Rewrite lint_expectations in a single pass. #127313

cjgillot · 2024-07-04T10:48:15Z

This PR aims at reducing the perf regression from #120924 (comment) with drive-by simplifications.

Basically, instead of using the lint level builder, which is slow, this PR splits lint_expectations logic in 2:

listing the LintExpectations is done in shallow_lint_levels_on, on a per-owner basis;
building the unstable->stable expectation id map is done by iterating on attributes.

r? ghost for perf

cjgillot · 2024-07-04T10:48:38Z

Rewrite lint_expectations in a single pass. This PR aims at reducing the perf regression from rust-lang#120924 (comment) with drive-by simplifications. Basically, instead of using the lint level builder, which is slow, this PR splits `lint_expectations` logic in 2: - listing the `LintExpectations` is done in `shallow_lint_levels_on`, on a per-owner basis; - building the unstable->stable expectation id map is done by iterating on attributes. r? ghost for perf

bors · 2024-07-04T10:49:48Z

⌛ Trying commit 66a9985 with merge 809e06b...

bors · 2024-07-04T12:32:44Z

☀️ Try build successful - checks-actions
Build commit: 809e06b (809e06bc005b736706b134a829fb59e278eb6017)

rust-timer · 2024-07-04T13:51:17Z

Finished benchmarking commit (809e06b): 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)	0.3%	[0.2%, 0.5%]	15
Regressions ❌ (secondary)	0.7%	[0.3%, 1.4%]	12
Improvements ✅ (primary)	-0.5%	[-1.5%, -0.2%]	70
Improvements ✅ (secondary)	-0.7%	[-1.6%, -0.2%]	42
All ❌✅ (primary)	-0.4%	[-1.5%, 0.5%]	85

Max RSS (memory usage)

Results (primary -5.5%, secondary -2.2%)

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)	-5.5%	[-5.5%, -5.5%]	1
Improvements ✅ (secondary)	-2.2%	[-4.2%, -0.9%]	8
All ❌✅ (primary)	-5.5%	[-5.5%, -5.5%]	1

Cycles

Results (primary 1.5%, secondary 1.6%)

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)	2.9%	[2.1%, 4.0%]	11
Regressions ❌ (secondary)	1.6%	[0.5%, 2.1%]	7
Improvements ✅ (primary)	-1.3%	[-1.9%, -0.9%]	5
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	1.5%	[-1.9%, 4.0%]	16

Binary size

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

Bootstrap: 738.413s -> 735.235s (-0.43%)
Artifact size: 328.22 MiB -> 328.15 MiB (-0.02%)

cjgillot · 2024-07-04T23:05:59Z

r? compiler

cjgillot · 2024-07-04T23:08:07Z

Perf: the results are an improvement almost everywhere. The regressions (externs, unused-warnings) appear due to more calls to shallow_lint_levels_on and associated query overhead.

cjgillot · 2024-07-04T23:08:23Z

cc @xFrednet

cjgillot · 2024-08-31T16:17:58Z

r? compiler

jieyouxu

Thanks, this LGTM for the most part, but I'm not super sure about how the AttrId <-> HirId mapping is constructed.

jieyouxu · 2024-09-01T10:08:18Z

compiler/rustc_lint/src/expect.rs

+            // Some attributes appear multiple times in HIR, to ensure they are correctly taken
+            // into account where they matter. This means we cannot just associate the AttrId to
+            // the first HirId where we see it, but need to check it actually appears in a lint
+            // level.
+            // FIXME(cjgillot): Can this cause an attribute to appear in multiple expectation ids?


Discussion: this seems a little suspicious to me, in that I would've expected AttrId <-> HirId to be a bijective map. "Some attributes appear multiple times in HIR, to ensure they are correctly taken into account where they matter" almost sounds like a bug to me?

This FIXME makes me sad too, and triggered the discussion here: #127884 (comment)

This map is not bijective, but it is well-defined in one direction. From a pair (HirId, attr_index), you'll get a single AttrId, but the converse is not true.

Thanks for the clarification, that discussion was enlightening. I feel like this will cause a bug somewhere but I also can't immediately come up with an example for. As I can't come up with an example, I'm inclined to merge this as-is and see what cases we run into in practice through fuzzing and whatnot. And see if we can come up with a scheme that is more robust.

(Side remark: our lint infra and lint level computation is unfortunately incredibly convoluted)

tests/ui/error-codes/E0602.stderr

jieyouxu

I'm okay with accepting the FIXME even if it may cause bugs, because the logic before the rework is also not exactly watertight. The perf change seems to be mixed leaning towards more improvements so I don't feel like it's blocking.

jieyouxu · 2024-09-01T12:31:05Z

In any case, I think rewriting into a single pass also makes the logic clearer than before.

@bors r+

bors · 2024-09-01T12:31:07Z

📌 Commit ff1fc68 has been approved by jieyouxu

It is now in the queue for this repository.

jieyouxu · 2024-09-01T12:32:06Z

@bors rollup=never (perf significant)

bors · 2024-09-01T15:50:51Z

⌛ Testing commit ff1fc68 with merge a48861a...

bors · 2024-09-01T18:38:43Z

☀️ Test successful - checks-actions
Approved by: jieyouxu
Pushing a48861a to master...

rust-timer · 2024-09-01T20:17:58Z

Finished benchmarking commit (a48861a): comparison URL.

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

Next Steps: If you can justify the regressions found in this perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please open an issue or create a new PR that fixes the regressions, add a comment linking to the newly created issue or PR, and then add the perf-regression-triaged label to this PR.

@rustbot label: +perf-regression
cc @rust-lang/wg-compiler-performance

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)	0.4%	[0.3%, 0.6%]	10
Regressions ❌ (secondary)	0.8%	[0.4%, 1.5%]	10
Improvements ✅ (primary)	-0.5%	[-1.5%, -0.2%]	71
Improvements ✅ (secondary)	-0.8%	[-1.8%, -0.3%]	52
All ❌✅ (primary)	-0.4%	[-1.5%, 0.6%]	81

Max RSS (memory usage)

Results (primary 0.3%, secondary 0.2%)

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)	0.8%	[0.4%, 2.4%]	49
Regressions ❌ (secondary)	1.0%	[0.4%, 3.1%]	42
Improvements ✅ (primary)	-1.0%	[-3.3%, -0.4%]	18
Improvements ✅ (secondary)	-0.7%	[-1.9%, -0.4%]	39
All ❌✅ (primary)	0.3%	[-3.3%, 2.4%]	67

Cycles

Results (primary -0.4%, secondary -0.0%)

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)	1.0%	[0.4%, 4.2%]	22
Regressions ❌ (secondary)	1.2%	[0.4%, 5.2%]	57
Improvements ✅ (primary)	-0.8%	[-3.3%, -0.4%]	83
Improvements ✅ (secondary)	-1.2%	[-4.4%, -0.4%]	58
All ❌✅ (primary)	-0.4%	[-3.3%, 4.2%]	105

Binary size

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

Bootstrap: 790.965s -> 787.22s (-0.47%)
Artifact size: 338.44 MiB -> 338.38 MiB (-0.02%)

lqd · 2024-09-03T15:20:26Z

updated results:

Kobzol · 2024-09-03T15:56:21Z

Many more improvements than regressions.

@rustbot label: +perf-regression-triaged

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 Jul 4, 2024

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 Jul 4, 2024

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 Jul 4, 2024

cjgillot marked this pull request as ready for review July 4, 2024 23:05

rustbot assigned compiler-errors Jul 4, 2024

cjgillot mentioned this pull request Jul 19, 2024

#[expect(unused_imports)] does not work correctly on grouped imports #127884

Closed

cjgillot added 4 commits August 31, 2024 13:58

Use AttrId key for unstable<->stable expectation map.

4928b22

Rewrite lint_expectations in a single pass.

111b0a9

Remove attr_id from stable lint ids.

5f1f45b

Simplify lint source computation.

ff1fc68

cjgillot force-pushed the single-expect branch from 66a9985 to ff1fc68 Compare August 31, 2024 14:03

rustbot assigned jieyouxu and unassigned compiler-errors Aug 31, 2024

jieyouxu reviewed Sep 1, 2024

View reviewed changes

jieyouxu added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Sep 1, 2024

jieyouxu approved these changes Sep 1, 2024

View reviewed changes

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-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Sep 1, 2024

bors added the merged-by-bors This PR was explicitly merged by bors. label Sep 1, 2024

bors merged commit a48861a into rust-lang:master Sep 1, 2024
7 checks passed

rustbot added this to the 1.83.0 milestone Sep 1, 2024

cjgillot deleted the single-expect branch September 1, 2024 18:55

rustbot added the perf-regression-triaged The performance regression has been triaged. label Sep 3, 2024

Mark-Simulacrum mentioned this pull request Sep 16, 2024

Rescope temp lifetime in if-let into IfElse with migration lint #107251

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Rewrite lint_expectations in a single pass. #127313

Rewrite lint_expectations in a single pass. #127313

cjgillot commented Jul 4, 2024

cjgillot commented Jul 4, 2024

This comment has been minimized.

bors commented Jul 4, 2024

bors commented Jul 4, 2024

This comment has been minimized.

rust-timer commented Jul 4, 2024

cjgillot commented Jul 4, 2024

cjgillot commented Jul 4, 2024

cjgillot commented Jul 4, 2024

cjgillot commented Aug 31, 2024

jieyouxu left a comment

jieyouxu Sep 1, 2024 •

edited

Loading

cjgillot Sep 1, 2024

jieyouxu Sep 1, 2024 •

edited

Loading

jieyouxu Sep 1, 2024 •

edited

Loading

jieyouxu left a comment •

edited

Loading

jieyouxu commented Sep 1, 2024

bors commented Sep 1, 2024

jieyouxu commented Sep 1, 2024

bors commented Sep 1, 2024

bors commented Sep 1, 2024

rust-timer commented Sep 1, 2024

lqd commented Sep 3, 2024 •

edited

Loading

Kobzol commented Sep 3, 2024

Rewrite lint_expectations in a single pass. #127313

Rewrite lint_expectations in a single pass. #127313

Conversation

cjgillot commented Jul 4, 2024

cjgillot commented Jul 4, 2024

This comment has been minimized.

bors commented Jul 4, 2024

bors commented Jul 4, 2024

This comment has been minimized.

rust-timer commented Jul 4, 2024

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

cjgillot commented Jul 4, 2024

cjgillot commented Jul 4, 2024

cjgillot commented Jul 4, 2024

cjgillot commented Aug 31, 2024

jieyouxu left a comment

Choose a reason for hiding this comment

jieyouxu Sep 1, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jieyouxu Sep 1, 2024 • edited Loading

Choose a reason for hiding this comment

jieyouxu Sep 1, 2024 • edited Loading

Choose a reason for hiding this comment

jieyouxu left a comment • edited Loading

Choose a reason for hiding this comment

jieyouxu commented Sep 1, 2024

bors commented Sep 1, 2024

jieyouxu commented Sep 1, 2024

bors commented Sep 1, 2024

bors commented Sep 1, 2024

rust-timer commented Sep 1, 2024

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

lqd commented Sep 3, 2024 • edited Loading

Kobzol commented Sep 3, 2024

jieyouxu Sep 1, 2024 •

edited

Loading

jieyouxu Sep 1, 2024 •

edited

Loading

jieyouxu Sep 1, 2024 •

edited

Loading

jieyouxu left a comment •

edited

Loading

lqd commented Sep 3, 2024 •

edited

Loading