[CIG-LONG] Gale
Ritske S. Huismans
Ritske.Huismans at geo.uib.no
Thu Sep 25 10:50:05 PDT 2008
Hello Walter,
I am a bit stuck on the 3D Gale models. Do you have any more thoughts on
this?
best regards,
Ritske
-----Original Message-----
From: Ritske S. Huismans
Sent: 21. september 2008 21:45
To: 'Walter Landry'
Cc: cig-long at geodynamics.org
Subject: RE: [CIG-LONG] Gale
Hello Walter,
thanks for the reply.
The error message starts like this and continues.
[0]PETSC ERROR: [29]PETSC ERROR:
------------------------------------------------------------------------
[21]PETSC ERROR:
------------------------------------------------------------------------
[13]PETSC ERROR:
------------------------------------------------------------------------
[13]PETSC ERROR: Caught signal number 11 SEGV: Segmentation Violation,
probably memory access out of range
[13]PETSC ERROR: Try option -start_in_debugger or
-on_error_attach_debugger
[13]PETSC ERROR: or see
http://www.mcs.anl.gov/petsc/petsc-as/documentation/troubleshooting.html
#Signal[13]PETSC ERROR: or try http://valgrind.org on linux or man
libgmalloc on Apple to find memory corruption errors
[7]PETSC ERROR:
------------------------------------------------------------------------
[31]PETSC ERROR:
------------------------------------------------------------------------
[29]PETSC ERROR: Caught signal number 11 SEGV: Segmentation Violation,
probably memory access out of range
[29]PETSC ERROR: Try option -start_in_debugger or
-on_error_attach_debugger
[29]PETSC ERROR: or see
http://www.mcs.anl.gov/petsc/petsc-as/documentation/troubleshooting.html
#Signal[29]PETSC ERROR: or try http://valgrind.org on linux or man
libgmalloc on Apple to find memory corruption errors
[29]PETSC ERROR: likely location of problem given in stack below
and continues like this.
the 1 cpu job has the same error.
Thanks,
Ritske
"Ritske S. Huismans" <Ritske.Huismans at geo.uib.no> wrote:
> Hello,
>
> I am interested to use Gale for comparison with other ALE codes in
> 2D/3D.
> I have installed Gale on our system, it runs the basic tests fine.
>
> some questions / comments that I am happy to get some feedback on:
> 1. There are some small differences in model results from visual
> comparison with the Gale manual.
>
> 2. Most notably, the 2D Tibet example does not show plastic
localisation
> when I run it, not sure what may be different.
It has been a while since I ran it, and the implementation of Mohr
Coulomb has changed a lot since then. I should run it again.
> 3. I have tried to run the tibet3d.xml example but get memory failures
> of ptsc, even if I reduce the model resolution to
> something small, e.g. 16 x 16 x 4
> I have tried to run this last example at 256 cpu with 1 Gb/core. this
> should be enough.
It may be that petsc is catching a place where Gale is writing to
invalid memory. What is the exact error that you get? For the small
model, can you run it on one processor?
> I am especially interested in feedback how to get Gale running for 3D
> problems including plastic localisation.
> Thanks for any comments you can give in this regard,
I am out of town right now, so it will take me a few days to run these
models again.
Cheers,
Walter Landry
walter at geodynamics.org
More information about the CIG-LONG
mailing list