Re: RFC: Flaw in BitLocker, Apple's FileVault, TrueCrypt, and dm-crypt David H. Lipman wrote:
> In a paper published on the Internet, researchers show that data is
> vulnerable because encryption keys and passwords linger in the temporary
> memory of computers after machines lose power.
I found a really bigger vulnerability: The keys are in memory while the
computer is still powered on. One could simply connect some hardware to the
memory bugs and read it out directly...
Or could could attach a key logger and wait until the user enters the
password...
> "We then wrote programs to collect the contents of memory after the
> computers were rebooted."
Only applies to hardware reboots. If the computer is properly shut down, the
software simply zeros out the key in memory.
> Laptops are especially vulnerable to the attack when the machines are in
> lock, sleep, or hibernation modes, according to the report.
Hibernate? The hibernate file is stored on the encrypted disc...
> "This isn't a minor flaw; it is a fundamental limitation in the way these
> systems were designed."
No, it's a well known intangible limit known since at least 40 years:
Software cannot defend against an attacker which has physical access to the
system. |