derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
104 lines
6.4 KiB
C
104 lines
6.4 KiB
C
// RUN: %clang_cc1 -x c -flax-vector-conversions=none -ffreestanding %s -triple=x86_64-apple-darwin -target-feature +avx512bf16 -emit-llvm -o - -Wall -Werror | FileCheck %s
|
|
// RUN: %clang_cc1 -x c -flax-vector-conversions=none -ffreestanding %s -triple=i386-apple-darwin -target-feature +avx512bf16 -emit-llvm -o - -Wall -Werror | FileCheck %s
|
|
// RUN: %clang_cc1 -x c++ -flax-vector-conversions=none -ffreestanding %s -triple=x86_64-apple-darwin -target-feature +avx512bf16 -emit-llvm -o - -Wall -Werror | FileCheck %s
|
|
// RUN: %clang_cc1 -x c++ -flax-vector-conversions=none -ffreestanding %s -triple=i386-apple-darwin -target-feature +avx512bf16 -emit-llvm -o - -Wall -Werror | FileCheck %s
|
|
//
|
|
// RUN: %clang_cc1 -x c -flax-vector-conversions=none -ffreestanding %s -triple=x86_64-apple-darwin -target-feature +avx512bf16 -emit-llvm -o - -Wall -Werror -fexperimental-new-constant-interpreter | FileCheck %s
|
|
// RUN: %clang_cc1 -x c -flax-vector-conversions=none -ffreestanding %s -triple=i386-apple-darwin -target-feature +avx512bf16 -emit-llvm -o - -Wall -Werror -fexperimental-new-constant-interpreter | FileCheck %s
|
|
// RUN: %clang_cc1 -x c++ -flax-vector-conversions=none -ffreestanding %s -triple=x86_64-apple-darwin -target-feature +avx512bf16 -emit-llvm -o - -Wall -Werror -fexperimental-new-constant-interpreter | FileCheck %s
|
|
// RUN: %clang_cc1 -x c++ -flax-vector-conversions=none -ffreestanding %s -triple=i386-apple-darwin -target-feature +avx512bf16 -emit-llvm -o - -Wall -Werror -fexperimental-new-constant-interpreter | FileCheck %s
|
|
|
|
#include <immintrin.h>
|
|
#include "builtin_test_helpers.h"
|
|
|
|
float test_mm_cvtsbh_ss(__bf16 A) {
|
|
// CHECK-LABEL: test_mm_cvtsbh_ss
|
|
// CHECK: fpext bfloat %{{.*}} to float
|
|
// CHECK: ret float %{{.*}}
|
|
return _mm_cvtsbh_ss(A);
|
|
}
|
|
TEST_CONSTEXPR(_mm_cvtsbh_ss(-1.0f) == -1.0f);
|
|
|
|
__m512bh test_mm512_cvtne2ps_pbh(__m512 A, __m512 B) {
|
|
// CHECK-LABEL: test_mm512_cvtne2ps_pbh
|
|
// CHECK: call {{.*}}<32 x bfloat> @llvm.x86.avx512bf16.cvtne2ps2bf16.512(<16 x float> %{{.*}}, <16 x float> %{{.*}})
|
|
return _mm512_cvtne2ps_pbh(A, B);
|
|
}
|
|
|
|
__m512bh test_mm512_maskz_cvtne2ps_pbh(__m512 A, __m512 B, __mmask32 U) {
|
