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
211 lines
15 KiB
C
211 lines
15 KiB
C
// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --version 5
|
|
// RUN: %clang_cc1 -triple arm64-apple-ios7 -target-feature +neon -ffreestanding -flax-vector-conversions=none -emit-llvm -o - %s \
|
|
// RUN: | FileCheck --check-prefix=UNCONSTRAINED %s
|
|
// RUN: %clang_cc1 -triple arm64-apple-ios7 -target-feature +neon -ffreestanding -flax-vector-conversions=none -ffp-exception-behavior=strict -emit-llvm -o - %s \
|
|
// RUN: | FileCheck --check-prefix=CONSTRAINED %s
|
|
|
|
// REQUIRES: aarch64-registered-target
|
|
|
|
#include <arm_neon.h>
|
|
|
|
// UNCONSTRAINED-LABEL: define <2 x double> @rnd5(
|
|
// UNCONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0:[0-9]+]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[__P0_ADDR_I:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: [[__RET_I:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: [[REF_TMP_I:%.*]] = alloca <16 x i8>, align 16
|
|
// UNCONSTRAINED-NEXT: [[A_ADDR:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[A]], ptr [[A_ADDR]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = load <2 x double>, ptr [[A_ADDR]], align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[TMP0]], ptr [[__P0_ADDR_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP1:%.*]] = load <16 x i8>, ptr [[__P0_ADDR_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[VRNDZ_I:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x double>
|
|
// UNCONSTRAINED-NEXT: [[VRNDZ1_I:%.*]] = call <2 x double> @llvm.trunc.v2f64(<2 x double> [[VRNDZ_I]])
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[VRNDZ1_I]], ptr [[REF_TMP_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP2:%.*]] = load <2 x double>, ptr [[REF_TMP_I]], align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[TMP2]], ptr [[__RET_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP3:%.*]] = load <2 x double>, ptr [[__RET_I]], align 16
|
|
// UNCONSTRAINED-NEXT: ret <2 x double> [[TMP3]]
|
|
//
|
|
// CONSTRAINED-LABEL: define <2 x double> @rnd5(
|
|
// CONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0:[0-9]+]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[__P0_ADDR_I:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: [[__RET_I:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: [[REF_TMP_I:%.*]] = alloca <16 x i8>, align 16
|
|
// CONSTRAINED-NEXT: [[A_ADDR:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[A]], ptr [[A_ADDR]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = load <2 x double>, ptr [[A_ADDR]], align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[TMP0]], ptr [[__P0_ADDR_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP1:%.*]] = load <16 x i8>, ptr [[__P0_ADDR_I]], align 16
|
|
// CONSTRAINED-NEXT: [[VRNDZ_I:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x double>
|
|
// CONSTRAINED-NEXT: [[VRNDZ1_I:%.*]] = call <2 x double> @llvm.experimental.constrained.trunc.v2f64(<2 x double> [[VRNDZ_I]], metadata !"fpexcept.strict") #[[ATTR2:[0-9]+]]
|
|
// CONSTRAINED-NEXT: store <2 x double> [[VRNDZ1_I]], ptr [[REF_TMP_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP2:%.*]] = load <2 x double>, ptr [[REF_TMP_I]], align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[TMP2]], ptr [[__RET_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP3:%.*]] = load <2 x double>, ptr [[__RET_I]], align 16
|
|
// CONSTRAINED-NEXT: ret <2 x double> [[TMP3]]
|
|
//
|
|
float64x2_t rnd5(float64x2_t a) { return vrndq_f64(a); }
|
|
|
|
// UNCONSTRAINED-LABEL: define <2 x double> @rnd13(
|
|
// UNCONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[__P0_ADDR_I:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: [[__RET_I:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: [[REF_TMP_I:%.*]] = alloca <16 x i8>, align 16
|
|
// UNCONSTRAINED-NEXT: [[A_ADDR:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[A]], ptr [[A_ADDR]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = load <2 x double>, ptr [[A_ADDR]], align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[TMP0]], ptr [[__P0_ADDR_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP1:%.*]] = load <16 x i8>, ptr [[__P0_ADDR_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[VRNDM_I:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x double>
|
|
// UNCONSTRAINED-NEXT: [[VRNDM1_I:%.*]] = call <2 x double> @llvm.floor.v2f64(<2 x double> [[VRNDM_I]])
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[VRNDM1_I]], ptr [[REF_TMP_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP2:%.*]] = load <2 x double>, ptr [[REF_TMP_I]], align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[TMP2]], ptr [[__RET_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP3:%.*]] = load <2 x double>, ptr [[__RET_I]], align 16
