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def selectsbf_128 : X86Builtin<"_Vector<8, __bf16>(unsigned char, _Vector<8, __bf16>, _Vector<8, __bf16>)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def selectss_128 : X86Builtin<"_Vector<4, float>(unsigned char, _Vector<4, float>, _Vector<4, float>)">;
+ def selectsd_128 : X86Builtin<"_Vector<2, double>(unsigned char, _Vector<2, double>, _Vector<2, double>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def reduce_fadd_pd512 : X86Builtin<"double(double, _Vector<8, double>)">;
+ def reduce_fadd_ps512 : X86Builtin<"float(float, _Vector<16, float>)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def reduce_fadd_ph512 : X86Builtin<"_Float16(_Float16, _Vector<32, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def reduce_fadd_ph256 : X86Builtin<"_Float16(_Float16, _Vector<16, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def reduce_fadd_ph128 : X86Builtin<"_Float16(_Float16, _Vector<8, _Float16>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def reduce_fmax_pd512 : X86Builtin<"double(_Vector<8, double>)">;
+ def reduce_fmax_ps512 : X86Builtin<"float(_Vector<16, float>)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def reduce_fmax_ph512 : X86Builtin<"_Float16(_Vector<32, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def reduce_fmax_ph256 : X86Builtin<"_Float16(_Vector<16, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def reduce_fmax_ph128 : X86Builtin<"_Float16(_Vector<8, _Float16>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def reduce_fmin_pd512 : X86Builtin<"double(_Vector<8, double>)">;
+ def reduce_fmin_ps512 : X86Builtin<"float(_Vector<16, float>)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def reduce_fmin_ph512 : X86Builtin<"_Float16(_Vector<32, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def reduce_fmin_ph256 : X86Builtin<"_Float16(_Vector<16, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def reduce_fmin_ph128 : X86Builtin<"_Float16(_Vector<8, _Float16>)">;
+}
+
+let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def reduce_fmul_pd512 : X86Builtin<"double(double, _Vector<8, double>)">;
+ def reduce_fmul_ps512 : X86Builtin<"float(float, _Vector<16, float>)">;
+}
+
+let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def reduce_fmul_ph512 : X86Builtin<"_Float16(_Float16, _Vector<32, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def reduce_fmul_ph256 : X86Builtin<"_Float16(_Float16, _Vector<16, _Float16>)">;
+}
+
+let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def reduce_fmul_ph128 : X86Builtin<"_Float16(_Float16, _Vector<8, _Float16>)">;
+}
+
+let Features = "mwaitx", Attributes = [NoThrow] in {
+ def monitorx : X86Builtin<"void(void const *, unsigned int, unsigned int)">;
+ def mwaitx : X86Builtin<"void(unsigned int, unsigned int, unsigned int)">;
+}
+
+let Features = "waitpkg", Attributes = [NoThrow] in {
+ def umonitor : X86Builtin<"void(void const *)">;
+ def umwait : X86Builtin<"unsigned char(unsigned int, unsigned int, unsigned int)">;
+ def tpause : X86Builtin<"unsigned char(unsigned int, unsigned int, unsigned int)">;
+}
+
+let Features = "clzero", Attributes = [NoThrow] in {
+ def clzero : X86Builtin<"void(void *)">;
+}
+
+let Features = "cldemote", Attributes = [NoThrow] in {
+ def cldemote : X86Builtin<"void(void const *)">;
+}
+
+let Features = "movdiri", Attributes = [NoThrow] in {
+ def directstore_u32 : X86Builtin<"void(unsigned int *, unsigned int)">;
+}
+
+let Features = "movdir64b", Attributes = [NoThrow] in {
+ def movdir64b : X86Builtin<"void(void *, void const *)">;
+}
+
+let Features = "ptwrite", Attributes = [NoThrow] in {
+ def ptwrite32 : X86Builtin<"void(unsigned int)">;
+}
+
+let Features = "invpcid", Attributes = [NoThrow, Const] in {
+ def invpcid : X86Builtin<"void(unsigned int, void *)">;
+}
+
+let Features = "enqcmd", Attributes = [NoThrow] in {
+ def enqcmd : X86Builtin<"unsigned char(void *, void const *)">;
+ def enqcmds : X86Builtin<"unsigned char(void *, void const *)">;
+}
+
+let Features = "kl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def loadiwkey : X86Builtin<"void(_Vector<2, long long int>, _Vector<2, long long int>, _Vector<2, long long int>, unsigned int)">;
+ def encodekey128_u32 : X86Builtin<"unsigned int(unsigned int, _Vector<2, long long int>, void *)">;
+ def encodekey256_u32 : X86Builtin<"unsigned int(unsigned int, _Vector<2, long long int>, _Vector<2, long long int>, void *)">;
+ def aesenc128kl_u8 : X86Builtin<"unsigned char(_Vector<2, long long int *>, _Vector<2, long long int>, void const *)">;
+ def aesenc256kl_u8 : X86Builtin<"unsigned char(_Vector<2, long long int *>, _Vector<2, long long int>, void const *)">;
+ def aesdec128kl_u8 : X86Builtin<"unsigned char(_Vector<2, long long int *>, _Vector<2, long long int>, void const *)">;
+ def aesdec256kl_u8 : X86Builtin<"unsigned char(_Vector<2, long long int *>, _Vector<2, long long int>, void const *)">;
+}
+
+let Features = "kl,widekl", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def aesencwide128kl_u8 : X86Builtin<"unsigned char(_Vector<2, long long int *>, _Vector<2, long long int const *>, void const *)">;
+ def aesencwide256kl_u8 : X86Builtin<"unsigned char(_Vector<2, long long int *>, _Vector<2, long long int const *>, void const *)">;
+ def aesdecwide128kl_u8 : X86Builtin<"unsigned char(_Vector<2, long long int *>, _Vector<2, long long int const *>, void const *)">;
+ def aesdecwide256kl_u8 : X86Builtin<"unsigned char(_Vector<2, long long int *>, _Vector<2, long long int const *>, void const *)">;
+}
+
+let Features = "serialize", Attributes = [NoThrow] in {
+ def serialize : X86Builtin<"void()">;
+}
+
+let Features = "tsxldtrk", Attributes = [NoThrow] in {
+ def xsusldtrk : X86Builtin<"void()">;
+ def xresldtrk : X86Builtin<"void()">;
+}
+
+let Features = "raoint", Attributes = [NoThrow] in {
+ def aadd32 : X86Builtin<"void(void *, signed int)">;
+ def aand32 : X86Builtin<"void(void *, signed int)">;
+ def aor32 : X86Builtin<"void(void *, signed int)">;
+ def axor32 : X86Builtin<"void(void *, signed int)">;
+}
+
+let Header = "intrin.h", Languages = "ALL_MS_LANGUAGES", Attributes = [NoThrow, RequireDeclaration] in {
+ def _BitScanForward : X86LibBuiltin<"unsigned char(msuint32_t *, msuint32_t)">;
+ def _BitScanReverse : X86LibBuiltin<"unsigned char(msuint32_t *, msuint32_t)">;
+ def _ReadWriteBarrier : X86LibBuiltin<"void()">;
+ def _ReadBarrier : X86LibBuiltin<"void()">;
+ def _WriteBarrier : X86LibBuiltin<"void()">;
+ def __cpuid : X86LibBuiltin<"void(int *, int)">;
+ def __cpuidex : X86LibBuiltin<"void(int *, int, int)">;
+}
+
+let Header = "intrin.h", Languages = "ALL_MS_LANGUAGES", Attributes = [NoThrow, Const, RequireDeclaration] in {
+ def __emul : X86LibBuiltin<"long long int(int, int)">;
+ def __emulu : X86LibBuiltin<"unsigned long long int(unsigned int, unsigned int)">;
+}
+
+let Header = "intrin.h", Languages = "ALL_MS_LANGUAGES", Attributes = [NoThrow, RequireDeclaration] in {
+ def _AddressOfReturnAddress : X86LibBuiltin<"void *()">;
+ def __stosb : X86LibBuiltin<"void(unsigned char *, unsigned char, size_t)">;
+}
+
+let Header = "intrin.h", Languages = "ALL_MS_LANGUAGES", Attributes = [NoThrow, RequireDeclaration, NoReturn] in {
+ def __int2c : X86LibBuiltin<"void()">;
+ def __ud2 : X86LibBuiltin<"void()">;
+}
+
+let Header = "intrin.h", Languages = "ALL_MS_LANGUAGES", Attributes = [NoThrow, RequireDeclaration] in {
+ def __readfsbyte : X86LibBuiltin<"unsigned char(msuint32_t)">;
+ def __readfsword : X86LibBuiltin<"unsigned short(msuint32_t)">;
+ def __readfsdword : X86LibBuiltin<"msuint32_t(msuint32_t)">;
+ def __readfsqword : X86LibBuiltin<"unsigned long long int(msuint32_t)">;
+ def __readgsbyte : X86LibBuiltin<"unsigned char(msuint32_t)">;
+ def __readgsword : X86LibBuiltin<"unsigned short(msuint32_t)">;
+ def __readgsdword : X86LibBuiltin<"msuint32_t(msuint32_t)">;
+ def __readgsqword : X86LibBuiltin<"unsigned long long int(msuint32_t)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vdpphps128 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<8, _Float16>, _Vector<8, _Float16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vdpphps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<16, _Float16>, _Vector<16, _Float16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vdpphps512 : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<32, _Float16>, _Vector<32, _Float16>)">;
+ def vpdpbssd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+ def vpdpbssds512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+ def vpdpbsud512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+ def vpdpbsuds512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+ def vpdpbuud512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+ def vpdpbuuds512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vpdpwsud512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+ def vpdpwsuds512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+ def vpdpwusd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+ def vpdpwusds512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+ def vpdpwuud512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+ def vpdpwuuds512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Vector<16, int>, _Vector<16, int>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def mpsadbw512 : X86Builtin<"_Vector<32, short>(_Vector<64, char>, _Vector<64, char>, _Constant char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vaddpd256_round : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant int)">;
+ def vaddph256_round : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Constant int)">;
+ def vaddps256_round : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant int)">;
+ def vcmppd256_round_mask : X86Builtin<"unsigned char(_Vector<4, double>, _Vector<4, double>, _Constant int, unsigned char, _Constant int)">;
+ def vcmpph256_round_mask : X86Builtin<"unsigned short(_Vector<16, _Float16>, _Vector<16, _Float16>, _Constant int, unsigned short, _Constant int)">;
+ def vcmpps256_round_mask : X86Builtin<"unsigned char(_Vector<8, float>, _Vector<8, float>, _Constant int, unsigned char, _Constant int)">;
+ def vcvtdq2ph256_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, int>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def vcvtdq2ps256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, int>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vcvtpd2dq256_round_mask : X86Builtin<"_Vector<4, int>(_Vector<4, double>, _Vector<4, int>, unsigned char, _Constant int)">;
+ def vcvtpd2ph256_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<4, double>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def vcvtpd2ps256_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, double>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vcvtpd2qq256_round_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, double>, _Vector<4, long long int>, unsigned char, _Constant int)">;
+ def vcvtpd2udq256_round_mask : X86Builtin<"_Vector<4, unsigned int>(_Vector<4, double>, _Vector<4, unsigned int>, unsigned char, _Constant int)">;
+ def vcvtpd2uqq256_round_mask : X86Builtin<"_Vector<4, unsigned long long int>(_Vector<4, double>, _Vector<4, unsigned long long int>, unsigned char, _Constant int)">;
+ def vcvtph2dq256_round_mask : X86Builtin<"_Vector<8, int>(_Vector<8, _Float16>, _Vector<8, int>, unsigned char, _Constant int)">;
+ def vcvtph2pd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<8, _Float16>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vcvtph2psx256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, _Float16>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vcvtph2qq256_round_mask : X86Builtin<"_Vector<4, long long int>(_Vector<8, _Float16>, _Vector<4, long long int>, unsigned char, _Constant int)">;
+ def vcvtph2udq256_round_mask : X86Builtin<"_Vector<8, unsigned int>(_Vector<8, _Float16>, _Vector<8, unsigned int>, unsigned char, _Constant int)">;
+ def vcvtph2uqq256_round_mask : X86Builtin<"_Vector<4, unsigned long long int>(_Vector<8, _Float16>, _Vector<4, unsigned long long int>, unsigned char, _Constant int)">;
