Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!unmvax!pprg.unm.edu!hc!ames!think!barmar
From: barmar@think.COM (Barry Margolin)
Newsgroups: comp.unix.wizards
Subject: Re: bashing (was:  Re: C bug causes double panic)
Keywords: opinions & flames
Message-ID: <37945@think.UUCP>
Date: 25 Mar 89 05:47:04 GMT
References: <244@tree.UUCP> <9884@smoke.BRL.MIL> <2044@viper.Lynx.MN.Org> <13866@rpp386.Dallas.TX.US> <2927@kalliope.rice.edu>
Sender: news@think.UUCP
Reply-To: barmar@kulla.think.com.UUCP (Barry Margolin)
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 23

In article <2927@kalliope.rice.edu> donn@titan.rice.edu (Donn Baumgartner) writes:
>Well, a C program shouldn't be able to crash the system, but one cannot blame
>the operating system for a mess-up on the compiler's part (or lack or the
>appropriate hardware on the user's part - perhaps).

No user-mode program should be able to crash the system, so you can't
blame the compiler.  Your implication is that an assembler program IS
permitted to crash the system (since the output of the compiler is
something that could also have been generated by a properly working
assembler), and this is obviously silly.

If proper operation of the OS requires certain hardware fixtures (such
as an appropriate model of MMU) that the user has not installed, then
this can be a valid excuse for a user program being able to crash the
system.  But it would be nice if the OS checked for this and warned
"Warning! Incorrect MMU installed, incorrect programs may cause system
crash."

Barry Margolin
Thinking Machines Corp.

barmar@think.com
{uunet,harvard}!think!barmar