Rollup of 5 perf-sensitive pull requests by JonathanBrouwer · Pull Request #164931 · rust-lang/rust

JonathanBrouwer · 2026-10-25T20:18:23Z

Successful merges:

Add various #[inline] annotations for simple functions on hot paths that don't get inlined #163845 (Add various #[inline] annotations for simple functions on hot paths that don't get inlined)
Don't emit allow attributes in try/contract desugaring #164122 (Don't emit allow attributes in try/contract desugaring)
create fast path for attributes created in builtin macro expansions #164260 (create fast path for attributes created in builtin macro expansions)
Deduplicate DefId encodings in the query cache #164300 (Deduplicate DefId encodings in the query cache)
Revert "rustc_codegen_gcc subtree update" #164929 (Revert "rustc_codegen_gcc subtree update")

Failed merges:

Split LintLevelsBuilder::add into add_ast and add_hir #164790 (Split LintLevelsBuilder::add into add_ast and add_hir)

r? @ghost

Create a similar rollup

This avoids needing the overhead of a call before discovering that the `Cow` is already owned.

These show up in hot paths, and inlining should allow optimizing them.

These are simple wrappers but nonetheless don't get inlined and show up in hot paths.

These are simple wrappers.

…ree-sync, r=GuillaumeGomez" This reverts commit 36096d0, reversing changes made to c1ed5e8.

…Simulacrum Add various `#[inline]` annotations for simple functions on hot paths that don't get inlined All of these are simple, likely to be optimizable once inlined into their caller, and found on hot paths in profiles (I used `clap_derive` for the profiling) as non-inlined functions.

…=hanna-kruppe Don't emit allow attributes in try/contract desugaring

@nnethercote

…rouwer,nnethercote create fast path for attributes created in builtin macro expansions Emitting attributes in rustc_builtin_macros involves creating an `ast::Attribute` and token trees, just to parse that into a parsed attribute almost immediately. This PR sidesteps that by creating a new `ast::Attribute` variant that does not do this. I've taken care to only do this for the derive and test macros, so that they never re-enter the parser. This results in some moderate performance gains. It'll also take away almost all of the regressions in rust-lang#164217 (comment) (see rust-lang#164281 (comment) for a perf run of this and that PR stacked) r? @nnethercote

…r=beetrees Deduplicate DefId encodings in the query cache I used a similar approach for DefId encoding and decoding, based on the existing Symbol encoder and decoder. https://github.com/rust-lang/rust/blob/81900216ce9ee0b48f2c563c696bfca98cfba755/compiler/rustc_middle/src/query/on_disk_cache.rs#L818 https://github.com/rust-lang/rust/blob/81900216ce9ee0b48f2c563c696bfca98cfba755/compiler/rustc_middle/src/query/on_disk_cache.rs#L447

Revert "rustc_codegen_gcc subtree update" Reverts rust-lang#164869 because it broke main again See rust-lang#164921 (comment) See discussion in [#t-infra > Tree ops @ 💬](https://rust-lang.zulipchat.com/#narrow/channel/242791-t-infra/topic/Tree.20ops/near/634790354)

JonathanBrouwer · 2026-10-25T20:18:38Z

@bors r+ p=5 force

rust-bors · 2026-10-25T20:18:40Z

📌 Commit 90dd286 has been approved by JonathanBrouwer

It is now in the queue for this repository.

rust-bors · 2026-10-25T23:22:45Z

☀️ Test successful - CI
Approved by: JonathanBrouwer
Duration: 3h 3m 48s
Pushing 76456a6 to main...

github-actions · 2026-10-25T23:26:07Z

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 7d0f9b4 (parent) -> 76456a6 (this PR)

