Re: the cold-boot attack - a paper tiger?

On 29.05.2008 04:41, Phil wrote:
> 
> --- Phil <[email protected]> wrote:
> 
> > 
> > --- Jacob Appelbaum <[email protected]> wrote:
> > > 
> > > Because loop-aes *is* vulnerable to our attacks.
> > > 
> > > The keying material is in memory when we mount our
> > > attack. We were able
> > > to reliably extract keys required to decrypt the
> > > data on the disk.
> > > 
> > 
> > So I am right in saying that quitting X and
> > overwriting  free memory as root with a utility such
> > as smem after pulling down the loop will prevent key
> > recovery?
> > 
> PS:  If so, why doesn't Jari just overwrite the slab
> of memory containing the keys when pulling down the
> loop? (I previously assumed loop-aes did this).

You should read the e-mail Jari wrote.
loop-AES does kill the key-material.

But you forgot the whole point about the attack:
The attacker don't "soft-boot" the computer, he presses the reset-key 
where the currently running OS (and therefore loop-AES) doesn't get the 
change to kill the key-material!

And the attack also implies that YOU, personally, weren't able to 
interfere. 

When you are able to get the computer to soft-boot or switch-off 
reguarly, loop-AES gets the chance to kill the key-material.

Modern computers and i guess most modern Distributions intercept the 
Power-Off-Button via ACPI and instead of "just switch-off power" they 
initiate a regular shutdown and soft-power-off afterwards. At least 
that's what my Debian-SID does by default when the acpid is running.

So when someone storms into my room and i am able to press the 
power-off-button i'm on the safe-side as long as the person doesn't 
press the reset-key or yanks out the power-cord before loop-AES had the 
chance to kill the key-material.





Bis denn

-- 
Real Programmers consider "what you see is what you get" to be just as 
bad a concept in Text Editors as it is in women. No, the Real Programmer
wants a "you asked for it, you got it" text editor -- complicated, 
cryptic, powerful, unforgiving, dangerous.


-
Linux-crypto:  cryptography in and on the Linux system
Archive:       http://mail.nl.linux.org/linux-crypto/