It does indeed appear to be the ohci driver that's screwing up the
interrrupts somehow.

Without it in the kernel there are no "stray 6600 irq ..." messages
while probing the scsibus* devices and as a result the isp driver doesn't
get into a snit as I reported previously (as quoted below).

The NFS root mount doesn't quite work yet, though I don't really care to
try to fix it right away as I can now at least boot the INSTALL kernel
A-OK and from there I should be able to put a local root FS on a SCSI
disk and be able to work from that.

	md0: internal 4650 KB image area
	root on md0a dumps on md0b
	no file system for md0 (dev 0x600)
	cannot mount root, error = 79
	root device (default md0a): tlp0
	dump device: none
	file system (default generic): nfs
	root on tlp0
	nfs_boot: trying DHCP/BOOTP
	stray 6600 irq 16
	nfs_boot: timeout...
	nfs_boot: timeout...
	nfs_boot: timeout...
	cannot mount root, error = 60
	root device (default md0a): tlp0
	dump device: none
	file system (default nfs): 
	root on tlp0
	nfs_boot: trying DHCP/BOOTP
	nfs_boot: timeout...
	nfs_boot: timeout...
	nfs_boot: timeout...
	cannot mount root, error = 60
	root device (default md0a): 

Note that irq 16 logged as stray above is apparently from isp0.  I'm
quite used to seeing the odd stray interrupt from an isp(4) device go
unclaimed though -- it happens regularly on my AS4000 with a Qlogic
1020 controller, but it doesn't seem to cause any problems and the disk
attached to that controller can be accessed OK:

	isp0 at pci1 dev 0 function 0: QLogic 1020 Ultra Wide SCSI HBA
	isp0: interrupting at kn300 irq 40
	scsibus1 at isp0: 16 targets, 8 luns per target
	...
	scsibus1: waiting 2 seconds for devices to settle...
	stray kn300 irq 40
	sd0 at scsibus1 target 0 lun 0: <DEC, RZ29B    (C) DEC, 0016> SCSI2 0/direct fixed
	sd0: 4091 MB, 3708 cyl, 20 head, 113 sec, 512 bytes/sect x 8380080 sectors
	stray kn300 irq 40
	sd0: sync (100.0ns offset 12), 16-bit (20.000MB/s) transfers, tagged queueing
	stray kn300 irq 40

(Occasionally I even see a stray IRQ logged for com0 too.)

I'll probably still try pulling a bunch of the more recent isp(4)
changes up to my netbsd-1-6 tree if possible, but I think the removal of
the ohci and other USB drivers more or less solves all of the immediate
problems and I should now be able to get an install done on the onboard
SCSI disks and be ready to go from there.


[ On Monday, September 20, 2004 at 19:26:11 (-0400), Greg A. Woods wrote: ]
> Subject: Re: ES40 nearly boots 1.6.x now....
>
...
> isp0 at pci0 dev 3 function 0: QLogic Dual Port FC-AL and 2Gbps Fabric HBA
> isp0: interrupting at dec 6600 irq 16
> isp0: bad execution throttle of 0- using 16
...
> ohci0 at pci0 dev 19 function 0: Acer Labs M5237 USB Host Controller (rev. 0x03)
> ohci0: interrupting at isa irq 10
> ohci0: OHCI version 1.0, legacy support
...
> scsibus2: waiting 2 seconds for devices to settle...
> stray 6600 irq 16
> stray 6600 irq 16
> stray 6600 irq 16
> stray 6600 irq 16
> stray 6600 irq 16; stopped logging
> isp0: Interrupting Mailbox Command (0x69) Timeout
> isp0: Mailbox Command 'GET FW STATE' failed (TIMEOUT)
> isp0: Interrupting Mailbox Command (0x69) Timeout
> isp0: Mailbox Command 'GET FW STATE' failed (TIMEOUT)

-- 
						Greg A. Woods

+1 416 218-0098                  VE3TCP            RoboHack <woods@robohack.ca>
Planix, Inc. <woods@planix.com>          Secrets of the Weird <woods@weird.com>