Good News !
I have just checked this issue
with the NEW LAM (beta) realease 7.0:
everything is fine now !
Thanks
Jerome BENOIT wrote:
> Hello LAM,
>
> I have just tried LAM 5.6.9 patched with `the' path
> which fixes a bug in `tkill.c' encountred on dikless computer
> ( http://www.lam-mpi.org/MailArchives/lam/msg05820.php )
> with a hello-like program `phello' (see attachement).
>
> the following procedure works fine:
> 1] lamboot ; lamhalt ;
>
> This one
> 2] lamboot ; mpirun -n 10 phello; lamhalt;
> works when the number of processor is greater than one,
> oterwise I get the message:
> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
> lam_deregister_object (name):
> Warning: No such file or directory: deleting shared memory id's from
> registry
> <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
>
> but if I launch phello twice, LAM stucks the second time:
> lamboot ; mpirun -n 10 phello; mpirun -n 10 phello; .......
>
> If I do `lamclean' between the `mpirun' calls,
> evrything works (when the number of processor is greater than one,
> I get the same error message as before).
>
> I hope that helps,
> Jerome
>
>
> ------------------------------------------------------------------------
>
> // `phello.c' MPI C file
> //
> // last major modification 2003/01/05
> // last minor modification 2003/05/03
> //
>
> // mpicc -o phello phello.c
> // mpirun -np 5 phello
>
> #include <stdio.h>
> #include <mpi.h>
>
> int main(int narg, char *args[]){
> int rank,size;
> char ProcessorName[MPI_MAX_PROCESSOR_NAME];
> int ProcessorNameLength;
>
> MPI_Init(&narg,&args);
>
> MPI_Comm_rank(MPI_COMM_WORLD,&rank);
> MPI_Comm_size(MPI_COMM_WORLD,&size);
>
> // printf("Hello world! I am %d of %d\n",rank,size);
> MPI_Get_processor_name(ProcessorName,&ProcessorNameLength);
> printf("<%s>\n",ProcessorName);
>
> MPI_Finalize();
>
> return 0; }
>
> //
> // End of file `phello.c'.
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
|