|
|
// UNCONSTRAINED-NEXT: ret <2 x double> [[TMP3]]
|
|
//
|
|
// CONSTRAINED-LABEL: define <2 x double> @rnd13(
|
|
// CONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[__P0_ADDR_I:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: [[__RET_I:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: [[REF_TMP_I:%.*]] = alloca <16 x i8>, align 16
|
|
// CONSTRAINED-NEXT: [[A_ADDR:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[A]], ptr [[A_ADDR]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = load <2 x double>, ptr [[A_ADDR]], align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[TMP0]], ptr [[__P0_ADDR_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP1:%.*]] = load <16 x i8>, ptr [[__P0_ADDR_I]], align 16
|
|
// CONSTRAINED-NEXT: [[VRNDM_I:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x double>
|
|
// CONSTRAINED-NEXT: [[VRNDM1_I:%.*]] = call <2 x double> @llvm.experimental.constrained.floor.v2f64(<2 x double> [[VRNDM_I]], metadata !"fpexcept.strict") #[[ATTR2]]
|
|
// CONSTRAINED-NEXT: store <2 x double> [[VRNDM1_I]], ptr [[REF_TMP_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP2:%.*]] = load <2 x double>, ptr [[REF_TMP_I]], align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[TMP2]], ptr [[__RET_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP3:%.*]] = load <2 x double>, ptr [[__RET_I]], align 16
|
|
// CONSTRAINED-NEXT: ret <2 x double> [[TMP3]]
|
|
//
|
|
float64x2_t rnd13(float64x2_t a) { return vrndmq_f64(a); }
|
|
|
|
// UNCONSTRAINED-LABEL: define <2 x double> @rnd18(
|
|
// UNCONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[__P0_ADDR_I:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: [[__RET_I:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: [[REF_TMP_I:%.*]] = alloca <16 x i8>, align 16
|
|
// UNCONSTRAINED-NEXT: [[A_ADDR:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[A]], ptr [[A_ADDR]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = load <2 x double>, ptr [[A_ADDR]], align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[TMP0]], ptr [[__P0_ADDR_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP1:%.*]] = load <16 x i8>, ptr [[__P0_ADDR_I]], align 16
|
|
// 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: store <2 x double> [[VRNDP1_I]], ptr [[REF_TMP_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP2:%.*]] = load <2 x double>, ptr [[REF_TMP_I]], align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[TMP2]], ptr [[__RET_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP3:%.*]] = load <2 x double>, ptr [[__RET_I]], align 16
|
|
// UNCONSTRAINED-NEXT: ret <2 x double> [[TMP3]]
|
|
//
|
|
// CONSTRAINED-LABEL: define <2 x double> @rnd18(
|
|
// CONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[__P0_ADDR_I:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: [[__RET_I:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: [[REF_TMP_I:%.*]] = alloca <16 x i8>, align 16
|
|
// CONSTRAINED-NEXT: [[A_ADDR:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[A]], ptr [[A_ADDR]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = load <2 x double>, ptr [[A_ADDR]], align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[TMP0]], ptr [[__P0_ADDR_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP1:%.*]] = load <16 x i8>, ptr [[__P0_ADDR_I]], align 16
|
|
// 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: store <2 x double> [[VRNDP1_I]], ptr [[REF_TMP_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP2:%.*]] = load <2 x double>, ptr [[REF_TMP_I]], align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[TMP2]], ptr [[__RET_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP3:%.*]] = load <2 x double>, ptr [[__RET_I]], align 16
|
|
// CONSTRAINED-NEXT: ret <2 x double> [[TMP3]]
|
|
//
|
|
float64x2_t rnd18(float64x2_t a) { return vrndpq_f64(a); }
|
|
|
|
// UNCONSTRAINED-LABEL: define <2 x double> @rnd22(
|
|
// UNCONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[__P0_ADDR_I:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: [[__RET_I:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: [[REF_TMP_I:%.*]] = alloca <16 x i8>, align 16
|
|
// UNCONSTRAINED-NEXT: [[A_ADDR:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[A]], ptr [[A_ADDR]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = load <2 x double>, ptr [[A_ADDR]], align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[TMP0]], ptr [[__P0_ADDR_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP1:%.*]] = load <16 x i8>, ptr [[__P0_ADDR_I]], align 16
|
|
