v2 ignore parameter

include/ck/tensor_operation/gpu/device/impl/device_batched_mha_bwd_xdl_cshuffle_kloop_v2.hpp

@@ -185,8 +185,14 @@ __global__ void
 #else
     ignore = p_a_grid;
     ignore = p_b_grid;
+    ignore = p_z_grid;
     ignore = p_b1_grid;
     ignore = p_c_grid;
+    ignore = p_lse_grid;
+    ignore = p_ygrad_grid;
+    ignore = p_qgrad_grid;
+    ignore = p_kgrad_grid;
+    ignore = p_vgrad_grid;
     ignore = a_element_op;
     ignore = b_element_op;
     ignore = acc_element_op;
@@ -194,10 +200,15 @@ __global__ void
     ignore = c_element_op;
     ignore = a_grid_desc_ak0_m_ak1;
     ignore = b_grid_desc_bk0_n_bk1;
+    ignore = c_grid_desc_m0_n0_m1_n1_m2_n2_m3_n3_n4_n5;
     ignore = b1_grid_desc_bk0_n_bk1;
     ignore = c_grid_desc_mblock_mperblock_nblock_nperblock;
+    ignore = lse_grid_desc_m;
+    ignore = vgrad_grid_desc_n_o;
+    ignore = ygrad_grid_desc_m0_o_m1;
     ignore = block_2_ctile_map;
     ignore = batch_count;
+    ignore = mblock;
     ignore = compute_base_ptr_of_batch;
     ignore = c0_matrix_mask;
     ignore = p_drop;

include/ck/tensor_operation/gpu/device/impl/device_batched_mha_bwd_xdl_cshuffle_qloop_v2.hpp

@@ -190,8 +190,14 @@ __global__ void
 #else
     ignore = p_a_grid;
     ignore = p_b_grid;
+    ignore = p_z_grid;
     ignore = p_b1_grid;
     ignore = p_c_grid;
+    ignore = p_lse_grid;
+    ignore = p_ygrad_grid;
+    ignore = p_qgrad_grid;
+    ignore = p_kgrad_grid;
+    ignore = p_vgrad_grid;
     ignore = a_element_op;
     ignore = b_element_op;
     ignore = acc_element_op;
@@ -199,15 +205,21 @@ __global__ void
     ignore = c_element_op;
     ignore = a_grid_desc_ak0_m_ak1;
     ignore = b_grid_desc_bk0_n_bk1;
+    ignore = c_grid_desc_m0_n0_m1_n1_m2_n2_m3_m4_m5_n3;
     ignore = b1_grid_desc_bk0_n_bk1;
     ignore = c_grid_desc_mblock_mperblock_nblock_nperblock;
+    ignore = lse_grid_desc_m;
+    ignore = ygrad_grid_desc_m0_o_m1;
     ignore = block_2_ctile_map;
     ignore = batch_count;
+    ignore = nblock;
     ignore = compute_base_ptr_of_batch;
     ignore = c0_matrix_mask;
     ignore = p_drop;
     ignore = seed;
     ignore = offset;
+    ignore = raw_m_padded;
+    ignore = raw_n_padded;
 #endif // end of if (defined(__gfx908__) || defined(__gfx90a__))
 }
 

include/ck/tensor_operation/gpu/device/impl/device_batched_mha_fwd_xdl_cshuffle_v1.hpp

@@ -187,7 +187,7 @@ __global__ void
     ignore = compute_base_ptr_of_batch;
     ignore = c0_matrix_mask;
     ignore = p_dropout_in_16bits;
-    ignore = GemmAccDataType p_dropout_rescale;
+    ignore = p_dropout_rescale;
     ignore = seed;
     ignore = offset;
 #endif // end of if (defined(__gfx908__) || defined(__gfx90a__))

include/ck/tensor_operation/gpu/device/impl/device_batched_mha_fwd_xdl_cshuffle_v2.hpp

@@ -176,6 +176,8 @@ __global__ void
     ignore = p_b_grid;
     ignore = p_b1_grid;
     ignore = p_c_grid;
+    ignore = p_z_grid;
+    ignore = p_lse_grid;
     ignore = a_element_op;
     ignore = b_element_op;
     ignore = acc_element_op;
@@ -185,10 +187,19 @@ __global__ void
     ignore = b_grid_desc_bk0_n_bk1;
     ignore = b1_grid_desc_bk0_n_bk1;
     ignore = c_grid_desc_mblock_mperblock_nblock_nperblock;
+    ignore = z_grid_desc_m0_n0_m1_n1_m2_n2_m3_n3_n4_n5;
+    ignore = lse_grid_desc_m;
     ignore = block_2_ctile_map;
     ignore = batch_count;
+    ignore = mblock;
     ignore = compute_base_ptr_of_batch;
     ignore = c0_matrix_mask;
+    ignore = p_dropout_in_16bits;
+    ignore = p_dropout_rescale;
+    ignore = seed;
+    ignore = offset;
+    ignore = raw_m_padded;
+    ignore = raw_n_padded;
 #endif // end of if (defined(__gfx908__) || defined(__gfx90a__))
 }