[cryptography] [ramble] [tldr] Layered security where encryption is used?
natanael.l at gmail.com
Sun Jul 21 17:28:24 EDT 2013
Big fat disclaimer: I'm also not a crypto expert. Don't trust this without
verification from somebody who is.
To add to what CodesInChaos said, just in case you can't use different keys
for different types of data then just before encrypting it you could add a
HMAC of the plaintext with a tag (should be pregenerated random strings)
that specifies the permissions as the HMAC key. So when the server decrypts
it, it checks if it's HMAC'ed with the tag for "top secret" or for "can be
shared with the user". If you don't also add the "permissions tag" in
plaintext form, you'll get a small performance hit here if there's a lot of
different tags to test, and possibly it would also introduce an "oracle
attack" that reveals what type of tag it has through timing. This also
helps you against that known-plaintext attack against the first cipher
block which you said you fear (although simply using a proper IV works
equally well for that part).
And that would be in ADDITION to doing HMAC on the ciphertext or using an
authenticated encryption mode.
> I can definitely see an argument that this is a bunch of over-engineering
It's not. You should never even leave wiggle-room for attackers. At least
for as long as it doesn't cost you more than a successful attack would.
Also, no you can't just simplify crypto guidelines that way. Just can't be
done. There's too many potential sidechannels and flaws and misconceptions
that could screw it up. If it could be simplified that way then it would
have been already, and you would already have heard of it.
2013/7/21 CodesInChaos <codesinchaos at gmail.com>
> 1) If you want to prevent tampering, use a MAC, not a cipher. My
> recommendation is HMAC-SHA-2. Be sure to use a constant time equality check
> while verifying the MAC.
> 2) If you want to encrypt something symmetrically, use authenticated
> encryption. Either with a specialized mode, like AES-GCM or with an
> encrypt-then-mac scheme. Use a proper IV and don't forget to include it in
> the MAC.
> 3) Use separate keys for different uses. This avoids interactions between
> different parts of the software.
> If you want only a single key in the config, then don't use it
> directly. Instead derive a distinct key for each usage with a key
> derivation function.
> My recommendation for a KDF is HKDF with HMAC-SHA-2 as building block.
> cryptography mailing list
> cryptography at randombit.net
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the cryptography