Skip to content

Fix dyn -> param suggestion in struct ICEs - #138238

Merged
bors merged 2 commits into
rust-lang:masterfrom
compiler-errors:dyn-suggestion-in-struct
Mar 11, 2025
Merged

Fix dyn -> param suggestion in struct ICEs#138238
bors merged 2 commits into
rust-lang:masterfrom
compiler-errors:dyn-suggestion-in-struct

Conversation

@compiler-errors

Copy link
Copy Markdown
Contributor

Makes the logic from #138042 a bit less ICEy and more clean. Also fixes an incorrect suggestion when the struct already has generics. I'll point out the major changes and observations in the code.

Fixes#138229
Fixes#138211

r? nnethercote since you reviewed the original pr, or re-roll if you don't want to review this

@rustbotrustbot 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 8, 2025
@rustbot

Copy link
Copy Markdown
Collaborator

HIR ty lowering was modified

cc @fmease

let mut sugg = vec![];

let parent_id = tcx.hir_get_parent_item(self_ty.hir_id).def_id;
let parent_item = tcx.hir_node_by_def_id(parent_id).expect_item();

Copy link
Copy Markdown
ContributorAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

expect_item here is not valid, since there are other kinds of HIR parents that can own this dyn Trait, like extern items (or trait/impl items, closures, etc).

let parent_hir_id = tcx.parent_hir_id(self_ty.hir_id);
let parent_item = tcx.hir_get_parent_item(self_ty.hir_id).def_id;

let generics = match tcx.hir_node_by_def_id(parent_item) {

Copy link
Copy Markdown
ContributorAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I added a check that we're suggesting this on a field directly. dyn may show up in other positions like in a where clause or in an expr on a default field value.

if let hir::ItemKind::Enum(..) = parent_item.kind {
return false;
}
// Look at the direct HIR parent, since we care about the relationship between

Copy link
Copy Markdown
ContributorAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This was also expect_iteming unnecessarily. I took the logic and made it focus on the relationship between the ty and its parent node, so we don't suggest turning:

struct Foo { f: Box<Any>, /* more fields */ }

Into this:

struct Foo<T: Any> { f: Box<T>, /* more fields */ }

My general logic here is that if the ty's hir node parent is another ty, then it's probably just a misspelled dyn Trait. There is no perfect heuristic here, but I think it's less invasive to suggest the most naive thing here, which is just to insert dyn, unless we're very confident it will be wrong.

@@ -13,7 +13,6 @@ struct Foo2 {
//~^ ERROR expected a type, found a trait

Copy link
Copy Markdown
ContributorAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I moved this test out of suggestions/ since that's a kitchen sink


// If the parent item is a struct, check if self_ty is the last field.
if let hir::ItemKind::Struct(variant_data, _) = parent_item.kind {
if variant_data.fields().last().unwrap().ty.span != self_ty.span {

Copy link
Copy Markdown
ContributorAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We don't need to use span comparisons here, hir_id comparisons are more faithful.

Comment on lines +168 to +173
let param = "TUV"
.chars()
.map(|c| c.to_string())
.chain((0..).map(|i| format!("P{i}")))
.find(|s| !generics.params.iter().any(|param| param.name.ident().as_str() == s))
.expect("we definitely can find at least one param name to generate");

Copy link
Copy Markdown
ContributorAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This logic didn't account for preexisting generic args, so this tries its best to suggest T, U, V, P1, P2, ...

@nnethercote

Copy link
Copy Markdown
Contributor

@bors r+

Thank for fixing the problems.

@bors

bors commented Mar 9, 2025

Copy link
Copy Markdown
Collaborator

📌 Commit ceb0401 has been approved by nnethercote

It is now in the queue for this repository.

@borsbors 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 9, 2025
@nnethercote

Copy link
Copy Markdown
Contributor

cc @xizheyin

@compiler-errors

Copy link
Copy Markdown
ContributorAuthor

error path only so

@bors rollup

matthiaskrgr added a commit to matthiaskrgr/rust that referenced this pull request Mar 9, 2025
…struct, r=nnethercote
Fix dyn -> param suggestion in struct ICEs
Makes the logic from rust-lang#138042 a bit less ICEy and more clean. Also fixes an incorrect suggestion when the struct already has generics. I'll point out the major changes and observations in the code.
Fixesrust-lang#138229Fixesrust-lang#138211
r? nnethercote since you reviewed the original pr, or re-roll if you don't want to review this
bors added a commit to rust-lang-ci/rust that referenced this pull request Mar 10, 2025
…iaskrgr
Rollup of 6 pull requests
Successful merges:
- rust-lang#137279 (Make some invalid codegen attr errors structured/translatable)
- rust-lang#137793 (Stablize anonymous pipe)
- rust-lang#138238 (Fix dyn -> param suggestion in struct ICEs)
- rust-lang#138270 (chore: Fix some comments)
- rust-lang#138280 (fix ICE in pretty-printing `global_asm!`)
- rust-lang#138286 (triagebot.toml: Don't label `test/rustdoc-json` as A-rustdoc-search (…)
r? `@ghost`
`@rustbot` modify labels: rollup
matthiaskrgr added a commit to matthiaskrgr/rust that referenced this pull request Mar 10, 2025
…struct, r=nnethercote
Fix dyn -> param suggestion in struct ICEs
Makes the logic from rust-lang#138042 a bit less ICEy and more clean. Also fixes an incorrect suggestion when the struct already has generics. I'll point out the major changes and observations in the code.
Fixesrust-lang#138229Fixesrust-lang#138211
r? nnethercote since you reviewed the original pr, or re-roll if you don't want to review this
bors added a commit to rust-lang-ci/rust that referenced this pull request Mar 10, 2025
…iaskrgr
Rollup of 9 pull requests
Successful merges:
- rust-lang#136395 (Update to rand 0.9.0)
- rust-lang#137279 (Make some invalid codegen attr errors structured/translatable)
- rust-lang#137585 (Update documentation to consistently use 'm' in atomic synchronization example)
- rust-lang#137926 (Add a test for `-znostart-stop-gc` usage with LLD)
- rust-lang#138074 (Support `File::seek` for Hermit)
- rust-lang#138238 (Fix dyn -> param suggestion in struct ICEs)
- rust-lang#138270 (chore: Fix some comments)
- rust-lang#138280 (fix ICE in pretty-printing `global_asm!`)
- rust-lang#138286 (triagebot.toml: Don't label `test/rustdoc-json` as A-rustdoc-search (…)
r? `@ghost`
`@rustbot` modify labels: rollup
bors added a commit to rust-lang-ci/rust that referenced this pull request Mar 10, 2025
…iaskrgr
Rollup of 8 pull requests
Successful merges:
- rust-lang#136395 (Update to rand 0.9.0)
- rust-lang#137279 (Make some invalid codegen attr errors structured/translatable)
- rust-lang#137585 (Update documentation to consistently use 'm' in atomic synchronization example)
- rust-lang#137926 (Add a test for `-znostart-stop-gc` usage with LLD)
- rust-lang#138074 (Support `File::seek` for Hermit)
- rust-lang#138238 (Fix dyn -> param suggestion in struct ICEs)
- rust-lang#138270 (chore: Fix some comments)
- rust-lang#138286 (triagebot.toml: Don't label `test/rustdoc-json` as A-rustdoc-search (…)
r? `@ghost`
`@rustbot` modify labels: rollup
bors added a commit to rust-lang-ci/rust that referenced this pull request Mar 10, 2025
…iaskrgr
Rollup of 8 pull requests
Successful merges:
- rust-lang#136395 (Update to rand 0.9.0)
- rust-lang#137279 (Make some invalid codegen attr errors structured/translatable)
- rust-lang#137585 (Update documentation to consistently use 'm' in atomic synchronization example)
- rust-lang#137926 (Add a test for `-znostart-stop-gc` usage with LLD)
- rust-lang#138074 (Support `File::seek` for Hermit)
- rust-lang#138238 (Fix dyn -> param suggestion in struct ICEs)
- rust-lang#138270 (chore: Fix some comments)
- rust-lang#138286 (triagebot.toml: Don't label `test/rustdoc-json` as A-rustdoc-search (…)
r? `@ghost`
`@rustbot` modify labels: rollup
bors added a commit to rust-lang-ci/rust that referenced this pull request Mar 10, 2025
…iaskrgr
Rollup of 8 pull requests
Successful merges:
- rust-lang#136395 (Update to rand 0.9.0)
- rust-lang#137279 (Make some invalid codegen attr errors structured/translatable)
- rust-lang#137585 (Update documentation to consistently use 'm' in atomic synchronization example)
- rust-lang#137926 (Add a test for `-znostart-stop-gc` usage with LLD)
- rust-lang#138074 (Support `File::seek` for Hermit)
- rust-lang#138238 (Fix dyn -> param suggestion in struct ICEs)
- rust-lang#138270 (chore: Fix some comments)
- rust-lang#138286 (triagebot.toml: Don't label `test/rustdoc-json` as A-rustdoc-search (…)
r? `@ghost`
`@rustbot` modify labels: rollup
jieyouxu added a commit to jieyouxu/rust that referenced this pull request Mar 10, 2025
…struct, r=nnethercote
Fix dyn -> param suggestion in struct ICEs
Makes the logic from rust-lang#138042 a bit less ICEy and more clean. Also fixes an incorrect suggestion when the struct already has generics. I'll point out the major changes and observations in the code.
Fixesrust-lang#138229Fixesrust-lang#138211
r? nnethercote since you reviewed the original pr, or re-roll if you don't want to review this
@jieyouxujieyouxu mentioned this pull request Mar 10, 2025
bors added a commit to rust-lang-ci/rust that referenced this pull request Mar 11, 2025
…iaskrgr
Rollup of 8 pull requests
Successful merges:
- rust-lang#136395 (Update to rand 0.9.0)
- rust-lang#137279 (Make some invalid codegen attr errors structured/translatable)
- rust-lang#137585 (Update documentation to consistently use 'm' in atomic synchronization example)
- rust-lang#137926 (Add a test for `-znostart-stop-gc` usage with LLD)
- rust-lang#138074 (Support `File::seek` for Hermit)
- rust-lang#138238 (Fix dyn -> param suggestion in struct ICEs)
- rust-lang#138270 (chore: Fix some comments)
- rust-lang#138286 (triagebot.toml: Don't label `test/rustdoc-json` as A-rustdoc-search (…)
r? `@ghost`
`@rustbot` modify labels: rollup
@bors
bors merged commit 1ae083d into rust-lang:masterMar 11, 2025
@rustbotrustbot added this to the 1.87.0 milestone Mar 11, 2025
rust-timer added a commit to rust-lang-ci/rust that referenced this pull request Mar 11, 2025
Rollup merge of rust-lang#138238 - compiler-errors:dyn-suggestion-in-struct, r=nnethercote
Fix dyn -> param suggestion in struct ICEs
Makes the logic from rust-lang#138042 a bit less ICEy and more clean. Also fixes an incorrect suggestion when the struct already has generics. I'll point out the major changes and observations in the code.
Fixesrust-lang#138229Fixesrust-lang#138211
r? nnethercote since you reviewed the original pr, or re-roll if you don't want to review this
Sign up for freeto join this conversation on GitHub. Already have an account? Sign in to comment

Labels

S-waiting-on-borsStatus: Waiting on bors to run and complete tests. Bors will change the label on completion.T-compilerRelevant to the compiler team, which will review and decide on the PR/issue.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

ICE: compiler/rustc_hir_analysis/src/hir_ty_lowering/lint.rs None ICE: $ident: found TraitItem(TraitItem

4 participants

@compiler-errors@rustbot@nnethercote@bors