Hi,
I am having problems outputting fields like Ez, Hx etc. to HDF while
running C++ meep in MPI.
Here is a simple test program:
////////////////////////////// Test program starts
#include <meep.hpp>
#include <iostream>
using namespace meep;
double eps(const vec &p) {
return p.x()+p.y();
}
int main(int argc, char **argv) {
initialize mpi(argc, argv); // do this even for non-MPI Meep
double resolution = 20; // pixels per distance
volume v = vol2d(5,10, resolution); // 5x10 2d cell
structure s(v, eps, pml(1.0));
fields f(&s);
double freq = 0.3, fwidth = 0.1;
gaussian_src_time src(freq, fwidth);
f.add_point_source(Ez, src, vec(1.1, 2.3));
while (f.time() < f.last_source_time()) {
f.step();
}
f.output_hdf5(Dielectric, v.surroundings());
f.output_hdf5(Ez, v.surroundings());
return 0;
}
////////////////////////////// Test program ends
While running it normally, without MPI it will output both
eps-000150.00.h5 (with the structure) and ez-000150.00.h5 (with the ez
component)
and it will finish without problems. After that I am able to view the
ez-000150.00.h5 or convert it to VTK with h5tovtk.
BUT, while running it in MPI mode, with 2 processors for example it
stucks always on creation of the ez-000150.00.h5 file. Like if
one node was waiting for the other and the program never finishes.
What is strange is that outputting the structure works but
outputting fields does not.
Am I doing something wrong or is it an internal problem with mpi meep ?
Thomas Jefferson
_______________________________________________
meep-discuss mailing list
meep-discuss@ab-initio.mit.edu
http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
