fix(manual_assert): don't lint on if cfg!(..) { panic!(..) }

[ada4a]

Fixes #13883

changelog: [manual_assert]: don't lint on if cfg!(..) { panic!(..) }

[rustbot]

r? @Alexendoo

rustbot has assigned @Alexendoo.
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

[samueltardieu]

Is that really a common idiom? If this is used through the debug_panic crate for example, shouldn't we instead refrain from linting code expanded from an external macro when we detect this pattern?

r? samueltardieu

View changes since this review

[ada4a]

Is that really a common idiom?

That is what the issue author was suggesting at least.

If this is used through the debug_panic crate for example, shouldn't we instead refrain from linting code expanded from an external macro when we detect this pattern?

I think that's an orthogonal but good idea, yes.

[samueltardieu]

Is that really a common idiom?

That is what the issue author was suggesting at least.

Yes, but it would be better if it was backed up by some numbers. Otherwise, we are just preventing the lint from triggering in legitimate situations. If this is not that common, the lint can just be allowed by users who want to do this.

[ada4a]

In that case, would you want to go ask the issue author for some numbers/close the issue , while I implement the from-expansion check?

EDIT: there's a from-expansion check already in place:

rust-clippy/clippy_lints/src/manual_assert.rs

Line 40 in 2a6197e

&& !expr.span.from_expansion()

[samueltardieu]

@oriongonza ?

[rustbot]

☔ The latest upstream changes (possibly #13787) made this pull request unmergeable. Please resolve the merge conflicts.