Compute `is_late_bound_map` query separately from lifetime resolution #97415

cjgillot · 2022-05-26T07:43:53Z

This query is actually very simple, and is only useful for functions and method. It can be computed directly by fetching the HIR, with no need to embed it within the lifetime resolution visitor.

Based on #96296

rust-highfive · 2022-05-26T07:43:56Z

Some changes occurred in src/tools/clippy.

cc @rust-lang/clippy

rust-highfive · 2022-05-26T07:43:58Z

r? @estebank

(rust-highfive has picked a reviewer for you, use r? to override)

jackh726 · 2022-05-26T12:53:58Z

This is a good candidate to split out independently, if you can. I'd r+ quickly.

cjgillot · 2022-05-26T14:33:35Z

Splitting this means reimplementing ae05823. I'd like to avoid it, as #96296 FCP ends soon. I can remove d738f62 though.

estebank · 2022-05-26T22:41:27Z

src/test/ui/hygiene/hygienic-labels.rs

-        'x: loop { $e }
-        //~^ WARNING shadows a label name that is already in scope
-        //~| WARNING shadows a label name that is already in scope
-        //~| WARNING shadows a label name that is already in scope
-        //~| WARNING shadows a label name that is already in scope


Wait, this isn't a regression? The new behavior is what this should have always been, right?

If so, let's change these files to be // check-build

This is explicitly proposed to lang team in #96296.

estebank · 2022-05-26T22:43:36Z

src/test/ui/for-loop-while/label_break_value_invalid.stderr

-warning: label name `'a` shadows a label name that is already in scope
-  --> $DIR/label_break_value_invalid.rs:22:13
-   |
-LL |       let x: u8 = 'a: {
-   |                   -- first declared here
-...
-LL |               'a: {
-   |               ^^ label `'a` already in scope
-...
-LL |       let x: u8 = mac3!('b: {
-   |  _________________-
-LL | |         if true {
-LL | |             break 'a 3;
-LL | |         }
-LL | |         0
-LL | |     });
-   | |______- in this macro invocation
-   |
-   = note: this warning originates in the macro `mac3` (in Nightly builds, run with -Z macro-backtrace for more info)


These on the other hand, are a regression, right?

estebank · 2022-05-26T22:44:53Z

src/test/ui/regions/regions-name-duplicated.stderr

+error[E0392]: parameter `'a` is never used
+  --> $DIR/regions-name-duplicated.rs:1:12
+   |
+LL | struct Foo<'a, 'a> {
+   |            ^^ unused parameter
+   |
+   = help: consider removing `'a`, referring to it in a field, or using a marker such as `PhantomData`


I'd take this to be a regression... Could we expand E0392 to check for the duplicated case so that we don't emit this redundant error?

The change needs to be in a totally different place. I'll do that in a follow-up PR.

estebank · 2022-05-26T22:46:07Z

compiler/rustc_middle/src/hir/map/mod.rs

+            OwnerNode::ForeignItem(ForeignItem {
+                kind: ForeignItemKind::Fn(_, _, generics),
+                ..
+            })


Could this potentially break existing (invalid) code? Likely worth it to land anyways, but might need a crater run to gauge the fallout.

There is an inconsistency here between hir().get_generics and the custom implementation in generics_of. I am tempted to correct it by adding more cases to get_generics and use it in generics_of. I don't see how it could break code, but I'll need to review the calls before giving an definite answer.

estebank · 2022-05-26T22:56:36Z

src/test/ui/for-loop-while/label_break_value.rs

@@ -128,8 +128,8 @@ fn label_break_macro() {
        0
    };
    assert_eq!(x, 0);
-    let x: u8 = 'a: { //~ WARNING `'a` shadows a label
-        'b: { //~ WARNING `'b` shadows a label


Marc file as check-pass

It is already run-pass because it checks that the correct labels are selected in case of hygienic shadowing.

estebank · 2022-05-26T23:07:36Z

src/test/ui/for-loop-while/label_break_value_invalid.rs

@@ -20,14 +20,11 @@ fn lbv_macro_test_hygiene_respected() {
    macro_rules! mac3 {


mark file as check-pass

This file still errs on undeclared labels at lines 7 and 29.

estebank · 2022-05-26T23:45:31Z

compiler/rustc_resolve/src/late/diagnostics.rs

+                let name = self.tcx.item_name(def_id.to_def_id());
+                let span = self.tcx.def_ident_span(def_id.to_def_id())?;
+                Some((name, span))


Wouldn't this be the same as getting the ident and passing that through?

The point of this commit was to stop storing the ident of parameters in the ribs. As we only need them in the case of diagnostics, we fetch them at the last moment.
Anyway, this code will go away with #97313.

estebank

r=me with confirmation that all the tests that should be check-pass, are

estebank · 2022-06-01T00:56:33Z

r=me after making CI happy

bors · 2022-06-02T14:03:09Z

☔ The latest upstream changes (presumably #97644) made this pull request unmergeable. Please resolve the merge conflicts.