|
// CHECK-LABEL: test_mm512_maskz_cvtne2ps_pbh
|
|
// CHECK: call {{.*}}<32 x bfloat> @llvm.x86.avx512bf16.cvtne2ps2bf16.512(<16 x float> %{{.*}}, <16 x float> %{{.*}})
|
|
// CHECK: select <32 x i1> %{{.*}}, <32 x bfloat> %{{.*}}, <32 x bfloat> %{{.*}}
|
|
return _mm512_maskz_cvtne2ps_pbh(U, A, B);
|
|
}
|
|
|
|
__m512bh test_mm512_mask_cvtne2ps_pbh(__m512bh C, __mmask32 U, __m512 A, __m512 B) {
|
|
// CHECK-LABEL: test_mm512_mask_cvtne2ps_pbh
|
|
// CHECK: call {{.*}}<32 x bfloat> @llvm.x86.avx512bf16.cvtne2ps2bf16.512(<16 x float> %{{.*}}, <16 x float> %{{.*}})
|
|
// CHECK: select <32 x i1> %{{.*}}, <32 x bfloat> %{{.*}}, <32 x bfloat> %{{.*}}
|
|
return _mm512_mask_cvtne2ps_pbh(C, U, A, B);
|
|
}
|
|
|
|
__m256bh test_mm512_cvtneps_pbh(__m512 A) {
|
|
// CHECK-LABEL: test_mm512_cvtneps_pbh
|
|
// CHECK: call {{.*}}<16 x bfloat> @llvm.x86.avx512bf16.cvtneps2bf16.512(<16 x float> %{{.*}})
|
|
return _mm512_cvtneps_pbh(A);
|
|
}
|
|
|
|
__m256bh test_mm512_mask_cvtneps_pbh(__m256bh C, __mmask16 U, __m512 A) {
|
|
// CHECK-LABEL: test_mm512_mask_cvtneps_pbh
|
|
// CHECK: call {{.*}}<16 x bfloat> @llvm.x86.avx512bf16.cvtneps2bf16.512(<16 x float> %{{.*}})
|
|
// CHECK: select <16 x i1> %{{.*}}, <16 x bfloat> %{{.*}}, <16 x bfloat> %{{.*}}
|
|
return _mm512_mask_cvtneps_pbh(C, U, A);
|
|
}
|
|
|
|
__m256bh test_mm512_maskz_cvtneps_pbh(__m512 A, __mmask16 U) {
|
|
// CHECK-LABEL: test_mm512_maskz_cvtneps_pbh
|
|
// CHECK: call {{.*}}<16 x bfloat> @llvm.x86.avx512bf16.cvtneps2bf16.512(<16 x float> %{{.*}})
|
|
// CHECK: select <16 x i1> %{{.*}}, <16 x bfloat> %{{.*}}, <16 x bfloat> %{{.*}}
|
|
return _mm512_maskz_cvtneps_pbh(U, A);
|
|
}
|
|
|
|
__m512 test_mm512_dpbf16_ps(__m512 D, __m512bh A, __m512bh B) {
|
|
// CHECK-LABEL: test_mm512_dpbf16_ps
|
|
// CHECK: call {{.*}}<16 x float> @llvm.x86.avx512bf16.dpbf16ps.512(<16 x float> %{{.*}}, <32 x bfloat> %{{.*}}, <32 x bfloat> %{{.*}})
|
|
return _mm512_dpbf16_ps(D, A, B);
|
|
}
|
|
|
|
__m512 test_mm512_maskz_dpbf16_ps(__m512 D, __m512bh A, __m512bh B, __mmask16 U) {
|
|
// CHECK-LABEL: test_mm512_maskz_dpbf16_ps
|
|
// CHECK: call {{.*}}<16 x float> @llvm.x86.avx512bf16.dpbf16ps.512(<16 x float> %{{.*}}, <32 x bfloat> %{{.*}}, <32 x bfloat> %{{.*}})
|
|
// CHECK: select <16 x i1> %{{.*}}, <16 x float> %{{.*}}, <16 x float> %{{.*}}
