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
262 lines
19 KiB
C
262 lines
19 KiB
C
// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --version 5
|
|
// RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon -target-cpu cyclone \
|
|
// RUN: -disable-O0-optnone -emit-llvm -o - %s | opt -S -passes=mem2reg,sroa \
|
|
// RUN: | FileCheck --check-prefix=UNCONSTRAINED %s
|
|
// RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon -target-cpu cyclone \
|
|
// RUN: -ffp-exception-behavior=strict \
|
|
// RUN: -disable-O0-optnone -emit-llvm -o - %s | opt -S -passes=mem2reg,sroa \
|
|
// RUN: | FileCheck --check-prefix=CONSTRAINED %s
|
|
|
|
// REQUIRES: aarch64-registered-target
|
|
|
|
// Test new aarch64 intrinsics and types but constrained
|
|
|
|
#include <arm_neon.h>
|
|
|
|
// UNCONSTRAINED-LABEL: define dso_local float @test_vfmas_lane_f32(
|
|
// UNCONSTRAINED-SAME: float noundef [[A:%.*]], float noundef [[B:%.*]], <2 x float> noundef [[C:%.*]]) #[[ATTR0:[0-9]+]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[EXTRACT:%.*]] = extractelement <2 x float> [[C]], i32 1
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = call float @llvm.fma.f32(float [[B]], float [[EXTRACT]], float [[A]])
|
|
// UNCONSTRAINED-NEXT: ret float [[TMP0]]
|
|
//
|
|
// CONSTRAINED-LABEL: define dso_local float @test_vfmas_lane_f32(
|
|
// CONSTRAINED-SAME: float noundef [[A:%.*]], float noundef [[B:%.*]], <2 x float> noundef [[C:%.*]]) #[[ATTR0:[0-9]+]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[EXTRACT:%.*]] = extractelement <2 x float> [[C]], i32 1
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = call float @llvm.experimental.constrained.fma.f32(float [[B]], float [[EXTRACT]], float [[A]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR2:[0-9]+]]
|
|
// CONSTRAINED-NEXT: ret float [[TMP0]]
|
|
//
|
|
float32_t test_vfmas_lane_f32(float32_t a, float32_t b, float32x2_t c) {
|
|
return vfmas_lane_f32(a, b, c, 1);
|
|
}
|
|
|
|
// UNCONSTRAINED-LABEL: define dso_local double @test_vfmad_lane_f64(
|
|
// UNCONSTRAINED-SAME: double noundef [[A:%.*]], double noundef [[B:%.*]], <1 x double> noundef [[C:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[EXTRACT:%.*]] = extractelement <1 x double> [[C]], i32 0
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = call double @llvm.fma.f64(double [[B]], double [[EXTRACT]], double [[A]])
|
|
// UNCONSTRAINED-NEXT: ret double [[TMP0]]
|
|
//
|
|
// CONSTRAINED-LABEL: define dso_local double @test_vfmad_lane_f64(
|
|
// CONSTRAINED-SAME: double noundef [[A:%.*]], double noundef [[B:%.*]], <1 x double> noundef [[C:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[EXTRACT:%.*]] = extractelement <1 x double> [[C]], i32 0
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = call double @llvm.experimental.constrained.fma.f64(double [[B]], double [[EXTRACT]], double [[A]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR2]]
|
|
// CONSTRAINED-NEXT: ret double [[TMP0]]
|
|
//
|
|
float64_t test_vfmad_lane_f64(float64_t a, float64_t b, float64x1_t c) {
|
|
return vfmad_lane_f64(a, b, c, 0);
|
|
}
|
|
|
|
// UNCONSTRAINED-LABEL: define dso_local double @test_vfmad_laneq_f64(
