[petsc-users] MUMPS causing segmentation fault

[Satish Balay]

>>>>>>
Configure options --prefix=/Users/manav/Documents/codes/numerical_lib/petsc/petsc_with_external_libs/petsc-3.5.3/../ --CC=mpicc-openmpi-mp --CXX=mpicxx-openmpi-mp --FC=mpif90-openmpi-mp --with-clanguage=c++ --with-fortran=0 --with-mpi-include=/opt/local/include/openmpi-mp --with-mpi-lib="[/opt/local/lib/openmpi-mp/libmpi_cxx.dylib,/opt/local/lib/openmpi-mp/libmpi.dylib]" --with-mpiexec=/opt/local/bin/mpiexec-openmpi-mp --with-x=0 --with-debugging=0 --with-lapack-lib=/usr/lib/liblapack.dylib --with-blas-lib=/usr/lib/libblas.dylib --download-superlu=yes --download-superlu_dist=yes --download-suitesparse=yes --download-mumps=yes --download-scalapack=yes --with-parmetis-dir=/opt/local/ --with-metis-dir=/opt/local --with-scalapack-dir=/opt/local
<<<<

I'm not sure which version of metis/parmetis you have - but mumps has
issues with the latest version.

Suggest using --download-metis --download-parmetis instead.

Satish


On Fri, 6 Feb 2015, Manav Bhatia wrote:

> Hi, 
>    I am trying to run MUMPS with the following command line options: 
>  -ksp_type preonly -pc_type lu -pc_factor_mat_solver_package mumps -mat_mumps_icntl_4 4 -info
> 
>   This works fine with ex2.c.
> 
>    My code, which runs fine with -ksp_type preonly -pc_type lu throws a segmentation fault error with the command line
> options listed above. The output of the code is attached. 
> 
>    Is there anything that seems obviously wrong here? 
> 
> Thanks,
> Manav
> 
> 
>