Re: A quote from Crypto-Gram

From: Mok-Kong Shen (mok-kong.shen_at_t-online.de)
Date: 08/17/04 

Date: Tue, 17 Aug 2004 15:31:00 +0200

Juergen Nieveler wrote:

[snip]
> It's still unclear to me just HOW one could manipulate the microcode to
> get access to data. Manipulating the RNG so that it always gives the
> same fairly long sequence, that's something I could see, but the
> microcode in the processor affects the commands a processor knows - it
> cannot differentiate between applications, and it would need a covert
> channel to send out the intercepted information. The processor doesn't
> do anything by itself, it needs an application to tell it what to do...
> the sequence of the commands isn't affected by the microcode, nor can
> it insert any commands into the application.

That's apparently beyond the knowledge of most of the group
(and me in particular). Presumably that's the same also for
the majority of sophisticated techniques currently employed
by the writers of malware. Establishing a cover channel is
what trojans do, if I don't err. Anyway, it appears to be
prudent in my view to take that potential risk into account,
in case one is in a category susceptible to sufficiently
high risks (targets of opponents having immense resources).
A nobody like me certainly needn't bother with risks of
attacks via microcodes (but perhaps instead via operating
systems), at least in the near future.

M. K. Shen
--------------------------------------

[OT] BTW, here is a recent news concerning security of PDAs:

    http://www.nwfusion.com/news/2004/080904pdavirus.html

Relevant Pages

  • Re: A quote from Crypto-Gram
    ... > It's still unclear to me just HOW one could manipulate the microcode to ... > microcode in the processor affects the commands a processor knows - it ... Every PC on startup executes it's own POST. ...
    (sci.crypt)
  • Re: A quote from Crypto-Gram
    ... > Does anybody with experience really think that they could modify microcode ... > cryptogram article referred to) to manipulate a RNG or backdoor (as is ... It may be hard to manipulate a RNG directly. ...
    (sci.crypt)
  • Re: A quote from Crypto-Gram
    ... >> It's still unclear to me just HOW one could manipulate the microcode ... It seems that the role and scope of microcode is ... means - you could for example reprogram an ADD instruction to be a SUB ... You might remember the i386 chip, ...
    (sci.crypt)
  • Re: A quote from Crypto-Gram
    ... On 2004-08-17, Juergen Nieveler wrote: ... > It's still unclear to me just HOW one could manipulate the microcode to ...
    (sci.crypt)
  • Re: A quote from Crypto-Gram
    ... Juergen Nieveler wrote: ... that manipulation of microcode is ... > It's still unclear to me just HOW one could manipulate the microcode to ... > microcode in the processor affects the commands a processor knows - it ...
    (sci.crypt)
Quantcast