Nmap Development mailing list archives

Re: Bug?: hang on socket_lock in p2p-conficker.nse


From: Patrick Donnelly <batrick () batbytes com>
Date: Tue, 28 Feb 2012 21:24:44 -0500

On Mon, Feb 27, 2012 at 11:41 PM, Patrick Donnelly <batrick () batbytes com> wrote:
On Feb 27, 2012 12:12 PM, "Daniel Miller" <bonsaiviking () gmail com> wrote:
Command line: sudo nmap -v -A -oA full-%y%m%d --script 'all and not dos
and not qscan and not fuzzer and not external' --reason --log-errors
192.168.1.0/24

I have run this twice with same result: execution hangs for hours/days on
this line in p2p-conficker. Any guidance as to how to debug this would be
great.

I can tell you that the hang is in the actual connect function not
socket_lock. The first stack frame would say socket_lock instead of connect
otherwise. I suspect the problem is the connection attempt is not timing
out? I can't look at the code right now from my phone.

So in the case that the connect function successfully yielded (which
is why it's in the waiting queue), the problem should be found in the
nsock library (or a really obscure NSE engine bug). I would suggest
adding --script-trace and --packet-trace (yes, both) to see if we ever
get the SYNACK back.

-- 
- Patrick Donnelly
_______________________________________________
Sent through the nmap-dev mailing list
http://cgi.insecure.org/mailman/listinfo/nmap-dev
Archived at http://seclists.org/nmap-dev/


Current thread: