CUTLASS 2.6.1 - functional and performance enhancements to strided DGRAD, fixes, and tuning

* cutlass 2.6 update

* remove debug prints

* cutlass 2.6.1 (minor update)

* Updated CHANGELOG.

* Minor edit to readme to indicate patch version.

* Minor edit to readme.

Co-authored-by:  Haicheng Wu <haichengw@nvidia.com>, Andrew Kerr <akerr@nvidia.com>

include/cutlass/gemm/device/gemm_array.h

@@ -445,14 +445,6 @@ public:
       if (result != cudaSuccess) {
         return Status::kErrorInternal;
       }
-
-      result = cudaFuncSetAttribute(
-          Kernel<GemmKernel>,
-          cudaFuncAttributePreferredSharedMemoryCarveout, 100);
-
-      if (result != cudaSuccess) {
-        return Status::kErrorInternal;
-      }
     }
 
     cutlass::Kernel<GemmKernel><<<grid, block, smem_size, stream>>>(params_);