Folder experiment: Micro-optimize RegionEraserVisitor

[compiler-errors]

NOTE: This is one of a series of perf experiments that I've come up with while sick in bed. I'm assigning them to lqd b/c you're a good reviewer and you'll hopefully be awake when these experiments finish, lol.

r? lqd

The region eraser is very hot, so let's see if we can avoid erasing types (and visiting consts and preds that don't have region-ful types) unnecessarily.

[compiler-errors]

@bors try @rust-timer queue

[bors]

⌛ Trying commit e9dfc29 with merge fd6ccb9...

[bors]

💔 Test failed - checks-actions

[compiler-errors]

@bors try @rust-timer queue

[bors]

⌛ Trying commit 9434486 with merge a3e9172...

[bors]

☀️ Try build successful - checks-actions
Build commit: a3e9172 (a3e9172fc2064003b1878cdafd40dc687a7464ad)

[rust-timer]

Finished benchmarking commit (a3e9172): comparison URL.

Overall result: ✅ improvements - no 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.

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

Instruction count

This is the most reliable metric that we have; it was used to determine the overall result at the top of this comment. However, even this metric can sometimes exhibit noise.

	mean	range	count
Regressions ❌ (primary)	0.2%	[0.2%, 0.2%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-2.7%	[-34.8%, -0.2%]	16
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-2.5%	[-34.8%, 0.2%]	17

Max RSS (memory usage)

Results (primary 2.0%)

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)	2.0%	[1.2%, 2.6%]	3
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.0%	[1.2%, 2.6%]	3

Cycles

Results (primary -8.0%, secondary 3.2%)

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)	3.2%	[3.2%, 3.2%]	1
Improvements ✅ (primary)	-8.0%	[-15.2%, -0.7%]	2
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-8.0%	[-15.2%, -0.7%]	2

Binary size

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

Bootstrap: 777.539s -> 774.134s (-0.44%)
Artifact size: 365.91 MiB -> 365.91 MiB (-0.00%)

[compiler-errors]

35% improvement on nalgebra (but just one suite lol) is kinda crazy, surprising it doesn't generalize though

[lqd]

What the.

So we are currently spending 23% of the time in nalgebra checking for possible inlining cycles 🤯

[lqd]

badass

@bors r+

[bors]

📌 Commit 9434486 has been approved by lqd

It is now in the queue for this repository.

[bors]

⌛ Testing commit 9434486 with merge 0c478fd...

[bors]

☀️ Test successful - checks-actions
Approved by: lqd
Pushing 0c478fd to master...

[github-actions]

What is this?

This is an experimental post-merge analysis report that shows differences in test outcomes between the merged PR and its parent PR.

Comparing 0e9c3e5 (parent) -> 0c478fd (this PR)

Test differences

No test diffs found

Job duration changes

1. dist-apple-various: 5890.7s -> 7350.3s (24.8%)
2. dist-x86_64-apple: 8319.5s -> 9927.8s (19.3%)
3. x86_64-gnu-stable: 6609.2s -> 7116.7s (7.7%)
4. dist-armhf-linux: 4989.6s -> 5327.0s (6.8%)
5. dist-i686-msvc: 6751.6s -> 7115.2s (5.4%)
6. dist-i686-linux: 5995.8s -> 6266.7s (4.5%)
7. arm-android: 5249.4s -> 5378.1s (2.5%)
8. dist-aarch64-apple: 4478.4s -> 4584.5s (2.4%)
9. x86_64-gnu-llvm-18-3: 6765.7s -> 6909.7s (2.1%)
10. dist-x86_64-freebsd: 4854.6s -> 4948.2s (1.9%)

How to interpret the job duration changes?

Job durations can vary a lot, based on the actual runner instance
that executed the job, system noise, invalidated caches, etc. The table above is provided
mostly for t-infra members, for simpler debugging of potential CI slow-downs.

[rust-timer]

Finished benchmarking commit (0c478fd): comparison URL.

Overall result: ✅ improvements - no action needed

@rustbot label: -perf-regression

Instruction count

This is the most reliable metric that we have; it was used to determine the overall result at the top of this comment. However, even this metric can sometimes exhibit noise.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-4.7%	[-34.9%, -0.4%]	9
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-4.7%	[-34.9%, -0.4%]	9

Max RSS (memory usage)

Results (primary 2.5%, secondary -0.9%)

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)	2.5%	[2.4%, 2.5%]	2
Regressions ❌ (secondary)	2.6%	[2.4%, 2.7%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-4.3%	[-5.2%, -3.5%]	2
All ❌✅ (primary)	2.5%	[2.4%, 2.5%]	2

Cycles

Results (primary -8.0%, secondary 4.9%)

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)	4.9%	[4.9%, 4.9%]	2
Improvements ✅ (primary)	-8.0%	[-15.2%, -0.8%]	2
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-8.0%	[-15.2%, -0.8%]	2

Binary size

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

Bootstrap: 776.909s -> 775.534s (-0.18%)
Artifact size: 365.93 MiB -> 365.93 MiB (-0.00%)