Fix accessibleType for package object prefixes #18057
Merged
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
Making a package object explicit re-computes the denotations of an overloaded method. So it should not be done after we have pruned down those denotations by an accessibility test. We now do it before checking accessibility. Fixes scala#15821
smarter
approved these changes
Jun 26, 2023
odersky
added
the
backport:nominated
|
It's backport nominated since it fixes a security hole. Previously, a protected member of a package object could be accessed from outside the package. This would not be caught at runtime either since protected in Scala is widened to public in bytecode. So all we have are compile time checks, and these were letting this case through. |
This was referenced Jun 28, 2023
Kordyjan
added a commit
that referenced
this pull request
Jun 28, 2023
Kordyjan
added
backport:accepted
odersky
added a commit
to dotty-staging/dotty
that referenced
this pull request
Jul 4, 2023
For private toplevel definitions we expand `private` to `private[p]` where `p` is the enclosing package. This PR applies the same scheme to protected toplevel definitions. Rationale: scala#18057 fixed an issue where toplevel protected members were always accessible because explicit package object prefixes were added after the accessibility check was done, and would re-establish the previous members without doing an accessibility check. The fix was done by adding package objects first, then doing he rest of the checks. But that also means that protected toplevel objects now get checked as members of their synthetic package object instead of as members of their package. To avoid that we make the package explicit as qalifier. Thsi shouls also make specs2 compile again.
odersky
added a commit
to dotty-staging/dotty
that referenced
this pull request
Jul 4, 2023
For private toplevel definitions we expand `private` to `private[p]` where `p` is the enclosing package. This PR applies the same scheme to protected toplevel definitions. Rationale: scala#18057 fixed an issue where toplevel protected members were always accessible because explicit package object prefixes were added after the accessibility check was done, and would re-establish the previous members without doing an accessibility check. The fix was done by adding package objects first, then doing he rest of the checks. But that also means that protected toplevel objects now get checked as members of their synthetic package object instead of as members of their package. To avoid that we make the package explicit as qalifier. Thsi shouls also make specs2 compile again.
odersky
added a commit
to dotty-staging/dotty
that referenced
this pull request
Jul 4, 2023
We usually have an access rule that the access to a protected member `foo` in class `C` must be from somewhere nested in a subclass of `C`. But that fails if the member is accessed from a package object `p.package`. In that case, the access does not need to be in the same object, it just has to be in package `p`. This clause was previously missing and is now added. Why was this only recently discovered? scala#18057 fixed an issue where toplevel protected members were always accessible because explicit package object prefixes were added after the accessibility check was done, and would re-establish the previous members without doing an accessibility check. The fix was done by adding package objects first, then doing he rest of the checks. But that also means that protected toplevel objects now get checked as members of their synthetic package object instead of as members of their package. The change here also makes specs2 compile again.
Kordyjan
pushed a commit
to dotty-staging/dotty
that referenced
this pull request
Jul 5, 2023
We usually have an access rule that the access to a protected member `foo` in class `C` must be from somewhere nested in a subclass of `C`. But that fails if the member is accessed from a package object `p.package`. In that case, the access does not need to be in the same object, it just has to be in package `p`. This clause was previously missing and is now added. Why was this only recently discovered? scala#18057 fixed an issue where toplevel protected members were always accessible because explicit package object prefixes were added after the accessibility check was done, and would re-establish the previous members without doing an accessibility check. The fix was done by adding package objects first, then doing he rest of the checks. But that also means that protected toplevel objects now get checked as members of their synthetic package object instead of as members of their package. The change here also makes specs2 compile again.
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.
Making a package object explicit re-computes the denotations of an overloaded method. So it should not be done after we have pruned down those denotations by an accessibility test. We now do it before checking accessibility.
Fixes #15821