On Wed, 23 Feb 2000, Michael L. Hitch wrote:

> On Wed, 23 Feb 2000, Erik Bertelsen wrote:
> 
> > since last week-end (Friday's (I think) or Saturday's sup) I've seen my 5000/133
> > hang while building NetBSD after a few hour's uptime.
...
> > Any explanation ?
> 
>   Yes - some changes to locore_mips1.S resulted in the possibility of a
> register containing the exception PC to be clobbered, with the result that
> the exception exit returning to random locations.  I've verified this on
> my 5000/25, and have committed a fix.
> 
>   There's the possibility of the same problem in locore_mips3.S, but I
> have not had the chance to verify that yet.

  When I finally took a real look at locore_mips3.S, I realized that it
will not suffer from the same problem as locore_mips1.S.  The first thing
locore_mips3.S does after returning from trap() is to explicitly disable
interrupts by clearing the status register.

--
Michael L. Hitch			mhitch@montana.edu
Computer Consultant
Information Technology Center
Montana State University	Bozeman, MT	USA