Horton

Re: pageinspect: Hash index support

Alvaro Herrera <alvherre@2ndquadrant.com>

View thread View on postgr.es
From: Alvaro Herrera <alvherre@2ndquadrant.com>
To: Jeff Janes <jeff.janes@gmail.com>
Cc: Michael Paquier <michael.paquier@gmail.com>, Jesper Pedersen <jesper.pedersen@redhat.com>, PostgreSQL Hackers <pgsql-hackers@postgresql.org>
Date: 2016-09-26T20:22:33Z
Lists: pgsql-hackers 
Jeff Janes wrote:
> On Tue, Sep 20, 2016 at 12:19 AM, Michael Paquier <michael.paquier@gmail.com
> > wrote:
> 
> >
> > Note: the patch checks if a superuser is calling the new functions,
> > which is a good thing.
> >
> 
> If we only have the bytea functions and the user needs to supply the raw
> pages themselves, rather than having the function go get the raw page for
> you, is there any reason to restrict the interpretation function to super
> users?  I guess if we don't trust the C coded functions not to dereference
> bogus data in harmful ways?

Yeah, it'd likely be simple to manufacture a fake page that causes the
server to misbehave resulting in a security leak or at least DoS.

-- 
Álvaro Herrera                https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services

Commits

  1. Fix incorrect typecast.

  2. In pageinspect/hashfuncs.c, avoid crashes on alignment-picky machines.

  3. pageinspect: Support hash indexes.

  4. Add uuid to the set of types supported by contrib/btree_gist.

  5. Allow CREATE EXTENSION to follow extension update paths.