Subject: Re: Boot failure with Adaptec 2940UW present in NetBSD/macppc system
To: None <mw@blobulent.com>
From: Robert Elz <kre@munnari.OZ.AU>
List: tech-kern
Date: 10/13/2000 13:26:45
    Date:        Thu, 12 Oct 2000 22:08:49 -0400
    From:        Michael Wolfson <mw@costello.cnf.cornell.edu>
    Message-ID:  <v0313030fb60c1de7de3c@[10.0.1.2]>

  | Once it gets to the point of probing the SCSI busses for
  | devices, it always fails in pretty much the same way:

I am having very similar problems(timed out while idle) on an alpha,
but only when the controller gets busy.   See PR kern/11180

I have now pretty much determined that it is likely to be some kind of
timimg issue - I have AHC_DEBUG defined (which won't be easy for you to
do without being able to compile a new kernel) (0x19F for those who know)
and have been thrashing the controller as much as is possible with all
that debug noise coming out a 9600bps serial console.  No problems at
all in this setup (other than that the debug makes the system impossibly
slow of course).   I have managed to load about 800MB onto a raid mirror
(so 800MB onto each of two drives) in this setup - doing it with no
debug enabled and it died after about 8Mb...

kre