Re: Trying to add post-fbprintf per-architecture callback (i.e. for DMA'ing)

That is an interesting idea, yes. Especially how it would clear the way 
for other purposes. I'll take this into consideration. Sounds like a 
good way to make it more generic, and less like a hack.

Regards,
Martijn

On 7-6-2010 0:30, Kevin O'Connor wrote:
> On Sun, Jun 06, 2010 at 11:13:37PM +0200, Martijn Stolk wrote:
>    
>> The callback is specifically meant to be called after an
>> fbprintf(), so that a DMA transfer can be performed.
>>      
> Another random thought - feel free to ignore.
>
> Instead of setting a postFbCallback function, one could have a "putc"
> function pointer.  By default it would point to fb_putc(), but one
> could override it.  For the dma case it would be something like:
>
> static void __preload my_putc(fbinfo *fbi, char c) {
>      fb_putc(fbi, c);
>      if (c != '\n')
>          return;
>      ... do dma ...
> }
>
> The gain here is that one could use the same mechanism to send output
> over a serial port.  Which I think has been requested a few times.
>
> Also, I think 'fbi' could be passed into Mach->preHardwareShutdown and
> the framebuffer setup done there.
>
> Anyway - just random ideas.
>
> -Kevin
>