Fix some #[expect] lint interaction

[xFrednet]

Fixing the first few lints that aren't caught by #[expect]. The root cause of these examples was, that the lint was emitted at the wrong location.

---

changelog: none

r? @Jarcho

cc: rust-lang/rust#97660

[xFrednet]

Hey @Jarcho, could you review this. You can r+ this directly if you agree with the changes 🙃

@bors delegate=Jarcho

[bors]

✌️ @Jarcho can now approve this pull request

[Jarcho]

Taking a second look at async_yields_async I don't think anything needs to change. It's basically linting the signature of the body. I originally thought it might have an issue because it did something with the span of the final expression.

[Jarcho]

Everything else LGTM.

[xFrednet]

Well, while async_yields_async lints on the entire body, for the user it would be nice, if expect/allow also works on the statement that's shown in the lint message. Every problem you found with the expectation not working was also affecting other lint attributes, from what I can tell 🙃. Btw, thank you very much for checking these lints, that has been super valuable!

So far, I haven't got any conflicts (Made these changes last weekend or so). Should we merge it as it is, or should I add the other lint changes to this PR as well? I'm fine either way :)

[Jarcho]

I question whether that's a good span to lint on in general, but that can be a different discussion.

May as well merge now. Waiting just increases the chances of conflicts coming up. @bors r+

[bors]

📌 Commit 9d201d6 has been approved by Jarcho

[bors]

⌛ Testing commit 9d201d6 with merge b3c94c0...

[bors]

☀️ Test successful - checks-action_dev_test, checks-action_remark_test, checks-action_test
Approved by: Jarcho
Pushing b3c94c0 to master...