* Re: please kindly get back to me
@ 2002-06-04 4:26 Hank Leininger
2002-06-04 18:36 ` [OT] " J Sloan
0 siblings, 1 reply; 3+ messages in thread
From: Hank Leininger @ 2002-06-04 4:26 UTC (permalink / raw)
To: linux-kernel
On 2002-06-03, J Sloan <joe@tmsusa.com> wrote:
> The thing with linux/unix "virii" is, they
> are actually for the most part trojans -
> they've been in labs for years, the problem
> is that there is no suitable transport vector!
> You'd have to dupe an unwitting superuser
> (now there's a dangerous combination) into
> running the "virus" by hand - sort of like
> the "honor system" virus....
...You mean like, get them to run './configure' ?[1][2]
...Or installing an RPM with trojanned binaries or install-time scripts,
without checking a signature?[3]
Unfortunately that's all too easy. Viruses, no. Malware, you bet. We
can't get too complacent while laughing at the virus phenomenon.
[1] http://marc.theaimsgroup.com/?l=bugtraq&m=102233939226053&w=2
http://marc.theaimsgroup.com/?l=bugtraq&m=102285523803434&w=2
[2] They don't have to do this as root, either. If they do it from an
account that can escalate privileges (i.e. is allowed to su or sudo)
then it's game over anyway, albeit with more steps.
[3] And of course signatures are useless if the signer was owned first.
Probably major distros are reasonably safe[4], but not Joe Random who
produces packages and distributes them...
[4] They're not out to get you; they've already got you:
http://www.acm.org/classics/sep95/
--
Hank Leininger <hlein@progressive-comp.com>
ALL YOUR BASE ARE BELONG TO KEN THOMPSON
^ permalink raw reply [flat|nested] 3+ messages in thread* [OT] Re: please kindly get back to me
2002-06-04 4:26 please kindly get back to me Hank Leininger
@ 2002-06-04 18:36 ` J Sloan
2002-06-04 20:24 ` Alan Cox
0 siblings, 1 reply; 3+ messages in thread
From: J Sloan @ 2002-06-04 18:36 UTC (permalink / raw)
To: Hank Leininger; +Cc: linux kernel
Hank Leininger wrote:
>On 2002-06-03, J Sloan <joe@tmsusa.com> wrote:
>
>
>
>>The thing with linux/unix "virii" is, they
>>are actually for the most part trojans -
>>they've been in labs for years, the problem
>>is that there is no suitable transport vector!
>>
>>
>
>
>
>>You'd have to dupe an unwitting superuser
>>(now there's a dangerous combination) into
>>running the "virus" by hand - sort of like
>>the "honor system" virus....
>>
>>
>
>...You mean like, get them to run './configure' ?[1][2]
>...Or installing an RPM with trojanned binaries or install-time scripts,
>without checking a signature?[3]
>
>Unfortunately that's all too easy. Viruses, no. Malware, you bet. We
>can't get too complacent while laughing at the virus phenomenon.
>
>
Complacency is never a good idea - however,
let's give credit where credit is due - it's orders
of magnitude more difficult to do something like
this against a unix system - most script kiddies
will go for the easy targets (microsoft) instead
Joe
>
>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [OT] Re: please kindly get back to me
2002-06-04 18:36 ` [OT] " J Sloan
@ 2002-06-04 20:24 ` Alan Cox
0 siblings, 0 replies; 3+ messages in thread
From: Alan Cox @ 2002-06-04 20:24 UTC (permalink / raw)
To: J Sloan; +Cc: Hank Leininger, linux kernel
On Tue, 2002-06-04 at 19:36, J Sloan wrote:
> Complacency is never a good idea - however,
> let's give credit where credit is due - it's orders
> of magnitude more difficult to do something like
> this against a unix system - most script kiddies
> will go for the easy targets (microsoft) instead
Each of the major viruses has probably got one author singular.
There are ways of making systems much more resistant to attack including
viruses. Things like RSBAC and the NSA security modules help you get
into a situation where this kind of stuff doesn't occur
User1 gets a virus
User1 owns a binary root users
Root gets the virus
Splat
Because with a trust model it goes instead
User1 geta a virus
User1 owns a binary root users
User1 virus patches the binary
Root is refused permission to run the binary because it no longer has
a high enough integrity
Even before that the lack of people checking GPG keys on RPM and other
packages is disturbing.
Alan
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2002-06-04 19:19 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-06-04 4:26 please kindly get back to me Hank Leininger
2002-06-04 18:36 ` [OT] " J Sloan
2002-06-04 20:24 ` Alan Cox
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox