Re: Remove unused vacuum-truncate-auto injection point

From: Sami Imseih <samimseih(at)gmail(dot)com>
To: Michael Paquier <michael(at)paquier(dot)xyz>
Cc: pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Remove unused vacuum-truncate-auto injection point
Date: 2026-08-07 22:45:11
Message-ID: CAA5RZ0smjp0R3ZEtZRcCXHFKBkZ0XRCTu-XwkHVAqqZYqvJX4A@mail.gmail.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

> > This looks like an oversight from 661643dedad, perhaps to mirror
> > index_cleanup?
> >
> > ```
> > if (params.index_cleanup == VACOPTVALUE_AUTO)
> > INJECTION_POINT("vacuum-index-cleanup-auto", NULL);
> > ```
> >
> > Small cleanup attached.
>
> I am pretty sure that I have defined this one to be able to track more
> easily an inconsistency if someone reworks this code, so as it becomes
> possible to see if the value is set to "auto" when crossing this code
> path, or at least consider if this is a good idea when doing an index
> cleanup.

To be clear, I am referring to
```
if (params.truncate == VACOPTVALUE_AUTO)
INJECTION_POINT("vacuum-truncate-auto", NULL);
```
truncate can never be auto, and is asserted in vacuumlazy.c

```
Assert(params->truncate != VACOPTVALUE_UNSPECIFIED &&
params->truncate != VACOPTVALUE_AUTO);
```

> Leaving things as they are now is not that bad, IMO. That's
> just one opinion, I'm fine to be outvoted.

I have no strong opinion, but I just don't see this being useful, and
if we ever make truncate support auto, we will need to update
injection_points/sql/vacuum.sql to use the option.

--
Sami

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Masahiko Sawada 2026-08-07 23:48:26 Re: Per-table resync for logical replication subscriptions
Previous Message Michael Paquier 2026-08-07 22:25:26 Re: WAL compression setting after PostgreSQL LZ4 default change