[Intel-gfx] [PATCH 5/6] drm/i915: Use the uncached domain for the display planes v2

[Eric Anholt]

The simplest and common method for ensuring scanout coherency on all
chipsets is to mark the scanout buffers as uncached (and for
userspace to remember to flush the render cache every so often).

We can improve upon this for later generations by marking scanout
objects as GFDT and only flush those cachelines when required. However,
we start simple.

[v2: Move the set to uncached above the clflush.  Otherwise, we'd skip
the clflush and try to scan out data that was still sitting in the
cache.]

Signed-off-by: Eric Anholt <eric at anholt.net>
Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_gem.c |   13 +++++++++++++
 1 files changed, 13 insertions(+), 0 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 8389b03..de9a446 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -3062,6 +3062,19 @@ i915_gem_object_set_to_display_plane(struct drm_i915_gem_object *obj,
 			return ret;
 	}
 
+	/* The display engine is not coherent with the LLC cache on gen6.  As
+	 * a result, we make sure that the pinning that is about to occur is
+	 * done with uncached PTEs. This is lowest common denominator for all
+	 * chipsets.
+	 *
+	 * However for gen6+, we could do better by using the GFDT bit instead
+	 * of uncaching, which would allow us to flush all the LLC-cached data
+	 * with that bit in the PTE to main memory with just one PIPE_CONTROL.
+	 */
+	ret = i915_gem_object_set_cache_level(obj, I915_CACHE_NONE);
+	if (ret)
+		return ret;
+
 	i915_gem_object_flush_cpu_write_domain(obj);
 
 	old_read_domains = obj->base.read_domains;
-- 
1.7.4.1