https://github.com/aaronpuchert commented:
> If so, it might be better to diagnose the attribute itself than to produce a > capability expression that names a member Inner doesn't have. Agreed. The explicitly spelled-out wildcard is fine, but just taking the address of a non-static member of an enclosing class shouldn't be allowed. https://github.com/llvm/llvm-project/pull/231604 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
