Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sun, 18 Dec 2011 19:59:50 +0400
From: Solar Designer <solar@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: 1.7.9-jumbo

On Sun, Dec 18, 2011 at 06:48:33PM +0400, Solar Designer wrote:
> Only in file 1: PIX MD5:Raw
...
> Only in file 2: PIX MD5:Many salts
> Only in file 2: PIX MD5:Only one salt

0010-j8-pix-md5-thin.diff changed BENCHMARK_LENGTH from -1 to 0.
I think this is a bug since SALT_SIZE is 0.  We should revert this
format's BENCHMARK_LENGTH to -1.

In fact, maybe bench.c should treat benchmark_length of 0 the same as -1
in case salt_size is 0.

1.7.8-jumbo-8:

Benchmarking: PIX MD5 [pix-md5]... DONE
Raw:    3785K c/s real, 3785K c/s virtual

1.7.9-jumbo-5:

Benchmarking: PIX MD5 [SSE2i 10x4x3]... DONE
Many salts:     11395K c/s real, 11509K c/s virtual
Only one salt:  11503K c/s real, 11389K c/s virtual

That's nice speedup.  Too bad relbench does not see this because of the
BENCHMARK_LENGTH bug.

Alexander

Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.