| CalculateHasMainK0BlockLoop(index_t K0) | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | inlinestatic |
| CBlockClusterAdaptor typedef | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | |
| CGridDesc_MBlock_MPerBlock_NBlock_NPerBlock typedef | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | |
| CheckValidity(const AGridDesc_B_K0_M_K1 &a_b_k0_m_k1_grid_desc, const BGridDesc_B_K0_N_K1 &b_b_k0_n_k1_grid_desc, const CMNGridDesc &c_m_n_grid_desc, const Block2CTileMap &block_2_ctile_map) | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | inlinestatic |
| FloatAAdjusted typedef | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | |
| FloatBAdjusted typedef | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | |
| GetABlockDescriptor_Batch_K0PerBlock_MPerBlock_K1() | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | inlinestatic |
| GetABlockDescriptor_K0PerBlock_MPerBlock_K1() | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | inlinestatic |
| GetBBlockDescriptor_Batch_K0PerBlock_NPerBlock_K1() | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | inlinestatic |
| GetBBlockDescriptor_K0PerBlock_NPerBlock_K1() | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | inlinestatic |
| GetCBlockDescriptor_MBlock_MPerBlock_NBlock_NPerBlock() | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | inlinestatic |
| GetSharedMemoryNumberOfByte() | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | inlinestatic |
| GridwiseGemmPipe typedef | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | |
| I0 | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| I1 | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| I2 | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| I3 | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| I4 | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| I5 | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| I6 | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| I7 | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| IsValidCompilationParameter() | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | inlinestatic |
| K1 | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| M0PerBlock | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| M1Padding | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| M1PerBlock | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| MakeCBlockClusterAdaptor(const CMNGridDesc &c_m_n_grid_desc, index_t M01, index_t N01, index_t KBatch) | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | inlinestatic |
| MakeCGridDesc_MBlock_MPerBlock_NBlock_NPerBlock(const CMNGridDesc &c_m_n_grid_desc) | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | inlinestatic |
| N0PerBlock | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| N1Padding | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| N1PerBlock | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | static |
| Run(const FloatA *__restrict__ p_a_grid, const FloatB *__restrict__ p_b_grid, FloatC *__restrict__ p_c_grid, void *__restrict__ p_shared, const AGridDesc_B_K0_M_K1 &a_b_k0_m_k1_grid_desc, const BGridDesc_B_K0_N_K1 &b_b_k0_n_k1_grid_desc, const CGridDesc_MBlock_MPerBlock_NBlock_NPerBlock &c_grid_desc_mblock_mperblock_nblock_nperblock, const AElementwiseOperation &a_element_op, const BElementwiseOperation &b_element_op, const CElementwiseOperation &c_element_op, const CBlockClusterAdaptor &c_block_cluster_adaptor) | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | inlinestatic |
| ThisThreadBlock typedef | ck::GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_bwd_weight< BlockSize, FloatA, FloatB, FloatAcc, FloatC, CGlobalMemoryDataOperation, AGridDesc_B_K0_M_K1, BGridDesc_B_K0_N_K1, CMNGridDesc, AElementwiseOperation, BElementwiseOperation, CElementwiseOperation, MPerBlock, NPerBlock, K0PerBlock, MPerXdl, NPerXdl, K1Value, MRepeat, NRepeat, ABlockTransferThreadClusterLengths_K0_M_K1, ABlockTransferThreadClusterArrangeOrder, ABlockTransferSrcAccessOrder, ABlockTransferSrcVectorDim, ABlockTransferSrcScalarPerVector, ABlockTransferDstScalarPerVector_K1, AThreadTransferSrcResetCoordinateAfterRun, ABlockLdsExtraM, ABlockLdsM1PerBlock, ABlockLdsM0PerBlock, ABlockLdsM1Padding, BBlockTransferThreadClusterLengths_K0_N_K1, BBlockTransferThreadClusterArrangeOrder, BBlockTransferSrcAccessOrder, BBlockTransferSrcVectorDim, BBlockTransferSrcScalarPerVector, BBlockTransferDstScalarPerVector_K1, BThreadTransferSrcResetCoordinateAfterRun, BBlockLdsExtraN, BBlockLdsN1PerBlock, BBlockLdsN0PerBlock, BBlockLdsN1Padding, CShuffleMRepeatPerShuffle, CShuffleNRepeatPerShuffle, CBlockTransferScalarPerVector_NWaveNPerXDL, CBlockTransferClusterLengths_MBlock_MPerBlock_NBlock_NPerBlock, ABlockLdsExtraM1Wrw, BBlockLdsExtraN1Wrw, NumGemmKPrefetchStage, PipelineVer, ComputeTypeA, ComputeTypeB > | |