Location (URL)

https://doc.rust-lang.org/std/thread/struct.LocalKey.html#method.try_with

Summary

Currently it says:

If the key has been destroyed (which may happen if this is called in a destructor), this function will return an AccessError.

However I think it should say: "If the key has been destroyed (which may happen if this is called in a destructor), this function may return an AccessError."

Let me explain why I think this would be the correct documentation:

• LocalKey::with clearly documents that it does not always panic when the destructor has run and it uses LocalKey::try_with internally. If LocalKey::try_with really did detect all destroyed thread locals, then LocalKey::with could always guaratee a panic.
• The top level LocalKey docs seem to explain why previously destroyed thread locals would not always be possible to detect:

  On all platforms it’s possible for TLS to re-initialize other TLS slots during destruction.

It could be that I came to the wrong conclusion, but there is at least some inconsistency in the docs here.