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

Pull drm updates from Dave Airlie:
 "This is the main drm pull request for 3.10.

  Wierd bits:
   - OMAP drm changes required OMAP dss changes, in drivers/video, so I
     took them in here.
   - one more fbcon fix for font handover
   - VT switch avoidance in pm code
   - scatterlist helpers for gpu drivers - have acks from akpm

  Highlights:
   - qxl kms driver - driver for the spice qxl virtual GPU

  Nouveau:
   - fermi/kepler VRAM compression
   - GK110/nvf0 modesetting support.

  Tegra:
   - host1x core merged with 2D engine support

  i915:
   - vt switchless resume
   - more valleyview support
   - vblank fixes
   - modesetting pipe config rework

  radeon:
   - UVD engine support
   - SI chip tiling support
   - GPU registers initialisation from golden values.

  exynos:
   - device tree changes
   - fimc block support

  Otherwise:
   - bunches of fixes all over the place."

* 'drm-next' of git://people.freedesktop.org/~airlied/linux: (513 commits)
  qxl: update to new idr interfaces.
  drm/nouveau: fix build with nv50->nvc0
  drm/radeon: fix handling of v6 power tables
  drm/radeon: clarify family checks in pm table parsing
  drm/radeon: consolidate UVD clock programming
  drm/radeon: fix UPLL_REF_DIV_MASK definition
  radeon: add bo tracking debugfs
  drm/radeon: add new richland pci ids
  drm/radeon: add some new SI PCI ids
  drm/radeon: fix scratch reg handling for UVD fence
  drm/radeon: allocate SA bo in the requested domain
  drm/radeon: fix possible segfault when parsing pm tables
  drm/radeon: fix endian bugs in atom_allocate_fb_scratch()
  OMAPDSS: TFP410: return EPROBE_DEFER if the i2c adapter not found
  OMAPDSS: VENC: Add error handling for venc_probe_pdata
  OMAPDSS: HDMI: Add error handling for hdmi_probe_pdata
  OMAPDSS: RFBI: Add error handling for rfbi_probe_pdata
  OMAPDSS: DSI: Add error handling for dsi_probe_pdata
  OMAPDSS: SDI: Add error handling for sdi_probe_pdata
  OMAPDSS: DPI: Add error handling for dpi_probe_pdata
  ...

44 files changed, 11160 insertions, 0 deletions

diff --git a/drivers/gpu/host1x/Kconfig b/drivers/gpu/host1x/Kconfig
new file mode 100644
index 00000000000..ccfd42b2360
--- /dev/null
+++ b/drivers/gpu/host1x/Kconfig
@@ -0,0 +1,24 @@
+config TEGRA_HOST1X
+	tristate "NVIDIA Tegra host1x driver"
+	depends on ARCH_TEGRA || ARCH_MULTIPLATFORM
+	help
+	  Driver for the NVIDIA Tegra host1x hardware.
+
+	  The Tegra host1x module is the DMA engine for register access to
+	  Tegra's graphics- and multimedia-related modules. The modules served
+	  by host1x are referred to as clients. host1x includes some other
+	  functionality, such as synchronization.
+
+if TEGRA_HOST1X
+
+config TEGRA_HOST1X_FIREWALL
+	bool "Enable HOST1X security firewall"
+	default y
+	help
+	  Say yes if kernel should protect command streams from tampering.
+
+	  If unsure, choose Y.
+
+source "drivers/gpu/host1x/drm/Kconfig"
+
+endif
diff --git a/drivers/gpu/host1x/Makefile b/drivers/gpu/host1x/Makefile
new file mode 100644
index 00000000000..3b037b6e029
--- /dev/null
+++ b/drivers/gpu/host1x/Makefile
@@ -0,0 +1,20 @@
+ccflags-y = -Idrivers/gpu/host1x
+
+host1x-y = \
+	syncpt.o \
+	dev.o \
+	intr.o \
+	cdma.o \
+	channel.o \
+	job.o \
+	debug.o \
+	hw/host1x01.o
+
+ccflags-y += -Iinclude/drm
+ccflags-$(CONFIG_DRM_TEGRA_DEBUG) += -DDEBUG
+
+host1x-$(CONFIG_DRM_TEGRA) += drm/drm.o drm/fb.o drm/dc.o
+host1x-$(CONFIG_DRM_TEGRA) += drm/output.o drm/rgb.o drm/hdmi.o
+host1x-$(CONFIG_DRM_TEGRA) += drm/gem.o
+host1x-$(CONFIG_DRM_TEGRA) += drm/gr2d.o
+obj-$(CONFIG_TEGRA_HOST1X) += host1x.o
diff --git a/drivers/gpu/host1x/cdma.c b/drivers/gpu/host1x/cdma.c
new file mode 100644
index 00000000000..de72172d3b5
--- /dev/null
+++ b/drivers/gpu/host1x/cdma.c
@@ -0,0 +1,491 @@
+/*
+ * Tegra host1x Command DMA
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+
+#include <asm/cacheflush.h>
+#include <linux/device.h>
+#include <linux/dma-mapping.h>
+#include <linux/interrupt.h>
+#include <linux/kernel.h>
+#include <linux/kfifo.h>
+#include <linux/slab.h>
+#include <trace/events/host1x.h>
+
+#include "cdma.h"
+#include "channel.h"
+#include "dev.h"
+#include "debug.h"
+#include "host1x_bo.h"
+#include "job.h"
+
+/*
+ * push_buffer
+ *
+ * The push buffer is a circular array of words to be fetched by command DMA.
+ * Note that it works slightly differently to the sync queue; fence == pos
+ * means that the push buffer is full, not empty.
+ */
+
+#define HOST1X_PUSHBUFFER_SLOTS	512
+
+/*
+ * Clean up push buffer resources
+ */
+static void host1x_pushbuffer_destroy(struct push_buffer *pb)
+{
+	struct host1x_cdma *cdma = pb_to_cdma(pb);
+	struct host1x *host1x = cdma_to_host1x(cdma);
+
+	if (pb->phys != 0)
+		dma_free_writecombine(host1x->dev, pb->size_bytes + 4,
+				      pb->mapped, pb->phys);
+
+	pb->mapped = NULL;
+	pb->phys = 0;
+}
+
+/*
+ * Init push buffer resources
+ */
+static int host1x_pushbuffer_init(struct push_buffer *pb)
+{
+	struct host1x_cdma *cdma = pb_to_cdma(pb);
+	struct host1x *host1x = cdma_to_host1x(cdma);
+
+	pb->mapped = NULL;
+	pb->phys = 0;
+	pb->size_bytes = HOST1X_PUSHBUFFER_SLOTS * 8;
+
+	/* initialize buffer pointers */
+	pb->fence = pb->size_bytes - 8;
+	pb->pos = 0;
+
+	/* allocate and map pushbuffer memory */
+	pb->mapped = dma_alloc_writecombine(host1x->dev, pb->size_bytes + 4,
+					    &pb->phys, GFP_KERNEL);
+	if (!pb->mapped)
+		goto fail;
+
+	host1x_hw_pushbuffer_init(host1x, pb);
+
+	return 0;
+
+fail:
+	host1x_pushbuffer_destroy(pb);
+	return -ENOMEM;
+}
+
+/*
+ * Push two words to the push buffer
+ * Caller must ensure push buffer is not full
+ */
+static void host1x_pushbuffer_push(struct push_buffer *pb, u32 op1, u32 op2)
+{
+	u32 pos = pb->pos;
+	u32 *p = (u32 *)((u32)pb->mapped + pos);
+	WARN_ON(pos == pb->fence);
+	*(p++) = op1;
+	*(p++) = op2;
+	pb->pos = (pos + 8) & (pb->size_bytes - 1);
+}
+
+/*
+ * Pop a number of two word slots from the push buffer
+ * Caller must ensure push buffer is not empty
+ */
+static void host1x_pushbuffer_pop(struct push_buffer *pb, unsigned int slots)
+{
+	/* Advance the next write position */
+	pb->fence = (pb->fence + slots * 8) & (pb->size_bytes - 1);
+}
+
+/*
+ * Return the number of two word slots free in the push buffer
+ */
+static u32 host1x_pushbuffer_space(struct push_buffer *pb)
+{
+	return ((pb->fence - pb->pos) & (pb->size_bytes - 1)) / 8;
+}
+
+/*
+ * Sleep (if necessary) until the requested event happens
+ *   - CDMA_EVENT_SYNC_QUEUE_EMPTY : sync queue is completely empty.
+ *     - Returns 1
+ *   - CDMA_EVENT_PUSH_BUFFER_SPACE : there is space in the push buffer
+ *     - Return the amount of space (> 0)
+ * Must be called with the cdma lock held.
+ */
+unsigned int host1x_cdma_wait_locked(struct host1x_cdma *cdma,
+				     enum cdma_event event)
+{
+	for (;;) {
+		unsigned int space;
+
+		if (event == CDMA_EVENT_SYNC_QUEUE_EMPTY)
+			space = list_empty(&cdma->sync_queue) ? 1 : 0;
+		else if (event == CDMA_EVENT_PUSH_BUFFER_SPACE) {
+			struct push_buffer *pb = &cdma->push_buffer;
+			space = host1x_pushbuffer_space(pb);
+		} else {
+			WARN_ON(1);
+			return -EINVAL;
+		}
+
+		if (space)
+			return space;
+
+		trace_host1x_wait_cdma(dev_name(cdma_to_channel(cdma)->dev),
+				       event);
+
+		/* If somebody has managed to already start waiting, yield */
+		if (cdma->event != CDMA_EVENT_NONE) {
+			mutex_unlock(&cdma->lock);
+			schedule();
+			mutex_lock(&cdma->lock);
+			continue;
+		}
+		cdma->event = event;
+
+		mutex_unlock(&cdma->lock);
+		down(&cdma->sem);
+		mutex_lock(&cdma->lock);
+	}
+	return 0;
+}
+
+/*
+ * Start timer that tracks the time spent by the job.
+ * Must be called with the cdma lock held.
+ */
+static void cdma_start_timer_locked(struct host1x_cdma *cdma,
+				    struct host1x_job *job)
+{
+	struct host1x *host = cdma_to_host1x(cdma);
+
+	if (cdma->timeout.client) {
+		/* timer already started */
+		return;
+	}
+
+	cdma->timeout.client = job->client;
+	cdma->timeout.syncpt = host1x_syncpt_get(host, job->syncpt_id);
+	cdma->timeout.syncpt_val = job->syncpt_end;
+	cdma->timeout.start_ktime = ktime_get();
+
+	schedule_delayed_work(&cdma->timeout.wq,
+			      msecs_to_jiffies(job->timeout));
+}
+
+/*
+ * Stop timer when a buffer submission completes.
+ * Must be called with the cdma lock held.
+ */
+static void stop_cdma_timer_locked(struct host1x_cdma *cdma)
+{
+	cancel_delayed_work(&cdma->timeout.wq);
+	cdma->timeout.client = 0;
+}
+
+/*
+ * For all sync queue entries that have already finished according to the
+ * current sync point registers:
+ *  - unpin & unref their mems
+ *  - pop their push buffer slots
+ *  - remove them from the sync queue
+ * This is normally called from the host code's worker thread, but can be
+ * called manually if necessary.
+ * Must be called with the cdma lock held.
+ */
+static void update_cdma_locked(struct host1x_cdma *cdma)
+{
+	bool signal = false;
+	struct host1x *host1x = cdma_to_host1x(cdma);
+	struct host1x_job *job, *n;
+
+	/* If CDMA is stopped, queue is cleared and we can return */
+	if (!cdma->running)
+		return;
+
+	/*
+	 * Walk the sync queue, reading the sync point registers as necessary,
+	 * to consume as many sync queue entries as possible without blocking
+	 */
+	list_for_each_entry_safe(job, n, &cdma->sync_queue, list) {
+		struct host1x_syncpt *sp =
+			host1x_syncpt_get(host1x, job->syncpt_id);
+
+		/* Check whether this syncpt has completed, and bail if not */
+		if (!host1x_syncpt_is_expired(sp, job->syncpt_end)) {
+			/* Start timer on next pending syncpt */
+			if (job->timeout)
+				cdma_start_timer_locked(cdma, job);
+			break;
+		}
+
+		/* Cancel timeout, when a buffer completes */
+		if (cdma->timeout.client)
+			stop_cdma_timer_locked(cdma);
+
+		/* Unpin the memory */
+		host1x_job_unpin(job);
+
+		/* Pop push buffer slots */
+		if (job->num_slots) {
+			struct push_buffer *pb = &cdma->push_buffer;
+			host1x_pushbuffer_pop(pb, job->num_slots);
+			if (cdma->event == CDMA_EVENT_PUSH_BUFFER_SPACE)
+				signal = true;
+		}
+
+		list_del(&job->list);
+		host1x_job_put(job);
+	}
+
+	if (cdma->event == CDMA_EVENT_SYNC_QUEUE_EMPTY &&
+	    list_empty(&cdma->sync_queue))
+		signal = true;
+
+	if (signal) {
+		cdma->event = CDMA_EVENT_NONE;
+		up(&cdma->sem);
+	}
+}
+
+void host1x_cdma_update_sync_queue(struct host1x_cdma *cdma,
+				   struct device *dev)
+{
+	u32 restart_addr;
+	u32 syncpt_incrs;
+	struct host1x_job *job = NULL;
+	u32 syncpt_val;
+	struct host1x *host1x = cdma_to_host1x(cdma);
+
+	syncpt_val = host1x_syncpt_load(cdma->timeout.syncpt);
+
+	dev_dbg(dev, "%s: starting cleanup (thresh %d)\n",
+		__func__, syncpt_val);
+
+	/*
+	 * Move the sync_queue read pointer to the first entry that hasn't
+	 * completed based on the current HW syncpt value. It's likely there
+	 * won't be any (i.e. we're still at the head), but covers the case
+	 * where a syncpt incr happens just prior/during the teardown.
+	 */
+
+	dev_dbg(dev, "%s: skip completed buffers still in sync_queue\n",
+		__func__);
+
+	list_for_each_entry(job, &cdma->sync_queue, list) {
+		if (syncpt_val < job->syncpt_end)
+			break;
+
+		host1x_job_dump(dev, job);
+	}
+
+	/*
+	 * Walk the sync_queue, first incrementing with the CPU syncpts that
+	 * are partially executed (the first buffer) or fully skipped while
+	 * still in the current context (slots are also NOP-ed).
+	 *
+	 * At the point contexts are interleaved, syncpt increments must be
+	 * done inline with the pushbuffer from a GATHER buffer to maintain
+	 * the order (slots are modified to be a GATHER of syncpt incrs).
+	 *
+	 * Note: save in restart_addr the location where the timed out buffer
+	 * started in the PB, so we can start the refetch from there (with the
+	 * modified NOP-ed PB slots). This lets things appear to have completed
+	 * properly for this buffer and resources are freed.
+	 */
+
+	dev_dbg(dev, "%s: perform CPU incr on pending same ctx buffers\n",
+		__func__);
+
+	if (!list_empty(&cdma->sync_queue))
+		restart_addr = job->first_get;
+	else
+		restart_addr = cdma->last_pos;
+
+	/* do CPU increments as long as this context continues */
+	list_for_each_entry_from(job, &cdma->sync_queue, list) {
+		/* different context, gets us out of this loop */
+		if (job->client != cdma->timeout.client)
+			break;
+
+		/* won't need a timeout when replayed */
+		job->timeout = 0;
+
+		syncpt_incrs = job->syncpt_end - syncpt_val;
+		dev_dbg(dev, "%s: CPU incr (%d)\n", __func__, syncpt_incrs);
+
+		host1x_job_dump(dev, job);
+
+		/* safe to use CPU to incr syncpts */
+		host1x_hw_cdma_timeout_cpu_incr(host1x, cdma, job->first_get,
+						syncpt_incrs, job->syncpt_end,
+						job->num_slots);
+
+		syncpt_val += syncpt_incrs;
+	}
+
+	/* The following sumbits from the same client may be dependent on the
+	 * failed submit and therefore they may fail. Force a small timeout
+	 * to make the queue cleanup faster */
+
+	list_for_each_entry_from(job, &cdma->sync_queue, list)
+		if (job->client == cdma->timeout.client)
+			job->timeout = min_t(unsigned int, job->timeout, 500);
+
+	dev_dbg(dev, "%s: finished sync_queue modification\n", __func__);
+
+	/* roll back DMAGET and start up channel again */
+	host1x_hw_cdma_resume(host1x, cdma, restart_addr);
+}
+
+/*
+ * Create a cdma
+ */
+int host1x_cdma_init(struct host1x_cdma *cdma)
+{
+	int err;
+
+	mutex_init(&cdma->lock);
+	sema_init(&cdma->sem, 0);
+
+	INIT_LIST_HEAD(&cdma->sync_queue);
+
+	cdma->event = CDMA_EVENT_NONE;
+	cdma->running = false;
+	cdma->torndown = false;
+
+	err = host1x_pushbuffer_init(&cdma->push_buffer);
+	if (err)
+		return err;
+	return 0;
+}
+
+/*
+ * Destroy a cdma
+ */
+int host1x_cdma_deinit(struct host1x_cdma *cdma)
+{
+	struct push_buffer *pb = &cdma->push_buffer;
+	struct host1x *host1x = cdma_to_host1x(cdma);
+
+	if (cdma->running) {
+		pr_warn("%s: CDMA still running\n", __func__);
+		return -EBUSY;
+	}
+
+	host1x_pushbuffer_destroy(pb);
+	host1x_hw_cdma_timeout_destroy(host1x, cdma);
+
+	return 0;
+}
+
+/*
+ * Begin a cdma submit
+ */
+int host1x_cdma_begin(struct host1x_cdma *cdma, struct host1x_job *job)
+{
+	struct host1x *host1x = cdma_to_host1x(cdma);
+
+	mutex_lock(&cdma->lock);
+
+	if (job->timeout) {
+		/* init state on first submit with timeout value */
+		if (!cdma->timeout.initialized) {
+			int err;
+			err = host1x_hw_cdma_timeout_init(host1x, cdma,
+							  job->syncpt_id);
+			if (err) {
+				mutex_unlock(&cdma->lock);
+				return err;
+			}
+		}
+	}
+	if (!cdma->running)
+		host1x_hw_cdma_start(host1x, cdma);
+
+	cdma->slots_free = 0;
+	cdma->slots_used = 0;
+	cdma->first_get = cdma->push_buffer.pos;
+
+	trace_host1x_cdma_begin(dev_name(job->channel->dev));
+	return 0;
+}
+
+/*
+ * Push two words into a push buffer slot
+ * Blocks as necessary if the push buffer is full.
+ */
+void host1x_cdma_push(struct host1x_cdma *cdma, u32 op1, u32 op2)
+{
+	struct host1x *host1x = cdma_to_host1x(cdma);
+	struct push_buffer *pb = &cdma->push_buffer;
+	u32 slots_free = cdma->slots_free;
+
+	if (host1x_debug_trace_cmdbuf)
+		trace_host1x_cdma_push(dev_name(cdma_to_channel(cdma)->dev),
+				       op1, op2);
+
+	if (slots_free == 0) {
+		host1x_hw_cdma_flush(host1x, cdma);
+		slots_free = host1x_cdma_wait_locked(cdma,
+						CDMA_EVENT_PUSH_BUFFER_SPACE);
+	}
+	cdma->slots_free = slots_free - 1;
+	cdma->slots_used++;
+	host1x_pushbuffer_push(pb, op1, op2);
+}
+
+/*
+ * End a cdma submit
+ * Kick off DMA, add job to the sync queue, and a number of slots to be freed
+ * from the pushbuffer. The handles for a submit must all be pinned at the same
+ * time, but they can be unpinned in smaller chunks.
+ */
+void host1x_cdma_end(struct host1x_cdma *cdma,
+		     struct host1x_job *job)
+{
+	struct host1x *host1x = cdma_to_host1x(cdma);
+	bool idle = list_empty(&cdma->sync_queue);
+
+	host1x_hw_cdma_flush(host1x, cdma);
+
+	job->first_get = cdma->first_get;
+	job->num_slots = cdma->slots_used;
+	host1x_job_get(job);
+	list_add_tail(&job->list, &cdma->sync_queue);
+
+	/* start timer on idle -> active transitions */
+	if (job->timeout && idle)
+		cdma_start_timer_locked(cdma, job);
+
+	trace_host1x_cdma_end(dev_name(job->channel->dev));
+	mutex_unlock(&cdma->lock);
+}
+
+/*
+ * Update cdma state according to current sync point values
+ */
+void host1x_cdma_update(struct host1x_cdma *cdma)
+{
+	mutex_lock(&cdma->lock);
+	update_cdma_locked(cdma);
+	mutex_unlock(&cdma->lock);
+}
diff --git a/drivers/gpu/host1x/cdma.h b/drivers/gpu/host1x/cdma.h
new file mode 100644
index 00000000000..313c4b78434
--- /dev/null
+++ b/drivers/gpu/host1x/cdma.h
@@ -0,0 +1,100 @@
+/*
+ * Tegra host1x Command DMA
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __HOST1X_CDMA_H
+#define __HOST1X_CDMA_H
+
+#include <linux/sched.h>
+#include <linux/semaphore.h>
+#include <linux/list.h>
+
+struct host1x_syncpt;
+struct host1x_userctx_timeout;
+struct host1x_job;
+
+/*
+ * cdma
+ *
+ * This is in charge of a host command DMA channel.
+ * Sends ops to a push buffer, and takes responsibility for unpinning
+ * (& possibly freeing) of memory after those ops have completed.
+ * Producer:
+ *	begin
+ *		push - send ops to the push buffer
+ *	end - start command DMA and enqueue handles to be unpinned
+ * Consumer:
+ *	update - call to update sync queue and push buffer, unpin memory
+ */
+
+struct push_buffer {
+	u32 *mapped;			/* mapped pushbuffer memory */
+	dma_addr_t phys;		/* physical address of pushbuffer */
+	u32 fence;			/* index we've written */
+	u32 pos;			/* index to write to */
+	u32 size_bytes;
+};
+
+struct buffer_timeout {
+	struct delayed_work wq;		/* work queue */
+	bool initialized;		/* timer one-time setup flag */
+	struct host1x_syncpt *syncpt;	/* buffer completion syncpt */
+	u32 syncpt_val;			/* syncpt value when completed */
+	ktime_t start_ktime;		/* starting time */
+	/* context timeout information */
+	int client;
+};
+
+enum cdma_event {
+	CDMA_EVENT_NONE,		/* not waiting for any event */
+	CDMA_EVENT_SYNC_QUEUE_EMPTY,	/* wait for empty sync queue */
+	CDMA_EVENT_PUSH_BUFFER_SPACE	/* wait for space in push buffer */
+};
+
+struct host1x_cdma {
+	struct mutex lock;		/* controls access to shared state */
+	struct semaphore sem;		/* signalled when event occurs */
+	enum cdma_event event;		/* event that sem is waiting for */
+	unsigned int slots_used;	/* pb slots used in current submit */
+	unsigned int slots_free;	/* pb slots free in current submit */
+	unsigned int first_get;		/* DMAGET value, where submit begins */
+	unsigned int last_pos;		/* last value written to DMAPUT */
+	struct push_buffer push_buffer;	/* channel's push buffer */
+	struct list_head sync_queue;	/* job queue */
+	struct buffer_timeout timeout;	/* channel's timeout state/wq */
+	bool running;
+	bool torndown;
+};
+
+#define cdma_to_channel(cdma) container_of(cdma, struct host1x_channel, cdma)
+#define cdma_to_host1x(cdma) dev_get_drvdata(cdma_to_channel(cdma)->dev->parent)
+#define pb_to_cdma(pb) container_of(pb, struct host1x_cdma, push_buffer)
+
+int host1x_cdma_init(struct host1x_cdma *cdma);
+int host1x_cdma_deinit(struct host1x_cdma *cdma);
+void host1x_cdma_stop(struct host1x_cdma *cdma);
+int host1x_cdma_begin(struct host1x_cdma *cdma, struct host1x_job *job);
+void host1x_cdma_push(struct host1x_cdma *cdma, u32 op1, u32 op2);
+void host1x_cdma_end(struct host1x_cdma *cdma, struct host1x_job *job);
+void host1x_cdma_update(struct host1x_cdma *cdma);
+void host1x_cdma_peek(struct host1x_cdma *cdma, u32 dmaget, int slot,
+		      u32 *out);
+unsigned int host1x_cdma_wait_locked(struct host1x_cdma *cdma,
+				     enum cdma_event event);
+void host1x_cdma_update_sync_queue(struct host1x_cdma *cdma,
+				   struct device *dev);
+#endif
diff --git a/drivers/gpu/host1x/channel.c b/drivers/gpu/host1x/channel.c
new file mode 100644
index 00000000000..83ea51b9f0f
--- /dev/null
+++ b/drivers/gpu/host1x/channel.c
@@ -0,0 +1,126 @@
+/*
+ * Tegra host1x Channel
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/slab.h>
+#include <linux/module.h>
+
+#include "channel.h"
+#include "dev.h"
+#include "job.h"
+
+/* Constructor for the host1x device list */
+int host1x_channel_list_init(struct host1x *host)
+{
+	INIT_LIST_HEAD(&host->chlist.list);
+	mutex_init(&host->chlist_mutex);
+
+	if (host->info->nb_channels > BITS_PER_LONG) {
+		WARN(1, "host1x hardware has more channels than supported by the driver\n");
+		return -ENOSYS;
+	}
+
+	return 0;
+}
+
+int host1x_job_submit(struct host1x_job *job)
+{
+	struct host1x *host = dev_get_drvdata(job->channel->dev->parent);
+
+	return host1x_hw_channel_submit(host, job);
+}
+
+struct host1x_channel *host1x_channel_get(struct host1x_channel *channel)
+{
+	int err = 0;
+
+	mutex_lock(&channel->reflock);
+
+	if (channel->refcount == 0)
+		err = host1x_cdma_init(&channel->cdma);
+
+	if (!err)
+		channel->refcount++;
+
+	mutex_unlock(&channel->reflock);
+
+	return err ? NULL : channel;
+}
+
+void host1x_channel_put(struct host1x_channel *channel)
+{
+	mutex_lock(&channel->reflock);
+
+	if (channel->refcount == 1) {
+		struct host1x *host = dev_get_drvdata(channel->dev->parent);
+
+		host1x_hw_cdma_stop(host, &channel->cdma);
+		host1x_cdma_deinit(&channel->cdma);
+	}
+
+	channel->refcount--;
+
+	mutex_unlock(&channel->reflock);
+}
+
+struct host1x_channel *host1x_channel_request(struct device *dev)
+{
+	struct host1x *host = dev_get_drvdata(dev->parent);
+	int max_channels = host->info->nb_channels;
+	struct host1x_channel *channel = NULL;
+	int index, err;
+
+	mutex_lock(&host->chlist_mutex);
+
+	index = find_first_zero_bit(&host->allocated_channels, max_channels);
+	if (index >= max_channels)
+		goto fail;
+
+	channel = kzalloc(sizeof(*channel), GFP_KERNEL);
+	if (!channel)
+		goto fail;
+
+	err = host1x_hw_channel_init(host, channel, index);
+	if (err < 0)
+		goto fail;
+
+	/* Link device to host1x_channel */
+	channel->dev = dev;
+
+	/* Add to channel list */
+	list_add_tail(&channel->list, &host->chlist.list);
+
+	host->allocated_channels |= BIT(index);
+
+	mutex_unlock(&host->chlist_mutex);
+	return channel;
+
+fail:
+	dev_err(dev, "failed to init channel\n");
+	kfree(channel);
+	mutex_unlock(&host->chlist_mutex);
+	return NULL;
+}
+
+void host1x_channel_free(struct host1x_channel *channel)
+{
+	struct host1x *host = dev_get_drvdata(channel->dev->parent);
+
+	host->allocated_channels &= ~BIT(channel->id);
+	list_del(&channel->list);
+	kfree(channel);
+}
diff --git a/drivers/gpu/host1x/channel.h b/drivers/gpu/host1x/channel.h
new file mode 100644
index 00000000000..48723b8eea4
--- /dev/null
+++ b/drivers/gpu/host1x/channel.h
@@ -0,0 +1,52 @@
+/*
+ * Tegra host1x Channel
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __HOST1X_CHANNEL_H
+#define __HOST1X_CHANNEL_H
+
+#include <linux/io.h>
+
+#include "cdma.h"
+
+struct host1x;
+
+struct host1x_channel {
+	struct list_head list;
+
+	unsigned int refcount;
+	unsigned int id;
+	struct mutex reflock;
+	struct mutex submitlock;
+	void __iomem *regs;
+	struct device *dev;
+	struct host1x_cdma cdma;
+};
+
+/* channel list operations */
+int host1x_channel_list_init(struct host1x *host);
+
+struct host1x_channel *host1x_channel_request(struct device *dev);
+void host1x_channel_free(struct host1x_channel *channel);
+struct host1x_channel *host1x_channel_get(struct host1x_channel *channel);
+void host1x_channel_put(struct host1x_channel *channel);
+int host1x_job_submit(struct host1x_job *job);
+
+#define host1x_for_each_channel(host, channel)				\
+	list_for_each_entry(channel, &host->chlist.list, list)
+
+#endif
diff --git a/drivers/gpu/host1x/debug.c b/drivers/gpu/host1x/debug.c
new file mode 100644
index 00000000000..3ec7d77de24
--- /dev/null
+++ b/drivers/gpu/host1x/debug.c
@@ -0,0 +1,210 @@
+/*
+ * Copyright (C) 2010 Google, Inc.
+ * Author: Erik Gilling <konkers@android.com>
+ *
+ * Copyright (C) 2011-2013 NVIDIA Corporation
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ */
+
+#include <linux/debugfs.h>
+#include <linux/seq_file.h>
+#include <linux/uaccess.h>
+
+#include <linux/io.h>
+
+#include "dev.h"
+#include "debug.h"
+#include "channel.h"
+
+unsigned int host1x_debug_trace_cmdbuf;
+
+static pid_t host1x_debug_force_timeout_pid;
+static u32 host1x_debug_force_timeout_val;
+static u32 host1x_debug_force_timeout_channel;
+
+void host1x_debug_output(struct output *o, const char *fmt, ...)
+{
+	va_list args;
+	int len;
+
+	va_start(args, fmt);
+	len = vsnprintf(o->buf, sizeof(o->buf), fmt, args);
+	va_end(args);
+	o->fn(o->ctx, o->buf, len);
+}
+
+static int show_channels(struct host1x_channel *ch, void *data, bool show_fifo)
+{
+	struct host1x *m = dev_get_drvdata(ch->dev->parent);
+	struct output *o = data;
+
+	mutex_lock(&ch->reflock);
+	if (ch->refcount) {
+		mutex_lock(&ch->cdma.lock);
+		if (show_fifo)
+			host1x_hw_show_channel_fifo(m, ch, o);
+		host1x_hw_show_channel_cdma(m, ch, o);
+		mutex_unlock(&ch->cdma.lock);
+	}
+	mutex_unlock(&ch->reflock);
+
+	return 0;
+}
+
+static void show_syncpts(struct host1x *m, struct output *o)
+{
+	int i;
+	host1x_debug_output(o, "---- syncpts ----\n");
+	for (i = 0; i < host1x_syncpt_nb_pts(m); i++) {
+		u32 max = host1x_syncpt_read_max(m->syncpt + i);
+		u32 min = host1x_syncpt_load(m->syncpt + i);
+		if (!min && !max)
+			continue;
+		host1x_debug_output(o, "id %d (%s) min %d max %d\n",
+				    i, m->syncpt[i].name, min, max);
+	}
+
+	for (i = 0; i < host1x_syncpt_nb_bases(m); i++) {
+		u32 base_val;
+		base_val = host1x_syncpt_load_wait_base(m->syncpt + i);
+		if (base_val)
+			host1x_debug_output(o, "waitbase id %d val %d\n", i,
+					    base_val);
+	}
+
+	host1x_debug_output(o, "\n");
+}
+
+static void show_all(struct host1x *m, struct output *o)
+{
+	struct host1x_channel *ch;
+
+	host1x_hw_show_mlocks(m, o);
+	show_syncpts(m, o);
+	host1x_debug_output(o, "---- channels ----\n");
+
+	host1x_for_each_channel(m, ch)
+		show_channels(ch, o, true);
+}
+
+#ifdef CONFIG_DEBUG_FS
+static void show_all_no_fifo(struct host1x *host1x, struct output *o)
+{
+	struct host1x_channel *ch;
+
+	host1x_hw_show_mlocks(host1x, o);
+	show_syncpts(host1x, o);
+	host1x_debug_output(o, "---- channels ----\n");
+
+	host1x_for_each_channel(host1x, ch)
+		show_channels(ch, o, false);
+}
+
+static int host1x_debug_show_all(struct seq_file *s, void *unused)
+{
+	struct output o = {
+		.fn = write_to_seqfile,
+		.ctx = s
+	};
+	show_all(s->private, &o);
+	return 0;
+}
+
+static int host1x_debug_show(struct seq_file *s, void *unused)
+{
+	struct output o = {
+		.fn = write_to_seqfile,
+		.ctx = s
+	};
+	show_all_no_fifo(s->private, &o);
+	return 0;
+}
+
+static int host1x_debug_open_all(struct inode *inode, struct file *file)
+{
+	return single_open(file, host1x_debug_show_all, inode->i_private);
+}
+
+static const struct file_operations host1x_debug_all_fops = {
+	.open		= host1x_debug_open_all,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
+static int host1x_debug_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, host1x_debug_show, inode->i_private);
+}
+
+static const struct file_operations host1x_debug_fops = {
+	.open		= host1x_debug_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
+void host1x_debug_init(struct host1x *host1x)
+{
+	struct dentry *de = debugfs_create_dir("tegra-host1x", NULL);
+
+	if (!de)
+		return;
+
+	/* Store the created entry */
+	host1x->debugfs = de;
+
+	debugfs_create_file("status", S_IRUGO, de, host1x, &host1x_debug_fops);
+	debugfs_create_file("status_all", S_IRUGO, de, host1x,
+			    &host1x_debug_all_fops);
+
+	debugfs_create_u32("trace_cmdbuf", S_IRUGO|S_IWUSR, de,
+			   &host1x_debug_trace_cmdbuf);
+
+	host1x_hw_debug_init(host1x, de);
+
+	debugfs_create_u32("force_timeout_pid", S_IRUGO|S_IWUSR, de,
+			   &host1x_debug_force_timeout_pid);
+	debugfs_create_u32("force_timeout_val", S_IRUGO|S_IWUSR, de,
+			   &host1x_debug_force_timeout_val);
+	debugfs_create_u32("force_timeout_channel", S_IRUGO|S_IWUSR, de,
+			   &host1x_debug_force_timeout_channel);
+}
+
+void host1x_debug_deinit(struct host1x *host1x)
+{
+	debugfs_remove_recursive(host1x->debugfs);
+}
+#else
+void host1x_debug_init(struct host1x *host1x)
+{
+}
+void host1x_debug_deinit(struct host1x *host1x)
+{
+}
+#endif
+
+void host1x_debug_dump(struct host1x *host1x)
+{
+	struct output o = {
+		.fn = write_to_printk
+	};
+	show_all(host1x, &o);
+}
+
+void host1x_debug_dump_syncpts(struct host1x *host1x)
+{
+	struct output o = {
+		.fn = write_to_printk
+	};
+	show_syncpts(host1x, &o);
+}
diff --git a/drivers/gpu/host1x/debug.h b/drivers/gpu/host1x/debug.h
new file mode 100644
index 00000000000..4595b2e0799
--- /dev/null
+++ b/drivers/gpu/host1x/debug.h
@@ -0,0 +1,51 @@
+/*
+ * Tegra host1x Debug
+ *
+ * Copyright (c) 2011-2013 NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+#ifndef __HOST1X_DEBUG_H
+#define __HOST1X_DEBUG_H
+
+#include <linux/debugfs.h>
+#include <linux/seq_file.h>
+
+struct host1x;
+
+struct output {
+	void (*fn)(void *ctx, const char *str, size_t len);
+	void *ctx;
+	char buf[256];
+};
+
+static inline void write_to_seqfile(void *ctx, const char *str, size_t len)
+{
+	seq_write((struct seq_file *)ctx, str, len);
+}
+
+static inline void write_to_printk(void *ctx, const char *str, size_t len)
+{
+	pr_info("%s", str);
+}
+
+void __printf(2, 3) host1x_debug_output(struct output *o, const char *fmt, ...);
+
+extern unsigned int host1x_debug_trace_cmdbuf;
+
+void host1x_debug_init(struct host1x *host1x);
+void host1x_debug_deinit(struct host1x *host1x);
+void host1x_debug_dump(struct host1x *host1x);
+void host1x_debug_dump_syncpts(struct host1x *host1x);
+
+#endif
diff --git a/drivers/gpu/host1x/dev.c b/drivers/gpu/host1x/dev.c
new file mode 100644
index 00000000000..28e28a23d44
--- /dev/null
+++ b/drivers/gpu/host1x/dev.c
@@ -0,0 +1,246 @@
+/*
+ * Tegra host1x driver
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/module.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/clk.h>
+#include <linux/io.h>
+
+#define CREATE_TRACE_POINTS
+#include <trace/events/host1x.h>
+
+#include "dev.h"
+#include "intr.h"
+#include "channel.h"
+#include "debug.h"
+#include "hw/host1x01.h"
+#include "host1x_client.h"
+
+void host1x_set_drm_data(struct device *dev, void *data)
+{
+	struct host1x *host1x = dev_get_drvdata(dev);
+	host1x->drm_data = data;
+}
+
+void *host1x_get_drm_data(struct device *dev)
+{
+	struct host1x *host1x = dev_get_drvdata(dev);
+	return host1x->drm_data;
+}
+
+void host1x_sync_writel(struct host1x *host1x, u32 v, u32 r)
+{
+	void __iomem *sync_regs = host1x->regs + host1x->info->sync_offset;
+
+	writel(v, sync_regs + r);
+}
+
+u32 host1x_sync_readl(struct host1x *host1x, u32 r)
+{
+	void __iomem *sync_regs = host1x->regs + host1x->info->sync_offset;
+
+	return readl(sync_regs + r);
+}
+
+void host1x_ch_writel(struct host1x_channel *ch, u32 v, u32 r)
+{
+	writel(v, ch->regs + r);
+}
+
+u32 host1x_ch_readl(struct host1x_channel *ch, u32 r)
+{
+	return readl(ch->regs + r);
+}
+
+static const struct host1x_info host1x01_info = {
+	.nb_channels	= 8,
+	.nb_pts		= 32,
+	.nb_mlocks	= 16,
+	.nb_bases	= 8,
+	.init		= host1x01_init,
+	.sync_offset	= 0x3000,
+};
+
+static struct of_device_id host1x_of_match[] = {
+	{ .compatible = "nvidia,tegra30-host1x", .data = &host1x01_info, },
+	{ .compatible = "nvidia,tegra20-host1x", .data = &host1x01_info, },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, host1x_of_match);
+
+static int host1x_probe(struct platform_device *pdev)
+{
+	const struct of_device_id *id;
+	struct host1x *host;
+	struct resource *regs;
+	int syncpt_irq;
+	int err;
+
+	id = of_match_device(host1x_of_match, &pdev->dev);
+	if (!id)
+		return -EINVAL;
+
+	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!regs) {
+		dev_err(&pdev->dev, "failed to get registers\n");
+		return -ENXIO;
+	}
+
+	syncpt_irq = platform_get_irq(pdev, 0);
+	if (syncpt_irq < 0) {
+		dev_err(&pdev->dev, "failed to get IRQ\n");
+		return -ENXIO;
+	}
+
+	host = devm_kzalloc(&pdev->dev, sizeof(*host), GFP_KERNEL);
+	if (!host)
+		return -ENOMEM;
+
+	host->dev = &pdev->dev;
+	host->info = id->data;
+
+	/* set common host1x device data */
+	platform_set_drvdata(pdev, host);
+
+	host->regs = devm_ioremap_resource(&pdev->dev, regs);
+	if (IS_ERR(host->regs))
+		return PTR_ERR(host->regs);
+
+	if (host->info->init) {
+		err = host->info->init(host);
+		if (err)
+			return err;
+	}
+
+	host->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(host->clk)) {
+		dev_err(&pdev->dev, "failed to get clock\n");
+		err = PTR_ERR(host->clk);
+		return err;
+	}
+
+	err = host1x_channel_list_init(host);
+	if (err) {
+		dev_err(&pdev->dev, "failed to initialize channel list\n");
+		return err;
+	}
+
+	err = clk_prepare_enable(host->clk);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to enable clock\n");
+		return err;
+	}
+
+	err = host1x_syncpt_init(host);
+	if (err) {
+		dev_err(&pdev->dev, "failed to initialize syncpts\n");
+		return err;
+	}
+
+	err = host1x_intr_init(host, syncpt_irq);
+	if (err) {
+		dev_err(&pdev->dev, "failed to initialize interrupts\n");
+		goto fail_deinit_syncpt;
+	}
+
+	host1x_debug_init(host);
+
+	host1x_drm_alloc(pdev);
+
+	return 0;
+
+fail_deinit_syncpt:
+	host1x_syncpt_deinit(host);
+	return err;
+}
+
+static int __exit host1x_remove(struct platform_device *pdev)
+{
+	struct host1x *host = platform_get_drvdata(pdev);
+
+	host1x_intr_deinit(host);
+	host1x_syncpt_deinit(host);
+	clk_disable_unprepare(host->clk);
+
+	return 0;
+}
+
+static struct platform_driver tegra_host1x_driver = {
+	.probe = host1x_probe,
+	.remove = __exit_p(host1x_remove),
+	.driver = {
+		.owner = THIS_MODULE,
+		.name = "tegra-host1x",
+		.of_match_table = host1x_of_match,
+	},
+};
+
+static int __init tegra_host1x_init(void)
+{
+	int err;
+
+	err = platform_driver_register(&tegra_host1x_driver);
+	if (err < 0)
+		return err;
+
+#ifdef CONFIG_DRM_TEGRA
+	err = platform_driver_register(&tegra_dc_driver);
+	if (err < 0)
+		goto unregister_host1x;
+
+	err = platform_driver_register(&tegra_hdmi_driver);
+	if (err < 0)
+		goto unregister_dc;
+
+	err = platform_driver_register(&tegra_gr2d_driver);
+	if (err < 0)
+		goto unregister_hdmi;
+#endif
+
+	return 0;
+
+#ifdef CONFIG_DRM_TEGRA
+unregister_hdmi:
+	platform_driver_unregister(&tegra_hdmi_driver);
+unregister_dc:
+	platform_driver_unregister(&tegra_dc_driver);
+unregister_host1x:
+	platform_driver_unregister(&tegra_host1x_driver);
+	return err;
+#endif
+}
+module_init(tegra_host1x_init);
+
+static void __exit tegra_host1x_exit(void)
+{
+#ifdef CONFIG_DRM_TEGRA
+	platform_driver_unregister(&tegra_gr2d_driver);
+	platform_driver_unregister(&tegra_hdmi_driver);
+	platform_driver_unregister(&tegra_dc_driver);
+#endif
+	platform_driver_unregister(&tegra_host1x_driver);
+}
+module_exit(tegra_host1x_exit);
+
+MODULE_AUTHOR("Thierry Reding <thierry.reding@avionic-design.de>");
+MODULE_AUTHOR("Terje Bergstrom <tbergstrom@nvidia.com>");
+MODULE_DESCRIPTION("Host1x driver for Tegra products");
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/host1x/dev.h b/drivers/gpu/host1x/dev.h
new file mode 100644
index 00000000000..a1607d6e135
--- /dev/null
+++ b/drivers/gpu/host1x/dev.h
@@ -0,0 +1,308 @@
+/*
+ * Copyright (c) 2012-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef HOST1X_DEV_H
+#define HOST1X_DEV_H
+
+#include <linux/platform_device.h>
+#include <linux/device.h>
+
+#include "channel.h"
+#include "syncpt.h"
+#include "intr.h"
+#include "cdma.h"
+#include "job.h"
+
+struct host1x_syncpt;
+struct host1x_channel;
+struct host1x_cdma;
+struct host1x_job;
+struct push_buffer;
+struct output;
+struct dentry;
+
+struct host1x_channel_ops {
+	int (*init)(struct host1x_channel *channel, struct host1x *host,
+		    unsigned int id);
+	int (*submit)(struct host1x_job *job);
+};
+
+struct host1x_cdma_ops {
+	void (*start)(struct host1x_cdma *cdma);
+	void (*stop)(struct host1x_cdma *cdma);
+	void (*flush)(struct  host1x_cdma *cdma);
+	int (*timeout_init)(struct host1x_cdma *cdma, u32 syncpt_id);
+	void (*timeout_destroy)(struct host1x_cdma *cdma);
+	void (*freeze)(struct host1x_cdma *cdma);
+	void (*resume)(struct host1x_cdma *cdma, u32 getptr);
+	void (*timeout_cpu_incr)(struct host1x_cdma *cdma, u32 getptr,
+				 u32 syncpt_incrs, u32 syncval, u32 nr_slots);
+};
+
+struct host1x_pushbuffer_ops {
+	void (*init)(struct push_buffer *pb);
+};
+
+struct host1x_debug_ops {
+	void (*debug_init)(struct dentry *de);
+	void (*show_channel_cdma)(struct host1x *host,
+				  struct host1x_channel *ch,
+				  struct output *o);
+	void (*show_channel_fifo)(struct host1x *host,
+				  struct host1x_channel *ch,
+				  struct output *o);
+	void (*show_mlocks)(struct host1x *host, struct output *output);
+
+};
+
+struct host1x_syncpt_ops {
+	void (*restore)(struct host1x_syncpt *syncpt);
+	void (*restore_wait_base)(struct host1x_syncpt *syncpt);
+	void (*load_wait_base)(struct host1x_syncpt *syncpt);
+	u32 (*load)(struct host1x_syncpt *syncpt);
+	void (*cpu_incr)(struct host1x_syncpt *syncpt);
+	int (*patch_wait)(struct host1x_syncpt *syncpt, void *patch_addr);
+};
+
+struct host1x_intr_ops {
+	int (*init_host_sync)(struct host1x *host, u32 cpm,
+		void (*syncpt_thresh_work)(struct work_struct *work));
+	void (*set_syncpt_threshold)(
+		struct host1x *host, u32 id, u32 thresh);
+	void (*enable_syncpt_intr)(struct host1x *host, u32 id);
+	void (*disable_syncpt_intr)(struct host1x *host, u32 id);
+	void (*disable_all_syncpt_intrs)(struct host1x *host);
+	int (*free_syncpt_irq)(struct host1x *host);
+};
+
+struct host1x_info {
+	int	nb_channels;		/* host1x: num channels supported */
+	int	nb_pts;			/* host1x: num syncpoints supported */
+	int	nb_bases;		/* host1x: num syncpoints supported */
+	int	nb_mlocks;		/* host1x: number of mlocks */
+	int	(*init)(struct host1x *); /* initialize per SoC ops */
+	int	sync_offset;
+};
+
+struct host1x {
+	const struct host1x_info *info;
+
+	void __iomem *regs;
+	struct host1x_syncpt *syncpt;
+	struct device *dev;
+	struct clk *clk;
+
+	struct mutex intr_mutex;
+	struct workqueue_struct *intr_wq;
+	int intr_syncpt_irq;
+
+	const struct host1x_syncpt_ops *syncpt_op;
+	const struct host1x_intr_ops *intr_op;
+	const struct host1x_channel_ops *channel_op;
+	const struct host1x_cdma_ops *cdma_op;
+	const struct host1x_pushbuffer_ops *cdma_pb_op;
+	const struct host1x_debug_ops *debug_op;
+
+	struct host1x_syncpt *nop_sp;
+
+	struct mutex chlist_mutex;
+	struct host1x_channel chlist;
+	unsigned long allocated_channels;
+	unsigned int num_allocated_channels;
+
+	struct dentry *debugfs;
+
+	void *drm_data;
+};
+
+void host1x_sync_writel(struct host1x *host1x, u32 r, u32 v);
+u32 host1x_sync_readl(struct host1x *host1x, u32 r);
+void host1x_ch_writel(struct host1x_channel *ch, u32 r, u32 v);
+u32 host1x_ch_readl(struct host1x_channel *ch, u32 r);
+
+static inline void host1x_hw_syncpt_restore(struct host1x *host,
+					    struct host1x_syncpt *sp)
+{
+	host->syncpt_op->restore(sp);
+}
+
+static inline void host1x_hw_syncpt_restore_wait_base(struct host1x *host,
+						      struct host1x_syncpt *sp)
+{
+	host->syncpt_op->restore_wait_base(sp);
+}
+
+static inline void host1x_hw_syncpt_load_wait_base(struct host1x *host,
+						   struct host1x_syncpt *sp)
+{
+	host->syncpt_op->load_wait_base(sp);
+}
+
+static inline u32 host1x_hw_syncpt_load(struct host1x *host,
+					struct host1x_syncpt *sp)
+{
+	return host->syncpt_op->load(sp);
+}
+
+static inline void host1x_hw_syncpt_cpu_incr(struct host1x *host,
+					     struct host1x_syncpt *sp)
+{
+	host->syncpt_op->cpu_incr(sp);
+}
+
+static inline int host1x_hw_syncpt_patch_wait(struct host1x *host,
+					      struct host1x_syncpt *sp,
+					      void *patch_addr)
+{
+	return host->syncpt_op->patch_wait(sp, patch_addr);
+}
+
+static inline int host1x_hw_intr_init_host_sync(struct host1x *host, u32 cpm,
+			void (*syncpt_thresh_work)(struct work_struct *))
+{
+	return host->intr_op->init_host_sync(host, cpm, syncpt_thresh_work);
+}
+
+static inline void host1x_hw_intr_set_syncpt_threshold(struct host1x *host,
+						       u32 id, u32 thresh)
+{
+	host->intr_op->set_syncpt_threshold(host, id, thresh);
+}
+
+static inline void host1x_hw_intr_enable_syncpt_intr(struct host1x *host,
+						     u32 id)
+{
+	host->intr_op->enable_syncpt_intr(host, id);
+}
+
+static inline void host1x_hw_intr_disable_syncpt_intr(struct host1x *host,
+						      u32 id)
+{
+	host->intr_op->disable_syncpt_intr(host, id);
+}
+
+static inline void host1x_hw_intr_disable_all_syncpt_intrs(struct host1x *host)
+{
+	host->intr_op->disable_all_syncpt_intrs(host);
+}
+
+static inline int host1x_hw_intr_free_syncpt_irq(struct host1x *host)
+{
+	return host->intr_op->free_syncpt_irq(host);
+}
+
+static inline int host1x_hw_channel_init(struct host1x *host,
+					 struct host1x_channel *channel,
+					 int chid)
+{
+	return host->channel_op->init(channel, host, chid);
+}
+
+static inline int host1x_hw_channel_submit(struct host1x *host,
+					   struct host1x_job *job)
+{
+	return host->channel_op->submit(job);
+}
+
+static inline void host1x_hw_cdma_start(struct host1x *host,
+					struct host1x_cdma *cdma)
+{
+	host->cdma_op->start(cdma);
+}
+
+static inline void host1x_hw_cdma_stop(struct host1x *host,
+				       struct host1x_cdma *cdma)
+{
+	host->cdma_op->stop(cdma);
+}
+
+static inline void host1x_hw_cdma_flush(struct host1x *host,
+					struct host1x_cdma *cdma)
+{
+	host->cdma_op->flush(cdma);
+}
+
+static inline int host1x_hw_cdma_timeout_init(struct host1x *host,
+					      struct host1x_cdma *cdma,
+					      u32 syncpt_id)
+{
+	return host->cdma_op->timeout_init(cdma, syncpt_id);
+}
+
+static inline void host1x_hw_cdma_timeout_destroy(struct host1x *host,
+						  struct host1x_cdma *cdma)
+{
+	host->cdma_op->timeout_destroy(cdma);
+}
+
+static inline void host1x_hw_cdma_freeze(struct host1x *host,
+					 struct host1x_cdma *cdma)
+{
+	host->cdma_op->freeze(cdma);
+}
+
+static inline void host1x_hw_cdma_resume(struct host1x *host,
+					 struct host1x_cdma *cdma, u32 getptr)
+{
+	host->cdma_op->resume(cdma, getptr);
+}
+
+static inline void host1x_hw_cdma_timeout_cpu_incr(struct host1x *host,
+						   struct host1x_cdma *cdma,
+						   u32 getptr,
+						   u32 syncpt_incrs,
+						   u32 syncval, u32 nr_slots)
+{
+	host->cdma_op->timeout_cpu_incr(cdma, getptr, syncpt_incrs, syncval,
+					nr_slots);
+}
+
+static inline void host1x_hw_pushbuffer_init(struct host1x *host,
+					     struct push_buffer *pb)
+{
+	host->cdma_pb_op->init(pb);
+}
+
+static inline void host1x_hw_debug_init(struct host1x *host, struct dentry *de)
+{
+	if (host->debug_op && host->debug_op->debug_init)
+		host->debug_op->debug_init(de);
+}
+
+static inline void host1x_hw_show_channel_cdma(struct host1x *host,
+					       struct host1x_channel *channel,
+					       struct output *o)
+{
+	host->debug_op->show_channel_cdma(host, channel, o);
+}
+
+static inline void host1x_hw_show_channel_fifo(struct host1x *host,
+					       struct host1x_channel *channel,
+					       struct output *o)
+{
+	host->debug_op->show_channel_fifo(host, channel, o);
+}
+
+static inline void host1x_hw_show_mlocks(struct host1x *host, struct output *o)
+{
+	host->debug_op->show_mlocks(host, o);
+}
+
+extern struct platform_driver tegra_hdmi_driver;
+extern struct platform_driver tegra_dc_driver;
+extern struct platform_driver tegra_gr2d_driver;
+
+#endif
diff --git a/drivers/gpu/host1x/drm/Kconfig b/drivers/gpu/host1x/drm/Kconfig
new file mode 100644
index 00000000000..69853a4de40
--- /dev/null
+++ b/drivers/gpu/host1x/drm/Kconfig
@@ -0,0 +1,29 @@
+config DRM_TEGRA
+	bool "NVIDIA Tegra DRM"
+	depends on DRM
+	select DRM_KMS_HELPER
+	select FB_SYS_FILLRECT
+	select FB_SYS_COPYAREA
+	select FB_SYS_IMAGEBLIT
+	help
+	  Choose this option if you have an NVIDIA Tegra SoC.
