ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-03 13:11:25 +00:00
Code Issues Packages Projects Releases Wiki Activity

[CK_TILE][REGRESSION] Correct blockSize in Generic2dBlockShape (c254f… (#2837)

Browse Source 
* [CK_TILE][REGRESSION] Correct blockSize in Generic2dBlockShape (c254f3d7b4 )

WarpPerBlock_M * WarpPerBlock_N are not equal with ThreadPerBlock_M * ThreadPerBlock_N /warpSize. we should calculate BlockSize from WarpPerBlock_M * WarpPerBlock_N

To compatible with wave32, function GetBlockSize is added to calculate correct size in host side.

* fix blocksize for all kernel related with generic2dblockshap

* remove constexpr for blocks
This commit is contained in:
linqunAMD
2025-09-16 23:47:55 +08:00
committed by GitHub
parent 671adb59c5
commit b7a806f244
10 changed files with 63 additions and 26 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
test/ck_tile/add_rmsnorm2d_rdquant/instances/add_rmsnorm2d_rdquant_fwd_instance_common.hpp
View File

@@ -58,7 +58,7 @@ float add_rmsnorm2d_rdquant_fwd_(const S& s, A a)
using Kernel = ck_tile::AddRmsnorm2dRdquantFwd<Pipeline>;
const dim3 grids = Kernel::GridSize(a);
constexpr dim3 blocks = Kernel::BlockSize();
const dim3 blocks = Kernel::BlockSize();
constexpr ck_tile::index_t kBlockPerCu = 1;
auto kargs = Kernel::MakeKargs(a);

2
test/ck_tile/moe_smoothquant/instances/moe_smoothquant_instance_common.hpp
View File

@@ -53,7 +53,7 @@ float moe_smoothquant_(const S& s, A a)
using Kernel = ck_tile::MoeSmoothquant<Pipeline>;
const dim3 grids = Kernel::GridSize(a);
constexpr dim3 blocks = Kernel::BlockSize();
const dim3 blocks = Kernel::BlockSize();
constexpr ck_tile::index_t kBlockPerCu = 1;
auto kargs = Kernel::MakeKargs(a);

2
test/ck_tile/rmsnorm2d/generate.py
View File

@@ -201,7 +201,7 @@ float rmsnorm2d_fwd_(const S& s, A a)
using Kernel = ck_tile::Rmsnorm2dFwd<Pipeline, Epilogue>;
const dim3 grids = Kernel::GridSize(a);
constexpr dim3 blocks = Kernel::BlockSize();
const dim3 blocks = Kernel::BlockSize();
constexpr ck_tile::index_t kBlockPerCu = 1;
auto kargs = Kernel::MakeKargs(a);

2
test/ck_tile/smoothquant/instances/smoothquant_instance_common.hpp
View File

@@ -49,7 +49,7 @@ float smoothquant_(const S& s, A a)
using Kernel = ck_tile::Smoothquant<Pipeline>;
const dim3 grids = Kernel::GridSize(a);
constexpr dim3 blocks = Kernel::BlockSize();
const dim3 blocks = Kernel::BlockSize();
constexpr ck_tile::index_t kBlockPerCu = 1;
auto kargs = Kernel::MakeKargs(a);