[vtk-developers] Valgrind and other dashboard issues

[David Partyka]

Hi David,

Your observations do not fall on deaf ears. I try to fill that role as time
permits as well, though more so off list. We will track down the sources of
those changes and have them resolve the leaks. Especially considering that
often those things are "low hanging fruit" and often easily corrected.

On Wed, Jun 29, 2011 at 12:03 PM, David Gobbi <david.gobbi at gmail.com> wrote:

> Hi All,
>
> There seem to be some valgrind issues (and dashboard warnings, test
> failures) that have been around for a long time.  It used to be the
> case that when people submitted code, they were given a stern warning
> if their code broke the dashboard, usually Bill Lorensen was the one
> cracking the whip.  These days, dashboard issues don't seem to draw
> much attention.
>
> For valgrind, specifically, these uninitialized memory
> reads/conditionals have been around for ages:
>
> Polyhedron/MVC: 5 UMRs, 1 UMC
> ReebGraph:  2 UMRs, 56 UMCs, 1 mismatched deallocate (!!!!)
> ZsweepConcavities: 1 UMR, 1 UMC
>
> The responsibility for fixing these issues should be shared between
> the person who submitted the code, and the person/group who accepted
> the code.  But most importantly, the VTK community as a whole should
> be more proactive in chasing after people to get these issues fixed.
>
> This might be an issue for the ARB to discuss.
>
>  - David
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Follow this link to subscribe/unsubscribe:
> http://www.vtk.org/mailman/listinfo/vtk-developers
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/vtk-developers/attachments/20110630/d284a420/attachment.html>