On Fri, 10 Mar 2006, Tim Rightnour wrote:

>
> On 08-Mar-2006 Jochen Kunz wrote:
>> I have to admit to be quite clueless in the bootloader department. How
>> hard is it to build a bootloader that reads a kernel image file from
>> FFS? I suppose most of FFS handling is in some kind of MI lib. So it may
>> be "just a bit if glue code" to do the actual IO via firmware calls. Is
>> there a firmware interface for device independent IO? (All I know is a
>> bit OpenFirmWare and the hp700/hppa PDC.)
>
> So.. I've done some investigation on the bootloader, and come up with something
> rather disturbing.
>
> The PReP spec quoteth:
>        "Conventional firmware must not export any entry point for call-back to
> firmware"
>
> This means we are essentially fscked.  When they say "conventional firmware"
> they mean prep firmware, as opposed to actual Open Firmware.  A second stage
> bootloader would essentially need a standalone scsi driver, (possibly requiring
> a PCI layer) a standalone ISA IDE driver, and maybe even network support.
>
> All of this means.. I think we are back to my original idea here.  We just need
> to make mkbootimage as painless as possible to use, and train sysinst to use it.

 	Another option would be to support something like kloader(4), then
 	your 'bootloader' becomes a kernel. Its unlinkely the benefits
 	of that outweigh the effort though...

-- 
 		David/absolute       -- www.NetBSD.org: No hype required --