Tighten `fn_decl_span` for async blocks #127058

compiler-errors · 2024-06-27T19:00:22Z

Tightens the span of async {} blocks in diagnostics, and subsequently async closures and async fns, by actually setting the fn_decl_span correctly. This is kinda a follow-up on #125078, but it fixes the problem in a more general way.

I think the diagnostics are significantly improved, since we no longer have a bunch of overlapping spans. I'll point out one caveat where I think the diagnostic may get a bit more confusing, but where I don't think it matters.

r? @estebank or @oli-obk or someone else on wg-diag or compiler i dont really care lol

compiler-errors · 2024-06-27T19:04:59Z

tests/ui/async-await/issues/issue-78938-async-block.stderr

-LL | |         game_loop(Arc::clone(&room_ref))
-   | |                               -------- `room_ref` is borrowed here
-LL | |     });
-   | |_____^ may outlive borrowed value `room_ref`


Here's the only place that the diagnostic arguably got worse. I kinda don't think it matters, though, because the primary message still says:

async block may outlive the current function, but it borrows room_ref, which is owned by the current function

So it's still clear that the problem is with an async block.

If you care enough about this for it to get fixed, then I suggest we change the label to state what may outlive the borrowed value, i.e.:

LL | let gameloop_handle = spawn(async { | ^^^^^ async block may outlive borrowed value `room_ref`

rustbot · 2024-06-27T19:19:27Z

Some changes occurred in src/tools/clippy

cc @rust-lang/clippy

Some changes occurred in src/tools/rustfmt

cc @rust-lang/rustfmt

oli-obk · 2024-06-27T20:03:18Z

r? @oli-obk

@bors rollup

bors · 2024-06-27T20:03:20Z

📌 Commit 789ee88 has been approved by oli-obk

It is now in the queue for this repository.

estebank · 2024-06-27T20:08:54Z

tests/ui/impl-trait/issues/issue-78722.stderr

@@ -8,7 +8,7 @@ LL |         let f: F = async { 1 };
   = help: add `#![feature(const_async_blocks)]` to the crate attributes to enable
   = note: this compiler was built on YYYY-MM-DD; consider upgrading it if it is out of date

-error[E0271]: expected `{async block@$DIR/issue-78722.rs:10:13: 10:21}` to be a future that resolves to `u8`, but it resolves to `()`
+error[E0271]: expected `{async block@$DIR/issue-78722.rs:10:13: 10:18}` to be a future that resolves to `u8`, but it resolves to `()`


Note to self: we should special case the wording here to be something like

Suggested change

error[E0271]: expected `{async block@$DIR/issue-78722.rs:10:13: 10:18}` to be a future that resolves to `u8`, but it resolves to `()`

error[E0271]: expected async block to be a future that resolves to `u8`, but it resolves to `()`

estebank · 2024-06-27T20:10:44Z

tests/ui/coroutine/gen_block_is_no_future.stderr

+error[E0277]: `{gen block@$DIR/gen_block_is_no_future.rs:5:5: 5:8}` is not a future
  --> $DIR/gen_block_is_no_future.rs:4:13
   |
 LL | fn foo() -> impl std::future::Future {
-   |             ^^^^^^^^^^^^^^^^^^^^^^^^ `{gen block@$DIR/gen_block_is_no_future.rs:5:5: 5:21}` is not a future
+   |             ^^^^^^^^^^^^^^^^^^^^^^^^ `{gen block@$DIR/gen_block_is_no_future.rs:5:5: 5:8}` is not a future
   |
-   = help: the trait `Future` is not implemented for `{gen block@$DIR/gen_block_is_no_future.rs:5:5: 5:21}`
+   = help: the trait `Future` is not implemented for `{gen block@$DIR/gen_block_is_no_future.rs:5:5: 5:8}`


We need to figure out how to point at the gen/async block in E0277s because the error as is isn't as clear as it should be: we point at the span that introduced the obligation, not where the obligation wasn't fulfilled.

This is a general problem with opaques.

we point at the span that introduced the obligation

This is the span that introduces the obligation. You're thinking about the span that introduces the expression that assigns the hidden type of the opaque. Determining what that span is is difficult to impossible without hacks that I would rather not try to land, because they will almost certainly regress with the new solver.

estebank · 2024-06-27T20:11:48Z

tests/ui/async-await/issue-86507.stderr

@@ -13,7 +13,7 @@ note: captured value is not `Send` because `&` references cannot be sent unless
   |
 LL |                     let x = x;
   |                             ^ has type `&T` which is not `Send`, because `T` is not `Sync`
-   = note: required for the cast from `Pin<Box<{async block@$DIR/issue-86507.rs:18:17: 20:18}>>` to `Pin<Box<(dyn Future<Output = ()> + Send + 'async_trait)>>`
+   = note: required for the cast from `Pin<Box<{async block@$DIR/issue-86507.rs:18:17: 18:27}>>` to `Pin<Box<(dyn Future<Output = ()> + Send + 'async_trait)>>`


Going through all of there errors, I wonder if mentioning the end of the span is necessary in messages...

estebank · 2024-06-27T20:12:03Z

tests/ui/async-await/issue-74072-lifetime-name-annotations.stderr

+LL |     (async move || {
+   |                  - return type of async closure is &'1 i32
+...
 LL |         let y = &*x;
   |                 --- `*x` is borrowed here
 LL |         *x += 1;
   |         ^^^^^^^ `*x` is assigned to here but it was already borrowed
 LL |         y
   |         - returning this value requires that `*x` is borrowed for `'1`
-LL |     })()
-   |     - return type of async closure is &'1 i32


estebank · 2024-06-27T20:12:41Z

tests/ui/async-await/async-closures/wrong-fn-kind.stderr

+LL |     needs_async_fn(async || {
+   |     -------------- ^^^^^^^^
+   |     |              |
+   |     |              cannot borrow as mutable
+   |     |              in this closure
+   |     expects `Fn` instead of `FnMut`


Maybe we should now remove the "in this closure" label, as it is no longer relevant?

…s, r=oli-obk Tighten `fn_decl_span` for async blocks Tightens the span of `async {}` blocks in diagnostics, and subsequently async closures and async fns, by actually setting the `fn_decl_span` correctly. This is kinda a follow-up on rust-lang#125078, but it fixes the problem in a more general way. I think the diagnostics are significantly improved, since we no longer have a bunch of overlapping spans. I'll point out one caveat where I think the diagnostic may get a bit more confusing, but where I don't think it matters. r? `@estebank` or `@oli-obk` or someone else on wg-diag or compiler i dont really care lol

…s, r=oli-obk Tighten `fn_decl_span` for async blocks Tightens the span of `async {}` blocks in diagnostics, and subsequently async closures and async fns, by actually setting the `fn_decl_span` correctly. This is kinda a follow-up on rust-lang#125078, but it fixes the problem in a more general way. I think the diagnostics are significantly improved, since we no longer have a bunch of overlapping spans. I'll point out one caveat where I think the diagnostic may get a bit more confusing, but where I don't think it matters. r? ``@estebank`` or ``@oli-obk`` or someone else on wg-diag or compiler i dont really care lol

…s, r=oli-obk Tighten `fn_decl_span` for async blocks Tightens the span of `async {}` blocks in diagnostics, and subsequently async closures and async fns, by actually setting the `fn_decl_span` correctly. This is kinda a follow-up on rust-lang#125078, but it fixes the problem in a more general way. I think the diagnostics are significantly improved, since we no longer have a bunch of overlapping spans. I'll point out one caveat where I think the diagnostic may get a bit more confusing, but where I don't think it matters. r? ```@estebank``` or ```@oli-obk``` or someone else on wg-diag or compiler i dont really care lol

…iaskrgr Rollup of 6 pull requests Successful merges: - rust-lang#124741 (patchable-function-entry: Add unstable compiler flag and attribute) - rust-lang#126470 (make cargo submodule optional) - rust-lang#126701 (ignore `llvm::Lld` if lld is not enabled) - rust-lang#126956 (core: avoid `extern type`s in formatting infrastructure) - rust-lang#126970 (Simplify `str::clone_into`) - rust-lang#127058 (Tighten `fn_decl_span` for async blocks) r? `@ghost` `@rustbot` modify labels: rollup

…iaskrgr Rollup of 6 pull requests Successful merges: - rust-lang#124741 (patchable-function-entry: Add unstable compiler flag and attribute) - rust-lang#126470 (make cargo submodule optional) - rust-lang#126956 (core: avoid `extern type`s in formatting infrastructure) - rust-lang#126970 (Simplify `str::clone_into`) - rust-lang#127022 (Support fetching `Attribute` of items.) - rust-lang#127058 (Tighten `fn_decl_span` for async blocks) r? `@ghost` `@rustbot` modify labels: rollup

