rewrite ensure_drop_params_and_item_params_correspond

[lcnr]

actually relating types here seems like it's overkill

[lcnr]

was ignored by highfive

r? rust-lang/compiler

[nagisa]

I wonder if try works with conflicts in place...

@bors try @rust-timer queue

[rust-timer]

Awaiting bors try build completion.

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

[nagisa]

Largely LGTM, though I'm not familiar enough with dropck to tell if there aren't any hidden potholes we need to be looking out for by making such a change.

[lcnr]

r? @nikomatsakis for 2b982e006bf568b70c7aaa04745b3bc5e2de81d5 in that case

[nikomatsakis]

Hmm.

So I don't like the syntactic comparison for equality that is throughout this code -- I rather prefer the older code, and I would rather express the correctness conditions logically. Certainly in a-mir-formality that is what I expect us to do. (Something like: for all instances of the struct, the drop impl conditions are met.)

That said, the code is cleaner this way, and the only case I could come up with where using syntactic checks would be a problem seems to already be rejected. I consider that a bug, but it's pre-existing and more of a problem with the other drop code.

struct Foo<T>
where
    T: Iterator,
    T::Item: Send,
{
    t: T,
}

impl<T, I> Drop for Foo<T>
where
    T: Iterator<Item = I>,
    I: Send,
{
    fn drop(&mut self) { }
}

fn main() { }

Do we have a test like this one? If not, can we add one? Otherwise, r=me.

[nikomatsakis]

@lcnr and I discussed and we will move this test to another PR, r=me as is

[lcnr]

@bors r=nikomatsakis

[bors]

📌 Commit 4a82bc9 has been approved by nikomatsakis

[bors]

⌛ Testing commit 4a82bc9 with merge 512a328...

[bors]

☀️ Test successful - checks-actions
Approved by: nikomatsakis
Pushing 512a328 to master...

[rust-timer]

Finished benchmarking commit (512a328): comparison url.

Instruction count

This benchmark run did not return any relevant results for this metric.

Max RSS (memory usage)

Results

• Primary benchmarks: mixed results
• Secondary benchmarks: mixed results

	Regressions 😿 (primary)	Regressions 😿 (secondary)	Improvements 🎉 (primary)	Improvements 🎉 (secondary)	All 😿 🎉 (primary)
count1	1	2	1	1	2
mean2	5.7%	2.0%	-5.2%	-3.6%	0.3%
max	5.7%	2.4%	-5.2%	-3.6%	5.7%

Cycles

Results

• Primary benchmarks: 🎉 relevant improvements found
• Secondary benchmarks: mixed results

	Regressions 😿 (primary)	Regressions 😿 (secondary)	Improvements 🎉 (primary)	Improvements 🎉 (secondary)	All 😿 🎉 (primary)
count1	0	1	2	1	2
mean2	N/A	4.0%	-2.0%	-4.6%	-2.0%
max	N/A	4.0%	-2.3%	-4.6%	-2.3%

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

@rustbot label: -perf-regression

Footnotes

1. number of relevant changes ↩ ↩²

2. the arithmetic mean of the percent change ↩ ↩²