On Nov 13, 2011, at 4:47 PM, Jed Brown wrote: > On Sun, Nov 13, 2011 at 16:44, Barry Smith <bsmith at mcs.anl.gov> wrote: > Why do it anyway different than DMDA? Better consistent and not a great way > then 10 different inconsistent and slightly different and better and worse > ways. > \ > The DMSliced implementation uses the template global vector for other things.
But DMDA does not? So it could be that DMSliced is just too much of the original design and needs updating as DMDA was modernised > We can change that so that it doesn't hold an extra circular reference, but I > think that's sidestepping the problem. Regardless, memory leaks make PETSc unusable so you need to fix the problem. Barry
