This problem is solved now. After deinstalling all other MPI distributions on
that machine openMPI 1.4.1 now works perfectly well.
Thanks very very much for your advises!!
________________________________________
Von: users-boun...@open-mpi.org [users-boun...@open-mpi.org] im Auftrag von
Eugene Loh [eugene....@oracle.com]
Gesendet: Freitag, 23. April 2010 17:06
An: Open MPI Users
Betreff: Re: [OMPI users] mpirun -np 4 hello_world; on a eight processor shared
memory machine produces wrong output
Great. That suggests mpirun is OMPI, as you suggested it was, and that
the different process ranks are being set up correctly by mpirun.
But can you also confirm that you're using the OMPI mpicc or mpif90?
Invoke the mpicc with a full path name, perhaps -- that is,
/usr/users/pankatz/OPENmpi/bin/mpicc -- or stick a
#ifdef OPEN_MPI
printf("using Open MPI\n");
#else
printf("not using Open MPI\n");
#endif
into your source code.
Pankatz, Klaus wrote:
> Allright, I've ran a mpirun -np 4 env. And I see OMPI_COMM_WORLD_RANK
> 0 to 3. So far so good.
> OMPI_COMM_WORLD_SIZE=4 everytime, I think thats correct.
> OMPI_MCA_mpi_yield_when_idle=0 everytime zero
> OMPI_MCA_orte_app_num=0 everytime zero
>
> Am 23.04.2010 um 14:54 schrieb Terry Dontje:
>
>> Ok can you do an "mpirun -np 4 env" you should
>> seeOMPI_COMM_WORLD_RANK range 0 thru 3. I am curious if you even see
>> OMPI_* env-vars and if you do is this one 0 for all procs?
>>
>> Pankatz, Klaus wrote:
>>
>>>Yeah, I sure that I use the right mpirun.
>>>which mpirun leads to /usr/users/pankatz/OPENmpi/bin/mpirun which is the
>>>right one.
>>>________________________________________
>>>Von: users-boun...@open-mpi.org [users-boun...@open-mpi.org] im Auftrag von
>>>Terry Dontje [terry.don...@oracle.com]
>>>Gesendet: Freitag, 23. April 2010 14:29
>>>An: Open MPI Users
>>>Betreff: Re: [OMPI users] mpirun -np 4 hello_world; on a eight processor
>>>shared memory machine produces wrong output
>>>
>>>This looks like you are using an mpirun or mpiexec from mvapich to run an
>>>executable compiled with OMPI. Can you make sure that you are using the
>>>right mpirun?
>>>
>>>--td
>>>
>>>Pankatz, Klaus wrote:
>>>
>>>Yes, I did that.
>>>
>>>It ist basically the same problem with a Fortran version of this little
>>>program. With that I used the mpif90 command of openMPI.
>>>________________________________________
>>>Von: users-boun...@open-mpi.org<mailto:users-boun...@open-mpi.org>
>>>[users-boun...@open-mpi.org<mailto:users-boun...@open-mpi.org>] im Auftrag
>>>von Reuti [re...@staff.uni-marburg.de<mailto:re...@staff.uni-marburg.de>]
>>>Gesendet: Freitag, 23. April 2010 14:15
>>>An: Open MPI Users
>>>Betreff: Re: [OMPI users] mpirun -np 4 hello_world; on a eight processor
>>>shared memory machine produces wrong output
>>>
>>>Hi,
>>>
>>>Am 23.04.2010 um 14:06 schrieb Pankatz, Klaus:
>>>
>>>
>>>
>>>Hi all,
>>>
>>>there's a problem with openMPI on my machine. When I simply try to run this
>>>little hello_world-program on multiple processors, the output isn't as
>>>expected:
>>>*****
>>>C code:
>>>#include <mpi.h>
>>>#include <stdio.h>
>>>#include <unistd.h>
>>>int main(int argc, char **argv)
>>>{
>>> int size,rank;
>>> char hostname[50];
>>> MPI_Init(&argc,&argv);
>>> MPI_Comm_rank(MPI_COMM_WORLD, &rank); //Who am I?
>>> MPI_Comm_size(MPI_COMM_WORLD, &size); //How many processes?
>>> gethostname (hostname, 50);
>>> printf ("Hello World! I'm number %2d of %2d running on host %s\n",
>>>rank, size, hostname);
>>> MPI_Finalize();
>>> return 0;
>>>}
>>>****
>>>
>>>Command: mpirun -np 4 a.out
>>>
>>>
>>>
>>>the mpirun (better, use: mpiexec) is the one from the Open MPI, and you also
>>>used its version mpicc to compile the program?
>>>
>>>-- Reuti
>>>
>>>
>>>
>>>
>>>Output:
>>>Hello World! I'm number 0 of 1 running on host marvin
>>>Hello World! I'm number 0 of 1 running on host marvin
>>>Hello World! I'm number 0 of 1 running on host marvin
>>>Hello World! I'm number 0 of 1 running on host marvin
>>>
>>>It should be more or less:
>>>Hello World! I'm number 1 of 4 running on host marvin
>>>Hello World! I'm number 2 of 4 running on host marvin
>>>....
>>>
>>>OpenMPI-version 1.4.1 compiled with Lahey Fortran 95 (lf95).
>>>OpenMPI was compiled "out of the box" only changing to the Lahey compiler
>>>with a setenv $FC lf95
>>>
>>>The System: Linux marvin 2.6.27.6-1 #1 SMP Sat Nov 15 20:19:04 CET 2008
>>>x86_64 GNU/Linux
>>>
>>>Compiler: Lahey/Fujitsu Linux64 Fortran Compiler Release L8.10a
>>>
>>>
_______________________________________________
users mailing list
us...@open-mpi.org
http://www.open-mpi.org/mailman/listinfo.cgi/users
