Merge branch 'drm-intel-fixes' of git://people.freedesktop.org/~danvet/drm-intel into drm-next

Daniel writes:
A few leftover fixes for 3.8:
- VIC support for hdmi infoframes with the associated drm helper, fixes
  some black TVs (Paulo Zanoni)
- Modeset state check (and fixup if the BIOS messed with the hw) for
  lid-open. modeset-rework fallout. Somehow the original reporter went
  awol, so this stalled for way too long until we've found a new
  victim^Wreporter with broken BIOS.
- seqno wrap fixes from Mika and Chris.
- Some minor fixes all over from various people.
- Another race fix in the pageflip vs. unpin code from Chris.
- hsw vga resume support and a few more fdi link fixes (only used for vga
  on hsw) from Paulo.
- Regression fix for DMAR from Zhenyu Wang - I've scavenged memory from my
  DMAR for a while and it broke right away :(
- Regression fix from Takashi Iwai for ivb lvds - some w/a needs to be
  (partially) moved back into place. Note that these are regressions in
  -next.
- One more fix for ivb 3 pipe support - it now actually seems to work.

* 'drm-intel-fixes' of git://people.freedesktop.org/~danvet/drm-intel: (25 commits)
  drm/i915: Fix missed needs_dmar setting
  drm/i915: Fix shifted screen on top of LVDS on IVY laptop
  drm/i915: disable cpt phase pointer fdi rx workaround
  drm/i915: set the LPT FDI RX polarity reversal bit when needed
  drm/i915: add lpt_init_pch_refclk
  drm/i915: add support for mPHY destination on intel_sbi_{read, write}
  drm/i915: reject modes the LPT FDI receiver can't handle
  drm/i915: fix hsw_fdi_link_train "retry" code
  drm/i915: Close race between processing unpin task and queueing the flip
  drm/i915: fixup l3 parity sysfs access check
  drm/i915: Clear the existing watermarks for g4x when modifying the cursor sr
  drm/i915: do not access BLC_PWM_CTL2 on pre-gen4 hardware
  drm/i915: Don't allow ring tail to reach the same cacheline as head
  drm/i915: Decouple the object from the unbound list before freeing pages
  drm/i915: Set sync_seqno properly after seqno wrap
  drm/i915: Include the last semaphore sync point in the error-state
  drm/i915: Rearrange code to only have a single method for waiting upon the ring
  drm/i915: Simplify flushing activity on the ring
  drm/i915: Preallocate next seqno before touching the ring
  drm/i915: force restore on lid open
  ...

23 files changed, 590 insertions, 280 deletions

diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
index 484c36a4b7a..5a3770fbd77 100644
--- a/drivers/gpu/drm/drm_edid.c
+++ b/drivers/gpu/drm/drm_edid.c
@@ -2079,3 +2079,22 @@ int drm_add_modes_noedid(struct drm_connector *connector,
 	return num_modes;
 }
 EXPORT_SYMBOL(drm_add_modes_noedid);
+
+/**
+ * drm_mode_cea_vic - return the CEA-861 VIC of a given mode
+ * @mode: mode
+ *
+ * RETURNS:
+ * The VIC number, 0 in case it's not a CEA-861 mode.
+ */
+uint8_t drm_mode_cea_vic(const struct drm_display_mode *mode)
+{
+	uint8_t i;
+
+	for (i = 0; i < drm_num_cea_modes; i++)
+		if (drm_mode_equal(mode, &edid_cea_modes[i]))
+			return i + 1;
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_mode_cea_vic);
diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 4568e7d8a06..e6a11ca85ea 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -317,7 +317,7 @@ static int i915_gem_pageflip_info(struct seq_file *m, void *data)
 			seq_printf(m, "No flip due on pipe %c (plane %c)\n",
 				   pipe, plane);
 		} else {
-			if (!work->pending) {
+			if (atomic_read(&work->pending) < INTEL_FLIP_COMPLETE) {
 				seq_printf(m, "Flip queued on pipe %c (plane %c)\n",
 					   pipe, plane);
 			} else {
@@ -328,7 +328,7 @@ static int i915_gem_pageflip_info(struct seq_file *m, void *data)
 				seq_printf(m, "Stall check enabled, ");
 			else
 				seq_printf(m, "Stall check waiting for page flip ioctl, ");
-			seq_printf(m, "%d prepares\n", work->pending);
+			seq_printf(m, "%d prepares\n", atomic_read(&work->pending));
 
 			if (work->old_fb_obj) {
 				struct drm_i915_gem_object *obj = work->old_fb_obj;
@@ -655,10 +655,12 @@ static void i915_ring_error_state(struct seq_file *m,
 	if (INTEL_INFO(dev)->gen >= 6) {
 		seq_printf(m, "  RC PSMI: 0x%08x\n", error->rc_psmi[ring]);
 		seq_printf(m, "  FAULT_REG: 0x%08x\n", error->fault_reg[ring]);
-		seq_printf(m, "  SYNC_0: 0x%08x\n",
-			   error->semaphore_mboxes[ring][0]);
-		seq_printf(m, "  SYNC_1: 0x%08x\n",
-			   error->semaphore_mboxes[ring][1]);
+		seq_printf(m, "  SYNC_0: 0x%08x [last synced 0x%08x]\n",
+			   error->semaphore_mboxes[ring][0],
+			   error->semaphore_seqno[ring][0]);
+		seq_printf(m, "  SYNC_1: 0x%08x [last synced 0x%08x]\n",
+			   error->semaphore_mboxes[ring][1],
+			   error->semaphore_seqno[ring][1]);
 	}
 	seq_printf(m, "  seqno: 0x%08x\n", error->seqno[ring]);
 	seq_printf(m, "  waiting: %s\n", yesno(error->waiting[ring]));
diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c
index a48e4910ea2..8f63cd5de4b 100644
--- a/drivers/gpu/drm/i915/i915_dma.c
+++ b/drivers/gpu/drm/i915/i915_dma.c
@@ -141,7 +141,7 @@ void i915_kernel_lost_context(struct drm_device * dev)
 
 	ring->head = I915_READ_HEAD(ring) & HEAD_ADDR;
 	ring->tail = I915_READ_TAIL(ring) & TAIL_ADDR;
-	ring->space = ring->head - (ring->tail + 8);
+	ring->space = ring->head - (ring->tail + I915_RING_FREE_SPACE);
 	if (ring->space < 0)
 		ring->space += ring->size;
 
@@ -592,10 +592,8 @@ static int i915_dispatch_flip(struct drm_device * dev)
 
 static int i915_quiescent(struct drm_device *dev)
 {
-	struct intel_ring_buffer *ring = LP_RING(dev->dev_private);
-
 	i915_kernel_lost_context(dev);
-	return intel_wait_ring_idle(ring);
+	return intel_ring_idle(LP_RING(dev->dev_private));
 }
 
 static int i915_flush_ioctl(struct drm_device *dev, void *data,
@@ -1045,7 +1043,7 @@ static int i915_set_status_page(struct drm_device *dev, void *data,
 {
 	drm_i915_private_t *dev_priv = dev->dev_private;
 	drm_i915_hws_addr_t *hws = data;
-	struct intel_ring_buffer *ring = LP_RING(dev_priv);
+	struct intel_ring_buffer *ring;
 
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		return -ENODEV;
@@ -1065,6 +1063,7 @@ static int i915_set_status_page(struct drm_device *dev, void *data,
 
 	DRM_DEBUG_DRIVER("set status page addr 0x%08x\n", (u32)hws->addr);
 
+	ring = LP_RING(dev_priv);
 	ring->status_page.gfx_addr = hws->addr & (0x1ffff<<12);
 
 	dev_priv->dri1.gfx_hws_cpu_addr =
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 6745c7f976d..530db83ef32 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -554,8 +554,7 @@ static int __i915_drm_thaw(struct drm_device *dev)
 
 	/* KMS EnterVT equivalent */
 	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
-		if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
-			ironlake_init_pch_refclk(dev);
+		intel_init_pch_refclk(dev);
 
 		mutex_lock(&dev->struct_mutex);
 		dev_priv->mm.suspended = 0;
@@ -564,7 +563,7 @@ static int __i915_drm_thaw(struct drm_device *dev)
 		mutex_unlock(&dev->struct_mutex);
 
 		intel_modeset_init_hw(dev);
-		intel_modeset_setup_hw_state(dev);
+		intel_modeset_setup_hw_state(dev, false);
 		drm_irq_install(dev);
 	}
 
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 87c06f97fa8..557843dd4b2 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -197,6 +197,7 @@ struct drm_i915_error_state {
 	u32 instdone[I915_NUM_RINGS];
 	u32 acthd[I915_NUM_RINGS];
 	u32 semaphore_mboxes[I915_NUM_RINGS][I915_NUM_RINGS - 1];
+	u32 semaphore_seqno[I915_NUM_RINGS][I915_NUM_RINGS - 1];
 	u32 rc_psmi[I915_NUM_RINGS]; /* sleep state */
 	/* our own tracking of ring head and tail */
 	u32 cpu_ring_head[I915_NUM_RINGS];
@@ -381,6 +382,11 @@ enum intel_pch {
 	PCH_LPT,	/* Lynxpoint PCH */
 };
 
+enum intel_sbi_destination {
+	SBI_ICLK,
+	SBI_MPHY,
+};
+
 #define QUIRK_PIPEA_FORCE (1<<0)
 #define QUIRK_LVDS_SSC_DISABLE (1<<1)
 #define QUIRK_INVERT_BRIGHTNESS (1<<2)
@@ -909,6 +915,8 @@ typedef struct drm_i915_private {
 	bool hw_contexts_disabled;
 	uint32_t hw_context_size;
 
+	bool fdi_rx_polarity_reversed;
+
 	struct i915_suspend_saved_registers regfile;
 
 	/* Old dri1 support infrastructure, beware the dragons ya fools entering
@@ -1417,8 +1425,7 @@ int __must_check i915_mutex_lock_interruptible(struct drm_device *dev);
 int i915_gem_object_sync(struct drm_i915_gem_object *obj,
 			 struct intel_ring_buffer *to);
 void i915_gem_object_move_to_active(struct drm_i915_gem_object *obj,
-				    struct intel_ring_buffer *ring,
-				    u32 seqno);
+				    struct intel_ring_buffer *ring);
 
 int i915_gem_dumb_create(struct drm_file *file_priv,
 			 struct drm_device *dev,
@@ -1436,7 +1443,7 @@ i915_seqno_passed(uint32_t seq1, uint32_t seq2)
 	return (int32_t)(seq1 - seq2) >= 0;
 }
 
