Clarify what -D warnings or -F warnings does

[tbu-]

They set all lints currently on the warning level to deny or forbid,
respectively.

[rust-highfive]

r? @pnkfelix

(rust_highfive has picked a reviewer for you, use r? to override)

[tbu-]

Tested what it actually does by running rustc on the following file

pub fn main() {
Box::new(());
}

with the lint box-pointers.

[shepmaster]

Ping from triage @pnkfelix — will you have some time to review this?

[nikomatsakis]

I think we can do a bit better.

[nikomatsakis]

Hmm, I don't find this phrasing very clear. "The warning level" sounds to me like something that one can control independently. Perhaps "all lints set to warn", or "all lints that are set to issue warnings"?

[nikomatsakis]

If you have a lint that defaults to allow, but you set it explicitly to warn, and then you do #[deny(warnings)], is that lint then denied?

[tbu-]

Hmm, I don't find this phrasing very clear.

Yea, I struggled to find a good wording. Thanks for the suggestions, I used the second one for now.

If you have a lint that defaults to allow, but you set it explicitly to warn, and then you do #[deny(warnings)], is that lint then denied?

Yes.

#![deny(warnings)]
fn main() {
Box::new(());
}

$ rustc -W box-pointers a.rs
error: type uses owned (Box type) pointers: std::boxed::Box<()>
--> box.rs:3:5
|
3 | Box::new(());
| ^^^^^^^^^^^^
|
note: lint level defined here
--> box.rs:1:9
|
1 | #![deny(warnings)]
| ^^^^^^^^
= note: #[deny(box_pointers)] implied by #[deny(warnings)]
error: aborting due to previous error

[nikomatsakis]

r? @nikomatsakis

[nikomatsakis]

@bors r+ rollup

[bors]

📌 Commit e1e1dcc has been approved by nikomatsakis

[bors]

⌛ Testing commit e1e1dcc with merge 3de30dd00edbc89613e93900a48ddfa4f8eb06a7...

[kennytm]

@bors retry — Prioritize for rollup (which included this PR...)