+ def vcvtph2uw256_round_mask : X86Builtin<"_Vector<16, unsigned short>(_Vector<16, _Float16>, _Vector<16, unsigned short>, unsigned short, _Constant int)">;
+ def vcvtph2w256_round_mask : X86Builtin<"_Vector<16, short>(_Vector<16, _Float16>, _Vector<16, short>, unsigned short, _Constant int)">;
+ def vcvtps2dq256_round_mask : X86Builtin<"_Vector<8, int>(_Vector<8, float>, _Vector<8, int>, unsigned char, _Constant int)">;
+ def vcvtps2pd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<4, float>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vcvtps2phx256_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, float>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def vcvtps2qq256_round_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, float>, _Vector<4, long long int>, unsigned char, _Constant int)">;
+ def vcvtps2udq256_round_mask : X86Builtin<"_Vector<8, unsigned int>(_Vector<8, float>, _Vector<8, unsigned int>, unsigned char, _Constant int)">;
+ def vcvtps2uqq256_round_mask : X86Builtin<"_Vector<4, unsigned long long int>(_Vector<4, float>, _Vector<4, unsigned long long int>, unsigned char, _Constant int)">;
+ def vcvtqq2pd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<4, long long int>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vcvtqq2ph256_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<4, long long int>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def vcvtqq2ps256_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, long long int>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vcvttpd2dq256_round_mask : X86Builtin<"_Vector<4, int>(_Vector<4, double>, _Vector<4, int>, unsigned char, _Constant int)">;
+ def vcvttpd2qq256_round_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, double>, _Vector<4, long long int>, unsigned char, _Constant int)">;
+ def vcvttpd2udq256_round_mask : X86Builtin<"_Vector<4, unsigned int>(_Vector<4, double>, _Vector<4, unsigned int>, unsigned char, _Constant int)">;
+ def vcvttpd2uqq256_round_mask : X86Builtin<"_Vector<4, unsigned long long int>(_Vector<4, double>, _Vector<4, unsigned long long int>, unsigned char, _Constant int)">;
+ def vcvttph2dq256_round_mask : X86Builtin<"_Vector<8, int>(_Vector<8, _Float16>, _Vector<8, int>, unsigned char, _Constant int)">;
+ def vcvttph2qq256_round_mask : X86Builtin<"_Vector<4, long long int>(_Vector<8, _Float16>, _Vector<4, long long int>, unsigned char, _Constant int)">;
+ def vcvttph2udq256_round_mask : X86Builtin<"_Vector<8, unsigned int>(_Vector<8, _Float16>, _Vector<8, unsigned int>, unsigned char, _Constant int)">;
+ def vcvttph2uqq256_round_mask : X86Builtin<"_Vector<4, unsigned long long int>(_Vector<8, _Float16>, _Vector<4, unsigned long long int>, unsigned char, _Constant int)">;
+ def vcvttph2uw256_round_mask : X86Builtin<"_Vector<16, unsigned short>(_Vector<16, _Float16>, _Vector<16, unsigned short>, unsigned short, _Constant int)">;
+ def vcvttph2w256_round_mask : X86Builtin<"_Vector<16, short>(_Vector<16, _Float16>, _Vector<16, short>, unsigned short, _Constant int)">;
+ def vcvttps2dq256_round_mask : X86Builtin<"_Vector<8, int>(_Vector<8, float>, _Vector<8, int>, unsigned char, _Constant int)">;
+ def vcvttps2qq256_round_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, float>, _Vector<4, long long int>, unsigned char, _Constant int)">;
+ def vcvttps2udq256_round_mask : X86Builtin<"_Vector<8, unsigned int>(_Vector<8, float>, _Vector<8, unsigned int>, unsigned char, _Constant int)">;
+ def vcvttps2uqq256_round_mask : X86Builtin<"_Vector<4, unsigned long long int>(_Vector<4, float>, _Vector<4, unsigned long long int>, unsigned char, _Constant int)">;
+ def vcvtudq2ph256_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, unsigned int>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def vcvtudq2ps256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, unsigned int>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vcvtuqq2pd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<4, unsigned long long int>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vcvtuqq2ph256_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<4, unsigned long long int>, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def vcvtuqq2ps256_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, unsigned long long int>, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vcvtuw2ph256_round_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, unsigned short>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vcvtw2ph256_round_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, short>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vdivpd256_round : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant int)">;
+ def vdivph256_round : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Constant int)">;
+ def vdivps256_round : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant int)">;
+ def vfcmaddcph256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfcmaddcph256_round_maskz : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfcmaddcph256_round_mask3 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfcmulcph256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfixupimmpd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, long long int>, _Constant int, unsigned char, _Constant int)">;
+ def vfixupimmpd256_round_maskz : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, long long int>, _Constant int, unsigned char, _Constant int)">;
+ def vfixupimmps256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, int>, _Constant int, unsigned char, _Constant int)">;
+ def vfixupimmps256_round_maskz : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, int>, _Constant int, unsigned char, _Constant int)">;
+ def vfmaddpd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vfmaddpd256_round_maskz : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vfmaddpd256_round_mask3 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vfmaddph256_round_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vfmaddph256_round_maskz : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vfmaddph256_round_mask3 : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vfmaddps256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfmaddps256_round_maskz : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfmaddps256_round_mask3 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfmaddcph256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfmaddcph256_round_maskz : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfmaddcph256_round_mask3 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfmaddsubpd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vfmaddsubpd256_round_maskz : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vfmaddsubpd256_round_mask3 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vfmaddsubph256_round_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vfmaddsubph256_round_maskz : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vfmaddsubph256_round_mask3 : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vfmaddsubps256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfmaddsubps256_round_maskz : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfmaddsubps256_round_mask3 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfmsubpd256_round_mask3 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vfmsubph256_round_mask3 : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vfmsubps256_round_mask3 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfmsubaddpd256_round_mask3 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vfmsubaddph256_round_mask3 : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vfmsubaddps256_round_mask3 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vfmulcph256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vgetexppd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vgetexpph256_round_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vgetexpps256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vgetmantpd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Constant int, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vgetmantph256_round_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Constant int, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vgetmantps256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Constant int, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vmaxpd256_round : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant int)">;
+ def vmaxph256_round : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Constant int)">;
+ def vmaxps256_round : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant int)">;
+ def vminpd256_round : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant int)">;
+ def vminph256_round : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Constant int)">;
+ def vminps256_round : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant int)">;
+ def vmulpd256_round : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant int)">;
+ def vmulph256_round : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Constant int)">;
+ def vmulps256_round : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant int)">;
+ def vrangepd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant int, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vrangeps256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant int, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vreducepd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Constant int, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vreduceph256_round_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Constant int, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vreduceps256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Constant int, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vrndscalepd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Constant int, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vrndscaleph256_round_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Constant int, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vrndscaleps256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Constant int, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vscalefpd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Vector<4, double>, unsigned char, _Constant int)">;
+ def vscalefph256_round_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+ def vscalefps256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Vector<8, float>, unsigned char, _Constant int)">;
+ def vsqrtpd256_round : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Constant int)">;
+ def vsqrtph256_round : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Constant int)">;
+ def vsqrtps256_round : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Constant int)">;
+ def vsubpd256_round : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant int)">;
+ def vsubph256_round : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Constant int)">;
+ def vsubps256_round : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant int)">;
+}
+
+let Features = "avxvnniint16|avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vpdpwsud128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avxvnniint16|avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vpdpwsud256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avxvnniint16|avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vpdpwsuds128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avxvnniint16|avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vpdpwsuds256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avxvnniint16|avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vpdpwusd128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avxvnniint16|avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vpdpwusd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avxvnniint16|avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vpdpwusds128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avxvnniint16|avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vpdpwusds256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avxvnniint16|avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vpdpwuud128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avxvnniint16|avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vpdpwuud256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avxvnniint16|avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vpdpwuuds128 : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Vector<4, int>, _Vector<4, int>)">;
