Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 24 Apr 2012 21:12:07 -0300
From: Claudio André <claudioandre.br@...il.com>
To: john-dev@...ts.openwall.com
Subject: Re: SHA-256 code from Bitcoin miners

Hi, I've been looking at these miners for a while. But since i'm not a 
criptographic guy, i haven't found (easy) good ideas.

But, now looking at some recent opencl files in this "cgminer", i 
noticed there are some ideas and/or probably the correct syntax of 
something i failed to use in the past.

I'll try the opencl usage part and i'll report.

Cheers


Em 24-04-2012 20:36, Solar Designer escreveu:
> Hi,
>
> I think we should look at and consider for reuse suitably licensed
> SHA-256 implementations from Bitcoin miners.  Here's one:
>
> https://github.com/ckolivas/cgminer
>
> It includes 4-way SIMD implementations of SHA-256 for SSE2 and AltiVec,
> two optimized OpenCL kernels (one of them for GCN), and more.  (In fact,
> I found it by searching the web for GCN instruction names, which
> suggests that the kernel was optimized with specific generated
> instructions in consideration.)
>
> While this miner as a whole has recently been relicensed to GPLv3 (as
> far as I understand), most of the individual files we're interested in
> have their own license statements on them and are under much more
> relaxed licenses.  So we should be able to take them for modification
> and reuse in JtR.
>
> diablo120328.cl (non-GCN) is under GPLv3+, but perhaps we can start with
> an older version, which maybe was still under GPLv2 (I did not check yet).
>
> Alexander

Powered by blists - more mailing lists

Your e-mail address:

Powered by Openwall GNU/*/Linux - Powered by OpenVZ