> 
> > 1) Minimal copying (all page moves would be virtual.)
> 
> On some systems (typically those using inverted page tables or
> virtually-addressed caches; I'm thinking in particular of the IBM RT
> and the HP PA...), random page-flipping like this can be more
> expensive than copying, because you can't have a page mapped at
> "random" virtual addresses at the same time in different processes, so
> you have to flush one mapping, flush/purge any references in the cache
> to that page, and take a sharing fault when you need the original
> mapping again.
> 
Hmmm... that is unfortunate.  However, what I was really implying
that we would use vm_page_rename to "virtually" copy the pages, as
opposed to remapping (or copying) the pages.  The pages in the
filesystem would not necessarily be mapped into a process's space (but
might be directly.)  (One of the kind-of evil things about MFS is that
page copying is being done when not needed, as opposed to vm_page_rename
operations.)

John