Rollup merge of rust-lang#127058 - compiler-errors:tighten-async-spans, r=oli-obk Tighten `fn_decl_span` for async blocks Tightens the span of `async {}` blocks in diagnostics, and subsequently async closures and async fns, by actually setting the `fn_decl_span` correctly. This is kinda a follow-up on rust-lang#125078, but it fixes the problem in a more general way. I think the diagnostics are significantly improved, since we no longer have a bunch of overlapping spans. I'll point out one caveat where I think the diagnostic may get a bit more confusing, but where I don't think it matters. r? ````@estebank```` or ````@oli-obk```` or someone else on wg-diag or compiler i dont really care lol

…s, r=oli-obk Tighten `fn_decl_span` for async blocks Tightens the span of `async {}` blocks in diagnostics, and subsequently async closures and async fns, by actually setting the `fn_decl_span` correctly. This is kinda a follow-up on rust-lang#125078, but it fixes the problem in a more general way. I think the diagnostics are significantly improved, since we no longer have a bunch of overlapping spans. I'll point out one caveat where I think the diagnostic may get a bit more confusing, but where I don't think it matters. r? ````@estebank```` or ````@oli-obk```` or someone else on wg-diag or compiler i dont really care lol

rustbot assigned estebank Jun 27, 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. WG-trait-system-refactor The Rustc Trait System Refactor Initiative (-Znext-solver) labels Jun 27, 2024

compiler-errors commented Jun 27, 2024

View reviewed changes

This comment has been minimized.

Sign in to view

Tighten spans for async blocks

789ee88

compiler-errors force-pushed the tighten-async-spans branch from 696e2ed to 789ee88 Compare June 27, 2024 19:19

rustbot assigned oli-obk and unassigned estebank Jun 27, 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 Jun 27, 2024

estebank approved these changes Jun 27, 2024

View reviewed changes

jhpratt mentioned this pull request Jun 28, 2024

Rollup of 12 pull requests #127064

Closed

matthiaskrgr mentioned this pull request Jun 28, 2024

Rollup of 7 pull requests #127065

Closed

matthiaskrgr mentioned this pull request Jun 28, 2024

Rollup of 6 pull requests #127067

Closed

matthiaskrgr mentioned this pull request Jun 28, 2024

Rollup of 6 pull requests #127076

Merged

bors merged commit 89a0cfe into rust-lang:master Jun 28, 2024
6 checks passed

rustbot added this to the 1.81.0 milestone Jun 28, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Tighten `fn_decl_span` for async blocks #127058

Tighten `fn_decl_span` for async blocks #127058

compiler-errors commented Jun 27, 2024

compiler-errors Jun 27, 2024

This comment has been minimized.

rustbot commented Jun 27, 2024

oli-obk commented Jun 27, 2024

bors commented Jun 27, 2024

estebank Jun 27, 2024

estebank Jun 27, 2024

compiler-errors Jun 27, 2024

estebank Jun 27, 2024

estebank Jun 27, 2024

estebank Jun 27, 2024

	error[E0271]: expected `{async block@$DIR/issue-78722.rs:10:13: 10:18}` to be a future that resolves to `u8`, but it resolves to `()`
	error[E0271]: expected async block to be a future that resolves to `u8`, but it resolves to `()`

Tighten fn_decl_span for async blocks #127058

Tighten fn_decl_span for async blocks #127058

Conversation

compiler-errors commented Jun 27, 2024

compiler-errors Jun 27, 2024

Choose a reason for hiding this comment

This comment has been minimized.

rustbot commented Jun 27, 2024

oli-obk commented Jun 27, 2024

bors commented Jun 27, 2024

estebank Jun 27, 2024

Choose a reason for hiding this comment

estebank Jun 27, 2024

Choose a reason for hiding this comment

compiler-errors Jun 27, 2024

Choose a reason for hiding this comment

estebank Jun 27, 2024

Choose a reason for hiding this comment

estebank Jun 27, 2024

Choose a reason for hiding this comment

estebank Jun 27, 2024

Choose a reason for hiding this comment

Tighten `fn_decl_span` for async blocks #127058

Tighten `fn_decl_span` for async blocks #127058