+}
+
+let Features = "avxvnniint16|avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vpdpwuuds256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>, _Vector<8, int>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvttsd2sis32 : X86Builtin<"int(_Vector<2, double>, _Constant int)">;
+ def vcvttsd2usis32 : X86Builtin<"unsigned int(_Vector<2, double>, _Constant int)">;
+ def vcvttss2sis32 : X86Builtin<"int(_Vector<4, float>, _Constant int)">;
+ def vcvttss2usis32 : X86Builtin<"unsigned int(_Vector<4, float>, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttpd2dqs128_mask : X86Builtin<"_Vector<4, int>(_Vector<2, double>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttpd2dqs256_round_mask : X86Builtin<"_Vector<4, int>(_Vector<4, double>, _Vector<4, int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttpd2dqs512_round_mask : X86Builtin<"_Vector<8, int>(_Vector<8, double>, _Vector<8, int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttpd2udqs128_mask : X86Builtin<"_Vector<4, int>(_Vector<2, double>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttpd2udqs256_round_mask : X86Builtin<"_Vector<4, int>(_Vector<4, double>, _Vector<4, int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttpd2udqs512_round_mask : X86Builtin<"_Vector<8, int>(_Vector<8, double>, _Vector<8, int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttpd2qqs128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<2, double>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttpd2qqs256_round_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, double>, _Vector<4, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttpd2qqs512_round_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, double>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttpd2uqqs128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<2, double>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttpd2uqqs256_round_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, double>, _Vector<4, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttpd2uqqs512_round_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, double>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttps2dqs128_mask : X86Builtin<"_Vector<4, int>(_Vector<4, float>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttps2dqs256_round_mask : X86Builtin<"_Vector<8, int>(_Vector<8, float>, _Vector<8, int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttps2dqs512_round_mask : X86Builtin<"_Vector<16, int>(_Vector<16, float>, _Vector<16, int>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttps2udqs128_mask : X86Builtin<"_Vector<4, int>(_Vector<4, float>, _Vector<4, int>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttps2udqs256_round_mask : X86Builtin<"_Vector<8, int>(_Vector<8, float>, _Vector<8, int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttps2udqs512_round_mask : X86Builtin<"_Vector<16, int>(_Vector<16, float>, _Vector<16, int>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttps2qqs128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<4, float>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttps2qqs256_round_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, float>, _Vector<4, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttps2qqs512_round_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, float>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttps2uqqs128_mask : X86Builtin<"_Vector<2, long long int>(_Vector<4, float>, _Vector<2, long long int>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttps2uqqs256_round_mask : X86Builtin<"_Vector<4, long long int>(_Vector<4, float>, _Vector<4, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttps2uqqs512_round_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, float>, _Vector<8, long long int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vbcstnebf162ps128 : X86Builtin<"_Vector<4, float>(__bf16 const *)">;
+}
+
+let Features = "avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vbcstnebf162ps256 : X86Builtin<"_Vector<8, float>(__bf16 const *)">;
+}
+
+let Features = "avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vbcstnesh2ps128 : X86Builtin<"_Vector<4, float>(_Float16 const *)">;
+}
+
+let Features = "avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vbcstnesh2ps256 : X86Builtin<"_Vector<8, float>(_Float16 const *)">;
+}
+
+let Features = "avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtneebf162ps128 : X86Builtin<"_Vector<4, float>(_Vector<8, __bf16 const *>)">;
+}
+
+let Features = "avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtneebf162ps256 : X86Builtin<"_Vector<8, float>(_Vector<16, __bf16 const *>)">;
+}
+
+let Features = "avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtneeph2ps128 : X86Builtin<"_Vector<4, float>(_Vector<8, _Float16 const *>)">;
+}
+
+let Features = "avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtneeph2ps256 : X86Builtin<"_Vector<8, float>(_Vector<16, _Float16 const *>)">;
+}
+
+let Features = "avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtneobf162ps128 : X86Builtin<"_Vector<4, float>(_Vector<8, __bf16 const *>)">;
+}
+
+let Features = "avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtneobf162ps256 : X86Builtin<"_Vector<8, float>(_Vector<16, __bf16 const *>)">;
+}
+
+let Features = "avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtneoph2ps128 : X86Builtin<"_Vector<4, float>(_Vector<8, _Float16 const *>)">;
+}
+
+let Features = "avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtneoph2ps256 : X86Builtin<"_Vector<8, float>(_Vector<16, _Float16 const *>)">;
+}
+
+let Features = "avx512bf16,avx512vl|avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtneps2bf16128 : X86Builtin<"_Vector<8, __bf16>(_Vector<4, float>)">;
+}
+
+let Features = "avx512bf16,avx512vl|avxneconvert", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtneps2bf16256 : X86Builtin<"_Vector<8, __bf16>(_Vector<8, float>)">;
+}
+
+let Features = "sha512", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vsha512msg1 : X86Builtin<"_Vector<4, unsigned long long int>(_Vector<4, unsigned long long int>, _Vector<2, unsigned long long int>)">;
+ def vsha512msg2 : X86Builtin<"_Vector<4, unsigned long long int>(_Vector<4, unsigned long long int>, _Vector<4, unsigned long long int>)">;
+ def vsha512rnds2 : X86Builtin<"_Vector<4, unsigned long long int>(_Vector<4, unsigned long long int>, _Vector<4, unsigned long long int>, _Vector<2, unsigned long long int>)">;
+}
+
+let Header = "intrin.h", Languages = "ALL_MS_LANGUAGES", Attributes = [NoThrow, RequireDeclaration] in {
+ def _InterlockedAnd64 : X86LibBuiltin<"int64_t(int64_t volatile *, int64_t)">;
+ def _InterlockedDecrement64 : X86LibBuiltin<"int64_t(int64_t volatile *)">;
+ def _InterlockedExchange64 : X86LibBuiltin<"int64_t(int64_t volatile *, int64_t)">;
+ def _InterlockedExchangeAdd64 : X86LibBuiltin<"int64_t(int64_t volatile *, int64_t)">;
+ def _InterlockedExchangeSub64 : X86LibBuiltin<"int64_t(int64_t volatile *, int64_t)">;
+ def _InterlockedIncrement64 : X86LibBuiltin<"int64_t(int64_t volatile *)">;
+ def _InterlockedOr64 : X86LibBuiltin<"int64_t(int64_t volatile *, int64_t)">;
+ def _InterlockedXor64 : X86LibBuiltin<"int64_t(int64_t volatile *, int64_t)">;
+}
+
+let Features = "sm3", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vsm3msg1 : X86Builtin<"_Vector<4, unsigned int>(_Vector<4, unsigned int>, _Vector<4, unsigned int>, _Vector<4, unsigned int>)">;
+ def vsm3msg2 : X86Builtin<"_Vector<4, unsigned int>(_Vector<4, unsigned int>, _Vector<4, unsigned int>, _Vector<4, unsigned int>)">;
+ def vsm3rnds2 : X86Builtin<"_Vector<4, unsigned int>(_Vector<4, unsigned int>, _Vector<4, unsigned int>, _Vector<4, unsigned int>, _Constant unsigned int)">;
+}
+
+let Features = "sm4", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vsm4key4128 : X86Builtin<"_Vector<4, unsigned int>(_Vector<4, unsigned int>, _Vector<4, unsigned int>)">;
+}
+
+let Features = "sm4", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vsm4key4256 : X86Builtin<"_Vector<8, unsigned int>(_Vector<8, unsigned int>, _Vector<8, unsigned int>)">;
+}
+
+let Features = "sm4", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vsm4rnds4128 : X86Builtin<"_Vector<4, unsigned int>(_Vector<4, unsigned int>, _Vector<4, unsigned int>)">;
+}
+
+let Features = "sm4", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vsm4rnds4256 : X86Builtin<"_Vector<8, unsigned int>(_Vector<8, unsigned int>, _Vector<8, unsigned int>)">;
+}
+
+let Features = "avx10.2-512,sm4", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vsm4key4512 : X86Builtin<"_Vector<16, unsigned int>(_Vector<16, unsigned int>, _Vector<16, unsigned int>)">;
+ def vsm4rnds4512 : X86Builtin<"_Vector<16, unsigned int>(_Vector<16, unsigned int>, _Vector<16, unsigned int>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vminmaxnepbf16128 : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Vector<8, __bf16>, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vminmaxnepbf16256 : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Vector<16, __bf16>, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vminmaxnepbf16512 : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Vector<32, __bf16>, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vminmaxpd128_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Constant int, _Vector<2, double>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vminmaxpd256_round_mask : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<4, double>, _Constant int, _Vector<4, double>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vminmaxpd512_round_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, _Constant int, _Vector<8, double>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vminmaxph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Constant int, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vminmaxph256_round_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, _Float16>, _Vector<16, _Float16>, _Constant int, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vminmaxph512_round_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<32, _Float16>, _Vector<32, _Float16>, _Constant int, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vminmaxps128_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Constant int, _Vector<4, float>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vminmaxps256_round_mask : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<8, float>, _Constant int, _Vector<8, float>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vminmaxps512_round_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Constant int, _Vector<16, float>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vminmaxsd_round_mask : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Constant int, _Vector<2, double>, unsigned char, _Constant int)">;
+ def vminmaxsh_round_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, _Vector<8, _Float16>, _Constant int, _Vector<8, _Float16>, unsigned char, _Constant int)">;
+ def vminmaxss_round_mask : X86Builtin<"_Vector<4, float>(_Vector<4, float>, _Vector<4, float>, _Constant int, _Vector<4, float>, unsigned char, _Constant int)">;