-u32 i915_gem_next_request_seqno(struct intel_ring_buffer *ring);
+extern int i915_gem_get_seqno(struct drm_device *dev, u32 *seqno);
 
 int __must_check i915_gem_object_get_fence(struct drm_i915_gem_object *obj);
 int __must_check i915_gem_object_put_fence(struct drm_i915_gem_object *obj);
@@ -1652,11 +1659,12 @@ extern void intel_modeset_init(struct drm_device *dev);
 extern void intel_modeset_gem_init(struct drm_device *dev);
 extern void intel_modeset_cleanup(struct drm_device *dev);
 extern int intel_modeset_vga_set_state(struct drm_device *dev, bool state);
-extern void intel_modeset_setup_hw_state(struct drm_device *dev);
+extern void intel_modeset_setup_hw_state(struct drm_device *dev,
+					 bool force_restore);
 extern bool intel_fbc_enabled(struct drm_device *dev);
 extern void intel_disable_fbc(struct drm_device *dev);
 extern bool ironlake_set_drps(struct drm_device *dev, u8 val);
-extern void ironlake_init_pch_refclk(struct drm_device *dev);
+extern void intel_init_pch_refclk(struct drm_device *dev);
 extern void gen6_set_rps(struct drm_device *dev, u8 val);
 extern void intel_detect_pch(struct drm_device *dev);
 extern int intel_trans_dp_port_sel(struct drm_crtc *crtc);
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index b0016bb6563..c1f691958f8 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1696,10 +1696,14 @@ i915_gem_object_put_pages(struct drm_i915_gem_object *obj)
 	if (obj->pages_pin_count)
 		return -EBUSY;
 
+	/* ->put_pages might need to allocate memory for the bit17 swizzle
+	 * array, hence protect them from being reaped by removing them from gtt
+	 * lists early. */
+	list_del(&obj->gtt_list);
+
 	ops->put_pages(obj);
 	obj->pages = NULL;
 
-	list_del(&obj->gtt_list);
 	if (i915_gem_object_is_purgeable(obj))
 		i915_gem_object_truncate(obj);
 
@@ -1857,11 +1861,11 @@ i915_gem_object_get_pages(struct drm_i915_gem_object *obj)
 
 void
 i915_gem_object_move_to_active(struct drm_i915_gem_object *obj,
-			       struct intel_ring_buffer *ring,
-			       u32 seqno)
+			       struct intel_ring_buffer *ring)
 {
 	struct drm_device *dev = obj->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 seqno = intel_ring_get_seqno(ring);
 
 	BUG_ON(ring == NULL);
 	obj->ring = ring;
@@ -1922,26 +1926,54 @@ i915_gem_object_move_to_inactive(struct drm_i915_gem_object *obj)
 	WARN_ON(i915_verify_lists(dev));
 }
 
-static u32
-i915_gem_get_seqno(struct drm_device *dev)
+static int
+i915_gem_handle_seqno_wrap(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	u32 seqno = dev_priv->next_seqno;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_ring_buffer *ring;
+	int ret, i, j;
 
-	/* reserve 0 for non-seqno */
-	if (++dev_priv->next_seqno == 0)
-		dev_priv->next_seqno = 1;
+	/* The hardware uses various monotonic 32-bit counters, if we
+	 * detect that they will wraparound we need to idle the GPU
+	 * and reset those counters.
+	 */
+	ret = 0;
+	for_each_ring(ring, dev_priv, i) {
+		for (j = 0; j < ARRAY_SIZE(ring->sync_seqno); j++)
+			ret |= ring->sync_seqno[j] != 0;
+	}
+	if (ret == 0)
+		return ret;
+
+	ret = i915_gpu_idle(dev);
+	if (ret)
+		return ret;
+
+	i915_gem_retire_requests(dev);
+	for_each_ring(ring, dev_priv, i) {
+		for (j = 0; j < ARRAY_SIZE(ring->sync_seqno); j++)
+			ring->sync_seqno[j] = 0;
+	}
 
-	return seqno;
+	return 0;
 }
 
-u32
-i915_gem_next_request_seqno(struct intel_ring_buffer *ring)
+int
+i915_gem_get_seqno(struct drm_device *dev, u32 *seqno)
 {
-	if (ring->outstanding_lazy_request == 0)
-		ring->outstanding_lazy_request = i915_gem_get_seqno(ring->dev);
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	/* reserve 0 for non-seqno */
+	if (dev_priv->next_seqno == 0) {
+		int ret = i915_gem_handle_seqno_wrap(dev);
+		if (ret)
+			return ret;
+
+		dev_priv->next_seqno = 1;
+	}
 
-	return ring->outstanding_lazy_request;
+	*seqno = dev_priv->next_seqno++;
+	return 0;
 }
 
 int
@@ -1952,7 +1984,6 @@ i915_add_request(struct intel_ring_buffer *ring,
 	drm_i915_private_t *dev_priv = ring->dev->dev_private;
 	struct drm_i915_gem_request *request;
 	u32 request_ring_position;
-	u32 seqno;
 	int was_empty;
 	int ret;
 
@@ -1971,7 +2002,6 @@ i915_add_request(struct intel_ring_buffer *ring,
 	if (request == NULL)
 		return -ENOMEM;
 
-	seqno = i915_gem_next_request_seqno(ring);
 
 	/* Record the position of the start of the request so that
 	 * should we detect the updated seqno part-way through the
@@ -1980,15 +2010,13 @@ i915_add_request(struct intel_ring_buffer *ring,
 	 */
 	request_ring_position = intel_ring_get_tail(ring);
 
-	ret = ring->add_request(ring, &seqno);
+	ret = ring->add_request(ring);
 	if (ret) {
 		kfree(request);
 		return ret;
 	}
 
-	trace_i915_gem_request_add(ring, seqno);
-
-	request->seqno = seqno;
+	request->seqno = intel_ring_get_seqno(ring);
 	request->ring = ring;
 	request->tail = request_ring_position;
 	request->emitted_jiffies = jiffies;
@@ -2006,6 +2034,7 @@ i915_add_request(struct intel_ring_buffer *ring,
 		spin_unlock(&file_priv->mm.lock);
 	}
 
+	trace_i915_gem_request_add(ring, request->seqno);
 	ring->outstanding_lazy_request = 0;
 
 	if (!dev_priv->mm.suspended) {
@@ -2022,7 +2051,7 @@ i915_add_request(struct intel_ring_buffer *ring,
 	}
 
 	if (out_seqno)
-		*out_seqno = seqno;
+		*out_seqno = request->seqno;
 	return 0;
 }
 
@@ -2120,7 +2149,6 @@ void
 i915_gem_retire_requests_ring(struct intel_ring_buffer *ring)
 {
 	uint32_t seqno;
-	int i;
 
 	if (list_empty(&ring->request_list))
 		return;
@@ -2129,10 +2157,6 @@ i915_gem_retire_requests_ring(struct intel_ring_buffer *ring)
 
 	seqno = ring->get_seqno(ring, true);
 
-	for (i = 0; i < ARRAY_SIZE(ring->sync_seqno); i++)
-		if (seqno >= ring->sync_seqno[i])
-			ring->sync_seqno[i] = 0;
-
 	while (!list_empty(&ring->request_list)) {
 		struct drm_i915_gem_request *request;
 
@@ -2377,7 +2401,11 @@ i915_gem_object_sync(struct drm_i915_gem_object *obj,
 
 	ret = to->sync_to(to, from, seqno);
 	if (!ret)
-		from->sync_seqno[idx] = seqno;
+		/* We use last_read_seqno because sync_to()
+		 * might have just caused seqno wrap under
+		 * the radar.
+		 */
+		from->sync_seqno[idx] = obj->last_read_seqno;
 
 	return ret;
 }
@@ -2460,14 +2488,6 @@ i915_gem_object_unbind(struct drm_i915_gem_object *obj)
 	return 0;
 }
 
-static int i915_ring_idle(struct intel_ring_buffer *ring)
-{
-	if (list_empty(&ring->active_list))
-		return 0;
-
-	return i915_wait_seqno(ring, i915_gem_next_request_seqno(ring));
-}
-
 int i915_gpu_idle(struct drm_device *dev)
 {
 	drm_i915_private_t *dev_priv = dev->dev_private;
@@ -2480,7 +2500,7 @@ int i915_gpu_idle(struct drm_device *dev)
 		if (ret)
 			return ret;
 
-		ret = i915_ring_idle(ring);
+		ret = intel_ring_idle(ring);
 		if (ret)
 			return ret;
 	}
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index 0e510df80d7..a3f06bcad55 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -410,9 +410,8 @@ static int do_switch(struct i915_hw_context *to)
 	 * MI_SET_CONTEXT instead of when the next seqno has completed.
 	 */
 	if (from_obj != NULL) {
-		u32 seqno = i915_gem_next_request_seqno(ring);
 		from_obj->base.read_domains = I915_GEM_DOMAIN_INSTRUCTION;
-		i915_gem_object_move_to_active(from_obj, ring, seqno);
+		i915_gem_object_move_to_active(from_obj, ring);
 		/* As long as MI_SET_CONTEXT is serializing, ie. it flushes the
 		 * whole damn pipeline, we don't need to explicitly mark the
 		 * object dirty. The only exception is that the context must be
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index 48e4317e72d..ee8f97f0539 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -713,8 +713,7 @@ validate_exec_list(struct drm_i915_gem_exec_object2 *exec,
 
 static void
 i915_gem_execbuffer_move_to_active(struct list_head *objects,
-				   struct intel_ring_buffer *ring,
-				   u32 seqno)
+				   struct intel_ring_buffer *ring)
 {
 	struct drm_i915_gem_object *obj;
 
@@ -726,10 +725,10 @@ i915_gem_execbuffer_move_to_active(struct list_head *objects,
 		obj->base.write_domain = obj->base.pending_write_domain;
 		obj->fenced_gpu_access = obj->pending_fenced_gpu_access;
 
-		i915_gem_object_move_to_active(obj, ring, seqno);
+		i915_gem_object_move_to_active(obj, ring);
 		if (obj->base.write_domain) {
 			obj->dirty = 1;
-			obj->last_write_seqno = seqno;
+			obj->last_write_seqno = intel_ring_get_seqno(ring);
 			if (obj->pin_count) /* check for potential scanout */
 				intel_mark_fb_busy(obj);
 		}
@@ -789,7 +788,6 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	struct intel_ring_buffer *ring;
 	u32 ctx_id = i915_execbuffer2_get_context_id(*args);
 	u32 exec_start, exec_len;
-	u32 seqno;
 	u32 mask;
 	u32 flags;
 	int ret, mode, i;
@@ -994,22 +992,6 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	if (ret)
 		goto err;
 
-	seqno = i915_gem_next_request_seqno(ring);
-	for (i = 0; i < ARRAY_SIZE(ring->sync_seqno); i++) {
-		if (seqno < ring->sync_seqno[i]) {
-			/* The GPU can not handle its semaphore value wrapping,
-			 * so every billion or so execbuffers, we need to stall
-			 * the GPU in order to reset the counters.
-			 */
-			ret = i915_gpu_idle(dev);
-			if (ret)
-				goto err;
-			i915_gem_retire_requests(dev);
-
-			BUG_ON(ring->sync_seqno[i]);
-		}
-	}
-
 	ret = i915_switch_context(ring, file, ctx_id);
 	if (ret)
 		goto err;
@@ -1035,8 +1017,6 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 			goto err;
 	}
 
-	trace_i915_gem_ring_dispatch(ring, seqno, flags);
-
 	exec_start = batch_obj->gtt_offset + args->batch_start_offset;
 	exec_len = args->batch_len;
 	if (cliprects) {
@@ -1060,7 +1040,9 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 			goto err;
 	}
 
-	i915_gem_execbuffer_move_to_active(&objects, ring, seqno);
+	trace_i915_gem_ring_dispatch(ring, intel_ring_get_seqno(ring), flags);
+
+	i915_gem_execbuffer_move_to_active(&objects, ring);
 	i915_gem_execbuffer_retire_commands(dev, file, ring);
 
 err:
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index f7ac61ee150..2c150dee78a 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -639,6 +639,10 @@ int i915_gem_gtt_init(struct drm_device *dev)
 	if (!pci_set_dma_mask(dev->pdev, DMA_BIT_MASK(40)))
 		pci_set_consistent_dma_mask(dev->pdev, DMA_BIT_MASK(40));
 
+#ifdef CONFIG_INTEL_IOMMU
+	dev_priv->mm.gtt->needs_dmar = 1;
+#endif
+
 	/* For GEN6+ the PTEs for the ggtt live at 2MB + BAR0 */
 	gtt_bus_addr = pci_resource_start(dev->pdev, 0) + (2<<20);
 	dev_priv->mm.gtt->gma_bus_addr = pci_resource_start(dev->pdev, 2);
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 2604867e6b7..a4dc97f8b9f 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -1120,6 +1120,8 @@ static void i915_record_ring_state(struct drm_device *dev,
 			= I915_READ(RING_SYNC_0(ring->mmio_base));
 		error->semaphore_mboxes[ring->id][1]
 			= I915_READ(RING_SYNC_1(ring->mmio_base));
+		error->semaphore_seqno[ring->id][0] = ring->sync_seqno[0];
+		error->semaphore_seqno[ring->id][1] = ring->sync_seqno[1];
 	}
 
 	if (INTEL_INFO(dev)->gen >= 4) {
@@ -1464,7 +1466,9 @@ static void i915_pageflip_stall_check(struct drm_device *dev, int pipe)
 	spin_lock_irqsave(&dev->event_lock, flags);
 	work = intel_crtc->unpin_work;
 
-	if (work == NULL || work->pending || !work->enable_stall_check) {
+	if (work == NULL ||
+	    atomic_read(&work->pending) >= INTEL_FLIP_COMPLETE ||
+	    !work->enable_stall_check) {
 		/* Either the pending flip IRQ arrived, or we're too early. Don't check */
 		spin_unlock_irqrestore(&dev->event_lock, flags);
 		return;
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 97fbd9d1823..3f75cfaf1c3 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -3843,7 +3843,9 @@
 #define  FDI_PHASE_SYNC_EN(pipe) (1<<(FDIA_PHASE_SYNC_SHIFT_EN - ((pipe) * 2)))
 #define  FDI_BC_BIFURCATION_SELECT	(1 << 12)
 #define SOUTH_CHICKEN2		0xc2004
-#define  DPLS_EDP_PPS_FIX_DIS	(1<<0)
+#define  FDI_MPHY_IOSFSB_RESET_STATUS	(1<<13)
+#define  FDI_MPHY_IOSFSB_RESET_CTL	(1<<12)
+#define  DPLS_EDP_PPS_FIX_DIS		(1<<0)
 
 #define _FDI_RXA_CHICKEN         0xc200c
 #define _FDI_RXB_CHICKEN         0xc2010
@@ -3915,6 +3917,7 @@
 #define  FDI_FS_ERRC_ENABLE		(1<<27)
 #define  FDI_FE_ERRC_ENABLE		(1<<26)
 #define  FDI_DP_PORT_WIDTH_X8           (7<<19)
+#define  FDI_RX_POLARITY_REVERSED_LPT	(1<<16)
 #define  FDI_8BPC                       (0<<16)
 #define  FDI_10BPC                      (1<<16)
 #define  FDI_6BPC                       (2<<16)
@@ -4534,6 +4537,10 @@
 #define SBI_ADDR			0xC6000
 #define SBI_DATA			0xC6004
 #define SBI_CTL_STAT			0xC6008
+#define  SBI_CTL_DEST_ICLK		(0x0<<16)
+#define  SBI_CTL_DEST_MPHY		(0x1<<16)
+#define  SBI_CTL_OP_IORD		(0x2<<8)
+#define  SBI_CTL_OP_IOWR		(0x3<<8)
 #define  SBI_CTL_OP_CRRD		(0x6<<8)
 #define  SBI_CTL_OP_CRWR		(0x7<<8)
 #define  SBI_RESPONSE_FAIL		(0x1<<1)
@@ -4551,10 +4558,12 @@
 #define   SBI_SSCDIVINTPHASE_PROPAGATE		(1<<0)
 #define  SBI_SSCCTL				0x020c
 #define  SBI_SSCCTL6				0x060C
+#define   SBI_SSCCTL_PATHALT			(1<<3)
 #define   SBI_SSCCTL_DISABLE			(1<<0)
 #define  SBI_SSCAUXDIV6				0x0610
 #define   SBI_SSCAUXDIV_FINALDIV2SEL(x)		((x)<<4)
 #define  SBI_DBUFF0				0x2a00
+#define   SBI_DBUFF0_ENABLE			(1<<0)
 
 /* LPT PIXCLK_GATE */
 #define PIXCLK_GATE			0xC6020
diff --git a/drivers/gpu/drm/i915/i915_sysfs.c b/drivers/gpu/drm/i915/i915_sysfs.c
index 3bf51d58319..9462081b1e6 100644
--- a/drivers/gpu/drm/i915/i915_sysfs.c
+++ b/drivers/gpu/drm/i915/i915_sysfs.c
@@ -97,7 +97,7 @@ static struct attribute_group rc6_attr_group = {
 
 static int l3_access_valid(struct drm_device *dev, loff_t offset)
 {
-	if (!IS_IVYBRIDGE(dev))
+	if (!HAS_L3_GPU_CACHE(dev))
 		return -EPERM;
 
 	if (offset % 4 != 0)
diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c
index 331af3bc689..9293878ec7e 100644
--- a/drivers/gpu/drm/i915/intel_crt.c
+++ b/drivers/gpu/drm/i915/intel_crt.c
@@ -198,6 +198,11 @@ static int intel_crt_mode_valid(struct drm_connector *connector,
 	if (mode->clock > max_clock)
 		return MODE_CLOCK_HIGH;
 
+	/* The FDI receiver on LPT only supports 8bpc and only has 2 lanes. */
+	if (HAS_PCH_LPT(dev) &&
+	    (ironlake_get_lanes_required(mode->clock, 270000, 24) > 2))
+		return MODE_CLOCK_HIGH;
+
 	return MODE_OK;
 }
 
@@ -793,4 +798,12 @@ void intel_crt_init(struct drm_device *dev)
 	crt->force_hotplug_required = 0;
 
 	dev_priv->hotplug_supported_mask |= CRT_HOTPLUG_INT_STATUS;
+
+	/*
+	 * TODO: find a proper way to discover whether we need to set the
+	 * polarity reversal bit or not, instead of relying on the BIOS.
+	 */
+	if (HAS_PCH_LPT(dev))
+		dev_priv->fdi_rx_polarity_reversed =
+		     !!(I915_READ(_FDI_RXA_CTL) & FDI_RX_POLARITY_REVERSED_LPT);
 }
diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c
index 852012b6fc5..4bad0f72401 100644
--- a/drivers/gpu/drm/i915/intel_ddi.c
+++ b/drivers/gpu/drm/i915/intel_ddi.c
@@ -138,6 +138,19 @@ static const long hsw_ddi_buf_ctl_values[] = {
 	DDI_BUF_EMP_800MV_3_5DB_HSW
 };
 
+static void intel_wait_ddi_buf_idle(struct drm_i915_private *dev_priv,
+				    enum port port)
+{
+	uint32_t reg = DDI_BUF_CTL(port);
+	int i;
+
+	for (i = 0; i < 8; i++) {
+		udelay(1);
+		if (I915_READ(reg) & DDI_BUF_IS_IDLE)
+			return;
+	}
+	DRM_ERROR("Timeout waiting for DDI BUF %c idle bit\n", port_name(port));
+}
 
 /* Starting with Haswell, different DDI ports can work in FDI mode for
  * connection to the PCH-located connectors. For this, it is necessary to train
@@ -167,6 +180,8 @@ void hsw_fdi_link_train(struct drm_crtc *crtc)
 	/* Enable the PCH Receiver FDI PLL */
 	rx_ctl_val = FDI_RX_PLL_ENABLE | FDI_RX_ENHANCE_FRAME_ENABLE |
 		     ((intel_crtc->fdi_lanes - 1) << 19);
+	if (dev_priv->fdi_rx_polarity_reversed)
+		rx_ctl_val |= FDI_RX_POLARITY_REVERSED_LPT;
 	I915_WRITE(_FDI_RXA_CTL, rx_ctl_val);
 	POSTING_READ(_FDI_RXA_CTL);
 	udelay(220);
@@ -231,18 +246,30 @@ void hsw_fdi_link_train(struct drm_crtc *crtc)
 			return;
 		}
 
+		temp = I915_READ(DDI_BUF_CTL(PORT_E));
+		temp &= ~DDI_BUF_CTL_ENABLE;
+		I915_WRITE(DDI_BUF_CTL(PORT_E), temp);
+		POSTING_READ(DDI_BUF_CTL(PORT_E));
+
 		/* Disable DP_TP_CTL and FDI_RX_CTL and retry */
-		I915_WRITE(DP_TP_CTL(PORT_E),
-			   I915_READ(DP_TP_CTL(PORT_E)) & ~DP_TP_CTL_ENABLE);
+		temp = I915_READ(DP_TP_CTL(PORT_E));
+		temp &= ~(DP_TP_CTL_ENABLE | DP_TP_CTL_LINK_TRAIN_MASK);
+		temp |= DP_TP_CTL_LINK_TRAIN_PAT1;
+		I915_WRITE(DP_TP_CTL(PORT_E), temp);
+		POSTING_READ(DP_TP_CTL(PORT_E));
+
+		intel_wait_ddi_buf_idle(dev_priv, PORT_E);
 
 		rx_ctl_val &= ~FDI_RX_ENABLE;
 		I915_WRITE(_FDI_RXA_CTL, rx_ctl_val);
+		POSTING_READ(_FDI_RXA_CTL);
 
 		/* Reset FDI_RX_MISC pwrdn lanes */
 		temp = I915_READ(_FDI_RXA_MISC);
 		temp &= ~(FDI_RX_PWRDN_LANE1_MASK | FDI_RX_PWRDN_LANE0_MASK);
 		temp |= FDI_RX_PWRDN_LANE1_VAL(2) | FDI_RX_PWRDN_LANE0_VAL(2);
 		I915_WRITE(_FDI_RXA_MISC, temp);
+		POSTING_READ(_FDI_RXA_MISC);
 	}
 
 	DRM_ERROR("FDI link training failed!\n");
@@ -1222,20 +1249,6 @@ static void intel_ddi_pre_enable(struct intel_encoder *intel_encoder)
 	}
 }
 
-static void intel_wait_ddi_buf_idle(struct drm_i915_private *dev_priv,
-				    enum port port)
-{
-	uint32_t reg = DDI_BUF_CTL(port);
-	int i;
-
-	for (i = 0; i < 8; i++) {
-		udelay(1);
-		if (I915_READ(reg) & DDI_BUF_IS_IDLE)
-			return;
-	}
-	DRM_ERROR("Timeout waiting for DDI BUF %c idle bit\n", port_name(port));
-}
-
 static void intel_ddi_post_disable(struct intel_encoder *intel_encoder)
 {
 	struct drm_encoder *encoder = &intel_encoder->base;
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index de51489de23..5d127e06895 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -1506,24 +1506,26 @@ static void intel_disable_pll(struct drm_i915_private *dev_priv, enum pipe pipe)
 
 /* SBI access */
 static void
-intel_sbi_write(struct drm_i915_private *dev_priv, u16 reg, u32 value)
+intel_sbi_write(struct drm_i915_private *dev_priv, u16 reg, u32 value,
+		enum intel_sbi_destination destination)
 {
 	unsigned long flags;
+	u32 tmp;
 
 	spin_lock_irqsave(&dev_priv->dpio_lock, flags);
-	if (wait_for((I915_READ(SBI_CTL_STAT) & SBI_BUSY) == 0,
-				100)) {
+	if (wait_for((I915_READ(SBI_CTL_STAT) & SBI_BUSY) == 0, 100)) {
 		DRM_ERROR("timeout waiting for SBI to become ready\n");
 		goto out_unlock;
 	}
 
-	I915_WRITE(SBI_ADDR,
-			(reg << 16));
-	I915_WRITE(SBI_DATA,
-			value);
-	I915_WRITE(SBI_CTL_STAT,
-			SBI_BUSY |
-			SBI_CTL_OP_CRWR);
+	I915_WRITE(SBI_ADDR, (reg << 16));
+	I915_WRITE(SBI_DATA, value);
+
+	if (destination == SBI_ICLK)
+		tmp = SBI_CTL_DEST_ICLK | SBI_CTL_OP_CRWR;
+	else
+		tmp = SBI_CTL_DEST_MPHY | SBI_CTL_OP_IOWR;
+	I915_WRITE(SBI_CTL_STAT, SBI_BUSY | tmp);
 
 	if (wait_for((I915_READ(SBI_CTL_STAT) & (SBI_BUSY | SBI_RESPONSE_FAIL)) == 0,
 				100)) {
@@ -1536,23 +1538,25 @@ out_unlock:
 }
 
 static u32
-intel_sbi_read(struct drm_i915_private *dev_priv, u16 reg)
+intel_sbi_read(struct drm_i915_private *dev_priv, u16 reg,
+	       enum intel_sbi_destination destination)
 {
 	unsigned long flags;
 	u32 value = 0;
 
 	spin_lock_irqsave(&dev_priv->dpio_lock, flags);
-	if (wait_for((I915_READ(SBI_CTL_STAT) & SBI_BUSY) == 0,
-				100)) {
+	if (wait_for((I915_READ(SBI_CTL_STAT) & SBI_BUSY) == 0, 100)) {
 		DRM_ERROR("timeout waiting for SBI to become ready\n");
 		goto out_unlock;
 	}
 
-	I915_WRITE(SBI_ADDR,
-			(reg << 16));
-	I915_WRITE(SBI_CTL_STAT,
-			SBI_BUSY |
-			SBI_CTL_OP_CRRD);
+	I915_WRITE(SBI_ADDR, (reg << 16));
+
+	if (destination == SBI_ICLK)
+		value = SBI_CTL_DEST_ICLK | SBI_CTL_OP_CRRD;
+	else
+		value = SBI_CTL_DEST_MPHY | SBI_CTL_OP_IORD;
+	I915_WRITE(SBI_CTL_STAT, value | SBI_BUSY);
 
 	if (wait_for((I915_READ(SBI_CTL_STAT) & (SBI_BUSY | SBI_RESPONSE_FAIL)) == 0,
 				100)) {
@@ -2424,18 +2428,6 @@ static void intel_fdi_normal_train(struct drm_crtc *crtc)
 			   FDI_FE_ERRC_ENABLE);
 }
 
-static void cpt_phase_pointer_enable(struct drm_device *dev, int pipe)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 flags = I915_READ(SOUTH_CHICKEN1);
-
-	flags |= FDI_PHASE_SYNC_OVR(pipe);
-	I915_WRITE(SOUTH_CHICKEN1, flags); /* once to unlock... */
-	flags |= FDI_PHASE_SYNC_EN(pipe);
-	I915_WRITE(SOUTH_CHICKEN1, flags); /* then again to enable */
-	POSTING_READ(SOUTH_CHICKEN1);
-}
-
 static void ivb_modeset_global_resources(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -2610,8 +2602,6 @@ static void gen6_fdi_link_train(struct drm_crtc *crtc)
 	POSTING_READ(reg);
 	udelay(150);
 
-	cpt_phase_pointer_enable(dev, pipe);
-
 	for (i = 0; i < 4; i++) {
 		reg = FDI_TX_CTL(pipe);
 		temp = I915_READ(reg);
@@ -2744,8 +2734,6 @@ static void ivb_manual_fdi_link_train(struct drm_crtc *crtc)
 	POSTING_READ(reg);
 	udelay(150);
 
-	cpt_phase_pointer_enable(dev, pipe);
-
 	for (i = 0; i < 4; i++) {
 		reg = FDI_TX_CTL(pipe);
 		temp = I915_READ(reg);
@@ -2884,17 +2872,6 @@ static void ironlake_fdi_pll_disable(struct intel_crtc *intel_crtc)
 	udelay(100);
 }
 
-static void cpt_phase_pointer_disable(struct drm_device *dev, int pipe)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 flags = I915_READ(SOUTH_CHICKEN1);
-
-	flags &= ~(FDI_PHASE_SYNC_EN(pipe));
-	I915_WRITE(SOUTH_CHICKEN1, flags); /* once to disable... */
-	flags &= ~(FDI_PHASE_SYNC_OVR(pipe));
-	I915_WRITE(SOUTH_CHICKEN1, flags); /* then again to lock */
-	POSTING_READ(SOUTH_CHICKEN1);
-}
 static void ironlake_fdi_disable(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
@@ -2921,8 +2898,6 @@ static void ironlake_fdi_disable(struct drm_crtc *crtc)
 	/* Ironlake workaround, disable clock pointer after downing FDI */
 	if (HAS_PCH_IBX(dev)) {
 		I915_WRITE(FDI_RX_CHICKEN(pipe), FDI_RX_PHASE_SYNC_POINTER_OVR);
-	} else if (HAS_PCH_CPT(dev)) {
-		cpt_phase_pointer_disable(dev, pipe);
 	}
 
 	/* still set train pattern 1 */
@@ -3024,8 +2999,9 @@ static void lpt_program_iclkip(struct drm_crtc *crtc)
 
 	/* Disable SSCCTL */
 	intel_sbi_write(dev_priv, SBI_SSCCTL6,
-				intel_sbi_read(dev_priv, SBI_SSCCTL6) |
-					SBI_SSCCTL_DISABLE);
+			intel_sbi_read(dev_priv, SBI_SSCCTL6, SBI_ICLK) |
+				SBI_SSCCTL_DISABLE,
+			SBI_ICLK);
 
 	/* 20MHz is a corner case which is out of range for the 7-bit divisor */
 	if (crtc->mode.clock == 20000) {
@@ -3066,33 +3042,25 @@ static void lpt_program_iclkip(struct drm_crtc *crtc)
 			phaseinc);
 
 	/* Program SSCDIVINTPHASE6 */
-	temp = intel_sbi_read(dev_priv, SBI_SSCDIVINTPHASE6);
+	temp = intel_sbi_read(dev_priv, SBI_SSCDIVINTPHASE6, SBI_ICLK);
 	temp &= ~SBI_SSCDIVINTPHASE_DIVSEL_MASK;
 	temp |= SBI_SSCDIVINTPHASE_DIVSEL(divsel);
 	temp &= ~SBI_SSCDIVINTPHASE_INCVAL_MASK;
 	temp |= SBI_SSCDIVINTPHASE_INCVAL(phaseinc);
 	temp |= SBI_SSCDIVINTPHASE_DIR(phasedir);
 	temp |= SBI_SSCDIVINTPHASE_PROPAGATE;
-
-	intel_sbi_write(dev_priv,
-			SBI_SSCDIVINTPHASE6,
-			temp);
+	intel_sbi_write(dev_priv, SBI_SSCDIVINTPHASE6, temp, SBI_ICLK);
 
 	/* Program SSCAUXDIV */
-	temp = intel_sbi_read(dev_priv, SBI_SSCAUXDIV6);
+	temp = intel_sbi_read(dev_priv, SBI_SSCAUXDIV6, SBI_ICLK);
 	temp &= ~SBI_SSCAUXDIV_FINALDIV2SEL(1);
 	temp |= SBI_SSCAUXDIV_FINALDIV2SEL(auxdiv);
-	intel_sbi_write(dev_priv,
-			SBI_SSCAUXDIV6,
-			temp);
-
+	intel_sbi_write(dev_priv, SBI_SSCAUXDIV6, temp, SBI_ICLK);
 
 	/* Enable modulator and associated divider */
-	temp = intel_sbi_read(dev_priv, SBI_SSCCTL6);
+	temp = intel_sbi_read(dev_priv, SBI_SSCCTL6, SBI_ICLK);
 	temp &= ~SBI_SSCCTL_DISABLE;
-	intel_sbi_write(dev_priv,
-			SBI_SSCCTL6,
-			temp);
+	intel_sbi_write(dev_priv, SBI_SSCCTL6, temp, SBI_ICLK);
 
 	/* Wait for initialization time */
 	udelay(24);
@@ -4878,10 +4846,7 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc,
 	return ret;
 }
 
-/*
- * Initialize reference clocks when the driver loads
- */
-void ironlake_init_pch_refclk(struct drm_device *dev)
+static void ironlake_init_pch_refclk(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_mode_config *mode_config = &dev->mode_config;
@@ -4995,6 +4960,182 @@ void ironlake_init_pch_refclk(struct drm_device *dev)
 	}
 }
 
+/* Sequence to enable CLKOUT_DP for FDI usage and configure PCH FDI I/O. */
+static void lpt_init_pch_refclk(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_mode_config *mode_config = &dev->mode_config;
+	struct intel_encoder *encoder;
+	bool has_vga = false;
+	bool is_sdv = false;
+	u32 tmp;
+
+	list_for_each_entry(encoder, &mode_config->encoder_list, base.head) {
+		switch (encoder->type) {
+		case INTEL_OUTPUT_ANALOG:
+			has_vga = true;
+			break;
+		}
+	}
+
+	if (!has_vga)
+		return;
+
+	/* XXX: Rip out SDV support once Haswell ships for real. */
+	if (IS_HASWELL(dev) && (dev->pci_device & 0xFF00) == 0x0C00)
+		is_sdv = true;
+
+	tmp = intel_sbi_read(dev_priv, SBI_SSCCTL, SBI_ICLK);
+	tmp &= ~SBI_SSCCTL_DISABLE;
+	tmp |= SBI_SSCCTL_PATHALT;
+	intel_sbi_write(dev_priv, SBI_SSCCTL, tmp, SBI_ICLK);
+
+	udelay(24);
+
+	tmp = intel_sbi_read(dev_priv, SBI_SSCCTL, SBI_ICLK);
+	tmp &= ~SBI_SSCCTL_PATHALT;
+	intel_sbi_write(dev_priv, SBI_SSCCTL, tmp, SBI_ICLK);
+
+	if (!is_sdv) {
+		tmp = I915_READ(SOUTH_CHICKEN2);
+		tmp |= FDI_MPHY_IOSFSB_RESET_CTL;
+		I915_WRITE(SOUTH_CHICKEN2, tmp);
+
+		if (wait_for_atomic_us(I915_READ(SOUTH_CHICKEN2) &
+				       FDI_MPHY_IOSFSB_RESET_STATUS, 100))
+			DRM_ERROR("FDI mPHY reset assert timeout\n");
+
+		tmp = I915_READ(SOUTH_CHICKEN2);
+		tmp &= ~FDI_MPHY_IOSFSB_RESET_CTL;
+		I915_WRITE(SOUTH_CHICKEN2, tmp);
+
+		if (wait_for_atomic_us((I915_READ(SOUTH_CHICKEN2) &
+				        FDI_MPHY_IOSFSB_RESET_STATUS) == 0,
+				       100))
+			DRM_ERROR("FDI mPHY reset de-assert timeout\n");
+	}
+
+	tmp = intel_sbi_read(dev_priv, 0x8008, SBI_MPHY);
+	tmp &= ~(0xFF << 24);
+	tmp |= (0x12 << 24);
+	intel_sbi_write(dev_priv, 0x8008, tmp, SBI_MPHY);
+
+	if (!is_sdv) {
+		tmp = intel_sbi_read(dev_priv, 0x808C, SBI_MPHY);
+		tmp &= ~(0x3 << 6);
+		tmp |= (1 << 6) | (1 << 0);
+		intel_sbi_write(dev_priv, 0x808C, tmp, SBI_MPHY);
+	}
+
+	if (is_sdv) {
+		tmp = intel_sbi_read(dev_priv, 0x800C, SBI_MPHY);
+		tmp |= 0x7FFF;
+		intel_sbi_write(dev_priv, 0x800C, tmp, SBI_MPHY);
+	}
+
+	tmp = intel_sbi_read(dev_priv, 0x2008, SBI_MPHY);
+	tmp |= (1 << 11);
+	intel_sbi_write(dev_priv, 0x2008, tmp, SBI_MPHY);
+
+	tmp = intel_sbi_read(dev_priv, 0x2108, SBI_MPHY);
+	tmp |= (1 << 11);
+	intel_sbi_write(dev_priv, 0x2108, tmp, SBI_MPHY);
+
+	if (is_sdv) {
+		tmp = intel_sbi_read(dev_priv, 0x2038, SBI_MPHY);
+		tmp |= (0x3F << 24) | (0xF << 20) | (0xF << 16);
+		intel_sbi_write(dev_priv, 0x2038, tmp, SBI_MPHY);
+
+		tmp = intel_sbi_read(dev_priv, 0x2138, SBI_MPHY);
+		tmp |= (0x3F << 24) | (0xF << 20) | (0xF << 16);
+		intel_sbi_write(dev_priv, 0x2138, tmp, SBI_MPHY);
+
+		tmp = intel_sbi_read(dev_priv, 0x203C, SBI_MPHY);
+		tmp |= (0x3F << 8);
+		intel_sbi_write(dev_priv, 0x203C, tmp, SBI_MPHY);
+
+		tmp = intel_sbi_read(dev_priv, 0x213C, SBI_MPHY);
+		tmp |= (0x3F << 8);
+		intel_sbi_write(dev_priv, 0x213C, tmp, SBI_MPHY);
+	}
+
+	tmp = intel_sbi_read(dev_priv, 0x206C, SBI_MPHY);
+	tmp |= (1 << 24) | (1 << 21) | (1 << 18);
+	intel_sbi_write(dev_priv, 0x206C, tmp, SBI_MPHY);
+
+	tmp = intel_sbi_read(dev_priv, 0x216C, SBI_MPHY);
+	tmp |= (1 << 24) | (1 << 21) | (1 << 18);
+	intel_sbi_write(dev_priv, 0x216C, tmp, SBI_MPHY);
+
+	if (!is_sdv) {
+		tmp = intel_sbi_read(dev_priv, 0x2080, SBI_MPHY);
+		tmp &= ~(7 << 13);
+		tmp |= (5 << 13);
+		intel_sbi_write(dev_priv, 0x2080, tmp, SBI_MPHY);
+
+		tmp = intel_sbi_read(dev_priv, 0x2180, SBI_MPHY);
+		tmp &= ~(7 << 13);
+		tmp |= (5 << 13);
+		intel_sbi_write(dev_priv, 0x2180, tmp, SBI_MPHY);
+	}
+
+	tmp = intel_sbi_read(dev_priv, 0x208C, SBI_MPHY);
+	tmp &= ~0xFF;
+	tmp |= 0x1C;
+	intel_sbi_write(dev_priv, 0x208C, tmp, SBI_MPHY);
+
+	tmp = intel_sbi_read(dev_priv, 0x218C, SBI_MPHY);
+	tmp &= ~0xFF;
+	tmp |= 0x1C;
+	intel_sbi_write(dev_priv, 0x218C, tmp, SBI_MPHY);
+
+	tmp = intel_sbi_read(dev_priv, 0x2098, SBI_MPHY);
+	tmp &= ~(0xFF << 16);
+	tmp |= (0x1C << 16);
+	intel_sbi_write(dev_priv, 0x2098, tmp, SBI_MPHY);
+
+	tmp = intel_sbi_read(dev_priv, 0x2198, SBI_MPHY);
+	tmp &= ~(0xFF << 16);
+	tmp |= (0x1C << 16);
+	intel_sbi_write(dev_priv, 0x2198, tmp, SBI_MPHY);
+
+	if (!is_sdv) {
+		tmp = intel_sbi_read(dev_priv, 0x20C4, SBI_MPHY);
+		tmp |= (1 << 27);
+		intel_sbi_write(dev_priv, 0x20C4, tmp, SBI_MPHY);
+
+		tmp = intel_sbi_read(dev_priv, 0x21C4, SBI_MPHY);
+		tmp |= (1 << 27);
+		intel_sbi_write(dev_priv, 0x21C4, tmp, SBI_MPHY);
+
+		tmp = intel_sbi_read(dev_priv, 0x20EC, SBI_MPHY);
+		tmp &= ~(0xF << 28);
+		tmp |= (4 << 28);
+		intel_sbi_write(dev_priv, 0x20EC, tmp, SBI_MPHY);
+
+		tmp = intel_sbi_read(dev_priv, 0x21EC, SBI_MPHY);
+		tmp &= ~(0xF << 28);
+		tmp |= (4 << 28);
+		intel_sbi_write(dev_priv, 0x21EC, tmp, SBI_MPHY);
+	}
+
+	/* ULT uses SBI_GEN0, but ULT doesn't have VGA, so we don't care. */
+	tmp = intel_sbi_read(dev_priv, SBI_DBUFF0, SBI_ICLK);
+	tmp |= SBI_DBUFF0_ENABLE;
+	intel_sbi_write(dev_priv, SBI_DBUFF0, tmp, SBI_ICLK);
+}
+
+/*
+ * Initialize reference clocks when the driver loads
+ */
+void intel_init_pch_refclk(struct drm_device *dev)
+{
+	if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
+		ironlake_init_pch_refclk(dev);
+	else if (HAS_PCH_LPT(dev))
+		lpt_init_pch_refclk(dev);
+}
+
 static int ironlake_get_refclk(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
@@ -5239,6 +5380,17 @@ static bool ironlake_check_fdi_lanes(struct intel_crtc *intel_crtc)
 	}
 }
 
+int ironlake_get_lanes_required(int target_clock, int link_bw, int bpp)
+{
+	/*
+	 * Account for spread spectrum to avoid
+	 * oversubscribing the link. Max center spread
+	 * is 2.5%; use 5% for safety's sake.
+	 */
+	u32 bps = target_clock * bpp * 21 / 20;
+	return bps / (link_bw * 8) + 1;
+}
+
 static void ironlake_set_m_n(struct drm_crtc *crtc,
 			     struct drm_display_mode *mode,
 			     struct drm_display_mode *adjusted_mode)
@@ -5292,15 +5444,9 @@ static void ironlake_set_m_n(struct drm_crtc *crtc,
 	else
 		target_clock = adjusted_mode->clock;
 
-	if (!lane) {
-		/*
-		 * Account for spread spectrum to avoid
-		 * oversubscribing the link. Max center spread
-		 * is 2.5%; use 5% for safety's sake.
-		 */
-		u32 bps = target_clock * intel_crtc->bpp * 21 / 20;
-		lane = bps / (link_bw * 8) + 1;
-	}
+	if (!lane)
+		lane = ironlake_get_lanes_required(target_clock, link_bw,
+						   intel_crtc->bpp);
 
 	intel_crtc->fdi_lanes = lane;
 
@@ -6940,11 +7086,18 @@ static void do_intel_finish_page_flip(struct drm_device *dev,
 
 	spin_lock_irqsave(&dev->event_lock, flags);
 	work = intel_crtc->unpin_work;
-	if (work == NULL || !work->pending) {
+
+	/* Ensure we don't miss a work->pending update ... */
+	smp_rmb();
+
+	if (work == NULL || atomic_read(&work->pending) < INTEL_FLIP_COMPLETE) {
 		spin_unlock_irqrestore(&dev->event_lock, flags);
 		return;
 	}
 
+	/* and that the unpin work is consistent wrt ->pending. */
+	smp_rmb();
+
 	intel_crtc->unpin_work = NULL;
 
 	if (work->event)
@@ -6988,16 +7141,25 @@ void intel_prepare_page_flip(struct drm_device *dev, int plane)
 		to_intel_crtc(dev_priv->plane_to_crtc_mapping[plane]);
 	unsigned long flags;
 
+	/* NB: An MMIO update of the plane base pointer will also
+	 * generate a page-flip completion irq, i.e. every modeset
+	 * is also accompanied by a spurious intel_prepare_page_flip().
+	 */
 	spin_lock_irqsave(&dev->event_lock, flags);
-	if (intel_crtc->unpin_work) {
-		if ((++intel_crtc->unpin_work->pending) > 1)
-			DRM_ERROR("Prepared flip multiple times\n");
-	} else {
-		DRM_DEBUG_DRIVER("preparing flip with no unpin work?\n");
-	}
+	if (intel_crtc->unpin_work)
+		atomic_inc_not_zero(&intel_crtc->unpin_work->pending);
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 }
 
+inline static void intel_mark_page_flip_active(struct intel_crtc *intel_crtc)
+{
+	/* Ensure that the work item is consistent when activating it ... */
+	smp_wmb();
+	atomic_set(&intel_crtc->unpin_work->pending, INTEL_FLIP_PENDING);
+	/* and that it is marked active as soon as the irq could fire. */
+	smp_wmb();
+}
+
 static int intel_gen2_queue_flip(struct drm_device *dev,
 				 struct drm_crtc *crtc,
 				 struct drm_framebuffer *fb,
@@ -7031,6 +7193,8 @@ static int intel_gen2_queue_flip(struct drm_device *dev,
 	intel_ring_emit(ring, fb->pitches[0]);
 	intel_ring_emit(ring, obj->gtt_offset + intel_crtc->dspaddr_offset);
 	intel_ring_emit(ring, 0); /* aux display base address, unused */
+
+	intel_mark_page_flip_active(intel_crtc);
 	intel_ring_advance(ring);
 	return 0;
 
@@ -7071,6 +7235,7 @@ static int intel_gen3_queue_flip(struct drm_device *dev,
 	intel_ring_emit(ring, obj->gtt_offset + intel_crtc->dspaddr_offset);
 	intel_ring_emit(ring, MI_NOOP);
 
+	intel_mark_page_flip_active(intel_crtc);
 	intel_ring_advance(ring);
 	return 0;
 
@@ -7117,6 +7282,8 @@ static int intel_gen4_queue_flip(struct drm_device *dev,
 	pf = 0;
 	pipesrc = I915_READ(PIPESRC(intel_crtc->pipe)) & 0x0fff0fff;
 	intel_ring_emit(ring, pf | pipesrc);
+
+	intel_mark_page_flip_active(intel_crtc);
 	intel_ring_advance(ring);
 	return 0;
 
@@ -7159,6 +7326,8 @@ static int intel_gen6_queue_flip(struct drm_device *dev,
 	pf = 0;
 	pipesrc = I915_READ(PIPESRC(intel_crtc->pipe)) & 0x0fff0fff;
 	intel_ring_emit(ring, pf | pipesrc);
+
+	intel_mark_page_flip_active(intel_crtc);
 	intel_ring_advance(ring);
 	return 0;
 
@@ -7213,6 +7382,8 @@ static int intel_gen7_queue_flip(struct drm_device *dev,
 	intel_ring_emit(ring, (fb->pitches[0] | obj->tiling_mode));
 	intel_ring_emit(ring, obj->gtt_offset + intel_crtc->dspaddr_offset);
 	intel_ring_emit(ring, (MI_NOOP));
+
+	intel_mark_page_flip_active(intel_crtc);
 	intel_ring_advance(ring);
 	return 0;
 
@@ -8394,8 +8565,7 @@ static void intel_setup_outputs(struct drm_device *dev)
 			intel_encoder_clones(encoder);
 	}
 
-	if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
-		ironlake_init_pch_refclk(dev);
+	intel_init_pch_refclk(dev);
 
 	drm_helper_move_panel_connectors_to_head(dev);
 }
@@ -8999,7 +9169,8 @@ static void intel_sanitize_encoder(struct intel_encoder *encoder)
 
 /* Scan out the current hw modeset state, sanitizes it and maps it into the drm
  * and i915 state tracking structures. */
-void intel_modeset_setup_hw_state(struct drm_device *dev)
+void intel_modeset_setup_hw_state(struct drm_device *dev,
+				  bool force_restore)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum pipe pipe;
@@ -9098,7 +9269,15 @@ void intel_modeset_setup_hw_state(struct drm_device *dev)
 		intel_sanitize_crtc(crtc);
 	}
 
-	intel_modeset_update_staged_output_state(dev);
+	if (force_restore) {
+		for_each_pipe(pipe) {
+			crtc = to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
+			intel_set_mode(&crtc->base, &crtc->base.mode,
+				       crtc->base.x, crtc->base.y, crtc->base.fb);
+		}
+	} else {
+		intel_modeset_update_staged_output_state(dev);
+	}
 
 	intel_modeset_check_state(dev);
 
@@ -9111,7 +9290,7 @@ void intel_modeset_gem_init(struct drm_device *dev)
 
 	intel_setup_overlay(dev);
 
-	intel_modeset_setup_hw_state(dev);
+	intel_modeset_setup_hw_state(dev, false);
 }
 
 void intel_modeset_cleanup(struct drm_device *dev)
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 522061ca068..8a1bd4a3ad0 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -401,7 +401,10 @@ struct intel_unpin_work {
 	struct drm_i915_gem_object *old_fb_obj;
 	struct drm_i915_gem_object *pending_flip_obj;
 	struct drm_pending_vblank_event *event;
-	int pending;
+	atomic_t pending;
+#define INTEL_FLIP_INACTIVE	0
+#define INTEL_FLIP_PENDING	1
+#define INTEL_FLIP_COMPLETE	2
 	bool enable_stall_check;
 };
 
@@ -556,6 +559,7 @@ intel_pipe_to_cpu_transcoder(struct drm_i915_private *dev_priv,
 			     enum pipe pipe);
 extern void intel_wait_for_vblank(struct drm_device *dev, int pipe);
 extern void intel_wait_for_pipe_off(struct drm_device *dev, int pipe);
+extern int ironlake_get_lanes_required(int target_clock, int link_bw, int bpp);
 
 struct intel_load_detect_pipe {
 	struct drm_framebuffer *release_fb;
diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c
index 5c279b48df9..2ee9821b9d9 100644
--- a/drivers/gpu/drm/i915/intel_hdmi.c
+++ b/drivers/gpu/drm/i915/intel_hdmi.c
@@ -340,6 +340,8 @@ static void intel_hdmi_set_avi_infoframe(struct drm_encoder *encoder,
 	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK)
 		avi_if.body.avi.YQ_CN_PR |= DIP_AVI_PR_2;
 
+	avi_if.body.avi.VIC = drm_mode_cea_vic(adjusted_mode);
+
 	intel_set_infoframe(encoder, &avi_if);
 }
 
diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c
index 81502e8be26..b9a660a5367 100644
--- a/drivers/gpu/drm/i915/intel_lvds.c
+++ b/drivers/gpu/drm/i915/intel_lvds.c
@@ -532,7 +532,7 @@ static int intel_lid_notify(struct notifier_block *nb, unsigned long val,
 	dev_priv->modeset_on_lid = 0;
 
 	mutex_lock(&dev->mode_config.mutex);
-	intel_modeset_check_state(dev);
+	intel_modeset_setup_hw_state(dev, true);
 	mutex_unlock(&dev->mode_config.mutex);
 
 	return NOTIFY_OK;
diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c
index c758ad27747..bee8cb6108a 100644
--- a/drivers/gpu/drm/i915/intel_panel.c
+++ b/drivers/gpu/drm/i915/intel_panel.c
@@ -130,8 +130,9 @@ static int is_backlight_combination_mode(struct drm_device *dev)
 	return 0;
 }
 
-static u32 i915_read_blc_pwm_ctl(struct drm_i915_private *dev_priv)
+static u32 i915_read_blc_pwm_ctl(struct drm_device *dev)
 {
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 val;
 
 	/* Restore the CTL value if it lost, e.g. GPU reset */
@@ -141,21 +142,22 @@ static u32 i915_read_blc_pwm_ctl(struct drm_i915_private *dev_priv)
 		if (dev_priv->regfile.saveBLC_PWM_CTL2 == 0) {
 			dev_priv->regfile.saveBLC_PWM_CTL2 = val;
 		} else if (val == 0) {
-			I915_WRITE(BLC_PWM_PCH_CTL2,
-				   dev_priv->regfile.saveBLC_PWM_CTL2);
 			val = dev_priv->regfile.saveBLC_PWM_CTL2;
+			I915_WRITE(BLC_PWM_PCH_CTL2, val);
 		}
 	} else {
 		val = I915_READ(BLC_PWM_CTL);
 		if (dev_priv->regfile.saveBLC_PWM_CTL == 0) {
 			dev_priv->regfile.saveBLC_PWM_CTL = val;
-			dev_priv->regfile.saveBLC_PWM_CTL2 = I915_READ(BLC_PWM_CTL2);
+			if (INTEL_INFO(dev)->gen >= 4)
+				dev_priv->regfile.saveBLC_PWM_CTL2 =
+					I915_READ(BLC_PWM_CTL2);
 		} else if (val == 0) {
-			I915_WRITE(BLC_PWM_CTL,
-				   dev_priv->regfile.saveBLC_PWM_CTL);
-			I915_WRITE(BLC_PWM_CTL2,
-				   dev_priv->regfile.saveBLC_PWM_CTL2);
 			val = dev_priv->regfile.saveBLC_PWM_CTL;
+			I915_WRITE(BLC_PWM_CTL, val);
+			if (INTEL_INFO(dev)->gen >= 4)
+				I915_WRITE(BLC_PWM_CTL2,
+					   dev_priv->regfile.saveBLC_PWM_CTL2);
 		}
 	}
 
