[Intel-gfx] [PATCH 06/13] drm/i915: Add an interface to dynamically change the cache level

[Chris Wilson]

[anholt v2: Don't forget that when going from cached to uncached, we
haven't been tracking the write domain from the CPU perspective, since
we haven't needed it for GPU coherency.]

[ickle v3: We also need to make sure we relinquish any fences on older
chipsets and clear the GTT for sane domain tracking.]

Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
Signed-off-by: Eric Anholt <eric at anholt.net>
---
 drivers/gpu/drm/i915/i915_drv.h         |    5 +++
 drivers/gpu/drm/i915/i915_gem.c         |   60 +++++++++++++++++++++++++++++++
 drivers/gpu/drm/i915/i915_gem_gtt.c     |    7 +---
 drivers/gpu/drm/i915/intel_ringbuffer.c |    6 ++-
 4 files changed, 71 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 4f63d17..61ccbeb 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1196,9 +1196,14 @@ void i915_gem_release(struct drm_device *dev, struct drm_file *file);
 uint32_t
 i915_gem_get_unfenced_gtt_alignment(struct drm_i915_gem_object *obj);
 
+int i915_gem_object_set_cache_level(struct drm_i915_gem_object *obj,
+				    enum i915_cache_level cache_level);
+
 /* i915_gem_gtt.c */
 void i915_gem_restore_gtt_mappings(struct drm_device *dev);
 int __must_check i915_gem_gtt_bind_object(struct drm_i915_gem_object *obj);
+void i915_gem_gtt_rebind_object(struct drm_i915_gem_object *obj,
+				enum i915_cache_level cache_level);
 void i915_gem_gtt_unbind_object(struct drm_i915_gem_object *obj);
 
 /* i915_gem_evict.c */
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 5d3e69f..b92510c 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -3034,6 +3034,66 @@ i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
 	return 0;
 }
 
+int i915_gem_object_set_cache_level(struct drm_i915_gem_object *obj,
+				    enum i915_cache_level cache_level)
+{
+	int ret;
+
+	if (obj->cache_level == cache_level)
+		return 0;
+
+	if (obj->pin_count) {
+		DRM_DEBUG("can not change the cache level of pinned objects\n");
+		return -EBUSY;
+	}
+
+	if (obj->gtt_space) {
+		ret = i915_gem_object_finish_gpu(obj);
+		if (ret)
+			return ret;
+
+		i915_gem_object_finish_gtt(obj);
+
+		/* Before SandyBridge, you could not use tiling or fence
+		 * registers with snooped memory, so relinquish any fences
+		 * currently pointing to our region in the aperture.
+		 */
+		if (INTEL_INFO(obj->base.dev)->gen < 6) {
+			ret = i915_gem_object_put_fence(obj);
+			if (ret)
+				return ret;
+		}
+
+		i915_gem_gtt_rebind_object(obj, cache_level);
+	}
+
+	if (cache_level == I915_CACHE_NONE) {
+		u32 old_read_domains, old_write_domain;
+
+		/* If we're coming from LLC cached, then we haven't
+		 * actually been tracking whether the data is in the
+		 * CPU cache or not, since we only allow one bit set
+		 * in obj->write_domain and have been skipping the clflushes.
+		 * Just set it to the CPU cache for now.
+		 */
+		WARN_ON(obj->base.write_domain & ~I915_GEM_DOMAIN_CPU);
+		WARN_ON(obj->base.read_domains & ~I915_GEM_DOMAIN_CPU);
+
+		old_read_domains = obj->base.read_domains;
+		old_write_domain = obj->base.write_domain;
+
+		obj->base.read_domains = I915_GEM_DOMAIN_CPU;
+		obj->base.write_domain = I915_GEM_DOMAIN_CPU;
+
+		trace_i915_gem_object_change_domain(obj,
+						    old_read_domains,
+						    old_write_domain);
+	}
+
+	obj->cache_level = cache_level;
+	return 0;
+}
+
 /*
  * Prepare buffer for display plane. Use uninterruptible for possible flush
  * wait, as in modesetting process we're not supposed to be interrupted.
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 4032289..3453f6c 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -29,9 +29,6 @@
 #include "i915_trace.h"
 #include "intel_drv.h"
 
-static void i915_gem_gtt_rebind_object(struct drm_i915_gem_object *obj,
-				       enum i915_cache_level cache_level);
-
 /* XXX kill agp_type! */
 static uint32_t cache_level_to_agp_type(struct drm_device *dev,
 					enum i915_cache_level cache_level)
@@ -93,8 +90,8 @@ int i915_gem_gtt_bind_object(struct drm_i915_gem_object *obj)
 	return 0;
 }
 
-static void i915_gem_gtt_rebind_object(struct drm_i915_gem_object *obj,
-				       enum i915_cache_level cache_level)
+void i915_gem_gtt_rebind_object(struct drm_i915_gem_object *obj,
+				enum i915_cache_level cache_level)
 {
 	struct drm_device *dev = obj->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index eab2565..f15d80f 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -236,7 +236,8 @@ init_pipe_control(struct intel_ring_buffer *ring)
 		ret = -ENOMEM;
 		goto err;
 	}
-	obj->cache_level = I915_CACHE_LLC;
+
+	i915_gem_object_set_cache_level(obj, I915_CACHE_LLC);
 
 	ret = i915_gem_object_pin(obj, 4096, true);
 	if (ret)
@@ -759,7 +760,8 @@ static int init_status_page(struct intel_ring_buffer *ring)
 		ret = -ENOMEM;
 		goto err;
 	}
-	obj->cache_level = I915_CACHE_LLC;
+
+	i915_gem_object_set_cache_level(obj, I915_CACHE_LLC);
 
 	ret = i915_gem_object_pin(obj, 4096, true);
 	if (ret != 0) {
-- 
1.7.4.1