121 lines · plain
1; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals all --version 42; RUN: opt < %s -passes=amdgpu-sw-lower-lds -amdgpu-asan-instrument-lds=false -S -mtriple=amdgcn-amd-amdhsa | FileCheck %s3 4; Test to check if static and dynamic LDS accesses are lowered correctly in kernel.5@lds_1 = internal addrspace(3) global [1 x i8] poison, align 46@lds_2 = internal addrspace(3) global [1 x i32] poison, align 87@lds_3 = external addrspace(3) global [0 x i8], align 48@lds_4 = external addrspace(3) global [0 x i8], align 89 10;.11; CHECK: @llvm.amdgcn.sw.lds.k0 = internal addrspace(3) global ptr poison, no_sanitize_address, align 8, !absolute_symbol [[META0:![0-9]+]]12; CHECK: @llvm.amdgcn.k0.dynlds = external addrspace(3) global [0 x i8], no_sanitize_address, align 8, !absolute_symbol [[META1:![0-9]+]]13; CHECK: @llvm.amdgcn.sw.lds.k0.md = internal addrspace(1) global %llvm.amdgcn.sw.lds.k0.md.type { %llvm.amdgcn.sw.lds.k0.md.item { i32 0, i32 8, i32 32 }, %llvm.amdgcn.sw.lds.k0.md.item { i32 32, i32 1, i32 32 }, %llvm.amdgcn.sw.lds.k0.md.item { i32 64, i32 4, i32 32 }, %llvm.amdgcn.sw.lds.k0.md.item { i32 96, i32 0, i32 32 }, %llvm.amdgcn.sw.lds.k0.md.item { i32 128, i32 0, i32 32 } }, no_sanitize_address14;.15define amdgpu_kernel void @k0() sanitize_address {16; CHECK-LABEL: define amdgpu_kernel void @k0(17; CHECK-SAME: ) #[[ATTR0:[0-9]+]] {18; CHECK-NEXT: WId:19; CHECK-NEXT: [[TMP0:%.*]] = call i32 @llvm.amdgcn.workitem.id.x()20; CHECK-NEXT: [[TMP1:%.*]] = call i32 @llvm.amdgcn.workitem.id.y()21; CHECK-NEXT: [[TMP2:%.*]] = call i32 @llvm.amdgcn.workitem.id.z()22; CHECK-NEXT: [[TMP3:%.*]] = or i32 [[TMP0]], [[TMP1]]23; CHECK-NEXT: [[TMP4:%.*]] = or i32 [[TMP3]], [[TMP2]]24; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i32 [[TMP4]], 025; CHECK-NEXT: br i1 [[TMP5]], label [[MALLOC:%.*]], label [[TMP21:%.*]]26; CHECK: Malloc:27; CHECK-NEXT: [[TMP9:%.*]] = load i32, ptr addrspace(1) getelementptr inbounds ([[LLVM_AMDGCN_SW_LDS_K0_MD_TYPE:%.*]], ptr addrspace(1) @llvm.amdgcn.sw.lds.k0.md, i32 0, i32 2, i32 0), align 428; CHECK-NEXT: [[TMP7:%.*]] = load i32, ptr addrspace(1) getelementptr inbounds ([[LLVM_AMDGCN_SW_LDS_K0_MD_TYPE]], ptr addrspace(1) @llvm.amdgcn.sw.lds.k0.md, i32 0, i32 2, i32 2), align 429; CHECK-NEXT: [[TMP8:%.*]] = add i32 [[TMP9]], [[TMP7]]30; CHECK-NEXT: [[TMP6:%.*]] = call ptr addrspace(4) @llvm.amdgcn.implicitarg.ptr()31; CHECK-NEXT: [[TMP10:%.*]] = getelementptr inbounds ptr addrspace(4), ptr addrspace(4) [[TMP6]], i64 1532; CHECK-NEXT: store i32 [[TMP8]], ptr addrspace(1) getelementptr inbounds ([[LLVM_AMDGCN_SW_LDS_K0_MD_TYPE]], ptr addrspace(1) @llvm.amdgcn.sw.lds.k0.md, i32 0, i32 3, i32 0), align 433; CHECK-NEXT: [[TMP11:%.*]] = load i32, ptr addrspace(4) [[TMP10]], align 434; CHECK-NEXT: store i32 [[TMP11]], ptr addrspace(1) getelementptr inbounds ([[LLVM_AMDGCN_SW_LDS_K0_MD_TYPE]], ptr addrspace(1) @llvm.amdgcn.sw.lds.k0.md, i32 0, i32 3, i32 1), align 435; CHECK-NEXT: [[TMP12:%.*]] = add i32 [[TMP11]], 736; CHECK-NEXT: [[TMP13:%.*]] = udiv i32 [[TMP12]], 837; CHECK-NEXT: [[TMP14:%.