ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-04 21:51:28 +00:00
Code Issues Packages Projects Releases Wiki Activity

Replace buffer load/store intrinsics with builtins (#1876)

Browse Source 
* replace buffer load/store intrinsics with builtins

* fix clang format

* replace buffer load/store intrinsics with built-ins in ck_tile

* fix clang format

* add switch between buffer intrinsics and built-ins

* change the builtins threshold to clang20

* fix clang format

* fix some compilation errors

* revert changes in ck_tile

* revert changes in ck_tile

* delete all root files and folders when CI completes

* try changing the username in CI

* fix groovy syntax

* add user and group id info to ci dockers

* change ownership of all files in CI to jenkins at the end

* update changelog
This commit is contained in:
Illia Silin
2025-03-05 14:33:28 -08:00
committed by GitHub
parent 4814db3905
commit a88bf76ecc
11 changed files with 3792 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
include/ck_tile/ops/fused_moe/pipeline/fused_moegemm_pipeline_flatmm_uk.hpp
View File

@@ -207,6 +207,7 @@ struct FusedMoeGemmPipeline_FlatmmUk
threadIdx.x % (BlockShape::Block_K0 / kAlignmentA) * kAlignmentA;
},
number<row_ids_a.size()>{});
auto a_res =
make_wave_buffer_resource(reinterpret_cast<const ADataType*>(kargs.a_ptr),
kargs.num_tokens * kargs.stride_token * sizeof(ADataType));
@@ -318,10 +319,10 @@ struct FusedMoeGemmPipeline_FlatmmUk
{0, 0},
dist_);
}();
auto o_res =
make_wave_buffer_resource(reinterpret_cast<const ODataType*>(kargs.o_ptr),
kargs.num_tokens * kargs.stride_token * sizeof(ODataType));
auto row_coords_o = GetRowCoords_O(sorted_tile_id * BlockShape::Block_M0);
auto w_scale = GetWeightScale(
row_coords_o, reinterpret_cast<const TopkWeightDataType*>(kargs.sorted_weight_ptr));