Bwd Qloop_PT1&Split K in LDS/V in Vgpr

include/ck/tensor_operation/gpu/device/impl/device_batched_mha_bwd_xdl_cshuffle_qloop_v1_protro.hpp

@@ -15,7 +15,7 @@
 #include "ck/tensor_operation/gpu/device/masking_specialization.hpp"
 #include "ck/tensor_operation/gpu/device/matrix_padder.hpp"
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
-#include "ck/tensor_operation/gpu/grid/gridwise_batched_mha_bwd_xdl_cshuffle_qloop_b2t_v1_protro2.hpp"
+#include "ck/tensor_operation/gpu/grid/gridwise_batched_mha_bwd_xdl_cshuffle_qloop_b2t_v1_protro3.hpp"
 #include "ck/tensor_operation/operator_transform/transform_contraction_to_gemm.hpp"
 #include "ck/host_utility/device_prop.hpp"
 #include "ck/host_utility/kernel_launch.hpp"
@@ -255,6 +255,7 @@ template <index_t NumDimG,
           index_t KPerBlock, // Gemm0KPerBlock
           index_t Gemm1NPerBlock,
           index_t Gemm1KPerBlock,
+          index_t Gemm2KPerBlock,
           index_t AK1,
           index_t BK1,
           index_t B1K1,
@@ -665,6 +666,7 @@ struct DeviceBatchedMultiheadAttentionBackward_Qloop_Xdl_CShuffle_V1
         KPerBlock,
         Gemm1NPerBlock,
         Gemm1KPerBlock,
+        Gemm2KPerBlock,
         AK1,
         BK1,
         B1K1,
@@ -1283,6 +1285,7 @@ struct DeviceBatchedMultiheadAttentionBackward_Qloop_Xdl_CShuffle_V1
             << MPerBlock << ", "
             << Gemm1NPerBlock << ", "
             << Gemm1KPerBlock << ", "
+            << Gemm2KPerBlock << ", "
             << B1K1 << ", "
             << getGemmSpecializationString(GemmSpec) << ", "
             << "ASpec" << getTensorSpecializationString(ASpec) << ", "