Optimize Clippy with PGO by Kobzol · Pull Request #159642 · rust-lang/rust

Kobzol · 2026-07-21T08:31:08Z

After rustdoc and Cargo, this seemed like an obvious target for PGO.

Perf results: https://perf.rust-lang.org/compare.html?start=390279b302ca98ae270f434100ae3730531d1246&end=caf67342a544d6eab7a40ebf0d02e802f62008a4&stat=instructions:u

The Clippy benchmarks are not that fast, the full time duration is now ~3h 20m. Maybe we should wait for Mark's new try build hardware before landing this, but it is borderline.

Kobzol · 2026-07-21T08:31:45Z

@bors try

Optimize Clippy with PGO

Kobzol · 2026-07-21T08:32:07Z

@rust-timer queue profiles=Check,Debug,Opt,Doc,Clippy

rust-bors · 2026-07-21T10:17:27Z

💔 Test for 2dc0eb1 failed: CI. Failed job:

try - dist-x86_64-linux-quick (web logs, enhanced plaintext logs)

Kobzol · 2026-07-21T11:16:04Z

@bors try

Optimize Clippy with PGO

rust-bors · 2026-07-21T13:36:35Z

☀️ Try build successful (CI)
Build commit: 3542ef2 (3542ef25e63b309ca2515102fb81bc590cbabcef)
Base parent: cde3f8a (cde3f8aee5a30928872e2438e3f18238a3fa306c)

rust-timer · 2026-07-21T20:52:12Z

Finished benchmarking commit (3542ef2): comparison URL.

Overall result: no relevant changes - no action needed

Benchmarking means the PR may be perf-sensitive. Consider adding rollup=never if this change is not fit for rolling up.

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

Instruction count

This perf run didn't have relevant results for this metric.

Max RSS (memory usage)

Results (primary -2.1%, secondary -0.8%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	5.8%	[5.8%, 5.8%]	1
Improvements ✅ (primary)	-2.1%	[-2.1%, -2.1%]	1
Improvements ✅ (secondary)	-3.0%	[-4.1%, -2.2%]	3
All ❌✅ (primary)	-2.1%	[-2.1%, -2.1%]	1

Cycles

Results (primary 2.5%, secondary -3.2%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	2.5%	[2.5%, 2.5%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-3.2%	[-3.2%, -3.2%]	1
All ❌✅ (primary)	2.5%	[2.5%, 2.5%]	1

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 486.164s -> 485.416s (-0.15%)
Artifact size: 387.87 MiB -> 387.92 MiB (0.01%)

Kobzol · 2026-07-22T05:48:06Z

@bors try jobs=dist-x86_64-linux @rust-timer queue

Optimize Clippy with PGO try-job: dist-x86_64-linux

rust-bors · 2026-07-22T09:05:04Z

☀️ Try build successful (CI)
Build commit: 9e8b7aa (9e8b7aa75dfcaaee85081452174dfaf8fb2283aa)
Base parent: b9ee8d8 (b9ee8d8bcb1e4c7905a183a37e6d984c4b9a8468)

Kobzol · 2026-07-22T09:11:21Z

I forgot to include the Clippy benchmarks 🤦‍♂️

@bors try jobs=dist-x86_64-linux

Optimize Clippy with PGO try-job: dist-x86_64-linux

Kobzol · 2026-07-22T09:13:38Z

On https://github.com/It4innovations/hyperqueue, this is locally a 7-8% wall-time win, nice!

Kobzol · 2026-07-22T09:19:11Z

Local benchmark for Clippy icounts for syn, serde and Cargo:

Nice!

rust-timer · 2026-07-22T09:45:23Z

Finished benchmarking commit (9e8b7aa): comparison URL.

Overall result: no relevant changes - no action needed

Benchmarking means the PR may be perf-sensitive. Consider adding rollup=never if this change is not fit for rolling up.

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

Instruction count

This perf run didn't have relevant results for this metric.

