ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-19 04:19:36 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge commit '4a49dac7c6fff9ffe4d275bed761a79e51188f3c' into develop

Browse Source
This commit is contained in:
assistant-librarian[bot]
2025-08-28 13:19:19 +00:00
parent b36f983558
commit c8234bd387
6 changed files with 30 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
include/ck_tile/ops/fused_moe/kernel/fused_moegemm_kernel.hpp
View File

@@ -1,5 +1,5 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved.
#pragma once
@@ -130,7 +130,7 @@ struct FusedMoeGemmKernel
// static_assert(kBlockPerCu > 0);
using BlockShape = typename Pipeline::BlockShape; // this is FusedMoeGemmShape
static constexpr index_t BlockSize_ = BlockShape::BlockSize;
static constexpr index_t kBlockSize = BlockShape::BlockSize;
using ADataType = typename Pipeline::Problem::ADataType;
using GDataType = typename Pipeline::Problem::GDataType;
@@ -231,7 +231,7 @@ struct FusedMoeGemmKernel
return Partitioner::GridSize(max_num_tokens_padded, hargs.intermediate_size);
}
CK_TILE_HOST static constexpr auto BlockSize() { return dim3(BlockSize_); }
CK_TILE_HOST static constexpr auto BlockSize() { return dim3(kBlockSize); }
CK_TILE_HOST_DEVICE static constexpr index_t GetSmemSize() { return Pipeline::GetSmemSize(); }

4
include/ck_tile/ops/gemm.hpp
View File

@@ -1,5 +1,5 @@
// Copyright © Advanced Micro Devices, Inc., or its affiliates.
// SPDX-License-Identifier: MIT
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved.
#pragma once

4
include/ck_tile/ops/grouped_convolution/kernel/grouped_convolution_backward_data_kernel.hpp
View File

@@ -487,7 +487,7 @@ struct GroupedConvolutionBackwardDataKernel
using GemmDsLayout = remove_cvref_t<typename EpiloguePipeline::DsLayout>;
static constexpr index_t NumDTensor = GroupedConvTraitsType_::NumDTensor;
static constexpr index_t KernelBlockSize = GemmPipeline::BlockSize;
static constexpr index_t kBlockSize = GemmPipeline::BlockSize;
using InDataType = remove_cvref_t<typename GemmPipeline::ADataType>;
using WeiDataType = remove_cvref_t<typename GemmPipeline::BDataType>;
@@ -530,7 +530,7 @@ struct GroupedConvolutionBackwardDataKernel
return dim3(kargs.grid_size_, kargs.GemmBatch, kargs.k_batch);
}
CK_TILE_HOST static constexpr auto BlockSize() { return dim3(KernelBlockSize); }
CK_TILE_HOST static constexpr auto BlockSize() { return dim3(kBlockSize); }
CK_TILE_HOST static constexpr GroupedConvBwdDataKernelArgsSpecialized
MakeKernelArgs(const GroupedConvBwdDataHostArgs& hostArgs)