On Tue, Apr 19, 2011 at 22:44, Ethan Coon <ecoon@lanl.gov> wrote:

The short answer -- make the z-size = s. It's not that huge of a hit --
the local arrays are 6*NX*NY*ndofs instead of 5*NX*NY*ndofs.

Indeed, though it is potentially about twice as much work because you only have to evaluate residuals on the owned nodes.

Otherwise, I'll try to implement the special-special case of
DMDA_BOUNDARY_PERIODIC, P=1, zl=1, and s > 1 soon.