http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52548
--- Comment #4 from Richard Guenther <rguenth at gcc dot gnu.org> 2012-03-21 11:33:25 UTC --- Created attachment 26940 --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=26940 patch Patch. Not sure what to do about PA_IN - can we even end up with any invalid expressions in it, given we have a "clean" ANTIC_IN already? Well, I guess I have to understand what partial-antic computes ;)
