EricWF added a comment.
In https://reviews.llvm.org/D41977#975403, @rsmith wrote:
> This will still diagnose valid and reasonable programs, such as:
>
> struct ConvertToRef { operator int&(); };
> std::tuple<int&> t = {ConvertToRef()};
>
>
> ... on compilers that don't provide the trait. You could maybe try to work
> around that by checking to see if the type has a member `.operator int&()`.
> But perhaps it's better to remove the non-conforming check entirely, at least
> in the case where you can't reasonably get it right.
I agree with both your comments. I'll remove the non-conforming check.
https://reviews.llvm.org/D41977
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
