https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122905

--- Comment #11 from Nathaniel Shead <nshead at gcc dot gnu.org> ---
(In reply to Jakub Jelinek from comment #9)
> Why does the r16-3233 change anything for class scope unnamed enums?  Those
> always have the linkage of the containing class, don't they?  If it would be
> restricted just to namespace scope unnamed enums, there would be no ODR
> violation.
> Or if there is some reason for that, we might want some flag on the enum and
> make the LTO ODR checking ignore mismatches of no DECL_NAME vs. DECL_NAME of
> the first enumerator.

This was only really intended to apply to namespace-scope enums.  The wording
change didn't specify this though and I overlooked this, but I agree that the
correct fix should probably be to just add `TYPE_NAMESPACE_SCOPE_P`.

Reply via email to