From:	"Charles Sandmann" <sandmann AT clio DOT rice DOT edu>
Newsgroups:	comp.os.msdos.djgpp
Subject:	Re: CWSDPMI - Found solution but, not the problem...
Date:	Sun, 7 Jun 1998 12:42:19
Organization:	Aspen Technology, Inc.
Lines:	22
Message-ID:	<357a8aab.sandmann@clio.rice.edu>
References:	<35794471 DOT 50439830 AT news DOT telepac DOT pt>
Reply-To:	sandmann AT clio DOT rice DOT edu
NNTP-Posting-Host:	dmcap2.aco.aspentech.com
To:	djgpp AT delorie DOT com
DJ-Gateway:	from newsgroup comp.os.msdos.djgpp

> Some days ago, i post a problem because i couldn't map the video
> memory while using any vbe/af implementation. I was using CWSDPMI R4
> but, two days ago, i was having problems using vesa modes provided by
> sdd 6.52 in dos mode with the new cpu k6 but for rest i start playing
> quake I, when i changed from dos into win98 (also i changed my SO in
> the same time i changed the cpu) i was talking about djgpp and then, i
> realised that quake I was made with djgpp so it must have been using
> CWSDPMI but the release was R1... I changed CWSDPMI R4 by CWSDPMI R1
> and i have no more trouble...

There were some changes in the device mapping code between R1 and R4 -
some to make the code force following the DPMI specification more closely
a requirement, some other bug fixes for devices in certain memory ranges.

For example, if the device window returned is in the lower 1Mb range, 
R1 would allow you to map it (incorrectly, since it's already mapped...)
while R4 shows an error status.

While there might be an error in R4 (who knows?) I suspect if you look 
closely at the return values from each call and compare them to the DPMI
specification, you will find your code needs a slight modification to 
work properly.

- Raw text -