Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Wed, 29 Jun 2011 21:21:52 +0400
From: Vasiliy Kulikov <>
To: Linus Torvalds <>
Cc: Andrew Morton <>,,
Subject: Re: [Security] CVE request: kernel: taskstats/procfs io infoleak
 (was: taskstats authorized_keys presence infoleak PoC)

On Wed, Jun 29, 2011 at 17:10 +0400, Vasiliy Kulikov wrote:
> On Wed, Jun 29, 2011 at 15:11 +0400, Vasiliy Kulikov wrote:
> > 2) as you say here:
> > 
> > 
> > If CONST is known and CONTROLLABLE is controlled by an attacker then he
> > may find C1 and C1+1 generating X kb - 1 and (X+1) kb traffic,
> (X+1) kb - 1 and (X+1) kb of course, they are rounded to X and X+1 kbs,
> respectively.

OK, what I've explored:

For the same ssh if try to log and send pubkey/password auth requests:

read = C1 + (C2 + X)*A + C3*B

    where 1 <= A+B <= 6, 0 < A, 0 <= B
    A - number of pubkey requests
    B - number of password requests
    C1, C2, C3 - system dependant constants

Trying all possible pairs (A,B) I get a set of rounded read_characters.
Comparing it with generated table of all possible lengthes and possible
inputs (A,B) I learn an interval of possible authorized_keys files
sizes.  For my system I can learn privkey length because for all
possible key len values (768, 1024, 2048) the intervals are different.

So, with rounded read_characters value it's possible to learn privkey

Not a password length, but already something.

Vasiliy Kulikov - bringing security into open computing environments

Powered by blists - more mailing lists

Your e-mail address:

Please check out the Open Source Software Security Wiki, which is counterpart to this mailing list.

Powered by Openwall GNU/*/Linux - Powered by OpenVZ