>> NMI: system interrupts: 40000000<VME=0,SBUS=0,ME>

> Well - I do see it on my LX as well, close in time to crashes with
> "lockmgr: no context".  [...]  Martin Husemann <martin@duskware.de>
> said he suspected hardware problems like an SBus card not seated
> firmly in the slot, and I was inclined to agree although I already
> had re-seated everything, but now doubts arise.

Yeah.  When I was getting it, it was in a hardware configuration that
didn't have _anything_ in _any_ SBus slots - the only hardware involved
was what was on the motherboard.

Also, my machine didn't crash.  I could get that message, and provided
I didn't keep playing with the framebuffer, everything could run
normally, apparently indefinitely.  Even using the framebuffer usually
worked, and when it didn't, it crashed only the userland program, not
the whole machine.

Weird.  Looking at intreg.h, the ME bit (the only bit that's set) is
commented as "Module Error (async)".  There is code in intr.c to handle
module errors if a module error handler has been set, but this is not
done at all in the rev I'm running and unless I've missed something is
not done for an LX even in -current (it's done only for the Viking,
which appears to equal the SuperSPARC).  Otherwise module errors,
whatever they are, are ignored.

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B