drm/i915: Use i915_gem_object_get_dma_address() to populate rotated vmas
authorVille Syrjälä <ville.syrjala@linux.intel.com>
Tue, 16 Oct 2018 15:04:13 +0000 (18:04 +0300)
committerVille Syrjälä <ville.syrjala@linux.intel.com>
Thu, 18 Oct 2018 16:35:47 +0000 (19:35 +0300)
commit73f522bad1794f7144caff1f46f1ffb31966a386
tree15e1b9e5487d9c94bfcffe831546bb438e22c50b
parent198a2a2f1aea3eb6ea033cfec26da9a637405289
drm/i915: Use i915_gem_object_get_dma_address() to populate rotated vmas

Replace the kvmalloc_array() with i915_gem_object_get_dma_address() when
populating rotated vmas. One random access mechanism ought to be enough
for everyone?

To calculate the size of the radix tree I think we can do
something like this (assuming 64bit pointers):
 num_pages = obj_size / 4096
 tree_height = ceil(log64(num_pages))
 num_nodes = sum(64^n, n, 0, tree_height-1)
 tree_size = num_nodes * 576

If we compare that with the object size we should get a relative
overhead of around .2% to 1% for reasonable sized objects,
which framebuffers tend to be.

Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Suggested-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20181016150413.11577-1-ville.syrjala@linux.intel.com
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
drivers/gpu/drm/i915/i915_gem_gtt.c