On Wed, Dec 10, 2003 at 06:15:38PM -0500, Sean Davis wrote:
> Now if I could only figure out what the root cause of the "sleep sleeps
> forever" bug, but I have no idea where to start looking for that.

This is pretty hard to debug, since it only happens rarely and you need
working DDB access and someone to diagnose it on a machine where it happens.

The current theory (backed by some DDB'ing/instrumentation of a machine in
broken state quite some time ago) is: kern_timeout.c:softclock() does not get
called, despite kern_clock.c:hardclock() doing
softintr_schedule(softclock_si) in the broken state.

I have no idea why that would happen though.

Martin