ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-02 04:31:25 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix for Add the API to load SGPR (#2913)

Browse Source 
* Revert "Revert "[CK-Tile] Add the API to load SGPR  (#2878)" (#2904)"

This reverts commit f161b5b738.

* Fix: sgpr minor issue

* cyclic dependency resolved

* clang formatted

* removing unused variable

* clang formatted

---------

Co-authored-by: Illia Silin <98187287+illsilin@users.noreply.github.com>
This commit is contained in:
Khushbu Agarwal
2025-09-25 10:32:42 -07:00
committed by GitHub
parent 64e61b8647
commit b56e5d1d79
41 changed files with 224 additions and 173 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
include/ck_tile/ops/topk_softmax/kernel/topk_softmax_kernel.hpp
View File

@@ -96,9 +96,9 @@ struct TopkSoftmaxKernel
if(block_row_id > kargs.num_rows)
return;
index_t block_os_inp = __builtin_amdgcn_readfirstlane(block_row_id * kargs.stride_input);
index_t block_os_out = __builtin_amdgcn_readfirstlane(block_row_id * kargs.stride_output);
index_t num_rows_rem = __builtin_amdgcn_readfirstlane(kargs.num_rows - block_row_id);
index_t block_os_inp = amd_wave_read_first_lane(block_row_id * kargs.stride_input);
index_t block_os_out = amd_wave_read_first_lane(block_row_id * kargs.stride_output);
index_t num_rows_rem = amd_wave_read_first_lane(kargs.num_rows - block_row_id);
const auto input_window = [&]() {
const InputType* p_input =