[PERF] see if my fork of stringdex affects perf

[yotamofek]

cc @GuillaumeGomez and @notriddle

please disregard the actual changes in my fork on stringdex, it's all "wip" commits and a bunch of different changes I was experimenting with..
I just wanna see if it actually has an effect on the rustc benchmarks (not just the stringdex benchmarks) to see if it's worth pursuing

[Kobzol]

@bors try @rust-timer queue

[rust-bors]

💥 Test timed out after 21600s

[notriddle]

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: cc4f770 (cc4f7703e067e4e62f404b76a3a1bd7c1d67b4ec, parent: c7a635f33c5fbd1ead110243a2f4a5f0561d79b0)

[rust-timer]

Finished benchmarking commit (cc4f770): comparison URL.

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

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.

Next Steps: If you can justify the regressions found in this try perf run, please do so in sufficient writing along with @rustbot label: +perf-regression-triaged. If not, please fix the regressions and do another perf run. If its results are neutral or positive, the label will be automatically removed.

@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
Regressions ❌ (secondary)	0.6%	[0.6%, 0.6%]	1
Improvements ✅ (primary)	-0.4%	[-0.9%, -0.1%]	12
Improvements ✅ (secondary)	-0.5%	[-0.9%, -0.3%]	3
All ❌✅ (primary)	-0.4%	[-0.9%, -0.1%]	12

Max RSS (memory usage)

Results (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)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-4.0%	[-4.0%, -4.0%]	1
All ❌✅ (primary)	-	-	0

Cycles

Results (primary -2.4%, secondary 2.0%)

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)	2.0%	[2.0%, 2.0%]	1
Improvements ✅ (primary)	-2.4%	[-2.4%, -2.4%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-2.4%	[-2.4%, -2.4%]	1

Binary size

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

Bootstrap: 472.842s -> 473.067s (0.05%)
Artifact size: 388.69 MiB -> 388.71 MiB (0.01%)

[JonathanBrouwer]

Performance looks excellent, the coercions result is clearly bimodal noise

[GuillaumeGomez]

@yotamofek: Do you want to merge this one already?

[yotamofek]

Nah, my changes to stringdex are a complete mess at the moment, I think it'll be better to first get a few of them up as PRs and have @notriddle go over them, and then released to crates.io. No point in depending on a fork when it's a crate that's owned by a project member.
(unless they don't have the capacity to review and/or work on stringdex)

[notriddle]

It's fine. Send me the PR on GitLab, and I'll look at them ASAP.

[yotamofek]

Still a WIP, but I've started PR-ing my changes and notriddle has merged a few of them already, and I'm interested in the interim perf results. Would love another perf run please 🙏
(the stringdex dep now points at the upstream repo, and not my fork)

[JonathanBrouwer]

@bors try @rust-timer queue

[rust-bors]

Unknown argument "cance". Run @bors2 help to see available commands.

[yotamofek]

@bors try cancel

[rust-bors]

Try build cancelled. Cancelled workflows:

• https://github.com/rust-lang/rust/actions/runs/19099060804

[yotamofek]

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: d73a1bd (d73a1bddf44384c37e074411122571b40f44c339, parent: 8e0b68e63cd2b7b6d18474fe6f49df6fb1570c25)

[rust-timer]

Finished benchmarking commit (d73a1bd): comparison URL.

Overall result: ✅ improvements - 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

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)	-0.5%	[-1.1%, -0.2%]	13
Improvements ✅ (secondary)	-0.5%	[-1.3%, -0.1%]	4
All ❌✅ (primary)	-0.5%	[-1.1%, -0.2%]	13

Max RSS (memory usage)

Results (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)	-	-	0
Improvements ✅ (secondary)	-3.4%	[-3.4%, -3.4%]	1
All ❌✅ (primary)	-	-	0

Cycles

Results (secondary 2.4%)

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)	3.9%	[2.1%, 7.2%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.1%	[-2.1%, -2.1%]	1
All ❌✅ (primary)	-	-	0

Binary size

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

Bootstrap: 475.57s -> 475.357s (-0.04%)
Artifact size: 390.76 MiB -> 390.77 MiB (0.00%)

[yotamofek]

I'm quite happy with these results, they're even better than the first perf run, but now all my changes to stringdex are properly separated into commits, and mostly merged.

@notriddle , once you approve the last PR, I think it would be a good point to release a new version of stringdex and call it a win. :)

[JonathanBrouwer]

Yeah this performance improvement is awesome to see, thanks a lot for your effort!

[yotamofek]

@bors try @rust-timer queue

Pushed a new version that takes a branch of stringdex that merges both PRs that are currently open (!11 and !12). Also added another commit to !11 since last perf run, which required some adaptations on the rustdoc side.

Gonna take advantage of the fact that the perf queue is really quiet right now to get another reading. Mostly for the sake of my curiosity, wanna see if I managed to squeeze out a few more green numbers until notriddle gets a chance to properly review everything :)

[rust-bors]

☀️ Try build successful (CI)
Build commit: 5d905a8 (5d905a831657432c7b62345e6e710cc0ada801ef, parent: 642c19bfc3a5c1de985bf5d0cc8207ac9d22708a)

[rust-timer]

Finished benchmarking commit (5d905a8): comparison URL.

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

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.

Next Steps: If you can justify the regressions found in this try perf run, please do so in sufficient writing along with @rustbot label: +perf-regression-triaged. If not, please fix the regressions and do another perf run. If its results are neutral or positive, the label will be automatically removed.

@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
Regressions ❌ (secondary)	0.1%	[0.1%, 0.1%]	2
Improvements ✅ (primary)	-0.5%	[-1.1%, -0.2%]	13
Improvements ✅ (secondary)	-0.5%	[-1.3%, -0.1%]	4
All ❌✅ (primary)	-0.5%	[-1.1%, -0.2%]	13

Max RSS (memory usage)

Results (primary -1.7%, 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)	-	-	0
Regressions ❌ (secondary)	3.4%	[3.4%, 3.4%]	1
Improvements ✅ (primary)	-1.7%	[-1.7%, -1.7%]	1
Improvements ✅ (secondary)	-3.8%	[-5.3%, -1.7%]	11
All ❌✅ (primary)	-1.7%	[-1.7%, -1.7%]	1

Cycles

Results (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)	-	-	0
Regressions ❌ (secondary)	7.1%	[3.6%, 10.6%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-6.2%	[-8.7%, -2.6%]	4
All ❌✅ (primary)	-	-	0

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: 474.467s -> 474.025s (-0.09%)
Artifact size: 390.76 MiB -> 390.90 MiB (0.03%)

[yotamofek]

Looks a tiny bit better when only looking at "Doc" profile (changes in other profiles are probably noise):

Previous run (avg. -0.49) vs last run (avg. -0.50).

[bors]

☔ The latest upstream changes (presumably #148573) made this pull request unmergeable. Please resolve the merge conflicts.