[OLPC-devel] Re: belt and suspenders...

[Ronald G Minnich]

John R. wrote:

> Yeah, I see what you're saying, corner cases are a pain and it's hard
> to know what's there until one gets neck deep into it. Is the proposed
> requirement to boot from any usb-storage device?  Can we just qualify
> 10 or 20 different makes/models and call it a day? Or would such a
> driver really have to boot from anything that calls itself
> usb-storage?

I think we're going to get into a world of trouble if we put a 
not-very-complete stack in linuxbios. Let's let linux do what it does so 
well. If Linux is not bootable from BIOS ROM, then we're screwed, pure 
and simple; the BIOS is corrupted at that point.

> If we can construct the requirement as "load the first so many sectors
> from a random sample of usb flash or hard drives formatted in a
> particular way to RAM when nothing else is connected to the bus" I
> think it's more likely to be doable.

That was an early proposal I had for the linux part, but we determined 
that we'd just go with a file system.

I still think that our effort would be better spent, if we assume that 
linux is the bios and it has to not be corrupt. Again, if you are in a 
state where the bios has gotten corrupted, all bets are off.

Also, work needs doing: I need to write a script (or you do :-) that 
takes a NAND file system image in tar.gz format on a usb stick, and 
fixes up the NAND and writes to the NAND. it would be really nice to have.

At the same time, what you are proposing could be generally very useful, 
even outside the domain of OLPC, and you could well prove me wrong -- I 
don't want to discourage you from trying it!

thanks

ron