Re: generic plans and "initial" pruning

Amit Langote <amitlangote09@gmail.com>

From: Amit Langote <amitlangote09@gmail.com>

To: Thom Brown <thom@linux.com>

Cc: Daniel Gustafsson <daniel@yesql.se>, Tom Lane <tgl@sss.pgh.pa.us>, Alvaro Herrera <alvherre@alvh.no-ip.org>, Andres Freund <andres@anarazel.de>, David Rowley <dgrowleyml@gmail.com>, Jacob Champion <jchampion@timescale.com>, PostgreSQL Hackers <pgsql-hackers@postgresql.org>, Robert Haas <robertmhaas@gmail.com>

Date: 2023-08-03T08:37:39Z

Lists: pgsql-hackers

Attachments

v44-0004-Set-inFromCl-to-false-in-child-table-RTEs.patch (application/octet-stream) patch v44-0004
v44-0006-Track-opened-range-table-relations-in-a-List-in-.patch (application/octet-stream) patch v44-0006
v44-0002-Refactoring-to-move-ExecutorStart-calls-to-be-ne.patch (application/octet-stream) patch v44-0002
v44-0003-Add-field-to-store-parent-relids-to-Append-Merge.patch (application/octet-stream) patch v44-0003
v44-0005-Delay-locking-of-child-tables-in-cached-plans-un.patch (application/octet-stream) patch v44-0005
v44-0001-Make-PlanState-tree-cleanup-non-recursive.patch (application/octet-stream) patch v44-0001

On Wed, Aug 2, 2023 at 10:39 PM Amit Langote <amitlangote09@gmail.com> wrote:
> Having extracted the ExecEndNode() change, I'm also starting to feel
> inclined to extract a couple of other bits from the main patch as
> separate patches, such as moving the ExecutorStart() call from
> PortalRun() to PortalStart() for the multi-query portals.  I'll do
> that in the next version.

Here's a patch set where the refactoring to move the ExecutorStart()
calls to be closer to GetCachedPlan() (for the call sites that use a
CachedPlan) is extracted into a separate patch, 0002.  Its commit
message notes an aspect of this refactoring that I feel a bit nervous
about -- needing to also move the CommandCounterIncrement() call from
the loop in PortalRunMulti() to PortalStart() which now does
ExecutorStart() for the PORTAL_MULTI_QUERY case.

-- 
Thanks, Amit Langote
EDB: http://www.enterprisedb.com

Commits

Stamp 19beta1.
- 4b0bf0788b06 19 (unreleased) cited
Revert "Don't lock partitions pruned by initial pruning"
- 1722d5eb05d8 18.0 landed
Ensure first ModifyTable rel initialized if all are pruned
- 28317de723b6 18.0 cited
Fix bug in cbc127917 to handle nested Append correctly
- cbb9086c9ef6 18.0 landed
Remove unstable test suite added by 525392d57
- 4f1b6e5bb4fe 18.0 landed
Don't lock partitions pruned by initial pruning
- 525392d5727f 18.0 landed
Fix an oversight in cbc127917 to handle MERGE correctly
- 75dfde13639a 18.0 landed
Track unpruned relids to avoid processing pruned relations
- cbc127917e04 18.0 landed
Perform runtime initial pruning outside ExecInitNode()
- d47cbf474ecb 18.0 landed
Move PartitionPruneInfo out of plan nodes into PlannedStmt
- bb3ec16e14de 18.0 landed
Fix setrefs.c's failure to do expression processing on prune steps.
- bf826ea06297 18.0 cited
Remove obsolete executor cleanup code
- d060e921ea5a 17.0 landed
Revert "Move PartitionPruneInfo out of plan nodes into PlannedStmt"
- 5472743d9e85 16.0 landed
Move PartitioPruneInfo out of plan nodes into PlannedStmt
- ec386948948c 16.0 landed
Refactor and cleanup runtime partition prune code a little
- 297daa9d4353 15.0 landed
Remove some unnecessary fields from Plan trees.
- 52ed730d511b 12.0 cited
Remove more redundant relation locking during executor startup.
- f2343653f5b2 12.0 cited
Shut down Gather's children before shutting down Gather itself.
- acf555bc53ac 10.0 cited