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
106 lines
5.8 KiB
C
106 lines
5.8 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 \
|
|
// RUN: -disable-O0-optnone -emit-llvm -o - %s \
|
|
// RUN: | opt -S -passes=mem2reg,sroa | FileCheck --check-prefix=UNCONSTRAINED %s
|
|
// RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon \
|
|
// RUN: -ffp-exception-behavior=strict \
|
|
// RUN: -disable-O0-optnone -emit-llvm -o - %s \
|
|
// RUN: | opt -S -passes=mem2reg,sroa | 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 <2 x double> @test_vrndaq_f64(
|
|
// UNCONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0:[0-9]+]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <2 x double> [[A]] to <2 x i64>
|
|
// UNCONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <2 x i64> [[TMP0]] to <16 x i8>
|
|
// UNCONSTRAINED-NEXT: [[VRNDA_I:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x double>
|
|
// UNCONSTRAINED-NEXT: [[VRNDA1_I:%.*]] = call <2 x double> @llvm.round.v2f64(<2 x double> [[VRNDA_I]])
|
|
// UNCONSTRAINED-NEXT: ret <2 x double> [[VRNDA1_I]]
|
|
//
|
|
// CONSTRAINED-LABEL: define dso_local <2 x double> @test_vrndaq_f64(
|
|
// CONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0:[0-9]+]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <2 x double> [[A]] to <2 x i64>
|
|
// CONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <2 x i64> [[TMP0]] to <16 x i8>
|
|
// CONSTRAINED-NEXT: [[VRNDA_I:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x double>
|
|
// CONSTRAINED-NEXT: [[VRNDA1_I:%.*]] = call <2 x double> @llvm.experimental.constrained.round.v2f64(<2 x double> [[VRNDA_I]], metadata !"fpexcept.strict") #[[ATTR2:[0-9]+]]
|
|
// CONSTRAINED-NEXT: ret <2 x double> [[VRNDA1_I]]
|
|
//
|
|
float64x2_t test_vrndaq_f64(float64x2_t a) {
|
|
return vrndaq_f64(a);
|
|
}
|
|
|
|
|
|
// UNCONSTRAINED-LABEL: define dso_local <2 x double> @test_vrndpq_f64(
|
|
// UNCONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <2 x double> [[A]] to <2 x i64>
|
|
// UNCONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <2 x i64> [[TMP0]] to <16 x i8>
|
|
// UNCONSTRAINED-NEXT: [[VRNDP_I:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x double>
|
|
// UNCONSTRAINED-NEXT: [[VRNDP1_I:%.*]] = call <2 x double> @llvm.ceil.v2f64(<2 x double> [[VRNDP_I]])
|
|
// UNCONSTRAINED-NEXT: ret <2 x double> [[VRNDP1_I]]
|
|
//
|
|
// CONSTRAINED-LABEL: define dso_local <2 x double> @test_vrndpq_f64(
|
|
// CONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <2 x double> [[A]] to <2 x i64>
|
|
// CONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <2 x i64> [[TMP0]] to <16 x i8>
|
|
// CONSTRAINED-NEXT: [[VRNDP_I:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x double>
|
|
// CONSTRAINED-NEXT: [[VRNDP1_I:%.*]] = call <2 x double> @llvm.experimental.constrained.ceil.v2f64(<2 x double> [[VRNDP_I]], metadata !"fpexcept.strict") #[[ATTR2]]
|
|
// CONSTRAINED-NEXT: ret <2 x double> [[VRNDP1_I]]
|
|
//
|
|
float64x2_t test_vrndpq_f64(float64x2_t a) {
|
|
return vrndpq_f64(a);
|
|
}
|
|
|
|
|
|
// UNCONSTRAINED-LABEL: define dso_local <4 x float> @test_vsqrtq_f32(
|
|
// UNCONSTRAINED-SAME: <4 x float> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <4 x float> [[A]] to <4 x i32>
|
|
// UNCONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <4 x i32> [[TMP0]] to <16 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP2:%.*]] = bitcast <16 x i8> [[TMP1]] to <4 x float>
|
|
// UNCONSTRAINED-NEXT: [[VSQRT_I:%.*]] = call <4 x float> @llvm.sqrt.v4f32(<4 x float> [[TMP2]])
|
|
// UNCONSTRAINED-NEXT: ret <4 x float> [[VSQRT_I]]
|
|
//
|
|
// CONSTRAINED-LABEL: define dso_local <4 x float> @test_vsqrtq_f32(
|
|
// CONSTRAINED-SAME: <4 x float> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <4 x float> [[A]] to <4 x i32>
|
|
// CONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <4 x i32> [[TMP0]] to <16 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP2:%.*]] = bitcast <16 x i8> [[TMP1]] to <4 x float>
|
|
// CONSTRAINED-NEXT: [[VSQRT_I:%.*]] = call <4 x float> @llvm.experimental.constrained.sqrt.v4f32(<4 x float> [[TMP2]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR2]]
|
|
// CONSTRAINED-NEXT: ret <4 x float> [[VSQRT_I]]
|
|
//
|
|
float32x4_t test_vsqrtq_f32(float32x4_t a) {
|
|
return vsqrtq_f32(a);
|
|
}
|
|
|
|
|
|
// UNCONSTRAINED-LABEL: define dso_local <2 x double> @test_vsqrtq_f64(
|
|
// UNCONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <2 x double> [[A]] to <2 x i64>
|
|
// UNCONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <2 x i64> [[TMP0]] to <16 x i8>
|
|
// UNCONSTRAINED-NEXT: [[TMP2:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x double>
|
|
// UNCONSTRAINED-NEXT: [[VSQRT_I:%.*]] = call <2 x double> @llvm.sqrt.v2f64(<2 x double> [[TMP2]])
|
|
// UNCONSTRAINED-NEXT: ret <2 x double> [[VSQRT_I]]
|
|
//
|
|
// CONSTRAINED-LABEL: define dso_local <2 x double> @test_vsqrtq_f64(
|
|
// CONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = bitcast <2 x double> [[A]] to <2 x i64>
|
|
// CONSTRAINED-NEXT: [[TMP1:%.*]] = bitcast <2 x i64> [[TMP0]] to <16 x i8>
|
|
// CONSTRAINED-NEXT: [[TMP2:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x double>
|
|
// CONSTRAINED-NEXT: [[VSQRT_I:%.*]] = call <2 x double> @llvm.experimental.constrained.sqrt.v2f64(<2 x double> [[TMP2]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR2]]
|
|
// CONSTRAINED-NEXT: ret <2 x double> [[VSQRT_I]]
|
|
//
|
|
float64x2_t test_vsqrtq_f64(float64x2_t a) {
|
|
return vsqrtq_f64(a);
|
|
}
|