Dries Kimpe wrote:
> It looks as if the problem is not really due to Open MPI, but to the
> included ROM-IO:
In the end it turned out to be a combination of a newer Open MPI with an older ROMIO ...
Quoting Rob Ross (on the pnetcdf mailing list):
The problem is that the OpenMPI group takes ROMIO as-is from an older release. Then they pass
MPI_COMBINER_SUBARRAY as-is to ROMIO, which in that release it does not understand. Then ROMIO blows up.
I'll try recompiling Open MPI with a newer version of ROMIO.