[mpich-discuss] Fatal error in PMPI_Comm_rank

Hong-Jun Kim hongjun at mcs.anl.gov
Fri Feb 25 10:28:20 CST 2011


I have the same errors with full path of "mpiexec" and even from serial running as follows.

-------------------------------
[hongjun at login01 algs]$ /home/hongjun/lib/mpich2-1.3.2p1/bin/mpiexec -l -n 2 ./test_parallelmesh
0: Fatal error in PMPI_Comm_rank: Invalid communicator, error stack:
0: PMPI_Comm_rank(109): MPI_Comm_rank(comm=0x0, rank=0x7fff9f927344) failed
0: PMPI_Comm_rank(66).: Invalid communicator
1: Fatal error in PMPI_Comm_rank: Invalid communicator, error stack:
1: PMPI_Comm_rank(109): MPI_Comm_rank(comm=0x0, rank=0x7fffb17b5e74) failed
1: PMPI_Comm_rank(66).: Invalid communicator
rank 0 in job 99  login01.cosmea.mcs.anl.gov_58699   caused collective abort of all ranks
  exit status of rank 0: return code 1 



[hongjun at login01 algs]$ ./test_parallelmesh 
Fatal error in PMPI_Comm_rank: Invalid communicator, error stack:
PMPI_Comm_rank(109): MPI_Comm_rank(comm=0x0, rank=0x7fff418044d4) failed
PMPI_Comm_rank(66).: Invalid communicator
-------------------------------

-----------------------------
Hong-Jun Kim
Post-doc researcher
MCS, Argonne National Laboratory
9700 S. Cass Ave. B240/R2147
Argonne, IL 60439
630-252-4791
hongjun at mcs.anl.gov
-----------------------------

