Introduce a minimum CGU size in non-incremental builds. #112448

nnethercote · 2023-06-09T05:56:05Z

Because tiny CGUs slow down compilation and result in worse generated code.

nnethercote · 2023-06-09T05:59:40Z

bors · 2023-06-09T05:59:51Z

⌛ Trying commit d7903614aaba46d87a0288e6eea386edb08d4c30 with merge 01c64cbb288ae7a44bb43912291d2b7b6e47da7d...

nnethercote · 2023-06-09T06:30:35Z

I got some tests/run-make/sepcomp-* failures locally. I've now fixed them, but I'll wait for the perf run to complete before uploading the relevant changes.

bors · 2023-06-09T07:17:28Z

☀️ Try build successful - checks-actions
Build commit: 01c64cbb288ae7a44bb43912291d2b7b6e47da7d (01c64cbb288ae7a44bb43912291d2b7b6e47da7d)

rust-timer · 2023-06-09T09:09:21Z

Finished benchmarking commit (01c64cbb288ae7a44bb43912291d2b7b6e47da7d): comparison URL.

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. While you can manually mark this PR as fit for rollup, we strongly recommend not doing so since this PR may lead to changes in compiler perf.

Next Steps: If you can justify the regressions found in this try perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please fix the regressions and do another perf run. If the next run shows neutral or positive results, the label will be automatically removed.

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

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	2.2%	[2.0%, 2.4%]	6
Regressions ❌ (secondary)	1.3%	[0.4%, 2.2%]	31
Improvements ✅ (primary)	-1.9%	[-7.6%, -0.3%]	20
Improvements ✅ (secondary)	-8.4%	[-46.0%, -0.2%]	21
All ❌✅ (primary)	-1.0%	[-7.6%, 2.4%]	26

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	1.9%	[0.9%, 2.8%]	2
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-2.7%	[-7.5%, -0.4%]	17
Improvements ✅ (secondary)	-5.1%	[-18.6%, -1.5%]	15
All ❌✅ (primary)	-2.2%	[-7.5%, 2.8%]	19

Cycles

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	1.8%	[1.6%, 2.1%]	6
Regressions ❌ (secondary)	1.7%	[1.3%, 2.1%]	12
Improvements ✅ (primary)	-3.8%	[-8.0%, -1.5%]	15
Improvements ✅ (secondary)	-11.2%	[-47.0%, -1.6%]	17
All ❌✅ (primary)	-2.2%	[-8.0%, 2.1%]	21

Binary size

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	1.8%	[0.5%, 2.8%]	16
Regressions ❌ (secondary)	2.5%	[0.2%, 2.8%]	71
Improvements ✅ (primary)	-2.1%	[-13.5%, -0.1%]	18
Improvements ✅ (secondary)	-7.1%	[-48.2%, -0.0%]	18
All ❌✅ (primary)	-0.3%	[-13.5%, 2.8%]	34

Bootstrap: 649.387s -> 650.164s (0.12%)

nnethercote · 2023-06-09T09:59:08Z

Hmm. This should be a slam dunk, based on logic and local measurements. The results here are great (or "no change") across the board for almost every metric (even the obscure ones) except walltime, where it's very mixed, with a few more regressions than improvements. Which is puzzling, because this really should be a clear win, especially for cases like unicode-normalization, libc, and html5ever, where a bunch of tiny CGUs are eliminated.

Also, lots of changes for incremental cases, which shouldn't be happening because this doesn't change incremental compilation at all. Including lots of changes to binary size for incremental cases, which makes no sense. (In comparison, the "object file" measurements look as expected. What's the difference between binary size (a.k.a. "linked artifact") and "object file"?)

nnethercote · 2023-06-09T13:58:03Z

@bors try @rust-timer queue

bors · 2023-06-09T13:58:14Z

⌛ Trying commit 45b08a53a870b118a159e65b710f0a8f3dfb1126 with merge 6532381723f53907d56ef7c7f21d2c3a77f7340a...

bors · 2023-06-09T15:33:07Z

☀️ Try build successful - checks-actions
Build commit: 6532381723f53907d56ef7c7f21d2c3a77f7340a (6532381723f53907d56ef7c7f21d2c3a77f7340a)

rust-timer · 2023-06-09T19:06:31Z

Finished benchmarking commit (6532381723f53907d56ef7c7f21d2c3a77f7340a): comparison URL.

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. While you can manually mark this PR as fit for rollup, we strongly recommend not doing so since this PR may lead to changes in compiler perf.

Next Steps: If you can justify the regressions found in this try perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please fix the regressions and do another perf run. If the next run shows neutral or positive results, the label will be automatically removed.

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

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	1.8%	[0.8%, 2.5%]	9
Regressions ❌ (secondary)	1.1%	[0.3%, 2.2%]	41
Improvements ✅ (primary)	-1.9%	[-7.2%, -0.3%]	20
Improvements ✅ (secondary)	-8.4%	[-46.0%, -0.4%]	21
All ❌✅ (primary)	-0.7%	[-7.2%, 2.5%]	29

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	1.4%	[0.9%, 2.0%]	2
Regressions ❌ (secondary)	3.2%	[1.5%, 4.9%]	3
Improvements ✅ (primary)	-3.2%	[-6.0%, -1.3%]	8
Improvements ✅ (secondary)	-6.8%	[-19.9%, -2.1%]	9
All ❌✅ (primary)	-2.3%	[-6.0%, 2.0%]	10

Cycles

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	1.8%	[1.6%, 2.0%]	5
Regressions ❌ (secondary)	1.6%	[1.2%, 2.0%]	12
Improvements ✅ (primary)	-3.5%	[-8.6%, -0.7%]	17
Improvements ✅ (secondary)	-9.6%	[-46.5%, -1.7%]	21
All ❌✅ (primary)	-2.3%	[-8.6%, 2.0%]	22

Binary size

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	1.8%	[0.5%, 2.8%]	16
Regressions ❌ (secondary)	2.5%	[0.2%, 2.8%]	71
Improvements ✅ (primary)	-1.8%	[-13.5%, -0.1%]	22
Improvements ✅ (secondary)	-7.2%	[-48.3%, -0.0%]	18
All ❌✅ (primary)	-0.3%	[-13.5%, 2.8%]	38

Bootstrap: 647.29s -> 647.896s (0.09%)

nikic · 2023-06-09T19:28:25Z

@nnethercote Could the fact that the CGUs now get sorted by size even if there are less than the limit make a difference?

nnethercote · 2023-06-09T23:34:32Z

@nnethercote Could the fact that the CGUs now get sorted by size even if there are less than the limit make a difference?

I had a similar thought, but this runs shortly afterwards, sorting the CGUs by name. Also, I did some debug logging and the CGU contents were identical before and after. Even worse, I'm getting big discrepancies between instruction counts measurements from Cachegrind vs perf. It's all very puzzling, I'll have to investigate more next week.

bors · 2023-06-13T09:41:10Z

⌛ Trying commit 9a516e047a6016474be5811bb66323cb5370e7e5 with merge 244b1111491e0e0b819b9a9aedb433964d59c0dd...

bors · 2023-06-13T11:14:44Z

☀️ Try build successful - checks-actions
Build commit: 244b1111491e0e0b819b9a9aedb433964d59c0dd (244b1111491e0e0b819b9a9aedb433964d59c0dd)

bors · 2023-06-13T11:14:44Z

☀️ Try build successful - checks-actions
Build commit: 244b1111491e0e0b819b9a9aedb433964d59c0dd (244b1111491e0e0b819b9a9aedb433964d59c0dd)

rust-timer · 2023-06-13T12:30:08Z

Finished benchmarking commit (244b1111491e0e0b819b9a9aedb433964d59c0dd): comparison URL.

Overall result: ✅ improvements - no action needed

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. While you can manually mark this PR as fit for rollup, we strongly recommend not doing so since this PR may lead to changes in compiler perf.

