[Intel-gfx] [PATCH] intel: Add a new bufmgr alloc function to get BOs ready for rendering to.

[Eric Anholt]

On Thu, 2009-02-19 at 10:53 +0000, Chris Wilson wrote:
> On Wed, 2009-02-18 at 13:13 -0800, Eric Anholt wrote:
> > This avoids using the oldest BO in the BO cache and waiting for it to be
> > idle before we turn around and render to it with the GPU.  Thanks to
> > Chris Wilson for pointing out how silly we were being.
> 
> Meh - usually it's the other away around :-)
> 
> Using the MRU buffer is a nice refinement - at the very least it
> preserves the oldest buffer for when we absolutely need an idle buffer.
> My immediate observation is that this will interact badly (on i915 class
> h/w) with tiled buffers if we subsequently need to change tiling. Not
> that I think this is likely with the current users, but...

We should probably replace the set_tiling(bo, TILING_NONE) in the
add-to-reuse-cache path with "if you're currently tiled, you get freed
instead" -- right now you've already hurt the performance before you get
to the point of reusing the buffer.

Though I suppose this only matters for !IS_I965G.

-- 
Eric Anholt
eric at anholt.net                         eric.anholt at intel.com


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part
URL: <http://lists.freedesktop.org/archives/intel-gfx/attachments/20090221/e6da7d28/attachment.sig>