Test differences

Show 54 test diffs

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

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard 76456a688bc0ed897fa23a677d4ddb5b3fdb575a --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

test-x86_64-msvc-ext2: 1h 3m -> 1h 49m (+71.4%)
test-i686-gnu-2: 1h 3m -> 1h 49m (+71.0%)
test-x86_64-msvc-ext3: 1h 8m -> 1h 53m (+65.8%)
test-i686-gnu-nopt-2: 1h 28m -> 2h 25m (+65.1%)
test-x86_64-mingw-2: 1h 43m -> 2h 48m (+62.1%)
test-x86_64-rust-for-linux: 34m 46s -> 53m 36s (+54.2%)
dist-loongarch64-musl: 1h 19m -> 2h 2m (+54.0%)
test-i686-msvc: 1h 34m -> 2h 20m (+48.7%)
test-i686-gnu-1: 1h 29m -> 2h 13m (+48.5%)
dist-armv7-linux: 1h 5m -> 1h 36m (+47.3%)

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-bors · 2026-10-26T00:29:04Z

📌 Perf builds for each rolled up PR:

PR#	Message	Perf Build Sha
#163845	Add various `#[inline]` annotations for simple functions on…	`06a58da5923c185c34e3853fd55231eedb906798` (link)
#164122	Don't emit allow attributes in try/contract desugaring	`eeff8b41823572a7bea6e97cd298f99f40dc6aa3` (link)
#164260	create fast path for attributes created in builtin macro ex…	`5d7c38d55377bccc1b2e7c738a36506dd19c1fef` (link)
#164300	Deduplicate DefId encodings in the query cache	`69cf1614881cbb8da5bb9f31876f2f52c6cb2d76` (link)
#164929	Revert "rustc_codegen_gcc subtree update"	`359bf16b79413f5e9b0791853524659cc651df5e` (link)

parent commit: 7d0f9b4ec5

In the case of a perf regression, run the following command with the SHAs of each PR you suspect might be the cause: @rust-timer triage $SHA $SHA $SHA..., or run @rust-timer triage all to benchmark all rollup members.

rust-timer · 2026-10-26T00:33:41Z

Finished benchmarking commit (76456a6): comparison URL.

Overall result: ❌✅ regressions and improvements - please read:

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)	0.7%	[0.1%, 2.5%]	88
Regressions ❌ (secondary)	0.7%	[0.1%, 2.2%]	64
Improvements ✅ (primary)	-0.8%	[-12.7%, -0.1%]	110
Improvements ✅ (secondary)	-1.0%	[-3.4%, -0.1%]	36
All ❌✅ (primary)	-0.2%	[-12.7%, 2.5%]	198

Max RSS (memory usage)

Results (primary -1.3%, secondary -1.6%)

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

	mean	range	count
Regressions ❌ (primary)	1.3%	[0.6%, 2.6%]	5
Regressions ❌ (secondary)	2.4%	[0.6%, 5.0%]	4
Improvements ✅ (primary)	-1.6%	[-3.6%, -0.4%]	56
Improvements ✅ (secondary)	-2.8%	[-5.4%, -0.6%]	13
All ❌✅ (primary)	-1.3%	[-3.6%, 2.6%]	61

Cycles

Results (primary -2.1%, secondary -0.3%)

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

	mean	range	count
Regressions ❌ (primary)	1.4%	[0.8%, 2.9%]	4
Regressions ❌ (secondary)	2.5%	[0.5%, 5.2%]	10
Improvements ✅ (primary)	-2.8%	[-13.1%, -0.5%]	22
Improvements ✅ (secondary)	-3.2%	[-6.0%, -0.6%]	10
All ❌✅ (primary)	-2.1%	[-13.1%, 2.9%]	26

Binary size

Results (primary -1.2%, 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)	2.5%	[2.5%, 2.5%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-1.3%	[-2.7%, -0.1%]	37
Improvements ✅ (secondary)	-0.6%	[-1.1%, -0.1%]	9
All ❌✅ (primary)	-1.2%	[-2.7%, 2.5%]	38

Bootstrap: 490.188s -> 489.982s (-0.04%)
Artifact size: 405.91 MiB -> 405.87 MiB (-0.01%)

