| Cyclops Tensor Framework
    parallel arithmetic on multidimensional arrays | 
performs replication along a dimension, generates 2.5D algs More...
#include <sum_tsr.h>


| Public Member Functions | |
| void | run () | 
| void | print () | 
| int64_t | mem_fp () | 
| returns the number of bytes of buffer space needed  More... | |
| tsum * | clone () | 
| tsum_replicate (tsum *other) | |
| ~tsum_replicate () | |
| tsum_replicate (summation const *s, int const *phys_mapped, int64_t blk_sz_A, int64_t blk_sz_B) | |
|  Public Member Functions inherited from CTF_int::tsum | |
| virtual | ~tsum () | 
| tsum (tsum *other) | |
| tsum (summation const *s) | |
| Data Fields | |
| int64_t | size_A | 
| int64_t | size_B | 
| int | ncdt_A | 
| int | ncdt_B | 
| CommData ** | cdt_A | 
| CommData ** | cdt_B | 
| tsum * | rec_tsum | 
|  Data Fields inherited from CTF_int::tsum | |
| char * | A | 
| algstrct const * | sr_A | 
| char const * | alpha | 
| char * | B | 
| algstrct const * | sr_B | 
| char const * | beta | 
| void * | buffer | 
performs replication along a dimension, generates 2.5D algs
| CTF_int::tsum_replicate::tsum_replicate | ( | tsum * | other | ) | 
Definition at line 193 of file sum_tsr.cxx.
References CTF_int::tsum::clone(), ncdt_A, ncdt_B, rec_tsum, size_A, and size_B.
Referenced by clone().
| CTF_int::tsum_replicate::~tsum_replicate | ( | ) | 
Definition at line 179 of file sum_tsr.cxx.
References CTF_int::cdealloc(), and CTF_int::tsum_virt::rec_tsum.
| CTF_int::tsum_replicate::tsum_replicate | ( | summation const * | s, | 
| int const * | phys_mapped, | ||
| int64_t | blk_sz_A, | ||
| int64_t | blk_sz_B | ||
| ) | 
Definition at line 203 of file sum_tsr.cxx.
References CTF_int::summation::A, CTF_int::alloc_ptr(), ASSERT, CTF_int::summation::B, cdt_A, cdt_B, CTF_int::topology::dim_comm, ncdt_A, ncdt_B, CTF_int::topology::order, size_A, size_B, and CTF_int::tensor::topo.
| 
 | virtual | 
Reimplemented from CTF_int::tsum.
Definition at line 243 of file sum_tsr.cxx.
References tsum_replicate().
| 
 | virtual | 
returns the number of bytes of buffer space needed
Reimplemented from CTF_int::tsum.
Definition at line 247 of file sum_tsr.cxx.
| 
 | virtual | 
Reimplemented from CTF_int::tsum.
Definition at line 162 of file sum_tsr.cxx.
References ctf.core::np(), CTF_int::tsum::print(), and CTF_int::tsum_virt::rec_tsum.
| 
 | virtual | 
Reimplemented from CTF_int::tsum.
Definition at line 251 of file sum_tsr.cxx.
References CTF_int::tsum::A, CTF_int::algstrct::addid(), CTF_int::algstrct::addmop(), CTF_int::CommData::allred(), CTF_int::tsum::alpha, CTF_int::tsum::B, CTF_int::CommData::bcast(), CTF_int::tsum::beta, CTF_int::cdealloc(), cdt_A, cdt_B, CTF_int::algstrct::mdtype(), CTF_int::algstrct::mulid(), ncdt_A, ncdt_B, CTF_int::CommData::rank, rec_tsum, CTF_int::tsum::run(), CTF_int::algstrct::set(), size_A, size_B, CTF_int::tsum::sr_A, and CTF_int::tsum::sr_B.
| CommData** CTF_int::tsum_replicate::cdt_A | 
Definition at line 129 of file sum_tsr.h.
Referenced by run(), and tsum_replicate().
| CommData** CTF_int::tsum_replicate::cdt_B | 
Definition at line 130 of file sum_tsr.h.
Referenced by run(), and tsum_replicate().
| int CTF_int::tsum_replicate::ncdt_A | 
Definition at line 126 of file sum_tsr.h.
Referenced by run(), and tsum_replicate().
| int CTF_int::tsum_replicate::ncdt_B | 
Definition at line 127 of file sum_tsr.h.
Referenced by run(), and tsum_replicate().
| tsum* CTF_int::tsum_replicate::rec_tsum | 
Definition at line 132 of file sum_tsr.h.
Referenced by CTF_int::summation::estimate_time(), run(), and tsum_replicate().
| int64_t CTF_int::tsum_replicate::size_A | 
Definition at line 124 of file sum_tsr.h.
Referenced by run(), and tsum_replicate().
| int64_t CTF_int::tsum_replicate::size_B | 
Definition at line 125 of file sum_tsr.h.
Referenced by run(), and tsum_replicate().