Re: Status of gfortran transition
Greetings, and thanks so much for your helpful reply, consisten labor
for Debian, and thoughtful attention to these matters!
Matthias Klose <doko@cs.tu-berlin.de> writes:
> Changing the soname to a version which might be used by upstream for
> other purposes doesn't make sense. Please read our proposed transition
> plan (that resembles all other transitions which introduce changes not
> caused by the package itself):
>
> - do NOT change the soname
> - rename refblas3 to libblas3
> - libblas3 conflicts/replaces refblas3.
>
> This way you cannot have packages (archive or third party) installed
> together.
OK, there is a problem here. The existing packages that depend on the
blas interface have the following runtime dependency:
atlas3-base | refblas3 | libblas.so.3
(defined in the shlibs.local) The latter is a virtual package.
lapack3, for example, acquires this dependency via linking in -lblas at
the last step. So installing libblas3 will replace refblas3, but make
anything linking against it or lapack3 segfault :-(. If the runtime
is made to conflict/replace, then we need to conflict/replace on
everything linked to it via some equivalent mechanism, I'd think, but
I'm sure I'm missing something here.
This just bit me, as I was building lapack, which requires a blas lib
dev, and it was using an atlas version, which was build against g77.
These are runtime incompatible, and that is what it appears sonames
are for.
I agree it might not be aesthetically pleasing to have an soname
differing from upstream, but I cannot see any other reason not to do
so. In fact, blas has no version number, so we make up an soname. It
would be ideal IMHO if the soname could be set to 3gf, but I'm not
getting too hopeful here (please tell me I'm overly pessimistic :-().
Take care,
>
> Matthias
>
> Camm Maguire writes:
> > Greetings!
> >
> > OK, just found out (should have known this before) that gfortran
> > compiled liblapack (for example) will segfault when run against g77
> > compiled blas. It seems that an soname bump is called for here. If
> > so, is the gf appendage needed?
> >
> > Would this not work:
> >
> > Real packages:
> >
> > refblas3 -> libblas4 et.al.
> > lapack3 -> liblapack4 et.al.
> >
> > Virtual packages:
> >
> > libblas.so.3 -> libblas.so.4, or libblas4 if that is legal
> > libblas-3.so -> libblas-4.so, or libblas4-dev if that works
> >
> > Then only conflict/replace on the -dev packages, allowing older g77
> > runtime to coexist with newer gfortran. In this case, I don't see why
> > we can't just upload to unstable.
> >
> > Please object if this is wrong in any way -- it appears best to me at
> > the moment so I'll try heading in that direction.
> >
> > Take care,
> >
> > Matthias Klose <doko@cs.tu-berlin.de> writes:
> >
> > > Kumar Appaiah writes:
> > > > On 17/10/2007, Kamaraju S Kusumanchi wrote:
> > > > > I think the main hold up is due to the atlas and lapack libraries. They
> > > > > still depend upon g77. Would anyone be interested to NMU them or should the
> > > > > NMUs be done only to fix RC bugs?
> > >
> > > If you do prepare NMUs, please upload those to experimental first; we
> > > didn't hear back from Camm now for some weeks, so the best thing would
> > > be to go ahead. As nearly all other stuff depends on refblas, lapack
> > > and atlas, these should be done before uploading anything to unstable.
> > >
> > > > I think we should move the discussion to debian-devel. I am mailing
> > > > there, will you also please follow the thread?
> > >
> > > It would be nice to keep the discussion in one place. some people from
> > > the gcc fortran list might listen here, and it's quiet enough that
> > > they can follow here (which is not the case for debian-devel).
> > >
> > > Matthias
> > >
> > > PS: Riku, didn't you start with the atlas packages?
> > >
> > >
> > >
> > >
> > >
> >
> > --
> > Camm Maguire camm@enhanced.com
> > ==========================================================================
> > "The earth is but one country, and mankind its citizens." -- Baha'u'llah
>
>
>
--
Camm Maguire camm@enhanced.com
==========================================================================
"The earth is but one country, and mankind its citizens." -- Baha'u'llah
Reply to: