Re: Move the injection_points extension to contrib?

From: Antonin Houska <ah(at)cybertec(dot)at>
To: "Hayato Kuroda (Fujitsu)" <kuroda(dot)hayato(at)fujitsu(dot)com>
Cc: "pgsql-hackers(at)lists(dot)postgresql(dot)org" <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: Move the injection_points extension to contrib?
Date: 2025-07-08 15:50:59
Message-ID: 34269.1751989859@localhost
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hayato Kuroda (Fujitsu) <kuroda(dot)hayato(at)fujitsu(dot)com> wrote:

> Dear Antonin,
>
> > Having the extension in contrib would be useful in cases a 3rd party extension
> > uses injection points in its regression tests. In particular, it's more
> > practical for CI to install the "injection_points" extension by installing the
> > "contrib" binary package than by building the whole server from source. (AFAIK
> > the src/modules/injection_points directory is currently not included in any
> > package.)
>
> IIRC, one of the motivation why we put src/test/modules is to ensure the flexiblity.
> Even contrib modules must follow the rule [1]. E.g., ABI must be kept and API
> change should be considered carefully - commits like [2] may be restricted.
> Also We may even have to consider that whether the name of injection points
> should be kept or not.
>
> How do you feel?
>
> [1]: https://www.postgresql.org/docs/devel/xfunc-c.html#XFUNC-GUIDANCE-API-COMPATIBILITY
> [2]: https://github.com/postgres/postgres/commit/f4af4515bb5f3591d49bc16b6cb8ddbf66f98374

ok, I assume you mean that the requirement for ABI/API stability would make it
hard to include tests for fixes like [2] in minor releases. Thanks for
explanation.

--
Antonin Houska
Web: https://www.cybertec-postgresql.com

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Cédric Villemain 2025-07-08 16:06:00 Re: Adding basic NUMA awareness - Preliminary feedback and outline for an extensible approach
Previous Message Bertrand Drouvot 2025-07-08 15:25:26 Re: Adding wait events statistics