Add `ErrorGuaranteed` to `ast::ExprKind::Err` #120586

ShE3py · 2024-02-02T18:35:44Z

See #119967 for context

      \
       \
          _~^~^~_
      \) /  o o  \ (/
        '_   -   _'
        / '-----' \

r? fmease

rustbot · 2024-02-02T18:35:53Z

Some changes occurred in src/tools/clippy

cc @rust-lang/clippy

compiler-errors · 2024-02-02T18:40:20Z

compiler/rustc_ast_lowering/src/expr.rs

+                ExprKind::Err(guar) => hir::ExprKind::Err(*guar),
+
+                ExprKind::Dummy => {
+                    span_bug!(e.span, "lowered ExprKind::Dummy")


rustbot · 2024-02-02T20:38:50Z

Some changes occurred in src/tools/rustfmt

cc @rust-lang/rustfmt

fmease

Very nice, thanks for working on this!

I have a couple of minor suggestions but there are still some parser changes I need to review. ~~Note that I probably won't approve this until CI works again~~ (oh, it runs again!). The tree is closed anyway.

~~You also need to to update src/tools/rustfmt~~ (ah, you just pushed them haha). I see that you've already updated Clippy, amazing!

compiler/rustc_ast/src/ast.rs

compiler/rustc_builtin_macros/src/concat_bytes.rs

compiler/rustc_builtin_macros/src/deriving/default.rs

fmease · 2024-02-02T20:17:01Z

compiler/rustc_expand/src/mbe/macro_rules.rs

-                    return false;
-                }
-            }
+            TokenTree::Delimited(.., del) => check_lhs_no_empty_seq(sess, &del.tts)?,


compiler/rustc_parse/src/parser/expr.rs

compiler/rustc_session/src/errors.rs

compiler/rustc_parse/src/parser/expr.rs

compiler/rustc_parse/src/parser/stmt.rs

compiler/rustc_builtin_macros/src/concat_bytes.rs

ShE3py · 2024-02-03T16:21:44Z

Refactored some code, also took the liberty to remove the ast:: invasion in the builtin macros I modified.

@rustbot review

flip1995 · 2024-02-05T09:26:35Z

src/tools/clippy/clippy_utils/src/ast_utils.rs

@@ -143,7 +143,7 @@ pub fn eq_expr(l: &Expr, r: &Expr) -> bool {
    match (&l.kind, &r.kind) {
        (Paren(l), _) => eq_expr(l, r),
        (_, Paren(r)) => eq_expr(l, r),
-        (Err, Err) => true,
+        (Err(_), Err(_)) => true,


(Dummy, Dummy) is missing from this match statement.

To mimic the old behavior (Err(_), Dummy) and (Dummy, Err(_)) may also be needed, and I wonder if (Dummy, _) | (_, Dummy) => true makes sense.

But the only two occurences of ExprKind::Dummy are in

<Expr as DummyAstNode>::dummy() which isn't actually called (I can replace its body with panic!() and both ./x test --stage 1 tests/ui/ and ./x check --stage 2 passes successfully)

rust/compiler/rustc_ast/src/mut_visit.rs

Lines 1635 to 1645 in 8c0b4f6

impl DummyAstNode for Expr {

fn dummy() -> Self {

Expr {

id: DUMMY_NODE_ID,

kind: ExprKind::Err,

span: Default::default(),

attrs: Default::default(),

tokens: Default::default(),

}

}

}

Parser::maybe_recover_trailing_expr() for avoiding creating a P<Expr>

rust/compiler/rustc_parse/src/parser/pat.rs

Lines 388 to 394 in 8c0b4f6

// Parse `?`, `.f`, `(arg0, arg1, ...)` or `[expr]` until they've all been eaten.

if let Ok(expr) = snapshot

.parse_expr_dot_or_call_with(

self.mk_expr_err(pat_span), // equivalent to transforming the parsed pattern into an `Expr`

pat_span,

AttrVec::new(),

)

So ast_utils.rs's eq_expr cannot actually have an ExprKind::Dummy, so this case can't be tested in the testsuite. I was tempted to put a panic!() to hold the choice until later if someone uses Dummy and it reaches Clippy, but if it's not tested I'd rather have a wrong diagnostic than a panic in prod.

