lkml.org 
[lkml]   [1998]   [Feb]   [12]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: MILO vs BIOS on Linux/Intel.
Date
Dave Cinege wrote:
> This is brings up a good point. Do we even need the BIOS and all it's stupity
> in the first place? Couldn't something like GRUB with real mode config
> programs replace it entirly?

Good luck booting from any SCSI disk, let alone initializing all the chips
on your mainboard ;-)

The good thing about the BIOS is that it knows about the real hardware and
all of its idiosyncrasies. The bad thing about it is that its design is
largely limited by what people perceived as necessary twenty years ago.

If you really want to replace the BIOS, you need to
- obtain complete documentation on your mainboard chipset
- obtain complete documentation on any other peripherals on your
mainboard if you need/intend to initialize them
- find out how your mainboard manufacturer put things together
- repeat the same procedure for any cards which bring their own BIOS
(SCSI, video, etc.)
- obtain complete documentation on how you query SIMMs/DIMMs/etc.
(including common bugs)
- write something that properly initializes your system (most likely
providing a limited BIOS-like interface so that you can deal with the
boot loader later; existing Linux boot loader typically only use very
few different BIOS functions)
- throw in your boot loader functionality. Now you have a solution for
one specific system in its current configuration.
- extend this to the most common chipsets and board designs
- get things tested with various memory configurations, etc.
- add some backward-compatibility interface for people needing DOS or
one of its offsprings
- get this tested with the usual operating systems
- implement some clever auto-configuration mechanism
- announce availability of the first GPLed BIOS to the world
- update things whenever a new CPU/chipset/board design/card with
BIOS/OS/OS version hits the market

Now all this isn't too different from general device driver implementation
and maintenance, except that (a) you have to be nice to DOS and friends
too, and (b) "chipset foo doesn't work with this release" will be very
happy news for the first one who finds out and who's probably left with a
system that - without some surgery - won't even get as far as accepting a
boot floppy with a new fixed BIOS. (b) leads to a chicken-and-egg problem:
very few people will be able to afford testing unstable versions, so it'll
be hard to know when/if things are stable. And releasing unstable things
as stable will get you loads of hate mail (sent from a friend's PC ;-)

To summarize: it's hard to do, you don't gain a lot over the existing
solutions, you'll have difficulties finding testers, and people will get
very upset at even your slightest mistake.

Special-case solutions aside, I don't see us getting rid of the BIOS
anytime soon.

- Werner

--
_________________________________________________________________________
/ Werner Almesberger, DI-ICA,EPFL,CH werner.almesberger@lrc.di.epfl.ch /
/_IN_R_131__Tel_+41_21_693_6621__Fax_+41_21_693_6610_____________________/

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu

\
 
 \ /
  Last update: 2005-03-22 13:41    [W:0.075 / U:0.008 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site