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
89 lines
4.7 KiB
LLVM
89 lines
4.7 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
|
|
; RUN: opt -passes=slp-vectorizer -S -slp-revec -slp-max-reg-size=1024 -slp-threshold=-1000 %s | FileCheck %s
|
|
|
|
; The 4 stores can be re-vectorised, make sure the poison sources
|
|
; are safely handled when trying to vectorise [ %0, poison, poison, %1 ]
|
|
define void @test_missing_lanes_1_2(ptr %ptr, i32 %val0, i32 %val1) {
|
|
; CHECK-LABEL: @test_missing_lanes_1_2(
|
|
; CHECK-NEXT: entry:
|
|
; CHECK-NEXT: [[TMP0:%.*]] = insertelement <4 x i32> <i32 poison, i32 0, i32 0, i32 0>, i32 [[VAL0:%.*]], i32 0
|
|
; CHECK-NEXT: [[TMP1:%.*]] = insertelement <4 x i32> <i32 poison, i32 0, i32 0, i32 0>, i32 [[VAL1:%.*]], i32 0
|
|
; CHECK-NEXT: [[GETELEMENTPTR0:%.*]] = getelementptr i32, ptr [[PTR:%.*]], i64 0
|
|
; CHECK-NEXT: store <4 x i32> [[TMP0]], ptr [[GETELEMENTPTR0]], align 4
|
|
; CHECK-NEXT: [[GETELEMENTPTR1:%.*]] = getelementptr i32, ptr [[PTR]], i64 4
|
|
; CHECK-NEXT: store <8 x i32> poison, ptr [[GETELEMENTPTR1]], align 4
|
|
; CHECK-NEXT: [[GETELEMENTPTR3:%.*]] = getelementptr i32, ptr [[PTR]], i64 12
|
|
; CHECK-NEXT: store <4 x i32> [[TMP1]], ptr [[GETELEMENTPTR3]], align 4
|
|
; CHECK-NEXT: ret void
|
|
;
|
|
entry:
|
|
%0 = insertelement <4 x i32> <i32 poison, i32 0, i32 0, i32 0>, i32 %val0, i32 0
|
|
%1 = insertelement <4 x i32> <i32 poison, i32 0, i32 0, i32 0>, i32 %val1, i32 0
|
|
|
|
%getelementptr0 = getelementptr i32, ptr %ptr, i64 0
|
|
store <4 x i32> %0, ptr %getelementptr0, align 4
|
|
%getelementptr1 = getelementptr i32, ptr %ptr, i64 4
|
|
store <4 x i32> poison, ptr %getelementptr1, align 4
|
|
%getelementptr2 = getelementptr i32, ptr %ptr, i64 8
|
|
store <4 x i32> poison, ptr %getelementptr2, align 4
|
|
%getelementptr3 = getelementptr i32, ptr %ptr, i64 12
|
|
store <4 x i32> %1, ptr %getelementptr3, align 4
|
|
|
|
ret void
|
|
}
|
|
|
|
; The 4 stores can be re-vectorised, make sure the poison sources
|
|
; are safely handled when trying to vectorise [ %0, poison, %1, poison ]
|
|
define void @test_missing_lanes_1_3(ptr %ptr, i32 %val0, i32 %val1) {
|
|
; CHECK-LABEL: @test_missing_lanes_1_3(
|
|
; CHECK-NEXT: entry:
|
|
; CHECK-NEXT: [[TMP0:%.*]] = insertelement <4 x i32> <i32 poison, i32 0, i32 0, i32 0>, i32 [[VAL0:%.*]], i32 0
|
|
; CHECK-NEXT: [[TMP1:%.*]] = insertelement <4 x i32> <i32 poison, i32 0, i32 0, i32 0>, i32 [[VAL1:%.*]], i32 0
|
|
; CHECK-NEXT: [[GETELEMENTPTR0:%.*]] = getelementptr i32, ptr [[PTR:%.*]], i64 0
