Newsgroups: comp.windows.ms
Path: utzoo!utgpu!watserv1!sunee!gpsteffl
From: gpsteffl@sunee.waterloo.edu (Glenn Patrick Steffler)
Subject: Re: Win3.0 - More questions
Message-ID: <1990Jun14.025353.22089@sunee.waterloo.edu>
Organization: Gerbils On Speed Inc.
References: <9906@rouge.usl.edu> <1990Jun12.140525.7945@cbnewsk.att.com> <1990Jun13.152605.14354@sj.ate.slb.com>
Date: Thu, 14 Jun 90 02:53:53 GMT
Lines: 30

In article <1990Jun13.152605.14354@sj.ate.slb.com> poffen@sj.ate.slb.com (Russ Poffenberger) writes:
>
>The statement that non-windows apps must reside entirely in RAM is probably
>correct. It may be able to swap an ENTIRE virtual session to virtual memory,
>but I don't know for sure.

The only memory which needs to remain resident for a DOS app would be
code which handles interrupts.  Otherwise it could be paged (of
course!).  Since Win386 handles all interrupts itself, and the DOS app
only sees the virtual interrupt, it would follow that most of the
memory allocated to the DOS app would not need to be resident.

The number of DOS apps which can be run would most likely be limited to
the amount of memory required to store the real-mode interrupt vector
handling code for each app.  DOS need not appear in every VM as a
memory hunk, rather it could appear in only one position in memory with
some clever memory mapping (logical mapping) on the 386.  I understand
this to be the case.  Otherwise 16Kb for every DOS app would need to
remain resident in the memory.

>Russ Poffenberger               DOMAIN: poffen@sj.ate.slb.com
>Schlumberger Technologies       UUCP:   {uunet,decwrl,amdahl}!sjsca4!poffen
>1601 Technology Drive		CIS:	72401,276
>San Jose, Ca. 95110             (408)437-5254

-- 
Co-Op    _____           "Bo doesn't know software" - George Brett
Scum   _/|__Q_\___ 
U of   | ww--+----#\    "Catch the mystery catch the spit!" - Tom Saywer (Rush)
Loo'91 ~~()~~~~~()~~                                     Glenn Patrick Steffler