ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-06-29 19:28:33 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add buffer_load_lds synchronization

Browse Source
This commit is contained in:
Feng Shijie
2025-09-11 07:06:52 +00:00
parent 1c05d3be89
commit 2e36b3c3ff
1 changed files with 1 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
include/ck_tile/ops/flatmm/pipeline/mixed_prec_flatmm_pipeline_agmem_bgmem_creg_v1.hpp
View File

@@ -679,6 +679,7 @@ struct F16xMXF4FlatmmPipelineAGmemBGmemCRegV1
async_load_tile(lds_tile_a, dram_tile_a);
};
auto prefill_lds_a_stage2 = [&](auto lds_tile_a) {
async_load_fence();
// data has been stored in lds, no need more operation.
static_assert(std::is_same_v<AElementFunction, identity>,
"buffer_load_lds don't support element func fot A before mfma");