On Mon, 2009-02-16 at 13:49 -0500, Sean DeNigris wrote:
> Hi all!  The following code seems like it should be open to session  
> fixation attacks, but is not.  Why?!
> 
> This is the beginning of the private page...
> <?php
> session_start();
> if (!isset($_SESSION['user']))
> {
>       header("Location: http://[address of login page]?requestedpage=[token  
> for this page]");
>       exit();
> }
> ....
> 
> If an attacker caused a known user to request the above page with ? 
> PHPSESSID=1234, the session_start would then register 1234 as the  
> current session
> 
> This is from the login page...
> <?php
> if($_POST['[a posted form var]'])
> {
>       // check submitted credentials against known users
>       $status = authenticate(...);
>       // if  user/pass combination is correct
>       if ($status == 1)
>       {
>               // initiate a session
>               session_start();
>       
>               // register some session variables
>               $_SESSION['XXXXXX] = filter($_POST['XX']);
> 
>               // redirect to protected page
>               header("Location: ...[requested page]);
>               exit();
>       }
> }
> 
> When the user logged in above, the session_start would use the session  
> cookie from the first session_start above and have a validated session  
> with an SID known to the attacker.
> 
> However, the top snippet does not cause an SID to be recorded in a  
> cookie, but the bottom one does.  Hence, the attack is prevented, but  
> why?
> 
> Thanks, cheers!
> 
> - Sean
> 
Erm, is this a trick question or your homework?


Ash
www.ashleysheridan.co.uk


-- 
PHP General Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php

Reply via email to