[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Mon, 8 Jun 2015 18:12:46 +0300
From: Solar Designer <solar@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: Kai's weekly report #6
Kai,
On Mon, Jun 08, 2015 at 08:28:41PM +0800, Kai Zhao wrote:
> > OpenCL and CUDA, please. And not only by afl, but also by fuzz.pl.
>
> Yes, also CUDA. But I think it would be better to test CUDA next week,
> since there are many things to do this week, such as improve fuzz.pl,
> also it takes several days to fuzz OpenCL by afl.
I feel that this "several days" figure is arbitrary. You should be
able to turn days into hours by starting to run parallel jobs, finally,
and by using tmpfs for session and pot files.
When you're keeping a machine with GPUs a few percent busy for a long
time, preventing us from benchmarking stuff on it, this is just
unreasonable. I'd prefer that you keep it 100% loaded for a shorter
period of time instead.
You just have to learn how to run parallel jobs.
Regarding fuzz.pl:
> As you said, add false positives detection and more replacement
> characters (add '*' and '#'). Also maybe I can try other mutate methods,
> such as insert string at somewhere in the hash, change case.
Sounds good.
Thanks,
Alexander
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.
