5.3 beta1 is no go for me.

[Wilko Bulte]

On Tue, Aug 31, 2004 at 11:22:42AM +0200, Bernd Walter wrote..
> On Mon, Aug 30, 2004 at 10:56:02AM -0400, John Baldwin wrote:
> > On Monday 30 August 2004 10:06 am, Poul-Henning Kamp wrote:
> > > In message <20040830122945.GL59909 at cicely12.cicely.de>, Bernd Walter
> > > writes:
> > >
> > > It does boot the 5.0-RC1 CD I have lying around:
> > 
> > I "fixed" the eisa code to actually include eisaconf.c on Alpha in between 5.1 
> > and 5.2 I think, so the problem is likely in the eisa code rather than the 
> > PCI bus numbering.
> 
> I know the EISA addition very well - I don't think anyone ever had been
> running an AS1000A since that.
> But after reducing the probed slot numbers Poul gets over the EISA part
> and there's no other point where EISA by itself should have influences.
> On the other hand we are tracking 13 month of -current development.
> Not unlikely that unrelated things have changed that broke this
> platform - don't you think that holes in PCI numbering could be special
> to that platform and there were enough PCI and bus changes in the
> last 13 month?
> I can only judge the situation from the boot output and what might
> come next - more guessing right now than anything else.
> That includes probing pci2 and isa, but not eisa.
> 
> Does anyone know how to find out which code triggered the trap,
> whithout spreading printf?
> I'm not very familar with SRM debugging features.

I am not aware of any SRM debugging functionality so..

-- 
Wilko Bulte				wilko at FreeBSD.org