phc-discussions - Re: [PHC] hash encryption

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20160323172126.GC6565@openwall.com>
Date: Wed, 23 Mar 2016 20:21:26 +0300
From: Solar Designer <solar@...nwall.com>
To: discussions@...sword-hashing.net
Subject: Re: [PHC] hash encryption

On Wed, Mar 23, 2016 at 07:54:56AM +0000, Jean-Philippe Aumasson wrote:
> What are your requirements wrt key size,

128-bit or larger

> block size,

256-bit

> mode,

N/A since it's just one block.  Or we can say it's ECB.

> speed?

The faster, the better.  It should be reasonable, but doesn't have to be
fast.  What I posted should be fine performance-wise: it's 4 or 7
SHA-256's, whereas many more are computed to expand yescrypt's inputs
into B and to compress the final B into yescrypt's output, given a
typical r=8 or so.

> With all the crypto in yescrypt, we'll probably find a decent scheme based
> on existing code :-)

Sure, and thanks, but my choice of building upon SHA-256 is also
motivated by these considerations:

Being able to (continue to) say that yescrypt's basic cryptographic
security relies solely on NIST-approved crypto (and on well-studied
constructions around such crypto primitives), including for this
optional feature.  The other option would have been to bring in AES, but
I don't like that.

Allowing for yescrypt implementations to use third-party implementations
of SHA-256, such as provided by a library or by a programming language.
This rules out building upon SHA-256 internals as opposed to upon
SHA-256 as a whole.

Makes sense?

Alexander