[cxx-abi-dev] Decltype mangling and type dependency

[David Vandevoorde]

On Jun 22, 2010, at 3:15 PM, Jason Merrill wrote:

> On 06/21/2010 05:51 PM, David Vandevoorde wrote:
>> But that's an error: The mangling doesn't matter at that point, no?
> 
> Not in this case, certainly, since the return type isn't mangled.
> 
>> I didn't think there is a valid case where access is the only disambiguator between two entities (from a mangling perspective, in particular).  Did I get that wrong?
> 
> Hmm.  In general, errors are relevant to mangling because of SFINAE, but I'm not sure if that is possible in this case (even assuming that access control violations get added to SFINAE).


I thought we (the C++ committee) deliberately did _not_ put access control under SFINAE?


>  Let's see:
> 
> template <class T> struct C;
> 
> class A
> {
>  int i;
>  friend struct C<int>;
> } a;
> 
> class B
> {
>  int i;
>  friend struct C<float>;
> } b;
> 
> template <class T>
> struct C
> {
>  template <class U> decltype (a.i) f() { } // #1
>  template <class U> decltype (b.i) f() { } // #2

This (#2) declaration is invalid because it's a redeclaration of #1.  (And if they weren't they'd have distinct signatures.)


> };
> 
> int main()
> {
>  C<int>().f<int>();     // calls #1
>  C<float>().f<float>(); // calls #2
> }
> 
> OK, looks like we can create a situation where we have two different function templates with the same signature but different access restrictions.


I still don't see it: Even ignoring the redeclaration error and the fact that access errors aren't under SFINAE, don't these two have different signatures since they're members of distinct class types (C<int> vs. C<float>)?

	Daveed