First off, the settings assigned to the drive are calculated bitwise, so
that
having a jumper on 0 gives the device SCSI ID 1, having it on 1 gives it
ID 2, and having it on 2 gives it ID 4.  Since you start out with
probe-scsi and
dmesg showing the disk at target 5, I assume there are jumpers on
positions 0 and 2 (4 + 1 = 5). I would try "boot /sbus/esp/sd@5,0" as my
first guess, as I don't remember how Sun assigns the numbers in "boot
diskX" to SCSI targets, though I know it's not just the ID number.

Maybe someone else can help by refreshing my memory? :-)

Matt Gozo

Peter W Waterland wrote:
> 
> Hi there :) I am new to sparcs maybe someone can
> help me out.TIA..
> 
> I have a sparc IPC w/ 36mb of ram. rom rev 2.6
> serial #1264830
> sbus slot 2 has a cgthree on it.
> 
> My main problem is this: i cant boot from the
> internal 1 gig scsi after a successful install
> of openbsd (i know its not netbsd..but if you
> can help me out..i will gladly try netbsd in the
> IPX i have) via FTP from my freebsd box.  I can
> successfully mount sd0 to check if the files are
> actually there with the install floppy..and they
> are. Which makes me think the drive is ok.
> Dmesg from the boot floppy brings this up:
> sd0 at scsibus0 target 5 lun 0: <HP C2247-300,
> 0BE4> scsi2 0/direct fixed.
> sd0: 1003mb, 2051cyl,13 head, 77 sec 512 bytes
> persec, 2054864 sec total.
> 
> 'probe-scsi' at the openboot version 2 prom does
> this:
> target 5
> unit 0 disk HP C2247-300 0BE4
> 
> so even the openboot prom sees it :)
> 
> however..when i try 'boot disk' or 'boot
> disk0/1/2/3' or boot /sbus/esp/sd@3,0 or 0,0 or
> a few other wild stabs in the dark i get:
> 
> boot device: /sbus/esp@0,800000/sd@3,0   file
> and args: bsd
> scsi device 3,0 is not responding
> cant open boot device.
> 
> This is obviously reducing my enjoyment of the
> little lunchbox :). 'test diskx' yields a return
> code of -1.
> 
> I have no experience of scsi before..but the
> drive is connected directly to the mobo in the
> sparc. The ID on the hdd can be changed by
> jumpers as far as i can see..altering it from
> 0/1/2 just seems to move the buggering target
> around on 'probe-scsi' and dmesg.
> 
> Please someone put me out of my misery on this
> and i'll buy /send you a pint ;)
> 
> Pete.