JonathanBrouwer · 2026-10-26T01:42:20Z

@rust-timer triage all

rust-timer · 2026-10-26T01:42:24Z

Running triage with 42 benchmarks

Triage only executes the benchmarks on rollup members, that were changed significantly on the rollup.
For this rollup, these benchmarks are:

await-call-tree
bitmaps-3.2.1
cargo-0.87.1
clap_derive-4.5.32
clap_derive-4.5.32-nll
coercions
cranelift-codegen-0.119.0
ctfe-stress-5
deeply-nested-multi
derive
diesel-2.2.10
diesel-2.2.10-new-solver
enum-stress
externs
eza-0.21.2
html5ever-0.31.0
hyper-1.6.0
image-0.25.6
issue-58319
issue-88862
libc-0.2.172
many-assoc-items
nalgebra-0.33.0
regex-automata-0.4.8
regression-31157
ripgrep-14.1.1
ripgrep-14.1.1-nll
ripgrep-14.1.1-tiny
serde-1.0.219
serde-1.0.219-nll
serde_derive-1.0.219
stm32f4-0.15.1
syn-2.0.101
syn-2.0.101-new-solver
tokio-1.53.1
tt-muncher
tuple-stress
typenum-1.18.0
ucd
unicode-normalization-0.1.24
wf-projection-stress-65510
wg-grammar

#163845 `06a58da` Add various `#[inline]` annotations for simple functions on hot paths that don't get inlined

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)	0.8%	[0.4%, 2.0%]	5
Regressions ❌ (secondary)	0.3%	[0.2%, 0.4%]	4
Improvements ✅ (primary)	-2.1%	[-12.8%, -0.2%]	19
Improvements ✅ (secondary)	-1.4%	[-3.4%, -0.3%]	3
All ❌✅ (primary)	-1.5%	[-12.8%, 2.0%]	24

Max RSS (memory usage)

Results (primary 0.3%, 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)	0.8%	[0.4%, 1.5%]	16
Regressions ❌ (secondary)	0.7%	[0.6%, 0.9%]	3
Improvements ✅ (primary)	-1.6%	[-3.3%, -0.9%]	4
Improvements ✅ (secondary)	-2.1%	[-2.1%, -2.1%]	1
All ❌✅ (primary)	0.3%	[-3.3%, 1.5%]	20

Cycles

Results (primary -2.1%, secondary -0.5%)

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

	mean	range	count
Regressions ❌ (primary)	1.2%	[0.6%, 2.9%]	4
Regressions ❌ (secondary)	1.3%	[0.7%, 2.2%]	3
Improvements ✅ (primary)	-3.2%	[-13.5%, -0.4%]	13
Improvements ✅ (secondary)	-1.6%	[-3.3%, -0.5%]	5
All ❌✅ (primary)	-2.1%	[-13.5%, 2.9%]	17

Binary size

Results (primary -1.2%, 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)	2.5%	[2.5%, 2.5%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-1.3%	[-2.7%, -0.1%]	38
Improvements ✅ (secondary)	-0.6%	[-1.1%, -0.1%]	9
All ❌✅ (primary)	-1.2%	[-2.7%, 2.5%]	39

#164122 `eeff8b4` Don't emit allow attributes in try/contract desugaring

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%]	1
Improvements ✅ (primary)	-0.7%	[-1.9%, -0.1%]	103
Improvements ✅ (secondary)	-0.4%	[-0.6%, -0.1%]	23
All ❌✅ (primary)	-0.7%	[-1.9%, -0.1%]	103

Max RSS (memory usage)

Results (primary 0.1%, secondary 0.4%)

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

	mean	range	count
Regressions ❌ (primary)	1.0%	[0.5%, 1.7%]	13
Regressions ❌ (secondary)	0.7%	[0.5%, 1.2%]	7
Improvements ✅ (primary)	-1.1%	[-3.3%, -0.5%]	10
Improvements ✅ (secondary)	-1.4%	[-1.4%, -1.4%]	1
All ❌✅ (primary)	0.1%	[-3.3%, 1.7%]	23

Cycles

Results (primary -0.9%, secondary 1.4%)

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

	mean	range	count
Regressions ❌ (primary)	0.8%	[0.4%, 1.1%]	4
Regressions ❌ (secondary)	2.2%	[0.6%, 5.1%]	5
Improvements ✅ (primary)	-1.2%	[-3.2%, -0.4%]	20
Improvements ✅ (secondary)	-0.5%	[-0.5%, -0.4%]	2
All ❌✅ (primary)	-0.9%	[-3.2%, 1.1%]	24

Binary size

This perf run didn't have relevant results for this metric.

#164260 `5d7c38d` create fast path for attributes created in builtin macro expansions

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)	-0.4%	[-1.3%, -0.1%]	76
Improvements ✅ (secondary)	-1.3%	[-3.9%, -0.2%]	26
All ❌✅ (primary)	-0.4%	[-1.3%, -0.1%]	76

Max RSS (memory usage)

Results (primary 0.4%, secondary 1.1%)

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

	mean	range	count
Regressions ❌ (primary)	0.9%	[0.4%, 2.0%]	12
Regressions ❌ (secondary)	1.3%	[0.5%, 4.2%]	8
Improvements ✅ (primary)	-1.3%	[-2.9%, -0.5%]	4
Improvements ✅ (secondary)	-0.7%	[-0.7%, -0.7%]	1
All ❌✅ (primary)	0.4%	[-2.9%, 2.0%]	16

Cycles

Results (primary -0.6%, secondary -1.9%)

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

	mean	range	count
Regressions ❌ (primary)	0.5%	[0.4%, 0.5%]	5
Regressions ❌ (secondary)	1.2%	[0.5%, 2.7%]	4
Improvements ✅ (primary)	-1.0%	[-2.1%, -0.4%]	17
Improvements ✅ (secondary)	-3.6%	[-6.8%, -0.5%]	7
All ❌✅ (primary)	-0.6%	[-2.1%, 0.5%]	22

Binary size

This perf run didn't have relevant results for this metric.

#164300 `69cf161` Deduplicate DefId encodings in the query cache

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)	0.8%	[0.1%, 1.6%]	149
Regressions ❌ (secondary)	0.7%	[0.1%, 2.2%]	81
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.8%	[0.1%, 1.6%]	149

Max RSS (memory usage)

Results (primary -1.1%, secondary -2.2%)

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

	mean	range	count
Regressions ❌ (primary)	0.9%	[0.4%, 1.8%]	10
Regressions ❌ (secondary)	0.8%	[0.4%, 1.6%]	4
Improvements ✅ (primary)	-1.7%	[-6.5%, -0.5%]	37
Improvements ✅ (secondary)	-3.4%	[-5.2%, -0.9%]	10
All ❌✅ (primary)	-1.1%	[-6.5%, 1.8%]	47

Cycles

Results (primary 0.0%, secondary 0.8%)

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

	mean	range	count
Regressions ❌ (primary)	0.8%	[0.5%, 1.2%]	8
Regressions ❌ (secondary)	1.5%	[1.0%, 2.4%]	4
Improvements ✅ (primary)	-0.9%	[-2.2%, -0.4%]	7
Improvements ✅ (secondary)	-0.6%	[-0.7%, -0.5%]	2
All ❌✅ (primary)	0.0%	[-2.2%, 1.2%]	15

Binary size

This perf run didn't have relevant results for this metric.

#164929 `359bf16` Revert "rustc_codegen_gcc subtree update"

Queued 359bf16 with parent 7d0f9b4, future comparison URL.
There are currently 15 preceding artifacts in the queue.
It will probably take at least ~5.9 hours until the benchmark run finishes.

joshtriplett and others added 14 commits October 5, 2026 14:46

Inline Cow::to_mut

d3cefda

This avoids needing the overhead of a call before discovering that the `Cow` is already owned.

Inline Vec fold operations

48c4946

These show up in hot paths, and inlining should allow optimizing them.

Inline Copied fold operations

3e91052

These are simple wrappers but nonetheless don't get inlined and show up in hot paths.

