https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87311
--- Comment #4 from Vincent Lefèvre <vincent-gcc at vinc17 dot net> --- (In reply to Richard Biener from comment #3) > Yes, I think we have quite some opportunistic replacement of traps by > "correct" results (according to twos complement). Yes, that's "correct" according to two's complement, but this may also be a real overflow that should have been caught (assuming -fwrapv hasn't been provided).