Linked by Thom Holwerda on Mon 25th Dec 2006 13:05 UTC, submitted by Matt D.
"Let's be honest here. When most of us think of open source and free software, we really aren't thinking of something flashed to a BIOS ROM chip. And yet when it comes to the Linux BIOS project - that's exactly what's happening here. The LinuxBIOS is a project designed to work out any perceived shortcomings from existing BIOS options distributed on today's motherboards. Started back in 1999, it has been in development steadily for quite some time now."
Thread beginning with comment 196202
To view parent comment, click here.
To read all comments associated with this story, please click here.
To view parent comment, click here.
To read all comments associated with this story, please click here.
Recent Original Stories
Recent Comments
Headlines
Random Comments
Random Stories
Random OS Link
Member since:
2005-11-16
http://linuxbios.org/ADLO
Apparently it can.
Apperently, you didn't read this page. Here's some of the highlights:
"What bootloaders work with ADLO?
Currently known to work are:
* LILO
* NTLDR (Windows 2000)
There is no technical reason why it couldn't work with *BSD, (Free)Dos, and other versions of Windows after some tweaking.
This means that other OSs (like FreeBSD and WindowsXP) currently won't boot (but it's expected that they will boot once more of ADLO is implemented).
"Out-of-the box ADLO probally won't boot unless you are using the exact mainboard that the ADLO project uses"
This means you need to hack ADLO to make it suit your motherboard/chipset (and I guess that there is no hooks in LinuxBIOS that ADLO can use for BIOS shadow control).
Lastly, if you follow the link on that page to WindowsXP information (http://www.missl.cs.umd.edu/winint/index2.html) you'll see an analysis of what would need to be supported by ADLO to get WindowsXP working (with no indication of which parts are already supported). If you read carefully you'll see they're looking into how much doesn't need to be supported (including ACPI, BIOS32, PnP BIOS, APM, the BIOS data area and the Extended BIOS data area).
If ADLO leaves all of that out it'd be much easier to finish implementing ADLO and (according to that page) WindowsXP should still boot. However, in this case WindowsXP wouldn't be able to do any power management and may have trouble with device resource usage - I severely doubt it'd be as correct as a standard BIOS unless most of this is included (but including most of it is a major undertaking that would need to be redone for each motherboard).
LinuxBIOS claims that it only takes 3 seconds from power on to a linux console, how nice would that be?
On their home page they do claim it only takes 3 seconds from power on to a linux console. On other pages (e.g. http://linuxbios.org/FAQ#Why_do_we_need_LinuxBIOS_for_cluster_maint...) they claim that 3 seconds is the current record (I assume that's for a cluster node, not an embedded system, as 3 seconds would be far too slow for the fastest embedded system). This means that the average computer is slower than 3 seconds (and LinuxBIOS could actually be slower than your current BIOS, especially if you're booting ADLO instead of Linux).
It'd be nice to know where this 3 second record came from too. Imagine a computer with a small amount of RAM, no hard disks and no other devices except for one ethernet card being booted using network boot (with the boot files pre-cached by the server), and no system services running on it. To me, 3 seconds seems too slow for this (and yes, all of this is possible for a node in a cluster)...