Use PGO for Cargo by Kobzol · Pull Request #159149 · rust-lang/rust

Kobzol · 2026-07-11T20:45:53Z

I downloaded the built Cargo and compared it with the parent commit locally on bors (which is of course not in the PGO training set). The results are quite nice:

Benchmark 1 (26 runs): cargo +babf6ea4aab9ed78ea9400b25b0e7e79a6870f29 test --no-run
  measurement          mean ± σ            min … max           outliers         delta
  wall_time           192ms ± 6.95ms     188ms …  226ms          1 ( 4%)        0%
  peak_rss           96.3MB ±  354KB    95.6MB … 96.9MB          0 ( 0%)        0%
  cpu_cycles          454M  ± 5.92M      448M  …  480M           1 ( 4%)        0%
  instructions        797M  ± 5.68M      796M  …  825M           1 ( 4%)        0%
  cache_references   32.7M  ±  352K     32.0M  … 34.0M           1 ( 4%)        0%
  cache_misses       8.82M  ± 86.9K     8.67M  … 9.10M           1 ( 4%)        0%
  branch_misses      3.76M  ± 41.3K     3.72M  … 3.93M           1 ( 4%)        0%

Benchmark 2 (28 runs): cargo +b2aa8c0474895b96828df9c09182952caf6f6450 test --no-run
  measurement          mean ± σ            min … max           outliers         delta
  wall_time           185ms ± 7.07ms     175ms …  208ms          1 ( 4%)        ⚡-  3.9% ±  2.0%
  peak_rss           90.9MB ±  207KB    90.4MB … 91.3MB          1 ( 4%)        ⚡-  5.6% ±  0.2%
  cpu_cycles          420M  ± 9.31M      406M  …  440M           0 ( 0%)        ⚡-  7.7% ±  0.9%
  instructions        736M  ± 5.47M      735M  …  764M           1 ( 4%)        ⚡-  7.6% ±  0.4%
  cache_references   26.2M  ±  406K     25.7M  … 28.0M           2 ( 7%)        ⚡- 19.9% ±  0.6%
  cache_misses       6.01M  ± 92.7K     5.85M  … 6.28M           1 ( 4%)        ⚡- 31.8% ±  0.6%
  branch_misses      3.47M  ± 55.9K     3.40M  … 3.69M           1 ( 4%)        ⚡-  7.6% ±  0.7%

~7.5% icount improvement, that's pretty cool.

Kobzol · 2026-07-11T20:46:59Z

@bors try @rust-timer queue

Use PGO for Cargo

rust-bors · 2026-07-11T21:26:06Z

💔 Test for 0109abd failed: CI. Failed job:

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

Kobzol · 2026-07-11T21:32:13Z

@bors try

Use PGO for Cargo

rust-bors · 2026-07-11T22:11:07Z

💔 Test for 7353424 failed: CI. Failed job:

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

Kobzol · 2026-07-12T09:00:28Z

@bors try

Use PGO for Cargo

rust-bors · 2026-07-12T09:42:52Z

💔 Test for 1f59755 failed: CI. Failed job:

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

Kobzol · 2026-07-12T11:31:58Z

@bors try

Use PGO for Cargo

rust-bors · 2026-07-12T12:13:54Z

💔 Test for cf8ac50 failed: CI. Failed job:

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

Kobzol · 2026-07-12T12:28:35Z

@bors try

Use PGO for Cargo

rust-bors · 2026-07-12T13:09:16Z

💔 Test for b222323 failed: CI. Failed job:

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

Kobzol · 2026-07-12T13:37:24Z

@bors try

Use PGO for Cargo

rust-bors · 2026-07-12T16:15:51Z

☀️ Try build successful (CI)
Build commit: b2aa8c0 (b2aa8c0474895b96828df9c09182952caf6f6450)
Base parent: babf6ea (babf6ea4aab9ed78ea9400b25b0e7e79a6870f29)

rust-timer · 2026-07-12T20:25:03Z

Finished benchmarking commit (b2aa8c0): comparison URL.

Overall result: ❌✅ regressions and improvements - please read:

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.

Next, please: If you can, justify the regressions found in this try perf run in writing along with @rustbot label: +perf-regression-triaged. If not, fix the regressions and do another perf run. Neutral or positive results will clear the label automatically.

@bors rollup=never
@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)	0.5%	[0.3%, 0.9%]	11
Regressions ❌ (secondary)	0.6%	[0.1%, 4.5%]	31
Improvements ✅ (primary)	-2.6%	[-6.4%, -0.2%]	34
Improvements ✅ (secondary)	-2.5%	[-8.2%, -0.2%]	51
All ❌✅ (primary)	-1.9%	[-6.4%, 0.9%]	45

Max RSS (memory usage)

Results (primary 4.3%, secondary 3.7%)

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

	mean	range	count
Regressions ❌ (primary)	4.3%	[3.7%, 4.9%]	2
Regressions ❌ (secondary)	3.7%	[2.2%, 5.4%]	4
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	4.3%	[3.7%, 4.9%]	2

Cycles

Results (primary -3.3%, secondary -3.1%)

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

	mean	range	count
Regressions ❌ (primary)	3.4%	[3.4%, 3.4%]	1
Regressions ❌ (secondary)	7.0%	[3.2%, 14.0%]	3
Improvements ✅ (primary)	-3.7%	[-5.4%, -2.2%]	18
Improvements ✅ (secondary)	-4.3%	[-7.5%, -2.6%]	25
All ❌✅ (primary)	-3.3%	[-5.4%, 3.4%]	19

Binary size

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

Bootstrap: 490.525s -> 491.031s (0.10%)
Artifact size: 389.30 MiB -> 390.59 MiB (0.33%)

Kobzol · 2026-07-12T20:40:26Z

@bors try jobs=dist-x86_64-linux

Trying full CI duration.

Use PGO for Cargo try-job: dist-x86_64-linux

rust-bors · 2026-07-13T00:22:10Z

☀️ Try build successful (CI)
Build commit: efbd9eb (efbd9ebd1dbc681e6096b386f0aeb8f616a52574)
Base parent: 77cf889 (77cf889bc178ddb44d6a1c78e5a820b5abb31d8d)

Kobzol added 6 commits July 10, 2026 16:45

Allow setting PGO for rustdoc separately from rustc

6a86307

Use separate rustdoc PGO in opt-dist

cbc5347

Run PGO on Doc benchmarks too

b496c39

Build rustdoc in opt-dist

f32bc9b

Separate rustc and rustdoc PGO profiles

b2e3489

Allow configuring PGO for Cargo

1300fc0

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 11, 2026