On Thu, Jun 24, 2010 at 08:47:09PM +0100, Edd Barrett wrote:
> On Wed, Jun 23, 2010 at 04:56:09PM -0600, Daniel Melameth wrote:
> > While most of us already know how the subject rings true, I still found the
> > following from REBOL's CTO's public blog post interesting nonetheless (I've
> > never used REBOL):
> >
> > "This was an interesting build, because it exposed a unique bug due to the
> > more secure methods of memory allocation on OpenBSD. Debugging it took some
> > time but was worth the effort. The bug has now been fixed and will be part
> > of the A100 releases for all platforms."
> >
> > The minor blog post is available at http://www.rebol.net/r3blogs/0321.html.
>
> Also I have found some bugs in both mutt and tex live with OpenBSD's
> J malloc flag.
>
> Great stuff
Who am I to disagree ;-)
In 4.7, there's a new S flag, which switches on in one go the options
(JFG) that help debugging even more, at the cost of some performance.
But I'd like to stress that even without flags malloc catches a lot of
bugs already. For example randomization of chunk and page allocation
and actually unmapping of unused pages happens even without options.
-Otto
