Add impl const for Clone

[Randl]

#142757

[rustbot]

r? @Mark-Simulacrum

rustbot has assigned @Mark-Simulacrum.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[compiler-errors]

This is fine, though it doesn't fix other things like (T,): [const] Clone.

It would be nice if you added tests for this, either in tests/ui/traits/const-traits or as a library test.

[compiler-errors]

@rustbot author

r? compiler-errors

[rustbot]

Reminder, once the PR becomes ready for a review, use @rustbot ready.

[Randl]

There is a problem that I'm not sure is a bug, which locks the constifying of other impls. If I add derive_const to something, in lib I get an error

error: const stability on the impl does not match the const stability on the trait
   --> library/core/src/ascii/ascii_char.rs:59:16
    |
59  |   #[derive_const(Clone)]
    |                  ^^^^^ in this derive macro expansion
    |
   ::: library/core/src/clone.rs:225:1
    |
225 | / pub macro Clone($item:item) {
226 | |     /* compiler built-in */
227 | | }
    | |_- in this expansion of `#[derive(Clone)]`
    |
note: this impl is (implicitly) stable...
   --> library/core/src/ascii/ascii_char.rs:59:16
    |
59  |   #[derive_const(Clone)]
    |                  ^^^^^ in this derive macro expansion
    |
   ::: library/core/src/clone.rs:225:1
    |
225 | / pub macro Clone($item:item) {
226 | |     /* compiler built-in */
227 | | }
    | |_- in this expansion of `#[derive(Clone)]`
note: ...but the trait is unstable
   --> library/core/src/clone.rs:162:11
    |
162 | pub trait Clone: Sized {
    |           ^^^^^

even though I add

#[rustc_const_unstable(feature = "const_clone", issue = "142757")]
#[derive_const(Clone)]

I think the interplay of const stability and derive_const is bugged, but I'm not sure

[Randl]

@rustbot ready

[Randl]

@rustbot ready

[Randl]

I've added some more impls

[Randl]

Couldn't find the Clone impl for tuple tho

[Randl]

rebased on master for the new syntax

[compiler-errors]

As I mentioned in the other PR, could you limit the scope of this?

#144847 (comment)

Specifically, I'd like to just see us constifying a few things at a time. Other constifications for more library-specific types should probably have a library reviewer instead and be done on a per-type basis.

[compiler-errors]

@rustbot author

[Randl]

Can you be a bit more specific so we're on the same page?
This implements Clone for primitives, references, pointers, ControlFlow, tuples, ranges, slices. These all feel like a stuff we'd want to have const. PhantomData, bignum, and niche types probably should be dropped from this PR?

[rustbot]

This PR was rebased onto a different master commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.

[Randl]

@rustbot ready

I've dropped the impls I mentioned in previous comment