Replace `TypeWalker` usage with `TypeVisitor` in `wf.rs` #122150

ShoyuVanilla · 2024-03-07T17:19:53Z

rustbot · 2024-03-07T17:20:02Z

rustbot has assigned @davidtwco.
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

ShoyuVanilla · 2024-03-07T17:31:20Z

tests/ui/const-generics/generic_const_exprs/const_kind_expr/issue_114151.rs

@@ -20,6 +20,7 @@ where
    //~^^^ ERROR: function takes 1 generic argument but 2 generic arguments were supplied
    //~^^^^ ERROR: unconstrained generic constant
    //~^^^^^ ERROR: unconstrained generic constant `{const expr}`
+    //~^^^^^^ ERROR: unconstrained generic constant `{const expr}`


The test output is changed, but I think that this would be correct because L + 1 + L is actually nested const expr.

The difference might be came from the difference in some recursive visiting behaviour between TypeWalker and Const::super_visit_with;

rust/compiler/rustc_middle/src/ty/walk.rs

Lines 205 to 236 in 52f8aec

GenericArgKind::Const(parent_ct) => {

stack.push(parent_ct.ty().into());

match parent_ct.kind() {

ty::ConstKind::Infer(_)

| ty::ConstKind::Param(_)

| ty::ConstKind::Placeholder(_)

| ty::ConstKind::Bound(..)

| ty::ConstKind::Value(_)

| ty::ConstKind::Error(_) => {}

ty::ConstKind::Expr(expr) => match expr {

ty::Expr::UnOp(_, v) => push_inner(stack, v.into()),

ty::Expr::Binop(_, l, r) => {

push_inner(stack, r.into());

push_inner(stack, l.into())

}

ty::Expr::FunctionCall(func, args) => {

for a in args.iter().rev() {

push_inner(stack, a.into());

}

push_inner(stack, func.into());

}

ty::Expr::Cast(_, c, t) => {

push_inner(stack, t.into());

push_inner(stack, c.into());

}

},

ty::ConstKind::Unevaluated(ct) => {

stack.extend(ct.args.iter().rev());

}

}

rust/compiler/rustc_middle/src/ty/structural_impls.rs

Lines 762 to 779 in 52f8aec

impl<'tcx> TypeSuperVisitable<TyCtxt<'tcx>> for ty::Const<'tcx> {

fn super_visit_with<V: TypeVisitor<TyCtxt<'tcx>>>(&self, visitor: &mut V) -> V::Result {

try_visit!(self.ty().visit_with(visitor));

match self.kind() {

ConstKind::Param(p) => p.visit_with(visitor),

ConstKind::Infer(i) => i.visit_with(visitor),

ConstKind::Bound(d, b) => {

try_visit!(d.visit_with(visitor));

b.visit_with(visitor)

}

ConstKind::Placeholder(p) => p.visit_with(visitor),

ConstKind::Unevaluated(uv) => uv.visit_with(visitor),

ConstKind::Value(v) => v.visit_with(visitor),

ConstKind::Error(e) => e.visit_with(visitor),

ConstKind::Expr(e) => e.visit_with(visitor),

}

}

}

the former doesn't track down some inner values.
It might possible to getting the same behaviour by changing implementation of impl TypeVisitor for WfPredicates, but it would be unintuitive and since this output seems better - at least I think so 😅 - I think that it would be good in this way.

compiler-errors · 2024-03-07T17:42:56Z

r? lcnr

lcnr

thanks, r=me after nits

lcnr · 2024-03-08T08:39:51Z

compiler/rustc_trait_selection/src/traits/wf.rs

-            }
-
-            debug!(?self.out);
+        match arg.unpack() {


arg.visit_with(self)

lcnr · 2024-03-08T08:42:14Z

compiler/rustc_trait_selection/src/traits/wf.rs

+                // let the visitor iterate into the argument/return
+                // types appearing in the fn signature


Suggested change

// let the visitor iterate into the argument/return

// types appearing in the fn signature

// Let the visitor iterate into the argument/return

// types appearing in the fn signature.

lcnr · 2024-03-08T08:44:15Z

compiler/rustc_trait_selection/src/traits/wf.rs

+        c.super_visit_with(self)
+    }
+}
+


can you explicitly impl fn visit_predicate and ICE there? We should never check that predicates are well formed and handle wf for clauses separately

lcnr · 2024-03-08T12:08:19Z

@bors try @rust-timer queue

checking for perf

bors · 2024-03-08T12:09:30Z

⌛ Trying commit 6721b39 with merge 96b719d...

