intel_vs_better

RHS Linux User (humbubba@raptor.cqi.com)
Sun, 20 Sep 1998 00:51:29 -0400 (EDT)


I'm browsing the web trying to figure out why Intel wants to be involved in the
creation of a driver interface specification.
And why they and x86 and non-x86 unix vendors would like to see Linux involved.
Would this be like IEEE 1275-1994 ? Open Firmware?
( I mean, aside from being Intel-specific.)
Millions of SPARCStations and PPC's of various flavors, including Motorola
boxen, have done well with this platform-independant facility. Is that a
problem?

The IEEE makes ( much of) it's money selling hardcopy versions of thier specs,
which are otherwise completely public AFAICT, as far as use or licensing. ANSI
is similar, and the ANSI Forth spec which is closely related to Open Firmware is
over $100-- US. In the case of ANS Forth there's a copy of a late draft version
of the standard on the net, but I haven't seen such a generous slip-up from
IEEE. Sun has some online stuff on OpenBoot though. But it's a standard, it's
clearly stated that anyone can implement it. Like RS-232.

It seems Linuxppc depends on Open Firmware but mklinux does not. Looking for
clues as to how O.F. relates to a Linux kernel I find the following on
yahoo....
..................................................................
/* $Id: mp.c,v 1.5 1996/04/04 16:31:06 tridge Exp $
* mp.c: OpenBoot Prom Multiprocessor support routines. Don't call
* these on a UP or else you will halt and catch fire. ;)
*
* Copyright (C) 1995 David S. Miller (davem@caip.rutgers.edu)
*/

#include <asm/openprom.h>
#include <asm/oplib.h>

/* Start cpu with prom-tree node 'cpunode' using context described
* by 'ctable_reg' in context 'ctx' at program counter 'pc'.
*
* XXX Have to look into what the return values mean. XXX

.............................................................

openprom.h looks very interesting. Is that a declaration of a bunch of Forth
words?

The Sun response to Mudge's recent Phrack artictle about "Boy Howdy! Root the
Sun Workstation you're sitting in front of!" showed up in yahoo for Openboot
also. The interesting thing I vaguely recall from that article is that Sun's
OpenBoot ( , which apparently came before Open Firmware) gives one the
ability to write most of the kernel variables in a running ( or halted, I don't
recall) Solaris. Not quite dd if=file > /proc, but you could write
the equivalent in Forth in minutes.

So it seems that O.F. is like a much broader kernel "command line" interface.
Am I in the ballpark? Would such a thing HAVE to be in ROM?
How much driver twiddling does it do that an x86 Linux kernel does itself?

Further web snips appended for grins.
( and yes, my Forth-style use of parentheses IS a subliminal message :o)

Are the ISA dodads in the following snip whole drivers?
I think one thing I don't get is, is there a Forth IN an O.F. compliant kernel?
A Forth "kernel" ( kernal?) WITHIN a unix kernel is a requirement to use
Forth sourcecode drivers. y/n ?
..........................................................
http://playground.sun.com/pub/1275/bindings/devices/html/8042-1_0d.html

CHRP(TM) Device Bindings to
IEEE 1275-1994
Standard for Boot (Initialization,
Configuration) Firmware

* CHRP ISA Keyboard/Mouse Controller Device Binding
* CHRP ISA Audio Device Binding
* CHRP ISA Floppy Controller Device Binding
* CHRP ISA DMA Controller Device Binding
* CHRP ISA Interrupt Controller Device Binding
* CHRP Linear Frame Buffer Display Device Binding
* CHRP ISA Parallel Port Device Binding
* CHRP ISA Serial Port Device Binding
* CHRP VGA Display Device Binding

http://playground.sun.com/pub/1275/bindings/devices/html/8042-1_0d.html
............................................................

............................................................

Mitch Bradley

Chief Technical Officer

BEEE - Vanderbilt University

CPGS - Cambridge University

MSEE - Stanford University

Bradley joined Sun Microsystems in 1982, handling a variety of
technical roles, including hardware design, operating system software
work and system-level troubleshooting. In 1986 Bradley conceived and
began development of OpenBootŪ firmware, which first shipped on Sun's
SPARCstation(tm) 1 system. Bradley led Sun's firmware development team
from 1986 until 1993, during which time Sun shipped approximately one
million copies of OpenBoot(tm) firmware and licensed its source code
to numerous other SPARC system vendors. Also during that time, Bradley
organized and led the IEEE 1275 working group to develop the IEEE
Standard 1275-1994.

Bradley resigned from Sun to focus on supporting the many companies
who were adopting Open Firmware. In response to the growing popularity
of Open Firmware and the resulting increased need for service and
support, Bradley founded FirmWorks in 1994.

Bradley is chairman of the Open Firmware Working Group and vice
chairman of the X3J14 ANS Forth Technical Committee.

Back to What is FirmWorks?
Back to FirmWorks Home Page

OpenBoot is a trademark of Sun Microsystems, Inc.
SPARCstation is a trademark of Sun Microsystems, Inc.
_________________________________________________________________

Last Modified: February 9, 1996
.................................................................

BTW, my latest Linuxisms are a 2.88 meg broot image for PS2's and a tweak of
rawrite that does 2.88 meg floppies.
http://glycerine.itsmm.uni.edu/mca/index.html
in the "raw dowload"s link I think, or ftp to pub/Incoming.

Rick Hohensee http://cqi.com/~humbubba
colorg on EFnet IRC #linux chanop
Forth C Linux Perl graphics music Md., USA
This is your brain on colorg --> (@#*%@#() <---~~~_()()(
Any questions?

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/