News for hackers. Stuff that matters.
Re:Multitun – Tunnel arbitrary traffic through an innocuous ... by ryan-c on Thursday November 20, 2014 @10:30PM (#8639671)
The use of SHA-2-224 is kinda weird.

Based on a quick skim:

* no kdf on the password, just SHA-2-224(password)

* uses the same key and iv (which always starts as 0) in both directions

* uses CFB mode without a MAC, which has malleability issues (IIRC you can flip bits in a ciphertext block, and the same bits will be flipped in the plaintext, but the rest of the stream will be garbage).

* the authentication used is pretty sketchy (the key is fixed at 16 bytes and PW_LEN 15, so ljust shouldn't be doing anything):

    if tmp_data[:len(self.key)].ljust(PW_LEN,'0') != self.key.ljust(PW_LEN, '0'):
        log.msg("Remote unauthorized", logLevel=logging.INFO)
Reminds me of this: https://www.cs.auckland.ac.nz/~pgut001/pubs/linux_vpn.txt
Re:Multitun – Tunnel arbitrary traffic through an innocuous ... by rapidally2 on Friday November 21, 2014 @01:14AM (#8640068)
The IV doesn't start as zero; it starts as a random number.
Re:Multitun – Tunnel arbitrary traffic through an innocuous ... by ryan-c on Friday November 21, 2014 @02:32AM (#8640218)
Oh, you are correct. It's chosen by the client here:

https://github.com/covertcodes/multitun/blob/master/multitun...