Rollup of 9 pull requests by tgross35 · Pull Request #142794 · rust-lang/rust

tgross35 · 2025-06-20T17:36:03Z

Successful merges:

Add trim_prefix and trim_suffix methods for both slice and str types. #142331 (Add trim_prefix and trim_suffix methods for both slice and str types.)
Rework #[cold] attribute parser #142491 (Rework #[cold] attribute parser)
Fix missing docs in rustc_attr_parsing #142494 (Fix missing docs in rustc_attr_parsing)
Better template for #[repr] attributes #142495 (Better template for #[repr] attributes)
Fix random failure when JS code is executed when the whole file was not read yet #142497 (Fix random failure when JS code is executed when the whole file was not read yet)
Ensure copy* intrinsics also perform the static self-init checks #142575 (Ensure copy* intrinsics also perform the static self-init checks)
Refactor Translator #142650 (Refactor Translator)
mbe: Refactor transcription #142713 (mbe: Refactor transcription)
rustdoc: Remove FormatRenderer::cache #142755 (rustdoc: Remove FormatRenderer::cache)

r? @ghost
@rustbot modify labels: rollup

Implements `trim_prefix` and `trim_suffix` methods for both `slice` and `str` types which remove at most one occurrence of a prefix/suffix while always returning a string/slice (rather than Option), enabling easy method chaining.

…ot read yet

… of `search.js`

Be more consistent with the otherwise top-down organization of this file.

Introduce `MacroTcbCtx` that holds everything relevant to transcription. This allows for the following changes: * Split `transcribe_sequence` and `transcribe_metavar` out of the heavily nested `transcribe` * Split `metavar_expr_concat` out of `transcribe_metavar_expr` This is a nonfunctional change.

We only called it it one place, which isn't generic and can be replaced with a field access.

Add `trim_prefix` and `trim_suffix` methods for both `slice` and `str` types. Implements `trim_prefix` and `trim_suffix` methods for both `slice` and `str` types, which remove at most one occurrence of a prefix/suffix while always returning a string/slice (rather than Option), enabling easy method chaining. ## Tracking issue rust-lang#142312 ## API ```rust impl str { pub fn trim_prefix<P: Pattern>(&self, prefix: P) -> &str; pub fn trim_suffix<P: Pattern>(&self, suffix: P) -> &str where for<'a> P::Searcher<'a>: ReverseSearcher<'a>; } impl<T> [T] { pub fn trim_prefix<P: SlicePattern<Item = T> + ?Sized>(&self, prefix: &P) -> &[T] where T: PartialEq; pub fn trim_suffix<P: SlicePattern<Item = T> + ?Sized>(&self, suffix: &P) -> &[T] where T: PartialEq; } ``` ## Examples ```rust // Method chaining assert_eq!(" <https://example.com/> ".trim().trim_prefix('<').trim_suffix('>').trim(), "https://example.com/"); // Slices let v = &[10, 40, 30]; assert_eq!(v.trim_prefix(&[10]), &[40, 30][..]); ``` ## ACP Originally proposed in rust-lang/libs-team#597

Rework #[cold] attribute parser r? `@oli-obk`

…docs, r=oli-obk Fix missing docs in `rustc_attr_parsing`

… r=oli-obk Better template for `#[repr]` attributes

…ailure, r=lolbinarycat Fix random failure when JS code is executed when the whole file was not read yet Very randomly (and rarely), when I arrived on a page with `?search=something` in the URL, I got this error: ![Screenshot From 2025-06-14 11-27-46](https://github.com/user-attachments/assets/4b61b067-4e80-49c1-9a45-cff1509bf86a) Moving the `initSearch` function at the bottom to ensure everything has been loaded fixes the issue. PS: Sorry for the noise. Pushed to the wrong branch and rust-lang#142496 closed. ><

Ensure copy* intrinsics also perform the static self-init checks fixes rust-lang#142532 r? `@RalfJung`

…petrochenkov Refactor Translator My main motivation was to simplify the usage of `SilentEmitter` for users like rustfmt. A few refactoring opportunities arose along the way. * Replace `Translate` trait with `Translator` struct * Replace `Emitter: Translate` with `Emitter::translator` * Split `SilentEmitter` into `FatalOnlyEmitter` and `SilentEmitter`

…r=petrochenkov mbe: Refactor transcription Introduce `MacroTcbCtx` that holds everything relevant to transcription. This allows for the following changes: * Split `transcribe_sequence` and `transcribe_metavar` out of the heavily nested `transcribe` * Split `metavar_expr_concat` out of `transcribe_metavar_expr` This is a nonfunctional change.

