ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-06-30 19:57:40 +00:00
Code Issues Packages Projects Releases Wiki Activity

remove schedule barrier

Browse Source
This commit is contained in:
aska-0096
2025-05-13 08:49:09 +00:00
parent 4f7be0f2b5
commit 656d32038b
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_blockscale_b_preshuffle_v3.hpp
View File

@@ -786,7 +786,7 @@ struct BlockwiseGemmXdlops_pipeline_blockscale_bpreshuffle_v3<BlockGemmPipelineS
// We need new compiler to enable this feature
// HotLoopScheduler();
__builtin_amdgcn_sched_barrier(0);
// __builtin_amdgcn_sched_barrier(0);
};
LoopFunc(I0, I1);
@@ -1034,7 +1034,7 @@ struct BlockwiseGemmXdlops_pipeline_blockscale_bpreshuffle_v3<BlockGemmPipelineS
});
// Let's leak last MFMA block to epilogue region, cover the potential lds-shuffle
// latency
// // __builtin_amdgcn_sched_barrier(0);
// __builtin_amdgcn_sched_barrier(0);
}
else
{