Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Wed, 02 Oct 2013 19:22:04 +0200
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: HMAC MD5/SHA1 optimizations and OMP (was: rakp format)

On 2013-10-02 02:07, magnum wrote:
> On 2013-10-01 23:40, magnum wrote:
>> However, now that I look at it we could obviously do the two initial
>> ipad/opad SHA1's in set_key() too (or leave them in crypt_all() but
>> avoid doing them again if keys did not change). Not sure how we could
>> miss that for all these years (in the HMAC-formats). This too would only
>> gain the "many salts" case but it'd be a good boost. I'll have a look at
>> it.
>
> Done. 65-90% boost. OMP scales better on intel than AMD, not sure why.
> Maybe it can be tweaked.

Did the same to our old HMAC formats, and added OMP support:

MD5 Before:
Benchmarking: HMAC-MD5 [MD5 128/128 AVX 12x]... DONE
Many salts:	10506K c/s real, 10506K c/s virtual
Only one salt:	6261K c/s real, 6261K c/s virtual

Now:
Benchmarking: HMAC-MD5 [MD5 128/128 AVX 12x]... DONE
Many salts:	20198K c/s real, 20198K c/s virtual
Only one salt:	6210K c/s real, 6210K c/s virtual

Benchmarking: HMAC-MD5 [MD5 128/128 AVX 12x]... (4xOMP) DONE
Many salts:	67043K c/s real, 19100K c/s virtual
Only one salt:	8282K c/s real, 5050K c/s virtual

SHA1 Before:
Benchmarking: HMAC-SHA1 [SHA1 128/128 AVX 4x]... DONE
Many salts:	5585K c/s real, 5585K c/s virtual
Only one salt:	3615K c/s real, 3615K c/s virtual

Now:
Benchmarking: HMAC-SHA1 [SHA1 128/128 AVX 4x]... DONE
Many salts:	10532K c/s real, 10637K c/s virtual
Only one salt:	3574K c/s real, 3574K c/s virtual

Benchmarking: HMAC-SHA1 [SHA1 128/128 AVX 4x]... (4xOMP) DONE
Many salts:	34070K c/s real, 9933K c/s virtual
Only one salt:	6758K c/s real, 3520K c/s virtual

magnum

Powered by blists - more mailing lists

Your e-mail address:

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