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

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-2.2%	[-7.7%, -0.3%]	19
Improvements ✅ (secondary)	-8.5%	[-46.0%, -0.5%]	22
All ❌✅ (primary)	-2.2%	[-7.7%, -0.3%]	19

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	2.2%	[2.2%, 2.2%]	1
Improvements ✅ (primary)	-3.5%	[-6.7%, -1.3%]	10
Improvements ✅ (secondary)	-7.2%	[-18.7%, -1.9%]	8
All ❌✅ (primary)	-3.5%	[-6.7%, -1.3%]	10

Cycles

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	4.0%	[4.0%, 4.2%]	3
Improvements ✅ (primary)	-3.9%	[-8.8%, -0.6%]	16
Improvements ✅ (secondary)	-10.9%	[-46.6%, -1.9%]	18
All ❌✅ (primary)	-3.9%	[-8.8%, -0.6%]	16

Binary size

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-1.9%	[-13.3%, -0.1%]	22
Improvements ✅ (secondary)	-7.2%	[-48.3%, -0.0%]	18
All ❌✅ (primary)	-1.9%	[-13.3%, -0.1%]	22

Bootstrap: 649.134s -> 650.1s (0.15%)

compiler/rustc_session/src/session.rs

wesleywiser · 2023-06-13T22:43:42Z

src/doc/rustc/src/codegen-options/index.md

+split into. (The exact number will depend on the size and structure of the
+crate's source code.) It takes an integer greater than 0.


I could see the parenthetical statement being interpreted as "we may create more CGUs than the value you specified". Maybe this formulation is more clear on that point?

Suggested change

split into. (The exact number will depend on the size and structure of the

crate's source code.) It takes an integer greater than 0.

split into. (Fewer code generation units may be created depending on the

size and structure of the crate's source code.) It takes an integer

greater than 0.

I'll remove the parenthetical. The important part of the change is specifying that it's a maximum.

wesleywiser · 2023-06-13T22:48:46Z

Another interesting result from your final commit, adding up all of the CGU sizes shows:

Crate	Before	After
html5ever	19,138	17,584
tt-muncher	9,097	9,075

My guess is that this is due to fewer duplicated inline functions in different CGUs. This might be another statistic worth reporting on.

For example, we go from this: ``` FINAL (4059 items, total_size=232342; 16 CGUs, max_size=39608, min_size=5468, max_size/min_size=7.2): - CGU[0] regex.f2ff11e98f8b05c7-cgu.0 (318 items, size=39608): - fn ... - fn ... ``` to this: ``` FINAL - unique items: 2726 (1459 root + 1267 inlined), unique size: 201214 (146046 root + 55168 inlined) - placed items: 4059 (1459 root + 2600 inlined), placed size: 232342 (146046 root + 86296 inlined) - placed/unique items ratio: 1.49, placed/unique size ratio: 1.15 - CGUs: 16, mean size: 14521.4, sizes: [39608, 31122, 20318, 20236, 16268, 13777, 12310, 10531, 10205, 9810, 9250, 9065 (x2), 7785, 7524, 5468] - CGU[0] - regex.f2ff11e98f8b05c7-cgu.0, size: 39608 - items: 318, mean size: 124.6, sizes: [28395, 3418, 558, 485, 259, 228, 176, 166, 146, 118, 117 (x3), 114 (x5), 113 (x3), 101, 84, 82, 77, 76, 72, 71 (x2), 66, 65, 62, 61, 59 (x2), 57, 55, 54 (x2), 53 (x4), 52 (x5), 51 (x4), 50, 48, 47, 46, 45 (x3), 44, 43 (x5), 42, 40, 38 (x4), 37, 35, 34 (x2), 32 (x2), 31, 30, 28 (x2), 27 (x2), 26 (x3), 24 (x2), 23 (x3), 22 (x2), 21, 20, 16 (x4), 15, 13 (x7), 12 (x3), 11 (x6), 10, 9 (x2), 8 (x4), 7 (x8), 6 (x38), 5 (x21), 4 (x7), 3 (x45), 2 (x63), 1 (x13)] - fn ... - fn ... ``` This is a lot more information, distinguishing between root items and inlined items, showing how much duplication there is of inlined items, plus the full range of sizes for CGUs and items within CGUs. All of which is really helpful when analyzing this stuff and trying different CGU formation algorithms.

