Re: [OMPI users] Open-MPI 1.2 and GM

Thanks for the response, I was hoping I'd just messed up something simple. Your advice took care of my issues. On 27/03/07 14:15 -0400, George Bosilca wrote: > Justin, > > There is no GM MTL. Therefore, the first mpirun allow the use of > every available BTL, while the second one don't allow

Re: [OMPI users] Open-MPI 1.2 and GM

Justin, There is no GM MTL. Therefore, the first mpirun allow the use of every available BTL, while the second one don't allow intra-node communications or self. The correct mpirun command line should be: mpirun -np 4 --mca btl gm,self ... george. On Mar 27, 2007, at 12:18 PM, Justin Br

Re: [OMPI users] Issues with Get/Put and IRecv

Well, mpich2 and mvapich2 are working smoothly for my app. mpich2 under gige is also giving ~2X the performance of openmpi during the working cases for openmpi. After the paper deadline, I'll attempt to package up a simple test case and send it to the list. Thanks! -Mike Mike Houston wrote

Re: [OMPI users] [Re: Memory leak in openmpi-1.2?]

Hello Mr. Van der Vlies, We are currently looking into this problem and will send out an email as soon as we recognize something and fix it. Thank you, > Subject: Re: [OMPI users] Memory leak in openmpi-1.2? > Date: Tue, 27 Mar 2007 13:58:15 +0200 > From: Bas van der Vlies > Reply-To: Open MPI

[OMPI users] Open-MPI 1.2 and GM

Having a user who requires some of the features of gfortran in 4.1.2, I recently began building a new image. The issue is that "-mca btl gm" fails while "-mca mtl gm" works. I have not yet done any benchmarking, as I was wondering if the move to mtl is part of the upgrade. Below are the package

Re: [OMPI users] very long linking time with mixed-language libraries

> -Original Message- > From: users-boun...@open-mpi.org [mailto:users-boun...@open-mpi.org] On > Behalf Of Jeff Squyres > > I notice that you are using the "medium" sized F90 bindings. Do > these FAQ entries help? > > http://www.open-mpi.org/faq/?category=mpi-apps#f90-mpi-slow-compiles >

Re: [OMPI users] MPI processes swapping out

I tried the trunk version with "--mca btl tcp,self". Essentially system time changes to idle time, since empty polling is being replaced by blocking (right?). Page faults go to 0 though. It is interesting since you can see what is going on now, with distinct phases of user time and idle time (slee

Re: [OMPI users] very long linking time with mixed-language libraries

I notice that you are using the "medium" sized F90 bindings. Do these FAQ entries help? http://www.open-mpi.org/faq/?category=mpi-apps#f90-mpi-slow-compiles http://www.open-mpi.org/faq/?category=building#f90-bindings-slow-compile On Mar 27, 2007, at 2:21 AM, de Almeida, Valmor F. wrote: H

Re: [OMPI users] Memory leak in openmpi-1.2?

Bas van der Vlies wrote: Hello, We are testing openmpi version 1.2 on Debian etch with openib. Some of our users are using scalapack/blacs that are running for a long time use a lot of mpi_comm functions. we have made a small C example that test if the mpi can handle this situation (see at

[OMPI users] Memory leak in openmpi-1.2?

Hello, We are testing openmpi version 1.2 on Debian etch with openib. Some of our users are using scalapack/blacs that are running for a long time use a lot of mpi_comm functions. we have made a small C example that test if the mpi can handle this situation (see attach file). When we run thi

[OMPI users] very long linking time with mixed-language libraries

Hello, I am using mpic++ to create a program that combines c++ and f90 libraries. The libraries are created with mpic++ and mpif90. OpenMPI-1.2 was built using gcc-4.1.1. (below follows the output of ompi_info. The final linking stage takes quite a long time compared to the creation of the librar