Mar 2022

1 commit
Fix header inclusion order in pg_receivewal.c
18 Mar 2022Michael Paquiermaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 100% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

Justin Pryzby proposed using <lz4frame.h> in #54; Michael Paquier also corrected its placement before PostgreSQL headers in #55. The commit makes precisely these changes in pg_receivewal.c, with no attached patchset.

  • Justin Pryzbysaw the final versionLLM analysis, 100% sureReported the incorrect quoted library-header inclusion and proposed the adopted angle-bracket form in #54, before the commit.

Nov 2021

1 commit
Remove useless LZ4 system call on failure when writing file header
24 Nov 2021Michael Paquiermaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 100% surewritten from the discussion, no matching patch2 of 2 reviewers saw the final version

Deletes LZ4F_compressEnd() from the header-write failure path in walmethods.c, as questioned in #44 and #48 and accepted in #49. No attached patchset contains this follow-up deletion.

  • Jeevan Ladhesaw the final versionLLM analysis, 100% sureIdentified the unnecessary LZ4F_compressEnd() call in #44 and pressed the point in #47, leading to its removal.
  • Robert Haassaw the final versionLLM analysis, 100% sureExplained in #48 why freeing the compression context makes the preceding finalization unnecessary on this error path.

Nov 2021

2 commits
Add support for LZ4 compression in pg_receivewal
5 Nov 2021Michael Paquiermaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 100% suremodified before commitlight changesfrom patchset #411 of 4 reviewers saw the final version

Matches the LZ4 implementation, documentation, and TAP tests in #41, with comment edits, a shorter unsupported-build diagnostic, and equivalent reordering of the loop condition. Michael Paquier confirmed application and testing in #42.

  • Michael Paquiersaw the final versiontrailerReviewed and revised the implementation throughout, then committed the final version after testing all zlib/LZ4 build combinations, as reported in #42.
  • Jian Guosaw an earlier versiontrailerReported the missing file-descriptor close in #19, before the later replacement of header-based validation with decompression.
  • Magnus Hagandersaw an earlier versiontrailerRecommended --compression-method in #4 and supported default-only LZ4 compression in #16, before substantial implementation changes.
  • Dilip Kumarsaw an earlier versiontrailerReviewed option naming, acceleration, and dictionary support in #3 and #13, before the segment-validation rework.
Rework compression options of pg_receivewal
4 Nov 2021Michael Paquiermaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 100% surepart of a posted patchsetmoderate changesfrom patchset #371 of 4 reviewers saw the final version

Commits the first patch of the two-patch series in #37, leaving LZ4 support pending. Committer changes include switch-based option validation, including zlib.h instead of redefining its constant, and renaming compression fields to compression_level, as described in #38.

  • Michael Paquiersaw the final versiontrailerDirected the option and filename-detection refactor, then reviewed, adjusted, and committed its final form as reported in #38.
  • Jian Guosaw an earlier versiontrailerReviewed the combined patch in #19 by identifying a descriptor leak in its LZ4 portion, before the split and subsequent refactor changes.
  • Magnus Hagandersaw an earlier versiontrailerHelped choose the option name and defended gzip as the user-facing format name in #33, before the committer reworked option validation.
  • Dilip Kumarsaw an earlier versiontrailerReviewed compression-option naming and semantics in #3 and #13, before the callback, parsing, and validation changes.

Sep 2021

1 commit
Clarify some errors in pg_receivewal when closing WAL segments
17 Sep 2021Michael Paquiermaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #221 of 1 reviewer saw the final version

Extracts the close_walfile() filename-reporting changes from #22 in receivelog.c, with minor statement placement and whitespace changes. Michael Paquier explicitly confirmed this separate application in #23.

  • Michael Paquiersaw the final versioncommittercommitted the patch

Credits across 5 commits

Author
Georgios Kokolatos3
Reviewer
Michael Paquiercommitter3
Dilip Kumar2
Jian Guo2
Magnus Hagander2
Reported by
Justin Pryzby1
Committer
Michael Paquier5