@@ -164,10 +166,9 @@ static u32 i915_read_blc_pwm_ctl(struct drm_i915_private *dev_priv)
 
 static u32 _intel_panel_get_max_backlight(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 max;
 
-	max = i915_read_blc_pwm_ctl(dev_priv);
+	max = i915_read_blc_pwm_ctl(dev);
 
 	if (HAS_PCH_SPLIT(dev)) {
 		max >>= 16;
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 9e619ada056..496caa73eb7 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -1325,10 +1325,11 @@ static void valleyview_update_wm(struct drm_device *dev)
 		   (planeb_wm << DSPFW_PLANEB_SHIFT) |
 		   planea_wm);
 	I915_WRITE(DSPFW2,
-		   (I915_READ(DSPFW2) & DSPFW_CURSORA_MASK) |
+		   (I915_READ(DSPFW2) & ~DSPFW_CURSORA_MASK) |
 		   (cursora_wm << DSPFW_CURSORA_SHIFT));
 	I915_WRITE(DSPFW3,
-		   (I915_READ(DSPFW3) | (cursor_sr << DSPFW_CURSOR_SR_SHIFT)));
+		   (I915_READ(DSPFW3) & ~DSPFW_CURSOR_SR_MASK) |
+		   (cursor_sr << DSPFW_CURSOR_SR_SHIFT));
 }
 
 static void g4x_update_wm(struct drm_device *dev)
@@ -1374,11 +1375,11 @@ static void g4x_update_wm(struct drm_device *dev)
 		   (planeb_wm << DSPFW_PLANEB_SHIFT) |
 		   planea_wm);
 	I915_WRITE(DSPFW2,
-		   (I915_READ(DSPFW2) & DSPFW_CURSORA_MASK) |
+		   (I915_READ(DSPFW2) & ~DSPFW_CURSORA_MASK) |
 		   (cursora_wm << DSPFW_CURSORA_SHIFT));
 	/* HPLL off in SR has some issues on G4x... disable it */
 	I915_WRITE(DSPFW3,
-		   (I915_READ(DSPFW3) & ~DSPFW_HPLL_SR_EN) |
+		   (I915_READ(DSPFW3) & ~(DSPFW_HPLL_SR_EN | DSPFW_CURSOR_SR_MASK)) |
 		   (cursor_sr << DSPFW_CURSOR_SR_SHIFT));
 }
 
@@ -2647,6 +2648,7 @@ static void ironlake_enable_rc6(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
+	bool was_interruptible;
 	int ret;
 
 	/* rc6 disabled by default due to repeated reports of hanging during
@@ -2661,6 +2663,9 @@ static void ironlake_enable_rc6(struct drm_device *dev)
 	if (ret)
 		return;
 
+	was_interruptible = dev_priv->mm.interruptible;
+	dev_priv->mm.interruptible = false;
+
 	/*
 	 * GPU can automatically power down the render unit if given a page
 	 * to save state.
@@ -2668,6 +2673,7 @@ static void ironlake_enable_rc6(struct drm_device *dev)
 	ret = intel_ring_begin(ring, 6);
 	if (ret) {
 		ironlake_teardown_rc6(dev);
+		dev_priv->mm.interruptible = was_interruptible;
 		return;
 	}
 
@@ -2688,7 +2694,8 @@ static void ironlake_enable_rc6(struct drm_device *dev)
 	 * does an implicit flush, combined with MI_FLUSH above, it should be
 	 * safe to assume that renderctx is valid
 	 */
-	ret = intel_wait_ring_idle(ring);
+	ret = intel_ring_idle(ring);
+	dev_priv->mm.interruptible = was_interruptible;
 	if (ret) {
 		DRM_ERROR("failed to enable ironlake power power savings\n");
 		ironlake_teardown_rc6(dev);
@@ -3440,6 +3447,11 @@ static void cpt_init_clock_gating(struct drm_device *dev)
 	I915_WRITE(SOUTH_DSPCLK_GATE_D, PCH_DPLSUNIT_CLOCK_GATE_DISABLE);
 	I915_WRITE(SOUTH_CHICKEN2, I915_READ(SOUTH_CHICKEN2) |
 		   DPLS_EDP_PPS_FIX_DIS);
+	/* The below fixes the weird display corruption, a few pixels shifted
+	 * downward, on (only) LVDS of some HP laptops with IVY.
+	 */
+	for_each_pipe(pipe)
+		I915_WRITE(TRANS_CHICKEN2(pipe), TRANS_CHICKEN2_TIMING_OVERRIDE);
 	/* WADP0ClockGatingDisable */
 	for_each_pipe(pipe) {
 		I915_WRITE(TRANS_CHICKEN1(pipe),
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 987eb5fdaf3..2346b920bd8 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -45,7 +45,7 @@ struct pipe_control {
 
 static inline int ring_space(struct intel_ring_buffer *ring)
 {
-	int space = (ring->head & HEAD_ADDR) - (ring->tail + 8);
+	int space = (ring->head & HEAD_ADDR) - (ring->tail + I915_RING_FREE_SPACE);
 	if (space < 0)
 		space += ring->size;
 	return space;
@@ -555,12 +555,11 @@ static void render_ring_cleanup(struct intel_ring_buffer *ring)
 
 static void
 update_mboxes(struct intel_ring_buffer *ring,
-	    u32 seqno,
-	    u32 mmio_offset)
+	      u32 mmio_offset)
 {
 	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
 	intel_ring_emit(ring, mmio_offset);
-	intel_ring_emit(ring, seqno);
+	intel_ring_emit(ring, ring->outstanding_lazy_request);
 }
 
 /**
@@ -573,8 +572,7 @@ update_mboxes(struct intel_ring_buffer *ring,
  * This acts like a signal in the canonical semaphore.
  */
 static int
-gen6_add_request(struct intel_ring_buffer *ring,
-		 u32 *seqno)
+gen6_add_request(struct intel_ring_buffer *ring)
 {
 	u32 mbox1_reg;
 	u32 mbox2_reg;
@@ -587,13 +585,11 @@ gen6_add_request(struct intel_ring_buffer *ring,
 	mbox1_reg = ring->signal_mbox[0];
 	mbox2_reg = ring->signal_mbox[1];
 
-	*seqno = i915_gem_next_request_seqno(ring);
-
-	update_mboxes(ring, *seqno, mbox1_reg);
-	update_mboxes(ring, *seqno, mbox2_reg);
+	update_mboxes(ring, mbox1_reg);
+	update_mboxes(ring, mbox2_reg);
 	intel_ring_emit(ring, MI_STORE_DWORD_INDEX);
 	intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
-	intel_ring_emit(ring, *seqno);
+	intel_ring_emit(ring, ring->outstanding_lazy_request);
 	intel_ring_emit(ring, MI_USER_INTERRUPT);
 	intel_ring_advance(ring);
 
@@ -650,10 +646,8 @@ do {									\
 } while (0)
 
 static int
-pc_render_add_request(struct intel_ring_buffer *ring,
-		      u32 *result)
+pc_render_add_request(struct intel_ring_buffer *ring)
 {
-	u32 seqno = i915_gem_next_request_seqno(ring);
 	struct pipe_control *pc = ring->private;
 	u32 scratch_addr = pc->gtt_offset + 128;
 	int ret;
@@ -674,7 +668,7 @@ pc_render_add_request(struct intel_ring_buffer *ring,
 			PIPE_CONTROL_WRITE_FLUSH |
 			PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE);
 	intel_ring_emit(ring, pc->gtt_offset | PIPE_CONTROL_GLOBAL_GTT);
-	intel_ring_emit(ring, seqno);
+	intel_ring_emit(ring, ring->outstanding_lazy_request);
 	intel_ring_emit(ring, 0);
 	PIPE_CONTROL_FLUSH(ring, scratch_addr);
 	scratch_addr += 128; /* write to separate cachelines */
@@ -693,11 +687,10 @@ pc_render_add_request(struct intel_ring_buffer *ring,
 			PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE |
 			PIPE_CONTROL_NOTIFY);
 	intel_ring_emit(ring, pc->gtt_offset | PIPE_CONTROL_GLOBAL_GTT);
-	intel_ring_emit(ring, seqno);
+	intel_ring_emit(ring, ring->outstanding_lazy_request);
 	intel_ring_emit(ring, 0);
 	intel_ring_advance(ring);
 
-	*result = seqno;
 	return 0;
 }
 
@@ -885,25 +878,20 @@ bsd_ring_flush(struct intel_ring_buffer *ring,
 }
 
 static int
-i9xx_add_request(struct intel_ring_buffer *ring,
-		 u32 *result)
+i9xx_add_request(struct intel_ring_buffer *ring)
 {
-	u32 seqno;
 	int ret;
 
 	ret = intel_ring_begin(ring, 4);
 	if (ret)
 		return ret;
 
-	seqno = i915_gem_next_request_seqno(ring);
-
 	intel_ring_emit(ring, MI_STORE_DWORD_INDEX);
 	intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
-	intel_ring_emit(ring, seqno);
+	intel_ring_emit(ring, ring->outstanding_lazy_request);
 	intel_ring_emit(ring, MI_USER_INTERRUPT);
 	intel_ring_advance(ring);
 
-	*result = seqno;
 	return 0;
 }
 
@@ -1110,6 +1098,7 @@ static int intel_init_ring_buffer(struct drm_device *dev,
 	INIT_LIST_HEAD(&ring->active_list);
 	INIT_LIST_HEAD(&ring->request_list);
 	ring->size = 32 * PAGE_SIZE;
+	memset(ring->sync_seqno, 0, sizeof(ring->sync_seqno));
 
 	init_waitqueue_head(&ring->irq_queue);
 
