[Octopus-devel] [Octopus-users] push pop subs

[David Strubbe]

The ultimate technique is to make pop_sub tell the name of the routine, and
then check that the right one is being popped. This however will take a lot
of effort to add everywhere...

On Tue, Mar 23, 2010 at 3:46 AM, matthieu verstraete <
Matthieu.Verstraete at ulg.ac.be> wrote:

> Hi David,
>
> ok, I have already run it on my copy of 3.2.0, but didn't know just
> anyone could re-commit to it. However I still have some missing
> push_subs which go undetected and stop my debug run. Do you have a
> technique to find them?
>
> Matthieu
>
> On Tue, Mar 23, 2010 at 1:17 AM, David Strubbe <dstrubbe at berkeley.edu>
> wrote:
> > Matthieu,
> >
> > That's really helpful that you made the script. You can actually make
> > whatever changes you think are needed to 3.2.x by checking out that
> branch
> > from the SVN repository, making the alterations, and committing, just as
> > with the trunk. I think the branch is automatically built into the
> tarball
> > nightly or something.
> >
> > David
> >
> > On Mon, Mar 22, 2010 at 4:08 AM, matthieu verstraete
> > <Matthieu.Verstraete at ulg.ac.be> wrote:
> >>
> >> Hello all,
> >>
> >> I am having some problems with 3.2.0 in parallel on a cvos amd cluster
> >> (ifort 10.0 compiled with openmpi)
> >>
> >> 1) I get a bunch of
> >> Block "ParallelizationGroupRanks" already defined
> >> messages and then the code segfaults (could be unrelated to the above)
> >>
> >> 2) When I try debug mode it halts
> >>
> >> * O    233.225631            0.000926  783936 |
> >> gauge_field.gauge_field_init
> >>  Too few recursion levels
> >>
> >> Could someone backport the push_sub pop_subs which have been corrected
> >> in the svn version to the 3.2.0.tar.gz version as well?
> >>
> >> I have made a little script to find the unpaired push pops in the
> >> src/* tree (attached)
> >> Copied to and run from the octopus root, it reports offending
> >> file+routine name. I have eliminated all false positives for now,
> >> though there could be some undetected errors, certainly. It would be
> >> nice if this were run systematically on the trunk to keep the debug
> >> mode functional (a buildbot step perhaps).
> >>
> >> ciao ciao
> >>
> >> Matthieu
> >>
> >>
> >>
> >> --
> >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> >> Prof. Matthieu Verstraete
> >>
> >> Universite de Liège
> >> Institut de Physique, Bat. B5, 3/7
> >> Allée du 6 aout, 17
> >> B- 4000 Sart Tilman, Liège
> >> Belgium
> >>
> >> Associate member European Theoretical Spectroscopy Facility (ETSF
> >> http://www.etsf.eu)
> >>
> >> Phone : +32 4 366 90 17
> >> Fax   : +32 4 366 36 29
> >>
> >> Mail : matthieu.verstraete at ulg.ac.be
> >>          matthieu.jean.verstraete at gmail.com
> >>
> >> _______________________________________________
> >> Octopus-users mailing list
> >> Octopus-users at tddft.org
> >> http://www.tddft.org/mailman/listinfo/octopus-users
> >>
> >
> >
>
>
>
> --
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Prof. Matthieu Verstraete
>
> Universite de Liège
> Institut de Physique, Bat. B5, 3/7
> Allée du 6 aout, 17
> B- 4000 Sart Tilman, Liège
> Belgium
>
> Associate member European Theoretical Spectroscopy Facility (ETSF
> http://www.etsf.eu)
>
> Phone : +32 4 366 90 17
> Fax   : +32 4 366 36 29
>
> Mail : matthieu.verstraete at ulg.ac.be
>          matthieu.jean.verstraete at gmail.com
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.tddft.org/pipermail/octopus-devel/attachments/20100323/ba955848/attachment.html