Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <cfdd5e8c65878261051873414ed141d4@smtp.hushmail.com>
Date: Sat, 25 Apr 2015 22:20:12 +0200
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: [GSoC] John the Ripper support for PHC finalists

On 2015-04-25 21:20, Solar Designer wrote:
> Also, I think you're over-using "unsigned long", such as for loop
> variables.  This may have performance impact on GPUs.  You should prefer
> "uint" except where a wider type is actually needed.

I have seen significant slowdowns from using long on GPU where not 
needed. All current GPUs are 32-bit. Not sure if the compiler will ever 
optimize the use of long down to int - except for loops.

> You may also try "#pragma unroll" and "#pragma unroll N" before loops,
> see e.g. cryptsha512_kernel_GPU.cl for examples.

Do try that, but I am yet to see it making any difference at all. I 
think Claudio disagrees with me.

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.