|
|
// UNCONSTRAINED-SAME: double noundef [[A:%.*]], double noundef [[B:%.*]], <2 x double> noundef [[C:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[EXTRACT:%.*]] = extractelement <2 x double> [[C]], i32 1
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = call double @llvm.fma.f64(double [[B]], double [[EXTRACT]], double [[A]])
|
|
// UNCONSTRAINED-NEXT: ret double [[TMP0]]
|
|
//
|
|
// CONSTRAINED-LABEL: define dso_local double @test_vfmad_laneq_f64(
|
|
// CONSTRAINED-SAME: double noundef [[A:%.*]], double noundef [[B:%.*]], <2 x double> noundef [[C:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[EXTRACT:%.*]] = extractelement <2 x double> [[C]], i32 1
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = call double @llvm.experimental.constrained.fma.f64(double [[B]], double [[EXTRACT]], double [[A]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR2]]
|
|
// CONSTRAINED-NEXT: ret double [[TMP0]]
|
|
//
|
|
float64_t test_vfmad_laneq_f64(float64_t a, float64_t b, float64x2_t c) {
|
|
return vfmad_laneq_f64(a, b, c, 1);
|
|
}
|
|
|
|
// UNCONSTRAINED-LABEL: define dso_local float @test_vfmss_lane_f32(
|
|
// UNCONSTRAINED-SAME: float noundef [[A:%.*]], float noundef [[B:%.*]], <2 x float> noundef [[C:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[FNEG:%.*]] = fneg float [[B]]
|
|
// UNCONSTRAINED-NEXT: [[EXTRACT:%.*]] = extractelement <2 x float> [[C]], i32 1
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = call float @llvm.fma.f32(float [[FNEG]], float [[EXTRACT]], float [[A]])
|
|
// UNCONSTRAINED-NEXT: ret float [[TMP0]]
|
|
//
|
|
// CONSTRAINED-LABEL: define dso_local float @test_vfmss_lane_f32(
|
|
// CONSTRAINED-SAME: float noundef [[A:%.*]], float noundef [[B:%.*]], <2 x float> noundef [[C:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[FNEG:%.*]] = fneg float [[B]]
|
|
// CONSTRAINED-NEXT: [[EXTRACT:%.*]] = extractelement <2 x float> [[C]], i32 1
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = call float @llvm.experimental.constrained.fma.f32(float [[FNEG]], float [[EXTRACT]], float [[A]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR2]]
|
|
// CONSTRAINED-NEXT: ret float [[TMP0]]
|
|
//
|
|
float32_t test_vfmss_lane_f32(float32_t a, float32_t b, float32x2_t c) {
|
|
return vfmss_lane_f32(a, b, c, 1);
|
|
}
|
|
|
|
// UNCONSTRAINED-LABEL: define dso_local <1 x double> @test_vfma_lane_f64(
|
|
// UNCONSTRAINED-SAME: <1 x double> noundef [[A:%.*]], <1 x double> noundef [[B:%.*]], <1 x double> noundef [[V:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <1 x double> [[A]] to i64
|
|
// UNCONSTRAINED-NEXT: [[__S0_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP0]], i32 0
|
|
// UNCONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <1 x double> [[B]] to i64
|
|
// UNCONSTRAINED-NEXT: [[__S1_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP1]], i32 0
|
|
// UNCONSTRAINED-NEXT: [[TMP2:%.*]] = bitcast <1 x double> [[V]] to i64
|
|
// UNCONSTRAINED-NEXT: [[__S2_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP2]], i32 0
|
|