|
|
; CHECK-NEXT: store <4 x i32> [[TMP0]], ptr [[GETELEMENTPTR0]], align 4
|
|
; CHECK-NEXT: [[GETELEMENTPTR1:%.*]] = getelementptr i32, ptr [[PTR]], i64 4
|
|
; CHECK-NEXT: [[TMP2:%.*]] = shufflevector <4 x i32> [[TMP1]], <4 x i32> poison, <8 x i32> <i32 poison, i32 poison, i32 poison, i32 poison, i32 0, i32 1, i32 2, i32 3>
|
|
; CHECK-NEXT: store <8 x i32> [[TMP2]], ptr [[GETELEMENTPTR1]], align 4
|
|
; CHECK-NEXT: [[GETELEMENTPTR3:%.*]] = getelementptr i32, ptr [[PTR]], i64 12
|
|
; CHECK-NEXT: store <4 x i32> poison, ptr [[GETELEMENTPTR3]], align 4
|
|
; CHECK-NEXT: ret void
|
|
;
|
|
entry:
|
|
%0 = insertelement <4 x i32> <i32 poison, i32 0, i32 0, i32 0>, i32 %val0, i32 0
|
|
%1 = insertelement <4 x i32> <i32 poison, i32 0, i32 0, i32 0>, i32 %val1, i32 0
|
|
|
|
%getelementptr0 = getelementptr i32, ptr %ptr, i64 0
|
|
store <4 x i32> %0, ptr %getelementptr0, align 4
|
|
%getelementptr1 = getelementptr i32, ptr %ptr, i64 4
|
|
store <4 x i32> poison, ptr %getelementptr1, align 4
|
|
%getelementptr2 = getelementptr i32, ptr %ptr, i64 8
|
|
store <4 x i32> %1, ptr %getelementptr2, align 4
|
|
%getelementptr3 = getelementptr i32, ptr %ptr, i64 12
|
|
store <4 x i32> poison, ptr %getelementptr3, align 4
|
|
|
|
ret void
|
|
}
|
|
|
|
; This could be re-vectorised to use a store <8 x i32> instruction.
|
|
define void @test_valid_value_operands(ptr %ptr, i32 %val0, i32 %val1) {
|
|
; CHECK-LABEL: @test_valid_value_operands(
|
|
; CHECK-NEXT: entry:
|
|
; CHECK-NEXT: [[TMP0:%.*]] = insertelement <4 x i32> <i32 poison, i32 0, i32 0, i32 0>, i32 [[VAL0:%.*]], i32 0
|
|
; CHECK-NEXT: [[TMP1:%.*]] = insertelement <4 x i32> <i32 poison, i32 0, i32 0, i32 0>, i32 [[VAL1:%.*]], i32 0
|
|
; CHECK-NEXT: [[GETELEMENTPTR0:%.*]] = getelementptr i32, ptr [[PTR:%.*]], i64 0
|
|
; CHECK-NEXT: store <4 x i32> [[TMP0]], ptr [[GETELEMENTPTR0]], align 4
|
|
; CHECK-NEXT: [[GETELEMENTPTR1:%.*]] = getelementptr i32, ptr [[PTR]], i64 4
|
|
; CHECK-NEXT: store <4 x i32> [[TMP1]], ptr [[GETELEMENTPTR1]], align 4
|
|
; CHECK-NEXT: ret void
|
|
;
|
|
entry:
|
|
%0 = insertelement <4 x i32> <i32 poison, i32 0, i32 0, i32 0>, i32 %val0, i32 0
|
|
%1 = insertelement <4 x i32> <i32 poison, i32 0, i32 0, i32 0>, i32 %val1, i32 0
|
|
|
|
%getelementptr0 = getelementptr i32, ptr %ptr, i64 0
|
|
store <4 x i32> %0, ptr %getelementptr0, align 4
|
|
%getelementptr1 = getelementptr i32, ptr %ptr, i64 4
|
|
store <4 x i32> %1, ptr %getelementptr1, align 4
|
|
ret void
|
|
}
|