Re: installboot is broken

[Johnny Billquist <bqt%softjar.se@localhost>]

On 2013-03-24 17:33, Matt Thomas wrote:
> On Mar 24, 2013, at 9:32 AM, Johnny Billquist <bqt%softjar.se@localhost> wrote:
>
> > On 2013-03-24 16:39, Anders Magnusson wrote:
> > > On 03/24/2013 03:36 PM, Martin Husemann wrote:
> > > > On Sun, Mar 24, 2013 at 09:21:39AM +0100, Martin Husemann wrote:
> > > > > I didn't have old enough disklabel sources around, will check there.
> > > > It looks to me that installboot does exactly what disklabel -B did after
> > > > this commit:
> > > >
> > > > revision 1.17
> > > > date: 1995/04/23 19:27:59;  author: ragge;  state: Exp;  lines: +3 -3
> > > > Changed to use only one bootblock file on VAX.
> > > >
> > > > to sbin/disklabel/Makefile.
> > > Vax only have one bootblock which is installed in the first 16 sectors
> > > on the disk.
> > > The boot block is the same for all different disk types. Which is nice :-)
> >
> > Right. However, the real problem is that installboot does not use block 0 from xxboot. 
> > installboot instead have a structure defined with some "magic" fields, and 
> > installboot only changes one or two values in the boot block when you run it. It does 
> > copy block 1 and up from xxboot, but it totally skips block 0 of xxboot.
>
> The reason for that is not overwrite the disklabel.

Yes. However, block 0 of xxboot has been designed to work together with 
the disk label. A large part of block 0 is zeroes, and worked around in 
the code, so that the disklabel can coexist.

        Johnny

--
Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt%softjar.se@localhost             ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol