http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50310
--- Comment #7 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-09-07 08:19:22 UTC --- (In reply to comment #2) > On this second testcase it started with > http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=171236 Still a latent issue - the above hid it probably as well, while with -fno-trapping-math it should have been exposed even without that patch.