Improve cgu size estimate

[csmoe]

View all comments

This patch improved walltime with -3% and rss with -120MB in my local benchmark of compiling rust-analyzer repo.

May I request a perf-run r?@bjorn3

[rustbot]

r? @mati865

rustbot has assigned @mati865.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: compiler
• compiler expanded to 73 candidates
• Random selection from 18 candidates

[lqd]

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: 4f3300d (4f3300d930eca8fd67617858ee4961c9cc52c522, parent: a31c27a887b40df16ab9dfb8c9f7924636092509)

[rust-timer]

Finished benchmarking commit (4f3300d): comparison URL.

Overall result: ❌✅ regressions and improvements - BENCHMARK(S) FAILED

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

❗ ❗ ❗ ❗ ❗
Warning ⚠️: The following benchmark(s) failed to build:

• Job failure

❗ ❗ ❗ ❗ ❗

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)	1.9%	[0.2%, 8.2%]	13
Regressions ❌ (secondary)	2.2%	[0.0%, 7.5%]	10
Improvements ✅ (primary)	-1.5%	[-3.4%, -0.7%]	4
Improvements ✅ (secondary)	-1.1%	[-2.2%, -0.1%]	2
All ❌✅ (primary)	1.1%	[-3.4%, 8.2%]	17

Max RSS (memory usage)

Results (primary 2.0%, 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)	3.1%	[2.2%, 4.6%]	9
Regressions ❌ (secondary)	3.0%	[1.7%, 4.6%]	7
Improvements ✅ (primary)	-2.9%	[-3.0%, -2.9%]	2
Improvements ✅ (secondary)	-3.0%	[-3.7%, -2.3%]	4
All ❌✅ (primary)	2.0%	[-3.0%, 4.6%]	11

Cycles

Results (primary 2.7%, secondary 4.0%)

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

	mean	range	count
Regressions ❌ (primary)	3.5%	[1.8%, 8.5%]	7
Regressions ❌ (secondary)	4.0%	[2.6%, 8.0%]	8
Improvements ✅ (primary)	-3.0%	[-3.0%, -3.0%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.7%	[-3.0%, 8.5%]	8

Binary size

Results (primary 0.6%, secondary 1.7%)

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

	mean	range	count
Regressions ❌ (primary)	1.3%	[0.0%, 4.7%]	24
Regressions ❌ (secondary)	2.2%	[0.0%, 15.2%]	22
Improvements ✅ (primary)	-0.9%	[-4.4%, -0.1%]	11
Improvements ✅ (secondary)	-3.0%	[-6.0%, -0.1%]	2
All ❌✅ (primary)	0.6%	[-4.4%, 4.7%]	35

Bootstrap: missing data
Artifact size: 400.56 MiB -> 400.65 MiB (0.02%)

[csmoe]

The job failed, May I have a retry? @lqd

rustc-perf did generated a report, but not sure if it's the expected one. walltime improved with -5% in mean.

Error while benchmarking job BenchmarkJob {
    id: 36134,
    target: X86_64UnknownLinuxGnu,
    backend: Llvm,
    profile: Opt,
    request_tag: "4f3300d930eca8fd67617858ee4961c9cc52c522",
    benchmark_set: BenchmarkSet(
        0,
    ),
    created_at: 2026-05-18T09:36:03.125470Z,
    status: InProgress {
        started_at: 2026-05-18T10:09:13.752235Z,
        collector_name: "x64-1",
    },
    deque_counter: 4,
    kind: Rustc,
    is_optional: false,
}

Job failed after being dequeued for 3 times

[lqd]

Sure, I can queue another run for a different commit, otherwise the runner will say that sha1 is already benchmarked. So you may need to rebase or add another empty commit, something like that.

[lqd]

I'll cc @Kobzol on that job failure though.

[rustbot]

This PR was rebased onto a different main commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.

[lqd]

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: 9776c7e (9776c7efb2325c383d1fddfe1df97a52abbd644f, parent: 5ea817c65e4896167300b7d2550781b98da9901a)

[rust-timer]

Finished benchmarking commit (9776c7e): 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)	1.9%	[0.2%, 8.1%]	13
Regressions ❌ (secondary)	2.2%	[0.0%, 7.5%]	10
Improvements ✅ (primary)	-1.6%	[-3.5%, -0.7%]	4
Improvements ✅ (secondary)	-1.3%	[-2.4%, -0.2%]	2
All ❌✅ (primary)	1.1%	[-3.5%, 8.1%]	17

Max RSS (memory usage)

Results (primary 1.6%, secondary 2.5%)

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

	mean	range	count
Regressions ❌ (primary)	2.9%	[1.3%, 4.8%]	9
Regressions ❌ (secondary)	3.2%	[0.7%, 7.5%]	7
Improvements ✅ (primary)	-2.4%	[-3.4%, -0.6%]	3
Improvements ✅ (secondary)	-2.6%	[-2.6%, -2.6%]	1
All ❌✅ (primary)	1.6%	[-3.4%, 4.8%]	12

Cycles

Results (primary 2.7%, secondary 3.5%)

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

	mean	range	count
Regressions ❌ (primary)	3.5%	[2.0%, 8.7%]	7
Regressions ❌ (secondary)	4.2%	[2.3%, 9.5%]	8
Improvements ✅ (primary)	-3.0%	[-3.0%, -3.0%]	1
Improvements ✅ (secondary)	-2.5%	[-2.5%, -2.5%]	1
All ❌✅ (primary)	2.7%	[-3.0%, 8.7%]	8

Binary size

Results (primary 0.7%, secondary 1.8%)

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

	mean	range	count
Regressions ❌ (primary)	1.3%	[0.0%, 4.7%]	24
Regressions ❌ (secondary)	2.2%	[0.0%, 15.2%]	22
Improvements ✅ (primary)	-1.0%	[-4.4%, -0.3%]	10
Improvements ✅ (secondary)	-6.0%	[-6.0%, -6.0%]	1
All ❌✅ (primary)	0.7%	[-4.4%, 4.7%]	34

Bootstrap: 511.688s -> 512.164s (0.09%)
Artifact size: 400.54 MiB -> 402.63 MiB (0.52%)

[Kobzol]

Rustc itself (the bootstrap benchmark) failed to build on the first try, not sure why, I won't have access to rustc-perf runner logs this week. It could have been a transient issue downloading some rustc dependency from the web.

[csmoe]

Another perf-run requested

[Kobzol]

@bors try @rust-timer queue

[rust-timer]

Awaiting bors try build completion.

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

[rust-bors]

⌛ Trying commit 295e11c with merge aa28d85…

To cancel the try build, run the command @bors try cancel.

Workflow: https://github.com/rust-lang/rust/actions/runs/26096412514