dolfin team mailing list archive

["Matthew Knepley" <knepley@xxxxxxxxx>]

It is possible that Dolfin is not linking against:

 $PETSC_DIR/lib/$PETSC_ARCH/libmpiuni.a

   Matt

On 11/29/06, Naboulsi, Samir K CTR USAF ASC/HP
<Samir.Naboulsi@xxxxxxxxxxxx> wrote:
> Dear All,
>
>
> As new user, I tried to compile the demo case poisson problem. IF I
> compile the Poisson.form using ffc and do "make" I get Poisson.h:76:
> error: 'const class dolfin::Cell' has no member named 'vertexID'  which
> created by ffc. I am using ffc-0.3.3
>
> If I use the Poisson.h that came with dolfin 0.6.2-1, I get a different
> error that has to do with PETSc running MPI
>
> /workspace/naboulsk/FX/petsc-2.3.2-p4/lib/linux-gnu-c-debug/libpetscksp.
> a(bjacobi.o)(.text+0x12a): In function `PCSetUp_BJacobi':
> /workspace/naboulsk/FX/petsc-2.3.2-p4/src/ksp/pc/impls/bjacobi/bjacobi.c
> :26: undefined reference to `MPIUNI_TMP'
> /workspace/naboulsk/FX/petsc-2.3.2-p4/lib/linux-gnu-c-debug/libpetscksp.
> a(bjacobi.o)(.text+0x197):/workspace/naboulsk/FX/petsc-2.3.2-p4/src/ksp/
> pc/impls/bjacobi/bjacobi.c:27: undefined reference to `MPIUNI_TMP'
>
>
> I have compiled PETSc disabling MPI (i.e. I am using single processor
> machine) and also when compiled Dolfin. i.e.
>
> >> ./config/configure.py --prefix=/workspace/naboulsk/FX/petsc-2.3.2-p4
> --with-c++-support --with-fortran --with-mpi=0
> --with-blas-lapack-dir=/usr/lib
>
> >> ./configure.local --enable-debug
> --with-boost=/workspace/naboulsk/FX/local/include/boost-1_33_1
> --enable-petsc --with-petsc-dir=/workspace/naboulsk/FX/petsc-2.3.2-p4
> --enable-pydolfin --disable-mpi
>
>
> How can I fix these problems?
> Thanks
> Sam
>
>
> _______________________________________________
> DOLFIN-dev mailing list
> DOLFIN-dev@xxxxxxxxxx
> http://www.fenics.org/mailman/listinfo/dolfin-dev


--
"Failure has a thousand explanations. Success doesn't need one" -- Sir
Alec Guiness