…try> Replace `TypeWalker` usage with `TypeVisitor` in `wf.rs` Resolves rust-lang#121693

bors · 2024-03-08T13:36:54Z

☀️ Try build successful - checks-actions
Build commit: 96b719d (96b719dcf319bd0b616754be35dc7c575118e6ca)

rust-timer · 2024-03-08T14:51:46Z

Finished benchmarking commit (96b719d): 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
Regressions ❌ (secondary)	0.7%	[0.2%, 1.3%]	6
Improvements ✅ (primary)	-0.3%	[-0.5%, -0.2%]	34
Improvements ✅ (secondary)	-0.3%	[-0.4%, -0.3%]	4
All ❌✅ (primary)	-0.3%	[-0.5%, -0.2%]	34

Max RSS (memory usage)

Results

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)	-7.8%	[-7.8%, -7.8%]	1
Improvements ✅ (secondary)	-2.3%	[-2.3%, -2.3%]	1
All ❌✅ (primary)	-7.8%	[-7.8%, -7.8%]	1

Cycles

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

Binary size

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

Bootstrap: 652.903s -> 648.597s (-0.66%)
Artifact size: 172.55 MiB -> 172.57 MiB (0.01%)

lcnr · 2024-03-08T15:21:14Z

unlike the type walker, visitors don't use a cache by default. It feels likely that caching the walking part of wf is not necessary.

Given the mostly positive perf impact

@bors r+ rollup=never

bors · 2024-03-08T15:21:16Z

📌 Commit 6721b39 has been approved by lcnr

It is now in the queue for this repository.

bors · 2024-03-09T12:02:28Z

⌛ Testing commit 6721b39 with merge b054da8...

bors · 2024-03-09T14:17:26Z

☀️ Test successful - checks-actions
Approved by: lcnr
Pushing b054da8 to master...

rust-timer · 2024-03-09T15:58:17Z

Finished benchmarking commit (b054da8): 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
Regressions ❌ (secondary)	0.7%	[0.3%, 1.3%]	6
Improvements ✅ (primary)	-0.3%	[-0.4%, -0.2%]	6
Improvements ✅ (secondary)	-0.2%	[-0.3%, -0.2%]	3
All ❌✅ (primary)	-0.3%	[-0.4%, -0.2%]	6

Max RSS (memory usage)

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

Cycles

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

Binary size

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

Bootstrap: 649.534s -> 649.076s (-0.07%)
Artifact size: 172.59 MiB -> 172.58 MiB (-0.00%)

pnkfelix · 2024-03-12T18:37:37Z

visiting for weekly performance triage

six (secondary) regressions (to variants of unify-linearly and regression-31157) were anticipated during development
we were also expecting a broader set of 34 primary improvements. But the mean primary improvement we expected was -0.3%, which was unchanged when this actually landed.
so, marking as triaged

@rustbot label: +perf-regression-triaged

rustbot assigned davidtwco Mar 7, 2024

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 Mar 7, 2024

ShoyuVanilla commented Mar 7, 2024

View reviewed changes

ShoyuVanilla marked this pull request as ready for review March 7, 2024 17:38

rustbot assigned lcnr and unassigned davidtwco Mar 7, 2024

lcnr reviewed Mar 8, 2024

View reviewed changes

Replace TypeWalker usage with TypeVisitor

6721b39

ShoyuVanilla force-pushed the replace-typewalker branch from dfdf227 to 6721b39 Compare March 8, 2024 11:50

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 Mar 8, 2024

bors added a commit to rust-lang-ci/rust that referenced this pull request Mar 8, 2024

Auto merge of rust-lang#122150 - ShoyuVanilla:replace-typewalker, r=<…

96b719d

…try> Replace `TypeWalker` usage with `TypeVisitor` in `wf.rs` Resolves rust-lang#121693

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 Mar 8, 2024

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-review Status: Awaiting review from the assignee but also interested parties. labels Mar 8, 2024

bors added the merged-by-bors This PR was explicitly merged by bors. label Mar 9, 2024

bors merged commit b054da8 into rust-lang:master Mar 9, 2024
12 checks passed

rustbot added this to the 1.78.0 milestone Mar 9, 2024

ShoyuVanilla deleted the replace-typewalker branch March 9, 2024 14:29

bors mentioned this pull request Mar 9, 2024

Fix representation when printing abstract consts #119212

Merged

rustbot added the perf-regression-triaged The performance regression has been triaged. label Mar 12, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Replace `TypeWalker` usage with `TypeVisitor` in `wf.rs` #122150

