[0035] LinAlg: Remove Stride and Layout params from Thread scoped InterlockedAccumulate (#849)

The thread-scope overload wasn't supposed to have Stride and Layout,
since the Layout must be OuterProductOptimal for thread scope. This
change removes these two parameters from the version of the function in
Appendix 1: HLSL Header.

Fixes #850

Author: tex3d

proposals/0035-linalg-matrix.md

@@ -199,8 +199,7 @@ class Matrix<ComponentTy, M, N, Use, MatrixScope::Thread> {
   template <MatrixUseEnum UseLocal = Use>
   typename hlsl::enable_if<Use == MatrixUse::Accumulator && UseLocal == Use,
                            void>::type
-  InterlockedAccumulate(RWByteAddressBuffer Res, uint StartOffset, uint Stride,
-                        MatrixLayoutEnum Layout,
+  InterlockedAccumulate(RWByteAddressBuffer Res, uint StartOffset,
                         uint Align = sizeof(ElementType));
 };
 
@@ -2069,8 +2068,7 @@ class Matrix<ComponentTy, M, N, Use, MatrixScope::Thread> {
   template <MatrixUseEnum UseLocal = Use>
   typename hlsl::enable_if<Use == MatrixUse::Accumulator && UseLocal == Use,
                            void>::type
-  InterlockedAccumulate(RWByteAddressBuffer Res, uint StartOffset, uint Stride,
-                        MatrixLayoutEnum Layout,
+  InterlockedAccumulate(RWByteAddressBuffer Res, uint StartOffset,
                         uint Align = sizeof(ElementType));
 };
 
@@ -2234,6 +2232,6 @@ void OuterProdAccum() {
   vector<float16_t, 8> VecB = (vector<float16_t, 8>)0;
   MatrixAccumTy MatAcc = OuterProduct<ComponentType::F16>(VecA, VecB);
 
-  MatAcc.InterlockedAccumulate(Buf, 0, 0, MatrixLayout::OuterProductOptimal);
+  MatAcc.InterlockedAccumulate(Buf, 0);
 }
 ```