Re: [OMPI users] Error with multiple MPI runs inside one Slurm allocation (with QLogic PSM)

2012-05-02 Thread Ralph Castain
Okay, your tests confirmed my suspicions. Slurm isn't doing any binding at all - that's why your srun of get-allowed-cpu-ompi showed no bindings. I don't see anything in your cmds that would tell slurm to bind us to anything. All your salloc did was tell slurm what to allocate - that doesn't imp

Re: [OMPI users] MPI with RDMA CM

2012-05-02 Thread Jeff Squyres
On May 2, 2012, at 6:32 AM, Shanthini Kannan wrote: > I am new to OpenMPI. > Do I definitely need IPoIB for OpenMPI to work? No. > If I include RDMA CM to connect my nodes, isn't that sufficient? It's been quite a while, but IIRC, RDMA CM requires IPoIB. FWIW: Open MPI actually defaults to usi

Re: [OMPI users] Error with multiple MPI runs inside one Slurm allocation (with QLogic PSM)

2012-05-02 Thread Rémi Palancher
On Fri, 27 Apr 2012 08:56:15 -0600, Ralph Castain wrote: Couple of things: 1. please do send the output from ompi_info You can find them attached to this email. 2. please send the slurm envars from your allocation - i.e., after you do your salloc. Here is an example: $ salloc -N 2 -n 20 -

[OMPI users] MPI with RDMA CM

2012-05-02 Thread Shanthini Kannan
Hello, I am new to OpenMPI. Do I definitely need IPoIB for OpenMPI to work? If I include RDMA CM to connect my nodes, isn't that sufficient? Any help will be greatly appreciated. Thanks! Shan

Re: [OMPI users] Segmentation fault on master task

2012-05-02 Thread Eduardo Morras
At 08:51 02/05/2012, you wrote: Hi, I am trying to execute following code on cluster. run_kernel is a cuda call with signature int run_kernel(int array[],int nelements, int taskid, char hostname[]); ... deleted code mysum = run_kernel(&onearray[2000], chunksize, taskid, myname); ...

[OMPI users] Segmentation fault on master task

2012-05-02 Thread Rohan Deshpande
Hi, I am trying to execute following code on cluster. run_kernel is a cuda call with signature int run_kernel(int array[],int nelements, int taskid, char hostname[]); My MPI code is #include "mpi.h" #include #include #include #define MASTER0 int *masterarray; int *onearray; int