https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77436
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|UNCONFIRMED |RESOLVED
Resolution|--- |INVALID
--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
This is undefined behavior and here is why:
> Since the lower limit is the additive inverse of the upper limit, all summands
> must cancel each other out and the result must be zero (this is still true
> with
> overflow)
No that is not true.
If you were adding ((-(INT_MAX-1)) + (INT_MAX-1)) + (-(INT_MAX-2)) +
(INT_MAX-2) in that order, there would be no overflow but since you are adding
in the following order:
((-(INT_MAX-1)) + (-(INT_MAX-2))
There is an overflow right away.
