sql/catalog/tabledesc: fixed a bug in maybeAddConstraintIDs
#85778
Conversation
|
This change is |
Previously, when we RunPostDeserializationChanges, one of the step is to fill in constraint IDs if they are not already set, which is done in function `maybeAddConstraintIDs`. That function however is buggy in that, if there is constraint mutation on the table descriptor, we directly go assign a constraint ID (from `tbl.nextConstraintID`) to it, without checking whether it already has a non-zero constraint ID or not. This PR fixes that. Release note (bug fix): Fixed a bug in post deserialization changes where we might incorrectly change constraint ID of a constraint that lives in the mutation slice of a table descriptor.
Xiang-Gu
force-pushed
the
fix_maybe_add_constraint_IDs_during_post_deserialization_change
branch
from
00f0991 to
be2ddb2
Compare
There was a problem hiding this comment.
Reviewed 2 of 7 files at r1.
Reviewable status:complete! 1 of 0 LGTMs obtained (waiting on @ajwerner and @Xiang-Gu)
-- commits line 12 at r1:
I don't think this needs a release note because we never released this bug.
There was a problem hiding this comment.
Reviewable status:
Previously, ajwerner wrote…
I don't think this needs a release note because we never released this bug.
removed.
|
TFTR! bors r+ |
|
Build failed (retrying...): |
|
Build failed (retrying...): |
|
Build failed: |
|
The merge failed due to a flaky test, retrying bors r+ |
|
Build failed (retrying...): |
|
Build failed: |
|
It's sad to see merge failure again but it seems unrelated to this change, retrying again, |
|
Build succeeded: |
Xiang-Gu
deleted the
fix_maybe_add_constraint_IDs_during_post_deserialization_change
branch
Previously, when we RunPostDeserializationChanges, one of the step is
to fill in constraint IDs if they are not already set, which is done
in function
maybeAddConstraintIDs. That function however is buggyin that, if there is constraint mutation on the table descriptor, we
directly go assign a constraint ID (from
tbl.nextConstraintID) to it,without checking whether it already has a non-zero constraint ID or not.
This PR fixes that.
Release note: None