Can anyone tell me the best layer and surfaces configuration to reduce
cpu usage.


- Minimise the number of times you copy the image
- Use hardware acceleration wherever possible
- Avoid having the CPU work on things in video memory

Based on your description, I don't think you need a separate surface.  For
best performance, simply lock the video layer surface itself, copy your data
straight into its back buffer, unlock it and flip it.  That will require one
system->video memory copy (which is unavoidable since your image is in
system memory to start with) and nothing more.  Currently, you have a
system->system copy followed by a system->video copy (neither of which is
accelerated).

Mark
_______________________________________________
directfb-users mailing list
[email protected]
http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-users

Reply via email to