Skip to content

Commit ae290e7

Browse files
farzonlfarzonl
committed
[DirectX] Array GEPs need two indices
partially fixes 142836 - Update DXILFlattenArrays.cpp GEPs to use two indicies since they are array GEPs - Update flatten test cases - This change reduces dxv bitcast validation errors by 364 (Total now is 1070x) - This change reduces dxv out of bounds validation errors by 124 (Total is now 24) - We are also able to successfully compile 4 more shaders
1 parent 2158036 commit ae290e7

File tree

5 files changed

+65
-70
lines changed

5 files changed

+65
-70
lines changed

‎llvm/lib/Target/DirectX/DXILFlattenArrays.cpp

Lines changed: 3 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -272,8 +272,9 @@ bool DXILFlattenArraysVisitor::visitGetElementPtrInstInGEPChainBase(
272272

273273
ArrayType *FlattenedArrayType = GEPInfo.ParentArrayType;
274274
Value *FlatGEP =
275-
Builder.CreateGEP(FlattenedArrayType, GEPInfo.ParendOperand, FlatIndex,
276-
GEP.getName() + ".flat", GEP.isInBounds());
275+
Builder.CreateGEP(FlattenedArrayType, GEPInfo.ParendOperand,
276+
{Builder.getInt32(0), FlatIndex},
277+
GEP.getName() + ".flat", GEP.getNoWrapFlags());
277278

278279
GEP.replaceAllUsesWith(FlatGEP);
279280
GEP.eraseFromParent();

‎llvm/test/CodeGen/DirectX/flatten-array.ll

Lines changed: 9 additions & 11 deletions
Original file line numberDiff line numberDiff line change
@@ -31,7 +31,7 @@ define void @alloca_4d_test () {
3131
; CHECK-LABEL: gep_2d_test
3232
define void @gep_2d_test () {
3333
; CHECK: [[a:%.*]] = alloca [9 x i32], align 4
34-
; CHECK-COUNT-9: getelementptr inbounds [9 x i32], ptr [[a]], i32 {{[0-8]}}
34+
; CHECK-COUNT-9: getelementptr inbounds [9 x i32], ptr [[a]], i32 0, i32 {{[0-8]}}
3535
; CHECK-NEXT: ret void
3636
%1 = alloca [3 x [3 x i32]], align 4
3737
%g2d0 = getelementptr inbounds [3 x [3 x i32]], [3 x [3 x i32]]* %1, i32 0, i32 0
@@ -53,7 +53,7 @@ define void @gep_2d_test () {
5353
; CHECK-LABEL: gep_3d_test
5454
define void @gep_3d_test () {
5555
; CHECK: [[a:%.*]] = alloca [8 x i32], align 4
56-
; CHECK-COUNT-8: getelementptr inbounds [8 x i32], ptr [[a]], i32 {{[0-7]}}
56+
; CHECK-COUNT-8: getelementptr inbounds [8 x i32], ptr [[a]], i32 0, i32 {{[0-7]}}
5757
; CHECK-NEXT: ret void
5858
%1 = alloca [2 x[2 x [2 x i32]]], align 4
5959
%g3d0 = getelementptr inbounds [2 x[2 x [2 x i32]]], [2 x[2 x [2 x i32]]]* %1, i32 0, i32 0
@@ -76,7 +76,7 @@ define void @gep_3d_test () {
7676
; CHECK-LABEL: gep_4d_test
7777
define void @gep_4d_test () {
7878
; CHECK: [[a:%.*]] = alloca [16 x i32], align 4
79-
; CHECK-COUNT-16: getelementptr inbounds [16 x i32], ptr [[a]], i32 {{[0-9]|1[0-5]}}
79+
; CHECK-COUNT-16: getelementptr inbounds [16 x i32], ptr [[a]], i32 0, i32 {{[0-9]|1[0-5]}}
8080
; CHECK-NEXT: ret void
8181
%1 = alloca [2x[2 x[2 x [2 x i32]]]], align 4
8282
%g4d0 = getelementptr inbounds [2x[2 x[2 x [2 x i32]]]], [2x[2 x[2 x [2 x i32]]]]* %1, i32 0, i32 0
@@ -123,8 +123,7 @@ define void @gep_4d_test () {
123123
@b = internal global [2 x [3 x [4 x i32]]] zeroinitializer, align 16
124124

