remove `visit_clobber` and `DummyAstNode` #141430

fee1-dead · 2025-05-23T09:49:32Z

used for one specific niche purpose, turns out we can just remove them and inline dummy values into callsites. this cleans them up.

rustbot · 2025-05-23T09:49:36Z

r? @oli-obk

rustbot has assigned @oli-obk.
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

fmease · 2025-05-23T10:00:09Z

FYI: Something similar was tried in an early version of PR #121829. Discussion: #121829 (comment), not sure if still relevant / relevant in general. I only skimmed your & the other PR.

fee1-dead · 2025-05-23T10:12:12Z

not sure if still relevant / relevant in general

seems irrelevant to me. There has been a year since the previous comment saying there could be additional uses of DummyAstNode but new use still has not appeared.

If we want to get a dummy of something why should we use a trait? just put it e.g. as an inherent method on P<Expr> would suffice too.

fee1-dead · 2025-05-23T12:05:18Z

🤦 Sorry for the spam. I really should stop being lazy and do more local testing before pushing.

fmease · 2025-05-23T12:06:08Z

Nah, stage2 tests are always a pain ^^'

petrochenkov · 2025-05-23T13:15:04Z

I don't think this is an improvement, invocation collector works with nodes generically, abstracting a number of small operations for every node, including clobbering.
trait InvocationCollectorNode is mostly used for that, but InvocationCollectorNode + DummyAstNode for some subset as well.
The current scheme also allows to do the clobbering only when strictly necessary (unlike this PR).
I'd rather prefer to keep things as is.

DummyAstNode and visit_clobber can be moved to expand.rs though.
Or even merged into InvocationCollectorNode with panic as the default implementation, but that would shift some errors from compile time to runtime.
Maybe some specific impls of DummyAstNode can also be removed now.

petrochenkov · 2025-05-23T13:17:38Z

compiler/rustc_expand/src/expand.rs

-                            }
-                        });
+                        if let Some(expr) = opt_expr.take() {
+                            *opt_expr = vis.filter_map_expr(expr)


This change is good, it's simpler and visit_clobber is not used generically here.

petrochenkov · 2025-05-23T13:18:49Z

tests/ui-fulldeps/pprust-expr-roundtrip.rs

-                tokens: None,
-            })
-        });
+        let expr = std::mem::replace(e, Expr::dummy());


This also looks better without visit_clobber.

petrochenkov · 2025-05-23T13:20:31Z

r? @petrochenkov
@rustbot author

rustbot · 2025-05-23T13:20:34Z

Requested reviewer is already assigned to this pull request.

Please choose another assignee.

fee1-dead · 2025-05-23T14:16:01Z

I don't think this is an improvement, invocation collector works with nodes generically, abstracting a number of small operations for every node, including clobbering.

As I said previously, I don't think having that generic behavior is necessary.

visit_node only needs the input type to implement DummyAstNode for clobbering because it takes in a &mut Node, when it doesn't have to. When it simply returns the new node created, clobbering is no longer needed.

The bridge from MutVisitor to InvocationCollectorNode is the clobbering. This PR makes the clobbering explicit and specific to each bridged type. InvocationCollectorNode does not need clobbering.

petrochenkov · 2025-05-23T14:53:34Z

@bors try @rust-timer queue

remove `visit_clobber` and `DummyAstNode` used for one specific niche purpose, turns out we can just remove them and inline dummy values into callsites. this cleans them up.

bors · 2025-05-23T14:54:44Z

⌛ Trying commit 477d572 with merge ec406cc...

bors · 2025-05-23T17:03:13Z

☀️ Try build successful - checks-actions
Build commit: ec406cc (ec406cc9a2d563186bb1b283394a034f064344ec)

rust-timer · 2025-05-23T19:02:30Z

Finished benchmarking commit (ec406cc): comparison URL.

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

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 the most reliable metric that we have; it was used to determine the overall result at the top of this comment. However, even this metric can sometimes exhibit noise.

	mean	range	count
Regressions ❌ (primary)	0.3%	[0.1%, 0.7%]	85
Regressions ❌ (secondary)	0.3%	[0.1%, 0.9%]	36
Improvements ✅ (primary)	-0.2%	[-0.2%, -0.2%]	2
Improvements ✅ (secondary)	-0.3%	[-0.4%, -0.1%]	7
All ❌✅ (primary)	0.2%	[-0.2%, 0.7%]	87

Max RSS (memory usage)

Results (primary -0.6%, secondary -4.8%)

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%	[1.9%, 1.9%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-1.8%	[-1.9%, -1.7%]	2
Improvements ✅ (secondary)	-4.8%	[-4.8%, -4.8%]	1
All ❌✅ (primary)	-0.6%	[-1.9%, 1.9%]	3

Cycles

Results (primary 2.0%, secondary 2.1%)

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)	2.0%	[2.0%, 2.0%]	1
Regressions ❌ (secondary)	2.1%	[2.1%, 2.1%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.0%	[2.0%, 2.0%]	1

Binary size

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

Bootstrap: 776.992s -> 778.523s (0.20%)
Artifact size: 365.49 MiB -> 365.48 MiB (-0.00%)

petrochenkov · 2025-05-23T21:14:21Z

#141430 (comment)

Yeah, that's why the clobbering should be inside visit_node, and not done eagerly.

fee1-dead · 2025-05-24T04:03:48Z

Yeah, that's why the clobbering should be inside visit_node, and not done eagerly.

ah, alright. I'll bring it back to rustc_expand, keeping the other cleanups then.

used for one specific niche purpose. move `DummyAstNode` to `rustc_expand` as that is where it is used.

fee1-dead · 2025-05-24T07:23:08Z

@bors try @rust-timer queue

bors · 2025-05-24T07:24:18Z

⌛ Trying commit fd6eaac with merge 25ef124...

remove `visit_clobber` and `DummyAstNode` used for one specific niche purpose, turns out we can just remove them and inline dummy values into callsites. this cleans them up.

bors · 2025-05-24T09:33:03Z

☀️ Try build successful - checks-actions
Build commit: 25ef124 (25ef124a4c5f0a3a2acf384ca51af8c53d050940)

rust-timer · 2025-05-24T11:03:07Z

Finished benchmarking commit (25ef124): comparison URL.

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

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 the most reliable metric that we have; it was used to determine the overall result at the top of this comment. However, even this metric can sometimes exhibit noise.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	0.3%	[0.2%, 0.4%]	9
Improvements ✅ (primary)	-0.2%	[-0.2%, -0.1%]	3
Improvements ✅ (secondary)	-0.3%	[-0.4%, -0.2%]	8
All ❌✅ (primary)	-0.2%	[-0.2%, -0.1%]	3

Max RSS (memory usage)

Results (primary 1.8%)

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%	[1.5%, 2.0%]	2
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	1.8%	[1.5%, 2.0%]	2

Cycles

Results (primary 2.3%)

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)	2.3%	[2.3%, 2.3%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.3%	[2.3%, 2.3%]	1

Binary size

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

Bootstrap: 777.008s -> 777.565s (0.07%)
Artifact size: 366.29 MiB -> 366.27 MiB (-0.01%)

fee1-dead · 2025-05-24T11:27:50Z

Not sure why this would be a regression, would you mind reviewing and let me know what you think? @petrochenkov

@rustbot ready

rustbot assigned oli-obk May 23, 2025

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 May 23, 2025