seemingly. here is the output of ring: bash-3.2$ ~/lib/openmpi-1.3.3/MAC/bin/mpicxx ring_cxx.cc bash-3.2$ ~/lib/openmpi-1.3.3/MAC/bin/mpirun -np 2 a.out Process 0 sending 10 to 1, tag 201 (2 processes in ring) Process 0 sent to 1 Process 0 decremented value: 9 Process 0 decremented value: 8 Process 0 decremented value: 7 Process 0 decremented value: 6 Process 0 decremented value: 5 Process 0 decremented value: 4 Process 0 decremented value: 3 Process 0 decremented value: 2 Process 0 decremented value: 1 Process 0 decremented value: 0 Process 0 exiting Process 1 exiting
and here is the output of hello: bash-3.2$ ~/lib/openmpi-1.3.3/MAC/bin/mpicc hello_c.c bash-3.2$ ~/lib/openmpi-1.3.3/MAC/bin/mpirun -np 2 hello_ bash-3.2$ ~/lib/openmpi-1.3.3/MAC/bin/mpirun -np 2 a.out Hello, world, I am 0 of 2 Hello, world, I am 1 of 2 I presume this output is correct? I guess the issue I have lies elsewhere then? Andrew On 09 Dec 2009, at 5:44 PM, Jeff Squyres wrote: > Can you run simple MPI applications, like sending a message around in a ring? > > On Dec 9, 2009, at 10:18 AM, Andrew McBride wrote: > >> Thanks for your quick response Ralph. >> >> The errors I get is now are of a completely different nature and have to do >> with, presumably, calling delete on an unallocated pointer. Now, this >> probably has little to do with openmpi and more to do with compilers used to >> create openmpi? >> >> I used gcc version 4.5.0 20090910 when compiling openmpi. >> >> Does anyone have any ideas? >> >> Regards >> Andrew >> >> Start testing: Dec 09 15:53 SAST >> ---------------------------------------------------------- >> 1/534 Testing: Teuchos_BLAS_test_MPI_1 >> 1/534 Test: Teuchos_BLAS_test_MPI_1 >> Command: "/Users/andrewmcbride/lib/openmpi-1.3.3/MAC/bin/mpiexec" "-np" "1" >> "/Users/andrewmcbride/lib/trilinos-10.0.2-Source/MAC_SL/packages/teuchos/test/BLAS/Teuchos_BLAS_test.exe" >> "-v" >> Directory: >> /Users/andrewmcbride/lib/trilinos-10.0.2-Source/MAC_SL/packages/teuchos/test/BLAS >> "Teuchos_BLAS_test_MPI_1" start time: Dec 09 15:53 SAST >> Output: >> ---------------------------------------------------------- >> Teuchos_BLAS_test.exe(72504) malloc: *** error for object 0x100727c00: >> pointer being freed was not allocated >> *** set a breakpoint in malloc_error_break to debug >> [macs-mac:72504] *** Process received signal *** >> [macs-mac:72504] Signal: Abort trap (6) >> >> On 09 Dec 2009, at 3:32 PM, Ralph Castain wrote: >> >>> You need to set your LD_LIBRARY_PATH to ~/lib/openmpi-1.3.3/MAC/lib, and >>> your PATH to ~/lib/openmpi-1.3.3/MAC/bin >>> >>> It should then run fine. >>> >>> On Wed, Dec 9, 2009 at 6:29 AM, Andrew McBride <andrew.mcbr...@uct.ac.za> >>> wrote: >>> Hi >>> >>> I've installed trilinos using the openmpi 1.3.3 libraries. I'm configuring >>> openmpi as follows: >>> /configure CXX=/usr/local/bin/g++ CC=/usr/local/bin/gcc >>> F77=/usr/local/bin/gfortran - >>> prefix=/Users/andrewmcbride/lib/openmpi-1.3.3/MAC >>> >>> Trilinos compiles without problem but the test fail (see below). I'm >>> running a Mac with OSX10.6 (snow leopard). The mpi tests seem to run fine: >>> >>> bash-3.2$ ~/lib/openmpi-1.3.3/MAC/bin/mpicc hello_c.c >>> bash-3.2$ ~/lib/openmpi-1.3.3/MAC/bin/mpirun -np 2 hello_ >>> bash-3.2$ ~/lib/openmpi-1.3.3/MAC/bin/mpirun -np 2 a.out >>> Hello, world, I am 0 of 2 >>> Hello, world, I am 1 of 2 >>> >>> I'm convinced that the problem has to do with the paths and different >>> versions of mpi lurking on the mac. I don't want to use the version of >>> openmpi that comes bundled with the mac for a different reason. >>> >>> Any help would be most appreciated >>> >>> Andrew >>> >>> >>> Start testing: Dec 09 12:18 SAST >>> ---------------------------------------------------------- >>> 1/534 Testing: Teuchos_BLAS_test_MPI_1 >>> 1/534 Test: Teuchos_BLAS_test_MPI_1 >>> Command: "/Users/andrewmcbride/lib/openmpi-1.3.3/MAC/bin/mpiexec" "-np" "1" >>> "/Users/andrewmcbride/lib/trilinos-10.0.2-Source/MAC_SL/packages/teuchos/test/BLAS/Teuchos_BLAS_test.exe" >>> "-v" >>> Directory: >>> /Users/andrewmcbride/lib/trilinos-10.0.2-Source/MAC_SL/packages/teuchos/test/BLAS >>> "Teuchos_BLAS_test_MPI_1" start time: Dec 09 12:18 SAST >>> Output: >>> ---------------------------------------------------------- >>> [macs-mac.local:71058] [[INVALID],INVALID] ORTE_ERROR_LOG: Not found in >>> file runtime/orte_init.c at line 125 >>> -------------------------------------------------------------------------- >>> It looks like orte_init failed for some reason; your parallel process is >>> likely to abort. There are many reasons that a parallel process can >>> fail during orte_init; some of which are due to configuration or >>> environment problems. This failure appears to be an internal failure; >>> here's some additional information (which may only be relevant to an >>> Open MPI developer): >>> >>> orte_ess_base_select failed >>> --> Returned value Not found (-13) instead of ORTE_SUCCESS >>> -------------------------------------------------------------------------- >>> *** An error occurred in MPI_Init >>> *** before MPI was initialized >>> *** MPI_ERRORS_ARE_FATAL (your MPI job will now abort) >>> [macs-mac.local:71058] Abort before MPI_INIT completed successfully; not >>> able to guarantee that all other processes were killed! >>> -------------------------------------------------------------------------- >>> >>> >>> _______________________________________________ >>> users mailing list >>> us...@open-mpi.org >>> http://www.open-mpi.org/mailman/listinfo.cgi/users >>> >>> _______________________________________________ >>> users mailing list >>> us...@open-mpi.org >>> http://www.open-mpi.org/mailman/listinfo.cgi/users >> >> _______________________________________________ >> users mailing list >> us...@open-mpi.org >> http://www.open-mpi.org/mailman/listinfo.cgi/users > > > -- > Jeff Squyres > jsquy...@cisco.com > > > _______________________________________________ > users mailing list > us...@open-mpi.org > http://www.open-mpi.org/mailman/listinfo.cgi/users
