Looks good..., now I just need to get my system back up.....
Kevin Lawton wrote:
> While I'm thinking about performance...
>
> Some stuff I'll be working on next:
>
> - When the emulation attempts native segment:offset accesses,
> it always copies (and modifies slightly) the guest segment
> to the monitor GDT. There's no need to do this for every access.
> Just manage a virtualized copy of each guest segment in empty monitor
> GDT slots for guest segment reloads. So can attempt native access
> at will, with no additional descriptor write.
>
> Some other stuff:
>
> - Three words which describe latched graphics mode (640x480x16 VGA).
> ACK, ACK, ACK!!! :^) I suppose I could move the VGA latch logic
> over to monitor space - that'd help. But we'll need something
> better in the graphics department anyways, for resolution/colordepths.
My Junior Design project is going to be a VGA card (hopefully PCI), so I'll
use the downtime I've got until I can build the thing trying the stuff out in
the VGA/SVGA department. I'd like to do the VGA/SVGA/XGA emulation as a
separate thread from the host/monitor stuff for speed reasons--and it might be
a good idea to do disk I/O the same way--any comments?
>
>
> - Special guest OS specific drivers. (video/disk/network) This might
> be a good time to get some of these subprojects fired up. Anybody
> interested in this?
It should be noted the for compatability we should keep things as industry
standard as possible--keeping all of the default behavior modes intact, for
instance.
>
>
> -Kevin
>
> --
> =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
> Kevin Lawton [EMAIL PROTECTED]
> MandrakeSoft, Inc. Plex86 developer
> http://www.linux-mandrake.com/ http://www.plex86.org/