The computation is actually much simpler, and can be done by directly fetching the HIR for the `FnDecl` and its generics.

cjgillot · 2022-06-03T10:07:57Z

@bors r=estebank

bors · 2022-06-03T10:08:00Z

📌 Commit ba40fe9 has been approved by estebank

…bank Compute `is_late_bound_map` query separately from lifetime resolution This query is actually very simple, and is only useful for functions and method. It can be computed directly by fetching the HIR, with no need to embed it within the lifetime resolution visitor. Based on rust-lang#96296

Rollup of 3 pull requests Successful merges: - rust-lang#97415 (Compute `is_late_bound_map` query separately from lifetime resolution) - rust-lang#97471 (Provide more context when denying invalid type params ) - rust-lang#97681 (Add more eslint checks) Failed merges: - rust-lang#97446 (Make hir().get_generics and generics_of consistent.) r? `@ghost` `@rustbot` modify labels: rollup

…bank Compute `is_late_bound_map` query separately from lifetime resolution This query is actually very simple, and is only useful for functions and method. It can be computed directly by fetching the HIR, with no need to embed it within the lifetime resolution visitor. Based on rust-lang#96296

Fail gracefully when encountering an HRTB in APIT. Fixes rust-lang#96954 ~The first commit will be merged as part of rust-lang#97415

rust-highfive assigned estebank May 26, 2022

rustbot added T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. T-rustdoc Relevant to the rustdoc team, which will review and decide on the PR/issue. labels May 26, 2022

rust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label May 26, 2022

estebank reviewed May 26, 2022

View reviewed changes

estebank approved these changes May 26, 2022

View reviewed changes

cjgillot added S-blocked Status: Marked as blocked ❌ on something else such as an RFC or other implementation work. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels May 27, 2022

cjgillot force-pushed the is-late-bound-solo branch from 8470fca to 0385024 Compare May 29, 2022 18:50

cjgillot added 2 commits June 3, 2022 12:03

Manipulate lifetimes by LocalDefId for region resolution.

b1294e8

Compute is_late_bound in a separate query.

ba40fe9

The computation is actually much simpler, and can be done by directly fetching the HIR for the `FnDecl` and its generics.

cjgillot force-pushed the is-late-bound-solo branch from 0385024 to ba40fe9 Compare June 3, 2022 10:05

cjgillot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-blocked Status: Marked as blocked ❌ on something else such as an RFC or other implementation work. labels Jun 3, 2022

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 Jun 3, 2022

cjgillot mentioned this pull request Jun 3, 2022

Fail gracefully when encountering an HRTB in APIT. #97683

Merged

This was referenced Jun 3, 2022

Rollup of 4 pull requests #97689

Closed

Rollup of 3 pull requests #97694

Merged

bors merged commit 53ab3b2 into rust-lang:master Jun 3, 2022

rustbot added this to the 1.63.0 milestone Jun 3, 2022

cjgillot deleted the is-late-bound-solo branch June 3, 2022 20:48

cuviper mentioned this pull request Jul 13, 2022

Incremental compilation bug with async fn method capturing multiple lifetimes #98890

Closed

lqd mentioned this pull request Sep 16, 2022

Random lifetime may not live long enough from other part of the code #96645

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Compute `is_late_bound_map` query separately from lifetime resolution #97415

Compute `is_late_bound_map` query separately from lifetime resolution #97415

cjgillot commented May 26, 2022

rust-highfive commented May 26, 2022

rust-highfive commented May 26, 2022

jackh726 commented May 26, 2022

cjgillot commented May 26, 2022

estebank May 26, 2022 •

edited

Loading

cjgillot May 26, 2022

estebank May 26, 2022

cjgillot May 26, 2022

estebank May 26, 2022

cjgillot May 27, 2022

estebank May 26, 2022

cjgillot May 26, 2022

estebank May 26, 2022

cjgillot May 26, 2022

estebank May 26, 2022

cjgillot May 27, 2022

estebank May 26, 2022

cjgillot May 29, 2022

estebank left a comment •

edited

Loading

estebank commented Jun 1, 2022

bors commented Jun 2, 2022

cjgillot commented Jun 3, 2022

bors commented Jun 3, 2022

		@@ -20,14 +20,11 @@ fn lbv_macro_test_hygiene_respected() {
		macro_rules! mac3 {

Compute is_late_bound_map query separately from lifetime resolution #97415

Compute is_late_bound_map query separately from lifetime resolution #97415

Conversation

cjgillot commented May 26, 2022

rust-highfive commented May 26, 2022

rust-highfive commented May 26, 2022

jackh726 commented May 26, 2022

cjgillot commented May 26, 2022

estebank May 26, 2022 • edited Loading

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

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

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

Choose a reason for hiding this comment

Choose a reason for hiding this comment

estebank left a comment • edited Loading

Choose a reason for hiding this comment

estebank commented Jun 1, 2022

bors commented Jun 2, 2022

cjgillot commented Jun 3, 2022

bors commented Jun 3, 2022

Compute `is_late_bound_map` query separately from lifetime resolution #97415

Compute `is_late_bound_map` query separately from lifetime resolution #97415

estebank May 26, 2022 •

edited

Loading

estebank left a comment •

edited

Loading