derek/TheRock
1
0 Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity
Files
1a182e2dac58879f215c0fad0f5f78c76dcafe56
TheRock/compiler/amd-llvm/clang/lib/Headers/avx10_2satcvtdsintrin.h
T
derek.holloway 7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
Details
CodeQL / Analyze (python) (push) Canceled after 0s
Details
pre-commit / pre-commit (push) Canceled after 0s
Details
Multi-Arch CI / setup (push) Canceled after 0s
Details
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Details
Setup (Multi-Arch) / setup (push) Canceled after 0s
Details
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Details
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Details
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Details
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Details
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Details
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
Details
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Details
Multi-Arch CI / CI Summary (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Details
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Details
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Details
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Details
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Details
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Details
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Details
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
Details
Flatten subprojects into a single base monorepo
2026-09-02 16:34:47 -07:00

377 lines
15 KiB
C
Raw Blame History

/*===----------- avx10_2satcvtdsintrin.h - AVX512SATCVTDS intrinsics --------===
*
* 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
*
*===-----------------------------------------------------------------------===
*/
#ifndef __IMMINTRIN_H
#error \
"Never use <avx10_2satcvtdsintrin.h> directly; include <immintrin.h> instead."
#endif // __IMMINTRIN_H
#ifndef __AVX10_2SATCVTDSINTRIN_H
#define __AVX10_2SATCVTDSINTRIN_H
/* Define the default attributes for the functions in this file. */
#define __DEFAULT_FN_ATTRS256 \
__attribute__((__always_inline__, __nodebug__, __target__("avx10.2"), \
__min_vector_width__(256)))
#define __DEFAULT_FN_ATTRS128 \
__attribute__((__always_inline__, __nodebug__, __target__("avx10.2"), \
__min_vector_width__(128)))
#define _mm_cvtts_roundsd_i32(__A, __R) \
((int)__builtin_ia32_vcvttsd2sis32((__v2df)(__m128)(__A), (const int)(__R)))
#define _mm_cvtts_roundsd_si32(__A, __R) \
((int)__builtin_ia32_vcvttsd2sis32((__v2df)(__m128d)(__A), (const int)(__R)))
#define _mm_cvtts_roundsd_u32(__A, __R) \
((unsigned int)__builtin_ia32_vcvttsd2usis32((__v2df)(__m128d)(__A), \
(const int)(__R)))
#define _mm_cvtts_roundss_i32(__A, __R) \
((int)__builtin_ia32_vcvttss2sis32((__v4sf)(__m128)(__A), (const int)(__R)))
#define _mm_cvtts_roundss_si32(__A, __R) \
((int)__builtin_ia32_vcvttss2sis32((__v4sf)(__m128)(__A), (const int)(__R)))
#define _mm_cvtts_roundss_u32(__A, __R) \
((unsigned int)__builtin_ia32_vcvttss2usis32((__v4sf)(__m128)(__A), \
(const int)(__R)))
#ifdef __x86_64__
#define _mm_cvtts_roundss_u64(__A, __R) \
((unsigned long long)__builtin_ia32_vcvttss2usis64((__v4sf)(__m128)(__A), \
(const int)(__R)))
#define _mm_cvtts_roundsd_u64(__A, __R) \
((unsigned long long)__builtin_ia32_vcvttsd2usis64((__v2df)(__m128d)(__A), \
(const int)(__R)))
#define _mm_cvtts_roundss_i64(__A, __R) \
((long long)__builtin_ia32_vcvttss2sis64((__v4sf)(__m128)(__A), \
(const int)(__R)))
#define _mm_cvtts_roundss_si64(__A, __R) \
((long long)__builtin_ia32_vcvttss2sis64((__v4sf)(__m128)(__A), \
(const int)(__R)))
