Fix loading Ds tensors.

include/ck/tensor_operation/gpu/grid/gridwise_gemm_multiple_d_xdl_splitk_cshuffle_v2.hpp

@@ -1303,14 +1303,14 @@ class GridwiseGemmMultipleD_xdl_splitk_cshuffle_v2
         });
 
         // TODO: on MI300 we could use NonTemporal load, MI200 streaming mode?
-        auto ds_grid_buf = generate_tuple(
+        const auto ds_grid_buf = generate_tuple(
             [&](auto i) {
                 return make_dynamic_buffer<AddressSpaceEnum::Global>(
                     p_ds_grid[i], ds_grid_desc_m0m1_n0n1n2[i].GetElementSpaceSize());
             },
             Number<NumDTensor>{});
 
-        constexpr auto ds_thread_buf = generate_tuple(
+        auto ds_thread_buf = generate_tuple(
             [&](auto i) {
                 using DDataType = remove_cvref_t<tuple_element_t<i.value, DsDataType>>;
                 return StaticBuffer<AddressSpaceEnum::Vgpr, DDataType, ScalarPerVector, true>{};
@@ -1353,7 +1353,7 @@ class GridwiseGemmMultipleD_xdl_splitk_cshuffle_v2
                 using SliceLengths = Sequence<I1, I1, I1, I1, ScalarPerVector>;
                 return ThreadwiseTensorSliceTransfer_v2<DDataType,
                                                         DDataType,
-                                                        decltype(ds_grid_desc_m0m1_n0n1n2(i)),
+                                                        decltype(ds_grid_desc_m0m1_n0n1n2[i]),
                                                         decltype(d_vgpr_buf_desc),
                                                         SliceLengths,
                                                         Sequence<0, 1, 2, 3, 4>,
@@ -1361,7 +1361,7 @@ class GridwiseGemmMultipleD_xdl_splitk_cshuffle_v2
                                                         ScalarPerVector,
                                                         1,
                                                         false>{
-                    ds_grid_desc_m_n(i),
+                    ds_grid_desc_m0m1_n0n1n2[i],
                     make_multi_index(
                         block_work_idx[I0],
                         thread_m_cluster_id * workspace_thread_desc_m0m1_n0n1n2.GetLength(I1),
@@ -1410,8 +1410,8 @@ class GridwiseGemmMultipleD_xdl_splitk_cshuffle_v2
             static_for<0, NIter, 1>{}([&](auto n_idx) {
                 // load multiple Ds:
                 static_for<0, NumDTensor, 1>{}([&](auto d_idx) {
-                    ds_grid_load(d_idx).Run(ds_grid_desc_m0m1_n0n1n2(d_idx),
-                                            ds_grid_buf(d_idx),
+                    ds_grid_load(d_idx).Run(ds_grid_desc_m0m1_n0n1n2[d_idx],
+                                            ds_grid_buf[d_idx],
                                             d_vgpr_buf_desc,
                                             make_tuple(I0, I0, I0, I0, I0),
                                             ds_thread_buf(d_idx));