[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <40da77d9bfd21372e12c8490e2ccdaf3@smtp.hushmail.com>
Date: Sat, 11 Aug 2012 14:22:40 +0200
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: bf-opencl fails self-test on CPU
On 2012-08-11 09:42, Sayantan Datta wrote:
> On Sat, Aug 11, 2012 at 8:17 AM, Solar Designer <solar@...nwall.com> wrote:
>> Any idea why bf-opencl fails self-test on CPU (with AMD's SDK)? Will it
>> succeed with some other settings in opencl_bf_std.h maybe?
>
> I guess it is due to the lack of LDS on CPU. I'm not sure though but I'll
> find out. Also is it necessary to run the bf-opencl on CPU? We might need a
> little modified kernel for that.
Some people have 96 cores or more, and OpenCL might scale better than
OMP (not sure about this case though). I would like all OpenCL formats
to at least run on CPU, and ideally they should even be somewhat
optimized for CPU - at least in terms of picking work group sizes. The
RAR format tries to do this.
You can use eg. #ifdef DEVICE_IS_CPU in kernel, and some of the plenty
of helper functions from common-opencl.c in host code. And you could
pick LWS/GWS depending on device (and/or number of cores) - or at least
get them from john.conf.
magnum
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.
