<div class="gmail_quote">On Mon, Aug 27, 2012 at 6:08 PM, Dmitry Karpeev <span dir="ltr"><<a href="mailto:karpeev@mcs.anl.gov" target="_blank">karpeev@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"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="gmail_quote"><div>2. Show me the part of this code that uses jac->head->ksp?</div></div>
</blockquote></div><div>Incidentally, regarding the use of jac->head->ksp: </div><div>True, there is no use of jac->head->ksp -- the Schur complements "inner" solver is being used, </div>
<div>but that's a conceptual mistake, as far as I can tell. It doesn't matter in petsc-3.3 where the inner </div><div>and outer A00 solvers are identical under all circumstances. Not so in petsc-dev -- we recently went</div>
<div>through all that trouble of introducing ways of setting up the inner and outer solvers separately -- and</div><div>now the outer solver (jac->head->ksp) isn't even being used. I imagine that's just simple oversight,</div>
<div>but once it's corrected, jac->head->ksp will have to be set up even in the Schur case. </div></blockquote></div><br><div>As I understand it, Matt is planning to complete this feature at some point.</div>