----- Original Message -----
> From: "Rajeev Thakur" <thakur at mcs.anl.gov>
> To: mpich-discuss at mcs.anl.gov
> Sent: Friday, February 25, 2011 10:20:15 AM
> Subject: Re: [mpich-discuss] Fatal error in PMPI_Comm_rank
> Make sure your mpiexec is also from the same place; i.e. look in the
> test_parallelmesh script and give the full path if necessary.
> 
> On Feb 25, 2011, at 10:18 AM, Hong-Jun Kim wrote:
> 
> > Thanks for your reply.
> >
> > However, I am using "mpicxx" to build my program and it is coming
> > from the right place as in the following build command line.
> >
> > -----------------------------
> > /home/hongjun/lib/mpich2-1.3.2p1/bin/mpicxx -pipe -Wall -g -o
> > test_parallelmesh test_parallelmesh-test_parallelmesh.o
> > -L/home/hongjun/lib/CGM/lib -L/home/hongjun/lib/OpenCASCADE6.3.0/lib
> > -L/home/hongjun/lib/mpich2-1.3.2p1/lib -L/usr/lib/gc
> > c/x86_64-redhat-linux/4.1.2
> > -L/usr/lib/gcc/x86_64-redhat-linux/4.1.2/../../../.. /lib64
> > -L/lib/../lib64 -L/usr/lib/../lib64
> > -----------------------------
> >
> > Actually, this program worked for me before, so I have no idea why
> > this problem is coming now.
> >
> > -----------------------------
> > Hong-Jun Kim
> > Post-doc researcher
> > MCS, Argonne National Laboratory
> > 9700 S. Cass Ave. B240/R2147
> > Argonne, IL 60439
> > 630-252-4791
> > hongjun at mcs.anl.gov
> > -----------------------------
> >
> > ----- Original Message -----
> >> From: "Darius Buntinas" <buntinas at mcs.anl.gov>
> >> To: mpich-discuss at mcs.anl.gov
> >> Sent: Thursday, February 24, 2011 9:11:16 PM
> >> Subject: Re: [mpich-discuss] Fatal error in PMPI_Comm_rank
> >> It looks like you might have compiled using an mpi.h of another MPI
> >> implementation. Make sure you use the compiler wrappers (e.g.,
> >> mpicc)
> >> instead of using gcc directly. This will make sure that the correct
> >> header files are used. Also, you should check to make sure the
> >> mpicc
> >> you're using is from the correct MPI implementation too (i.e.,
> >> "which
> >> mpicc").
> >>
> >> -d
> >>
> >> On Feb 24, 2011, at 7:29 PM, Hong-Jun Kim wrote:
> >>
> >>> Hello,
> >>>
> >>> I compiled my program with "mpich2-1.3.2p1" without any problems
> >>> as
> >>> at the end of this e-mail.
> >>> But, if I execute it, I have some errors I can not understand as
> >>> follows.
> >>> Could any body give me any advice about this problem?
> >>> Thanks.
> >>>
> >>> -------------------------------------------
> >>> [hongjun at login01 algs]$ ./test_parallelmesh
> >>> Fatal error in PMPI_Comm_rank: Invalid communicator, error stack:
> >>> PMPI_Comm_rank(109): MPI_Comm_rank(comm=0x0, rank=0x7fffde36da34)
> >>> failed
> >>> PMPI_Comm_rank(66).: Invalid communicator
> >>> [hongjun at login01 algs]$ mpiexec -l -n 2 ./test_parallelmesh
> >>> 0: Fatal error in PMPI_Comm_rank: Invalid communicator, error
> >>> stack:
> >>> 0: PMPI_Comm_rank(109): MPI_Comm_rank(comm=0x0,
> >>> rank=0x7fffee4e1464)
> >>> failed
> >>> 0: PMPI_Comm_rank(66).: Invalid communicator
> >>> 1: Fatal error in PMPI_Comm_rank: Invalid communicator, error
> >>> stack:
> >>> 1: PMPI_Comm_rank(109): MPI_Comm_rank(comm=0x0,
> >>> rank=0x7fff6055a9c4)
> >>> failed
> >>> 1: PMPI_Comm_rank(66).: Invalid communicator
> >>> rank 1 in job 98 login01.cosmea.mcs.anl.gov_58699 caused
> >>> collective
> >>> abort of
> >>> all ranks
> >>> exit status of rank 1: return code 1
> >>> -------------------------------------------
> >>> -----------------------------
> >>> /home/hongjun/lib/mpich2-1.3.2p1/bin/mpicxx -pipe -Wall -g -o
> >>> test_parallelmesh
> >>> test_parallelmesh-test_parallelmesh.o -L/home/hongjun/lib/CGM/lib
> >>> -L/home/hongj
> >>> un/lib/OpenCASCADE6.3.0/lib -L/home/hongjun/lib/mpich2-1.3.2p1/lib
> >>> -L/usr/lib/gc
> >>> c/x86_64-redhat-linux/4.1.2
> >>> -L/usr/lib/gcc/x86_64-redhat-linux/4.1.2/../../../..
> >>> /lib64 -L/lib/../lib64 -L/usr/lib/../lib64
> >>> ../../src/.libs/libMeshKit.a -L/home/
> >>> hongjun/lib/Lasso/lib -L/home/hongjun/lib/MOAB/lib
> >>> -L/home/hongjun/lib/camal.5.1
> >>> .0-Linux64/lib/Linux64 ../../camal/.libs/libMKcamal.a
> >>> -lCMLEdgeMesher -lCMLTet -
> >>> lCMLSweep -lCMLDelTri -lCMLTri -lCMLSurfMapper -lCMLPave
> >>> -lCMLSizer
> >>> -lc -lgcc_s
> >>> -lmpichcxx -lmpich -lrt -lstdc++
> >>> /home/hongjun/lib/mpich2-1.3.2p1/lib/libopa.so
> >>> -lpthread /home/hongjun/lib/mpich2-1.3.2p1/lib/libmpl.so -ldl
> >>> /home/hongjun/lib/
> >>> Lasso/lib/libiRel.a /home/hongjun/lib/CGM/lib/libiGeom.a
> >>> /home/hongjun/lib/MOAB/
> >>> lib/libiMesh.a /home/hongjun/lib/MOAB/lib/libMOAB.a
> >>> /home/hongjun/lib/CGM/lib/li
> >>> bcgm.a /home/hongjun/lib/OpenCASCADE6.3.0/lib/libTKSTL.so
> >>> /home/hongjun/lib/Open
> >>> CASCADE6.3.0/lib/libTKSTEP.so -lm -Wl,--rpath
> >>> -Wl,/home/hongjun/lib/mpich2-1.3.2p1/lib -Wl,--rpath -Wl,/home/h
> >>> ongjun/lib/OpenCASCADE6.3.0/lib -Wl,--rpath
> >>> -Wl,/home/hongjun/lib/mpich2-1.3.2p1
> >>> /lib -Wl,--rpath -Wl,/home/hongjun/lib/OpenCASCADE6.3.0/lib
> >>> -----------------------------
> >>>
> >>>
> >>>
> >>> -----------------------------
> >>> Hong-Jun Kim
> >>> Post-doc researcher
> >>> MCS, Argonne National Laboratory
> >>> 9700 S. Cass Ave. B240/R2147
> >>> Argonne, IL 60439
> >>> 630-252-4791
> >>> hongjun at mcs.anl.gov
> >>> -----------------------------
> >>>
> >>> _______________________________________________
> >>> mpich-discuss mailing list
> >>> mpich-discuss at mcs.anl.gov
> >>> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
> >>
> >> _______________________________________________
> >> mpich-discuss mailing list
> >> mpich-discuss at mcs.anl.gov
> >> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
> > _______________________________________________
> > mpich-discuss mailing list
> > mpich-discuss at mcs.anl.gov
> > https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
> 
> _______________________________________________
> mpich-discuss mailing list
> mpich-discuss at mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss


More information about the mpich-discuss mailing list