On Fri, Dec 9, 2011 at 10:31 AM, Alexander Grayver <span dir="ltr"><<a href="mailto:agrayver@gfz-potsdam.de">agrayver@gfz-potsdam.de</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Hi dev-team,<br>
<br>
I have this code:<br>
<br>
  call KSPCreate(comm3d,ksp,ierr);<u></u>CHKERRQ(ierr)<br>
  call KSPSetOperators(ksp,A,A,<u></u>DIFFERENT_NONZERO_PATTERN,<u></u>ierr);CHKERRQ(ierr)<br>
  call KSPSetType(ksp,KSPPREONLY,<u></u>ierr);CHKERRQ(ierr)<br>
  call KSPGetPC(ksp,pc,ierr);CHKERRQ(<u></u>ierr)<br>
  call PCSetType(pc,PCCHOLESKY,ierr);<u></u>CHKERRQ(ierr)<br>
  call PCFactorSetMatSolverPackage(<u></u>pc,MATSOLVERMUMPS,ierr);<u></u>CHKERRQ(ierr)<br>
  call PCFactorSetUpMatSolverPackage(<u></u>pc,ierr);CHKERRQ(ierr)<br>
  call PCFactorGetMatrix(pc,F,ierr);<u></u>CHKERRQ(ierr)<br>
<br>
which works well under petsc-3.2-p5, but produces error with petsc-dev:<br></blockquote><div><br></div><div>I believe that Cholesky only works with SBAIJ not AIJ.</div><div><br></div><div>   Matt</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">

[0]PETSC ERROR: --------------------- Error Message ------------------------------<u></u>------<br>
[0]PETSC ERROR: No support for this operation for this object type!<br>
[0]PETSC ERROR: Matrix format mpiaij does not have a built-in PETSc CHOLESKY!<br>
[0]PETSC ERROR: ------------------------------<u></u>------------------------------<u></u>------------<br>
[0]PETSC ERROR: Petsc Development HG revision: a4f22f4fcb371a016e046e537076dc<u></u>fd2ce5087f  HG Date: Fri Dec 09 09:08:30 2011 -0600<br>
[0]PETSC ERROR: See docs/changes/index.html for recent updates.<br>
[0]PETSC ERROR: See docs/faq.html for hints about trouble shooting.<br>
[0]PETSC ERROR: See docs/index.html for manual pages.<br>
[0]PETSC ERROR: ------------------------------<u></u>------------------------------<u></u>------------<br>
[0]PETSC ERROR: /home/model on a openmpi-i named node233 by agrayver Fri Dec  9 17:23:30 2011<br>
[0]PETSC ERROR: Libraries linked from /home/lib/petsc-dev/openmpi-<u></u>intel-complex-release-f-ds/lib<br>
[0]PETSC ERROR: Configure run at Fri Dec  9 16:51:10 2011<br>
[0]PETSC ERROR: Configure options --download-metis --download-mumps --download-parmetis --download-superlu_dist --with-blacs-include=/opt/<u></u>intel/Compiler/11.1/072/mkl/<u></u>include --with-blacs-lib=/opt/intel/<u></u>Compiler/11.1/072/mkl/lib/<u></u>em64t/libmkl_blacs_openmpi_<u></u>lp64.a --with-blas-lapack-dir=/opt/<u></u>intel/Compiler/11.1/072/mkl/<u></u>lib/em64t --with-debugging=0 --with-fortran-interfaces=1 --with-fortran-kernels=1 --with-mpi-dir=/opt/mpi/intel/<u></u>openmpi-1.4.2 --with-petsc-arch=openmpi-<u></u>intel-complex-release-f-ds --with-precision=double --with-scalapack-include=/opt/<u></u>intel/Compiler/11.1/072/mkl/<u></u>include --with-scalapack-lib=/opt/<u></u>intel/Compiler/11.1/072/mkl/<u></u>lib/em64t/libmkl_scalapack_<u></u>lp64.a --with-scalar-type=complex --with-x=0 PETSC_ARCH=openmpi-intel-<u></u>complex-release-f-ds<br>

[0]PETSC ERROR: ------------------------------<u></u>------------------------------<u></u>------------<br>
[0]PETSC ERROR: MatGetFactor() line 3943 in /home/lib/petsc-dev/src/mat/<u></u>interface/matrix.c<br>
[0]PETSC ERROR: PCFactorSetUpMatSolverPackage_<u></u>Factor() line 17 in /home/lib/petsc-dev/src/ksp/<u></u>pc/impls/factor/factimpl.c<br>
[0]PETSC ERROR: PCFactorSetUpMatSolverPackage(<u></u>) line 26 in /home/lib/petsc-dev/src/ksp/<u></u>pc/impls/factor/factor.c<br>
<br>
Any idea?<br>
<br>
Regards,<br>
Alexander<br>
</blockquote></div><br><br clear="all"><div><br></div>-- <br>What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>
-- Norbert Wiener<br>