On Jun 24, 2013, at 9:25 , David Blaikie <[email protected]> wrote:

> 
> On Jun 24, 2013 9:17 AM, "Jordan Rose" <[email protected]> wrote:
> >
> > Couldn't this create problems for very big positive enums, i.e. those 
> > larger than 2^63?
> 
> Yep, that's the 'still a problem' I mentioned.
> 
> > Not that this isn't strictly better than what we had before.
> 
> Actually its not strictly better - I traded off one valid range for another. 
> It just happens to be the one we now get right is the same as GCC, which is 
> nice. GCC and clang both get the case you mentioned wrong in different ways 
> so far as I can tell
> 
Well, abstractly no. In practice I would guess people use negative enums more 
often than really big positive enums. ;-)

_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits

Reply via email to