<div dir="ltr"><div class="gmail_quote">On Wed, Mar 30, 2011 at 18:43, I-Kamenetskiy, Dmitry S <span dir="ltr"><<a href="mailto:dmitry.s.kamenetskiy@boeing.com">dmitry.s.kamenetskiy@boeing.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
I have run in to the following issue with the PETSc SNES
usage. The logic of the SNESSolve() routine in my code is that the solution
vector x should include ghost points at the boundary of subdomains in the
distributed case.</blockquote></div><br><div>Not true, the solvers take "global" vectors, each degree of freedom appears exactly once on exactly one process. Ghost points are typically needed inside of residual evaluation and Jacobian assembly, but they should not be present in the vector you give to the solver.</div>
</div>