https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122908
--- Comment #1 from Bernardo Negri <b.gomes.negri at gmail dot com> --- Might be related to bug #121574, but I think this is a different issue. That bug was a request for GCC to ignore "errors" in code that does not strictly conform to the spec but was still functional. I do not think GCC's behavior in this case conforms to spec, but there is a strong chance I am wrong.
