clang format

src/targets/gpu/device/argmax.cpp

@@ -30,8 +30,7 @@ void argmax(hipStream_t stream, const argument& result, const argument& arg, int
             block_size *= 2;
         }
 
-        launch(
-            stream, batch_shape.elements() * block_size, block_size)([=](auto idx) __device__ {
+        launch(stream, batch_shape.elements() * block_size, block_size)([=](auto idx) __device__ {
             size_t thr_idx = idx.local;
             size_t blk_idx = idx.group;
             using type     = device_type<std::remove_cv_t<typename decltype(output)::value_type>>;
@@ -47,7 +46,7 @@ void argmax(hipStream_t stream, const argument& result, const argument& arg, int
             lds_index[max_block_size] = 0;
             for(size_t i = thr_idx; i < round_item_num; i += block_size)
             {
-                if (i < batch_item_num)
+                if(i < batch_item_num)
                 {
                     data_idx[axis]     = i;
                     lds_index[thr_idx] = i;
@@ -61,7 +60,7 @@ void argmax(hipStream_t stream, const argument& result, const argument& arg, int
                 remaining_item_num -= block_size;
             }
 
-            if (thr_idx == 0)
+            if(thr_idx == 0)
             {
                 output[batch_idx] = lds_index[max_block_size];
             }

src/targets/gpu/device/argmin.cpp

@@ -30,8 +30,7 @@ void argmin(hipStream_t stream, const argument& result, const argument& arg, int
             block_size *= 2;
         }
 
-        launch(
-            stream, batch_shape.elements() * block_size, block_size)([=](auto idx) __device__ {
+        launch(stream, batch_shape.elements() * block_size, block_size)([=](auto idx) __device__ {
             size_t thr_idx = idx.local;
             size_t blk_idx = idx.group;
             using type     = device_type<std::remove_cv_t<typename decltype(output)::value_type>>;
@@ -47,7 +46,7 @@ void argmin(hipStream_t stream, const argument& result, const argument& arg, int
             lds_index[max_block_size] = 0;
             for(size_t i = thr_idx; i < round_item_num; i += block_size)
             {
-                if (i < batch_item_num)
+                if(i < batch_item_num)
                 {
                     data_idx[axis]     = i;
                     lds_index[thr_idx] = i;
@@ -61,7 +60,7 @@ void argmin(hipStream_t stream, const argument& result, const argument& arg, int
                 remaining_item_num -= block_size;
             }
 
-            if (thr_idx == 0)
+            if(thr_idx == 0)
             {
                 output[batch_idx] = lds_index[max_block_size];
             }

src/targets/gpu/include/migraphx/gpu/device/reduce_opers.hpp

@@ -38,7 +38,12 @@ inline __device__ void reduce_max(T* data_ptr, size_t block_size, size_t thr_idx
 }
 
 template <class T>
-inline __device__ void reduce_argmax(T* data_ptr, int64_t* index_ptr, size_t block_size, size_t thr_idx, size_t item_num, size_t max_index)
+inline __device__ void reduce_argmax(T* data_ptr,
+                                     int64_t* index_ptr,
+                                     size_t block_size,
+                                     size_t thr_idx,
+                                     size_t item_num,
+                                     size_t max_index)
 {
     while(true)
     {
@@ -46,7 +51,7 @@ inline __device__ void reduce_argmax(T* data_ptr, int64_t* index_ptr, size_t blo
         auto size   = item_num / 2;
         for(size_t i = thr_idx; i < size; i += block_size)
         {
-            if (data_ptr[i] < data_ptr[i + stride])
+            if(data_ptr[i] < data_ptr[i + stride])
             {
                 data_ptr[i]  = data_ptr[i + stride];
                 index_ptr[i] = index_ptr[i + stride];
@@ -61,7 +66,7 @@ inline __device__ void reduce_argmax(T* data_ptr, int64_t* index_ptr, size_t blo
 
     if(thr_idx == 0)
     {
-        if (data_ptr[max_index] < data_ptr[0])
+        if(data_ptr[max_index] < data_ptr[0])
         {
             data_ptr[max_index]  = data_ptr[0];
             index_ptr[max_index] = index_ptr[0];
@@ -72,7 +77,8 @@ inline __device__ void reduce_argmax(T* data_ptr, int64_t* index_ptr, size_t blo
 }
 
 template <class T>
-inline __device__ void reduce_argmin(T* data_ptr, int64_t* index_ptr, size_t block_size, size_t thr_idx, size_t item_num)
+inline __device__ void
+reduce_argmin(T* data_ptr, int64_t* index_ptr, size_t block_size, size_t thr_idx, size_t item_num)
 {
     size_t min_index = item_num;
     while(true)
@@ -81,7 +87,7 @@ inline __device__ void reduce_argmin(T* data_ptr, int64_t* index_ptr, size_t blo
         auto size   = item_num / 2;
         for(size_t i = thr_idx; i < size; i += block_size)
         {
-            if (data_ptr[i] > data_ptr[i + stride])
+            if(data_ptr[i] > data_ptr[i + stride])
             {
                 data_ptr[i]  = data_ptr[i + stride];
                 index_ptr[i] = index_ptr[i + stride];
@@ -96,7 +102,7 @@ inline __device__ void reduce_argmin(T* data_ptr, int64_t* index_ptr, size_t blo
 
     if(thr_idx == 0)
     {
-        if (data_ptr[min_index] > data_ptr[0])
+        if(data_ptr[min_index] > data_ptr[0])
         {
             data_ptr[min_index]  = data_ptr[0];
             index_ptr[min_index] = index_ptr[0];