From owner-freebsd-hackers  Sun Mar 31 13:58:14 1996
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id NAA01097
          for hackers-outgoing; Sun, 31 Mar 1996 13:58:14 -0800 (PST)
Received: from mail.think.com (Mail1.Think.COM [131.239.33.245])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id NAA01091
          for <hackers@freefall.freebsd.org>; Sun, 31 Mar 1996 13:58:11 -0800 (PST)
Received: from Early-Bird-1.Think.COM by mail.think.com; Sun, 31 Mar 96 16:58:07 -0500
Received: from compound ([206.10.99.151]) by Early-Bird.Think.COM; Sun, 31 Mar 96 16:58:03 EST
Received: (from alk@localhost) by compound (8.6.12/8.6.112) id PAA26843; Sun, 31 Mar 1996 15:59:25 -0600
Date: Sun, 31 Mar 1996 15:59:25 -0600
Message-Id: <199603312159.PAA26843@compound>
From: Tony Kimball <alk@Think.COM>
To: hackers@freefall.freebsd.org
Subject: random traps
Sender: owner-hackers@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk


I once had a program which might do the stability of FreeBSD a world
of good, if applied in earnest.  That program generated random
syscalls.  It would reliably crash Ultrix or SunOS 4.0.x or 4.1.[01]
within 5-50 seconds (depending on the OS more strongly than chance:-),
but it would run forever under SunOS 4.1.[23] (or more precisely would
run repeatedly ad nauseam without crashing the box).  I haven't tried
it in a few years, on more modern systems, but the degree of
resistance to this abuse was at the time almost perfectly correlative
to my intuitive notion of OS quality.  So... I wonder whether the
intuitive quality of FreeBSD might not be given a hand up by such a
treatment...

Just a thought.

//alk