#define _mm_cvtts_roundsd_si64(__A, __R) \
((long long)__builtin_ia32_vcvttsd2sis64((__v2df)(__m128d)(__A), \
(const int)(__R)))
#define _mm_cvtts_roundsd_i64(__A, __R) \
((long long)__builtin_ia32_vcvttsd2sis64((__v2df)(__m128d)(__A), \
(const int)(__R)))
#endif /* __x86_64__ */
// 128 Bit : Double -> int
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_cvtts_pd_epi32(__m128d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2dqs128_mask(
(__v2df)__A, (__v4si)(__m128i)_mm_undefined_si128(), (__mmask8)(-1)));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_mask_cvtts_pd_epi32(__m128i __W, __mmask8 __U, __m128d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2dqs128_mask((__v2df)__A, (__v4si)__W,
__U));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_maskz_cvtts_pd_epi32(__mmask16 __U, __m128d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2dqs128_mask(
(__v2df)__A, (__v4si)(__m128i)_mm_setzero_si128(), __U));
}
// 256 Bit : Double -> int
static __inline__ __m128i __DEFAULT_FN_ATTRS256
_mm256_cvtts_pd_epi32(__m256d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2dqs256_mask(
(__v4df)__A, (__v4si)_mm_undefined_si128(), (__mmask8)-1));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS256
_mm256_mask_cvtts_pd_epi32(__m128i __W, __mmask8 __U, __m256d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2dqs256_mask((__v4df)__A, (__v4si)__W,
__U));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS256
_mm256_maskz_cvtts_pd_epi32(__mmask8 __U, __m256d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2dqs256_mask(
(__v4df)__A, (__v4si)_mm_setzero_si128(), __U));
}
// 128 Bit : Double -> uint
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_cvtts_pd_epu32(__m128d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2udqs128_mask(
(__v2df)__A, (__v4si)(__m128i)_mm_undefined_si128(), (__mmask8)(-1)));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_mask_cvtts_pd_epu32(__m128i __W, __mmask8 __U, __m128d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2udqs128_mask(
(__v2df)__A, (__v4si)(__m128i)__W, (__mmask8)__U));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_maskz_cvtts_pd_epu32(__mmask8 __U, __m128d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2udqs128_mask(
(__v2df)__A, (__v4si)(__m128i)_mm_setzero_si128(), __U));
}
// 256 Bit : Double -> uint
static __inline__ __m128i __DEFAULT_FN_ATTRS256
_mm256_cvtts_pd_epu32(__m256d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2udqs256_mask(
(__v4df)__A, (__v4si)_mm_undefined_si128(), (__mmask8)-1));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS256
_mm256_mask_cvtts_pd_epu32(__m128i __W, __mmask8 __U, __m256d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2udqs256_mask((__v4df)__A, (__v4si)__W,
__U));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS256
_mm256_maskz_cvtts_pd_epu32(__mmask8 __U, __m256d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2udqs256_mask(
(__v4df)__A, (__v4si)_mm_setzero_si128(), __U));
}
// 128 Bit : Double -> long
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_cvtts_pd_epi64(__m128d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2qqs128_mask(
(__v2df)__A, (__v2di)_mm_undefined_si128(), (__mmask8)-1));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_mask_cvtts_pd_epi64(__m128i __W, __mmask8 __U, __m128d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2qqs128_mask((__v2df)__A, (__v2di)__W,
(__mmask8)__U));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_maskz_cvtts_pd_epi64(__mmask8 __U, __m128d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2qqs128_mask(
(__v2df)__A, (__v2di)_mm_setzero_si128(), (__mmask8)__U));
}
