[Bug c++/50115] Integer test optimised away at -O2 by the VRP

[redi at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50115

--- Comment #10 from Jonathan Wakely <redi at gcc dot gnu.org> 2011-08-18 13:13:22 UTC ---
actually, technically unsigned integers cannot "overflow", they wrap, with
well-defined behaviour (as it says, "unsigned arithmetic does not overflow")

but in layman's terms, yes, you are correct