Fix handling of matrix in StoreVertexOutput (#3441)

lib/HLSL/HLSignatureLower.cpp

@@ -676,8 +676,11 @@ void replaceMatStWithStOutputs(CallInst *CI, HLMatLoadStoreOpcode matOp,
       for (unsigned r = 0; r < MatTy.getNumRows(); r++) {
         unsigned matIdx = MatTy.getColumnMajorIndex(r, c);
         Value *Elt = LocalBuilder.CreateExtractElement(Val, matIdx);
-        LocalBuilder.CreateCall(ldStFunc,
-                                { OpArg, ID, colIdx, columnConsts[r], Elt });
+
+        SmallVector<Value*, 6> argList = {OpArg, ID, colIdx, columnConsts[r], Elt};
+        if (vertexOrPrimID)
+          argList.emplace_back(vertexOrPrimID);
+        LocalBuilder.CreateCall(ldStFunc, argList);
       }
     }
   } else {
@@ -687,8 +690,11 @@ void replaceMatStWithStOutputs(CallInst *CI, HLMatLoadStoreOpcode matOp,
       for (unsigned c = 0; c < MatTy.getNumColumns(); c++) {
         unsigned matIdx = MatTy.getRowMajorIndex(r, c);
         Value *Elt = LocalBuilder.CreateExtractElement(Val, matIdx);
-        LocalBuilder.CreateCall(ldStFunc,
-                                { OpArg, ID, rowIdx, columnConsts[c], Elt });
+
+        SmallVector<Value*, 6> argList = {OpArg, ID, rowIdx, columnConsts[c], Elt};
+        if (vertexOrPrimID)
+          argList.emplace_back(vertexOrPrimID);
+        LocalBuilder.CreateCall(ldStFunc, argList);
       }
     }
   }

tools/clang/test/HLSLFileCheck/shader_targets/mesh/matVertexStOutput_colmajor.hlsl

