It appears that F-Secure
http://www.f-secure.com/weblog/archives/00001949.html

Trend
http://countermeasures.trendmicro.eu/you-just-cant-trust-a-drunk/

Sophos
http://www.sophos.com/blogs/gc/g/2010/05/11/khobe-vulnerability-game-security-software/

and ESET
http://www.eset.com/blog/2010/05/11/khobe-wan-these-arent-the-droids-youre-looking-for

have posted their 'is-the-game-over' type response.

Juha-Matti

Nick FitzGerald [[email protected]] kirjoitti: 
> Rich Kulawiec wrote:
> 
> > I'm not qualified to evaluate this research on its technical merits,
> > but I believe that some of you are.
> 
> It's a race attack against a classic TOCTTOU (pr. "tock-too"; time-of-
> check-to-time-of-use) vuln.
> 
> The advisory's authors apparently don't know that terminology, but it's 
> a class of security vulnerability that has been known for about as long 
> as we've known about security vulnerabilities.  IIRC (never actually 
> laid eyes on the report myself) this is one of the categories in the 
> (in)famous RISOS Project (Research In Secured Operating Systems) 
> reports from the early 70s.  The typical "fix" to avoid such 
> possibilities is use of a critical section (it's why they were 
> invented, I think) or to make special atomic functions that are 
> effeectively chains of "smaller" functions. Neither is 
> reasonable/possible here -- as I understand the advisory, the code that 
> needs protection against this TOCTTOU can be arbitrarily pre-empted by 
> the scheduler and it would (probably) take significant re-architecting 
> of Windows to provide an atomic function for this special anti-malware 
> purpose (and that would have to be made non-pre-emptible).
> 
> The advisory's authors suggest they have a solution, but they only make 
> that information available to their paying clients.
> 
> 
> 
> Regards,
> 
> Nick FitzGerald
> 

_______________________________________________
Fun and Misc security discussion for OT posts.
https://linuxbox.org/cgi-bin/mailman/listinfo/funsec
Note: funsec is a public and open mailing list.

Reply via email to