phc-discussions - Re: [PHC] omegacrypt and timing

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]

Date: Fri, 19 Sep 2014 23:06:13 +0200
From: Markus Duermuth <markus.duermuth@....de>
To: discussions@...sword-hashing.net
Subject: Re: [PHC] omegacrypt and timing

On 17.09.2014 12:30, Bill Cox wrote:
> On 09/17/2014 05:53 AM, Krisztián Pintér wrote:
>> i picked omegacrypt because i just spotted this. out of the 3, 
>> antcrypt seems to be the worst, the other two can get away with 
>> some blinding (dummy operations), but you can't compensate for 
>> differences in floating point, as the timing is just unknown and 
>> depends on zillions of factors, including the data itself.

Honestly, I'm not worried too much about the resulting timing leak.  It
is correct that the single branches can have vastly different timing,
but when you observe the overall timing for one password what you will
see is the sum of a very large number of (pseudo-)random timings.  One
should check mean and variance of the resulting runtime, but I would be
_very_ surprised if this carries enough useful information.

> Of the four, AntCrypt does the best job of making it's case for using
> data-based branching for SIMD defense.  I think OmegaCrypt, POMELO,

Thank you :-)

> POMELO, I'd drop an calls to existing cryptographic hashes, and depend

I don't see a good reason to totally stop using hash functions.  Calling
a hash function at least once (in the correct way, of course), and
refraining from doing something really bad outside the hash function,
guarantees one-wayness of the construction and you have one less thing
to worry about.  Apart from that you can use them to easily implement
nice features such as strength-extensibility (by collapsing and then
expanding the state again), like AntCrypt is doing.  But I agree that
you should/can use them sparingly.

Cheers,
Markus