Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Wed, 29 Oct 2014 13:56:05 +0300
From: Aleksey Cherepanov <lyosha@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: [RFC] Johnny further development proposal

Hi Shinnok,

On Wed, Oct 08, 2014 at 09:58:27AM +0300, Shinnok wrote:
> 2. Fix any outstanding bugs or crashes (crash on exit while john is running,
> pause not working, etc..)

I hope there are no outstanding bugs or crashes at the moment. I think
the crash on exit was due to multithreading. I avoided threading in
2012. So I think Johnny does not crash anymore, right?

The threading was intended to unfreeze ui when john produces a lot of
output. But there was a problem: the main load is not from the output
itself but it is from handling the output (including placing it into
the table). I've added hash table and the other way to get output
(using `john --show` and temporary file) so the handling is really
fast. It could be improved further but one should move the whole
handling into other thread, it looks unnecessary complex.

`john --show` needs more explanations: it prints hashes in john's
canonical form (as they are stored in .pot file) but we have
original form in the table (these forms are not always equal). To
solve this, we craft file with hashes that has original hash in 2
fields: in hash position and in gecos position. `john --show` prints
gecos unmodified so we get original hash and password.

These are methods:
void MainWindow::callJohnShow()
void MainWindow::readJohnShow()

Temporary file is populated in
void MainWindow::showJohnStarted()

Hm, what if we have two "original" forms of 1 hash?... I have to try.

Thanks!

-- 
Regards,
Aleksey Cherepanov

Powered by blists - more mailing lists

Your e-mail address:

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