What does that really mean? Can you please give me a real life example?
Attack example 1: Cross-Site Request Forgery (CSRF) with an HTML form
On page at evil.com the attacker has put:
<form method="post" action="http://bank.com/trasfer">
<input type="hidden" name="to" value="ciro">
<input type="hidden" name="ammount" value="100000000">
<input type="submit" value="CLICK TO CLAIM YOUR PRIZE!!!">
</form>
Without further security measures:
- the request does get sent. The SOP does not forbid this request from being sent.
- it includes authentication cookies from
bank.com which log you in. As mentioned at "Are all cross-origin requests forbidden?" below, sending cookies across domains allows you for example to click a <a link to bank.com and go to that website already logged-in, which is what users generally want.
It is the synchronizer token pattern, alone, even without the SOP, prevents this from working.
Synchronizer token pattern
For every form on bank.com, the developers generate a one time random sequence as a hidden parameter, and only accept the request if the server gets the parameter.
E.g., Rails' HTML helpers automatically add an authenticity_token parameter to the HTML, so the legitimate form would look like:
<form action="http://bank.com/transfer" method="post">
<p><input type="hidden" name="authenticity_token"
value="j/DcoJ2VZvr7vdf8CHKsvjdlDbmiizaOb5B8DMALg6s=" ></p>
<p><input type="hidden" name="to" value="ciro"></p>
<p><input type="hidden" name="ammount" value="100000000"></p>
<p><button type="submit">Send 100000000$ to Ciro.</button></p>
</form>
as mentioned at: https://stackoverflow.com/questions/941594/understanding-the-rails-authenticity-token/26895980#26895980
So if evil.com makes a post single request, they would never guess that token, and the server would reject the transaction!
See also: synchronizer token pattern at OWASP.
Attack example 2: Cross-Site Request Forgery (CSRF) with JavaScript AJAX
But then, what prevents the evil.com from making 2 requests with JavaScript, just like a legitimate browser would do:
- XHR GET for the token
- XHR POST containing the good token
so evil.com would try something like this (jQuery because lazy):
$.ajax({
url: 'http://bank.com/transfer',
type: "GET",
xhrFields: {withCredentials: true},
});
// Parse HTML reply and extract token.
$.ajax({
url: 'http://bank.com/transfer',
type: "POST",
data: {
to: 'ciro',
ammount: '100000000',
authenticity_token: extracted_token
},
xhrFields: {withCredentials: true},
});
Here the attacker used withCredentials: true because XHR does not send cross request cookies without it, see also: Why are cookies sent with HTML page's cross domain requests but not with JS's XHR?
This is where the SOP comes into play. Although the GET and POST do actually send the authenticated request just like the HTML form, the sender's browser prevents the JavaScript code from reading the HTML reply back, because the request was sent to a separate domain!
The Chromium developer console shows an error for it of type:
Access to XMLHttpRequest at 'http://bank.com' from origin 'http://evil.com' has been blocked by CORS policy: No 'Access-Control-Allow-Origin' header is present on the requested resource.
which has been asked at: https://stackoverflow.com/questions/20035101/why-does-my-javascript-code-get-a-no-access-control-allow-origin-header-is-pr
Why not just not send cross request cookies instead?
What if implementations had a rule like: "allow any request, but only send cookies on current domain XHR"? Wouldn't that be simpler?
But that would still allow for another type of attack: when authentication is based not on cookies, but on source (IP) of the request!
For example, you are in your company's intranet and from there you can access an internal server, which is not visible from the outside and serves secret data.
Are all cross-origin requests forbidden?
Even forgetting CORS, no, we do them every day!
The only thing that can't ever be allowed is reading the result of such request back in JavaScript, as that would beat the synchronizer token pattern.
From MDN:
Cross-origin writes are typically allowed: links, redirects and form submissions.
[My comment]: e.g., when you click a link, you often expected to go logged in to the website, and that requires making an authenticated GET request that returns the new page.
Cross-origin embedding is typically allowed: images, external CSS and Javascript, iframes.
Cross-origin reads are typically not allowed: XHR (example above), iframe read.
However, read access is often leaked by embedding. For example you can read the width and height of an embedded image, the actions of an embedded script, or the availability of an embedded resource (and thus possibly if the user is logged in or not on a given domain)
Other prevention approaches
- check if certain headers is present e.g.
X-Requested-With:
- check the value of the
Origin header: Why is the synchronizer token pattern preferred over the origin header check to prevent CSRF
- re-authentication: ask user for password again. This should be done for every critical operation (bank login and money transfers, password changes in most websites), in case your site ever gets XSSed. The downside is that the user has to type the password multiple times, which is tiresome, and increases the chances of keylogging / shoulder surfing.
Other prevention approaches: JWT
JSON Web Token is quite a popular alternative to cookies + synchronizer token pattern circa 2020.
What this method does is:
- store a signed token in
window.localStorage
- whenever you want to make an authenticated request to the server, send a header
Authentication: <token>. Note that this can only be done from JavaScript.
This method works because unlike cookies, localStorage is only available when you make requests from the website itself (through JavaScript), thus dispensing the synchronizer token.
Then, when users first visit the website, they are initially logged off, and a dummy loading page shows.
Then the browser runs the JavaScript is just received from the server, reads localStorage (now that we are on the correct domain already) and sends an authenticated GET request to an API path to get only the data without HTML, usually as JSON.
Finally the JavaScript renders that data on the browser.
This approach has become particularly popular due to the popularity of Single Page Applications, where the simplest implementation approach is this two-step get dummy page then populate it with the API data.
So this basically carries the tradeoffs:
- advantages:
- simpler to implement since no synchronizer on every form
- the usual SPA advantages: you get only data after the initial request, not HTML tags
- disadvantages:
- the usual SPA disadvantages:
- during first load the user might see annoying loading dummy page elements
- the website is not visible without JavaScript
See also
