https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70920
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|UNCONFIRMED |NEW
Last reconfirmed| |2016-05-03
Ever confirmed|0 |1
--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Confirmed. Not sure about the non-zero constants though. At least
non-equality pointer compares are only valid for pointers to the same object.
This can be easily fixed with a new match.pd rule like
(for cmp (ne eq)
(simplify
(cmp (convert @0) INTEGER_CST@1)
(if (POINTER_TYPE_P (TREE_TYPE (@0)))
(cmp @0 (convert @1)))))
I wonder about the reverse transform when a is an integer and the cast is to
a pointer type. I suppose we should strip the cast there as well.