125125
define void @global_gep_load() {
126-
; CHECK: [[GEP_PTR:%.*]] = getelementptr inbounds [24 x i32], ptr @a.1dim, i32 6
127-
; CHECK: load i32, ptr [[GEP_PTR]], align 4
126+
; CHECK: load i32, ptr getelementptr inbounds ([24 x i32], ptr @a.1dim, i32 0, i32 6), align 4
128127
; CHECK-NEXT: ret void
129128
%1 = getelementptr inbounds [2 x [3 x [4 x i32]]], [2 x [3 x [4 x i32]]]* @a, i32 0, i32 0
130129
%2 = getelementptr inbounds [3 x [4 x i32]], [3 x [4 x i32]]* %1, i32 0, i32 1
@@ -142,11 +141,11 @@ define void @global_gep_load_index(i32 %row, i32 %col, i32 %timeIndex) {
142141
; CHECK-NEXT: [[TMP4:%.*]] = add i32 [[TMP2]], [[TMP3]]
143142
; CHECK-NEXT: [[TMP5:%.*]] = mul i32 [[ROW]], 12
144143
; CHECK-NEXT: [[TMP6:%.*]] = add i32 [[TMP4]], [[TMP5]]
145-
; CHECK-NEXT: [[DOTFLAT:%.*]] = getelementptr inbounds [24 x i32], ptr @a.1dim, i32 [[TMP6]]
144+
; CHECK-NEXT: %.flat = getelementptr inbounds [24 x i32], ptr @a.1dim, i32 0, i32 %6
146145
; CHECK-NOT: getelementptr inbounds [2 x [3 x [4 x i32]]]{{.*}}
147146
; CHECK-NOT: getelementptr inbounds [3 x [4 x i32]]{{.*}}
148147
; CHECK-NOT: getelementptr inbounds [4 x i32]{{.*}}
149-
; CHECK-NEXT: [[TMP7:%.*]] = load i32, ptr [[DOTFLAT]], align 4
148+
; CHECK-NEXT: %7 = load i32, ptr %.flat, align 4
150149
; CHECK-NEXT: ret void
151150
;
152151
%1 = getelementptr inbounds [2 x [3 x [4 x i32]]], [2 x [3 x [4 x i32]]]* @a, i32 0, i32 %row
@@ -163,11 +162,11 @@ define void @global_incomplete_gep_chain(i32 %row, i32 %col) {
163162
; CHECK-NEXT: [[TMP2:%.*]] = add i32 0, [[TMP1]]
164163
; CHECK-NEXT: [[TMP3:%.*]] = mul i32 [[ROW]], 3
165164
; CHECK-NEXT: [[TMP4:%.*]] = add i32 [[TMP2]], [[TMP3]]
166-
; CHECK-NEXT: [[DOTFLAT:%.*]] = getelementptr inbounds [24 x i32], ptr @a.1dim, i32 [[TMP4]]
165+
; CHECK-NEXT: %.flat = getelementptr inbounds [24 x i32], ptr @a.1dim, i32 0, i32 %4
167166
; CHECK-NOT: getelementptr inbounds [2 x [3 x [4 x i32]]]{{.*}}
168167
; CHECK-NOT: getelementptr inbounds [3 x [4 x i32]]{{.*}}
169168
; CHECK-NOT: getelementptr inbounds [4 x i32]{{.*}}
170-
; CHECK-NEXT: [[TMP5:%.*]] = load i32, ptr [[DOTFLAT]], align 4
169+
; CHECK-NEXT: %5 = load i32, ptr %.flat, align 4
171170
; CHECK-NEXT: ret void
172171
;
173172
%1 = getelementptr inbounds [2 x [3 x [4 x i32]]], [2 x [3 x [4 x i32]]]* @a, i32 0, i32 %row
@@ -177,8 +176,7 @@ define void @global_incomplete_gep_chain(i32 %row, i32 %col) {
177176
}
178177

