Given the following code: [playground]
#[derive(Default)]
struct T {}
struct U {}The current output is:
warning: struct `U` is never constructed
--> src/lib.rs:4:8
|
4 | struct U {}
| ^
|
= note: `#[warn(dead_code)]` on by default
Ideally the output should look like:
warning: struct `T` is never constructed
--> src/lib.rs:2:8
|
2 | struct T {}
| ^
|
= note: `#[warn(dead_code)]` on by default
warning: struct `U` is never constructed
--> src/lib.rs:4:8
|
4 | struct U {}
| ^
After the derive, the code looks roughly like
struct T {}
#[automatically_derived]
impl Default for T {
#[inline]
fn default() -> T { T {} }
}
struct U {}As the default implementation is both 1) itself unused and 2) #[automatically_derived], ideally it should not count as a use for suppressing the dead_code lint on T.
The derived implementations for Clone, Debug, PartialEq, and Hash also show this behavior.
I seem to recall a previous change to #[automatically_derived] to change the interaction between the unused code lint and derived implementations, but could not find it offhand.
Given the following code: [playground]
The current output is:
Ideally the output should look like:
After the derive, the code looks roughly like
As the default implementation is both 1) itself unused and 2)
#[automatically_derived], ideally it should not count as a use for suppressing thedead_codelint onT.The derived implementations for
Clone,Debug,PartialEq, andHashalso show this behavior.I seem to recall a previous change to
#[automatically_derived]to change the interaction between the unused code lint and derived implementations, but could not find it offhand.