Re: should we allow users with a predefined role to access pg_backend_memory_contexts view and pg_log_backend_memory_contexts function?

On Thu, 2021-10-14 at 13:43 -0400, Stephen Frost wrote:
> I'm not sure that it's really inconsistent- if you want the function
> to
> run as someone else, define it as SECURITY DEFINER and it will.

There are two issues:

1. Does having permissions to read a view give the reader the ability
to execute the function as a part of reading the view?

Here it seems like we should allow the user to execute the function
that's a part of the view. If it's doing something that performs
another permission check, then it could fail, but at least they'd be
able to execute it. That seems consistent with the ability to read
tables as a part of reading the view.

2. If the function is executed, is it SECURITY INVOKER or SECURITY
DEFINER?

I think here the answer is SECURITY INVOKER. SECURITY DEFINER doesn't
even really make sense, because the definer might not be the owner of
the view. Maybe we need a concept where the function is executed as
neither the invoker or the definer, but as the owner of the view (or
something else), which sounds appealing, but sounds more like a new
feature.

Regards,
Jeff Davis