1; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --tool llc --version 5 2; RUN: llc %s -mtriple=dxil-pc-shadermodel6.3-library --filetype=asm -o - | FileCheck %s 3 4; Make sure we can load groupshared, static vectors and arrays of vectors 5 6@"arrayofVecData" = local_unnamed_addr addrspace(3) global [2 x <3 x float>] zeroinitializer, align 16 7@"vecData" = external addrspace(3) global <4 x i32>, align 4 8@staticArrayOfVecData = internal global [3 x <4 x i32>] [<4 x i32> <i32 1, i32 2, i32 3, i32 4>, <4 x i32> <i32 5, i32 6, i32 7, i32 8>, <4 x i32> <i32 9, i32 10, i32 11, i32 12>], align 4 9@"groushared2dArrayofVectors" = local_unnamed_addr addrspace(3) global [3 x [ 3 x <4 x i32>]] zeroinitializer, align 16 10 11; CHECK: @arrayofVecData.scalarized.1dim = local_unnamed_addr addrspace(3) global [6 x float] zeroinitializer, align 16 12; CHECK: @vecData.scalarized = external addrspace(3) global [4 x i32], align 4 13; CHECK: @staticArrayOfVecData.scalarized.1dim = internal global [12 x i32] [i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12], align 4 14; CHECK: @groushared2dArrayofVectors.scalarized.1dim = local_unnamed_addr addrspace(3) global [36 x i32] zeroinitializer, align 16 15 16; CHECK-NOT: @arrayofVecData 17; CHECK-NOT: @arrayofVecData.scalarized 18; CHECK-NOT: @vecData 19; CHECK-NOT: @staticArrayOfVecData 20; CHECK-NOT: @staticArrayOfVecData.scalarized 21; CHECK-NOT: @groushared2dArrayofVectors 22; CHECK-NOT: @groushared2dArrayofVectors.scalarized 23 24define <4 x i32> @load_array_vec_test() #0 { 25; CHECK-LABEL: define <4 x i32> @load_array_vec_test( 26; CHECK-SAME: ) #[[ATTR0:[0-9]+]] { 27; CHECK-NEXT: [[TMP1:%.*]] = bitcast ptr addrspace(3) @arrayofVecData.scalarized.1dim to ptr addrspace(3) 28; CHECK-NEXT: [[TMP2:%.*]] = load i32, ptr addrspace(3) [[TMP1]], align 4 29; CHECK-NEXT: [[TMP3:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) @arrayofVecData.scalarized.1dim, i32 1) to ptr addrspace(3) 30; CHECK-NEXT: [[TMP4:%.*]] = load i32, ptr addrspace(3) [[TMP3]], align 4 31; CHECK-NEXT: [[TMP5:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) @arrayofVecData.scalarized.1dim, i32 2) to ptr addrspace(3) 32; CHECK-NEXT: [[TMP6:%.*]] = load i32, ptr addrspace(3) [[TMP5]], align 4 33; CHECK-NEXT: [[TMP7:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) @arrayofVecData.scalarized.1dim, i32 3) to ptr addrspace(3) 34; CHECK-NEXT: [[TMP8:%.*]] = load i32, ptr addrspace(3) [[TMP7]], align 4 35; CHECK-NEXT: [[TMP11:%.*]] = bitcast ptr addrspace(3) getelementptr inbounds ([6 x float], ptr addrspace(3) @arrayofVecData.scalarized.1dim, i32 1) to ptr addrspace(3) 36; CHECK-NEXT: [[TMP12:%.*]] = load i32, ptr addrspace(3) [[TMP11]], align 4 37; CHECK-NEXT: [[DOTI12:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) getelementptr inbounds ([6 x float], ptr addrspace(3) @arrayofVecData.scalarized.1dim, i32 1), i32 1) to ptr addrspace(3) 38; CHECK-NEXT: [[DOTI13:%.*]] = load i32, ptr addrspace(3) [[DOTI12]], align 4 39; CHECK-NEXT: [[DOTI24:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) getelementptr inbounds ([6 x float], ptr addrspace(3) @arrayofVecData.scalarized.1dim, i32 1), i32 2) to ptr addrspace(3) 40; CHECK-NEXT: [[DOTI25:%.*]] = load i32, ptr addrspace(3) [[DOTI24]], align 4 41; CHECK-NEXT: [[DOTI36:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) getelementptr inbounds ([6 x float], ptr addrspace(3) @arrayofVecData.scalarized.1dim, i32 1), i32 3) to ptr addrspace(3) 42; CHECK-NEXT: [[DOTI37:%.*]] = load i32, ptr addrspace(3) [[DOTI36]], align 4 43; CHECK-NEXT: [[DOTI08:%.*]] = add i32 [[TMP2]], [[TMP12]] 44; CHECK-NEXT: [[DOTI19:%.*]] = add i32 [[TMP4]], [[DOTI13]] 45; CHECK-NEXT: [[DOTI210:%.*]] = add i32 [[TMP6]], [[DOTI25]] 46; CHECK-NEXT: [[DOTI311:%.*]] = add i32 [[TMP8]], [[DOTI37]] 47; CHECK-NEXT: [[DOTUPTO015:%.*]] = insertelement <4 x i32> poison, i32 [[DOTI08]], i64 0 48; CHECK-NEXT: [[DOTUPTO116:%.*]] = insertelement <4 x i32> [[DOTUPTO015]], i32 [[DOTI19]], i64 1 49; CHECK-NEXT: [[DOTUPTO217:%.*]] = insertelement <4 x i32> [[DOTUPTO116]], i32 [[DOTI210]], i64 2 50; CHECK-NEXT: [[TMP16:%.*]] = insertelement <4 x i32> [[DOTUPTO217]], i32 [[DOTI311]], i64 3 51; CHECK-NEXT: ret <4 x i32> [[TMP16]] 52; 53 %1 = load <4 x i32>, <4 x i32> addrspace(3)* getelementptr inbounds ([2 x <4 x i32>], [2 x <4 x i32>] addrspace(3)* @"arrayofVecData", i32 0, i32 0), align 4 54 %2 = load <4 x i32>, <4 x i32> addrspace(3)* getelementptr inbounds ([2 x <4 x i32>], [2 x <4 x i32>] addrspace(3)* @"arrayofVecData", i32 0, i32 1), align 4 55 %3 = add <4 x i32> %1, %2 56 ret <4 x i32> %3 57} 58 59 60define <4 x i32> @load_vec_test() #0 { 61; CHECK-LABEL: define <4 x i32> @load_vec_test( 62; CHECK-SAME: ) #[[ATTR0]] { 63; CHECK-NEXT: [[TMP1:%.*]] = bitcast ptr addrspace(3) @vecData.scalarized to ptr addrspace(3) 64; CHECK-NEXT: [[TMP2:%.*]] = load i32, ptr addrspace(3) [[TMP1]], align 4 65; CHECK-NEXT: [[TMP3:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) @vecData.scalarized, i32 1) to ptr addrspace(3) 66; CHECK-NEXT: [[TMP4:%.*]] = load i32, ptr addrspace(3) [[TMP3]], align 4 67; CHECK-NEXT: [[TMP5:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) @vecData.scalarized, i32 2) to ptr addrspace(3) 68; CHECK-NEXT: [[TMP6:%.*]] = load i32, ptr addrspace(3) [[TMP5]], align 4 69; CHECK-NEXT: [[TMP7:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) @vecData.scalarized, i32 3) to ptr addrspace(3) 70; CHECK-NEXT: [[TMP8:%.*]] = load i32, ptr addrspace(3) [[TMP7]], align 4 71; CHECK-NEXT: [[DOTUPTO0:%.*]] = insertelement <4 x i32> poison, i32 [[TMP2]], i64 0 72; CHECK-NEXT: [[DOTUPTO1:%.*]] = insertelement <4 x i32> [[DOTUPTO0]], i32 [[TMP4]], i64 1 73; CHECK-NEXT: [[DOTUPTO2:%.*]] = insertelement <4 x i32> [[DOTUPTO1]], i32 [[TMP6]], i64 2 74; CHECK-NEXT: [[TMP9:%.*]] = insertelement <4 x i32> [[DOTUPTO2]], i32 [[TMP8]], i64 3 75; CHECK-NEXT: ret <4 x i32> [[TMP9]] 76; 77 %1 = load <4 x i32>, <4 x i32> addrspace(3)* @"vecData", align 4 78 ret <4 x i32> %1 79} 80 81define <4 x i32> @load_static_array_of_vec_test(i32 %index) #0 { 82; CHECK-LABEL: define <4 x i32> @load_static_array_of_vec_test( 83; CHECK-SAME: i32 [[INDEX:%.*]]) #[[ATTR0]] { 84; CHECK-NEXT: [[DOTFLAT:%.*]] = getelementptr inbounds [12 x i32], ptr @staticArrayOfVecData.scalarized.1dim, i32 [[INDEX]] 85; CHECK-NEXT: [[TMP1:%.*]] = bitcast ptr [[DOTFLAT]] to ptr 86; CHECK-NEXT: [[TMP2:%.*]] = load i32, ptr [[TMP1]], align 4 87; CHECK-NEXT: [[TMP3:%.*]] = bitcast ptr [[DOTFLAT]] to ptr 88; CHECK-NEXT: [[DOTFLAT_I1:%.*]] = getelementptr i32, ptr [[TMP3]], i32 1 89; CHECK-NEXT: [[DOTI1:%.*]] = load i32, ptr [[DOTFLAT_I1]], align 4 90; CHECK-NEXT: [[TMP4:%.*]] = bitcast ptr [[DOTFLAT]] to ptr 91; CHECK-NEXT: [[DOTFLAT_I2:%.*]] = getelementptr i32, ptr [[TMP4]], i32 2 92; CHECK-NEXT: [[DOTI2:%.*]] = load i32, ptr [[DOTFLAT_I2]], align 4 93; CHECK-NEXT: [[TMP5:%.*]] = bitcast ptr [[DOTFLAT]] to ptr 94; CHECK-NEXT: [[DOTFLAT_I3:%.*]] = getelementptr i32, ptr [[TMP5]], i32 3 95; CHECK-NEXT: [[DOTI3:%.*]] = load i32, ptr [[DOTFLAT_I3]], align 4 96; CHECK-NEXT: [[DOTUPTO0:%.*]] = insertelement <4 x i32> poison, i32 [[TMP2]], i64 0 97; CHECK-NEXT: [[DOTUPTO1:%.*]] = insertelement <4 x i32> [[DOTUPTO0]], i32 [[DOTI1]], i64 1 98; CHECK-NEXT: [[DOTUPTO2:%.*]] = insertelement <4 x i32> [[DOTUPTO1]], i32 [[DOTI2]], i64 2 99; CHECK-NEXT: [[TMP6:%.*]] = insertelement <4 x i32> [[DOTUPTO2]], i32 [[DOTI3]], i64 3 100; CHECK-NEXT: ret <4 x i32> [[TMP6]] 101; 102 %3 = getelementptr inbounds [3 x <4 x i32>], [3 x <4 x i32>]* @staticArrayOfVecData, i32 0, i32 %index 103 %4 = load <4 x i32>, <4 x i32>* %3, align 4 104 ret <4 x i32> %4 105} 106 107define <4 x i32> @multid_load_test() #0 { 108; CHECK-LABEL: define <4 x i32> @multid_load_test( 109; CHECK-SAME: ) #[[ATTR0]] { 110; CHECK-NEXT: [[TMP1:%.*]] = bitcast ptr addrspace(3) @groushared2dArrayofVectors.scalarized.1dim to ptr addrspace(3) 111; CHECK-NEXT: [[TMP2:%.*]] = load i32, ptr addrspace(3) [[TMP1]], align 4 112; CHECK-NEXT: [[TMP3:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) @groushared2dArrayofVectors.scalarized.1dim, i32 1) to ptr addrspace(3) 113; CHECK-NEXT: [[TMP4:%.