If this can really not occur here, then I suggest to add a debug_assertion or an unreachable! in this function. I'd like this to be an exhaustive match.

fmease · 2024-02-08T12:03:56Z

Just to give an update: I'm at 35 / 39 files viewed with pending review comments. Finishing my review later.

compiler/rustc_expand/src/base.rs

nnethercote · 2024-02-14T10:48:51Z

Oh dear, I started working on this today, not realizing that someone else was working on it. I also did TyKind::Err (#121072) and LitKind::Err (mostly done, but no PR yet).

I didn't get all the way to fixing ErrKind::Err, but I think it's possible to do this without introducing ErrKind::Dummy, which would reduce the size of the diff.

compiler/rustc_session/src/errors.rs

fmease

Just posting my pending review comments. I still need to look at some of these files.

compiler/rustc_parse/src/validate_attr.rs

compiler/rustc_builtin_macros/src/source_util.rs

fmease · 2024-02-08T00:51:46Z

compiler/rustc_expand/src/base.rs

@@ -1065,7 +1071,7 @@ pub struct ExtCtxt<'a> {
    pub sess: &'a Session,
    pub ecfg: expand::ExpansionConfig<'a>,
    pub num_standard_library_imports: usize,
-    pub reduced_recursion_limit: Option<Limit>,
+    pub reduced_recursion_limit: Option<(Limit, ErrorGuaranteed)>,


Wondering if this change is strictly required or if this can stay an Option<Limit> unless it forces us to introduce delayed_bug somewhere?

rust/compiler/rustc_expand/src/expand.rs

Lines 639 to 648 in 8c0b1fc

if !recursion_limit.value_within_limit(self.cx.current_expansion.depth) {

if self.cx.reduced_recursion_limit.is_none() {

self.error_recursion_limit_reached();

}

// Reduce the recursion limit by half each time it triggers.

self.cx.reduced_recursion_limit = Some(recursion_limit / 2);

return ExpandResult::Ready(invoc.fragment_kind.dummy(invoc.span()));

}

The error is only emitted once (MacroExpander::error_recursion_limit_reached) and AstFragmentKind::dummy requires a guar.

It's possible to use DiagCtxt::has_errors().unwrap(), but we lose the guarantee that if the recursion limit was reduced, then an error was emitted.

compiler/rustc_parse/src/parser/item.rs

fmease · 2024-02-08T01:10:32Z

compiler/rustc_parse/src/parser/pat.rs

@@ -388,7 +388,7 @@ impl<'a> Parser<'a> {
        // Parse `?`, `.f`, `(arg0, arg1, ...)` or `[expr]` until they've all been eaten.
        if let Ok(expr) = snapshot
            .parse_expr_dot_or_call_with(
-                self.mk_expr_err(pat_span), // equivalent to transforming the parsed pattern into an `Expr`
+                self.mk_expr_dummy(pat_span), // equivalent to transforming the parsed pattern into an `Expr`


Oh, it's unfortunate that we have to use ExprKind::Dummy here, I wish we wouldn't introduce more uses of it but I don't see a better option here either.

fmease · 2024-02-14T11:36:00Z

[…] I think it's possible to do this without introducing ErrKind::Dummy, which would reduce the size of the diff.

@nnethercote How would you go about updating trait DummyAstNode and everything that uses it without ExprKind::Dummy?

bors · 2024-02-14T14:15:54Z

☔ The latest upstream changes (presumably #121078) made this pull request unmergeable. Please resolve the merge conflicts.

nnethercote · 2024-02-14T19:53:22Z

@nnethercote How would you go about updating trait DummyAstNode and everything that uses it without ExprKind::Dummy?

Use ExprKind::Tup(ThinVec::new()) wherever possible, and only use ExprKind::Err if absolutely necessary. DummyAstNode for PatKind and (soon) TyKind use a non-error value. And DummyResult::raw_expr is another example that uses ExprKind::Tup(ThinVec::new()) where possible.