*]] = mul i32 [[TMP13]], 838; CHECK-NEXT: store i32 [[TMP14]], ptr addrspace(1) getelementptr inbounds ([[LLVM_AMDGCN_SW_LDS_K0_MD_TYPE]], ptr addrspace(1) @llvm.amdgcn.sw.lds.k0.md, i32 0, i32 3, i32 2), align 439; CHECK-NEXT: [[TMP15:%.*]] = add i32 [[TMP8]], [[TMP14]]40; CHECK-NEXT: store i32 [[TMP15]], ptr addrspace(1) getelementptr inbounds ([[LLVM_AMDGCN_SW_LDS_K0_MD_TYPE]], ptr addrspace(1) @llvm.amdgcn.sw.lds.k0.md, i32 0, i32 4, i32 0), align 441; CHECK-NEXT: [[TMP31:%.*]] = load i32, ptr addrspace(4) [[TMP10]], align 442; CHECK-NEXT: store i32 [[TMP31]], ptr addrspace(1) getelementptr inbounds ([[LLVM_AMDGCN_SW_LDS_K0_MD_TYPE]], ptr addrspace(1) @llvm.amdgcn.sw.lds.k0.md, i32 0, i32 4, i32 1), align 443; CHECK-NEXT: [[TMP17:%.*]] = add i32 [[TMP31]], 744; CHECK-NEXT: [[TMP18:%.*]] = udiv i32 [[TMP17]], 845; CHECK-NEXT: [[TMP19:%.*]] = mul i32 [[TMP18]], 846; CHECK-NEXT: store i32 [[TMP19]], ptr addrspace(1) getelementptr inbounds ([[LLVM_AMDGCN_SW_LDS_K0_MD_TYPE]], ptr addrspace(1) @llvm.amdgcn.sw.lds.k0.md, i32 0, i32 4, i32 2), align 447; CHECK-NEXT: [[TMP32:%.*]] = add i32 [[TMP15]], [[TMP19]]48; CHECK-NEXT: [[TMP30:%.*]] = zext i32 [[TMP32]] to i6449; CHECK-NEXT: [[TMP22:%.*]] = call ptr @llvm.returnaddress(i32 0)50; CHECK-NEXT: [[TMP23:%.*]] = ptrtoint ptr [[TMP22]] to i6451; CHECK-NEXT: [[TMP39:%.*]] = call i64 @__asan_malloc_impl(i64 [[TMP30]], i64 [[TMP23]])52; CHECK-NEXT: [[TMP20:%.*]] = inttoptr i64 [[TMP39]] to ptr addrspace(1)53; CHECK-NEXT: store ptr addrspace(1) [[TMP20]], ptr addrspace(3) @llvm.amdgcn.sw.lds.k0, align 854; CHECK-NEXT: [[TMP40:%.*]] = getelementptr inbounds i8, ptr addrspace(1) [[TMP20]], i64 855; CHECK-NEXT: [[TMP41:%.*]] = ptrtoint ptr addrspace(1) [[TMP40]] to i6456; CHECK-NEXT: call void @__asan_poison_region(i64 [[TMP41]], i64 24)57; CHECK-NEXT: [[TMP57:%.*]] = getelementptr inbounds i8, ptr addrspace(1) [[TMP20]], i64 3358; CHECK-NEXT: [[TMP111:%.*]] = ptrtoint ptr addrspace(1) [[TMP57]] to i6459; CHECK-NEXT: call void @__asan_poison_region(i64 [[TMP111]], i64 31)60; CHECK-NEXT: [[TMP112:%.*]] = getelementptr inbounds i8, ptr addrspace(1) [[TMP20]], i64 6861; CHECK-NEXT: [[TMP113:%.*]] = ptrtoint ptr addrspace(1) [[TMP112]] to i6462; CHECK-NEXT: call void @__asan_poison_region(i64 [[TMP113]], i64 28)63; CHECK-NEXT: br label [[TMP21]]64; CHECK: 32:65; CHECK-NEXT: [[XYZCOND:%.*]] = phi i1 [ false, [[WID:%.*]] ], [ true, [[MALLOC]] ]66; CHECK-NEXT: call void @llvm.amdgcn.s.barrier()67; CHECK-NEXT: [[TMP35:%.*]] = load ptr addrspace(1), ptr addrspace(3) @llvm.amdgcn.sw.lds.k0, align 868; CHECK-NEXT: [[TMP24:%.*]] = load i32, ptr addrspace(1) getelementptr inbounds ([[LLVM_AMDGCN_SW_LDS_K0_MD_TYPE]], ptr addrspace(1) @llvm.amdgcn.sw.lds.k0.md, i32 0, i32 1, i32 0), align 469; CHECK-NEXT: [[TMP25:%.*]] = getelementptr inbounds i8, ptr addrspace(3) @llvm.amdgcn.sw.lds.k0, i32 [[TMP24]]70; CHECK-NEXT: [[TMP26:%.*]] = load i32, ptr addrspace(1) getelementptr inbounds ([[LLVM_AMDGCN_SW_LDS_K0_MD_TYPE]], ptr addrspace(1) @llvm.amdgcn.sw.lds.k0.md, i32 0, i32 2, i32 0), align 471; CHECK-NEXT: [[TMP27:%.