Thanks,<div>Alan<br><br><div class="gmail_quote">On Sat, Apr 7, 2012 at 3:49 PM, Jed Brown <span dir="ltr"><<a href="mailto:jedbrown@mcs.anl.gov">jedbrown@mcs.anl.gov</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im"><div class="gmail_quote">On Sat, Apr 7, 2012 at 15:41, Alan Wei <span dir="ltr"><<a href="mailto:zhenglun.wei@gmail.com" target="_blank">zhenglun.wei@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
I got a problem with /ksp/ksp/example/tutorial. I was trying to execute this example with petsc-dev in a share memory dell workstation with 8 cores. The parameter for DMDACreate2d is 500, 500, which means it generates a 1997*1997 2d grid. It is fine to run it with single process with either "./ex29" or "mpiexec -np 1 ./exe29". However, it came across a error when I tempted to run it with more than 1 process. The error message is like this, when I was trying to execute it with 4 processes. I was thinking about if the 'stack size' is the problem, so I enlarge it to 1000000000 with 'ulimit' command. However, the same error message keeps showing up.</blockquote>
</div><br></div><div>This custom viewer (in the example) is broken in parallel and has always been broken in parallel. I don't know who wrote it, but it should be removed since there is a binary VTK PetscViewer now.</div>
<div>
<br></div><div>Matt, this has come up a few times before and I thought you defended that viewer. Is there anything it does that VecView with PETSCVIEWERVTK does not?</div>
</blockquote></div><br></div>