http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54844
--- Comment #3 from Paolo Carlini <paolo.carlini at oracle dot com> 2012-10-08 12:10:26 UTC --- REAL_TYPE remains unhandled in tsubst_copy. First blush however, I'm not sure if we shouldn't be passing REAL_TYPEs at all, or it's just a trivial matter of adding to the switch.