…aumeGomez rustdoc: Remove `FormatRenderer::cache` We only called it it one place, which isn't generic and can be replaced with a field access.

tgross35 · 2025-06-20T20:29:03Z

Ah you answered right before me, don't mind me either then 😆

bors · 2025-06-20T23:09:51Z

⌛ Testing commit 61f4918 with merge 15c701f...

bors · 2025-06-21T02:04:15Z

☀️ Test successful - checks-actions
Approved by: tgross35
Pushing 15c701f to master...

rust-timer · 2025-06-21T02:06:48Z

📌 Perf builds for each rolled up PR:

PR#	Message	Perf Build Sha
#142331	Add `trim_prefix` and `trim_suffix` methods for both `slice…	`79ed30102bd770b06326970da3b4878abbddf95b` (link)
#142491	Rework #[cold] attribute parser	`8b18a200eeeda37682c24943b879cbaf143edc20` (link)
#142494	Fix missing docs in `rustc_attr_parsing`	`2228a644b92354dcf92d1721fa0d8323a6b6843c` (link)
#142495	Better template for `#[repr]` attributes	`9ed520219e1601b9a934fad72587556e9bffa24a` (link)
#142497	Fix random failure when JS code is executed when the whole …	`c78a1d814221fe60327ddba3834b766481dcac2b` (link)
#142575	Ensure copy* intrinsics also perform the static self-init c…	`543e757cbf569b43fe5482931286df0a8e9ab298` (link)
#142650	Refactor Translator	`62dadd2f9b7510ff67bcb3f6485d5bdf5f744959` (link)
#142713	mbe: Refactor transcription	`7c928508c11816b71001f101aba6b5ffa0beca87` (link)
#142755	rustdoc: Remove `FormatRenderer::cache`	`5024179adaf907aa6792c12cced0877466f370c4` (link)

previous master: 5526a2f47c

In the case of a perf regression, run the following command for each PR you suspect might be the cause: @rust-timer build $SHA

github-actions · 2025-06-21T02:07:24Z

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 5526a2f (parent) -> 15c701f (this PR)

Test differences

Show 184 test diffs

Stage 1

[ui] tests/ui/attributes/expected-word.rs: [missing] -> pass (J0)
[ui] tests/ui/statics/read_before_init.rs: [missing] -> pass (J0)

Stage 2

[ui] tests/ui/attributes/expected-word.rs: [missing] -> pass (J1)
[ui] tests/ui/statics/read_before_init.rs: [missing] -> pass (J1)

Additionally, 180 doctest diffs were found. These are ignored, as they are noisy.

Job group index

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard 15c701fbc995eb6c5b3a86021c18185f8eee020d --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

x86_64-apple-1: 8266.9s -> 6243.9s (-24.5%)
dist-aarch64-apple: 6044.5s -> 4719.8s (-21.9%)
dist-apple-various: 6674.5s -> 5442.3s (-18.5%)
mingw-check-1: 1573.0s -> 1861.0s (18.3%)
x86_64-msvc-1: 8357.9s -> 9655.2s (15.5%)
aarch64-gnu-debug: 3447.4s -> 3960.6s (14.9%)
x86_64-rust-for-linux: 2544.7s -> 2910.3s (14.4%)
i686-gnu-2: 5339.8s -> 6029.5s (12.9%)
i686-gnu-nopt-1: 7091.4s -> 7915.6s (11.6%)
mingw-check-2: 1838.9s -> 2052.2s (11.6%)

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 · 2025-06-21T05:11:32Z

Finished benchmarking commit (15c701f): comparison URL.

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

Our benchmarks found a performance regression caused by this PR.
This might be an actual regression, but it can also be just noise.

Next Steps:

If the regression was expected or you think it can be justified,
please write a comment with sufficient written justification, and add
@rustbot label: +perf-regression-triaged to it, to mark the regression as triaged.
If you think that you know of a way to resolve the regression, try to create
a new PR with a fix for the regression.
If you do not understand the regression or you think that it is just noise,
you can ask the @rust-lang/wg-compiler-performance working group for help (members of this group
were already notified of this PR).

