http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47049
--- Comment #12 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-02-08 13:58:17 UTC --- It is a regression in the sense that e.g. the #c10 testcase with 4.4 resulted in a bunch of errors, but not an ICE, while it ICEs in 4.6.
