On Fri, Jun 10, 2011 at 19:21, Roy Stogner <[email protected]> wrote:
> It sounds as if we ought to allow a couple optional bool arguments to
> the PetscVector(Vec) constructor?
>
Yeah, and of course you could have a way to change them later.
> There's no way for that method to
> efficiently infer the state of a Vec (you'd need all-to-all
> communication equivalent to a redundant close() just to determine that
> there was nothing for close() to do), but most callers of that method
> ought to be able to tell it whether assembly and/or ghost updates are
> necessary.
>
Yes, but note that there are two useful local-to-global semantics:
ADD:
The usual finite element assembly,
VecGhostUpdateBegin(g,ADD_VALUES,SCATTER_REVERSE);
INSERT:
Off-process values are all identical and you only want to keep one. Used,
for example, for non-penalty enforcement of Dirichlet boundary conditions.
You must already manage these semantics, therefore you have enough
information for close() to do the right thing.
------------------------------------------------------------------------------
EditLive Enterprise is the world's most technically advanced content
authoring tool. Experience the power of Track Changes, Inline Image
Editing and ensure content is compliant with Accessibility Checking.
http://p.sf.net/sfu/ephox-dev2dev
_______________________________________________
Libmesh-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/libmesh-devel