179178
define void @global_gep_store() {
180-
; CHECK: [[GEP_PTR:%.*]] = getelementptr inbounds [24 x i32], ptr @b.1dim, i32 13
181-
; CHECK: store i32 1, ptr [[GEP_PTR]], align 4
179+
; CHECK: store i32 1, ptr getelementptr inbounds ([24 x i32], ptr @b.1dim, i32 0, i32 13), align 4
182180
; CHECK-NEXT: ret void
183181
%1 = getelementptr inbounds [2 x [3 x [4 x i32]]], [2 x [3 x [4 x i32]]]* @b, i32 0, i32 1
184182
%2 = getelementptr inbounds [3 x [4 x i32]], [3 x [4 x i32]]* %1, i32 0, i32 0

‎llvm/test/CodeGen/DirectX/flatten-bug-117273.ll

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -7,10 +7,10 @@
77

88
define internal void @main() {
99
; CHECK-LABEL: define internal void @main() {
10-
; CHECK-NEXT: [[ENTRY:.*:]]
11-
; CHECK-NEXT: [[TMP0:%.*]] = getelementptr [24 x float], ptr @ZerroInitArr.1dim, i32 1
10+
; CHECK-NEXT: entry:
11+
; CHECK-NEXT: [[TMP0:%.*]] = getelementptr [24 x float], ptr @ZerroInitArr.1dim, i32 0, i32 1
1212
; CHECK-NEXT: [[DOTI0:%.*]] = load float, ptr [[TMP0]], align 16
13-
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr [24 x float], ptr @ZerroInitArr.1dim, i32 2
13+
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr [24 x float], ptr @ZerroInitArr.1dim, i32 0, i32 2
1414
; CHECK-NEXT: [[DOTI03:%.*]] = load float, ptr [[TMP1]], align 16
1515
; CHECK-NEXT: ret void
1616
;

