Chris Torek wrote:
> We have seen X11R5 wedge under both SunOS and BSD; this sounds like something
> identical.  If you kill the server, everything is OK, so it seems like a bug
> in the server - except that when we traced this under SunOS, it appeared to
> be a kernel bug (but then why did it persist under BSD?).

No - this is far nastier.  It happened again tonight (I was testing LBX, the
Low-Bandwidth X extension).

The machine dies.  At first I thought it was just wedged.  Then I got a bright
idea (for once (-: ).  I thought, "Gee, last time I did a blind `L1-A' and a
`reset', and it came back".  Now I wondered if maybe I was already at the
PROM, and didn't realize it (screen goes black, remember).  So I (blindly)
typed (just) "boot", and voila, lo and behold my screen came back on (thank
GOODNESS these old PROMs don't do an automatic "reset" on "boot") and I found
something rather grisly still on-screen:

	Watchdog reset
	Instruction Access Exception
	ok

Holy double bus errors, Batman!

I can see it now "R6 Xsun server - first `killer' app for NetBSD/SPARC ... 
Film at eleven" ...  (-:

Being on cranial auto-pilot, I then typed

	ok sync

and was reminded "This ain't SunOS, bucko":

	No callback routine has been installed
	ok

As I said the first time:

> P.S. Can one boot a 1.0 system under DDB?  If so, what is required to do so?
>      (ddb(4) isn't terribly forthcoming on the subject)

	- Greg