Not logged in
Log in now
Create an account
Subscribe to LWN
Dividing the Linux desktop
LWN.net Weekly Edition for June 13, 2013
A report from pgCon 2013
Little things that matter in language design
LWN.net Weekly Edition for June 6, 2013
> I think it goes to show it would be simpler just to use Linux in place of DXE.
Isn't there any actual "LinuxBIOS" project out there?
> That's probably naive, but what's the problem? Is Linux too big? Too slow?
Not in the right place at the right time? GPLed?
Maybe a bit of all these.
The Extensible Firmware Interface vs early Linux
Posted Aug 15, 2011 14:23 UTC (Mon) by giraffedata (subscriber, #1954)
You favor USB but other people/platforms favor something else. Anything else.
It doesn't matter to me if it's USB, and I'm not aware of anyone proposing anything else.
The essential element of the concept is a separate removable storage device just for the bootloader. I can't think of anything other than a USB memory stick that would be practical for that, but if there is something, I'm all for it.
Posted Aug 15, 2011 22:08 UTC (Mon) by marcH (subscriber, #57642)
Posted Aug 18, 2011 19:36 UTC (Thu) by giraffedata (subscriber, #1954)
Posted Aug 22, 2011 12:41 UTC (Mon) by marcH (subscriber, #57642)
Granted, you can imagine dream and modular firmware code that can be selectively trimmed down on a per-platform basis. EFI might get there but why would you do this for (expensive) PC motherboards? A development effort to lose some of your customers?!
By the way USB + managed flash is among the most expensive solutions; this matters for low cost embedded systems.
Posted Aug 22, 2011 19:53 UTC (Mon) by giraffedata (subscriber, #1954)
A significant number of the interfaces in this list are used for booting. If you "simplify" your firmware design by hardcoding to any of these you will make users of the others unhappy.
I think you're pointing out the dilemma of standardization. You pick one of many paths in order to reap the benefits of uniformity, but at the cost of going down a path that isn't ideal for some, or even all, particular cases. Of course, we standardize all the time and companies that were relying on the protocol that didn't get chosen suck it up and switch.
A development effort to lose some of your customers?!
Unless you're talking about losing customers because it costs more, you've misunderstood the proposal, because customers can still use all of those boot protocols -- the ROM loads from the USB device the bootloader that knows how to load Windows from a SATA drive.
I agree my scheme is not appropriate for embedded systems. It would add very little and cost a lot.
Posted Aug 22, 2011 22:35 UTC (Mon) by marcH (subscriber, #57642)
Yes something like that.
> you've misunderstood the proposal, because customers can still use all of those boot protocols -- the ROM loads from the USB device the bootloader that knows how to load Windows from a SATA drive.
I think I understood the proposal; it looks like we have different customers in mind. I am considering the "ROM loads from X" part while you are considering the "bootloader that knows..." part.
Copyright © 2013, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds