AIC7XXX

[Andy Kennedy]

I originally sent this message to the original writer of the aic7xxx
driver.  I was advised by him to get the newest version of the driver, I
downloaded 5.0.19.  The problem wasn't solved, but here is a bit more to
the problem:  The problem, as described below, occurs generally in the
_MORNING_ and not during the heat of the day.  The drives are unactive at
night (due to us being a daytime only station) when most all of the
equipment is off the air.  There is no heat problem.  In the afternoon,
there is heat, MUCH heat, but no problem.

I have an AHA-2940 with the following devices hung from it:  4 Seagate 9.1
GB Barracuda UW, 1 Seagate 9.1 GB Barracuda Fast SCSI II (still upset with
Insight about that), 1 Seagate (Conner) 4.1 GB tape backup, 1 Pioneer 6
Disc changer, also I have a HP CDR which is not connected because when I
do connect it, it causes a kernel panic due to a SCSI0 : time out which
then tries a reset, which never can (why, I don't know) then the kernel
freezes.  I haven't tried to reconnect this device because I cannot aford
having this computer down for the 25 min that it takes to reboot after a
kernel panic (the 9 gigs have to have FSCHK run on them and they are all
full, take a while).  What happens is this:  a request goes out for a file
from one of the HD's.  Normal operation, all other drives on that
particular side of the channel (the UW side) cause the SCSI device to time
out.  On Linux 2.0.27 with the aic7xxx driver version 4.0 I didn't have
this problem (much, only every once in a while).  Nothing has changed
except the driver version.  This didn't just start happening, though.  I
upgraded to Linux 2.0.33 about a week or two ago, then the problem began.
In the beginning, the problem was just with the Pioneer.  I looked through
the sr.c file and found that the MAX_RETRIES was set to 3.  I changed this
to 6 and that gave the changer enough time to change trays.  I have had no
problems from this device since then.  Now, however, the Seagate drives
are going nuts every morning.  What is wrong?  What can I do to correct
the problem?  I know that there is not a problem with the connections
because I have mounted the whole system int two cases inside an 18" rack.
The rack is never disturbed.  Since the problem didn't start untill the
new version of Linux I concluded that it must be a problem with either (a)
some library or new function of Linux 2.0.33 or (b) something new in the
latest aic7xxx driver.  Any Ideas?


wuff,
andy kennedy


To Unsubscribe: send mail to majordomo at FreeBSD.org
with "unsubscribe freebsd-aic7xxx" in the body of the message