Re: Restricting Direct Access to a C Function in PostgreSQL

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Heikki Linnakangas <hlinnaka(at)iki(dot)fi>
Cc: Pavel Stehule <pavel(dot)stehule(at)gmail(dot)com>, Ayush Vatsa <ayushvatsa1810(at)gmail(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Restricting Direct Access to a C Function in PostgreSQL
Date: 2024-08-11 15:29:08
Message-ID: 3748638.1723390148@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Heikki Linnakangas <hlinnaka(at)iki(dot)fi> writes:
> Sounds complicated. I would go with the GRANT approach. Make my_func() a
> SECURITY DEFINER function, and revoke access to my_func_extended() for
> all other roles.

+1

> Another option to consider is to not expose my_func_extended() at the
> SQL level in the first place, and rewrite my_func() in C. Dunno how
> complicated the logic in my_func() is, if that makes sense.

Another way to think about that is "push down into C the part of
my_func() that you feel is necessary to make my_func_extended()
safely callable". Personally I'd probably change my_func_extended()
itself to do that, but if you feel a need to leave it alone, you
could write a C wrapper function. Anyway my point is you might
not have to move *all* of my_func()'s functionality into C. Think
about what it is exactly that makes you feel it's unsafe to call
my_func_extended() directly.

regards, tom lane

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2024-08-11 16:33:30 Re: tiny step toward threading: reduce dependence on setlocale()
Previous Message Alena Rybakina 2024-08-11 13:58:54 Re: Vacuum statistics