warn on must_use use on async fn's

[guswynn]

As referenced in #78149

This only works on async fn's for now, I can also look into if I can get Box<dyn Future> and impl Future working at this level (hir)

[rust-highfive]

r? @wesleywiser

(rust-highfive has picked a reviewer for you, use r? to override)

[wesleywiser]

cc @eholk since this is an async-polish item

Before r+ing, I need to double check if we need to get the ok from the lang team before adding new warnings

[eholk]

Looks good to me!

[wesleywiser]

Nominating for lang team discussion. This PR adds a new warning that detects #[must_use] attributes applied to async fns which currently does nothing.

warning: `must_use` attribute on `async` functions applies to the anonymous Future returned by the function, not the value within.
  --> $DIR/unused-async.rs:5:1
   |
LL |   #[must_use]
   |   ^^^^^^^^^^^
LL |
LL | / async fn test() -> i32 {
LL | |     1
LL | | }
   | |_- this attribute does nothing, the Futures returned by async functions are already `must_use`

[Mark-Simulacrum]

I suspect this might just be an oversight in this PR, but it seems like this should be an unused_attributes lint, not an unconditional warning (and a custom message is still fine).

If so, then I suspect it need not receive T-lang approval... though that is a little murky these days :)

[Mark-Simulacrum]

T-lang discussed this in today's meeting and agreed that this can move ahead. We may eventually want to make the must_use have behavior related to return of the .await'ing the resulting future, but that's not happening immediately (if ever), so we can move ahead here in the meantime.

We briefly discussed my concern around unused_attributes and generally felt positive about its use here, I think.

[guswynn]

@wesleywiser what review is outstanding?

[wesleywiser]

@guswynn I believe this is still waiting on Mark's feedback to be addressed:

it seems like this should be an unused_attributes lint, not an unconditional warning (and a custom message is still fine).

[guswynn]

@wesleywiser sorry about the delay, was on a break, this should now be ready!

[wesleywiser]

Looks great, thanks @guswynn!

[wesleywiser]

@bors r+

[bors]

📌 Commit 958de5a has been approved by wesleywiser

[wesleywiser]

Blessed tests.

@bors r+

[bors]

📌 Commit 83ce771 has been approved by wesleywiser

[guswynn]

@wesleywiser you are a legend for blessing those tests, sorry this fell off my radar!

[wesleywiser]

@guswynn Thanks for working on it!! 🎉