diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c
index 2b9725f..f221631 100644
--- a/drivers/gpu/drm/i915/i915_dma.c
+++ b/drivers/gpu/drm/i915/i915_dma.c
@@ -1827,19 +1827,11 @@
 
 int i915_driver_open(struct drm_device *dev, struct drm_file *file)
 {
-	struct drm_i915_file_private *file_priv;
+	int ret;
 
-	DRM_DEBUG_DRIVER("\n");
-	file_priv = kzalloc(sizeof(*file_priv), GFP_KERNEL);
-	if (!file_priv)
-		return -ENOMEM;
-
-	file->driver_priv = file_priv;
-
-	spin_lock_init(&file_priv->mm.lock);
-	INIT_LIST_HEAD(&file_priv->mm.request_list);
-
-	idr_init(&file_priv->context_idr);
+	ret = i915_gem_open(dev, file);
+	if (ret)
+		return ret;
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 79bbcf9..d1c6b92 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -844,9 +844,6 @@
 	struct work_struct work;
 	u32 pm_iir;
 
-	/* On vlv we need to manually drop to Vmin with a delayed work. */
-	struct delayed_work vlv_work;
-
 	/* The below variables an all the rps hw state are protected by
 	 * dev->struct mutext. */
 	u8 cur_delay;
@@ -965,6 +962,15 @@
 	struct delayed_work retire_work;
 
 	/**
+	 * When we detect an idle GPU, we want to turn on
+	 * powersaving features. So once we see that there
+	 * are no more requests outstanding and no more
+	 * arrive within a small period of time, we fire
+	 * off the idle_work.
+	 */
+	struct delayed_work idle_work;
+
+	/**
 	 * Are we in a non-interruptible section of code like
 	 * modesetting?
 	 */
@@ -1597,13 +1603,17 @@
 };
 
 struct drm_i915_file_private {
+	struct drm_i915_private *dev_priv;
+
 	struct {
 		spinlock_t lock;
 		struct list_head request_list;
+		struct delayed_work idle_work;
 	} mm;
 	struct idr context_idr;
 
 	struct i915_ctx_hang_stats hang_stats;
+	atomic_t rps_wait_boost;
 };
 
 #define INTEL_INFO(dev)	(to_i915(dev)->info)
@@ -1953,7 +1963,7 @@
 	}
 }
 
-void i915_gem_retire_requests(struct drm_device *dev);
+bool i915_gem_retire_requests(struct drm_device *dev);
 void i915_gem_retire_requests_ring(struct intel_ring_buffer *ring);
 int __must_check i915_gem_check_wedge(struct i915_gpu_error *error,
 				      bool interruptible);
@@ -2004,6 +2014,7 @@
 void i915_gem_detach_phys_object(struct drm_device *dev,
 				 struct drm_i915_gem_object *obj);
 void i915_gem_free_all_phys_object(struct drm_device *dev);
+int i915_gem_open(struct drm_device *dev, struct drm_file *file);
 void i915_gem_release(struct drm_device *dev, struct drm_file *file);
 
 uint32_t
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 53e3151..19ecfa8 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -982,6 +982,14 @@
 	return test_bit(ring->id, &dev_priv->gpu_error.missed_irq_rings);
 }
 
+static bool can_wait_boost(struct drm_i915_file_private *file_priv)
+{
+	if (file_priv == NULL)
+		return true;
+
+	return !atomic_xchg(&file_priv->rps_wait_boost, true);
+}
+
 /**
  * __wait_seqno - wait until execution of seqno has finished
  * @ring: the ring expected to report seqno
@@ -1002,7 +1010,9 @@
  */
 static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 			unsigned reset_counter,
-			bool interruptible, struct timespec *timeout)
+			bool interruptible,
+			struct timespec *timeout,
+			struct drm_i915_file_private *file_priv)
 {
 	drm_i915_private_t *dev_priv = ring->dev->dev_private;
 	struct timespec before, now;
@@ -1017,6 +1027,14 @@
 
 	timeout_jiffies = timeout ? timespec_to_jiffies_timeout(timeout) : 1;
 
+	if (dev_priv->info->gen >= 6 && can_wait_boost(file_priv)) {
+		gen6_rps_boost(dev_priv);
+		if (file_priv)
+			mod_delayed_work(dev_priv->wq,
+					 &file_priv->mm.idle_work,
+					 msecs_to_jiffies(100));
+	}
+
 	if (!(dev_priv->gpu_error.test_irq_rings & intel_ring_flag(ring)) &&
 	    WARN_ON(!ring->irq_get(ring)))
 		return -ENODEV;
@@ -1116,7 +1134,7 @@
 
 	return __wait_seqno(ring, seqno,
 			    atomic_read(&dev_priv->gpu_error.reset_counter),
-			    interruptible, NULL);
+			    interruptible, NULL, NULL);
 }
 
 static int
@@ -1166,6 +1184,7 @@
  */
 static __must_check int
 i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj,
+					    struct drm_file *file,
 					    bool readonly)
 {
 	struct drm_device *dev = obj->base.dev;
@@ -1192,7 +1211,7 @@
 
 	reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter);
 	mutex_unlock(&dev->struct_mutex);
-	ret = __wait_seqno(ring, seqno, reset_counter, true, NULL);
+	ret = __wait_seqno(ring, seqno, reset_counter, true, NULL, file->driver_priv);
 	mutex_lock(&dev->struct_mutex);
 	if (ret)
 		return ret;
@@ -1241,7 +1260,7 @@
 	 * We will repeat the flush holding the lock in the normal manner
 	 * to catch cases where we are gazumped.
 	 */
-	ret = i915_gem_object_wait_rendering__nonblocking(obj, !write_domain);
+	ret = i915_gem_object_wait_rendering__nonblocking(obj, file, !write_domain);
 	if (ret)
 		goto unref;
 
@@ -2162,6 +2181,7 @@
 		i915_queue_hangcheck(ring->dev);
 
 		if (was_empty) {
+			cancel_delayed_work_sync(&dev_priv->mm.idle_work);
 			queue_delayed_work(dev_priv->wq,
 					   &dev_priv->mm.retire_work,
 					   round_jiffies_up_relative(HZ));
@@ -2183,10 +2203,8 @@
 		return;
 
 	spin_lock(&file_priv->mm.lock);
-	if (request->file_priv) {
-		list_del(&request->client_list);
-		request->file_priv = NULL;
-	}
+	list_del(&request->client_list);
+	request->file_priv = NULL;
 	spin_unlock(&file_priv->mm.lock);
 }
 
@@ -2450,57 +2468,53 @@
 	WARN_ON(i915_verify_lists(ring->dev));
 }
 
-void
+bool
 i915_gem_retire_requests(struct drm_device *dev)
 {
 	drm_i915_private_t *dev_priv = dev->dev_private;
 	struct intel_ring_buffer *ring;
+	bool idle = true;
 	int i;
 
-	for_each_ring(ring, dev_priv, i)
+	for_each_ring(ring, dev_priv, i) {
 		i915_gem_retire_requests_ring(ring);
+		idle &= list_empty(&ring->request_list);
+	}
+
+	if (idle)
+		mod_delayed_work(dev_priv->wq,
+				   &dev_priv->mm.idle_work,
+				   msecs_to_jiffies(100));
+
+	return idle;
 }
 
 static void
 i915_gem_retire_work_handler(struct work_struct *work)
 {
-	drm_i915_private_t *dev_priv;
-	struct drm_device *dev;
-	struct intel_ring_buffer *ring;
+	struct drm_i915_private *dev_priv =
+		container_of(work, typeof(*dev_priv), mm.retire_work.work);
+	struct drm_device *dev = dev_priv->dev;
 	bool idle;
-	int i;
-
-	dev_priv = container_of(work, drm_i915_private_t,
-				mm.retire_work.work);
-	dev = dev_priv->dev;
 
 	/* Come back later if the device is busy... */
-	if (!mutex_trylock(&dev->struct_mutex)) {
+	idle = false;
+	if (mutex_trylock(&dev->struct_mutex)) {
+		idle = i915_gem_retire_requests(dev);
+		mutex_unlock(&dev->struct_mutex);
+	}
+	if (!idle)
 		queue_delayed_work(dev_priv->wq, &dev_priv->mm.retire_work,
 				   round_jiffies_up_relative(HZ));
-		return;
-	}
+}
 
-	i915_gem_retire_requests(dev);
+static void
+i915_gem_idle_work_handler(struct work_struct *work)
+{
+	struct drm_i915_private *dev_priv =
+		container_of(work, typeof(*dev_priv), mm.idle_work.work);
 
