Skip to content

GitLab

Commit f881fa70 authored by Qianfeng Zhang's avatar Qianfeng Zhang
Browse files

Use __builtin_amdgcn_sched_barrier(0x7f) in the pipeline

parent 660261a2
Hide whitespace changes
Inline Side-by-side
Showing with 1 addition and 1 deletion
include/ck_tile/ops/fmha/pipeline/block_fmha_pipeline_qr_ks_vs_async.hpp
View file @ f881fa70
...@@ -666,7 +666,7 @@ struct BlockFmhaPipelineQRKSVSAsync ...@@ -666,7 +666,7 @@ struct BlockFmhaPipelineQRKSVSAsync
smem_ptr, seqlen_k_start + i_total_loops * kN0, p_compute, randval_dram_window); smem_ptr, seqlen_k_start + i_total_loops * kN0, p_compute, randval_dram_window);
} }
__builtin_amdgcn_sched_barrier(0); __builtin_amdgcn_sched_barrier(0x7f);
if constexpr(std::is_same_v<VLayout, ck_tile::tensor_layout::gemm::RowMajor>) if constexpr(std::is_same_v<VLayout, ck_tile::tensor_layout::gemm::RowMajor>)
{ {
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment