Subject: Re: Lockups on 1.3D sparc & i386
To: William O Ferry <WOFerry+@CMU.EDU>
From: Scott Ellis <>
List: current-users
Date: 02/27/1998 09:29:19
On Thu, Feb 26, 1998 at 09:26:33PM -0500, William O Ferry wrote:
>     I've built new kernels from sources sup'ed on 2/24, for a SPARC and
> an i386, both with and without UVM.  Since compiling these kernels, I
> have been experiencing rather regular lockups on both my SPARC and my
> i386.  Unfortunately I cannot give much information on this problem. 
> The boxes seem to lock while sitting completely idle, it does not seem
> to be caused by a certain amount of load or memory usage or anything
> like that.  The SPARC locks up hard enough that even Stop-A (L1-A) does
> not bring up the boot prompt.  The PC stops responding to pings or such,
> but typing on the keyboard un-blanks the pcvt console, though no text
> appears on the screen, and the caps lock light and such don't function
> (i.e. it's completely locked up, but it did unblank when I typed a key. 
> It did not reblank afterwards).

I had a similar experience with a UVM kernel on i386.  The machine was
essentially idle (ircd and mail were the only 'active' processes), and
had been up for around 2 days.  I have a Mac connected to the serial port,
and had accidentally been spamming the serial port (i.e., getty) with
Appletalk packets.  When I realized what was happening, I couldn't login
via the serial port, and went to the console to find that the keyboard
was dead (i.e., capslock didn't work), and xdm was locked up (X was dead,
cursor didn't move, etc.).  On reboot, I couldn't find anything of
interest in the logs to point out what went wrong. ;-/

This was on NetBSD/i386-current as of around 2/19/98, UVM and PMAP_NEW.

  //    Scott Ellis     //   //   //
// WARNING: This signature warps  time and space in its vicinity    //