[mpich-discuss] coupling

[Huiwei Lu]

It seems to me the application has went wrong and called MPI_Abort to exit.
Have you tried to run some other MPI programs to see whether MPI workg? e.g. cpi.c under mpi_src_path/examples/

    mpirun -n p 2 ./cpi

Also, could you please also send us the mpi.log you generated to help diagnose the error.

And, if there is no specific reason. You are more than welcomed to use the latest release of MPICH, —currently MPICH 3.0.4 at http://www.mpich.org/, as it has more new features and gets better supported.

--
Huiwei Lu
http://www.mcs.anl.gov/~huiweilu/

On Oct 27, 2013, at 3:39 AM, fereshteh komijani <fereshtehkomijani at gmail.com> wrote:

> dear
> i am a newbie of this forum. i am coupling ROMS and  SWAN model which need MCT and MPICH libraries. therefore MCT has been installed by
> Code:
> ./configure CC=gcc FC=gfortran CPPFLAGS=-I/home/usr/mct/include LDFLAGS=-L/home/usr/mctcal/lib --prefix=/home/usr/mct
> make install
> and mpich2 release 1.4.1p1 by
> Code:
> /configure FC=gfortran CC=gcc  --prefix=/home/<user>/mpich2_install_new 2>&1 |tee c.txt
> make 2>&1 |tee m.txt
> make install  2>&1 |tee mi.txt
> but for running coupled model, when i set 1 cpu for SWAN model and 1 for ROMS model (therefore 1 node for each of them) by typing
> code:
> mpirun -np 2 ./oceanG coupling_inlet-test.in>mpi.log 
> it replies:
> application called MPI_Abort(comm=0x84000002, 4) - process 0
> would you please tell me how i can solved this problem.
> cheers
> fereshte
> **Angel**
>  
>  
> _______________________________________________
> discuss mailing list     discuss at mpich.org
> To manage subscription options or unsubscribe:
> https://lists.mpich.org/mailman/listinfo/discuss