// UNCONSTRAINED-NEXT: [[TMP3:%.*]] = bitcast <1 x i64> [[__S0_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP4:%.*]] = bitcast <1 x i64> [[__S1_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP5:%.*]] = bitcast <1 x i64> [[__S2_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP6:%.*]] = bitcast <8 x i8> [[TMP5]] to <1 x double>
|
|
// UNCONSTRAINED-NEXT: [[LANE:%.*]] = shufflevector <1 x double> [[TMP6]], <1 x double> [[TMP6]], <1 x i32> zeroinitializer
|
|
// UNCONSTRAINED-NEXT: [[FMLA:%.*]] = bitcast <8 x i8> [[TMP4]] to <1 x double>
|
|
// UNCONSTRAINED-NEXT: [[FMLA1:%.*]] = bitcast <8 x i8> [[TMP3]] to <1 x double>
|
|
// UNCONSTRAINED-NEXT: [[FMLA2:%.*]] = call <1 x double> @llvm.fma.v1f64(<1 x double> [[FMLA]], <1 x double> [[LANE]], <1 x double> [[FMLA1]])
|
|
// UNCONSTRAINED-NEXT: ret <1 x double> [[FMLA2]]
|
|
//
|
|
// CONSTRAINED-LABEL: define dso_local <1 x double> @test_vfma_lane_f64(
|
|
// CONSTRAINED-SAME: <1 x double> noundef [[A:%.*]], <1 x double> noundef [[B:%.*]], <1 x double> noundef [[V:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <1 x double> [[A]] to i64
|
|
// CONSTRAINED-NEXT: [[__S0_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP0]], i32 0
|
|
// CONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <1 x double> [[B]] to i64
|
|
// CONSTRAINED-NEXT: [[__S1_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP1]], i32 0
|
|
// CONSTRAINED-NEXT: [[TMP2:%.*]] = bitcast <1 x double> [[V]] to i64
|
|
// CONSTRAINED-NEXT: [[__S2_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP2]], i32 0
|
|
// CONSTRAINED-NEXT: [[TMP3:%.*]] = bitcast <1 x i64> [[__S0_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP4:%.*]] = bitcast <1 x i64> [[__S1_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP5:%.*]] = bitcast <1 x i64> [[__S2_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP6:%.*]] = bitcast <8 x i8> [[TMP5]] to <1 x double>
|
|
// CONSTRAINED-NEXT: [[LANE:%.*]] = shufflevector <1 x double> [[TMP6]], <1 x double> [[TMP6]], <1 x i32> zeroinitializer
|
|
// CONSTRAINED-NEXT: [[FMLA:%.*]] = bitcast <8 x i8> [[TMP4]] to <1 x double>
|
|
// CONSTRAINED-NEXT: [[FMLA1:%.*]] = bitcast <8 x i8> [[TMP3]] to <1 x double>
|
|
// CONSTRAINED-NEXT: [[FMLA2:%.*]] = call <1 x double> @llvm.experimental.constrained.fma.v1f64(<1 x double> [[FMLA]], <1 x double> [[LANE]], <1 x double> [[FMLA1]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR2]]
|
|
// CONSTRAINED-NEXT: ret <1 x double> [[FMLA2]]
|
|
//
|
|
float64x1_t test_vfma_lane_f64(float64x1_t a, float64x1_t b, float64x1_t v) {
|
|
return vfma_lane_f64(a, b, v, 0);
|
|
}
|
|
|
|
// UNCONSTRAINED-LABEL: define dso_local <1 x double> @test_vfms_lane_f64(
|
|
// UNCONSTRAINED-SAME: <1 x double> noundef [[A:%.*]], <1 x double> noundef [[B:%.*]], <1 x double> noundef [[V:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <1 x double> [[A]] to i64
|
|
// UNCONSTRAINED-NEXT: [[__S0_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP0]], i32 0
|
|
// UNCONSTRAINED-NEXT: [[FNEG:%.*]] = fneg <1 x double> [[B]]
|
|
// UNCONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <1 x double> [[FNEG]] to i64
|
|
// UNCONSTRAINED-NEXT: [[__S1_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP1]], i32 0
