[dune-pdelab] dune-pdelab release 2.0: problems with dnaplfv example from howto

[Rebecca de Cuveland]

Hey everyone,
thanks to Christian, Jö and Steffen I was able to solve the first problem.
My parmetis version was too old, I now installed 4.0.3 and with that 
dnaplfv compiles again.
If the code in dune-istl stays like that we should probably add a 
comment somewhere about the required version of parmetis.

Nevertheless the code now compiles but I still get a segfault when 
trying to run it.
Bye,
Rebecca

-- 
Rebecca de Cuveland
Universität Heidelberg
Interdisziplinäres Zentrum für Wissenschaftliches Rechnen
Im Neuenheimer Feld 368, 69120 Heidelberg
Tel: 06221/548890   Fax: 06221/548884
e-mail: <rebecca.cuveland at iwr.uni-heidelberg.de>

On 16.07.2014 11:09, Rebecca de Cuveland wrote:
> Hey PDELab!
> It´s been almost one year since I last worked with dune and pdelab but 
> finally I am back :-)
>
> As a first step I tried to install the new pdelab version 2.0.
> I followed the instructions on http://www.dune-project.org/pdelab/
> and used the dune core modules with tag v2.3.1.
> Then there is dune-typetree with v2.3.1 and dune-pdelab and 
> dune-pdelab-howto with v2.0.0
>
> My dunecontrol.opts file reads:
> CONFIGURE_FLAGS="--with-hdf5=/home/rebecca/Software/hdf5 
> --enable-parallel CXXFLAGS='-Wall -g -O0 -fno-strict-aliasing'"
>
> Everything runs through until the example:
> dune-pdelab-howto/src/two-phase/dnalplfv.cc
>
> There I get the following error message:
> dnaplfv.cc:646: instantiated from here
> /home/rebecca/dune2014_pdelab_2.0/dune-istl/dune/istl/repartition.hh:1357: 
> error: cannot convert ‘Dune::idxtype*’ to ‘idxtype*’ for argument ‘2’ 
> to ‘void ParMETIS_V3_PartKway(idxtype*, idxtype*, idxtype*, idxtype*, 
> idxtype*, int*, int*, int*, int*, float*, float*, int*, int*, 
> idxtype*, ompi_communicator_t**)’
>
> There seems to be some problem with the parallel AMG version of v2.3.1 
> or with the setup of the code.
> Does anybody know what the problem is? I heard there were problems 
> with v2.3.0 but that they should be resolved with v2.3.1.
>
> When I don´t use the AMG solver, but the commented 
> ISTLBackend_OVLP_BCGS_SSORk the compilation runs through, but when I 
> try to run the program with
> ./dnaplfv 1 200 20
> I immediatelty get a segmentation fault.
>
> I think there´s something wrong with the adaption to the new backend, 
> but I don´t know what.
> I would really appreciate it, if someone took the time to help me, as 
> this example is quite similar to my code and I would really like to 
> use the new pdelab/dune.
>
> By the way, I also tried the most recent version of pdelab and the 
> core modules (from about one week ago), there I had the same problem.
> I also found some small bugs in the new pdelab and will be happy to 
> contribute again if I get my code up and running :-)
>
> Thanks for your time and help,
> Rebecca
>
>
>