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

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)	2.9%	[2.9%, 2.9%]	1
Regressions ❌ (secondary)	0.8%	[0.2%, 1.2%]	10
Improvements ✅ (primary)	-0.2%	[-0.2%, -0.2%]	4
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.4%	[-0.2%, 2.9%]	5

Max RSS (memory usage)

Results (primary -3.0%, secondary 2.8%)

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)	5.2%	[4.6%, 5.9%]	2
Improvements ✅ (primary)	-3.0%	[-3.0%, -3.0%]	1
Improvements ✅ (secondary)	-2.0%	[-2.0%, -2.0%]	1
All ❌✅ (primary)	-3.0%	[-3.0%, -3.0%]	1

Cycles

Results (primary 2.8%, secondary 0.6%)

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

	mean	range	count
Regressions ❌ (primary)	2.8%	[2.8%, 2.8%]	1
Regressions ❌ (secondary)	2.0%	[2.0%, 2.1%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.3%	[-2.3%, -2.3%]	1
All ❌✅ (primary)	2.8%	[2.8%, 2.8%]	1

Binary size

Results (primary 1.1%)

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

	mean	range	count
Regressions ❌ (primary)	1.1%	[1.1%, 1.1%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	1.1%	[1.1%, 1.1%]	1

Bootstrap: 692.994s -> 690.617s (-0.34%)
Artifact size: 371.93 MiB -> 371.83 MiB (-0.02%)

tgross35 · 2025-06-21T05:17:14Z

Good chance it's #142713 based on regressions in macro expansion

tgross35 · 2025-06-21T08:05:40Z

^ The above has identical regression results

rylev · 2025-06-24T11:38:06Z

@tgross35 any thoughts on how to win back the performance lost here?

tgross35 · 2025-06-24T17:15:02Z

I have been doing some experiments at #142815, but unfortunately haven't had any luck. I'm going to keep trying a few things but am a bit pessimistic here - the PR was only refactoring, which is making me think that using early return from smaller functions is just getting us worse codegen than the huge match statement.

tgross35 · 2025-06-25T05:45:42Z

I still don't really know how to interpret perf results well, but is it possible the clap_derive regression is a bit spurious? The graphs on the diff seem to indicate more time spent in the backend, which doesn't make any sense for a frontend-only refactoring, and it seems to have come back down from the blip here https://perf.rust-lang.org/index.html?start=2025-06-20&end=2025-06-21&benchmark=clap_derive-4.5.32&profile=opt&scenario=full&stat=instructions%3Au&kind=raw.

The tt muncher benchmarks seem more legit, and make sense with the change. I'd like to recommend triaging as accepted here; ignoring Clap, this is only a pretty small regression on a heavily recursive macro. Unfortunately I'm not having any luck getting perf back since #142815 isn't really successful.

Still trying one last thing out at #142990.

panstromek · 2025-06-25T07:35:46Z

Yea, clap_derive is bimodal, that regression is spurious.

The tt_muncher isn't and it's a bit suspicous when it's caused by a refactor. Whether to accept or not I guess depends on what's the purpose of that PR (I don't have any context for that). I assume it's laying the groundwork for some feature? If that's the case, it seems fine (we're talking about ~10ms on pretty sensitive benchmarks here IIUC).

tgross35 · 2025-06-25T08:07:48Z

Whether to accept or not I guess depends on what's the purpose of that PR (I don't have any context for that). I assume it's laying the groundwork for some feature?

Indeed; #142713 can stand on its own as a cleanup of messy code, but I have some more significant changes to metavariable expressions based on it. Since #[inline(always)] didn't seem to make any difference, my guess is that LLVM is having some trouble optimizing through the context struct everything got collected to. Which is pretty unfortunate, but I don't think there is a clean way to work around it (short of reverting).

panstromek · 2025-06-25T08:25:59Z

Ok. If you plan to make more changes in this area, then the perf impact will probably change anyway, so I think we can accept this and keep an eye on perf for future changes.

@rustbot label: +perf-regression-triaged

Rollup of 9 pull requests Successful merges: - rust-lang#142331 (Add `trim_prefix` and `trim_suffix` methods for both `slice` and `str` types.) - rust-lang#142491 (Rework #[cold] attribute parser) - rust-lang#142494 (Fix missing docs in `rustc_attr_parsing`) - rust-lang#142495 (Better template for `#[repr]` attributes) - rust-lang#142497 (Fix random failure when JS code is executed when the whole file was not read yet) - rust-lang#142575 (Ensure copy* intrinsics also perform the static self-init checks) - rust-lang#142650 (Refactor Translator) - rust-lang#142713 (mbe: Refactor transcription) - rust-lang#142755 (rustdoc: Remove `FormatRenderer::cache`) r? `@ghost` `@rustbot` modify labels: rollup

deven and others added 24 commits June 11, 2025 00:27

Add trim_prefix and trim_suffix for slice and str.