|
|
// UNCONSTRAINED-NEXT: [[TMP2:%.*]] = bitcast <1 x double> [[V]] to i64
|
|
// UNCONSTRAINED-NEXT: [[__S2_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP2]], i32 0
|
|
// UNCONSTRAINED-NEXT: [[TMP3:%.*]] = bitcast <1 x i64> [[__S0_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP4:%.*]] = bitcast <1 x i64> [[__S1_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP5:%.*]] = bitcast <1 x i64> [[__S2_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP6:%.*]] = bitcast <8 x i8> [[TMP5]] to <1 x double>
|
|
// UNCONSTRAINED-NEXT: [[LANE:%.*]] = shufflevector <1 x double> [[TMP6]], <1 x double> [[TMP6]], <1 x i32> zeroinitializer
|
|
// UNCONSTRAINED-NEXT: [[FMLA:%.*]] = bitcast <8 x i8> [[TMP4]] to <1 x double>
|
|
// UNCONSTRAINED-NEXT: [[FMLA1:%.*]] = bitcast <8 x i8> [[TMP3]] to <1 x double>
|
|
// UNCONSTRAINED-NEXT: [[FMLA2:%.*]] = call <1 x double> @llvm.fma.v1f64(<1 x double> [[FMLA]], <1 x double> [[LANE]], <1 x double> [[FMLA1]])
|
|
// UNCONSTRAINED-NEXT: ret <1 x double> [[FMLA2]]
|
|
//
|
|
// CONSTRAINED-LABEL: define dso_local <1 x double> @test_vfms_lane_f64(
|
|
// CONSTRAINED-SAME: <1 x double> noundef [[A:%.*]], <1 x double> noundef [[B:%.*]], <1 x double> noundef [[V:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <1 x double> [[A]] to i64
|
|
// CONSTRAINED-NEXT: [[__S0_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP0]], i32 0
|
|
// CONSTRAINED-NEXT: [[FNEG:%.*]] = fneg <1 x double> [[B]]
|
|
// CONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <1 x double> [[FNEG]] to i64
|
|
// CONSTRAINED-NEXT: [[__S1_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP1]], i32 0
|
|
// CONSTRAINED-NEXT: [[TMP2:%.*]] = bitcast <1 x double> [[V]] to i64
|
|
// CONSTRAINED-NEXT: [[__S2_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP2]], i32 0
|
|
// CONSTRAINED-NEXT: [[TMP3:%.*]] = bitcast <1 x i64> [[__S0_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP4:%.*]] = bitcast <1 x i64> [[__S1_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP5:%.*]] = bitcast <1 x i64> [[__S2_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP6:%.*]] = bitcast <8 x i8> [[TMP5]] to <1 x double>
|
|
// CONSTRAINED-NEXT: [[LANE:%.*]] = shufflevector <1 x double> [[TMP6]], <1 x double> [[TMP6]], <1 x i32> zeroinitializer
|
|
// CONSTRAINED-NEXT: [[FMLA:%.*]] = bitcast <8 x i8> [[TMP4]] to <1 x double>
|
|
// CONSTRAINED-NEXT: [[FMLA1:%.*]] = bitcast <8 x i8> [[TMP3]] to <1 x double>
|
|
// CONSTRAINED-NEXT: [[FMLA2:%.*]] = call <1 x double> @llvm.experimental.constrained.fma.v1f64(<1 x double> [[FMLA]], <1 x double> [[LANE]], <1 x double> [[FMLA1]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR2]]
|
|
// CONSTRAINED-NEXT: ret <1 x double> [[FMLA2]]
|
|
//
|
|
float64x1_t test_vfms_lane_f64(float64x1_t a, float64x1_t b, float64x1_t v) {
|
|
return vfms_lane_f64(a, b, v, 0);
|
|
}
|
|
|
|
// UNCONSTRAINED-LABEL: define dso_local <1 x double> @test_vfma_laneq_f64(
|
|
// UNCONSTRAINED-SAME: <1 x double> noundef [[A:%.*]], <1 x double> noundef [[B:%.*]], <2 x double> noundef [[V:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <1 x double> [[A]] to i64
|
|
