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
133 lines
6.8 KiB
C
133 lines
6.8 KiB
C
// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --version 5
|
|
// RUN: %clang_cc1 -flax-vector-conversions=none -ffreestanding %s -triple=x86_64-- -target-feature +avx -disable-O0-optnone -emit-llvm -o - | opt -S -passes=mem2reg | FileCheck %s
|
|
// RUN: %clang_cc1 -flax-vector-conversions=none -ffreestanding %s -triple=i386-- -target-feature +avx -disable-O0-optnone -emit-llvm -o - | opt -S -passes=mem2reg | FileCheck %s
|
|
|
|
|
|
#include <immintrin.h>
|
|
|
|
//
|
|
// Test LLVM IR codegen of cmpXY instructions
|
|
//
|
|
|
|
// CHECK-LABEL: define dso_local <2 x double> @test_cmp_sd(
|
|
// CHECK-SAME: <2 x double> noundef [[A:%.*]], <2 x double> noundef [[B:%.*]]) #[[ATTR0:[0-9]+]] {
|
|
// CHECK-NEXT: [[ENTRY:.*:]]
|
|
// CHECK-NEXT: [[TMP0:%.*]] = call <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double> [[A]], <2 x double> [[B]], i8 13)
|
|
// CHECK-NEXT: ret <2 x double> [[TMP0]]
|
|
//
|
|
__m128d test_cmp_sd(__m128d a, __m128d b) {
|
|
// Expects that the third argument in LLVM IR is immediate expression
|
|
return _mm_cmp_sd(a, b, _CMP_GE_OS);
|
|
}
|
|
|
|
// CHECK-LABEL: define dso_local <4 x float> @test_cmp_ss(
|
|
// CHECK-SAME: <4 x float> noundef [[A:%.*]], <4 x float> noundef [[B:%.*]]) #[[ATTR0]] {
|
|
// CHECK-NEXT: [[ENTRY:.*:]]
|
|
// CHECK-NEXT: [[TMP0:%.*]] = call <4 x float> @llvm.x86.sse.cmp.ss(<4 x float> [[A]], <4 x float> [[B]], i8 13)
|
|
// CHECK-NEXT: ret <4 x float> [[TMP0]]
|
|
//
|
|
__m128 test_cmp_ss(__m128 a, __m128 b) {
|
|
// Expects that the third argument in LLVM IR is immediate expression
|
|
return _mm_cmp_ss(a, b, _CMP_GE_OS);
|
|
}
|
|
|
|
// CHECK-LABEL: define dso_local <4 x float> @test_cmpgt_ss(
|
|
// CHECK-SAME: <4 x float> noundef [[A:%.*]], <4 x float> noundef [[B:%.*]]) #[[ATTR0]] {
|
|
// CHECK-NEXT: [[ENTRY:.*:]]
|
|
// CHECK-NEXT: [[TMP0:%.*]] = call <4 x float> @llvm.x86.sse.cmp.ss(<4 x float> [[B]], <4 x float> [[A]], i8 1)
|
|
// CHECK-NEXT: [[SHUFFLE_I:%.*]] = shufflevector <4 x float> [[A]], <4 x float> [[TMP0]], <4 x i32> <i32 4, i32 1, i32 2, i32 3>
|
|
// CHECK-NEXT: ret <4 x float> [[SHUFFLE_I]]
|
|
//
|
|
__m128 test_cmpgt_ss(__m128 a, __m128 b) {
|
|
return _mm_cmpgt_ss(a, b);
|
|
}
|
|
|
|
// CHECK-LABEL: define dso_local <4 x float> @test_cmpge_ss(
|
|
// CHECK-SAME: <4 x float> noundef [[A:%.*]], <4 x float> noundef [[B:%.*]]) #[[ATTR0]] {
|
|
// CHECK-NEXT: [[ENTRY:.*:]]
