[petsc-dev] Weird interaction between VecDuplicate and VecSetBlockSize

[Jed Brown]

On Wed, Apr 4, 2012 at 19:17, Barry Smith <bsmith at mcs.anl.gov> wrote:

>  And this makes a difference how?


I just wanted to clarify the current state. It was very annoying to check
whether the matrix type was one for which MatSetBlockSize() could be
called, so I made it always safe. (That was quite a while ago.)

I don't have a problem with your proposed restructuring.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20120404/6d5bace0/attachment.html>