[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Sun, 13 Jan 2013 11:00:47 +0100
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: Re: ssha opencl kernel update
On 13 Jan, 2013, at 9:49 , mayuxiang <losywar@....com> wrote:
>> void prepare_msg(__global uchar *s, char *dest, __global uchar *salt, int
> blocksize) {
>
> I want to know how to use this kernel, it it what the data_info、 salt and
> pain_key represent. Can you help me, thank you.
That kernel has evolved since the old thread you were quoting. It's still not very optimized. Most recent version here:
https://github.com/magnumripper/JohnTheRipper/blob/unstable-jumbo/src/opencl/ssha_kernel.cl
data_info provided information of plaintext array size and global work size. Since the former never change in a session, we now instead pass -DPLAINTEXT_LENGTH=<length> to the run-time compiler. This potentially mean better optimization from the compiler. Passing global worksize was never a good idea, get_global_size(0) is the obvious choice for that.
In https://github.com/magnumripper/JohnTheRipper/tree/unstable-jumbo/src/opencl there are other implementations (eg. iterated sha1) that may be better depending on your needs.
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.
