TileLoadThreadGroup Struct Reference

TileLoadThreadGroup Struct Reference#

Composable Kernel: ck::GridwiseGemm_k0mk1_k0nk1_mn_xdl_waveletmodel_cshuffle< ABDataType, FloatGemmAcc, EDataTypeShuffle, EDataType, AElementwiseOperation, BElementwiseOperation, EElementwiseOperation, CGlobalMemoryDataOperation, AGridDesc_M_K, BGridDesc_N_K, EGridDesc_M_N, NumGemmKPrefetchStage, TileLoadThreadGroupSize, TileMathThreadGroupSize, MPerBlock, NPerBlock, KPerBlock, AK1Value, BK1Value, MPerXdl, NPerXdl, MXdlPerWave, NXdlPerWave, ABlockTransferThreadClusterLengths_AK0_M_AK1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_AK1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, BBlockTransferThreadClusterLengths_BK0_N_BK1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_BK1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, CShuffleMXdlPerWavePerShuffle, CShuffleNXdlPerWavePerShuffle, CShuffleBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, CShuffleBlockTransferScalarPerVector_NPerBlock >::TileLoadThreadGroup Struct Reference
Static Public Member Functions | List of all members
ck::GridwiseGemm_k0mk1_k0nk1_mn_xdl_waveletmodel_cshuffle< ABDataType, FloatGemmAcc, EDataTypeShuffle, EDataType, AElementwiseOperation, BElementwiseOperation, EElementwiseOperation, CGlobalMemoryDataOperation, AGridDesc_M_K, BGridDesc_N_K, EGridDesc_M_N, NumGemmKPrefetchStage, TileLoadThreadGroupSize, TileMathThreadGroupSize, MPerBlock, NPerBlock, KPerBlock, AK1Value, BK1Value, MPerXdl, NPerXdl, MXdlPerWave, NXdlPerWave, ABlockTransferThreadClusterLengths_AK0_M_AK1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_AK1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, BBlockTransferThreadClusterLengths_BK0_N_BK1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_BK1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, CShuffleMXdlPerWavePerShuffle, CShuffleNXdlPerWavePerShuffle, CShuffleBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, CShuffleBlockTransferScalarPerVector_NPerBlock >::TileLoadThreadGroup Struct Reference

#include <gridwise_gemm_xdl_waveletmodel_cshuffle.hpp>

Static Public Member Functions

static __device__ constexpr index_t GetNumOfThread ()
static __device__ constexpr bool IsBelong ()
static __device__ index_t GetThreadId ()

Member Function Documentation

◆ GetNumOfThread()

template<typename ABDataType, typename FloatGemmAcc, typename EDataTypeShuffle, typename EDataType, typename AElementwiseOperation, typename BElementwiseOperation, typename EElementwiseOperation, InMemoryDataOperationEnum CGlobalMemoryDataOperation, typename AGridDesc_M_K, typename BGridDesc_N_K, typename EGridDesc_M_N, index_t NumGemmKPrefetchStage, index_t TileLoadThreadGroupSize, index_t TileMathThreadGroupSize, index_t MPerBlock, index_t NPerBlock, index_t KPerBlock, index_t AK1Value, index_t BK1Value, index_t MPerXdl, index_t NPerXdl, index_t MXdlPerWave, index_t NXdlPerWave, typename ABlockTransferThreadClusterLengths_AK0_M_AK1, typename ABlockTransferThreadClusterArrangeOrder, typename ABlockTransferSrcAccessOrder, index_t ABlockTransferSrcVectorDim, index_t ABlockTransferSrcScalarPerVector, index_t ABlockTransferDstScalarPerVector_AK1, bool AThreadTransferSrcResetCoordinateAfterRun, index_t ABlockLdsExtraM, typename BBlockTransferThreadClusterLengths_BK0_N_BK1, typename BBlockTransferThreadClusterArrangeOrder, typename BBlockTransferSrcAccessOrder, index_t BBlockTransferSrcVectorDim, index_t BBlockTransferSrcScalarPerVector, index_t BBlockTransferDstScalarPerVector_BK1, bool BThreadTransferSrcResetCoordinateAfterRun, index_t BBlockLdsExtraN, index_t CShuffleMXdlPerWavePerShuffle, index_t CShuffleNXdlPerWavePerShuffle, typename CShuffleBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, index_t CShuffleBlockTransferScalarPerVector_NPerBlock>
__device__ constexpr index_t ck::GridwiseGemm_k0mk1_k0nk1_mn_xdl_waveletmodel_cshuffle< ABDataType, FloatGemmAcc, EDataTypeShuffle, EDataType, AElementwiseOperation, BElementwiseOperation, EElementwiseOperation, CGlobalMemoryDataOperation, AGridDesc_M_K, BGridDesc_N_K, EGridDesc_M_N, NumGemmKPrefetchStage, TileLoadThreadGroupSize, TileMathThreadGroupSize, MPerBlock, NPerBlock, KPerBlock, AK1Value, BK1Value, MPerXdl, NPerXdl, MXdlPerWave, NXdlPerWave, ABlockTransferThreadClusterLengths_AK0_M_AK1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_AK1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, BBlockTransferThreadClusterLengths_BK0_N_BK1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_BK1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, CShuffleMXdlPerWavePerShuffle, CShuffleNXdlPerWavePerShuffle, CShuffleBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, CShuffleBlockTransferScalarPerVector_NPerBlock >::TileLoadThreadGroup::GetNumOfThread ( )
inlinestaticconstexpr

