Re: Open MPI 1.3 transition
On Sun, 31 May 2009, Manuel Prinz wrote:
> On Sat, May 30, 2009 5:29 pm, Adeodato Simó wrote:
> > Okay, thanks. I went ahead and scheduled the required Bin-NMUs.
>
> Thanks!
>
> > Unfortunately, most of them failed, seemingly due to an error in
> > alternatives handling:
> >
> > https://buildd.debian.org/~luk/status/package.php?p=arpack,blacs-mpi,boost1.37,boost1.38,gmsh,gpivtools,gromacs,hdf5,hypre,illuminator,libgpiv,libmesh,music,openmpi,paraview,parmetis,petsc,rmpi,scalapack,slepc,spooles
> >
> > Then there's a weird failure of boost1.38/amd64 with mpic++ segfaulting,
> > and I didn't look at more stuff.
> >
> > Could you check what's wrong with the packaging?
>
> I will have a look at it ASAP. Unfortunately, it's not possible for me to
> fix it before Tuesday. I'll upload a fixed version then.
>
> Raphael's suggestion is probably the solution to the problem. I already
> suspected something like that. Interestingly, it did not show up when I
> rebuild the packages (with pbuilder).
That's expected, the bad alternatives are cruft left-over from old
upgrades... at one point in the past they were real master alternatives
and then they got enslaved with another one without being properly removed
before-hand.
I don't know the history of the package so I have no idea which
alternatives are concerned. If many packages are affected, it might make
sense to factorize the cleanup code in dpkg's preinst directly.
Cheers,
--
Raphaël Hertzog
Contribuez à Debian et gagnez un cahier de l'admin Debian Lenny :
http://www.ouaza.com/wp/2009/03/02/contribuer-a-debian-gagner-un-livre/
Reply to: