d8677b100d
forgot an << endl;
2016-12-10 14:06:42 +01:00
b32cf1fb62
clarified a warning for MPI
2016-12-10 12:35:36 +01:00
ab682cc0bd
A few fixes for MPI
...
main.cpp:
1. Check return value of ParseFDTDSetup and exit if false
2. Use exit instead of return. These are almost identical. But
in my OpenMPI installation the process with teh highes rank
segfaults at the end when using return. This is not the case
with exit. Probably some C++ cleanup problem (destructors).
openems.cpp:
Give Parse_XML_FDTDSetup a deterministic return value.
openems_fdtd_mpi.cpp:
1. Remove the word "only" in an error message because there can
also be too many processes.
2. Fix the indexing variables for SetSplitPos in SetupMPI. Otherwise
more than one split results in an out-of-range exception and
unexpected behavior.
RunOpenEMS_MPI.m:
Apply Settings.MPI.GlobalArgs also to multi-host scenarios.
2016-12-02 19:03:35 +01:00
d5cbfec2be
MPI: fix FDTD setup for non-MPI
...
Signed-off-by: Thorsten Liebig <Thorsten.Liebig@gmx.de>
2016-01-31 13:32:17 +01:00
d75069cfef
MPI: update for MPI usage
...
Signed-off-by: Thorsten Liebig <Thorsten.Liebig@gmx.de>
2016-01-27 18:37:07 +01:00
274800054f
silence some warnings
...
Signed-off-by: Thorsten Liebig <Thorsten.Liebig@gmx.de>
2013-02-07 11:25:34 +01:00
d64e43a73c
run post processing after final time evaluation
...
Signed-off-by: Thorsten Liebig <Thorsten.Liebig@gmx.de>
2012-11-06 12:46:54 +01:00
2f228f2161
update & fix to FDTD simulation statistics
...
Signed-off-by: Thorsten Liebig <Thorsten.Liebig@gmx.de>
2012-09-25 11:24:25 +02:00
2966ba2902
new option to dump FDTD simulation statistics
...
Signed-off-by: Thorsten Liebig <Thorsten.Liebig@gmx.de>
2012-09-24 15:16:47 +02:00
e20121540f
reorganization and cleanup of excitation handling
...
Signed-off-by: Thorsten Liebig <Thorsten.Liebig@gmx.de>
2012-07-16 17:15:10 +02:00
2cbdf2d2b2
timestep length warning only for gaussian pulse excitation type
2012-01-17 11:31:53 +01:00
488772de25
removed misleading timestep progress information, warning about to few timesteps instead
...
Signed-off-by: Thorsten Liebig <Thorsten.Liebig@gmx.de>
2012-01-17 10:22:16 +01:00
c9357823ff
fix in openEMS MPI setup
2011-11-08 10:15:36 +01:00
b34848f323
new engine interface using sse & speed enhancement in energy estimate
...
Note: cylindrical engine interface now based on sse engine interface
--> similar to sse engine & cylindrical engine
2011-11-07 12:07:55 +01:00
01e8a8011a
fixed argument parsing & always redirect MPI output
2011-04-13 16:21:59 +02:00
91ba3a6f6f
MPI: lib info update & compilation fix
2011-04-08 12:26:32 +02:00
62acf5f1b3
ProcessFields now using new VTK_File_IO class for vtk dumps
...
Todo:
- HDF5_File_IO
- replace all old vtk dump method usage and remove them
2011-04-01 15:29:56 +02:00
06eb82832e
MPI: operator saves process table
2011-03-28 10:26:08 +02:00
005eb3a4f6
MPI: added an additional line at the upper MPI-bound for correct material averaging
...
This also fixes the current send and receive issue
--> Engine_MPI::SendReceiveCurrents()
2011-03-15 09:41:29 +01:00
edb40489d7
MPI: tell MPI-operator about global mesh and its position in it
2011-03-14 16:20:33 +01:00
f630abf813
MPI: external libs info reflects MPI usage
2011-03-10 10:39:08 +01:00
9c061e9d84
MPI: redirect output for ranks>0 to file
...
changes to main.cpp:
read global command line arguments first to allow for immediate output redirection
Signed-off-by: Thorsten Liebig <Thorsten.Liebig@gmx.de>
2011-03-10 10:39:08 +01:00
75fb5a0cc9
MPI: fix in handling SplitN_? equal 1 or smaller
2011-03-10 10:39:08 +01:00
8d09a7ee28
MPI: allow requested number of splits
2011-03-10 10:39:08 +01:00
eb7524b6f6
MPI: Setup MPI now allows split in all directions with defitions read from xml-file
...
Warning: depends on API changes in CSXCAD!
2011-03-10 10:39:08 +01:00
14e12f9138
MPI: new methode to define neighbors
2011-03-10 10:39:08 +01:00
729a92c5e1
MPI: move mpi mesh split to openEMS_FDTD_MPI
2011-03-10 10:39:08 +01:00
b395546483
MPI: some fixes in handling the MPI engine
2011-03-10 10:39:07 +01:00
3b402e976b
MPI: disable/rename processings if they are in conflict
2011-03-10 10:39:07 +01:00
e2628740fc
new openEMS_FDTD_MPI for better MPI handling
2011-03-10 10:39:07 +01:00
Georg Michel
Thorsten Liebig