Subject: Re: Core's, panic's, and hard lockups
To: Henry B. Hotz <hotz@jpl.nasa.gov>
From: Leonard Chung <leonard@ssl.berkeley.edu>
List: port-mac68k
Date: 11/02/2000 17:38:29
--=====================_32250694==_.ALT
Content-Type: text/plain; charset="us-ascii"; format=flowed
At 08:47 AM 11/1/2000, you wrote:
>Not a guarantee because NetBSD operates the controller differently and
>hence has different sensitivities to timing problems. Even the two
>different NetBSD drivers have different sensitivities to what may be a
>timing problem.
Good point.
>They have been looking for a SCSI analyzer for less than $2K used for some
>time now so they can debug the problem. Your setup would be a good test
>case, but I don't think test cases are the only thing needed. The bug is
>believed to be in the MI SCSI code since it affects both the ncr and sbc
>drivers and I think there are some minor problems on other ports which are
>believed to be related.
I heard the problem may be due to having a Quantum drive in the chain.
However, I used to have a Conner 40MB drive previously running 1.4 about 2
years ago. I recall it would core then also, but the problem was nowhere as
severe. Do you think regressing back to the Quantum would fix things a bit?
Or could the added instability be due to changes since then in 1.4.2?
Leonard
--
Leonard Chung - <leonard@ssl.berkeley.edu>
SETI@home - The Search for Extraterrestrial Intelligence @ home
http://www.setiathome.ssl.berkeley.edu
--=====================_32250694==_.ALT
Content-Type: text/html; charset="us-ascii"
<html>
At 08:47 AM 11/1/2000, you wrote:<br>
<blockquote type=cite class=cite cite>Not a guarantee because NetBSD
operates the controller differently and hence has different sensitivities
to timing problems. Even the two different NetBSD drivers have
different sensitivities to what may be a timing
problem.</blockquote><br>
Good point.<br>
<br>
<br>
<br>
<blockquote type=cite class=cite cite>They have been looking for a SCSI
analyzer for less than $2K used for some time now so they can debug the
problem. Your setup would be a good test case, but I don't think
test cases are the only thing needed. The bug is believed to be in
the MI SCSI code since it affects both the ncr and sbc drivers and I
think there are some minor problems on other ports which are believed to
be related.</blockquote><br>
<br>
I heard the problem may be due to having a Quantum drive in the chain.
However, I used to have a Conner 40MB drive previously running 1.4 about
2 years ago. I recall it would core then also, but the problem was
nowhere as severe. Do you think regressing back to the Quantum would fix
things a bit? Or could the added instability be due to changes since then
in 1.4.2?<br>
<br>
Leonard<br>
<br>
<x-sigsep><p></x-sigsep>
--<br>
Leonard Chung - <leonard@ssl.berkeley.edu><br>
SETI@home - The Search for Extraterrestrial Intelligence @ home<br>
<font color="#0000FF"><u><a href="http://www.setiathome.ssl.berkeley.edu/" eudora="autourl">http://www.setiathome.ssl.berkeley.edu</a></font></u></html>
--=====================_32250694==_.ALT--