#include <gridwise_elementwise_2d.hpp>
|
static __device__ void | Run (const InGridDescTuple &in_grid_desc_tuple, const OutGridDescTuple &out_grid_desc_tuple, const InDataTypePointerTuple &p_in_global_tuple, const OutDataTypePointerTuple &p_out_global_tuple, const Block2TileMap &block_2_tile_map, const ElementwiseOperation &elementwise_op, const index_t block_id=get_block_1d_id()) |
|
◆ PassThroughOp
template<typename InGridDescTuple , typename OutGridDescTuple , typename InDataTypePointerTuple , typename OutDataTypePointerTuple , typename Block2TileMap , typename ElementwiseOperation , index_t BlockSize, index_t M0PerBlock, index_t M1PerBlock, index_t M0PerThread, index_t M1PerThread, typename ThreadClusterArrangeOrder , typename InScalarPerVectorSeq , typename OutScalarPerVectorSeq , index_t SrcVectorDim, index_t DstVectorDim>
using ck::GridwiseElementwise< InGridDescTuple, OutGridDescTuple, InDataTypePointerTuple, OutDataTypePointerTuple, Block2TileMap, ElementwiseOperation, BlockSize, M0PerBlock, M1PerBlock, M0PerThread, M1PerThread, ThreadClusterArrangeOrder, InScalarPerVectorSeq, OutScalarPerVectorSeq, SrcVectorDim, DstVectorDim >::PassThroughOp = tensor_operation::element_wise::PassThrough |
◆ Run()
template<typename InGridDescTuple , typename OutGridDescTuple , typename InDataTypePointerTuple , typename OutDataTypePointerTuple , typename Block2TileMap , typename ElementwiseOperation , index_t BlockSize, index_t M0PerBlock, index_t M1PerBlock, index_t M0PerThread, index_t M1PerThread, typename ThreadClusterArrangeOrder , typename InScalarPerVectorSeq , typename OutScalarPerVectorSeq , index_t SrcVectorDim, index_t DstVectorDim>
static __device__ void ck::GridwiseElementwise< InGridDescTuple, OutGridDescTuple, InDataTypePointerTuple, OutDataTypePointerTuple, Block2TileMap, ElementwiseOperation, BlockSize, M0PerBlock, M1PerBlock, M0PerThread, M1PerThread, ThreadClusterArrangeOrder, InScalarPerVectorSeq, OutScalarPerVectorSeq, SrcVectorDim, DstVectorDim >::Run |
( |
const InGridDescTuple & |
in_grid_desc_tuple, |
|
|
const OutGridDescTuple & |
out_grid_desc_tuple, |
|
|
const InDataTypePointerTuple & |
p_in_global_tuple, |
|
|
const OutDataTypePointerTuple & |
p_out_global_tuple, |
|
|
const Block2TileMap & |
block_2_tile_map, |
|
|
const ElementwiseOperation & |
elementwise_op, |
|
|
const index_t |
block_id = get_block_1d_id() |
|
) |
| |
|
inlinestatic |
◆ I0
template<typename InGridDescTuple , typename OutGridDescTuple , typename InDataTypePointerTuple , typename OutDataTypePointerTuple , typename Block2TileMap , typename ElementwiseOperation , index_t BlockSize, index_t M0PerBlock, index_t M1PerBlock, index_t M0PerThread, index_t M1PerThread, typename ThreadClusterArrangeOrder , typename InScalarPerVectorSeq , typename OutScalarPerVectorSeq , index_t SrcVectorDim, index_t DstVectorDim>
constexpr auto ck::GridwiseElementwise< InGridDescTuple, OutGridDescTuple, InDataTypePointerTuple, OutDataTypePointerTuple, Block2TileMap, ElementwiseOperation, BlockSize, M0PerBlock, M1PerBlock, M0PerThread, M1PerThread, ThreadClusterArrangeOrder, InScalarPerVectorSeq, OutScalarPerVectorSeq, SrcVectorDim, DstVectorDim >::I0 = Number<0>{} |
|
staticconstexpr |
◆ I1
template<typename InGridDescTuple , typename OutGridDescTuple , typename InDataTypePointerTuple , typename OutDataTypePointerTuple , typename Block2TileMap , typename ElementwiseOperation , index_t BlockSize, index_t M0PerBlock, index_t M1PerBlock, index_t M0PerThread, index_t M1PerThread, typename ThreadClusterArrangeOrder , typename InScalarPerVectorSeq , typename OutScalarPerVectorSeq , index_t SrcVectorDim, index_t DstVectorDim>
constexpr auto ck::GridwiseElementwise< InGridDescTuple, OutGridDescTuple, InDataTypePointerTuple, OutDataTypePointerTuple, Block2TileMap, ElementwiseOperation, BlockSize, M0PerBlock, M1PerBlock, M0PerThread, M1PerThread, ThreadClusterArrangeOrder, InScalarPerVectorSeq, OutScalarPerVectorSeq, SrcVectorDim, DstVectorDim >::I1 = Number<1>{} |
|
staticconstexpr |
◆ NumInput
template<typename InGridDescTuple , typename OutGridDescTuple , typename InDataTypePointerTuple , typename OutDataTypePointerTuple , typename Block2TileMap , typename ElementwiseOperation , index_t BlockSize, index_t M0PerBlock, index_t M1PerBlock, index_t M0PerThread, index_t M1PerThread, typename ThreadClusterArrangeOrder , typename InScalarPerVectorSeq , typename OutScalarPerVectorSeq , index_t SrcVectorDim, index_t DstVectorDim>
constexpr index_t ck::GridwiseElementwise< InGridDescTuple, OutGridDescTuple, InDataTypePointerTuple, OutDataTypePointerTuple, Block2TileMap, ElementwiseOperation, BlockSize, M0PerBlock, M1PerBlock, M0PerThread, M1PerThread, ThreadClusterArrangeOrder, InScalarPerVectorSeq, OutScalarPerVectorSeq, SrcVectorDim, DstVectorDim >::NumInput = InDataTypePointerTuple::Size() |
|
staticconstexpr |
◆ NumOutput
template<typename InGridDescTuple , typename OutGridDescTuple , typename InDataTypePointerTuple , typename OutDataTypePointerTuple , typename Block2TileMap , typename ElementwiseOperation , index_t BlockSize, index_t M0PerBlock, index_t M1PerBlock, index_t M0PerThread, index_t M1PerThread, typename ThreadClusterArrangeOrder , typename InScalarPerVectorSeq , typename OutScalarPerVectorSeq , index_t SrcVectorDim, index_t DstVectorDim>
constexpr index_t ck::GridwiseElementwise< InGridDescTuple, OutGridDescTuple, InDataTypePointerTuple, OutDataTypePointerTuple, Block2TileMap, ElementwiseOperation, BlockSize, M0PerBlock, M1PerBlock, M0PerThread, M1PerThread, ThreadClusterArrangeOrder, InScalarPerVectorSeq, OutScalarPerVectorSeq, SrcVectorDim, DstVectorDim >::NumOutput = OutDataTypePointerTuple::Size() |
|
staticconstexpr |
The documentation for this struct was generated from the following file:
- /home/docs/checkouts/readthedocs.org/user_builds/advanced-micro-devices-composable-kernel/checkouts/docs-6.4.3/include/ck/tensor_operation/gpu/grid/gridwise_elementwise_2d.hpp