We are testing for a specific line when looking for this patch:
test ! -z "`grep 'filename, LT_LAZY_OR_NOW' opal/libltdl/
loaders/dlopen.c`"; then
If this line is different in your dlopen.c, then it doesn't find it
and therefore autogen.sh doesn't patch it.
Did you already patc
> Ok so I ask the mpirun masters how would you do the following:
>
> I submit a job with torque (we use --with-tm) like the following:
>
> nodes=4:ppn=2
>
> My desired outcome is to have 1 mpi process per 2 cpus and use
> threaded blas (or my own OpenMP take your pick)
For the above, -np 4 --bynod
Ok so I ask the mpirun masters how would you do the following:
I submit a job with torque (we use --with-tm) like the following:
nodes=4:ppn=2
My desired outcome is to have 1 mpi process per 2 cpus and use
threaded blas (or my own OpenMP take your pick)
Our cluster has some 4 core machines
The rank 0 received SIGTERM ... I wonder how this happens. Usually, we
don't just send SIGTERM around without a good reason. One of these
reasons might be that we detected a segfault, but then there is some
output.
Is that the complete output you get from your run ?
Thanks,
george.
Josh,
At this moment I´m working in the uncoordinated checkpoint, and
probably I´ll have some tools to collect data from the process and
environment and probably from the application.
About the application I´m considering the possibility to do something
like this (MPI_Checkpoint??).
Leonardo Fia
Not at the moment.
This would be a neat addition to Open MPI if application developers
see a need for it. There are many issues surrounding this type of a
feature (like any feature). Most of them surround what an application
expects and requires from such an API. One such question is whethe
Are there plans to provide an API that would allow a fault tolerant
enabled program to invoke checkpointing directly?
-Wayne
We have the checkpoint/restart working now. Turns out that the BLCR
kernel mods were installed incorrectly.
Thanks for the help.
-Wayne
-Original Message-
From: users-boun...@open-mpi.org [mailto:users-boun...@open-mpi.org] On
Behalf Of Josh Hursey
Sent: Monday, January 28, 2008 6:57 PM
George Bosilca wrote:
Both cases should work just fine. In fact as long as there is only one
execution flow using MPI functions, the user will not face any problems.
I compiled my mpi fortran code using the -mp option to verify that the
mpi code would still run. I get this message when I run
I've solved the problem by adding the flag RTLD_GLOBAL in the call to
dlopen() in function "sys_dl_open (loader_data, filename)"
(opal/libltdl/ltdl.c)
It seems that I need this flag. However when I run autogen.sh, I get the
following:
** Adjusting libltdl for OMPI :-(
++ patching for argz bu
Hi
Sorry to bring this subject up again -
but i have a problem getting xterms
running for all of my processes (for debugging purposes).
There are actually two problem involved:
display, and paths.
my ssh is set up so that X forwarding is allowed,
and, indeed,
ssh nano_00 xterm
opens an xterm fr
11 matches
Mail list logo