// 256 Bit : Double -> long
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_cvtts_pd_epi64(__m256d __A) {
return ((__m256i)__builtin_ia32_vcvttpd2qqs256_mask(
(__v4df)__A, (__v4di)_mm256_undefined_si256(), (__mmask8)-1));
}
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_mask_cvtts_pd_epi64(__m256i __W, __mmask8 __U, __m256d __A) {
return ((__m256i)__builtin_ia32_vcvttpd2qqs256_mask((__v4df)__A, (__v4di)__W,
__U));
}
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_maskz_cvtts_pd_epi64(__mmask8 __U, __m256d __A) {
return ((__m256i)__builtin_ia32_vcvttpd2qqs256_mask(
(__v4df)__A, (__v4di)_mm256_setzero_si256(), __U));
}
// 128 Bit : Double -> ulong
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_cvtts_pd_epu64(__m128d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2uqqs128_mask(
(__v2df)__A, (__v2di)_mm_undefined_si128(), (__mmask8)-1));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_mask_cvtts_pd_epu64(__m128i __W, __mmask8 __U, __m128d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2uqqs128_mask((__v2df)__A, (__v2di)__W,
(__mmask8)__U));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_maskz_cvtts_pd_epu64(__mmask8 __U, __m128d __A) {
return ((__m128i)__builtin_ia32_vcvttpd2uqqs128_mask(
(__v2df)__A, (__v2di)_mm_setzero_si128(), (__mmask8)__U));
}
// 256 Bit : Double -> ulong
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_cvtts_pd_epu64(__m256d __A) {
return ((__m256i)__builtin_ia32_vcvttpd2uqqs256_mask(
(__v4df)__A, (__v4di)_mm256_undefined_si256(), (__mmask8)-1));
}
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_mask_cvtts_pd_epu64(__m256i __W, __mmask8 __U, __m256d __A) {
return ((__m256i)__builtin_ia32_vcvttpd2uqqs256_mask((__v4df)__A, (__v4di)__W,
__U));
}
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_maskz_cvtts_pd_epu64(__mmask8 __U, __m256d __A) {
return ((__m256i)__builtin_ia32_vcvttpd2uqqs256_mask(
(__v4df)__A, (__v4di)_mm256_setzero_si256(), __U));
}
// 128 Bit : float -> int
static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtts_ps_epi32(__m128 __A) {
return ((__m128i)__builtin_ia32_vcvttps2dqs128_mask(
(__v4sf)__A, (__v4si)(__m128i)_mm_undefined_si128(), (__mmask8)(-1)));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_mask_cvtts_ps_epi32(__m128i __W, __mmask8 __U, __m128 __A) {
return ((__m128i)__builtin_ia32_vcvttps2dqs128_mask((__v4sf)__A, (__v4si)__W,
(__mmask8)__U));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_maskz_cvtts_ps_epi32(__mmask8 __U, __m128 __A) {
return ((__m128i)__builtin_ia32_vcvttps2dqs128_mask(
(__v4sf)__A, (__v4si)(__m128i)_mm_setzero_si128(), (__mmask8)__U));
}
// 256 Bit : float -> int
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_cvtts_ps_epi32(__m256 __A) {
return ((__m256i)__builtin_ia32_vcvttps2dqs256_mask(
(__v8sf)__A, (__v8si)_mm256_undefined_si256(), (__mmask8)-1));
}
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_mask_cvtts_ps_epi32(__m256i __W, __mmask8 __U, __m256 __A) {
return ((__m256i)__builtin_ia32_vcvttps2dqs256_mask((__v8sf)__A, (__v8si)__W,
__U));
}
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_maskz_cvtts_ps_epi32(__mmask8 __U, __m256 __A) {
return ((__m256i)__builtin_ia32_vcvttps2dqs256_mask(
(__v8sf)__A, (__v8si)_mm256_setzero_si256(), __U));
}
// 128 Bit : float -> uint
static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtts_ps_epu32(__m128 __A) {
return ((__m128i)__builtin_ia32_vcvttps2udqs128_mask(
(__v4sf)__A, (__v4si)(__m128i)_mm_undefined_si128(), (__mmask8)(-1)));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_mask_cvtts_ps_epu32(__m128i __W, __mmask8 __U, __m128 __A) {
return ((__m128i)__builtin_ia32_vcvttps2udqs128_mask((__v4sf)__A, (__v4si)__W,