// UNCONSTRAINED-NEXT: [[__S0_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP0]], i32 0
|
|
// UNCONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <1 x double> [[B]] to i64
|
|
// UNCONSTRAINED-NEXT: [[__S1_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP1]], i32 0
|
|
// UNCONSTRAINED-NEXT: [[TMP2:%.*]] = bitcast <2 x double> [[V]] to <2 x i64>
|
|
// UNCONSTRAINED-NEXT: [[TMP3:%.*]] = bitcast <1 x i64> [[__S0_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP4:%.*]] = bitcast <1 x i64> [[__S1_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP5:%.*]] = bitcast <2 x i64> [[TMP2]] to <16 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP6:%.*]] = bitcast <8 x i8> [[TMP3]] to double
|
|
// UNCONSTRAINED-NEXT: [[TMP7:%.*]] = bitcast <8 x i8> [[TMP4]] to double
|
|
// UNCONSTRAINED-NEXT: [[TMP8:%.*]] = bitcast <16 x i8> [[TMP5]] to <2 x double>
|
|
// UNCONSTRAINED-NEXT: [[EXTRACT:%.*]] = extractelement <2 x double> [[TMP8]], i32 0
|
|
// UNCONSTRAINED-NEXT: [[TMP9:%.*]] = call double @llvm.fma.f64(double [[TMP7]], double [[EXTRACT]], double [[TMP6]])
|
|
// UNCONSTRAINED-NEXT: [[TMP10:%.*]] = bitcast double [[TMP9]] to <1 x double>
|
|
// UNCONSTRAINED-NEXT: ret <1 x double> [[TMP10]]
|
|
//
|
|
// CONSTRAINED-LABEL: define dso_local <1 x double> @test_vfma_laneq_f64(
|
|
// CONSTRAINED-SAME: <1 x double> noundef [[A:%.*]], <1 x double> noundef [[B:%.*]], <2 x double> noundef [[V:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <1 x double> [[A]] to i64
|
|
// CONSTRAINED-NEXT: [[__S0_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP0]], i32 0
|
|
// CONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <1 x double> [[B]] to i64
|
|
// CONSTRAINED-NEXT: [[__S1_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP1]], i32 0
|
|
// CONSTRAINED-NEXT: [[TMP2:%.*]] = bitcast <2 x double> [[V]] to <2 x i64>
|
|
// CONSTRAINED-NEXT: [[TMP3:%.*]] = bitcast <1 x i64> [[__S0_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP4:%.*]] = bitcast <1 x i64> [[__S1_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP5:%.*]] = bitcast <2 x i64> [[TMP2]] to <16 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP6:%.*]] = bitcast <8 x i8> [[TMP3]] to double
|
|
// CONSTRAINED-NEXT: [[TMP7:%.*]] = bitcast <8 x i8> [[TMP4]] to double
|
|
// CONSTRAINED-NEXT: [[TMP8:%.*]] = bitcast <16 x i8> [[TMP5]] to <2 x double>
|
|
// CONSTRAINED-NEXT: [[EXTRACT:%.*]] = extractelement <2 x double> [[TMP8]], i32 0
|
|
// CONSTRAINED-NEXT: [[TMP9:%.*]] = call double @llvm.experimental.constrained.fma.f64(double [[TMP7]], double [[EXTRACT]], double [[TMP6]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR2]]
|
|
// CONSTRAINED-NEXT: [[TMP10:%.*]] = bitcast double [[TMP9]] to <1 x double>
|
|
// CONSTRAINED-NEXT: ret <1 x double> [[TMP10]]
|
|
//
|
|
float64x1_t test_vfma_laneq_f64(float64x1_t a, float64x1_t b, float64x2_t v) {
|
|
return vfma_laneq_f64(a, b, v, 0);
|
|
}
|
|
|
|
// UNCONSTRAINED-LABEL: define dso_local <1 x double> @test_vfms_laneq_f64(
|
|
// UNCONSTRAINED-SAME: <1 x double> noundef [[A:%.*]], <1 x double> noundef [[B:%.*]], <2 x double> noundef [[V:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <1 x double> [[A]] to i64
