format

include/ck/tensor_operation/gpu/element/unary_element_wise_operation.hpp

@@ -66,7 +66,6 @@ __device__ inline half2_t pki4_to_half2(pk_i4_t q)
 
     return amd_assembly_pk_add_f16(bit_cast<half2_t>(lo), bit_cast<half2_t>(SUB));
 #endif
-
 }
 
 struct PassThroughPack8
@@ -87,10 +86,14 @@ struct PassThroughPack8
         vector_type<half_t, 8> dst;
         vector_type<pk_i4_t, 4> src{x};
 
-        dst.template AsType<half2_t>()(Number<0>{}) = pki4_to_half2(src.template AsType<pk_i4_t>()[Number<0>{}]);
-        dst.template AsType<half2_t>()(Number<1>{}) = pki4_to_half2(src.template AsType<pk_i4_t>()[Number<1>{}]);
-        dst.template AsType<half2_t>()(Number<2>{}) = pki4_to_half2(src.template AsType<pk_i4_t>()[Number<2>{}]);
-        dst.template AsType<half2_t>()(Number<3>{}) = pki4_to_half2(src.template AsType<pk_i4_t>()[Number<3>{}]);
+        dst.template AsType<half2_t>()(Number<0>{}) =
+            pki4_to_half2(src.template AsType<pk_i4_t>()[Number<0>{}]);
+        dst.template AsType<half2_t>()(Number<1>{}) =
+            pki4_to_half2(src.template AsType<pk_i4_t>()[Number<1>{}]);
+        dst.template AsType<half2_t>()(Number<2>{}) =
+            pki4_to_half2(src.template AsType<pk_i4_t>()[Number<2>{}]);
+        dst.template AsType<half2_t>()(Number<3>{}) =
+            pki4_to_half2(src.template AsType<pk_i4_t>()[Number<3>{}]);
 
         y          = dst.template AsType<half8_t>()[Number<0>{}];
 #endif

include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v3.hpp

@@ -1370,7 +1370,6 @@ struct GridwiseGemm_xdl_cshuffle_v3
                                                                          c_thread_buf,
                                                                          num_k_block_main_loop);
 
-
         // shuffle C and write out
         {
             static_assert(MXdlPerWave % CShuffleMXdlPerWavePerShuffle == 0 &&

include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp

@@ -1025,8 +1025,7 @@ struct ThreadwiseTensorSliceTransfer_v4
 
         if constexpr(is_same_v<remove_cvref_t<SrcData>, pk_i4_t>)
         {
-            static_assert(SrcScalarPerVector % PackedSize == 0,
-                    "pk data N cannot be 1");
+            static_assert(SrcScalarPerVector % PackedSize == 0, "pk data N cannot be 1");
         }
     }
 
@@ -1126,8 +1125,9 @@ struct ThreadwiseTensorSliceTransfer_v4
 
             using src_vector_t = typename decltype(src_tmp_vector)::type;
 
-            //const bool is_src_valid = coordinate_has_valid_offset_assuming_visible_index_is_valid(
-                //src_desc, src_data_coord);
+            // const bool is_src_valid =
+            // coordinate_has_valid_offset_assuming_visible_index_is_valid( src_desc,
+            // src_data_coord);
             const bool is_src_valid = true;
 
             // copy data from src_buf into src_tmp_vector

include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v3r1.hpp

@@ -82,12 +82,11 @@ struct ThreadwiseTensorSliceTransfer_v3r1
             static_assert(is_same_v<remove_cvref_t<SrcData>, remove_cvref_t<DstData>>,
                           "SrcData != DstData");
 
-            static_assert(SrcScalarPerVector_ % PackedSize == 0 && DstScalarPerVector_ % PackedSize == 0,
+            static_assert(SrcScalarPerVector_ % PackedSize == 0 &&
+                              DstScalarPerVector_ % PackedSize == 0,
                           "SrcScalarPerVector_ and DstScalarPerVector_ cannot be 1");
 
-            static_assert(
-                    SrcVectorDim == DstVectorDim,
-                    "pk_i4_t does not support transpose");
+            static_assert(SrcVectorDim == DstVectorDim, "pk_i4_t does not support transpose");
         }
     }
 
@@ -875,8 +874,8 @@ struct ThreadwiseTensorSliceTransfer_v3r1
 
     private:
     static constexpr auto src_thread_scratch_desc_ = decltype(GetSrcThreadScratchDescriptor()){};
-    //static constexpr auto src_oob_thread_scratch_desc_ =
-        //decltype(GetSrcThreadScratchDescriptor()){};
+    // static constexpr auto src_oob_thread_scratch_desc_ =
+    // decltype(GetSrcThreadScratchDescriptor()){};
     static constexpr auto dst_thread_scratch_desc_ = decltype(GetDstThreadScratchDescriptor()){};
 
     using SrcThreadScratch =