(__mmask8)__U));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_maskz_cvtts_ps_epu32(__mmask8 __U, __m128 __A) {
return ((__m128i)__builtin_ia32_vcvttps2udqs128_mask(
(__v4sf)__A, (__v4si)_mm_setzero_si128(), (__mmask8)__U));
}
// 256 Bit : float -> uint
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_cvtts_ps_epu32(__m256 __A) {
return ((__m256i)__builtin_ia32_vcvttps2udqs256_mask(
(__v8sf)__A, (__v8si)_mm256_undefined_si256(), (__mmask8)-1));
}
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_mask_cvtts_ps_epu32(__m256i __W, __mmask8 __U, __m256 __A) {
return ((__m256i)__builtin_ia32_vcvttps2udqs256_mask((__v8sf)__A, (__v8si)__W,
__U));
}
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_maskz_cvtts_ps_epu32(__mmask8 __U, __m256 __A) {
return ((__m256i)__builtin_ia32_vcvttps2udqs256_mask(
(__v8sf)__A, (__v8si)_mm256_setzero_si256(), __U));
}
// 128 bit : float -> long
static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtts_ps_epi64(__m128 __A) {
return ((__m128i)__builtin_ia32_vcvttps2qqs128_mask(
(__v4sf)__A, (__v2di)_mm_undefined_si128(), (__mmask8)-1));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_mask_cvtts_ps_epi64(__m128i __W, __mmask8 __U, __m128 __A) {
return ((__m128i)__builtin_ia32_vcvttps2qqs128_mask(
(__v4sf)__A, (__v2di)(__m128i)__W, (__mmask8)__U));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_maskz_cvtts_ps_epi64(__mmask8 __U, __m128 __A) {
return ((__m128i)__builtin_ia32_vcvttps2qqs128_mask(
(__v4sf)__A, (__v2di)_mm_setzero_si128(), (__mmask8)__U));
}
// 256 bit : float -> long
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_cvtts_ps_epi64(__m128 __A) {
return ((__m256i)__builtin_ia32_vcvttps2qqs256_mask(
(__v4sf)__A, (__v4di)_mm256_undefined_si256(), (__mmask8)-1));
}
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_mask_cvtts_ps_epi64(__m256i __W, __mmask8 __U, __m128 __A) {
return ((__m256i)__builtin_ia32_vcvttps2qqs256_mask((__v4sf)__A, (__v4di)__W,
__U));
}
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_maskz_cvtts_ps_epi64(__mmask8 __U, __m128 __A) {
return ((__m256i)__builtin_ia32_vcvttps2qqs256_mask(
(__v4sf)__A, (__v4di)_mm256_setzero_si256(), __U));
}
// 128 bit : float -> ulong
static __inline__ __m128i __DEFAULT_FN_ATTRS128 _mm_cvtts_ps_epu64(__m128 __A) {
return ((__m128i)__builtin_ia32_vcvttps2uqqs128_mask(
(__v4sf)__A, (__v2di)_mm_undefined_si128(), (__mmask8)-1));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_mask_cvtts_ps_epu64(__m128i __W, __mmask8 __U, __m128 __A) {
return ((__m128i)__builtin_ia32_vcvttps2uqqs128_mask(
(__v4sf)__A, (__v2di)(__m128i)__W, (__mmask8)__U));
}
static __inline__ __m128i __DEFAULT_FN_ATTRS128
_mm_maskz_cvtts_ps_epu64(__mmask8 __U, __m128 __A) {
return ((__m128i)__builtin_ia32_vcvttps2uqqs128_mask(
(__v4sf)__A, (__v2di)_mm_setzero_si128(), (__mmask8)__U));
}
// 256 bit : float -> ulong
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_cvtts_ps_epu64(__m128 __A) {
return ((__m256i)__builtin_ia32_vcvttps2uqqs256_mask(
(__v4sf)__A, (__v4di)_mm256_undefined_si256(), (__mmask8)-1));
}
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_mask_cvtts_ps_epu64(__m256i __W, __mmask8 __U, __m128 __A) {
return ((__m256i)__builtin_ia32_vcvttps2uqqs256_mask((__v4sf)__A, (__v4di)__W,
__U));
}
static __inline__ __m256i __DEFAULT_FN_ATTRS256
_mm256_maskz_cvtts_ps_epu64(__mmask8 __U, __m128 __A) {
return ((__m256i)__builtin_ia32_vcvttps2uqqs256_mask(
(__v4sf)__A, (__v4di)_mm256_setzero_si256(), __U));
}
#undef __DEFAULT_FN_ATTRS128
#undef __DEFAULT_FN_ATTRS256
#endif // __AVX10_2SATCVTDSINTRIN_H
Reference in new issue View Git Blame Copy Permalink