minor bug in sort_inner_and_outer()

From: Tomas Vondra <tomas(dot)vondra(at)enterprisedb(dot)com>
To: PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: minor bug in sort_inner_and_outer()
Date: 2022-01-14 00:48:33
Message-ID: bf3a6ea1-a7d8-7211-0669-189d5c169374@enterprisedb.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi,

While looking at sort_inner_and_outer() I was rather confused what is
stored in all_pathkeys, because the code does this:

List *all_pathkeys;

...

all_pathkeys = select_outer_pathkeys_for_merge(root,
extra->mergeclause_list,
joinrel);

foreach(l, all_pathkeys)
{
List *front_pathkey = (List *) lfirst(l);
...

/* Make a pathkey list with this guy first */
if (l != list_head(all_pathkeys))
outerkeys = lcons(front_pathkey,
...);
else
...

which seems to suggest all_pathkeys is a list of lists, because why else
would front_pathkey be a (List *). But that doesn't seem to be the case,
front_pathkey is actually a PathKey, not a List, as demonstrated by gdb:

(gdb) p *front_pathkey
$2 = {type = T_PathKey, length = 0, ...}

Maybe it's some clever list-fu that I can't comprehend, but I guess it's
a bug present since ~2004. It's benign because we only ever pass the
front_pathkey to lcons() which does not really care.

regards

--
Tomas Vondra
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

Attachment Content-Type Size
sort-inner-and-outer-fix.patch text/x-patch 492 bytes

Browse pgsql-hackers by date

  From Date Subject
Next Message Amit Kapila 2022-01-14 02:16:56 Re: row filtering for logical replication
Previous Message Tomas Vondra 2022-01-14 00:39:30 Re: PATCH: generate fractional cheapest paths in generate_orderedappend_path