-	/* Send a periodic flush down the ring so we don't hold onto GEM
-	 * objects indefinitely.
-	 */
-	idle = true;
-	for_each_ring(ring, dev_priv, i) {
-		if (ring->gpu_caches_dirty)
-			i915_add_request(ring, NULL);
-
-		idle &= list_empty(&ring->request_list);
-	}
-
-	if (!dev_priv->ums.mm_suspended && !idle)
-		queue_delayed_work(dev_priv->wq, &dev_priv->mm.retire_work,
-				   round_jiffies_up_relative(HZ));
-	if (idle)
-		intel_mark_idle(dev);
-
-	mutex_unlock(&dev->struct_mutex);
+	intel_mark_idle(dev_priv->dev);
 }
 
 /**
@@ -2598,7 +2612,7 @@
 	reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter);
 	mutex_unlock(&dev->struct_mutex);
 
-	ret = __wait_seqno(ring, seqno, reset_counter, true, timeout);
+	ret = __wait_seqno(ring, seqno, reset_counter, true, timeout, file->driver_priv);
 	if (timeout)
 		args->timeout_ns = timespec_to_ns(timeout);
 	return ret;
@@ -3809,7 +3823,7 @@
 	if (seqno == 0)
 		return 0;
 
-	ret = __wait_seqno(ring, seqno, reset_counter, true, NULL);
+	ret = __wait_seqno(ring, seqno, reset_counter, true, NULL, NULL);
 	if (ret == 0)
 		queue_delayed_work(dev_priv->wq, &dev_priv->mm.retire_work, 0);
 
@@ -4279,6 +4293,7 @@
 
 	/* Cancel the retire work handler, which should be idle now. */
 	cancel_delayed_work_sync(&dev_priv->mm.retire_work);
+	cancel_delayed_work_sync(&dev_priv->mm.idle_work);
 
 	return 0;
 }
@@ -4612,6 +4627,8 @@
 		INIT_LIST_HEAD(&dev_priv->fence_regs[i].lru_list);
 	INIT_DELAYED_WORK(&dev_priv->mm.retire_work,
 			  i915_gem_retire_work_handler);
+	INIT_DELAYED_WORK(&dev_priv->mm.idle_work,
+			  i915_gem_idle_work_handler);
 	init_waitqueue_head(&dev_priv->gpu_error.reset_queue);
 
 	/* On GEN3 we really need to make sure the ARB C3 LP bit is set */
@@ -4836,6 +4853,8 @@
 {
 	struct drm_i915_file_private *file_priv = file->driver_priv;
 
+	cancel_delayed_work_sync(&file_priv->mm.idle_work);
+
 	/* Clean up our request list when the client is going away, so that
 	 * later retire_requests won't dereference our soon-to-be-gone
 	 * file_priv.
@@ -4853,6 +4872,38 @@
 	spin_unlock(&file_priv->mm.lock);
 }
 
+static void
+i915_gem_file_idle_work_handler(struct work_struct *work)
+{
+	struct drm_i915_file_private *file_priv =
+		container_of(work, typeof(*file_priv), mm.idle_work.work);
+
+	atomic_set(&file_priv->rps_wait_boost, false);
+}
+
+int i915_gem_open(struct drm_device *dev, struct drm_file *file)
+{
+	struct drm_i915_file_private *file_priv;
+
+	DRM_DEBUG_DRIVER("\n");
+
+	file_priv = kzalloc(sizeof(*file_priv), GFP_KERNEL);
+	if (!file_priv)
+		return -ENOMEM;
+
+	file->driver_priv = file_priv;
+	file_priv->dev_priv = dev->dev_private;
+
+	spin_lock_init(&file_priv->mm.lock);
+	INIT_LIST_HEAD(&file_priv->mm.request_list);
+	INIT_DELAYED_WORK(&file_priv->mm.idle_work,
+			  i915_gem_file_idle_work_handler);
+
+	idr_init(&file_priv->context_idr);
+
+	return 0;
+}
+
 static bool mutex_is_locked_by(struct mutex *mutex, struct task_struct *task)
 {
 	if (!mutex_is_locked(mutex))
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 05c05a6..6ee5572 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -859,17 +859,6 @@
 			gen6_set_rps(dev_priv->dev, new_delay);
 	}
 
-	if (IS_VALLEYVIEW(dev_priv->dev)) {
-		/*
-		 * On VLV, when we enter RC6 we may not be at the minimum
-		 * voltage level, so arm a timer to check.  It should only
-		 * fire when there's activity or once after we've entered
-		 * RC6, and then won't be re-armed until the next RPS interrupt.
-		 */
-		mod_delayed_work(dev_priv->wq, &dev_priv->rps.vlv_work,
-				 msecs_to_jiffies(100));
-	}
-
 	mutex_unlock(&dev_priv->rps.hw_lock);
 }
 
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index d9f698e..169d682 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -7726,6 +7726,9 @@
 
 		intel_decrease_pllclock(crtc);
 	}
+
+	if (dev_priv->info->gen >= 6)
+		gen6_rps_idle(dev->dev_private);
 }
 
 void intel_mark_fb_busy(struct drm_i915_gem_object *obj,
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index e6db0c6..922c5d7 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -825,4 +825,7 @@
 /* intel_tv.c */
 void intel_tv_init(struct drm_device *dev);
 
+void gen6_rps_idle(struct drm_i915_private *dev_priv);
+void gen6_rps_boost(struct drm_i915_private *dev_priv);
+
 #endif /* __INTEL_DRV_H__ */
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 9753bd9..ca063c2 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -3345,6 +3345,26 @@
 	trace_intel_gpu_freq_change(val * 50);
 }
 
+void gen6_rps_idle(struct drm_i915_private *dev_priv)
+{
+	mutex_lock(&dev_priv->rps.hw_lock);
+	if (dev_priv->info->is_valleyview)
+		valleyview_set_rps(dev_priv->dev, dev_priv->rps.min_delay);
+	else
+		gen6_set_rps(dev_priv->dev, dev_priv->rps.min_delay);
+	mutex_unlock(&dev_priv->rps.hw_lock);
+}
+
+void gen6_rps_boost(struct drm_i915_private *dev_priv)
+{
+	mutex_lock(&dev_priv->rps.hw_lock);
+	if (dev_priv->info->is_valleyview)
+		valleyview_set_rps(dev_priv->dev, dev_priv->rps.max_delay);
+	else
+		gen6_set_rps(dev_priv->dev, dev_priv->rps.max_delay);
+	mutex_unlock(&dev_priv->rps.hw_lock);
+}
+
 /*
  * Wait until the previous freq change has completed,
  * or the timeout elapsed, and then update our notion
@@ -3734,24 +3754,6 @@
 	return vlv_punit_read(dev_priv, PUNIT_REG_GPU_LFM) & 0xff;
 }
 
-static void vlv_rps_timer_work(struct work_struct *work)
-{
-	drm_i915_private_t *dev_priv = container_of(work, drm_i915_private_t,
-						    rps.vlv_work.work);
-
-	/*
-	 * Timer fired, we must be idle.  Drop to min voltage state.
-	 * Note: we use RPe here since it should match the
-	 * Vmin we were shooting for.  That should give us better
-	 * perf when we come back out of RC6 than if we used the
-	 * min freq available.
-	 */
-	mutex_lock(&dev_priv->rps.hw_lock);
-	if (dev_priv->rps.cur_delay > dev_priv->rps.rpe_delay)
-		valleyview_set_rps(dev_priv->dev, dev_priv->rps.rpe_delay);
-	mutex_unlock(&dev_priv->rps.hw_lock);
-}
-
 static void valleyview_setup_pctx(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -3894,8 +3896,6 @@
 				      dev_priv->rps.rpe_delay),
 			 dev_priv->rps.rpe_delay);
 
-	INIT_DELAYED_WORK(&dev_priv->rps.vlv_work, vlv_rps_timer_work);
-
 	valleyview_set_rps(dev_priv->dev, dev_priv->rps.rpe_delay);
 
 	gen6_enable_rps_interrupts(dev);
@@ -4635,8 +4635,6 @@
 	} else if (INTEL_INFO(dev)->gen >= 6) {
 		cancel_delayed_work_sync(&dev_priv->rps.delayed_resume_work);
 		cancel_work_sync(&dev_priv->rps.work);
-		if (IS_VALLEYVIEW(dev))
-			cancel_delayed_work_sync(&dev_priv->rps.vlv_work);
 		mutex_lock(&dev_priv->rps.hw_lock);
 		if (IS_VALLEYVIEW(dev))
 			valleyview_disable_rps(dev);