------- Additional Comments From pinskia at gcc dot gnu dot org 2005-05-08 19:08 ------- Hmm, we do optimize this with ashiftrt but not with ashift. We also fold ((1 << a) & 1) != 0 into ((1 >> a & 1) != 0) which gets optimized on the RTL level.
-- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15350
