MPI_Abort after IF refactor
This issue is related to MR 157 from @Raphael
I get some MPI_Abort
when the code run in parallel. I git bisect
ed the code and the problem is somewhere in this merge (v13.0
is fine).
More background: When I run a (rather complex) code with more than 2 processes I get MPI_Abort
. I've run a parallel debugger (DDT) on CSCS cluster, and the offending line seems to be physics/electrics.cc:120 (function electrics.cc:set_elec_tissue_properties
), where I get param_globals::gregion[i].name
not defined on one process. So strdup
fails. However, it is weird, so maybe the error is somewhere else. Please note that other tests from the tutorial work just fine, also with more than 2 processes.
The python code is not mine but from @la816 , and hosted here: https://bitbucket.org/carpusers/af_inducibility/src/master/
I'm new to openCARP and Luca's code, so I cannot really say what is done differently with respect to regression tests or the tutorials. In case I can share the parameter file and the call to openCARP.