netbsd-4 on i386, source as of about a week ago.

It has dropped to db> from several locations, each time with a message like:

kernel: supervisor trap double fault, code = 0
Stopped in pid 759.1 (nmbd) at netbsd:Xintr_lapic_ltimer+0x7: \
    mov   %gs,0(%esp)

It isn't just in nmbd as noted above -- I've had it drop in siside0 and in
other user apps.  The backtrace shows no steps other than the same call
above.

Is this something that has been addressed by recent updates or is this worth
chasing?


Also, I booted this system with a non-ACPI kernel and the fxp0's wouldn't
come out with a dev timed out message.

peter