Max RSS (memory usage)

This perf run didn't have relevant results for this metric.

Cycles

Results (secondary 0.1%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	2.6%	[2.4%, 2.8%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-4.8%	[-4.8%, -4.8%]	1
All ❌✅ (primary)	-	-	0

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 486.905s -> 487.72s (0.17%)
Artifact size: 387.85 MiB -> 387.77 MiB (-0.02%)

rust-bors · 2026-07-22T12:30:27Z

☀️ Try build successful (CI)
Build commit: def8243 (def824372694ad2e8ae59f6d0d7243060a71e843)
Base parent: 1af98b7 (1af98b7cdf86932b76182a387f52f1be6f0c1be0)

Kobzol · 2026-07-22T13:13:55Z

@rust-timer build def8243 profiles=Clippy

Kobzol · 2026-07-22T13:22:49Z

@bors try

Starting a fast try build, which should not build/optimize Clippy.

Optimize Clippy with PGO

rust-timer · 2026-07-22T14:18:22Z

Finished benchmarking commit (def8243): comparison URL.

Overall result: no relevant changes - no action needed

Benchmarking means the PR may be perf-sensitive. Consider adding rollup=never if this change is not fit for rolling up.

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

Instruction count

This perf run didn't have relevant results for this metric.

Max RSS (memory usage)

This perf run didn't have relevant results for this metric.

Cycles

This perf run didn't have relevant results for this metric.

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 485.602s -> 486.897s (0.27%)
Artifact size: 387.89 MiB -> 388.56 MiB (0.17%)

rust-bors · 2026-07-22T15:28:58Z

☀️ Try build successful (CI)
Build commit: f45359a (f45359a5a36c708cbbb13409d5671c016e102c5b)
Base parent: c188565 (c188565812a9b414ddf02f3c7d89e84f1ba53f7d)

Kobzol · 2026-07-23T06:52:06Z

@bors try jobs=dist-x86_64-linux @rust-timer queue profiles=Clippy

Optimize Clippy with PGO try-job: dist-x86_64-linux

rust-bors · 2026-07-23T10:10:08Z

☀️ Try build successful (CI)
Build commit: caf6734 (caf67342a544d6eab7a40ebf0d02e802f62008a4)
Base parent: 390279b (390279b302ca98ae270f434100ae3730531d1246)

rust-timer · 2026-07-23T10:56:23Z

Finished benchmarking commit (caf6734): comparison URL.

Overall result: ✅ improvements - no action needed

Benchmarking means the PR may be perf-sensitive. It's automatically marked not fit for rolling up. Overriding is possible but disadvised: it risks changing compiler perf.

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

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-4.6%	[-16.6%, -0.9%]	79
Improvements ✅ (secondary)	-4.4%	[-17.3%, -0.4%]	83
All ❌✅ (primary)	-4.6%	[-16.6%, -0.9%]	79

Max RSS (memory usage)

Results (primary -1.1%, secondary -2.1%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-1.1%	[-3.7%, -0.4%]	78
Improvements ✅ (secondary)	-2.1%	[-4.1%, -0.7%]	97
All ❌✅ (primary)	-1.1%	[-3.7%, -0.4%]	78

Cycles

Results (primary -5.3%, secondary -5.5%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	0.7%	[0.5%, 0.8%]	2
Improvements ✅ (primary)	-5.3%	[-23.9%, -0.9%]	83
Improvements ✅ (secondary)	-5.6%	[-24.8%, -0.5%]	92
All ❌✅ (primary)	-5.3%	[-23.9%, -0.9%]	83

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 486.724s -> 487.519s (0.16%)
Artifact size: 387.61 MiB -> 387.59 MiB (-0.01%)

Allow configuring PGO for Clippy

26d9c65

rustbot added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. T-bootstrap Relevant to the bootstrap subteam: Rust's build system (x.py and src/bootstrap) labels Jul 21, 2026