https://gcc.gnu.org/bugzilla/show_bug.cgi?id=48665
--- Comment #17 from Jonathan Wakely <redi at gcc dot gnu.org> --- (In reply to David Blaikie from comment #15) > "The compiler still reuses the same representation for const/volatile and > for some attributes, sometimes misinterpreting one for the other." - sounds > like rejecting this valid code is a workaround for that choice of > representation? That comment you quoted is explaining why the diagnostic printed "void () __attribute__((const))" insetad of "void () const", not why the code should be rejected. It's not rejected because of a limitation in GCC.
