pcp
[Top] [All Lists]

Re: Checking PCP archives - RFC

To: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Subject: Re: Checking PCP archives - RFC
From: fche@xxxxxxxxxx (Frank Ch. Eigler)
Date: Thu, 23 May 2013 23:28:48 -0400
Cc: pcp@xxxxxxxxxxx
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <519D2BE3.9010107@xxxxxxxxxxxxxxxx> (Ken McDonell's message of "Thu, 23 May 2013 06:34:43 +1000")
References: <519AC94B.9020904@xxxxxxxxxxxxxxxx> <y0mfvxgl3r3.fsf@xxxxxxxx> <519C0AA9.5010706@xxxxxxxxxxxxxxxx> <20130522132819.GJ28935@xxxxxxxxxx> <519D2BE3.9010107@xxxxxxxxxxxxxxxx>
User-agent: Gnus/5.1008 (Gnus v5.10.8) Emacs/21.4 (gnu/linux)
Ken McDonell <kenj@xxxxxxxxxxxxxxxx> writes:

> [...]
>> (Yes, fuzzing can include structured data, by teaching it the grammar of
>> PCP archives but then messing with the productions randomly.)
>
> Frank do you have a pointer to an available toolkit that would be
> suitable for this sort of effort?  I've read about, but never used
> fuzzers.  

Yeah, I'm in the same boat.  One might reinvent the wheel by
hand-coded error-generators like mkbadlen, we could investigate tools
like http://peachfuzzer.com/, wherein an XMLy model is made of the
data format, then the tool generates a gajillion slightly-wrong ones.
The fact that PCP archives span more than one file may pose a problem.
OTOH, the same toolset can also be configured to do fuzzing of the
wire protocol (!).


- FChE

<Prev in Thread] Current Thread [Next in Thread>