Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

Pull drm fixes from Dave Airlie:
 "This feels larger than I'd like but its for three reasons.

   a) amdkfd finalising the API more, this is a new feature introduced
      last merge window, and I'd prefer to make the tweaks to the API
      before it first gets into a stable release.

   b) radeon regression required splitting an internal API to fix
      properly, so it just changed a few more lines

   c) vmwgfx fix changes a lock from a mutex->spin lock, this is fallout
      from the new sleep checking.

  Otherwise there is just some tda998x fixes"

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
  drm/radeon: Remove rdev->gart.pages_addr array
  drm/radeon: Restore GART table contents after pinning it in VRAM v3
  drm/radeon: Split off gart_get_page_entry ASIC hook from set_page_entry
  drm/amdkfd: Fix bug in call to init_pipelines()
  drm/amdkfd: Fix bug in pipelines initialization
  drm/radeon: Don't increment pipe_id in kgd_init_pipeline
  drm/i2c: tda998x: set the CEC I2C address based on the slave I2C address
  drm/vmwgfx: Replace the hw mutex with a hw spinlock
  drm/amdkfd: Allow user to limit only queues per device
  drm/amdkfd: PQM handle queue creation fault
  drm: tda998x: Fix EDID read timeout on HDMI connect
  drm: tda998x: Protect the page register

29 files changed, 315 insertions, 188 deletions

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index 633532a2e7ec..25bc47f3c1cf 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -26,6 +26,7 @@
 #include <linux/slab.h>
 #include "kfd_priv.h"
 #include "kfd_device_queue_manager.h"
+#include "kfd_pm4_headers.h"
 
 #define MQD_SIZE_ALIGNED 768
 
@@ -169,9 +170,8 @@ bool kgd2kfd_device_init(struct kfd_dev *kfd,
 	kfd->shared_resources = *gpu_resources;
 
 	/* calculate max size of mqds needed for queues */
-	size = max_num_of_processes *
-		max_num_of_queues_per_process *
-		kfd->device_info->mqd_size_aligned;
+	size = max_num_of_queues_per_device *
+			kfd->device_info->mqd_size_aligned;
 
 	/* add another 512KB for all other allocations on gart */
 	size += 512 * 1024;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
index 30c8fda9622e..0d8694f015c1 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -183,6 +183,13 @@ static int create_queue_nocpsch(struct device_queue_manager *dqm,
 
 	mutex_lock(&dqm->lock);
 
+	if (dqm->total_queue_count >= max_num_of_queues_per_device) {
+		pr_warn("amdkfd: Can't create new usermode queue because %d queues were already created\n",
+				dqm->total_queue_count);
+		mutex_unlock(&dqm->lock);
+		return -EPERM;
+	}
+
 	if (list_empty(&qpd->queues_list)) {
 		retval = allocate_vmid(dqm, qpd, q);
 		if (retval != 0) {
@@ -207,6 +214,14 @@ static int create_queue_nocpsch(struct device_queue_manager *dqm,
 	list_add(&q->list, &qpd->queues_list);
 	dqm->queue_count++;
 
+	/*
+	 * Unconditionally increment this counter, regardless of the queue's
+	 * type or whether the queue is active.
+	 */
+	dqm->total_queue_count++;
+	pr_debug("Total of %d queues are accountable so far\n",
+			dqm->total_queue_count);
+
 	mutex_unlock(&dqm->lock);
 	return 0;
 }
@@ -326,6 +341,15 @@ static int destroy_queue_nocpsch(struct device_queue_manager *dqm,
 	if (list_empty(&qpd->queues_list))
 		deallocate_vmid(dqm, qpd, q);
 	dqm->queue_count--;
+
+	/*
+	 * Unconditionally decrement this counter, regardless of the queue's
+	 * type
+	 */
+	dqm->total_queue_count--;
+	pr_debug("Total of %d queues are accountable so far\n",
+			dqm->total_queue_count);
+
 out:
 	mutex_unlock(&dqm->lock);
 	return retval;
@@ -541,10 +565,14 @@ static int init_pipelines(struct device_queue_manager *dqm,
 
 	for (i = 0; i < pipes_num; i++) {
 		inx = i + first_pipe;
+		/*
+		 * HPD buffer on GTT is allocated by amdkfd, no need to waste
+		 * space in GTT for pipelines we don't initialize
+		 */
 		pipe_hpd_addr = dqm->pipelines_addr + i * CIK_HPD_EOP_BYTES;
 		pr_debug("kfd: pipeline address %llX\n", pipe_hpd_addr);
 		/* = log2(bytes/4)-1 */
-		kfd2kgd->init_pipeline(dqm->dev->kgd, i,
+		kfd2kgd->init_pipeline(dqm->dev->kgd, inx,
 				CIK_HPD_EOP_BYTES_LOG2 - 3, pipe_hpd_addr);
 	}
 
