[ follow-ups to port-arm? ]

On Thursday, October 16, 2003, at 07:41  AM, Richard Earnshaw wrote:

> But then, so would having to decode every Thumb instruction.

Well, I guess it basically boils down to "which one is faster?"

Maybe there's a compromise possible, here... on permission faults, you 
still always know it was a write.  Translation faults is the tricky 
case.

         -- Jason R. Thorpe <thorpej@wasabisystems.com>