[Debian-ppc64-devel] Re: Package: installation-reports

[Sven Luther]

On Wed, Nov 09, 2005 at 11:13:31PM +0100, Frank van Berkel wrote:
> 
> Hi Sven,
> 
> On Wednesday 09 November 2005 12:06, Sven Luther wrote:
> 
> > It should be listed together with ext2/ext3/swap and so on.
> >
> > Can you check if partman-prep is installed ? I think so because prep was
> > listed in your partman log, and this worked fine on my prep box. And the
> > list has :
> >
> > XB-Subarchitecture: chrp_rs6k prep
> >
> > are you sure you are chrp_rs6k, and not plain chrp ?
> 
> I think I made a mistake telling you that I was sure it's 
> chrp_rs6k....sowwy ;-s 
> After a double check I think you're right, look here:
> 
> /proc/cpuinfo: processor        : 0
> /proc/cpuinfo: cpu              : POWER5 (gr)
> /proc/cpuinfo: clock            : 1654.344000MHz
> /proc/cpuinfo: revision : 2.1
> /proc/cpuinfo: 
> /proc/cpuinfo: processor        : 1
> /proc/cpuinfo: cpu              : POWER5 (gr)
> /proc/cpuinfo: clock            : 1654.344000MHz
> /proc/cpuinfo: revision : 2.1
> /proc/cpuinfo: 
> /proc/cpuinfo: timebase : 207048000
> /proc/cpuinfo: machine          : CHRP IBM,9406-520
>                                              ^^^^^^
> 
> I think it's plain chrp. Explains also why I can't install partman-prep with 
> anna-install because it's a wrong subarchitecture. But it needs the PREP 
> partition to get it booted (it's also the redhat way). Is there a way to get 
> partman-prep installed?

Indeed, and i believe that is your problem :

        { "CHRP IBM,7028-6C4", "chrp_rs6k" },
        { "CHRP IBM,7029-6E3", "chrp_rs6k" },
        { "CHRP IBM,7038-6M2", "chrp_rs6k" },
        { "CHRP IBM,7039-651", "chrp_rs6k" },
        { "CHRP IBM,7040-681", "chrp_rs6k" },
        { "CHRP IBM,7044-270", "chrp_rs6k" },
        { "CHRP IBM,7046-B50", "chrp_rs6k" },
        { "CHRP IBM,8842-21X", "chrp_rs6k" },
        { "CHRP IBM,9111-520", "chrp_rs6k" },
        { "CHRP IBM,9113-550", "chrp_rs6k" },
        { "CHRP IBM,9117-570", "chrp_rs6k" },
        { "CHRP IBM,9124-720", "chrp_rs6k" },

I think this rule is ridiculous, and that all CHRP IBM can be marked as
chrp_rs6k, or maybe we should change this to chrp_ibm, or something.

Mmm, will investigate and fic.