[cap-talk] Another "core" principle - Brinkmann

[Bill Frantz]

shap at eros-os.com (Jonathan S. Shapiro) on Friday, December 29, 2006 wrote:

> 2. In OS-based systems, virtualization breaks down for memory object
>    capabilities. It is not possible to fully virtualize a "page",
>    because the hardware MMU cannot process entry capabilities.
>
>I can already here Charlie Landau warming up his keyboard on point (2).
>Yes, we can have a keeper on a single page and play games with mappings,
>though I quibble that virtualizing a page by using a page doesn't seem
>very much like virtualization. The real failure is that the keeper lacks
>the ability to respond to a fault by saying "single step the process
>that trapped and invoke me again, meanwhile don't give anybody else
>access to this page".

I think it was the Motorola 88000 which permitted the page fault handler
to return the word being fetched and allow the process to continue
without actually providing a page.  (It could also simulate stores.) 
When the KeyKOS folk were working with this machine, we thought we could
see a way to virtualize pages.  As I remember there were some problems
with some of the instructions.  The double compare and swap could access
two separate pages, which in KeyKOS might have separate keepers.  Having
to invoke two separate keepers to allow the program to proceed kept us
from actually implementing anything.

Cheers - Bill

-----------------------------------------------------------------------
Bill Frantz        | gets() remains as a monument | Periwinkle 
(408)356-8506      | to C's continuing support of | 16345 Englewood Ave
www.pwpconsult.com | buffer overruns.             | Los Gatos, CA 95032