Cyclops Tensor Framework
parallel arithmetic on multidimensional arrays
|
#include "spctr_2d_general.h"
#include "../tensor/untyped_tensor.h"
#include "../mapping/mapping.h"
#include "../shared/util.h"
#include <climits>
Go to the source code of this file.
Namespaces | |
CTF_int | |
Functions | |
char * | CTF_int::bcast_step (int edge_len, char *A, bool is_sparse_A, bool move_A, algstrct const *sr_A, int64_t b_A, int64_t s_A, char *buf_A, CommData *cdt_A, int64_t ctr_sub_lda_A, int64_t ctr_lda_A, int nblk_A, int64_t const *size_blk_A, int &new_nblk_A, int64_t *&new_size_blk_A, int64_t *offsets_A, int ib) |
char * | CTF_int::reduce_step_pre (int edge_len, char *C, bool is_sparse_C, bool move_C, algstrct const *sr_C, int64_t b_C, int64_t s_C, char *buf_C, CommData *cdt_C, int64_t ctr_sub_lda_C, int64_t ctr_lda_C, int nblk_C, int64_t const *size_blk_C, int &new_nblk_C, int64_t *&new_size_blk_C, int64_t *offsets_C, int ib, char const *&rec_beta) |
void | CTF_int::reduce_step_post (int edge_len, char *C, bool is_sparse_C, bool move_C, algstrct const *sr_C, int64_t b_C, int64_t s_C, char *buf_C, CommData *cdt_C, int64_t ctr_sub_lda_C, int64_t ctr_lda_C, int nblk_C, int64_t *size_blk_C, int &new_nblk_C, int64_t *&new_size_blk_C, int64_t *offsets_C, int ib, char const *&rec_beta, char const *beta, char *&up_C, char *&new_C, int n_new_C_grps, int &i_new_C_grp, char **new_C_grps) |