Fix typo and improve documentation for E0632

[FabianWolff]

Edit: After #85520 (comment), this PR has been boiled down to just an extended description for E0632 and a fixed typo.

[rust-highfive]

Some changes occurred in diagnostic error codes

cc @GuillaumeGomez

[rust-highfive]

r? @jackh726

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

[cjgillot]

You may need to filter out the usage of the const generic parameter to specify a parameter's type:

fn foo<const N: usize>(_: [i32; N], _: impl Copy) {}

struct Boo<const N: i32>{ ... }
fn bar<const N: i32>(_: Boo<i32>, _: impl Debug) {}

[FabianWolff]

You may need to filter out the usage of the const generic parameter to specify a parameter's type:

You are right, of course. Thanks for pointing this out, it should be fixed now.

[bors]

☔ The latest upstream changes (presumably #86588) made this pull request unmergeable. Please resolve the merge conflicts.

[jackh726]

Sorry for the delay in review here. This looks great. I have two thoughts:

1. It'll be interesting if Implement a explicit_generic_args_with_impl_trait feature gate #86176 lands, since that feature gate essentially means this lint isn't needed anymore. But I think in the meantime, it's worth landing this.
2. I'd like a perf run after this is rebase. It is an extra ast pass, so I expect maybe a little bit of a regression; but I want to put a number on it before it lands.

Other than that, r=me after perf

[FabianWolff]

I'd like a perf run after this is rebase.

Rebased on what? I did a rebase about half an hour ago on the current master branch.

[jackh726]

Oh whoops. Did see that it was already rebased. 🤦‍♀️

@bors try @rust-timer queue

[rust-timer]

Awaiting bors try build completion.

@rustbot label: +S-waiting-on-perf

[bors]

⌛ Trying commit 3ff3c614cefb484ad6b209d4518be0b37cbb578d with merge d52bd326fc9e84dc83a60bfbaa9155ee6dc47cc3...

[bors]

☀️ Try build successful - checks-actions
Build commit: d52bd326fc9e84dc83a60bfbaa9155ee6dc47cc3 (d52bd326fc9e84dc83a60bfbaa9155ee6dc47cc3)

[rust-timer]

Queued d52bd326fc9e84dc83a60bfbaa9155ee6dc47cc3 with parent 456a032, future comparison URL.

[rust-timer]

Finished benchmarking try commit (d52bd326fc9e84dc83a60bfbaa9155ee6dc47cc3): comparison url.

Summary: ERROR categorizing benchmark run!

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. Please note that if the perf results are neutral, you should likely undo the rollup=never given below by specifying rollup- to bors.

Importantly, though, if the results of this run are non-neutral do not roll this PR up -- it will mask other regressions or improvements in the roll up.

@bors rollup=never
@rustbot label: +S-waiting-on-review -S-waiting-on-perf -perf-regression

[FabianWolff]

I'm not sure what to make of these results (the comparison URL link seems to be broken: "could not find start commit for bound ..."), but in any case, I have added an early return for functions that don't have const parameters, and if no impl Trait argument types were found. I've also made the search for impl Trait arguments more robust to catch cases like Option<impl Trait>.

[GuillaumeGomez]

I guess it's a bug. Let's run another one and hope it works this time.

@bors try @rust-timer queue

[rust-timer]

Awaiting bors try build completion.

@rustbot label: +S-waiting-on-perf

[bors]

⌛ Trying commit 06824f63ec4c3787719b7a7b7ad0e066540bc307 with merge f2285179271f79ab0fc43ae08cbd2d1d727c515f...

[bors]

☀️ Try build successful - checks-actions
Build commit: f2285179271f79ab0fc43ae08cbd2d1d727c515f (f2285179271f79ab0fc43ae08cbd2d1d727c515f)

[rust-timer]

Queued f2285179271f79ab0fc43ae08cbd2d1d727c515f with parent 17ea490, future comparison URL.

[rust-timer]

Finished benchmarking try commit (f2285179271f79ab0fc43ae08cbd2d1d727c515f): comparison url.

Summary: This benchmark run did not return any significant changes.

If you disagree with this performance assessment, please file an issue in rust-lang/rustc-perf.

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. Please note that if the perf results are neutral, you should likely undo the rollup=never given below by specifying rollup- to bors.

Importantly, though, if the results of this run are non-neutral do not roll this PR up -- it will mask other regressions or improvements in the roll up.

@bors rollup=never
@rustbot label: +S-waiting-on-review -S-waiting-on-perf -perf-regression

[jackh726]

@FabianWolff I guess the policy is a compiler MCP and lang FCP for new lints (see https://rust-lang.zulipchat.com/#narrow/stream/131828-t-compiler/topic/New.20lint.20.28declaring.20an.20uncallable.20function.29.20.28.2385520.29/near/244202217).

@petrochenkov noted that the cost/benefit of this is too high

Can you make an MCP if you think this is worth it and see where that goes?

[FabianWolff]

@FabianWolff I guess the policy is a compiler MCP and lang FCP for new lints

Ah, I didn't know it is such a big deal. No, I don't think it is worth it here.

I have removed all of my changes except for a fixed typo and the extended description for E0632 now; how do you feel about merging just that?

[jackh726]

@bors r+ rollup

[bors]

📌 Commit 69c8573 has been approved by jackh726