@@ -560,7 +588,7 @@ static int init_scheduler(struct device_queue_manager *dqm)
 
 	pr_debug("kfd: In %s\n", __func__);
 
-	retval = init_pipelines(dqm, get_pipes_num(dqm), KFD_DQM_FIRST_PIPE);
+	retval = init_pipelines(dqm, get_pipes_num(dqm), get_first_pipe(dqm));
 	if (retval != 0)
 		return retval;
 
@@ -752,6 +780,21 @@ static int create_kernel_queue_cpsch(struct device_queue_manager *dqm,
 	pr_debug("kfd: In func %s\n", __func__);
 
 	mutex_lock(&dqm->lock);
+	if (dqm->total_queue_count >= max_num_of_queues_per_device) {
+		pr_warn("amdkfd: Can't create new kernel queue because %d queues were already created\n",
+				dqm->total_queue_count);
+		mutex_unlock(&dqm->lock);
+		return -EPERM;
+	}
+
+	/*
+	 * Unconditionally increment this counter, regardless of the queue's
+	 * type or whether the queue is active.
+	 */
+	dqm->total_queue_count++;
+	pr_debug("Total of %d queues are accountable so far\n",
+			dqm->total_queue_count);
+
 	list_add(&kq->list, &qpd->priv_queue_list);
 	dqm->queue_count++;
 	qpd->is_debug = true;
@@ -775,6 +818,13 @@ static void destroy_kernel_queue_cpsch(struct device_queue_manager *dqm,
 	dqm->queue_count--;
 	qpd->is_debug = false;
 	execute_queues_cpsch(dqm, false);
+	/*
+	 * Unconditionally decrement this counter, regardless of the queue's
+	 * type.
+	 */
+	dqm->total_queue_count++;
+	pr_debug("Total of %d queues are accountable so far\n",
+			dqm->total_queue_count);
 	mutex_unlock(&dqm->lock);
 }
 
@@ -793,6 +843,13 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q,
 
 	mutex_lock(&dqm->lock);
 
+	if (dqm->total_queue_count >= max_num_of_queues_per_device) {
+		pr_warn("amdkfd: Can't create new usermode queue because %d queues were already created\n",
+				dqm->total_queue_count);
+		retval = -EPERM;
+		goto out;
+	}
+
 	mqd = dqm->get_mqd_manager(dqm, KFD_MQD_TYPE_CIK_CP);
 	if (mqd == NULL) {
 		mutex_unlock(&dqm->lock);
@@ -810,6 +867,15 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q,
 		retval = execute_queues_cpsch(dqm, false);
 	}
 
+	/*
+	 * Unconditionally increment this counter, regardless of the queue's
+	 * type or whether the queue is active.
+	 */
+	dqm->total_queue_count++;
+
+	pr_debug("Total of %d queues are accountable so far\n",
+			dqm->total_queue_count);
+
 out:
 	mutex_unlock(&dqm->lock);
 	return retval;
@@ -930,6 +996,14 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm,
 
 	mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj);
 
+	/*
+	 * Unconditionally decrement this counter, regardless of the queue's
+	 * type
+	 */
+	dqm->total_queue_count--;
+	pr_debug("Total of %d queues are accountable so far\n",
+			dqm->total_queue_count);
+
 	mutex_unlock(&dqm->lock);
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
index c3f189e8ae35..52035bf0c1cb 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
@@ -130,6 +130,7 @@ struct device_queue_manager {
 	struct list_head	queues;
 	unsigned int		processes_count;
 	unsigned int		queue_count;
+	unsigned int		total_queue_count;
 	unsigned int		next_pipe_to_allocate;
 	unsigned int		*allocated_queues;
 	unsigned int		vmid_bitmap;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_module.c b/drivers/gpu/drm/amd/amdkfd/kfd_module.c
index 95d5af138e6e..a8be6df85347 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_module.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_module.c
@@ -50,15 +50,10 @@ module_param(sched_policy, int, 0444);
 MODULE_PARM_DESC(sched_policy,
 	"Kernel cmdline parameter that defines the amdkfd scheduling policy");
 
