------- Comment #3 from pinskia at gcc dot gnu dot org 2006-12-14 03:59 ------- (In reply to comment #2) > but only if we don't honor NaNs or signed zeros.
Some of these optimizations can be done even without worrying about NaNs because of -fcx-limited-range . -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30172
