Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Thu, 8 Oct 2015 01:55:00 +0300
From: Solar Designer <>
Subject: Re: nVidia Maxwell support (especially descrypt)?

On Thu, Oct 08, 2015 at 12:33:44AM +0200, magnum wrote:
> On 2015-10-07 23:37, Solar Designer wrote:
> >On Wed, Oct 07, 2015 at 06:54:20PM +0200, magnum wrote:
> >>BTW we now also use LOP3.LUT for many MD4, MD5 and SHA-2 OpenCL formats.
> >>Some driver bug prevented me for using it in SHA-1 with nvidia 352.39
> >>(the code is there, just disabled) and md5crypt disable it because of
> >>performance regression (still to be investigated). Some formats show a
> >>fine boost but none as much as DEScrypt.
> >
> >... with our guess on why lower boost being that LOP3.LUT was often
> >used anyway, introduced in the PTX to ISA translation.
> Right - actually I did not expect any boost, I just did it because I 
> could. But I'm puzzled by the md5crypt regression,

Yeah, that's a bit puzzling.  While our md5crypt-opencl uses around 600
bytes of private memory, there doesn't appear to be any spilling when
targeting Titan X now, so I would expect it to be computation-bound on
that GPU.  Yet it runs much slower than expected (albeit faster than it
does on other cards).

> and the SHA-1 problem 
> (fails self-test). Perhaps the latter should be reported to nvidia, I 
> think they would actually look into it.

Please give this a try if you have time.  Thanks!


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.