+
+	  To compile this driver as a module, choose M here: the module
+	  will be called tegra-drm.
+
+if DRM_TEGRA
+
+config DRM_TEGRA_STAGING
+	bool "Enable HOST1X interface"
+	depends on STAGING
+	help
+	  Say yes if HOST1X should be available for userspace DRM users.
+
+	  If unsure, choose N.
+
+config DRM_TEGRA_DEBUG
+	bool "NVIDIA Tegra DRM debug support"
+	help
+	  Say yes here to enable debugging support.
+
+endif
diff --git a/drivers/gpu/host1x/drm/dc.c b/drivers/gpu/host1x/drm/dc.c
new file mode 100644
index 00000000000..1e2060324f0
--- /dev/null
+++ b/drivers/gpu/host1x/drm/dc.c
@@ -0,0 +1,1200 @@
+/*
+ * Copyright (C) 2012 Avionic Design GmbH
+ * Copyright (C) 2012 NVIDIA CORPORATION.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/clk.h>
+#include <linux/debugfs.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/clk/tegra.h>
+
+#include "host1x_client.h"
+#include "dc.h"
+#include "drm.h"
+#include "gem.h"
+
+struct tegra_plane {
+	struct drm_plane base;
+	unsigned int index;
+};
+
+static inline struct tegra_plane *to_tegra_plane(struct drm_plane *plane)
+{
+	return container_of(plane, struct tegra_plane, base);
+}
+
+static int tegra_plane_update(struct drm_plane *plane, struct drm_crtc *crtc,
+			      struct drm_framebuffer *fb, int crtc_x,
+			      int crtc_y, unsigned int crtc_w,
+			      unsigned int crtc_h, uint32_t src_x,
+			      uint32_t src_y, uint32_t src_w, uint32_t src_h)
+{
+	struct tegra_plane *p = to_tegra_plane(plane);
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+	struct tegra_dc_window window;
+	unsigned int i;
+
+	memset(&window, 0, sizeof(window));
+	window.src.x = src_x >> 16;
+	window.src.y = src_y >> 16;
+	window.src.w = src_w >> 16;
+	window.src.h = src_h >> 16;
+	window.dst.x = crtc_x;
+	window.dst.y = crtc_y;
+	window.dst.w = crtc_w;
+	window.dst.h = crtc_h;
+	window.format = tegra_dc_format(fb->pixel_format);
+	window.bits_per_pixel = fb->bits_per_pixel;
+
+	for (i = 0; i < drm_format_num_planes(fb->pixel_format); i++) {
+		struct tegra_bo *bo = tegra_fb_get_plane(fb, i);
+
+		window.base[i] = bo->paddr + fb->offsets[i];
+
+		/*
+		 * Tegra doesn't support different strides for U and V planes
+		 * so we display a warning if the user tries to display a
+		 * framebuffer with such a configuration.
+		 */
+		if (i >= 2) {
+			if (fb->pitches[i] != window.stride[1])
+				DRM_ERROR("unsupported UV-plane configuration\n");
+		} else {
+			window.stride[i] = fb->pitches[i];
+		}
+	}
+
+	return tegra_dc_setup_window(dc, p->index, &window);
+}
+
+static int tegra_plane_disable(struct drm_plane *plane)
+{
+	struct tegra_dc *dc = to_tegra_dc(plane->crtc);
+	struct tegra_plane *p = to_tegra_plane(plane);
+	unsigned long value;
+
+	value = WINDOW_A_SELECT << p->index;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_WINDOW_HEADER);
+
+	value = tegra_dc_readl(dc, DC_WIN_WIN_OPTIONS);
+	value &= ~WIN_ENABLE;
+	tegra_dc_writel(dc, value, DC_WIN_WIN_OPTIONS);
+
+	tegra_dc_writel(dc, WIN_A_UPDATE << p->index, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, WIN_A_ACT_REQ << p->index, DC_CMD_STATE_CONTROL);
+
+	return 0;
+}
+
+static void tegra_plane_destroy(struct drm_plane *plane)
+{
+	tegra_plane_disable(plane);
+	drm_plane_cleanup(plane);
+}
+
+static const struct drm_plane_funcs tegra_plane_funcs = {
+	.update_plane = tegra_plane_update,
+	.disable_plane = tegra_plane_disable,
+	.destroy = tegra_plane_destroy,
+};
+
+static const uint32_t plane_formats[] = {
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_UYVY,
+	DRM_FORMAT_YUV420,
+	DRM_FORMAT_YUV422,
+};
+
+static int tegra_dc_add_planes(struct drm_device *drm, struct tegra_dc *dc)
+{
+	unsigned int i;
+	int err = 0;
+
+	for (i = 0; i < 2; i++) {
+		struct tegra_plane *plane;
+
+		plane = devm_kzalloc(drm->dev, sizeof(*plane), GFP_KERNEL);
+		if (!plane)
+			return -ENOMEM;
+
+		plane->index = 1 + i;
+
+		err = drm_plane_init(drm, &plane->base, 1 << dc->pipe,
+				     &tegra_plane_funcs, plane_formats,
+				     ARRAY_SIZE(plane_formats), false);
+		if (err < 0)
+			return err;
+	}
+
+	return 0;
+}
+
+static int tegra_dc_set_base(struct tegra_dc *dc, int x, int y,
+			     struct drm_framebuffer *fb)
+{
+	struct tegra_bo *bo = tegra_fb_get_plane(fb, 0);
+	unsigned long value;
+
+	tegra_dc_writel(dc, WINDOW_A_SELECT, DC_CMD_DISPLAY_WINDOW_HEADER);
+
+	value = fb->offsets[0] + y * fb->pitches[0] +
+		x * fb->bits_per_pixel / 8;
+
+	tegra_dc_writel(dc, bo->paddr + value, DC_WINBUF_START_ADDR);
+	tegra_dc_writel(dc, fb->pitches[0], DC_WIN_LINE_STRIDE);
+
+	value = GENERAL_UPDATE | WIN_A_UPDATE;
+	tegra_dc_writel(dc, value, DC_CMD_STATE_CONTROL);
+
+	value = GENERAL_ACT_REQ | WIN_A_ACT_REQ;
+	tegra_dc_writel(dc, value, DC_CMD_STATE_CONTROL);
+
+	return 0;
+}
+
+void tegra_dc_enable_vblank(struct tegra_dc *dc)
+{
+	unsigned long value, flags;
+
+	spin_lock_irqsave(&dc->lock, flags);
+
+	value = tegra_dc_readl(dc, DC_CMD_INT_MASK);
+	value |= VBLANK_INT;
+	tegra_dc_writel(dc, value, DC_CMD_INT_MASK);
+
+	spin_unlock_irqrestore(&dc->lock, flags);
+}
+
+void tegra_dc_disable_vblank(struct tegra_dc *dc)
+{
+	unsigned long value, flags;
+
+	spin_lock_irqsave(&dc->lock, flags);
+
+	value = tegra_dc_readl(dc, DC_CMD_INT_MASK);
+	value &= ~VBLANK_INT;
+	tegra_dc_writel(dc, value, DC_CMD_INT_MASK);
+
+	spin_unlock_irqrestore(&dc->lock, flags);
+}
+
+static void tegra_dc_finish_page_flip(struct tegra_dc *dc)
+{
+	struct drm_device *drm = dc->base.dev;
+	struct drm_crtc *crtc = &dc->base;
+	unsigned long flags, base;
+	struct tegra_bo *bo;
+
+	if (!dc->event)
+		return;
+
+	bo = tegra_fb_get_plane(crtc->fb, 0);
+
+	/* check if new start address has been latched */
+	tegra_dc_writel(dc, READ_MUX, DC_CMD_STATE_ACCESS);
+	base = tegra_dc_readl(dc, DC_WINBUF_START_ADDR);
+	tegra_dc_writel(dc, 0, DC_CMD_STATE_ACCESS);
+
+	if (base == bo->paddr + crtc->fb->offsets[0]) {
+		spin_lock_irqsave(&drm->event_lock, flags);
+		drm_send_vblank_event(drm, dc->pipe, dc->event);
+		drm_vblank_put(drm, dc->pipe);
+		dc->event = NULL;
+		spin_unlock_irqrestore(&drm->event_lock, flags);
+	}
+}
+
+void tegra_dc_cancel_page_flip(struct drm_crtc *crtc, struct drm_file *file)
+{
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+	struct drm_device *drm = crtc->dev;
+	unsigned long flags;
+
+	spin_lock_irqsave(&drm->event_lock, flags);
+
+	if (dc->event && dc->event->base.file_priv == file) {
+		dc->event->base.destroy(&dc->event->base);
+		drm_vblank_put(drm, dc->pipe);
+		dc->event = NULL;
+	}
+
+	spin_unlock_irqrestore(&drm->event_lock, flags);
+}
+
+static int tegra_dc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
+			      struct drm_pending_vblank_event *event)
+{
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+	struct drm_device *drm = crtc->dev;
+
+	if (dc->event)
+		return -EBUSY;
+
+	if (event) {
+		event->pipe = dc->pipe;
+		dc->event = event;
+		drm_vblank_get(drm, dc->pipe);
+	}
+
+	tegra_dc_set_base(dc, 0, 0, fb);
+	crtc->fb = fb;
+
+	return 0;
+}
+
+static const struct drm_crtc_funcs tegra_crtc_funcs = {
+	.page_flip = tegra_dc_page_flip,
+	.set_config = drm_crtc_helper_set_config,
+	.destroy = drm_crtc_cleanup,
+};
+
+static void tegra_crtc_disable(struct drm_crtc *crtc)
+{
+	struct drm_device *drm = crtc->dev;
+	struct drm_plane *plane;
+
+	list_for_each_entry(plane, &drm->mode_config.plane_list, head) {
+		if (plane->crtc == crtc) {
+			tegra_plane_disable(plane);
+			plane->crtc = NULL;
+
+			if (plane->fb) {
+				drm_framebuffer_unreference(plane->fb);
+				plane->fb = NULL;
+			}
+		}
+	}
+}
+
+static bool tegra_crtc_mode_fixup(struct drm_crtc *crtc,
+				  const struct drm_display_mode *mode,
+				  struct drm_display_mode *adjusted)
+{
+	return true;
+}
+
+static inline u32 compute_dda_inc(unsigned int in, unsigned int out, bool v,
+				  unsigned int bpp)
+{
+	fixed20_12 outf = dfixed_init(out);
+	fixed20_12 inf = dfixed_init(in);
+	u32 dda_inc;
+	int max;
+
+	if (v)
+		max = 15;
+	else {
+		switch (bpp) {
+		case 2:
+			max = 8;
+			break;
+
+		default:
+			WARN_ON_ONCE(1);
+			/* fallthrough */
+		case 4:
+			max = 4;
+			break;
+		}
+	}
+
+	outf.full = max_t(u32, outf.full - dfixed_const(1), dfixed_const(1));
+	inf.full -= dfixed_const(1);
+
+	dda_inc = dfixed_div(inf, outf);
+	dda_inc = min_t(u32, dda_inc, dfixed_const(max));
+
+	return dda_inc;
+}
+
+static inline u32 compute_initial_dda(unsigned int in)
+{
+	fixed20_12 inf = dfixed_init(in);
+	return dfixed_frac(inf);
+}
+
+static int tegra_dc_set_timings(struct tegra_dc *dc,
+				struct drm_display_mode *mode)
+{
+	/* TODO: For HDMI compliance, h & v ref_to_sync should be set to 1 */
+	unsigned int h_ref_to_sync = 0;
+	unsigned int v_ref_to_sync = 0;
+	unsigned long value;
+
+	tegra_dc_writel(dc, 0x0, DC_DISP_DISP_TIMING_OPTIONS);
+
+	value = (v_ref_to_sync << 16) | h_ref_to_sync;
+	tegra_dc_writel(dc, value, DC_DISP_REF_TO_SYNC);
+
+	value = ((mode->vsync_end - mode->vsync_start) << 16) |
+		((mode->hsync_end - mode->hsync_start) <<  0);
+	tegra_dc_writel(dc, value, DC_DISP_SYNC_WIDTH);
+
+	value = ((mode->vtotal - mode->vsync_end) << 16) |
+		((mode->htotal - mode->hsync_end) <<  0);
+	tegra_dc_writel(dc, value, DC_DISP_BACK_PORCH);
+
+	value = ((mode->vsync_start - mode->vdisplay) << 16) |
+		((mode->hsync_start - mode->hdisplay) <<  0);
+	tegra_dc_writel(dc, value, DC_DISP_FRONT_PORCH);
+
+	value = (mode->vdisplay << 16) | mode->hdisplay;
+	tegra_dc_writel(dc, value, DC_DISP_ACTIVE);
+
+	return 0;
+}
+
+static int tegra_crtc_setup_clk(struct drm_crtc *crtc,
+				struct drm_display_mode *mode,
+				unsigned long *div)
+{
+	unsigned long pclk = mode->clock * 1000, rate;
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+	struct tegra_output *output = NULL;
+	struct drm_encoder *encoder;
+	long err;
+
+	list_for_each_entry(encoder, &crtc->dev->mode_config.encoder_list, head)
+		if (encoder->crtc == crtc) {
+			output = encoder_to_output(encoder);
+			break;
+		}
+
+	if (!output)
+		return -ENODEV;
+
+	/*
+	 * This assumes that the display controller will divide its parent
+	 * clock by 2 to generate the pixel clock.
+	 */
+	err = tegra_output_setup_clock(output, dc->clk, pclk * 2);
+	if (err < 0) {
+		dev_err(dc->dev, "failed to setup clock: %ld\n", err);
+		return err;
+	}
+
+	rate = clk_get_rate(dc->clk);
+	*div = (rate * 2 / pclk) - 2;
+
+	DRM_DEBUG_KMS("rate: %lu, div: %lu\n", rate, *div);
+
+	return 0;
+}
+
+static bool tegra_dc_format_is_yuv(unsigned int format, bool *planar)
+{
+	switch (format) {
+	case WIN_COLOR_DEPTH_YCbCr422:
+	case WIN_COLOR_DEPTH_YUV422:
+		if (planar)
+			*planar = false;
+
+		return true;
+
+	case WIN_COLOR_DEPTH_YCbCr420P:
+	case WIN_COLOR_DEPTH_YUV420P:
+	case WIN_COLOR_DEPTH_YCbCr422P:
+	case WIN_COLOR_DEPTH_YUV422P:
+	case WIN_COLOR_DEPTH_YCbCr422R:
+	case WIN_COLOR_DEPTH_YUV422R:
+	case WIN_COLOR_DEPTH_YCbCr422RA:
+	case WIN_COLOR_DEPTH_YUV422RA:
+		if (planar)
+			*planar = true;
+
+		return true;
+	}
+
+	return false;
+}
+
+int tegra_dc_setup_window(struct tegra_dc *dc, unsigned int index,
+			  const struct tegra_dc_window *window)
+{
+	unsigned h_offset, v_offset, h_size, v_size, h_dda, v_dda, bpp;
+	unsigned long value;
+	bool yuv, planar;
+
+	/*
+	 * For YUV planar modes, the number of bytes per pixel takes into
+	 * account only the luma component and therefore is 1.
+	 */
+	yuv = tegra_dc_format_is_yuv(window->format, &planar);
+	if (!yuv)
+		bpp = window->bits_per_pixel / 8;
+	else
+		bpp = planar ? 1 : 2;
+
+	value = WINDOW_A_SELECT << index;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_WINDOW_HEADER);
+
+	tegra_dc_writel(dc, window->format, DC_WIN_COLOR_DEPTH);
+	tegra_dc_writel(dc, 0, DC_WIN_BYTE_SWAP);
+
+	value = V_POSITION(window->dst.y) | H_POSITION(window->dst.x);
+	tegra_dc_writel(dc, value, DC_WIN_POSITION);
+
+	value = V_SIZE(window->dst.h) | H_SIZE(window->dst.w);
+	tegra_dc_writel(dc, value, DC_WIN_SIZE);
+
+	h_offset = window->src.x * bpp;
+	v_offset = window->src.y;
+	h_size = window->src.w * bpp;
+	v_size = window->src.h;
+
+	value = V_PRESCALED_SIZE(v_size) | H_PRESCALED_SIZE(h_size);
+	tegra_dc_writel(dc, value, DC_WIN_PRESCALED_SIZE);
+
+	/*
+	 * For DDA computations the number of bytes per pixel for YUV planar
+	 * modes needs to take into account all Y, U and V components.
+	 */
+	if (yuv && planar)
+		bpp = 2;
+
+	h_dda = compute_dda_inc(window->src.w, window->dst.w, false, bpp);
+	v_dda = compute_dda_inc(window->src.h, window->dst.h, true, bpp);
+
+	value = V_DDA_INC(v_dda) | H_DDA_INC(h_dda);
+	tegra_dc_writel(dc, value, DC_WIN_DDA_INC);
+
+	h_dda = compute_initial_dda(window->src.x);
+	v_dda = compute_initial_dda(window->src.y);
+
+	tegra_dc_writel(dc, h_dda, DC_WIN_H_INITIAL_DDA);
+	tegra_dc_writel(dc, v_dda, DC_WIN_V_INITIAL_DDA);
+
+	tegra_dc_writel(dc, 0, DC_WIN_UV_BUF_STRIDE);
+	tegra_dc_writel(dc, 0, DC_WIN_BUF_STRIDE);
+
+	tegra_dc_writel(dc, window->base[0], DC_WINBUF_START_ADDR);
+
+	if (yuv && planar) {
+		tegra_dc_writel(dc, window->base[1], DC_WINBUF_START_ADDR_U);
+		tegra_dc_writel(dc, window->base[2], DC_WINBUF_START_ADDR_V);
+		value = window->stride[1] << 16 | window->stride[0];
+		tegra_dc_writel(dc, value, DC_WIN_LINE_STRIDE);
+	} else {
+		tegra_dc_writel(dc, window->stride[0], DC_WIN_LINE_STRIDE);
+	}
+
+	tegra_dc_writel(dc, h_offset, DC_WINBUF_ADDR_H_OFFSET);
+	tegra_dc_writel(dc, v_offset, DC_WINBUF_ADDR_V_OFFSET);
+
+	value = WIN_ENABLE;
+
+	if (yuv) {
+		/* setup default colorspace conversion coefficients */
+		tegra_dc_writel(dc, 0x00f0, DC_WIN_CSC_YOF);
+		tegra_dc_writel(dc, 0x012a, DC_WIN_CSC_KYRGB);
+		tegra_dc_writel(dc, 0x0000, DC_WIN_CSC_KUR);
+		tegra_dc_writel(dc, 0x0198, DC_WIN_CSC_KVR);
+		tegra_dc_writel(dc, 0x039b, DC_WIN_CSC_KUG);
+		tegra_dc_writel(dc, 0x032f, DC_WIN_CSC_KVG);
+		tegra_dc_writel(dc, 0x0204, DC_WIN_CSC_KUB);
+		tegra_dc_writel(dc, 0x0000, DC_WIN_CSC_KVB);
+
+		value |= CSC_ENABLE;
+	} else if (window->bits_per_pixel < 24) {
+		value |= COLOR_EXPAND;
+	}
+
+	tegra_dc_writel(dc, value, DC_WIN_WIN_OPTIONS);
+
+	/*
+	 * Disable blending and assume Window A is the bottom-most window,
+	 * Window C is the top-most window and Window B is in the middle.
+	 */
+	tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_NOKEY);
+	tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_1WIN);
+
+	switch (index) {
+	case 0:
+		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_2WIN_X);
+		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_2WIN_Y);
+		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_3WIN_XY);
+		break;
+
+	case 1:
+		tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_2WIN_X);
+		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_2WIN_Y);
+		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_3WIN_XY);
+		break;
+
+	case 2:
+		tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_2WIN_X);
+		tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_2WIN_Y);
+		tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_3WIN_XY);
+		break;
+	}
+
+	tegra_dc_writel(dc, WIN_A_UPDATE << index, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, WIN_A_ACT_REQ << index, DC_CMD_STATE_CONTROL);
+
+	return 0;
+}
+
+unsigned int tegra_dc_format(uint32_t format)
+{
+	switch (format) {
+	case DRM_FORMAT_XBGR8888:
+		return WIN_COLOR_DEPTH_R8G8B8A8;
+
+	case DRM_FORMAT_XRGB8888:
+		return WIN_COLOR_DEPTH_B8G8R8A8;
+
+	case DRM_FORMAT_RGB565:
+		return WIN_COLOR_DEPTH_B5G6R5;
+
+	case DRM_FORMAT_UYVY:
+		return WIN_COLOR_DEPTH_YCbCr422;
+
+	case DRM_FORMAT_YUV420:
+		return WIN_COLOR_DEPTH_YCbCr420P;
+
+	case DRM_FORMAT_YUV422:
+		return WIN_COLOR_DEPTH_YCbCr422P;
+
+	default:
+		break;
+	}
+
+	WARN(1, "unsupported pixel format %u, using default\n", format);
+	return WIN_COLOR_DEPTH_B8G8R8A8;
+}
+
+static int tegra_crtc_mode_set(struct drm_crtc *crtc,
+			       struct drm_display_mode *mode,
+			       struct drm_display_mode *adjusted,
+			       int x, int y, struct drm_framebuffer *old_fb)
+{
+	struct tegra_bo *bo = tegra_fb_get_plane(crtc->fb, 0);
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+	struct tegra_dc_window window;
+	unsigned long div, value;
+	int err;
+
+	drm_vblank_pre_modeset(crtc->dev, dc->pipe);
+
+	err = tegra_crtc_setup_clk(crtc, mode, &div);
+	if (err) {
+		dev_err(dc->dev, "failed to setup clock for CRTC: %d\n", err);
+		return err;
+	}
+
+	/* program display mode */
+	tegra_dc_set_timings(dc, mode);
+
+	value = DE_SELECT_ACTIVE | DE_CONTROL_NORMAL;
+	tegra_dc_writel(dc, value, DC_DISP_DATA_ENABLE_OPTIONS);
+
+	value = tegra_dc_readl(dc, DC_COM_PIN_OUTPUT_POLARITY(1));
+	value &= ~LVS_OUTPUT_POLARITY_LOW;
+	value &= ~LHS_OUTPUT_POLARITY_LOW;
+	tegra_dc_writel(dc, value, DC_COM_PIN_OUTPUT_POLARITY(1));
+
+	value = DISP_DATA_FORMAT_DF1P1C | DISP_ALIGNMENT_MSB |
+		DISP_ORDER_RED_BLUE;
+	tegra_dc_writel(dc, value, DC_DISP_DISP_INTERFACE_CONTROL);
+
+	tegra_dc_writel(dc, 0x00010001, DC_DISP_SHIFT_CLOCK_OPTIONS);
+
+	value = SHIFT_CLK_DIVIDER(div) | PIXEL_CLK_DIVIDER_PCD1;
+	tegra_dc_writel(dc, value, DC_DISP_DISP_CLOCK_CONTROL);
+
+	/* setup window parameters */
+	memset(&window, 0, sizeof(window));
+	window.src.x = 0;
+	window.src.y = 0;
+	window.src.w = mode->hdisplay;
+	window.src.h = mode->vdisplay;
+	window.dst.x = 0;
+	window.dst.y = 0;
+	window.dst.w = mode->hdisplay;
+	window.dst.h = mode->vdisplay;
+	window.format = tegra_dc_format(crtc->fb->pixel_format);
+	window.bits_per_pixel = crtc->fb->bits_per_pixel;
+	window.stride[0] = crtc->fb->pitches[0];
+	window.base[0] = bo->paddr;
+
+	err = tegra_dc_setup_window(dc, 0, &window);
+	if (err < 0)
+		dev_err(dc->dev, "failed to enable root plane\n");
+
+	return 0;
+}
+
+static int tegra_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
+				    struct drm_framebuffer *old_fb)
+{
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+
+	return tegra_dc_set_base(dc, x, y, crtc->fb);
+}
+
+static void tegra_crtc_prepare(struct drm_crtc *crtc)
+{
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+	unsigned int syncpt;
+	unsigned long value;
+
+	/* hardware initialization */
+	tegra_periph_reset_deassert(dc->clk);
+	usleep_range(10000, 20000);
+
+	if (dc->pipe)
+		syncpt = SYNCPT_VBLANK1;
+	else
+		syncpt = SYNCPT_VBLANK0;
+
+	/* initialize display controller */
+	tegra_dc_writel(dc, 0x00000100, DC_CMD_GENERAL_INCR_SYNCPT_CNTRL);
+	tegra_dc_writel(dc, 0x100 | syncpt, DC_CMD_CONT_SYNCPT_VSYNC);
+
+	value = WIN_A_UF_INT | WIN_B_UF_INT | WIN_C_UF_INT | WIN_A_OF_INT;
+	tegra_dc_writel(dc, value, DC_CMD_INT_TYPE);
+
+	value = WIN_A_UF_INT | WIN_B_UF_INT | WIN_C_UF_INT |
+		WIN_A_OF_INT | WIN_B_OF_INT | WIN_C_OF_INT;
+	tegra_dc_writel(dc, value, DC_CMD_INT_POLARITY);
+
+	value = PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+		PW4_ENABLE | PM0_ENABLE | PM1_ENABLE;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
+	value = tegra_dc_readl(dc, DC_CMD_DISPLAY_COMMAND);
+	value |= DISP_CTRL_MODE_C_DISPLAY;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_COMMAND);
+
+	/* initialize timer */
+	value = CURSOR_THRESHOLD(0) | WINDOW_A_THRESHOLD(0x20) |
+		WINDOW_B_THRESHOLD(0x20) | WINDOW_C_THRESHOLD(0x20);
+	tegra_dc_writel(dc, value, DC_DISP_DISP_MEM_HIGH_PRIORITY);
+
+	value = CURSOR_THRESHOLD(0) | WINDOW_A_THRESHOLD(1) |
+		WINDOW_B_THRESHOLD(1) | WINDOW_C_THRESHOLD(1);
+	tegra_dc_writel(dc, value, DC_DISP_DISP_MEM_HIGH_PRIORITY_TIMER);
+
+	value = VBLANK_INT | WIN_A_UF_INT | WIN_B_UF_INT | WIN_C_UF_INT;
+	tegra_dc_writel(dc, value, DC_CMD_INT_ENABLE);
+
+	value = WIN_A_UF_INT | WIN_B_UF_INT | WIN_C_UF_INT;
+	tegra_dc_writel(dc, value, DC_CMD_INT_MASK);
+}
+
+static void tegra_crtc_commit(struct drm_crtc *crtc)
+{
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+	unsigned long value;
+
+	value = GENERAL_UPDATE | WIN_A_UPDATE;
+	tegra_dc_writel(dc, value, DC_CMD_STATE_CONTROL);
+
+	value = GENERAL_ACT_REQ | WIN_A_ACT_REQ;
+	tegra_dc_writel(dc, value, DC_CMD_STATE_CONTROL);
+
+	drm_vblank_post_modeset(crtc->dev, dc->pipe);
+}
+
+static void tegra_crtc_load_lut(struct drm_crtc *crtc)
+{
+}
+
+static const struct drm_crtc_helper_funcs tegra_crtc_helper_funcs = {
+	.disable = tegra_crtc_disable,
+	.mode_fixup = tegra_crtc_mode_fixup,
+	.mode_set = tegra_crtc_mode_set,
+	.mode_set_base = tegra_crtc_mode_set_base,
+	.prepare = tegra_crtc_prepare,
+	.commit = tegra_crtc_commit,
+	.load_lut = tegra_crtc_load_lut,
+};
+
+static irqreturn_t tegra_dc_irq(int irq, void *data)
+{
+	struct tegra_dc *dc = data;
+	unsigned long status;
+
+	status = tegra_dc_readl(dc, DC_CMD_INT_STATUS);
+	tegra_dc_writel(dc, status, DC_CMD_INT_STATUS);
+
+	if (status & FRAME_END_INT) {
+		/*
+		dev_dbg(dc->dev, "%s(): frame end\n", __func__);
+		*/
+	}
+
+	if (status & VBLANK_INT) {
+		/*
+		dev_dbg(dc->dev, "%s(): vertical blank\n", __func__);
+		*/
+		drm_handle_vblank(dc->base.dev, dc->pipe);
+		tegra_dc_finish_page_flip(dc);
+	}
+
+	if (status & (WIN_A_UF_INT | WIN_B_UF_INT | WIN_C_UF_INT)) {
+		/*
+		dev_dbg(dc->dev, "%s(): underflow\n", __func__);
+		*/
+	}
+
+	return IRQ_HANDLED;
+}
+
+static int tegra_dc_show_regs(struct seq_file *s, void *data)
+{
+	struct drm_info_node *node = s->private;
+	struct tegra_dc *dc = node->info_ent->data;
+
+#define DUMP_REG(name)						\
+	seq_printf(s, "%-40s %#05x %08lx\n", #name, name,	\
+		   tegra_dc_readl(dc, name))
+
+	DUMP_REG(DC_CMD_GENERAL_INCR_SYNCPT);
+	DUMP_REG(DC_CMD_GENERAL_INCR_SYNCPT_CNTRL);
+	DUMP_REG(DC_CMD_GENERAL_INCR_SYNCPT_ERROR);
+	DUMP_REG(DC_CMD_WIN_A_INCR_SYNCPT);
+	DUMP_REG(DC_CMD_WIN_A_INCR_SYNCPT_CNTRL);
+	DUMP_REG(DC_CMD_WIN_A_INCR_SYNCPT_ERROR);
+	DUMP_REG(DC_CMD_WIN_B_INCR_SYNCPT);
+	DUMP_REG(DC_CMD_WIN_B_INCR_SYNCPT_CNTRL);
+	DUMP_REG(DC_CMD_WIN_B_INCR_SYNCPT_ERROR);
+	DUMP_REG(DC_CMD_WIN_C_INCR_SYNCPT);
+	DUMP_REG(DC_CMD_WIN_C_INCR_SYNCPT_CNTRL);
+	DUMP_REG(DC_CMD_WIN_C_INCR_SYNCPT_ERROR);
+	DUMP_REG(DC_CMD_CONT_SYNCPT_VSYNC);
+	DUMP_REG(DC_CMD_DISPLAY_COMMAND_OPTION0);
+	DUMP_REG(DC_CMD_DISPLAY_COMMAND);
+	DUMP_REG(DC_CMD_SIGNAL_RAISE);
+	DUMP_REG(DC_CMD_DISPLAY_POWER_CONTROL);
+	DUMP_REG(DC_CMD_INT_STATUS);
+	DUMP_REG(DC_CMD_INT_MASK);
+	DUMP_REG(DC_CMD_INT_ENABLE);
+	DUMP_REG(DC_CMD_INT_TYPE);
+	DUMP_REG(DC_CMD_INT_POLARITY);
+	DUMP_REG(DC_CMD_SIGNAL_RAISE1);
+	DUMP_REG(DC_CMD_SIGNAL_RAISE2);
+	DUMP_REG(DC_CMD_SIGNAL_RAISE3);
+	DUMP_REG(DC_CMD_STATE_ACCESS);
+	DUMP_REG(DC_CMD_STATE_CONTROL);
+	DUMP_REG(DC_CMD_DISPLAY_WINDOW_HEADER);
+	DUMP_REG(DC_CMD_REG_ACT_CONTROL);
+	DUMP_REG(DC_COM_CRC_CONTROL);
+	DUMP_REG(DC_COM_CRC_CHECKSUM);
+	DUMP_REG(DC_COM_PIN_OUTPUT_ENABLE(0));
+	DUMP_REG(DC_COM_PIN_OUTPUT_ENABLE(1));
+	DUMP_REG(DC_COM_PIN_OUTPUT_ENABLE(2));
+	DUMP_REG(DC_COM_PIN_OUTPUT_ENABLE(3));
+	DUMP_REG(DC_COM_PIN_OUTPUT_POLARITY(0));
+	DUMP_REG(DC_COM_PIN_OUTPUT_POLARITY(1));
+	DUMP_REG(DC_COM_PIN_OUTPUT_POLARITY(2));
+	DUMP_REG(DC_COM_PIN_OUTPUT_POLARITY(3));
+	DUMP_REG(DC_COM_PIN_OUTPUT_DATA(0));
+	DUMP_REG(DC_COM_PIN_OUTPUT_DATA(1));
+	DUMP_REG(DC_COM_PIN_OUTPUT_DATA(2));
+	DUMP_REG(DC_COM_PIN_OUTPUT_DATA(3));
+	DUMP_REG(DC_COM_PIN_INPUT_ENABLE(0));
+	DUMP_REG(DC_COM_PIN_INPUT_ENABLE(1));
+	DUMP_REG(DC_COM_PIN_INPUT_ENABLE(2));
+	DUMP_REG(DC_COM_PIN_INPUT_ENABLE(3));
+	DUMP_REG(DC_COM_PIN_INPUT_DATA(0));
+	DUMP_REG(DC_COM_PIN_INPUT_DATA(1));
+	DUMP_REG(DC_COM_PIN_OUTPUT_SELECT(0));
+	DUMP_REG(DC_COM_PIN_OUTPUT_SELECT(1));
+	DUMP_REG(DC_COM_PIN_OUTPUT_SELECT(2));
+	DUMP_REG(DC_COM_PIN_OUTPUT_SELECT(3));
+	DUMP_REG(DC_COM_PIN_OUTPUT_SELECT(4));
+	DUMP_REG(DC_COM_PIN_OUTPUT_SELECT(5));
+	DUMP_REG(DC_COM_PIN_OUTPUT_SELECT(6));
+	DUMP_REG(DC_COM_PIN_MISC_CONTROL);
+	DUMP_REG(DC_COM_PIN_PM0_CONTROL);
+	DUMP_REG(DC_COM_PIN_PM0_DUTY_CYCLE);
+	DUMP_REG(DC_COM_PIN_PM1_CONTROL);
+	DUMP_REG(DC_COM_PIN_PM1_DUTY_CYCLE);
+	DUMP_REG(DC_COM_SPI_CONTROL);
+	DUMP_REG(DC_COM_SPI_START_BYTE);
+	DUMP_REG(DC_COM_HSPI_WRITE_DATA_AB);
+	DUMP_REG(DC_COM_HSPI_WRITE_DATA_CD);
+	DUMP_REG(DC_COM_HSPI_CS_DC);
+	DUMP_REG(DC_COM_SCRATCH_REGISTER_A);
+	DUMP_REG(DC_COM_SCRATCH_REGISTER_B);
+	DUMP_REG(DC_COM_GPIO_CTRL);
+	DUMP_REG(DC_COM_GPIO_DEBOUNCE_COUNTER);
+	DUMP_REG(DC_COM_CRC_CHECKSUM_LATCHED);
+	DUMP_REG(DC_DISP_DISP_SIGNAL_OPTIONS0);
+	DUMP_REG(DC_DISP_DISP_SIGNAL_OPTIONS1);
+	DUMP_REG(DC_DISP_DISP_WIN_OPTIONS);
+	DUMP_REG(DC_DISP_DISP_MEM_HIGH_PRIORITY);
+	DUMP_REG(DC_DISP_DISP_MEM_HIGH_PRIORITY_TIMER);
+	DUMP_REG(DC_DISP_DISP_TIMING_OPTIONS);
+	DUMP_REG(DC_DISP_REF_TO_SYNC);
+	DUMP_REG(DC_DISP_SYNC_WIDTH);
+	DUMP_REG(DC_DISP_BACK_PORCH);
+	DUMP_REG(DC_DISP_ACTIVE);
+	DUMP_REG(DC_DISP_FRONT_PORCH);
+	DUMP_REG(DC_DISP_H_PULSE0_CONTROL);
+	DUMP_REG(DC_DISP_H_PULSE0_POSITION_A);
+	DUMP_REG(DC_DISP_H_PULSE0_POSITION_B);
+	DUMP_REG(DC_DISP_H_PULSE0_POSITION_C);
+	DUMP_REG(DC_DISP_H_PULSE0_POSITION_D);
+	DUMP_REG(DC_DISP_H_PULSE1_CONTROL);
+	DUMP_REG(DC_DISP_H_PULSE1_POSITION_A);
+	DUMP_REG(DC_DISP_H_PULSE1_POSITION_B);
+	DUMP_REG(DC_DISP_H_PULSE1_POSITION_C);
+	DUMP_REG(DC_DISP_H_PULSE1_POSITION_D);
+	DUMP_REG(DC_DISP_H_PULSE2_CONTROL);
+	DUMP_REG(DC_DISP_H_PULSE2_POSITION_A);
+	DUMP_REG(DC_DISP_H_PULSE2_POSITION_B);
+	DUMP_REG(DC_DISP_H_PULSE2_POSITION_C);
+	DUMP_REG(DC_DISP_H_PULSE2_POSITION_D);
+	DUMP_REG(DC_DISP_V_PULSE0_CONTROL);
+	DUMP_REG(DC_DISP_V_PULSE0_POSITION_A);
+	DUMP_REG(DC_DISP_V_PULSE0_POSITION_B);
+	DUMP_REG(DC_DISP_V_PULSE0_POSITION_C);
+	DUMP_REG(DC_DISP_V_PULSE1_CONTROL);
+	DUMP_REG(DC_DISP_V_PULSE1_POSITION_A);
+	DUMP_REG(DC_DISP_V_PULSE1_POSITION_B);
+	DUMP_REG(DC_DISP_V_PULSE1_POSITION_C);
+	DUMP_REG(DC_DISP_V_PULSE2_CONTROL);
+	DUMP_REG(DC_DISP_V_PULSE2_POSITION_A);
+	DUMP_REG(DC_DISP_V_PULSE3_CONTROL);
+	DUMP_REG(DC_DISP_V_PULSE3_POSITION_A);
+	DUMP_REG(DC_DISP_M0_CONTROL);
+	DUMP_REG(DC_DISP_M1_CONTROL);
+	DUMP_REG(DC_DISP_DI_CONTROL);
+	DUMP_REG(DC_DISP_PP_CONTROL);
+	DUMP_REG(DC_DISP_PP_SELECT_A);
+	DUMP_REG(DC_DISP_PP_SELECT_B);
+	DUMP_REG(DC_DISP_PP_SELECT_C);
+	DUMP_REG(DC_DISP_PP_SELECT_D);
+	DUMP_REG(DC_DISP_DISP_CLOCK_CONTROL);
+	DUMP_REG(DC_DISP_DISP_INTERFACE_CONTROL);
+	DUMP_REG(DC_DISP_DISP_COLOR_CONTROL);
+	DUMP_REG(DC_DISP_SHIFT_CLOCK_OPTIONS);
+	DUMP_REG(DC_DISP_DATA_ENABLE_OPTIONS);
+	DUMP_REG(DC_DISP_SERIAL_INTERFACE_OPTIONS);
+	DUMP_REG(DC_DISP_LCD_SPI_OPTIONS);
+	DUMP_REG(DC_DISP_BORDER_COLOR);
+	DUMP_REG(DC_DISP_COLOR_KEY0_LOWER);
+	DUMP_REG(DC_DISP_COLOR_KEY0_UPPER);
+	DUMP_REG(DC_DISP_COLOR_KEY1_LOWER);
+	DUMP_REG(DC_DISP_COLOR_KEY1_UPPER);
+	DUMP_REG(DC_DISP_CURSOR_FOREGROUND);
+	DUMP_REG(DC_DISP_CURSOR_BACKGROUND);
+	DUMP_REG(DC_DISP_CURSOR_START_ADDR);
+	DUMP_REG(DC_DISP_CURSOR_START_ADDR_NS);
+	DUMP_REG(DC_DISP_CURSOR_POSITION);
+	DUMP_REG(DC_DISP_CURSOR_POSITION_NS);
+	DUMP_REG(DC_DISP_INIT_SEQ_CONTROL);
+	DUMP_REG(DC_DISP_SPI_INIT_SEQ_DATA_A);
+	DUMP_REG(DC_DISP_SPI_INIT_SEQ_DATA_B);
+	DUMP_REG(DC_DISP_SPI_INIT_SEQ_DATA_C);
+	DUMP_REG(DC_DISP_SPI_INIT_SEQ_DATA_D);
+	DUMP_REG(DC_DISP_DC_MCCIF_FIFOCTRL);
+	DUMP_REG(DC_DISP_MCCIF_DISPLAY0A_HYST);
+	DUMP_REG(DC_DISP_MCCIF_DISPLAY0B_HYST);
+	DUMP_REG(DC_DISP_MCCIF_DISPLAY1A_HYST);
+	DUMP_REG(DC_DISP_MCCIF_DISPLAY1B_HYST);
+	DUMP_REG(DC_DISP_DAC_CRT_CTRL);
+	DUMP_REG(DC_DISP_DISP_MISC_CONTROL);
+	DUMP_REG(DC_DISP_SD_CONTROL);
+	DUMP_REG(DC_DISP_SD_CSC_COEFF);
+	DUMP_REG(DC_DISP_SD_LUT(0));
+	DUMP_REG(DC_DISP_SD_LUT(1));
+	DUMP_REG(DC_DISP_SD_LUT(2));
+	DUMP_REG(DC_DISP_SD_LUT(3));
+	DUMP_REG(DC_DISP_SD_LUT(4));
+	DUMP_REG(DC_DISP_SD_LUT(5));
+	DUMP_REG(DC_DISP_SD_LUT(6));
+	DUMP_REG(DC_DISP_SD_LUT(7));
+	DUMP_REG(DC_DISP_SD_LUT(8));
+	DUMP_REG(DC_DISP_SD_FLICKER_CONTROL);
+	DUMP_REG(DC_DISP_DC_PIXEL_COUNT);
+	DUMP_REG(DC_DISP_SD_HISTOGRAM(0));
+	DUMP_REG(DC_DISP_SD_HISTOGRAM(1));
+	DUMP_REG(DC_DISP_SD_HISTOGRAM(2));
+	DUMP_REG(DC_DISP_SD_HISTOGRAM(3));
+	DUMP_REG(DC_DISP_SD_HISTOGRAM(4));
+	DUMP_REG(DC_DISP_SD_HISTOGRAM(5));
+	DUMP_REG(DC_DISP_SD_HISTOGRAM(6));
+	DUMP_REG(DC_DISP_SD_HISTOGRAM(7));
+	DUMP_REG(DC_DISP_SD_BL_TF(0));
+	DUMP_REG(DC_DISP_SD_BL_TF(1));
+	DUMP_REG(DC_DISP_SD_BL_TF(2));
+	DUMP_REG(DC_DISP_SD_BL_TF(3));
+	DUMP_REG(DC_DISP_SD_BL_CONTROL);
+	DUMP_REG(DC_DISP_SD_HW_K_VALUES);
+	DUMP_REG(DC_DISP_SD_MAN_K_VALUES);
+	DUMP_REG(DC_WIN_WIN_OPTIONS);
+	DUMP_REG(DC_WIN_BYTE_SWAP);
+	DUMP_REG(DC_WIN_BUFFER_CONTROL);
+	DUMP_REG(DC_WIN_COLOR_DEPTH);
+	DUMP_REG(DC_WIN_POSITION);
+	DUMP_REG(DC_WIN_SIZE);
+	DUMP_REG(DC_WIN_PRESCALED_SIZE);
+	DUMP_REG(DC_WIN_H_INITIAL_DDA);
+	DUMP_REG(DC_WIN_V_INITIAL_DDA);
+	DUMP_REG(DC_WIN_DDA_INC);
+	DUMP_REG(DC_WIN_LINE_STRIDE);
+	DUMP_REG(DC_WIN_BUF_STRIDE);
+	DUMP_REG(DC_WIN_UV_BUF_STRIDE);
+	DUMP_REG(DC_WIN_BUFFER_ADDR_MODE);
+	DUMP_REG(DC_WIN_DV_CONTROL);
+	DUMP_REG(DC_WIN_BLEND_NOKEY);
+	DUMP_REG(DC_WIN_BLEND_1WIN);
+	DUMP_REG(DC_WIN_BLEND_2WIN_X);
+	DUMP_REG(DC_WIN_BLEND_2WIN_Y);
+	DUMP_REG(DC_WIN_BLEND_3WIN_XY);
+	DUMP_REG(DC_WIN_HP_FETCH_CONTROL);
+	DUMP_REG(DC_WINBUF_START_ADDR);
+	DUMP_REG(DC_WINBUF_START_ADDR_NS);
+	DUMP_REG(DC_WINBUF_START_ADDR_U);
+	DUMP_REG(DC_WINBUF_START_ADDR_U_NS);
+	DUMP_REG(DC_WINBUF_START_ADDR_V);
+	DUMP_REG(DC_WINBUF_START_ADDR_V_NS);
+	DUMP_REG(DC_WINBUF_ADDR_H_OFFSET);
+	DUMP_REG(DC_WINBUF_ADDR_H_OFFSET_NS);
+	DUMP_REG(DC_WINBUF_ADDR_V_OFFSET);
+	DUMP_REG(DC_WINBUF_ADDR_V_OFFSET_NS);
+	DUMP_REG(DC_WINBUF_UFLOW_STATUS);
+	DUMP_REG(DC_WINBUF_AD_UFLOW_STATUS);
+	DUMP_REG(DC_WINBUF_BD_UFLOW_STATUS);
+	DUMP_REG(DC_WINBUF_CD_UFLOW_STATUS);
+
+#undef DUMP_REG
+
+	return 0;
+}
+
+static struct drm_info_list debugfs_files[] = {
+	{ "regs", tegra_dc_show_regs, 0, NULL },
+};
+
+static int tegra_dc_debugfs_init(struct tegra_dc *dc, struct drm_minor *minor)
+{
+	unsigned int i;
+	char *name;
+	int err;
+
+	name = kasprintf(GFP_KERNEL, "dc.%d", dc->pipe);
+	dc->debugfs = debugfs_create_dir(name, minor->debugfs_root);
+	kfree(name);
+
+	if (!dc->debugfs)
+		return -ENOMEM;
+
+	dc->debugfs_files = kmemdup(debugfs_files, sizeof(debugfs_files),
+				    GFP_KERNEL);
+	if (!dc->debugfs_files) {
+		err = -ENOMEM;
+		goto remove;
+	}
+
+	for (i = 0; i < ARRAY_SIZE(debugfs_files); i++)
+		dc->debugfs_files[i].data = dc;
+
+	err = drm_debugfs_create_files(dc->debugfs_files,
+				       ARRAY_SIZE(debugfs_files),
+				       dc->debugfs, minor);
+	if (err < 0)
+		goto free;
+
+	dc->minor = minor;
+
+	return 0;
+
+free:
+	kfree(dc->debugfs_files);
+	dc->debugfs_files = NULL;
+remove:
+	debugfs_remove(dc->debugfs);
+	dc->debugfs = NULL;
+
+	return err;
+}
+
+static int tegra_dc_debugfs_exit(struct tegra_dc *dc)
+{
+	drm_debugfs_remove_files(dc->debugfs_files, ARRAY_SIZE(debugfs_files),
+				 dc->minor);
+	dc->minor = NULL;
+
+	kfree(dc->debugfs_files);
+	dc->debugfs_files = NULL;
+
+	debugfs_remove(dc->debugfs);
+	dc->debugfs = NULL;
+
+	return 0;
+}
+
+static int tegra_dc_drm_init(struct host1x_client *client,
+			     struct drm_device *drm)
+{
+	struct tegra_dc *dc = host1x_client_to_dc(client);
+	int err;
+
+	dc->pipe = drm->mode_config.num_crtc;
+
+	drm_crtc_init(drm, &dc->base, &tegra_crtc_funcs);
+	drm_mode_crtc_set_gamma_size(&dc->base, 256);
+	drm_crtc_helper_add(&dc->base, &tegra_crtc_helper_funcs);
+
+	err = tegra_dc_rgb_init(drm, dc);
+	if (err < 0 && err != -ENODEV) {
+		dev_err(dc->dev, "failed to initialize RGB output: %d\n", err);
+		return err;
+	}
+
+	err = tegra_dc_add_planes(drm, dc);
+	if (err < 0)
+		return err;
+
+	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
+		err = tegra_dc_debugfs_init(dc, drm->primary);
+		if (err < 0)
+			dev_err(dc->dev, "debugfs setup failed: %d\n", err);
+	}
+
+	err = devm_request_irq(dc->dev, dc->irq, tegra_dc_irq, 0,
+			       dev_name(dc->dev), dc);
+	if (err < 0) {
+		dev_err(dc->dev, "failed to request IRQ#%u: %d\n", dc->irq,
+			err);
+		return err;
+	}
+
+	return 0;
+}
+
+static int tegra_dc_drm_exit(struct host1x_client *client)
+{
+	struct tegra_dc *dc = host1x_client_to_dc(client);
+	int err;
+
+	devm_free_irq(dc->dev, dc->irq, dc);
+
+	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
+		err = tegra_dc_debugfs_exit(dc);
+		if (err < 0)
+			dev_err(dc->dev, "debugfs cleanup failed: %d\n", err);
+	}
+
+	err = tegra_dc_rgb_exit(dc);
+	if (err) {
+		dev_err(dc->dev, "failed to shutdown RGB output: %d\n", err);
+		return err;
+	}
+
+	return 0;
+}
+
+static const struct host1x_client_ops dc_client_ops = {
+	.drm_init = tegra_dc_drm_init,
+	.drm_exit = tegra_dc_drm_exit,
+};
+
+static int tegra_dc_probe(struct platform_device *pdev)
+{
+	struct host1x_drm *host1x = host1x_get_drm_data(pdev->dev.parent);
+	struct resource *regs;
+	struct tegra_dc *dc;
+	int err;
+
+	dc = devm_kzalloc(&pdev->dev, sizeof(*dc), GFP_KERNEL);
+	if (!dc)
+		return -ENOMEM;
+
+	spin_lock_init(&dc->lock);
+	INIT_LIST_HEAD(&dc->list);
+	dc->dev = &pdev->dev;
+
+	dc->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(dc->clk)) {
+		dev_err(&pdev->dev, "failed to get clock\n");
+		return PTR_ERR(dc->clk);
+	}
+
+	err = clk_prepare_enable(dc->clk);
+	if (err < 0)
+		return err;
+
+	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!regs) {
+		dev_err(&pdev->dev, "failed to get registers\n");
+		return -ENXIO;
+	}
+
+	dc->regs = devm_ioremap_resource(&pdev->dev, regs);
+	if (IS_ERR(dc->regs))
+		return PTR_ERR(dc->regs);
+
+	dc->irq = platform_get_irq(pdev, 0);
+	if (dc->irq < 0) {
+		dev_err(&pdev->dev, "failed to get IRQ\n");
+		return -ENXIO;
+	}
+
+	INIT_LIST_HEAD(&dc->client.list);
+	dc->client.ops = &dc_client_ops;
+	dc->client.dev = &pdev->dev;
+
+	err = tegra_dc_rgb_probe(dc);
+	if (err < 0 && err != -ENODEV) {
+		dev_err(&pdev->dev, "failed to probe RGB output: %d\n", err);
+		return err;
+	}
+
+	err = host1x_register_client(host1x, &dc->client);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to register host1x client: %d\n",
+			err);
+		return err;
+	}
+
+	platform_set_drvdata(pdev, dc);
+
+	return 0;
+}
+
+static int tegra_dc_remove(struct platform_device *pdev)
+{
+	struct host1x_drm *host1x = host1x_get_drm_data(pdev->dev.parent);
+	struct tegra_dc *dc = platform_get_drvdata(pdev);
+	int err;
+
+	err = host1x_unregister_client(host1x, &dc->client);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to unregister host1x client: %d\n",
+			err);
+		return err;
+	}
+
+	clk_disable_unprepare(dc->clk);
+
+	return 0;
+}
+
+static struct of_device_id tegra_dc_of_match[] = {
+	{ .compatible = "nvidia,tegra30-dc", },
+	{ .compatible = "nvidia,tegra20-dc", },
+	{ },
+};
+
+struct platform_driver tegra_dc_driver = {
+	.driver = {
+		.name = "tegra-dc",
+		.owner = THIS_MODULE,
+		.of_match_table = tegra_dc_of_match,
+	},
+	.probe = tegra_dc_probe,
+	.remove = tegra_dc_remove,
+};
diff --git a/drivers/gpu/host1x/drm/dc.h b/drivers/gpu/host1x/drm/dc.h
new file mode 100644
index 00000000000..79eaec9aac7
--- /dev/null
+++ b/drivers/gpu/host1x/drm/dc.h
@@ -0,0 +1,400 @@
+/*
+ * Copyright (C) 2012 Avionic Design GmbH
+ * Copyright (C) 2012 NVIDIA CORPORATION.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef TEGRA_DC_H
+#define TEGRA_DC_H 1
+
+#define DC_CMD_GENERAL_INCR_SYNCPT		0x000
+#define DC_CMD_GENERAL_INCR_SYNCPT_CNTRL	0x001
+#define DC_CMD_GENERAL_INCR_SYNCPT_ERROR	0x002
+#define DC_CMD_WIN_A_INCR_SYNCPT		0x008
+#define DC_CMD_WIN_A_INCR_SYNCPT_CNTRL		0x009
+#define DC_CMD_WIN_A_INCR_SYNCPT_ERROR		0x00a
+#define DC_CMD_WIN_B_INCR_SYNCPT		0x010
+#define DC_CMD_WIN_B_INCR_SYNCPT_CNTRL		0x011
+#define DC_CMD_WIN_B_INCR_SYNCPT_ERROR		0x012
+#define DC_CMD_WIN_C_INCR_SYNCPT		0x018
+#define DC_CMD_WIN_C_INCR_SYNCPT_CNTRL		0x019
+#define DC_CMD_WIN_C_INCR_SYNCPT_ERROR		0x01a
+#define DC_CMD_CONT_SYNCPT_VSYNC		0x028
+#define DC_CMD_DISPLAY_COMMAND_OPTION0		0x031
+#define DC_CMD_DISPLAY_COMMAND			0x032
+#define DISP_CTRL_MODE_STOP (0 << 5)
+#define DISP_CTRL_MODE_C_DISPLAY (1 << 5)
+#define DISP_CTRL_MODE_NC_DISPLAY (2 << 5)
+#define DC_CMD_SIGNAL_RAISE			0x033
+#define DC_CMD_DISPLAY_POWER_CONTROL		0x036
+#define PW0_ENABLE (1 <<  0)
+#define PW1_ENABLE (1 <<  2)
+#define PW2_ENABLE (1 <<  4)
+#define PW3_ENABLE (1 <<  6)
+#define PW4_ENABLE (1 <<  8)
+#define PM0_ENABLE (1 << 16)
+#define PM1_ENABLE (1 << 18)
+
+#define DC_CMD_INT_STATUS			0x037
+#define DC_CMD_INT_MASK				0x038
