tidy up

src/include/common.hip.hpp

@@ -62,10 +62,9 @@ __host__ __device__ constexpr T min(T x, Ts... xs)
 
     return x < y ? x : y;
 }
-}// namespace mod_conv
+} // namespace mod_conv
 
 #if DEVICE_BACKEND_HIP
 // cast a pointer of LDS to its address
 extern "C" __attribute__((address_space(3))) void* __to_local(void* p)[[hc]];
 #endif
-

src/include/gridwise_convolution_implicit_gemm_v2_chwn_cyxk_khwn.hip.hpp

@@ -287,10 +287,10 @@ struct GridwiseConvolutionImplicitGemm_v2_chwn_cyxk_khwn
 #elif 0
                     blockwise_gemm.Run_asm
 #endif
-                    (p_wei_block + wei_cyxk_block_desc.Get1dIndex(0, y, x, 0),
-                     p_in_block + y * Wi + x,
-                     p_out_thread,
-                     f_accum);
+                        (p_wei_block + wei_cyxk_block_desc.Get1dIndex(0, y, x, 0),
+                         p_in_block + y * Wi + x,
+                         p_out_thread,
+                         f_accum);
                 }
             }
         }

src/include/gridwise_convolution_implicit_gemm_v2_chwn_cyxk_khwn_lds_double_buffer.hip.hpp

@@ -36,11 +36,14 @@ template <index_t GridSize,
           index_t WeiBlockCopyDataPerRead>
 struct GridwiseConvolutionImplicitGemm_v2_chwn_cyxk_khwn_lds_double_buffer
 {
-    __host__ __device__ constexpr GridwiseConvolutionImplicitGemm_v2_chwn_cyxk_khwn_lds_double_buffer() {}
+    __host__
+        __device__ constexpr GridwiseConvolutionImplicitGemm_v2_chwn_cyxk_khwn_lds_double_buffer()
+    {
+    }
 
     __device__ void Run(const Float* const __restrict__ p_in_global,
-                               const Float* const __restrict__ p_wei_global,
-                               Float* const __restrict__ p_out_global) const
+                        const Float* const __restrict__ p_wei_global,
+                        Float* const __restrict__ p_out_global) const
     {
         constexpr auto I0 = Number<0>{};
         constexpr auto I1 = Number<1>{};