[clang] [Clang] Implement C++26’s P2893R3 ‘Variadic friends’ (PR #101448)

[Vlad Serebrennikov via cfe-commits]

Endilll wrote:

> I would recommend creating an issue and see how ewg handles the new information before reaping out a lot of code.

Sure, we don't necessarily need to change our implementation as CWG continues to consider the issue, but we need to update the test and the status for bookkeeping purposes. That's only fair when we write tests against moving targets (open issues).

https://github.com/llvm/llvm-project/pull/101448