|
|
return _mm512_maskz_dpbf16_ps(U, D, A, B);
|
|
}
|
|
|
|
__m512 test_mm512_mask_dpbf16_ps(__m512 D, __m512bh A, __m512bh B, __mmask16 U) {
|
|
// CHECK-LABEL: test_mm512_mask_dpbf16_ps
|
|
// CHECK: call {{.*}}<16 x float> @llvm.x86.avx512bf16.dpbf16ps.512(<16 x float> %{{.*}}, <32 x bfloat> %{{.*}}, <32 x bfloat> %{{.*}})
|
|
// CHECK: select <16 x i1> %{{.*}}, <16 x float> %{{.*}}, <16 x float> %{{.*}}
|
|
return _mm512_mask_dpbf16_ps(D, U, A, B);
|
|
}
|
|
|
|
__m512 test_mm512_cvtpbh_ps(__m256bh A) {
|
|
// CHECK-LABEL: test_mm512_cvtpbh_ps
|
|
// CHECK: fpext <16 x bfloat> %{{.*}} to <16 x float>
|
|
return _mm512_cvtpbh_ps(A);
|
|
}
|
|
TEST_CONSTEXPR(match_m512(_mm512_cvtpbh_ps((__m256bh){-0.0f, 1.0f, -2.0f, 4.0f, -8.0f, 16.0f, -32.0f, 64.0f, -128.0f, -0.5f, 0.25f, -0.125f, -4.0f, 2.0f, -1.0f, 0.0f}), -0.0f, 1.0f, -2.0f, 4.0f, -8.0f, 16.0f, -32.0f, 64.0f, -128.0f, -0.5f, 0.25f, -0.125f, -4.0f, 2.0f, -1.0f, 0.0f));
|
|
|
|
__m512 test_mm512_maskz_cvtpbh_ps(__mmask16 M, __m256bh A) {
|
|
// CHECK-LABEL: test_mm512_maskz_cvtpbh_ps
|
|
// CHECK: fpext <16 x bfloat> %{{.*}} to <16 x float>
|
|
// CHECK: select <16 x i1> %{{.*}}, <16 x float> %{{.*}}, <16 x float> %{{.*}}
|
|
return _mm512_maskz_cvtpbh_ps(M, A);
|
|
}
|
|
TEST_CONSTEXPR(match_m512(_mm512_maskz_cvtpbh_ps(0xA753, (__m256bh){-0.0f, 1.0f, -2.0f, 4.0f, -8.0f, 16.0f, -32.0f, 64.0f, -128.0f, -0.5f, 0.25f, -0.125f, -4.0f, 2.0f, -1.0f, 0.0f}), -0.0f, 1.0f, 0.0f, 0.0f, -8.0f, 0.0f, -32.0f, 0.0f, -128.0f, -0.5f, 0.25f, 0.0f, 0.0f, 2.0f, 0.0f, 0.0f));
|
|
|
|
__m512 test_mm512_mask_cvtpbh_ps(__m512 S, __mmask16 M, __m256bh A) {
|
|
// CHECK-LABEL: test_mm512_mask_cvtpbh_ps
|
|
// CHECK: fpext <16 x bfloat> %{{.*}} to <16 x float>
|
|
// CHECK: select <16 x i1> %{{.*}}, <16 x float> %{{.*}}, <16 x float> %{{.*}}
|
|
return _mm512_mask_cvtpbh_ps(S, M, A);
|
|
}
|
|
TEST_CONSTEXPR(match_m512(_mm512_mask_cvtpbh_ps((__m512){ 99.0f, 99.0f, 99.0f, 99.0f, 99.0f, 99.0f, 99.0f, 99.0f, 99.0f, 99.0f, 99.0f, 99.0f, 99.0f, 99.0f, 99.0f, 99.0f }, 0xA753, (__m256bh){-0.0f, 1.0f, -2.0f, 4.0f, -8.0f, 16.0f, -32.0f, 64.0f, -128.0f, -0.5f, 0.25f, -0.125f, -4.0f, 2.0f, -1.0f, 0.0f}), -0.0f, 1.0f, 99.0f, 99.0f, -8.0f, 99.0f, -32.0f, 99.0f, -128.0f, -0.5f, 0.25f, 99.0f, 99.0f, 2.0f, 99.0f, 0.0f));
|