Avoid a few more allocations in write_shared.rs

[yotamofek]

Inspired by #141421 , avoids a few Vec, PathBuf and String allocations in write_shared.rs. I don't think these will show up on benchmarks, but are still worthwhile IMHO.
Also includes a few small cleanups.
r? nnethercote - if you'd like :)

[rustbot]

nnethercote is not on the review rotation at the moment.
They may take a while to respond.

[yotamofek]

Oh, maybe r? @GuillaumeGomez then?

[GuillaumeGomez]

Looks promising, thanks! Let's check perf impact.

@bors try @rust-timer queue

[bors]

⌛ Trying commit 56fe2f6 with merge 91cf0c3...

[bors]

☀️ Try build successful - checks-actions
Build commit: 91cf0c3 (91cf0c3afad8861a1760cf5bae082ee32c4375b5)

[nnethercote]

I like the cleanups in commits 1, 3, 4, and 5.

I'm ambivalent about the introduction of Cow in commits 2 and 6. Optimized code is often more complex that non-optimized code, and if the optimizations don't have measurable effects then they aren't worth it. I.e. I disagree with "I don't think these will show up on benchmarks, but are still worthwhile IMHO." I did #141421 by looking at memory profiles (from DHAT) and avoiding the hot allocation paths.

[rust-timer]

Finished benchmarking commit (91cf0c3): comparison URL.

Overall result: no relevant changes - no action needed

Benchmarking this pull request means it may be perf-sensitive – we'll automatically label it not fit for rolling up. You can override this, but we strongly advise not to, due to possible changes in compiler perf.

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

Instruction count

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

Max RSS (memory usage)

Results (secondary 2.7%)

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)	2.7%	[2.7%, 2.7%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Cycles

Results (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)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.5%	[-2.5%, -2.5%]	1
All ❌✅ (primary)	-	-	0

Binary size

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

Bootstrap: 692.01s -> 691.962s (-0.01%)
Artifact size: 372.13 MiB -> 372.12 MiB (-0.00%)

[nnethercote]

The perf CI run is neutral. r=me for commits 1, 3, 4, and 5.

[yotamofek]

@nnethercote thanks for reviewing! Rebased without those two commits.
Just wanted to say that I didn't mean that it's worthwhile to make these changes even if they don't improve perf (because like you said - they just add complexity), what I meant was that the benchmark suite doesn't necessarily cover all workloads, and that some improvements might be too minute to register, but will have a real-life impact when added up. i.e., benchmarks are never perfect. :)

[nnethercote]

Thanks!

@bors r+

[bors]

📌 Commit 2bf66e3 has been approved by nnethercote

It is now in the queue for this repository.

[bors]

⌛ Testing commit 2bf66e3 with merge fa2f355...

[bors]

☀️ Test successful - checks-actions
Approved by: nnethercote
Pushing fa2f355 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 d4e1159 (parent) -> fa2f355 (this PR)

Test differences

No test diffs found

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard fa2f3552dae918294c8572b2a07fed2746722dd3 --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

1. x86_64-apple-2: 4878.1s -> 3715.6s (-23.8%)
2. x86_64-apple-1: 6230.9s -> 7013.4s (12.6%)
3. x86_64-msvc-1: 8401.0s -> 9110.6s (8.4%)
4. dist-apple-various: 6204.2s -> 5756.3s (-7.2%)
5. dist-x86_64-musl: 7106.4s -> 7557.0s (6.3%)
6. dist-aarch64-apple: 5706.0s -> 5391.6s (-5.5%)
7. x86_64-msvc-2: 6651.0s -> 7016.5s (5.5%)
8. dist-x86_64-apple: 7078.8s -> 7399.7s (4.5%)
9. aarch64-gnu: 6389.9s -> 6128.6s (-4.1%)
10. x86_64-gnu-aux: 6101.3s -> 5853.9s (-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]

Finished benchmarking commit (fa2f355): comparison URL.

Overall result: no relevant changes - no action needed

@rustbot label: -perf-regression

Instruction count

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

Max RSS (memory usage)

Results (primary -1.7%, secondary -3.4%)

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)	-1.7%	[-2.2%, -1.2%]	2
Improvements ✅ (secondary)	-3.4%	[-5.6%, -1.0%]	6
All ❌✅ (primary)	-1.7%	[-2.2%, -1.2%]	2

Cycles

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

Binary size

Results (secondary 0.0%)

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)	0.0%	[0.0%, 0.0%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Bootstrap: 691.957s -> 689.982s (-0.29%)
Artifact size: 371.85 MiB -> 371.93 MiB (0.02%)