Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <551C8141.9080606@openwall.com>
Date: Thu, 02 Apr 2015 02:37:37 +0300
From: Alexander Cherepanov <ch3root@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: Advice on proposal: John the Ripper jumbo robustness

On 2015-04-01 03:10, magnum wrote:
> On 2015-04-01 01:51, Alexander Cherepanov wrote:
>> On 2015-03-29 14:08, Kai Zhao wrote:
>>>> it might be more of a distraction from focusing the actual fuzzing jobs.
>>>
>>> Yes, I agree with you. It would be better to focus the actual fuzzing
>>> job.
>>
>> The first thing to do with fuzzing john is making john start fast.
>>
>> I've taken a look at it. There are two easy things here: instrumenting
>> only a fuzzed format and john.c make john ~5x faster, removing most
>> content from john.conf makes john ~7x faster. That's 35x boost.
>>
>> Solar, magnum, any other easy ways to make john start faster?
>
> Registering the dynamic formats usually end up near top in (too) short
> profiling runs, even when not using dynamic. Did you trim the
> dynamic*conf files too?

I think so. They are .include'd into john.conf so emptying it bypasses 
all dynamic*conf.

-- 
Alexander Cherepanov

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.