Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sun, 18 Dec 2011 17:03:50 +0100
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: 1.7.9-jumbo

On 12/18/2011 04:19 PM, Solar Designer wrote:
> On Sun, Dec 18, 2011 at 06:48:33PM +0400, Solar Designer wrote:
>> Minimum:                        0.81602 real, 0.81602 virtual
> ...
>> Also, someone could want to identify the format that became 18% slower
>> and see if this is reproducible and if it can be avoided (in a future
>> version).
>
> It's CRC-32.  On 1.7.8-jumbo-8 we had:
>
> Benchmarking: CRC-32 [32/64]... DONE
> Many salts:     63225K c/s real, 63225K c/s virtual
> Only one salt:  28983K c/s real, 28696K c/s virtual
>
> 1.7.9-jumbo-5 gives only:
>
> Benchmarking: CRC-32 [32/64]... DONE
> Many salts:     51593K c/s real, 51593K c/s virtual
> Only one salt:  27557K c/s real, 27557K c/s virtual
>
> (same machine, same compiler, same make target, no load).

The change is that in 1.7.8, we always had 16K keys_per_crypt, and in 
1.7.9 we have 8K x number of threads (the latter is capped to 4, because 
it does not scale further). I did that on some AMD hardware using 
gcc-4.5 and I did not see this performance drop (on the contrary).

The fastest formats will always show these variations depending on 
hardware and exact compiler version but we could possibly change this to 
16K x number of threads and see if it helps.

magnum

Powered by blists - more mailing lists

Your e-mail address:

Powered by Openwall GNU/*/Linux - Powered by OpenVZ