Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 07 Apr 2000 12:18:17 +0000
From:      Gustavo V G C Rios <kernel@tdnet.com.br>
To:        Alfred Perlstein <bright@wintelcom.net>
Cc:        freebsd-hackers@FreeBSD.ORG
Subject:   Re: Is traditional unixes kernel really stable ?
Message-ID:  <38EDD209.421EF9B0@tdnet.com.br>
References:  <38ED128C.22C3AA28@tdnet.com.br> <20000406192206.N22104@fw.wintelcom.net> <38ED233E.74716D02@tdnet.com.br> <20000406230234.B4381@fw.wintelcom.net>

next in thread | previous in thread | raw e-mail | index | archive | help
Alfred Perlstein wrote:
> 
> Some archs (such as i386) allow the OS to set page protections and
> io permission bitmaps that effectively can pretect against problems
> with drivers touching incorrect IO ranges, however...
> 
> >
> > Worse yet: What about hardware buggy devices?
> > This could case the entiry system to crash, isn't it ?
> 
> Yes, incorrectly programmed hardware either by firmware (on
> chip/board) or by drivers can cause crashes and hardware damage.
> 

That's the point!
Why not a different approach ?
Why not starting a microkernel arch? The microkernel would basically do
just feel tasks, like:

IPC: managing and routing messages.
Process scheduling.
First level interrupt handling.


All other tasks would run in like any other user process, like a fyle
system daemon, process daemon , internet daemon (not inetd), and, of
course, device drivers programs.

This design, would not let a system crash due to device drivers problems
or even bad hardware desgin.

What all you think about that ?


-- 
If you're happy, you're successful.


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?38EDD209.421EF9B0>