https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85518
--- Comment #4 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Actually not fully. While it adds mangling for _Float{16,32,64,128}, it
doesn't add mangling for _Float{32,64,128}x.
https://itanium-cxx-abi.github.io/cxx-abi/abi.html#mangling doesn't have
anything for those and there is no need to add _Float{32,64,128}x support to
the C++ FE right now.
Perhaps the *x suffixed builtins should be C only?
