dolfin team mailing list archive

["Garth N. Wells" <gnw20@xxxxxxxxx>]

On 16 September 2011 19:27, Johan Hake <johan.hake@xxxxxxxxx> wrote:
> On Friday September 16 2011 09:34:26 Garth N. Wells wrote:
>> On 16 September 2011 17:31, Johan Hake <johan.hake@xxxxxxxxx> wrote:
>> > Nice!
>> >
>> > cpp demo works for me (Btw: the plotting is wrong in the cpp demo), but
>> > python breaks with the following assert:
>> >
>> > [hake@bamse] ../neumann-poisson/python > mpirun -n 2 python demo_neumann-
>> > poisson.py
>> > Process 0: Number of global vertices: 4225
>> > Process 0: Number of global cells: 8192
>> > Process 1: Partitioned mesh, edge cut is 89.
>> > Process 0: Partitioned mesh, edge cut is 89.
>> > python:
>> > /home/hake/bzr/fenics/dolfin/work/dolfin/fem/DofMapBuilder.cpp:236:
>> > static void
>> > dolfin::DofMapBuilder::compute_ownership(dolfin::DofMapBuilder::set&,
>> > dolfin::DofMapBuilder::set&, dolfin::DofMapBuilder::set&, const
>> > dolfin::DofMap&, const dolfin::DofMapBuilder::set&, const dolfin::Mesh&):
>> > Assertion `MPI::sum(_owned_dim) == dofmap.global_dimension()' failed.
>> > [bamse:06853] *** Process received signal ***
>> > [bamse:06853] Signal: Aborted (6)
>> > [bamse:06853] Signal code:  (-6)
>> > python:
>> > /home/hake/bzr/fenics/dolfin/work/dolfin/fem/DofMapBuilder.cpp:236:
>> > static void
>> > dolfin::DofMapBuilder::compute_ownership(dolfin::DofMapBuilder::set&,
>> > dolfin::DofMapBuilder::set&, dolfin::DofMapBuilder::set&, const
>> > dolfin::DofMap&, const dolfin::DofMapBuilder::set&, const dolfin::Mesh&):
>> > Assertion `MPI::sum(_owned_dim) == dofmap.global_dimension()' failed.
>> > [bamse:06854] *** Process received signal ***
>> >
>> > Any suggestions?
>>
>> Have you updated FFC and run instant-clean?
>
> Details...
>
> But it worked fine after your suggestions ;)
>
> How are the global dofs distributed? Are they present on all processes, or are
> they present on just one process?
>

They are 'owned' by process 0.

Garth

> Johan
>
>> Garth
>>
>> > Johan
>> >
>> > On Friday September 16 2011 05:12:14 noreply@xxxxxxxxxxxxx wrote:
>> >> ------------------------------------------------------------
>> >> revno: 6201
>> >> committer: Garth N. Wells <gnw20@xxxxxxxxx>
>> >> branch nick: dolfin-wells
>> >> timestamp: Fri 2011-09-16 12:43:32 +0100
>> >> message:
>> >>   Add support for global dofs (e.g. Lagrange multipliers) in parallel.
>> >> modified:
>> >>   demo/undocumented/neumann-poisson/cpp/main.cpp
>> >>   demo/undocumented/neumann-poisson/python/demo_neumann-poisson.py
>> >>   dolfin/fem/DofMapBuilder.cpp
>> >>   dolfin/fem/DofMapBuilder.h
>> >>   test/regression/test.py
>> >>
>> >>
>> >> --
>> >> lp:dolfin
>> >> https://code.launchpad.net/~dolfin-core/dolfin/main
>> >>
>> >> Your team DOLFIN Core Team is subscribed to branch lp:dolfin.
>> >> To unsubscribe from this branch go to
>> >> https://code.launchpad.net/~dolfin-core/dolfin/main/+edit-subscription
>> >
>> > _______________________________________________
>> > Mailing list: https://launchpad.net/~dolfin
>> > Post to     : dolfin@xxxxxxxxxxxxxxxxxxx
>> > Unsubscribe : https://launchpad.net/~dolfin
>> > More help   : https://help.launchpad.net/ListHelp
>