Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Sun, 20 Sep 2015 07:58:58 +0300
From: Solar Designer <solar@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: current large hash file speeds and bottlenecks (was: ldr_split_line() performance regression)

On Sun, Sep 20, 2015 at 07:46:14AM +0300, Solar Designer wrote:
> Here's further speedup for --show:
> 
> real    0m12.012s
> user    0m11.607s
> sys     0m0.449s

And even further:

real    0m11.524s
user    0m11.169s
sys     0m0.450s

Change this line:

		hash <<= 3; extra <<= 2;

to:

		hash <<= 1; extra <<= 1;

and this line:

			extra ^= extra >> CRACKED_HASH_LOG;

to:

			extra ^= extra >> (CRACKED_HASH_LOG - 1);

Sorry I was too quick to post the previous patch, only to realize that
typical hash encodings are long enough that we don't need to shift
quicker to achieve the full hash size.

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.