+ def vcvtnebf162ibs128 : X86Builtin<"_Vector<8, unsigned short>(_Vector<8, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtnebf162ibs256 : X86Builtin<"_Vector<16, unsigned short>(_Vector<16, __bf16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtnebf162ibs512 : X86Builtin<"_Vector<32, unsigned short>(_Vector<32, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtnebf162iubs128 : X86Builtin<"_Vector<8, unsigned short>(_Vector<8, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtnebf162iubs256 : X86Builtin<"_Vector<16, unsigned short>(_Vector<16, __bf16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtnebf162iubs512 : X86Builtin<"_Vector<32, unsigned short>(_Vector<32, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtph2ibs128_mask : X86Builtin<"_Vector<8, unsigned short>(_Vector<8, _Float16>, _Vector<8, unsigned short>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtph2ibs256_mask : X86Builtin<"_Vector<16, unsigned short>(_Vector<16, _Float16>, _Vector<16, unsigned short>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtph2ibs512_mask : X86Builtin<"_Vector<32, unsigned short>(_Vector<32, _Float16>, _Vector<32, unsigned short>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtph2iubs128_mask : X86Builtin<"_Vector<8, unsigned short>(_Vector<8, _Float16>, _Vector<8, unsigned short>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtph2iubs256_mask : X86Builtin<"_Vector<16, unsigned short>(_Vector<16, _Float16>, _Vector<16, unsigned short>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtph2iubs512_mask : X86Builtin<"_Vector<32, unsigned short>(_Vector<32, _Float16>, _Vector<32, unsigned short>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtps2ibs128_mask : X86Builtin<"_Vector<4, unsigned int>(_Vector<4, float>, _Vector<4, unsigned int>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtps2ibs256_mask : X86Builtin<"_Vector<8, unsigned int>(_Vector<8, float>, _Vector<8, unsigned int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtps2ibs512_mask : X86Builtin<"_Vector<16, unsigned int>(_Vector<16, float>, _Vector<16, unsigned int>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtps2iubs128_mask : X86Builtin<"_Vector<4, unsigned int>(_Vector<4, float>, _Vector<4, unsigned int>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtps2iubs256_mask : X86Builtin<"_Vector<8, unsigned int>(_Vector<8, float>, _Vector<8, unsigned int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtps2iubs512_mask : X86Builtin<"_Vector<16, unsigned int>(_Vector<16, float>, _Vector<16, unsigned int>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttnebf162ibs128 : X86Builtin<"_Vector<8, unsigned short>(_Vector<8, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttnebf162ibs256 : X86Builtin<"_Vector<16, unsigned short>(_Vector<16, __bf16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttnebf162ibs512 : X86Builtin<"_Vector<32, unsigned short>(_Vector<32, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttnebf162iubs128 : X86Builtin<"_Vector<8, unsigned short>(_Vector<8, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttnebf162iubs256 : X86Builtin<"_Vector<16, unsigned short>(_Vector<16, __bf16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttnebf162iubs512 : X86Builtin<"_Vector<32, unsigned short>(_Vector<32, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttph2ibs128_mask : X86Builtin<"_Vector<8, unsigned short>(_Vector<8, _Float16>, _Vector<8, unsigned short>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttph2ibs256_mask : X86Builtin<"_Vector<16, unsigned short>(_Vector<16, _Float16>, _Vector<16, unsigned short>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttph2ibs512_mask : X86Builtin<"_Vector<32, unsigned short>(_Vector<32, _Float16>, _Vector<32, unsigned short>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttph2iubs128_mask : X86Builtin<"_Vector<8, unsigned short>(_Vector<8, _Float16>, _Vector<8, unsigned short>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttph2iubs256_mask : X86Builtin<"_Vector<16, unsigned short>(_Vector<16, _Float16>, _Vector<16, unsigned short>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttph2iubs512_mask : X86Builtin<"_Vector<32, unsigned short>(_Vector<32, _Float16>, _Vector<32, unsigned short>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttps2ibs128_mask : X86Builtin<"_Vector<4, unsigned int>(_Vector<4, float>, _Vector<4, unsigned int>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttps2ibs256_mask : X86Builtin<"_Vector<8, unsigned int>(_Vector<8, float>, _Vector<8, unsigned int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttps2ibs512_mask : X86Builtin<"_Vector<16, unsigned int>(_Vector<16, float>, _Vector<16, unsigned int>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvttps2iubs128_mask : X86Builtin<"_Vector<4, unsigned int>(_Vector<4, float>, _Vector<4, unsigned int>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvttps2iubs256_mask : X86Builtin<"_Vector<8, unsigned int>(_Vector<8, float>, _Vector<8, unsigned int>, unsigned char, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvttps2iubs512_mask : X86Builtin<"_Vector<16, unsigned int>(_Vector<16, float>, _Vector<16, unsigned int>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvt2ps2phx128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<4, float>, _Vector<4, float>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcvt2ps2phx256_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<8, float>, _Vector<8, float>, _Vector<16, _Float16>, unsigned short, _Constant int)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcvt2ps2phx512_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<16, float>, _Vector<16, float>, _Vector<32, _Float16>, unsigned int, _Constant int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtbiasph2bf8_128_mask : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<8, _Float16>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtbiasph2bf8_256_mask : X86Builtin<"_Vector<16, char>(_Vector<32, char>, _Vector<16, _Float16>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtbiasph2bf8_512_mask : X86Builtin<"_Vector<32, char>(_Vector<64, char>, _Vector<32, _Float16>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtbiasph2bf8s_128_mask : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<8, _Float16>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtbiasph2bf8s_256_mask : X86Builtin<"_Vector<16, char>(_Vector<32, char>, _Vector<16, _Float16>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtbiasph2bf8s_512_mask : X86Builtin<"_Vector<32, char>(_Vector<64, char>, _Vector<32, _Float16>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtbiasph2hf8_128_mask : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<8, _Float16>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtbiasph2hf8_256_mask : X86Builtin<"_Vector<16, char>(_Vector<32, char>, _Vector<16, _Float16>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtbiasph2hf8_512_mask : X86Builtin<"_Vector<32, char>(_Vector<64, char>, _Vector<32, _Float16>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtbiasph2hf8s_128_mask : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<8, _Float16>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtbiasph2hf8s_256_mask : X86Builtin<"_Vector<16, char>(_Vector<32, char>, _Vector<16, _Float16>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtbiasph2hf8s_512_mask : X86Builtin<"_Vector<32, char>(_Vector<64, char>, _Vector<32, _Float16>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtne2ph2bf8_128 : X86Builtin<"_Vector<16, char>(_Vector<8, _Float16>, _Vector<8, _Float16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtne2ph2bf8_256 : X86Builtin<"_Vector<32, char>(_Vector<16, _Float16>, _Vector<16, _Float16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtne2ph2bf8_512 : X86Builtin<"_Vector<64, char>(_Vector<32, _Float16>, _Vector<32, _Float16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtne2ph2bf8s_128 : X86Builtin<"_Vector<16, char>(_Vector<8, _Float16>, _Vector<8, _Float16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtne2ph2bf8s_256 : X86Builtin<"_Vector<32, char>(_Vector<16, _Float16>, _Vector<16, _Float16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtne2ph2bf8s_512 : X86Builtin<"_Vector<64, char>(_Vector<32, _Float16>, _Vector<32, _Float16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtne2ph2hf8_128 : X86Builtin<"_Vector<16, char>(_Vector<8, _Float16>, _Vector<8, _Float16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtne2ph2hf8_256 : X86Builtin<"_Vector<32, char>(_Vector<16, _Float16>, _Vector<16, _Float16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtne2ph2hf8_512 : X86Builtin<"_Vector<64, char>(_Vector<32, _Float16>, _Vector<32, _Float16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtne2ph2hf8s_128 : X86Builtin<"_Vector<16, char>(_Vector<8, _Float16>, _Vector<8, _Float16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtne2ph2hf8s_256 : X86Builtin<"_Vector<32, char>(_Vector<16, _Float16>, _Vector<16, _Float16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtne2ph2hf8s_512 : X86Builtin<"_Vector<64, char>(_Vector<32, _Float16>, _Vector<32, _Float16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvthf8_2ph128_mask : X86Builtin<"_Vector<8, _Float16>(_Vector<16, char>, _Vector<8, _Float16>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvthf8_2ph256_mask : X86Builtin<"_Vector<16, _Float16>(_Vector<16, char>, _Vector<16, _Float16>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvthf8_2ph512_mask : X86Builtin<"_Vector<32, _Float16>(_Vector<32, char>, _Vector<32, _Float16>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtneph2bf8_128_mask : X86Builtin<"_Vector<16, char>(_Vector<8, _Float16>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtneph2bf8_256_mask : X86Builtin<"_Vector<16, char>(_Vector<16, _Float16>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtneph2bf8_512_mask : X86Builtin<"_Vector<32, char>(_Vector<32, _Float16>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtneph2bf8s_128_mask : X86Builtin<"_Vector<16, char>(_Vector<8, _Float16>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtneph2bf8s_256_mask : X86Builtin<"_Vector<16, char>(_Vector<16, _Float16>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtneph2bf8s_512_mask : X86Builtin<"_Vector<32, char>(_Vector<32, _Float16>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtneph2hf8_128_mask : X86Builtin<"_Vector<16, char>(_Vector<8, _Float16>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtneph2hf8_256_mask : X86Builtin<"_Vector<16, char>(_Vector<16, _Float16>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtneph2hf8_512_mask : X86Builtin<"_Vector<32, char>(_Vector<32, _Float16>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vcvtneph2hf8s_128_mask : X86Builtin<"_Vector<16, char>(_Vector<8, _Float16>, _Vector<16, char>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vcvtneph2hf8s_256_mask : X86Builtin<"_Vector<16, char>(_Vector<16, _Float16>, _Vector<16, char>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vcvtneph2hf8s_512_mask : X86Builtin<"_Vector<32, char>(_Vector<32, _Float16>, _Vector<32, char>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def loadsbf16128_mask : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16 const *>, _Vector<8, __bf16>, unsigned char)">;
+ def storesbf16128_mask : X86Builtin<"void(_Vector<8, __bf16 *>, _Vector<8, __bf16>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vaddnepbf16128 : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Vector<8, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vaddnepbf16256 : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Vector<16, __bf16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vaddnepbf16512 : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Vector<32, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vdivnepbf16128 : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Vector<8, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vdivnepbf16256 : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Vector<16, __bf16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vdivnepbf16512 : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Vector<32, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vmaxpbf16128 : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Vector<8, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vmaxpbf16256 : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Vector<16, __bf16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vmaxpbf16512 : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Vector<32, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vminpbf16128 : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Vector<8, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vminpbf16256 : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Vector<16, __bf16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vminpbf16512 : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Vector<32, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vmulnepbf16128 : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Vector<8, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vmulnepbf16256 : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Vector<16, __bf16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vmulnepbf16512 : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Vector<32, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vsubnepbf16128 : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Vector<8, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vsubnepbf16256 : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Vector<16, __bf16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vsubnepbf16512 : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Vector<32, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcomsbf16eq : X86Builtin<"int(_Vector<8, __bf16>, _Vector<8, __bf16>)">;
+ def vcomsbf16lt : X86Builtin<"int(_Vector<8, __bf16>, _Vector<8, __bf16>)">;
+ def vcomsbf16neq : X86Builtin<"int(_Vector<8, __bf16>, _Vector<8, __bf16>)">;
+ def vcomsbf16ge : X86Builtin<"int(_Vector<8, __bf16>, _Vector<8, __bf16>)">;
+ def vcomsbf16gt : X86Builtin<"int(_Vector<8, __bf16>, _Vector<8, __bf16>)">;
+ def vcomsbf16le : X86Builtin<"int(_Vector<8, __bf16>, _Vector<8, __bf16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vcmppbf16512_mask : X86Builtin<"unsigned int(_Vector<32, __bf16>, _Vector<32, __bf16>, _Constant int, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vcmppbf16256_mask : X86Builtin<"unsigned short(_Vector<16, __bf16>, _Vector<16, __bf16>, _Constant int, unsigned short)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcmppbf16128_mask : X86Builtin<"unsigned char(_Vector<8, __bf16>, _Vector<8, __bf16>, _Constant int, unsigned char)">;
+ def vfpclasspbf16128_mask : X86Builtin<"unsigned char(_Vector<8, __bf16>, _Constant int, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vfpclasspbf16256_mask : X86Builtin<"unsigned short(_Vector<16, __bf16>, _Constant int, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vfpclasspbf16512_mask : X86Builtin<"unsigned int(_Vector<32, __bf16>, _Constant int, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vscalefpbf16128_mask : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Vector<8, __bf16>, _Vector<8, __bf16>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vscalefpbf16256_mask : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Vector<16, __bf16>, _Vector<16, __bf16>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vscalefpbf16512_mask : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Vector<32, __bf16>, _Vector<32, __bf16>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vrcppbf16128_mask : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Vector<8, __bf16>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vrcppbf16256_mask : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Vector<16, __bf16>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vrcppbf16512_mask : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Vector<32, __bf16>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vgetexppbf16128_mask : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Vector<8, __bf16>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vgetexppbf16256_mask : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Vector<16, __bf16>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vgetexppbf16512_mask : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Vector<32, __bf16>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vrsqrtpbf16128_mask : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Vector<8, __bf16>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vrsqrtpbf16256_mask : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Vector<16, __bf16>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vrsqrtpbf16512_mask : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Vector<32, __bf16>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vreducenepbf16128_mask : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Constant int, _Vector<8, __bf16>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vreducenepbf16256_mask : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Constant int, _Vector<16, __bf16>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vreducenepbf16512_mask : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Constant int, _Vector<32, __bf16>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vrndscalenepbf16_128_mask : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Constant int, _Vector<8, __bf16>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vrndscalenepbf16_256_mask : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Constant int, _Vector<16, __bf16>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vrndscalenepbf16_mask : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Constant int, _Vector<32, __bf16>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vgetmantpbf16128_mask : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Constant int, _Vector<8, __bf16>, unsigned char)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vgetmantpbf16256_mask : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Constant int, _Vector<16, __bf16>, unsigned short)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vgetmantpbf16512_mask : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Constant int, _Vector<32, __bf16>, unsigned int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vsqrtnepbf16 : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vsqrtnepbf16256 : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>)">;
+}
+
+let Features = "avx10.2-512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in {
+ def vsqrtnepbf16512 : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>)">;
+ def vfmaddnepbh512 : X86Builtin<"_Vector<32, __bf16>(_Vector<32, __bf16>, _Vector<32, __bf16>, _Vector<32, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vfmaddnepbh256 : X86Builtin<"_Vector<16, __bf16>(_Vector<16, __bf16>, _Vector<16, __bf16>, _Vector<16, __bf16>)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vfmaddnepbh128 : X86Builtin<"_Vector<8, __bf16>(_Vector<8, __bf16>, _Vector<8, __bf16>, _Vector<8, __bf16>)">;
+}
diff --git a/clang/include/clang/Basic/TargetBuiltins.h b/clang/include/clang/Basic/TargetBuiltins.h
index a14fd2c4b224d8..556332dd4b217b 100644
--- a/clang/include/clang/Basic/TargetBuiltins.h
+++ b/clang/include/clang/Basic/TargetBuiltins.h
@@ -124,8 +124,6 @@ namespace clang {
enum {
LastTIBuiltin = clang::Builtin::FirstTSBuiltin - 1,
#define BUILTIN(ID, TYPE, ATTRS) BI##ID,
-#include "clang/Basic/BuiltinsX86.def"
-#define BUILTIN(ID, TYPE, ATTRS) BI##ID,
#include "clang/Basic/BuiltinsX86.inc"
FirstX86_64Builtin,
LastX86CommonBuiltin = FirstX86_64Builtin - 1,
diff --git a/clang/lib/Basic/Targets/X86.cpp b/clang/lib/Basic/Targets/X86.cpp
index 1b16888a0711b5..7e5a5c78aa6b58 100644
--- a/clang/lib/Basic/Targets/X86.cpp
+++ b/clang/lib/Basic/Targets/X86.cpp
@@ -24,14 +24,6 @@ namespace clang {
namespace targets {
static constexpr Builtin::Info BuiltinInfoX86[] = {
-#define BUILTIN(ID, TYPE, ATTRS) \
- {#ID, TYPE, ATTRS, nullptr, HeaderDesc::NO_HEADER, ALL_LANGUAGES},
-#define TARGET_BUILTIN(ID, TYPE, ATTRS, FEATURE) \
- {#ID, TYPE, ATTRS, FEATURE, HeaderDesc::NO_HEADER, ALL_LANGUAGES},
-#define TARGET_HEADER_BUILTIN(ID, TYPE, ATTRS, HEADER, LANGS, FEATURE) \
- {#ID, TYPE, ATTRS, FEATURE, HeaderDesc::HEADER, LANGS},
-#include "clang/Basic/BuiltinsX86.def"
-
#define BUILTIN(ID, TYPE, ATTRS) \
{#ID, TYPE, ATTRS, nullptr, HeaderDesc::NO_HEADER, ALL_LANGUAGES},
#define TARGET_BUILTIN(ID, TYPE, ATTRS, FEATURE) \
diff --git a/clang/utils/TableGen/ClangBuiltinsEmitter.cpp b/clang/utils/TableGen/ClangBuiltinsEmitter.cpp
index 6c3604adc92b99..b08b02d9dcecaa 100644
--- a/clang/utils/TableGen/ClangBuiltinsEmitter.cpp
+++ b/clang/utils/TableGen/ClangBuiltinsEmitter.cpp
@@ -25,6 +25,7 @@ enum class BuiltinType {
LibBuiltin,
LangBuiltin,
TargetBuiltin,
+ TargetLibBuiltin,
};
class PrototypeParser {
@@ -108,9 +109,15 @@ class PrototypeParser {
} else if (T.consume_back("&")) {
ParseType(T);
Type += "&";
+ } else if (T.consume_front("long long")) {
+ Type += "O";
+ ParseType(T);
} else if (T.consume_front("long")) {
Type += "L";
ParseType(T);
+ } else if (T.consume_front("signed")) {
+ Type += "S";
+ ParseType(T);
} else if (T.consume_front("unsigned")) {
Type += "U";
ParseType(T);
@@ -155,6 +162,7 @@ class PrototypeParser {
.Case("__fp16", "h")
.Case("__int128_t", "LLLi")
.Case("_Float16", "x")
+ .Case("__bf16", "y")
.Case("bool", "b")
.Case("char", "c")
.Case("constant_CFString", "F")
@@ -262,6 +270,9 @@ void EmitBuiltinDef(raw_ostream &OS, StringRef Substitution,
case BuiltinType::TargetBuiltin:
OS << "TARGET_BUILTIN";
break;
+ case BuiltinType::TargetLibBuiltin:
+ OS << "TARGET_HEADER_BUILTIN";
+ break;
}
OS << "(" << Spelling;
@@ -279,6 +290,12 @@ void EmitBuiltinDef(raw_ostream &OS, StringRef Substitution,
OS << ", " << Builtin->getValueAsString("Languages");
break;
}
+ case BuiltinType::TargetLibBuiltin: {
+ OS << ", ";
+ HeaderNameParser{Builtin}.Print(OS);
+ OS << ", " << Builtin->getValueAsString("Languages");
+ [[fallthrough]];
+ }
case BuiltinType::TargetBuiltin:
OS << ", \"" << Builtin->getValueAsString("Features") << "\"";
break;
@@ -331,6 +348,8 @@ void EmitBuiltin(raw_ostream &OS, const Record *Builtin) {
BT = BuiltinType::AtomicBuiltin;
} else if (Builtin->isSubClassOf("LangBuiltin")) {
BT = BuiltinType::LangBuiltin;
+ } else if (Builtin->isSubClassOf("TargetLibBuiltin")) {
+ BT = BuiltinType::TargetLibBuiltin;
} else if (Builtin->isSubClassOf("TargetBuiltin")) {
BT = BuiltinType::TargetBuiltin;
} else if (Builtin->isSubClassOf("LibBuiltin")) {
@@ -367,6 +386,10 @@ void clang::EmitClangBuiltins(const RecordKeeper &Records, raw_ostream &OS) {
#if defined(BUILTIN) && !defined(TARGET_BUILTIN)
# define TARGET_BUILTIN(ID, TYPE, ATTRS, FEATURE) BUILTIN(ID, TYPE, ATTRS)
#endif
+
+#if defined(BUILTIN) && !defined(TARGET_HEADER_BUILTIN)
+# define TARGET_HEADER_BUILTIN(ID, TYPE, ATTRS, HEADER, LANG, FEATURE) BUILTIN(ID, TYPE, ATTRS)
+#endif
)c++";
// AtomicBuiltins are order dependent
@@ -390,5 +413,6 @@ void clang::EmitClangBuiltins(const RecordKeeper &Records, raw_ostream &OS) {
#undef LIBBUILTIN
#undef LANGBUILTIN
#undef TARGET_BUILTIN
+#undef TARGET_HEADER_BUILTIN
)c++";
}
>From 37ecefb81179a9f721aad09582444910cafd0914 Mon Sep 17 00:00:00 2001
From: Chandler Carruth <chandlerc at gmail.com>
Date: Tue, 24 Dec 2024 08:41:49 +0000
Subject: [PATCH 2/2] Bulk port 64-bit x86 builtins to TableGen
This PR follows #120831 (the PR contains both, only review the last
commit here as the other commit will be reviewed on the other PR).
Similar to that PR, this does a very mechanical port of X86 builtins to
TableGen. There is a *lot* of improvement available here to use TableGen
more effectively and collapse repeated structures. But those can now be
follow-up PRs that restructure *within* the `.td` file.
The current structure produces a file that exactly matches the original
X-macros except for the differences outlined in #120831:
- Horizontal whitespace
- `long long` types now use `long long` outside of OpenCL, but switch to
`long` in OpenCL (if relevant at all).
Otherwise, only the order of builtins change, and no tests regress.