‎llvm/test/CodeGen/DirectX/llc-vector-load-scalarize.ll

Lines changed: 47 additions & 51 deletions
Original file line numberDiff line numberDiff line change
@@ -32,23 +32,23 @@ define <4 x i32> @load_array_vec_test() #0 {
3232
; CHECK-NEXT: [[TMP6:%.*]] = load i32, ptr addrspace(3) [[TMP5]], align 4
3333
; CHECK-NEXT: [[TMP7:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) @arrayofVecData.scalarized.1dim, i32 3) to ptr addrspace(3)
3434
; 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]], i32 0
48-
; CHECK-NEXT: [[DOTUPTO116:%.*]] = insertelement <4 x i32> [[DOTUPTO015]], i32 [[DOTI19]], i32 1
49-
; CHECK-NEXT: [[DOTUPTO217:%.*]] = insertelement <4 x i32> [[DOTUPTO116]], i32 [[DOTI210]], i32 2
50-
; CHECK-NEXT: [[TMP16:%.*]] = insertelement <4 x i32> [[DOTUPTO217]], i32 [[DOTI311]], i32 3
51-
; CHECK-NEXT: ret <4 x i32> [[TMP16]]
35+
; CHECK-NEXT: [[TMP9:%.*]] = bitcast ptr addrspace(3) getelementptr inbounds ([6 x float], ptr addrspace(3) @arrayofVecData.scalarized.1dim, i32 0, i32 1) to ptr addrspace(3)
36+
; CHECK-NEXT: [[TMP10:%.*]] = load i32, ptr addrspace(3) [[TMP9]], align 4
37+
; CHECK-NEXT: [[TMP11:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) getelementptr inbounds ([6 x float], ptr addrspace(3) @arrayofVecData.scalarized.1dim, i32 0, i32 1), i32 1) to ptr addrspace(3)
38+
; CHECK-NEXT: [[TMP12:%.*]] = load i32, ptr addrspace(3) [[TMP11]], align 4
39+
; CHECK-NEXT: [[TMP13:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) getelementptr inbounds ([6 x float], ptr addrspace(3) @arrayofVecData.scalarized.1dim, i32 0, i32 1), i32 2) to ptr addrspace(3)
40+
; CHECK-NEXT: [[TMP14:%.*]] = load i32, ptr addrspace(3) [[TMP13]], align 4
41+
; CHECK-NEXT: [[TMP15:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) getelementptr inbounds ([6 x float], ptr addrspace(3) @arrayofVecData.scalarized.1dim, i32 0, i32 1), i32 3) to ptr addrspace(3)
42+
; CHECK-NEXT: [[TMP16:%.*]] = load i32, ptr addrspace(3) [[TMP15]], align 4
43+
; CHECK-NEXT: [[DOTI05:%.*]] = add i32 [[TMP2]], [[TMP10]]
44+
; CHECK-NEXT: [[DOTI16:%.*]] = add i32 [[TMP4]], [[TMP12]]
45+
; CHECK-NEXT: [[DOTI27:%.*]] = add i32 [[TMP6]], [[TMP14]]
46+
; CHECK-NEXT: [[DOTI38:%.*]] = add i32 [[TMP8]], [[TMP16]]
47+
; CHECK-NEXT: [[DOTUPTO01215:%.*]] = insertelement <4 x i32> poison, i32 [[DOTI05]], i32 0
48+
; CHECK-NEXT: [[DOTUPTO11316:%.*]] = insertelement <4 x i32> [[DOTUPTO01215]], i32 [[DOTI16]], i32 1
49+
; CHECK-NEXT: [[DOTUPTO21417:%.*]] = insertelement <4 x i32> [[DOTUPTO11316]], i32 [[DOTI27]], i32 2
50+
; CHECK-NEXT: [[TMP17:%.*]] = insertelement <4 x i32> [[DOTUPTO21417]], i32 [[DOTI38]], i32 3
51+
; CHECK-NEXT: ret <4 x i32> [[TMP17]]
5252
;
5353
%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
5454
%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
@@ -81,23 +81,19 @@ define <4 x i32> @load_vec_test() #0 {
8181
define <4 x i32> @load_static_array_of_vec_test(i32 %index) #0 {
8282
; CHECK-LABEL: define <4 x i32> @load_static_array_of_vec_test(
8383
; 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]], i32 0
97-
; CHECK-NEXT: [[DOTUPTO1:%.*]] = insertelement <4 x i32> [[DOTUPTO0]], i32 [[DOTI1]], i32 1
98-
; CHECK-NEXT: [[DOTUPTO2:%.*]] = insertelement <4 x i32> [[DOTUPTO1]], i32 [[DOTI2]], i32 2
99-
; CHECK-NEXT: [[TMP6:%.*]] = insertelement <4 x i32> [[DOTUPTO2]], i32 [[DOTI3]], i32 3
100-
; CHECK-NEXT: ret <4 x i32> [[TMP6]]
84+
; CHECK-NEXT: [[DOTFLAT:%.*]] = getelementptr inbounds [12 x i32], ptr @staticArrayOfVecData.scalarized.1dim, i32 0, i32 [[INDEX]]
85+
; CHECK-NEXT: [[DOTI0:%.*]] = load i32, ptr [[DOTFLAT]], align 4
86+
; CHECK-NEXT: [[DOTFLAT_I1:%.*]] = getelementptr i32, ptr [[DOTFLAT]], i32 1
87+
; CHECK-NEXT: [[DOTI1:%.*]] = load i32, ptr [[DOTFLAT_I1]], align 4
88+
; CHECK-NEXT: [[DOTFLAT_I2:%.*]] = getelementptr i32, ptr [[DOTFLAT]], i32 2
89+
; CHECK-NEXT: [[DOTI2:%.*]] = load i32, ptr [[DOTFLAT_I2]], align 4
90+
; CHECK-NEXT: [[DOTFLAT_I3:%.*]] = getelementptr i32, ptr [[DOTFLAT]], i32 3
91+
; CHECK-NEXT: [[DOTI3:%.*]] = load i32, ptr [[DOTFLAT_I3]], align 4
92+
; CHECK-NEXT: [[DOTUPTO01:%.*]] = insertelement <4 x i32> poison, i32 [[DOTI0]], i32 0
93+
; CHECK-NEXT: [[DOTUPTO12:%.*]] = insertelement <4 x i32> [[DOTUPTO01]], i32 [[DOTI1]], i32 1
94+
; CHECK-NEXT: [[DOTUPTO23:%.*]] = insertelement <4 x i32> [[DOTUPTO12]], i32 [[DOTI2]], i32 2
95+
; CHECK-NEXT: [[TMP1:%.*]] = insertelement <4 x i32> [[DOTUPTO23]], i32 [[DOTI3]], i32 3
96+
; CHECK-NEXT: ret <4 x i32> [[TMP1]]
10197
;
10298
%3 = getelementptr inbounds [3 x <4 x i32>], [3 x <4 x i32>]* @staticArrayOfVecData, i32 0, i32 %index
10399
%4 = load <4 x i32>, <4 x i32>* %3, align 4
@@ -115,23 +111,23 @@ define <4 x i32> @multid_load_test() #0 {
115111
; CHECK-NEXT: [[TMP6:%.*]] = load i32, ptr addrspace(3) [[TMP5]], align 4
116112
; CHECK-NEXT: [[TMP7:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) @groushared2dArrayofVectors.scalarized.1dim, i32 3) to ptr addrspace(3)
117113
; 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]], i32 0
131-
; CHECK-NEXT: [[DOTUPTO116:%.*]] = insertelement <4 x i32> [[DOTUPTO015]], i32 [[DOTI19]], i32 1
132-
; CHECK-NEXT: [[DOTUPTO217:%.*]] = insertelement <4 x i32> [[DOTUPTO116]], i32 [[DOTI210]], i32 2
133-
; CHECK-NEXT: [[TMP16:%.*]] = insertelement <4 x i32> [[DOTUPTO217]], i32 [[DOTI311]], i32 3
134-
; CHECK-NEXT: ret <4 x i32> [[TMP16]]
114+
; CHECK-NEXT: [[TMP9:%.*]] = bitcast ptr addrspace(3) getelementptr inbounds ([36 x i32], ptr addrspace(3) @groushared2dArrayofVectors.scalarized.1dim, i32 0, i32 1) to ptr addrspace(3)
115+
; CHECK-NEXT: [[TMP10:%.*]] = load i32, ptr addrspace(3) [[TMP9]], align 4
116+
; CHECK-NEXT: [[TMP11:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) getelementptr inbounds ([36 x i32], ptr addrspace(3) @groushared2dArrayofVectors.scalarized.1dim, i32 0, i32 1), i32 1) to ptr addrspace(3)
117+
; CHECK-NEXT: [[TMP12:%.*]] = load i32, ptr addrspace(3) [[TMP11]], align 4
118+
; CHECK-NEXT: [[TMP13:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) getelementptr inbounds ([36 x i32], ptr addrspace(3) @groushared2dArrayofVectors.scalarized.1dim, i32 0, i32 1), i32 2) to ptr addrspace(3)
119+
; CHECK-NEXT: [[TMP14:%.*]] = load i32, ptr addrspace(3) [[TMP13]], align 4
120+
; CHECK-NEXT: [[TMP15:%.*]] = bitcast ptr addrspace(3) getelementptr (i32, ptr addrspace(3) getelementptr inbounds ([36 x i32], ptr addrspace(3) @groushared2dArrayofVectors.scalarized.1dim, i32 0, i32 1), i32 3) to ptr addrspace(3)
121+
; CHECK-NEXT: [[TMP16:%.*]] = load i32, ptr addrspace(3) [[TMP15]], align 4
122+
; CHECK-NEXT: [[DOTI05:%.*]] = add i32 [[TMP2]], [[TMP10]]
123+
; CHECK-NEXT: [[DOTI16:%.*]] = add i32 [[TMP4]], [[TMP12]]
124+
; CHECK-NEXT: [[DOTI27:%.*]] = add i32 [[TMP6]], [[TMP14]]
125+
; CHECK-NEXT: [[DOTI38:%.*]] = add i32 [[TMP8]], [[TMP16]]
126+
; CHECK-NEXT: [[DOTUPTO01215:%.*]] = insertelement <4 x i32> poison, i32 [[DOTI05]], i32 0
127+
; CHECK-NEXT: [[DOTUPTO11316:%.*]] = insertelement <4 x i32> [[DOTUPTO01215]], i32 [[DOTI16]], i32 1
128+
; CHECK-NEXT: [[DOTUPTO21417:%.*]] = insertelement <4 x i32> [[DOTUPTO11316]], i32 [[DOTI27]], i32 2
129+
; CHECK-NEXT: [[TMP17:%.*]] = insertelement <4 x i32> [[DOTUPTO21417]], i32 [[DOTI38]], i32 3
130+
; CHECK-NEXT: ret <4 x i32> [[TMP17]]
135131
;
136132
%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
137133
%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

