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 on the error (#1956)

Browse Source
This commit is contained in:
Thomas Ning
2025-03-07 13:43:52 -08:00
committed by GitHub
parent 0e8e711ec8
commit 9d51d17dd0
2 changed files with 7 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuffle_v1.hpp
View File

@@ -211,11 +211,12 @@ struct BlockwiseGemmXdlops_pipeline_bpreshuffle_v1<BlockGemmPipelineScheduler::I
});
// A local
static_for<0, num_ds_read_inst_a / 2 * mfma_interleave, 1>{}([&](auto i) {
ignore = i;
__builtin_amdgcn_sched_group_barrier(0x008, 1, 0); // MFMA
__builtin_amdgcn_sched_group_barrier(0x100, 2 / mfma_interleave, 0); // DS read
});
static_for<0, MPerXDL == 32 ? num_ds_read_inst_a / 2 : num_ds_read_inst_a, 1>{}(
[&](auto i) {
ignore = i;
__builtin_amdgcn_sched_group_barrier(0x008, 1, 0); // MFMA
__builtin_amdgcn_sched_group_barrier(0x100, MPerXDL == 32 ? 2 : 1, 0); // DS read
});
}
template <bool HasMainLoop,