Horton

Re: Sequence Access Methods, round two

Michael Paquier <michael@paquier.xyz>

View thread
From: Michael Paquier <michael@paquier.xyz>
To: Kirill Reshke <reshkekirill@gmail.com>
Cc: Peter Eisentraut <peter@eisentraut.org>, Peter Smith <smithpb2250@gmail.com>, Postgres hackers <pgsql-hackers@lists.postgresql.org>
Date: 2025-08-21T02:57:58Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Refactor init_params() in sequence.c to not use FormData_pg_sequence_data

  2. Fix comment thinko in sequence.c

  3. Group more closely cache updates for backends in sequence.c

  4. Introduce sequence_*() access functions

Attachments

On Tue, Aug 19, 2025 at 10:14:33AM +0500, Kirill Reshke wrote:
> I played with patch sets some more, trying to break things in nasty
> ways (something like CREATE OPERATOR FAMILY ff USING seqlocal or other
> incorrect ddl).
> But I had no success. I will try some more later, maybe I will find something

Thanks.

> PFA diff which describes the access method of sequence on \d+ psql
> meta-command likewise we do it for tables.

Yes, that may be useful, but I don't think that this should use the
psql variable to hide table AMs.  I am rebasing a new patch set, v18,
with a couple of changes:
- Added your feedback about psql.
- I have put more thoughts into the code shared between the in-core
sequence method and the snowflake one, and looked at reducing the
duplication between the two.  At the end, I have introduced a new
header called sequence_page.h, which is able to reduce the work for
AMs when these rely on a single page through the addition of macros
able to initialize and read sequence pages.  In this patch set, this
new part is labelled with 0006.
- Fixed a few more things, like comments.
--
Michael