Quoting Michel D�nzer ([EMAIL PROTECTED]):
> On Wed, 2003-08-06 at 21:43, Denis Oliver Kropp wrote:
> > 
> > When DRI has the graphics card lock for more than 100 ms, DirectFB 
> > uses software rendering even if hardware rendering was available. 
> 
> I'm curious: How do you avoid conflicts between software rendering and
> the hardware engine? Do you simply assume they never touch the same
> framebuffer areas? (Couldn't there be conflicts even when that
> assumption holds?)

Each buffer has access flags indicating that read/write access
was done by software/hardware. Each time a buffer is locked the
flags are examined and updated. Depending on the last access
the transition from/to CPU/GPU read/write access is handled by
waiting for the GPU to be idle (if GPU wrote) or by flushing the
GPUs read cache (if CPU wrote).

-- 
Best regards,
  Denis Oliver Kropp

.------------------------------------------.
| DirectFB - Hardware accelerated graphics |
| http://www.directfb.org/                 |
"------------------------------------------"

                            Convergence GmbH


-------------------------------------------------------
This SF.Net email sponsored by: Free pre-built ASP.NET sites including
Data Reports, E-commerce, Portals, and Forums are available now.
Download today and enter to win an XBOX or Visual Studio .NET.
http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/01
_______________________________________________
Dri-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dri-devel

Reply via email to