Because tiny CGUs make compilation less efficient *and* result in worse generated code. We don't do this when the number of CGUs is explicitly given, because there are times when the requested number is very important, as described in some comments within the commit. So the commit also introduces a `CodegenUnits` type that distinguishes between default values and user-specified values. This change has a roughly neutral effect on walltimes across the rustc-perf benchmarks; there are some speedups and some slowdowns. But it has significant wins for most other metrics on numerous benchmarks, including instruction counts, cycles, binary size, and max-rss. It also reduces parallelism, which is good for reducing jobserver competition when multiple rustc processes are running at the same time. It's smaller benchmarks that benefit the most; larger benchmarks already have CGUs that are all larger than the minimum size. Here are some example before/after CGU sizes for opt builds. - html5ever - CGUs: 16, mean size: 1196.1, sizes: [3908, 2992, 1706, 1652, 1572, 1136, 1045, 948, 946, 938, 579, 471, 443, 327, 286, 189] - CGUs: 4, mean size: 4396.0, sizes: [6706, 3908, 3490, 3480] - libc - CGUs: 12, mean size: 35.3, sizes: [163, 93, 58, 53, 37, 8, 2 (x6)] - CGUs: 1, mean size: 424.0, sizes: [424] - tt-muncher - CGUs: 5, mean size: 1819.4, sizes: [8508, 350, 198, 34, 7] - CGUs: 1, mean size: 9075.0, sizes: [9075] Note that CGUs of size 100,000+ aren't unusual in larger programs.

nnethercote · 2023-06-14T00:58:30Z

My guess is that this is due to fewer duplicated inline functions in different CGUs. This might be another statistic worth reporting on.

Yes, that's definitely a big part of the improvement here. It's hard to report on because merging currently happens before inlining, so you can't make this measurement without two versions of the compiler, one doing the tiny CGU merging and one not.

nnethercote · 2023-06-14T01:00:07Z

I have addressed the review comments. I also made some very small tweaks to the debug printing, but nothing needing additional review. Thanks!

@bors r=wesleywiser

bors · 2023-06-14T01:00:08Z

📌 Commit 7c3ce02 has been approved by wesleywiser

It is now in the queue for this repository.

bors · 2023-06-14T02:34:20Z

⌛ Testing commit 7c3ce02 with merge fa8762b...

bors · 2023-06-14T05:37:01Z

☀️ Test successful - checks-actions
Approved by: wesleywiser
Pushing fa8762b to master...

rust-timer · 2023-06-14T06:53:04Z

Finished benchmarking commit (fa8762b): comparison URL.

Overall result: ✅ improvements - no action needed

@rustbot label: -perf-regression

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-2.1%	[-7.8%, -0.3%]	19
Improvements ✅ (secondary)	-8.5%	[-45.9%, -0.4%]	22
All ❌✅ (primary)	-2.1%	[-7.8%, -0.3%]	19

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-3.6%	[-6.4%, -1.8%]	10
Improvements ✅ (secondary)	-7.1%	[-18.9%, -2.4%]	9
All ❌✅ (primary)	-3.6%	[-6.4%, -1.8%]	10

Cycles

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	5.0%	[4.4%, 5.3%]	3
Improvements ✅ (primary)	-3.6%	[-8.6%, -0.8%]	18
Improvements ✅ (secondary)	-11.6%	[-46.6%, -2.1%]	17
All ❌✅ (primary)	-3.6%	[-8.6%, -0.8%]	18

Binary size

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-1.9%	[-13.3%, -0.1%]	22
Improvements ✅ (secondary)	-7.3%	[-48.2%, -0.0%]	18
All ❌✅ (primary)	-1.9%	[-13.3%, -0.1%]	22

Bootstrap: 648.99s -> 648.277s (-0.11%)

rustbot assigned wesleywiser Jun 9, 2023

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. labels Jun 9, 2023

This comment has been minimized.

Sign in to view

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Jun 9, 2023

This comment has been minimized.

Sign in to view

rustbot added perf-regression Performance regression. and removed S-waiting-on-perf Status: Waiting on a perf run to be completed. labels Jun 9, 2023

nnethercote force-pushed the no-tiny-cgus branch from d790361 to 45b08a5 Compare June 9, 2023 09:50

This comment has been minimized.

Sign in to view

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Jun 9, 2023

This comment has been minimized.

Sign in to view

rustbot removed the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Jun 9, 2023

nnethercote force-pushed the no-tiny-cgus branch from 45b08a5 to 6b456f8 Compare June 10, 2023 09:41

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Jun 13, 2023

This comment has been minimized.

