On Wed, Nov 23, 2005 at 07:07:21PM +0100, Bernd Eckenfels wrote: > In article <[🔎] 5B97FFB7F28E3E4294B049C968CBF47BE46E1B@SERVER-EX02A.INTERN.BBNED.NL> you wrote: > > Well, obviously it is not a _security_ bug, since it has nothing to do > > with security. > ... Without looking at the bug in detail you cannot tell for sure. A DoS condition can become a remote exploit. (...) > However it doesnt matter, you are right: critical application crashes > (especially if triggerable by untrusted peers) are critical enough to be > fixed anyway. AND crashes often have the potential to be exploitable > (stacksmashing?). Indeed (to the last sentence) for a recent example take a look at CAN-2005-1790, which was a DoS condition for IE 5.5 and 6.x, it was initially labeled as "DoS - low risk" and has been later found to be exploitable and allow for remote compromise: http://www.computerterrorism.com/research/ie/ct21-11-2005 Regards Javier
Attachment:
signature.asc
Description: Digital signature