‎llvm/test/CodeGen/DirectX/scalar-bug-117273.ll

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -7,14 +7,14 @@
77
; Function Attrs: alwaysinline convergent mustprogress norecurse nounwind
88
define internal void @main() #1 {
99
; CHECK-LABEL: define internal void @main() {
10-
; CHECK-NEXT: [[ENTRY:.*:]]
11-
; CHECK-NEXT: [[TMP0:%.*]] = getelementptr inbounds [24 x float], ptr @StaticArr.scalarized.1dim, i32 1
10+
; CHECK-NEXT: entry:
11+
; CHECK-NEXT: [[TMP0:%.*]] = getelementptr inbounds [24 x float], ptr @StaticArr.scalarized.1dim, i32 0, i32 1
1212
; CHECK-NEXT: [[DOTI0:%.*]] = load float, ptr [[TMP0]], align 16
1313
; CHECK-NEXT: [[DOTI1:%.*]] = getelementptr float, ptr [[TMP0]], i32 1
1414
; CHECK-NEXT: [[DOTI11:%.*]] = load float, ptr [[DOTI1]], align 4
1515
; CHECK-NEXT: [[DOTI2:%.*]] = getelementptr float, ptr [[TMP0]], i32 2
1616
; CHECK-NEXT: [[DOTI22:%.*]] = load float, ptr [[DOTI2]], align 8
17-
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds [24 x float], ptr @StaticArr.scalarized.1dim, i32 2
17+
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds [24 x float], ptr @StaticArr.scalarized.1dim, i32 0, i32 2
1818
; CHECK-NEXT: [[DOTI03:%.*]] = load float, ptr [[TMP1]], align 16
1919
; CHECK-NEXT: [[DOTI14:%.*]] = getelementptr float, ptr [[TMP1]], i32 1
2020
; CHECK-NEXT: [[DOTI15:%.*]] = load float, ptr [[DOTI14]], align 4

0 commit comments

Comments
 (0)