

- the compiler does not 8-byte align vlong/uvlongs.  Because of this
   there is some code in l.s that writes these in two-halves rather
   than a single 8-byte write.  Should be fixed when the compiler is
   updated.
- we dont make use of the FP dirty bits.  Doing so could possibly
   simplify FP active/inactive handling.
- when we go MP, are we sharing mmu ctx across processors?  If
   so we will have to be more careful during ctx dealloc!
- still using TSO.  should eventually support a higher performance ordering.

- seperating traps and interrupts would be more efficient --
   especially fast-tracking the linear mapping of RAM above KZERO
   and the remapping of 0x60 interrupts to softints.
	- should have a fast trap handler for 0x60, remapping softint.
	- should have trap entries for each different way of getting
	   a fault addr, jump to a common handler which replaces
	   kernel tlb on fast path, otherwise vectors to normal handler.
- traps are a mess right now.  rethink after we get everything figured out.
- traplinks are very simplistic.  We could tweak some performance here
   if we needed it.
- scsi stuff isnt completely tested, also not using wide scsi features
   now.  Needs cleanup.


TODO
- true 64bit compiler support and kernel support.
- graphics, keyboard, normal console.
- SMP
