http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56355
--- Comment #4 from Marc Glisse <glisse at gcc dot gnu.org> 2013-03-20 16:19:33 UTC --- (In reply to comment #3) > Jeff Law mentions that it would be good to add this "squares are nonnegative" > information to VRP, not just tree_binary_nonnegative_warnv_p. > http://gcc.gnu.org/ml/gcc-patches/2013-03/msg00690.html Er, actually, since my patch, VRP does manage to deduce a non-negative range for squares, although, sadly, when computing x*x, it doesn't manage to merge the information from the interval of x and this non-negative property.