<div dir="ltr"><div>I just had a look into it. The problem is the switch to CollectiveCommunication. Backwards-compatibility is given in all cases, but in the case when FakeMPIHelper::getCommunicator() was given to the constructor. This can indeed be fixed by your patch, so I think its a good idea.<br><br></div>On the other hand, I will modify the tests to use the new constructors. That will make the problem vanish, too (but not the question of FULL backwards compatibility).<br></div><div class="gmail_extra"><br><div class="gmail_quote">On Fri, Oct 3, 2014 at 12:08 PM, Markus Blatt <span dir="ltr"><<a href="mailto:markus@dr-blatt.de" target="_blank">markus@dr-blatt.de</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On Fri, Oct 03, 2014 at 11:54:41AM +0200, Dominic Kempf wrote:<br>
> Actually, that is nothing new at all. Its been that way "forever".<br>
><br>
<br>
</span>That does not make it more intuitive. The problem was revealed by the<br>
new constructors:<br>
<br>
<a href="http://plagula.de:8010/builders/dune-grid%20test/builds/106/steps/check/logs/stdio" target="_blank">http://plagula.de:8010/builders/dune-grid%20test/builds/106/steps/check/logs/stdio</a><br>
<div><div class="h5"><br>
> In the sequential case, one wants to use the default<br>
> CollectiveCommunication (as defined in collectivecommunication.hh). To get<br>
> that default implementation, a template parameter is to be given, that<br>
> ensures that there is no specialization, that actually does something. Yasp<br>
> uses its own name to do that. Any other name other than MPI_Comm<br>
> would do.<br>
> I will have a look into vtktest to check what is wrong there.<br>
><br>
><br>
> On Fri, Oct 3, 2014 at 11:23 AM, Markus Blatt <<a href="mailto:markus@dr-blatt.de">markus@dr-blatt.de</a>> wrote:<br>
><br>
> > New commit, appeared at Fri Oct 3 11:23:45 2014 +0200<br>
> > as part of the following ref changes:<br>
> ><br>
> > branch refs/heads/master updated from a3ff58a -> f6165ab<br>
> ><br>
> > Browsable version:<br>
> > <a href="http://cgit.dune-project.org/repositories/dune-grid/commit/?id=f6165aba501863dce958719ebbc2103ae8a646da" target="_blank">http://cgit.dune-project.org/repositories/dune-grid/commit/?id=f6165aba501863dce958719ebbc2103ae8a646da</a><br>
> ><br>
> > ======================================================================<br>
> ><br>
> > commit f6165aba501863dce958719ebbc2103ae8a646da<br>
> > Author: Markus Blatt <<a href="mailto:markus@dr-blatt.de">markus@dr-blatt.de</a>><br>
> > Date: Fri Oct 3 11:23:02 2014 +0200<br>
> ><br>
> > [yaspgrid,bugfix] Use canonical communicator type for non-MPI case<br>
> ><br>
> > If people write programs the DUNE way then they will end up with<br>
> > Dune::FakeMPIHelper::MPICommunicator as their communicator type.<br>
> > Unfortunately Yaspgrid uses YaspGrid<dim, Coordinates> which is far<br>
> > from being inituitive and broke vtktest. This patch resorts to using<br>
> > the canonical choice, Dune::FakeMPIHelper::MPICommunicator, and fixes<br>
> > the test.<br>
> ><br>
> > dune/grid/yaspgrid.hh | 4 ++--<br>
> > 1 file changed, 2 insertions(+), 2 deletions(-)<br>
> ><br>
> ><br>
> ><br>
> > diff --git a/dune/grid/yaspgrid.hh b/dune/grid/yaspgrid.hh<br>
> > index f54fe6f..832ab7c 100644<br>
> > --- a/dune/grid/yaspgrid.hh<br>
> > +++ b/dune/grid/yaspgrid.hh<br>
> > @@ -89,7 +89,7 @@ namespace Dune {<br>
> > #if HAVE_MPI<br>
> > typedef CollectiveCommunication<MPI_Comm> CCType;<br>
> > #else<br>
> > - typedef CollectiveCommunication<Dune::YaspGrid<dim, Coordinates> ><br>
> > CCType;<br>
> > + typedef CollectiveCommunication<FakeMPIHelper::MPICommunicator><br>
> > CCType;<br>
> > #endif<br>
> ><br>
> > typedef GridTraits<dim, //<br>
> > dimension of the grid<br>
> > @@ -167,7 +167,7 @@ namespace Dune {<br>
> > #ifdef HAVE_MPI<br>
> > typedef CollectiveCommunication<MPI_Comm> CollectiveCommunicationType;<br>
> > #else<br>
> > - typedef CollectiveCommunication<YaspGrid<dim, Coordinates> ><br>
> > CollectiveCommunicationType;<br>
> > + typedef CollectiveCommunication<FakeMPIHelper::MPICommunicator><br>
> > CollectiveCommunicationType;<br>
> > #endif<br>
> ><br>
> > #ifndef DOXYGEN<br>
> ><br>
> > _______________________________________________<br>
> > Dune-Commit mailing list<br>
> > <a href="mailto:Dune-Commit@dune-project.org">Dune-Commit@dune-project.org</a><br>
> > <a href="http://lists.dune-project.org/mailman/listinfo/dune-commit" target="_blank">http://lists.dune-project.org/mailman/listinfo/dune-commit</a><br>
> ><br>
<br>
</div></div>> _______________________________________________<br>
> Dune-devel mailing list<br>
> <a href="mailto:Dune-devel@dune-project.org">Dune-devel@dune-project.org</a><br>
> <a href="http://lists.dune-project.org/mailman/listinfo/dune-devel" target="_blank">http://lists.dune-project.org/mailman/listinfo/dune-devel</a><br>
<span class="HOEnZb"><font color="#888888"><br>
<br>
--<br>
Do you need more support with DUNE or HPC in general?<br>
<br>
Dr. Markus Blatt - HPC-Simulation-Software & Services <a href="http://www.dr-blatt.de" target="_blank">http://www.dr-blatt.de</a><br>
Hans-Bunte-Str. 8-10, 69123 Heidelberg, Germany<br>
Tel.: <a href="tel:%2B49%20%280%29%20160%2097590858" value="+4916097590858">+49 (0) 160 97590858</a><br>
</font></span></blockquote></div><br></div>