6 #ifndef GHOST_TSMM_INPLACE_H
7 #define GHOST_TSMM_INPLACE_H
83 ghost_densemat *w,
const char *transw,
void *alpha,
void *beta,
int reduce,
int printerror);
ghost_implementation
Possible implementations of a CPU function.
Definition: types.h:323
Header file for type definitions.
int ncolsin
The number of columns for the input densemat.
Definition: tsmm_inplace.h:22
ghost_error ghost_tsmm_inplace_valid(ghost_densemat *x, ghost_densemat *v, const char *transv, ghost_densemat *w, const char *transw, void *alpha, void *beta, int reduce, int printerror)
Check whether TSMM-inplace can be applied instead of GEMM with the given arguments.
Definition: tsmm_inplace.cpp:45
ghost_error
Error return type.
Definition: error.h:23
ghost_alignment
Possible alignments of data access.
Definition: types.h:315
ghost_implementation impl
Definition: tsmm_inplace.h:28
ghost_datatype dt
The data type of the densemats.
Definition: tsmm_inplace.h:18
ghost_datatype
Available primitive data types.
Definition: types.h:248
int ncolsout
The number of columns for the output densemat.
Definition: tsmm_inplace.h:26
ghost_error ghost_tsmm_inplace(ghost_densemat *x, ghost_densemat *w, void *alpha, void *beta)
Multiply a distributed dense tall skinny matrix with a redundant dense matrix in-place.
Definition: tsmm_inplace.cpp:116
ghost_error(* ghost_tsmm_inplace_kernel)(ghost_densemat *, ghost_densemat *, void *, void *)
A tsmm-inplace kernel function.
Definition: tsmm_inplace.h:35
ghost_alignment alignment
Definition: tsmm_inplace.h:29
Definition: tsmm_inplace.h:13
A dense vector/matrix.
Definition: densemat.h:226