Sign in to view

rustbot removed S-waiting-on-perf Status: Waiting on a perf run to be completed. perf-regression Performance regression. labels Jun 13, 2023

nnethercote force-pushed the no-tiny-cgus branch from 9a516e0 to 6f00c8d Compare June 13, 2023 22:40

wesleywiser approved these changes Jun 13, 2023

View reviewed changes

nnethercote force-pushed the no-tiny-cgus branch from 6f00c8d to 50acba5 Compare June 14, 2023 00:23

nnethercote force-pushed the no-tiny-cgus branch from 50acba5 to 7c3ce02 Compare June 14, 2023 00:57

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Jun 14, 2023

bors added the merged-by-bors This PR was explicitly merged by bors. label Jun 14, 2023

bors merged commit fa8762b into rust-lang:master Jun 14, 2023
12 checks passed

rustbot added this to the 1.72.0 milestone Jun 14, 2023

nnethercote deleted the no-tiny-cgus branch June 14, 2023 06:52

lqd mentioned this pull request Jun 18, 2023

improve cgu schedule with greedy partition #112766

Closed

Mark-Simulacrum mentioned this pull request Jan 14, 2024

[partitioning] Don't create CGUs below a certain size #47318

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Introduce a minimum CGU size in non-incremental builds. #112448

Introduce a minimum CGU size in non-incremental builds. #112448

nnethercote commented Jun 9, 2023

nnethercote commented Jun 9, 2023

This comment has been minimized.

bors commented Jun 9, 2023

nnethercote commented Jun 9, 2023 •

edited

This comment has been minimized.

bors commented Jun 9, 2023

This comment has been minimized.

rust-timer commented Jun 9, 2023

nnethercote commented Jun 9, 2023

This comment has been minimized.

nnethercote commented Jun 9, 2023

This comment has been minimized.

bors commented Jun 9, 2023

bors commented Jun 9, 2023

This comment has been minimized.

rust-timer commented Jun 9, 2023

nikic commented Jun 9, 2023

nnethercote commented Jun 9, 2023

bors commented Jun 13, 2023

bors commented Jun 13, 2023

bors commented Jun 13, 2023

This comment has been minimized.

rust-timer commented Jun 13, 2023

wesleywiser Jun 13, 2023

nnethercote Jun 14, 2023

wesleywiser commented Jun 13, 2023

nnethercote commented Jun 14, 2023

nnethercote commented Jun 14, 2023

bors commented Jun 14, 2023

bors commented Jun 14, 2023

bors commented Jun 14, 2023

rust-timer commented Jun 14, 2023

		split into. (The exact number will depend on the size and structure of the
		crate's source code.) It takes an integer greater than 0.

Introduce a minimum CGU size in non-incremental builds. #112448

Introduce a minimum CGU size in non-incremental builds. #112448

Conversation

nnethercote commented Jun 9, 2023

nnethercote commented Jun 9, 2023

This comment has been minimized.

bors commented Jun 9, 2023

nnethercote commented Jun 9, 2023 • edited

This comment has been minimized.

bors commented Jun 9, 2023

This comment has been minimized.

rust-timer commented Jun 9, 2023

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

nnethercote commented Jun 9, 2023

This comment has been minimized.

nnethercote commented Jun 9, 2023

This comment has been minimized.

bors commented Jun 9, 2023

bors commented Jun 9, 2023

This comment has been minimized.

rust-timer commented Jun 9, 2023

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

nikic commented Jun 9, 2023

nnethercote commented Jun 9, 2023

bors commented Jun 13, 2023

bors commented Jun 13, 2023

bors commented Jun 13, 2023

This comment has been minimized.

rust-timer commented Jun 13, 2023

Overall result: ✅ improvements - no action needed

Choose a reason for hiding this comment

Choose a reason for hiding this comment

wesleywiser commented Jun 13, 2023

nnethercote commented Jun 14, 2023

nnethercote commented Jun 14, 2023

bors commented Jun 14, 2023

bors commented Jun 14, 2023

bors commented Jun 14, 2023

rust-timer commented Jun 14, 2023

Overall result: ✅ improvements - no action needed

nnethercote commented Jun 9, 2023 •

edited