Subject: Re: SCSI reset hang on sparc64 1.5X and sunos binary
To: None <jdc@coris.demon.co.uk, port-sparc64@netbsd.org>
From: None <eeh@netbsd.org>
List: port-sparc64
Date: 09/23/2001 17:55:46
|
| > 	I've just had a couple of hangs on my Ultra1 (the first time
| > 	its misbehaved under 1.5X).
| > 
| > 	Both times were while running the SunOS 4.x netscape binary, and
| > 	both resulted in a complete hang short of L1+A
| > 
| > 	The last dmesg entries were:
|
| Hmm, I was wondering if my Ultra 1 had a hardware problem.  I often (every
| couple of days) get a lockup, with the trace :
|
| 	zsc_intr_hard
| 	zs_hard
| 	intr_list_handler
| 	sparc_intr_retry
| 	pmap_enter
| 	uvm_fault
| 	data_access_error
| 	data_error
|
| I don't get anything written to the message buffer.  Processes running keep
| going, but no disk access is possible.  When I try an run through the
| debugger, I end up with a completely wedged machine.  All I've managed to
| see is that it's continuously looping through the serial interrupt routines.

Remember, hitting a key to enter DDB goes through the serial interrupt handler.

However, a data_access_error is a likely indication of bad memory or another 
hardware error.

Eduardo