SPARCs Re: tlug: Digital Nursery

[Tim Meggs <tim@example.com>]

--------------------------------------------------------
tlug note from Tim Meggs <tim@example.com>
--------------------------------------------------------
From: "Stephen J. Turnbull" <turnbull@example.com>
Subject: Re: tlug: Digital Nursery 
Date: Fri, 27 Jun 1997 16:51:50 +0900

> Um, why dontcha talk to John Little?  Sounds like a perfect
> application for an excess Sparc muthaboad.  I might be willing to give 
> mine up....  But then I will need it in October.  (^^;

Speaking of SPARCs, that reminds me of the qn. i have been meaning to send 
to the list for the last week, or so.....

I have a SLC at home (yes, it is a John Little special:), but recently it has
been giving a WATCHDOG ERROR (at the h/w, or f/w level it seems) and dumping
me out of Linux, back to the ROM BOOT prompt.

The wierd thing is that it *only* happens the first time I boot the machine
from power on - normally withing say 5-10mins.   If after beeing dumped back 
to the ROM prompt I reboot with:  bsd(), or boot sd()  then the system will 
run fine for hours [note I don't leave it on 24hrs/day as it heats up the 
room too much and is noisey so several hours is the most i've ever tested]

There were a few things that I did just before I observed the problem:

*  Transfered the system onto a different physical HD.

*  Upgraded the kernel from 2.0.18, to 2.0.27

That is when I first observed the problem.   But as I was planning to go to
2.0.30, I thought that might clear it up.   But there was no change.
Currently i am running 2.0.30.

Do you think I missed doing something in the kernel upgrade, or HD swap over?

FWIW I have used check mem at the boot prompt (after the crash), but it said 
all was OK.   Maybe I should run it shortly after power-on.

What I have not yet tried is booting the SLC with the HD off, then manually
doing the bsd() from the ROM prompt to see if it does/doesn't crash.

The actual message I get when it crashes varies, two examples are below:

* Watchdog reset Window Underflow Type

* Watchdog reset Illegal Instruction Type

Do you think it could be a h/w problem that is only observed withing 5, or so
mins of start up (unlikely).   Or do you think that maybe there is a WDT
flag/reg. somewhere that is not being correctly initialised with my std boot
proceedure - but is set/reset to the correct state after the crash??

tim-confused :/
-----------------------------------------------------------------
a word from the sponsor will appear below
-----------------------------------------------------------------
The TLUG mailing list is proudly sponsored by TWICS - Japan's First
Public-Access Internet System.  Now offering 20,000 yen/year flat
rate Internet access with no time charges.  Full line of corporate
Internet and intranet products are available.   info@example.com
Tel: 03-3351-5977   Fax: 03-3353-6096