@@ -0,0 +1,189 @@
+// RUN: %dxc -E main -DMAT1x1 -Zpc -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT1x1
+// RUN: %dxc -E main -DMAT1x2 -Zpc -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT1x2
+// RUN: %dxc -E main -DMAT2x1 -Zpc -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT2x1
+// RUN: %dxc -E main -DMAT2x2 -Zpc -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT2x2
+// RUN: %dxc -E main -DMAT2x3 -Zpc -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT2x3
+// RUN: %dxc -E main -DMAT3x2 -Zpc -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT3x2
+// RUN: %dxc -E main -DMAT3x3 -Zpc -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT3x3
+// RUN: %dxc -E main -DMAT3x4 -Zpc -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT3x4
+// RUN: %dxc -E main -DMAT4x3 -Zpc -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT4x3
+// RUN: %dxc -E main -DMAT4x4 -Zpc -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT4x4
+
+// Regression test to check that store vertex output for matrix works fine
+
+#ifdef MAT1x1
+#define TY float1x1
+#endif
+
+#ifdef MAT1x2
+#define TY float1x2
+#endif
+
+#ifdef MAT2x1
+#define TY float2x1
+#endif
+
+#ifdef MAT2x2
+#define TY float2x2
+#endif
+
+#ifdef MAT2x3
+#define TY float2x3
+#endif
+
+#ifdef MAT3x2
+#define TY float3x2
+#endif
+
+#ifdef MAT3x3
+#define TY float3x3
+#endif
+
+#ifdef MAT3x4
+#define TY float3x4
+#endif
+
+#ifdef MAT4x3
+#define TY float4x3
+#endif
+
+#ifdef MAT4x4
+#define TY float4x4
+#endif
+
+struct VertexOutput
+{
+		TY test : TEXCOORD0;
+};
+
+[NumThreads(64, 1, 1)]
+[OutputTopology("triangle")]
+void main(out vertices VertexOutput verts[1])
+{
+ 
+ // CHK_MAT1x1: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+#ifdef MAT1x1  
+	verts[0].test = TY(1);
+#endif  
+
+  // CHK_MAT1x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT1x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 2.000000e+00, i32 0)
+#ifdef MAT1x2
+	verts[0].test = TY(float2(1, 2));
+#endif
+
+  // CHK_MAT2x1: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT2x1: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 2.000000e+00, i32 0)
+#ifdef MAT2x1
+	verts[0].test = TY(float2(1, 2)); 
+#endif
+
+
+  // CHK_MAT2x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT2x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 3.000000e+00, i32 0)
+  // CHK_MAT2x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 2.000000e+00, i32 0)
+  // CHK_MAT2x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 4.000000e+00, i32 0)  
+#ifdef MAT2x2
+	verts[0].test = TY(float2(1, 2), 
+                     float2(3, 4));
+#endif
+
+  // CHK_MAT2x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT2x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 4.000000e+00, i32 0)
+  // CHK_MAT2x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 2.000000e+00, i32 0)
+  // CHK_MAT2x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 5.000000e+00, i32 0)
+  // CHK_MAT2x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 0, float 3.000000e+00, i32 0) 
+  // CHK_MAT2x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 1, float 6.000000e+00, i32 0)                     
+#ifdef MAT2x3
+	verts[0].test = TY(float3(1, 2, 3),
+                     float3(4, 5, 6));
+#endif
+
+  // CHK_MAT3x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT3x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 3.000000e+00, i32 0)
+  // CHK_MAT3x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 2, float 5.000000e+00, i32 0)
+  // CHK_MAT3x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 2.000000e+00, i32 0)  
+  // CHK_MAT3x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 4.000000e+00, i32 0)  
+  // CHK_MAT3x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 2, float 6.000000e+00, i32 0)                     
+#ifdef MAT3x2
+	verts[0].test = TY(float2(1, 2),
+                     float2(3, 4),
+                     float2(5, 6));
+#endif
+
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 4.000000e+00, i32 0)
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 2, float 7.000000e+00, i32 0)  
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 2.000000e+00, i32 0)
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 5.000000e+00, i32 0)
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 2, float 8.000000e+00, i32 0)
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 0, float 3.000000e+00, i32 0)  
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 1, float 6.000000e+00, i32 0) 
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 2, float 9.000000e+00, i32 0)                     
+#ifdef MAT3x3
+	verts[0].test = TY(float3(1, 2, 3),
+                     float3(4, 5, 6),
+                     float3(7, 8, 9));
+#endif
+
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 5.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 2, float 9.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 2.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 6.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 2, float 1.000000e+01, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 0, float 3.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 1, float 7.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 2, float 1.100000e+01, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 0, float 4.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 1, float 8.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 2, float 1.200000e+01, i32 0)                     
+#ifdef MAT3x4
+	verts[0].test = TY(float4(1, 2,  3,  4),
+                     float4(5, 6,  7,  8),
+                     float4(9, 10, 11, 12));
+#endif
+
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 4.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 2, float 7.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 3, float 1.000000e+01, i32 0)  
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 2.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 5.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 2, float 8.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 3, float 1.100000e+01, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 0, float 3.000000e+00, i32 0)  
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 1, float 6.000000e+00, i32 0)  
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 2, float 9.000000e+00, i32 0)  
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 3, float 1.200000e+01, i32 0)                     
+#ifdef MAT4x3
+	verts[0].test = TY(float3(1,  2,  3),
+                     float3(4,  5,  6),
+                     float3(7,  8,  9),
+                     float3(10, 11, 12));
+#endif
+ 
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 5.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 2, float 9.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 3, float 1.300000e+01, i32 0)  
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 2.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 6.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 2, float 1.000000e+01, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 3, float 1.400000e+01, i32 0)  
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 0, float 3.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 1, float 7.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 2, float 1.100000e+01, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 3, float 1.500000e+01, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 0, float 4.000000e+00, i32 0)  
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 1, float 8.000000e+00, i32 0) 
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 2, float 1.200000e+01, i32 0) 
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 3, float 1.600000e+01, i32 0) 
+#ifdef MAT4x4
+ 	verts[0].test = TY(float4(1,  2,  3,  4), 
+                     float4(5,  6,  7,  8), 
+                     float4(9,  10, 11, 12),
+                     float4(13, 14, 15, 16));
+#endif
+  SetMeshOutputCounts(0, 0);
+}

tools/clang/test/HLSLFileCheck/shader_targets/mesh/matVertexStOutput_rowmajor.hlsl

