fixed merge conflict after merge with develop

include/ck_tile/ops/epilogue.hpp

@@ -3,5 +3,6 @@
 
 #pragma once
 
+#include "ck_tile/ops/epilogue/cshuffle_epilogue.hpp"
 #include "ck_tile/ops/epilogue/default_2d_epilogue.hpp"
 #include "ck_tile/ops/common/tensor_layout.hpp"

include/ck_tile/ops/fmha/block/block_masking.hpp

@@ -308,9 +308,9 @@ struct SimplifiedGenericAttentionMask
     {
         auto [origin_start, origin_end] = GetTileRangeAlongX(i_y, height, width);
 
-        const index_t x_per_split = ck_tile::max(1, x_total / num_splits);
+        const index_t x_per_split = ck_tile::max(1, integer_divide_ceil(x_total, num_splits));
         const index_t split_start = x_per_split * i_split;
-        const index_t split_end = (i_split == num_splits - 1 ? x_total : split_start + x_per_split);
+        const index_t split_end   = split_start + x_per_split;
 
         return ck_tile::make_tuple(ck_tile::max(origin_start, split_start),
                                    ck_tile::min(origin_end, split_end));

include/ck_tile/ops/fmha/kernel/fmha_fwd_splitkv_combine_tile_partitioner.hpp

@@ -13,21 +13,20 @@ struct FmhaFwdSplitKVCombineTilePartitioner
     static constexpr ck_tile::index_t kM0 = kM0_;
     static constexpr ck_tile::index_t kN1 = kN1_;
 
-    CK_TILE_HOST static constexpr auto GridSize(ck_tile::index_t batch_size_,
-                                                ck_tile::index_t nhead_,
-                                                ck_tile::index_t seqlen_q_,
-                                                ck_tile::index_t hdim_v_)
+    CK_TILE_HOST static constexpr auto GridSize(ck_tile::index_t batch_size,
+                                                ck_tile::index_t nhead,
+                                                ck_tile::index_t max_seqlen_q,
+                                                ck_tile::index_t hdim_v)
     {
         // TODO: this may need tuning
-        return dim3(ck_tile::integer_divide_ceil(seqlen_q_, kM0) *
-                        ck_tile::integer_divide_ceil(hdim_v_, kN1),
-                    nhead_,
-                    batch_size_);
+        return dim3(ck_tile::integer_divide_ceil(max_seqlen_q, kM0) *
+                        ck_tile::integer_divide_ceil(hdim_v, kN1),
+                    nhead,
+                    batch_size);
     }
 
     CK_TILE_DEVICE auto operator()(ck_tile::index_t /*seqlen_q*/, ck_tile::index_t hdim_v)
     {
-        // const index_t num_tile_m0 = seqlen_q / kM0;
         const index_t num_tile_n1 = ck_tile::integer_divide_ceil(hdim_v, kN1);
 
         const index_t i_block = blockIdx.x;

include/ck_tile/ops/fmha/kernel/fmha_fwd_splitkv_tile_partitioner.hpp

@@ -20,12 +20,12 @@ struct FmhaFwdSplitKVTilePartitioner
 
     __host__ static constexpr auto GridSize(ck_tile::index_t batch_size,
                                             ck_tile::index_t nhead,
-                                            ck_tile::index_t seqlen_q,
+                                            ck_tile::index_t max_seqlen_q,
                                             ck_tile::index_t hdim_v,
                                             ck_tile::index_t num_splits)
     {
         // TODO: this may need tuning
-        return dim3(ck_tile::integer_divide_ceil(seqlen_q, kM0) *
+        return dim3(ck_tile::integer_divide_ceil(max_seqlen_q, kM0) *
                         ck_tile::integer_divide_ceil(hdim_v, kN1),
                     nhead * num_splits,
                     batch_size);

include/ck_tile/ops/gemm.hpp

@@ -23,12 +23,14 @@
 #include "ck_tile/ops/gemm/block/block_gemm_problem.hpp"
 #include "ck_tile/ops/gemm/kernel/gemm_kernel.hpp"
 #include "ck_tile/ops/gemm/kernel/gemm_tile_partitioner.hpp"
-#include "ck_tile/ops/gemm/pipeline/block_gemm_pipeline_agmem_bgmem_creg_v1.hpp"
-#include "ck_tile/ops/gemm/pipeline/block_gemm_pipeline_agmem_bgmem_creg_v1_default_policy.hpp"
-#include "ck_tile/ops/gemm/pipeline/block_gemm_pipeline_agmem_bgmem_creg_v2.hpp"
-#include "ck_tile/ops/gemm/pipeline/block_gemm_pipeline_agmem_bgmem_creg_v2_default_policy.hpp"
-#include "ck_tile/ops/gemm/pipeline/block_gemm_pipeline_problem.hpp"
+#include "ck_tile/ops/gemm/pipeline/gemm_universal_pipeline_ag_bg_cr_policy.hpp"
+#include "ck_tile/ops/gemm/pipeline/gemm_pipeline_agmem_bgmem_creg_v1.hpp"
+#include "ck_tile/ops/gemm/pipeline/gemm_pipeline_agmem_bgmem_creg_v1_default_policy.hpp"
+#include "ck_tile/ops/gemm/pipeline/gemm_pipeline_agmem_bgmem_creg_v2.hpp"
+#include "ck_tile/ops/gemm/pipeline/gemm_pipeline_agmem_bgmem_creg_v2_default_policy.hpp"
+#include "ck_tile/ops/gemm/pipeline/gemm_pipeline_problem.hpp"
 #include "ck_tile/ops/gemm/pipeline/tile_gemm_shape.hpp"
+#include "ck_tile/ops/gemm/pipeline/tile_gemm_traits.hpp"
 #include "ck_tile/ops/gemm/warp/warp_gemm.hpp"
 #include "ck_tile/ops/gemm/warp/warp_gemm_attribute_mfma.hpp"
 #include "ck_tile/ops/gemm/warp/warp_gemm_attribute_mfma_impl.hpp"