From: "Steve Patton" <leaphe AT softhome DOT net>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: Accessing VESA 1.2
Date: Thu, 28 May 1998 19:17:58 -0600
Organization: AT&T WorldNet Services
Lines: 37
Message-ID: <6kl27p$glh@bgtnsc03.worldnet.att.net>
References: <19980528232445 DOT AAA7077 AT ppp101 DOT cartsys DOT com>
NNTP-Posting-Host: 12.67.32.223
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Precedence: bulk

You will recieve a speed gain by calling the procedure directly (well, by
using DPMI to call it rather), than to use DPMI to call the interrupt, which
will change to real mode anyway, then will call the interrupt procedure,
which will have to do a series of checks, then call the procedure.  But I
understand your reasoning, since both ways you have to switch to real-mode.

--
- Steve
http://home.att.net/~pattonl

Nate Eldredge wrote in message
<19980528232445 DOT AAA7077 AT ppp101 DOT cartsys DOT com>...
>At 05:24  5/28/1998 GMT, Nicholas Law wrote:
>>Hi all!
>>
>>I know to access the linear framebuffer under VESA 2.0. However, several
of
>>the users of my software have requested a version that runs under VESA
1.2.
>>I can access the video memory fine, but only using _slow_ interrupt based
>>memory window page select...
>>
>>In 16bit code I used to call the vesa interrupt to get the memory address
>>of a vesa function to do this...
>>
>>How can I access this function in DJGPP?
>
>Well, you can use `__dpmi_simulate_real_mode_procedure_retf' to call the
>function, but I don't know whether you'll see significant speed gains.
>
>Nate Eldredge
>nate AT cartsys DOT com
>
>
>