🚀 go-pugleaf

In article <BvFunJ.E1H@gate.demon.co.uk> paul@seer.demon.co.uk
	(Paul Lynch) writes:
>In article <1992Sep30.152220.18225@pilhuhn.ka.sub.org>  
>jr@pilhuhn.ka.sub.org (Jochen Richter) writes:
>> The user program CRASHME that is available from some ftp-servers
>> crashes NS 3.0 in less than a second!!! That is the depressing 
>> result of my tests. It's a shame that NS hasn't been tested with
>> this in many magazines demonstrated user program. It produces
>> many different system exceptions that have to be handled properly
>> by the kernel...
>> After the crash you can't even reach NMI! You have to use 
>> <Command><Alternate>*.
>> 
>> Shame!!!
>> 
>
>So what.  Almost every Unix system is crashed in < 2 seconds by Crashme.
>

If I remember correctly, a fair number of Unix systems -aren't-
crashed by CrashMe.  Anyway, what CrashMe does is fork as many
processes as it can and write to memory randomly, in the hopes of
destroying something important that the o/s wasn't protecting
properly. I would assume that the problem NeXTs have with this is the
same old one that comes up when you fork() too many processes or
threads--the panic out of nowhere and lock-up that was discussed
earlier in the thread about what happens when you try to write a
multi-threaded app with, say, 200 threads...

matt