rustc: don't visit lifetime parameters through visit_lifetime. #51215
Conversation
rust-highfive
added
the
S-waiting-on-review
| @@ -733,7 +733,16 @@ pub fn walk_ty_param_bound<'v, V: Visitor<'v>>(visitor: &mut V, bound: &'v TyPar | |||
| pub fn walk_generic_param<'v, V: Visitor<'v>>(visitor: &mut V, param: &'v GenericParam) { | |||
| match *param { | |||
| GenericParam::Lifetime(ref ld) => { | |||
| visitor.visit_lifetime(&ld.lifetime); | |||
| visitor.visit_id(ld.lifetime.id); | |||
There was a problem hiding this comment.
Wait, is there no visit_lifetime_def? Can we make one?
There was a problem hiding this comment.
I guess we don't have visit_type_def. Seems a shame, but ok.
There was a problem hiding this comment.
It's visit_generic_param - handling just one kind of parameter is generally a bug.
|
|
||
| for param in &generics.params { | ||
| match *param { | ||
| hir::GenericParam::Lifetime(ref lifetime_def) => { | ||
| if !lifetime_def.bounds.is_empty() { | ||
| // `'a: 'b` means both `'a` and `'b` are referenced | ||
| appears_in_where_clause.visit_generic_param(param); | ||
| appears_in_where_clause.regions.insert(lifetime_def.lifetime.name); |
There was a problem hiding this comment.
if we had a visit_lifetime_def, we could handle this within the visit impl, which seems like it'd be nicer
There was a problem hiding this comment.
That'd be wrong, because it'd also apply on nested for's but we only want the outermost params. I almost moved this check to the loop below.
|
@bors r=nikomatsakis |
|
📌 Commit df45c4f has been approved by |
bors
added
S-waiting-on-bors
|
🔒 Merge conflict |
bors
added
S-waiting-on-author
eddyb
force-pushed
the
visit-for-a-lifetime
branch
from
df45c4f to
5c76b64
Compare
|
@bors r=nikomatsakis |
|
📌 Commit 5c76b64 has been approved by |
bors
added
S-waiting-on-bors
rustc: don't visit lifetime parameters through visit_lifetime. Ideally we'd also not use the `Lifetime` struct for parameters, but I'll leave that to @varkor (#48149). Fixes #51185 (discovered while auditing all the `visit_lifetime` implementations). r? @nikomatsakis
|
☀️ Test successful - status-appveyor, status-travis |
Ideally we'd also not use the
Lifetimestruct for parameters, but I'll leave that to @varkor (#48149).Fixes #51185 (discovered while auditing all the
visit_lifetimeimplementations).r? @nikomatsakis