This indicates that your LAM installation does not have the gm RPI
module built (this should also be evident if you run "laminfo" -- you
will not see the gm rpi module listed).
Specifically, you told the test suite to run with the gm module, but
that module does not exist in your installation, so it aborted.
If you want to build LAM with gm support, you must use the --with-gm
argument to LAM's configure to tell it where the gm header files and
libraries are. For example:
./configure --with-gm=/opt/gm .....
replacing /opt/gm with whatever the relevant path is on your system, of
course. If you look through the output of configure, you'll see when
the gm module configures itself. It will announce whether it intends
to build or not -- and if it decides not to build, it should show the
reason why (e.g., it could not find gm.h).
Hope that helps.
On Nov 28, 2005, at 7:43 PM, yicheng wrote:
> Hi Forks,
>
> After I installed LAM-7.0.4 on my cluster, I run “make –k check
> MODELS=”gm” ” command from lam test package, and I got bunches of the
> following errors, I don’t know if it’s ok, can someone help me out?
>
> Making check in reporting
> make[1]: Entering directory
> `/home/sun/Desktop/lamtests-7.0.4/reporting'
> make[1]: Nothing to be done for `check'.
> make[1]: Leaving directory `/home/sun/Desktop/lamtests-7.0.4/reporting'
> Making check in ccl
> make[1]: Entering directory `/home/sun/Desktop/lamtests-7.0.4/ccl'
> make check-TESTS
> make[2]: Entering directory `/home/sun/Desktop/lamtests-7.0.4/ccl'
> mpirun -x TEST -s h C -ssi rpi gm
> /home/sun/Desktop/lamtests-7.0.4/ccl/./allgath
> er
> -----------------------------------------------------------------------
> ------
> The rpi module named "gm" could not be found.
>
> This typically means that you misspelled the desired module name, used
> the wrong name entirely, or the module has decided that it does not
> want to run in this environment.
> -----------------------------------------------------------------------
> ------
> -----------------------------------------------------------------------
> ------
> The rpi module named "gm" could not be found.
>
> This typically means that you misspelled the desired module name, used
> the wrong name entirely, or the module has decided that it does not
> want to run in this environment.
> -----------------------------------------------------------------------
> ------
> -----------------------------------------------------------------------
> ------
> The RPI SSI subsystem failed to open during MPI_INIT. This is a fatal
> error; I must abort.
>
> This occurred on host cluster.hpc.ric.org (n0).
> The PID of failed process was 15341 (MPI_COMM_WORLD rank: 0)
> -----------------------------------------------------------------------
> ------
> -----------------------------------------------------------------------
> ------
> The RPI SSI subsystem failed to open during MPI_INIT. This is a fatal
> error; I must abort.
>
> This occurred on host cluster.hpc.ric.org (n0).
> The PID of failed process was 15342 (MPI_COMM_WORLD rank: 1)
> -----------------------------------------------------------------------
> ------
> -----------------------------------------------------------------------
> ------
> It seems that [at least] one of the processes that was started with
> mpirun did not invoke MPI_INIT before quitting (it is possible that
> more than one process did not invoke MPI_INIT -- mpirun was only
> notified of the first one, which was on node n4236525).
>
> mpirun can *only* be used with MPI programs (i.e., programs that
> invoke MPI_INIT and MPI_FINALIZE). You can use the "lamexec" program
> to run non-MPI programs over the lambooted nodes.
> -----------------------------------------------------------------------
> ------
> PASS: allgather
> mpirun -x TEST -s h C -ssi rpi gm
> /home/sun/Desktop/lamtests-7.0.4/ccl/./allredu
> ce
> -----------------------------------------------------------------------
> ------
> The rpi module named "gm" could not be found.
>
> This typically means that you misspelled the desired module name, used
> the wrong name entirely, or the module has decided that it does not
> want to run in this environment.
> -----------------------------------------------------------------------
> ------
> -----------------------------------------------------------------------
> ------
> The RPI SSI subsystem failed to open during MPI_INIT. This is a fatal
> error; I must abort.
>
> This occurred on host cluster.hpc.ric.org (n0).
>
>
> Thanks!
> Yicheng
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
--
{+} Jeff Squyres
{+} The Open MPI Project
{+} http://www.open-mpi.org/
|
