Comments (4)
from nosurf.
I'm not using a copy of the server's key and that is the problem. The CSRF tokens are not encrypted or validated in anyway because there is no server key in use.
Looking at this more, I don't think this code works at all.
Take a look at maskToken() and unmaskToken() and oneTimePad():
https://github.com/justinas/nosurf/blob/master/crypto.go#L8-L54
maskToken() and unmaskToken() are called without arguments yet generate an internal key. That key is passed to oneTimePad() which claims it modifies the data in place. Yet it doesn't take a pointer...
Notice how the tests test masking but don't actually compare against a pre-calculated masked value?
https://github.com/justinas/nosurf/blob/master/crypto_test.go#L54-L83
Really it should take a key on init like gorilla csrf.
from nosurf.
Hi @jolan,
You are not entirely incorrect, but your bypass of the token check hinges on one very important aspect: in your example, you are able to set the cookie on the request. However, in a natural scenario (an HTML form on attackers page), the attacker is able to manipulate POST parameters, but not able to set a cookie value with your host. The check of the token received in the form against the one stored in the cookie does indeed happen:
if !verifyToken(realToken, sentToken) {
ctxSetReason(r, ErrBadToken)
h.handleFailure(w, r)
return
}
I see no way of preventing the client from setting the cookie to whatever they like without introducing a stateful storage mechanism to nosurf. But again, this can not happen in a CSRF attack.
Now token authenticity is a feature that has been suggested before, but it would somewhat change nosurf's API and there are already other libraries that provide this, so I never saw a point to make another clone. Nosurf remains a less full-fledged solution, for better or worse.
That key is passed to oneTimePad() which claims it modifies the data in place. Yet it doesn't take a pointer...
Go slices are reference types, so in this case and in fact most cases a pointer is not needed: the function does not receive a copy of the slice's data but a slice header consisting of a pointer to the same underlying array as the caller and integers indicating the length and capacity of the slice. See this Go blog article for more info on slice internals.
Notice how the tests test masking but don't actually compare against a pre-calculated masked value?
oneTimePad()
is tested separately and has a test that actually compares its result to a precomputed value. See this test
I hope I answered your worries. I will close the issue as I think this behavior of nosurf poses no threat in a CSRF scenario, but feel free to re-open it if you still think it is a security problem in any way.
from nosurf.
Yep sorry I realized my mistake a few minutes after hitting the Comment button. I had a debug Printf in the wrong place which made me misinterpret. 😊
from nosurf.
Related Issues (20)
- Filtering out safe methods and excluded paths HOT 2
- Token value error HOT 6
- Send a response body in defaultFailureHandler HOT 2
- please consider updating tags or deleting the current tag HOT 3
- Remove Referer check HOT 1
- Possible flaw HOT 2
- SetBaseCookie not having effect HOT 8
- Wiki page for newbies doubts and problems
- Doubts about many cookies and many Path for a single domain. HOT 13
- Prevent form resubmit HOT 1
- Logging HOT 1
- Combining Session and CSRF cookie HOT 3
- Is this normal behavior? HOT 5
- RegenerateToken generates two CSRF cookies when no previous CSRF cookie was set
- How does nosurf OTP protect against BREACH? HOT 5
- Inappropriate key in call to `context.WithValue`
- Ability to handle multiple cookies in context HOT 5
- Installing command of `nosurf` is missing in README.md
- CSRF failed with bad request HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from nosurf.