Inline Cloned fold operations

d136cc4

These are simple wrappers.

Inline fold operations for Chain and Map, for performance

39c2880

deduplicate DefId encodings in the query cache

8190021

Don't emit allow attributes in try/contract desugaring

e1663b6

create fast path for attributes created in builtin macro expansions

db2155b

Revert "Rollup merge of rust-lang#164869 - GuillaumeGomez:cg_gcc-subt…

ceeb0e0

…ree-sync, r=GuillaumeGomez" This reverts commit 36096d0, reversing changes made to c1ed5e8.

Rollup merge of rust-lang#164122 - mejrs:no-new_ast_lowering-attrs, r…

0a4720d

…=hanna-kruppe Don't emit allow attributes in try/contract desugaring

rust-bors Bot added the rollup A PR which is a rollup label Oct 25, 2026

rust-bors Bot 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 Oct 25, 2026

This comment has been minimized.

Sign in to view

JonathanBrouwer changed the title ~~Rollup of 5 pull requests~~ Rollup of 5 perf-sensitive pull requests Oct 25, 2026

rust-bors Bot added merged-by-bors This PR was explicitly merged by bors. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Oct 25, 2026

rust-bors Bot merged commit 76456a6 into rust-lang:main Oct 25, 2026
15 checks passed

rustbot added this to the 1.101.0 milestone Oct 25, 2026

rustbot added the perf-regression Performance regression. label Oct 26, 2026

Navigation Menu

Repository navigation

Rollup of 5 perf-sensitive pull requests - #164931

Rollup of 5 perf-sensitive pull requests#164931
rust-bors[bot] merged 14 commits into
rust-lang:mainfrom
JonathanBrouwer:rollup-evgAZRd

JonathanBrouwer commented Oct 25, 2026

Uh oh!

JonathanBrouwer commented Oct 25, 2026

Uh oh!

rust-bors Bot commented Oct 25, 2026

Uh oh!

This comment has been minimized.

rust-bors Bot commented Oct 25, 2026

Uh oh!

Uh oh!

github-actions Bot commented Oct 25, 2026

Uh oh!

rust-bors Bot commented Oct 26, 2026

Uh oh!

rust-timer commented Oct 26, 2026

Uh oh!

JonathanBrouwer commented Oct 26, 2026

Uh oh!

rust-timer commented Oct 26, 2026 •

edited

Loading

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

Conversation

JonathanBrouwer commented Oct 25, 2026

Uh oh!

JonathanBrouwer commented Oct 25, 2026

Uh oh!

rust-bors Bot commented Oct 25, 2026

Uh oh!

This comment has been minimized.

rust-bors Bot commented Oct 25, 2026

Uh oh!

Uh oh!

github-actions Bot commented Oct 25, 2026

Test differences

Job duration changes

Uh oh!

rust-bors Bot commented Oct 26, 2026

Uh oh!

rust-timer commented Oct 26, 2026

Overall result: ❌✅ regressions and improvements - please read:

Uh oh!

JonathanBrouwer commented Oct 26, 2026

Uh oh!

rust-timer commented Oct 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

#163845 06a58da Add various #[inline] annotations for simple functions on hot paths that don't get inlined

#164122 eeff8b4 Don't emit allow attributes in try/contract desugaring

#164260 5d7c38d create fast path for attributes created in builtin macro expansions

#164300 69cf161 Deduplicate DefId encodings in the query cache

#164929 359bf16 Revert "rustc_codegen_gcc subtree update"

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

rust-timer commented Oct 26, 2026 •

edited

Loading

#163845 `06a58da` Add various `#[inline]` annotations for simple functions on hot paths that don't get inlined

#164122 `eeff8b4` Don't emit allow attributes in try/contract desugaring

#164260 `5d7c38d` create fast path for attributes created in builtin macro expansions

#164300 `69cf161` Deduplicate DefId encodings in the query cache

#164929 `359bf16` Revert "rustc_codegen_gcc subtree update"