Re: Weird Crash - "WAITING_TO_RUNNING"

[Nathan <nathan.stocks () gmail com>]

On Tue, Nov 23, 2010 at 2:59 PM, David Fifield <david () bamsoftware com> wrote:
> I looked over the code and I can't find what might be causing the change
> in run time. I attached four different versions of nse_main.lua that
> make incremental changes. Please try running each of the four and report
> the times you get. Just copy each one over nse_main.lua in turn.
>
> David Fifield

Sorry for the delay, it's been a very busy work week.  Here are the
measurements you asked for with your custom versions of nse_main.lua.
Please note that the values will fluctuate ~2% from run to run, but
within about that margin of error the runs were very consistent with
themselves.  Also, these were all run with nmap 5.35DC1, with
nse_main.lua replaced by the file indicated.

nse_main.lua.1: 53732 kB RAM, 5m40s

nse_main.lua.2: 53588 kB RAM, 5m21s

nse_main.lua.3: 55512 kB RAM, 5m27s

nse_main.lua.4: 54352 kB RAM, 5m22s

Since none of those seem to display any problems, I went and ran a
bunch of other things:

1) Here's the nse_main.lua pulled from 5.36TEST2 and copied over into 5.35DC1

54448 kB RAM, 5m23s

2) Same as #1, but with the CONCURRENCY_LIMIT lowered to 100

54312 kB RAM, 5m32.126s

3) Here's the vanilla nmap 5.36TEST2 on the same machine (when I run
this one, I compile it but don't install it, then I run it from the
directory it was compiled in).

94532 kB RAM, 14m48s


So what really mystifies me is #1 up there.  It's the nse_main.lua
from 5.36TEST2 with a CONCURRENCY_LIMIT of 1000, and yet it seems to
run fine.  And then that same file running under nmap 5.36TEST2
performs horribly.  Perhaps the problem is with some other change in
5.36TEST2, and not in nse_main.lua.

Please let me know if you would like me to try anything else out.

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