does libc++ support c++17 special math functions?

[adapt-L]

https://en.cppreference.com/w/cpp/numeric/special_functions

These seem not to work for me unless I use libstdc++

[philnik777]

Not yet. We talked about this on the last libc++ meeting. See https://docs.google.com/document/d/1yC-3bE6pefVizUhfiUs1Shv11EzuL0KOklGSLqto6dk/edit?pli=1#heading=h.1qj3hyp88s2w for the discussion.

[h-vetinari]

I had noted in #52642:

For example, trying to find the current status of what's blocking C++17 conformance is not an easy task; e.g., searches for "P0226" or "Mathematical Special Functions" turn up nothing, although there were and are PRs for it. [since then also another]

I still think it would be good to have separate issues for these unimplemented papers (like clang does, mostly), so that it's easier to find out the current status.

Not yet. We talked about this on the last libc++ meeting.

Cool to see some movement on this!

[mordante]

I had noted in #52642:

For example, trying to find the current status of what's blocking C++17 conformance is not an easy task; e.g., searches for "P0226" or "Mathematical Special Functions" turn up nothing, although there were and are PRs for it. [since then also another]

I still think it would be good to have separate issues for these unimplemented papers (like clang does, mostly), so that it's easier to find out the current status.

I strongly disagree it's hard to find out. I don't know what you mean with Clang's status; I assume you mean https://clang.llvm.org/cxx_status.html. Libc++ has similar pages per language version. For example, for C++17 there is https://libcxx.llvm.org/Status/Cxx17.html.

Since this is not a bug, I'm closing the issue.

[h-vetinari]

I strongly disagree it's hard to find out.

As a prolific contributor, you're not a good stand-in for the experience of a less experienced user coming to the LLVM repo to find out the status of a given unimplemented paper. Even before the shutdown of phabricator, it would have been a challenge to find all attempts related to P0226, completely aside from having to switch systems to have any chance of success. This is even more the case now with phab in cryostasis and everything else being on Github.

I still think it would be good to have separate issues for these unimplemented papers (like clang does, mostly), so that it's easier to find out the current status.

I don't know what you mean with Clang's status; I assume you mean https://clang.llvm.org/cxx_status.html.

No, I meant the github issues that are separate from https://clang.llvm.org/cxx_status.html.

In any case, I'm glad to see that things seem to be moving in this direction after all: #79219.