public inbox for [email protected]  

help / color / mirror / Atom feed

From: Andrew Dunstan <[email protected]>
To: Mark Dilger <[email protected]>
To: Tom Lane <[email protected]>
Cc: Joshua Brindle <[email protected]>
Cc: Robert Haas <[email protected]>
Cc: Jeff Davis <[email protected]>
Cc: PostgreSQL-development <[email protected]>
Cc: Joe Conway <[email protected]>
Subject: Re: Granting SET and ALTER SYSTE privileges for GUCs
Date: Wed, 16 Mar 2022 07:43:28 -0400
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>
References: <[email protected]>
	<[email protected]>
	<CA+TgmobUt_tSF-9tz-eRvWzT1pbe_ssvxfnPniGGEgLzems40Q@mail.gmail.com>
	<[email protected]>
	<[email protected]>
	<[email protected]>
	<[email protected]>
	<[email protected]>
	<CAGB+Vh4_UJsYHCWborm-y4CFt8gVvY_KbcNrpWcpXAk-GcwOtA@mail.gmail.com>
	<[email protected]>
	<[email protected]>
	<[email protected]>
	<[email protected]>
	<[email protected]>
	<[email protected]>
	<[email protected]>
	<[email protected]>
	<[email protected]>


On 3/15/22 16:59, Mark Dilger wrote:
>> On Mar 6, 2022, at 3:27 PM, Tom Lane <[email protected]> wrote:
>>
>> Mark Dilger <[email protected]> writes:
>>> The existing patch allows grants on unknown gucs, because it can't know what guc an upgrade script will introduce, and the grant statement may need to execute before the guc exists.
>> Yeah, that's the problematic case.  It might mostly work to assume that
>> an unknown GUC has an empty default ACL.  This could fail to retain the
>> default PUBLIC SET permission if it later turns out the GUC is USERSET
> On further reflection, I concluded this isn't needed.  No current extension, whether in-core or third party, expects to be able to create a new GUC and then grant or revoke permissions on it.  They can already specify the guc context (PGC_USERS, etc).  Introducing a feature that depends on the dubious assumption that unrecognized GUCs will turn out to be USERSET doesn't seem warranted.

Agreed.


>
> The patch attributes all grants of setting privileges to the bootstrap superuser.  Only superusers can grant or revoke privileges on settings, and all settings are implicitly owned by the bootstrap superuser because there is no explicit owner associated with settings.  Consequently, select_best_grantor(some_superuser, ..., BOOTSTRAP_SUPERUSERID, ...) always chooses the bootstrap superuser.  I don't see a problem with this, but wouldn't mind a second opinion.  Some people might find it surprising when viewing the pg_setting_acl.setacl field.


I think it's OK as long as we document it. An alternative might be to
invent a pseudo-superuser called, say, 'postgres_system', but that seems
like overkill to solve what is in effect a cosmetic problem.

Generally I think this is now in fairly good shape, I've played with it
and it seems to do what I expect in every case, and the things I found
surprising are gone.


cheers


andrew


--
Andrew Dunstan
EDB: https://www.enterprisedb.com

view thread (24+ messages)  latest in thread

reply

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Reply to all the recipients using the --to and --cc options:
  reply via email

  To: [email protected]
  Cc: [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected]
  Subject: Re: Granting SET and ALTER SYSTE privileges for GUCs
  In-Reply-To: <[email protected]>

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox