Re: nmap -P0 reboots Windows XP

[Casey Williams <Lists () aviditysoftware com>]

On 12:54 Wed 03/08/06 Mar    , Loris Degioanni wrote:
> Let's try to clarify a bit.
>
> IRQL_NOT_LESS_OR_EQUAL is one of the most common Windows kernel 
> bugchecks, and  doesn't normally have anything to do with corrupted 
> memory. It just tells you "one of the component of the OS kernel, most 
> probably a driver, did something wrong, and therefore the OS will stop".

Excuse me, I should have spoken more clearly.  I didn't mean that I thought it was a memory 
issue.  What I meant was that I thought something was going awry during the driver unload 
process.  What would go wrong?  I'm not sure, I'm not knowledgable about such things, however 
here's what I know: When I start my program, and close it without scanning it will never crash.  
However if I DO scan something (thus initializing the winpcap driver), it will 
crash...intermitantly.  And it only crashed on exit from the application.  This seems to be the 
same issue that Kris is having with Nmap...
 
> The most probable cause of the error, if this happens while you use 
> nmap, is the kernel driver that nmap uses to send and receive raw 
> network frames, i.e. WinPcap's NPF.sys. In most cases, in fact, you can 
> see if the fault is in winpcap by checking if around the bottom of the 
> blue screen you see somewhere "npf.sys".

I didn't see anything on the BSOD that indicated any specific driver, I don't know how I 
would have missed it, but I'm sure it's possible. :)  And I agree, it's probably the winpcap 
driver causing the issue.

> If the fault is in winpcap, first of all you should make sure you are 
> using the latest version of the driver, which can be found at 
> http://www.winpcap.org/install/default.htm. If the problem persists with 
> the latest version, you can report the bug to the developers as 
> explained at http://www.winpcap.org/contact.htm, under "Need to report a 
> bug?".

Yep, I have the latest libraries, and I'd like to file a bug report, but I don't have any 
solid proof of who's to blame, and since we're on the topic I'm posting my finding here in hopes 
to be of some help.  Sorry if I'm being a nuisance. :)

--
-C



_______________________________________________
Sent through the nmap-dev mailing list
http://cgi.insecure.org/mailman/listinfo/nmap-dev