Kernels I built for a NetBSD/pmax snapshot seem to repeatably hang
processes when I type "sync".  This causes shutdown and reboot to
fail.  reboot -n, however, works.  This happened for kernels built
from march 14/15 source.  It didn't happen for kernels built on march
5 from March 4 sup source.

I only have FFS and NFS filesystems mounted.  ps shows that update
and any sync processes are waiting on "thrd_s" (thread_sleep?)

Has anyone seen anything similar on this or any other platform?  I
haven't touched any of the pmax kernel in some time. I've also
upgraded the version of GNU ld I'm using, but pmax kernels are still
linked with a binutils-2.6 kernel, so I don't see anything
port-specific thats' changed.  I tend to blame some of the recent VM
changes, but it's just a guess.

thanks
--Jonathan

(ps: can we either limit waitchannel names to something that fit in
ps's "WCHAN" column, or fix ps -awwl to print the entire waitchannel
name?)