Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sun, 15 Mar 2015 05:23:43 +0300
From: Solar Designer <solar@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: AVX2

On Sun, Mar 15, 2015 at 04:20:25AM +0300, Solar Designer wrote:
> On Sun, Mar 15, 2015 at 04:09:32AM +0300, Solar Designer wrote:
> > On our i7-4770K, before:
> > 
> > solar@...l:~/j/john-1.8.0.4.orig/src$ ../run/john -te -form=descrypt
> > Will run 8 OpenMP threads
> > Benchmarking: descrypt, traditional crypt(3) [DES 128/128 AVX-16]... DONE
> > Many salts:     25087K c/s real, 3139K c/s virtual
> > Only one salt:  20375K c/s real, 2550K c/s virtual
> > 
> > After:
> > 
> > solar@...l:~/j/john-1.8.0.4-avx2/src$ ../run/john -te -form=descrypt
> > Will run 8 OpenMP threads
> > Benchmarking: descrypt, traditional crypt(3) [DES 256/256 AVX2-16]... DONE
> > Many salts:     43489K c/s real, 5443K c/s virtual
> > Only one salt:  30002K c/s real, 3755K c/s virtual
> 
> and even:
> 
> Benchmarking: descrypt, traditional crypt(3) [DES 256/256 AVX2-16]... DONE
> Many salts:     44820K c/s real, 5609K c/s virtual
> Only one salt:  32001K c/s real, 4006K c/s virtual
> 
> with DES_bs_cpt reduced from 32 to 16.

and slightly higher single-salt speed:

Benchmarking: descrypt, traditional crypt(3) [DES 256/256 AVX2-16]... DONE
Many salts:     44820K c/s real, 5609K c/s virtual
Only one salt:  32420K c/s real, 4056K c/s virtual

with:

#define vshl1(dst, src) \
	(dst) = _mm256_add_epi8((src), (src))

(Obviously, the cumulative speed for --fork=8 should be higher yet.)

I've also tried various mixed builds: 256+256, 256+128, 256+64 MMX,
256+64 scalar, 256+64 MMX+64 scalar.  All of them are slower.  (I had
previously tried them for 256-bit AVX.  Now re-tried them for AVX2.)

Alexander

Powered by blists - more mailing lists

Your e-mail address:

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