rustdoc: Speed up `TypeAliasPart::get` #141421

nnethercote · 2025-05-23T05:50:16Z

Big speedups here on a couple of the benchmarks.

r? @GuillaumeGomez

nnethercote · 2025-05-23T06:55:28Z

@bors try @rust-timer queue

rustdoc: Speed up `TypeAliasPart::get` r? `@ghost`

bors · 2025-05-23T06:56:39Z

⌛ Trying commit e53bbf9 with merge 36aa0e9...

bors · 2025-05-23T09:05:00Z

☀️ Try build successful - checks-actions
Build commit: 36aa0e9 (36aa0e98dcc96e3382dd186aef9b16a4d214bfe0)

rust-timer · 2025-05-23T11:42:43Z

Finished benchmarking commit (36aa0e9): 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
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-19.5%	[-78.7%, -0.3%]	8
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-19.5%	[-78.7%, -0.3%]	8

Max RSS (memory usage)

Results (primary 11.5%, secondary 7.3%)

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)	11.5%	[11.5%, 11.5%]	1
Regressions ❌ (secondary)	7.3%	[7.3%, 7.3%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	11.5%	[11.5%, 11.5%]	1

Cycles

Results (primary -47.4%)

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)	-	-	0
Improvements ✅ (primary)	-47.4%	[-77.2%, -6.1%]	3
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-47.4%	[-77.2%, -6.1%]	3

Binary size

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

Bootstrap: 778.581s -> 777.715s (-0.11%)
Artifact size: 365.44 MiB -> 365.46 MiB (0.00%)

nnethercote · 2025-05-23T12:32:05Z

Details in the individual commits.

GuillaumeGomez · 2025-05-23T12:33:56Z

It's super cool, well done!

@bors r+ rollup=iffy

bors · 2025-05-23T12:33:59Z

📌 Commit e53bbf9 has been approved by GuillaumeGomez

It is now in the queue for this repository.

bors · 2025-05-23T12:33:59Z

🌲 The tree is currently closed for pull requests below priority 100. This pull request will be tested once the tree is reopened.

src/librustdoc/html/render/write_shared.rs

Kobzol · 2025-05-23T18:03:13Z

Awesome result! I wonder if we could somehow automate finding situations where we do a bunch of stuff and then throw away the result without using it. It almost sounds like a simple version of that could be a Clippy lint (this variable is unused in a dataflow path), but that would probably have too many false positives. Or some runtime instrumentation pass that marks variables unused (or only Dropped) after being created :)

The comparison against `text` seems to be unnecessary.

This is a huge perf win for rustdoc on the `typenum` and `nalgebra` benchmarks, because the `else` branch doesn't get hit much.

nnethercote · 2025-05-23T21:50:39Z

It almost sounds like a simple version of that could be a Clippy lint (this variable is unused in a dataflow path), but that would probably have too many false positives

Sounds like a great idea. @blyxyas suggested the same thing to me.

nnethercote · 2025-05-23T21:51:30Z

@bors r=GuillaumeGomez rollup=never

bors · 2025-05-23T21:51:32Z

📌 Commit dfe8fe8 has been approved by GuillaumeGomez

It is now in the queue for this repository.

bors · 2025-05-24T03:37:16Z

⌛ Testing commit dfe8fe8 with merge 105354d...

bors · 2025-05-24T06:48:41Z

☀️ Test successful - checks-actions
Approved by: GuillaumeGomez
Pushing 105354d to master...

github-actions · 2025-05-24T06:51:25Z

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 80c3498 (parent) -> 105354d (this PR)

Test differences

No test diffs found

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard 105354db48223b7bf799b05d4dd3a412aca7bd05 --output-dir test-dashboard

And then open test-dashboard/index.html in your browser to see an overview of all executed tests.

Job duration changes

x86_64-apple-1: 5743.2s -> 7847.0s (36.6%)
dist-apple-various: 6695.7s -> 8706.1s (30.0%)
aarch64-apple: 3926.6s -> 3171.5s (-19.2%)
dist-armv7-linux: 5403.1s -> 6092.2s (12.8%)
dist-x86_64-apple: 7707.3s -> 8288.8s (7.5%)
dist-aarch64-linux: 5328.0s -> 5689.8s (6.8%)
x86_64-rust-for-linux: 2741.4s -> 2557.4s (-6.7%)
x86_64-msvc-ext1: 7372.7s -> 6960.1s (-5.6%)
dist-s390x-linux: 5432.4s -> 5154.9s (-5.1%)
x86_64-gnu-aux: 4467.3s -> 4282.1s (-4.1%)

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 · 2025-05-24T08:19:17Z

Finished benchmarking commit (105354d): comparison URL.

Overall result: ❌✅ regressions and improvements - please read the text below

Our benchmarks found a performance regression caused by this PR.
This might be an actual regression, but it can also be just noise.