-int max_num_of_processes = KFD_MAX_NUM_OF_PROCESSES_DEFAULT;
-module_param(max_num_of_processes, int, 0444);
-MODULE_PARM_DESC(max_num_of_processes,
-	"Kernel cmdline parameter that defines the amdkfd maximum number of supported processes");
-
-int max_num_of_queues_per_process = KFD_MAX_NUM_OF_QUEUES_PER_PROCESS_DEFAULT;
-module_param(max_num_of_queues_per_process, int, 0444);
-MODULE_PARM_DESC(max_num_of_queues_per_process,
-	"Kernel cmdline parameter that defines the amdkfd maximum number of supported queues per process");
+int max_num_of_queues_per_device = KFD_MAX_NUM_OF_QUEUES_PER_DEVICE_DEFAULT;
+module_param(max_num_of_queues_per_device, int, 0444);
+MODULE_PARM_DESC(max_num_of_queues_per_device,
+	"Maximum number of supported queues per device (1 = Minimum, 4096 = default)");
 
 bool kgd2kfd_init(unsigned interface_version,
 		  const struct kfd2kgd_calls *f2g,
@@ -100,16 +95,10 @@ static int __init kfd_module_init(void)
 	}
 
 	/* Verify module parameters */
-	if ((max_num_of_processes < 0) ||
-		(max_num_of_processes > KFD_MAX_NUM_OF_PROCESSES)) {
-		pr_err("kfd: max_num_of_processes must be between 0 to KFD_MAX_NUM_OF_PROCESSES\n");
-		return -1;
-	}
-
-	if ((max_num_of_queues_per_process < 0) ||
-		(max_num_of_queues_per_process >
-			KFD_MAX_NUM_OF_QUEUES_PER_PROCESS)) {
-		pr_err("kfd: max_num_of_queues_per_process must be between 0 to KFD_MAX_NUM_OF_QUEUES_PER_PROCESS\n");
+	if ((max_num_of_queues_per_device < 0) ||
+		(max_num_of_queues_per_device >
+			KFD_MAX_NUM_OF_QUEUES_PER_DEVICE)) {
+		pr_err("kfd: max_num_of_queues_per_device must be between 0 to KFD_MAX_NUM_OF_QUEUES_PER_DEVICE\n");
 		return -1;
 	}
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_pasid.c b/drivers/gpu/drm/amd/amdkfd/kfd_pasid.c
index 4c25ef504f79..6cfe7f1f18cf 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_pasid.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_pasid.c
@@ -30,7 +30,7 @@ static DEFINE_MUTEX(pasid_mutex);
 
 int kfd_pasid_init(void)
 {
-	pasid_limit = max_num_of_processes;
+	pasid_limit = KFD_MAX_NUM_OF_PROCESSES;
 
 	pasid_bitmap = kcalloc(BITS_TO_LONGS(pasid_limit), sizeof(long), GFP_KERNEL);
 	if (!pasid_bitmap)
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index b3dc13c83169..96dc10e8904a 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -52,20 +52,19 @@
 #define kfd_alloc_struct(ptr_to_struct)	\
 	((typeof(ptr_to_struct)) kzalloc(sizeof(*ptr_to_struct), GFP_KERNEL))
 
-/* Kernel module parameter to specify maximum number of supported processes */
-extern int max_num_of_processes;
-
-#define KFD_MAX_NUM_OF_PROCESSES_DEFAULT 32
 #define KFD_MAX_NUM_OF_PROCESSES 512
+#define KFD_MAX_NUM_OF_QUEUES_PER_PROCESS 1024
 
 /*
- * Kernel module parameter to specify maximum number of supported queues
- * per process
+ * Kernel module parameter to specify maximum number of supported queues per
+ * device
  */
-extern int max_num_of_queues_per_process;
+extern int max_num_of_queues_per_device;
 
-#define KFD_MAX_NUM_OF_QUEUES_PER_PROCESS_DEFAULT 128
-#define KFD_MAX_NUM_OF_QUEUES_PER_PROCESS 1024
+#define KFD_MAX_NUM_OF_QUEUES_PER_DEVICE_DEFAULT 4096
+#define KFD_MAX_NUM_OF_QUEUES_PER_DEVICE		\
+	(KFD_MAX_NUM_OF_PROCESSES *			\
+			KFD_MAX_NUM_OF_QUEUES_PER_PROCESS)
 
