All of lore.kernel.org
 help / color / mirror / Atom feed
From: "cto@itechfrontiers.com" <cto@itechfrontiers.com>
To: Ethan Heidrick <ethanheidrick@gmail.com>
Cc: Sanjai Narain <narain@research.telcordia.com>, selinux@tycho.nsa.gov
Subject: Re: SELinux and Stuxnet
Date: Tue, 22 Feb 2011 22:41:47 -0500	[thread overview]
Message-ID: <4D6481FB.9050105@itechfrontiers.com> (raw)
In-Reply-To: <AANLkTikKu4P866h+U3JeTMtjynxt_uOYEFyOn6V_Lgw+@mail.gmail.com>

 > ...  taking advantage of the
 > communication of processes in the tree?

to begin with virtual memory handling, LSM, network stack, signaling, 
polling, devices and etc

There are more easier techniques for compromise, such as adding kernel 
module, device driver and etc. (root kit) after getting elevated access 
to a daemon with enough privileges (system calls) or elevated access

or even easier one, in example a daemon with elevated (root) access can 
do almost any system/kernel call,  access to /proc , direct access to 
devices, memory (raw), sockets and etc.

as a target for compromise you can always rely on Daemons and users,

Mandatory Access Control can be used to avoid elevation of access by 
unprivileged users/daemons, such elevation of privileges can be easy in 
example with finding a window of vulnerability in a daemon running on a 
system, some daemons need to run under root and create sub processes 
with dropped privileges, even RPC calls or shared memory are  good 
points to start with

SELinux is not just about protecting kernel, it is about adding 
Mandatory Access Control /Role Based Access Control to Linux that by 
default has DAC

it is a matter of context based security, containing/sandboxing users, 
daemons and resources

However another purpose of SELinux is MLS (MultiLevel Security) 
-classified, unclassified and etc-

it is not just about protecting kernel tree, it is about not trusting 
users, daemons and processes in utilizing the resources, and by far 
restricting them to a predefined security context

in that view anything run on the system can be treated as a possible 
point of threat to some extent

also you cannot trust a user on a DAC based system, how do you enforce 
user not to in example chmod 777 his folder?

So as you may know anything on the system can be a concern for MAC and 
SELinux,


it actually depends on the goal and that particular system, it is not 
always necessary to gain root privilege when your target daemon is 
running on an unprivileged user (system wide)

Thus actually almost anything; even codes not from the Linux kernel as 
you can see

Best,

Patrick K.


On 2/22/2011 9:54 PM, Ethan Heidrick wrote:
> The policies that are written are concerned with architecture based
> penetration of an interconnecting tree (kernal), where SeLinux is
> concerned, What "blocks" of code would be advantageous for an attacker
> concerned with such penetration techniques such as key functions
> {algorithms} and  data compression encryption taking advantage of the
> communication of processes in the tree?
>
>
> On Tue, Feb 22, 2011 at 2:13 PM, cto@itechfrontiers.com
> <mailto:cto@itechfrontiers.com> <cto@itechfrontiers.com
> <mailto:cto@itechfrontiers.com>> wrote:
>
>     Ethan,
>
>     What are you talking about?
>
>
>     Patrick K.
>
>
>     On 2/22/2011 4:47 PM, Ethan Heidrick wrote:
>
>         IE: infrastructure is process based on detecting such side
>         channeling
>         attacks excuse the pun, but revising SeLinux security
>         authorization if
>         that is what you are suggesting would create an independent node of
>         programmable patches directed specific technique.
>
>         Where would an node discrimination in the coding be "hazardous"
>         for such
>         red team analysis for penetration?
>
>         On Tue, Feb 22, 2011 at 9:54 AM, cto@itechfrontiers.com
>         <mailto:cto@itechfrontiers.com>
>         <mailto:cto@itechfrontiers.com <mailto:cto@itechfrontiers.com>>
>         <cto@itechfrontiers.com <mailto:cto@itechfrontiers.com>
>
>         <mailto:cto@itechfrontiers.com <mailto:cto@itechfrontiers.com>>>
>         wrote:
>
>             Need to add it myself, that human being is also error-prone,
>
>             i.e. last message I meant "waives" and wrote "waves"
>
>             such errors happen even in development, in software and in
>         security
>
>
>
>             On 2/22/2011 12:43 PM, cto@itechfrontiers.com
>         <mailto:cto@itechfrontiers.com>
>         <mailto:cto@itechfrontiers.com <mailto:cto@itechfrontiers.com>>
>         wrote:
>               Sanjai,
>
>               Security is a complex business, I'm afraid that SELINUX is an
>             attempt to
>               simplify part of this job at least,
>
>               The more secure you want to make a system the more complex
>             naturally it
>               becomes,
>
>               however complexity is enemy of security by itself,
>
>               There is somewhat a dilemma, a paradox in here, I'm afraid it
>             cannot be
>               oversimplified as regular users would become security
>         experts or such
>               simplification waves the need for security specialists
>
>               Best,
>
>               Patrick K.
>
>
>
>                 On 2/22/2011 12:19 PM, Sanjai Narain wrote:
>
>                     Hi Patrick: Thanks for your note. I understand that
>         SELinux
>                     does not
>                     directly apply to Stuxnet since it targeted Windows.
>         However, my
>                     question was conceptually motivated: whether mandatory
>                     access control
>                     could have contained the impact of this worm, had it
>         been
>                     available. I
>                     had thought that the answer is yes but wanted to
>         find out
>                     from other
>                     experts. I believe you concur. Now, if only we could
>         make
>                     SELinux a lot
>                     easier to use..... this is where one of my interests
>         lie. --
>                     Sanjai
>
>
>                     On 2/22/2011 11:53 AM, cto@itechfrontiers.com
>         <mailto:cto@itechfrontiers.com>
>         <mailto:cto@itechfrontiers.com <mailto:cto@itechfrontiers.com>>
>         wrote:
>
>                         On 1/30/2011 7:39 PM, cto@itechfrontiers.com
>         <mailto:cto@itechfrontiers.com>
>         <mailto:cto@itechfrontiers.com <mailto:cto@itechfrontiers.com>>
>         wrote:
>
>                             Hello,
>
>                             Stuxnet is a Windows Worm, and SELinux is
>         Mandatory
>                             Access Control for
>                             Linux
>
>                             on Linux SELinux can reduce the impact of
>         such worms
>                             if targeting Linux
>                             boxes, but it is not a preemptive mechanism
>         for not
>                             having any kind of
>                             compromise due to any vulnerability,
>         Although if you
>                             protect your
>                             system
>                             and targeted processes you may have reach
>         the goal
>                             of containing the
>                             impact of possible compromises
>
>
>                             Best,
>
>                             Patrick K.
>
>                             On 1/30/2011 5:20 PM, Sanjai Narain wrote:
>
>                                 Has there been thinking on whether
>                                 SELinux-hardened machines can avoid
>                                 the spread of Stuxnet-like worms?
>         Thanks. --Sanjai
>
>
>
>                             --
>                             This message was distributed to subscribers
>         of the
>                             selinux mailing
>                             list.
>                             If you no longer wish to subscribe, send mail to
>         majordomo@tycho.nsa.gov <mailto:majordomo@tycho.nsa.gov>
>         <mailto:majordomo@tycho.nsa.gov <mailto:majordomo@tycho.nsa.gov>>
>
>                             with
>                             the words "unsubscribe selinux" without
>         quotes as
>                             the message.
>
>
>                         Sanjai,
>
>                         SELinux is Mandatory Access Control for Linux
>
>                         Stuxnet only compromises Windows, SCADA and PLC 7
>                         systems (Siemens
>                         systems)
>
>                         it is a worm, for a worm to compromise a system
>         you need
>                         to have
>                         certain vulnerabilities
>
>                         It cannot compromise Linux (the same way); as
>         that worm
>                         has been
>                         designed for particular purposes and taking
>         advantages
>                         of Windows
>                         vulnerabilities
>
>                         If you mean protecting a network using Linux
>         front ends
>                         or inline
>                         systems Like IPS systems that's another story
>         which is
>                         irrelevant to
>                         SELINUX actually (although an IPS system -Intrusion
>                         Prevention system-
>                         on Linux can take advantages of SELINUX)
>
>                         in brief , theoretically in case of a worm for
>         Linux, it
>                         could be
>                         contained if SELINUX is effectively used.
>
>                         in practice Stuxnet is for Windows
>
>                         Best,
>
>                         Patrick K.
>
>
>
>
>                 --
>                 This message was distributed to subscribers of the selinux
>                 mailing list.
>                 If you no longer wish to subscribe, send mail to
>         majordomo@tycho.nsa.gov <mailto:majordomo@tycho.nsa.gov>
>         <mailto:majordomo@tycho.nsa.gov <mailto:majordomo@tycho.nsa.gov>>
>
>                 with
>                 the words "unsubscribe selinux" without quotes as the
>         message.
>
>
>
>             --
>             This message was distributed to subscribers of the selinux
>         mailing list.
>             If you no longer wish to subscribe, send mail to
>         majordomo@tycho.nsa.gov <mailto:majordomo@tycho.nsa.gov>
>         <mailto:majordomo@tycho.nsa.gov
>         <mailto:majordomo@tycho.nsa.gov>> with
>
>             the words "unsubscribe selinux" without quotes as the message.
>
>
>
>


--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.

      reply	other threads:[~2011-02-23  3:41 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-01-26 18:16 SeLinux Policy design question Ger Lawlor (gelawlor)
2011-01-26 20:16 ` Dominick Grift
2011-01-26 20:35   ` Ethan Heidrick
2011-01-30 22:20 ` SELinux and Stuxnet Sanjai Narain
2011-01-31  0:39   ` cto
2011-02-22 16:53     ` cto
2011-02-22 17:19       ` Sanjai Narain
2011-02-22 17:43         ` cto
2011-02-22 17:54           ` cto
2011-02-22 21:47             ` Ethan Heidrick
2011-02-22 22:13               ` cto
2011-02-23  2:54                 ` Ethan Heidrick
2011-02-23  3:41                   ` cto [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=4D6481FB.9050105@itechfrontiers.com \
    --to=cto@itechfrontiers.com \
    --cc=ethanheidrick@gmail.com \
    --cc=narain@research.telcordia.com \
    --cc=selinux@tycho.nsa.gov \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.