[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CA+E3k90sxA667AU0w7TvKf_3FzzQQuinG4j4JSQej6ib2NoNTQ@mail.gmail.com>
Date: Sun, 31 Aug 2025 14:47:46 -0800
From: Royce Williams <royce@...ho.org>
To: john-users@...ts.openwall.com
Subject: Re: suppressing potfile / minimizing writes?
On Thu, Aug 28, 2025 at 12:21 PM Solar Designer <solar@...nwall.com> wrote:
> On Tue, Aug 26, 2025 at 09:22:07PM -0800, Royce Williams wrote:
> > I would like to do the equivalent of disabling potfile writing entirely,
> or
> > write it to /dev/null. This is for various benchmarking tasks, in which I
> > do not want I/O write speed to affect the results.
> >
> > I tried --pot=/dev/null, but john attempts to chmod the potfile prior to
> > running, which fails on Linux so john aborts. I then tried a FIFO to do
> the
> > equivalent, which john explicitly disallows.
> >
> > How can I accomplish this?
>
> We do not have a feature like this, sorry. Maybe you'd contribute it.
>
This is interesting! Naively, there are a couple of options.
One approach would be to detect when filesystem objects that do not support
chmod are invoked, and skip such steps, and make other modifications so
that such special files could be used for output.
Another option might be to simply detect when specific, well-known
filesystem objects (/dev/null on Unix-likes, NUL on Windows, etc.) are used
for --pot, and simply ... return early on attempts to write to them,
thereby skipping the filesystem work entirely.
Which of these would be preferable? Or would some other choice be better?
Royce
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.
