Subject: Re: vbl handler in grfabs_cc.c
To: None <chopps@emunix.emich.edu, rhialto@mbfys.kun.nl>
From: Olaf Seibert <rhialto@mbfys.kun.nl>
List: amiga-dev
Date: 03/17/1994 11:02:57
chopps@emunix.emich.edu (Chris Hopps) wrote:
> I think your talking about the copperlist terminating to early?  If so
> all my lists end with
> 
> 	CWAIT(255, 255);
> 	CWAIT(255, 255);
> 
> Which handles the extra lines I belive.  If not I am all ears.  I

I have been looking at my copperlists (in AGA and ECS mode) and
they seem to end with a single WAIT 255,127.

> learned amiga HW programming when starting to work on NEtBSD so its
> very possible I have overlooked certain "undocumented" features of
> the HW :^)

Anyway, in an attempt to be useful for a change: in your code you
comment if you could borrow somebodies A4000. Well perhaps I can offer
the next best thing: you provide an ECS-mode dblpal/dblntsc (-hires-
no-flicker) (i.e., 31 KHz 640x512 noninterlaced), and I (try to ;-)
do the AGA extension.

Looking at the copperlists for my wb screen, both in ECS and AGA mode,
and at least the copperlists look mostly like non-interlaced 512 lines
copperlists. I suppose the hard part is to get the variable sync rates
right - the ECS bit in the HRM is a bit vague here.

The difference between ECS and AGA isn't too much from a register
level interface: there is no need for the weird colour register
setup (except that you must do some tricks to access the extra
colour precision bits and the colour registers >= 32), and a single
extra store in the FMODE register to speed up the bitplane DMA (4x).
Without that, 2-plane dblpal eats all the chip mem bandwidth, which
is really horrible. It only needs 8-byte aligned bitplanes, which
is probably guaranteed already.

For some aga info, I looked at the file 'agaguide.lha', somewhere on
aminet. Interesting to read, except that it doesn't specify the new
limits on DDFSTRT and DDFSTOP, but the values for those in my aga-mode
copperlist are lower than in my ecs-mode copperlist. Perhaps it is
also some align-to-multiple-of-8 case, though.

> Chris.
-Olaf.
--
___ Olaf 'Rhialto' Seibert       D787B44DFC896063 4CBB95A5BD1DAA96
\X/ There are no lemurs in this post	      rhialto@mbfys.kun.nl

------------------------------------------------------------------------------