Re: [PATCH 0/5] obstacks again

[Alan Modra]

On Wed, Oct 29, 2014 at 12:33:19AM -0700, Paul Eggert wrote:
> Thanks for doing all this.  The gnulib patches are good as far as they go,
> but they need one more change: alignments should also change from int to
> size_t. The first attached gnulib patch does that, plus it fixes a
> longstanding integer overflow bug that can occur with large alignments (plus
> large sizes).  While we're in the neighborhood we should be using C11's
> alignof rather than reinventing that particular wheel; the second attached
> gnulib patch does that. I've installed your five gnulib patches plus the two
> attached patches into gnulib.

One thing though, I didn't put the ChangeLog diffs in the patch as I
usually add them when committing.

> Two things for the glibc patch.  First, the updated gnulib patches need to
> be merged into the glibc patch.  Second, the manual needs to be updated to
> match the revised API induced by all these patches.

Ow!  The manual reflects obstacks as they were 20 years ago.

Hmm, looking over it I see a misfeature that I've removed.

@cindex shrinking objects
You can use @code{obstack_blank} with a negative size argument to make
the current object smaller.  Just don't try to shrink it beyond zero
length---there's no telling what will happen if you do that.

It is no longer possible to shrink an obstack with obstack_blank (but
you can still do that with obstack_blank_fast).

-- 
Alan Modra
Australia Development Lab, IBM