[Intel-gfx] [ANNOUNCE] xf86-video-intel 2.8.0

[Kristian Høgsberg]

On Mon, Aug 3, 2009 at 12:13 PM, Eric Anholt<eric at anholt.net> wrote:
> We would certainly love to keep the GPU running flat out, but right now
> we have no tearing and we're going to stick with that until we get
> something better in.  jbarnes and krh at least have been working on
> adding DRI2 pageflipping support, so if your compositor uses SwapBuffers
> as opposed to CopySubBuffer (the previous recommendation for
> compositors, sorry for changing our minds), the GPU will be able to keep
> running because the page flip happens independently of the ringbuffer
> operation.  I'm guessing this will be in the 2.6.32 timeframe.

Yup, the asynchronous glXSwapBuffer page flipping looks like it's
done.  I've been landing the patches in Fedora but will push them
upstream soon (the Fedora 12 feature deadline was a little too close,
so I had to push it out downstream first).  This work has the benefit
that it doesn't block any other application from using the GPU while
we're waiting for the page flip to happen.  Like the
"wait-for-scanline-and-copy" approach, glXSwapBuffers doesn't block on
the vertical retrace to happen but returns immediately, allowing the
compositor to handle events or other CPU bound tasks before rendering
the next frame.  I'll propose it for the 2.6.32 merge window and land
the userspace bits (libdrm, mesa, xserver, dri2proto,
xf86-video-{intel,...}) once we get the page flip ioctl upstream.

cheers,
Kristian