home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!think.com!yale.edu!ira.uka.de!math.fu-berlin.de!uni-paderborn.de!chandler
- From: chandler@uni-paderborn.de (Martin Grote)
- Newsgroups: comp.sys.amiga.programmer
- Subject: Re: physical memory protection with MMU
- Date: 12 Nov 1992 17:28:37 GMT
- Organization: Uni-GH Paderborn, Germany
- Lines: 36
- Distribution: world
- Message-ID: <1du485INNdbn@uni-paderborn.de>
- References: <1992Oct26.162400@uni-paderborn.de> <heinz.04a3@edohwg.adsp.sub.org> <1992Nov10.154421.18725@cbis.ece.drexel.edu> <1dtp0rINN9jk@uni-paderborn.de>
- NNTP-Posting-Host: greco.uni-paderborn.de
- Keywords: mmu, memory, task
-
- tron@uni-paderborn.de (Matthias Scheler) writes:
- >> [stuff emphasizing memory protection deleted]
- > It would be nice, but you would loose 99.9999% of the software.
-
- So here we are at the beginning again. My original suggestion was:
- > As the memory protection would have to be optional anyway because not all
- > machines have an MMU, make it optional for all users!
- > [...]
- > Just open a window asking the user what to do, e.g.
- > 'Kill task - Ignore forever - Ignore'.
- > [...]
-
- > The advantages would be that all new software would definitely have to avoid
- > MMU hits and the user could still use old buggy software if she really needs
- > to ...
-
-
- Besides, you write:
- > you would loose 99.9999% of the software
- ^^^^^^^^
- Hmm. Ever went to a statistics lesson here? :^>
- So let's say 90%. But as I wrote before:
-
- > The advantages would be that all new software would definitely have to avoid
- > MMU hits and the user could still use old buggy software if she really needs
- > to ...
-
- However, to focus on my original intention again:
- - is it impossible / too difficult?
- - any disadvantages (does it have slow down the system)?
- - if not, how about implementing it?
-
- --
- Martin Grote
- chandler@uni-paderborn.de
-
-