---
clang/include/clang/Basic/BuiltinsX86.td | 17 +-
clang/include/clang/Basic/BuiltinsX86Base.td | 28 ++
clang/include/clang/Basic/BuiltinsX86_64.td | 485 +++++++++++++++++++
clang/include/clang/Basic/CMakeLists.txt | 4 +
clang/include/clang/Basic/TargetBuiltins.h | 2 +-
clang/lib/Basic/Targets/X86.cpp | 2 +-
6 files changed, 520 insertions(+), 18 deletions(-)
create mode 100644 clang/include/clang/Basic/BuiltinsX86Base.td
create mode 100644 clang/include/clang/Basic/BuiltinsX86_64.td
diff --git a/clang/include/clang/Basic/BuiltinsX86.td b/clang/include/clang/Basic/BuiltinsX86.td
index be2802f3908ff3..1f134ce24108ef 100644
--- a/clang/include/clang/Basic/BuiltinsX86.td
+++ b/clang/include/clang/Basic/BuiltinsX86.td
@@ -10,22 +10,7 @@
//
//===----------------------------------------------------------------------===//
-include "clang/Basic/BuiltinsBase.td"
-
-class X86Builtin<string prototype> : TargetBuiltin {
- let Spellings = ["__builtin_ia32_" # NAME];
- let Prototype = prototype;
-}
-
-class X86NoPrefixBuiltin<string prototype> : TargetBuiltin {
- let Spellings = [NAME];
- let Prototype = prototype;
-}
-
-class X86LibBuiltin<string prototype> : TargetLibBuiltin {
- let Spellings = [NAME];
- let Prototype = prototype;
-}
+include "clang/Basic/BuiltinsX86Base.td"
def rdpmc : X86Builtin<"unsigned long long int(int)">;
def rdtsc : X86Builtin<"unsigned long long int()">;
diff --git a/clang/include/clang/Basic/BuiltinsX86Base.td b/clang/include/clang/Basic/BuiltinsX86Base.td
new file mode 100644
index 00000000000000..b92b3f3fe0c77a
--- /dev/null
+++ b/clang/include/clang/Basic/BuiltinsX86Base.td
@@ -0,0 +1,28 @@
+//===--- BuiltinsX86Base.td - X86 Builtin function classes ------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// This file defines the X86-specific builtin function classes.
+//
+//===----------------------------------------------------------------------===//
+
+include "clang/Basic/BuiltinsBase.td"
+
+class X86Builtin<string prototype> : TargetBuiltin {
+ let Spellings = ["__builtin_ia32_" # NAME];
+ let Prototype = prototype;
+}
+
+class X86NoPrefixBuiltin<string prototype> : TargetBuiltin {
+ let Spellings = [NAME];
+ let Prototype = prototype;
+}
+
+class X86LibBuiltin<string prototype> : TargetLibBuiltin {
+ let Spellings = [NAME];
+ let Prototype = prototype;
+}
diff --git a/clang/include/clang/Basic/BuiltinsX86_64.td b/clang/include/clang/Basic/BuiltinsX86_64.td
new file mode 100644
index 00000000000000..a6c6ef80eac212
--- /dev/null
+++ b/clang/include/clang/Basic/BuiltinsX86_64.td
@@ -0,0 +1,485 @@
+//===--- BuiltinsX86_64.td - X86-64 Builtin function database ---*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// This file defines the X86-64-specific builtin function database.
+//
+//===----------------------------------------------------------------------===//
+
+include "clang/Basic/BuiltinsX86Base.td"
+
+let Header = "intrin.h", Languages = "ALL_MS_LANGUAGES", Attributes = [NoThrow, RequireDeclaration] in {
+ def _BitScanForward64 : X86LibBuiltin<"unsigned char(msuint32_t *, unsigned long long int)">;
+ def _BitScanReverse64 : X86LibBuiltin<"unsigned char(msuint32_t *, unsigned long long int)">;
+}
+
+let Header = "intrin.h", Languages = "ALL_MS_LANGUAGES", Attributes = [NoThrow, Const, RequireDeclaration] in {
+ def __mulh : X86LibBuiltin<"long long int(long long int, long long int)">;
+ def __umulh : X86LibBuiltin<"unsigned long long int(unsigned long long int, unsigned long long int)">;
+ def _mul128 : X86LibBuiltin<"long long int(long long int, long long int, long long int *)">;
+ def _umul128 : X86LibBuiltin<"unsigned long long int(unsigned long long int, unsigned long long int, unsigned long long int *)">;
+}
+
+let Header = "intrin.h", Languages = "ALL_MS_LANGUAGES", Attributes = [NoThrow, RequireDeclaration] in {
+ def __faststorefence : X86LibBuiltin<"void()">;
+}
+
+let Header = "intrin.h", Languages = "ALL_MS_LANGUAGES", Attributes = [NoThrow, Const, RequireDeclaration] in {
+ def __shiftleft128 : X86LibBuiltin<"unsigned long long int(unsigned long long int, unsigned long long int, unsigned char)">;
+ def __shiftright128 : X86LibBuiltin<"unsigned long long int(unsigned long long int, unsigned long long int, unsigned char)">;
+}
+
+let Features = "cx16", Header = "intrin.h", Languages = "ALL_MS_LANGUAGES", Attributes = [NoThrow, RequireDeclaration] in {
+ def _InterlockedCompareExchange128 : X86LibBuiltin<"unsigned char(long long int volatile *, long long int, long long int, long long int *)">;
+}
+
+let Attributes = [NoThrow] in {
+ def readeflags_u64 : X86Builtin<"unsigned long long int()">;
+ def writeeflags_u64 : X86Builtin<"void(unsigned long long int)">;
+}
+
+let Features = "sse", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtss2si64 : X86Builtin<"long long int(_Vector<4, float>)">;
+ def cvttss2si64 : X86Builtin<"long long int(_Vector<4, float>)">;
+}
+
+let Features = "sse2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def cvtsd2si64 : X86Builtin<"long long int(_Vector<2, double>)">;
+ def cvttsd2si64 : X86Builtin<"long long int(_Vector<2, double>)">;
+}
+
+let Features = "sse2", Attributes = [NoThrow] in {
+ def movnti64 : X86Builtin<"void(long long int *, long long int)">;
+}
+
+let Features = "sse4.1", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vec_set_v2di : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, long long int, _Constant int)">;
+}
+
+let Features = "crc32", Attributes = [NoThrow, Const] in {
+ def crc32di : X86Builtin<"unsigned long long int(unsigned long long int, unsigned long long int)">;
+}
+
+let Features = "avx", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
+ def vec_ext_v4di : X86Builtin<"long long int(_Vector<4, long long int>, _Constant int)">;
+ def vec_set_v4di : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, long long int, _Constant int)">;
+}
+
+let Features = "fsgsbase", Attributes = [NoThrow] in {
+ def rdfsbase32 : X86Builtin<"unsigned int()">;
+ def rdfsbase64 : X86Builtin<"unsigned long long int()">;
+ def rdgsbase32 : X86Builtin<"unsigned int()">;
+ def rdgsbase64 : X86Builtin<"unsigned long long int()">;
+ def wrfsbase32 : X86Builtin<"void(unsigned int)">;
+ def wrfsbase64 : X86Builtin<"void(unsigned long long int)">;
+ def wrgsbase32 : X86Builtin<"void(unsigned int)">;
+ def wrgsbase64 : X86Builtin<"void(unsigned long long int)">;
+}
+
+let Features = "fxsr", Attributes = [NoThrow] in {
+ def fxrstor64 : X86Builtin<"void(void *)">;
+ def fxsave64 : X86Builtin<"void(void *)">;
+}
+
+let Features = "xsave", Attributes = [NoThrow] in {
+ def xsave64 : X86Builtin<"void(void *, unsigned long long int)">;
+ def xrstor64 : X86Builtin<"void(void *, unsigned long long int)">;
+}
+
+let Features = "xsaveopt", Attributes = [NoThrow] in {
+ def xsaveopt64 : X86Builtin<"void(void *, unsigned long long int)">;
+}
+
+let Features = "xsaves", Attributes = [NoThrow] in {
+ def xrstors64 : X86Builtin<"void(void *, unsigned long long int)">;
+}
+
+let Features = "xsavec", Attributes = [NoThrow] in {
+ def xsavec64 : X86Builtin<"void(void *, unsigned long long int)">;
+}
+
+let Features = "xsaves", Attributes = [NoThrow] in {
+ def xsaves64 : X86Builtin<"void(void *, unsigned long long int)">;
+}
+
+let Features = "shstk", Attributes = [NoThrow] in {
+ def incsspq : X86Builtin<"void(unsigned long long int)">;
+ def rdsspq : X86Builtin<"unsigned long long int(unsigned long long int)">;
+ def wrssq : X86Builtin<"void(unsigned long long int, void *)">;
+ def wrussq : X86Builtin<"void(unsigned long long int, void *)">;
+}
+
+let Attributes = [NoThrow, Constexpr] in {
+ def addcarryx_u64 : X86Builtin<"unsigned char(unsigned char, unsigned long long int, unsigned long long int, unsigned long long int *)">;
+ def subborrow_u64 : X86Builtin<"unsigned char(unsigned char, unsigned long long int, unsigned long long int, unsigned long long int *)">;
+}
+
+let Features = "rdrnd", Attributes = [NoThrow] in {
+ def rdrand64_step : X86Builtin<"unsigned int(unsigned long long int *)">;
+}
+
+let Features = "rdseed", Attributes = [NoThrow] in {
+ def rdseed64_step : X86Builtin<"unsigned int(unsigned long long int *)">;
+}
+
+let Features = "lzcnt", Attributes = [NoThrow, Const, Constexpr] in {
+ def lzcnt_u64 : X86Builtin<"unsigned long long int(unsigned long long int)">;
+}
+
+let Features = "bmi", Attributes = [NoThrow, Const, Constexpr] in {
+ def bextr_u64 : X86Builtin<"unsigned long long int(unsigned long long int, unsigned long long int)">;
+}
+
+let Attributes = [NoThrow, Const, Constexpr] in {
+ def tzcnt_u64 : X86Builtin<"unsigned long long int(unsigned long long int)">;
+}
+
+let Features = "bmi2", Attributes = [NoThrow, Const, Constexpr] in {
+ def bzhi_di : X86Builtin<"unsigned long long int(unsigned long long int, unsigned long long int)">;
+ def pdep_di : X86Builtin<"unsigned long long int(unsigned long long int, unsigned long long int)">;
+ def pext_di : X86Builtin<"unsigned long long int(unsigned long long int, unsigned long long int)">;
+}
+
+let Features = "tbm", Attributes = [NoThrow, Const, Constexpr] in {
+ def bextri_u64 : X86Builtin<"unsigned long long int(unsigned long long int, _Constant unsigned long long int)">;
+}
+
+let Features = "lwp", Attributes = [NoThrow] in {
+ def lwpins64 : X86Builtin<"unsigned char(unsigned long long int, unsigned int, _Constant unsigned int)">;
