ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-07-15 03:30:11 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix the bug

Browse Source
This commit is contained in:
aska-0096
2025-08-12 08:46:06 +00:00
parent 96d24497f5
commit bcc05eee62
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
include/ck_tile/ops/gemm/block/block_gemm_areg_breg_creg_v1.hpp
View File

@@ -258,14 +258,14 @@ struct BlockGemmARegBRegCRegV1
// read A warp tensor from A Block window // read A warp tensor from A Block window
AWarpTensor a_warp_tensor; AWarpTensor a_warp_tensor;
a_warp_tensor.get_thread_buffer() = a_block_tensor.get_y_sliced_thread_data( a_warp_tensor.get_thread_buffer() = a_block_tensor.get_y_sliced_thread_data(
merge_sequences(sequence<mIter, kIter>{}, a_warp_y_index_zeros), merge_sequences(sequence<kIter, mIter>{}, a_warp_y_index_zeros),
merge_sequences(sequence<1, 1>{}, a_warp_y_lengths)); merge_sequences(sequence<1, 1>{}, a_warp_y_lengths));
static_for<0, NIterPerWarp, 1>{}([&](auto nIter) { static_for<0, NIterPerWarp, 1>{}([&](auto nIter) {
// read B warp tensor from B block tensor // read B warp tensor from B block tensor
BWarpTensor b_warp_tensor; BWarpTensor b_warp_tensor;
b_warp_tensor.get_thread_buffer() = b_block_tensor.get_y_sliced_thread_data( b_warp_tensor.get_thread_buffer() = b_block_tensor.get_y_sliced_thread_data(
merge_sequences(sequence<nIter, kIter>{}, b_warp_y_index_zeros), merge_sequences(sequence<kIter, nIter>{}, b_warp_y_index_zeros),
merge_sequences(sequence<1, 1>{}, b_warp_y_lengths)); merge_sequences(sequence<1, 1>{}, b_warp_y_lengths));
// read C warp tensor from C block tensor // read C warp tensor from C block tensor