Message-Id: <335FC288.50A4@canvaslink.com>
Date: Thu, 24 Apr 1997 16:28:56 -0400
From: Tom Grandgent <tgrand AT canvaslink DOT com>
Reply-To: tgrand AT canvaslink DOT com
Organization: Canvas Link, Inc.
Mime-Version: 1.0
To: Roger Noss <noss AT pupk DOT unibe DOT ch>
Cc: djgpp AT delorie DOT com
Subject: Re: timer interrupt - sample code anyone?
References: <1 DOT 5 DOT 4 DOT 32 DOT 19970424175545 DOT 0068dd24 AT ubeclu DOT unibe DOT ch>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Precedence: bulk

Roger Noss wrote:
> 
> At 11:42 AM 4/24/97 -0400, you wrote:
> >
> >However, it's pretty easy to find vertical retrace.  This should wait
> >for the vertical retrace:
> >
> >while (inportb(0x3DA) & 1);
> 
> Um, I hate to drag out this false lead on the thread, but my VGA specs show
> that bit 1 of 0x3DA is horizontal retrace.  Today I (think I) succeeded with
> "& 8", for bit 3.  But the funny thing is, in NT3.51 I measured 69 Hz in
> full screen and 72.3 Hz in a window, *** regardless of whether I set Control
> Panel -> Display to 60 Hz or 72 Hz (rebooting, of course) ***.  Does anyone
> understand this??  I realize that this is no longer about djgpp, so what is
> the proper discussion forum for this kind of question?
> 
> Roger Noss

Hmm, interesting!  I got that from Allegro's vsync_out() function, but
perhaps I didn't pay enough attention to what I was doing.  Now that I
look more closely, that code is supposed to wait until the VGA is not
in either a horizontal or vertical retrace.  There is also a function
called vsync_in() though, which does the same thing except checking bit
8.

Then there is a function called _vga_vsync(), which simply calls
_vsync_out() and then _vsync_in().  So I think this is how Allegro is
waiting until the monitor has finished drawing a frame... (at least in
VGA mode anyway.)

Now, I also went back and looked at my old waitforretrace() macro that
I used to use back in Borland C++ 3.1.  It is: while (!(inp(0x3DA) & 8))
And this always worked just fine for me to provide smooth animation..

Apparently there is a bit more to this than I originally thought (?).
And as for your readings in NT, are you sure you're really being able
to access the -real- 0x3CA, or could NT be secretly doing some emulation
or something?  I say this only because I remember that OS/2 used to have
a DOS session setting that was something like
VERTICAL_RETRACE_EMULATION.

	Tom Grandgent
	tgrand AT canvaslink DOT com
	Canvas Link, Inc.