clang format

src/targets/gpu/device/logsoftmax.cpp

@@ -41,7 +41,7 @@ void logsoftmax(hipStream_t stream, argument result, argument arg, int axis)
             // load data to lds and compute the batch max
             size_t remaining_item_num = batch_item_num;
             size_t round_item_num     = (batch_item_num + block_size - 1) / block_size * block_size;
-            lds_data[max_block_size]      = input[0];
+            lds_data[max_block_size]  = input[0];
             for(size_t i = thr_idx; i < round_item_num; i += block_size)
             {
                 if(i < batch_item_num)
@@ -61,7 +61,7 @@ void logsoftmax(hipStream_t stream, argument result, argument arg, int axis)
             __syncthreads();
 
             lds_data[max_block_size] = 0;
-            remaining_item_num   = batch_item_num;
+            remaining_item_num       = batch_item_num;
             for(size_t i = thr_idx; i < round_item_num; i += block_size)
             {
                 if(i < batch_item_num)

src/targets/gpu/device/softmax.cpp

@@ -41,7 +41,7 @@ void softmax(hipStream_t stream, argument result, argument arg, int axis)
             // load data to lds and compute the batch max
             size_t remaining_item_num = batch_item_num;
             size_t round_item_num     = (batch_item_num + block_size - 1) / block_size * block_size;
-            lds_data[max_block_size]      = input[0];
+            lds_data[max_block_size]  = input[0];
             for(size_t i = thr_idx; i < round_item_num; i += block_size)
             {
                 if(i < batch_item_num)
@@ -62,7 +62,7 @@ void softmax(hipStream_t stream, argument result, argument arg, int axis)
             __syncthreads();
 
             lds_data[max_block_size] = 0;
-            remaining_item_num   = batch_item_num;
+            remaining_item_num       = batch_item_num;
             for(size_t i = thr_idx; i < round_item_num; i += block_size)
             {
                 if(i < batch_item_num)

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

@@ -11,7 +11,8 @@ namespace gpu {
 namespace device {
 
 template <class T>
-inline __device__ void reduce_max(T* data_ptr, size_t block_size, size_t thr_idx, size_t item_num, size_t max_index)
+inline __device__ void
+reduce_max(T* data_ptr, size_t block_size, size_t thr_idx, size_t item_num, size_t max_index)
 {
     while(true)
     {
@@ -38,7 +39,8 @@ inline __device__ void reduce_max(T* data_ptr, size_t block_size, size_t thr_idx
 }
 
 template <class T>
-inline __device__ void reduce_min(T* data_ptr, size_t block_size, size_t thr_idx, size_t item_num, size_t min_index)
+inline __device__ void
+reduce_min(T* data_ptr, size_t block_size, size_t thr_idx, size_t item_num, size_t min_index)
 {
     while(true)
     {
@@ -65,7 +67,8 @@ inline __device__ void reduce_min(T* data_ptr, size_t block_size, size_t thr_idx
 }
 
 template <class T>
-inline __device__ void reduce_sum(T* data_ptr, size_t block_size, size_t thr_idx, size_t item_num, size_t sum_index)
+inline __device__ void
+reduce_sum(T* data_ptr, size_t block_size, size_t thr_idx, size_t item_num, size_t sum_index)
 {
     while(true)
     {