Because I didn't fully convert ExprKind::Err myself I'm not 100% sure, but it really seems like "use Tup(ThinVec::new()) wherever possible and Err(guar) everywhere else" should be enough.

compiler/rustc_builtin_macros/src/concat.rs

compiler/rustc_expand/src/base.rs

compiler/rustc_builtin_macros/src/concat_bytes.rs

…ease Add `ErrorGuaranteed` to `ast::LitKind::Err`, `token::LitKind::Err`. Similar to recent work doing the same for `ExprKind::Err` (rust-lang#120586) and `TyKind::Err` (rust-lang#121109). r? `@oli-obk`

bors · 2024-02-26T09:49:45Z

💔 Test failed - checks-actions

fmease · 2024-02-26T10:02:15Z

FAILED: lib/liblldMachO.a
C:\a\rust\rust\mingw64\bin\ranlib.exe: could not create temporary file whilst writing archive: no more archived files

@bors retry (spurious?)

Add `ErrorGuaranteed` to `ast::ExprKind::Err` See rust-lang#119967 for context ``` \ \ _~^~^~_ \) / o o \ (/ '_ - _' / '-----' \ ``` r? fmease

bors · 2024-02-26T12:07:20Z

⌛ Testing commit 34eae07 with merge 99eba8f...

bors · 2024-02-26T12:10:50Z

💔 Test failed - checks-actions

rust-log-analyzer · 2024-02-26T12:11:45Z

A job failed! Check out the build log: (web) (plain)

Click to see the possible cause of the failure (guessed by this bot)

#1 [internal] booting buildkit
#1 pulling image moby/buildkit:buildx-stable-1
#1 pulling image moby/buildkit:buildx-stable-1 0.2s done
#1 creating container buildx_buildkit_inspiring_archimedes0
#1 30.03 error: failed to list workers: Unavailable: connection error: desc = "transport: error while dialing: dial unix /run/buildkit/buildkitd.sock: connect: no such file or directory"
#1 ERROR: exit code 1
------
 > [internal] booting buildkit:
 > [internal] booting buildkit:
30.03 error: failed to list workers: Unavailable: connection error: desc = "transport: error while dialing: dial unix /run/buildkit/buildkitd.sock: connect: no such file or directory"
ERROR: exit code 1
Command failed. Attempt 2/5:
Command failed. Attempt 2/5:
error: dial unix /run/buildkit/buildkitd.sock: connect: no such file or directory
ERROR: listing workers for Build: failed to list workers: Unavailable: connection error: desc = "error reading server preface: EOF"
Command failed. Attempt 3/5:
error: dial unix /run/buildkit/buildkitd.sock: connect: no such file or directory
ERROR: listing workers for Build: failed to list workers: Unavailable: connection error: desc = "error reading server preface: EOF"
Command failed. Attempt 4/5:
error: dial unix /run/buildkit/buildkitd.sock: connect: no such file or directory
ERROR: listing workers for Build: failed to list workers: Unavailable: connection error: desc = "error reading server preface: EOF"
Command failed. Attempt 5/5:
error: dial unix /run/buildkit/buildkitd.sock: connect: no such file or directory
ERROR: listing workers for Build: failed to list workers: Unavailable: connection error: desc = "error reading server preface: EOF"
##[error]Process completed with exit code 1.
Post job cleanup.

fmease · 2024-02-26T12:12:23Z

@bors retry

bors · 2024-02-26T14:43:19Z

⌛ Testing commit 34eae07 with merge b79db43...

bors · 2024-02-26T17:13:07Z

☀️ Test successful - checks-actions
Approved by: fmease
Pushing b79db43 to master...

…ease Add `ErrorGuaranteed` to `ast::LitKind::Err`, `token::LitKind::Err`. Similar to recent work doing the same for `ExprKind::Err` (rust-lang#120586) and `TyKind::Err` (rust-lang#121109). r? `@oli-obk`

rust-timer · 2024-02-26T18:30:00Z

Finished benchmarking commit (b79db43): comparison URL.

