Content-Type: multipart/alternative;
 boundary="------------0PMfUN0Nj31EFVbUEIQtaIQr"
Message-ID: <9b10b847-067e-45cf-bbf7-7802b6303fc0@gmail.com>
Date: Sat, 30 Aug 2025 16:00:01 +0300
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: Correction of RowMark Removal During Sel-Join Elimination
To: Alexander Korotkov <aekorotkov@gmail.com>,
 Andrei Lepikhov <lepihov@gmail.com>
Cc: Greg Sabino Mullane <htamfids@gmail.com>,
 PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
References: <18c6bd6c-6d2a-419a-b0da-dfedef34b585@gmail.com>
 <CAKAnmmJzSGzTRaLATNZPErByO3-j6MwsK--qPSiRLJi4kB05HQ@mail.gmail.com>
 <e7c8173c-91be-41a3-a390-2ca84fbc5147@gmail.com>
 <CAPpHfdsP_abX5hjnxiRJ6guU9P1FH4EdNTMksoGD05tTG1o9vA@mail.gmail.com>
Content-Language: en-US
From: Alexander Lakhin <exclusion@gmail.com>
In-Reply-To: 
 <CAPpHfdsP_abX5hjnxiRJ6guU9P1FH4EdNTMksoGD05tTG1o9vA@mail.gmail.com>
Archived-At: 
 <https://www.postgresql.org/message-id/9b10b847-067e-45cf-bbf7-7802b6303fc0%40gmail.com>
Precedence: bulk

This is a multi-part message in MIME format.
--------------0PMfUN0Nj31EFVbUEIQtaIQr
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit

Hello Alexander,

24.08.2025 03:44, Alexander Korotkov wrote:
> Thank you for catching this.  And thank you for the fix.  I think it
> worth separating fix and refactoring.  This helps to understand what
> exactly the fix is by looking at the patch.  I also edited commit
> message.  I'm going to push this if no objections.

Please try the following script:
create table t1(i1 int primary key);
create table t2 (i2 int, check (i2 is not null));

set constraint_exclusion = 'on';

select 1 from t1 right join t2 on i1 = i2 where (select true);

which triggers (starting from 5f6f951f8):
Core was generated by `postgres: law regression [local] SELECT                                       '.
Program terminated with signal SIGSEGV, Segmentation fault.
#0  0x000058dd6a7b6014 in var_is_nonnullable (root=0x58dd78d505f8, var=0x58dd78d75610, use_rel_info=false) at clauses.c:4242
4242                    if (rte->inh && rte->relkind != RELKIND_PARTITIONED_TABLE)
(gdb) bt
#0  0x000058dd6a7b6014 in var_is_nonnullable (root=0x58dd78d505f8, var=0x58dd78d75610, use_rel_info=false) at clauses.c:4242
#1  0x000058dd6a7b4f74 in eval_const_expressions_mutator (node=0x58dd78d75590, context=0x7fff875a9f00) at clauses.c:3556
#2  0x000058dd6a7b2935 in eval_const_expressions (root=0x58dd78d505f8, node=0x58dd78d75590) at clauses.c:2272
#3  0x000058dd6a7c857d in get_relation_constraints (root=0x58dd78d505f8, relationObjectId=16385, rel=0x58dd78d738e0, 
include_noinherit=true, include_notnull=true, include_partition=true) at plancat.c:1419
#4  0x000058dd6a7c8fdb in relation_excluded_by_constraints (root=0x58dd78d505f8, rel=0x58dd78d738e0, rte=0x58dd78c6b968) 
at plancat.c:1808
#5  0x000058dd6a73675e in set_rel_size (root=0x58dd78d505f8, rel=0x58dd78d738e0, rti=2, rte=0x58dd78c6b968) at 
allpaths.c:364
#6  0x000058dd6a73664c in set_base_rel_sizes (root=0x58dd78d505f8) at allpaths.c:322
#7  0x000058dd6a73631e in make_one_rel (root=0x58dd78d505f8, joinlist=0x58dd78d74a90) at allpaths.c:183
...

Discovered with SQLsmith.

Best regards,
Alexander
--------------0PMfUN0Nj31EFVbUEIQtaIQr
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">Hello Alexander,<br>
      <br>
      24.08.2025 03:44, Alexander Korotkov wrote:<span
      style="white-space: pre-wrap">
</span></div>
    <blockquote type="cite"
cite="mid:CAPpHfdsP_abX5hjnxiRJ6guU9P1FH4EdNTMksoGD05tTG1o9vA@mail.gmail.com">
      <pre wrap="" class="moz-quote-pre">Thank you for catching this.  And thank you for the fix.  I think it
worth separating fix and refactoring.  This helps to understand what
exactly the fix is by looking at the patch.  I also edited commit
message.  I'm going to push this if no objections.
</pre>
    </blockquote>
    <br>
    Please try the following script:<br>
    create table t1(i1 int primary key);<br>
    create table t2 (i2 int, check (i2 is not null));<br>
    <br>
    set constraint_exclusion = 'on';<br>
    <br>
    select 1 from t1 right join t2 on i1 = i2 where (select true);<br>
    <br>
    which triggers (starting from 5f6f951f8):<br>
    Core was generated by `postgres: law regression [local]
    SELECT                                       '.<br>
    Program terminated with signal SIGSEGV, Segmentation fault.<br>
    #0  0x000058dd6a7b6014 in var_is_nonnullable (root=0x58dd78d505f8,
    var=0x58dd78d75610, use_rel_info=false) at clauses.c:4242<br>
    4242                    if (rte-&gt;inh &amp;&amp; rte-&gt;relkind
    != RELKIND_PARTITIONED_TABLE)<br>
    (gdb) bt<br>
    #0  0x000058dd6a7b6014 in var_is_nonnullable (root=0x58dd78d505f8,
    var=0x58dd78d75610, use_rel_info=false) at clauses.c:4242<br>
    #1  0x000058dd6a7b4f74 in eval_const_expressions_mutator
    (node=0x58dd78d75590, context=0x7fff875a9f00) at clauses.c:3556<br>
    #2  0x000058dd6a7b2935 in eval_const_expressions
    (root=0x58dd78d505f8, node=0x58dd78d75590) at clauses.c:2272<br>
    #3  0x000058dd6a7c857d in get_relation_constraints
    (root=0x58dd78d505f8, relationObjectId=16385, rel=0x58dd78d738e0,
    include_noinherit=true, include_notnull=true,
    include_partition=true) at plancat.c:1419<br>
    #4  0x000058dd6a7c8fdb in relation_excluded_by_constraints
    (root=0x58dd78d505f8, rel=0x58dd78d738e0, rte=0x58dd78c6b968) at
    plancat.c:1808<br>
    #5  0x000058dd6a73675e in set_rel_size (root=0x58dd78d505f8,
    rel=0x58dd78d738e0, rti=2, rte=0x58dd78c6b968) at allpaths.c:364<br>
    #6  0x000058dd6a73664c in set_base_rel_sizes (root=0x58dd78d505f8)
    at allpaths.c:322<br>
    #7  0x000058dd6a73631e in make_one_rel (root=0x58dd78d505f8,
    joinlist=0x58dd78d74a90) at allpaths.c:183<br>
    ...<br>
    <br>
    Discovered with SQLsmith.<br>
    <br>
    Best regards,<br>
    Alexander<br>
  </body>
</html>

--------------0PMfUN0Nj31EFVbUEIQtaIQr--