This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [patch] Unify bitmap interface.


On 10/29/12, Diego Novillo <dnovillo@google.com> wrote:
> On Oct 29, 2012 Lawrence Crowl <crowl@googlers.com> wrote:
> > The sbitmap popcount function is only used in ebitmap, which is
> > itself not used.  If we do anything, removing them might be the
> > thing to do.
>
> Yes, please.

Separate patch, please.

> > The bitmap.h iterators set and the sbitmap.h iterator set
> > intersect in one element.  I would rather treat the iterator
> > conversion as a separate patch.
>
> OK, but this would be for 4.8, right?

Yes, depending on when stage one closes.

> > Richard said in an earlier mail, "I'd rather not mix this with
> > any kind of further C++-ification (that is introduction of
> > member functions or operator overloads)."  So, exactly what
> > are you all after?
>
> I'm after full unification.  The patch seemed to stop short of
> doing that.  If you are going to do follow-up patches, that would
> be fine.  But stage 1 is closing, and I'd like to have this done
> for 4.8.  I want to minimize partial transitions.  We have too
> many of those already.

The intent is for follow-up patches.

> > Yes, I meant that.  There are no functional changes here.
> > Why is contrib/config-list.mk not sufficient?
>
> Just to make sure.  Testing on ppc should be fast, for example.

Okay.

-- 
Lawrence Crowl


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]