|
|
// CHECK-NEXT: [[TMP0:%.*]] = call <4 x float> @llvm.x86.sse.cmp.ss(<4 x float> [[B]], <4 x float> [[A]], i8 2)
|
|
// CHECK-NEXT: [[SHUFFLE_I:%.*]] = shufflevector <4 x float> [[A]], <4 x float> [[TMP0]], <4 x i32> <i32 4, i32 1, i32 2, i32 3>
|
|
// CHECK-NEXT: ret <4 x float> [[SHUFFLE_I]]
|
|
//
|
|
__m128 test_cmpge_ss(__m128 a, __m128 b) {
|
|
return _mm_cmpge_ss(a, b);
|
|
}
|
|
|
|
// CHECK-LABEL: define dso_local <4 x float> @test_cmpngt_ss(
|
|
// CHECK-SAME: <4 x float> noundef [[A:%.*]], <4 x float> noundef [[B:%.*]]) #[[ATTR0]] {
|
|
// CHECK-NEXT: [[ENTRY:.*:]]
|
|
// CHECK-NEXT: [[TMP0:%.*]] = call <4 x float> @llvm.x86.sse.cmp.ss(<4 x float> [[B]], <4 x float> [[A]], i8 5)
|
|
// CHECK-NEXT: [[SHUFFLE_I:%.*]] = shufflevector <4 x float> [[A]], <4 x float> [[TMP0]], <4 x i32> <i32 4, i32 1, i32 2, i32 3>
|
|
// CHECK-NEXT: ret <4 x float> [[SHUFFLE_I]]
|
|
//
|
|
__m128 test_cmpngt_ss(__m128 a, __m128 b) {
|
|
return _mm_cmpngt_ss(a, b);
|
|
}
|
|
|
|
// CHECK-LABEL: define dso_local <4 x float> @test_cmpnge_ss(
|
|
// CHECK-SAME: <4 x float> noundef [[A:%.*]], <4 x float> noundef [[B:%.*]]) #[[ATTR0]] {
|
|
// CHECK-NEXT: [[ENTRY:.*:]]
|
|
// CHECK-NEXT: [[TMP0:%.*]] = call <4 x float> @llvm.x86.sse.cmp.ss(<4 x float> [[B]], <4 x float> [[A]], i8 6)
|
|
// CHECK-NEXT: [[SHUFFLE_I:%.*]] = shufflevector <4 x float> [[A]], <4 x float> [[TMP0]], <4 x i32> <i32 4, i32 1, i32 2, i32 3>
|
|
// CHECK-NEXT: ret <4 x float> [[SHUFFLE_I]]
|
|
//
|
|
__m128 test_cmpnge_ss(__m128 a, __m128 b) {
|
|
return _mm_cmpnge_ss(a, b);
|
|
}
|
|
|
|
// CHECK-LABEL: define dso_local <2 x double> @test_cmpgt_sd(
|
|
// CHECK-SAME: <2 x double> noundef [[A:%.*]], <2 x double> noundef [[B:%.*]]) #[[ATTR0]] {
|
|
// CHECK-NEXT: [[ENTRY:.*:]]
|
|
// CHECK-NEXT: [[TMP0:%.*]] = call <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double> [[B]], <2 x double> [[A]], i8 1)
|
|
// CHECK-NEXT: [[VECEXT_I:%.*]] = extractelement <2 x double> [[TMP0]], i32 0
|
|
// CHECK-NEXT: [[VECINIT_I:%.*]] = insertelement <2 x double> poison, double [[VECEXT_I]], i32 0
|
|
// CHECK-NEXT: [[VECEXT1_I:%.*]] = extractelement <2 x double> [[A]], i32 1
|
|
// CHECK-NEXT: [[VECINIT2_I:%.*]] = insertelement <2 x double> [[VECINIT_I]], double [[VECEXT1_I]], i32 1
|
|
// CHECK-NEXT: ret <2 x double> [[VECINIT2_I]]
|
|
//
|
|
__m128d test_cmpgt_sd(__m128d a, __m128d b) {
|
|
return _mm_cmpgt_sd(a, b);
|
|
}
|
|
|
|
// CHECK-LABEL: define dso_local <2 x double> @test_cmpge_sd(