Replace `TypeWalker` usage with `TypeVisitor` in `wf.rs` #122150

ShoyuVanilla commented Mar 7, 2024

rustbot commented Mar 7, 2024

ShoyuVanilla Mar 7, 2024

compiler-errors commented Mar 7, 2024

lcnr left a comment

lcnr Mar 8, 2024

lcnr Mar 8, 2024

lcnr Mar 8, 2024

lcnr commented Mar 8, 2024

This comment has been minimized.

bors commented Mar 8, 2024

bors commented Mar 8, 2024

This comment has been minimized.

rust-timer commented Mar 8, 2024

lcnr commented Mar 8, 2024

bors commented Mar 8, 2024

bors commented Mar 9, 2024

bors commented Mar 9, 2024

rust-timer commented Mar 9, 2024

pnkfelix commented Mar 12, 2024 •

edited

Loading

	GenericArgKind::Const(parent_ct) => {
	stack.push(parent_ct.ty().into());
	match parent_ct.kind() {
	ty::ConstKind::Infer(_)
	\| ty::ConstKind::Param(_)
	\| ty::ConstKind::Placeholder(_)
	\| ty::ConstKind::Bound(..)
	\| ty::ConstKind::Value(_)
	\| ty::ConstKind::Error(_) => {}

	ty::ConstKind::Expr(expr) => match expr {
	ty::Expr::UnOp(_, v) => push_inner(stack, v.into()),
	ty::Expr::Binop(_, l, r) => {
	push_inner(stack, r.into());
	push_inner(stack, l.into())
	}
	ty::Expr::FunctionCall(func, args) => {
	for a in args.iter().rev() {
	push_inner(stack, a.into());
	}
	push_inner(stack, func.into());
	}
	ty::Expr::Cast(_, c, t) => {
	push_inner(stack, t.into());
	push_inner(stack, c.into());
	}
	},

	ty::ConstKind::Unevaluated(ct) => {
	stack.extend(ct.args.iter().rev());
	}
	}

	impl<'tcx> TypeSuperVisitable<TyCtxt<'tcx>> for ty::Const<'tcx> {
	fn super_visit_with<V: TypeVisitor<TyCtxt<'tcx>>>(&self, visitor: &mut V) -> V::Result {
	try_visit!(self.ty().visit_with(visitor));
	match self.kind() {
	ConstKind::Param(p) => p.visit_with(visitor),
	ConstKind::Infer(i) => i.visit_with(visitor),
	ConstKind::Bound(d, b) => {
	try_visit!(d.visit_with(visitor));
	b.visit_with(visitor)
	}
	ConstKind::Placeholder(p) => p.visit_with(visitor),
	ConstKind::Unevaluated(uv) => uv.visit_with(visitor),
	ConstKind::Value(v) => v.visit_with(visitor),
	ConstKind::Error(e) => e.visit_with(visitor),
	ConstKind::Expr(e) => e.visit_with(visitor),
	}
	}
	}

		// let the visitor iterate into the argument/return
		// types appearing in the fn signature

Replace TypeWalker usage with TypeVisitor in wf.rs #122150

Replace TypeWalker usage with TypeVisitor in wf.rs #122150

Conversation

ShoyuVanilla commented Mar 7, 2024

rustbot commented Mar 7, 2024

Choose a reason for hiding this comment

compiler-errors commented Mar 7, 2024

lcnr left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

lcnr commented Mar 8, 2024

This comment has been minimized.

bors commented Mar 8, 2024

bors commented Mar 8, 2024

This comment has been minimized.

rust-timer commented Mar 8, 2024

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

lcnr commented Mar 8, 2024

bors commented Mar 8, 2024

bors commented Mar 9, 2024

bors commented Mar 9, 2024

rust-timer commented Mar 9, 2024

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

pnkfelix commented Mar 12, 2024 • edited Loading

Replace `TypeWalker` usage with `TypeVisitor` in `wf.rs` #122150

Replace `TypeWalker` usage with `TypeVisitor` in `wf.rs` #122150

pnkfelix commented Mar 12, 2024 •

edited

Loading