ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-03 13:11:25 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix bug where scaling may not be applied in some code path (#526)

Browse Source 
* fix bug where scaling may not be applied in some code path

* more test

* revert accidental example code changes
This commit is contained in:
Anthony Chang
2022-12-03 01:43:34 +08:00
committed by GitHub
parent 23ecf0fa9e
commit d156709432
3 changed files with 15 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
include/ck/tensor_operation/gpu/grid/gridwise_batched_gemm_softmax_gemm_xdl_cshuffle_v1.hpp
View File

@@ -796,6 +796,11 @@ struct GridwiseBatchedGemmSoftmaxGemm_Xdl_CShuffle
}
});
}
else
{
static_for<0, acc_thread_buf.Size(), 1>{}(
[&](auto i) { acc_element_op(acc_thread_buf(i), acc_thread_buf[i]); });
}
block_sync_lds(); // wait for lds read in gemm0 blockwise gemm