On Tue, Dec 17, 2002 at 11:20:40AM +0200, Teemu Rinta-aho wrote:
> On Mon, 16 Dec 2002, Manuel Bouyer wrote:
> 
> > > when the problem occurs, there are no wd0 or wd1 detected, only
> >
> > wd0 and wd1 *are* detected, otherwise the real drive woudln't end on wd2
> 
> How come there are no prints on the screen? Can I add some
> options somewhere to make it visible?

You should have them visible after:
pciide0: secondary channel wired to compatibility mode

> 
> > OK, so it finds 2 gost drives on the second channel of the VIA ide controller.
> > Do you have a way to disable it in BIOS ?
> 
> Is this ghost drive finding a known feature?

This happened in the past, and I made changes to the probe logic that fixed
it at that time. It seems it didn't catch all cases.

> I can disable the second
> channel on the VIA. I will try it tonight. Thanks for pointing this out,
> I was only concentrating on the Promise and didn't think of this.
> 
> > You can hardwire wd0 to pciide1 channel 0 in the kernel config file.
> > This means you'll have to create 2 entries, pciide0 and pciide1 in the
> > config file, instead of pciide*
> 
> OK, so you consider it as a hardware bug? I will try the patch Alexander

No, it a bug in the probe. But as it's hard to reproduce, it will also
probably be hard to fix.

--
Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
     NetBSD: 23 ans d'experience feront toujours la difference
--