Vulnerability Development mailing list archives

Re: CR II - winME? confirmation? (Slightly OT)


From: "Jordan" <jordanf () home com>
Date: Thu, 9 Aug 2001 12:39:31 -0700

I've held off on posting about this for quite some time cause I thought it
would go away...but this topic seems to still be alive, so here goes.

Quoted from the wonderful analysis of the second version of Code Red by the
brilliant Eeye folks and Mr Levy.

"This worm, like the original Code Red worm, will only exploit Windows 2000
web servers because it overwrites EIP with a jmp that is only correct under
Windows 2000. Under NT4.0 etc... that offset is different so, the process
will simply crash instead of allowing the worm to infect the system and
spread."

Read this paragraph over and over and over again, and if you still have any
questions about whether it'll work on ME, NT4, OSX or any other operating
system, read it again. I'm amazed that no one's posted this yet...Oh yeah,
and obviously a web server of some sort must be running, the worm propagates
by GET requests, which'll have no effect on a server that doesn't process
them...hard to see how anyone could be confused on this front...

If we're talking strictly about exploiting an unchecked buffer in idq.dll,
that's one topic, and of course it's possible that it'll work on any system
with a vulnerable idq.dll, but if we're talking about a specific worm, say
CR2, then at least read the analysis before posting about it...

Jordan Frank
jordanf () home com


----- Original Message -----
From: "Ken Pfeil" <Ken () infosec101 org>
To: "Meritt James" <meritt_james () bah com>; "kam" <kam () aversion net>
Cc: "Amer Karim" <amerk () telus net>; "VULN-DEV List"
<VULN-DEV () securityfocus com>
Sent: Wednesday, August 08, 2001 10:32 AM
Subject: RE: CR II - winME? confirmation? (Slightly OT)


Nope. If IIS is not running, there is no delivery mechanism for the
overflow
to be delivered on. If the mappings are not present, the overflow cannot
take place to the vulnerable ISAPI .dll's, and if you are patched with
MS01-033 you do not have vulnerable .dll's.

Plain and simple:
If users can establish a web session under IIS, you have not applied the
patch, AND the mappings are present- you are vulnerable.

-----Original Message-----
From: Meritt James [mailto:meritt_james () bah com]
Sent: Wednesday, August 08, 2001 9:28 AM
To: kam
Cc: Amer Karim; VULN-DEV List
Subject: Re: CR II - winME? confirmation? (Slightly OT)


"running" or "installed"?  It is my understanding that the vulnerability
exists if the files and mapping are there no matter the process state of
the IIS server.  Is my understanding incorrect?

Jim

kam wrote:

Without IIS running, an attacker has no means of exploiting the
vulnerable
file. With no access to the file, the vulnerability does not exist. If
they're running IIS, then there is a hole which they can exploit. Even
though it comes installed by default on 2000, it's not a risk
until you turn
on your web services.

kam

----- Original Message -----
From: "Amer Karim" <amerk () telus net>
To: "VULN-DEV List" <VULN-DEV () SECURITYFOCUS COM>
Sent: Tuesday, August 07, 2001 10:03 AM
Subject: Re: CR II - winME? confirmation? (Slightly OT)

Hi All,

All the advisories about CR state that only IIS servers are
vulnerable.
However, it's my understanding that the unchecked buffer in
idq.dll was
the
source of that vulnerability.  If that's the case, then why have the
advisories not included Win2K systems (all flavours) since idq.dll
is
installed by default as part of the indexing service on all these
systems -
regardless of whether they are using the service or not?
Wouldn't that
make
ANY system with the indexing service on it just as vulnerable
as systems
with IIS? Am I overlooking something obvious here?

Regards,
Amer Karim
Nautilis Information Systems
e-mail: amerk () telus net, mamerk () hotmail com




--
James W. Meritt, CISSP, CISA
Booz, Allen & Hamilton
phone: (410) 684-6566




Current thread: