> > Other suggestions.  (Note I'll be rummaging around other platform pmap's to 
> > see how things are done)
> 
> Make use of the new pmap_update() calls that Jason has added to defer 
> options when possible until they are really needed (hopefully when the 
> pmap in question isn't the current one).  Build work lists when we can and 
> then manage them during the update calls.
> 
> Investigate caching the L1/2 tables to improve performance.  May require 
> restructuring most of the code...
> 
> Look at the way i386 pmap organises things -- it should be possible to 
> significantly reduce the number of calls to pmap_pde/pmap_pte (which is 
> currently huge)

Implement the pmap stealing functions so that we stand less chance of 
failing in critical conditions.

Fix the code so that it can fail (rather than panic) when failure is now 
permitted.

Fix the L1/L2 page reclamation routines so that when a process is swapped 
out it really is completely swapped out.

R.