[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20130922042458.GA25362@openwall.com>
Date: Sun, 22 Sep 2013 08:24:58 +0400
From: Solar Designer <solar@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: ZedBoard / Parallella: bcrypt
Katja,
On Sat, Sep 21, 2013 at 06:40:34PM +0200, Katja Malvoni wrote:
> On Sat, Sep 21, 2013 at 5:22 PM, Katja Malvoni <kmalvoni@...il.com> wrote:
>
> > I don't know yet, I'm trying 125 MHz at the moment.
>
> It also fails timing analysis but both bcrypt format work. Performance is
> 50.5 c/s which doesn't make sense (expected is 45 c/s + communication
> overhead).
Curious.
What's the highest clock rate that ISE reports as meeting timing
constraints for your bcrypt core on this Zynq device?
Regarding observed speeds not matching expectations for a given clock
rate, perhaps you can use higher bcrypt cost settings to reduce the
effect the overhead has, and then you'll be able to calculate the actual
clock rate from the observed c/s rate. Maybe this will give a clue.
For example, $2a$10 should be taking almost a second, and you can go
even higher than that for greater precision.
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.