 #define KFD_KERNEL_QUEUE_SIZE 2048
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
index 47526780d736..f37cf5efe642 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
@@ -54,11 +54,11 @@ static int find_available_queue_slot(struct process_queue_manager *pqm,
 	pr_debug("kfd: in %s\n", __func__);
 
 	found = find_first_zero_bit(pqm->queue_slot_bitmap,
-			max_num_of_queues_per_process);
+			KFD_MAX_NUM_OF_QUEUES_PER_PROCESS);
 
 	pr_debug("kfd: the new slot id %lu\n", found);
 
-	if (found >= max_num_of_queues_per_process) {
+	if (found >= KFD_MAX_NUM_OF_QUEUES_PER_PROCESS) {
 		pr_info("amdkfd: Can not open more queues for process with pasid %d\n",
 				pqm->process->pasid);
 		return -ENOMEM;
@@ -76,7 +76,7 @@ int pqm_init(struct process_queue_manager *pqm, struct kfd_process *p)
 
 	INIT_LIST_HEAD(&pqm->queues);
 	pqm->queue_slot_bitmap =
-			kzalloc(DIV_ROUND_UP(max_num_of_queues_per_process,
+			kzalloc(DIV_ROUND_UP(KFD_MAX_NUM_OF_QUEUES_PER_PROCESS,
 					BITS_PER_BYTE), GFP_KERNEL);
 	if (pqm->queue_slot_bitmap == NULL)
 		return -ENOMEM;
@@ -203,6 +203,7 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 		pqn->kq = NULL;
 		retval = dev->dqm->create_queue(dev->dqm, q, &pdd->qpd,
 						&q->properties.vmid);
+		pr_debug("DQM returned %d for create_queue\n", retval);
 		print_queue(q);
 		break;
 	case KFD_QUEUE_TYPE_DIQ:
@@ -222,7 +223,7 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 	}
 
 	if (retval != 0) {
-		pr_err("kfd: error dqm create queue\n");
+		pr_debug("Error dqm create queue\n");
 		goto err_create_queue;
 	}
 
@@ -241,7 +242,10 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 err_create_queue:
 	kfree(pqn);
 err_allocate_pqn:
+	/* check if queues list is empty unregister process from device */
 	clear_bit(*qid, pqm->queue_slot_bitmap);
+	if (list_empty(&pqm->queues))
+		dev->dqm->unregister_process(dev->dqm, &pdd->qpd);
 	return retval;
 }
 
diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
index d4762799351d..a9041d1a8ff0 100644
--- a/drivers/gpu/drm/i2c/tda998x_drv.c
+++ b/drivers/gpu/drm/i2c/tda998x_drv.c
@@ -32,6 +32,8 @@
 struct tda998x_priv {
 	struct i2c_client *cec;
 	struct i2c_client *hdmi;
+	struct mutex mutex;
+	struct delayed_work dwork;
 	uint16_t rev;
 	uint8_t current_page;
 	int dpms;
@@ -402,9 +404,10 @@ reg_read_range(struct tda998x_priv *priv, uint16_t reg, char *buf, int cnt)
 	uint8_t addr = REG2ADDR(reg);
 	int ret;
 
+	mutex_lock(&priv->mutex);
 	ret = set_page(priv, reg);
 	if (ret < 0)
-		return ret;
+		goto out;
 
 	ret = i2c_master_send(client, &addr, sizeof(addr));
 	if (ret < 0)
@@ -414,10 +417,12 @@ reg_read_range(struct tda998x_priv *priv, uint16_t reg, char *buf, int cnt)
 	if (ret < 0)
 		goto fail;
 
-	return ret;
+	goto out;
 
 fail:
 	dev_err(&client->dev, "Error %d reading from 0x%x\n", ret, reg);
+out:
+	mutex_unlock(&priv->mutex);
 	return ret;
 }
 
@@ -431,13 +436,16 @@ reg_write_range(struct tda998x_priv *priv, uint16_t reg, uint8_t *p, int cnt)
 	buf[0] = REG2ADDR(reg);
 	memcpy(&buf[1], p, cnt);
 
