Re: MPI implementations in squeeze
On 01/12/09 at 21:44 +0100, Sylvestre Ledru wrote:
> Le mardi 01 décembre 2009 à 21:20 +0100, Lucas Nussbaum a écrit :
> > On 01/12/09 at 11:33 +0100, Manuel Prinz wrote:
> > > Am Montag, den 09.11.2009, 16:47 -0800 schrieb Nicholas Breen:
> > > > * should we start filing wishlist bugs asking packagers not to build against
> > > > MPICH (1) and LAM?
> > >
> > > If noone objects by the end of the week, I will file the bugs then. The
> > > number of affected packages is really low and the severity wishlist, so
> > > I guess we do not need to go through the usual MBF hassle, or do we?
> >
> > So the plan would be to use OpenMPI on all architectures where it is
> > supported, and MPICH2 on the other ones, using mpi-default?
> To me, this would be the best solution.
OK, let's do that. We can always revisit this decision after the squeeze
release, but it's important to have something sane for squeeze.
Sylvestre, can you take the lead on that, as you proposed? The required
steps, I think are:
- change mpi-default to use OpenMPI + MPICH2
- identify the packages that currently use mpi-default need binNMUs or
full source uploads, and file bugs with severity: serious
- identify packages that don't use mpi-default yet, but use one of the
MPI implementations directly, and file bugs:
severity: serious if they use Mpich1 or LAM
severity: normal if they use OpenMPI or MPICH2, asking them to switch
to mpi-default
Thanks
--
| Lucas Nussbaum
| lucas@lucas-nussbaum.net http://www.lucas-nussbaum.net/ |
| jabber: lucas@nussbaum.fr GPG: 1024D/023B3F4F |
Reply to: