Please note, this is a STATIC archive of website developer.mozilla.org from 03 Nov 2016, cach3.com does not collect or store any user information, there is no "phishing" involved.

Problems integrating with CRSF protection

Persona is no longer actively developed by Mozilla. Mozilla has committed to operational and security support of the persona.org services until November 30th, 2016.

On November 30th, 2016, Mozilla will shut down the persona.org services. Persona.org and related domains will be taken offline.

If you run a website that relies on Persona, you need to implement an alternative login solution for your users before this date.

For more information, see this guide to migrating your site away from Persona:

https://wiki.mozilla.org/Identity/Persona_Shutdown_Guidelines_for_Reliers

If you implement a particular type of protection against CSRF (Cross-Site Request Forgery) login attacks, you will experience a problem when a user has multiple pages open on your site, and then tries to log in using one of them. This document explains what the problem is, and how to resolve it.

CSRF and CSRF login attacks

In a normal CSRF attack, the user has already logged into the target website (for example, their bank). The user then loads a maliciously-constructed page on another site. This page sends an HTTP request to the target website: the request has some valuable payoff (for example, it transfers money to the attacker). The target website allows the request because it assumes that since the user is logged into the site, then the request was made by the user.

In a CSRF login attack, the user loads a maliciously-constructed page that logs the user into the target site as the attacker: the target site then sets a session cookie in the user's browser. The attacker is able to access the account later, and retrieve any information that the site collected about that user.

Protection against CSRF: CSRF tokens derived from session IDs

A common protection against normal CSRF attacks is for websites to derive a token from the browser's session ID, and embed it in the pages they serve: then POST requests have to include the token, which is checked on the server. This means POST request can't be made from other domains, because they can't get access to the token used to authenticate them.

With CSRF login attacks, of course, the user isn't logged into the site already, so the website sets up an initial session as soon as the user visits the page. This is used to generate the CSRF token until the user logs in. Once the user logs in, a new session ID is generated, and the CSRF regenerated accordingly.

The problem with Persona

The problem with Persona arises when the user is not yet logged in, and has loaded pages from your website in two separate tabs, A and B. The pages contain the same CSRF token, derived from the initial session ID. Then the user logs in from tab A, and the website generates a new session ID, and thus a new CSRF token. But tab B still has the old page loaded, containing the old, now invalid, token.

Under normal circumstances this would not usually matter: as soon tab B reloads the page, or the user navigates to a new page, then the embedded CSRF token is updated. But when Persona has generated an assertion in response to a call to navigator.id.request(), Persona calls onlogin for every tab which has that website loaded, not just the single tab that requested the assertion. As soon as tab B's onlogin handler is called, it attempts to POST the assertion to the server using the old CSRF token - and the server raises a CSRF error.

The solution

The solution is for the onlogin handler to send a GET request which fetches a fresh CSRF token and checks whether the user is already logged in. If the user is already logged in, then the handler just has to reload the page. Otherwise it uses the new CSRF token to POST the assertion to the server.

Document Tags and Contributors

Tags: 
 Contributors to this page: wbamberg, Sheppy
 Last updated by: wbamberg,