Johannes Ring wrote:
> On Tue, December 30, 2008 10:44, Anders Logg wrote:
>> On Tue, Dec 30, 2008 at 10:29:10AM +0100, Ilmar Wilbers wrote:
>>> Anders Logg wrote:
>>>> On Tue, Dec 30, 2008 at 10:24:52AM +0100, Ilmar Wilbers wrote:
>>>>> The buildbots have these warnings all the time. But they are all
>>> green
>>>>> now, by the way.
>>>>>
>>>>> ilmar
>>>> Not linux_64. It looks red from here.
>>>>
>>> Technically yellow :-)
>>>
>>> There was a problem this night, the only step that failed was
>>> instant-clean, but this was because my home folder was unavailable, some
>>> machine must have been down. All the compiling and running steps were
>>> successfull.
>>>
>>> ilmar
>> Looks green now! :-)
> 
> Even though the buildbots are all green, I still get the error below when
> PETSc is enabled. If the solution is to require PETSc to be configured
> with the option '--with-clanguage=C++', then the PETSc package in Debian
> and Ubuntu will be useless. This means that I have to build Debian/Ubuntu
> packages for DOLFIN without support for PETSc.
>

We can revert to using C-only PETSc functions.

Garth

> Johannes
> 
> mpic++ -o dolfin/la/PETScVector.os -c -Wall -pipe -ansi -Werror -DDEBUG -g
> -DNDEBUG -O2 -DHAS_MPI=1 -DMPICH_IGNORE_CXX_SEEK
> -DPACKAGE_VERSION=\"0.8.1\" -DHAS_PETSC=1 -DHAS_SLEPC=1 -DHAS_SCOTCH=1
> -DHAS_UMFPACK=1 -DHAS_GTS=1 -DHAS_TRILINOS=1 -DHAS_CHOLMOD=1 -DHAS_MTL4=1
> -DHAS_PARMETIS=1 -fPIC -I. -I/opt/simula/include -I/usr/include
> -I/usr/include/libxml2 -I/usr/lib/petsc/bmake/linux-gnu-c-opt
> -I/usr/lib/petsc/include -I/usr/include -I/usr/lib/openmpi/include
> -I/usr/lib/openmpi/lib -Idolfin -I/usr/local/src/slepc-2.3.3
> -I/usr/local/src/slepc-2.3.3/include
> -I/usr/lib/petsc/bmake/linux-gnu-c-opt -I/usr/lib/petsc/include
> -I/usr/include -I/usr/lib/openmpi/include -I/usr/lib/openmpi/lib -Idolfin
> -I/usr/local/src/scotch_5.1/include -I/usr/include/suitesparse
> -I/usr/include -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include
> -I/opt/local/include -I/usr/include/suitesparse -I/usr/local/src/mtl4
> -I/usr/include dolfin/la/PETScVector.cpp
> dolfin/la/PETScVector.cpp: In member function 'virtual void
> dolfin::PETScVector::resize(dolfin::uint)':
> dolfin/la/PETScVector.cpp:82: error: cannot convert '_p_Vec**' to
> 'ompi_communicator_t*' for argument '1' to 'PetscErrorCode
> VecCreate(ompi_communicator_t*, _p_Vec**)'
> /usr/lib/petsc/include/petscvec.h: In member function 'virtual
> dolfin::uint dolfin::PETScVector::size() const':
> /usr/lib/petsc/include/petscvec.h:356: error: too few arguments to
> function 'PetscErrorCode VecGetSize(_p_Vec*, PetscInt*)'
> dolfin/la/PETScVector.cpp:174: error: at this point in file
> /usr/lib/petsc/include/petscvec.h: In member function 'virtual double
> dolfin::PETScVector::inner(const dolfin::GenericVector&) const':
> /usr/lib/petsc/include/petscvec.h:80: error: too few arguments to function
> 'PetscErrorCode VecDot(_p_Vec*, _p_Vec*, PetscScalar*)'
> dolfin/la/PETScVector.cpp:241: error: at this point in file
> /usr/lib/petsc/include/petscvec.h: In member function 'virtual double
> dolfin::PETScVector::norm(dolfin::NormType) const':
> /usr/lib/petsc/include/petscvec.h:155: error: too few arguments to
> function 'PetscErrorCode VecNorm(_p_Vec*, NormType, PetscReal*)'
> dolfin/la/PETScVector.cpp:263: error: at this point in file
> /usr/lib/petsc/include/petscvec.h:155: error: too few arguments to
> function 'PetscErrorCode VecNorm(_p_Vec*, NormType, PetscReal*)'
> dolfin/la/PETScVector.cpp:265: error: at this point in file
> /usr/lib/petsc/include/petscvec.h:155: error: too few arguments to
> function 'PetscErrorCode VecNorm(_p_Vec*, NormType, PetscReal*)'
> dolfin/la/PETScVector.cpp:267: error: at this point in file
> scons: *** [dolfin/la/PETScVector.os] Error 1
> scons: building terminated because of errors.
> 
> 
> _______________________________________________
> DOLFIN-dev mailing list
> [email protected]
> http://www.fenics.org/mailman/listinfo/dolfin-dev


_______________________________________________
DOLFIN-dev mailing list
[email protected]
http://www.fenics.org/mailman/listinfo/dolfin-dev

Reply via email to