That is very very strange. I have no idea.

It's strange on two counts- one that it goes to the empty controller. And,
second, that you get this aborted command.

This looks more like a misdirected interrupt (kn300) than a Qlogic
problem. Can you build a DEBUG kernel and publish the full boot messages?

You know, it's too bad that Jason nuked the messages I used to have in the
kernel- the stray interrupts could then be specifically bound to a
specific bus and PCI slot.


On Mon, 13 Dec 1999, David Hopper wrote:

> I have two Qlogic 1040B's:  an isp0 with three Seagate 4-gig drives, and
> an isp1 that isn't hooked up to anything.  The system is on the first
> disk, sd0a, on isp0, and I'm trying to stripe the remaining two disks at
> sd1a and sd2a, again, all three disks on isp0.  I get the following
> errors when I run raidctl -C /etc/raidcfg raid0:
> 
> isp1: command for target 4 lun 0 was aborted
> stray kn300 irq 8
> [...]
> ad infinitum: the process continues to attempt the RAID creation,
> failing each time on the empty controller.  Why is it hitting this
> controller??
> 
> Here's my /etc/raidcfg:
> 
> START disks
> /dev/sd1a
> /dev/sd2a
> 
> START layout
> 32 1 1 0
> 
> START queue
> fifo 100
> 
> I'm using the platform .997, Core 1.11 Qlogic driver from the 12/11/99
> sources, and a hand-rolled kernel with the DEC_KN300 and RAID defines.
> 
> Any thoughts?
> dhop
> 
> 
> 
>