home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.linux
- Path: sparky!uunet!haven.umd.edu!wam.umd.edu!joel
- From: joel@wam.umd.edu (Joel M. Hoffman)
- Subject: Re: DOS emulation
- Message-ID: <1993Jan7.224043.23029@wam.umd.edu>
- Sender: usenet@wam.umd.edu (USENET News system)
- Nntp-Posting-Host: rac2.wam.umd.edu
- Organization: University of Maryland, College Park
- References: <79132@hydra.gatech.EDU> <1993Jan5.174435.3669@wam.umd.edu> <1993Jan7.184329.8050@bernina.ethz.ch>
- Date: Thu, 7 Jan 1993 22:40:43 GMT
- Lines: 37
-
- In article <1993Jan7.184329.8050@bernina.ethz.ch> almesber@nessie.cs.id.ethz.ch (Werner Almesberger) writes:
- >In article <1993Jan5.174435.3669@wam.umd.edu> joel@wam.umd.edu (Joel M. Hoffman) writes:
- >> I see two possibilities: One, giving the DOS program full control of
- >> the screen, for SVGA graphics. The program could do whatever it
- >> wanted, and so the VC code would have to be augmented to save/restore
- >> the complete state of the video card. This would be a good thing,
- >[...]
- >> The only problem with this approach is that I don't think
- >> graphics programs could run in the background.
- >
- >If you mean "detaching from a display", why not ? It's not necessarily
- >different from switching a VC. VCs aren't things that really exist in
- >hardware either.
- >
- >If you mean "switching to a different VC", why not ? The display memory
- >could be mapped to the physical frame buffer while the respective VC is
- >in the "foreground" which is copied to a buffer in user space when
- >switching to a different VC. Now, that buffer memory could be mapped to
- >the display memory and the DOS program could happily continue to
- >scribble to what it thinks is its screen. It would even run faster ;-)
- >
-
- It's not that simple, becuase of the registers. Currently, each VC
- emulates a video buffer, which means that as long as a process is only
- writing to the text part of a video card, everything is fine. X, on
- the other hand, does lots of complex things with the card, and so
- doesn't like the (current) VC scheme. To really get things running,
- so that DOS processes that like to use the entire video card (not to
- mention X) can run either in the background or on a non-selected VC
- (which, as you mention, is basically the same thing), the VC code
- would have to emulate the entire video card, not just the text buffer.
-
- I think that's a good goal to shoot for, for lots of reasons. But I
- certainly don't know enough even to begin.
-
- -Joel
- (joel@wam.umd.edu)
-