phc-discussions - Re: [PHC] Lyra2 initial review

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: <20140417190515.GC10725@openwall.com>
Date: Thu, 17 Apr 2014 23:05:15 +0400
From: Solar Designer <solar@...nwall.com>
To: discussions@...sword-hashing.net
Subject: Re: [PHC] Lyra2 initial review

On Thu, Apr 17, 2014 at 02:40:15PM -0400, Bill Cox wrote:
> In theory I meant to simply divide the memory (2GiB) by the total runtime
> (including memory allocation).  D'oh!  Somehow I'm unable to use a
> calculator properly...  your numbers below are correct, though on my
> development machine, I do not trust the split between users and sys time.
>  I see it vary randomly all over the place, while the total runtime remains
> fairly constant.  That's why I don't try to factor out the memory
> allocation overhead.

Yes, the user vs. system time doesn't reliably correspond to hashing vs.
memory allocation overhead.  Not on my systems, too.  To determine the
overhead-less speeds, we need to measure time within the program or/and
iterate the hash computation many times while keeping the allocation.

Alexander