"fix" a regression in the 5.2 merge #3004
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
mshinwell
pushed a commit
that referenced
this pull request
Sep 20, 2024
* add failing test * "fix" by restoring old behavior (cherry picked from commit 8db5026)
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Outdated suggestions cannot be applied.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 adds a test that newly fails on the 5.2 branch, and a fix, in two commits.
The test passes upstream and seems fine, so I think this is a regression in the merge. But I don't think the fix here (restoring a piece of the previous implementation of checking
Pexp_constraints in typecore) is correct. I'm just posting this so @mshinwell can use it to make more progress and so I can point to it for discussion.What is happening is that
loosen_arrow_modesintype_argumentis updating the scope of the type, and the fact that the previous implementation was throwing away the old type meant we ignored that. That seems wrong, instead, I thinktype_argumentshouldn't need to use the equation for mode checking in this example, and so this program typechecks in 5.1 for the wrong reason.Will discuss with modes people.