@@ -0,0 +1,188 @@
+// RUN: %dxc -E main -DMAT1x1 -Zpr -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT1x1
+// RUN: %dxc -E main -DMAT1x2 -Zpr -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT1x2
+// RUN: %dxc -E main -DMAT2x1 -Zpr -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT2x1
+// RUN: %dxc -E main -DMAT2x2 -Zpr -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT2x2
+// RUN: %dxc -E main -DMAT2x3 -Zpr -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT2x3
+// RUN: %dxc -E main -DMAT3x2 -Zpr -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT3x2
+// RUN: %dxc -E main -DMAT3x3 -Zpr -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT3x3
+// RUN: %dxc -E main -DMAT3x4 -Zpr -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT3x4
+// RUN: %dxc -E main -DMAT4x3 -Zpr -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT4x3
+// RUN: %dxc -E main -DMAT4x4 -Zpr -T ms_6_5 %s | FileCheck %s -check-prefix=CHK_MAT4x4
+
+// Regression test to check that store vertex output for matrix works fine
+
+#ifdef MAT1x1
+#define TY float1x1
+#endif
+
+#ifdef MAT1x2
+#define TY float1x2
+#endif
+
+#ifdef MAT2x1
+#define TY float2x1
+#endif
+
+#ifdef MAT2x2
+#define TY float2x2
+#endif
+
+#ifdef MAT2x3
+#define TY float2x3
+#endif
+
+#ifdef MAT3x2
+#define TY float3x2
+#endif
+
+#ifdef MAT3x3
+#define TY float3x3
+#endif
+
+#ifdef MAT3x4
+#define TY float3x4
+#endif
+
+#ifdef MAT4x3
+#define TY float4x3
+#endif
+
+#ifdef MAT4x4
+#define TY float4x4
+#endif
+
+struct VertexOutput
+{
+		TY test : TEXCOORD0;
+};
+
+[NumThreads(64, 1, 1)]
+[OutputTopology("triangle")]
+void main(out vertices VertexOutput verts[1])
+{
+ 
+ // CHK_MAT1x1: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+#ifdef MAT1x1  
+	verts[0].test = TY(1);
+#endif  
+
+  // CHK_MAT1x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT1x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 2.000000e+00, i32 0)
+#ifdef MAT1x2
+	verts[0].test = TY(float2(1, 2));
+#endif
+
+  // CHK_MAT2x1: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT2x1: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 2.000000e+00, i32 0)
+#ifdef MAT2x1
+	verts[0].test = TY(float2(1, 2)); 
+#endif
+
+  // CHK_MAT2x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT2x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 2.000000e+00, i32 0)
+  // CHK_MAT2x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 3.000000e+00, i32 0)
+  // CHK_MAT2x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 4.000000e+00, i32 0)  
+#ifdef MAT2x2
+	verts[0].test = TY(float2(1, 2), 
+                     float2(3, 4));
+#endif
+
+  // CHK_MAT2x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT2x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 2.000000e+00, i32 0)
+  // CHK_MAT2x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 2, float 3.000000e+00, i32 0)
+  // CHK_MAT2x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 4.000000e+00, i32 0)
+  // CHK_MAT2x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 5.000000e+00, i32 0)
+  // CHK_MAT2x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 2, float 6.000000e+00, i32 0)                     
+#ifdef MAT2x3
+	verts[0].test = TY(float3(1, 2, 3),
+                     float3(4, 5, 6));
+#endif
+
+  // CHK_MAT3x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT3x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 2.000000e+00, i32 0)
+  // CHK_MAT3x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 3.000000e+00, i32 0)
+  // CHK_MAT3x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 4.000000e+00, i32 0)
+  // CHK_MAT3x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 0, float 5.000000e+00, i32 0)
+  // CHK_MAT3x2: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 1, float 6.000000e+00, i32 0)                     
+#ifdef MAT3x2
+	verts[0].test = TY(float2(1, 2),
+                     float2(3, 4),
+                     float2(5, 6));
+#endif
+
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 2.000000e+00, i32 0)
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 2, float 3.000000e+00, i32 0)
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 4.000000e+00, i32 0)
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 5.000000e+00, i32 0)
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 2, float 6.000000e+00, i32 0)
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 0, float 7.000000e+00, i32 0)
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 1, float 8.000000e+00, i32 0)
+  // CHK_MAT3x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 2, float 9.000000e+00, i32 0)                     
+#ifdef MAT3x3
+	verts[0].test = TY(float3(1, 2, 3),
+                     float3(4, 5, 6),
+                     float3(7, 8, 9));
+#endif
+
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 2.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 2, float 3.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 3, float 4.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 5.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 6.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 2, float 7.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 3, float 8.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 0, float 9.000000e+00, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 1, float 1.000000e+01, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 2, float 1.100000e+01, i32 0)
+  // CHK_MAT3x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 3, float 1.200000e+01, i32 0)                     
+#ifdef MAT3x4
+	verts[0].test = TY(float4(1, 2,  3,  4),
+                     float4(5, 6,  7,  8),
+                     float4(9, 10, 11, 12));
+#endif
+
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 2.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 2, float 3.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 4.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 5.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 2, float 6.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 0, float 7.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 1, float 8.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 2, float 9.000000e+00, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 0, float 1.000000e+01, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 1, float 1.100000e+01, i32 0)
+  // CHK_MAT4x3: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 2, float 1.200000e+01, i32 0)                     
+#ifdef MAT4x3
+	verts[0].test = TY(float3(1,  2,  3),
+                     float3(4,  5,  6),
+                     float3(7,  8,  9),
+                     float3(10, 11, 12));
+#endif
+ 
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 0, float 1.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 1, float 2.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 2, float 3.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 0, i8 3, float 4.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 0, float 5.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 1, float 6.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 2, float 7.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 1, i8 3, float 8.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 0, float 9.000000e+00, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 1, float 1.000000e+01, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 2, float 1.100000e+01, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 2, i8 3, float 1.200000e+01, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 0, float 1.300000e+01, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 1, float 1.400000e+01, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 2, float 1.500000e+01, i32 0)
+  // CHK_MAT4x4: call void @dx.op.storeVertexOutput.f32(i32 171, i32 0, i32 3, i8 3, float 1.600000e+01, i32 0) 
+#ifdef MAT4x4
+ 	verts[0].test = TY(float4(1,  2,  3,  4), 
+                     float4(5,  6,  7,  8), 
+                     float4(9,  10, 11, 12),
+                     float4(13, 14, 15, 16));
+#endif
+  SetMeshOutputCounts(0, 0);
+}