+	mutex_lock(&priv->mutex);
 	ret = set_page(priv, reg);
 	if (ret < 0)
-		return;
+		goto out;
 
 	ret = i2c_master_send(client, buf, cnt + 1);
 	if (ret < 0)
 		dev_err(&client->dev, "Error %d writing to 0x%x\n", ret, reg);
+out:
+	mutex_unlock(&priv->mutex);
 }
 
 static int
@@ -459,13 +467,16 @@ reg_write(struct tda998x_priv *priv, uint16_t reg, uint8_t val)
 	uint8_t buf[] = {REG2ADDR(reg), val};
 	int ret;
 
+	mutex_lock(&priv->mutex);
 	ret = set_page(priv, reg);
 	if (ret < 0)
-		return;
+		goto out;
 
 	ret = i2c_master_send(client, buf, sizeof(buf));
 	if (ret < 0)
 		dev_err(&client->dev, "Error %d writing to 0x%x\n", ret, reg);
+out:
+	mutex_unlock(&priv->mutex);
 }
 
 static void
@@ -475,13 +486,16 @@ reg_write16(struct tda998x_priv *priv, uint16_t reg, uint16_t val)
 	uint8_t buf[] = {REG2ADDR(reg), val >> 8, val};
 	int ret;
 
+	mutex_lock(&priv->mutex);
 	ret = set_page(priv, reg);
 	if (ret < 0)
-		return;
+		goto out;
 
 	ret = i2c_master_send(client, buf, sizeof(buf));
 	if (ret < 0)
 		dev_err(&client->dev, "Error %d writing to 0x%x\n", ret, reg);
+out:
+	mutex_unlock(&priv->mutex);
 }
 
 static void
@@ -536,6 +550,17 @@ tda998x_reset(struct tda998x_priv *priv)
 	reg_write(priv, REG_MUX_VP_VIP_OUT, 0x24);
 }
 
+/* handle HDMI connect/disconnect */
+static void tda998x_hpd(struct work_struct *work)
+{
+	struct delayed_work *dwork = to_delayed_work(work);
+	struct tda998x_priv *priv =
+			container_of(dwork, struct tda998x_priv, dwork);
+
+	if (priv->encoder && priv->encoder->dev)
+		drm_kms_helper_hotplug_event(priv->encoder->dev);
+}
+
 /*
  * only 2 interrupts may occur: screen plug/unplug and EDID read
  */
@@ -559,8 +584,7 @@ static irqreturn_t tda998x_irq_thread(int irq, void *data)
 		priv->wq_edid_wait = 0;
 		wake_up(&priv->wq_edid);
 	} else if (cec != 0) {			/* HPD change */
-		if (priv->encoder && priv->encoder->dev)
-			drm_helper_hpd_irq_event(priv->encoder->dev);
+		schedule_delayed_work(&priv->dwork, HZ/10);
 	}
 	return IRQ_HANDLED;
 }
@@ -1170,8 +1194,10 @@ static void tda998x_destroy(struct tda998x_priv *priv)
 	/* disable all IRQs and free the IRQ handler */
 	cec_write(priv, REG_CEC_RXSHPDINTENA, 0);
 	reg_clear(priv, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD);
-	if (priv->hdmi->irq)
+	if (priv->hdmi->irq) {
 		free_irq(priv->hdmi->irq, priv);
+		cancel_delayed_work_sync(&priv->dwork);
+	}
 
 	i2c_unregister_device(priv->cec);
 }
@@ -1255,6 +1281,7 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 	struct device_node *np = client->dev.of_node;
 	u32 video;
 	int rev_lo, rev_hi, ret;
+	unsigned short cec_addr;
 
 	priv->vip_cntrl_0 = VIP_CNTRL_0_SWAP_A(2) | VIP_CNTRL_0_SWAP_B(3);
 	priv->vip_cntrl_1 = VIP_CNTRL_1_SWAP_C(0) | VIP_CNTRL_1_SWAP_D(1);
@@ -1262,12 +1289,16 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 
 	priv->current_page = 0xff;
 	priv->hdmi = client;
-	priv->cec = i2c_new_dummy(client->adapter, 0x34);
+	/* CEC I2C address bound to TDA998x I2C addr by configuration pins */
+	cec_addr = 0x34 + (client->addr & 0x03);
+	priv->cec = i2c_new_dummy(client->adapter, cec_addr);
 	if (!priv->cec)
 		return -ENODEV;
 
 	priv->dpms = DRM_MODE_DPMS_OFF;
 
+	mutex_init(&priv->mutex);	/* protect the page access */
+
 	/* wake up the device: */
 	cec_write(priv, REG_CEC_ENAMODS,
 			CEC_ENAMODS_EN_RXSENS | CEC_ENAMODS_EN_HDMI);
@@ -1323,8 +1354,9 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 	if (client->irq) {
 		int irqf_trigger;
 
-		/* init read EDID waitqueue */
+		/* init read EDID waitqueue and HDP work */
 		init_waitqueue_head(&priv->wq_edid);
+		INIT_DELAYED_WORK(&priv->dwork, tda998x_hpd);
 
 		/* clear pending interrupts */
 		reg_read(priv, REG_INT_FLAGS_0);
diff --git a/drivers/gpu/drm/radeon/cik_sdma.c b/drivers/gpu/drm/radeon/cik_sdma.c
index a0133c74f4cf..42cd0cffe210 100644
--- a/drivers/gpu/drm/radeon/cik_sdma.c
+++ b/drivers/gpu/drm/radeon/cik_sdma.c
@@ -816,7 +816,6 @@ void cik_sdma_vm_write_pages(struct radeon_device *rdev,
 		for (; ndw > 0; ndw -= 2, --count, pe += 8) {
 			if (flags & R600_PTE_SYSTEM) {
 				value = radeon_vm_map_gart(rdev, addr);
-				value &= 0xFFFFFFFFFFFFF000ULL;
 			} else if (flags & R600_PTE_VALID) {
 				value = addr;
 			} else {
diff --git a/drivers/gpu/drm/radeon/ni_dma.c b/drivers/gpu/drm/radeon/ni_dma.c
index 4be2bb7cbef3..ce787a9f12c0 100644
--- a/drivers/gpu/drm/radeon/ni_dma.c
+++ b/drivers/gpu/drm/radeon/ni_dma.c
@@ -372,7 +372,6 @@ void cayman_dma_vm_write_pages(struct radeon_device *rdev,
 		for (; ndw > 0; ndw -= 2, --count, pe += 8) {
 			if (flags & R600_PTE_SYSTEM) {
 				value = radeon_vm_map_gart(rdev, addr);
-				value &= 0xFFFFFFFFFFFFF000ULL;
 			} else if (flags & R600_PTE_VALID) {
 				value = addr;
 			} else {
diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c
index 74f06d540591..279801ca5110 100644
--- a/drivers/gpu/drm/radeon/r100.c
+++ b/drivers/gpu/drm/radeon/r100.c
@@ -644,6 +644,7 @@ int r100_pci_gart_init(struct radeon_device *rdev)
 		return r;
 	rdev->gart.table_size = rdev->gart.num_gpu_pages * 4;
 	rdev->asic->gart.tlb_flush = &r100_pci_gart_tlb_flush;
+	rdev->asic->gart.get_page_entry = &r100_pci_gart_get_page_entry;
 	rdev->asic->gart.set_page = &r100_pci_gart_set_page;
 	return radeon_gart_table_ram_alloc(rdev);
 }
@@ -681,11 +682,16 @@ void r100_pci_gart_disable(struct radeon_device *rdev)
 	WREG32(RADEON_AIC_HI_ADDR, 0);
 }
 
+uint64_t r100_pci_gart_get_page_entry(uint64_t addr, uint32_t flags)
+{
+	return addr;
+}
+
 void r100_pci_gart_set_page(struct radeon_device *rdev, unsigned i,
-			    uint64_t addr, uint32_t flags)
+			    uint64_t entry)
 {
 	u32 *gtt = rdev->gart.ptr;
-	gtt[i] = cpu_to_le32(lower_32_bits(addr));
+	gtt[i] = cpu_to_le32(lower_32_bits(entry));
 }
 
 void r100_pci_gart_fini(struct radeon_device *rdev)
diff --git a/drivers/gpu/drm/radeon/r300.c b/drivers/gpu/drm/radeon/r300.c
index 064ad5569cca..08d68f3e13e9 100644
--- a/drivers/gpu/drm/radeon/r300.c
+++ b/drivers/gpu/drm/radeon/r300.c
@@ -73,11 +73,8 @@ void rv370_pcie_gart_tlb_flush(struct radeon_device *rdev)
 #define R300_PTE_WRITEABLE (1 << 2)
 #define R300_PTE_READABLE  (1 << 3)
 