*]] = getelementptr inbounds i8, ptr addrspace(3) @llvm.amdgcn.sw.lds.k0, i32 [[TMP26]]72; CHECK-NEXT: [[TMP28:%.*]] = load i32, ptr addrspace(1) getelementptr inbounds ([[LLVM_AMDGCN_SW_LDS_K0_MD_TYPE]], ptr addrspace(1) @llvm.amdgcn.sw.lds.k0.md, i32 0, i32 3, i32 0), align 473; CHECK-NEXT: [[TMP29:%.*]] = getelementptr inbounds i8, ptr addrspace(3) @llvm.amdgcn.sw.lds.k0, i32 [[TMP28]]74; CHECK-NEXT: [[TMP33:%.*]] = load i32, ptr addrspace(1) getelementptr inbounds ([[LLVM_AMDGCN_SW_LDS_K0_MD_TYPE]], ptr addrspace(1) @llvm.amdgcn.sw.lds.k0.md, i32 0, i32 4, i32 0), align 475; CHECK-NEXT: [[TMP34:%.*]] = getelementptr inbounds i8, ptr addrspace(3) @llvm.amdgcn.sw.lds.k0, i32 [[TMP33]]76; CHECK-NEXT: call void @llvm.donothing() [ "ExplicitUse"(ptr addrspace(3) @llvm.amdgcn.k0.dynlds) ]77; CHECK-NEXT: [[TMP42:%.*]] = ptrtoint ptr addrspace(3) [[TMP25]] to i3278; CHECK-NEXT: [[TMP43:%.*]] = getelementptr inbounds i8, ptr addrspace(1) [[TMP35]], i32 [[TMP42]]79; CHECK-NEXT: store i8 7, ptr addrspace(1) [[TMP43]], align 480; CHECK-NEXT: [[TMP59:%.*]] = ptrtoint ptr addrspace(3) [[TMP27]] to i3281; CHECK-NEXT: [[TMP60:%.*]] = getelementptr inbounds i8, ptr addrspace(1) [[TMP35]], i32 [[TMP59]]82; CHECK-NEXT: store i32 8, ptr addrspace(1) [[TMP60]], align 883; CHECK-NEXT: [[TMP77:%.*]] = ptrtoint ptr addrspace(3) [[TMP29]] to i3284; CHECK-NEXT: [[TMP78:%.*]] = getelementptr inbounds i8, ptr addrspace(1) [[TMP35]], i32 [[TMP77]]85; CHECK-NEXT: store i8 7, ptr addrspace(1) [[TMP78]], align 486; CHECK-NEXT: [[TMP94:%.*]] = ptrtoint ptr addrspace(3) [[TMP34]] to i3287; CHECK-NEXT: [[TMP95:%.*]] = getelementptr inbounds i8, ptr addrspace(1) [[TMP35]], i32 [[TMP94]]88; CHECK-NEXT: store i8 8, ptr addrspace(1) [[TMP95]], align 889; CHECK-NEXT: br label [[CONDFREE1:%.*]]90; CHECK: CondFree:91; CHECK-NEXT: call void @llvm.amdgcn.s.barrier()92; CHECK-NEXT: br i1 [[XYZCOND]], label [[FREE:%.*]], label [[END:%.*]]93; CHECK: Free:94; CHECK-NEXT: [[TMP36:%.*]] = call ptr @llvm.returnaddress(i32 0)95; CHECK-NEXT: [[TMP37:%.*]] = ptrtoint ptr [[TMP36]] to i6496; CHECK-NEXT: [[TMP38:%.*]] = ptrtoint ptr addrspace(1) [[TMP35]] to i6497; CHECK-NEXT: call void @__asan_free_impl(i64 [[TMP38]], i64 [[TMP37]])98; CHECK-NEXT: br label [[END]]99; CHECK: End:100; CHECK-NEXT: ret void101;102 store i8 7, ptr addrspace(3) @lds_1, align 4103 store i32 8, ptr addrspace(3) @lds_2, align 8104 store i8 7, ptr addrspace(3) @lds_3, align 4105 store i8 8, ptr addrspace(3) @lds_4, align 8106 ret void107}108 109!llvm.module.flags = !{!0}110!0 = !{i32 4, !"nosanitize_address", i32 1}111 112;.113; CHECK: attributes #[[ATTR0]] = { sanitize_address "amdgpu-lds-size"="8,8" }114; CHECK: attributes #[[ATTR1:[0-9]+]] = { nocallback nofree nosync nounwind willreturn memory(none) }115; CHECK: attributes #[[ATTR2:[0-9]+]] = { nocallback nofree nosync nounwind speculatable willreturn memory(none) }116; CHECK: attributes #[[ATTR3:[0-9]+]] = { convergent nocallback nofree nounwind willreturn }117;.118; CHECK: [[META0]] = !{i32 0, i32 1}119; CHECK: [[META1]] = !{i32 8, i32 9}120;.121