simplify some `proc_macro` things by cyrgani · Pull Request #157271 · rust-lang/rust

cyrgani · 2026-06-01T20:04:37Z

Each commit should be reviewable on its own. Locally, this also resulted in some slightly better perf results.

rustbot · 2026-06-01T20:04:41Z

rust-analyzer is developed in its own repository. If possible, consider making this change to rust-lang/rust-analyzer instead.

cc @rust-lang/rust-analyzer

rustbot · 2026-06-01T20:04:44Z

r? @petrochenkov

rustbot has assigned @petrochenkov.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

Owners of files modified in this PR: @petrochenkov

Kobzol · 2026-06-01T20:05:46Z

@bors try @rust-timer queue

simplify some `proc_macro` things

rust-bors · 2026-06-01T22:20:48Z

☀️ Try build successful (CI)
Build commit: aba743d (aba743d72e1a2d02bf787988ac919c3c416f243c, parent: c0bb140a37c81cf59a0b40c21c9413059644e294)

rust-timer · 2026-06-01T23:01:38Z

Finished benchmarking commit (aba743d): comparison URL.

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

Benchmarking means the PR may be perf-sensitive. It's automatically marked not fit for rolling up. Overriding is possible but disadvised: it risks changing compiler perf.

Next, please: If you can, justify the regressions found in this try perf run in writing along with @rustbot label: +perf-regression-triaged. If not, fix the regressions and do another perf run. Neutral or positive results will clear the label automatically.

@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)	0.4%	[0.1%, 1.1%]	27
Regressions ❌ (secondary)	0.2%	[0.2%, 0.2%]	1
Improvements ✅ (primary)	-0.8%	[-1.9%, -0.2%]	13
Improvements ✅ (secondary)	-1.9%	[-4.2%, -0.1%]	13
All ❌✅ (primary)	-0.0%	[-1.9%, 1.1%]	40

Max RSS (memory usage)

Results (primary 2.4%)

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

	mean	range	count
Regressions ❌ (primary)	2.4%	[2.4%, 2.4%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.4%	[2.4%, 2.4%]	1

Cycles

Results (primary 0.4%, secondary -5.6%)

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

	mean	range	count
Regressions ❌ (primary)	1.9%	[1.7%, 2.1%]	2
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-2.7%	[-2.7%, -2.7%]	1
Improvements ✅ (secondary)	-5.6%	[-5.6%, -5.6%]	1
All ❌✅ (primary)	0.4%	[-2.7%, 2.1%]	3

Binary size

Results (primary -0.9%, secondary -2.1%)

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)	-0.9%	[-1.3%, -0.4%]	17
Improvements ✅ (secondary)	-2.1%	[-3.0%, -0.4%]	7
All ❌✅ (primary)	-0.9%	[-1.3%, -0.4%]	17

Bootstrap: 511.375s -> 517.888s (1.27%)
Artifact size: 400.63 MiB -> 400.68 MiB (0.01%)

bjorn3 · 2026-06-02T08:20:06Z


-    fn ts_drop(&mut self, stream: Self::TokenStream) {
-        drop(stream);
-    }


Wouldn't this cause all tokenstreams to be unnecessarily retained on the server side until the end of the proc macro invocation even if the client drops it earlier?

View changes since the review

Yeah, the motivation is missing for this change now - what did this drop do, what happens now when it's removed, why it's good.

The main reason for this change is that currently all TokenStream implementations are wrapped in an Arc or Rc and I would not expect that to change. So none of them need anything special happening to drop them.
It seemed wasteful to me to have to go through the bridge just to drop the value considering this and the assumption that in most of these cases, dropping a token stream would just decrease the Arcs refcount by 1 without actually freeing the underlying value already. So I expected the perf gain from no longer going through the bridge to drop them to outweigh the cost of storing them a little longer. But I may very well be mistaken here.

The server side needs to be notified somehow that the client side dropped the TokenStream. Otherwise it will stick around in the OwnedStore until the proc macro finishes. This notifying is done by passing it to ts_drop.

petrochenkov · 2026-06-02T10:09:48Z

-    // We have the array method separate from extending from a slice. This is
-    // because in the case of small arrays, codegen can be more efficient
-    // (avoiding a memmove call). With extend_from_slice, LLVM at least
-    // currently is not able to make that optimization.


Any specific reason to remove this optimization?
It would probably be better to benchmark it separately in any case.

View changes since the review

I've removed this change from this PR. It's not too important either.

petrochenkov · 2026-06-02T10:11:06Z

    len: usize,
    capacity: usize,
-    reserve: extern "C" fn(Buffer, usize) -> Buffer,
-    drop: extern "C" fn(Buffer),


Why is it better to directly manage Buffer memory without using Vec?

View changes since the review

Locally, this showed some promising improvements of instruction counts for some crates. I've removed it from this PR again for now since it seems to affect others negatively, its now in #157451.

mark `Encode`, `Decode`, `Mark` impls as `#[inline]` Related to rust-lang#157271 (though it was not a part of it). This looked pretty promising locally.

Rollup merge of #157450 - cyrgani:pm-inline, r=petrochenkov mark `Encode`, `Decode`, `Mark` impls as `#[inline]` Related to #157271 (though it was not a part of it). This looked pretty promising locally.

rustbot assigned petrochenkov Jun 1, 2026

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 1, 2026

This comment has been minimized.

Sign in to view

rust-bors Bot pushed a commit that referenced this pull request Jun 1, 2026

Auto merge of #157271 - cyrgani:pm-cleaning, r=<try>

aba743d

simplify some `proc_macro` things

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 1, 2026

bjorn3 reviewed Jun 2, 2026

View reviewed changes

petrochenkov reviewed Jun 2, 2026

View reviewed changes

petrochenkov 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 Jun 2, 2026

This comment has been minimized.

Sign in to view

cyrgani added 5 commits June 3, 2026 20:09

remove some redundant closures

beb1d9e

remove ts_drop bridge method

bd45858

simplify Buffer

1354363

rename CustomDerives trait_name to name

be8f3eb

set issue = none for proc_macro_internals

750ea38

cyrgani mentioned this pull request Jun 4, 2026

mark Encode, Decode, Mark impls as #[inline] #157450

Merged

rust-cloud-vms Bot force-pushed the pm-cleaning branch from 7f9f32a to 750ea38 Compare June 4, 2026 20:45

cyrgani mentioned this pull request Jun 5, 2026

[PERF] directly manage Buffer memory without using Vec #157451

Draft

Uh oh!

Conversation

cyrgani commented Jun 1, 2026

Uh oh!

rustbot commented Jun 1, 2026

Uh oh!

rustbot commented Jun 1, 2026

Uh oh!

Kobzol commented Jun 1, 2026

Uh oh!

This comment has been minimized.

This comment has been minimized.

rust-bors Bot commented Jun 1, 2026

Uh oh!

This comment has been minimized.

rust-timer commented Jun 1, 2026

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

Uh oh!

bjorn3 Jun 2, 2026 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

petrochenkov Jun 2, 2026 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

petrochenkov Jun 2, 2026 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

This comment has been minimized.

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

bjorn3 Jun 2, 2026 •

edited by rustbot

Loading

petrochenkov Jun 2, 2026 •

edited by rustbot

Loading

petrochenkov Jun 2, 2026 •

edited by rustbot

Loading