-void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i,
-			      uint64_t addr, uint32_t flags)
+uint64_t rv370_pcie_gart_get_page_entry(uint64_t addr, uint32_t flags)
 {
-	void __iomem *ptr = rdev->gart.ptr;
-
 	addr = (lower_32_bits(addr) >> 8) |
 		((upper_32_bits(addr) & 0xff) << 24);
 	if (flags & RADEON_GART_PAGE_READ)
@@ -86,10 +83,18 @@ void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i,
 		addr |= R300_PTE_WRITEABLE;
 	if (!(flags & RADEON_GART_PAGE_SNOOP))
 		addr |= R300_PTE_UNSNOOPED;
+	return addr;
+}
+
+void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i,
+			      uint64_t entry)
+{
+	void __iomem *ptr = rdev->gart.ptr;
+
 	/* on x86 we want this to be CPU endian, on powerpc
 	 * on powerpc without HW swappers, it'll get swapped on way
 	 * into VRAM - so no need for cpu_to_le32 on VRAM tables */
-	writel(addr, ((void __iomem *)ptr) + (i * 4));
+	writel(entry, ((void __iomem *)ptr) + (i * 4));
 }
 
 int rv370_pcie_gart_init(struct radeon_device *rdev)
@@ -109,6 +114,7 @@ int rv370_pcie_gart_init(struct radeon_device *rdev)
 		DRM_ERROR("Failed to register debugfs file for PCIE gart !\n");
 	rdev->gart.table_size = rdev->gart.num_gpu_pages * 4;
 	rdev->asic->gart.tlb_flush = &rv370_pcie_gart_tlb_flush;
+	rdev->asic->gart.get_page_entry = &rv370_pcie_gart_get_page_entry;
 	rdev->asic->gart.set_page = &rv370_pcie_gart_set_page;
 	return radeon_gart_table_vram_alloc(rdev);
 }
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index 54529b837afa..3f2a8d3febca 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -242,6 +242,7 @@ bool radeon_get_bios(struct radeon_device *rdev);
  * Dummy page
  */
 struct radeon_dummy_page {
+	uint64_t	entry;
 	struct page	*page;
 	dma_addr_t	addr;
 };
@@ -645,7 +646,7 @@ struct radeon_gart {
 	unsigned			num_cpu_pages;
 	unsigned			table_size;
 	struct page			**pages;
-	dma_addr_t			*pages_addr;
+	uint64_t			*pages_entry;
 	bool				ready;
 };
 
