On Feb 20, 2014, at 7:10 AM, Jeff Squyres (jsquyres) wrote:
> For all of these, I'm using the openshmem test suite that is now committed to
> the ompi-svn SVN repo. I don't know if the errors are with the tests or with
> oshmem itself.
>
> 1. I'm running the oshmem test suite at 32 processes
On Feb 20, 2014, at 10:44 AM, "Jeff Squyres (jsquyres)"
wrote:
> Yes, I've added them to my Cisco MTT ini files in the ompi-svn repo.
Err... I meant ompi-tests SVN repo. :-)
> Look in cisco/mtt/usnic/usnic-trunk.ini and usnic-v1.7.ini.
>
> All relevant sections have "oshmem" in them.
--
Yes, I've added them to my Cisco MTT ini files in the ompi-svn repo. Look in
cisco/mtt/usnic/usnic-trunk.ini and usnic-v1.7.ini.
All relevant sections have "oshmem" in them.
Most are copied from the Mellanox examples, but I made a few
tweaks/improvements here and there. I also anticipate adju
Could you send along the relevant mtt .ini sections?
On Feb 20, 2014, at 7:10 AM, Jeff Squyres (jsquyres) wrote:
> For all of these, I'm using the openshmem test suite that is now committed to
> the ompi-svn SVN repo. I don't know if the errors are with the tests or with
> oshmem itself.
>
For all of these, I'm using the openshmem test suite that is now committed to
the ompi-svn SVN repo. I don't know if the errors are with the tests or with
oshmem itself.
1. I'm running the oshmem test suite at 32 processes across 2 16-core servers.
I'm seeing a segv in "examples/shmem_2dheat.
The Fortran programs in the oshmem test suite don't compile because my_pe and
num_pes are already declared in OMPI's shmem.fh.
To be fair, I asked Mellanox to put those declarations in shmem.fh because I
thought it was crazy that all applications would have to declare them.
Apparently, the shme
