Skip to content

Clean up GPDB_14_MERGE_FIXME comments checked by testing - #2293

Open
chenjinbao1989 wants to merge 3 commits into
apache:mainfrom
chenjinbao1989:fixme-cleanup-comments
Open

Clean up GPDB_14_MERGE_FIXME comments checked by testing#2293
chenjinbao1989 wants to merge 3 commits into
apache:mainfrom
chenjinbao1989:fixme-cleanup-comments

Conversation

@chenjinbao1989

@chenjinbao1989 chenjinbao1989 commented Nov 12, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Comment-only change.

* explain.c: intoClause isn't an optimizer problem. Without this
  assignment, EXPLAIN ANALYZE CREATE TABLE AS EXECUTE crashes the server,
  because ExplainOneQuery() doesn't run on that path.
* procarray.c: the local xmin is safe for definitely_needed. An xid
  between the two bounds makes GlobalVis call ComputeXidHorizons(), which
  takes distributed snapshots into account.
* utility.c: CREATE TABLE LIKE ... INCLUDING ALL with legacy subpartition
  syntax creates the indexes and constraints on all partitions, on the
  coordinator and the segments, so expanding statements at runtime works.

@chenjinbao1989 chenjinbao1989 changed the title Fixme cleanup comments Clean up GPDB_14_MERGE_FIXME comments checked by testing Nov 12, 2026
my-ship-it
my-ship-it approved these changes Nov 12, 2026
View reviewed changes

@my-ship-it my-ship-it left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM, thanks

@chenjinbao1989
chenjinbao1989 enabled auto-merge (rebase) November 12, 2026 21:11
@chenjinbao1989 @reshke
Remove stale GPDB_14_MERGE_FIXME comments and dead code
9294e0d 
These FIXMEs have either been resolved already or match upstream:

* nodeAgg.c: hash_mem_multiplier is already applied; drop the #if 0
  hash_mem_limit.
* functioncmds.c: procForm is already refetched from the new tuple;
  name the include_out_arguments argument of func_get_detail().
* execAmi.c: upstream lists T_TidRangeScan together with T_SeqScan too.
* typecmds.c: upstream also uses typeNamespace for the multirange array.
* distributedlog.c: the code already returns the more accurate value.
* cdbmutate.c: ModifyTable no longer has a plans list; its subplan is
  lefttree, which the generic branch walks.
* pg_dump.c: upstream does not check postDataSchemaOnly for
  DO_TABLE_ATTACH either.
* copyfrom.c, copyto.c: a setvbuf() failure only affects performance.

Also remove dead code: the #if 0 fetch_database_tuple() in vacuum.c,
the unused REINDEX_REL_RECURSING_PARTITIONED_TABLE flag, leftover
pg_pltemplate comments in gp_indexing.h, and the unused
binary_upgrade_next_pg_type_oid variable (its declaration is already
under #if 0 in binary_upgrade.h).
@chenjinbao1989 @reshke
Clean up resolved GPDB_14_MERGE_FIXME comments
9f9657b 
Comment-only change, no behavior change. Each FIXME was checked against
the current code, then removed or turned into a plain "GPDB:" note.

Resolved, so removed:

* seclabel.c: these object types already return false.
* tablecmds.c: foreign tables are already checked first.
* createplan.c: ModifyTablePath has a single subpath.
* CTranslatorDXLToPlStmt.cpp: splitUpdate is set, and the executor does
  not read partColsUpdated.
* dependency.c: every object type tracked in pg_stat_last_* has a
  matching drop tracker.
* decode.c, pg_control.h: 0xC0 is XLOG_NEXTRELFILENODE.
* parse_partition_gp.c: DETACH CONCURRENTLY is disabled, so no partition
  can be pending detach and omit_detached makes no difference.
* pg_dump.c: the multirange function signature matches, the additional
  function columns are queried unconditionally, and the commented-out
  asserts were inverted.
* copyfrom.c: CopyInitDataParser() has two callers.
* brin_bloom_optimizer.out, brin_multi_optimizer*.out: the plans do use
  a bitmap index scan, contrary to the comment.

Intentional differences from upstream, so now documented as such: the
horizon comments in procarray.c, analyze.c, nbtree.c, pathnodes.h,
pg_proc.dat, varsup_test.c, pg_dump.c and type_sanity.

gram.y: the keyword precedence list does not have to follow
bare_label_keyword. Removing it causes shift/reduce conflicts, and a new
conflicting keyword fails the build because of %expect 0. Postfix
operators no longer exist, so describe the actual reason.
@chenjinbao1989 @reshke
Clean up GPDB_14_MERGE_FIXME comments checked by testing
2676813 
Comment-only change.

* explain.c: intoClause isn't an optimizer problem. Without this
  assignment, EXPLAIN ANALYZE CREATE TABLE AS EXECUTE crashes the server,
  because ExplainOneQuery() doesn't run on that path.
* procarray.c: the local xmin is safe for definitely_needed. An xid
  between the two bounds makes GlobalVis call ComputeXidHorizons(), which
  takes distributed snapshots into account.
* utility.c: CREATE TABLE LIKE ... INCLUDING ALL with legacy subpartition
  syntax creates the indexes and constraints on all partitions, on the
  coordinator and the segments, so expanding statements at runtime works.
@reshke
reshke force-pushed the fixme-cleanup-comments branch from d5925cf to 2676813 Compare November 14, 2026 14:54
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@leborchuk leborchuk leborchuk approved these changes
@my-ship-it my-ship-it my-ship-it approved these changes

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@chenjinbao1989 @leborchuk @my-ship-it