[petsc-dev] Mat Type: HYPRE

Kong, Fande fande.kong at inl.gov
Tue Oct 3 17:03:37 CDT 2017


Hi ALL,

There is a new Mat type: HYPRE in PETSc. I would like to try it out because
it may save the memory if use a hypre preconditioner. When use a native
PETSc type such as MPIAIJ, we need to keep an extra copy of Jacobian matrix
usually for the hypre preconditioner.

I was wondering if or not Mat HYPRE is complete. There are some errors when
I am tring to use it. The code works fine with MPIAIJ.


[0]PETSC ERROR: --------------------- Error Message
-------------------------------------------------------------- [0]PETSC
ERROR: Error in external library [0]PETSC ERROR: Error in
HYPRE_IJMatrixAddToValues(): error code 1 [0]PETSC ERROR: See
http://www.mcs.anl.gov/petsc/documentation/faq.html for trouble shooting.
[0]PETSC ERROR: Petsc Development GIT revision: v3.8-2-gd4ad237  GIT Date:
2017-10-03 12:37:25 -0600 [0]PETSC ERROR: ../../../moose_test-opt on a
arch-darwin-c-debug-master named FN604208 by kongf Tue Oct  3 15:54:15 2017
[0]PETSC ERROR: Configure options --with-debugging=yes
--with-shared-libraries=1 --with-mpi=1 --download-fblaslapack=1
--download-metis=1 --download-parmetis=1 --download-superlu_dist=1
--download-hypre=1 --download-mumps=1 --download-scalapack=1
PETSC_ARCH=arch-darwin-c-debug-master [0]PETSC ERROR: #1
MatSetValues_HYPRE() line 1014 in
/Users/kongf/projects/petsc/src/mat/impls/hypre/mhypre.c [0]PETSC ERROR: #2
MatSetValues() line 1270 in
/Users/kongf/projects/petsc/src/mat/interface/matrix.c application called
MPI_Abort(MPI_COMM_WORLD, 1) - process 0 [unset]: write_line error; fd=-1
buf=:cmd=abort exitcode=1 : system msg for write_line failure : Bad file
descriptor


Fande,
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20171003/fcb1b85e/attachment.html>


More information about the petsc-dev mailing list