@@ -1186,7 +1175,7 @@ void intel_cleanup_ring_buffer(struct intel_ring_buffer *ring)
 
 	/* Disable the ring buffer. The ring must be idle at this point */
 	dev_priv = ring->dev->dev_private;
-	ret = intel_wait_ring_idle(ring);
+	ret = intel_ring_idle(ring);
 	if (ret)
 		DRM_ERROR("failed to quiesce %s whilst cleaning up: %d\n",
 			  ring->name, ret);
@@ -1205,28 +1194,6 @@ void intel_cleanup_ring_buffer(struct intel_ring_buffer *ring)
 	cleanup_status_page(ring);
 }
 
-static int intel_wrap_ring_buffer(struct intel_ring_buffer *ring)
-{
-	uint32_t __iomem *virt;
-	int rem = ring->size - ring->tail;
-
-	if (ring->space < rem) {
-		int ret = intel_wait_ring_buffer(ring, rem);
-		if (ret)
-			return ret;
-	}
-
-	virt = ring->virtual_start + ring->tail;
-	rem /= 4;
-	while (rem--)
-		iowrite32(MI_NOOP, virt++);
-
-	ring->tail = 0;
-	ring->space = ring_space(ring);
-
-	return 0;
-}
-
 static int intel_ring_wait_seqno(struct intel_ring_buffer *ring, u32 seqno)
 {
 	int ret;
@@ -1260,7 +1227,7 @@ static int intel_ring_wait_request(struct intel_ring_buffer *ring, int n)
 		if (request->tail == -1)
 			continue;
 
-		space = request->tail - (ring->tail + 8);
+		space = request->tail - (ring->tail + I915_RING_FREE_SPACE);
 		if (space < 0)
 			space += ring->size;
 		if (space >= n) {
@@ -1295,7 +1262,7 @@ static int intel_ring_wait_request(struct intel_ring_buffer *ring, int n)
 	return 0;
 }
 
-int intel_wait_ring_buffer(struct intel_ring_buffer *ring, int n)
+static int ring_wait_for_space(struct intel_ring_buffer *ring, int n)
 {
 	struct drm_device *dev = ring->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1338,6 +1305,60 @@ int intel_wait_ring_buffer(struct intel_ring_buffer *ring, int n)
 	return -EBUSY;
 }
 
+static int intel_wrap_ring_buffer(struct intel_ring_buffer *ring)
+{
+	uint32_t __iomem *virt;
+	int rem = ring->size - ring->tail;
+
+	if (ring->space < rem) {
+		int ret = ring_wait_for_space(ring, rem);
+		if (ret)
+			return ret;
+	}
+
+	virt = ring->virtual_start + ring->tail;
+	rem /= 4;
+	while (rem--)
+		iowrite32(MI_NOOP, virt++);
+
+	ring->tail = 0;
+	ring->space = ring_space(ring);
+
+	return 0;
+}
+
+int intel_ring_idle(struct intel_ring_buffer *ring)
+{
+	u32 seqno;
+	int ret;
+
+	/* We need to add any requests required to flush the objects and ring */
+	if (ring->outstanding_lazy_request) {
+		ret = i915_add_request(ring, NULL, NULL);
+		if (ret)
+			return ret;
+	}
+
+	/* Wait upon the last request to be completed */
+	if (list_empty(&ring->request_list))
+		return 0;
+
+	seqno = list_entry(ring->request_list.prev,
+			   struct drm_i915_gem_request,
+			   list)->seqno;
+
+	return i915_wait_seqno(ring, seqno);
+}
+
+static int
+intel_ring_alloc_seqno(struct intel_ring_buffer *ring)
+{
+	if (ring->outstanding_lazy_request)
+		return 0;
+
+	return i915_gem_get_seqno(ring->dev, &ring->outstanding_lazy_request);
+}
+
 int intel_ring_begin(struct intel_ring_buffer *ring,
 		     int num_dwords)
 {
@@ -1349,6 +1370,11 @@ int intel_ring_begin(struct intel_ring_buffer *ring,
 	if (ret)
 		return ret;
 
+	/* Preallocate the olr before touching the ring */
+	ret = intel_ring_alloc_seqno(ring);
+	if (ret)
+		return ret;
+
 	if (unlikely(ring->tail + n > ring->effective_size)) {
 		ret = intel_wrap_ring_buffer(ring);
 		if (unlikely(ret))
@@ -1356,7 +1382,7 @@ int intel_ring_begin(struct intel_ring_buffer *ring,
 	}
 
 	if (unlikely(ring->space < n)) {
-		ret = intel_wait_ring_buffer(ring, n);
+		ret = ring_wait_for_space(ring, n);
 		if (unlikely(ret))
 			return ret;
 	}
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 5af65b89765..526182ed0c6 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -1,6 +1,17 @@
 #ifndef _INTEL_RINGBUFFER_H_
 #define _INTEL_RINGBUFFER_H_
 
+/*
+ * Gen2 BSpec "1. Programming Environment" / 1.4.4.6 "Ring Buffer Use"
+ * Gen3 BSpec "vol1c Memory Interface Functions" / 2.3.4.5 "Ring Buffer Use"
+ * Gen4+ BSpec "vol1c Memory Interface and Command Stream" / 5.3.4.5 "Ring Buffer Use"
+ *
+ * "If the Ring Buffer Head Pointer and the Tail Pointer are on the same
+ * cacheline, the Head Pointer must not be greater than the Tail
+ * Pointer."
+ */
+#define I915_RING_FREE_SPACE 64
+
 struct  intel_hw_status_page {
 	u32		*page_addr;
 	unsigned int	gfx_addr;
@@ -70,8 +81,7 @@ struct  intel_ring_buffer {
 	int __must_check (*flush)(struct intel_ring_buffer *ring,
 				  u32	invalidate_domains,
 				  u32	flush_domains);
-	int		(*add_request)(struct intel_ring_buffer *ring,
-				       u32 *seqno);
+	int		(*add_request)(struct intel_ring_buffer *ring);
 	/* Some chipsets are not quite as coherent as advertised and need
 	 * an expensive kick to force a true read of the up-to-date seqno.
 	 * However, the up-to-date seqno is not always required and the last
@@ -188,24 +198,16 @@ intel_read_status_page(struct intel_ring_buffer *ring,
 
 void intel_cleanup_ring_buffer(struct intel_ring_buffer *ring);
 
-int __must_check intel_wait_ring_buffer(struct intel_ring_buffer *ring, int n);
-static inline int intel_wait_ring_idle(struct intel_ring_buffer *ring)
-{
-	return intel_wait_ring_buffer(ring, ring->size - 8);
-}
-
 int __must_check intel_ring_begin(struct intel_ring_buffer *ring, int n);
-
 static inline void intel_ring_emit(struct intel_ring_buffer *ring,
 				   u32 data)
 {
 	iowrite32(data, ring->virtual_start + ring->tail);
 	ring->tail += 4;
 }
-
 void intel_ring_advance(struct intel_ring_buffer *ring);
+int __must_check intel_ring_idle(struct intel_ring_buffer *ring);
 
-u32 intel_ring_get_seqno(struct intel_ring_buffer *ring);
 int intel_ring_flush_all_caches(struct intel_ring_buffer *ring);
 int intel_ring_invalidate_all_caches(struct intel_ring_buffer *ring);
 
@@ -221,6 +223,12 @@ static inline u32 intel_ring_get_tail(struct intel_ring_buffer *ring)
 	return ring->tail;
 }
 
+static inline u32 intel_ring_get_seqno(struct intel_ring_buffer *ring)
+{
+	BUG_ON(ring->outstanding_lazy_request == 0);
+	return ring->outstanding_lazy_request;
+}
+
 static inline void i915_trace_irq_get(struct intel_ring_buffer *ring, u32 seqno)
 {
 	if (ring->trace_irq_seqno == 0 && ring->irq_get(ring))
diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c
index 4b07401540e..c275bf0fa36 100644
--- a/drivers/gpu/drm/i915/intel_sdvo.c
+++ b/drivers/gpu/drm/i915/intel_sdvo.c
@@ -509,7 +509,7 @@ out:
 static bool intel_sdvo_read_response(struct intel_sdvo *intel_sdvo,
 				     void *response, int response_len)
 {
-	u8 retry = 5;
+	u8 retry = 15; /* 5 quick checks, followed by 10 long checks */
 	u8 status;
 	int i;
 
@@ -522,14 +522,27 @@ static bool intel_sdvo_read_response(struct intel_sdvo *intel_sdvo,
 	 * command to be complete.
 	 *
 	 * Check 5 times in case the hardware failed to read the docs.
+	 *
+	 * Also beware that the first response by many devices is to
+	 * reply PENDING and stall for time. TVs are notorious for
+	 * requiring longer than specified to complete their replies.
+	 * Originally (in the DDX long ago), the delay was only ever 15ms
+	 * with an additional delay of 30ms applied for TVs added later after
+	 * many experiments. To accommodate both sets of delays, we do a
+	 * sequence of slow checks if the device is falling behind and fails
+	 * to reply within 5*15µs.
 	 */
 	if (!intel_sdvo_read_byte(intel_sdvo,
 				  SDVO_I2C_CMD_STATUS,
 				  &status))
 		goto log_fail;
 
-	while (status == SDVO_CMD_STATUS_PENDING && retry--) {
-		udelay(15);
+	while (status == SDVO_CMD_STATUS_PENDING && --retry) {
+		if (retry < 10)
+			msleep(15);
+		else
+			udelay(15);
+
 		if (!intel_sdvo_read_byte(intel_sdvo,
 					  SDVO_I2C_CMD_STATUS,
 					  &status))
@@ -1535,15 +1548,9 @@ intel_sdvo_detect(struct drm_connector *connector, bool force)
 	struct intel_sdvo_connector *intel_sdvo_connector = to_intel_sdvo_connector(connector);
 	enum drm_connector_status ret;
 
-	if (!intel_sdvo_write_cmd(intel_sdvo,
-				  SDVO_CMD_GET_ATTACHED_DISPLAYS, NULL, 0))
-		return connector_status_unknown;
-
-	/* add 30ms delay when the output type might be TV */
-	if (intel_sdvo->caps.output_flags & SDVO_TV_MASK)
-		msleep(30);
-
-	if (!intel_sdvo_read_response(intel_sdvo, &response, 2))
+	if (!intel_sdvo_get_value(intel_sdvo,
+				  SDVO_CMD_GET_ATTACHED_DISPLAYS,
+				  &response, 2))
 		return connector_status_unknown;
 
 	DRM_DEBUG_KMS("SDVO response %d %d [%x]\n",