<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Hi Zacharie,<br>
The Petsc options have changed from version 3.0.0 on, I think that's
what causes the problem.<br>
With my version, your problem works.<br>
<br>
-mat_type umfpack<br>
becomes thus<br>
-pc_factor_mat_solver_package umfpack <br>
<br>
By the way, you should give a try to mumps. For that, download the
petsc 3.1-p*, configure it with option --download-mumps=ifneeded,
and recompile with openmpi. It is quicker even when serial.<br>
And then your option for using a direct solver will be: <br>
-pc_factor_mat_solver_package mumps<br>
<br>
Regards,<br>
Ruth<br>
<br>
<br>
On 07/10/10 16:11, Zacharie De Grève wrote:
<blockquote
cite="mid:201010071611.59806.zacharie.degreve@umons.ac.be"
type="cite">
<pre wrap="">
Dear GetDP users and developers,
I have some difficulties to solve the model contained in the enclosed .pro and
.msh files (2D magnetodynamics, a-v formulation with circuit relations).
With my previous configuration some months ago, I had no problems to solve it :
* getdp-2.0.0-cvs-20090612 compiled with petsc 2.3.3, and the following
.petsrc file :
-ksp_type preonly
-pc_type lu
-mat_type umfpack
* gmsh 2.3.0 (binary)
* Kubuntu 8.04
I decided a few days ago to upgrade everything :
* binary of getdp-2.0.0, already linked with petsc 3.0.0 and UMFPACK solvers
if I am right
* most recent versions of blas and liblapack
* gmsh 2.4.2 (binary)
* Kubuntu 10.04
Unfortunately, when I try to solve the enclosed model, with a :
getdp test.pro -msh TFO.msh -cal,
I get the following "zero pivot in LU factorization" error :
Info : Running 'getdp test.pro -msh TFO.msh -cal' [1 node(s)]
Info : Started on Thu Oct 7 14:04:25 2010
Info : Increasing process stack size (8192 kB < 16 MB)
...
P r e - P r o c e s s i n g . . .
Info : Loding Pre-Processing data 'test.pre'
Info : (CPU = 0.052003s, Mem = 0Mb)
E n d P r e - P r o c e s s i n g
P r o c e s s i n g . . .
Info : Generate[Sys_Mag]
Info : (CPU = 0.348021s, Mem = 0Mb)
Info : Solve[Sys_Mag]
Info : N: 4408
[0]PETSC ERROR: --------------------- Error Message
---------------------------------
[0]PETSC ERROR: Detected zero pivot in LU factorization
see <a class="moz-txt-link-freetext" href="http://www.mcs.anl.gov/petsc/petsc">http://www.mcs.anl.gov/petsc/petsc</a>-
as/documentation/troubleshooting.html#ZeroPivot!
[0]PETSC ERROR: Row 1427 has missing diagonal in factored matrix
try running with -pc_factor_nonzeros_along_diagonal or -
pc_factor_diagonal_fill!
[0]PETSC ERROR:
------------------------------------------------------------------------
[0]PETSC ERROR: Petsc Release Version 3.0.0, Patch 7, Mon Jul 6 11:33:34 CDT
2009
[0]PETSC ERROR: See docs/changes/index.html for recent updates.
[0]PETSC ERROR: See docs/faq.html for hints about trouble shooting.
[0]PETSC ERROR: See docs/index.html for manual pages.
[0]PETSC ERROR:
------------------------------------------------------------------------
[0]PETSC ERROR: test.pro on a umfpack-c named zdegreve by
zdegreve Thu Oct 7 14:04:25 2010
[0]PETSC ERROR: Libraries linked from /home/geuzaine/src/petsc-3.0.0-
p7/umfpack-cxx-opt/lib
[0]PETSC ERROR: Configure run at Mon Jul 27 22:43:37 2009
[0]PETSC ERROR: Configure options --with-debugging=0 --with-scalar-type=complex
--with-clanguage=cxx --with-shared=0 --with-mpi=0 --with-x=0 --with-fortran=1
--with-umfpack=1 --download-umfpack=ifneeded
[0]PETSC ERROR:
------------------------------------------------------------------------
[0]PETSC ERROR: MatILUFactorSymbolic_SeqAIJ() line 1335 in
src/mat/impls/aij/seq/aijfact.c
[0]PETSC ERROR: MatILUFactorSymbolic() line 5308 in src/mat/interface/matrix.c
[0]PETSC ERROR: PCSetUp_ILU() line 293 in src/ksp/pc/impls/factor/ilu/ilu.c
[0]PETSC ERROR: PCSetUp() line 794 in src/ksp/pc/interface/precon.c
[0]PETSC ERROR: KSPSetUp() line 237 in src/ksp/ksp/interface/itfunc.c
[0]PETSC ERROR: KSPSolve() line 353 in src/ksp/ksp/interface/itfunc.c
[0]PETSC ERROR: User provided function() line 69 in
unknowndirectory/LinAlg_PETSC.cpp
I have tried to add the options -pc_factor_nonzeros_along_diagonal or -
pc_factor_diagonal_fill as it is suggested, but it still does not work.
Do you have any ideas on how to fix it ?
Thank you very much in advance for your help !
Zacharie De Grève
PhD Student
University of Mons - Belgium
</pre>
<pre wrap="">
<fieldset class="mimeAttachmentHeader"></fieldset>
_______________________________________________
getdp mailing list
<a class="moz-txt-link-abbreviated" href="mailto:getdp@geuz.org">getdp@geuz.org</a>
<a class="moz-txt-link-freetext" href="http://www.geuz.org/mailman/listinfo/getdp">http://www.geuz.org/mailman/listinfo/getdp</a>
</pre>
</blockquote>
<br>
<pre class="moz-signature" cols="72">--
Dr. Ir. Ruth V. Sabariego
University of Liege, Dept. of Electrical Engineering & Computer Science,
Applied & Computational Electromagnetics (ACE),
phone: +32-4-3663737 - fax: +32-4-3662910 - <a class="moz-txt-link-freetext" href="http://ace.montefiore.ulg.ac.be/">http://ace.montefiore.ulg.ac.be/</a>
</pre>
</body>
</html>