32cb8f1

Implements `trim_prefix` and `trim_suffix` methods for both `slice` and `str` types which remove at most one occurrence of a prefix/suffix while always returning a string/slice (rather than Option), enabling easy method chaining.

Fix random failure when JS code is executed when the whole file was n…

6ca4639

…ot read yet

Add comment to prevent having code other than initSearch at the end…

2eccf18

… of `search.js`

better template for repr attributes

815f0b7

fixup missing docs in attr parsing

7f49de2

mbe: Move transcribe_metavar_expr directly after transcribe

496cbe1

Be more consistent with the otherwise top-down organization of this file.

Extract Translator struct

07b9bb1

Rename SilentEmitter -> FatalOnlyEmitter

316f63b

Extract SilentEmitter

3388d83

rustdoc: Remove FormatRenderer::cache

aab941f

We only called it it one place, which isn't generic and can be replaced with a field access.

Add regression test

e159cf0

Ensure copy* intrinsics also perform the static self-init checks

cfc22cf

cold

de0fd27

expected word diagnostic test

b9107a8

Rollup merge of rust-lang#142491 - jdonszelmann:cold, r=oli-obk

f0374dc

Rework #[cold] attribute parser r? `@oli-obk`

Rollup merge of rust-lang#142494 - jdonszelmann:missing-attr-parsing-…

d2a00eb

…docs, r=oli-obk Fix missing docs in `rustc_attr_parsing`

Rollup merge of rust-lang#142495 - jdonszelmann:better-repr-template,…

1e2ac53

… r=oli-obk Better template for `#[repr]` attributes

Rollup merge of rust-lang#142575 - oli-obk:sneaky-self-init, r=RalfJung

5d6ecf5

Ensure copy* intrinsics also perform the static self-init checks fixes rust-lang#142532 r? `@RalfJung`

Rollup merge of rust-lang#142755 - aDotInTheVoid:rdj-shattrs, r=Guill…

61f4918

…aumeGomez rustdoc: Remove `FormatRenderer::cache` We only called it it one place, which isn't generic and can be replaced with a field access.

bors added the merged-by-bors This PR was explicitly merged by bors. label Jun 21, 2025

bors merged commit 15c701f into rust-lang:master Jun 21, 2025
11 checks passed

rustbot added this to the 1.89.0 milestone Jun 21, 2025

tgross35 deleted the rollup-iae7okj branch June 21, 2025 02:06

rustbot added the perf-regression Performance regression. label Jun 21, 2025

tgross35 mentioned this pull request Jun 21, 2025

mbe: Refactor transcription #142713

Merged

rustbot added the perf-regression-triaged The performance regression has been triaged. label Jun 25, 2025

Distortedlogic mentioned this pull request Jun 25, 2025

rust-analyzer shows false positive E0308 type mismatch errors with PyO3 macros rust-lang/rust-analyzer#20068

Closed

Uh oh!

Conversation

tgross35 commented Jun 20, 2025 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

tgross35 commented Jun 20, 2025

Uh oh!

bors commented Jun 20, 2025

Uh oh!

bors commented Jun 21, 2025

Uh oh!

Uh oh!

rust-timer commented Jun 21, 2025

Uh oh!

github-actions bot commented Jun 21, 2025

Test differences

Stage 1

Stage 2

Job duration changes

Uh oh!

rust-timer commented Jun 21, 2025

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

Instruction count

Max RSS (memory usage)

Cycles

Binary size

Uh oh!

tgross35 commented Jun 21, 2025

Uh oh!

tgross35 commented Jun 21, 2025

Uh oh!

rylev commented Jun 24, 2025

Uh oh!

tgross35 commented Jun 24, 2025

Uh oh!

tgross35 commented Jun 25, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

panstromek commented Jun 25, 2025

Uh oh!

tgross35 commented Jun 25, 2025

Uh oh!

panstromek commented Jun 25, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

12 participants

tgross35 commented Jun 20, 2025 •

edited by rustbot

Loading

tgross35 commented Jun 25, 2025 •

edited

Loading