`Vec::is_empty` is not marked as `#[inline]`

[EnricoGhiorzi]

Similar methods in Vec are annotated with #[inline] including Vec::len but Vec::is_empty is not. Looking at the relevant policy it seems to me is_empty satisfies all the requirements to be inlined. Is this deliberate or it could be beneficial to add it?

To be more precise, I am referring to this function: https://doc.rust-lang.org/nightly/src/alloc/vec/mod.rs.html#3040.
Curiously, this other implementation of is_empty for slices is inline: https://doc.rust-lang.org/nightly/src/core/slice/mod.rs.html#136

[xtqqczze]

The policy specifically states: "You shouldn't need #[inline]: On methods that have any generics in scope."

[Human9000-bit]

@rustbot label A-collections C-enhancement -needs-triage

[xtqqczze]

The policy specifically states: "You shouldn't need #[inline]: On methods that have any generics in scope."

Nevertheless, this could be beneficial, see #102539

[EnricoGhiorzi]

The policy specifically states: "You shouldn't need #[inline]: On methods that have any generics in scope."

That's true, I saw no generics in the function signature and forgot that there was one in scope.

[RalfJung]

Sufficiently small functions are automatically marked as inline precisely so that we don't have to spam the attribute everywhere, so adding the attribute is probably a NOP. Cc @saethlin

[saethlin]

I added that logic to reduce the chance that you get goofy codegen if you forget the attribute, not to reduce use of the attribute. The "sufficiently small" criteria is unintuitive IMO, for example it didn't occur to me that it should special-case intrinsics: #145910 and periodically I get a mention on the community Discord because people expect that functions that just delegate to a big function will be automatically marked, but they are deliberately not.

[saethlin]

In this case specifically I do not have a strong opinion on using the attribute explicitly.