On Fri, Feb 19, 2010 at 10:05:06AM +0200, Lars Nooden wrote:
> Otto Moerbeek wrote:
> > On Thu, Feb 18, 2010 at 09:44:17PM +0200, Lars Nooden wrote:
> >
> >> On Thu, 18 Feb 2010, Otto Moerbeek wrote:
> >>> Retry with boot -k tftp://..., as suggested by the error message.
> >>> Also PMON sometimes gets confused, and a power cycle is needed (using the
> >>> reset button is not enough in all cases).
> >> Thanks. I had misinterpreted the message and put the -k as an
> >> argument for bsd.rd
> >>
> >> Boots bsd.rd fine now. There are a great many 'spurious interrupt
> >> 4' messages during the installation process. The ext2 boot
> >> partition seems to still needed for booting. I tried to dig out
> >> some linux netboot for that but couldn't find anything that supports
> >> fuloong yet. Ended up using dd to make the ext2 partition.
> >>
> >> It boots bsd current just fine now via the ext2 partition.
> >>
> >> /Lars
> >
> > The sprurious interrupts will be solved if you update to current.
>
> Ok. I'm probably trailing a version behind. According to dmesg it is
> with GENERIC #90 from 17 Feb at 16:59. I'll take another look with the
> next snapshot!
Yes, development is fast. You could also update your source tree and
rebuild the kernel if you want the latest. A new snap with the new
installer code will hit the mirrors soon I hope. This snap also will
have the spurious interrupt fix.
>
> > The lasy days I spent on working at the install procedure. The code I
> > am about to commit is able to create a small ext2 partition or use an
> > existing ext2 one to install the bootloader on. The kernel the wil be
> > read from ffs.
>
> Sweet. Can I ask how, priori to that, you solved the ext2 problem?
>
> /Lars
I ported newfs_ext2fs.
-Otto