Mail Archives: djgpp/1999/08/10/19:06:04
From: | "Stefan Viljoen" <rylan AT intekom DOT com>
|
Newsgroups: | comp.os.msdos.djgpp
|
Subject: | Re: calling a pointer
|
Date: | Mon, 9 Aug 1999 23:26:01 +0300
|
Organization: | The South African Internet Exchange
|
Lines: | 32
|
Message-ID: | <7onh9l$3ne$6@ctb-nnrp1.saix.net>
|
References: | <37ADD81B DOT 4840 AT iquest DOT net>
|
NNTP-Posting-Host: | pc36-01-p09.nt.saix.net
|
X-Trace: | ctb-nnrp1.saix.net 934234229 3822 155.239.194.9 (9 Aug 1999 21:30:29 GMT)
|
X-Complaints-To: | abuse AT saix DOT net
|
NNTP-Posting-Date: | 9 Aug 1999 21:30:29 GMT
|
X-Priority: | 3
|
X-MSMail-Priority: | Normal
|
X-Newsreader: | Microsoft Outlook Express 5.00.2014.211
|
X-MimeOLE: | Produced By Microsoft MimeOLE V5.00.2014.211
|
To: | djgpp AT delorie DOT com
|
DJ-Gateway: | from newsgroup comp.os.msdos.djgpp
|
Reply-To: | djgpp AT delorie DOT com
|
Mike Nansgar <dkw AT iquest DOT net> wrote in message
news:37ADD81B DOT 4840 AT iquest DOT net...
> I am writing VESA VBE 2.0 routines using a combination of DJGPP and
> NASM. It seems as if my code (which calls the NEAR p-mode code) half
> works; it does set the page because when I write to the video, it sets
> the right pixels. But immediately after setting them it crashes with a
> GPF at my 'rep stosd' instruction! The same 'rep stosd' works perfectly
> when used with the set bank interrupt. Any help would be greatly
> appreciated. Thanks in advance!
I have encountered something similar when I was converting demo code from
real mode TASM to protected mode DJGPP + NASM - the problem was that my rep
stosd instructions worked perfectly, but as soon as I linked in Allegro into
the converted program, it started cracking up with a GPF at the rep stosd
just as your program is doing. If I remember right it had something to do
with allegro changing the descriptor in DS (Shawn?) - I seem to remember
fixing it by recording the DS selector in a variable, and switching this
back before a rep stosd, and then replacing it again afterward. Some
instructions that I dropped totally in the real / protected mode switch
included some compound instructions that required implied registers that I
could not use in protected mode like in real mode - I broke these up into
simpler instructions, like a mov and an inc for something like stos...
--
This message courtesy of
RylanNet rylan AT intekom DOT co DOT za
http://home.intekom.com/rylan/
--
StarWars for ever!.
- Raw text -