[Intel-gfx] inter-engine read-read optimisations

Lionel Landwerlin lionel.g.landwerlin at intel.com
Thu Mar 26 09:09:09 PDT 2015


Chris,

Here is another iteration/rebase on top of nightly of your read-read
optimisations patch.

The main change is replacing the refcount for the active field of the
drm_i915_gem_object with a bit field (1 bit per engine).

This was tested with ChromiumOS by using a bunch of webgl demos
running at the same time as a couple of 1080p videos decoded using
libva.

Cheers,

-
Lionel


More information about the Intel-gfx mailing list