Re: Soon-to-be-broken regression test case

From: David Rowley <david(dot)rowley(at)2ndquadrant(dot)com>
To: Alvaro Herrera <alvherre(at)2ndquadrant(dot)com>
Cc: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, Robert Haas <robertmhaas(at)gmail(dot)com>, Jeevan Chalke <jeevan(dot)chalke(at)enterprisedb(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: Soon-to-be-broken regression test case
Date: 2018-10-11 19:45:37
Message-ID: CAKJS1f_0A42gP94CWduAamThYWdY6=VeUZ0u4RvnOPbDiDWA1w@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 12 October 2018 at 05:52, Alvaro Herrera <alvherre(at)2ndquadrant(dot)com> wrote:
> On 2018-Oct-11, Tom Lane wrote:
>
>> Hm, I'm not seeing any regression test result changes there. However,
>> if you're just executing queries and not EXPLAIN'ing them, it's possible
>> something unwanted is happening under the hood.
>
> Hmm, no, the explains are there. Here's one example -- maybe your new
> planner smarts do not change these plans for some reason (I note that
> you mentioned EXISTS in your OP, which this one does not use; I further
> note that we don't use EXISTS anywhere in partition_prune.sql, which
> probably amounts to uncovered cases):
>
> prepare ab_q2 (int, int) as
> select a from ab where a between $1 and $2 and b < (select 3);

I guess if we ever did something to break that then we'd need to not
do anything when there are volatile functions present. If people are
writing that then probably they're doing so to trick the planner,
perhaps to hide some stats that get outdated easily. I'd imagine we'd
upset more people than we'd please.

--
David Rowley http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Christoph Berg 2018-10-11 19:49:27 Re: Debian mips: Failed test 'Check expected t_009_tbl data on standby'
Previous Message Andres Freund 2018-10-11 19:04:24 Re: [HACKERS] removing abstime, reltime, tinterval.c, spi/timetravel