|
|
// UNCONSTRAINED-NEXT: [[__S0_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP0]], i32 0
|
|
// UNCONSTRAINED-NEXT: [[FNEG:%.*]] = fneg <1 x double> [[B]]
|
|
// UNCONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <1 x double> [[FNEG]] to i64
|
|
// UNCONSTRAINED-NEXT: [[__S1_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP1]], i32 0
|
|
// UNCONSTRAINED-NEXT: [[TMP2:%.*]] = bitcast <2 x double> [[V]] to <2 x i64>
|
|
// UNCONSTRAINED-NEXT: [[TMP3:%.*]] = bitcast <1 x i64> [[__S0_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP4:%.*]] = bitcast <1 x i64> [[__S1_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP5:%.*]] = bitcast <2 x i64> [[TMP2]] to <16 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP6:%.*]] = bitcast <8 x i8> [[TMP3]] to double
|
|
// UNCONSTRAINED-NEXT: [[TMP7:%.*]] = bitcast <8 x i8> [[TMP4]] to double
|
|
// UNCONSTRAINED-NEXT: [[TMP8:%.*]] = bitcast <16 x i8> [[TMP5]] to <2 x double>
|
|
// UNCONSTRAINED-NEXT: [[EXTRACT:%.*]] = extractelement <2 x double> [[TMP8]], i32 0
|
|
// UNCONSTRAINED-NEXT: [[TMP9:%.*]] = call double @llvm.fma.f64(double [[TMP7]], double [[EXTRACT]], double [[TMP6]])
|
|
// UNCONSTRAINED-NEXT: [[TMP10:%.*]] = bitcast double [[TMP9]] to <1 x double>
|
|
// UNCONSTRAINED-NEXT: ret <1 x double> [[TMP10]]
|
|
//
|
|
// CONSTRAINED-LABEL: define dso_local <1 x double> @test_vfms_laneq_f64(
|
|
// CONSTRAINED-SAME: <1 x double> noundef [[A:%.*]], <1 x double> noundef [[B:%.*]], <2 x double> noundef [[V:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <1 x double> [[A]] to i64
|
|
// CONSTRAINED-NEXT: [[__S0_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP0]], i32 0
|
|
// CONSTRAINED-NEXT: [[FNEG:%.*]] = fneg <1 x double> [[B]]
|
|
// CONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <1 x double> [[FNEG]] to i64
|
|
// CONSTRAINED-NEXT: [[__S1_SROA_0_0_VEC_INSERT:%.*]] = insertelement <1 x i64> undef, i64 [[TMP1]], i32 0
|
|
// CONSTRAINED-NEXT: [[TMP2:%.*]] = bitcast <2 x double> [[V]] to <2 x i64>
|
|
// CONSTRAINED-NEXT: [[TMP3:%.*]] = bitcast <1 x i64> [[__S0_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP4:%.*]] = bitcast <1 x i64> [[__S1_SROA_0_0_VEC_INSERT]] to <8 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP5:%.*]] = bitcast <2 x i64> [[TMP2]] to <16 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP6:%.*]] = bitcast <8 x i8> [[TMP3]] to double
|
|
// CONSTRAINED-NEXT: [[TMP7:%.*]] = bitcast <8 x i8> [[TMP4]] to double
|
|
// CONSTRAINED-NEXT: [[TMP8:%.*]] = bitcast <16 x i8> [[TMP5]] to <2 x double>
|
|
// CONSTRAINED-NEXT: [[EXTRACT:%.*]] = extractelement <2 x double> [[TMP8]], i32 0
|
|
// CONSTRAINED-NEXT: [[TMP9:%.*]] = call double @llvm.experimental.constrained.fma.f64(double [[TMP7]], double [[EXTRACT]], double [[TMP6]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR2]]
|
|
// CONSTRAINED-NEXT: [[TMP10:%.*]] = bitcast double [[TMP9]] to <1 x double>
|
|
// CONSTRAINED-NEXT: ret <1 x double> [[TMP10]]
|
|
//
|
|
float64x1_t test_vfms_laneq_f64(float64x1_t a, float64x1_t b, float64x2_t v) {
|
|
return vfms_laneq_f64(a, b, v, 0);
|
|
}
|
|
|