+#define DC_CMD_INT_ENABLE			0x039
+#define DC_CMD_INT_TYPE				0x03a
+#define DC_CMD_INT_POLARITY			0x03b
+#define CTXSW_INT     (1 << 0)
+#define FRAME_END_INT (1 << 1)
+#define VBLANK_INT    (1 << 2)
+#define WIN_A_UF_INT  (1 << 8)
+#define WIN_B_UF_INT  (1 << 9)
+#define WIN_C_UF_INT  (1 << 10)
+#define WIN_A_OF_INT  (1 << 14)
+#define WIN_B_OF_INT  (1 << 15)
+#define WIN_C_OF_INT  (1 << 16)
+
+#define DC_CMD_SIGNAL_RAISE1			0x03c
+#define DC_CMD_SIGNAL_RAISE2			0x03d
+#define DC_CMD_SIGNAL_RAISE3			0x03e
+
+#define DC_CMD_STATE_ACCESS			0x040
+#define READ_MUX  (1 << 0)
+#define WRITE_MUX (1 << 2)
+
+#define DC_CMD_STATE_CONTROL			0x041
+#define GENERAL_ACT_REQ (1 <<  0)
+#define WIN_A_ACT_REQ   (1 <<  1)
+#define WIN_B_ACT_REQ   (1 <<  2)
+#define WIN_C_ACT_REQ   (1 <<  3)
+#define GENERAL_UPDATE  (1 <<  8)
+#define WIN_A_UPDATE    (1 <<  9)
+#define WIN_B_UPDATE    (1 << 10)
+#define WIN_C_UPDATE    (1 << 11)
+#define NC_HOST_TRIG    (1 << 24)
+
+#define DC_CMD_DISPLAY_WINDOW_HEADER		0x042
+#define WINDOW_A_SELECT (1 << 4)
+#define WINDOW_B_SELECT (1 << 5)
+#define WINDOW_C_SELECT (1 << 6)
+
+#define DC_CMD_REG_ACT_CONTROL			0x043
+
+#define DC_COM_CRC_CONTROL			0x300
+#define DC_COM_CRC_CHECKSUM			0x301
+#define DC_COM_PIN_OUTPUT_ENABLE(x) (0x302 + (x))
+#define DC_COM_PIN_OUTPUT_POLARITY(x) (0x306 + (x))
+#define LVS_OUTPUT_POLARITY_LOW (1 << 28)
+#define LHS_OUTPUT_POLARITY_LOW (1 << 30)
+#define DC_COM_PIN_OUTPUT_DATA(x) (0x30a + (x))
+#define DC_COM_PIN_INPUT_ENABLE(x) (0x30e + (x))
+#define DC_COM_PIN_INPUT_DATA(x) (0x312 + (x))
+#define DC_COM_PIN_OUTPUT_SELECT(x) (0x314 + (x))
+
+#define DC_COM_PIN_MISC_CONTROL			0x31b
+#define DC_COM_PIN_PM0_CONTROL			0x31c
+#define DC_COM_PIN_PM0_DUTY_CYCLE		0x31d
+#define DC_COM_PIN_PM1_CONTROL			0x31e
+#define DC_COM_PIN_PM1_DUTY_CYCLE		0x31f
+
+#define DC_COM_SPI_CONTROL			0x320
+#define DC_COM_SPI_START_BYTE			0x321
+#define DC_COM_HSPI_WRITE_DATA_AB		0x322
+#define DC_COM_HSPI_WRITE_DATA_CD		0x323
+#define DC_COM_HSPI_CS_DC			0x324
+#define DC_COM_SCRATCH_REGISTER_A		0x325
+#define DC_COM_SCRATCH_REGISTER_B		0x326
+#define DC_COM_GPIO_CTRL			0x327
+#define DC_COM_GPIO_DEBOUNCE_COUNTER		0x328
+#define DC_COM_CRC_CHECKSUM_LATCHED		0x329
+
+#define DC_DISP_DISP_SIGNAL_OPTIONS0		0x400
+#define H_PULSE_0_ENABLE (1 <<  8)
+#define H_PULSE_1_ENABLE (1 << 10)
+#define H_PULSE_2_ENABLE (1 << 12)
+
+#define DC_DISP_DISP_SIGNAL_OPTIONS1		0x401
+
+#define DC_DISP_DISP_WIN_OPTIONS		0x402
+#define HDMI_ENABLE (1 << 30)
+
+#define DC_DISP_DISP_MEM_HIGH_PRIORITY		0x403
+#define CURSOR_THRESHOLD(x)   (((x) & 0x03) << 24)
+#define WINDOW_A_THRESHOLD(x) (((x) & 0x7f) << 16)
+#define WINDOW_B_THRESHOLD(x) (((x) & 0x7f) <<  8)
+#define WINDOW_C_THRESHOLD(x) (((x) & 0xff) <<  0)
+
+#define DC_DISP_DISP_MEM_HIGH_PRIORITY_TIMER	0x404
+#define CURSOR_DELAY(x)   (((x) & 0x3f) << 24)
+#define WINDOW_A_DELAY(x) (((x) & 0x3f) << 16)
+#define WINDOW_B_DELAY(x) (((x) & 0x3f) <<  8)
+#define WINDOW_C_DELAY(x) (((x) & 0x3f) <<  0)
+
+#define DC_DISP_DISP_TIMING_OPTIONS		0x405
+#define VSYNC_H_POSITION(x) ((x) & 0xfff)
+
+#define DC_DISP_REF_TO_SYNC			0x406
+#define DC_DISP_SYNC_WIDTH			0x407
+#define DC_DISP_BACK_PORCH			0x408
+#define DC_DISP_ACTIVE				0x409
+#define DC_DISP_FRONT_PORCH			0x40a
+#define DC_DISP_H_PULSE0_CONTROL		0x40b
+#define DC_DISP_H_PULSE0_POSITION_A		0x40c
+#define DC_DISP_H_PULSE0_POSITION_B		0x40d
+#define DC_DISP_H_PULSE0_POSITION_C		0x40e
+#define DC_DISP_H_PULSE0_POSITION_D		0x40f
+#define DC_DISP_H_PULSE1_CONTROL		0x410
+#define DC_DISP_H_PULSE1_POSITION_A		0x411
+#define DC_DISP_H_PULSE1_POSITION_B		0x412
+#define DC_DISP_H_PULSE1_POSITION_C		0x413
+#define DC_DISP_H_PULSE1_POSITION_D		0x414
+#define DC_DISP_H_PULSE2_CONTROL		0x415
+#define DC_DISP_H_PULSE2_POSITION_A		0x416
+#define DC_DISP_H_PULSE2_POSITION_B		0x417
+#define DC_DISP_H_PULSE2_POSITION_C		0x418
+#define DC_DISP_H_PULSE2_POSITION_D		0x419
+#define DC_DISP_V_PULSE0_CONTROL		0x41a
+#define DC_DISP_V_PULSE0_POSITION_A		0x41b
+#define DC_DISP_V_PULSE0_POSITION_B		0x41c
+#define DC_DISP_V_PULSE0_POSITION_C		0x41d
+#define DC_DISP_V_PULSE1_CONTROL		0x41e
+#define DC_DISP_V_PULSE1_POSITION_A		0x41f
+#define DC_DISP_V_PULSE1_POSITION_B		0x420
+#define DC_DISP_V_PULSE1_POSITION_C		0x421
+#define DC_DISP_V_PULSE2_CONTROL		0x422
+#define DC_DISP_V_PULSE2_POSITION_A		0x423
+#define DC_DISP_V_PULSE3_CONTROL		0x424
+#define DC_DISP_V_PULSE3_POSITION_A		0x425
+#define DC_DISP_M0_CONTROL			0x426
+#define DC_DISP_M1_CONTROL			0x427
+#define DC_DISP_DI_CONTROL			0x428
+#define DC_DISP_PP_CONTROL			0x429
+#define DC_DISP_PP_SELECT_A			0x42a
+#define DC_DISP_PP_SELECT_B			0x42b
+#define DC_DISP_PP_SELECT_C			0x42c
+#define DC_DISP_PP_SELECT_D			0x42d
+
+#define PULSE_MODE_NORMAL    (0 << 3)
+#define PULSE_MODE_ONE_CLOCK (1 << 3)
+#define PULSE_POLARITY_HIGH  (0 << 4)
+#define PULSE_POLARITY_LOW   (1 << 4)
+#define PULSE_QUAL_ALWAYS    (0 << 6)
+#define PULSE_QUAL_VACTIVE   (2 << 6)
+#define PULSE_QUAL_VACTIVE1  (3 << 6)
+#define PULSE_LAST_START_A   (0 << 8)
+#define PULSE_LAST_END_A     (1 << 8)
+#define PULSE_LAST_START_B   (2 << 8)
+#define PULSE_LAST_END_B     (3 << 8)
+#define PULSE_LAST_START_C   (4 << 8)
+#define PULSE_LAST_END_C     (5 << 8)
+#define PULSE_LAST_START_D   (6 << 8)
+#define PULSE_LAST_END_D     (7 << 8)
+
+#define PULSE_START(x) (((x) & 0xfff) <<  0)
+#define PULSE_END(x)   (((x) & 0xfff) << 16)
+
+#define DC_DISP_DISP_CLOCK_CONTROL		0x42e
+#define PIXEL_CLK_DIVIDER_PCD1  (0 << 8)
+#define PIXEL_CLK_DIVIDER_PCD1H (1 << 8)
+#define PIXEL_CLK_DIVIDER_PCD2  (2 << 8)
+#define PIXEL_CLK_DIVIDER_PCD3  (3 << 8)
+#define PIXEL_CLK_DIVIDER_PCD4  (4 << 8)
+#define PIXEL_CLK_DIVIDER_PCD6  (5 << 8)
+#define PIXEL_CLK_DIVIDER_PCD8  (6 << 8)
+#define PIXEL_CLK_DIVIDER_PCD9  (7 << 8)
+#define PIXEL_CLK_DIVIDER_PCD12 (8 << 8)
+#define PIXEL_CLK_DIVIDER_PCD16 (9 << 8)
+#define PIXEL_CLK_DIVIDER_PCD18 (10 << 8)
+#define PIXEL_CLK_DIVIDER_PCD24 (11 << 8)
+#define PIXEL_CLK_DIVIDER_PCD13 (12 << 8)
+#define SHIFT_CLK_DIVIDER(x)    ((x) & 0xff)
+
+#define DC_DISP_DISP_INTERFACE_CONTROL		0x42f
+#define DISP_DATA_FORMAT_DF1P1C    (0 << 0)
+#define DISP_DATA_FORMAT_DF1P2C24B (1 << 0)
+#define DISP_DATA_FORMAT_DF1P2C18B (2 << 0)
+#define DISP_DATA_FORMAT_DF1P2C16B (3 << 0)
+#define DISP_DATA_FORMAT_DF2S      (4 << 0)
+#define DISP_DATA_FORMAT_DF3S      (5 << 0)
+#define DISP_DATA_FORMAT_DFSPI     (6 << 0)
+#define DISP_DATA_FORMAT_DF1P3C24B (7 << 0)
+#define DISP_DATA_FORMAT_DF1P3C18B (8 << 0)
+#define DISP_ALIGNMENT_MSB         (0 << 8)
+#define DISP_ALIGNMENT_LSB         (1 << 8)
+#define DISP_ORDER_RED_BLUE        (0 << 9)
+#define DISP_ORDER_BLUE_RED        (1 << 9)
+
+#define DC_DISP_DISP_COLOR_CONTROL		0x430
+#define BASE_COLOR_SIZE666     (0 << 0)
+#define BASE_COLOR_SIZE111     (1 << 0)
+#define BASE_COLOR_SIZE222     (2 << 0)
+#define BASE_COLOR_SIZE333     (3 << 0)
+#define BASE_COLOR_SIZE444     (4 << 0)
+#define BASE_COLOR_SIZE555     (5 << 0)
+#define BASE_COLOR_SIZE565     (6 << 0)
+#define BASE_COLOR_SIZE332     (7 << 0)
+#define BASE_COLOR_SIZE888     (8 << 0)
+#define DITHER_CONTROL_DISABLE (0 << 8)
+#define DITHER_CONTROL_ORDERED (2 << 8)
+#define DITHER_CONTROL_ERRDIFF (3 << 8)
+
+#define DC_DISP_SHIFT_CLOCK_OPTIONS		0x431
+
+#define DC_DISP_DATA_ENABLE_OPTIONS		0x432
+#define DE_SELECT_ACTIVE_BLANK  (0 << 0)
+#define DE_SELECT_ACTIVE        (1 << 0)
+#define DE_SELECT_ACTIVE_IS     (2 << 0)
+#define DE_CONTROL_ONECLK       (0 << 2)
+#define DE_CONTROL_NORMAL       (1 << 2)
+#define DE_CONTROL_EARLY_EXT    (2 << 2)
+#define DE_CONTROL_EARLY        (3 << 2)
+#define DE_CONTROL_ACTIVE_BLANK (4 << 2)
+
+#define DC_DISP_SERIAL_INTERFACE_OPTIONS	0x433
+#define DC_DISP_LCD_SPI_OPTIONS			0x434
+#define DC_DISP_BORDER_COLOR			0x435
+#define DC_DISP_COLOR_KEY0_LOWER		0x436
+#define DC_DISP_COLOR_KEY0_UPPER		0x437
+#define DC_DISP_COLOR_KEY1_LOWER		0x438
+#define DC_DISP_COLOR_KEY1_UPPER		0x439
+
+#define DC_DISP_CURSOR_FOREGROUND		0x43c
+#define DC_DISP_CURSOR_BACKGROUND		0x43d
+
+#define DC_DISP_CURSOR_START_ADDR		0x43e
+#define DC_DISP_CURSOR_START_ADDR_NS		0x43f
+
+#define DC_DISP_CURSOR_POSITION			0x440
+#define DC_DISP_CURSOR_POSITION_NS		0x441
+
+#define DC_DISP_INIT_SEQ_CONTROL		0x442
+#define DC_DISP_SPI_INIT_SEQ_DATA_A		0x443
+#define DC_DISP_SPI_INIT_SEQ_DATA_B		0x444
+#define DC_DISP_SPI_INIT_SEQ_DATA_C		0x445
+#define DC_DISP_SPI_INIT_SEQ_DATA_D		0x446
+
+#define DC_DISP_DC_MCCIF_FIFOCTRL		0x480
+#define DC_DISP_MCCIF_DISPLAY0A_HYST		0x481
+#define DC_DISP_MCCIF_DISPLAY0B_HYST		0x482
+#define DC_DISP_MCCIF_DISPLAY1A_HYST		0x483
+#define DC_DISP_MCCIF_DISPLAY1B_HYST		0x484
+
+#define DC_DISP_DAC_CRT_CTRL			0x4c0
+#define DC_DISP_DISP_MISC_CONTROL		0x4c1
+#define DC_DISP_SD_CONTROL			0x4c2
+#define DC_DISP_SD_CSC_COEFF			0x4c3
+#define DC_DISP_SD_LUT(x)			(0x4c4 + (x))
+#define DC_DISP_SD_FLICKER_CONTROL		0x4cd
+#define DC_DISP_DC_PIXEL_COUNT			0x4ce
+#define DC_DISP_SD_HISTOGRAM(x)			(0x4cf + (x))
+#define DC_DISP_SD_BL_PARAMETERS		0x4d7
+#define DC_DISP_SD_BL_TF(x)			(0x4d8 + (x))
+#define DC_DISP_SD_BL_CONTROL			0x4dc
+#define DC_DISP_SD_HW_K_VALUES			0x4dd
+#define DC_DISP_SD_MAN_K_VALUES			0x4de
+
+#define DC_WIN_CSC_YOF				0x611
+#define DC_WIN_CSC_KYRGB			0x612
+#define DC_WIN_CSC_KUR				0x613
+#define DC_WIN_CSC_KVR				0x614
+#define DC_WIN_CSC_KUG				0x615
+#define DC_WIN_CSC_KVG				0x616
+#define DC_WIN_CSC_KUB				0x617
+#define DC_WIN_CSC_KVB				0x618
+
+#define DC_WIN_WIN_OPTIONS			0x700
+#define COLOR_EXPAND (1 <<  6)
+#define CSC_ENABLE   (1 << 18)
+#define WIN_ENABLE   (1 << 30)
+
+#define DC_WIN_BYTE_SWAP			0x701
+#define BYTE_SWAP_NOSWAP  (0 << 0)
+#define BYTE_SWAP_SWAP2   (1 << 0)
+#define BYTE_SWAP_SWAP4   (2 << 0)
+#define BYTE_SWAP_SWAP4HW (3 << 0)
+
+#define DC_WIN_BUFFER_CONTROL			0x702
+#define BUFFER_CONTROL_HOST  (0 << 0)
+#define BUFFER_CONTROL_VI    (1 << 0)
+#define BUFFER_CONTROL_EPP   (2 << 0)
+#define BUFFER_CONTROL_MPEGE (3 << 0)
+#define BUFFER_CONTROL_SB2D  (4 << 0)
+
+#define DC_WIN_COLOR_DEPTH			0x703
+#define WIN_COLOR_DEPTH_P1              0
+#define WIN_COLOR_DEPTH_P2              1
+#define WIN_COLOR_DEPTH_P4              2
+#define WIN_COLOR_DEPTH_P8              3
+#define WIN_COLOR_DEPTH_B4G4R4A4        4
+#define WIN_COLOR_DEPTH_B5G5R5A         5
+#define WIN_COLOR_DEPTH_B5G6R5          6
+#define WIN_COLOR_DEPTH_AB5G5R5         7
+#define WIN_COLOR_DEPTH_B8G8R8A8       12
+#define WIN_COLOR_DEPTH_R8G8B8A8       13
+#define WIN_COLOR_DEPTH_B6x2G6x2R6x2A8 14
+#define WIN_COLOR_DEPTH_R6x2G6x2B6x2A8 15
+#define WIN_COLOR_DEPTH_YCbCr422       16
+#define WIN_COLOR_DEPTH_YUV422         17
+#define WIN_COLOR_DEPTH_YCbCr420P      18
+#define WIN_COLOR_DEPTH_YUV420P        19
+#define WIN_COLOR_DEPTH_YCbCr422P      20
+#define WIN_COLOR_DEPTH_YUV422P        21
+#define WIN_COLOR_DEPTH_YCbCr422R      22
+#define WIN_COLOR_DEPTH_YUV422R        23
+#define WIN_COLOR_DEPTH_YCbCr422RA     24
+#define WIN_COLOR_DEPTH_YUV422RA       25
+
+#define DC_WIN_POSITION				0x704
+#define H_POSITION(x) (((x) & 0x1fff) <<  0)
+#define V_POSITION(x) (((x) & 0x1fff) << 16)
+
+#define DC_WIN_SIZE				0x705
+#define H_SIZE(x) (((x) & 0x1fff) <<  0)
+#define V_SIZE(x) (((x) & 0x1fff) << 16)
+
+#define DC_WIN_PRESCALED_SIZE			0x706
+#define H_PRESCALED_SIZE(x) (((x) & 0x7fff) <<  0)
+#define V_PRESCALED_SIZE(x) (((x) & 0x1fff) << 16)
+
+#define DC_WIN_H_INITIAL_DDA			0x707
+#define DC_WIN_V_INITIAL_DDA			0x708
+#define DC_WIN_DDA_INC				0x709
+#define H_DDA_INC(x) (((x) & 0xffff) <<  0)
+#define V_DDA_INC(x) (((x) & 0xffff) << 16)
+
+#define DC_WIN_LINE_STRIDE			0x70a
+#define DC_WIN_BUF_STRIDE			0x70b
+#define DC_WIN_UV_BUF_STRIDE			0x70c
+#define DC_WIN_BUFFER_ADDR_MODE			0x70d
+#define DC_WIN_DV_CONTROL			0x70e
+
+#define DC_WIN_BLEND_NOKEY			0x70f
+#define DC_WIN_BLEND_1WIN			0x710
+#define DC_WIN_BLEND_2WIN_X			0x711
+#define DC_WIN_BLEND_2WIN_Y			0x712
+#define DC_WIN_BLEND_3WIN_XY			0x713
+
+#define DC_WIN_HP_FETCH_CONTROL			0x714
+
+#define DC_WINBUF_START_ADDR			0x800
+#define DC_WINBUF_START_ADDR_NS			0x801
+#define DC_WINBUF_START_ADDR_U			0x802
+#define DC_WINBUF_START_ADDR_U_NS		0x803
+#define DC_WINBUF_START_ADDR_V			0x804
+#define DC_WINBUF_START_ADDR_V_NS		0x805
+
+#define DC_WINBUF_ADDR_H_OFFSET			0x806
+#define DC_WINBUF_ADDR_H_OFFSET_NS		0x807
+#define DC_WINBUF_ADDR_V_OFFSET			0x808
+#define DC_WINBUF_ADDR_V_OFFSET_NS		0x809
+
+#define DC_WINBUF_UFLOW_STATUS			0x80a
+
+#define DC_WINBUF_AD_UFLOW_STATUS		0xbca
+#define DC_WINBUF_BD_UFLOW_STATUS		0xdca
+#define DC_WINBUF_CD_UFLOW_STATUS		0xfca
+
+/* synchronization points */
+#define SYNCPT_VBLANK0 26
+#define SYNCPT_VBLANK1 27
+
+#endif /* TEGRA_DC_H */
diff --git a/drivers/gpu/host1x/drm/drm.c b/drivers/gpu/host1x/drm/drm.c
new file mode 100644
index 00000000000..2b561c9118c
--- /dev/null
+++ b/drivers/gpu/host1x/drm/drm.c
@@ -0,0 +1,640 @@
+/*
+ * Copyright (C) 2012 Avionic Design GmbH
+ * Copyright (C) 2012-2013 NVIDIA CORPORATION.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/module.h>
+#include <linux/of_address.h>
+#include <linux/of_platform.h>
+
+#include <linux/dma-mapping.h>
+#include <asm/dma-iommu.h>
+
+#include <drm/drm.h>
+#include <drm/drmP.h>
+
+#include "host1x_client.h"
+#include "dev.h"
+#include "drm.h"
+#include "gem.h"
+#include "syncpt.h"
+
+#define DRIVER_NAME "tegra"
+#define DRIVER_DESC "NVIDIA Tegra graphics"
+#define DRIVER_DATE "20120330"
+#define DRIVER_MAJOR 0
+#define DRIVER_MINOR 0
+#define DRIVER_PATCHLEVEL 0
+
+struct host1x_drm_client {
+	struct host1x_client *client;
+	struct device_node *np;
+	struct list_head list;
+};
+
+static int host1x_add_drm_client(struct host1x_drm *host1x,
+				 struct device_node *np)
+{
+	struct host1x_drm_client *client;
+
+	client = kzalloc(sizeof(*client), GFP_KERNEL);
+	if (!client)
+		return -ENOMEM;
+
+	INIT_LIST_HEAD(&client->list);
+	client->np = of_node_get(np);
+
+	list_add_tail(&client->list, &host1x->drm_clients);
+
+	return 0;
+}
+
+static int host1x_activate_drm_client(struct host1x_drm *host1x,
+				      struct host1x_drm_client *drm,
+				      struct host1x_client *client)
+{
+	mutex_lock(&host1x->drm_clients_lock);
+	list_del_init(&drm->list);
+	list_add_tail(&drm->list, &host1x->drm_active);
+	drm->client = client;
+	mutex_unlock(&host1x->drm_clients_lock);
+
+	return 0;
+}
+
+static int host1x_remove_drm_client(struct host1x_drm *host1x,
+				    struct host1x_drm_client *client)
+{
+	mutex_lock(&host1x->drm_clients_lock);
+	list_del_init(&client->list);
+	mutex_unlock(&host1x->drm_clients_lock);
+
+	of_node_put(client->np);
+	kfree(client);
+
+	return 0;
+}
+
+static int host1x_parse_dt(struct host1x_drm *host1x)
+{
+	static const char * const compat[] = {
+		"nvidia,tegra20-dc",
+		"nvidia,tegra20-hdmi",
+		"nvidia,tegra20-gr2d",
+		"nvidia,tegra30-dc",
+		"nvidia,tegra30-hdmi",
+		"nvidia,tegra30-gr2d",
+	};
+	unsigned int i;
+	int err;
+
+	for (i = 0; i < ARRAY_SIZE(compat); i++) {
+		struct device_node *np;
+
+		for_each_child_of_node(host1x->dev->of_node, np) {
+			if (of_device_is_compatible(np, compat[i]) &&
+			    of_device_is_available(np)) {
+				err = host1x_add_drm_client(host1x, np);
+				if (err < 0)
+					return err;
+			}
+		}
+	}
+
+	return 0;
+}
+
+int host1x_drm_alloc(struct platform_device *pdev)
+{
+	struct host1x_drm *host1x;
+	int err;
+
+	host1x = devm_kzalloc(&pdev->dev, sizeof(*host1x), GFP_KERNEL);
+	if (!host1x)
+		return -ENOMEM;
+
+	mutex_init(&host1x->drm_clients_lock);
+	INIT_LIST_HEAD(&host1x->drm_clients);
+	INIT_LIST_HEAD(&host1x->drm_active);
+	mutex_init(&host1x->clients_lock);
+	INIT_LIST_HEAD(&host1x->clients);
+	host1x->dev = &pdev->dev;
+
+	err = host1x_parse_dt(host1x);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to parse DT: %d\n", err);
+		return err;
+	}
+
+	host1x_set_drm_data(&pdev->dev, host1x);
+
+	return 0;
+}
+
+int host1x_drm_init(struct host1x_drm *host1x, struct drm_device *drm)
+{
+	struct host1x_client *client;
+
+	mutex_lock(&host1x->clients_lock);
+
+	list_for_each_entry(client, &host1x->clients, list) {
+		if (client->ops && client->ops->drm_init) {
+			int err = client->ops->drm_init(client, drm);
+			if (err < 0) {
+				dev_err(host1x->dev,
+					"DRM setup failed for %s: %d\n",
+					dev_name(client->dev), err);
+				return err;
+			}
+		}
+	}
+
+	mutex_unlock(&host1x->clients_lock);
+
+	return 0;
+}
+
+int host1x_drm_exit(struct host1x_drm *host1x)
+{
+	struct platform_device *pdev = to_platform_device(host1x->dev);
+	struct host1x_client *client;
+
+	if (!host1x->drm)
+		return 0;
+
+	mutex_lock(&host1x->clients_lock);
+
+	list_for_each_entry_reverse(client, &host1x->clients, list) {
+		if (client->ops && client->ops->drm_exit) {
+			int err = client->ops->drm_exit(client);
+			if (err < 0) {
+				dev_err(host1x->dev,
+					"DRM cleanup failed for %s: %d\n",
+					dev_name(client->dev), err);
+				return err;
+			}
+		}
+	}
+
+	mutex_unlock(&host1x->clients_lock);
+
+	drm_platform_exit(&tegra_drm_driver, pdev);
+	host1x->drm = NULL;
+
+	return 0;
+}
+
+int host1x_register_client(struct host1x_drm *host1x,
+			   struct host1x_client *client)
+{
+	struct host1x_drm_client *drm, *tmp;
+	int err;
+
+	mutex_lock(&host1x->clients_lock);
+	list_add_tail(&client->list, &host1x->clients);
+	mutex_unlock(&host1x->clients_lock);
+
+	list_for_each_entry_safe(drm, tmp, &host1x->drm_clients, list)
+		if (drm->np == client->dev->of_node)
+			host1x_activate_drm_client(host1x, drm, client);
+
+	if (list_empty(&host1x->drm_clients)) {
+		struct platform_device *pdev = to_platform_device(host1x->dev);
+
+		err = drm_platform_init(&tegra_drm_driver, pdev);
+		if (err < 0) {
+			dev_err(host1x->dev, "drm_platform_init(): %d\n", err);
+			return err;
+		}
+	}
+
+	return 0;
+}
+
+int host1x_unregister_client(struct host1x_drm *host1x,
+			     struct host1x_client *client)
+{
+	struct host1x_drm_client *drm, *tmp;
+	int err;
+
+	list_for_each_entry_safe(drm, tmp, &host1x->drm_active, list) {
+		if (drm->client == client) {
+			err = host1x_drm_exit(host1x);
+			if (err < 0) {
+				dev_err(host1x->dev, "host1x_drm_exit(): %d\n",
+					err);
+				return err;
+			}
+
+			host1x_remove_drm_client(host1x, drm);
+			break;
+		}
+	}
+
+	mutex_lock(&host1x->clients_lock);
+	list_del_init(&client->list);
+	mutex_unlock(&host1x->clients_lock);
+
+	return 0;
+}
+
+static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
+{
+	struct host1x_drm *host1x;
+	int err;
+
+	host1x = host1x_get_drm_data(drm->dev);
+	drm->dev_private = host1x;
+	host1x->drm = drm;
+
+	drm_mode_config_init(drm);
+
+	err = host1x_drm_init(host1x, drm);
+	if (err < 0)
+		return err;
+
+	err = drm_vblank_init(drm, drm->mode_config.num_crtc);
+	if (err < 0)
+		return err;
+
+	err = tegra_drm_fb_init(drm);
+	if (err < 0)
+		return err;
+
+	drm_kms_helper_poll_init(drm);
+
+	return 0;
+}
+
+static int tegra_drm_unload(struct drm_device *drm)
+{
+	drm_kms_helper_poll_fini(drm);
+	tegra_drm_fb_exit(drm);
+
+	drm_mode_config_cleanup(drm);
+
+	return 0;
+}
+
+static int tegra_drm_open(struct drm_device *drm, struct drm_file *filp)
+{
+	struct host1x_drm_file *fpriv;
+
+	fpriv = kzalloc(sizeof(*fpriv), GFP_KERNEL);
+	if (!fpriv)
+		return -ENOMEM;
+
+	INIT_LIST_HEAD(&fpriv->contexts);
+	filp->driver_priv = fpriv;
+
+	return 0;
+}
+
+static void host1x_drm_context_free(struct host1x_drm_context *context)
+{
+	context->client->ops->close_channel(context);
+	kfree(context);
+}
+
+static void tegra_drm_lastclose(struct drm_device *drm)
+{
+	struct host1x_drm *host1x = drm->dev_private;
+
+	tegra_fbdev_restore_mode(host1x->fbdev);
+}
+
+#ifdef CONFIG_DRM_TEGRA_STAGING
+static bool host1x_drm_file_owns_context(struct host1x_drm_file *file,
+					 struct host1x_drm_context *context)
+{
+	struct host1x_drm_context *ctx;
+
+	list_for_each_entry(ctx, &file->contexts, list)
+		if (ctx == context)
+			return true;
+
+	return false;
+}
+
+static int tegra_gem_create(struct drm_device *drm, void *data,
+			    struct drm_file *file)
+{
+	struct drm_tegra_gem_create *args = data;
+	struct tegra_bo *bo;
+
+	bo = tegra_bo_create_with_handle(file, drm, args->size,
+					 &args->handle);
+	if (IS_ERR(bo))
+		return PTR_ERR(bo);
+
+	return 0;
+}
+
+static int tegra_gem_mmap(struct drm_device *drm, void *data,
+			  struct drm_file *file)
+{
+	struct drm_tegra_gem_mmap *args = data;
+	struct drm_gem_object *gem;
+	struct tegra_bo *bo;
+
+	gem = drm_gem_object_lookup(drm, file, args->handle);
+	if (!gem)
+		return -EINVAL;
+
+	bo = to_tegra_bo(gem);
+
+	args->offset = tegra_bo_get_mmap_offset(bo);
+
+	drm_gem_object_unreference(gem);
+
+	return 0;
+}
+
+static int tegra_syncpt_read(struct drm_device *drm, void *data,
+			     struct drm_file *file)
+{
+	struct drm_tegra_syncpt_read *args = data;
+	struct host1x *host = dev_get_drvdata(drm->dev);
+	struct host1x_syncpt *sp = host1x_syncpt_get(host, args->id);
+
+	if (!sp)
+		return -EINVAL;
+
+	args->value = host1x_syncpt_read_min(sp);
+	return 0;
+}
+
+static int tegra_syncpt_incr(struct drm_device *drm, void *data,
+			     struct drm_file *file)
+{
+	struct drm_tegra_syncpt_incr *args = data;
+	struct host1x *host = dev_get_drvdata(drm->dev);
+	struct host1x_syncpt *sp = host1x_syncpt_get(host, args->id);
+
+	if (!sp)
+		return -EINVAL;
+
+	host1x_syncpt_incr(sp);
+	return 0;
+}
+
+static int tegra_syncpt_wait(struct drm_device *drm, void *data,
+			     struct drm_file *file)
+{
+	struct drm_tegra_syncpt_wait *args = data;
+	struct host1x *host = dev_get_drvdata(drm->dev);
+	struct host1x_syncpt *sp = host1x_syncpt_get(host, args->id);
+
+	if (!sp)
+		return -EINVAL;
+
+	return host1x_syncpt_wait(sp, args->thresh, args->timeout,
+				  &args->value);
+}
+
+static int tegra_open_channel(struct drm_device *drm, void *data,
+			      struct drm_file *file)
+{
+	struct drm_tegra_open_channel *args = data;
+	struct host1x_client *client;
+	struct host1x_drm_context *context;
+	struct host1x_drm_file *fpriv = file->driver_priv;
+	struct host1x_drm *host1x = drm->dev_private;
+	int err = -ENODEV;
+
+	context = kzalloc(sizeof(*context), GFP_KERNEL);
+	if (!context)
+		return -ENOMEM;
+
+	list_for_each_entry(client, &host1x->clients, list)
+		if (client->class == args->client) {
+			err = client->ops->open_channel(client, context);
+			if (err)
+				break;
+
+			context->client = client;
+			list_add(&context->list, &fpriv->contexts);
+			args->context = (uintptr_t)context;
+			return 0;
+		}
+
+	kfree(context);
+	return err;
+}
+
+static int tegra_close_channel(struct drm_device *drm, void *data,
+			       struct drm_file *file)
+{
+	struct drm_tegra_close_channel *args = data;
+	struct host1x_drm_file *fpriv = file->driver_priv;
+	struct host1x_drm_context *context =
+		(struct host1x_drm_context *)(uintptr_t)args->context;
+
+	if (!host1x_drm_file_owns_context(fpriv, context))
+		return -EINVAL;
+
+	list_del(&context->list);
+	host1x_drm_context_free(context);
+
+	return 0;
+}
+
+static int tegra_get_syncpt(struct drm_device *drm, void *data,
+			    struct drm_file *file)
+{
+	struct drm_tegra_get_syncpt *args = data;
+	struct host1x_drm_file *fpriv = file->driver_priv;
+	struct host1x_drm_context *context =
+		(struct host1x_drm_context *)(uintptr_t)args->context;
+	struct host1x_syncpt *syncpt;
+
+	if (!host1x_drm_file_owns_context(fpriv, context))
+		return -ENODEV;
+
+	if (args->index >= context->client->num_syncpts)
+		return -EINVAL;
+
+	syncpt = context->client->syncpts[args->index];
+	args->id = host1x_syncpt_id(syncpt);
+
+	return 0;
+}
+
+static int tegra_submit(struct drm_device *drm, void *data,
+			struct drm_file *file)
+{
+	struct drm_tegra_submit *args = data;
+	struct host1x_drm_file *fpriv = file->driver_priv;
+	struct host1x_drm_context *context =
+		(struct host1x_drm_context *)(uintptr_t)args->context;
+
+	if (!host1x_drm_file_owns_context(fpriv, context))
+		return -ENODEV;
+
+	return context->client->ops->submit(context, args, drm, file);
+}
+#endif
+
+static struct drm_ioctl_desc tegra_drm_ioctls[] = {
+#ifdef CONFIG_DRM_TEGRA_STAGING
+	DRM_IOCTL_DEF_DRV(TEGRA_GEM_CREATE, tegra_gem_create, DRM_UNLOCKED | DRM_AUTH),
+	DRM_IOCTL_DEF_DRV(TEGRA_GEM_MMAP, tegra_gem_mmap, DRM_UNLOCKED),
+	DRM_IOCTL_DEF_DRV(TEGRA_SYNCPT_READ, tegra_syncpt_read, DRM_UNLOCKED),
+	DRM_IOCTL_DEF_DRV(TEGRA_SYNCPT_INCR, tegra_syncpt_incr, DRM_UNLOCKED),
+	DRM_IOCTL_DEF_DRV(TEGRA_SYNCPT_WAIT, tegra_syncpt_wait, DRM_UNLOCKED),
+	DRM_IOCTL_DEF_DRV(TEGRA_OPEN_CHANNEL, tegra_open_channel, DRM_UNLOCKED),
+	DRM_IOCTL_DEF_DRV(TEGRA_CLOSE_CHANNEL, tegra_close_channel, DRM_UNLOCKED),
+	DRM_IOCTL_DEF_DRV(TEGRA_GET_SYNCPT, tegra_get_syncpt, DRM_UNLOCKED),
+	DRM_IOCTL_DEF_DRV(TEGRA_SUBMIT, tegra_submit, DRM_UNLOCKED),
+#endif
+};
+
+static const struct file_operations tegra_drm_fops = {
+	.owner = THIS_MODULE,
+	.open = drm_open,
+	.release = drm_release,
+	.unlocked_ioctl = drm_ioctl,
+	.mmap = tegra_drm_mmap,
+	.poll = drm_poll,
+	.fasync = drm_fasync,
+	.read = drm_read,
+#ifdef CONFIG_COMPAT
+	.compat_ioctl = drm_compat_ioctl,
+#endif
+	.llseek = noop_llseek,
+};
+
+static struct drm_crtc *tegra_crtc_from_pipe(struct drm_device *drm, int pipe)
+{
+	struct drm_crtc *crtc;
+
+	list_for_each_entry(crtc, &drm->mode_config.crtc_list, head) {
+		struct tegra_dc *dc = to_tegra_dc(crtc);
+
+		if (dc->pipe == pipe)
+			return crtc;
+	}
+
+	return NULL;
+}
+
+static u32 tegra_drm_get_vblank_counter(struct drm_device *dev, int crtc)
+{
+	/* TODO: implement real hardware counter using syncpoints */
+	return drm_vblank_count(dev, crtc);
+}
+
+static int tegra_drm_enable_vblank(struct drm_device *drm, int pipe)
+{
+	struct drm_crtc *crtc = tegra_crtc_from_pipe(drm, pipe);
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+
+	if (!crtc)
+		return -ENODEV;
+
+	tegra_dc_enable_vblank(dc);
+
+	return 0;
+}
+
+static void tegra_drm_disable_vblank(struct drm_device *drm, int pipe)
+{
+	struct drm_crtc *crtc = tegra_crtc_from_pipe(drm, pipe);
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+
+	if (crtc)
+		tegra_dc_disable_vblank(dc);
+}
+
+static void tegra_drm_preclose(struct drm_device *drm, struct drm_file *file)
+{
+	struct host1x_drm_file *fpriv = file->driver_priv;
+	struct host1x_drm_context *context, *tmp;
+	struct drm_crtc *crtc;
+
+	list_for_each_entry(crtc, &drm->mode_config.crtc_list, head)
+		tegra_dc_cancel_page_flip(crtc, file);
+
+	list_for_each_entry_safe(context, tmp, &fpriv->contexts, list)
+		host1x_drm_context_free(context);
+
+	kfree(fpriv);
+}
+
+#ifdef CONFIG_DEBUG_FS
+static int tegra_debugfs_framebuffers(struct seq_file *s, void *data)
+{
+	struct drm_info_node *node = (struct drm_info_node *)s->private;
+	struct drm_device *drm = node->minor->dev;
+	struct drm_framebuffer *fb;
+
+	mutex_lock(&drm->mode_config.fb_lock);
+
+	list_for_each_entry(fb, &drm->mode_config.fb_list, head) {
+		seq_printf(s, "%3d: user size: %d x %d, depth %d, %d bpp, refcount %d\n",
+			   fb->base.id, fb->width, fb->height, fb->depth,
+			   fb->bits_per_pixel,
+			   atomic_read(&fb->refcount.refcount));
+	}
+
+	mutex_unlock(&drm->mode_config.fb_lock);
+
+	return 0;
+}
+
+static struct drm_info_list tegra_debugfs_list[] = {
+	{ "framebuffers", tegra_debugfs_framebuffers, 0 },
+};
+
+static int tegra_debugfs_init(struct drm_minor *minor)
+{
+	return drm_debugfs_create_files(tegra_debugfs_list,
+					ARRAY_SIZE(tegra_debugfs_list),
+					minor->debugfs_root, minor);
+}
+
+static void tegra_debugfs_cleanup(struct drm_minor *minor)
+{
+	drm_debugfs_remove_files(tegra_debugfs_list,
+				 ARRAY_SIZE(tegra_debugfs_list), minor);
+}
+#endif
+
+struct drm_driver tegra_drm_driver = {
+	.driver_features = DRIVER_BUS_PLATFORM | DRIVER_MODESET | DRIVER_GEM,
+	.load = tegra_drm_load,
+	.unload = tegra_drm_unload,
+	.open = tegra_drm_open,
+	.preclose = tegra_drm_preclose,
+	.lastclose = tegra_drm_lastclose,
+
+	.get_vblank_counter = tegra_drm_get_vblank_counter,
+	.enable_vblank = tegra_drm_enable_vblank,
+	.disable_vblank = tegra_drm_disable_vblank,
+
+#if defined(CONFIG_DEBUG_FS)
+	.debugfs_init = tegra_debugfs_init,
+	.debugfs_cleanup = tegra_debugfs_cleanup,
+#endif
+
+	.gem_free_object = tegra_bo_free_object,
+	.gem_vm_ops = &tegra_bo_vm_ops,
+	.dumb_create = tegra_bo_dumb_create,
+	.dumb_map_offset = tegra_bo_dumb_map_offset,
+	.dumb_destroy = tegra_bo_dumb_destroy,
+
+	.ioctls = tegra_drm_ioctls,
+	.num_ioctls = ARRAY_SIZE(tegra_drm_ioctls),
+	.fops = &tegra_drm_fops,
+
+	.name = DRIVER_NAME,
+	.desc = DRIVER_DESC,
+	.date = DRIVER_DATE,
+	.major = DRIVER_MAJOR,
+	.minor = DRIVER_MINOR,
+	.patchlevel = DRIVER_PATCHLEVEL,
+};
diff --git a/drivers/gpu/host1x/drm/drm.h b/drivers/gpu/host1x/drm/drm.h
new file mode 100644
index 00000000000..02ce020f257
--- /dev/null
+++ b/drivers/gpu/host1x/drm/drm.h
@@ -0,0 +1,271 @@
+/*
+ * Copyright (C) 2012 Avionic Design GmbH
+ * Copyright (C) 2012-2013 NVIDIA CORPORATION.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef HOST1X_DRM_H
+#define HOST1X_DRM_H 1
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_fixed.h>
+#include <uapi/drm/tegra_drm.h>
+
+#include "host1x.h"
+
+struct tegra_fb {
+	struct drm_framebuffer base;
+	struct tegra_bo **planes;
+	unsigned int num_planes;
+};
+
+struct tegra_fbdev {
+	struct drm_fb_helper base;
+	struct tegra_fb *fb;
+};
+
+struct host1x_drm {
+	struct drm_device *drm;
+	struct device *dev;
+	void __iomem *regs;
+	struct clk *clk;
+	int syncpt;
+	int irq;
+
+	struct mutex drm_clients_lock;
+	struct list_head drm_clients;
+	struct list_head drm_active;
+
+	struct mutex clients_lock;
+	struct list_head clients;
+
+	struct tegra_fbdev *fbdev;
+};
+
+struct host1x_client;
+
+struct host1x_drm_context {
+	struct host1x_client *client;
+	struct host1x_channel *channel;
+	struct list_head list;
+};
+
+struct host1x_client_ops {
+	int (*drm_init)(struct host1x_client *client, struct drm_device *drm);
+	int (*drm_exit)(struct host1x_client *client);
+	int (*open_channel)(struct host1x_client *client,
+			    struct host1x_drm_context *context);
+	void (*close_channel)(struct host1x_drm_context *context);
+	int (*submit)(struct host1x_drm_context *context,
+		      struct drm_tegra_submit *args, struct drm_device *drm,
+		      struct drm_file *file);
+};
+
+struct host1x_drm_file {
+	struct list_head contexts;
+};
+
+struct host1x_client {
+	struct host1x_drm *host1x;
+	struct device *dev;
+
+	const struct host1x_client_ops *ops;
+
+	enum host1x_class class;
+	struct host1x_channel *channel;
+
+	struct host1x_syncpt **syncpts;
+	unsigned int num_syncpts;
+
+	struct list_head list;
+};
+
+extern int host1x_drm_init(struct host1x_drm *host1x, struct drm_device *drm);
+extern int host1x_drm_exit(struct host1x_drm *host1x);
+
+extern int host1x_register_client(struct host1x_drm *host1x,
+				  struct host1x_client *client);
+extern int host1x_unregister_client(struct host1x_drm *host1x,
+				    struct host1x_client *client);
+
+struct tegra_output;
+
+struct tegra_dc {
+	struct host1x_client client;
+	spinlock_t lock;
+
+	struct host1x_drm *host1x;
+	struct device *dev;
+
+	struct drm_crtc base;
+	int pipe;
+
+	struct clk *clk;
+
+	void __iomem *regs;
+	int irq;
+
+	struct tegra_output *rgb;
+
+	struct list_head list;
+
+	struct drm_info_list *debugfs_files;
+	struct drm_minor *minor;
+	struct dentry *debugfs;
+
+	/* page-flip handling */
+	struct drm_pending_vblank_event *event;
+};
+
+static inline struct tegra_dc *host1x_client_to_dc(struct host1x_client *client)
+{
+	return container_of(client, struct tegra_dc, client);
+}
+
+static inline struct tegra_dc *to_tegra_dc(struct drm_crtc *crtc)
+{
+	return container_of(crtc, struct tegra_dc, base);
+}
+
+static inline void tegra_dc_writel(struct tegra_dc *dc, unsigned long value,
+				   unsigned long reg)
+{
+	writel(value, dc->regs + (reg << 2));
+}
+
+static inline unsigned long tegra_dc_readl(struct tegra_dc *dc,
+					   unsigned long reg)
+{
+	return readl(dc->regs + (reg << 2));
+}
+
+struct tegra_dc_window {
+	struct {
+		unsigned int x;
+		unsigned int y;
+		unsigned int w;
+		unsigned int h;
+	} src;
+	struct {
+		unsigned int x;
+		unsigned int y;
+		unsigned int w;
+		unsigned int h;
+	} dst;
+	unsigned int bits_per_pixel;
+	unsigned int format;
+	unsigned int stride[2];
+	unsigned long base[3];
+};
+
+/* from dc.c */
+extern unsigned int tegra_dc_format(uint32_t format);
+extern int tegra_dc_setup_window(struct tegra_dc *dc, unsigned int index,
+				 const struct tegra_dc_window *window);
+extern void tegra_dc_enable_vblank(struct tegra_dc *dc);
+extern void tegra_dc_disable_vblank(struct tegra_dc *dc);
+extern void tegra_dc_cancel_page_flip(struct drm_crtc *crtc,
+				      struct drm_file *file);
+
+struct tegra_output_ops {
+	int (*enable)(struct tegra_output *output);
+	int (*disable)(struct tegra_output *output);
+	int (*setup_clock)(struct tegra_output *output, struct clk *clk,
+			   unsigned long pclk);
+	int (*check_mode)(struct tegra_output *output,
+			  struct drm_display_mode *mode,
+			  enum drm_mode_status *status);
+};
+
+enum tegra_output_type {
+	TEGRA_OUTPUT_RGB,
+	TEGRA_OUTPUT_HDMI,
+};
+
+struct tegra_output {
+	struct device_node *of_node;
+	struct device *dev;
+
+	const struct tegra_output_ops *ops;
+	enum tegra_output_type type;
+
+	struct i2c_adapter *ddc;
+	const struct edid *edid;
+	unsigned int hpd_irq;
+	int hpd_gpio;
+
+	struct drm_encoder encoder;
+	struct drm_connector connector;
+};
+
+static inline struct tegra_output *encoder_to_output(struct drm_encoder *e)
+{
+	return container_of(e, struct tegra_output, encoder);
+}
+
+static inline struct tegra_output *connector_to_output(struct drm_connector *c)
+{
+	return container_of(c, struct tegra_output, connector);
+}
+
+static inline int tegra_output_enable(struct tegra_output *output)
+{
+	if (output && output->ops && output->ops->enable)
+		return output->ops->enable(output);
+
+	return output ? -ENOSYS : -EINVAL;
+}
+
+static inline int tegra_output_disable(struct tegra_output *output)
+{
+	if (output && output->ops && output->ops->disable)
+		return output->ops->disable(output);
+
+	return output ? -ENOSYS : -EINVAL;
+}
+
+static inline int tegra_output_setup_clock(struct tegra_output *output,
+					   struct clk *clk, unsigned long pclk)
+{
+	if (output && output->ops && output->ops->setup_clock)
+		return output->ops->setup_clock(output, clk, pclk);
+
+	return output ? -ENOSYS : -EINVAL;
+}
+
+static inline int tegra_output_check_mode(struct tegra_output *output,
+					  struct drm_display_mode *mode,
+					  enum drm_mode_status *status)
+{
+	if (output && output->ops && output->ops->check_mode)
+		return output->ops->check_mode(output, mode, status);
+
+	return output ? -ENOSYS : -EINVAL;
+}
+
+/* from rgb.c */
+extern int tegra_dc_rgb_probe(struct tegra_dc *dc);
+extern int tegra_dc_rgb_init(struct drm_device *drm, struct tegra_dc *dc);
+extern int tegra_dc_rgb_exit(struct tegra_dc *dc);
+
+/* from output.c */
+extern int tegra_output_parse_dt(struct tegra_output *output);
+extern int tegra_output_init(struct drm_device *drm, struct tegra_output *output);
+extern int tegra_output_exit(struct tegra_output *output);
+
+/* from fb.c */
+struct tegra_bo *tegra_fb_get_plane(struct drm_framebuffer *framebuffer,
+				    unsigned int index);
+extern int tegra_drm_fb_init(struct drm_device *drm);
+extern void tegra_drm_fb_exit(struct drm_device *drm);
+extern void tegra_fbdev_restore_mode(struct tegra_fbdev *fbdev);
+
+extern struct drm_driver tegra_drm_driver;
+
+#endif /* HOST1X_DRM_H */
diff --git a/drivers/gpu/host1x/drm/fb.c b/drivers/gpu/host1x/drm/fb.c
new file mode 100644
index 00000000000..979a3e32b78
--- /dev/null
+++ b/drivers/gpu/host1x/drm/fb.c
@@ -0,0 +1,374 @@
+/*
+ * Copyright (C) 2012-2013 Avionic Design GmbH
+ * Copyright (C) 2012 NVIDIA CORPORATION.  All rights reserved.
+ *
+ * Based on the KMS/FB CMA helpers
+ *   Copyright (C) 2012 Analog Device Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/module.h>
+
+#include "drm.h"
+#include "gem.h"
+
+static inline struct tegra_fb *to_tegra_fb(struct drm_framebuffer *fb)
+{
+	return container_of(fb, struct tegra_fb, base);
+}
+
+static inline struct tegra_fbdev *to_tegra_fbdev(struct drm_fb_helper *helper)
+{
+	return container_of(helper, struct tegra_fbdev, base);
+}
+
+struct tegra_bo *tegra_fb_get_plane(struct drm_framebuffer *framebuffer,
+				    unsigned int index)
+{
+	struct tegra_fb *fb = to_tegra_fb(framebuffer);
+
+	if (index >= drm_format_num_planes(framebuffer->pixel_format))
+		return NULL;
+
+	return fb->planes[index];
+}
+
+static void tegra_fb_destroy(struct drm_framebuffer *framebuffer)
+{
+	struct tegra_fb *fb = to_tegra_fb(framebuffer);
+	unsigned int i;
+
+	for (i = 0; i < fb->num_planes; i++) {
+		struct tegra_bo *bo = fb->planes[i];
+
+		if (bo)
+			drm_gem_object_unreference_unlocked(&bo->gem);
+	}
+
+	drm_framebuffer_cleanup(framebuffer);
+	kfree(fb->planes);
+	kfree(fb);
+}
+
+static int tegra_fb_create_handle(struct drm_framebuffer *framebuffer,
+				  struct drm_file *file, unsigned int *handle)
+{
+	struct tegra_fb *fb = to_tegra_fb(framebuffer);
+
+	return drm_gem_handle_create(file, &fb->planes[0]->gem, handle);
+}
+
+static struct drm_framebuffer_funcs tegra_fb_funcs = {
+	.destroy = tegra_fb_destroy,
+	.create_handle = tegra_fb_create_handle,
+};
+
+static struct tegra_fb *tegra_fb_alloc(struct drm_device *drm,
+				       struct drm_mode_fb_cmd2 *mode_cmd,
+				       struct tegra_bo **planes,
+				       unsigned int num_planes)
+{
+	struct tegra_fb *fb;
+	unsigned int i;
+	int err;
+
+	fb = kzalloc(sizeof(*fb), GFP_KERNEL);
+	if (!fb)
+		return ERR_PTR(-ENOMEM);
+
+	fb->planes = kzalloc(num_planes * sizeof(*planes), GFP_KERNEL);
+	if (!fb->planes)
+		return ERR_PTR(-ENOMEM);
+
+	fb->num_planes = num_planes;
+
+	drm_helper_mode_fill_fb_struct(&fb->base, mode_cmd);
+
+	for (i = 0; i < fb->num_planes; i++)
+		fb->planes[i] = planes[i];
+
+	err = drm_framebuffer_init(drm, &fb->base, &tegra_fb_funcs);
+	if (err < 0) {
+		dev_err(drm->dev, "failed to initialize framebuffer: %d\n",
+			err);
+		kfree(fb->planes);
+		kfree(fb);
+		return ERR_PTR(err);
+	}
+
+	return fb;
+}
+
+static struct drm_framebuffer *tegra_fb_create(struct drm_device *drm,
+					       struct drm_file *file,
+					       struct drm_mode_fb_cmd2 *cmd)
+{
+	unsigned int hsub, vsub, i;
+	struct tegra_bo *planes[4];
+	struct drm_gem_object *gem;
+	struct tegra_fb *fb;
+	int err;
+
+	hsub = drm_format_horz_chroma_subsampling(cmd->pixel_format);
+	vsub = drm_format_vert_chroma_subsampling(cmd->pixel_format);
+
+	for (i = 0; i < drm_format_num_planes(cmd->pixel_format); i++) {
+		unsigned int width = cmd->width / (i ? hsub : 1);
+		unsigned int height = cmd->height / (i ? vsub : 1);
+		unsigned int size, bpp;
+
+		gem = drm_gem_object_lookup(drm, file, cmd->handles[i]);
+		if (!gem) {
+			err = -ENXIO;
+			goto unreference;
+		}
+
+		bpp = drm_format_plane_cpp(cmd->pixel_format, i);
+
+		size = (height - 1) * cmd->pitches[i] +
+		       width * bpp + cmd->offsets[i];
+
+		if (gem->size < size) {
+			err = -EINVAL;
+			goto unreference;
+		}
+
+		planes[i] = to_tegra_bo(gem);
+	}
+
+	fb = tegra_fb_alloc(drm, cmd, planes, i);
+	if (IS_ERR(fb)) {
+		err = PTR_ERR(fb);
+		goto unreference;
+	}
+
+	return &fb->base;
+
+unreference:
+	while (i--)
+		drm_gem_object_unreference_unlocked(&planes[i]->gem);
+
+	return ERR_PTR(err);
+}
+
+static struct fb_ops tegra_fb_ops = {
+	.owner = THIS_MODULE,
+	.fb_fillrect = sys_fillrect,
+	.fb_copyarea = sys_copyarea,
+	.fb_imageblit = sys_imageblit,
+	.fb_check_var = drm_fb_helper_check_var,
+	.fb_set_par = drm_fb_helper_set_par,
+	.fb_blank = drm_fb_helper_blank,
+	.fb_pan_display = drm_fb_helper_pan_display,
+	.fb_setcmap = drm_fb_helper_setcmap,
+};
+
+static int tegra_fbdev_probe(struct drm_fb_helper *helper,
+			     struct drm_fb_helper_surface_size *sizes)
+{
+	struct tegra_fbdev *fbdev = to_tegra_fbdev(helper);
+	struct drm_device *drm = helper->dev;
+	struct drm_mode_fb_cmd2 cmd = { 0 };
+	unsigned int bytes_per_pixel;
+	struct drm_framebuffer *fb;
+	unsigned long offset;
+	struct fb_info *info;
+	struct tegra_bo *bo;
+	size_t size;
+	int err;
+
+	bytes_per_pixel = DIV_ROUND_UP(sizes->surface_bpp, 8);
+
+	cmd.width = sizes->surface_width;
+	cmd.height = sizes->surface_height;
+	cmd.pitches[0] = sizes->surface_width * bytes_per_pixel;
+	cmd.pixel_format = drm_mode_legacy_fb_format(sizes->surface_bpp,
+						     sizes->surface_depth);
+
+	size = cmd.pitches[0] * cmd.height;
+
+	bo = tegra_bo_create(drm, size);
+	if (IS_ERR(bo))
+		return PTR_ERR(bo);
+
+	info = framebuffer_alloc(0, drm->dev);
+	if (!info) {
+		dev_err(drm->dev, "failed to allocate framebuffer info\n");
+		tegra_bo_free_object(&bo->gem);
+		return -ENOMEM;
+	}
+
+	fbdev->fb = tegra_fb_alloc(drm, &cmd, &bo, 1);
+	if (IS_ERR(fbdev->fb)) {
+		dev_err(drm->dev, "failed to allocate DRM framebuffer\n");
+		err = PTR_ERR(fbdev->fb);
+		goto release;
+	}
+
+	fb = &fbdev->fb->base;
+	helper->fb = fb;
+	helper->fbdev = info;
+
+	info->par = helper;
+	info->flags = FBINFO_FLAG_DEFAULT;
+	info->fbops = &tegra_fb_ops;
+
+	err = fb_alloc_cmap(&info->cmap, 256, 0);
+	if (err < 0) {
+		dev_err(drm->dev, "failed to allocate color map: %d\n", err);
+		goto destroy;
+	}
+
+	drm_fb_helper_fill_fix(info, fb->pitches[0], fb->depth);
+	drm_fb_helper_fill_var(info, helper, fb->width, fb->height);
+
+	offset = info->var.xoffset * bytes_per_pixel +
+		 info->var.yoffset * fb->pitches[0];
+
+	drm->mode_config.fb_base = (resource_size_t)bo->paddr;
+	info->screen_base = bo->vaddr + offset;
+	info->screen_size = size;
+	info->fix.smem_start = (unsigned long)(bo->paddr + offset);
+	info->fix.smem_len = size;
+
+	return 0;
+
+destroy:
+	drm_framebuffer_unregister_private(fb);
+	tegra_fb_destroy(fb);
+release:
+	framebuffer_release(info);
+	return err;
+}
+
+static struct drm_fb_helper_funcs tegra_fb_helper_funcs = {
+	.fb_probe = tegra_fbdev_probe,
+};
+
+static struct tegra_fbdev *tegra_fbdev_create(struct drm_device *drm,
+					      unsigned int preferred_bpp,
+					      unsigned int num_crtc,
+					      unsigned int max_connectors)
+{
+	struct drm_fb_helper *helper;
+	struct tegra_fbdev *fbdev;
+	int err;
+
+	fbdev = kzalloc(sizeof(*fbdev), GFP_KERNEL);
+	if (!fbdev) {
+		dev_err(drm->dev, "failed to allocate DRM fbdev\n");
+		return ERR_PTR(-ENOMEM);
+	}
+
+	fbdev->base.funcs = &tegra_fb_helper_funcs;
+	helper = &fbdev->base;
+
+	err = drm_fb_helper_init(drm, &fbdev->base, num_crtc, max_connectors);
+	if (err < 0) {
+		dev_err(drm->dev, "failed to initialize DRM FB helper\n");
+		goto free;
+	}
+
+	err = drm_fb_helper_single_add_all_connectors(&fbdev->base);
+	if (err < 0) {
+		dev_err(drm->dev, "failed to add connectors\n");
+		goto fini;
+	}
+
+	drm_helper_disable_unused_functions(drm);
+
+	err = drm_fb_helper_initial_config(&fbdev->base, preferred_bpp);
+	if (err < 0) {
+		dev_err(drm->dev, "failed to set initial configuration\n");
+		goto fini;
+	}
+
+	return fbdev;
+
+fini:
+	drm_fb_helper_fini(&fbdev->base);
+free:
+	kfree(fbdev);
+	return ERR_PTR(err);
+}
+
+static void tegra_fbdev_free(struct tegra_fbdev *fbdev)
+{
+	struct fb_info *info = fbdev->base.fbdev;
+
+	if (info) {
+		int err;
+
+		err = unregister_framebuffer(info);
+		if (err < 0)
+			DRM_DEBUG_KMS("failed to unregister framebuffer\n");
+
+		if (info->cmap.len)
+			fb_dealloc_cmap(&info->cmap);
+
+		framebuffer_release(info);
+	}
+
+	if (fbdev->fb) {
+		drm_framebuffer_unregister_private(&fbdev->fb->base);
+		tegra_fb_destroy(&fbdev->fb->base);
+	}
+
+	drm_fb_helper_fini(&fbdev->base);
+	kfree(fbdev);
+}
+
+static void tegra_fb_output_poll_changed(struct drm_device *drm)
+{
+	struct host1x_drm *host1x = drm->dev_private;
+
+	if (host1x->fbdev)
+		drm_fb_helper_hotplug_event(&host1x->fbdev->base);
+}
+
+static const struct drm_mode_config_funcs tegra_drm_mode_funcs = {
+	.fb_create = tegra_fb_create,
+	.output_poll_changed = tegra_fb_output_poll_changed,
+};
+
+int tegra_drm_fb_init(struct drm_device *drm)
+{
+	struct host1x_drm *host1x = drm->dev_private;
+	struct tegra_fbdev *fbdev;
+
+	drm->mode_config.min_width = 0;
+	drm->mode_config.min_height = 0;
+
+	drm->mode_config.max_width = 4096;
+	drm->mode_config.max_height = 4096;
+
+	drm->mode_config.funcs = &tegra_drm_mode_funcs;
+
+	fbdev = tegra_fbdev_create(drm, 32, drm->mode_config.num_crtc,
+				   drm->mode_config.num_connector);
+	if (IS_ERR(fbdev))
+		return PTR_ERR(fbdev);
+
+	host1x->fbdev = fbdev;
+
+	return 0;
+}
+
+void tegra_drm_fb_exit(struct drm_device *drm)
+{
+	struct host1x_drm *host1x = drm->dev_private;
+
+	tegra_fbdev_free(host1x->fbdev);
+}
+
+void tegra_fbdev_restore_mode(struct tegra_fbdev *fbdev)
+{
+	if (fbdev) {
+		drm_modeset_lock_all(fbdev->base.dev);
+		drm_fb_helper_restore_fbdev_mode(&fbdev->base);
+		drm_modeset_unlock_all(fbdev->base.dev);
+	}
+}
diff --git a/drivers/gpu/host1x/drm/gem.c b/drivers/gpu/host1x/drm/gem.c
new file mode 100644
index 00000000000..c5e9a9b494c
--- /dev/null
+++ b/drivers/gpu/host1x/drm/gem.c
@@ -0,0 +1,270 @@
+/*
+ * NVIDIA Tegra DRM GEM helper functions
+ *
+ * Copyright (C) 2012 Sascha Hauer, Pengutronix
+ * Copyright (C) 2013 NVIDIA CORPORATION, All rights reserved.
+ *
+ * Based on the GEM/CMA helpers
+ *
+ * Copyright (c) 2011 Samsung Electronics Co., Ltd.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/mm.h>
+#include <linux/slab.h>
+#include <linux/mutex.h>
+#include <linux/export.h>
+#include <linux/dma-mapping.h>
+
+#include <drm/drmP.h>
+#include <drm/drm.h>
+
+#include "gem.h"
+
+static inline struct tegra_bo *host1x_to_drm_bo(struct host1x_bo *bo)
+{
+	return container_of(bo, struct tegra_bo, base);
+}
+
+static void tegra_bo_put(struct host1x_bo *bo)
+{
+	struct tegra_bo *obj = host1x_to_drm_bo(bo);
+	struct drm_device *drm = obj->gem.dev;
+
+	mutex_lock(&drm->struct_mutex);
+	drm_gem_object_unreference(&obj->gem);
+	mutex_unlock(&drm->struct_mutex);
+}
+
+static dma_addr_t tegra_bo_pin(struct host1x_bo *bo, struct sg_table **sgt)
+{
+	struct tegra_bo *obj = host1x_to_drm_bo(bo);
+
+	return obj->paddr;
+}
+
+static void tegra_bo_unpin(struct host1x_bo *bo, struct sg_table *sgt)
+{
+}
+
+static void *tegra_bo_mmap(struct host1x_bo *bo)
+{
+	struct tegra_bo *obj = host1x_to_drm_bo(bo);
+
+	return obj->vaddr;
+}
+
+static void tegra_bo_munmap(struct host1x_bo *bo, void *addr)
+{
+}
+
+static void *tegra_bo_kmap(struct host1x_bo *bo, unsigned int page)
+{
+	struct tegra_bo *obj = host1x_to_drm_bo(bo);
+
+	return obj->vaddr + page * PAGE_SIZE;
+}
+
+static void tegra_bo_kunmap(struct host1x_bo *bo, unsigned int page,
+			    void *addr)
+{
+}
+
+static struct host1x_bo *tegra_bo_get(struct host1x_bo *bo)
+{
+	struct tegra_bo *obj = host1x_to_drm_bo(bo);
+	struct drm_device *drm = obj->gem.dev;
+
+	mutex_lock(&drm->struct_mutex);
+	drm_gem_object_reference(&obj->gem);
+	mutex_unlock(&drm->struct_mutex);
+
+	return bo;
+}
+
+const struct host1x_bo_ops tegra_bo_ops = {
+	.get = tegra_bo_get,
+	.put = tegra_bo_put,
+	.pin = tegra_bo_pin,
+	.unpin = tegra_bo_unpin,
+	.mmap = tegra_bo_mmap,
+	.munmap = tegra_bo_munmap,
+	.kmap = tegra_bo_kmap,
+	.kunmap = tegra_bo_kunmap,
+};
+
+static void tegra_bo_destroy(struct drm_device *drm, struct tegra_bo *bo)
+{
+	dma_free_writecombine(drm->dev, bo->gem.size, bo->vaddr, bo->paddr);
+}
+
+unsigned int tegra_bo_get_mmap_offset(struct tegra_bo *bo)
+{
+	return (unsigned int)bo->gem.map_list.hash.key << PAGE_SHIFT;
+}
+
+struct tegra_bo *tegra_bo_create(struct drm_device *drm, unsigned int size)
+{
+	struct tegra_bo *bo;
+	int err;
+
+	bo = kzalloc(sizeof(*bo), GFP_KERNEL);
+	if (!bo)
+		return ERR_PTR(-ENOMEM);
+
+	host1x_bo_init(&bo->base, &tegra_bo_ops);
+	size = round_up(size, PAGE_SIZE);
+
+	bo->vaddr = dma_alloc_writecombine(drm->dev, size, &bo->paddr,
+					   GFP_KERNEL | __GFP_NOWARN);
+	if (!bo->vaddr) {
+		dev_err(drm->dev, "failed to allocate buffer with size %u\n",
+			size);
+		err = -ENOMEM;
+		goto err_dma;
+	}
+
+	err = drm_gem_object_init(drm, &bo->gem, size);
+	if (err)
+		goto err_init;
+
+	err = drm_gem_create_mmap_offset(&bo->gem);
+	if (err)
+		goto err_mmap;
+
+	return bo;
+
+err_mmap:
+	drm_gem_object_release(&bo->gem);
+err_init:
+	tegra_bo_destroy(drm, bo);
+err_dma:
+	kfree(bo);
+
+	return ERR_PTR(err);
+
+}
+
+struct tegra_bo *tegra_bo_create_with_handle(struct drm_file *file,
+					    struct drm_device *drm,
+					    unsigned int size,
+					    unsigned int *handle)
+{
+	struct tegra_bo *bo;
+	int ret;
+
+	bo = tegra_bo_create(drm, size);
+	if (IS_ERR(bo))
+		return bo;
+
+	ret = drm_gem_handle_create(file, &bo->gem, handle);
+	if (ret)
+		goto err;
+
+	drm_gem_object_unreference_unlocked(&bo->gem);
+
+	return bo;
+
+err:
+	tegra_bo_free_object(&bo->gem);
+	return ERR_PTR(ret);
+}
+
+void tegra_bo_free_object(struct drm_gem_object *gem)
+{
+	struct tegra_bo *bo = to_tegra_bo(gem);
+
+	if (gem->map_list.map)
+		drm_gem_free_mmap_offset(gem);
+
+	drm_gem_object_release(gem);
+	tegra_bo_destroy(gem->dev, bo);
+
+	kfree(bo);
+}
+
+int tegra_bo_dumb_create(struct drm_file *file, struct drm_device *drm,
+			 struct drm_mode_create_dumb *args)
+{
+	int min_pitch = DIV_ROUND_UP(args->width * args->bpp, 8);
+	struct tegra_bo *bo;
+
+	if (args->pitch < min_pitch)
+		args->pitch = min_pitch;
+
+	if (args->size < args->pitch * args->height)
+		args->size = args->pitch * args->height;
+
+	bo = tegra_bo_create_with_handle(file, drm, args->size,
+					    &args->handle);
+	if (IS_ERR(bo))
+		return PTR_ERR(bo);
+
+	return 0;
+}
+
+int tegra_bo_dumb_map_offset(struct drm_file *file, struct drm_device *drm,
+			     uint32_t handle, uint64_t *offset)
+{
+	struct drm_gem_object *gem;
+	struct tegra_bo *bo;
+
+	mutex_lock(&drm->struct_mutex);
+
+	gem = drm_gem_object_lookup(drm, file, handle);
+	if (!gem) {
+		dev_err(drm->dev, "failed to lookup GEM object\n");
+		mutex_unlock(&drm->struct_mutex);
+		return -EINVAL;
+	}
+
+	bo = to_tegra_bo(gem);
+
+	*offset = tegra_bo_get_mmap_offset(bo);
+
+	drm_gem_object_unreference(gem);
+
+	mutex_unlock(&drm->struct_mutex);
+
+	return 0;
+}
+
+const struct vm_operations_struct tegra_bo_vm_ops = {
+	.open = drm_gem_vm_open,
+	.close = drm_gem_vm_close,
+};
+
+int tegra_drm_mmap(struct file *file, struct vm_area_struct *vma)
+{
+	struct drm_gem_object *gem;
+	struct tegra_bo *bo;
+	int ret;
+
+	ret = drm_gem_mmap(file, vma);
+	if (ret)
+		return ret;
+
+	gem = vma->vm_private_data;
+	bo = to_tegra_bo(gem);
+
+	ret = remap_pfn_range(vma, vma->vm_start, bo->paddr >> PAGE_SHIFT,
+			      vma->vm_end - vma->vm_start, vma->vm_page_prot);
+	if (ret)
+		drm_gem_vm_close(vma);
+
+	return ret;
+}
+
+int tegra_bo_dumb_destroy(struct drm_file *file, struct drm_device *drm,
+			  unsigned int handle)
+{
+	return drm_gem_handle_delete(file, handle);
+}
diff --git a/drivers/gpu/host1x/drm/gem.h b/drivers/gpu/host1x/drm/gem.h
new file mode 100644
index 00000000000..34de2b486eb
--- /dev/null
+++ b/drivers/gpu/host1x/drm/gem.h
@@ -0,0 +1,59 @@
+/*
+ * Tegra host1x GEM implementation
+ *
+ * Copyright (c) 2012-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __HOST1X_GEM_H
+#define __HOST1X_GEM_H
+
+#include <drm/drm.h>
+#include <drm/drmP.h>
+
+#include "host1x_bo.h"
+
+struct tegra_bo {
+	struct drm_gem_object gem;
+	struct host1x_bo base;
+	dma_addr_t paddr;
+	void *vaddr;
+};
+
+static inline struct tegra_bo *to_tegra_bo(struct drm_gem_object *gem)
+{
+	return container_of(gem, struct tegra_bo, gem);
+}
+
+extern const struct host1x_bo_ops tegra_bo_ops;
+
+struct tegra_bo *tegra_bo_create(struct drm_device *drm, unsigned int size);
+struct tegra_bo *tegra_bo_create_with_handle(struct drm_file *file,
+					    struct drm_device *drm,
+					    unsigned int size,
+					    unsigned int *handle);
+void tegra_bo_free_object(struct drm_gem_object *gem);
+unsigned int tegra_bo_get_mmap_offset(struct tegra_bo *bo);
+int tegra_bo_dumb_create(struct drm_file *file, struct drm_device *drm,
+			 struct drm_mode_create_dumb *args);
+int tegra_bo_dumb_map_offset(struct drm_file *file, struct drm_device *drm,
+			     uint32_t handle, uint64_t *offset);
+int tegra_bo_dumb_destroy(struct drm_file *file, struct drm_device *drm,
+			  unsigned int handle);
+
+int tegra_drm_mmap(struct file *file, struct vm_area_struct *vma);
+
+extern const struct vm_operations_struct tegra_bo_vm_ops;
+
+#endif
diff --git a/drivers/gpu/host1x/drm/gr2d.c b/drivers/gpu/host1x/drm/gr2d.c
new file mode 100644
index 00000000000..6a45ae090ee
--- /dev/null
+++ b/drivers/gpu/host1x/drm/gr2d.c
@@ -0,0 +1,339 @@
+/*
+ * drivers/video/tegra/host/gr2d/gr2d.c
+ *
+ * Tegra Graphics 2D
+ *
+ * Copyright (c) 2012-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/export.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/clk.h>
+
+#include "channel.h"
+#include "drm.h"
+#include "gem.h"
+#include "job.h"
+#include "host1x.h"
+#include "host1x_bo.h"
+#include "host1x_client.h"
+#include "syncpt.h"
+
+struct gr2d {
+	struct host1x_client client;
+	struct clk *clk;
+	struct host1x_channel *channel;
+	unsigned long *addr_regs;
+};
+
+static inline struct gr2d *to_gr2d(struct host1x_client *client)
+{
+	return container_of(client, struct gr2d, client);
+}
+
+static int gr2d_is_addr_reg(struct device *dev, u32 class, u32 reg);
+
+static int gr2d_client_init(struct host1x_client *client,
+			    struct drm_device *drm)
+{
+	return 0;
+}
+
+static int gr2d_client_exit(struct host1x_client *client)
+{
+	return 0;
+}
+
+static int gr2d_open_channel(struct host1x_client *client,
+			     struct host1x_drm_context *context)
+{
+	struct gr2d *gr2d = to_gr2d(client);
+
+	context->channel = host1x_channel_get(gr2d->channel);
+
+	if (!context->channel)
+		return -ENOMEM;
+
+	return 0;
+}
+
+static void gr2d_close_channel(struct host1x_drm_context *context)
+{
+	host1x_channel_put(context->channel);
+}
+
+static struct host1x_bo *host1x_bo_lookup(struct drm_device *drm,
+					  struct drm_file *file,
+					  u32 handle)
+{
+	struct drm_gem_object *gem;
+	struct tegra_bo *bo;
+
+	gem = drm_gem_object_lookup(drm, file, handle);
+	if (!gem)
+		return 0;
+
+	mutex_lock(&drm->struct_mutex);
+	drm_gem_object_unreference(gem);
+	mutex_unlock(&drm->struct_mutex);
+
+	bo = to_tegra_bo(gem);
+	return &bo->base;
+}
+
+static int gr2d_submit(struct host1x_drm_context *context,
+		       struct drm_tegra_submit *args, struct drm_device *drm,
+		       struct drm_file *file)
+{
+	struct host1x_job *job;
+	unsigned int num_cmdbufs = args->num_cmdbufs;
+	unsigned int num_relocs = args->num_relocs;
+	unsigned int num_waitchks = args->num_waitchks;
+	struct drm_tegra_cmdbuf __user *cmdbufs =
+		(void * __user)(uintptr_t)args->cmdbufs;
+	struct drm_tegra_reloc __user *relocs =
+		(void * __user)(uintptr_t)args->relocs;
+	struct drm_tegra_waitchk __user *waitchks =
+		(void * __user)(uintptr_t)args->waitchks;
+	struct drm_tegra_syncpt syncpt;
+	int err;
+
+	/* We don't yet support other than one syncpt_incr struct per submit */
+	if (args->num_syncpts != 1)
+		return -EINVAL;
+
+	job = host1x_job_alloc(context->channel, args->num_cmdbufs,
+			       args->num_relocs, args->num_waitchks);
+	if (!job)
+		return -ENOMEM;
+
+	job->num_relocs = args->num_relocs;
+	job->num_waitchk = args->num_waitchks;
+	job->client = (u32)args->context;
+	job->class = context->client->class;
+	job->serialize = true;
+
+	while (num_cmdbufs) {
+		struct drm_tegra_cmdbuf cmdbuf;
+		struct host1x_bo *bo;
+
+		err = copy_from_user(&cmdbuf, cmdbufs, sizeof(cmdbuf));
+		if (err)
+			goto fail;
+
+		bo = host1x_bo_lookup(drm, file, cmdbuf.handle);
+		if (!bo)
+			goto fail;
+
+		host1x_job_add_gather(job, bo, cmdbuf.words, cmdbuf.offset);
+		num_cmdbufs--;
+		cmdbufs++;
+	}
+
+	err = copy_from_user(job->relocarray, relocs,
+			     sizeof(*relocs) * num_relocs);
+	if (err)
+		goto fail;
+
+	while (num_relocs--) {
+		struct host1x_reloc *reloc = &job->relocarray[num_relocs];
+		struct host1x_bo *cmdbuf, *target;
+
+		cmdbuf = host1x_bo_lookup(drm, file, (u32)reloc->cmdbuf);
+		target = host1x_bo_lookup(drm, file, (u32)reloc->target);
+
+		reloc->cmdbuf = cmdbuf;
+		reloc->target = target;
+
+		if (!reloc->target || !reloc->cmdbuf)
+			goto fail;
+	}
+
+	err = copy_from_user(job->waitchk, waitchks,
+			     sizeof(*waitchks) * num_waitchks);
+	if (err)
+		goto fail;
+
+	err = copy_from_user(&syncpt, (void * __user)(uintptr_t)args->syncpts,
+			     sizeof(syncpt));
+	if (err)
+		goto fail;
+
+	job->syncpt_id = syncpt.id;
+	job->syncpt_incrs = syncpt.incrs;
+	job->timeout = 10000;
+	job->is_addr_reg = gr2d_is_addr_reg;
+
+	if (args->timeout && args->timeout < 10000)
+		job->timeout = args->timeout;
+
+	err = host1x_job_pin(job, context->client->dev);
+	if (err)
+		goto fail;
+
+	err = host1x_job_submit(job);
+	if (err)
+		goto fail_submit;
+
+	args->fence = job->syncpt_end;
+
+	host1x_job_put(job);
+	return 0;
+
+fail_submit:
+	host1x_job_unpin(job);
+fail:
+	host1x_job_put(job);
+	return err;
+}
+
+static struct host1x_client_ops gr2d_client_ops = {
+	.drm_init = gr2d_client_init,
+	.drm_exit = gr2d_client_exit,
+	.open_channel = gr2d_open_channel,
+	.close_channel = gr2d_close_channel,
+	.submit = gr2d_submit,
+};
+
+static void gr2d_init_addr_reg_map(struct device *dev, struct gr2d *gr2d)
+{
+	const u32 gr2d_addr_regs[] = {0x1a, 0x1b, 0x26, 0x2b, 0x2c, 0x2d, 0x31,
+				      0x32, 0x48, 0x49, 0x4a, 0x4b, 0x4c};
+	unsigned long *bitmap;
+	int i;
+
+	bitmap = devm_kzalloc(dev, DIV_ROUND_UP(256, BITS_PER_BYTE),
+			      GFP_KERNEL);
+
+	for (i = 0; i < ARRAY_SIZE(gr2d_addr_regs); ++i) {
+		u32 reg = gr2d_addr_regs[i];
+		bitmap[BIT_WORD(reg)] |= BIT_MASK(reg);
+	}
+
+	gr2d->addr_regs = bitmap;
+}
+
+static int gr2d_is_addr_reg(struct device *dev, u32 class, u32 reg)
+{
+	struct gr2d *gr2d = dev_get_drvdata(dev);
+
+	switch (class) {
+	case HOST1X_CLASS_HOST1X:
+		return reg == 0x2b;
+	case HOST1X_CLASS_GR2D:
+	case HOST1X_CLASS_GR2D_SB:
+		reg &= 0xff;
+		if (gr2d->addr_regs[BIT_WORD(reg)] & BIT_MASK(reg))
+			return 1;
+	default:
+		return 0;
+	}
+}
+
+static const struct of_device_id gr2d_match[] = {
+	{ .compatible = "nvidia,tegra30-gr2d" },
+	{ .compatible = "nvidia,tegra20-gr2d" },
+	{ },
+};
+
+static int gr2d_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct host1x_drm *host1x = host1x_get_drm_data(dev->parent);
+	int err;
+	struct gr2d *gr2d = NULL;
+	struct host1x_syncpt **syncpts;
+
+	gr2d = devm_kzalloc(dev, sizeof(*gr2d), GFP_KERNEL);
+	if (!gr2d)
+		return -ENOMEM;
+
+	syncpts = devm_kzalloc(dev, sizeof(*syncpts), GFP_KERNEL);
+	if (!syncpts)
+		return -ENOMEM;
+
+	gr2d->clk = devm_clk_get(dev, NULL);
+	if (IS_ERR(gr2d->clk)) {
+		dev_err(dev, "cannot get clock\n");
+		return PTR_ERR(gr2d->clk);
+	}
+
+	err = clk_prepare_enable(gr2d->clk);
+	if (err) {
+		dev_err(dev, "cannot turn on clock\n");
+		return err;
+	}
+
+	gr2d->channel = host1x_channel_request(dev);
+	if (!gr2d->channel)
+		return -ENOMEM;
+
+	*syncpts = host1x_syncpt_request(dev, 0);
+	if (!(*syncpts)) {
+		host1x_channel_free(gr2d->channel);
+		return -ENOMEM;
+	}
+
+	gr2d->client.ops = &gr2d_client_ops;
+	gr2d->client.dev = dev;
+	gr2d->client.class = HOST1X_CLASS_GR2D;
+	gr2d->client.syncpts = syncpts;
+	gr2d->client.num_syncpts = 1;
+
+	err = host1x_register_client(host1x, &gr2d->client);
+	if (err < 0) {
+		dev_err(dev, "failed to register host1x client: %d\n", err);
+		return err;
+	}
+
+	gr2d_init_addr_reg_map(dev, gr2d);
+
+	platform_set_drvdata(pdev, gr2d);
+
+	return 0;
+}
+
+static int __exit gr2d_remove(struct platform_device *pdev)
+{
+	struct host1x_drm *host1x = host1x_get_drm_data(pdev->dev.parent);
+	struct gr2d *gr2d = platform_get_drvdata(pdev);
+	unsigned int i;
+	int err;
+
+	err = host1x_unregister_client(host1x, &gr2d->client);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to unregister client: %d\n", err);
+		return err;
+	}
+
+	for (i = 0; i < gr2d->client.num_syncpts; i++)
+		host1x_syncpt_free(gr2d->client.syncpts[i]);
+
+	host1x_channel_free(gr2d->channel);
+	clk_disable_unprepare(gr2d->clk);
+
+	return 0;
+}
+
+struct platform_driver tegra_gr2d_driver = {
+	.probe = gr2d_probe,
+	.remove = __exit_p(gr2d_remove),
+	.driver = {
+		.owner = THIS_MODULE,
+		.name = "gr2d",
+		.of_match_table = gr2d_match,
+	}
+};
diff --git a/drivers/gpu/host1x/drm/hdmi.c b/drivers/gpu/host1x/drm/hdmi.c
new file mode 100644
index 00000000000..01097da09f7
--- /dev/null
+++ b/drivers/gpu/host1x/drm/hdmi.c
@@ -0,0 +1,1313 @@
+/*
+ * Copyright (C) 2012 Avionic Design GmbH
+ * Copyright (C) 2012 NVIDIA CORPORATION.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/clk.h>
+#include <linux/debugfs.h>
+#include <linux/gpio.h>
+#include <linux/hdmi.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/regulator/consumer.h>
+#include <linux/clk/tegra.h>
+
+#include <drm/drm_edid.h>
+
+#include "hdmi.h"
+#include "drm.h"
+#include "dc.h"
+#include "host1x_client.h"
+
+struct tegra_hdmi {
+	struct host1x_client client;
+	struct tegra_output output;
+	struct device *dev;
+
+	struct regulator *vdd;
+	struct regulator *pll;
+
+	void __iomem *regs;
+	unsigned int irq;
+
+	struct clk *clk_parent;
+	struct clk *clk;
+
+	unsigned int audio_source;
+	unsigned int audio_freq;
+	bool stereo;
+	bool dvi;
+
+	struct drm_info_list *debugfs_files;
+	struct drm_minor *minor;
+	struct dentry *debugfs;
+};
+
+static inline struct tegra_hdmi *
+host1x_client_to_hdmi(struct host1x_client *client)
+{
+	return container_of(client, struct tegra_hdmi, client);
+}
+
+static inline struct tegra_hdmi *to_hdmi(struct tegra_output *output)
+{
+	return container_of(output, struct tegra_hdmi, output);
+}
+
+#define HDMI_AUDIOCLK_FREQ 216000000
+#define HDMI_REKEY_DEFAULT 56
+
+enum {
+	AUTO = 0,
+	SPDIF,
+	HDA,
+};
+
+static inline unsigned long tegra_hdmi_readl(struct tegra_hdmi *hdmi,
+					     unsigned long reg)
+{
+	return readl(hdmi->regs + (reg << 2));
+}
+
+static inline void tegra_hdmi_writel(struct tegra_hdmi *hdmi, unsigned long val,
+				     unsigned long reg)
+{
+	writel(val, hdmi->regs + (reg << 2));
+}
+
+struct tegra_hdmi_audio_config {
+	unsigned int pclk;
+	unsigned int n;
+	unsigned int cts;
+	unsigned int aval;
+};
+
+static const struct tegra_hdmi_audio_config tegra_hdmi_audio_32k[] = {
+	{  25200000, 4096,  25200, 24000 },
+	{  27000000, 4096,  27000, 24000 },
+	{  74250000, 4096,  74250, 24000 },
+	{ 148500000, 4096, 148500, 24000 },
+	{         0,    0,      0,     0 },
+};
+
+static const struct tegra_hdmi_audio_config tegra_hdmi_audio_44_1k[] = {
+	{  25200000, 5880,  26250, 25000 },
+	{  27000000, 5880,  28125, 25000 },
+	{  74250000, 4704,  61875, 20000 },
+	{ 148500000, 4704, 123750, 20000 },
+	{         0,    0,      0,     0 },
+};
+
+static const struct tegra_hdmi_audio_config tegra_hdmi_audio_48k[] = {
+	{  25200000, 6144,  25200, 24000 },
+	{  27000000, 6144,  27000, 24000 },
+	{  74250000, 6144,  74250, 24000 },
+	{ 148500000, 6144, 148500, 24000 },
+	{         0,    0,      0,     0 },
+};
+
+static const struct tegra_hdmi_audio_config tegra_hdmi_audio_88_2k[] = {
+	{  25200000, 11760,  26250, 25000 },
+	{  27000000, 11760,  28125, 25000 },
+	{  74250000,  9408,  61875, 20000 },
+	{ 148500000,  9408, 123750, 20000 },
+	{         0,     0,      0,     0 },
+};
+
+static const struct tegra_hdmi_audio_config tegra_hdmi_audio_96k[] = {
+	{  25200000, 12288,  25200, 24000 },
+	{  27000000, 12288,  27000, 24000 },
+	{  74250000, 12288,  74250, 24000 },
+	{ 148500000, 12288, 148500, 24000 },
+	{         0,     0,      0,     0 },
+};
+
+static const struct tegra_hdmi_audio_config tegra_hdmi_audio_176_4k[] = {
+	{  25200000, 23520,  26250, 25000 },
+	{  27000000, 23520,  28125, 25000 },
+	{  74250000, 18816,  61875, 20000 },
+	{ 148500000, 18816, 123750, 20000 },
+	{         0,     0,      0,     0 },
+};
+
+static const struct tegra_hdmi_audio_config tegra_hdmi_audio_192k[] = {
+	{  25200000, 24576,  25200, 24000 },
+	{  27000000, 24576,  27000, 24000 },
+	{  74250000, 24576,  74250, 24000 },
+	{ 148500000, 24576, 148500, 24000 },
+	{         0,     0,      0,     0 },
+};
+
+struct tmds_config {
+	unsigned int pclk;
+	u32 pll0;
+	u32 pll1;
+	u32 pe_current;
+	u32 drive_current;
+};
+
+static const struct tmds_config tegra2_tmds_config[] = {
+	{ /* slow pixel clock modes */
+		.pclk = 27000000,
+		.pll0 = SOR_PLL_BG_V17_S(3) | SOR_PLL_ICHPMP(1) |
+			SOR_PLL_RESISTORSEL | SOR_PLL_VCOCAP(0) |
+			SOR_PLL_TX_REG_LOAD(3),
+		.pll1 = SOR_PLL_TMDS_TERM_ENABLE,
+		.pe_current = PE_CURRENT0(PE_CURRENT_0_0_mA) |
+			PE_CURRENT1(PE_CURRENT_0_0_mA) |
+			PE_CURRENT2(PE_CURRENT_0_0_mA) |
+			PE_CURRENT3(PE_CURRENT_0_0_mA),
+		.drive_current = DRIVE_CURRENT_LANE0(DRIVE_CURRENT_7_125_mA) |
+			DRIVE_CURRENT_LANE1(DRIVE_CURRENT_7_125_mA) |
+			DRIVE_CURRENT_LANE2(DRIVE_CURRENT_7_125_mA) |
+			DRIVE_CURRENT_LANE3(DRIVE_CURRENT_7_125_mA),
+	},
+	{ /* high pixel clock modes */
+		.pclk = UINT_MAX,
+		.pll0 = SOR_PLL_BG_V17_S(3) | SOR_PLL_ICHPMP(1) |
+			SOR_PLL_RESISTORSEL | SOR_PLL_VCOCAP(1) |
+			SOR_PLL_TX_REG_LOAD(3),
+		.pll1 = SOR_PLL_TMDS_TERM_ENABLE | SOR_PLL_PE_EN,
+		.pe_current = PE_CURRENT0(PE_CURRENT_6_0_mA) |
+			PE_CURRENT1(PE_CURRENT_6_0_mA) |
+			PE_CURRENT2(PE_CURRENT_6_0_mA) |
+			PE_CURRENT3(PE_CURRENT_6_0_mA),
+		.drive_current = DRIVE_CURRENT_LANE0(DRIVE_CURRENT_7_125_mA) |
+			DRIVE_CURRENT_LANE1(DRIVE_CURRENT_7_125_mA) |
+			DRIVE_CURRENT_LANE2(DRIVE_CURRENT_7_125_mA) |
+			DRIVE_CURRENT_LANE3(DRIVE_CURRENT_7_125_mA),
+	},
+};
+
+static const struct tmds_config tegra3_tmds_config[] = {
+	{ /* 480p modes */
+		.pclk = 27000000,
+		.pll0 = SOR_PLL_BG_V17_S(3) | SOR_PLL_ICHPMP(1) |
+			SOR_PLL_RESISTORSEL | SOR_PLL_VCOCAP(0) |
+			SOR_PLL_TX_REG_LOAD(0),
+		.pll1 = SOR_PLL_TMDS_TERM_ENABLE,
+		.pe_current = PE_CURRENT0(PE_CURRENT_0_0_mA) |
+			PE_CURRENT1(PE_CURRENT_0_0_mA) |
+			PE_CURRENT2(PE_CURRENT_0_0_mA) |
+			PE_CURRENT3(PE_CURRENT_0_0_mA),
+		.drive_current = DRIVE_CURRENT_LANE0(DRIVE_CURRENT_5_250_mA) |
+			DRIVE_CURRENT_LANE1(DRIVE_CURRENT_5_250_mA) |
+			DRIVE_CURRENT_LANE2(DRIVE_CURRENT_5_250_mA) |
+			DRIVE_CURRENT_LANE3(DRIVE_CURRENT_5_250_mA),
+	}, { /* 720p modes */
+		.pclk = 74250000,
+		.pll0 = SOR_PLL_BG_V17_S(3) | SOR_PLL_ICHPMP(1) |
+			SOR_PLL_RESISTORSEL | SOR_PLL_VCOCAP(1) |
+			SOR_PLL_TX_REG_LOAD(0),
+		.pll1 = SOR_PLL_TMDS_TERM_ENABLE | SOR_PLL_PE_EN,
+		.pe_current = PE_CURRENT0(PE_CURRENT_5_0_mA) |
+			PE_CURRENT1(PE_CURRENT_5_0_mA) |
+			PE_CURRENT2(PE_CURRENT_5_0_mA) |
+			PE_CURRENT3(PE_CURRENT_5_0_mA),
+		.drive_current = DRIVE_CURRENT_LANE0(DRIVE_CURRENT_5_250_mA) |
+			DRIVE_CURRENT_LANE1(DRIVE_CURRENT_5_250_mA) |
+			DRIVE_CURRENT_LANE2(DRIVE_CURRENT_5_250_mA) |
+			DRIVE_CURRENT_LANE3(DRIVE_CURRENT_5_250_mA),
+	}, { /* 1080p modes */
+		.pclk = UINT_MAX,
+		.pll0 = SOR_PLL_BG_V17_S(3) | SOR_PLL_ICHPMP(1) |
+			SOR_PLL_RESISTORSEL | SOR_PLL_VCOCAP(3) |
+			SOR_PLL_TX_REG_LOAD(0),
+		.pll1 = SOR_PLL_TMDS_TERM_ENABLE | SOR_PLL_PE_EN,
+		.pe_current = PE_CURRENT0(PE_CURRENT_5_0_mA) |
+			PE_CURRENT1(PE_CURRENT_5_0_mA) |
+			PE_CURRENT2(PE_CURRENT_5_0_mA) |
+			PE_CURRENT3(PE_CURRENT_5_0_mA),
+		.drive_current = DRIVE_CURRENT_LANE0(DRIVE_CURRENT_5_250_mA) |
+			DRIVE_CURRENT_LANE1(DRIVE_CURRENT_5_250_mA) |
+			DRIVE_CURRENT_LANE2(DRIVE_CURRENT_5_250_mA) |
+			DRIVE_CURRENT_LANE3(DRIVE_CURRENT_5_250_mA),
+	},
+};
+
+static const struct tegra_hdmi_audio_config *
+tegra_hdmi_get_audio_config(unsigned int audio_freq, unsigned int pclk)
+{
+	const struct tegra_hdmi_audio_config *table;
+
+	switch (audio_freq) {
+	case 32000:
+		table = tegra_hdmi_audio_32k;
+		break;
+
+	case 44100:
+		table = tegra_hdmi_audio_44_1k;
+		break;
+
+	case 48000:
+		table = tegra_hdmi_audio_48k;
+		break;
+
+	case 88200:
+		table = tegra_hdmi_audio_88_2k;
+		break;
+
+	case 96000:
+		table = tegra_hdmi_audio_96k;
+		break;
+
+	case 176400:
+		table = tegra_hdmi_audio_176_4k;
+		break;
+
+	case 192000:
+		table = tegra_hdmi_audio_192k;
+		break;
+
+	default:
+		return NULL;
+	}
+
+	while (table->pclk) {
+		if (table->pclk == pclk)
+			return table;
+
+		table++;
+	}
+
+	return NULL;
+}
+
+static void tegra_hdmi_setup_audio_fs_tables(struct tegra_hdmi *hdmi)
+{
+	const unsigned int freqs[] = {
+		32000, 44100, 48000, 88200, 96000, 176400, 192000
+	};
+	unsigned int i;
+
+	for (i = 0; i < ARRAY_SIZE(freqs); i++) {
+		unsigned int f = freqs[i];
+		unsigned int eight_half;
+		unsigned long value;
+		unsigned int delta;
+
+		if (f > 96000)
+			delta = 2;
+		else if (f > 480000)
+			delta = 6;
+		else
+			delta = 9;
+
+		eight_half = (8 * HDMI_AUDIOCLK_FREQ) / (f * 128);
+		value = AUDIO_FS_LOW(eight_half - delta) |
+			AUDIO_FS_HIGH(eight_half + delta);
+		tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_AUDIO_FS(i));
+	}
+}
+
+static int tegra_hdmi_setup_audio(struct tegra_hdmi *hdmi, unsigned int pclk)
+{
+	struct device_node *node = hdmi->dev->of_node;
+	const struct tegra_hdmi_audio_config *config;
+	unsigned int offset = 0;
+	unsigned long value;
+
+	switch (hdmi->audio_source) {
+	case HDA:
+		value = AUDIO_CNTRL0_SOURCE_SELECT_HDAL;
+		break;
+
+	case SPDIF:
+		value = AUDIO_CNTRL0_SOURCE_SELECT_SPDIF;
+		break;
+
+	default:
+		value = AUDIO_CNTRL0_SOURCE_SELECT_AUTO;
+		break;
+	}
+
+	if (of_device_is_compatible(node, "nvidia,tegra30-hdmi")) {
+		value |= AUDIO_CNTRL0_ERROR_TOLERANCE(6) |
+			 AUDIO_CNTRL0_FRAMES_PER_BLOCK(0xc0);
+		tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_AUDIO_CNTRL0);
+	} else {
+		value |= AUDIO_CNTRL0_INJECT_NULLSMPL;
+		tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_AUDIO_CNTRL0);
+
+		value = AUDIO_CNTRL0_ERROR_TOLERANCE(6) |
+			AUDIO_CNTRL0_FRAMES_PER_BLOCK(0xc0);
+		tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_AUDIO_CNTRL0);
+	}
+
+	config = tegra_hdmi_get_audio_config(hdmi->audio_freq, pclk);
+	if (!config) {
+		dev_err(hdmi->dev, "cannot set audio to %u at %u pclk\n",
+			hdmi->audio_freq, pclk);
+		return -EINVAL;
+	}
+
+	tegra_hdmi_writel(hdmi, 0, HDMI_NV_PDISP_HDMI_ACR_CTRL);
+
+	value = AUDIO_N_RESETF | AUDIO_N_GENERATE_ALTERNATE |
+		AUDIO_N_VALUE(config->n - 1);
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_AUDIO_N);
+
+	tegra_hdmi_writel(hdmi, ACR_SUBPACK_N(config->n) | ACR_ENABLE,
+			  HDMI_NV_PDISP_HDMI_ACR_0441_SUBPACK_HIGH);
+
+	value = ACR_SUBPACK_CTS(config->cts);
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_HDMI_ACR_0441_SUBPACK_LOW);
+
+	value = SPARE_HW_CTS | SPARE_FORCE_SW_CTS | SPARE_CTS_RESET_VAL(1);
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_HDMI_SPARE);
+
+	value = tegra_hdmi_readl(hdmi, HDMI_NV_PDISP_AUDIO_N);
+	value &= ~AUDIO_N_RESETF;
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_AUDIO_N);
+
+	if (of_device_is_compatible(node, "nvidia,tegra30-hdmi")) {
+		switch (hdmi->audio_freq) {
+		case 32000:
+			offset = HDMI_NV_PDISP_SOR_AUDIO_AVAL_0320;
+			break;
+
+		case 44100:
+			offset = HDMI_NV_PDISP_SOR_AUDIO_AVAL_0441;
+			break;
+
+		case 48000:
+			offset = HDMI_NV_PDISP_SOR_AUDIO_AVAL_0480;
+			break;
+
+		case 88200:
+			offset = HDMI_NV_PDISP_SOR_AUDIO_AVAL_0882;
+			break;
+
+		case 96000:
+			offset = HDMI_NV_PDISP_SOR_AUDIO_AVAL_0960;
+			break;
+
+		case 176400:
+			offset = HDMI_NV_PDISP_SOR_AUDIO_AVAL_1764;
+			break;
+
+		case 192000:
+			offset = HDMI_NV_PDISP_SOR_AUDIO_AVAL_1920;
+			break;
+		}
+
+		tegra_hdmi_writel(hdmi, config->aval, offset);
+	}
+
+	tegra_hdmi_setup_audio_fs_tables(hdmi);
+
+	return 0;
+}
+
+static inline unsigned long tegra_hdmi_subpack(const u8 *ptr, size_t size)
+{
+	unsigned long value = 0;
+	size_t i;
+
+	for (i = size; i > 0; i--)
+		value = (value << 8) | ptr[i - 1];
+
+	return value;
+}
+
+static void tegra_hdmi_write_infopack(struct tegra_hdmi *hdmi, const void *data,
+				      size_t size)
+{
+	const u8 *ptr = data;
+	unsigned long offset;
+	unsigned long value;
+	size_t i, j;
+
+	switch (ptr[0]) {
+	case HDMI_INFOFRAME_TYPE_AVI:
+		offset = HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_HEADER;
+		break;
+
+	case HDMI_INFOFRAME_TYPE_AUDIO:
+		offset = HDMI_NV_PDISP_HDMI_AUDIO_INFOFRAME_HEADER;
+		break;
+
+	case HDMI_INFOFRAME_TYPE_VENDOR:
+		offset = HDMI_NV_PDISP_HDMI_GENERIC_HEADER;
+		break;
+
+	default:
+		dev_err(hdmi->dev, "unsupported infoframe type: %02x\n",
+			ptr[0]);
+		return;
+	}
+
+	value = INFOFRAME_HEADER_TYPE(ptr[0]) |
+		INFOFRAME_HEADER_VERSION(ptr[1]) |
+		INFOFRAME_HEADER_LEN(ptr[2]);
+	tegra_hdmi_writel(hdmi, value, offset);
+	offset++;
+
+	/*
+	 * Each subpack contains 7 bytes, divided into:
+	 * - subpack_low: bytes 0 - 3
+	 * - subpack_high: bytes 4 - 6 (with byte 7 padded to 0x00)
+	 */
+	for (i = 3, j = 0; i < size; i += 7, j += 8) {
+		size_t rem = size - i, num = min_t(size_t, rem, 4);
+
+		value = tegra_hdmi_subpack(&ptr[i], num);
+		tegra_hdmi_writel(hdmi, value, offset++);
+
+		num = min_t(size_t, rem - num, 3);
+
+		value = tegra_hdmi_subpack(&ptr[i + 4], num);
+		tegra_hdmi_writel(hdmi, value, offset++);
+	}
+}
+
+static void tegra_hdmi_setup_avi_infoframe(struct tegra_hdmi *hdmi,
+					   struct drm_display_mode *mode)
+{
+	struct hdmi_avi_infoframe frame;
+	u8 buffer[17];
+	ssize_t err;
+
+	if (hdmi->dvi) {
+		tegra_hdmi_writel(hdmi, 0,
+				  HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_CTRL);
+		return;
+	}
+
+	err = drm_hdmi_avi_infoframe_from_display_mode(&frame, mode);
+	if (err < 0) {
+		dev_err(hdmi->dev, "failed to setup AVI infoframe: %zd\n", err);
+		return;
+	}
+
+	err = hdmi_avi_infoframe_pack(&frame, buffer, sizeof(buffer));
+	if (err < 0) {
+		dev_err(hdmi->dev, "failed to pack AVI infoframe: %zd\n", err);
+		return;
+	}
+
+	tegra_hdmi_write_infopack(hdmi, buffer, err);
+
+	tegra_hdmi_writel(hdmi, INFOFRAME_CTRL_ENABLE,
+			  HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_CTRL);
+}
+
+static void tegra_hdmi_setup_audio_infoframe(struct tegra_hdmi *hdmi)
+{
+	struct hdmi_audio_infoframe frame;
+	u8 buffer[14];
+	ssize_t err;
+
+	if (hdmi->dvi) {
+		tegra_hdmi_writel(hdmi, 0,
+				  HDMI_NV_PDISP_HDMI_AUDIO_INFOFRAME_CTRL);
+		return;
+	}
+
+	err = hdmi_audio_infoframe_init(&frame);
+	if (err < 0) {
+		dev_err(hdmi->dev, "failed to initialize audio infoframe: %d\n",
+			err);
+		return;
+	}
+
+	frame.channels = 2;
+
+	err = hdmi_audio_infoframe_pack(&frame, buffer, sizeof(buffer));
+	if (err < 0) {
+		dev_err(hdmi->dev, "failed to pack audio infoframe: %zd\n",
+			err);
+		return;
+	}
+
+	/*
+	 * The audio infoframe has only one set of subpack registers, so the
+	 * infoframe needs to be truncated. One set of subpack registers can
+	 * contain 7 bytes. Including the 3 byte header only the first 10
+	 * bytes can be programmed.
+	 */
+	tegra_hdmi_write_infopack(hdmi, buffer, min(10, err));
+
+	tegra_hdmi_writel(hdmi, INFOFRAME_CTRL_ENABLE,
+			  HDMI_NV_PDISP_HDMI_AUDIO_INFOFRAME_CTRL);
+}
+
+static void tegra_hdmi_setup_stereo_infoframe(struct tegra_hdmi *hdmi)
+{
+	struct hdmi_vendor_infoframe frame;
+	unsigned long value;
+	u8 buffer[10];
+	ssize_t err;
+
+	if (!hdmi->stereo) {
+		value = tegra_hdmi_readl(hdmi, HDMI_NV_PDISP_HDMI_GENERIC_CTRL);
+		value &= ~GENERIC_CTRL_ENABLE;
+		tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_HDMI_GENERIC_CTRL);
+		return;
+	}
+
+	memset(&frame, 0, sizeof(frame));
+
+	frame.type = HDMI_INFOFRAME_TYPE_VENDOR;
+	frame.version = 0x01;
+	frame.length = 6;
+
+	frame.data[0] = 0x03; /* regid0 */
+	frame.data[1] = 0x0c; /* regid1 */
+	frame.data[2] = 0x00; /* regid2 */
+	frame.data[3] = 0x02 << 5; /* video format */
+
+	/* TODO: 74 MHz limit? */
+	if (1) {
+		frame.data[4] = 0x00 << 4; /* 3D structure */
+	} else {
+		frame.data[4] = 0x08 << 4; /* 3D structure */
+		frame.data[5] = 0x00 << 4; /* 3D ext. data */
+	}
+
+	err = hdmi_vendor_infoframe_pack(&frame, buffer, sizeof(buffer));
+	if (err < 0) {
+		dev_err(hdmi->dev, "failed to pack vendor infoframe: %zd\n",
+			err);
+		return;
+	}
+
+	tegra_hdmi_write_infopack(hdmi, buffer, err);
+
+	value = tegra_hdmi_readl(hdmi, HDMI_NV_PDISP_HDMI_GENERIC_CTRL);
+	value |= GENERIC_CTRL_ENABLE;
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_HDMI_GENERIC_CTRL);
+}
+
+static void tegra_hdmi_setup_tmds(struct tegra_hdmi *hdmi,
+				  const struct tmds_config *tmds)
+{
+	unsigned long value;
+
+	tegra_hdmi_writel(hdmi, tmds->pll0, HDMI_NV_PDISP_SOR_PLL0);
+	tegra_hdmi_writel(hdmi, tmds->pll1, HDMI_NV_PDISP_SOR_PLL1);
+	tegra_hdmi_writel(hdmi, tmds->pe_current, HDMI_NV_PDISP_PE_CURRENT);
+
+	value = tmds->drive_current | DRIVE_CURRENT_FUSE_OVERRIDE;
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_LANE_DRIVE_CURRENT);
+}
+
+static int tegra_output_hdmi_enable(struct tegra_output *output)
+{
+	unsigned int h_sync_width, h_front_porch, h_back_porch, i, rekey;
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
+	struct drm_display_mode *mode = &dc->base.mode;
+	struct tegra_hdmi *hdmi = to_hdmi(output);
+	struct device_node *node = hdmi->dev->of_node;
+	unsigned int pulse_start, div82, pclk;
+	const struct tmds_config *tmds;
+	unsigned int num_tmds;
+	unsigned long value;
+	int retries = 1000;
+	int err;
+
+	pclk = mode->clock * 1000;
+	h_sync_width = mode->hsync_end - mode->hsync_start;
+	h_back_porch = mode->htotal - mode->hsync_end;
+	h_front_porch = mode->hsync_start - mode->hdisplay;
+
+	err = regulator_enable(hdmi->vdd);
+	if (err < 0) {
+		dev_err(hdmi->dev, "failed to enable VDD regulator: %d\n", err);
+		return err;
+	}
+
+	err = regulator_enable(hdmi->pll);
+	if (err < 0) {
+		dev_err(hdmi->dev, "failed to enable PLL regulator: %d\n", err);
+		return err;
+	}
+
+	/*
+	 * This assumes that the display controller will divide its parent
+	 * clock by 2 to generate the pixel clock.
+	 */
+	err = tegra_output_setup_clock(output, hdmi->clk, pclk * 2);
+	if (err < 0) {
+		dev_err(hdmi->dev, "failed to setup clock: %d\n", err);
+		return err;
+	}
+
+	err = clk_set_rate(hdmi->clk, pclk);
+	if (err < 0)
+		return err;
+
+	err = clk_enable(hdmi->clk);
+	if (err < 0) {
+		dev_err(hdmi->dev, "failed to enable clock: %d\n", err);
+		return err;
+	}
+
+	tegra_periph_reset_assert(hdmi->clk);
+	usleep_range(1000, 2000);
+	tegra_periph_reset_deassert(hdmi->clk);
+
+	tegra_dc_writel(dc, VSYNC_H_POSITION(1),
+			DC_DISP_DISP_TIMING_OPTIONS);
+	tegra_dc_writel(dc, DITHER_CONTROL_DISABLE | BASE_COLOR_SIZE888,
+			DC_DISP_DISP_COLOR_CONTROL);
+
+	/* video_preamble uses h_pulse2 */
+	pulse_start = 1 + h_sync_width + h_back_porch - 10;
+
+	tegra_dc_writel(dc, H_PULSE_2_ENABLE, DC_DISP_DISP_SIGNAL_OPTIONS0);
+
+	value = PULSE_MODE_NORMAL | PULSE_POLARITY_HIGH | PULSE_QUAL_VACTIVE |
+		PULSE_LAST_END_A;
+	tegra_dc_writel(dc, value, DC_DISP_H_PULSE2_CONTROL);
+
+	value = PULSE_START(pulse_start) | PULSE_END(pulse_start + 8);
+	tegra_dc_writel(dc, value, DC_DISP_H_PULSE2_POSITION_A);
+
+	value = VSYNC_WINDOW_END(0x210) | VSYNC_WINDOW_START(0x200) |
+		VSYNC_WINDOW_ENABLE;
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_HDMI_VSYNC_WINDOW);
+
+	if (dc->pipe)
+		value = HDMI_SRC_DISPLAYB;
+	else
+		value = HDMI_SRC_DISPLAYA;
+
+	if ((mode->hdisplay == 720) && ((mode->vdisplay == 480) ||
+					(mode->vdisplay == 576)))
+		tegra_hdmi_writel(hdmi,
+				  value | ARM_VIDEO_RANGE_FULL,
+				  HDMI_NV_PDISP_INPUT_CONTROL);
+	else
+		tegra_hdmi_writel(hdmi,
+				  value | ARM_VIDEO_RANGE_LIMITED,
+				  HDMI_NV_PDISP_INPUT_CONTROL);
+
+	div82 = clk_get_rate(hdmi->clk) / 1000000 * 4;
+	value = SOR_REFCLK_DIV_INT(div82 >> 2) | SOR_REFCLK_DIV_FRAC(div82);
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_REFCLK);
+
+	if (!hdmi->dvi) {
+		err = tegra_hdmi_setup_audio(hdmi, pclk);
+		if (err < 0)
+			hdmi->dvi = true;
+	}
+
+	if (of_device_is_compatible(node, "nvidia,tegra20-hdmi")) {
+		/*
+		 * TODO: add ELD support
+		 */
+	}
+
+	rekey = HDMI_REKEY_DEFAULT;
+	value = HDMI_CTRL_REKEY(rekey);
+	value |= HDMI_CTRL_MAX_AC_PACKET((h_sync_width + h_back_porch +
+					  h_front_porch - rekey - 18) / 32);
+
+	if (!hdmi->dvi)
+		value |= HDMI_CTRL_ENABLE;
+
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_HDMI_CTRL);
+
+	if (hdmi->dvi)
+		tegra_hdmi_writel(hdmi, 0x0,
+				  HDMI_NV_PDISP_HDMI_GENERIC_CTRL);
+	else
+		tegra_hdmi_writel(hdmi, GENERIC_CTRL_AUDIO,
+				  HDMI_NV_PDISP_HDMI_GENERIC_CTRL);
+
+	tegra_hdmi_setup_avi_infoframe(hdmi, mode);
+	tegra_hdmi_setup_audio_infoframe(hdmi);
+	tegra_hdmi_setup_stereo_infoframe(hdmi);
+
+	/* TMDS CONFIG */
+	if (of_device_is_compatible(node, "nvidia,tegra30-hdmi")) {
+		num_tmds = ARRAY_SIZE(tegra3_tmds_config);
+		tmds = tegra3_tmds_config;
+	} else {
+		num_tmds = ARRAY_SIZE(tegra2_tmds_config);
+		tmds = tegra2_tmds_config;
+	}
+
+	for (i = 0; i < num_tmds; i++) {
+		if (pclk <= tmds[i].pclk) {
+			tegra_hdmi_setup_tmds(hdmi, &tmds[i]);
+			break;
+		}
+	}
+
+	tegra_hdmi_writel(hdmi,
+			  SOR_SEQ_CTL_PU_PC(0) |
+			  SOR_SEQ_PU_PC_ALT(0) |
+			  SOR_SEQ_PD_PC(8) |
+			  SOR_SEQ_PD_PC_ALT(8),
+			  HDMI_NV_PDISP_SOR_SEQ_CTL);
+
+	value = SOR_SEQ_INST_WAIT_TIME(1) |
+		SOR_SEQ_INST_WAIT_UNITS_VSYNC |
+		SOR_SEQ_INST_HALT |
+		SOR_SEQ_INST_PIN_A_LOW |
+		SOR_SEQ_INST_PIN_B_LOW |
+		SOR_SEQ_INST_DRIVE_PWM_OUT_LO;
+
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_SEQ_INST(0));
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_SEQ_INST(8));
+
+	value = 0x1c800;
+	value &= ~SOR_CSTM_ROTCLK(~0);
+	value |= SOR_CSTM_ROTCLK(2);
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_CSTM);
+
+	tegra_dc_writel(dc, DISP_CTRL_MODE_STOP, DC_CMD_DISPLAY_COMMAND);
+	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+
+	/* start SOR */
+	tegra_hdmi_writel(hdmi,
+			  SOR_PWR_NORMAL_STATE_PU |
+			  SOR_PWR_NORMAL_START_NORMAL |
+			  SOR_PWR_SAFE_STATE_PD |
+			  SOR_PWR_SETTING_NEW_TRIGGER,
+			  HDMI_NV_PDISP_SOR_PWR);
+	tegra_hdmi_writel(hdmi,
+			  SOR_PWR_NORMAL_STATE_PU |
+			  SOR_PWR_NORMAL_START_NORMAL |
+			  SOR_PWR_SAFE_STATE_PD |
+			  SOR_PWR_SETTING_NEW_DONE,
+			  HDMI_NV_PDISP_SOR_PWR);
+
+	do {
+		BUG_ON(--retries < 0);
+		value = tegra_hdmi_readl(hdmi, HDMI_NV_PDISP_SOR_PWR);
+	} while (value & SOR_PWR_SETTING_NEW_PENDING);
+
+	value = SOR_STATE_ASY_CRCMODE_COMPLETE |
+		SOR_STATE_ASY_OWNER_HEAD0 |
+		SOR_STATE_ASY_SUBOWNER_BOTH |
+		SOR_STATE_ASY_PROTOCOL_SINGLE_TMDS_A |
+		SOR_STATE_ASY_DEPOL_POS;
+
+	/* setup sync polarities */
+	if (mode->flags & DRM_MODE_FLAG_PHSYNC)
+		value |= SOR_STATE_ASY_HSYNCPOL_POS;
+
+	if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+		value |= SOR_STATE_ASY_HSYNCPOL_NEG;
+
+	if (mode->flags & DRM_MODE_FLAG_PVSYNC)
+		value |= SOR_STATE_ASY_VSYNCPOL_POS;
+
+	if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+		value |= SOR_STATE_ASY_VSYNCPOL_NEG;
+
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_STATE2);
+
+	value = SOR_STATE_ASY_HEAD_OPMODE_AWAKE | SOR_STATE_ASY_ORMODE_NORMAL;
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_STATE1);
+
+	tegra_hdmi_writel(hdmi, 0, HDMI_NV_PDISP_SOR_STATE0);
+	tegra_hdmi_writel(hdmi, SOR_STATE_UPDATE, HDMI_NV_PDISP_SOR_STATE0);
+	tegra_hdmi_writel(hdmi, value | SOR_STATE_ATTACHED,
+			  HDMI_NV_PDISP_SOR_STATE1);
+	tegra_hdmi_writel(hdmi, 0, HDMI_NV_PDISP_SOR_STATE0);
+
+	tegra_dc_writel(dc, HDMI_ENABLE, DC_DISP_DISP_WIN_OPTIONS);
+
+	value = PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+		PW4_ENABLE | PM0_ENABLE | PM1_ENABLE;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
+	value = DISP_CTRL_MODE_C_DISPLAY;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_COMMAND);
+
+	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+
+	/* TODO: add HDCP support */
+
+	return 0;
+}
+
+static int tegra_output_hdmi_disable(struct tegra_output *output)
+{
+	struct tegra_hdmi *hdmi = to_hdmi(output);
+
+	tegra_periph_reset_assert(hdmi->clk);
+	clk_disable(hdmi->clk);
+	regulator_disable(hdmi->pll);
+	regulator_disable(hdmi->vdd);
+
+	return 0;
+}
+
+static int tegra_output_hdmi_setup_clock(struct tegra_output *output,
+					 struct clk *clk, unsigned long pclk)
+{
+	struct tegra_hdmi *hdmi = to_hdmi(output);
+	struct clk *base;
+	int err;
+
+	err = clk_set_parent(clk, hdmi->clk_parent);
+	if (err < 0) {
+		dev_err(output->dev, "failed to set parent: %d\n", err);
+		return err;
+	}
+
+	base = clk_get_parent(hdmi->clk_parent);
+
+	/*
+	 * This assumes that the parent clock is pll_d_out0 or pll_d2_out
+	 * respectively, each of which divides the base pll_d by 2.
+	 */
+	err = clk_set_rate(base, pclk * 2);
+	if (err < 0)
+		dev_err(output->dev,
+			"failed to set base clock rate to %lu Hz\n",
+			pclk * 2);
+
+	return 0;
+}
+
+static int tegra_output_hdmi_check_mode(struct tegra_output *output,
+					struct drm_display_mode *mode,
+					enum drm_mode_status *status)
+{
+	struct tegra_hdmi *hdmi = to_hdmi(output);
+	unsigned long pclk = mode->clock * 1000;
+	struct clk *parent;
+	long err;
+
+	parent = clk_get_parent(hdmi->clk_parent);
+
+	err = clk_round_rate(parent, pclk * 4);
+	if (err < 0)
+		*status = MODE_NOCLOCK;
+	else
+		*status = MODE_OK;
+
+	return 0;
+}
+
+static const struct tegra_output_ops hdmi_ops = {
+	.enable = tegra_output_hdmi_enable,
+	.disable = tegra_output_hdmi_disable,
+	.setup_clock = tegra_output_hdmi_setup_clock,
+	.check_mode = tegra_output_hdmi_check_mode,
+};
+
+static int tegra_hdmi_show_regs(struct seq_file *s, void *data)
+{
+	struct drm_info_node *node = s->private;
+	struct tegra_hdmi *hdmi = node->info_ent->data;
+
+#define DUMP_REG(name)						\
+	seq_printf(s, "%-56s %#05x %08lx\n", #name, name,	\
+		tegra_hdmi_readl(hdmi, name))
+
+	DUMP_REG(HDMI_CTXSW);
+	DUMP_REG(HDMI_NV_PDISP_SOR_STATE0);
+	DUMP_REG(HDMI_NV_PDISP_SOR_STATE1);
+	DUMP_REG(HDMI_NV_PDISP_SOR_STATE2);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_AN_MSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_AN_LSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_CN_MSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_CN_LSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_AKSV_MSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_AKSV_LSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_BKSV_MSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_BKSV_LSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_CKSV_MSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_CKSV_LSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_DKSV_MSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_DKSV_LSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_CTRL);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_CMODE);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_MPRIME_MSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_MPRIME_LSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_SPRIME_MSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_SPRIME_LSB2);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_SPRIME_LSB1);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_RI);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_CS_MSB);
+	DUMP_REG(HDMI_NV_PDISP_RG_HDCP_CS_LSB);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AUDIO_EMU0);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AUDIO_EMU_RDATA0);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AUDIO_EMU1);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AUDIO_EMU2);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AUDIO_INFOFRAME_CTRL);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AUDIO_INFOFRAME_STATUS);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AUDIO_INFOFRAME_HEADER);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AUDIO_INFOFRAME_SUBPACK0_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AUDIO_INFOFRAME_SUBPACK0_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_CTRL);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_STATUS);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_HEADER);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_SUBPACK0_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_SUBPACK0_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_SUBPACK1_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_SUBPACK1_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GENERIC_CTRL);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GENERIC_STATUS);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GENERIC_HEADER);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK0_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK0_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK1_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK1_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK2_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK2_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK3_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK3_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_CTRL);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_0320_SUBPACK_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_0320_SUBPACK_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_0441_SUBPACK_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_0441_SUBPACK_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_0882_SUBPACK_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_0882_SUBPACK_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_1764_SUBPACK_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_1764_SUBPACK_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_0480_SUBPACK_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_0480_SUBPACK_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_0960_SUBPACK_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_0960_SUBPACK_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_1920_SUBPACK_LOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_ACR_1920_SUBPACK_HIGH);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_CTRL);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_VSYNC_KEEPOUT);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_VSYNC_WINDOW);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GCP_CTRL);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GCP_STATUS);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_GCP_SUBPACK);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_CHANNEL_STATUS1);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_CHANNEL_STATUS2);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_EMU0);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_EMU1);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_EMU1_RDATA);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_SPARE);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_SPDIF_CHN_STATUS1);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_SPDIF_CHN_STATUS2);
+	DUMP_REG(HDMI_NV_PDISP_HDMI_HDCPRIF_ROM_CTRL);
+	DUMP_REG(HDMI_NV_PDISP_SOR_CAP);
+	DUMP_REG(HDMI_NV_PDISP_SOR_PWR);
+	DUMP_REG(HDMI_NV_PDISP_SOR_TEST);
+	DUMP_REG(HDMI_NV_PDISP_SOR_PLL0);
+	DUMP_REG(HDMI_NV_PDISP_SOR_PLL1);
+	DUMP_REG(HDMI_NV_PDISP_SOR_PLL2);
+	DUMP_REG(HDMI_NV_PDISP_SOR_CSTM);
+	DUMP_REG(HDMI_NV_PDISP_SOR_LVDS);
+	DUMP_REG(HDMI_NV_PDISP_SOR_CRCA);
+	DUMP_REG(HDMI_NV_PDISP_SOR_CRCB);
+	DUMP_REG(HDMI_NV_PDISP_SOR_BLANK);
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_CTL);
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(0));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(1));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(2));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(3));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(4));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(5));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(6));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(7));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(8));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(9));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(10));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(11));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(12));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(13));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(14));
+	DUMP_REG(HDMI_NV_PDISP_SOR_SEQ_INST(15));
+	DUMP_REG(HDMI_NV_PDISP_SOR_VCRCA0);
+	DUMP_REG(HDMI_NV_PDISP_SOR_VCRCA1);
+	DUMP_REG(HDMI_NV_PDISP_SOR_CCRCA0);
+	DUMP_REG(HDMI_NV_PDISP_SOR_CCRCA1);
+	DUMP_REG(HDMI_NV_PDISP_SOR_EDATAA0);
+	DUMP_REG(HDMI_NV_PDISP_SOR_EDATAA1);
+	DUMP_REG(HDMI_NV_PDISP_SOR_COUNTA0);
+	DUMP_REG(HDMI_NV_PDISP_SOR_COUNTA1);
+	DUMP_REG(HDMI_NV_PDISP_SOR_DEBUGA0);
+	DUMP_REG(HDMI_NV_PDISP_SOR_DEBUGA1);
+	DUMP_REG(HDMI_NV_PDISP_SOR_TRIG);
+	DUMP_REG(HDMI_NV_PDISP_SOR_MSCHECK);
+	DUMP_REG(HDMI_NV_PDISP_SOR_LANE_DRIVE_CURRENT);
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_DEBUG0);
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_DEBUG1);
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_DEBUG2);
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_FS(0));
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_FS(1));
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_FS(2));
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_FS(3));
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_FS(4));
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_FS(5));
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_FS(6));
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_PULSE_WIDTH);
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_THRESHOLD);
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_CNTRL0);
+	DUMP_REG(HDMI_NV_PDISP_AUDIO_N);
+	DUMP_REG(HDMI_NV_PDISP_HDCPRIF_ROM_TIMING);
+	DUMP_REG(HDMI_NV_PDISP_SOR_REFCLK);
+	DUMP_REG(HDMI_NV_PDISP_CRC_CONTROL);
+	DUMP_REG(HDMI_NV_PDISP_INPUT_CONTROL);
+	DUMP_REG(HDMI_NV_PDISP_SCRATCH);
+	DUMP_REG(HDMI_NV_PDISP_PE_CURRENT);
+	DUMP_REG(HDMI_NV_PDISP_KEY_CTRL);
+	DUMP_REG(HDMI_NV_PDISP_KEY_DEBUG0);
+	DUMP_REG(HDMI_NV_PDISP_KEY_DEBUG1);
+	DUMP_REG(HDMI_NV_PDISP_KEY_DEBUG2);
+	DUMP_REG(HDMI_NV_PDISP_KEY_HDCP_KEY_0);
+	DUMP_REG(HDMI_NV_PDISP_KEY_HDCP_KEY_1);
+	DUMP_REG(HDMI_NV_PDISP_KEY_HDCP_KEY_2);
+	DUMP_REG(HDMI_NV_PDISP_KEY_HDCP_KEY_3);
+	DUMP_REG(HDMI_NV_PDISP_KEY_HDCP_KEY_TRIG);
+	DUMP_REG(HDMI_NV_PDISP_KEY_SKEY_INDEX);
+	DUMP_REG(HDMI_NV_PDISP_SOR_AUDIO_CNTRL0);
+	DUMP_REG(HDMI_NV_PDISP_SOR_AUDIO_HDA_ELD_BUFWR);
+	DUMP_REG(HDMI_NV_PDISP_SOR_AUDIO_HDA_PRESENSE);
+
+#undef DUMP_REG
+
+	return 0;
+}
+
+static struct drm_info_list debugfs_files[] = {
+	{ "regs", tegra_hdmi_show_regs, 0, NULL },
+};
+
+static int tegra_hdmi_debugfs_init(struct tegra_hdmi *hdmi,
+				   struct drm_minor *minor)
+{
+	unsigned int i;
+	int err;
+
+	hdmi->debugfs = debugfs_create_dir("hdmi", minor->debugfs_root);
+	if (!hdmi->debugfs)
+		return -ENOMEM;
+
+	hdmi->debugfs_files = kmemdup(debugfs_files, sizeof(debugfs_files),
+				      GFP_KERNEL);
+	if (!hdmi->debugfs_files) {
+		err = -ENOMEM;
+		goto remove;
+	}
+
+	for (i = 0; i < ARRAY_SIZE(debugfs_files); i++)
+		hdmi->debugfs_files[i].data = hdmi;
+
+	err = drm_debugfs_create_files(hdmi->debugfs_files,
+				       ARRAY_SIZE(debugfs_files),
+				       hdmi->debugfs, minor);
+	if (err < 0)
+		goto free;
+
+	hdmi->minor = minor;
+
+	return 0;
+
+free:
+	kfree(hdmi->debugfs_files);
+	hdmi->debugfs_files = NULL;
+remove:
+	debugfs_remove(hdmi->debugfs);
+	hdmi->debugfs = NULL;
+
+	return err;
+}
+
+static int tegra_hdmi_debugfs_exit(struct tegra_hdmi *hdmi)
+{
+	drm_debugfs_remove_files(hdmi->debugfs_files, ARRAY_SIZE(debugfs_files),
+				 hdmi->minor);
+	hdmi->minor = NULL;
+
+	kfree(hdmi->debugfs_files);
+	hdmi->debugfs_files = NULL;
+
+	debugfs_remove(hdmi->debugfs);
+	hdmi->debugfs = NULL;
+
+	return 0;
+}
+
+static int tegra_hdmi_drm_init(struct host1x_client *client,
+			       struct drm_device *drm)
+{
+	struct tegra_hdmi *hdmi = host1x_client_to_hdmi(client);
+	int err;
+
+	hdmi->output.type = TEGRA_OUTPUT_HDMI;
+	hdmi->output.dev = client->dev;
+	hdmi->output.ops = &hdmi_ops;
+
+	err = tegra_output_init(drm, &hdmi->output);
+	if (err < 0) {
+		dev_err(client->dev, "output setup failed: %d\n", err);
+		return err;
+	}
+
+	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
+		err = tegra_hdmi_debugfs_init(hdmi, drm->primary);
+		if (err < 0)
+			dev_err(client->dev, "debugfs setup failed: %d\n", err);
+	}
+
+	return 0;
+}
+
+static int tegra_hdmi_drm_exit(struct host1x_client *client)
+{
+	struct tegra_hdmi *hdmi = host1x_client_to_hdmi(client);
+	int err;
+
+	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
+		err = tegra_hdmi_debugfs_exit(hdmi);
+		if (err < 0)
+			dev_err(client->dev, "debugfs cleanup failed: %d\n",
+				err);
+	}
+
+	err = tegra_output_disable(&hdmi->output);
+	if (err < 0) {
+		dev_err(client->dev, "output failed to disable: %d\n", err);
+		return err;
+	}
+
+	err = tegra_output_exit(&hdmi->output);
+	if (err < 0) {
+		dev_err(client->dev, "output cleanup failed: %d\n", err);
+		return err;
+	}
+
+	return 0;
+}
+
+static const struct host1x_client_ops hdmi_client_ops = {
+	.drm_init = tegra_hdmi_drm_init,
+	.drm_exit = tegra_hdmi_drm_exit,
+};
+
+static int tegra_hdmi_probe(struct platform_device *pdev)
+{
+	struct host1x_drm *host1x = host1x_get_drm_data(pdev->dev.parent);
+	struct tegra_hdmi *hdmi;
+	struct resource *regs;
+	int err;
+
+	hdmi = devm_kzalloc(&pdev->dev, sizeof(*hdmi), GFP_KERNEL);
+	if (!hdmi)
+		return -ENOMEM;
+
+	hdmi->dev = &pdev->dev;
+	hdmi->audio_source = AUTO;
+	hdmi->audio_freq = 44100;
+	hdmi->stereo = false;
+	hdmi->dvi = false;
+
+	hdmi->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(hdmi->clk)) {
+		dev_err(&pdev->dev, "failed to get clock\n");
+		return PTR_ERR(hdmi->clk);
+	}
+
+	err = clk_prepare(hdmi->clk);
+	if (err < 0)
+		return err;
+
+	hdmi->clk_parent = devm_clk_get(&pdev->dev, "parent");
+	if (IS_ERR(hdmi->clk_parent))
+		return PTR_ERR(hdmi->clk_parent);
+
+	err = clk_prepare(hdmi->clk_parent);
+	if (err < 0)
+		return err;
+
+	err = clk_set_parent(hdmi->clk, hdmi->clk_parent);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to setup clocks: %d\n", err);
+		return err;
+	}
+
+	hdmi->vdd = devm_regulator_get(&pdev->dev, "vdd");
+	if (IS_ERR(hdmi->vdd)) {
+		dev_err(&pdev->dev, "failed to get VDD regulator\n");
+		return PTR_ERR(hdmi->vdd);
+	}
+
+	hdmi->pll = devm_regulator_get(&pdev->dev, "pll");
+	if (IS_ERR(hdmi->pll)) {
+		dev_err(&pdev->dev, "failed to get PLL regulator\n");
+		return PTR_ERR(hdmi->pll);
+	}
+
+	hdmi->output.dev = &pdev->dev;
+
+	err = tegra_output_parse_dt(&hdmi->output);
+	if (err < 0)
+		return err;
+
+	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!regs)
+		return -ENXIO;
+
+	hdmi->regs = devm_ioremap_resource(&pdev->dev, regs);
+	if (IS_ERR(hdmi->regs))
+		return PTR_ERR(hdmi->regs);
+
+	err = platform_get_irq(pdev, 0);
+	if (err < 0)
+		return err;
+
+	hdmi->irq = err;
+
+	hdmi->client.ops = &hdmi_client_ops;
+	INIT_LIST_HEAD(&hdmi->client.list);
+	hdmi->client.dev = &pdev->dev;
+
+	err = host1x_register_client(host1x, &hdmi->client);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to register host1x client: %d\n",
+			err);
+		return err;
+	}
+
+	platform_set_drvdata(pdev, hdmi);
+
+	return 0;
+}
+
+static int tegra_hdmi_remove(struct platform_device *pdev)
+{
+	struct host1x_drm *host1x = host1x_get_drm_data(pdev->dev.parent);
+	struct tegra_hdmi *hdmi = platform_get_drvdata(pdev);
+	int err;
+
+	err = host1x_unregister_client(host1x, &hdmi->client);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to unregister host1x client: %d\n",
+			err);
+		return err;
+	}
+
+	clk_unprepare(hdmi->clk_parent);
+	clk_unprepare(hdmi->clk);
+
+	return 0;
+}
+
+static struct of_device_id tegra_hdmi_of_match[] = {
+	{ .compatible = "nvidia,tegra30-hdmi", },
+	{ .compatible = "nvidia,tegra20-hdmi", },
+	{ },
+};
+
+struct platform_driver tegra_hdmi_driver = {
+	.driver = {
+		.name = "tegra-hdmi",
+		.owner = THIS_MODULE,
+		.of_match_table = tegra_hdmi_of_match,
+	},
+	.probe = tegra_hdmi_probe,
+	.remove = tegra_hdmi_remove,
+};
diff --git a/drivers/gpu/host1x/drm/hdmi.h b/drivers/gpu/host1x/drm/hdmi.h
new file mode 100644
index 00000000000..52ac36e08cc
--- /dev/null
+++ b/drivers/gpu/host1x/drm/hdmi.h
@@ -0,0 +1,386 @@
+/*
+ * Copyright (C) 2012 Avionic Design GmbH
+ * Copyright (C) 2012 NVIDIA CORPORATION.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef TEGRA_HDMI_H
+#define TEGRA_HDMI_H 1
+
+/* register definitions */
+#define HDMI_CTXSW						0x00
+
+#define HDMI_NV_PDISP_SOR_STATE0				0x01
+#define SOR_STATE_UPDATE (1 << 0)
+
+#define HDMI_NV_PDISP_SOR_STATE1				0x02
+#define SOR_STATE_ASY_HEAD_OPMODE_AWAKE (2 << 0)
+#define SOR_STATE_ASY_ORMODE_NORMAL     (1 << 2)
+#define SOR_STATE_ATTACHED              (1 << 3)
+
+#define HDMI_NV_PDISP_SOR_STATE2				0x03
+#define SOR_STATE_ASY_OWNER_NONE         (0 <<  0)
+#define SOR_STATE_ASY_OWNER_HEAD0        (1 <<  0)
+#define SOR_STATE_ASY_SUBOWNER_NONE      (0 <<  4)
+#define SOR_STATE_ASY_SUBOWNER_SUBHEAD0  (1 <<  4)
+#define SOR_STATE_ASY_SUBOWNER_SUBHEAD1  (2 <<  4)
+#define SOR_STATE_ASY_SUBOWNER_BOTH      (3 <<  4)
+#define SOR_STATE_ASY_CRCMODE_ACTIVE     (0 <<  6)
+#define SOR_STATE_ASY_CRCMODE_COMPLETE   (1 <<  6)
+#define SOR_STATE_ASY_CRCMODE_NON_ACTIVE (2 <<  6)
+#define SOR_STATE_ASY_PROTOCOL_SINGLE_TMDS_A (1 << 8)
+#define SOR_STATE_ASY_PROTOCOL_CUSTOM        (15 << 8)
+#define SOR_STATE_ASY_HSYNCPOL_POS       (0 << 12)
+#define SOR_STATE_ASY_HSYNCPOL_NEG       (1 << 12)
+#define SOR_STATE_ASY_VSYNCPOL_POS       (0 << 13)
+#define SOR_STATE_ASY_VSYNCPOL_NEG       (1 << 13)
+#define SOR_STATE_ASY_DEPOL_POS          (0 << 14)
+#define SOR_STATE_ASY_DEPOL_NEG          (1 << 14)
+
+#define HDMI_NV_PDISP_RG_HDCP_AN_MSB				0x04
+#define HDMI_NV_PDISP_RG_HDCP_AN_LSB				0x05
+#define HDMI_NV_PDISP_RG_HDCP_CN_MSB				0x06
+#define HDMI_NV_PDISP_RG_HDCP_CN_LSB				0x07
+#define HDMI_NV_PDISP_RG_HDCP_AKSV_MSB				0x08
+#define HDMI_NV_PDISP_RG_HDCP_AKSV_LSB				0x09
+#define HDMI_NV_PDISP_RG_HDCP_BKSV_MSB				0x0a
+#define HDMI_NV_PDISP_RG_HDCP_BKSV_LSB				0x0b
+#define HDMI_NV_PDISP_RG_HDCP_CKSV_MSB				0x0c
+#define HDMI_NV_PDISP_RG_HDCP_CKSV_LSB				0x0d
+#define HDMI_NV_PDISP_RG_HDCP_DKSV_MSB				0x0e
+#define HDMI_NV_PDISP_RG_HDCP_DKSV_LSB				0x0f
+#define HDMI_NV_PDISP_RG_HDCP_CTRL				0x10
+#define HDMI_NV_PDISP_RG_HDCP_CMODE				0x11
+#define HDMI_NV_PDISP_RG_HDCP_MPRIME_MSB			0x12
+#define HDMI_NV_PDISP_RG_HDCP_MPRIME_LSB			0x13
+#define HDMI_NV_PDISP_RG_HDCP_SPRIME_MSB			0x14
+#define HDMI_NV_PDISP_RG_HDCP_SPRIME_LSB2			0x15
+#define HDMI_NV_PDISP_RG_HDCP_SPRIME_LSB1			0x16
+#define HDMI_NV_PDISP_RG_HDCP_RI				0x17
+#define HDMI_NV_PDISP_RG_HDCP_CS_MSB				0x18
+#define HDMI_NV_PDISP_RG_HDCP_CS_LSB				0x19
+#define HDMI_NV_PDISP_HDMI_AUDIO_EMU0				0x1a
+#define HDMI_NV_PDISP_HDMI_AUDIO_EMU_RDATA0			0x1b
+#define HDMI_NV_PDISP_HDMI_AUDIO_EMU1				0x1c
+#define HDMI_NV_PDISP_HDMI_AUDIO_EMU2				0x1d
+
+#define HDMI_NV_PDISP_HDMI_AUDIO_INFOFRAME_CTRL			0x1e
+#define HDMI_NV_PDISP_HDMI_AUDIO_INFOFRAME_STATUS		0x1f
+#define HDMI_NV_PDISP_HDMI_AUDIO_INFOFRAME_HEADER		0x20
+#define HDMI_NV_PDISP_HDMI_AUDIO_INFOFRAME_SUBPACK0_LOW		0x21
+#define HDMI_NV_PDISP_HDMI_AUDIO_INFOFRAME_SUBPACK0_HIGH	0x22
+#define HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_CTRL			0x23
+#define HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_STATUS			0x24
+#define HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_HEADER			0x25
+#define HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_SUBPACK0_LOW		0x26
+#define HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_SUBPACK0_HIGH		0x27
+#define HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_SUBPACK1_LOW		0x28
+#define HDMI_NV_PDISP_HDMI_AVI_INFOFRAME_SUBPACK1_HIGH		0x29
+
+#define INFOFRAME_CTRL_ENABLE (1 << 0)
+
+#define INFOFRAME_HEADER_TYPE(x)    (((x) & 0xff) <<  0)
+#define INFOFRAME_HEADER_VERSION(x) (((x) & 0xff) <<  8)
+#define INFOFRAME_HEADER_LEN(x)     (((x) & 0x0f) << 16)
+
+#define HDMI_NV_PDISP_HDMI_GENERIC_CTRL				0x2a
+#define GENERIC_CTRL_ENABLE (1 <<  0)
+#define GENERIC_CTRL_OTHER  (1 <<  4)
+#define GENERIC_CTRL_SINGLE (1 <<  8)
+#define GENERIC_CTRL_HBLANK (1 << 12)
+#define GENERIC_CTRL_AUDIO  (1 << 16)
+
+#define HDMI_NV_PDISP_HDMI_GENERIC_STATUS			0x2b
+#define HDMI_NV_PDISP_HDMI_GENERIC_HEADER			0x2c
+#define HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK0_LOW			0x2d
+#define HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK0_HIGH		0x2e
+#define HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK1_LOW			0x2f
+#define HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK1_HIGH		0x30
+#define HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK2_LOW			0x31
+#define HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK2_HIGH		0x32
+#define HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK3_LOW			0x33
+#define HDMI_NV_PDISP_HDMI_GENERIC_SUBPACK3_HIGH		0x34
+
+#define HDMI_NV_PDISP_HDMI_ACR_CTRL				0x35
+#define HDMI_NV_PDISP_HDMI_ACR_0320_SUBPACK_LOW			0x36
+#define HDMI_NV_PDISP_HDMI_ACR_0320_SUBPACK_HIGH		0x37
+#define HDMI_NV_PDISP_HDMI_ACR_0441_SUBPACK_LOW			0x38
+#define HDMI_NV_PDISP_HDMI_ACR_0441_SUBPACK_HIGH		0x39
+#define HDMI_NV_PDISP_HDMI_ACR_0882_SUBPACK_LOW			0x3a
+#define HDMI_NV_PDISP_HDMI_ACR_0882_SUBPACK_HIGH		0x3b
+#define HDMI_NV_PDISP_HDMI_ACR_1764_SUBPACK_LOW			0x3c
+#define HDMI_NV_PDISP_HDMI_ACR_1764_SUBPACK_HIGH		0x3d
+#define HDMI_NV_PDISP_HDMI_ACR_0480_SUBPACK_LOW			0x3e
+#define HDMI_NV_PDISP_HDMI_ACR_0480_SUBPACK_HIGH		0x3f
+#define HDMI_NV_PDISP_HDMI_ACR_0960_SUBPACK_LOW			0x40
+#define HDMI_NV_PDISP_HDMI_ACR_0960_SUBPACK_HIGH		0x41
+#define HDMI_NV_PDISP_HDMI_ACR_1920_SUBPACK_LOW			0x42
+#define HDMI_NV_PDISP_HDMI_ACR_1920_SUBPACK_HIGH		0x43
+
+#define ACR_SUBPACK_CTS(x) (((x) & 0xffffff) << 8)
+#define ACR_SUBPACK_N(x)   (((x) & 0xffffff) << 0)
+#define ACR_ENABLE         (1 << 31)
+
+#define HDMI_NV_PDISP_HDMI_CTRL					0x44
+#define HDMI_CTRL_REKEY(x)         (((x) & 0x7f) <<  0)
+#define HDMI_CTRL_MAX_AC_PACKET(x) (((x) & 0x1f) << 16)
+#define HDMI_CTRL_ENABLE           (1 << 30)
+
+#define HDMI_NV_PDISP_HDMI_VSYNC_KEEPOUT			0x45
+#define HDMI_NV_PDISP_HDMI_VSYNC_WINDOW				0x46
+#define VSYNC_WINDOW_END(x)   (((x) & 0x3ff) <<  0)
+#define VSYNC_WINDOW_START(x) (((x) & 0x3ff) << 16)
+#define VSYNC_WINDOW_ENABLE   (1 << 31)
+
+#define HDMI_NV_PDISP_HDMI_GCP_CTRL				0x47
+#define HDMI_NV_PDISP_HDMI_GCP_STATUS				0x48
+#define HDMI_NV_PDISP_HDMI_GCP_SUBPACK				0x49
+#define HDMI_NV_PDISP_HDMI_CHANNEL_STATUS1			0x4a
+#define HDMI_NV_PDISP_HDMI_CHANNEL_STATUS2			0x4b
+#define HDMI_NV_PDISP_HDMI_EMU0					0x4c
+#define HDMI_NV_PDISP_HDMI_EMU1					0x4d
+#define HDMI_NV_PDISP_HDMI_EMU1_RDATA				0x4e
+
+#define HDMI_NV_PDISP_HDMI_SPARE				0x4f
+#define SPARE_HW_CTS           (1 << 0)
+#define SPARE_FORCE_SW_CTS     (1 << 1)
+#define SPARE_CTS_RESET_VAL(x) (((x) & 0x7) << 16)
+
+#define HDMI_NV_PDISP_HDMI_SPDIF_CHN_STATUS1			0x50
+#define HDMI_NV_PDISP_HDMI_SPDIF_CHN_STATUS2			0x51
+#define HDMI_NV_PDISP_HDMI_HDCPRIF_ROM_CTRL			0x53
+#define HDMI_NV_PDISP_SOR_CAP					0x54
+#define HDMI_NV_PDISP_SOR_PWR					0x55
+#define SOR_PWR_NORMAL_STATE_PD     (0 <<  0)
+#define SOR_PWR_NORMAL_STATE_PU     (1 <<  0)
+#define SOR_PWR_NORMAL_START_NORMAL (0 <<  1)
+#define SOR_PWR_NORMAL_START_ALT    (1 <<  1)
+#define SOR_PWR_SAFE_STATE_PD       (0 << 16)
+#define SOR_PWR_SAFE_STATE_PU       (1 << 16)
+#define SOR_PWR_SETTING_NEW_DONE    (0 << 31)
+#define SOR_PWR_SETTING_NEW_PENDING (1 << 31)
+#define SOR_PWR_SETTING_NEW_TRIGGER (1 << 31)
+
+#define HDMI_NV_PDISP_SOR_TEST					0x56
+#define HDMI_NV_PDISP_SOR_PLL0					0x57
+#define SOR_PLL_PWR            (1 << 0)
+#define SOR_PLL_PDBG           (1 << 1)
+#define SOR_PLL_VCAPD          (1 << 2)
+#define SOR_PLL_PDPORT         (1 << 3)
+#define SOR_PLL_RESISTORSEL    (1 << 4)
+#define SOR_PLL_PULLDOWN       (1 << 5)
+#define SOR_PLL_VCOCAP(x)      (((x) & 0xf) <<  8)
+#define SOR_PLL_BG_V17_S(x)    (((x) & 0xf) << 12)
+#define SOR_PLL_FILTER(x)      (((x) & 0xf) << 16)
+#define SOR_PLL_ICHPMP(x)      (((x) & 0xf) << 24)
+#define SOR_PLL_TX_REG_LOAD(x) (((x) & 0xf) << 28)
+
+#define HDMI_NV_PDISP_SOR_PLL1					0x58
+#define SOR_PLL_TMDS_TERM_ENABLE (1 << 8)
+#define SOR_PLL_TMDS_TERMADJ(x)  (((x) & 0xf) <<  9)
+#define SOR_PLL_LOADADJ(x)       (((x) & 0xf) << 20)
+#define SOR_PLL_PE_EN            (1 << 28)
+#define SOR_PLL_HALF_FULL_PE     (1 << 29)
+#define SOR_PLL_S_D_PIN_PE       (1 << 30)
+
+#define HDMI_NV_PDISP_SOR_PLL2					0x59
+
+#define HDMI_NV_PDISP_SOR_CSTM					0x5a
+#define SOR_CSTM_ROTCLK(x) (((x) & 0xf) << 24)
+
+#define HDMI_NV_PDISP_SOR_LVDS					0x5b
+#define HDMI_NV_PDISP_SOR_CRCA					0x5c
+#define HDMI_NV_PDISP_SOR_CRCB					0x5d
+#define HDMI_NV_PDISP_SOR_BLANK					0x5e
+#define HDMI_NV_PDISP_SOR_SEQ_CTL				0x5f
+#define SOR_SEQ_CTL_PU_PC(x) (((x) & 0xf) <<  0)
+#define SOR_SEQ_PU_PC_ALT(x) (((x) & 0xf) <<  4)
+#define SOR_SEQ_PD_PC(x)     (((x) & 0xf) <<  8)
+#define SOR_SEQ_PD_PC_ALT(x) (((x) & 0xf) << 12)
+#define SOR_SEQ_PC(x)        (((x) & 0xf) << 16)
+#define SOR_SEQ_STATUS       (1 << 28)
+#define SOR_SEQ_SWITCH       (1 << 30)
+
+#define HDMI_NV_PDISP_SOR_SEQ_INST(x)				(0x60 + (x))
+
+#define SOR_SEQ_INST_WAIT_TIME(x)     (((x) & 0x3ff) << 0)
+#define SOR_SEQ_INST_WAIT_UNITS_VSYNC (2 << 12)
+#define SOR_SEQ_INST_HALT             (1 << 15)
+#define SOR_SEQ_INST_PIN_A_LOW        (0 << 21)
+#define SOR_SEQ_INST_PIN_A_HIGH       (1 << 21)
+#define SOR_SEQ_INST_PIN_B_LOW        (0 << 22)
+#define SOR_SEQ_INST_PIN_B_HIGH       (1 << 22)
+#define SOR_SEQ_INST_DRIVE_PWM_OUT_LO (1 << 23)
+
+#define HDMI_NV_PDISP_SOR_VCRCA0				0x72
+#define HDMI_NV_PDISP_SOR_VCRCA1				0x73
+#define HDMI_NV_PDISP_SOR_CCRCA0				0x74
+#define HDMI_NV_PDISP_SOR_CCRCA1				0x75
+#define HDMI_NV_PDISP_SOR_EDATAA0				0x76
+#define HDMI_NV_PDISP_SOR_EDATAA1				0x77
+#define HDMI_NV_PDISP_SOR_COUNTA0				0x78
+#define HDMI_NV_PDISP_SOR_COUNTA1				0x79
+#define HDMI_NV_PDISP_SOR_DEBUGA0				0x7a
+#define HDMI_NV_PDISP_SOR_DEBUGA1				0x7b
+#define HDMI_NV_PDISP_SOR_TRIG					0x7c
+#define HDMI_NV_PDISP_SOR_MSCHECK				0x7d
+
+#define HDMI_NV_PDISP_SOR_LANE_DRIVE_CURRENT			0x7e
+#define DRIVE_CURRENT_LANE0(x)      (((x) & 0x3f) <<  0)
+#define DRIVE_CURRENT_LANE1(x)      (((x) & 0x3f) <<  8)
+#define DRIVE_CURRENT_LANE2(x)      (((x) & 0x3f) << 16)
+#define DRIVE_CURRENT_LANE3(x)      (((x) & 0x3f) << 24)
+#define DRIVE_CURRENT_FUSE_OVERRIDE (1 << 31)
+
+#define DRIVE_CURRENT_1_500_mA  0x00
+#define DRIVE_CURRENT_1_875_mA  0x01
+#define DRIVE_CURRENT_2_250_mA  0x02
+#define DRIVE_CURRENT_2_625_mA  0x03
+#define DRIVE_CURRENT_3_000_mA  0x04
+#define DRIVE_CURRENT_3_375_mA  0x05
+#define DRIVE_CURRENT_3_750_mA  0x06
+#define DRIVE_CURRENT_4_125_mA  0x07
+#define DRIVE_CURRENT_4_500_mA  0x08
+#define DRIVE_CURRENT_4_875_mA  0x09
+#define DRIVE_CURRENT_5_250_mA  0x0a
+#define DRIVE_CURRENT_5_625_mA  0x0b
+#define DRIVE_CURRENT_6_000_mA  0x0c
+#define DRIVE_CURRENT_6_375_mA  0x0d
+#define DRIVE_CURRENT_6_750_mA  0x0e
+#define DRIVE_CURRENT_7_125_mA  0x0f
+#define DRIVE_CURRENT_7_500_mA  0x10
+#define DRIVE_CURRENT_7_875_mA  0x11
+#define DRIVE_CURRENT_8_250_mA  0x12
+#define DRIVE_CURRENT_8_625_mA  0x13
+#define DRIVE_CURRENT_9_000_mA  0x14
+#define DRIVE_CURRENT_9_375_mA  0x15
+#define DRIVE_CURRENT_9_750_mA  0x16
+#define DRIVE_CURRENT_10_125_mA 0x17
+#define DRIVE_CURRENT_10_500_mA 0x18
+#define DRIVE_CURRENT_10_875_mA 0x19
+#define DRIVE_CURRENT_11_250_mA 0x1a
+#define DRIVE_CURRENT_11_625_mA 0x1b
+#define DRIVE_CURRENT_12_000_mA 0x1c
+#define DRIVE_CURRENT_12_375_mA 0x1d
+#define DRIVE_CURRENT_12_750_mA 0x1e
+#define DRIVE_CURRENT_13_125_mA 0x1f
+#define DRIVE_CURRENT_13_500_mA 0x20
+#define DRIVE_CURRENT_13_875_mA 0x21
+#define DRIVE_CURRENT_14_250_mA 0x22
+#define DRIVE_CURRENT_14_625_mA 0x23
+#define DRIVE_CURRENT_15_000_mA 0x24
+#define DRIVE_CURRENT_15_375_mA 0x25
+#define DRIVE_CURRENT_15_750_mA 0x26
+#define DRIVE_CURRENT_16_125_mA 0x27
+#define DRIVE_CURRENT_16_500_mA 0x28
+#define DRIVE_CURRENT_16_875_mA 0x29
+#define DRIVE_CURRENT_17_250_mA 0x2a
+#define DRIVE_CURRENT_17_625_mA 0x2b
+#define DRIVE_CURRENT_18_000_mA 0x2c
+#define DRIVE_CURRENT_18_375_mA 0x2d
+#define DRIVE_CURRENT_18_750_mA 0x2e
+#define DRIVE_CURRENT_19_125_mA 0x2f
+#define DRIVE_CURRENT_19_500_mA 0x30
+#define DRIVE_CURRENT_19_875_mA 0x31
+#define DRIVE_CURRENT_20_250_mA 0x32
+#define DRIVE_CURRENT_20_625_mA 0x33
+#define DRIVE_CURRENT_21_000_mA 0x34
+#define DRIVE_CURRENT_21_375_mA 0x35
+#define DRIVE_CURRENT_21_750_mA 0x36
+#define DRIVE_CURRENT_22_125_mA 0x37
+#define DRIVE_CURRENT_22_500_mA 0x38
+#define DRIVE_CURRENT_22_875_mA 0x39
+#define DRIVE_CURRENT_23_250_mA 0x3a
+#define DRIVE_CURRENT_23_625_mA 0x3b
+#define DRIVE_CURRENT_24_000_mA 0x3c
+#define DRIVE_CURRENT_24_375_mA 0x3d
+#define DRIVE_CURRENT_24_750_mA 0x3e
+
+#define HDMI_NV_PDISP_AUDIO_DEBUG0				0x7f
+#define HDMI_NV_PDISP_AUDIO_DEBUG1				0x80
+#define HDMI_NV_PDISP_AUDIO_DEBUG2				0x81
+
+#define HDMI_NV_PDISP_AUDIO_FS(x)				(0x82 + (x))
+#define AUDIO_FS_LOW(x)  (((x) & 0xfff) <<  0)
+#define AUDIO_FS_HIGH(x) (((x) & 0xfff) << 16)
+
+#define HDMI_NV_PDISP_AUDIO_PULSE_WIDTH				0x89
+#define HDMI_NV_PDISP_AUDIO_THRESHOLD				0x8a
+#define HDMI_NV_PDISP_AUDIO_CNTRL0				0x8b
+#define AUDIO_CNTRL0_ERROR_TOLERANCE(x)  (((x) & 0xff) << 0)
+#define AUDIO_CNTRL0_SOURCE_SELECT_AUTO  (0 << 20)
+#define AUDIO_CNTRL0_SOURCE_SELECT_SPDIF (1 << 20)
+#define AUDIO_CNTRL0_SOURCE_SELECT_HDAL  (2 << 20)
+#define AUDIO_CNTRL0_FRAMES_PER_BLOCK(x) (((x) & 0xff) << 24)
+
+#define HDMI_NV_PDISP_AUDIO_N					0x8c
+#define AUDIO_N_VALUE(x)           (((x) & 0xfffff) << 0)
+#define AUDIO_N_RESETF             (1 << 20)
+#define AUDIO_N_GENERATE_NORMAL    (0 << 24)
+#define AUDIO_N_GENERATE_ALTERNATE (1 << 24)
+
+#define HDMI_NV_PDISP_HDCPRIF_ROM_TIMING			0x94
+#define HDMI_NV_PDISP_SOR_REFCLK				0x95
+#define SOR_REFCLK_DIV_INT(x)  (((x) & 0xff) << 8)
+#define SOR_REFCLK_DIV_FRAC(x) (((x) & 0x03) << 6)
+
+#define HDMI_NV_PDISP_CRC_CONTROL				0x96
+#define HDMI_NV_PDISP_INPUT_CONTROL				0x97
+#define HDMI_SRC_DISPLAYA       (0 << 0)
+#define HDMI_SRC_DISPLAYB       (1 << 0)
+#define ARM_VIDEO_RANGE_FULL    (0 << 1)
+#define ARM_VIDEO_RANGE_LIMITED (1 << 1)
+
+#define HDMI_NV_PDISP_SCRATCH					0x98
+#define HDMI_NV_PDISP_PE_CURRENT				0x99
+#define PE_CURRENT0(x) (((x) & 0xf) << 0)
+#define PE_CURRENT1(x) (((x) & 0xf) << 8)
+#define PE_CURRENT2(x) (((x) & 0xf) << 16)
+#define PE_CURRENT3(x) (((x) & 0xf) << 24)
+
+#define PE_CURRENT_0_0_mA 0x0
+#define PE_CURRENT_0_5_mA 0x1
+#define PE_CURRENT_1_0_mA 0x2
+#define PE_CURRENT_1_5_mA 0x3
+#define PE_CURRENT_2_0_mA 0x4
+#define PE_CURRENT_2_5_mA 0x5
+#define PE_CURRENT_3_0_mA 0x6
+#define PE_CURRENT_3_5_mA 0x7
+#define PE_CURRENT_4_0_mA 0x8
+#define PE_CURRENT_4_5_mA 0x9
+#define PE_CURRENT_5_0_mA 0xa
+#define PE_CURRENT_5_5_mA 0xb
+#define PE_CURRENT_6_0_mA 0xc
+#define PE_CURRENT_6_5_mA 0xd
+#define PE_CURRENT_7_0_mA 0xe
+#define PE_CURRENT_7_5_mA 0xf
+
+#define HDMI_NV_PDISP_KEY_CTRL					0x9a
+#define HDMI_NV_PDISP_KEY_DEBUG0				0x9b
+#define HDMI_NV_PDISP_KEY_DEBUG1				0x9c
+#define HDMI_NV_PDISP_KEY_DEBUG2				0x9d
+#define HDMI_NV_PDISP_KEY_HDCP_KEY_0				0x9e
+#define HDMI_NV_PDISP_KEY_HDCP_KEY_1				0x9f
+#define HDMI_NV_PDISP_KEY_HDCP_KEY_2				0xa0
+#define HDMI_NV_PDISP_KEY_HDCP_KEY_3				0xa1
+#define HDMI_NV_PDISP_KEY_HDCP_KEY_TRIG				0xa2
+#define HDMI_NV_PDISP_KEY_SKEY_INDEX				0xa3
+
+#define HDMI_NV_PDISP_SOR_AUDIO_CNTRL0				0xac
+#define AUDIO_CNTRL0_INJECT_NULLSMPL (1 << 29)
+#define HDMI_NV_PDISP_SOR_AUDIO_HDA_ELD_BUFWR			0xbc
+#define HDMI_NV_PDISP_SOR_AUDIO_HDA_PRESENSE			0xbd
+
+#define HDMI_NV_PDISP_SOR_AUDIO_AVAL_0320    0xbf
+#define HDMI_NV_PDISP_SOR_AUDIO_AVAL_0441    0xc0
+#define HDMI_NV_PDISP_SOR_AUDIO_AVAL_0882    0xc1
+#define HDMI_NV_PDISP_SOR_AUDIO_AVAL_1764    0xc2
+#define HDMI_NV_PDISP_SOR_AUDIO_AVAL_0480    0xc3
+#define HDMI_NV_PDISP_SOR_AUDIO_AVAL_0960    0xc4
+#define HDMI_NV_PDISP_SOR_AUDIO_AVAL_1920    0xc5
+#define HDMI_NV_PDISP_SOR_AUDIO_AVAL_DEFAULT 0xc5
+
+#endif /* TEGRA_HDMI_H */
diff --git a/drivers/gpu/host1x/drm/output.c b/drivers/gpu/host1x/drm/output.c
new file mode 100644
index 00000000000..8140fc6c34d
--- /dev/null
+++ b/drivers/gpu/host1x/drm/output.c
@@ -0,0 +1,272 @@
+/*
+ * Copyright (C) 2012 Avionic Design GmbH
+ * Copyright (C) 2012 NVIDIA CORPORATION.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/module.h>
+#include <linux/of_gpio.h>
+#include <linux/of_i2c.h>
+
+#include "drm.h"
+
+static int tegra_connector_get_modes(struct drm_connector *connector)
+{
+	struct tegra_output *output = connector_to_output(connector);
+	struct edid *edid = NULL;
+	int err = 0;
+
+	if (output->edid)
+		edid = kmemdup(output->edid, sizeof(*edid), GFP_KERNEL);
+	else if (output->ddc)
+		edid = drm_get_edid(connector, output->ddc);
+
+	drm_mode_connector_update_edid_property(connector, edid);
+
+	if (edid) {
+		err = drm_add_edid_modes(connector, edid);
+		kfree(edid);
+	}
+
+	return err;
+}
+
+static int tegra_connector_mode_valid(struct drm_connector *connector,
+				      struct drm_display_mode *mode)
+{
+	struct tegra_output *output = connector_to_output(connector);
+	enum drm_mode_status status = MODE_OK;
+	int err;
+
+	err = tegra_output_check_mode(output, mode, &status);
+	if (err < 0)
+		return MODE_ERROR;
+
+	return status;
+}
+
+static struct drm_encoder *
+tegra_connector_best_encoder(struct drm_connector *connector)
+{
+	struct tegra_output *output = connector_to_output(connector);
+
+	return &output->encoder;
+}
+
+static const struct drm_connector_helper_funcs connector_helper_funcs = {
+	.get_modes = tegra_connector_get_modes,
+	.mode_valid = tegra_connector_mode_valid,
+	.best_encoder = tegra_connector_best_encoder,
+};
+
+static enum drm_connector_status
+tegra_connector_detect(struct drm_connector *connector, bool force)
+{
+	struct tegra_output *output = connector_to_output(connector);
+	enum drm_connector_status status = connector_status_unknown;
+
+	if (gpio_is_valid(output->hpd_gpio)) {
+		if (gpio_get_value(output->hpd_gpio) == 0)
+			status = connector_status_disconnected;
+		else
+			status = connector_status_connected;
+	} else {
+		if (connector->connector_type == DRM_MODE_CONNECTOR_LVDS)
+			status = connector_status_connected;
+	}
+
+	return status;
+}
+
+static void tegra_connector_destroy(struct drm_connector *connector)
+{
+	drm_sysfs_connector_remove(connector);
+	drm_connector_cleanup(connector);
+}
+
+static const struct drm_connector_funcs connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.detect = tegra_connector_detect,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = tegra_connector_destroy,
+};
+
+static void tegra_encoder_destroy(struct drm_encoder *encoder)
+{
+	drm_encoder_cleanup(encoder);
+}
+
+static const struct drm_encoder_funcs encoder_funcs = {
+	.destroy = tegra_encoder_destroy,
+};
+
+static void tegra_encoder_dpms(struct drm_encoder *encoder, int mode)
+{
+}
+
+static bool tegra_encoder_mode_fixup(struct drm_encoder *encoder,
+				     const struct drm_display_mode *mode,
+				     struct drm_display_mode *adjusted)
+{
+	return true;
+}
+
+static void tegra_encoder_prepare(struct drm_encoder *encoder)
+{
+}
+
+static void tegra_encoder_commit(struct drm_encoder *encoder)
+{
+}
+
+static void tegra_encoder_mode_set(struct drm_encoder *encoder,
+				   struct drm_display_mode *mode,
+				   struct drm_display_mode *adjusted)
+{
+	struct tegra_output *output = encoder_to_output(encoder);
+	int err;
+
+	err = tegra_output_enable(output);
+	if (err < 0)
+		dev_err(encoder->dev->dev, "tegra_output_enable(): %d\n", err);
+}
+
+static const struct drm_encoder_helper_funcs encoder_helper_funcs = {
+	.dpms = tegra_encoder_dpms,
+	.mode_fixup = tegra_encoder_mode_fixup,
+	.prepare = tegra_encoder_prepare,
+	.commit = tegra_encoder_commit,
+	.mode_set = tegra_encoder_mode_set,
+};
+
+static irqreturn_t hpd_irq(int irq, void *data)
+{
+	struct tegra_output *output = data;
+
+	drm_helper_hpd_irq_event(output->connector.dev);
+
+	return IRQ_HANDLED;
+}
+
+int tegra_output_parse_dt(struct tegra_output *output)
+{
+	enum of_gpio_flags flags;
+	struct device_node *ddc;
+	size_t size;
+	int err;
+
+	if (!output->of_node)
+		output->of_node = output->dev->of_node;
+
+	output->edid = of_get_property(output->of_node, "nvidia,edid", &size);
+
+	ddc = of_parse_phandle(output->of_node, "nvidia,ddc-i2c-bus", 0);
+	if (ddc) {
+		output->ddc = of_find_i2c_adapter_by_node(ddc);
+		if (!output->ddc) {
+			err = -EPROBE_DEFER;
+			of_node_put(ddc);
+			return err;
+		}
+
+		of_node_put(ddc);
+	}
+
+	if (!output->edid && !output->ddc)
+		return -ENODEV;
+
+	output->hpd_gpio = of_get_named_gpio_flags(output->of_node,
+						   "nvidia,hpd-gpio", 0,
+						   &flags);
+
+	return 0;
+}
+
+int tegra_output_init(struct drm_device *drm, struct tegra_output *output)
+{
+	int connector, encoder, err;
+
+	if (gpio_is_valid(output->hpd_gpio)) {
+		unsigned long flags;
+
+		err = gpio_request_one(output->hpd_gpio, GPIOF_DIR_IN,
+				       "HDMI hotplug detect");
+		if (err < 0) {
+			dev_err(output->dev, "gpio_request_one(): %d\n", err);
+			return err;
+		}
+
+		err = gpio_to_irq(output->hpd_gpio);
+		if (err < 0) {
+			dev_err(output->dev, "gpio_to_irq(): %d\n", err);
+			goto free_hpd;
+		}
+
+		output->hpd_irq = err;
+
+		flags = IRQF_TRIGGER_RISING | IRQF_TRIGGER_FALLING |
+			IRQF_ONESHOT;
+
+		err = request_threaded_irq(output->hpd_irq, NULL, hpd_irq,
+					   flags, "hpd", output);
+		if (err < 0) {
+			dev_err(output->dev, "failed to request IRQ#%u: %d\n",
+				output->hpd_irq, err);
+			goto free_hpd;
+		}
+
+		output->connector.polled = DRM_CONNECTOR_POLL_HPD;
+	}
+
+	switch (output->type) {
+	case TEGRA_OUTPUT_RGB:
+		connector = DRM_MODE_CONNECTOR_LVDS;
+		encoder = DRM_MODE_ENCODER_LVDS;
+		break;
+
+	case TEGRA_OUTPUT_HDMI:
+		connector = DRM_MODE_CONNECTOR_HDMIA;
+		encoder = DRM_MODE_ENCODER_TMDS;
+		break;
+
+	default:
+		connector = DRM_MODE_CONNECTOR_Unknown;
+		encoder = DRM_MODE_ENCODER_NONE;
+		break;
+	}
+
+	drm_connector_init(drm, &output->connector, &connector_funcs,
+			   connector);
+	drm_connector_helper_add(&output->connector, &connector_helper_funcs);
+
+	drm_encoder_init(drm, &output->encoder, &encoder_funcs, encoder);
+	drm_encoder_helper_add(&output->encoder, &encoder_helper_funcs);
+
+	drm_mode_connector_attach_encoder(&output->connector, &output->encoder);
+	drm_sysfs_connector_add(&output->connector);
+
+	output->encoder.possible_crtcs = 0x3;
+
+	return 0;
+
+free_hpd:
+	gpio_free(output->hpd_gpio);
+
+	return err;
+}
+
+int tegra_output_exit(struct tegra_output *output)
+{
+	if (gpio_is_valid(output->hpd_gpio)) {
+		free_irq(output->hpd_irq, output);
+		gpio_free(output->hpd_gpio);
+	}
+
+	if (output->ddc)
+		put_device(&output->ddc->dev);
+
+	return 0;
+}
diff --git a/drivers/gpu/host1x/drm/rgb.c b/drivers/gpu/host1x/drm/rgb.c
new file mode 100644
index 00000000000..ed4416f2026
--- /dev/null
+++ b/drivers/gpu/host1x/drm/rgb.c
@@ -0,0 +1,228 @@
+/*
+ * Copyright (C) 2012 Avionic Design GmbH
+ * Copyright (C) 2012 NVIDIA CORPORATION.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/clk.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+
+#include "drm.h"
+#include "dc.h"
+
+struct tegra_rgb {
+	struct tegra_output output;
+	struct clk *clk_parent;
+	struct clk *clk;
+};
+
+static inline struct tegra_rgb *to_rgb(struct tegra_output *output)
+{
+	return container_of(output, struct tegra_rgb, output);
+}
+
+struct reg_entry {
+	unsigned long offset;
+	unsigned long value;
+};
+
+static const struct reg_entry rgb_enable[] = {
+	{ DC_COM_PIN_OUTPUT_ENABLE(0),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_ENABLE(1),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_ENABLE(2),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_ENABLE(3),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_POLARITY(0), 0x00000000 },
+	{ DC_COM_PIN_OUTPUT_POLARITY(1), 0x01000000 },
+	{ DC_COM_PIN_OUTPUT_POLARITY(2), 0x00000000 },
+	{ DC_COM_PIN_OUTPUT_POLARITY(3), 0x00000000 },
+	{ DC_COM_PIN_OUTPUT_DATA(0),     0x00000000 },
+	{ DC_COM_PIN_OUTPUT_DATA(1),     0x00000000 },
+	{ DC_COM_PIN_OUTPUT_DATA(2),     0x00000000 },
+	{ DC_COM_PIN_OUTPUT_DATA(3),     0x00000000 },
+	{ DC_COM_PIN_OUTPUT_SELECT(0),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_SELECT(1),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_SELECT(2),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_SELECT(3),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_SELECT(4),   0x00210222 },
+	{ DC_COM_PIN_OUTPUT_SELECT(5),   0x00002200 },
+	{ DC_COM_PIN_OUTPUT_SELECT(6),   0x00020000 },
+};
+
+static const struct reg_entry rgb_disable[] = {
+	{ DC_COM_PIN_OUTPUT_SELECT(6),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_SELECT(5),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_SELECT(4),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_SELECT(3),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_SELECT(2),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_SELECT(1),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_SELECT(0),   0x00000000 },
+	{ DC_COM_PIN_OUTPUT_DATA(3),     0xaaaaaaaa },
+	{ DC_COM_PIN_OUTPUT_DATA(2),     0xaaaaaaaa },
+	{ DC_COM_PIN_OUTPUT_DATA(1),     0xaaaaaaaa },
+	{ DC_COM_PIN_OUTPUT_DATA(0),     0xaaaaaaaa },
+	{ DC_COM_PIN_OUTPUT_POLARITY(3), 0x00000000 },
+	{ DC_COM_PIN_OUTPUT_POLARITY(2), 0x00000000 },
+	{ DC_COM_PIN_OUTPUT_POLARITY(1), 0x00000000 },
+	{ DC_COM_PIN_OUTPUT_POLARITY(0), 0x00000000 },
+	{ DC_COM_PIN_OUTPUT_ENABLE(3),   0x55555555 },
+	{ DC_COM_PIN_OUTPUT_ENABLE(2),   0x55555555 },
+	{ DC_COM_PIN_OUTPUT_ENABLE(1),   0x55150005 },
+	{ DC_COM_PIN_OUTPUT_ENABLE(0),   0x55555555 },
+};
+
+static void tegra_dc_write_regs(struct tegra_dc *dc,
+				const struct reg_entry *table,
+				unsigned int num)
+{
+	unsigned int i;
+
+	for (i = 0; i < num; i++)
+		tegra_dc_writel(dc, table[i].value, table[i].offset);
+}
+
+static int tegra_output_rgb_enable(struct tegra_output *output)
+{
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
+
+	tegra_dc_write_regs(dc, rgb_enable, ARRAY_SIZE(rgb_enable));
+
+	return 0;
+}
+
+static int tegra_output_rgb_disable(struct tegra_output *output)
+{
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
+
+	tegra_dc_write_regs(dc, rgb_disable, ARRAY_SIZE(rgb_disable));
+
+	return 0;
+}
+
+static int tegra_output_rgb_setup_clock(struct tegra_output *output,
+					struct clk *clk, unsigned long pclk)
+{
+	struct tegra_rgb *rgb = to_rgb(output);
+
+	return clk_set_parent(clk, rgb->clk_parent);
+}
+
+static int tegra_output_rgb_check_mode(struct tegra_output *output,
+				       struct drm_display_mode *mode,
+				       enum drm_mode_status *status)
+{
+	/*
+	 * FIXME: For now, always assume that the mode is okay. There are
+	 * unresolved issues with clk_round_rate(), which doesn't always
+	 * reliably report whether a frequency can be set or not.
+	 */
+
+	*status = MODE_OK;
+
+	return 0;
+}
+
+static const struct tegra_output_ops rgb_ops = {
+	.enable = tegra_output_rgb_enable,
+	.disable = tegra_output_rgb_disable,
+	.setup_clock = tegra_output_rgb_setup_clock,
+	.check_mode = tegra_output_rgb_check_mode,
+};
+
+int tegra_dc_rgb_probe(struct tegra_dc *dc)
+{
+	struct device_node *np;
+	struct tegra_rgb *rgb;
+	int err;
+
+	np = of_get_child_by_name(dc->dev->of_node, "rgb");
+	if (!np || !of_device_is_available(np))
+		return -ENODEV;
+
+	rgb = devm_kzalloc(dc->dev, sizeof(*rgb), GFP_KERNEL);
+	if (!rgb)
+		return -ENOMEM;
+
+	rgb->clk = devm_clk_get(dc->dev, NULL);
+	if (IS_ERR(rgb->clk)) {
+		dev_err(dc->dev, "failed to get clock\n");
+		return PTR_ERR(rgb->clk);
+	}
+
+	rgb->clk_parent = devm_clk_get(dc->dev, "parent");
+	if (IS_ERR(rgb->clk_parent)) {
+		dev_err(dc->dev, "failed to get parent clock\n");
+		return PTR_ERR(rgb->clk_parent);
+	}
+
+	err = clk_set_parent(rgb->clk, rgb->clk_parent);
+	if (err < 0) {
+		dev_err(dc->dev, "failed to set parent clock: %d\n", err);
+		return err;
+	}
+
+	rgb->output.dev = dc->dev;
+	rgb->output.of_node = np;
+
+	err = tegra_output_parse_dt(&rgb->output);
+	if (err < 0)
+		return err;
+
+	dc->rgb = &rgb->output;
+
+	return 0;
+}
+
+int tegra_dc_rgb_init(struct drm_device *drm, struct tegra_dc *dc)
+{
+	struct tegra_rgb *rgb = to_rgb(dc->rgb);
+	int err;
+
+	if (!dc->rgb)
+		return -ENODEV;
+
+	rgb->output.type = TEGRA_OUTPUT_RGB;
+	rgb->output.ops = &rgb_ops;
+
+	err = tegra_output_init(dc->base.dev, &rgb->output);
+	if (err < 0) {
+		dev_err(dc->dev, "output setup failed: %d\n", err);
+		return err;
+	}
+
+	/*
+	 * By default, outputs can be associated with each display controller.
+	 * RGB outputs are an exception, so we make sure they can be attached
+	 * to only their parent display controller.
+	 */
+	rgb->output.encoder.possible_crtcs = 1 << dc->pipe;
+
+	return 0;
+}
+
+int tegra_dc_rgb_exit(struct tegra_dc *dc)
+{
+	if (dc->rgb) {
+		int err;
+
+		err = tegra_output_disable(dc->rgb);
+		if (err < 0) {
+			dev_err(dc->dev, "output failed to disable: %d\n", err);
+			return err;
+		}
+
+		err = tegra_output_exit(dc->rgb);
+		if (err < 0) {
+			dev_err(dc->dev, "output cleanup failed: %d\n", err);
+			return err;
+		}
+
+		dc->rgb = NULL;
+	}
+
+	return 0;
+}
diff --git a/drivers/gpu/host1x/host1x.h b/drivers/gpu/host1x/host1x.h
new file mode 100644
index 00000000000..a2bc1e65e97
--- /dev/null
+++ b/drivers/gpu/host1x/host1x.h
@@ -0,0 +1,30 @@
+/*
+ * Tegra host1x driver
+ *
+ * Copyright (c) 2009-2013, NVIDIA Corporation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ */
+
+#ifndef __LINUX_HOST1X_H
+#define __LINUX_HOST1X_H
+
+enum host1x_class {
+	HOST1X_CLASS_HOST1X	= 0x1,
+	HOST1X_CLASS_GR2D	= 0x51,
+	HOST1X_CLASS_GR2D_SB    = 0x52
+};
+
+#endif
diff --git a/drivers/gpu/host1x/host1x_bo.h b/drivers/gpu/host1x/host1x_bo.h
new file mode 100644
index 00000000000..4c1f10bd773
--- /dev/null
+++ b/drivers/gpu/host1x/host1x_bo.h
@@ -0,0 +1,87 @@
+/*
+ * Tegra host1x Memory Management Abstraction header
+ *
+ * Copyright (c) 2012-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef _HOST1X_BO_H
+#define _HOST1X_BO_H
+
+struct host1x_bo;
+
+struct host1x_bo_ops {
+	struct host1x_bo *(*get)(struct host1x_bo *bo);
+	void (*put)(struct host1x_bo *bo);
+	dma_addr_t (*pin)(struct host1x_bo *bo, struct sg_table **sgt);
+	void (*unpin)(struct host1x_bo *bo, struct sg_table *sgt);
+	void *(*mmap)(struct host1x_bo *bo);
+	void (*munmap)(struct host1x_bo *bo, void *addr);
+	void *(*kmap)(struct host1x_bo *bo, unsigned int pagenum);
+	void (*kunmap)(struct host1x_bo *bo, unsigned int pagenum, void *addr);
+};
+
+struct host1x_bo {
+	const struct host1x_bo_ops *ops;
+};
+
+static inline void host1x_bo_init(struct host1x_bo *bo,
+				  const struct host1x_bo_ops *ops)
+{
+	bo->ops = ops;
+}
+
+static inline struct host1x_bo *host1x_bo_get(struct host1x_bo *bo)
+{
+	return bo->ops->get(bo);
+}
+
+static inline void host1x_bo_put(struct host1x_bo *bo)
+{
+	bo->ops->put(bo);
+}
+
+static inline dma_addr_t host1x_bo_pin(struct host1x_bo *bo,
+				       struct sg_table **sgt)
+{
+	return bo->ops->pin(bo, sgt);
+}
+
+static inline void host1x_bo_unpin(struct host1x_bo *bo, struct sg_table *sgt)
+{
+	bo->ops->unpin(bo, sgt);
+}
+
+static inline void *host1x_bo_mmap(struct host1x_bo *bo)
+{
+	return bo->ops->mmap(bo);
+}
+
+static inline void host1x_bo_munmap(struct host1x_bo *bo, void *addr)
+{
+	bo->ops->munmap(bo, addr);
+}
+
+static inline void *host1x_bo_kmap(struct host1x_bo *bo, unsigned int pagenum)
+{
+	return bo->ops->kmap(bo, pagenum);
+}
+
+static inline void host1x_bo_kunmap(struct host1x_bo *bo,
+				    unsigned int pagenum, void *addr)
+{
+	bo->ops->kunmap(bo, pagenum, addr);
+}
+
+#endif
diff --git a/drivers/gpu/host1x/host1x_client.h b/drivers/gpu/host1x/host1x_client.h
new file mode 100644
index 00000000000..9b85f10f4a4
--- /dev/null
+++ b/drivers/gpu/host1x/host1x_client.h
@@ -0,0 +1,35 @@
+/*
+ * Copyright (c) 2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef HOST1X_CLIENT_H
+#define HOST1X_CLIENT_H
+
+struct device;
+struct platform_device;
+
+#ifdef CONFIG_DRM_TEGRA
+int host1x_drm_alloc(struct platform_device *pdev);
+#else
+static inline int host1x_drm_alloc(struct platform_device *pdev)
+{
+	return 0;
+}
+#endif
+
+void host1x_set_drm_data(struct device *dev, void *data);
+void *host1x_get_drm_data(struct device *dev);
+
+#endif
diff --git a/drivers/gpu/host1x/hw/Makefile b/drivers/gpu/host1x/hw/Makefile
new file mode 100644
index 00000000000..9b50863a223
--- /dev/null
+++ b/drivers/gpu/host1x/hw/Makefile
@@ -0,0 +1,6 @@
+ccflags-y = -Idrivers/gpu/host1x
+
+host1x-hw-objs  = \
+	host1x01.o
+
+obj-$(CONFIG_TEGRA_HOST1X) += host1x-hw.o
diff --git a/drivers/gpu/host1x/hw/cdma_hw.c b/drivers/gpu/host1x/hw/cdma_hw.c
new file mode 100644
index 00000000000..590b69d91da
--- /dev/null
+++ b/drivers/gpu/host1x/hw/cdma_hw.c
@@ -0,0 +1,326 @@
+/*
+ * Tegra host1x Command DMA
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/slab.h>
+#include <linux/scatterlist.h>
+#include <linux/dma-mapping.h>
+
+#include "cdma.h"
+#include "channel.h"
+#include "dev.h"
+#include "debug.h"
+
+/*
+ * Put the restart at the end of pushbuffer memor
+ */
+static void push_buffer_init(struct push_buffer *pb)
+{
+	*(pb->mapped + (pb->size_bytes >> 2)) = host1x_opcode_restart(0);
+}
+
+/*
+ * Increment timedout buffer's syncpt via CPU.
+ */
+static void cdma_timeout_cpu_incr(struct host1x_cdma *cdma, u32 getptr,
+				u32 syncpt_incrs, u32 syncval, u32 nr_slots)
+{
+	struct host1x *host1x = cdma_to_host1x(cdma);
+	struct push_buffer *pb = &cdma->push_buffer;
+	u32 i;
+
+	for (i = 0; i < syncpt_incrs; i++)
+		host1x_syncpt_cpu_incr(cdma->timeout.syncpt);
+
+	/* after CPU incr, ensure shadow is up to date */
+	host1x_syncpt_load(cdma->timeout.syncpt);
+
+	/* NOP all the PB slots */
+	while (nr_slots--) {
+		u32 *p = (u32 *)((u32)pb->mapped + getptr);
+		*(p++) = HOST1X_OPCODE_NOP;
+		*(p++) = HOST1X_OPCODE_NOP;
+		dev_dbg(host1x->dev, "%s: NOP at 0x%x\n", __func__,
+			pb->phys + getptr);
+		getptr = (getptr + 8) & (pb->size_bytes - 1);
+	}
+	wmb();
+}
+
+/*
+ * Start channel DMA
+ */
+static void cdma_start(struct host1x_cdma *cdma)
+{
+	struct host1x_channel *ch = cdma_to_channel(cdma);
+
+	if (cdma->running)
+		return;
+
+	cdma->last_pos = cdma->push_buffer.pos;
+
+	host1x_ch_writel(ch, HOST1X_CHANNEL_DMACTRL_DMASTOP,
+			 HOST1X_CHANNEL_DMACTRL);
+
+	/* set base, put and end pointer */
+	host1x_ch_writel(ch, cdma->push_buffer.phys, HOST1X_CHANNEL_DMASTART);
+	host1x_ch_writel(ch, cdma->push_buffer.pos, HOST1X_CHANNEL_DMAPUT);
+	host1x_ch_writel(ch, cdma->push_buffer.phys +
+			 cdma->push_buffer.size_bytes + 4,
+			 HOST1X_CHANNEL_DMAEND);
+
+	/* reset GET */
+	host1x_ch_writel(ch, HOST1X_CHANNEL_DMACTRL_DMASTOP |
+			 HOST1X_CHANNEL_DMACTRL_DMAGETRST |
+			 HOST1X_CHANNEL_DMACTRL_DMAINITGET,
+			 HOST1X_CHANNEL_DMACTRL);
+
+	/* start the command DMA */
+	host1x_ch_writel(ch, 0, HOST1X_CHANNEL_DMACTRL);
+
+	cdma->running = true;
+}
+
+/*
+ * Similar to cdma_start(), but rather than starting from an idle
+ * state (where DMA GET is set to DMA PUT), on a timeout we restore
+ * DMA GET from an explicit value (so DMA may again be pending).
+ */
+static void cdma_timeout_restart(struct host1x_cdma *cdma, u32 getptr)
+{
+	struct host1x *host1x = cdma_to_host1x(cdma);
+	struct host1x_channel *ch = cdma_to_channel(cdma);
+
+	if (cdma->running)
+		return;
+
+	cdma->last_pos = cdma->push_buffer.pos;
+
+	host1x_ch_writel(ch, HOST1X_CHANNEL_DMACTRL_DMASTOP,
+			 HOST1X_CHANNEL_DMACTRL);
+
+	/* set base, end pointer (all of memory) */
+	host1x_ch_writel(ch, cdma->push_buffer.phys, HOST1X_CHANNEL_DMASTART);
+	host1x_ch_writel(ch, cdma->push_buffer.phys +
+			 cdma->push_buffer.size_bytes,
+			 HOST1X_CHANNEL_DMAEND);
+
+	/* set GET, by loading the value in PUT (then reset GET) */
+	host1x_ch_writel(ch, getptr, HOST1X_CHANNEL_DMAPUT);
+	host1x_ch_writel(ch, HOST1X_CHANNEL_DMACTRL_DMASTOP |
+			 HOST1X_CHANNEL_DMACTRL_DMAGETRST |
+			 HOST1X_CHANNEL_DMACTRL_DMAINITGET,
+			 HOST1X_CHANNEL_DMACTRL);
+
+	dev_dbg(host1x->dev,
+		"%s: DMA GET 0x%x, PUT HW 0x%x / shadow 0x%x\n", __func__,
+		host1x_ch_readl(ch, HOST1X_CHANNEL_DMAGET),
+		host1x_ch_readl(ch, HOST1X_CHANNEL_DMAPUT),
+		cdma->last_pos);
+
+	/* deassert GET reset and set PUT */
+	host1x_ch_writel(ch, HOST1X_CHANNEL_DMACTRL_DMASTOP,
+			 HOST1X_CHANNEL_DMACTRL);
+	host1x_ch_writel(ch, cdma->push_buffer.pos, HOST1X_CHANNEL_DMAPUT);
+
+	/* start the command DMA */
+	host1x_ch_writel(ch, 0, HOST1X_CHANNEL_DMACTRL);
+
+	cdma->running = true;
+}
+
+/*
+ * Kick channel DMA into action by writing its PUT offset (if it has changed)
+ */
+static void cdma_flush(struct host1x_cdma *cdma)
+{
+	struct host1x_channel *ch = cdma_to_channel(cdma);
+
+	if (cdma->push_buffer.pos != cdma->last_pos) {
+		host1x_ch_writel(ch, cdma->push_buffer.pos,
+				 HOST1X_CHANNEL_DMAPUT);
+		cdma->last_pos = cdma->push_buffer.pos;
+	}
+}
+
+static void cdma_stop(struct host1x_cdma *cdma)
+{
+	struct host1x_channel *ch = cdma_to_channel(cdma);
+
+	mutex_lock(&cdma->lock);
+	if (cdma->running) {
+		host1x_cdma_wait_locked(cdma, CDMA_EVENT_SYNC_QUEUE_EMPTY);
+		host1x_ch_writel(ch, HOST1X_CHANNEL_DMACTRL_DMASTOP,
+				 HOST1X_CHANNEL_DMACTRL);
+		cdma->running = false;
+	}
+	mutex_unlock(&cdma->lock);
+}
+
+/*
+ * Stops both channel's command processor and CDMA immediately.
+ * Also, tears down the channel and resets corresponding module.
+ */
+static void cdma_freeze(struct host1x_cdma *cdma)
+{
+	struct host1x *host = cdma_to_host1x(cdma);
+	struct host1x_channel *ch = cdma_to_channel(cdma);
+	u32 cmdproc_stop;
+
+	if (cdma->torndown && !cdma->running) {
+		dev_warn(host->dev, "Already torn down\n");
+		return;
+	}
+
+	dev_dbg(host->dev, "freezing channel (id %d)\n", ch->id);
+
+	cmdproc_stop = host1x_sync_readl(host, HOST1X_SYNC_CMDPROC_STOP);
+	cmdproc_stop |= BIT(ch->id);
+	host1x_sync_writel(host, cmdproc_stop, HOST1X_SYNC_CMDPROC_STOP);
+
+	dev_dbg(host->dev, "%s: DMA GET 0x%x, PUT HW 0x%x / shadow 0x%x\n",
+		__func__, host1x_ch_readl(ch, HOST1X_CHANNEL_DMAGET),
+		host1x_ch_readl(ch, HOST1X_CHANNEL_DMAPUT),
+		cdma->last_pos);
+
+	host1x_ch_writel(ch, HOST1X_CHANNEL_DMACTRL_DMASTOP,
+			 HOST1X_CHANNEL_DMACTRL);
+
+	host1x_sync_writel(host, BIT(ch->id), HOST1X_SYNC_CH_TEARDOWN);
+
+	cdma->running = false;
+	cdma->torndown = true;
+}
+
+static void cdma_resume(struct host1x_cdma *cdma, u32 getptr)
+{
+	struct host1x *host1x = cdma_to_host1x(cdma);
+	struct host1x_channel *ch = cdma_to_channel(cdma);
+	u32 cmdproc_stop;
+
+	dev_dbg(host1x->dev,
+		"resuming channel (id %d, DMAGET restart = 0x%x)\n",
+		ch->id, getptr);
+
+	cmdproc_stop = host1x_sync_readl(host1x, HOST1X_SYNC_CMDPROC_STOP);
+	cmdproc_stop &= ~(BIT(ch->id));
+	host1x_sync_writel(host1x, cmdproc_stop, HOST1X_SYNC_CMDPROC_STOP);
+
+	cdma->torndown = false;
+	cdma_timeout_restart(cdma, getptr);
+}
+
+/*
+ * If this timeout fires, it indicates the current sync_queue entry has
+ * exceeded its TTL and the userctx should be timed out and remaining
+ * submits already issued cleaned up (future submits return an error).
+ */
+static void cdma_timeout_handler(struct work_struct *work)
+{
+	struct host1x_cdma *cdma;
+	struct host1x *host1x;
+	struct host1x_channel *ch;
+
+	u32 syncpt_val;
+
+	u32 prev_cmdproc, cmdproc_stop;
+
+	cdma = container_of(to_delayed_work(work), struct host1x_cdma,
+			    timeout.wq);
+	host1x = cdma_to_host1x(cdma);
+	ch = cdma_to_channel(cdma);
+
+	host1x_debug_dump(cdma_to_host1x(cdma));
+
+	mutex_lock(&cdma->lock);
+
+	if (!cdma->timeout.client) {
+		dev_dbg(host1x->dev,
+			"cdma_timeout: expired, but has no clientid\n");
+		mutex_unlock(&cdma->lock);
+		return;
+	}
+
+	/* stop processing to get a clean snapshot */
+	prev_cmdproc = host1x_sync_readl(host1x, HOST1X_SYNC_CMDPROC_STOP);
+	cmdproc_stop = prev_cmdproc | BIT(ch->id);
+	host1x_sync_writel(host1x, cmdproc_stop, HOST1X_SYNC_CMDPROC_STOP);
+
+	dev_dbg(host1x->dev, "cdma_timeout: cmdproc was 0x%x is 0x%x\n",
+		prev_cmdproc, cmdproc_stop);
+
+	syncpt_val = host1x_syncpt_load(cdma->timeout.syncpt);
+
+	/* has buffer actually completed? */
+	if ((s32)(syncpt_val - cdma->timeout.syncpt_val) >= 0) {
+		dev_dbg(host1x->dev,
+			"cdma_timeout: expired, but buffer had completed\n");
+		/* restore */
+		cmdproc_stop = prev_cmdproc & ~(BIT(ch->id));
+		host1x_sync_writel(host1x, cmdproc_stop,
+				   HOST1X_SYNC_CMDPROC_STOP);
+		mutex_unlock(&cdma->lock);
+		return;
+	}
+
+	dev_warn(host1x->dev, "%s: timeout: %d (%s), HW thresh %d, done %d\n",
+		__func__, cdma->timeout.syncpt->id, cdma->timeout.syncpt->name,
+		syncpt_val, cdma->timeout.syncpt_val);
+
+	/* stop HW, resetting channel/module */
+	host1x_hw_cdma_freeze(host1x, cdma);
+
+	host1x_cdma_update_sync_queue(cdma, ch->dev);
+	mutex_unlock(&cdma->lock);
+}
+
+/*
+ * Init timeout resources
+ */
+static int cdma_timeout_init(struct host1x_cdma *cdma, u32 syncpt_id)
+{
+	INIT_DELAYED_WORK(&cdma->timeout.wq, cdma_timeout_handler);
+	cdma->timeout.initialized = true;
+
+	return 0;
+}
+
+/*
+ * Clean up timeout resources
+ */
+static void cdma_timeout_destroy(struct host1x_cdma *cdma)
+{
+	if (cdma->timeout.initialized)
+		cancel_delayed_work(&cdma->timeout.wq);
+	cdma->timeout.initialized = false;
+}
+
+static const struct host1x_cdma_ops host1x_cdma_ops = {
+	.start = cdma_start,
+	.stop = cdma_stop,
+	.flush = cdma_flush,
+
+	.timeout_init = cdma_timeout_init,
+	.timeout_destroy = cdma_timeout_destroy,
+	.freeze = cdma_freeze,
+	.resume = cdma_resume,
+	.timeout_cpu_incr = cdma_timeout_cpu_incr,
+};
+
+static const struct host1x_pushbuffer_ops host1x_pushbuffer_ops = {
+	.init = push_buffer_init,
+};
diff --git a/drivers/gpu/host1x/hw/channel_hw.c b/drivers/gpu/host1x/hw/channel_hw.c
new file mode 100644
index 00000000000..ee199623e36
--- /dev/null
+++ b/drivers/gpu/host1x/hw/channel_hw.c
@@ -0,0 +1,168 @@
+/*
+ * Tegra host1x Channel
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/slab.h>
+#include <trace/events/host1x.h>
+
+#include "host1x.h"
+#include "host1x_bo.h"
+#include "channel.h"
+#include "dev.h"
+#include "intr.h"
+#include "job.h"
+
+#define HOST1X_CHANNEL_SIZE 16384
+#define TRACE_MAX_LENGTH 128U
+
+static void trace_write_gather(struct host1x_cdma *cdma, struct host1x_bo *bo,
+			       u32 offset, u32 words)
+{
+	void *mem = NULL;
+
+	if (host1x_debug_trace_cmdbuf)
+		mem = host1x_bo_mmap(bo);
+
+	if (mem) {
+		u32 i;
+		/*
+		 * Write in batches of 128 as there seems to be a limit
+		 * of how much you can output to ftrace at once.
+		 */
+		for (i = 0; i < words; i += TRACE_MAX_LENGTH) {
+			trace_host1x_cdma_push_gather(
+				dev_name(cdma_to_channel(cdma)->dev),
+				(u32)bo, min(words - i, TRACE_MAX_LENGTH),
+				offset + i * sizeof(u32), mem);
+		}
+		host1x_bo_munmap(bo, mem);
+	}
+}
+
+static void submit_gathers(struct host1x_job *job)
+{
+	struct host1x_cdma *cdma = &job->channel->cdma;
+	unsigned int i;
+
+	for (i = 0; i < job->num_gathers; i++) {
+		struct host1x_job_gather *g = &job->gathers[i];
+		u32 op1 = host1x_opcode_gather(g->words);
+		u32 op2 = g->base + g->offset;
+		trace_write_gather(cdma, g->bo, g->offset, op1 & 0xffff);
+		host1x_cdma_push(cdma, op1, op2);
+	}
+}
+
+static int channel_submit(struct host1x_job *job)
+{
+	struct host1x_channel *ch = job->channel;
+	struct host1x_syncpt *sp;
+	u32 user_syncpt_incrs = job->syncpt_incrs;
+	u32 prev_max = 0;
+	u32 syncval;
+	int err;
+	struct host1x_waitlist *completed_waiter = NULL;
+	struct host1x *host = dev_get_drvdata(ch->dev->parent);
+
+	sp = host->syncpt + job->syncpt_id;
+	trace_host1x_channel_submit(dev_name(ch->dev),
+				    job->num_gathers, job->num_relocs,
+				    job->num_waitchk, job->syncpt_id,
+				    job->syncpt_incrs);
+
+	/* before error checks, return current max */
+	prev_max = job->syncpt_end = host1x_syncpt_read_max(sp);
+
+	/* get submit lock */
+	err = mutex_lock_interruptible(&ch->submitlock);
+	if (err)
+		goto error;
+
+	completed_waiter = kzalloc(sizeof(*completed_waiter), GFP_KERNEL);
+	if (!completed_waiter) {
+		mutex_unlock(&ch->submitlock);
+		err = -ENOMEM;
+		goto error;
+	}
+
+	/* begin a CDMA submit */
+	err = host1x_cdma_begin(&ch->cdma, job);
+	if (err) {
+		mutex_unlock(&ch->submitlock);
+		goto error;
+	}
+
+	if (job->serialize) {
+		/*
+		 * Force serialization by inserting a host wait for the
+		 * previous job to finish before this one can commence.
+		 */
+		host1x_cdma_push(&ch->cdma,
+				 host1x_opcode_setclass(HOST1X_CLASS_HOST1X,
+					host1x_uclass_wait_syncpt_r(), 1),
+				 host1x_class_host_wait_syncpt(job->syncpt_id,
+					host1x_syncpt_read_max(sp)));
+	}
+
+	syncval = host1x_syncpt_incr_max(sp, user_syncpt_incrs);
+
+	job->syncpt_end = syncval;
+
+	/* add a setclass for modules that require it */
+	if (job->class)
+		host1x_cdma_push(&ch->cdma,
+				 host1x_opcode_setclass(job->class, 0, 0),
+				 HOST1X_OPCODE_NOP);
+
+	submit_gathers(job);
+
+	/* end CDMA submit & stash pinned hMems into sync queue */
+	host1x_cdma_end(&ch->cdma, job);
+
+	trace_host1x_channel_submitted(dev_name(ch->dev), prev_max, syncval);
+
+	/* schedule a submit complete interrupt */
+	err = host1x_intr_add_action(host, job->syncpt_id, syncval,
+				     HOST1X_INTR_ACTION_SUBMIT_COMPLETE, ch,
+				     completed_waiter, NULL);
+	completed_waiter = NULL;
+	WARN(err, "Failed to set submit complete interrupt");
+
+	mutex_unlock(&ch->submitlock);
+
+	return 0;
+
+error:
+	kfree(completed_waiter);
+	return err;
+}
+
+static int host1x_channel_init(struct host1x_channel *ch, struct host1x *dev,
+			       unsigned int index)
+{
+	ch->id = index;
+	mutex_init(&ch->reflock);
+	mutex_init(&ch->submitlock);
+
+	ch->regs = dev->regs + index * HOST1X_CHANNEL_SIZE;
+	return 0;
+}
+
+static const struct host1x_channel_ops host1x_channel_ops = {
+	.init = host1x_channel_init,
+	.submit = channel_submit,
+};
diff --git a/drivers/gpu/host1x/hw/debug_hw.c b/drivers/gpu/host1x/hw/debug_hw.c
new file mode 100644
index 00000000000..334c038052f
--- /dev/null
+++ b/drivers/gpu/host1x/hw/debug_hw.c
@@ -0,0 +1,322 @@
+/*
+ * Copyright (C) 2010 Google, Inc.
+ * Author: Erik Gilling <konkers@android.com>
+ *
+ * Copyright (C) 2011-2013 NVIDIA Corporation
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ */
+
+#include <linux/debugfs.h>
+#include <linux/seq_file.h>
+#include <linux/mm.h>
+#include <linux/scatterlist.h>
+
+#include <linux/io.h>
+
+#include "dev.h"
+#include "debug.h"
+#include "cdma.h"
+#include "channel.h"
+#include "host1x_bo.h"
+
+#define HOST1X_DEBUG_MAX_PAGE_OFFSET 102400
+
+enum {
+	HOST1X_OPCODE_SETCLASS	= 0x00,
+	HOST1X_OPCODE_INCR	= 0x01,
+	HOST1X_OPCODE_NONINCR	= 0x02,
+	HOST1X_OPCODE_MASK	= 0x03,
+	HOST1X_OPCODE_IMM	= 0x04,
+	HOST1X_OPCODE_RESTART	= 0x05,
+	HOST1X_OPCODE_GATHER	= 0x06,
+	HOST1X_OPCODE_EXTEND	= 0x0e,
+};
+
+enum {
+	HOST1X_OPCODE_EXTEND_ACQUIRE_MLOCK	= 0x00,
+	HOST1X_OPCODE_EXTEND_RELEASE_MLOCK	= 0x01,
+};
+
+static unsigned int show_channel_command(struct output *o, u32 val)
+{
+	unsigned mask;
+	unsigned subop;
+
+	switch (val >> 28) {
+	case HOST1X_OPCODE_SETCLASS:
+		mask = val & 0x3f;
+		if (mask) {
+			host1x_debug_output(o, "SETCL(class=%03x, offset=%03x, mask=%02x, [",
+					    val >> 6 & 0x3ff,
+					    val >> 16 & 0xfff, mask);
+			return hweight8(mask);
+		} else {
+			host1x_debug_output(o, "SETCL(class=%03x)\n",
+					    val >> 6 & 0x3ff);
+			return 0;
+		}
+
+	case HOST1X_OPCODE_INCR:
+		host1x_debug_output(o, "INCR(offset=%03x, [",
+				    val >> 16 & 0xfff);
+		return val & 0xffff;
+
+	case HOST1X_OPCODE_NONINCR:
+		host1x_debug_output(o, "NONINCR(offset=%03x, [",
+				    val >> 16 & 0xfff);
+		return val & 0xffff;
+
+	case HOST1X_OPCODE_MASK:
+		mask = val & 0xffff;
+		host1x_debug_output(o, "MASK(offset=%03x, mask=%03x, [",
+				    val >> 16 & 0xfff, mask);
+		return hweight16(mask);
+
+	case HOST1X_OPCODE_IMM:
+		host1x_debug_output(o, "IMM(offset=%03x, data=%03x)\n",
+				    val >> 16 & 0xfff, val & 0xffff);
+		return 0;
+
+	case HOST1X_OPCODE_RESTART:
+		host1x_debug_output(o, "RESTART(offset=%08x)\n", val << 4);
+		return 0;
+
+	case HOST1X_OPCODE_GATHER:
+		host1x_debug_output(o, "GATHER(offset=%03x, insert=%d, type=%d, count=%04x, addr=[",
+				    val >> 16 & 0xfff, val >> 15 & 0x1,
+				    val >> 14 & 0x1, val & 0x3fff);
+		return 1;
+
+	case HOST1X_OPCODE_EXTEND:
+		subop = val >> 24 & 0xf;
+		if (subop == HOST1X_OPCODE_EXTEND_ACQUIRE_MLOCK)
+			host1x_debug_output(o, "ACQUIRE_MLOCK(index=%d)\n",
+					    val & 0xff);
+		else if (subop == HOST1X_OPCODE_EXTEND_RELEASE_MLOCK)
+			host1x_debug_output(o, "RELEASE_MLOCK(index=%d)\n",
+					    val & 0xff);
+		else
+			host1x_debug_output(o, "EXTEND_UNKNOWN(%08x)\n", val);
+		return 0;
+
+	default:
+		return 0;
+	}
+}
+
+static void show_gather(struct output *o, phys_addr_t phys_addr,
+			unsigned int words, struct host1x_cdma *cdma,
+			phys_addr_t pin_addr, u32 *map_addr)
+{
+	/* Map dmaget cursor to corresponding mem handle */
+	u32 offset = phys_addr - pin_addr;
+	unsigned int data_count = 0, i;
+
+	/*
+	 * Sometimes we're given different hardware address to the same
+	 * page - in these cases the offset will get an invalid number and
+	 * we just have to bail out.
+	 */
+	if (offset > HOST1X_DEBUG_MAX_PAGE_OFFSET) {
+		host1x_debug_output(o, "[address mismatch]\n");
+		return;
+	}
+
+	for (i = 0; i < words; i++) {
+		u32 addr = phys_addr + i * 4;
+		u32 val = *(map_addr + offset / 4 + i);
+
+		if (!data_count) {
+			host1x_debug_output(o, "%08x: %08x:", addr, val);
+			data_count = show_channel_command(o, val);
+		} else {
+			host1x_debug_output(o, "%08x%s", val,
+					    data_count > 0 ? ", " : "])\n");
+			data_count--;
+		}
+	}
+}
+
+static void show_channel_gathers(struct output *o, struct host1x_cdma *cdma)
+{
+	struct host1x_job *job;
+
+	list_for_each_entry(job, &cdma->sync_queue, list) {
+		int i;
+		host1x_debug_output(o, "\n%p: JOB, syncpt_id=%d, syncpt_val=%d, first_get=%08x, timeout=%d num_slots=%d, num_handles=%d\n",
+				    job, job->syncpt_id, job->syncpt_end,
+				    job->first_get, job->timeout,
+				    job->num_slots, job->num_unpins);
+
+		for (i = 0; i < job->num_gathers; i++) {
+			struct host1x_job_gather *g = &job->gathers[i];
+			u32 *mapped;
+
+			if (job->gather_copy_mapped)
+				mapped = (u32 *)job->gather_copy_mapped;
+			else
+				mapped = host1x_bo_mmap(g->bo);
+
+			if (!mapped) {
+				host1x_debug_output(o, "[could not mmap]\n");
+				continue;
+			}
+
+			host1x_debug_output(o, "    GATHER at %08x+%04x, %d words\n",
+					    g->base, g->offset, g->words);
+
+			show_gather(o, g->base + g->offset, g->words, cdma,
+				    g->base, mapped);
+
+			if (!job->gather_copy_mapped)
+				host1x_bo_munmap(g->bo, mapped);
+		}
+	}
+}
+
+static void host1x_debug_show_channel_cdma(struct host1x *host,
+					   struct host1x_channel *ch,
+					   struct output *o)
+{
+	struct host1x_cdma *cdma = &ch->cdma;
+	u32 dmaput, dmaget, dmactrl;
+	u32 cbstat, cbread;
+	u32 val, base, baseval;
+
+	dmaput = host1x_ch_readl(ch, HOST1X_CHANNEL_DMAPUT);
+	dmaget = host1x_ch_readl(ch, HOST1X_CHANNEL_DMAGET);
+	dmactrl = host1x_ch_readl(ch, HOST1X_CHANNEL_DMACTRL);
+	cbread = host1x_sync_readl(host, HOST1X_SYNC_CBREAD(ch->id));
+	cbstat = host1x_sync_readl(host, HOST1X_SYNC_CBSTAT(ch->id));
+
+	host1x_debug_output(o, "%d-%s: ", ch->id, dev_name(ch->dev));
+
+	if (HOST1X_CHANNEL_DMACTRL_DMASTOP_V(dmactrl) ||
+	    !ch->cdma.push_buffer.mapped) {
+		host1x_debug_output(o, "inactive\n\n");
+		return;
+	}
+
+	if (HOST1X_SYNC_CBSTAT_CBCLASS_V(cbstat) == HOST1X_CLASS_HOST1X &&
+	    HOST1X_SYNC_CBSTAT_CBOFFSET_V(cbstat) ==
+	    HOST1X_UCLASS_WAIT_SYNCPT)
+		host1x_debug_output(o, "waiting on syncpt %d val %d\n",
+				    cbread >> 24, cbread & 0xffffff);
+	else if (HOST1X_SYNC_CBSTAT_CBCLASS_V(cbstat) ==
+	   HOST1X_CLASS_HOST1X &&
+	   HOST1X_SYNC_CBSTAT_CBOFFSET_V(cbstat) ==
+	   HOST1X_UCLASS_WAIT_SYNCPT_BASE) {
+
+		base = (cbread >> 16) & 0xff;
+		baseval =
+			host1x_sync_readl(host, HOST1X_SYNC_SYNCPT_BASE(base));
+		val = cbread & 0xffff;
+		host1x_debug_output(o, "waiting on syncpt %d val %d (base %d = %d; offset = %d)\n",
+				    cbread >> 24, baseval + val, base,
+				    baseval, val);
+	} else
+		host1x_debug_output(o, "active class %02x, offset %04x, val %08x\n",
+				    HOST1X_SYNC_CBSTAT_CBCLASS_V(cbstat),
+				    HOST1X_SYNC_CBSTAT_CBOFFSET_V(cbstat),
+				    cbread);
+
+	host1x_debug_output(o, "DMAPUT %08x, DMAGET %08x, DMACTL %08x\n",
+			    dmaput, dmaget, dmactrl);
+	host1x_debug_output(o, "CBREAD %08x, CBSTAT %08x\n", cbread, cbstat);
+
+	show_channel_gathers(o, cdma);
+	host1x_debug_output(o, "\n");
+}
+
+static void host1x_debug_show_channel_fifo(struct host1x *host,
+					   struct host1x_channel *ch,
+					   struct output *o)
+{
+	u32 val, rd_ptr, wr_ptr, start, end;
+	unsigned int data_count = 0;
+
+	host1x_debug_output(o, "%d: fifo:\n", ch->id);
+
+	val = host1x_ch_readl(ch, HOST1X_CHANNEL_FIFOSTAT);
+	host1x_debug_output(o, "FIFOSTAT %08x\n", val);
+	if (HOST1X_CHANNEL_FIFOSTAT_CFEMPTY_V(val)) {
+		host1x_debug_output(o, "[empty]\n");
+		return;
+	}
+
+	host1x_sync_writel(host, 0x0, HOST1X_SYNC_CFPEEK_CTRL);
+	host1x_sync_writel(host, HOST1X_SYNC_CFPEEK_CTRL_ENA_F(1) |
+			   HOST1X_SYNC_CFPEEK_CTRL_CHANNR_F(ch->id),
+			   HOST1X_SYNC_CFPEEK_CTRL);
+
+	val = host1x_sync_readl(host, HOST1X_SYNC_CFPEEK_PTRS);
+	rd_ptr = HOST1X_SYNC_CFPEEK_PTRS_CF_RD_PTR_V(val);
+	wr_ptr = HOST1X_SYNC_CFPEEK_PTRS_CF_WR_PTR_V(val);
+
+	val = host1x_sync_readl(host, HOST1X_SYNC_CF_SETUP(ch->id));
+	start = HOST1X_SYNC_CF_SETUP_BASE_V(val);
+	end = HOST1X_SYNC_CF_SETUP_LIMIT_V(val);
+
+	do {
+		host1x_sync_writel(host, 0x0, HOST1X_SYNC_CFPEEK_CTRL);
+		host1x_sync_writel(host, HOST1X_SYNC_CFPEEK_CTRL_ENA_F(1) |
+				   HOST1X_SYNC_CFPEEK_CTRL_CHANNR_F(ch->id) |
+				   HOST1X_SYNC_CFPEEK_CTRL_ADDR_F(rd_ptr),
+				   HOST1X_SYNC_CFPEEK_CTRL);
+		val = host1x_sync_readl(host, HOST1X_SYNC_CFPEEK_READ);
+
+		if (!data_count) {
+			host1x_debug_output(o, "%08x:", val);
+			data_count = show_channel_command(o, val);
+		} else {
+			host1x_debug_output(o, "%08x%s", val,
+					    data_count > 0 ? ", " : "])\n");
+			data_count--;
+		}
+
+		if (rd_ptr == end)
+			rd_ptr = start;
+		else
+			rd_ptr++;
+	} while (rd_ptr != wr_ptr);
+
+	if (data_count)
+		host1x_debug_output(o, ", ...])\n");
+	host1x_debug_output(o, "\n");
+
+	host1x_sync_writel(host, 0x0, HOST1X_SYNC_CFPEEK_CTRL);
+}
+
+static void host1x_debug_show_mlocks(struct host1x *host, struct output *o)
+{
+	int i;
+
+	host1x_debug_output(o, "---- mlocks ----\n");
+	for (i = 0; i < host1x_syncpt_nb_mlocks(host); i++) {
+		u32 owner =
+			host1x_sync_readl(host, HOST1X_SYNC_MLOCK_OWNER(i));
+		if (HOST1X_SYNC_MLOCK_OWNER_CH_OWNS_V(owner))
+			host1x_debug_output(o, "%d: locked by channel %d\n",
+				i, HOST1X_SYNC_MLOCK_OWNER_CHID_F(owner));
+		else if (HOST1X_SYNC_MLOCK_OWNER_CPU_OWNS_V(owner))
+			host1x_debug_output(o, "%d: locked by cpu\n", i);
+		else
+			host1x_debug_output(o, "%d: unlocked\n", i);
+	}
+	host1x_debug_output(o, "\n");
+}
+
+static const struct host1x_debug_ops host1x_debug_ops = {
+	.show_channel_cdma = host1x_debug_show_channel_cdma,
+	.show_channel_fifo = host1x_debug_show_channel_fifo,
+	.show_mlocks = host1x_debug_show_mlocks,
+};
diff --git a/drivers/gpu/host1x/hw/host1x01.c b/drivers/gpu/host1x/hw/host1x01.c
new file mode 100644
index 00000000000..a14e91cd1e5
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x01.c
@@ -0,0 +1,42 @@
+/*
+ * Host1x init for T20 and T30 Architecture Chips
+ *
+ * Copyright (c) 2011-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+/* include hw specification */
+#include "hw/host1x01.h"
+#include "hw/host1x01_hardware.h"
+
+/* include code */
+#include "hw/cdma_hw.c"
+#include "hw/channel_hw.c"
+#include "hw/debug_hw.c"
+#include "hw/intr_hw.c"
+#include "hw/syncpt_hw.c"
+
+#include "dev.h"
+
+int host1x01_init(struct host1x *host)
+{
+	host->channel_op = &host1x_channel_ops;
+	host->cdma_op = &host1x_cdma_ops;
+	host->cdma_pb_op = &host1x_pushbuffer_ops;
+	host->syncpt_op = &host1x_syncpt_ops;
+	host->intr_op = &host1x_intr_ops;
+	host->debug_op = &host1x_debug_ops;
+
+	return 0;
+}
diff --git a/drivers/gpu/host1x/hw/host1x01.h b/drivers/gpu/host1x/hw/host1x01.h
new file mode 100644
index 00000000000..2706b674325
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x01.h
@@ -0,0 +1,25 @@
+/*
+ * Host1x init for T20 and T30 Architecture Chips
+ *
+ * Copyright (c) 2011-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+#ifndef HOST1X_HOST1X01_H
+#define HOST1X_HOST1X01_H
+
+struct host1x;
+
+int host1x01_init(struct host1x *host);
+
+#endif /* HOST1X_HOST1X01_H_ */
diff --git a/drivers/gpu/host1x/hw/host1x01_hardware.h b/drivers/gpu/host1x/hw/host1x01_hardware.h
new file mode 100644
index 00000000000..5f0fb866efa
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x01_hardware.h
@@ -0,0 +1,143 @@
+/*
+ * Tegra host1x Register Offsets for Tegra20 and Tegra30
+ *
+ * Copyright (c) 2010-2013 NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __HOST1X_HOST1X01_HARDWARE_H
+#define __HOST1X_HOST1X01_HARDWARE_H
+
+#include <linux/types.h>
+#include <linux/bitops.h>
+
+#include "hw_host1x01_channel.h"
+#include "hw_host1x01_sync.h"
+#include "hw_host1x01_uclass.h"
+
+static inline u32 host1x_class_host_wait_syncpt(
+	unsigned indx, unsigned threshold)
+{
+	return host1x_uclass_wait_syncpt_indx_f(indx)
+		| host1x_uclass_wait_syncpt_thresh_f(threshold);
+}
+
+static inline u32 host1x_class_host_load_syncpt_base(
+	unsigned indx, unsigned threshold)
+{
+	return host1x_uclass_load_syncpt_base_base_indx_f(indx)
+		| host1x_uclass_load_syncpt_base_value_f(threshold);
+}
+
+static inline u32 host1x_class_host_wait_syncpt_base(
+	unsigned indx, unsigned base_indx, unsigned offset)
+{
+	return host1x_uclass_wait_syncpt_base_indx_f(indx)
+		| host1x_uclass_wait_syncpt_base_base_indx_f(base_indx)
+		| host1x_uclass_wait_syncpt_base_offset_f(offset);
+}
+
+static inline u32 host1x_class_host_incr_syncpt_base(
+	unsigned base_indx, unsigned offset)
+{
+	return host1x_uclass_incr_syncpt_base_base_indx_f(base_indx)
+		| host1x_uclass_incr_syncpt_base_offset_f(offset);
+}
+
+static inline u32 host1x_class_host_incr_syncpt(
+	unsigned cond, unsigned indx)
+{
+	return host1x_uclass_incr_syncpt_cond_f(cond)
+		| host1x_uclass_incr_syncpt_indx_f(indx);
+}
+
+static inline u32 host1x_class_host_indoff_reg_write(
+	unsigned mod_id, unsigned offset, bool auto_inc)
+{
+	u32 v = host1x_uclass_indoff_indbe_f(0xf)
+		| host1x_uclass_indoff_indmodid_f(mod_id)
+		| host1x_uclass_indoff_indroffset_f(offset);
+	if (auto_inc)
+		v |= host1x_uclass_indoff_autoinc_f(1);
+	return v;
+}
+
+static inline u32 host1x_class_host_indoff_reg_read(
+	unsigned mod_id, unsigned offset, bool auto_inc)
+{
+	u32 v = host1x_uclass_indoff_indmodid_f(mod_id)
+		| host1x_uclass_indoff_indroffset_f(offset)
+		| host1x_uclass_indoff_rwn_read_v();
+	if (auto_inc)
+		v |= host1x_uclass_indoff_autoinc_f(1);
+	return v;
+}
+
+
+/* cdma opcodes */
+static inline u32 host1x_opcode_setclass(
+	unsigned class_id, unsigned offset, unsigned mask)
+{
+	return (0 << 28) | (offset << 16) | (class_id << 6) | mask;
+}
+
+static inline u32 host1x_opcode_incr(unsigned offset, unsigned count)
+{
+	return (1 << 28) | (offset << 16) | count;
+}
+
+static inline u32 host1x_opcode_nonincr(unsigned offset, unsigned count)
+{
+	return (2 << 28) | (offset << 16) | count;
+}
+
+static inline u32 host1x_opcode_mask(unsigned offset, unsigned mask)
+{
+	return (3 << 28) | (offset << 16) | mask;
+}
+
+static inline u32 host1x_opcode_imm(unsigned offset, unsigned value)
+{
+	return (4 << 28) | (offset << 16) | value;
+}
+
+static inline u32 host1x_opcode_imm_incr_syncpt(unsigned cond, unsigned indx)
+{
+	return host1x_opcode_imm(host1x_uclass_incr_syncpt_r(),
+		host1x_class_host_incr_syncpt(cond, indx));
+}
+
+static inline u32 host1x_opcode_restart(unsigned address)
+{
+	return (5 << 28) | (address >> 4);
+}
+
+static inline u32 host1x_opcode_gather(unsigned count)
+{
+	return (6 << 28) | count;
+}
+
+static inline u32 host1x_opcode_gather_nonincr(unsigned offset,	unsigned count)
+{
+	return (6 << 28) | (offset << 16) | BIT(15) | count;
+}
+
+static inline u32 host1x_opcode_gather_incr(unsigned offset, unsigned count)
+{
+	return (6 << 28) | (offset << 16) | BIT(15) | BIT(14) | count;
+}
+
+#define HOST1X_OPCODE_NOP host1x_opcode_nonincr(0, 0)
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x01_channel.h b/drivers/gpu/host1x/hw/hw_host1x01_channel.h
new file mode 100644
index 00000000000..b4bc7ca4e05
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x01_channel.h
@@ -0,0 +1,120 @@
+/*
+ * Copyright (c) 2012-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ */
+
+ /*
+  * Function naming determines intended use:
+  *
+  *     <x>_r(void) : Returns the offset for register <x>.
+  *
+  *     <x>_w(void) : Returns the word offset for word (4 byte) element <x>.
+  *
+  *     <x>_<y>_s(void) : Returns size of field <y> of register <x> in bits.
+  *
+  *     <x>_<y>_f(u32 v) : Returns a value based on 'v' which has been shifted
+  *         and masked to place it at field <y> of register <x>.  This value
+  *         can be |'d with others to produce a full register value for
+  *         register <x>.
+  *
+  *     <x>_<y>_m(void) : Returns a mask for field <y> of register <x>.  This
+  *         value can be ~'d and then &'d to clear the value of field <y> for
+  *         register <x>.
+  *
+  *     <x>_<y>_<z>_f(void) : Returns the constant value <z> after being shifted
+  *         to place it at field <y> of register <x>.  This value can be |'d
+  *         with others to produce a full register value for <x>.
+  *
+  *     <x>_<y>_v(u32 r) : Returns the value of field <y> from a full register
+  *         <x> value 'r' after being shifted to place its LSB at bit 0.
+  *         This value is suitable for direct comparison with other unshifted
+  *         values appropriate for use in field <y> of register <x>.
+  *
+  *     <x>_<y>_<z>_v(void) : Returns the constant value for <z> defined for
+  *         field <y> of register <x>.  This value is suitable for direct
+  *         comparison with unshifted values appropriate for use in field <y>
+  *         of register <x>.
+  */
+
+#ifndef __hw_host1x_channel_host1x_h__
+#define __hw_host1x_channel_host1x_h__
+
+static inline u32 host1x_channel_fifostat_r(void)
+{
+	return 0x0;
+}
+#define HOST1X_CHANNEL_FIFOSTAT \
+	host1x_channel_fifostat_r()
+static inline u32 host1x_channel_fifostat_cfempty_v(u32 r)
+{
+	return (r >> 10) & 0x1;
+}
+#define HOST1X_CHANNEL_FIFOSTAT_CFEMPTY_V(r) \
+	host1x_channel_fifostat_cfempty_v(r)
+static inline u32 host1x_channel_dmastart_r(void)
+{
+	return 0x14;
+}
+#define HOST1X_CHANNEL_DMASTART \
+	host1x_channel_dmastart_r()
+static inline u32 host1x_channel_dmaput_r(void)
+{
+	return 0x18;
+}
+#define HOST1X_CHANNEL_DMAPUT \
+	host1x_channel_dmaput_r()
+static inline u32 host1x_channel_dmaget_r(void)
+{
+	return 0x1c;
+}
+#define HOST1X_CHANNEL_DMAGET \
+	host1x_channel_dmaget_r()
+static inline u32 host1x_channel_dmaend_r(void)
+{
+	return 0x20;
+}
+#define HOST1X_CHANNEL_DMAEND \
+	host1x_channel_dmaend_r()
+static inline u32 host1x_channel_dmactrl_r(void)
+{
+	return 0x24;
+}
+#define HOST1X_CHANNEL_DMACTRL \
+	host1x_channel_dmactrl_r()
+static inline u32 host1x_channel_dmactrl_dmastop(void)
+{
+	return 1 << 0;
+}
+#define HOST1X_CHANNEL_DMACTRL_DMASTOP \
+	host1x_channel_dmactrl_dmastop()
+static inline u32 host1x_channel_dmactrl_dmastop_v(u32 r)
+{
+	return (r >> 0) & 0x1;
+}
+#define HOST1X_CHANNEL_DMACTRL_DMASTOP_V(r) \
+	host1x_channel_dmactrl_dmastop_v(r)
+static inline u32 host1x_channel_dmactrl_dmagetrst(void)
+{
+	return 1 << 1;
+}
+#define HOST1X_CHANNEL_DMACTRL_DMAGETRST \
+	host1x_channel_dmactrl_dmagetrst()
+static inline u32 host1x_channel_dmactrl_dmainitget(void)
+{
+	return 1 << 2;
+}
+#define HOST1X_CHANNEL_DMACTRL_DMAINITGET \
+	host1x_channel_dmactrl_dmainitget()
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x01_sync.h b/drivers/gpu/host1x/hw/hw_host1x01_sync.h
new file mode 100644
index 00000000000..ac704e57997
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x01_sync.h
@@ -0,0 +1,243 @@
+/*
+ * Copyright (c) 2012-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ */
+
+ /*
+  * Function naming determines intended use:
+  *
+  *     <x>_r(void) : Returns the offset for register <x>.
+  *
+  *     <x>_w(void) : Returns the word offset for word (4 byte) element <x>.
+  *
+  *     <x>_<y>_s(void) : Returns size of field <y> of register <x> in bits.
+  *
+  *     <x>_<y>_f(u32 v) : Returns a value based on 'v' which has been shifted
+  *         and masked to place it at field <y> of register <x>.  This value
+  *         can be |'d with others to produce a full register value for
+  *         register <x>.
+  *
+  *     <x>_<y>_m(void) : Returns a mask for field <y> of register <x>.  This
+  *         value can be ~'d and then &'d to clear the value of field <y> for
+  *         register <x>.
+  *
+  *     <x>_<y>_<z>_f(void) : Returns the constant value <z> after being shifted
+  *         to place it at field <y> of register <x>.  This value can be |'d
+  *         with others to produce a full register value for <x>.
+  *
+  *     <x>_<y>_v(u32 r) : Returns the value of field <y> from a full register
+  *         <x> value 'r' after being shifted to place its LSB at bit 0.
+  *         This value is suitable for direct comparison with other unshifted
+  *         values appropriate for use in field <y> of register <x>.
+  *
+  *     <x>_<y>_<z>_v(void) : Returns the constant value for <z> defined for
+  *         field <y> of register <x>.  This value is suitable for direct
+  *         comparison with unshifted values appropriate for use in field <y>
+  *         of register <x>.
+  */
+
+#ifndef __hw_host1x01_sync_h__
+#define __hw_host1x01_sync_h__
+
+#define REGISTER_STRIDE	4
+
+static inline u32 host1x_sync_syncpt_r(unsigned int id)
+{
+	return 0x400 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT(id) \
+	host1x_sync_syncpt_r(id)
+static inline u32 host1x_sync_syncpt_thresh_cpu0_int_status_r(unsigned int id)
+{
+	return 0x40 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(id) \
+	host1x_sync_syncpt_thresh_cpu0_int_status_r(id)
+static inline u32 host1x_sync_syncpt_thresh_int_disable_r(unsigned int id)
+{
+	return 0x60 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT_THRESH_INT_DISABLE(id) \
+	host1x_sync_syncpt_thresh_int_disable_r(id)
+static inline u32 host1x_sync_syncpt_thresh_int_enable_cpu0_r(unsigned int id)
+{
+	return 0x68 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT_THRESH_INT_ENABLE_CPU0(id) \
+	host1x_sync_syncpt_thresh_int_enable_cpu0_r(id)
+static inline u32 host1x_sync_cf_setup_r(unsigned int channel)
+{
+	return 0x80 + channel * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_CF_SETUP(channel) \
+	host1x_sync_cf_setup_r(channel)
+static inline u32 host1x_sync_cf_setup_base_v(u32 r)
+{
+	return (r >> 0) & 0x1ff;
+}
+#define HOST1X_SYNC_CF_SETUP_BASE_V(r) \
+	host1x_sync_cf_setup_base_v(r)
+static inline u32 host1x_sync_cf_setup_limit_v(u32 r)
+{
+	return (r >> 16) & 0x1ff;
+}
+#define HOST1X_SYNC_CF_SETUP_LIMIT_V(r) \
+	host1x_sync_cf_setup_limit_v(r)
+static inline u32 host1x_sync_cmdproc_stop_r(void)
+{
+	return 0xac;
+}
+#define HOST1X_SYNC_CMDPROC_STOP \
+	host1x_sync_cmdproc_stop_r()
+static inline u32 host1x_sync_ch_teardown_r(void)
+{
+	return 0xb0;
+}
+#define HOST1X_SYNC_CH_TEARDOWN \
+	host1x_sync_ch_teardown_r()
+static inline u32 host1x_sync_usec_clk_r(void)
+{
+	return 0x1a4;
+}
+#define HOST1X_SYNC_USEC_CLK \
+	host1x_sync_usec_clk_r()
+static inline u32 host1x_sync_ctxsw_timeout_cfg_r(void)
+{
+	return 0x1a8;
+}
+#define HOST1X_SYNC_CTXSW_TIMEOUT_CFG \
+	host1x_sync_ctxsw_timeout_cfg_r()
+static inline u32 host1x_sync_ip_busy_timeout_r(void)
+{
+	return 0x1bc;
+}
+#define HOST1X_SYNC_IP_BUSY_TIMEOUT \
+	host1x_sync_ip_busy_timeout_r()
+static inline u32 host1x_sync_mlock_owner_r(unsigned int id)
+{
+	return 0x340 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_MLOCK_OWNER(id) \
+	host1x_sync_mlock_owner_r(id)
+static inline u32 host1x_sync_mlock_owner_chid_f(u32 v)
+{
+	return (v & 0xf) << 8;
+}
+#define HOST1X_SYNC_MLOCK_OWNER_CHID_F(v) \
+	host1x_sync_mlock_owner_chid_f(v)
+static inline u32 host1x_sync_mlock_owner_cpu_owns_v(u32 r)
+{
+	return (r >> 1) & 0x1;
+}
+#define HOST1X_SYNC_MLOCK_OWNER_CPU_OWNS_V(r) \
+	host1x_sync_mlock_owner_cpu_owns_v(r)
+static inline u32 host1x_sync_mlock_owner_ch_owns_v(u32 r)
+{
+	return (r >> 0) & 0x1;
+}
+#define HOST1X_SYNC_MLOCK_OWNER_CH_OWNS_V(r) \
+	host1x_sync_mlock_owner_ch_owns_v(r)
+static inline u32 host1x_sync_syncpt_int_thresh_r(unsigned int id)
+{
+	return 0x500 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT_INT_THRESH(id) \
+	host1x_sync_syncpt_int_thresh_r(id)
+static inline u32 host1x_sync_syncpt_base_r(unsigned int id)
+{
+	return 0x600 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT_BASE(id) \
+	host1x_sync_syncpt_base_r(id)
+static inline u32 host1x_sync_syncpt_cpu_incr_r(unsigned int id)
+{
+	return 0x700 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT_CPU_INCR(id) \
+	host1x_sync_syncpt_cpu_incr_r(id)
+static inline u32 host1x_sync_cbread_r(unsigned int channel)
+{
+	return 0x720 + channel * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_CBREAD(channel) \
+	host1x_sync_cbread_r(channel)
+static inline u32 host1x_sync_cfpeek_ctrl_r(void)
+{
+	return 0x74c;
+}
+#define HOST1X_SYNC_CFPEEK_CTRL \
+	host1x_sync_cfpeek_ctrl_r()
+static inline u32 host1x_sync_cfpeek_ctrl_addr_f(u32 v)
+{
+	return (v & 0x1ff) << 0;
+}
+#define HOST1X_SYNC_CFPEEK_CTRL_ADDR_F(v) \
+	host1x_sync_cfpeek_ctrl_addr_f(v)
+static inline u32 host1x_sync_cfpeek_ctrl_channr_f(u32 v)
+{
+	return (v & 0x7) << 16;
+}
+#define HOST1X_SYNC_CFPEEK_CTRL_CHANNR_F(v) \
+	host1x_sync_cfpeek_ctrl_channr_f(v)
+static inline u32 host1x_sync_cfpeek_ctrl_ena_f(u32 v)
+{
+	return (v & 0x1) << 31;
+}
+#define HOST1X_SYNC_CFPEEK_CTRL_ENA_F(v) \
+	host1x_sync_cfpeek_ctrl_ena_f(v)
+static inline u32 host1x_sync_cfpeek_read_r(void)
+{
+	return 0x750;
+}
+#define HOST1X_SYNC_CFPEEK_READ \
+	host1x_sync_cfpeek_read_r()
+static inline u32 host1x_sync_cfpeek_ptrs_r(void)
+{
+	return 0x754;
+}
+#define HOST1X_SYNC_CFPEEK_PTRS \
+	host1x_sync_cfpeek_ptrs_r()
+static inline u32 host1x_sync_cfpeek_ptrs_cf_rd_ptr_v(u32 r)
+{
+	return (r >> 0) & 0x1ff;
+}
+#define HOST1X_SYNC_CFPEEK_PTRS_CF_RD_PTR_V(r) \
+	host1x_sync_cfpeek_ptrs_cf_rd_ptr_v(r)
+static inline u32 host1x_sync_cfpeek_ptrs_cf_wr_ptr_v(u32 r)
+{
+	return (r >> 16) & 0x1ff;
+}
+#define HOST1X_SYNC_CFPEEK_PTRS_CF_WR_PTR_V(r) \
+	host1x_sync_cfpeek_ptrs_cf_wr_ptr_v(r)
+static inline u32 host1x_sync_cbstat_r(unsigned int channel)
+{
+	return 0x758 + channel * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_CBSTAT(channel) \
+	host1x_sync_cbstat_r(channel)
+static inline u32 host1x_sync_cbstat_cboffset_v(u32 r)
+{
+	return (r >> 0) & 0xffff;
+}
+#define HOST1X_SYNC_CBSTAT_CBOFFSET_V(r) \
+	host1x_sync_cbstat_cboffset_v(r)
+static inline u32 host1x_sync_cbstat_cbclass_v(u32 r)
+{
+	return (r >> 16) & 0x3ff;
+}
+#define HOST1X_SYNC_CBSTAT_CBCLASS_V(r) \
+	host1x_sync_cbstat_cbclass_v(r)
+
+#endif /* __hw_host1x01_sync_h__ */
diff --git a/drivers/gpu/host1x/hw/hw_host1x01_uclass.h b/drivers/gpu/host1x/hw/hw_host1x01_uclass.h
new file mode 100644
index 00000000000..42f3ce19ca3
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x01_uclass.h
@@ -0,0 +1,174 @@
+/*
+ * Copyright (c) 2012-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ */
+
+ /*
+  * Function naming determines intended use:
+  *
+  *     <x>_r(void) : Returns the offset for register <x>.
+  *
+  *     <x>_w(void) : Returns the word offset for word (4 byte) element <x>.
+  *
+  *     <x>_<y>_s(void) : Returns size of field <y> of register <x> in bits.
+  *
+  *     <x>_<y>_f(u32 v) : Returns a value based on 'v' which has been shifted
+  *         and masked to place it at field <y> of register <x>.  This value
+  *         can be |'d with others to produce a full register value for
+  *         register <x>.
+  *
+  *     <x>_<y>_m(void) : Returns a mask for field <y> of register <x>.  This
+  *         value can be ~'d and then &'d to clear the value of field <y> for
+  *         register <x>.
+  *
+  *     <x>_<y>_<z>_f(void) : Returns the constant value <z> after being shifted
+  *         to place it at field <y> of register <x>.  This value can be |'d
+  *         with others to produce a full register value for <x>.
+  *
+  *     <x>_<y>_v(u32 r) : Returns the value of field <y> from a full register
+  *         <x> value 'r' after being shifted to place its LSB at bit 0.
+  *         This value is suitable for direct comparison with other unshifted
+  *         values appropriate for use in field <y> of register <x>.
+  *
+  *     <x>_<y>_<z>_v(void) : Returns the constant value for <z> defined for
+  *         field <y> of register <x>.  This value is suitable for direct
+  *         comparison with unshifted values appropriate for use in field <y>
+  *         of register <x>.
+  */
+
+#ifndef __hw_host1x_uclass_host1x_h__
+#define __hw_host1x_uclass_host1x_h__
+
+static inline u32 host1x_uclass_incr_syncpt_r(void)
+{
+	return 0x0;
+}
+#define HOST1X_UCLASS_INCR_SYNCPT \
+	host1x_uclass_incr_syncpt_r()
+static inline u32 host1x_uclass_incr_syncpt_cond_f(u32 v)
+{
+	return (v & 0xff) << 8;
+}
+#define HOST1X_UCLASS_INCR_SYNCPT_COND_F(v) \
+	host1x_uclass_incr_syncpt_cond_f(v)
+static inline u32 host1x_uclass_incr_syncpt_indx_f(u32 v)
+{
+	return (v & 0xff) << 0;
+}
+#define HOST1X_UCLASS_INCR_SYNCPT_INDX_F(v) \
+	host1x_uclass_incr_syncpt_indx_f(v)
+static inline u32 host1x_uclass_wait_syncpt_r(void)
+{
+	return 0x8;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT \
+	host1x_uclass_wait_syncpt_r()
+static inline u32 host1x_uclass_wait_syncpt_indx_f(u32 v)
+{
+	return (v & 0xff) << 24;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_INDX_F(v) \
+	host1x_uclass_wait_syncpt_indx_f(v)
+static inline u32 host1x_uclass_wait_syncpt_thresh_f(u32 v)
+{
+	return (v & 0xffffff) << 0;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_THRESH_F(v) \
+	host1x_uclass_wait_syncpt_thresh_f(v)
+static inline u32 host1x_uclass_wait_syncpt_base_r(void)
+{
+	return 0x9;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_BASE \
+	host1x_uclass_wait_syncpt_base_r()
+static inline u32 host1x_uclass_wait_syncpt_base_indx_f(u32 v)
+{
+	return (v & 0xff) << 24;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_BASE_INDX_F(v) \
+	host1x_uclass_wait_syncpt_base_indx_f(v)
+static inline u32 host1x_uclass_wait_syncpt_base_base_indx_f(u32 v)
+{
+	return (v & 0xff) << 16;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_BASE_BASE_INDX_F(v) \
+	host1x_uclass_wait_syncpt_base_base_indx_f(v)
+static inline u32 host1x_uclass_wait_syncpt_base_offset_f(u32 v)
+{
+	return (v & 0xffff) << 0;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_BASE_OFFSET_F(v) \
+	host1x_uclass_wait_syncpt_base_offset_f(v)
+static inline u32 host1x_uclass_load_syncpt_base_base_indx_f(u32 v)
+{
+	return (v & 0xff) << 24;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE_BASE_INDX_F(v) \
+	host1x_uclass_load_syncpt_base_base_indx_f(v)
+static inline u32 host1x_uclass_load_syncpt_base_value_f(u32 v)
+{
+	return (v & 0xffffff) << 0;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE_VALUE_F(v) \
+	host1x_uclass_load_syncpt_base_value_f(v)
+static inline u32 host1x_uclass_incr_syncpt_base_base_indx_f(u32 v)
+{
+	return (v & 0xff) << 24;
+}
+#define HOST1X_UCLASS_INCR_SYNCPT_BASE_BASE_INDX_F(v) \
+	host1x_uclass_incr_syncpt_base_base_indx_f(v)
+static inline u32 host1x_uclass_incr_syncpt_base_offset_f(u32 v)
+{
+	return (v & 0xffffff) << 0;
+}
+#define HOST1X_UCLASS_INCR_SYNCPT_BASE_OFFSET_F(v) \
+	host1x_uclass_incr_syncpt_base_offset_f(v)
+static inline u32 host1x_uclass_indoff_r(void)
+{
+	return 0x2d;
+}
+#define HOST1X_UCLASS_INDOFF \
+	host1x_uclass_indoff_r()
+static inline u32 host1x_uclass_indoff_indbe_f(u32 v)
+{
+	return (v & 0xf) << 28;
+}
+#define HOST1X_UCLASS_INDOFF_INDBE_F(v) \
+	host1x_uclass_indoff_indbe_f(v)
+static inline u32 host1x_uclass_indoff_autoinc_f(u32 v)
+{
+	return (v & 0x1) << 27;
+}
+#define HOST1X_UCLASS_INDOFF_AUTOINC_F(v) \
+	host1x_uclass_indoff_autoinc_f(v)
+static inline u32 host1x_uclass_indoff_indmodid_f(u32 v)
+{
+	return (v & 0xff) << 18;
+}
+#define HOST1X_UCLASS_INDOFF_INDMODID_F(v) \
+	host1x_uclass_indoff_indmodid_f(v)
+static inline u32 host1x_uclass_indoff_indroffset_f(u32 v)
+{
+	return (v & 0xffff) << 2;
+}
+#define HOST1X_UCLASS_INDOFF_INDROFFSET_F(v) \
+	host1x_uclass_indoff_indroffset_f(v)
+static inline u32 host1x_uclass_indoff_rwn_read_v(void)
+{
+	return 1;
+}
+#define HOST1X_UCLASS_INDOFF_INDROFFSET_F(v) \
+	host1x_uclass_indoff_indroffset_f(v)
+#endif
diff --git a/drivers/gpu/host1x/hw/intr_hw.c b/drivers/gpu/host1x/hw/intr_hw.c
new file mode 100644
index 00000000000..b592eef1efc
--- /dev/null
+++ b/drivers/gpu/host1x/hw/intr_hw.c
@@ -0,0 +1,143 @@
+/*
+ * Tegra host1x Interrupt Management
+ *
+ * Copyright (C) 2010 Google, Inc.
