On Nov 21, 2011, at 9:39 AM, Hui Zhang wrote: > >> On Nov 21, 2011, at 8:01 AM, Jed Brown wrote: >> >>> On Mon, Nov 21, 2011 at 04:52, Hui Zhang <buyong.huier at gmail.com> wrote: >>> When using DMDAGlobaltoLocal, how can I communicate only the boundary of the >>> ghosted domain (with stecil width larger than one)? >>> >>> Only the ghost values are communicated. The interior is all local. >> >> I think Hui means DMDALocalToLocalBegin() and DMDALocalToLocalEnd() but >> normally one doesn't need this with the PETSc solvers, one only needs the >> DMGlobalToLocalBegin/End() >> >> Barry >> > > Thank you all! > > I'm actually implementing domain decomposition methods with restricted > communication. > From my understanding, DMDALocalToLocalXX is also communicating all ghost > values of > the second Vec. However, in a restricted way, only the ghost points most far > away from > the interior are communicated.
Yes, we always communicate all ghost points, not just the boundary of the ghosted domain. Sorry I miss understood your earlier question. It may be possible to modify the DMDA to communicate only the boundary but would require mucking around with the details of the code a bit. Barry
