[comp.os.mach] 88k version of Mach?

sritchie@cs.ubc.ca (Stuart Ritchie) (06/28/91)

I recall reading sometime ago about someone doing an 88k port of Mach.

Just starting on some 88k kernel hacking myself, I would really like
to get in touch with people who have 88k experience down at the
kernel level.  I have many questions...

For example, how well does the "shootdown" algorithm work when applied to
the 88k?  Is context switching expensive (lots of registers to save and
restore)?  In MP environments, how expensive is cache snooping compared to
write-through?  How well does migrating tasks and threads from CPU to CPU
work (is it worthwhile)?

Eagerly awaiting your postings and replies,

Stuart