Jan 2023

1 commit
Fix and clarify function comment on LogicalTapeSetCreate.
23 Jan 2023Heikki Linnakangasmaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1REL_15_STABLE15.2
LLM analysis, 100% surelanded as postedfrom patchset #251 of 1 reviewer saw the final version

The src/backend/utils/sort/logtape.c diff has the same stable patch identity as #25, including the brief explanation of preallocation. #28 confirms that the attached changes were pushed.

  • Peter Eisentrautsaw the final versiontrailerPeter reported the stale API documentation in #24, endorsed the replacement wording in #26, and requested committing it in #27.

Oct 2021

2 commits
Refactor LogicalTapeSet/LogicalTape interface.
18 Oct 2021Heikki Linnakangasmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 99% surepart of a posted patchsetlight changesfrom patchset #201 of 3 reviewers saw the final version

This commit takes only the LogicalTape API refactoring component of #20, with an identical stable patch identity. It covers nodeAgg.c, logtape.c, tuplesort.c and their headers; the balanced-merge component landed separately.

  • Peter Geoghegansaw an earlier versiontrailerPeter questioned the pause/resume dependency and proposed clearer tape API terminology in #5, before the substantial hash-aggregation and parallel-tape adaptations in #11.
  • Zhihong Yuno review foundtrailerZhihong's comments in #18 concern the merge algorithm's version comment and buffer-size type, with no demonstrated review of this separate API refactoring.
  • John Naylorsaw the final versiontrailerJohn explicitly called the API patch a useful simplification and tested both final patches together in #21.
Replace polyphase merge algorithm with a simple balanced k-way merge.
18 Oct 2021Heikki Linnakangasmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 99% surepart of a posted patchsetlight changesfrom patchset #203 of 3 reviewers saw the final version

This commit takes only the balanced-merge component of the two-patch series in #20. Its changed lines in src/backend/utils/sort/tuplesort.c differ from that component only by correcting the historical version comment from PostgreSQL 14 to 15, as confirmed in #23.

  • Peter Geoghegansaw the final versiontrailerPeter supported and analyzed the original design in #4 and #5, then justified retaining signed memory accounting for the final version in #22.
  • Zhihong Yusaw the final versiontrailerZhihong identified the incorrect release number and questioned the buffer-size type in #18, after which only rebasing and the comment correction affected this component.
  • John Naylorsaw the final versiontrailerJohn reviewed both patches, reported performance tests, supported retaining int64, and marked the series ready subject to the release-number correction in #21.

Credits across 3 commits

Author
Heikki LinnakangasLLM3
Reviewer
John Naylor2
Peter Geoghegan2
Zhihong Yu2
Peter Eisentraut1
Committer
Heikki Linnakangas3

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