Add REDUNDANT_LIFETIMES lint to detect lifetimes which are semantically redundant

[compiler-errors]

There already is a UNUSED_LIFETIMES lint which is fired when we detect where clause bounds like where 'a: 'static, however, it doesn't use the full power of lexical region resolution to detect failures.

Right now UNUSED_LIFETIMES is an Allow lint, though presumably we could bump it to warn? I can (somewhat) easily implement a structured suggestion so this can be rustfix'd automatically, since we can just walk through the HIR body, replacing instances of the redundant lifetime.

Fixes #118376
r? lcnr

[compiler-errors]

The only wrinkle is that this approach may be detrimental to performance...

@bors try @rust-timer queue

[bors]

⌛ Trying commit b2c18d9 with merge 58aa0f0...

[bors]

💔 Test failed - checks-actions

[compiler-errors]

@bors try

[bors]

⌛ Trying commit d528fad with merge 7a2c7d7...

[bors]

☀️ Try build successful - checks-actions
Build commit: 7a2c7d7 (7a2c7d785725785a467e2dc491986536eef554b5)

[rust-timer]

Finished benchmarking commit (7a2c7d7): comparison URL.

Overall result: ❌ regressions - ACTION NEEDED

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. While you can manually mark this PR as fit for rollup, we strongly recommend not doing so since this PR may lead to changes in compiler perf.

Next Steps: If you can justify the regressions found in this try perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please fix the regressions and do another perf run. If the next run shows neutral or positive results, the label will be automatically removed.

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

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	0.5%	[0.2%, 1.4%]	32
Regressions ❌ (secondary)	0.5%	[0.0%, 0.8%]	16
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.5%	[0.2%, 1.4%]	32

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	3.9%	[3.9%, 3.9%]	1
Improvements ✅ (primary)	-2.3%	[-2.3%, -2.3%]	1
Improvements ✅ (secondary)	-3.0%	[-4.1%, -2.0%]	2
All ❌✅ (primary)	-2.3%	[-2.3%, -2.3%]	1

Cycles

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

Binary size

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

Bootstrap: 675.422s -> 674.91s (-0.08%)
Artifact size: 313.36 MiB -> 313.46 MiB (0.03%)

[aliemjay]

few notes:

• I think we should not lint if the "victim" lifetime is elided. The current impl lints in this case but it should not imo:

struct Ty<T: static>(T);
fn test(_: Ty<&str>) {}

• There is no need to fork the inference context and do region resolution for each check, you can simply use outlives_env.free_region_map.sub_free_regions() to check for equality.

• To avoid the perf impact, can you try to run the lint in enter_wf_checking_ctxt where we've already got the OutlivesEnvironment?

[compiler-errors]

I will try adding this into WF if perf still comes back bad...

@bors try @rust-timer queue

[bors]

⌛ Trying commit b09e25e with merge f5059fb...

[rfcbot]

The final comment period, with a disposition to merge, as per the review above, is now complete.

As the automated representative of the governance process, I would like to thank the author for their work and everyone else who contributed.

This will be merged soon.

[lcnr]

r=me after nits

[matthiaskrgr]

How different is this implementation from the preexisting clippy::needless_lifetimes lint?
https://rust-lang.github.io/rust-clippy/master/index.html#/needless_lifetimes

[lcnr]

this is about named lifetimes which are equal to other named lifetimes, e.g. in the followiuinig example 'b and 'a are actually equal due to implied bounds

struct Foo<'a>(&'a ());
impl<'a> Foo<'a> {
    fn use<'b: 'a>(&'b self, s: &'b str) {}
}

[lcnr]

thank you for implementing this ❤️ I would like to move both lifetime lints to WARN by default in a separate PR soon 🤔

@bors r+ rollup

[bors]

📌 Commit da2b714 has been approved by lcnr

It is now in the queue for this repository.