cuda_transpose_impl.h File Reference

Implementation for CUDA transposition kernels - proxy file to hide kernel header if unused. More...

#include "inc/util/cuda_span.h"
#include "inc/util/span.h"
#include "inc/util/scalars.h"
#include "inc/dense/cuda_transpose_impl_cukrn.h"

Include dependency graph for cuda_transpose_impl.h:

This graph shows which files directly or indirectly include this file:

Namespaces
	syten
	Syten namespace.
	syten::CudaDenseTensorImpl
	Implementation namespace for CUDA dense tensors.

Functions
template<Rank rank, typename Scalar >
void	syten::CudaDenseTensorImpl::cuda_transpose_kernel (CudaConstSpan< Scalar > inp, CudaMutSpan< Scalar > out, ConstSpan< IndexNumber > perm, ConstSpan< Index > dim, Conj do_conj)
	Wrapper around the CUDA transpose kernels which sets everything up such that the functions in cuda_transpose_impl_cukrn.h really only have to launch the kernels. More...

Detailed Description

Implementation for CUDA transposition kernels - proxy file to hide kernel header if unused.