CTR and disk sector encryption
Available news archives: comp.lang.tcl - comp.lang.python - comp.security.firewalls - sci.crypt - comp.lang.php - comp.lang.javascript
Google
 
Web news.hping.org


sci.crypt archive

CTR and disk sector encryption

From: <axellec@netcourrier.com>
Date: Thu Apr 06 2006 - 13:49:25 CEST

Hi all,

I am trying to make up my mind whether CTR can or cannot be used for
disk sector encryption, and what are the conditions.
To my understanding, people tend to disapprove with use of CTR for disk
encryption at sector level, because of the same counter must never be
reused. So, for example, the following is bad:
sector N: counter1 = N, counter2 = N+1, counter3=N+2
sector N+1: counter1 = N+1, counter2=N+2, counter3=N+3
because the value N+1 and N+2 are reused.

However, if I am on a system where I can ensure that counters are not
reused, is CTR safe for disk sector encryption, or are there attacks I
am not aware of ?

(example: if I can guarantee that:
sector N: counter1 = N, counter2 = N+1, counter3=N+2
sector N+1: counter1 = N+3, counter2=N+4, counter3=N+5)

Regards,
Axelle.
Received on Mon May 1 01:55:47 2006