It is unclear how you have actually mapped unknowns into your linear<br>
system. However, a global PETSc Mat would naively have only one copy<br>
of a given unknown, and thus the duplicate unknowns would indeed be<br>
out of synch. You have to do extra work to communicate the values after<br>
the solve.<br>
<br>
However, all this is handled for you if you use the DA object. It only handles<br>
logically Cartesian grids, but does all the ghosting. We will soon have an<br>
unstructured counterpart.<br>
<br>
Matt<br><br><div><span class="gmail_quote">On 5/17/06, <b class="gmail_sendername">Thomas Geenen</b> <<a href="mailto:geenen@gmail.com">geenen@gmail.com</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Dear Petsc users,<br><br>I am solving a system of equations Ax=b resulting from a Finite element<br>package. The parallelization strategy for this program is to minimize<br>communication. therefore each subdomain has its own copy of common nodal
<br>points between subdomains (local numbering). This strategy is already<br>implemented and I don't want to change that<br><br>I fill the matrix and right hand side with a call to MatSetValuesLocal and<br>VecSetValuesLocal. Some entries are filled on different subdomains (of course
<br>with the same value)<br><br>The solution for the first subdomain is correct. however the solution vector<br>on the second subdomain does not contain the solution on the common nodal<br>points it shares with subdomain 1.
<br>This is probably a feature but it is not consistent with my program setup.<br>The question is:<br>how can i tell petsc to return a solution for all the positions i filled with<br>VecSetValuesLocal?<br><br>Thanks in advance
<br>Thomas Geenen<br>ps I got the impression that the assemble routines also exchange common nodal<br>point info. It would be nice if this could be skipped as well.<br>Of course some extra communication could solve my problem but I would prefer a
<br>more elegant solution if one exists<br><br></blockquote></div><br><br clear="all"><br>-- <br>"Failure has a thousand explanations. Success doesn't need one" -- Sir Alec Guiness