Next Steps:

If the regression was expected or you think it can be justified,
please write a comment with sufficient written justification, and add
@rustbot label: +perf-regression-triaged to it, to mark the regression as triaged.
If you think that you know of a way to resolve the regression, try to create
a new PR with a fix for the regression.
If you do not understand the regression or you think that it is just noise,
you can ask the @rust-lang/wg-compiler-performance working group for help (members of this group
were already notified of this PR).

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

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)	0.2%	[0.2%, 0.2%]	1
Improvements ✅ (primary)	-19.5%	[-78.7%, -0.3%]	8
Improvements ✅ (secondary)	-0.2%	[-0.2%, -0.2%]	1
All ❌✅ (primary)	-19.5%	[-78.7%, -0.3%]	8

Max RSS (memory usage)

Results (secondary 2.8%)

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)	2.8%	[2.8%, 2.8%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Cycles

Results (primary -47.5%)

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)	-	-	0
Improvements ✅ (primary)	-47.5%	[-77.2%, -6.1%]	3
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-47.5%	[-77.2%, -6.1%]	3

Binary size

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

Bootstrap: 778.415s -> 777.008s (-0.18%)
Artifact size: 366.33 MiB -> 366.29 MiB (-0.01%)

lqd · 2025-05-24T08:59:55Z

Pure rustdoc change, so coercions noise.

@rustbot label: +perf-regression-triaged

rustbot added T-rustdoc Relevant to the rustdoc team, which will review and decide on the PR/issue. T-rustdoc-frontend Relevant to the rustdoc-frontend team, which will review and decide on the web UI/UX output. labels May 23, 2025

This comment has been minimized.

Sign in to view

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label May 23, 2025

bors added a commit to rust-lang-ci/rust that referenced this pull request May 23, 2025

Auto merge of rust-lang#141421 - nnethercote:TypeAliasPart-get, r=<try>

36aa0e9

rustdoc: Speed up `TypeAliasPart::get` r? `@ghost`

This comment has been minimized.

Sign in to view

rustbot removed the S-waiting-on-perf Status: Waiting on a perf run to be completed. label May 23, 2025

rustbot assigned GuillaumeGomez May 23, 2025

nnethercote marked this pull request as ready for review May 23, 2025 12:30

rustbot added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label May 23, 2025

GuillaumeGomez approved these changes May 23, 2025

View reviewed changes

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels May 23, 2025

yotamofek reviewed May 23, 2025

View reviewed changes

src/librustdoc/html/render/write_shared.rs Outdated Show resolved Hide resolved

nnethercote added 3 commits May 24, 2025 07:31

Simplify the "is some" test in TypeAliasPart::get.

a7baf4b

The comparison against `text` seems to be unnecessary.

Move code inside the else in TypeAliasPart::get.

e01f407

This is a huge perf win for rustdoc on the `typenum` and `nalgebra` benchmarks, because the `else` branch doesn't get hit much.

Simplify things a little more.

dfe8fe8

nnethercote force-pushed the TypeAliasPart-get branch from e53bbf9 to dfe8fe8 Compare May 23, 2025 21:48

bors added the merged-by-bors This PR was explicitly merged by bors. label May 24, 2025

bors merged commit 105354d into rust-lang:master May 24, 2025
7 checks passed

rustbot added this to the 1.89.0 milestone May 24, 2025

rustbot added the perf-regression Performance regression. label May 24, 2025

rustbot added the perf-regression-triaged The performance regression has been triaged. label May 24, 2025

nnethercote deleted the TypeAliasPart-get branch May 25, 2025 22:11

rustdoc: Speed up TypeAliasPart::get #141421

rustdoc: Speed up TypeAliasPart::get #141421

Conversation

nnethercote commented May 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

nnethercote commented May 23, 2025

Uh oh!

This comment has been minimized.

bors commented May 23, 2025

Uh oh!

bors commented May 23, 2025

Uh oh!

This comment has been minimized.

rust-timer commented May 23, 2025

Overall result: ✅ improvements - no action needed

Uh oh!

nnethercote commented May 23, 2025

Uh oh!

GuillaumeGomez commented May 23, 2025

Uh oh!

bors commented May 23, 2025

Uh oh!

bors commented May 23, 2025

Uh oh!

Uh oh!

Kobzol commented May 23, 2025

Uh oh!

nnethercote commented May 23, 2025

Uh oh!

nnethercote commented May 23, 2025

Uh oh!

bors commented May 23, 2025

Uh oh!

bors commented May 24, 2025

Uh oh!

bors commented May 24, 2025

Uh oh!

Uh oh!

github-actions bot commented May 24, 2025

Test differences

Job duration changes

Uh oh!

rust-timer commented May 24, 2025

Overall result: ❌✅ regressions and improvements - please read the text below

Uh oh!

lqd commented May 24, 2025

Uh oh!

Uh oh!

rustdoc: Speed up `TypeAliasPart::get` #141421

rustdoc: Speed up `TypeAliasPart::get` #141421

nnethercote commented May 23, 2025 •

edited

Loading