Re: Assert in pageinspect with NULL pages

Peter Geoghegan <pg@bowt.ie>

From: Peter Geoghegan <pg@bowt.ie>
To: Robert Haas <robertmhaas@gmail.com>
Cc: Michael Paquier <michael@paquier.xyz>, Alexander Lakhin <exclusion@gmail.com>, Justin Pryzby <pryzby@telsasoft.com>, Daria Lepikhova <d.lepikhova@postgrespro.ru>, PostgreSQL Hackers <pgsql-hackers@postgresql.org>
Date: 2022-03-27T21:36:54Z
Lists: pgsql-hackers
On Sun, Mar 27, 2022 at 2:02 PM Robert Haas <robertmhaas@gmail.com> wrote:
> On Sun, Mar 27, 2022 at 4:26 PM Peter Geoghegan <pg@bowt.ie> wrote:
> > We're not dealing
> > with adversarial page images here.
>
> I think it's bad that we have to make that assumption, considering
> that there's nothing whatever to keep users from supplying arbitrary
> page images to pageinspect.

Maybe it isn't strictly necessary for bt_page_items(), but making that
level of guarantee is really hard, and not particularly useful. And
that's the easy case for pageinspect: gist_page_items() takes a raw
bytea, and puts it through the underlying types output functions.

I think that it might actually be fundamentally impossible to
guarantee that that'll be safe, because we have no idea what the
output function might be doing. It's arbitrary user-defined code that
could easily be implemented in C. Combined with an arbitrary page
image.

> But I also agree that if we're unable or
> unwilling to make things perfect, it's still good to make them better.

I think that most of the functions can approach being perfectly
robust, with a little work. In practical terms they can almost
certainly be made so robust that no real user of bt_page_items() will
ever crash the server. Somebody that goes out of their way to do that
*might* find a way (even with the easier cases), but that doesn't
particularly concern me.

-- 
Peter Geoghegan



Commits

  1. pageinspect: Fix handling of all-zero pages

  2. pageinspect: Add more sanity checks to prevent out-of-bound reads

  3. pageinspect: Fix handling of page sizes and AM types

  4. Teach pageinspect about nbtree deduplication.