In D92044#2491583, @ldionne wrote:

@rarutyun Can you please rebase onto main so that the CI will trigger? I added the LLVM monorepo tag to the review.

Michael approved this review. Can we proceed?

Any progress on that review? Should I rebase it one more time?

In D92044#2416979, @ldionne wrote:

I also faintly remember that we already tried implementing this in libc++, didn't we?

Didn't find corresponding feature macro. Let me know if we need to do something with that.

I prepared the small change with adding the proper release version but I cannot upload the new diff. I guess that's because I am not author anymore @ldionne, is it possible to make me author again? I didn't find the way to do it myself, I guess I don't have permissions for that.

By the way currently the tests failed on GCC with friend approach.

I think this is pretty much good to go, however there's a specific way I'd like to structure the test harness to make the approach clearer. I'll commandeer this because it's a lot easier than explaining exactly what I mean, but LMK if you're OK with my approach. It's very very close to yours.

@ldionne any thoughts on the current state of the review?

@rarutyun Please look at the tests in D59886 and take anything that increases the coverage.

Hi @zoecarver. I am also sorry about duplicating the work. I looked at the cxx2a_status.html and saw that it is not taken. I created the review with just having the update of cxx2a_status.html with marking that feature as In progress. Unfortunately, I didn't have information who should be invited to the review. I knew only Marshall and added him. Seems nobody was informed I took this part of work. As you can see the review was created on September the 5th. After that it was very hard to find the time to finish the feature implementation. Finally I found the bandwidth and implemented all of this stuff for a couple of days.