Drew Northup wrote:

> This means that we are going to have to do
> one of several things: deal with the Win32 swapping algorithm & hope it
> doesn't kill us, write a user-space<-->hardware-space memory-mapping/locking
> driver/program, ...

Whatever physical memory a guest kernel driver uses, and needs
to pass the physical address of outside guest space, should be
locked down using the native guest OS services.

-Kevin

-- 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Kevin Lawton                        [EMAIL PROTECTED]
MandrakeSoft, Inc.                  Plex86 developer
http://www.linux-mandrake.com/      http://www.plex86.org/

Reply via email to