A82: Re: display port vs. graph_mem
[Prev][Next][Index][Thread]
A82: Re: display port vs. graph_mem
The only way to access the display is by using the ports to write each of
the bytes to display, so using the display controller directly should be
faster. The CR_GRBCopy is very optimized, so to increse speed by writing
directly to the port you need to use optimized routines too.
The advantage of using CR_GRBCopy is that one can use block copy to output
the bitmap to the LCD, which combined with writing column by column speeds
up the process.
If trying to use the display controller directly remember the following:
- Do not use the TI delay, it is way to long
- Use block copy to write sprites etc
- For most sprites write column by column
Dines
/----------------------------\
| Dines Justsen |
| dines@aub.dk |
| www.student.dtu.dk/~c958362|
\----------------------------/
----- Original Message -----
From: "Doug Torrance" <profzoom@hotmail.com>
To: <assembly-82@lists.ticalc.org>
Sent: Sunday, January 23, 2000 12:57 AM
Subject: A82: display port vs. graph_mem
>
> It seems like writing data to the GRAPH_MEM and doing a DISP_GRAPH or
> CR_GRBCopy works faster than writing directly to the display port byte by
> byte. Why is this? Don't routines that copy GRAPH_MEM to the screen end
up
> writing to the display port in the end? Or is there a way to somehow
> transfer all 768 bytes at once to the display port, without having to do
> them one by one?
>
> Thanks,
> Doug
> ______________________________________________________
> Get Your Private, Free Email at http://www.hotmail.com
>
>
References: