Re: [Full-disclosure] Web 2.0 backdoors made easy with MSIE XMLHttpRequest

2007-02-05 Thread Troy Cregger
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

 The 2005 text does briefly mention Accessing content / web-scanning 
 (take a look at Notes 1-3).
 
 So the problem is much older.

Well, that's Micro$loth for ya.

Amit Klein wrote:
 Michal Zalewski wrote:
 On Sat, 3 Feb 2007, Michal Zalewski wrote:

   
   xmlhttp.open(GET\thttp://dione.ids.pl/\tHTTP/1.0\n\n;, x,true);
 
 Funny enough, Paul Szabo was quick to point out that Amit Klein found the
 same vector that I used here for client-side backdoors in May 2006 (still
 not patched?! *shrieks in horror*), but for cache poisoning:

   IE + some popular forward proxy servers = XSS, defacement (browser cache 
 poisoning)
   http://www.securityfocus.com/archive/1/434931

 This is getting depressing. May 2006.

   
 Much worse. The basic technique already appeared in two of my previous 
 write-ups:
 
 Exploiting the XmlHttpRequest object in IE - Referrer spoofing, and a 
 lot more... (September 2005)
 http://www.securityfocus.com/archive/1/411585
 
 XS(T) attack variants which can, in some cases, eliminate the need for 
 TRACE /archive/107/308433/30/0/threaded (January 2003)
 http://www.securityfocus.com/archive/107/308433
 
 The 2005 text does briefly mention Accessing content / web-scanning 
 (take a look at Notes 1-3).
 
 So the problem is much older.
 
 Thanks,
 -Amit
 
 ___
 Full-Disclosure - We believe in it.
 Charter: http://lists.grok.org.uk/full-disclosure-charter.html
 Hosted and sponsored by Secunia - http://secunia.com/
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFFx0j9nBEWLrrYRl8RAuzDAJ9j7ucnNaYlD7qXuFOsvwirZ4+WGwCfaupB
6WNArFbS1cB+5CRbYvkh/go=
=4eMy
-END PGP SIGNATURE-

___
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/


Re: [Full-disclosure] Web 2.0 backdoors made easy with MSIE XMLHttpRequest

2007-02-04 Thread Amit Klein
Michal Zalewski wrote:
 On Sat, 3 Feb 2007, Michal Zalewski wrote:

   
   xmlhttp.open(GET\thttp://dione.ids.pl/\tHTTP/1.0\n\n;, x,true);
 

 Funny enough, Paul Szabo was quick to point out that Amit Klein found the
 same vector that I used here for client-side backdoors in May 2006 (still
 not patched?! *shrieks in horror*), but for cache poisoning:

   IE + some popular forward proxy servers = XSS, defacement (browser cache 
 poisoning)
   http://www.securityfocus.com/archive/1/434931

 This is getting depressing. May 2006.

   
Much worse. The basic technique already appeared in two of my previous 
write-ups:

Exploiting the XmlHttpRequest object in IE - Referrer spoofing, and a 
lot more... (September 2005)
http://www.securityfocus.com/archive/1/411585

XS(T) attack variants which can, in some cases, eliminate the need for 
TRACE /archive/107/308433/30/0/threaded (January 2003)
http://www.securityfocus.com/archive/107/308433

The 2005 text does briefly mention Accessing content / web-scanning 
(take a look at Notes 1-3).

So the problem is much older.

Thanks,
-Amit

___
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/


Re: [Full-disclosure] Web 2.0 backdoors made easy with MSIE XMLHttpRequest

2007-02-03 Thread Tyop?
On 2/3/07, Michal Zalewski [EMAIL PROTECTED] wrote:
 On Sat, 3 Feb 2007, Michal Zalewski wrote:
   xmlhttp.open(GET\thttp://dione.ids.pl/\tHTTP/1.0\n\n;, x,true);
 Funny enough, Paul Szabo was quick to point out that Amit Klein found the
 same vector that I used here for client-side backdoors in May 2006 (still
 not patched?! *shrieks in horror*), but for cache poisoning:
   IE + some popular forward proxy servers = XSS, defacement
 (browser cache poisoning)
   http://www.securityfocus.com/archive/1/434931

 This is getting depressing. May 2006.

but not really surprising, yes?

Remember browserfun#18 (Tuesday, July 18, 2006)
http://osvdb.org/27110
Metasploit, exploit in the wild like they said.

Patched in October. 3 months of real insecurity.
(^o^)

troll
Thx to Determina.
http://www.determina.com/security_center/security_advisories/securityadvisory_0day_09282.asp
/troll

--
Tyop? [Fr]
http://altmylife.blogspot.com

___
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/


Re: [Full-disclosure] Web 2.0 backdoors made easy with MSIE XMLHttpRequest

2007-02-03 Thread Michal Zalewski
On Sun, 4 Feb 2007, Tyop? wrote:

 This is getting depressing. May 2006.
 but not really surprising, yes?

No, though this bug is truly remarkable in that a quick fix, I'm quite
certain, amounts to changing != ' ' to  ' ' in the code.

That's two characters, and no chance for a negative impact on any
legitimate application, simply no way.

Oh, and actually,did I say May? It gets even better!

If you look at that paper, Amit initially noticed that \n and \t are not
filtered in September 2005 (17 months ago), and described it as a referrer
spoofing bug (granted, not an earth-shattering discovery).

He then followed up in May 2006 demonstrating how this can be used to do
local cache poisoning, which is kinda more problematic.

It's February 2007, the attack can be obviously used to do a really nasty
interactive firewall bypass attack in corporate environments - so... ugh.

At least they managed to fix it in IE7's new native XMLHttpRequest code,
which I bet happened by accident.

/mz

___
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/


Re: [Full-disclosure] Web 2.0 backdoors made easy with MSIE XMLHttpRequest

2007-02-03 Thread James Matthews

Yes this is bad!

On 2/3/07, Michal Zalewski [EMAIL PROTECTED] wrote:


On Sat, 3 Feb 2007, Michal Zalewski wrote:

   xmlhttp.open(GET\thttp://dione.ids.pl/\tHTTP/1.0\n\n;, x,true);

Funny enough, Paul Szabo was quick to point out that Amit Klein found the
same vector that I used here for client-side backdoors in May 2006 (still
not patched?! *shrieks in horror*), but for cache poisoning:

  IE + some popular forward proxy servers = XSS, defacement (browser
cache poisoning)
  http://www.securityfocus.com/archive/1/434931

This is getting depressing. May 2006.

/mz


___
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/





--
http://www.goldwatches.com
http://www.wazoozle.com
___
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/