Horton

Re: [SECURITY] DoS attack on backend possible (was: Re:

Florian Weimer <weimer@cert.uni-stuttgart.de>

View thread
From: Florian Weimer <Weimer@CERT.Uni-Stuttgart.DE>
To: Greg Copeland <greg@CopelandConsulting.Net>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Justin Clift <justin@postgresql.org>, PostgresSQL Hackers Mailing List <pgsql-hackers@postgresql.org>
Date: 2002-08-12T13:48:10Z
Lists: pgsql-hackers 
Greg Copeland <greg@CopelandConsulting.Net> writes:

> Well, if it's a buffer overrun, there is certainly potential for risks
> well beyond that of simply crashing the "be".

It's a buffer overrun, but the data has to pass through the date/time
parser in the backend, so it's not entirely obvious how you can
exploit this to run arbitrary code.

-- 
Florian Weimer 	                  Weimer@CERT.Uni-Stuttgart.DE
University of Stuttgart           http://CERT.Uni-Stuttgart.DE/people/fw/
RUS-CERT                          fax +49-711-685-5898