On Sat, 2005-03-12 at 17:12 +0000, Torgeir Veimo wrote:

On Sat, 2005-03-12 at 12:10 -0500, Michel D채nzer wrote:

quoted

On Sat, 2005-03-12 at 17:13 +0200, Ville Syrj채l채 wrote:

quoted

Also adding support for FB_ACTIVATE_VBL should be quite simple now.

You don't need an interrupt for that. You can choose via the
CRTC_OFFSET_CNTL:CRTC_OFFSET_FLIP_CNTL bit whether the update of the
CRTC offset should take effect immediately or on the next vertical
blank.

Personally, I'm not sure why something that needs this functionality
can't use the DRM...

My interest for adding this is video playback. It's kind of hard to do
through the DRM.

You don't have to use the DRM for anything but the wait for vsync; I can
see the effort of initializing the DRM to the point where this is
possible being prohibitive though.

The real point is that the usage of the interrupt needs to be
coordinated between the framebuffer device and the DRM. Something like
the first one to request the IRQ setting up a callback mechanism where
the other can hook into might be a possibility? Or the shared
mini-driver that has been proposed for years, but isn't liked by some...


-- 
Earthling Michel D채nzer      |     Debian (powerpc), X and DRI developer
Libre software enthusiast    |   http://svcs.affero.net/rm.php?r=daenzer


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_ide95&alloc_id396&op=click