◆ GetThreadId()

template<typename ABDataType, typename FloatGemmAcc, typename EDataTypeShuffle, typename EDataType, typename AElementwiseOperation, typename BElementwiseOperation, typename EElementwiseOperation, InMemoryDataOperationEnum CGlobalMemoryDataOperation, typename AGridDesc_M_K, typename BGridDesc_N_K, typename EGridDesc_M_N, index_t NumGemmKPrefetchStage, index_t TileLoadThreadGroupSize, index_t TileMathThreadGroupSize, index_t MPerBlock, index_t NPerBlock, index_t KPerBlock, index_t AK1Value, index_t BK1Value, index_t MPerXdl, index_t NPerXdl, index_t MXdlPerWave, index_t NXdlPerWave, typename ABlockTransferThreadClusterLengths_AK0_M_AK1, typename ABlockTransferThreadClusterArrangeOrder, typename ABlockTransferSrcAccessOrder, index_t ABlockTransferSrcVectorDim, index_t ABlockTransferSrcScalarPerVector, index_t ABlockTransferDstScalarPerVector_AK1, bool AThreadTransferSrcResetCoordinateAfterRun, index_t ABlockLdsExtraM, typename BBlockTransferThreadClusterLengths_BK0_N_BK1, typename BBlockTransferThreadClusterArrangeOrder, typename BBlockTransferSrcAccessOrder, index_t BBlockTransferSrcVectorDim, index_t BBlockTransferSrcScalarPerVector, index_t BBlockTransferDstScalarPerVector_BK1, bool BThreadTransferSrcResetCoordinateAfterRun, index_t BBlockLdsExtraN, index_t CShuffleMXdlPerWavePerShuffle, index_t CShuffleNXdlPerWavePerShuffle, typename CShuffleBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, index_t CShuffleBlockTransferScalarPerVector_NPerBlock>
__device__ index_t ck::GridwiseGemm_k0mk1_k0nk1_mn_xdl_waveletmodel_cshuffle< ABDataType, FloatGemmAcc, EDataTypeShuffle, EDataType, AElementwiseOperation, BElementwiseOperation, EElementwiseOperation, CGlobalMemoryDataOperation, AGridDesc_M_K, BGridDesc_N_K, EGridDesc_M_N, NumGemmKPrefetchStage, TileLoadThreadGroupSize, TileMathThreadGroupSize, MPerBlock, NPerBlock, KPerBlock, AK1Value, BK1Value, MPerXdl, NPerXdl, MXdlPerWave, NXdlPerWave, ABlockTransferThreadClusterLengths_AK0_M_AK1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_AK1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, BBlockTransferThreadClusterLengths_BK0_N_BK1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_BK1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, CShuffleMXdlPerWavePerShuffle, CShuffleNXdlPerWavePerShuffle, CShuffleBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, CShuffleBlockTransferScalarPerVector_NPerBlock >::TileLoadThreadGroup::GetThreadId ( )
inlinestatic

◆ IsBelong()

template<typename ABDataType, typename FloatGemmAcc, typename EDataTypeShuffle, typename EDataType, typename AElementwiseOperation, typename BElementwiseOperation, typename EElementwiseOperation, InMemoryDataOperationEnum CGlobalMemoryDataOperation, typename AGridDesc_M_K, typename BGridDesc_N_K, typename EGridDesc_M_N, index_t NumGemmKPrefetchStage, index_t TileLoadThreadGroupSize, index_t TileMathThreadGroupSize, index_t MPerBlock, index_t NPerBlock, index_t KPerBlock, index_t AK1Value, index_t BK1Value, index_t MPerXdl, index_t NPerXdl, index_t MXdlPerWave, index_t NXdlPerWave, typename ABlockTransferThreadClusterLengths_AK0_M_AK1, typename ABlockTransferThreadClusterArrangeOrder, typename ABlockTransferSrcAccessOrder, index_t ABlockTransferSrcVectorDim, index_t ABlockTransferSrcScalarPerVector, index_t ABlockTransferDstScalarPerVector_AK1, bool AThreadTransferSrcResetCoordinateAfterRun, index_t ABlockLdsExtraM, typename BBlockTransferThreadClusterLengths_BK0_N_BK1, typename BBlockTransferThreadClusterArrangeOrder, typename BBlockTransferSrcAccessOrder, index_t BBlockTransferSrcVectorDim, index_t BBlockTransferSrcScalarPerVector, index_t BBlockTransferDstScalarPerVector_BK1, bool BThreadTransferSrcResetCoordinateAfterRun, index_t BBlockLdsExtraN, index_t CShuffleMXdlPerWavePerShuffle, index_t CShuffleNXdlPerWavePerShuffle, typename CShuffleBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, index_t CShuffleBlockTransferScalarPerVector_NPerBlock>
__device__ constexpr bool ck::GridwiseGemm_k0mk1_k0nk1_mn_xdl_waveletmodel_cshuffle< ABDataType, FloatGemmAcc, EDataTypeShuffle, EDataType, AElementwiseOperation, BElementwiseOperation, EElementwiseOperation, CGlobalMemoryDataOperation, AGridDesc_M_K, BGridDesc_N_K, EGridDesc_M_N, NumGemmKPrefetchStage, TileLoadThreadGroupSize, TileMathThreadGroupSize, MPerBlock, NPerBlock, KPerBlock, AK1Value, BK1Value, MPerXdl, NPerXdl, MXdlPerWave, NXdlPerWave, ABlockTransferThreadClusterLengths_AK0_M_AK1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_AK1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, BBlockTransferThreadClusterLengths_BK0_N_BK1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_BK1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, CShuffleMXdlPerWavePerShuffle, CShuffleNXdlPerWavePerShuffle, CShuffleBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, CShuffleBlockTransferScalarPerVector_NPerBlock >::TileLoadThreadGroup::IsBelong ( )
inlinestaticconstexpr
The documentation for this struct was generated from the following file: