Toby, Thank you for your response. Step-17 and step-40 both worked, which was part of my confusion. I ended up recompiling everything using gcc and openmpi, and everything (SLEPc) works fine. At this point, this is a reasonable solution for me.
Best, Jon On Fri, Dec 9, 2011 at 5:07 AM, Toby D. Young <[email protected]> wrote: > > > Hello Jon, > > > Has anyone out been successful with step-36 and Intel products? From > > the error I'm getting, it seems like there may be a difference in the > > Intel and GNU implementation of a function somewhere, which is > > causing incorrect sized/initialized objects to be passed to the > > solver. Has anyone run into an issue like this before? > > I have no idea if anyone has tried running SLEPc with anything other > than the gnu compilers. I have not tried this with Intel compilers. > > Judging from the error (below), SLEPc has completed solving the > eigenspectrum problem and is having trouble recursively destroying the > EPS context - it seems to be deleting the mpi_communicator that > triggers the errors. > > So here's a question: Does step-17 work ok with intel for you? Since > step-36 is a serial code only, and both PETSc and SLEPc wrappers default > to a MPI_COMM_SELF as an mpi_communicator, the error may be being > triggered by PETSc alone. Having said that step-17 uses the > MPI_COMM_WORLD, I don't know imediately what the difference is in this > case. Is step-17 going to work in serial if you instantiate the code > with MPI_COMM_SELF instead of MPI_COMM_WORLD? (line 316 step-17.cc) > > Maybe there is a clue there(?) Sorry I can not test this myself, you > will have to dig around a little bit. > Please let me know if you find some clues, or even figure out what is > going wrong here. Thanks. > > Best, > Toby > > > > ------------------------------------------------------------------------ > > [0]PETSC ERROR: PetscCommDestroy() line 246 in src/sys/objects/tagm.c > > [0]PETSC ERROR: PetscSubcommDestroy() line 133 in > > src/sys/objects/subcomm.c [0]PETSC ERROR: PCDestroy_Redundant() line > > 243 in src/ksp/pc/impls/redundant/redundant.c [0]PETSC ERROR: > > PCDestroy() line 119 in src/ksp/pc/interface/precon.c [0]PETSC ERROR: > > KSPDestroy() line 755 in src/ksp/ksp/interface/itfunc.c [0]PETSC > > ERROR: STDestroy() line 161 in src/st/interface/stfunc.c [0]PETSC > > ERROR: EPSDestroy() line 682 in src/eps/interface/basic.c terminate > > called after throwing an instance of > > 'dealii::SLEPcWrappers::SolverBase::ExcSLEPcError' what(): > > -------------------------------------------------------- An error > > occurred in line <44> of file > > </software/ARL/deal.II/svn-gale/source/lac/slepc_solver.cc> in > > function dealii::SLEPcWrappers::SolverBase::SolverData::~SolverData() > > The violated condition was: ierr == 0 > > The name and call sequence of the exception was: > > ExcSLEPcError(ierr) > > Additional Information: > > An error with error number 64 occured while calling a SLEPc > > function -------------------------------------------------------- > > > > Aborted > > > -- > Toby D. Young > > Assistant Professor Philosophy & Physics > Polish Academy of Sciences > www: http://www.ippt.gov.pl/~tyoung > skype: stenografia > > _______________________________________________ > dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii >
_______________________________________________ dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
