https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88775
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|UNCONFIRMED |NEW
Last reconfirmed| |2019-01-10
Ever confirmed|0 |1
--- Comment #4 from Richard Biener <rguenth at gcc dot gnu.org> ---
(In reply to Richard Biener from comment #3)
> _4 = _3 > "hello";
> _5 = __builtin_constant_p (_4);
> if (_5 != 0)
> goto <bb 4>; [34.00%]
> else
> goto <bb 3>; [66.00%]
>
> <bb 3> [local count: 708669605]:
> __x.5_6 = (long unsigned int) "hello";
> __y.6_7 = (long unsigned int) _3;
> _8 = __x.5_6 < __y.6_7;
>
> <bb 4> [local count: 1073741824]:
> # _9 = PHI <_4(2), _8(3)>
> if (_9 != 0)
> goto <bb 8>; [50.00%]
> else
> goto <bb 5>; [50.00%]
which is a somewhat awkward pattern. I think that VN should really be
able to value-number _4 and _8 the same eliding this crap. Let me see
to do that.
OK, that helps _a lot_.