*]] = load i32, ptr addrspace(3) [[TMP3]], align 4 114; CHECK-NEXT: [[TMP5:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) @groushared2dArrayofVectors.scalarized.1dim, i32 2) to ptr addrspace(3) 115; CHECK-NEXT: [[TMP6:%.*]] = load i32, ptr addrspace(3) [[TMP5]], align 4 116; CHECK-NEXT: [[TMP7:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) @groushared2dArrayofVectors.scalarized.1dim, i32 3) to ptr addrspace(3) 117; CHECK-NEXT: [[TMP8:%.*]] = load i32, ptr addrspace(3) [[TMP7]], align 4 118; CHECK-NEXT: [[TMP11:%.*]] = bitcast ptr addrspace(3) getelementptr inbounds ([36 x i32], ptr addrspace(3) @groushared2dArrayofVectors.scalarized.1dim, i32 1) to ptr addrspace(3) 119; CHECK-NEXT: [[TMP12:%.*]] = load i32, ptr addrspace(3) [[TMP11]], align 4 120; CHECK-NEXT: [[DOTI12:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) getelementptr inbounds ([36 x i32], ptr addrspace(3) @groushared2dArrayofVectors.scalarized.1dim, i32 1), i32 1) to ptr addrspace(3) 121; CHECK-NEXT: [[DOTI13:%.*]] = load i32, ptr addrspace(3) [[DOTI12]], align 4 122; CHECK-NEXT: [[DOTI24:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) getelementptr inbounds ([36 x i32], ptr addrspace(3) @groushared2dArrayofVectors.scalarized.1dim, i32 1), i32 2) to ptr addrspace(3) 123; CHECK-NEXT: [[DOTI25:%.*]] = load i32, ptr addrspace(3) [[DOTI24]], align 4 124; CHECK-NEXT: [[DOTI36:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) getelementptr inbounds ([36 x i32], ptr addrspace(3) @groushared2dArrayofVectors.scalarized.1dim, i32 1), i32 3) to ptr addrspace(3) 125; CHECK-NEXT: [[DOTI37:%.*]] = load i32, ptr addrspace(3) [[DOTI36]], align 4 126; CHECK-NEXT: [[DOTI08:%.*]] = add i32 [[TMP2]], [[TMP12]] 127; CHECK-NEXT: [[DOTI19:%.*]] = add i32 [[TMP4]], [[DOTI13]] 128; CHECK-NEXT: [[DOTI210:%.*]] = add i32 [[TMP6]], [[DOTI25]] 129; CHECK-NEXT: [[DOTI311:%.*]] = add i32 [[TMP8]], [[DOTI37]] 130; CHECK-NEXT: [[DOTUPTO015:%.*]] = insertelement <4 x i32> poison, i32 [[DOTI08]], i64 0 131; CHECK-NEXT: [[DOTUPTO116:%.*]] = insertelement <4 x i32> [[DOTUPTO015]], i32 [[DOTI19]], i64 1 132; CHECK-NEXT: [[DOTUPTO217:%.*]] = insertelement <4 x i32> [[DOTUPTO116]], i32 [[DOTI210]], i64 2 133; CHECK-NEXT: [[TMP16:%.*]] = insertelement <4 x i32> [[DOTUPTO217]], i32 [[DOTI311]], i64 3 134; CHECK-NEXT: ret <4 x i32> [[TMP16]] 135; 136 %1 = load <4 x i32>, <4 x i32> addrspace(3)* getelementptr inbounds ([3 x [3 x <4 x i32>]], [3 x [3 x <4 x i32>]] addrspace(3)* @"groushared2dArrayofVectors", i32 0, i32 0, i32 0), align 4 137 %2 = load <4 x i32>, <4 x i32> addrspace(3)* getelementptr inbounds ([3 x [3 x <4 x i32>]], [3 x [3 x <4 x i32>]] addrspace(3)* @"groushared2dArrayofVectors", i32 0, i32 1, i32 1), align 4 138 %3 = add <4 x i32> %1, %2 139 ret <4 x i32> %3 140} 141 142attributes #0 = { convergent norecurse nounwind "hlsl.export"} 143