On Fri, Oct 27, 2000 at 07:13:17AM -0400, Joseph Sarkes wrote:
> > Ok, so I'm continuing to fight with this P-133 box. I've now added
> > an AHA 1540C (? something close to that) and three drives. This is
> > in addition to the two IDE drives and one IDE cdrom already installed.
> > 
> > I'm having a problem with the 4 gig IBM drive (the newest drive in the
> > box) where it starts making scary clicking noises and subsequently
> > the box crashes. 
> > 
> > You know the noise, the sound of a drive head reseeking on a sector,
> > perhaps reading from a bad sector. Scary stuff, but no messages are
> > returned on the console. 
> 
> My system does exactly the same thing when an ide cable comes
> loose. I can fix it by pressing on the connector. Obviously I
> either need a new cable, or the motherboard ide connnector is
> faulty. UDMA/66 ide does crc as I understand, so I'm not very
> worried when the system comes back by reseating the connector
> but regular ide may not be so forgiving, and may corrupt your
> data doing the same thing as I said above. Good luck.
 
Hmmm.

It might have helped if I had been a bit more specific. The ailing drive
is a SCSI drive. The IDE drives have been running fine. 

I'll check my cabling tonight to see if I actually do have loose connectors.

I'll also check and verify that I have parity enabled. I had to muck with
every knob in sight to get the machine to boot. My drives wouldn't probe
by NetBSD unless I disabled sync transfers in the SCSI card BIOS settings
(what's the proper term for that, anyway?).
 
> > 
> > 
> > Anyway, I can bring the box down with: 
> > # cd /local/src ; tar -cvvf /dev/null sys
> > 
> > The drive makes clicking noises and then shortly (meaning, before the
> > tar finishes) the box freezes. I'm dropped into ddb. Here's a ten finger
> > trace from ddb (partial):
> > 
> > db> trace
> > cpu_Debugger(c02568a0,c0448e14,c41e207c,c41e2150,c0448e001) at cpu_Debugger+0x4
> > aha_done(c0448e00,c41e2150,c0454010) at aha_done+0xae
> > aha_finish_ccbs(c448e00,c0442e60,c444c7f0,c444c7f0,c4460c811)at aha_finish_ccbs+0x142
> > aha_intr(c0448e00) at aha_intr+0x86
> > --- interrupt ---
> > 
> > I'm running 1.5_ALPHA2.
> > 
> > Does this sound like a drive problem or a cabling problem? I suspect the
> > drive is going north (I'm already in the South, so it must be going
> > north). Would an ailing drive cause the aha driver to hang? 
> > 
> > How about if SCSIDEBUG and AHA_DEBUG (I think that's it) were defined
> > when the kernel was built?
> > 
> > *sigh*
> > 
> > Thanks all.
-- 
"A method for inducing cats to exercise consists of directing a beam of
invisible light produced by a hand-held laser apparatus onto the floor ...
in the vicinity of the cat, then moving the laser ... in an irregular way
fascinating to cats,..." -- US patent 5443036, "Method of exercising a cat"