I have got a lot of trouble with 1.2G, and I'm wondering of there is
someone else having these troubles. I get very often panics because 
of "access fault without process" and such things; which I have never
gotten before. I have compiled without my checkins the latest weeks
and still have the problem, so I think that there are a change to the
common code that has triggered a bug in the vax code.
Anyway; the reason to this fault is that the kernel tries to tsleep()
when it's malloc()'ing at interrupt level. This in turn seems to be
because of the map where it tries to malloc() from is locked.
The reason I write this is that I have to go away for 1-2 weeks, and
won't have time to try to find the bug. So; I hope I have provided 
anyone who wants to search for it with enough info :-)

-- Ragge