Overall result: no relevant changes - no action needed

@rustbot label: -perf-regression

Instruction count

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

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)	3.1%	[2.7%, 3.5%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Cycles

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)	-	-	0
Improvements ✅ (secondary)	-1.3%	[-1.3%, -1.3%]	1
All ❌✅ (primary)	-	-	0

Binary size

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

Bootstrap: 649.794s -> 649.972s (0.03%)
Artifact size: 311.12 MiB -> 311.16 MiB (0.01%)

Properly emit `expected ;` on `#[attr] expr` Fixes rust-lang#121647 See rust-lang#118182, rust-lang#120586 --- r? estebank

Rollup merge of rust-lang#121651 - ShE3py:issue-121647, r=estebank Properly emit `expected ;` on `#[attr] expr` Fixes rust-lang#121647 See rust-lang#118182, rust-lang#120586 --- r? estebank

Add `ErrorGuaranteed` to `ast::ExprKind::Err` See rust-lang#119967 for context ``` \ \ _~^~^~_ \) / o o \ (/ '_ - _' / '-----' \ ``` r? fmease

rustbot assigned fmease Feb 2, 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 Feb 2, 2024

compiler-errors reviewed Feb 2, 2024

View reviewed changes

This comment has been minimized.

Sign in to view

fmease reviewed Feb 2, 2024

View reviewed changes

fmease 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 Feb 2, 2024

ShE3py commented Feb 3, 2024

View reviewed changes

compiler/rustc_builtin_macros/src/concat_bytes.rs Outdated Show resolved Hide resolved

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Feb 3, 2024

flip1995 reviewed Feb 5, 2024

View reviewed changes

nnethercote reviewed Feb 14, 2024

View reviewed changes

compiler/rustc_expand/src/base.rs Outdated Show resolved Hide resolved

nnethercote reviewed Feb 14, 2024

View reviewed changes

compiler/rustc_session/src/errors.rs Outdated Show resolved Hide resolved

fmease reviewed Feb 14, 2024

View reviewed changes

nnethercote reviewed Feb 14, 2024

View reviewed changes

compiler/rustc_builtin_macros/src/concat.rs Outdated Show resolved Hide resolved

This was referenced Feb 14, 2024

Add an ErrorGuaranteed to ast::TyKind::Err (attempt 2) #121109

Merged

Add ErrorGuaranteed to ast::LitKind::Err, token::LitKind::Err. #121120

Merged

nnethercote reviewed Feb 15, 2024

View reviewed changes

compiler/rustc_builtin_macros/src/concat.rs Outdated Show resolved Hide resolved

nnethercote reviewed Feb 15, 2024

View reviewed changes

compiler/rustc_expand/src/base.rs Outdated Show resolved Hide resolved

nnethercote reviewed Feb 15, 2024

View reviewed changes

compiler/rustc_builtin_macros/src/concat_bytes.rs Outdated Show resolved Hide resolved

bors added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Feb 26, 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 Feb 26, 2024

bors added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Feb 26, 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 Feb 26, 2024

bors added the merged-by-bors This PR was explicitly merged by bors. label Feb 26, 2024

bors merged commit b79db43 into rust-lang:master Feb 26, 2024
12 checks passed

rustbot added this to the 1.78.0 milestone Feb 26, 2024

This was referenced Feb 26, 2024

Diagnostic renaming #121489

Merged

Use ControlFlow in visitors. #121563

Merged

matthiaskrgr mentioned this pull request Feb 26, 2024

ICE: not a tail expression #121647

Closed

ShE3py mentioned this pull request Feb 26, 2024

Properly emit expected ; on #[attr] expr #121651

Merged

matthiaskrgr added a commit to matthiaskrgr/rust that referenced this pull request Feb 26, 2024

Rollup merge of rust-lang#121651 - ShE3py:issue-121647, r=estebank

ec5c5b7

Properly emit `expected ;` on `#[attr] expr` Fixes rust-lang#121647 See rust-lang#118182, rust-lang#120586 --- r? estebank

