Mar 2023

3 commits
Improve several permission-related error messages.
17 Mar 2023Peter Eisentrautmaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 100% surelanded as postedfrom patchset #641 of 4 reviewers saw the final version

The complete diff matches #64, including the retained CREATEROLE requirement in DropRole's detail message and the final file_fdw parameterization.

  • Tom Lanesaw an earlier versionLLM analysis, 96% sureTom recommended errdetail in #36, supplied the separate password explanation in #43, and supported omitting redundant superuser references in #49, before the patch expanded beyond user.c.
  • Alvaro Herrerasaw an earlier versionLLM analysis, 97% sureAlvaro proposed parameterized messages and clearer attribute wording in #37, then approved the patch while requesting the bootstrap-user error-code correction in #51, before the broader companion patch was added.
  • Robert Haassaw an earlier versionLLM analysis, 96% sureRobert questioned combining password and attribute failures in #39 and helped settle the superuser-message policy in #46 and #48, before the cleanup expanded.
  • Peter Eisentrautsaw the final versioncommittercommitted the patch
Integrate superuser check into has_rolreplication()
16 Mar 2023Peter Eisentrautmaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #591 of 1 reviewer saw the final version

This extracts the unchanged has_rolreplication superuser check and removal of redundant caller checks from #59, touching miscinit.c, slot.c, and postinit.c. Peter reports committing the non-message pieces separately in #60.

  • Peter Eisentrautsaw the final versioncommittercommitted the patch
Small code simplification
16 Mar 2023Peter Eisentrautmaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #594 of 4 reviewers saw the final version

This extracts the unchanged RenameRole simplification from #59, replacing a repeated GETSTRUCT cast with authform->rolsuper in src/backend/commands/user.c.

  • Tom Lanesaw the final versionLLM analysis, 80% sureTom gave the draft containing this simplification a favorable once-over in #36, and the extracted line remained unchanged through #59.
  • Alvaro Herrerasaw the final versionLLM analysis, 86% sureAlvaro said the user.c patch looked good in #51, and its RenameRole simplification was retained unchanged in #59.
  • Robert Haassaw the final versionLLM analysis, 78% sureRobert reviewed the user.c patch favorably apart from its password-error handling in #39, while the included RenameRole simplification remained unchanged.
  • Peter Eisentrautsaw the final versioncommittercommitted the patch

Jan 2023

4 commits
Adjust interaction of CREATEROLE with role properties.
24 Jan 2023Robert Haasmaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 100% surelanded as postedfrom patchset #83 of 3 reviewers saw the final version

The seven-file diff matches #8, including the CREATEROLE checks, exported have_createdb_privilege function, documentation, and regression tests.

  • Nathan Bossartsaw the final versionLLM analysis, 100% sureNathan endorsed the change and exercised its permission failures in #12, identifying error-message improvements as separate follow-up work.
  • Tushar Ahujasaw the final versionLLM analysis, 100% sureTushar reported successful testing of role attributes, pg_dumpall, and pg_basebackup with Neha in #33.
  • Neha Sharmasaw the final versionLLM analysis, 100% sureThe commit explicitly credits Neha Sharma as a reviewer, corroborated by Tushar's joint testing report in #33, although she did not post in this thread.
Add new GUC reserved_connections.
20 Jan 2023Robert Haasmaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 99% surepart of a posted patchsetlight changesfrom patchset #242 of 2 reviewers saw the final version

The diff exactly matches the reserved_connections feature component of the three-patch series in #24; the prerequisite rename and documentation correction landed separately.

  • Tushar Ahujasaw the final versionLLM analysis, 100% sureTushar tested connection exhaustion and found incorrect role names in #15, #16, and #17, then confirmed the latest patches fixed them in #26.
  • Robert Haassaw the final versioncommittercommitted the patch
Rename ReservedBackends variable to SuperuserReservedConnections.
20 Jan 2023Robert Haasmaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 99% surepart of a posted patchsetlight changesfrom patchset #242 of 2 reviewers saw the final version

The four-file diff exactly matches the variable-renaming component of #24, changing ReservedBackends to SuperuserReservedConnections without additional implementation edits.

  • Tushar Ahujasaw the final versionLLM analysis, 97% sureThe commit explicitly credits Tushar Ahuja, whose naming review and confirmation of the latest series appear in #15 and #26.
  • Robert Haassaw the final versioncommittercommitted the patch
Update docs and error message for superuser_reserved_connections.
20 Jan 2023Robert Haasmaster16beta1REL_19_STABLE16beta1REL_18_STABLE16beta1REL_17_STABLE16beta1REL_16_STABLE16beta1
LLM analysis, 99% surepart of a posted patchsetlight changesfrom patchset #242 of 2 reviewers saw the final version

The diff exactly matches the first component of #24, removing obsolete replication restrictions from doc/src/sgml/config.sgml and the connection-limit error in postinit.c.

  • Tushar Ahujasaw the final versionLLM analysis, 94% sureThe commit explicitly credits Tushar Ahuja as a reviewer, and #26 records his confirmation after testing the latest reserved-connections series.
  • Robert Haassaw the final versioncommittercommitted the patch

Credits across 7 commits

Author
Nathan Bossart3 LLM6
Robert HaasLLM1
Reviewer
Peter Eisentrautcommitter3
Robert Haascommitter3
Committer
Robert Haas4
Peter Eisentraut3

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