Content-Type: multipart/alternative;
 boundary="------------kV5S6HDi2lGzHeaVWjHBSQV0"
Message-ID: <b63acc7a-60bc-4595-81a9-fc6de22e367f@postgrespro.ru>
Date: Mon, 31 Mar 2025 06:21:43 +0300
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: Memoize ANTI and SEMI JOIN inner
To: David Rowley <dgrowleyml@gmail.com>
Cc: Andrei Lepikhov <lepihov@gmail.com>,
 PostgreSQL-development <pgsql-hackers@postgresql.org>
References: <60bf8d26-7c7e-4915-b544-afdb9020011d@gmail.com>
 <95586c5d-2632-4f1c-bfb9-4c391f260de4@gmail.com>
 <CAApHDvr2NRbjLXEUXvvKNCDcHfhC6UBL6ZmJzXZ29MBPi3+y8g@mail.gmail.com>
 <da135789-dcaf-49a6-9d05-25b68f1f72c5@gmail.com>
 <26e6dd23-ad56-4e00-8a24-c84b508267aa@postgrespro.ru>
 <CAApHDvqbVaiAV0xsk0G4W7mmN89vG-dZ1jTYQS=jT1041e82zA@mail.gmail.com>
Content-Language: en-US
From: Alena Rybakina <a.rybakina@postgrespro.ru>
In-Reply-To: 
 <CAApHDvqbVaiAV0xsk0G4W7mmN89vG-dZ1jTYQS=jT1041e82zA@mail.gmail.com>
Archived-At: 
 <https://www.postgresql.org/message-id/b63acc7a-60bc-4595-81a9-fc6de22e367f%40postgrespro.ru>
Precedence: bulk

This is a multi-part message in MIME format.
--------------kV5S6HDi2lGzHeaVWjHBSQV0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit

On 31.03.2025 06:04, David Rowley wrote:
> On Mon, 31 Mar 2025 at 15:33, Alena Rybakina<a.rybakina@postgrespro.ru>  wrote:
>> I believe it's worth asserting that both inner_unique and single_mode are not true at the same time — just as a safety check.
> add_paths_to_joinrel() just chooses not to populate inner_unique for
> SEMI and ANTI joins because, as of today's master, it's pretty
> pointless to determine that because the executor will short-circuit
> and skip to the next outer tuple for those join types anyway. I don't
> follow why having both these flags set would cause trouble. It seems
> perfectly legitimate that add_paths_to_joinrel() could choose to set
> the inner_unique flag for these join types, and if it did, the Assert
> you're proposing would fail for no good reason.
>
I tend to agree with you that someone might set this flag to true for 
these join types in the future.

However, is it necessary to check that extra->inner_unique must be false 
for SEMI/ANTI joins here, or am I missing something? It looks a little 
confusing at this point.

if(!extra->inner_unique&& (jointype== JOIN_SEMI|| jointype== JOIN_ANTI))

     single_mode= true;

--

Regards,
Alena Rybakina
Postgres Professional

--------------kV5S6HDi2lGzHeaVWjHBSQV0
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: 8bit

<!DOCTYPE html>
<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <div class="moz-cite-prefix">On 31.03.2025 06:04, David Rowley
      wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:CAApHDvqbVaiAV0xsk0G4W7mmN89vG-dZ1jTYQS=jT1041e82zA@mail.gmail.com">
      <pre class="moz-quote-pre" wrap="">On Mon, 31 Mar 2025 at 15:33, Alena Rybakina <a class="moz-txt-link-rfc2396E" href="mailto:a.rybakina@postgrespro.ru">&lt;a.rybakina@postgrespro.ru&gt;</a> wrote:
</pre>
      <blockquote type="cite">
        <pre class="moz-quote-pre" wrap="">I believe it's worth asserting that both inner_unique and single_mode are not true at the same time — just as a safety check.
</pre>
      </blockquote>
      <pre class="moz-quote-pre" wrap="">
add_paths_to_joinrel() just chooses not to populate inner_unique for
SEMI and ANTI joins because, as of today's master, it's pretty
pointless to determine that because the executor will short-circuit
and skip to the next outer tuple for those join types anyway. I don't
follow why having both these flags set would cause trouble. It seems
perfectly legitimate that add_paths_to_joinrel() could choose to set
the inner_unique flag for these join types, and if it did, the Assert
you're proposing would fail for no good reason.

</pre>
    </blockquote>
    I tend to agree with you that someone might set this flag to true
    for these join types in the future. <br>
    <p>However, is it necessary to check that extra-&gt;inner_unique
      must be false for SEMI/ANTI joins here, or am I missing something?
      It looks a little confusing at this point.</p>
    <p><span style="color: #0000ff;">if</span><span
        style="color: #000000;"> (!</span><span style="color: #808080;">extra</span><span
        style="color: #000000;">-&gt;</span><span
        style="color: #000000;">inner_unique</span><span
        style="color: #000000;"> &amp;&amp; (</span><span
        style="color: #808080;">jointype</span><span
        style="color: #000000;"> == </span><span
        style="color: #2f4f4f;">JOIN_SEMI</span><span
        style="color: #000000;"> || </span><span
        style="color: #000000;"></span><span style="color: #808080;">jointype</span><span
        style="color: #000000;"> == </span><span
        style="color: #2f4f4f;">JOIN_ANTI</span><span
        style="color: #000000;">))</span></p>
    <div><span style="color: #000000;"> </span><span
        style="color: #000000;">    single_mode</span><span
        style="color: #000000;"> = </span><span style="color: #bd63c5;">true</span><span
        style="color: #000000;">;</span></div>
    <pre class="moz-quote-pre" style="white-space: pre-wrap;" wrap="">
--
</pre>
    <pre class="moz-signature" cols="72">Regards,
Alena Rybakina
Postgres Professional</pre>
  </body>
</html>

--------------kV5S6HDi2lGzHeaVWjHBSQV0--