Dec 2024

4 commits
Fix further fallout from EXPLAIN ANALYZE BUFFERS change
11 Dec 2024David Rowleymaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 99% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

Adds BUFFERS OFF to contrib/pg_stat_statements/sql/level_tracking.sql and updates expected output, matching the outstanding failure described in #44 and the completion report in #50. No attached patchset contains these changes; #47 changes only rules.sgml.

  • Jelte Fennema-Niosaw the final versionLLM analysis, 94% sureBefore the fix was committed, Jelte reported the specific pg_stat_statements regression failure and supplied its failing-test output in #48.
Add missing BUFFERS OFF in regression tests, take 2
11 Dec 2024David Rowleymaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surewritten from the discussion, no matching patch

Adds missing BUFFERS OFF settings in matview.sql and misc_functions.sql, with matching expected-output changes, after testing with RELCACHE_FORCE_RELEASE and CATCACHE_FORCE_RELEASE. These unattached follow-up fixes are described retrospectively in #44.

Add missing BUFFERS OFF in select_into regression tests
11 Dec 2024David Rowleymaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surewritten from the discussion, no matching patch

Fixes two select_into.sql statements omitted from #41, together with their expected output, after buildfarm failures exposed planning-time buffer accesses. No attachment contains these additional hunks; David reports the follow-up fixes in #44.

Enable BUFFERS with EXPLAIN ANALYZE by default
11 Dec 2024David Rowleymaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surepart of a posted patchsetmoderate changesfrom patchset #413 of 3 reviewers saw the final version

The diff matches the complete first patch of #41 after normalization, but excludes the second patch changing auto_explain.log_buffers. David explicitly confirms that split outcome in #43, and the commit credits Guillaume and David as co-authors.

  • Robert Haassaw the final versiontrailerRobert assessed David's approach in #17 and reviewed Guillaume's implementation in #20; subsequent changes to the committed core feature were documentation and test adjustments.
  • Greg Sabino Mullanesaw the final versiontrailerGreg reviewed the patch and documentation in #22, endorsed revised wording in #24, and supplied richer example counts in #26, followed only by light changes to the committed feature.
  • Michael Christofidessaw the final versiontrailerMichael checked documentation coverage and reported positively in #39, before David's remaining documentation and regression-test polishing.

Credits across 4 commits

Author
David Rowley1
Guillaume Lelarge1
Reviewer
Greg Sabino Mullane1
Michael Christofides1
Robert Haas1
Committer
David Rowley4