ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-04-19 22:39:03 +00:00
Code Issues Packages Projects Releases Wiki Activity

clang-format (#1612)

Browse Source
This commit is contained in:
rocking
2024-10-30 23:13:30 +08:00
committed by GitHub
parent 24d996aae1
commit 7d9111545f
1 changed files with 9 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
include/ck_tile/ops/reduce/block/block_reduce.hpp
View File

@@ -16,8 +16,8 @@ namespace ck_tile {
// synchronize reduce result (cross lane reduction and broadcast on replicated dimension)
template <typename AccDistributedTensor_, typename ReduceFunc, bool WithBroadcast = true>
CK_TILE_DEVICE void block_tile_reduce_sync(AccDistributedTensor_& acc_tensor,
const ReduceFunc& reduce_func,
bool_constant<WithBroadcast> = {})
const ReduceFunc& reduce_func,
bool_constant<WithBroadcast> = {})
{
using Dstr = typename AccDistributedTensor_::StaticTileDistribution;
using DstrEncode = typename Dstr::DstrEncode;
@@ -116,7 +116,7 @@ CK_TILE_DEVICE void block_tile_reduce_sync(AccDistributedTensor_& acc_tensor,
*/
template <typename AccDistributedTensor_, typename ReduceFunc>
CK_TILE_DEVICE void block_tile_reduce_xor_sync(AccDistributedTensor_& acc_tensor,
const ReduceFunc& reduce_func)
const ReduceFunc& reduce_func)
{
using Dstr = typename AccDistributedTensor_::StaticTileDistribution;
using DstrEncode = typename Dstr::DstrEncode;
@@ -175,9 +175,9 @@ template <typename AccDistributedTensor_,
index_t... InReduceDims,
typename ReduceFunc>
CK_TILE_DEVICE void block_tile_reduce(AccDistributedTensor_& acc_tensor,
const InDistributedTensor_& in_tensor,
sequence<InReduceDims...>,
const ReduceFunc& reduce_func)
const InDistributedTensor_& in_tensor,
sequence<InReduceDims...>,
const ReduceFunc& reduce_func)
{
constexpr auto I0 = number<0>{};
constexpr auto I1 = number<1>{};
@@ -250,9 +250,9 @@ template <typename AccDataType_,
typename ReduceFunc,
typename InDataType_>
CK_TILE_DEVICE auto block_tile_reduce(const InDistributedTensor_& in_tensor,
sequence<InReduceDims...> in_reduce_dims,
const ReduceFunc& reduce_func,
const InDataType_& reduce_init)
sequence<InReduceDims...> in_reduce_dims,
const ReduceFunc& reduce_func,
const InDataType_& reduce_init)
{
using InDataType = typename InDistributedTensor_::DataType;
using AccDataType = remove_cvref_t<AccDataType_>;