Fix mark join decorrelation #20033
Merged
Fix mark join decorrelation #20033
Conversation
This file contains hidden or 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
Mark joins without correlation on the left side could not be handled correctly without this PR. Information regarding the correlated columns was simply lost, which could lead to wrong query behavior. This PR fixes this and introduces correct decorrelation of mark joins without correlation on the left side. ---- fix: duckdb#19504
8e752a0 to
55ae961
Compare
… side have correlation Kind of a weird case, maybe we should leave the join decorrelation early here. Will have to investigate some time.
lnkuiper
reviewed
Dec 4, 2025
Collaborator
lnkuiper
left a comment
There was a problem hiding this comment.
Thanks for the PR! Looks good to me. I just have one question, then this can be merged.
| } | ||
| } else if (join.join_type == JoinType::MARK) { | ||
| if (!left_has_correlation && right_has_correlation) { | ||
| // found a MARK join where the left side has no correlation |
Collaborator
There was a problem hiding this comment.
This covers the left/right {0, 1} correlation case. Are we sure all other cases are covered by the existing code?
Contributor
Author
There was a problem hiding this comment.
To be quite honest—I'm not 100% sure if we will need another special case, when the left and right sides are correlated, like is the case here:
SELECT query,
(WITH t AS MATERIALIZED (SELECT query)
SELECT *
FROM (SELECT query) AS _(x)
WHERE x IN (SELECT * FROM t)) AS broken
FROM (VALUES (('cat')), (('dog')), (('duck'))) AS queries(query);However, the query does return the correct result 🤔 So, we may be fine, I think?
Collaborator
There was a problem hiding this comment.
Thanks, that's good enough for me, for now. v1.4 is LTS, so if we run into another issue we can fix it, and we can reference this discussion.
2 tasks
github-actions bot
pushed a commit
to duckdb/duckdb-r
that referenced
this pull request
Dec 4, 2025
Fix mark join decorrelation (duckdb/duckdb#20033) Enable windows_arm64 arch for main extensions (duckdb/duckdb#20004)
github-actions bot
added a commit
to duckdb/duckdb-r
that referenced
this pull request
Dec 4, 2025
Fix mark join decorrelation (duckdb/duckdb#20033) Enable windows_arm64 arch for main extensions (duckdb/duckdb#20004) Co-authored-by: krlmlr <krlmlr@users.noreply.github.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
Mark joins without correlation on the left side could not be handled correctly without this PR. Information regarding the correlated columns was simply lost, which could lead to wrong query behavior. This PR fixes this and introduces correct decorrelation of mark joins without correlation on the left side.
Fix: #19504