I've seen a similar panic in a different context - I have a HyperSPARC
system here that I've been testing stuff for Paul on. It did the same thing
until the floppy stuff was removed.

By contrast, I booted a kernel made from yesterday's sources on my
SPARCclassic build system with no problems; it's busy building today's sup
right now.

	just another data point,

	Erik <fair@clock.org>