ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-04-19 22:39:03 +00:00
Code Issues Packages Projects Releases Wiki Activity

Support Wave32 in CK_TILE - Part 1 (#2594)

Browse Source 
* Support wave32/wave64 in CK_TILE - Part 1

* remove blocksize in kernel launch

* fix build error

* fix clang format

* fix clang format 2

* fix clang format 3

* fix fmha build error

* fix fmha build 2

* fix fmha build 3

* fix build error 4

* address review comment

* update change log

* replace KernelBlockSize with kBlockSize

* fix CI fail

* fix clang format

* address review comment and rebase code.

* fix universal test fail

---------

Co-authored-by: Lin, Qun <Quentin.Lin+amdeng@amd.com>
Co-authored-by: Thomas Ning <Thomas.Ning@amd.com>
This commit is contained in:
linqunAMD
2025-08-19 01:08:31 +08:00
committed by GitHub
parent 26d3300930
commit 9fcc1ee9fd
113 changed files with 610 additions and 531 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
example/ck_tile/20_grouped_convolution/grouped_convolution_backward_weight.cpp
View File

@@ -78,7 +78,6 @@ float grouped_conv_bwd_weight(const ck_tile::GroupedConvBwdWeightHostArgs& args,
typename GroupedConvTraitsType::ImplicitGemmDsLayout,
ck_tile::tensor_layout::gemm::RowMajor,
CDEElementWise,
CodegenPipelineProblem::kBlockSize,
TilePartitioner::MPerBlock,
TilePartitioner::NPerBlock,
M_Warp,
@@ -98,8 +97,8 @@ float grouped_conv_bwd_weight(const ck_tile::GroupedConvBwdWeightHostArgs& args,
ConvEpilogue>;
auto kargs = Kernel::MakeKernelArgs(args);
const dim3 grids = Kernel::GridSize(kargs);
constexpr dim3 blocks = Kernel::BlockSize();
const dim3 grids = Kernel::GridSize(kargs);
const dim3 blocks = Kernel::BlockSize();
if(!Kernel::IsSupportedArgument(kargs))
{
@@ -123,7 +122,7 @@ float grouped_conv_bwd_weight(const ck_tile::GroupedConvBwdWeightHostArgs& args,
float ave_time = ck_tile::launch_kernel_time_mask(
s,
Kernel::Preprocess(kargs, s),
ck_tile::make_kernel<blocks.x, kBlockPerCu>(Kernel{}, grids, blocks, 0, kargs));
ck_tile::make_kernel<kBlockPerCu>(Kernel{}, grids, blocks, 0, kargs));
return ave_time;
};

7
example/ck_tile/20_grouped_convolution/grouped_convolution_forward.cpp
View File

@@ -77,7 +77,6 @@ float grouped_conv_fwd(const ck_tile::GroupedConvFwdHostArgs& args, const ck_til
typename GroupedConvTraitsType::ImplicitGemmDsLayout,
ck_tile::tensor_layout::gemm::RowMajor,
CDEElementWise,
CodegenPipelineProblem::kBlockSize,
TilePartitioner::MPerBlock,
TilePartitioner::NPerBlock,
M_Warp,
@@ -97,8 +96,8 @@ float grouped_conv_fwd(const ck_tile::GroupedConvFwdHostArgs& args, const ck_til
ConvEpilogue>;
auto kargs = Kernel::MakeKernelArgs(args);
const dim3 grids = Kernel::GridSize(kargs);
constexpr dim3 blocks = Kernel::BlockSize();
const dim3 grids = Kernel::GridSize(kargs);
const dim3 blocks = Kernel::BlockSize();
if(!Kernel::IsSupportedArgument(kargs))
{
@@ -120,7 +119,7 @@ float grouped_conv_fwd(const ck_tile::GroupedConvFwdHostArgs& args, const ck_til
}
float ave_time = ck_tile::launch_kernel(
s, ck_tile::make_kernel<blocks.x, kBlockPerCu>(Kernel{}, grids, blocks, 0, kargs));
s, ck_tile::make_kernel<kBlockPerCu>(Kernel{}, grids, blocks, 0, kargs));
return ave_time;
};