Horton

Re: Significant Execution Time Difference Between PG13.14 and PG16.4 for Query on information_schema Tables.

David Rowley <dgrowleyml@gmail.com>

View thread View on postgr.es
From: David Rowley <dgrowleyml@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>, Richard Guo <guofenglinux@gmail.com>
Cc: nikhil raj <nikhilraj474@gmail.com>, pgsql-hackers@lists.postgresql.org
Date: 2024-08-27T23:03:08Z
Lists: pgsql-hackers, pgsql-general 
On Wed, 28 Aug 2024 at 09:52, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> The other problem with this is that it breaks one test case in
> memoize.sql: a query that formerly generated a memoize plan
> now does not use memoize.  I am not sure why not --- does that
> mean anything to you?

The reason it works in master is that get_memoize_path() calls
extract_lateral_vars_from_PHVs() and finds PlaceHolderVars to use as
the Memoize keys. With your patch PlannerInfo.placeholder_list is
empty.

The commit that made this work is 069d0ff02. Richard might be able to
explain better. I don't quite understand why RelOptInfo.lateral_vars
don't contain these in the first place.

David

Commits

  1. Avoid inserting PlaceHolderVars in cases where pre-v16 PG did not.

  2. Remove one memoize test case added by commit 069d0ff02.