// 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: store <2 x double> [[VRNDA1_I]], ptr [[REF_TMP_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP2:%.*]] = load <2 x double>, ptr [[REF_TMP_I]], align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[TMP2]], ptr [[__RET_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP3:%.*]] = load <2 x double>, ptr [[__RET_I]], align 16
|
|
// UNCONSTRAINED-NEXT: ret <2 x double> [[TMP3]]
|
|
//
|
|
// CONSTRAINED-LABEL: define <2 x double> @rnd22(
|
|
// CONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[__P0_ADDR_I:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: [[__RET_I:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: [[REF_TMP_I:%.*]] = alloca <16 x i8>, align 16
|
|
// CONSTRAINED-NEXT: [[A_ADDR:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[A]], ptr [[A_ADDR]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = load <2 x double>, ptr [[A_ADDR]], align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[TMP0]], ptr [[__P0_ADDR_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP1:%.*]] = load <16 x i8>, ptr [[__P0_ADDR_I]], align 16
|
|
// 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]]
|
|
// CONSTRAINED-NEXT: store <2 x double> [[VRNDA1_I]], ptr [[REF_TMP_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP2:%.*]] = load <2 x double>, ptr [[REF_TMP_I]], align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[TMP2]], ptr [[__RET_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP3:%.*]] = load <2 x double>, ptr [[__RET_I]], align 16
|
|
// CONSTRAINED-NEXT: ret <2 x double> [[TMP3]]
|
|
//
|
|
float64x2_t rnd22(float64x2_t a) { return vrndaq_f64(a); }
|
|
|
|
// UNCONSTRAINED-LABEL: define <2 x double> @rnd25(
|
|
// UNCONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// UNCONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// UNCONSTRAINED-NEXT: [[__P0_ADDR_I:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: [[__RET_I:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: [[REF_TMP_I:%.*]] = alloca <16 x i8>, align 16
|
|
// UNCONSTRAINED-NEXT: [[A_ADDR:%.*]] = alloca <2 x double>, align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[A]], ptr [[A_ADDR]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP0:%.*]] = load <2 x double>, ptr [[A_ADDR]], align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[TMP0]], ptr [[__P0_ADDR_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP1:%.*]] = load <16 x i8>, ptr [[__P0_ADDR_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[VRNDX_I:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x double>
|
|
// UNCONSTRAINED-NEXT: [[VRNDX1_I:%.*]] = call <2 x double> @llvm.rint.v2f64(<2 x double> [[VRNDX_I]])
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[VRNDX1_I]], ptr [[REF_TMP_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP2:%.*]] = load <2 x double>, ptr [[REF_TMP_I]], align 16
|
|
// UNCONSTRAINED-NEXT: store <2 x double> [[TMP2]], ptr [[__RET_I]], align 16
|
|
// UNCONSTRAINED-NEXT: [[TMP3:%.*]] = load <2 x double>, ptr [[__RET_I]], align 16
|
|
// UNCONSTRAINED-NEXT: ret <2 x double> [[TMP3]]
|
|
//
|
|
// CONSTRAINED-LABEL: define <2 x double> @rnd25(
|
|
// CONSTRAINED-SAME: <2 x double> noundef [[A:%.*]]) #[[ATTR0]] {
|
|
// CONSTRAINED-NEXT: [[ENTRY:.*:]]
|
|
// CONSTRAINED-NEXT: [[__P0_ADDR_I:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: [[__RET_I:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: [[REF_TMP_I:%.*]] = alloca <16 x i8>, align 16
|
|
// CONSTRAINED-NEXT: [[A_ADDR:%.*]] = alloca <2 x double>, align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[A]], ptr [[A_ADDR]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP0:%.*]] = load <2 x double>, ptr [[A_ADDR]], align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[TMP0]], ptr [[__P0_ADDR_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP1:%.*]] = load <16 x i8>, ptr [[__P0_ADDR_I]], align 16
|
|
// CONSTRAINED-NEXT: [[VRNDX_I:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x double>
|
|
// CONSTRAINED-NEXT: [[VRNDX1_I:%.*]] = call <2 x double> @llvm.experimental.constrained.rint.v2f64(<2 x double> [[VRNDX_I]], metadata !"round.tonearest", metadata !"fpexcept.strict") #[[ATTR2]]
|
|
// CONSTRAINED-NEXT: store <2 x double> [[VRNDX1_I]], ptr [[REF_TMP_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP2:%.*]] = load <2 x double>, ptr [[REF_TMP_I]], align 16
|
|
// CONSTRAINED-NEXT: store <2 x double> [[TMP2]], ptr [[__RET_I]], align 16
|
|
// CONSTRAINED-NEXT: [[TMP3:%.*]] = load <2 x double>, ptr [[__RET_I]], align 16
|
|
// CONSTRAINED-NEXT: ret <2 x double> [[TMP3]]
|
|
//
|
|
float64x2_t rnd25(float64x2_t a) { return vrndxq_f64(a); }
|
|
|