Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!ames!haven!decuac!e2big.mko.dec.com!bacchus.pa.dec.com!decwrl!sdd.hp.com!zaphod.mps.ohio-state.edu!usc!snorkelwacker!bloom-beacon!eru!luth!sunic!nuug!sigyn.idt.unit.no!sigyn.idt.unit.no!bjornmu
From: bjornmu@idt.unit.no (Bj|rn Munch)
Newsgroups: comp.lang.c
Subject: Re: Crash a RISC machine from user-mode code:
Message-ID: <1990Aug11.164932.9852@idt.unit.no>
Date: 11 Aug 90 16:49:32 GMT
References: <1826@mountn.dec.com> <49041@seismo.CSS.GOV> <30273.26c33972@ccavax.camb.com>
Sender: news@idt.unit.no (Usenet news admin)
Organization: Div of CS & T, Norwegian Institute of Technology
Lines: 25

In article <30273.26c33972@ccavax.camb.com>, merriman@ccavax.camb.com writes:
|> In article <49041@seismo.CSS.GOV>, 
|>     stead@beno.CSS.GOV (Richard Stead) writes:
|> [stuff deleted]
|> > Who could possibly care
|> > that a random instruction sequence crashes a risc box? 
|> 
|> Some villain wanting to mount a denial-of-service attack.

Or just about anyone running a buggy program.  It's bad enough that
the program crashes, you shouldn't have to risk (RISC :-)) bringing
the whole machine down or do unknown things to other processes.  And
there are always people running buggy programs, for example those who
are debugging them....

A compiler won't guarantee against executing "random" instructions.
You may forget to initialize a function pointer, or write past the end
of an array on the stack, thereby disrupting a return address.

Just my $.02 (15 |re) worth...

_______________________________________________________________________
Bj|rn Munch                  | Div. of Comp. Science & Telematics,
bjornmu@idt.unit.no          | Norwegian Institute of Technology (NTH),
PhD Student (well, soon...)  | Trondheim, Norway