Implement RFC 3373: Avoid non-local definitions in functions

[Urgau]

This PR implements RFC 3373: Avoid non-local definitions in functions.

[rustbot]

r? @petrochenkov

(rustbot has picked a reviewer for you, use r? to override)

[rustbot]

Some changes occurred in src/tools/clippy

cc @rust-lang/clippy

[rustbot]

Some changes occurred in src/tools/rustfmt

cc @rust-lang/rustfmt

[Urgau]

@Dylan-DPC I appreciate your enthusiasm, but the FCP isn't finished yet (no finished-final-comment-period The final comment period is finished for this PR / Issue. ), it will finish this saturday.

@bors r-

[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.

[Urgau]

@bors r=@WaffleLapkin

[bors]

📌 Commit 63469ab has been approved by WaffleLapkin

It is now in the queue for this repository.

[bors]

⌛ Testing commit 63469ab with merge b0d3e04...

[bors]

☀️ Test successful - checks-actions
Approved by: WaffleLapkin
Pushing b0d3e04 to master...

[rust-timer]

Finished benchmarking commit (b0d3e04): comparison URL.

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

Next Steps: If you can justify the regressions found in this perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please open an issue or create a new PR that fixes the regressions, add a comment linking to the newly created issue or PR, and then add the perf-regression-triaged label to this PR.

@rustbot label: +perf-regression
cc @rust-lang/wg-compiler-performance

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)	2.2%	[0.5%, 4.2%]	12
Regressions ❌ (secondary)	0.7%	[0.1%, 1.8%]	5
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.4%	[-0.8%, -0.1%]	27
All ❌✅ (primary)	2.2%	[0.5%, 4.2%]	12

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)	1.8%	[0.8%, 4.7%]	4
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.0%	[-2.0%, -2.0%]	1
All ❌✅ (primary)	1.8%	[0.8%, 4.7%]	4

Cycles

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)	2.7%	[2.3%, 3.1%]	6
Regressions ❌ (secondary)	1.0%	[1.0%, 1.0%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.7%	[2.3%, 3.1%]	6

Binary size

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

Bootstrap: 650.73s -> 651.083s (0.05%)
Artifact size: 311.11 MiB -> 311.22 MiB (0.03%)

[rylev]

For perf triage I'm going to mark this as triaged since #121625 is meant to try to address this. However, I hope that PR is currently only recovering a small fraction of the performance loss here. Given the complex nature of the change, I imagine a revert is not desirable though even if performance can't be gained back.

@rustbot label: +perf-regression-triaged

[Urgau]

@rylev I tried very hard in #121625 to fix the perf regression, but it when mostly nowhere. My latest (and most promising) attempt end up having a correctness issue.

At some point I was wondering why diesel was the only benchmark to show any regressions (even in incremental), just to realize that in the Crater run some older versions of the crate were affect by the lint; and indeed it now produces 155 warnings!

No wonder why my all attempts went no-where.

If only I could have gotten kcachegrind working earlier it would have screamed this to me; it shows that 99.39% the instructions in the lint is the diagnostic machinery.

So in summary, there is no perf-regression per se, rustc is now just doing way more work because it has to display more than 155 different warnings.

Details

For those curious: