Re: remove partColsUpdated - Mailing list pgsql-hackers

From Nathan Bossart
Subject Re: remove partColsUpdated
Date
Msg-id aPEapYAkDt8Y-QSy@nathan
Whole thread Raw
In response to Re: remove partColsUpdated  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
On Wed, Oct 15, 2025 at 08:18:41PM -0400, Tom Lane wrote:
> Chao Li <li.evan.chao@gmail.com> writes:
>> On Oct 15, 2025, at 23:50, Nathan Bossart <nathandbossart@gmail.com> wrote:
>>> This was first added in commit 2f17844 (v11), and AFAICT it was only used
>>> for a variable named update_tuple_routing_needed in ExecInitModifyTable(),
>>> which was removed by commit c5b7ba4 (v14).  Any objections to removing it
>>> now?
> 
>> Looks like this is only one assignment to it and nobody reads it. I
>> don’t see a reason to retain it.
> 
> I just had a look through https://codesearch.debian.net/ and couldn't
> find any evidence that any extensions are using it, so +1 to remove.

Thanks, will go commit it now.

>> Maybe back patch through 14?
> 
> Certainly not.  That would cause an ABI break for any extension that
> touches later fields in PlannerInfo or ModifyTable.  We don't expect
> extensions to get recompiled for minor releases.

Even if it wasn't an ABI break, there's very little upside to back-patching
small stuff like this that's not hurting anything.  The risk/reward ratio
is not favorable.

-- 
nathan



pgsql-hackers by date:

Previous
From: Jim Jones
Date:
Subject: Re: [PATCH] Add pg_get_trigger_ddl() to retrieve the CREATE TRIGGER statement
Next
From: Ashutosh Bapat
Date:
Subject: Re: Changing shared_buffers without restart