Segmentation fault in ffddm

General Discussion
1

Hi,

I have been trying to run a minimal example with mesh decomposition using ffddmbuildDmesh, but I run into a segmentation fault:

Minimal example code (test.edp):

load "msh3"
macro dimension 3// EOM
include "ffddm.idp"
mesh3 ThGlobal("inputmesh.mesh");
ffddmbuildDmesh(Lap, ThGlobal, mpiCommWorld)

I run the script using:
ff-mpirun -np 4 test.edp

Error log:

  [Lap] Building decomposition from mesh of 549080 elements

===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   PID 242 RUNNING AT 96ad2211bf64
=   EXIT CODE: 139
=   CLEANING UP REMAINING PROCESSES
=   YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
===================================================================================
YOUR APPLICATION TERMINATED WITH THE EXIT STRING: Segmentation fault (signal 11)
This typically refers to a problem with your application.
Please see the FAQ page for debugging suggestions

I’m currently using FreeFem v4.9 and I have encountered no error with ffddmbuildDmesh when running other example scripts (like https://github.com/FreeFem/FreeFem-sources/blob/master/examples/ffddm/diffusion-3d-minimal-direct.edp).

In the past I have successfully used this mesh (inputmesh.mesh) to solve equations in parallel using pARMS in FreeFem v3.6.
I was wondering if ffddmbuildDmesh doesn’t work with arbitrary meshes or if there is another issue with my mesh/code/FreeFem installation?

If anyone is willing to test things out on his side, the script (test.edp) and the mesh (inputmesh.mesh) are available here.

Thanks a lot in advance,
Kind regards,
Lazaros

2

Hello,

it appears that the following example also produces a segmentation fault, which would suggest that your mesh is defective in some way:

mesh3 ThGlobal("inputmesh.mesh");
fespace Uh(ThGlobal,P1);
Uh u = x;
plot(u);

I tried to clean/fix your mesh and understand where the problem lies, but with no success so far.

1 Like
3

Thank you for taking the time to investigate this! Indeed, one of the keywords in the .mesh file was incorrect (Triangle instead of Triangles). Thanks for your help!