May 2023

1 commit
Fix pgbench in prepared mode with an empty pipeline
25 May 2023Alvaro Herreramaster16beta2REL_19_STABLE16beta2REL_18_STABLE16beta2REL_17_STABLE16beta2REL_16_STABLE16beta2REL_15_STABLE15.4
LLM analysis, 100% suremodified before commitlight changesfrom patchset #250 of 1 reviewer saw the final version

Matches the allocation fix and TAP test in #25, with allocatePreparedFlags renamed to allocCStatePrepared and its introductory comment revised in src/bin/pgbench/pgbench.c.

  • Alexander Lakhinsaw an earlier versionLLM analysis, 91% sureReported a reproducible empty-pipeline crash and backtrace in #24 against the landed implementation from #22, prompting the allocation changes in #25.

Feb 2023

1 commit
pgbench: Prepare commands in pipelines in advance
21 Feb 2023Alvaro Herreramaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1REL_15_STABLE15.3REL_14_STABLE14.8
LLM analysis, 100% suremodified before commitmoderate changesfrom patchset #220 of 5 reviewers saw the final version

The development-branch commit retains the implementation and TAP test from #22. In src/bin/pgbench/pgbench.c, postprocess_sql_command() is adjusted so only QUERY_PREPARED generates a statement name before falling through to shared query parsing.

  • Fabien COELHOsaw an earlier versionLLM analysis, 100% sureReported successful compilation and tests in #2 and requested regression coverage and comment changes for #1, before the substantial rewrite in #19.
  • Daniel Gustafssonsaw an earlier versionLLM analysis, 99% sureReported the TAP failure of #4 in #5, leading to the test correction in #6 before the later redesign.
  • Kyotaro Horiguchisaw an earlier versionLLM analysis, 99% sureRequested a fuller explanation of preparation timing in #7 while reviewing #6, which was incorporated in #8 before the redesign.
  • Tom Lanesaw an earlier versionLLM analysis, 98% sureQuestioned whole-script preparation in #9 and proposed preparing only pipeline contents before pipeline execution, an approach developed in #19.
  • Andres Freundsaw an earlier versionLLM analysis, 99% sureReported reproducible core dumps from #19 in #20, followed by allocation, threading, and option-order fixes in #21 and #22.

Credits across 2 commits

Author
Alvaro HerreraLLM2
Reported by
Alexander Lakhin1
Yugo Nagata1
Committer
Alvaro Herrera2

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