Apr 2020

1 commit
Fix AddressSanitizer use-after-scope complaint.
30 Apr 2020Peter Geogheganmaster13beta1REL_19_STABLE13beta1REL_18_STABLE13beta1REL_17_STABLE13beta1REL_16_STABLE13beta1REL_15_STABLE13beta1REL_14_STABLE13beta1REL_13_STABLE13beta1
LLM analysis, 100% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

The two hunks in src/backend/access/nbtree/nbtinsert.c exactly match Peter Eisentraut's inline fix #20, moving upostingoff into the enclosing scope so its lifetime covers WAL registration. Peter Geoghegan confirmed committing that fix #22.

  • Peter Geoghegansaw the final versioncommittercommitted the patch

Mar 2020

3 commits
Silence nbtree.h cpluspluscheck warning.
2 Mar 2020Peter Geogheganmaster13beta1REL_19_STABLE13beta1REL_18_STABLE13beta1REL_17_STABLE13beta1REL_16_STABLE13beta1REL_15_STABLE13beta1REL_14_STABLE13beta1REL_13_STABLE13beta1
LLM analysis, 100% surewritten from the discussion, no matching patch

The commit adds the size_t cast in BTreeTupleSetPosting() proposed after Tom Lane reported the cpluspluscheck warning #15. No attachment contains this change, and Peter Geoghegan confirmed pushing it #16.

Add assertions to _bt_update_posting().
2 Mar 2020Peter Geogheganmaster13beta1REL_19_STABLE13beta1REL_18_STABLE13beta1REL_17_STABLE13beta1REL_16_STABLE13beta1REL_15_STABLE13beta1REL_14_STABLE13beta1REL_13_STABLE13beta1
LLM analysis, 100% surelanded as postedfrom patchset #120 of 1 reviewer saw the final version

Both hunks exactly match #12: assertions check the tuple size limit, MAXALIGN alignment, and the singleton heap TID in _bt_update_posting(). Peter Geoghegan confirmed the push #14.

  • Tom Lanesaw an earlier versionLLM analysis, 94% sureTom identified duplication and questioned size alignment in #10, prompting replacement of the restructuring patch #9 with the assertions in #12; his subsequent bare agreement #13 adds no substantive final-version review.
Remove dead code from _bt_update_posting().
1 Mar 2020Peter Geogheganmaster13beta1REL_19_STABLE13beta1REL_18_STABLE13beta1REL_17_STABLE13beta1REL_16_STABLE13beta1REL_15_STABLE13beta1REL_14_STABLE13beta1REL_13_STABLE13beta1
LLM analysis, 100% surewritten from the discussion, no matching patch

Peter Geoghegan identified the unreachable non-posting original-tuple case in #8; the commit removes that branch and obtains keysize directly from BTreeTupleGetPostingOffset(). It predates #9, whose base already includes this cleanup.

Credits across 4 commits

Author
Peter Eisentraut1
Peter GeogheganLLM1
Reviewer
Peter Geoghegancommitter1
Reported by
Peter Eisentraut1
Tom Lane1
Committer
Peter Geoghegan4

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