+ def lwpval64 : X86Builtin<"void(unsigned long long int, unsigned int, _Constant unsigned int)">;
+}
+
+let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtsd2si64 : X86Builtin<"long long int(_Vector<2, double>, _Constant int)">;
+ def vcvtsd2usi64 : X86Builtin<"unsigned long long int(_Vector<2, double>, _Constant int)">;
+ def vcvtss2si64 : X86Builtin<"long long int(_Vector<4, float>, _Constant int)">;
+ def vcvtss2usi64 : X86Builtin<"unsigned long long int(_Vector<4, float>, _Constant int)">;
+ def vcvttsd2si64 : X86Builtin<"long long int(_Vector<2, double>, _Constant int)">;
+ def vcvttsd2usi64 : X86Builtin<"unsigned long long int(_Vector<2, double>, _Constant int)">;
+ def vcvttss2si64 : X86Builtin<"long long int(_Vector<4, float>, _Constant int)">;
+ def vcvttss2usi64 : X86Builtin<"unsigned long long int(_Vector<4, float>, _Constant int)">;
+ def cvtsi2sd64 : X86Builtin<"_Vector<2, double>(_Vector<2, double>, long long int, _Constant int)">;
+ def cvtsi2ss64 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, long long int, _Constant int)">;
+ def cvtusi2sd64 : X86Builtin<"_Vector<2, double>(_Vector<2, double>, unsigned long long int, _Constant int)">;
+ def cvtusi2ss64 : X86Builtin<"_Vector<4, float>(_Vector<4, float>, unsigned long long int, _Constant int)">;
+}
+
+let Features = "avx512fp16", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvtsh2si64 : X86Builtin<"long long int(_Vector<8, _Float16>, _Constant int)">;
+ def vcvtsh2usi64 : X86Builtin<"unsigned long long int(_Vector<8, _Float16>, _Constant int)">;
+ def vcvtusi642sh : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, unsigned long long int, _Constant int)">;
+ def vcvtsi642sh : X86Builtin<"_Vector<8, _Float16>(_Vector<8, _Float16>, long long int, _Constant int)">;
+ def vcvttsh2si64 : X86Builtin<"long long int(_Vector<8, _Float16>, _Constant int)">;
+ def vcvttsh2usi64 : X86Builtin<"unsigned long long int(_Vector<8, _Float16>, _Constant int)">;
+}
+
+let Features = "movdiri", Attributes = [NoThrow] in {
+ def directstore_u64 : X86Builtin<"void(unsigned long int *, unsigned long int)">;
+}
+
+let Features = "avx10.2-256", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in {
+ def vcvttsd2sis64 : X86Builtin<"long long int(_Vector<2, double>, _Constant int)">;
+ def vcvttsd2usis64 : X86Builtin<"unsigned long long int(_Vector<2, double>, _Constant int)">;
+ def vcvttss2sis64 : X86Builtin<"long long int(_Vector<4, float>, _Constant int)">;
+ def vcvttss2usis64 : X86Builtin<"unsigned long long int(_Vector<4, float>, _Constant int)">;
+}
+
+let Features = "uintr", Attributes = [NoThrow] in {
+ def clui : X86Builtin<"void()">;
+ def stui : X86Builtin<"void()">;
+ def testui : X86Builtin<"unsigned char()">;
+ def senduipi : X86Builtin<"void(uint64_t)">;
+}
+
+let Features = "usermsr", Attributes = [NoThrow] in {
+ def urdmsr : X86Builtin<"unsigned long long int(unsigned long long int)">;
+ def uwrmsr : X86Builtin<"void(unsigned long long int, unsigned long long int)">;
+}
+
+let Features = "amx-tile", Attributes = [NoThrow] in {
+ def tile_loadconfig_internal : X86Builtin<"void(void const *)">;
+ def tileloadd64_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, void const *, size_t)">;
+}
+
+let Features = "amx-movrs", Attributes = [NoThrow] in {
+ def tileloaddrs64_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, void const *, size_t)">;
+}
+
+let Features = "amx-tile", Attributes = [NoThrow] in {
+ def tileloaddt164_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, void const *, size_t)">;
+}
+
+let Features = "amx-movrs", Attributes = [NoThrow] in {
+ def tileloaddrst164_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, void const *, size_t)">;
+}
+
+let Features = "amx-int8", Attributes = [NoThrow] in {
+ def tdpbssd_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+ def tdpbsud_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+ def tdpbusd_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+ def tdpbuud_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+}
+
+let Features = "amx-tile", Attributes = [NoThrow] in {
+ def tilestored64_internal : X86Builtin<"void(unsigned short, unsigned short, void *, size_t, _Vector<256, int>)">;
+ def tilezero_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short)">;
+}
+
+let Features = "amx-bf16", Attributes = [NoThrow] in {
+ def tdpbf16ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+}
+
+let Features = "amx-fp16", Attributes = [NoThrow] in {
+ def tdpfp16ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+}
+
+let Features = "amx-complex", Attributes = [NoThrow] in {
+ def tcmmimfp16ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+ def tcmmrlfp16ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+}
+
+let Features = "amx-transpose", Attributes = [NoThrow] in {
+ def t2rpntlvwz0_internal : X86Builtin<"void(unsigned short, unsigned short, unsigned short, _Vector<256, int *>, _Vector<256, int *>, void const *, size_t)">;
+}
+
+let Features = "amx-movrs,amx-transpose", Attributes = [NoThrow] in {
+ def t2rpntlvwz0rs_internal : X86Builtin<"void(unsigned short, unsigned short, unsigned short, _Vector<256, int *>, _Vector<256, int *>, void const *, size_t)">;
+}
+
+let Features = "amx-transpose", Attributes = [NoThrow] in {
+ def t2rpntlvwz0t1_internal : X86Builtin<"void(unsigned short, unsigned short, unsigned short, _Vector<256, int *>, _Vector<256, int *>, void const *, size_t)">;
+}
+
+let Features = "amx-movrs,amx-transpose", Attributes = [NoThrow] in {
+ def t2rpntlvwz0rst1_internal : X86Builtin<"void(unsigned short, unsigned short, unsigned short, _Vector<256, int *>, _Vector<256, int *>, void const *, size_t)">;
+}
+
+let Features = "amx-transpose", Attributes = [NoThrow] in {
+ def t2rpntlvwz1_internal : X86Builtin<"void(unsigned short, unsigned short, unsigned short, _Vector<256, int *>, _Vector<256, int *>, void const *, size_t)">;
+}
+
+let Features = "amx-movrs,amx-transpose", Attributes = [NoThrow] in {
+ def t2rpntlvwz1rs_internal : X86Builtin<"void(unsigned short, unsigned short, unsigned short, _Vector<256, int *>, _Vector<256, int *>, void const *, size_t)">;
+}
+
+let Features = "amx-transpose", Attributes = [NoThrow] in {
+ def t2rpntlvwz1t1_internal : X86Builtin<"void(unsigned short, unsigned short, unsigned short, _Vector<256, int *>, _Vector<256, int *>, void const *, size_t)">;
+}
+
+let Features = "amx-movrs,amx-transpose", Attributes = [NoThrow] in {
+ def t2rpntlvwz1rst1_internal : X86Builtin<"void(unsigned short, unsigned short, unsigned short, _Vector<256, int *>, _Vector<256, int *>, void const *, size_t)">;
+}
+
+let Features = "amx-transpose", Attributes = [NoThrow] in {
+ def ttransposed_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, _Vector<256, int>)">;
+}
+
+let Features = "amx-bf16,amx-transpose", Attributes = [NoThrow] in {
+ def ttdpbf16ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+}
+
+let Features = "amx-fp16,amx-transpose", Attributes = [NoThrow] in {
+ def ttdpfp16ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+}
+
+let Features = "amx-complex,amx-transpose", Attributes = [NoThrow] in {
+ def ttcmmimfp16ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+ def ttcmmrlfp16ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+ def tconjtcmmimfp16ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+ def tconjtfp16_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, _Vector<256, int>)">;
+}
+
+let Features = "amx-avx512,avx10.2-512", Attributes = [NoThrow] in {
+ def tcvtrowd2ps_internal : X86Builtin<"_Vector<16, float>(unsigned short, unsigned short, _Vector<256, int>, unsigned int)">;
+ def tcvtrowps2pbf16h_internal : X86Builtin<"_Vector<32, __bf16>(unsigned short, unsigned short, _Vector<256, int>, unsigned int)">;
+ def tcvtrowps2pbf16l_internal : X86Builtin<"_Vector<32, __bf16>(unsigned short, unsigned short, _Vector<256, int>, unsigned int)">;
+ def tcvtrowps2phh_internal : X86Builtin<"_Vector<32, _Float16>(unsigned short, unsigned short, _Vector<256, int>, unsigned int)">;
+ def tcvtrowps2phl_internal : X86Builtin<"_Vector<32, _Float16>(unsigned short, unsigned short, _Vector<256, int>, unsigned int)">;
+ def tilemovrow_internal : X86Builtin<"_Vector<16, int>(unsigned short, unsigned short, _Vector<256, int>, unsigned int)">;
+}
+
+let Features = "amx-tf32", Attributes = [NoThrow] in {
+ def tmmultf32ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+}
+
+let Features = "amx-tf32,amx-transpose", Attributes = [NoThrow] in {
+ def ttmmultf32ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+}
+
+let Features = "amx-fp8", Attributes = [NoThrow] in {
+ def tdpbf8ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+ def tdpbhf8ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+ def tdphbf8ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+ def tdphf8ps_internal : X86Builtin<"_Vector<256, int>(unsigned short, unsigned short, unsigned short, _Vector<256, int>, _Vector<256, int>, _Vector<256, int>)">;
+}
+
+let Features = "amx-tile", Attributes = [NoThrow] in {
+ def tile_loadconfig : X86Builtin<"void(void const *)">;
+ def tile_storeconfig : X86Builtin<"void(void const *)">;
+ def tilerelease : X86Builtin<"void()">;
+ def tilezero : X86Builtin<"void(unsigned char)">;