|
|
// CHECK-SAME: <2 x double> noundef [[A:%.*]], <2 x double> noundef [[B:%.*]]) #[[ATTR0]] {
|
|
// CHECK-NEXT: [[ENTRY:.*:]]
|
|
// CHECK-NEXT: [[TMP0:%.*]] = call <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double> [[B]], <2 x double> [[A]], i8 2)
|
|
// CHECK-NEXT: [[VECEXT_I:%.*]] = extractelement <2 x double> [[TMP0]], i32 0
|
|
// CHECK-NEXT: [[VECINIT_I:%.*]] = insertelement <2 x double> poison, double [[VECEXT_I]], i32 0
|
|
// CHECK-NEXT: [[VECEXT1_I:%.*]] = extractelement <2 x double> [[A]], i32 1
|
|
// CHECK-NEXT: [[VECINIT2_I:%.*]] = insertelement <2 x double> [[VECINIT_I]], double [[VECEXT1_I]], i32 1
|
|
// CHECK-NEXT: ret <2 x double> [[VECINIT2_I]]
|
|
//
|
|
__m128d test_cmpge_sd(__m128d a, __m128d b) {
|
|
return _mm_cmpge_sd(a, b);
|
|
}
|
|
|
|
// CHECK-LABEL: define dso_local <2 x double> @test_cmpngt_sd(
|
|
// CHECK-SAME: <2 x double> noundef [[A:%.*]], <2 x double> noundef [[B:%.*]]) #[[ATTR0]] {
|
|
// CHECK-NEXT: [[ENTRY:.*:]]
|
|
// CHECK-NEXT: [[TMP0:%.*]] = call <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double> [[B]], <2 x double> [[A]], i8 5)
|
|
// CHECK-NEXT: [[VECEXT_I:%.*]] = extractelement <2 x double> [[TMP0]], i32 0
|
|
// CHECK-NEXT: [[VECINIT_I:%.*]] = insertelement <2 x double> poison, double [[VECEXT_I]], i32 0
|
|
// CHECK-NEXT: [[VECEXT1_I:%.*]] = extractelement <2 x double> [[A]], i32 1
|
|
// CHECK-NEXT: [[VECINIT2_I:%.*]] = insertelement <2 x double> [[VECINIT_I]], double [[VECEXT1_I]], i32 1
|
|
// CHECK-NEXT: ret <2 x double> [[VECINIT2_I]]
|
|
//
|
|
__m128d test_cmpngt_sd(__m128d a, __m128d b) {
|
|
return _mm_cmpngt_sd(a, b);
|
|
}
|
|
|
|
// CHECK-LABEL: define dso_local <2 x double> @test_cmpnge_sd(
|
|
// CHECK-SAME: <2 x double> noundef [[A:%.*]], <2 x double> noundef [[B:%.*]]) #[[ATTR0]] {
|
|
// CHECK-NEXT: [[ENTRY:.*:]]
|
|
// CHECK-NEXT: [[TMP0:%.*]] = call <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double> [[B]], <2 x double> [[A]], i8 6)
|
|
// CHECK-NEXT: [[VECEXT_I:%.*]] = extractelement <2 x double> [[TMP0]], i32 0
|
|
// CHECK-NEXT: [[VECINIT_I:%.*]] = insertelement <2 x double> poison, double [[VECEXT_I]], i32 0
|
|
// CHECK-NEXT: [[VECEXT1_I:%.*]] = extractelement <2 x double> [[A]], i32 1
|
|
// CHECK-NEXT: [[VECINIT2_I:%.*]] = insertelement <2 x double> [[VECINIT_I]], double [[VECEXT1_I]], i32 1
|
|
// CHECK-NEXT: ret <2 x double> [[VECINIT2_I]]
|
|
//
|
|
__m128d test_cmpnge_sd(__m128d a, __m128d b) {
|
|
return _mm_cmpnge_sd(a, b);
|
|
}
|