compiler/middle/lint: Suggest #[expect(dead_code)] instead of #[allow(dead_code)]

[Turbo87]

Until now the compiler suggested

help: to override -W unused add #[allow(dead_code)]

when encountering e.g. an unused function in the codebase.

This PR changes the suggestion to #[expect(dead_code)], which will warn again once the code is being used somewhere. This is helpful because it lets the developer know that the attribute can now be removed again. Without this the codebase will eventually be littered with stray #[allow(dead_code)] attributes that are no longer serving any purpose.

This was suggested in https://bsky.app/profile/steveklabnik.com/post/3m2uh7pf6e225 and I agreed strongly enough to take a look and implement it 😆

[rustbot]

r? @petrochenkov

rustbot has assigned @petrochenkov.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[Turbo87]

r? @estebank

since we already talked about it :)

[Turbo87]

I scoped this to only the dead_code suggestion for now. It might make sense to extend it to others in the future too though.

[djc]

This is great! I wonder if it would be hard to make this conditional on the rust-version of the crate in question?

[estebank]

I wonder if it would be hard to make this conditional on the rust-version of the crate in question?

I don't think rustc has access to that information from cargo.

[Turbo87]

I noticed one small complication:

error: this lint expectation is unfulfilled
   --> src/coords.rs:100:14
    |
100 |     #[expect(dead_code)]
    |              ^^^^^^^^^
    |
    = note: `-D unfulfilled-lint-expectations` implied by `-D warnings`
    = help: to override `-D warnings` add `#[allow(unfulfilled_lint_expectations)]`

error: could not compile `openair` (lib test) due to 1 previous error

If the code is used in tests, but not from public API of a library, then the expect() will trigger a lint expectation is unfulfilled warning for the lib test compile target. with allow() this would not be an issue because it allows both used and unused cases.

I wonder whether we should suggest both expect() and allow() for dead_code instead?

[rustbot]

This PR was rebased onto a different master commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.

[Turbo87]

@estebank any thoughts on this? if you don't have time, should I find a different reviewer?