+}
+
+let Features = "amx-movrs,amx-transpose", Attributes = [NoThrow] in {
+ def t2rpntlvwz0rs : X86Builtin<"void(_Constant unsigned char, void const *, size_t)">;
+ def t2rpntlvwz0rst1 : X86Builtin<"void(_Constant unsigned char, void const *, size_t)">;
+ def t2rpntlvwz1rs : X86Builtin<"void(_Constant unsigned char, void const *, size_t)">;
+ def t2rpntlvwz1rst1 : X86Builtin<"void(_Constant unsigned char, void const *, size_t)">;
+}
+
+let Features = "amx-movrs", Attributes = [NoThrow] in {
+ def tileloaddrs64 : X86Builtin<"void(_Constant unsigned char, void const *, size_t)">;
+ def tileloaddrst164 : X86Builtin<"void(_Constant unsigned char, void const *, size_t)">;
+}
+
+let Features = "amx-tile", Attributes = [NoThrow] in {
+ def tileloadd64 : X86Builtin<"void(_Constant unsigned char, void const *, size_t)">;
+ def tileloaddt164 : X86Builtin<"void(_Constant unsigned char, void const *, size_t)">;
+ def tilestored64 : X86Builtin<"void(_Constant unsigned char, void *, size_t)">;
+}
+
+let Features = "amx-int8", Attributes = [NoThrow] in {
+ def tdpbssd : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+ def tdpbsud : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+ def tdpbusd : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+ def tdpbuud : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+}
+
+let Features = "amx-bf16", Attributes = [NoThrow] in {
+ def tdpbf16ps : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+}
+
+let Features = "ptwrite", Attributes = [NoThrow] in {
+ def ptwrite64 : X86Builtin<"void(unsigned long long int)">;
+}
+
+let Features = "amx-complex", Attributes = [NoThrow] in {
+ def tcmmimfp16ps : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+ def tcmmrlfp16ps : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+}
+
+let Features = "amx-transpose", Attributes = [NoThrow] in {
+ def t2rpntlvwz0 : X86Builtin<"void(_Constant unsigned char, void const *, size_t)">;
+ def t2rpntlvwz0t1 : X86Builtin<"void(_Constant unsigned char, void const *, size_t)">;
+ def t2rpntlvwz1 : X86Builtin<"void(_Constant unsigned char, void const *, size_t)">;
+ def t2rpntlvwz1t1 : X86Builtin<"void(_Constant unsigned char, void const *, size_t)">;
+ def ttransposed : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char)">;
+}
+
+let Features = "amx-bf16,amx-transpose", Attributes = [NoThrow] in {
+ def ttdpbf16ps : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+}
+
+let Features = "amx-fp16,amx-transpose", Attributes = [NoThrow] in {
+ def ttdpfp16ps : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+}
+
+let Features = "amx-complex,amx-transpose", Attributes = [NoThrow] in {
+ def ttcmmimfp16ps : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+ def ttcmmrlfp16ps : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+ def tconjtcmmimfp16ps : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+ def tconjtfp16 : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char)">;
+}
+
+let Features = "amx-avx512,avx10.2-512", Attributes = [NoThrow] in {
+ def tcvtrowd2ps : X86Builtin<"_Vector<16, float>(_Constant unsigned char, unsigned int)">;
+ def tcvtrowps2pbf16h : X86Builtin<"_Vector<32, __bf16>(_Constant unsigned char, unsigned int)">;
+ def tcvtrowps2pbf16l : X86Builtin<"_Vector<32, __bf16>(_Constant unsigned char, unsigned int)">;
+ def tcvtrowps2phh : X86Builtin<"_Vector<32, _Float16>(_Constant unsigned char, unsigned int)">;
+ def tcvtrowps2phl : X86Builtin<"_Vector<32, _Float16>(_Constant unsigned char, unsigned int)">;
+ def tilemovrow : X86Builtin<"_Vector<16, int>(_Constant unsigned char, unsigned int)">;
+}
+
+let Features = "amx-fp16", Attributes = [NoThrow] in {
+ def tdpfp16ps : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+}
+
+let Features = "amx-fp8", Attributes = [NoThrow] in {
+ def tdpbf8ps : X86Builtin<"void(_Constant unsigned char, unsigned _Constant char, unsigned _Constant char)">;
+ def tdpbhf8ps : X86Builtin<"void(_Constant unsigned char, unsigned _Constant char, unsigned _Constant char)">;
+ def tdphbf8ps : X86Builtin<"void(_Constant unsigned char, unsigned _Constant char, unsigned _Constant char)">;
+ def tdphf8ps : X86Builtin<"void(_Constant unsigned char, unsigned _Constant char, unsigned _Constant char)">;
+}
+
+let Features = "amx-tf32", Attributes = [NoThrow] in {
+ def tmmultf32ps : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+}
+
+let Features = "amx-tf32,amx-transpose", Attributes = [NoThrow] in {
+ def ttmmultf32ps : X86Builtin<"void(_Constant unsigned char, _Constant unsigned char, _Constant unsigned char)">;
+}
+
+let Features = "prefetchi", Attributes = [NoThrow, Const] in {
+ def prefetchi : X86Builtin<"void(void const *, unsigned int)">;
+}
+
+let Features = "cmpccxadd", Attributes = [NoThrow] in {
+ def cmpccxadd32 : X86Builtin<"signed int(void *, signed int, signed int, _Constant int)">;
+ def cmpccxadd64 : X86Builtin<"signed long long int(signed long long int *, signed long long int, signed long long int, _Constant int)">;
+}
+
+let Features = "raoint", Attributes = [NoThrow] in {
+ def aadd64 : X86Builtin<"void(void *, signed long long int)">;
+ def aand64 : X86Builtin<"void(void *, signed long long int)">;
+ def aor64 : X86Builtin<"void(void *, signed long long int)">;
+ def axor64 : X86Builtin<"void(void *, signed long long int)">;
+}
+
+let Features = "movrs", Attributes = [NoThrow] in {
+ def movrsqi : X86Builtin<"signed char(void const *)">;
+ def movrshi : X86Builtin<"signed short(void const *)">;
+ def movrssi : X86Builtin<"signed int(void const *)">;
+ def movrsdi : X86Builtin<"signed long long int(void const *)">;
+}
+
+let Features = "movrs,avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vmovrsb128 : X86Builtin<"_Vector<16, char>(_Vector<16, char const *>)">;
+}
+
+let Features = "movrs,avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vmovrsb256 : X86Builtin<"_Vector<32, char>(_Vector<32, char const *>)">;
+}
+
+let Features = "movrs,avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vmovrsb512 : X86Builtin<"_Vector<64, char>(_Vector<64, char const *>)">;
+}
+
+let Features = "movrs,avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vmovrsd128 : X86Builtin<"_Vector<4, int>(_Vector<4, int const *>)">;
+}
+
+let Features = "movrs,avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vmovrsd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int const *>)">;
+}
+
+let Features = "movrs,avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vmovrsd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int const *>)">;
+}
+
+let Features = "movrs,avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vmovrsq128 : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int const *>)">;
+}
+
+let Features = "movrs,avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vmovrsq256 : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int const *>)">;
+}
+
+let Features = "movrs,avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vmovrsq512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int const *>)">;
+}
+
+let Features = "movrs,avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<128>] in {
+ def vmovrsw128 : X86Builtin<"_Vector<8, short>(_Vector<8, short const *>)">;
+}
+
+let Features = "movrs,avx10.2-256", Attributes = [NoThrow, RequiredVectorWidth<256>] in {
+ def vmovrsw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short const *>)">;
+}
+
+let Features = "movrs,avx10.2-512", Attributes = [NoThrow, RequiredVectorWidth<512>] in {
+ def vmovrsw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short const *>)">;
+}
diff --git a/clang/include/clang/Basic/CMakeLists.txt b/clang/include/clang/Basic/CMakeLists.txt
index 76ac3367e23a66..1ccc73892fe6e2 100644
--- a/clang/include/clang/Basic/CMakeLists.txt
+++ b/clang/include/clang/Basic/CMakeLists.txt
@@ -64,6 +64,10 @@ clang_tablegen(BuiltinsX86.inc -gen-clang-builtins
SOURCE BuiltinsX86.td
TARGET ClangBuiltinsX86)
+clang_tablegen(BuiltinsX86_64.inc -gen-clang-builtins
+ SOURCE BuiltinsX86_64.td
+ TARGET ClangBuiltinsX86_64)
+
# ARM NEON and MVE
clang_tablegen(arm_neon.inc -gen-arm-neon-sema
SOURCE arm_neon.td
diff --git a/clang/include/clang/Basic/TargetBuiltins.h b/clang/include/clang/Basic/TargetBuiltins.h
index 556332dd4b217b..914be3691ee812 100644
--- a/clang/include/clang/Basic/TargetBuiltins.h
+++ b/clang/include/clang/Basic/TargetBuiltins.h
@@ -128,7 +128,7 @@ namespace clang {
FirstX86_64Builtin,
LastX86CommonBuiltin = FirstX86_64Builtin - 1,
#define BUILTIN(ID, TYPE, ATTRS) BI##ID,
-#include "clang/Basic/BuiltinsX86_64.def"
+#include "clang/Basic/BuiltinsX86_64.inc"
LastTSBuiltin
};
}
diff --git a/clang/lib/Basic/Targets/X86.cpp b/clang/lib/Basic/Targets/X86.cpp
index 7e5a5c78aa6b58..d2d92fb864c310 100644
--- a/clang/lib/Basic/Targets/X86.cpp
+++ b/clang/lib/Basic/Targets/X86.cpp
@@ -38,7 +38,7 @@ static constexpr Builtin::Info BuiltinInfoX86[] = {
{#ID, TYPE, ATTRS, FEATURE, HeaderDesc::NO_HEADER, ALL_LANGUAGES},
#define TARGET_HEADER_BUILTIN(ID, TYPE, ATTRS, HEADER, LANGS, FEATURE) \
{#ID, TYPE, ATTRS, FEATURE, HeaderDesc::HEADER, LANGS},
-#include "clang/Basic/BuiltinsX86_64.def"
+#include "clang/Basic/BuiltinsX86_64.inc"
};
static const char *const GCCRegNames[] = {
More information about the cfe-commits
mailing list