Feb 2019

1 commit
Delay lock acquisition for partitions until we route a tuple to them.
21 Feb 2019Robert Haasmaster12beta1REL_19_STABLE12beta1REL_18_STABLE12beta1REL_17_STABLE12beta1REL_16_STABLE12beta1REL_15_STABLE12beta1REL_14_STABLE12beta1REL_13_STABLE12beta1+1 more
LLM analysis, 100% surelanded as postedfrom patchset #435 of 5 reviewers saw the final version

All six diff hunks in src/backend/executor/execPartition.c match #43 exactly, apart from index-hash abbreviation; the commit message was rewritten. Robert Haas confirms committing that version in #45.

  • Tomas Vondrasaw the final versionLLM analysis, 100% sureBenchmarked the original patch on two machines and examined descriptor safety in #2, then reconfirmed performance and investigated deadlock cases in #16 and #18; subsequent changes were only comments.
  • Sho Katosaw the final versionLLM analysis, 100% sureReported concurrency benchmarks in #3 and confirmed all tests passed after correcting plan_cache_mode in #10; only rebasing and comment changes followed.
  • John Naylorsaw the final versionLLM analysis, 100% sureBenchmarked the original patch in #5 and investigated reduced gains with the rebased version in #13 and #15; the final revision changed only comments.
  • Tom Lanesaw the final versionLLM analysis, 100% sureExamined existing locks in #31 and requested comment improvements while checking the completeness of RowExclusiveLock replacements in #41, which #43 addressed without changing executable code.
  • Robert Haassaw the final versioncommittercommitted the patch

Credits across 1 commit

Author
David RowleyLLM1
Reviewer
Robert Haascommitter1
Committer
Robert Haas1

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