Slight inaccuracy in the rustc_mir documentation

[giltho]

Hello,

It feels like this shouldn't be an issue but I don't really know where to post it otherwise..
The documentation of eval_place in the MIR InterpCx says that there is a more efficient solution when the place is access for read called eval_place_for_read, but the latter does not exist anywhere.
I'm guessing it should be referring to eval_place_to_op, but I am not entirely sure. Also as a side question, I'd be curious to know why it is slower to access a place for write than it is for read.

I'm going to spend a bit of time reading that part of the code, if I ever find other tiny things like that, where should I report it ? It feels like this could be a small discussion in a chat followed by a tiny PR rather than an issue.

Thank you !

[scottmcm]

If you have a question while making a PR, on good resource is Zulip: https://rust-lang.zulipchat.com/#narrow/stream/182449-t-compiler.2Fhelp

Depending how easy a PR it is and whether you're reasonably confident in your deduction, you could also just make a PR changing it and the reviewer will confirm that the update is accurate.

[pierwill]

@giltho Your analysis looks correct to me. I'm working up a PR to fix these docs.

@rustbot claim

[pierwill]

I'm going to spend a bit of time reading that part of the code, if I ever find other tiny things like that, where should I report it ? It feels like this could be a small discussion in a chat followed by a tiny PR rather than an issue.

Issues like this are welcome, as are small PRs. :)