Cool , sounds exciting.
Kevin Lawton wrote:
> - "Kevin P. Lawton" <[EMAIL PROTECTED]>: Thu Apr 26 18:00:12 EDT 2001
> More work on dt-testbed/proto4. Implemented reverse address lookup
> table (tcode address to instruction address) capability. This is
> needed for proper exception handling when the code is integrated
> into the real VM environment.
>
> Next, I'm going to do some more instrumentation and analysis of the
> storage efficiencies of the constructs which are used to store
> tcode and the forward and reverse lookup tables.
>
> Some analysis and tweaking should prove useful. Then on to
> tying up some loose ends (stuff I marked with '+++' in the code),
> then moving on to integration of prototype into the main code.
>
> I need to update the Plex86 Interals Guide to reflect some of
> the details and changes I've made. But I'm going to wait until
> I get things tweaked first.
>
> Implementing the storage and lookup structures for a DT (dynamic translation)
> technique (even this brain-dead quasi-DT, only translate a limited set of
> instructions) is quite a balancing act of efficiencies of storage space
> and access time etc. And making it work well in a page-centric way
> so that translated code can be cached, and efficiently invalidated/revalidated
> across context switches in the guest OS. And...
>
> Good stuff. Can't wait to get to integration.
>
>
> -Kevin
>
> --
> =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
> Kevin Lawton [EMAIL PROTECTED]
> MandrakeSoft, Inc. Plex86 developer
> http://www.linux-mandrake.com/ http://www.plex86.org/
--
Drew Northup, N1XIM
