On 2/10/25 08:05, Matteo Malvestiti wrote:
*BUT*, as soon as I introduce a simple global refinement on the serial
triangulation, like:
1.5) tria_base.refine_global(5);
deali.II 9.5.0: Works fine
deal.II 9.6.0: ERROR - or actually goes in a deadlock during the call to
/tria_pft.create_triangulation(description);/
/
/
-------
I attach the simple demo script as a .cpp file.
Does anybody know something about this?
Matteo -- can you help debug this a bit further? Does the problem
already happen with two MPI processes, or how many do you need? Do you
actually need 5 global refinements, or what is the minimum so that the
problem happens?
Lastly, let's assume that you can get away with just two MPI processes.
Then it's not so difficult to run the two processes in a debugger. If
the program ends in a deadlock, can you obtain a backtrace for where the
program is stuck? (If you've never run a parallel program in a debugger,
check out lecture 41.25 at
https://www.math.colostate.edu/~bangerth/videos.html .)
Thanks already for producing a fairly minimal program -- this is exactly
what is necessary to debug the issue!
Best
Wolfgang
--
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see
https://groups.google.com/d/forum/dealii?hl=en
---
You received this message because you are subscribed to the Google Groups "deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to dealii+unsubscr...@googlegroups.com.
To view this discussion visit
https://groups.google.com/d/msgid/dealii/145c9f56-9f41-4e3b-8e6e-5eff607be8c3%40colostate.edu.