@@ -1847,8 +1848,9 @@ struct radeon_asic {
 	/* gart */
 	struct {
 		void (*tlb_flush)(struct radeon_device *rdev);
+		uint64_t (*get_page_entry)(uint64_t addr, uint32_t flags);
 		void (*set_page)(struct radeon_device *rdev, unsigned i,
-				 uint64_t addr, uint32_t flags);
+				 uint64_t entry);
 	} gart;
 	struct {
 		int (*init)(struct radeon_device *rdev);
@@ -2852,7 +2854,8 @@ static inline void radeon_ring_write(struct radeon_ring *ring, uint32_t v)
 #define radeon_vga_set_state(rdev, state) (rdev)->asic->vga_set_state((rdev), (state))
 #define radeon_asic_reset(rdev) (rdev)->asic->asic_reset((rdev))
 #define radeon_gart_tlb_flush(rdev) (rdev)->asic->gart.tlb_flush((rdev))
-#define radeon_gart_set_page(rdev, i, p, f) (rdev)->asic->gart.set_page((rdev), (i), (p), (f))
+#define radeon_gart_get_page_entry(a, f) (rdev)->asic->gart.get_page_entry((a), (f))
+#define radeon_gart_set_page(rdev, i, e) (rdev)->asic->gart.set_page((rdev), (i), (e))
 #define radeon_asic_vm_init(rdev) (rdev)->asic->vm.init((rdev))
 #define radeon_asic_vm_fini(rdev) (rdev)->asic->vm.fini((rdev))
 #define radeon_asic_vm_copy_pages(rdev, ib, pe, src, count) ((rdev)->asic->vm.copy_pages((rdev), (ib), (pe), (src), (count)))
diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c
index 121aff6a3b41..ed0e10eee2dc 100644
--- a/drivers/gpu/drm/radeon/radeon_asic.c
+++ b/drivers/gpu/drm/radeon/radeon_asic.c
@@ -159,11 +159,13 @@ void radeon_agp_disable(struct radeon_device *rdev)
 		DRM_INFO("Forcing AGP to PCIE mode\n");
 		rdev->flags |= RADEON_IS_PCIE;
 		rdev->asic->gart.tlb_flush = &rv370_pcie_gart_tlb_flush;
+		rdev->asic->gart.get_page_entry = &rv370_pcie_gart_get_page_entry;
 		rdev->asic->gart.set_page = &rv370_pcie_gart_set_page;
 	} else {
 		DRM_INFO("Forcing AGP to PCI mode\n");
 		rdev->flags |= RADEON_IS_PCI;
 		rdev->asic->gart.tlb_flush = &r100_pci_gart_tlb_flush;
+		rdev->asic->gart.get_page_entry = &r100_pci_gart_get_page_entry;
 		rdev->asic->gart.set_page = &r100_pci_gart_set_page;
 	}
 	rdev->mc.gtt_size = radeon_gart_size * 1024 * 1024;
@@ -199,6 +201,7 @@ static struct radeon_asic r100_asic = {
 	.mc_wait_for_idle = &r100_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &r100_pci_gart_tlb_flush,
+		.get_page_entry = &r100_pci_gart_get_page_entry,
 		.set_page = &r100_pci_gart_set_page,
 	},
 	.ring = {
@@ -265,6 +268,7 @@ static struct radeon_asic r200_asic = {
 	.mc_wait_for_idle = &r100_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &r100_pci_gart_tlb_flush,
+		.get_page_entry = &r100_pci_gart_get_page_entry,
 		.set_page = &r100_pci_gart_set_page,
 	},
 	.ring = {
@@ -359,6 +363,7 @@ static struct radeon_asic r300_asic = {
 	.mc_wait_for_idle = &r300_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &r100_pci_gart_tlb_flush,
+		.get_page_entry = &r100_pci_gart_get_page_entry,
 		.set_page = &r100_pci_gart_set_page,
 	},
 	.ring = {
@@ -425,6 +430,7 @@ static struct radeon_asic r300_asic_pcie = {
 	.mc_wait_for_idle = &r300_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rv370_pcie_gart_tlb_flush,
+		.get_page_entry = &rv370_pcie_gart_get_page_entry,
 		.set_page = &rv370_pcie_gart_set_page,
 	},
 	.ring = {
@@ -491,6 +497,7 @@ static struct radeon_asic r420_asic = {
 	.mc_wait_for_idle = &r300_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rv370_pcie_gart_tlb_flush,
+		.get_page_entry = &rv370_pcie_gart_get_page_entry,
 		.set_page = &rv370_pcie_gart_set_page,
 	},
 	.ring = {
@@ -557,6 +564,7 @@ static struct radeon_asic rs400_asic = {
 	.mc_wait_for_idle = &rs400_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rs400_gart_tlb_flush,
+		.get_page_entry = &rs400_gart_get_page_entry,
 		.set_page = &rs400_gart_set_page,
 	},
 	.ring = {
@@ -623,6 +631,7 @@ static struct radeon_asic rs600_asic = {
 	.mc_wait_for_idle = &rs600_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rs600_gart_tlb_flush,
+		.get_page_entry = &rs600_gart_get_page_entry,
 		.set_page = &rs600_gart_set_page,
 	},
 	.ring = {
@@ -691,6 +700,7 @@ static struct radeon_asic rs690_asic = {
 	.mc_wait_for_idle = &rs690_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rs400_gart_tlb_flush,
+		.get_page_entry = &rs400_gart_get_page_entry,
 		.set_page = &rs400_gart_set_page,
 	},
 	.ring = {
@@ -759,6 +769,7 @@ static struct radeon_asic rv515_asic = {
 	.mc_wait_for_idle = &rv515_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rv370_pcie_gart_tlb_flush,
+		.get_page_entry = &rv370_pcie_gart_get_page_entry,
 		.set_page = &rv370_pcie_gart_set_page,
 	},
 	.ring = {
@@ -825,6 +836,7 @@ static struct radeon_asic r520_asic = {
 	.mc_wait_for_idle = &r520_mc_wait_for_idle,
 	.gart = {
 		.tlb_flush = &rv370_pcie_gart_tlb_flush,
+		.get_page_entry = &am