Re: pg18: Virtual generated columns are not (yet) safe when superuser selects from them

Robert Haas <robertmhaas@gmail.com>

From: Robert Haas <robertmhaas@gmail.com>

To: jian he <jian.universality@gmail.com>

Cc: Peter Eisentraut <peter@eisentraut.org>, Feike Steenbergen <feikesteenbergen@gmail.com>, PostgreSQL mailing lists <pgsql-hackers@postgresql.org>

Date: 2025-06-05T19:15:53Z

Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

Restrict virtual columns to use built-in functions and types
- 0cd69b3d7ef3 18.0 landed
Fix virtual generated column type checking for ALTER TABLE
- 49fe1c83ecf3 18.0 landed
Expand virtual generated columns in the planner
- 1e4351af329f 18.0 cited

On Thu, Jun 5, 2025 at 11:19 AM jian he <jian.universality@gmail.com> wrote:
> I think it will work.
> because we already require the generated column expression to be
> immutable functions.
>
> The above functions you mentioned are all not immutable.

Hmm. I guess I have no evidence that we have built-in immutable
functions that would cause a problem here. I still think it's a bad
direction to go.

-- 
Robert Haas
EDB: http://www.enterprisedb.com