Tracking issue for cfg_if

[gnzlbg]

This issue tracks the stabilization of libcore's cfg_if! macro. See: #59443

[jonas-schievink]

Added in #59336, I think?

[SimonSapin]

#59443 was closed, and #59336 adds the macro but does not export it. Closing as there is no public unstable API to track here. Feel free to reopen if/when the macro is exported.

[RalfJung]

In the light of #65860 we might reconsider actually exposing this from libcore: fn-like macros are "more powerful" in some sense than attribute-based ones, in particular around syntax the compiler does not understand. We don't ship a fn-like cfg macro in libcore, but with the recent syntax feature gating issues I feel we should -- and cfg_if seems like a great choice for that.

[gnzlbg]

A cfg_if! like addition would require an RFC, and in the PR that proposed adding it a couple of people argued that they would prefer a cfg_match! instead, but nobody has put much work into exploring that. So if somebody wants to purse a solution to the problem, that would be a good place to get started.

[Centril]

Fwiw, while I prefer cfg_match!, I think cfg_if! is a decent solution as well.
cc @SimonSapin & @alexcrichton -- does this require an RFC or is a PR enough?

[RalfJung]

Agreed, match seems more Rust-y than an if-else chain.

[alexcrichton]

Given the contention in the past with cfg_if! it feels like its best to be safe and have an RFC for this.