StefanoPetrilli
|
b2230a6d6d
|
Improved critical region U
|
3 years ago |
|
StefanoPetrilli
|
f921aa47c4
|
Master region for tempCounter.clear()
(Might have solved stucked runs)
|
3 years ago |
|
StefanoPetrilli
|
532701031e
|
Extendend parallel region after SEND PACKET BUNDLE
Nothing parallelizable founded
|
3 years ago |
|
StefanoPetrilli
|
b079d71f30
|
Further optimizations PARALLEL_PROCESS_EXPOSED_VERTEX_B
|
3 years ago |
|
StefanoPetrilli
|
e2ca97ca47
|
Removed one critical region from PARALLEL_PROCESS_EXPOSED_VERTEX_B
|
3 years ago |
|
StefanoPetrilli
|
5bc4f2a080
|
PROCESS MATCHED VERTICES parallelization improvement
|
3 years ago |
|
StefanoPetrilli
|
2c8dc2ffdd
|
PROCESS MATCHED VERTICES parallelization draft
|
3 years ago |
|
StefanoPetrilli
|
f3d7b3ab5e
|
False sharing fix
|
3 years ago |
|
StefanoPetrilli
|
766ef320c2
|
Refactoring + critical(Mate)
|
3 years ago |
|
StefanoPetrilli
|
002239f5b6
|
False sharing fix
|
3 years ago |
|
StefanoPetrilli
|
70b7c4db55
|
PARALLEL_PROCESS_EXPOSED_VERTEX_B named critical sections
|
3 years ago |
|
StefanoPetrilli
|
2cac21b345
|
fix and reformatting
|
3 years ago |
|
StefanoPetrilli
|
6180f29f39
|
PARALLEL_COMPUTE_CANDIDATE_MATE_B is now paralle and correct
|
3 years ago |
|
StefanoPetrilli
|
b4bfdd83e5
|
computeCandidateMate and isAlreadyMatched
|
3 years ago |
|
StefanoPetrilli
|
1140669ea7
|
firstComputeCandidateMate
|
3 years ago |
|
StefanoPetrilli
|
919e2a2918
|
PARALLEL_PROCESS_EXPOSED_VERTEX_B is actually not parallelizable. Atleast not as I was doing.
|
3 years ago |
|
StefanoPetrilli
|
baffff3d93
|
Instable PARALLEL_PROCESS_EXPOSED_VERTEX_B
|
3 years ago |
|
StefanoPetrilli
|
25a603debe
|
PARALLEL_COMPUTE_CANDIDATE_MATE_B
|
3 years ago |
|
StefanoPetrilli
|
a20f0d47e7
|
Solved the static queue out of scope problem
|
3 years ago |
|
StefanoPetrilli
|
76e04ee997
|
The OMP and MPI version is now separated in two different files
|
3 years ago |
|
StefanoPetrilli
|
0a8debe43a
|
Single parallel regions with multiple for cycles
Added OMP for testing
|
3 years ago |
|
StefanoPetrilli
|
8f6dc5fac2
|
verGhostPtrInitialization is now parallelized
|
3 years ago |
|
StefanoPetrilli
|
7d40fde21d
|
verGhostIndInitialization and Ghost2LocalInitialization cycles parallelization
|
3 years ago |
|
StefanoPetrilli
|
1760afbe97
|
Time tracking in algoDistEdge
|
3 years ago |
|
StefanoPetrilli
|
60f90804d5
|
Time tracking in MatchBox
|
3 years ago |
Salvatore Filippone
|
e02df3725e
|
Bump version 1.0.1
|
3 years ago |
|
Salvatore Filippone
|
697f325df6
|
Fix use of SuperLU_Dist, configure checks and ifdefs
|
3 years ago |
|
Salvatore Filippone
|
425743939c
|
Fix for new SuperLU_Dist version, change configure
|
3 years ago |
|
Salvatore Filippone
|
7e48a0a742
|
Fix defines for SLUD v7
|
3 years ago |
|
Salvatore Filippone
|
87cdd76f8d
|
Fix spurious error notification with prec%descr
|
3 years ago |
|
Salvatore Filippone
|
a8f4009cb1
|
Take out spurious csize and maxnlev from parmatch aggregator object.
|
3 years ago |
|
Salvatore Filippone
|
794080e386
|
Fix target coarse size handling.
|
3 years ago |
|
Salvatore Filippone
|
818f7a78a0
|
Do not call %default on setting coarse_solve
|
3 years ago |
|
Salvatore Filippone
|
939d7c9a89
|
Do not invoke default() after setting KRM for coarse solver.
|
3 years ago |
|
Salvatore Filippone
|
af178daa84
|
Modify dump method to print base level matrix.
|
3 years ago |
|
Salvatore Filippone
|
5768238f66
|
Typographical fixes.
|
3 years ago |
|
Salvatore Filippone
|
9d11a99ed4
|
Fix settings in samples/PDEGEN
|
3 years ago |
|
Salvatore Filippone
|
af75364c54
|
Fix matchbox internal interface names.
|
3 years ago |
|
Salvatore Filippone
|
2542c0fda4
|
Do not print matching statistics
|
4 years ago |
|
Salvatore Filippone
|
8482067b52
|
Deactivate MINNRG
|
4 years ago |
|
Salvatore Filippone
|
7319dab30f
|
Deactivate MINNRG
|
4 years ago |
|
Salvatore Filippone
|
4bbba3ebd7
|
Fix interface inconsistencies
|
4 years ago |
|
Salvatore Filippone
|
1fa94d0372
|
Fix AS%FREE()
|
4 years ago |
|
Salvatore Filippone
|
555d7433b7
|
Redefine interface of prec%descr to get INFO
|
4 years ago |
Cirdans-Home
|
50951ef636
|
Fixed set of coarse matrix for BJAC
|
4 years ago |
|
Cirdans-Home
|
47eba23460
|
Added error check and defaults
|
4 years ago |
|
Salvatore Filippone
|
ee218171e7
|
New configure script
|
4 years ago |
|
Cirdans-Home
|
8d3ebba561
|
Removed deprecated MPI function
|
4 years ago |
|
Salvatore Filippone
|
1541da5fbf
|
Fix name of %linmap component
|
4 years ago |
|
Salvatore Filippone
|
6f0f5feb34
|
Fix for SERIAL_MPI compilation
|
4 years ago |