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add `ErrorGuaranteed` to `ast::ExprKind::Err` #120586

Add `ErrorGuaranteed` to `ast::ExprKind::Err` #120586

ShE3py commented Feb 2, 2024

rustbot commented Feb 2, 2024

compiler-errors Feb 2, 2024

This comment has been minimized.

rustbot commented Feb 2, 2024

fmease left a comment •

edited

Loading

fmease Feb 2, 2024

ShE3py commented Feb 3, 2024

flip1995 Feb 5, 2024

ShE3py Feb 8, 2024

flip1995 Feb 15, 2024 •

edited

Loading

fmease commented Feb 8, 2024 •

edited

Loading

nnethercote commented Feb 14, 2024 •

edited

Loading

fmease left a comment

fmease Feb 8, 2024 •

edited

Loading

ShE3py Feb 25, 2024

fmease Feb 8, 2024

fmease commented Feb 14, 2024

bors commented Feb 14, 2024

nnethercote commented Feb 14, 2024

bors commented Feb 26, 2024

fmease commented Feb 26, 2024

bors commented Feb 26, 2024

bors commented Feb 26, 2024

rust-log-analyzer commented Feb 26, 2024

fmease commented Feb 26, 2024

bors commented Feb 26, 2024

bors commented Feb 26, 2024

rust-timer commented Feb 26, 2024

	impl DummyAstNode for Expr {
	fn dummy() -> Self {
	Expr {
	id: DUMMY_NODE_ID,
	kind: ExprKind::Err,
	span: Default::default(),
	attrs: Default::default(),
	tokens: Default::default(),
	}
	}
	}

	// Parse `?`, `.f`, `(arg0, arg1, ...)` or `[expr]` until they've all been eaten.
	if let Ok(expr) = snapshot
	.parse_expr_dot_or_call_with(
	self.mk_expr_err(pat_span), // equivalent to transforming the parsed pattern into an `Expr`
	pat_span,
	AttrVec::new(),
	)

	if !recursion_limit.value_within_limit(self.cx.current_expansion.depth) {
	if self.cx.reduced_recursion_limit.is_none() {
	self.error_recursion_limit_reached();
	}

	// Reduce the recursion limit by half each time it triggers.
	self.cx.reduced_recursion_limit = Some(recursion_limit / 2);

	return ExpandResult::Ready(invoc.fragment_kind.dummy(invoc.span()));
	}

Add ErrorGuaranteed to ast::ExprKind::Err #120586

Add ErrorGuaranteed to ast::ExprKind::Err #120586

Conversation

ShE3py commented Feb 2, 2024

rustbot commented Feb 2, 2024

Choose a reason for hiding this comment

This comment has been minimized.

rustbot commented Feb 2, 2024

fmease left a comment • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ShE3py commented Feb 3, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

flip1995 Feb 15, 2024 • edited Loading

Choose a reason for hiding this comment

fmease commented Feb 8, 2024 • edited Loading

nnethercote commented Feb 14, 2024 • edited Loading

fmease left a comment

Choose a reason for hiding this comment

fmease Feb 8, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

fmease commented Feb 14, 2024

bors commented Feb 14, 2024

nnethercote commented Feb 14, 2024

bors commented Feb 26, 2024

fmease commented Feb 26, 2024

bors commented Feb 26, 2024

bors commented Feb 26, 2024

rust-log-analyzer commented Feb 26, 2024

fmease commented Feb 26, 2024

bors commented Feb 26, 2024

bors commented Feb 26, 2024

rust-timer commented Feb 26, 2024

Overall result: no relevant changes - no action needed

Add `ErrorGuaranteed` to `ast::ExprKind::Err` #120586

Add `ErrorGuaranteed` to `ast::ExprKind::Err` #120586

fmease left a comment •

edited

Loading

flip1995 Feb 15, 2024 •

edited

Loading

fmease commented Feb 8, 2024 •

edited

Loading

nnethercote commented Feb 14, 2024 •

edited

Loading

fmease Feb 8, 2024 •

edited

Loading