Continue to borrowck even if there were previous errors #120550

oli-obk · 2024-02-01T13:33:04Z

but only from the perspective of the whole compiler. Individual items should not get borrowcked if their MIR is tainted by errors.

r? @estebank @nnethercote

rustbot · 2024-02-01T13:33:15Z

Some changes occurred to MIR optimizations

cc @rust-lang/wg-mir-opt

Some changes occurred in src/tools/clippy

cc @rust-lang/clippy

oli-obk · 2024-02-01T14:13:23Z

compiler/rustc_passes/src/liveness.rs

-                self.cont_ln
-                    .get(&sc)
-                    .cloned()
-                    .unwrap_or_else(|| span_bug!(expr.span, "continue to unknown label"))
+                self.cont_ln.get(&sc).cloned().unwrap_or_else(|| {
+                    self.ir.tcx.dcx().span_delayed_bug(expr.span, "continue to unknown label");
+                    self.ir.add_live_node(ErrNode)
+                })


We can't (yet) know that we should not be doing liveness checks, because the error happens in check_mod_loops for which we have no path to forward a taint to here ("here" being check_liveness)

fwiw this (at least used to?) fire here I believe: #113379

Oh neat. More evidence that those ICEs I've caused are preexisting, but hard to trigger

oli-obk · 2024-02-01T14:13:48Z

compiler/rustc_mir_build/src/build/scope.rs

+            let Some(drops) = self.scopes.breakable_scopes[break_index].continue_drops.as_mut()
+            else {
+                self.tcx.dcx().span_delayed_bug(
+                    source_info.span,
+                    "unlabelled `continue` within labelled block",
+                );
+                self.cfg.terminate(block, source_info, TerminatorKind::Unreachable);
+
+                return self.cfg.start_new_block().unit();
+            };
+            drops


also part of the continue without target issue

estebank · 2024-02-01T17:58:35Z

tests/ui/underscore-lifetime/underscore-lifetime-binders.stderr

+error: lifetime may not live long enough
+  --> $DIR/underscore-lifetime-binders.rs:16:43
+   |
+LL | fn foo2(_: &'_ u8, y: &'_ u8) -> &'_ u8 { y }
+   |                       -                   ^ returning this value requires that `'1` must outlive `'static`
+   |                       |
+   |                       let's call the lifetime of this reference `'1`


This error is a shame, and it is because we use ReStatic as a fallback. That was a reasonable solution before this change, but after we should likely introduce a ReError and silence lifetime errors that involve them.

estebank · 2024-02-01T18:09:46Z

tests/ui/rfcs/rfc-0000-never_patterns/check.rs

@@ -15,18 +15,18 @@ fn no_arms_or_guards(x: Void) {
        //~^ ERROR a never pattern is always unreachable
        None => {}
    }
-    match None::<Void> {
+    match None::<Void> { //~ ERROR: `Some(_)` not covered
        Some(!) if true,


How is this not a parse error!?

It is, but we fall back to generating all arms up to and including the parse error arm.

I tried replacing the entire match with an error expression, but then we lose the information that we don't need a semicolon after it. I guess I could encode that fact in the error variant. I'll try that out tomorrow

I was just surprised that we didn't have an //~ERROR annotation in the test.

compiler/rustc_borrowck/src/lib.rs

nnethercote · 2024-02-01T21:19:25Z

What is the motivation for this PR?

oli-obk · 2024-02-01T21:34:38Z

compiler/rustc_hir_analysis/src/lib.rs

-    if let Some(reported) = tcx.dcx().has_errors() { Err(reported) } else { Ok(()) }
+    Ok(())


this is the motivation of this PR. removing another has_errors that affects compilation progress

tests/ui/closures/2229_closure_analysis/diagnostics/borrowck/borrowck-4.rs

estebank · 2024-02-01T22:24:24Z

tests/ui/consts/control-flow/loop.rs

    for i in 0..4 { //~ ERROR `for` is not allowed in a `const`
+        //~^ ERROR: cannot call
+        //~| ERROR: mutable references
+        //~| ERROR: cannot convert


We should deduplicate these errors in constants, but it's fine for now.

estebank

r=me

oli-obk · 2024-02-02T07:46:52Z

@bors r=estebank

bors · 2024-02-02T07:46:54Z

📌 Commit 6ff8054 has been approved by estebank

It is now in the queue for this repository.

bors · 2024-02-02T07:46:55Z

🌲 The tree is currently closed for pull requests below priority 100. This pull request will be tested once the tree is reopened.

oli-obk · 2024-02-02T21:00:58Z

@bors r-

oli-obk · 2024-02-02T21:03:39Z

@bors r=estebank

bors · 2024-02-02T21:03:41Z

📌 Commit 0f3976b has been approved by estebank

It is now in the queue for this repository.

…=compiler-errors fix ICE for deref coercions with type errors Follow-up to rust-lang#120895, where I made types with errors go through the full coercion code, which is necessary if we want to build MIR for bodies with errors (rust-lang#120550). The code for coercing `&T` to `&U` currently assumes that autoderef for `&T` will succeed for at least two steps (`&T` and `T`): https://github.com/rust-lang/rust/blob/b17491c8f6d555386104dfd82004c01bfef09c95/compiler/rustc_hir_typeck/src/coercion.rs#L339-L464 But for types with errors, we previously only returned the no-op autoderef step (`&{type error}` -> `&{type error}`) and then stopped early. This PR changes autoderef for types with errors to still go through the built-in derefs (e.g. `&&{type error}` -> `&{type error}` -> `{type error}`) and only stop early when it would have to go looking for `Deref` trait impls. fixes rust-lang#120945 r? `@compiler-errors` or compiler

…=compiler-errors fix ICE for deref coercions with type errors Follow-up to rust-lang#120895, where I made types with errors go through the full coercion code, which is necessary if we want to build MIR for bodies with errors (rust-lang#120550). The code for coercing `&T` to `&U` currently assumes that autoderef for `&T` will succeed for at least two steps (`&T` and `T`): https://github.com/rust-lang/rust/blob/b17491c8f6d555386104dfd82004c01bfef09c95/compiler/rustc_hir_typeck/src/coercion.rs#L339-L464 But for types with errors, we previously only returned the no-op autoderef step (`&{type error}` -> `&{type error}`) and then stopped early. This PR changes autoderef for types with errors to still go through the built-in derefs (e.g. `&&{type error}` -> `&{type error}` -> `{type error}`) and only stop early when it would have to go looking for `Deref` trait impls. fixes rust-lang#120945 r? ``@compiler-errors`` or compiler

Rollup merge of rust-lang#120972 - lukas-code:autoderef-type-error, r=compiler-errors fix ICE for deref coercions with type errors Follow-up to rust-lang#120895, where I made types with errors go through the full coercion code, which is necessary if we want to build MIR for bodies with errors (rust-lang#120550). The code for coercing `&T` to `&U` currently assumes that autoderef for `&T` will succeed for at least two steps (`&T` and `T`): https://github.com/rust-lang/rust/blob/b17491c8f6d555386104dfd82004c01bfef09c95/compiler/rustc_hir_typeck/src/coercion.rs#L339-L464 But for types with errors, we previously only returned the no-op autoderef step (`&{type error}` -> `&{type error}`) and then stopped early. This PR changes autoderef for types with errors to still go through the built-in derefs (e.g. `&&{type error}` -> `&{type error}` -> `{type error}`) and only stop early when it would have to go looking for `Deref` trait impls. fixes rust-lang#120945 r? ``@compiler-errors`` or compiler

…-errors fix ICE for deref coercions with type errors Follow-up to rust-lang/rust#120895, where I made types with errors go through the full coercion code, which is necessary if we want to build MIR for bodies with errors (rust-lang/rust#120550). The code for coercing `&T` to `&U` currently assumes that autoderef for `&T` will succeed for at least two steps (`&T` and `T`): https://github.com/rust-lang/rust/blob/b17491c8f6d555386104dfd82004c01bfef09c95/compiler/rustc_hir_typeck/src/coercion.rs#L339-L464 But for types with errors, we previously only returned the no-op autoderef step (`&{type error}` -> `&{type error}`) and then stopped early. This PR changes autoderef for types with errors to still go through the built-in derefs (e.g. `&&{type error}` -> `&{type error}` -> `{type error}`) and only stop early when it would have to go looking for `Deref` trait impls. fixes rust-lang/rust#120945 r? ``@compiler-errors`` or compiler

rustbot assigned estebank Feb 1, 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 1, 2024

oli-obk commented Feb 1, 2024

View reviewed changes

estebank reviewed Feb 1, 2024

View reviewed changes

oli-obk 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 1, 2024

nnethercote reviewed Feb 1, 2024

View reviewed changes

compiler/rustc_borrowck/src/lib.rs Outdated Show resolved Hide resolved

oli-obk commented Feb 1, 2024

View reviewed changes

This comment has been minimized.

Sign in to view

estebank reviewed Feb 1, 2024

View reviewed changes

tests/ui/closures/2229_closure_analysis/diagnostics/borrowck/borrowck-4.rs Outdated Show resolved Hide resolved

estebank reviewed Feb 1, 2024

View reviewed changes

estebank approved these changes Feb 1, 2024

View reviewed changes

oli-obk force-pushed the track_errors8 branch from a32abc9 to 6ff8054 Compare February 1, 2024 22:45

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-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Feb 2, 2024

This comment has been minimized.

Sign in to view

bors 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-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Feb 2, 2024

oli-obk force-pushed the track_errors8 branch from 6ff8054 to 0f3976b Compare February 2, 2024 21:03

lukas-code mentioned this pull request Feb 12, 2024

fix ICE for deref coercions with type errors #120972

Merged

This was referenced Feb 22, 2024

ICE: underef: index out of bounds #121472

Closed

ICE: cannot convert ReEarlyParam(DefId(0:11 ~ crash[8cfd]::{impl#0}::y::{opaque#0}::'_), 3, '_) to a region vid #121574

Open

ICE: begin <= end (12 <= 11) when slicing 'Self::Assoc<'_>' #121807

Closed

matthiaskrgr mentioned this pull request Mar 8, 2024

ICE: unhandled type: {type error} #122191

Closed

matthiaskrgr mentioned this pull request Mar 26, 2024

ICE: thir: LocalTableInContext: key not found #123077

Open

GrigorenkoPV mentioned this pull request Mar 26, 2024

ICE: expected bits of u8, got Ty(10_u32,) #123092

Closed

matthiaskrgr mentioned this pull request Apr 9, 2024

ICE: Unsigned value 0x100 does not fit in 8 bits #123690

Open

eggyal mentioned this pull request Apr 10, 2024

ICE: assertion left == right failed: non-Aggregate field with matching ABI but differing alignment #123710

Closed

matthiaskrgr mentioned this pull request Apr 14, 2024

ICE: rustc_abi the len is 0 but the index is 0 calculating offset #123917

Open

dtolnay mentioned this pull request Apr 16, 2024

In Rust 1.78+, too many errors get printed when building with neither "std" nor "alloc" serde-rs/json#1125

Closed

kadiwa4 mentioned this pull request Apr 24, 2024

regression: failed to normalize alias(opaque ...) #123276

Open

matthiaskrgr mentioned this pull request Apr 24, 2024

ICE with anonymous_lifetime_in_impl_trait feature: "called Option::unwrap() on a None value" #124340

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Continue to borrowck even if there were previous errors #120550

Continue to borrowck even if there were previous errors #120550

oli-obk commented Feb 1, 2024

rustbot commented Feb 1, 2024

oli-obk Feb 1, 2024

matthiaskrgr Feb 7, 2024

oli-obk Feb 7, 2024

oli-obk Feb 1, 2024

estebank Feb 1, 2024

estebank Feb 1, 2024

oli-obk Feb 1, 2024

oli-obk Feb 1, 2024

estebank Feb 1, 2024

nnethercote commented Feb 1, 2024

oli-obk Feb 1, 2024

nnethercote Feb 2, 2024

This comment has been minimized.

estebank Feb 1, 2024

estebank left a comment

oli-obk commented Feb 2, 2024

bors commented Feb 2, 2024

bors commented Feb 2, 2024

This comment has been minimized.

oli-obk commented Feb 2, 2024

oli-obk commented Feb 2, 2024

bors commented Feb 2, 2024

		if let Some(reported) = tcx.dcx().has_errors() { Err(reported) } else { Ok(()) }
		Ok(())

Continue to borrowck even if there were previous errors #120550

Continue to borrowck even if there were previous errors #120550

Conversation

oli-obk commented Feb 1, 2024

rustbot commented Feb 1, 2024

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

nnethercote commented Feb 1, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

This comment has been minimized.

Choose a reason for hiding this comment

estebank left a comment

Choose a reason for hiding this comment

oli-obk commented Feb 2, 2024

bors commented Feb 2, 2024

bors commented Feb 2, 2024

This comment has been minimized.

oli-obk commented Feb 2, 2024

oli-obk commented Feb 2, 2024

bors commented Feb 2, 2024