Jun 2023

1 commit
Add a test case for a316a3bc
30 Jun 2023Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1REL_16_STABLE16beta3
LLM analysis, 100% surelanded as postedfrom patchset #96

Both postgres_fdw test-file diffs match #96, with an expanded commit explanation. Git confirms this is the master commit, and #97 announces the landing on HEAD and PostgreSQL 16.

Feb 2023

2 commits
Correctly set userid of subquery relations' child rels
20 Feb 2023Alvaro Herreramaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 100% suremodified before commitlight changesfrom patchset #921 of 2 reviewers saw the final version

The build_simple_rel() condition and assertion match #92; the only source-diff alteration rewraps a comment. The test remains omitted, as confirmed in #93.

  • Justin Pryzbysaw an earlier versionLLM analysis, 94% sureJustin reviewed the initial userid-fix patch and supplied wording corrections in #78, before the condition was rewritten around reloptkind.
  • Alvaro Herrerasaw the final versioncommittercommitted the patch
Fix buggy recursion in flatten_rtes_walker().
13 Feb 2023Tom Lanemaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1Also in: ERROR: permission info at index 1 ....
LLM analysis, 99% surewritten from the discussion, no matching patch2 of 2 reviewers saw the final version

No attachment contains this save-and-restore fix for flatten_rtes_walker() or its join.sql test; #86 reports the failures and #88 announces the externally diagnosed fix. The commit credits Tender Wang, who has no message in this thread.

  • Justin Pryzbysaw the final versionLLM analysis, 97% sureJustin supplied two concrete sqlsmith reproductions of the permission-list context failure in #86, prompting this fix without an attached patchset.
  • Tom Lanesaw the final versioncommittercommitted the patch

Jan 2023

1 commit
Remove some dead code in selfuncs.c
19 Jan 2023Alvaro Herreramaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #841 of 2 reviewers saw the final version

Exactly matches the selfuncs.c and relnode.c cleanup component of #84, which incorporates #83; the test and userid bugfix components were left pending.

  • Justin Pryzbysaw an earlier versionLLM analysis, 99% sureJustin identified redundant userid recomputation in #76 and reviewed the first cleanup in #78, before Alvaro's broader hoisting refactor.
  • Alvaro Herrerasaw the final versioncommittercommitted the patch

Nov - Dec 2022

5 commits
Update outdated comment in ApplyRetrieveRule
7 Dec 2022Alvaro Herreramaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 100% surelanded as postedfrom patchset #721 of 1 reviewer saw the final version

The single ApplyRetrieveRule() comment hunk in src/backend/rewrite/rewriteHandler.c exactly matches #72; #75 confirms it was pushed.

  • Alvaro Herrerasaw the final versioncommittercommitted the patch
Rework query relation permission checking
6 Dec 2022Alvaro Herreramaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 100% suremodified before commitmoderate changesfrom patchset #691 of 5 reviewers saw the final version

Implements #69 with the original ExecutorCheckPerms_hook name restored, the executor permission-info wrapper removed, and CombineRangeTables() changed to update two destination pointers. These edits are explicitly described in #70 and #73, and the diff also includes a catalog-version bump.

  • Tom Lanesaw an earlier versionLLM analysis, 100% sureTom proposed separating permission data in #5 and rejected OID-based merging in #32, leading to substantial redesign before landing.
  • David Rowleysaw an earlier versionLLM analysis, 100% sureDavid examined the initial bitmap patch in #2 and supplied a detailed permission-list, userid, and serialization review in #23.
  • Zhihong Yusaw an earlier versionLLM analysis, 99% sureZhihong reviewed control flow, comments, naming, and mapping validity in #18 and #22, before the permission-list redesign.
  • Andres Freundsaw an earlier versionLLM analysis, 94% sureAndres reported parse-tree round-trip failures and incorrect generated-column results in #35, which prompted the serialization repair in #36.
  • Alvaro Herrerasaw the final versioncommittercommitted the patch
Generalize ri_RootToPartitionMap to use for non-partition children
2 Dec 2022Alvaro Herreramaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #661 of 2 reviewers saw the final version

Extracts the mapping prerequisite from #66, with ri_RootToPartitionMap renamed to ri_RootToChildMap and fields relocated within ResultRelInfo. The comments and formatting edits agree with the landing explanation in #67.

  • Zhihong Yusaw an earlier versionLLM analysis, 88% sureZhihong questioned the root-to-child map's validity-flag handling in #22, before it was redesigned to reuse the existing TupleConversionMap.
  • Alvaro Herrerasaw the final versioncommittercommitted the patch
Stop accessing checkAsUser via RTE in some cases
30 Nov 2022Alvaro Herreramaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #593 of 3 reviewers saw the final version

Extracts the checkAsUser prerequisite from #59, preserving its ForeignScan and RelOptInfo userid changes while replacing OID-as-boolean tests with OidIsValid() and updating comments. Alvaro announces that editing intention in #63.

  • Tom Lanesaw the final versionLLM analysis, 88% sureTom challenged the added ForeignScan.checkAsUser field in #32, and Amit justified the retained design in #34, with only light subsequent changes to this extracted component.
  • David Rowleysaw the final versionLLM analysis, 84% sureDavid reviewed RelOptInfo.userid's broadened use and placement in #23, and that component reached this prerequisite with only light subsequent edits.
  • Alvaro Herrerasaw the final versioncommittercommitted the patch
Add 'missing_ok' argument to build_attrmap_by_name
29 Nov 2022Alvaro Herreramaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #571 of 1 reviewer saw the final version

Extracts the missing_ok argument and caller adaptations from the main patch in #57, including attmap.c and attmap.h. The extracted implementation matches those hunks, and #58 explicitly announces its separate landing.

  • Alvaro Herrerasaw the final versioncommittercommitted the patch

Credits across 9 commits

Author
Amit Langote1 LLM8
Tender WangLLM1
Reviewer
Alvaro Herreracommitter7
Tom Lanecommitter1
Reported by
Justin Pryzby1
Committer
Alvaro Herrera7
Amit Langote1
Tom Lane1

N LLM means N of the credits come from LLM analysis, not from commit trailers.