change default block to ctile map parameter

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

@@ -118,7 +118,7 @@ struct BlockToCTileMap_M00_N0_M01Adapt
     __host__ __device__ BlockToCTileMap_M00_N0_M01Adapt() = default;
 
     __host__ __device__ BlockToCTileMap_M00_N0_M01Adapt(const CGridDesc_M_N& c_grid_desc_m_n,
-                                                        index_t M01 = 8)
+                                                        index_t M01 = 1)
         : M01_(M01), c_grid_desc_m_n_(c_grid_desc_m_n)
     {
     }

include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r4.hpp

@@ -268,7 +268,7 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4
         const CMNGridDesc& c_m_n_grid_desc, index_t /* M01 */, index_t /* N01 */, index_t KBatch)
     {
         return BlockToCTileMap_KSplit_M00_N0_M01Adapt<MPerBlock, NPerBlock, CMNGridDesc>(
-            c_m_n_grid_desc, 8, KBatch);
+            c_m_n_grid_desc, 1, KBatch);
     }
 
     using CM0N0M1N1M2M3M4N2GridDesc = decltype(MakeCM0N0M1N1M2M3M4N2GridDescriptor(CMNGridDesc{}));

include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r4r2.hpp

@@ -242,7 +242,7 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
         const CMNGridDesc& c_m_n_grid_desc, index_t /* M01 */, index_t /* N01 */, index_t KBatch)
     {
         return BlockToCTileMap_KSplit_M00_N0_M01Adapt<MPerBlock, NPerBlock, CMNGridDesc>(
-            c_m_n_grid_desc, 8, KBatch);
+            c_m_n_grid_desc, 1, KBatch);
     }
 
     __host__ __device__ static constexpr auto