Re: let's make the list of reportable GUCs configurable (was Re: Add %r substitution for psql prompts to show recovery status)

On Thu, Jul 11, 2019 at 3:16 PM Dave Cramer <pg(at)fastcrypt(dot)com> wrote:
> On Thu, 11 Jul 2019 at 15:07, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
>> On Thu, Jul 11, 2019 at 2:29 PM Dave Cramer <pg(at)fastcrypt(dot)com> wrote:
>> > So if I understand this correctly if user bob has altered his search path and there is a security-definer function called owned by him then
>> > the search path will be changed for the duration of the function and reported for every iteration? The implications of this are "interesting" to say the least.
>>
>> I don't believe that it matters what search path has been set using
>> ALTER USER bob.
>
> Why wouldn't it ???

Because the search_path used to execute a security definer function
has nothing to do with the search_path that would be applied to a new
session created by bob.

Unless I am confused. But you can easily test this. Just make a
security-definer function that prints its search_path and experiment.
I think what's going to happen is that it'll inherit the search_path
from the surrounding session unless you use ALTER FUNCTION .. SET, and
that ALTER USER will make no difference. Sometimes I'm wrong, though.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company