l4-hurd
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Broken dream of mine :(


From: Sam Mason
Subject: Re: Broken dream of mine :(
Date: Tue, 22 Sep 2009 17:39:29 +0100
User-agent: Mutt/1.5.13 (2006-08-11)

On Tue, Sep 22, 2009 at 05:19:36PM +0200, Michal Suchanek wrote:
> 2009/9/22 Sam Mason <address@hidden>:
> > If somebody breaks in and installs some malicious code then I want it to
> > break in the most obvious way possible.  The admin then reinstalls the
> > system and only when everything has been brought back to normal will the
> > system will be allowed back into the network.
> 
> How do you actually check for malicious code?

The obvious way is just to say that if any part of the TCB is unknown
then it's treated as malicious.

> The drm scheme only needs to protect a particular key store and
> integrity of a single application - the media player. This may be
> feasible even on Linux.

I'm not sure if I care about DRM or key stores at all.  All I think I
want TPM for is to verify that my TCB is one that's considered good.

> On the other hand, making sure that none of your documents are
> randomly sent over the network or overwritten  is hard, you can do
> that with a shell script or similar on most systems.
> 
> If they were drm protected media files there is no harm to the
> provider of the drm content, they can be still accesses only with the
> right keys and the right system and player.

I'm not interested in any of that.  The case I'm thinking of is a
network of computers using potentially sensitive data/code and I want to
make sure they're only running the "correct" code and neither code or
data are going fall under the control of an attacker.

> Different goals often require different tools.

Indeed they do, and I think that TPM should be able to increase my
confidence in this.  I'm not sure how much this will actually help
though.

> Even considering these possible variables there is still much less to
> check than with a TPM chip. You can also obtain information on the
> construction of the flash chip so you should be aware of possible
> pitfalls in advance.
> 
> It is also more feasible to get a custom BIOS than it is to get a
> custom TPM chip.

Yup, I'm starting to get the feeling you're right.  TPM is a fun bit of
research, but it's so far removed from any practical application that I'm
struggling to justify it.

-- 
  Sam  http://samason.me.uk/




reply via email to

[Prev in Thread] Current Thread [Next in Thread]