Commit 7da35f54 authored by Shucai Xiao's avatar Shucai Xiao
Browse files

clang format

parent 3e70d01b
...@@ -54,7 +54,8 @@ void logsoftmax(hipStream_t stream, const argument& result, const argument& arg, ...@@ -54,7 +54,8 @@ void logsoftmax(hipStream_t stream, const argument& result, const argument& arg,
auto item_num = (remaining_item_num > block_size) ? block_size : remaining_item_num; auto item_num = (remaining_item_num > block_size) ? block_size : remaining_item_num;
// reduce_max(lds_data, block_size, thr_idx, item_num, max_block_size); // reduce_max(lds_data, block_size, thr_idx, item_num, max_block_size);
block_reduce<type, max_op<type>>(lds_data, max_op<type>{}, block_size, thr_idx, item_num, max_block_size); block_reduce<type, max_op<type>>(
lds_data, max_op<type>{}, block_size, thr_idx, item_num, max_block_size);
remaining_item_num -= block_size; remaining_item_num -= block_size;
} }
...@@ -77,7 +78,8 @@ void logsoftmax(hipStream_t stream, const argument& result, const argument& arg, ...@@ -77,7 +78,8 @@ void logsoftmax(hipStream_t stream, const argument& result, const argument& arg,
auto item_num = (remaining_item_num > block_size) ? block_size : remaining_item_num; auto item_num = (remaining_item_num > block_size) ? block_size : remaining_item_num;
// reduce_sum(lds_data, block_size, thr_idx, item_num, max_block_size); // reduce_sum(lds_data, block_size, thr_idx, item_num, max_block_size);
block_reduce<type, sum_op<type>>(lds_data, sum_op<type>{}, block_size, thr_idx, item_num, max_block_size); block_reduce<type, sum_op<type>>(
lds_data, sum_op<type>{}, block_size, thr_idx, item_num, max_block_size);
remaining_item_num -= block_size; remaining_item_num -= block_size;
} }
......
...@@ -42,7 +42,7 @@ inline __device__ void block_reduce(T* data_ptr, ...@@ -42,7 +42,7 @@ inline __device__ void block_reduce(T* data_ptr,
auto size = item_num / 2; auto size = item_num / 2;
for(std::size_t i = thr_idx; i < size; i += block_size) for(std::size_t i = thr_idx; i < size; i += block_size)
{ {
//data_ptr[i] = ::max(to_hip_type(data_ptr[i]), to_hip_type(data_ptr[i + stride])); // data_ptr[i] = ::max(to_hip_type(data_ptr[i]), to_hip_type(data_ptr[i + stride]));
data_ptr[i] = op(data_ptr[i], data_ptr[i + stride]); data_ptr[i] = op(data_ptr[i], data_ptr[i + stride]);
} }
__syncthreads(); __syncthreads();
...@@ -62,7 +62,6 @@ inline __device__ void block_reduce(T* data_ptr, ...@@ -62,7 +62,6 @@ inline __device__ void block_reduce(T* data_ptr,
__syncthreads(); __syncthreads();
} }
template <class T> template <class T>
inline __device__ void reduce_max(T* data_ptr, inline __device__ void reduce_max(T* data_ptr,
std::size_t block_size, std::size_t block_size,
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment