Fix opaque types leaking rhs when inlined and found in type params (and a related stale symbol issue) #22655
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
jchyb
force-pushed
the
fix-i20449-stale-opaque
branch
from
a1755a6 to
4d937b2
Compare
dwijnand
requested changes
Feb 28, 2025
| @@ -399,7 +399,9 @@ class Inliner(val call: tpd.Tree)(using Context): | |||
| * type aliases, add proxy definitions to `opaqueProxies` that expose these aliases. | |||
| */ | |||
| private def addOpaqueProxies(tp: Type, span: Span, forThisProxy: Boolean)(using Context): Unit = | |||
| tp.foreachPart { | |||
| val foreachTpPart = | |||
| (p: Type => Unit) => if (forThisProxy) tp.foreachPartWithoutTypeParams(p) else tp.foreachPart(p) | |||
There was a problem hiding this comment.
I wouldn't introduce foreachPartWithoutTypeParams for one use case. Also, it's a confusing name, because you're skipping applied type arguments, rather than "type parameters".
| @@ -712,7 +712,7 @@ class Namer { typer: Typer => | |||
| enterSymbol(classConstructorCompanion(classSym.asClass)) | |||
| else | |||
| for moduleSym <- companionVals do | |||
| if moduleSym.is(Module) && !moduleSym.isDefinedInCurrentRun then | |||
| if moduleSym.lastKnownDenotation.is(Module) && !moduleSym.isDefinedInCurrentRun then | |||
There was a problem hiding this comment.
Can you add a little detail in a comment about this usage (with any references that would be helpful)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR fixes the 2 issues found in #20449, split into 2 commits.
The first commit fixes the stale symbol related issue found if the files from the issue minimization are compiled together. After suspending and retrying compilation, the classDefs that are defined directly in packages previously would sometimes not have companion objects regenerated, instead relying on the stale symbols from the previous run, causing them not to to pass the reallyExists check when looking for a specific ref. Now we make sure to go through lastKnwonDenotation, since the current one may not exists and may not point us to a Module flag when checking if to regenerate it.
The second commit fixes the opaque type alias rhs leaking in a macro. That was caused by building proxies for all parts of the type, including type arguments to opaque types - from the perspective of a type like Object[OpaqueType], the opaque type rhs should not be visible.