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/io.h>
+#include <asm/mach/irq.h>
+
+#include "intr.h"
+#include "dev.h"
+
+/*
+ * Sync point threshold interrupt service function
+ * Handles sync point threshold triggers, in interrupt context
+ */
+static void host1x_intr_syncpt_handle(struct host1x_syncpt *syncpt)
+{
+	unsigned int id = syncpt->id;
+	struct host1x *host = syncpt->host;
+
+	host1x_sync_writel(host, BIT_MASK(id),
+		HOST1X_SYNC_SYNCPT_THRESH_INT_DISABLE(BIT_WORD(id)));
+	host1x_sync_writel(host, BIT_MASK(id),
+		HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(BIT_WORD(id)));
+
+	queue_work(host->intr_wq, &syncpt->intr.work);
+}
+
+static irqreturn_t syncpt_thresh_isr(int irq, void *dev_id)
+{
+	struct host1x *host = dev_id;
+	unsigned long reg;
+	int i, id;
+
+	for (i = 0; i <= BIT_WORD(host->info->nb_pts); i++) {
+		reg = host1x_sync_readl(host,
+			HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(i));
+		for_each_set_bit(id, &reg, BITS_PER_LONG) {
+			struct host1x_syncpt *syncpt =
+				host->syncpt + (i * BITS_PER_LONG + id);
+			host1x_intr_syncpt_handle(syncpt);
+		}
+	}
+
+	return IRQ_HANDLED;
+}
+
+static void _host1x_intr_disable_all_syncpt_intrs(struct host1x *host)
+{
+	u32 i;
+
+	for (i = 0; i <= BIT_WORD(host->info->nb_pts); ++i) {
+		host1x_sync_writel(host, 0xffffffffu,
+			HOST1X_SYNC_SYNCPT_THRESH_INT_DISABLE(i));
+		host1x_sync_writel(host, 0xffffffffu,
+			HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(i));
+	}
+}
+
+static int _host1x_intr_init_host_sync(struct host1x *host, u32 cpm,
+	void (*syncpt_thresh_work)(struct work_struct *))
+{
+	int i, err;
+
+	host1x_hw_intr_disable_all_syncpt_intrs(host);
+
+	for (i = 0; i < host->info->nb_pts; i++)
+		INIT_WORK(&host->syncpt[i].intr.work, syncpt_thresh_work);
+
+	err = devm_request_irq(host->dev, host->intr_syncpt_irq,
+			       syncpt_thresh_isr, IRQF_SHARED,
+			       "host1x_syncpt", host);
+	if (IS_ERR_VALUE(err)) {
+		WARN_ON(1);
+		return err;
+	}
+
+	/* disable the ip_busy_timeout. this prevents write drops */
+	host1x_sync_writel(host, 0, HOST1X_SYNC_IP_BUSY_TIMEOUT);
+
+	/*
+	 * increase the auto-ack timout to the maximum value. 2d will hang
+	 * otherwise on Tegra2.
+	 */
+	host1x_sync_writel(host, 0xff, HOST1X_SYNC_CTXSW_TIMEOUT_CFG);
+
+	/* update host clocks per usec */
+	host1x_sync_writel(host, cpm, HOST1X_SYNC_USEC_CLK);
+
+	return 0;
+}
+
+static void _host1x_intr_set_syncpt_threshold(struct host1x *host,
+	u32 id, u32 thresh)
+{
+	host1x_sync_writel(host, thresh, HOST1X_SYNC_SYNCPT_INT_THRESH(id));
+}
+
+static void _host1x_intr_enable_syncpt_intr(struct host1x *host, u32 id)
+{
+	host1x_sync_writel(host, BIT_MASK(id),
+		HOST1X_SYNC_SYNCPT_THRESH_INT_ENABLE_CPU0(BIT_WORD(id)));
+}
+
+static void _host1x_intr_disable_syncpt_intr(struct host1x *host, u32 id)
+{
+	host1x_sync_writel(host, BIT_MASK(id),
+		HOST1X_SYNC_SYNCPT_THRESH_INT_DISABLE(BIT_WORD(id)));
+	host1x_sync_writel(host, BIT_MASK(id),
+		HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(BIT_WORD(id)));
+}
+
+static int _host1x_free_syncpt_irq(struct host1x *host)
+{
+	devm_free_irq(host->dev, host->intr_syncpt_irq, host);
+	flush_workqueue(host->intr_wq);
+	return 0;
+}
+
+static const struct host1x_intr_ops host1x_intr_ops = {
+	.init_host_sync = _host1x_intr_init_host_sync,
+	.set_syncpt_threshold = _host1x_intr_set_syncpt_threshold,
+	.enable_syncpt_intr = _host1x_intr_enable_syncpt_intr,
+	.disable_syncpt_intr = _host1x_intr_disable_syncpt_intr,
+	.disable_all_syncpt_intrs = _host1x_intr_disable_all_syncpt_intrs,
+	.free_syncpt_irq = _host1x_free_syncpt_irq,
+};
diff --git a/drivers/gpu/host1x/hw/syncpt_hw.c b/drivers/gpu/host1x/hw/syncpt_hw.c
new file mode 100644
index 00000000000..61174990102
--- /dev/null
+++ b/drivers/gpu/host1x/hw/syncpt_hw.c
@@ -0,0 +1,114 @@
+/*
+ * Tegra host1x Syncpoints
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/io.h>
+
+#include "dev.h"
+#include "syncpt.h"
+
+/*
+ * Write the current syncpoint value back to hw.
+ */
+static void syncpt_restore(struct host1x_syncpt *sp)
+{
+	struct host1x *host = sp->host;
+	int min = host1x_syncpt_read_min(sp);
+	host1x_sync_writel(host, min, HOST1X_SYNC_SYNCPT(sp->id));
+}
+
+/*
+ * Write the current waitbase value back to hw.
+ */
+static void syncpt_restore_wait_base(struct host1x_syncpt *sp)
+{
+	struct host1x *host = sp->host;
+	host1x_sync_writel(host, sp->base_val,
+			   HOST1X_SYNC_SYNCPT_BASE(sp->id));
+}
+
+/*
+ * Read waitbase value from hw.
+ */
+static void syncpt_read_wait_base(struct host1x_syncpt *sp)
+{
+	struct host1x *host = sp->host;
+	sp->base_val =
+		host1x_sync_readl(host, HOST1X_SYNC_SYNCPT_BASE(sp->id));
+}
+
+/*
+ * Updates the last value read from hardware.
+ */
+static u32 syncpt_load(struct host1x_syncpt *sp)
+{
+	struct host1x *host = sp->host;
+	u32 old, live;
+
+	/* Loop in case there's a race writing to min_val */
+	do {
+		old = host1x_syncpt_read_min(sp);
+		live = host1x_sync_readl(host, HOST1X_SYNC_SYNCPT(sp->id));
+	} while ((u32)atomic_cmpxchg(&sp->min_val, old, live) != old);
+
+	if (!host1x_syncpt_check_max(sp, live))
+		dev_err(host->dev, "%s failed: id=%u, min=%d, max=%d\n",
+			__func__, sp->id, host1x_syncpt_read_min(sp),
+			host1x_syncpt_read_max(sp));
+
+	return live;
+}
+
+/*
+ * Write a cpu syncpoint increment to the hardware, without touching
+ * the cache.
+ */
+static void syncpt_cpu_incr(struct host1x_syncpt *sp)
+{
+	struct host1x *host = sp->host;
+	u32 reg_offset = sp->id / 32;
+
+	if (!host1x_syncpt_client_managed(sp) &&
+	    host1x_syncpt_idle(sp)) {
+		dev_err(host->dev, "Trying to increment syncpoint id %d beyond max\n",
+			sp->id);
+		host1x_debug_dump(sp->host);
+		return;
+	}
+	host1x_sync_writel(host, BIT_MASK(sp->id),
+			   HOST1X_SYNC_SYNCPT_CPU_INCR(reg_offset));
+	wmb();
+}
+
+/* remove a wait pointed to by patch_addr */
+static int syncpt_patch_wait(struct host1x_syncpt *sp, void *patch_addr)
+{
+	u32 override = host1x_class_host_wait_syncpt(
+		HOST1X_SYNCPT_RESERVED, 0);
+
+	*((u32 *)patch_addr) = override;
+	return 0;
+}
+
+static const struct host1x_syncpt_ops host1x_syncpt_ops = {
+	.restore = syncpt_restore,
+	.restore_wait_base = syncpt_restore_wait_base,
+	.load_wait_base = syncpt_read_wait_base,
+	.load = syncpt_load,
+	.cpu_incr = syncpt_cpu_incr,
+	.patch_wait = syncpt_patch_wait,
+};
diff --git a/drivers/gpu/host1x/intr.c b/drivers/gpu/host1x/intr.c
new file mode 100644
index 00000000000..2491bf82e30
--- /dev/null
+++ b/drivers/gpu/host1x/intr.c
@@ -0,0 +1,354 @@
+/*
+ * Tegra host1x Interrupt Management
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/clk.h>
+#include <linux/interrupt.h>
+#include <linux/slab.h>
+#include <linux/irq.h>
+
+#include <trace/events/host1x.h>
+#include "channel.h"
+#include "dev.h"
+#include "intr.h"
+
+/* Wait list management */
+
+enum waitlist_state {
+	WLS_PENDING,
+	WLS_REMOVED,
+	WLS_CANCELLED,
+	WLS_HANDLED
+};
+
+static void waiter_release(struct kref *kref)
+{
+	kfree(container_of(kref, struct host1x_waitlist, refcount));
+}
+
+/*
+ * add a waiter to a waiter queue, sorted by threshold
+ * returns true if it was added at the head of the queue
+ */
+static bool add_waiter_to_queue(struct host1x_waitlist *waiter,
+				struct list_head *queue)
+{
+	struct host1x_waitlist *pos;
+	u32 thresh = waiter->thresh;
+
+	list_for_each_entry_reverse(pos, queue, list)
+		if ((s32)(pos->thresh - thresh) <= 0) {
+			list_add(&waiter->list, &pos->list);
+			return false;
+		}
+
+	list_add(&waiter->list, queue);
+	return true;
+}
+
+/*
+ * run through a waiter queue for a single sync point ID
+ * and gather all completed waiters into lists by actions
+ */
+static void remove_completed_waiters(struct list_head *head, u32 sync,
+			struct list_head completed[HOST1X_INTR_ACTION_COUNT])
+{
+	struct list_head *dest;
+	struct host1x_waitlist *waiter, *next, *prev;
+
+	list_for_each_entry_safe(waiter, next, head, list) {
+		if ((s32)(waiter->thresh - sync) > 0)
+			break;
+
+		dest = completed + waiter->action;
+
+		/* consolidate submit cleanups */
+		if (waiter->action == HOST1X_INTR_ACTION_SUBMIT_COMPLETE &&
+		    !list_empty(dest)) {
+			prev = list_entry(dest->prev,
+					  struct host1x_waitlist, list);
+			if (prev->data == waiter->data) {
+				prev->count++;
+				dest = NULL;
+			}
+		}
+
+		/* PENDING->REMOVED or CANCELLED->HANDLED */
+		if (atomic_inc_return(&waiter->state) == WLS_HANDLED || !dest) {
+			list_del(&waiter->list);
+			kref_put(&waiter->refcount, waiter_release);
+		} else
+			list_move_tail(&waiter->list, dest);
+	}
+}
+
+static void reset_threshold_interrupt(struct host1x *host,
+				      struct list_head *head,
+				      unsigned int id)
+{
+	u32 thresh =
+		list_first_entry(head, struct host1x_waitlist, list)->thresh;
+
+	host1x_hw_intr_set_syncpt_threshold(host, id, thresh);
+	host1x_hw_intr_enable_syncpt_intr(host, id);
+}
+
+static void action_submit_complete(struct host1x_waitlist *waiter)
+{
+	struct host1x_channel *channel = waiter->data;
+
+	host1x_cdma_update(&channel->cdma);
+
+	/*  Add nr_completed to trace */
+	trace_host1x_channel_submit_complete(dev_name(channel->dev),
+					     waiter->count, waiter->thresh);
+
+}
+
+static void action_wakeup(struct host1x_waitlist *waiter)
+{
+	wait_queue_head_t *wq = waiter->data;
+	wake_up(wq);
+}
+
+static void action_wakeup_interruptible(struct host1x_waitlist *waiter)
+{
+	wait_queue_head_t *wq = waiter->data;
+	wake_up_interruptible(wq);
+}
+
+typedef void (*action_handler)(struct host1x_waitlist *waiter);
+
+static action_handler action_handlers[HOST1X_INTR_ACTION_COUNT] = {
+	action_submit_complete,
+	action_wakeup,
+	action_wakeup_interruptible,
+};
+
+static void run_handlers(struct list_head completed[HOST1X_INTR_ACTION_COUNT])
+{
+	struct list_head *head = completed;
+	int i;
+
+	for (i = 0; i < HOST1X_INTR_ACTION_COUNT; ++i, ++head) {
+		action_handler handler = action_handlers[i];
+		struct host1x_waitlist *waiter, *next;
+
+		list_for_each_entry_safe(waiter, next, head, list) {
+			list_del(&waiter->list);
+			handler(waiter);
+			WARN_ON(atomic_xchg(&waiter->state, WLS_HANDLED) !=
+				WLS_REMOVED);
+			kref_put(&waiter->refcount, waiter_release);
+		}
+	}
+}
+
+/*
+ * Remove & handle all waiters that have completed for the given syncpt
+ */
+static int process_wait_list(struct host1x *host,
+			     struct host1x_syncpt *syncpt,
+			     u32 threshold)
+{
+	struct list_head completed[HOST1X_INTR_ACTION_COUNT];
+	unsigned int i;
+	int empty;
+
+	for (i = 0; i < HOST1X_INTR_ACTION_COUNT; ++i)
+		INIT_LIST_HEAD(completed + i);
+
+	spin_lock(&syncpt->intr.lock);
+
+	remove_completed_waiters(&syncpt->intr.wait_head, threshold,
+				 completed);
+
+	empty = list_empty(&syncpt->intr.wait_head);
+	if (empty)
+		host1x_hw_intr_disable_syncpt_intr(host, syncpt->id);
+	else
+		reset_threshold_interrupt(host, &syncpt->intr.wait_head,
+					  syncpt->id);
+
+	spin_unlock(&syncpt->intr.lock);
+
+	run_handlers(completed);
+
+	return empty;
+}
+
+/*
+ * Sync point threshold interrupt service thread function
+ * Handles sync point threshold triggers, in thread context
+ */
+
+static void syncpt_thresh_work(struct work_struct *work)
+{
+	struct host1x_syncpt_intr *syncpt_intr =
+		container_of(work, struct host1x_syncpt_intr, work);
+	struct host1x_syncpt *syncpt =
+		container_of(syncpt_intr, struct host1x_syncpt, intr);
+	unsigned int id = syncpt->id;
+	struct host1x *host = syncpt->host;
+
+	(void)process_wait_list(host, syncpt,
+				host1x_syncpt_load(host->syncpt + id));
+}
+
+int host1x_intr_add_action(struct host1x *host, u32 id, u32 thresh,
+			   enum host1x_intr_action action, void *data,
+			   struct host1x_waitlist *waiter, void **ref)
+{
+	struct host1x_syncpt *syncpt;
+	int queue_was_empty;
+
+	if (waiter == NULL) {
+		pr_warn("%s: NULL waiter\n", __func__);
+		return -EINVAL;
+	}
+
+	/* initialize a new waiter */
+	INIT_LIST_HEAD(&waiter->list);
+	kref_init(&waiter->refcount);
+	if (ref)
+		kref_get(&waiter->refcount);
+	waiter->thresh = thresh;
+	waiter->action = action;
+	atomic_set(&waiter->state, WLS_PENDING);
+	waiter->data = data;
+	waiter->count = 1;
+
+	syncpt = host->syncpt + id;
+
+	spin_lock(&syncpt->intr.lock);
+
+	queue_was_empty = list_empty(&syncpt->intr.wait_head);
+
+	if (add_waiter_to_queue(waiter, &syncpt->intr.wait_head)) {
+		/* added at head of list - new threshold value */
+		host1x_hw_intr_set_syncpt_threshold(host, id, thresh);
+
+		/* added as first waiter - enable interrupt */
+		if (queue_was_empty)
+			host1x_hw_intr_enable_syncpt_intr(host, id);
+	}
+
+	spin_unlock(&syncpt->intr.lock);
+
+	if (ref)
+		*ref = waiter;
+	return 0;
+}
+
+void host1x_intr_put_ref(struct host1x *host, u32 id, void *ref)
+{
+	struct host1x_waitlist *waiter = ref;
+	struct host1x_syncpt *syncpt;
+
+	while (atomic_cmpxchg(&waiter->state, WLS_PENDING, WLS_CANCELLED) ==
+	       WLS_REMOVED)
+		schedule();
+
+	syncpt = host->syncpt + id;
+	(void)process_wait_list(host, syncpt,
+				host1x_syncpt_load(host->syncpt + id));
+
+	kref_put(&waiter->refcount, waiter_release);
+}
+
+int host1x_intr_init(struct host1x *host, unsigned int irq_sync)
+{
+	unsigned int id;
+	u32 nb_pts = host1x_syncpt_nb_pts(host);
+
+	mutex_init(&host->intr_mutex);
+	host->intr_syncpt_irq = irq_sync;
+	host->intr_wq = create_workqueue("host_syncpt");
+	if (!host->intr_wq)
+		return -ENOMEM;
+
+	for (id = 0; id < nb_pts; ++id) {
+		struct host1x_syncpt *syncpt = host->syncpt + id;
+
+		spin_lock_init(&syncpt->intr.lock);
+		INIT_LIST_HEAD(&syncpt->intr.wait_head);
+		snprintf(syncpt->intr.thresh_irq_name,
+			 sizeof(syncpt->intr.thresh_irq_name),
+			 "host1x_sp_%02d", id);
+	}
+
+	host1x_intr_start(host);
+
+	return 0;
+}
+
+void host1x_intr_deinit(struct host1x *host)
+{
+	host1x_intr_stop(host);
+	destroy_workqueue(host->intr_wq);
+}
+
+void host1x_intr_start(struct host1x *host)
+{
+	u32 hz = clk_get_rate(host->clk);
+	int err;
+
+	mutex_lock(&host->intr_mutex);
+	err = host1x_hw_intr_init_host_sync(host, DIV_ROUND_UP(hz, 1000000),
+					    syncpt_thresh_work);
+	if (err) {
+		mutex_unlock(&host->intr_mutex);
+		return;
+	}
+	mutex_unlock(&host->intr_mutex);
+}
+
+void host1x_intr_stop(struct host1x *host)
+{
+	unsigned int id;
+	struct host1x_syncpt *syncpt = host->syncpt;
+	u32 nb_pts = host1x_syncpt_nb_pts(host);
+
+	mutex_lock(&host->intr_mutex);
+
+	host1x_hw_intr_disable_all_syncpt_intrs(host);
+
+	for (id = 0; id < nb_pts; ++id) {
+		struct host1x_waitlist *waiter, *next;
+
+		list_for_each_entry_safe(waiter, next,
+			&syncpt[id].intr.wait_head, list) {
+			if (atomic_cmpxchg(&waiter->state,
+			    WLS_CANCELLED, WLS_HANDLED) == WLS_CANCELLED) {
+				list_del(&waiter->list);
+				kref_put(&waiter->refcount, waiter_release);
+			}
+		}
+
+		if (!list_empty(&syncpt[id].intr.wait_head)) {
+			/* output diagnostics */
+			mutex_unlock(&host->intr_mutex);
+			pr_warn("%s cannot stop syncpt intr id=%d\n",
+				__func__, id);
+			return;
+		}
+	}
+
+	host1x_hw_intr_free_syncpt_irq(host);
+
+	mutex_unlock(&host->intr_mutex);
+}
diff --git a/drivers/gpu/host1x/intr.h b/drivers/gpu/host1x/intr.h
new file mode 100644
index 00000000000..2b8adf016a0
--- /dev/null
+++ b/drivers/gpu/host1x/intr.h
@@ -0,0 +1,102 @@
+/*
+ * Tegra host1x Interrupt Management
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __HOST1X_INTR_H
+#define __HOST1X_INTR_H
+
+#include <linux/interrupt.h>
+#include <linux/workqueue.h>
+
+struct host1x;
+
+enum host1x_intr_action {
+	/*
+	 * Perform cleanup after a submit has completed.
+	 * 'data' points to a channel
+	 */
+	HOST1X_INTR_ACTION_SUBMIT_COMPLETE = 0,
+
+	/*
+	 * Wake up a  task.
+	 * 'data' points to a wait_queue_head_t
+	 */
+	HOST1X_INTR_ACTION_WAKEUP,
+
+	/*
+	 * Wake up a interruptible task.
+	 * 'data' points to a wait_queue_head_t
+	 */
+	HOST1X_INTR_ACTION_WAKEUP_INTERRUPTIBLE,
+
+	HOST1X_INTR_ACTION_COUNT
+};
+
+struct host1x_syncpt_intr {
+	spinlock_t lock;
+	struct list_head wait_head;
+	char thresh_irq_name[12];
+	struct work_struct work;
+};
+
+struct host1x_waitlist {
+	struct list_head list;
+	struct kref refcount;
+	u32 thresh;
+	enum host1x_intr_action action;
+	atomic_t state;
+	void *data;
+	int count;
+};
+
+/*
+ * Schedule an action to be taken when a sync point reaches the given threshold.
+ *
+ * @id the sync point
+ * @thresh the threshold
+ * @action the action to take
+ * @data a pointer to extra data depending on action, see above
+ * @waiter waiter structure - assumes ownership
+ * @ref must be passed if cancellation is possible, else NULL
+ *
+ * This is a non-blocking api.
+ */
+int host1x_intr_add_action(struct host1x *host, u32 id, u32 thresh,
+	enum host1x_intr_action action, void *data,
+	struct host1x_waitlist *waiter, void **ref);
+
+/*
+ * Unreference an action submitted to host1x_intr_add_action().
+ * You must call this if you passed non-NULL as ref.
+ * @ref the ref returned from host1x_intr_add_action()
+ */
+void host1x_intr_put_ref(struct host1x *host, u32 id, void *ref);
+
+/* Initialize host1x sync point interrupt */
+int host1x_intr_init(struct host1x *host, unsigned int irq_sync);
+
+/* Deinitialize host1x sync point interrupt */
+void host1x_intr_deinit(struct host1x *host);
+
+/* Enable host1x sync point interrupt */
+void host1x_intr_start(struct host1x *host);
+
+/* Disable host1x sync point interrupt */
+void host1x_intr_stop(struct host1x *host);
+
+irqreturn_t host1x_syncpt_thresh_fn(void *dev_id);
+#endif
diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c
new file mode 100644
index 00000000000..f665d679031
--- /dev/null
+++ b/drivers/gpu/host1x/job.c
@@ -0,0 +1,603 @@
+/*
+ * Tegra host1x Job
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/dma-mapping.h>
+#include <linux/err.h>
+#include <linux/kref.h>
+#include <linux/module.h>
+#include <linux/scatterlist.h>
+#include <linux/slab.h>
+#include <linux/vmalloc.h>
+#include <trace/events/host1x.h>
+
+#include "channel.h"
+#include "dev.h"
+#include "host1x_bo.h"
+#include "job.h"
+#include "syncpt.h"
+
+struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
+				    u32 num_cmdbufs, u32 num_relocs,
+				    u32 num_waitchks)
+{
+	struct host1x_job *job = NULL;
+	unsigned int num_unpins = num_cmdbufs + num_relocs;
+	u64 total;
+	void *mem;
+
+	/* Check that we're not going to overflow */
+	total = sizeof(struct host1x_job) +
+		num_relocs * sizeof(struct host1x_reloc) +
+		num_unpins * sizeof(struct host1x_job_unpin_data) +
+		num_waitchks * sizeof(struct host1x_waitchk) +
+		num_cmdbufs * sizeof(struct host1x_job_gather) +
+		num_unpins * sizeof(dma_addr_t) +
+		num_unpins * sizeof(u32 *);
+	if (total > ULONG_MAX)
+		return NULL;
+
+	mem = job = kzalloc(total, GFP_KERNEL);
+	if (!job)
+		return NULL;
+
+	kref_init(&job->ref);
+	job->channel = ch;
+
+	/* Redistribute memory to the structs  */
+	mem += sizeof(struct host1x_job);
+	job->relocarray = num_relocs ? mem : NULL;
+	mem += num_relocs * sizeof(struct host1x_reloc);
+	job->unpins = num_unpins ? mem : NULL;
+	mem += num_unpins * sizeof(struct host1x_job_unpin_data);
+	job->waitchk = num_waitchks ? mem : NULL;
+	mem += num_waitchks * sizeof(struct host1x_waitchk);
+	job->gathers = num_cmdbufs ? mem : NULL;
+	mem += num_cmdbufs * sizeof(struct host1x_job_gather);
+	job->addr_phys = num_unpins ? mem : NULL;
+
+	job->reloc_addr_phys = job->addr_phys;
+	job->gather_addr_phys = &job->addr_phys[num_relocs];
+
+	return job;
+}
+
+struct host1x_job *host1x_job_get(struct host1x_job *job)
+{
+	kref_get(&job->ref);
+	return job;
+}
+
+static void job_free(struct kref *ref)
+{
+	struct host1x_job *job = container_of(ref, struct host1x_job, ref);
+
+	kfree(job);
+}
+
+void host1x_job_put(struct host1x_job *job)
+{
+	kref_put(&job->ref, job_free);
+}
+
+void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *bo,
+			   u32 words, u32 offset)
+{
+	struct host1x_job_gather *cur_gather = &job->gathers[job->num_gathers];
+
+	cur_gather->words = words;
+	cur_gather->bo = bo;
+	cur_gather->offset = offset;
+	job->num_gathers++;
+}
+
+/*
+ * NULL an already satisfied WAIT_SYNCPT host method, by patching its
+ * args in the command stream. The method data is changed to reference
+ * a reserved (never given out or incr) HOST1X_SYNCPT_RESERVED syncpt
+ * with a matching threshold value of 0, so is guaranteed to be popped
+ * by the host HW.
+ */
+static void host1x_syncpt_patch_offset(struct host1x_syncpt *sp,
+				       struct host1x_bo *h, u32 offset)
+{
+	void *patch_addr = NULL;
+
+	/* patch the wait */
+	patch_addr = host1x_bo_kmap(h, offset >> PAGE_SHIFT);
+	if (patch_addr) {
+		host1x_syncpt_patch_wait(sp,
+					 patch_addr + (offset & ~PAGE_MASK));
+		host1x_bo_kunmap(h, offset >> PAGE_SHIFT, patch_addr);
+	} else
+		pr_err("Could not map cmdbuf for wait check\n");
+}
+
+/*
+ * Check driver supplied waitchk structs for syncpt thresholds
+ * that have already been satisfied and NULL the comparison (to
+ * avoid a wrap condition in the HW).
+ */
+static int do_waitchks(struct host1x_job *job, struct host1x *host,
+		       struct host1x_bo *patch)
+{
+	int i;
+
+	/* compare syncpt vs wait threshold */
+	for (i = 0; i < job->num_waitchk; i++) {
+		struct host1x_waitchk *wait = &job->waitchk[i];
+		struct host1x_syncpt *sp =
+			host1x_syncpt_get(host, wait->syncpt_id);
+
+		/* validate syncpt id */
+		if (wait->syncpt_id > host1x_syncpt_nb_pts(host))
+			continue;
+
+		/* skip all other gathers */
+		if (patch != wait->bo)
+			continue;
+
+		trace_host1x_syncpt_wait_check(wait->bo, wait->offset,
+					       wait->syncpt_id, wait->thresh,
+					       host1x_syncpt_read_min(sp));
+
+		if (host1x_syncpt_is_expired(sp, wait->thresh)) {
+			dev_dbg(host->dev,
+				"drop WAIT id %d (%s) thresh 0x%x, min 0x%x\n",
+				wait->syncpt_id, sp->name, wait->thresh,
+				host1x_syncpt_read_min(sp));
+
+			host1x_syncpt_patch_offset(sp, patch, wait->offset);
+		}
+
+		wait->bo = NULL;
+	}
+
+	return 0;
+}
+
+static unsigned int pin_job(struct host1x_job *job)
+{
+	unsigned int i;
+
+	job->num_unpins = 0;
+
+	for (i = 0; i < job->num_relocs; i++) {
+		struct host1x_reloc *reloc = &job->relocarray[i];
+		struct sg_table *sgt;
+		dma_addr_t phys_addr;
+
+		reloc->target = host1x_bo_get(reloc->target);
+		if (!reloc->target)
+			goto unpin;
+
+		phys_addr = host1x_bo_pin(reloc->target, &sgt);
+		if (!phys_addr)
+			goto unpin;
+
+		job->addr_phys[job->num_unpins] = phys_addr;
+		job->unpins[job->num_unpins].bo = reloc->target;
+		job->unpins[job->num_unpins].sgt = sgt;
+		job->num_unpins++;
+	}
+
+	for (i = 0; i < job->num_gathers; i++) {
+		struct host1x_job_gather *g = &job->gathers[i];
+		struct sg_table *sgt;
+		dma_addr_t phys_addr;
+
+		g->bo = host1x_bo_get(g->bo);
+		if (!g->bo)
+			goto unpin;
+
+		phys_addr = host1x_bo_pin(g->bo, &sgt);
+		if (!phys_addr)
+			goto unpin;
+
+		job->addr_phys[job->num_unpins] = phys_addr;
+		job->unpins[job->num_unpins].bo = g->bo;
+		job->unpins[job->num_unpins].sgt = sgt;
+		job->num_unpins++;
+	}
+
+	return job->num_unpins;
+
+unpin:
+	host1x_job_unpin(job);
+	return 0;
+}
+
+static unsigned int do_relocs(struct host1x_job *job, struct host1x_bo *cmdbuf)
+{
+	int i = 0;
+	u32 last_page = ~0;
+	void *cmdbuf_page_addr = NULL;
+
+	/* pin & patch the relocs for one gather */
+	while (i < job->num_relocs) {
+		struct host1x_reloc *reloc = &job->relocarray[i];
+		u32 reloc_addr = (job->reloc_addr_phys[i] +
+			reloc->target_offset) >> reloc->shift;
+		u32 *target;
+
+		/* skip all other gathers */
+		if (!(reloc->cmdbuf && cmdbuf == reloc->cmdbuf)) {
+			i++;
+			continue;
+		}
+
+		if (last_page != reloc->cmdbuf_offset >> PAGE_SHIFT) {
+			if (cmdbuf_page_addr)
+				host1x_bo_kunmap(cmdbuf, last_page,
+						 cmdbuf_page_addr);
+
+			cmdbuf_page_addr = host1x_bo_kmap(cmdbuf,
+					reloc->cmdbuf_offset >> PAGE_SHIFT);
+			last_page = reloc->cmdbuf_offset >> PAGE_SHIFT;
+
+			if (unlikely(!cmdbuf_page_addr)) {
+				pr_err("Could not map cmdbuf for relocation\n");
+				return -ENOMEM;
+			}
+		}
+
+		target = cmdbuf_page_addr + (reloc->cmdbuf_offset & ~PAGE_MASK);
+		*target = reloc_addr;
+
+		/* mark this gather as handled */
+		reloc->cmdbuf = 0;
+	}
+
+	if (cmdbuf_page_addr)
+		host1x_bo_kunmap(cmdbuf, last_page, cmdbuf_page_addr);
+
+	return 0;
+}
+
+static int check_reloc(struct host1x_reloc *reloc, struct host1x_bo *cmdbuf,
+		       unsigned int offset)
+{
+	offset *= sizeof(u32);
+
+	if (reloc->cmdbuf != cmdbuf || reloc->cmdbuf_offset != offset)
+		return -EINVAL;
+
+	return 0;
+}
+
+struct host1x_firewall {
+	struct host1x_job *job;
+	struct device *dev;
+
+	unsigned int num_relocs;
+	struct host1x_reloc *reloc;
+
+	struct host1x_bo *cmdbuf_id;
+	unsigned int offset;
+
+	u32 words;
+	u32 class;
+	u32 reg;
+	u32 mask;
+	u32 count;
+};
+
+static int check_mask(struct host1x_firewall *fw)
+{
+	u32 mask = fw->mask;
+	u32 reg = fw->reg;
+
+	while (mask) {
+		if (fw->words == 0)
+			return -EINVAL;
+
+		if (mask & 1) {
+			if (fw->job->is_addr_reg(fw->dev, fw->class, reg)) {
+				bool bad_reloc = check_reloc(fw->reloc,
+							     fw->cmdbuf_id,
+							     fw->offset);
+				if (!fw->num_relocs || bad_reloc)
+					return -EINVAL;
+				fw->reloc++;
+				fw->num_relocs--;
+			}
+			fw->words--;
+			fw->offset++;
+		}
+		mask >>= 1;
+		reg++;
+	}
+
+	return 0;
+}
+
+static int check_incr(struct host1x_firewall *fw)
+{
+	u32 count = fw->count;
+	u32 reg = fw->reg;
+
+	while (fw) {
+		if (fw->words == 0)
+			return -EINVAL;
+
+		if (fw->job->is_addr_reg(fw->dev, fw->class, reg)) {
+			bool bad_reloc = check_reloc(fw->reloc, fw->cmdbuf_id,
+						     fw->offset);
+			if (!fw->num_relocs || bad_reloc)
+				return -EINVAL;
+			fw->reloc++;
+			fw->num_relocs--;
+		}
+		reg++;
+		fw->words--;
+		fw->offset++;
+		count--;
+	}
+
+	return 0;
+}
+
+static int check_nonincr(struct host1x_firewall *fw)
+{
+	int is_addr_reg = fw->job->is_addr_reg(fw->dev, fw->class, fw->reg);
+	u32 count = fw->count;
+
+	while (count) {
+		if (fw->words == 0)
+			return -EINVAL;
+
+		if (is_addr_reg) {
+			bool bad_reloc = check_reloc(fw->reloc, fw->cmdbuf_id,
+						     fw->offset);
+			if (!fw->num_relocs || bad_reloc)
+				return -EINVAL;
+			fw->reloc++;
+			fw->num_relocs--;
+		}
+		fw->words--;
+		fw->offset++;
+		count--;
+	}
+
+	return 0;
+}
+
+static int validate(struct host1x_job *job, struct device *dev,
+		    struct host1x_job_gather *g)
+{
+	u32 *cmdbuf_base;
+	int err = 0;
+	struct host1x_firewall fw;
+
+	fw.job = job;
+	fw.dev = dev;
+	fw.reloc = job->relocarray;
+	fw.num_relocs = job->num_relocs;
+	fw.cmdbuf_id = g->bo;
+
+	fw.offset = 0;
+	fw.class = 0;
+
+	if (!job->is_addr_reg)
+		return 0;
+
+	cmdbuf_base = host1x_bo_mmap(g->bo);
+	if (!cmdbuf_base)
+		return -ENOMEM;
+
+	fw.words = g->words;
+	while (fw.words && !err) {
+		u32 word = cmdbuf_base[fw.offset];
+		u32 opcode = (word & 0xf0000000) >> 28;
+
+		fw.mask = 0;
+		fw.reg = 0;
+		fw.count = 0;
+		fw.words--;
+		fw.offset++;
+
+		switch (opcode) {
+		case 0:
+			fw.class = word >> 6 & 0x3ff;
+			fw.mask = word & 0x3f;
+			fw.reg = word >> 16 & 0xfff;
+			err = check_mask(&fw);
+			if (err)
+				goto out;
+			break;
+		case 1:
+			fw.reg = word >> 16 & 0xfff;
+			fw.count = word & 0xffff;
+			err = check_incr(&fw);
+			if (err)
+				goto out;
+			break;
+
+		case 2:
+			fw.reg = word >> 16 & 0xfff;
+			fw.count = word & 0xffff;
+			err = check_nonincr(&fw);
+			if (err)
+				goto out;
+			break;
+
+		case 3:
+			fw.mask = word & 0xffff;
+			fw.reg = word >> 16 & 0xfff;
+			err = check_mask(&fw);
+			if (err)
+				goto out;
+			break;
+		case 4:
+		case 5:
+		case 14:
+			break;
+		default:
+			err = -EINVAL;
+			break;
+		}
+	}
+
+	/* No relocs should remain at this point */
+	if (fw.num_relocs)
+		err = -EINVAL;
+
+out:
+	host1x_bo_munmap(g->bo, cmdbuf_base);
+
+	return err;
+}
+
+static inline int copy_gathers(struct host1x_job *job, struct device *dev)
+{
+	size_t size = 0;
+	size_t offset = 0;
+	int i;
+
+	for (i = 0; i < job->num_gathers; i++) {
+		struct host1x_job_gather *g = &job->gathers[i];
+		size += g->words * sizeof(u32);
+	}
+
+	job->gather_copy_mapped = dma_alloc_writecombine(dev, size,
+							 &job->gather_copy,
+							 GFP_KERNEL);
+	if (!job->gather_copy_mapped) {
+		int err = PTR_ERR(job->gather_copy_mapped);
+		job->gather_copy_mapped = NULL;
+		return err;
+	}
+
+	job->gather_copy_size = size;
+
+	for (i = 0; i < job->num_gathers; i++) {
+		struct host1x_job_gather *g = &job->gathers[i];
+		void *gather;
+
+		gather = host1x_bo_mmap(g->bo);
+		memcpy(job->gather_copy_mapped + offset, gather + g->offset,
+		       g->words * sizeof(u32));
+		host1x_bo_munmap(g->bo, gather);
+
+		g->base = job->gather_copy;
+		g->offset = offset;
+		g->bo = NULL;
+
+		offset += g->words * sizeof(u32);
+	}
+
+	return 0;
+}
+
+int host1x_job_pin(struct host1x_job *job, struct device *dev)
+{
+	int err;
+	unsigned int i, j;
+	struct host1x *host = dev_get_drvdata(dev->parent);
+	DECLARE_BITMAP(waitchk_mask, host1x_syncpt_nb_pts(host));
+
+	bitmap_zero(waitchk_mask, host1x_syncpt_nb_pts(host));
+	for (i = 0; i < job->num_waitchk; i++) {
+		u32 syncpt_id = job->waitchk[i].syncpt_id;
+		if (syncpt_id < host1x_syncpt_nb_pts(host))
+			set_bit(syncpt_id, waitchk_mask);
+	}
+
+	/* get current syncpt values for waitchk */
+	for_each_set_bit(i, waitchk_mask, host1x_syncpt_nb_pts(host))
+		host1x_syncpt_load(host->syncpt + i);
+
+	/* pin memory */
+	err = pin_job(job);
+	if (!err)
+		goto out;
+
+	/* patch gathers */
+	for (i = 0; i < job->num_gathers; i++) {
+		struct host1x_job_gather *g = &job->gathers[i];
+
+		/* process each gather mem only once */
+		if (g->handled)
+			continue;
+
+		g->base = job->gather_addr_phys[i];
+
+		for (j = 0; j < job->num_gathers; j++)
+			if (job->gathers[j].bo == g->bo)
+				job->gathers[j].handled = true;
+
+		err = 0;
+
+		if (IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL))
+			err = validate(job, dev, g);
+
+		if (err)
+			dev_err(dev, "Job invalid (err=%d)\n", err);
+
+		if (!err)
+			err = do_relocs(job, g->bo);
+
+		if (!err)
+			err = do_waitchks(job, host, g->bo);
+
+		if (err)
+			break;
+	}
+
+	if (IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL) && !err) {
+		err = copy_gathers(job, dev);
+		if (err) {
+			host1x_job_unpin(job);
+			return err;
+		}
+	}
+
+out:
+	wmb();
+
+	return err;
+}
+
+void host1x_job_unpin(struct host1x_job *job)
+{
+	unsigned int i;
+
+	for (i = 0; i < job->num_unpins; i++) {
+		struct host1x_job_unpin_data *unpin = &job->unpins[i];
+		host1x_bo_unpin(unpin->bo, unpin->sgt);
+		host1x_bo_put(unpin->bo);
+	}
+	job->num_unpins = 0;
+
+	if (job->gather_copy_size)
+		dma_free_writecombine(job->channel->dev, job->gather_copy_size,
+				      job->gather_copy_mapped,
+				      job->gather_copy);
+}
+
+/*
+ * Debug routine used to dump job entries
+ */
+void host1x_job_dump(struct device *dev, struct host1x_job *job)
+{
+	dev_dbg(dev, "    SYNCPT_ID   %d\n", job->syncpt_id);
+	dev_dbg(dev, "    SYNCPT_VAL  %d\n", job->syncpt_end);
+	dev_dbg(dev, "    FIRST_GET   0x%x\n", job->first_get);
+	dev_dbg(dev, "    TIMEOUT     %d\n", job->timeout);
+	dev_dbg(dev, "    NUM_SLOTS   %d\n", job->num_slots);
+	dev_dbg(dev, "    NUM_HANDLES %d\n", job->num_unpins);
+}
diff --git a/drivers/gpu/host1x/job.h b/drivers/gpu/host1x/job.h
new file mode 100644
index 00000000000..fba45f20458
--- /dev/null
+++ b/drivers/gpu/host1x/job.h
@@ -0,0 +1,162 @@
+/*
+ * Tegra host1x Job
+ *
+ * Copyright (c) 2011-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __HOST1X_JOB_H
+#define __HOST1X_JOB_H
+
+struct host1x_job_gather {
+	u32 words;
+	dma_addr_t base;
+	struct host1x_bo *bo;
+	int offset;
+	bool handled;
+};
+
+struct host1x_cmdbuf {
+	u32 handle;
+	u32 offset;
+	u32 words;
+	u32 pad;
+};
+
+struct host1x_reloc {
+	struct host1x_bo *cmdbuf;
+	u32 cmdbuf_offset;
+	struct host1x_bo *target;
+	u32 target_offset;
+	u32 shift;
+	u32 pad;
+};
+
+struct host1x_waitchk {
+	struct host1x_bo *bo;
+	u32 offset;
+	u32 syncpt_id;
+	u32 thresh;
+};
+
+struct host1x_job_unpin_data {
+	struct host1x_bo *bo;
+	struct sg_table *sgt;
+};
+
+/*
+ * Each submit is tracked as a host1x_job.
+ */
+struct host1x_job {
+	/* When refcount goes to zero, job can be freed */
+	struct kref ref;
+
+	/* List entry */
+	struct list_head list;
+
+	/* Channel where job is submitted to */
+	struct host1x_channel *channel;
+
+	u32 client;
+
+	/* Gathers and their memory */
+	struct host1x_job_gather *gathers;
+	unsigned int num_gathers;
+
+	/* Wait checks to be processed at submit time */
+	struct host1x_waitchk *waitchk;
+	unsigned int num_waitchk;
+	u32 waitchk_mask;
+
+	/* Array of handles to be pinned & unpinned */
+	struct host1x_reloc *relocarray;
+	unsigned int num_relocs;
+	struct host1x_job_unpin_data *unpins;
+	unsigned int num_unpins;
+
+	dma_addr_t *addr_phys;
+	dma_addr_t *gather_addr_phys;
+	dma_addr_t *reloc_addr_phys;
+
+	/* Sync point id, number of increments and end related to the submit */
+	u32 syncpt_id;
+	u32 syncpt_incrs;
+	u32 syncpt_end;
+
+	/* Maximum time to wait for this job */
+	unsigned int timeout;
+
+	/* Index and number of slots used in the push buffer */
+	unsigned int first_get;
+	unsigned int num_slots;
+
+	/* Copy of gathers */
+	size_t gather_copy_size;
+	dma_addr_t gather_copy;
+	u8 *gather_copy_mapped;
+
+	/* Check if register is marked as an address reg */
+	int (*is_addr_reg)(struct device *dev, u32 reg, u32 class);
+
+	/* Request a SETCLASS to this class */
+	u32 class;
+
+	/* Add a channel wait for previous ops to complete */
+	bool serialize;
+};
+/*
+ * Allocate memory for a job. Just enough memory will be allocated to
+ * accomodate the submit.
+ */
+struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
+				    u32 num_cmdbufs, u32 num_relocs,
+				    u32 num_waitchks);
+
+/*
+ * Add a gather to a job.
+ */
+void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *mem_id,
+			   u32 words, u32 offset);
+
+/*
+ * Increment reference going to host1x_job.
+ */
+struct host1x_job *host1x_job_get(struct host1x_job *job);
+
+/*
+ * Decrement reference job, free if goes to zero.
+ */
+void host1x_job_put(struct host1x_job *job);
+
+/*
+ * Pin memory related to job. This handles relocation of addresses to the
+ * host1x address space. Handles both the gather memory and any other memory
+ * referred to from the gather buffers.
+ *
+ * Handles also patching out host waits that would wait for an expired sync
+ * point value.
+ */
+int host1x_job_pin(struct host1x_job *job, struct device *dev);
+
+/*
+ * Unpin memory related to job.
+ */
+void host1x_job_unpin(struct host1x_job *job);
+
+/*
+ * Dump contents of job to debug output.
+ */
+void host1x_job_dump(struct device *dev, struct host1x_job *job);
+
+#endif
diff --git a/drivers/gpu/host1x/syncpt.c b/drivers/gpu/host1x/syncpt.c
new file mode 100644
index 00000000000..4b493453e80
--- /dev/null
+++ b/drivers/gpu/host1x/syncpt.c
@@ -0,0 +1,387 @@
+/*
+ * Tegra host1x Syncpoints
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/slab.h>
+
+#include <trace/events/host1x.h>
+
+#include "syncpt.h"
+#include "dev.h"
+#include "intr.h"
+#include "debug.h"
+
+#define SYNCPT_CHECK_PERIOD (2 * HZ)
+#define MAX_STUCK_CHECK_COUNT 15
+
+static struct host1x_syncpt *_host1x_syncpt_alloc(struct host1x *host,
+						  struct device *dev,
+						  int client_managed)
+{
+	int i;
+	struct host1x_syncpt *sp = host->syncpt;
+	char *name;
+
+	for (i = 0; i < host->info->nb_pts && sp->name; i++, sp++)
+		;
+	if (sp->dev)
+		return NULL;
+
+	name = kasprintf(GFP_KERNEL, "%02d-%s", sp->id,
+			dev ? dev_name(dev) : NULL);
+	if (!name)
+		return NULL;
+
+	sp->dev = dev;
+	sp->name = name;
+	sp->client_managed = client_managed;
+
+	return sp;
+}
+
+u32 host1x_syncpt_id(struct host1x_syncpt *sp)
+{
+	return sp->id;
+}
+
+/*
+ * Updates the value sent to hardware.
+ */
+u32 host1x_syncpt_incr_max(struct host1x_syncpt *sp, u32 incrs)
+{
+	return (u32)atomic_add_return(incrs, &sp->max_val);
+}
+
+ /*
+ * Write cached syncpoint and waitbase values to hardware.
+ */
+void host1x_syncpt_restore(struct host1x *host)
+{
+	struct host1x_syncpt *sp_base = host->syncpt;
+	u32 i;
+
+	for (i = 0; i < host1x_syncpt_nb_pts(host); i++)
+		host1x_hw_syncpt_restore(host, sp_base + i);
+	for (i = 0; i < host1x_syncpt_nb_bases(host); i++)
+		host1x_hw_syncpt_restore_wait_base(host, sp_base + i);
+	wmb();
+}
+
+/*
+ * Update the cached syncpoint and waitbase values by reading them
+ * from the registers.
+  */
+void host1x_syncpt_save(struct host1x *host)
+{
+	struct host1x_syncpt *sp_base = host->syncpt;
+	u32 i;
+
+	for (i = 0; i < host1x_syncpt_nb_pts(host); i++) {
+		if (host1x_syncpt_client_managed(sp_base + i))
+			host1x_hw_syncpt_load(host, sp_base + i);
+		else
+			WARN_ON(!host1x_syncpt_idle(sp_base + i));
+	}
+
+	for (i = 0; i < host1x_syncpt_nb_bases(host); i++)
+		host1x_hw_syncpt_load_wait_base(host, sp_base + i);
+}
+
+/*
+ * Updates the cached syncpoint value by reading a new value from the hardware
+ * register
+ */
+u32 host1x_syncpt_load(struct host1x_syncpt *sp)
+{
+	u32 val;
+	val = host1x_hw_syncpt_load(sp->host, sp);
+	trace_host1x_syncpt_load_min(sp->id, val);
+
+	return val;
+}
+
+/*
+ * Get the current syncpoint base
+ */
+u32 host1x_syncpt_load_wait_base(struct host1x_syncpt *sp)
+{
+	u32 val;
+	host1x_hw_syncpt_load_wait_base(sp->host, sp);
+	val = sp->base_val;
+	return val;
+}
+
+/*
+ * Write a cpu syncpoint increment to the hardware, without touching
+ * the cache. Caller is responsible for host being powered.
+ */
+void host1x_syncpt_cpu_incr(struct host1x_syncpt *sp)
+{
+	host1x_hw_syncpt_cpu_incr(sp->host, sp);
+}
+
+/*
+ * Increment syncpoint value from cpu, updating cache
+ */
+void host1x_syncpt_incr(struct host1x_syncpt *sp)
+{
+	if (host1x_syncpt_client_managed(sp))
+		host1x_syncpt_incr_max(sp, 1);
+	host1x_syncpt_cpu_incr(sp);
+}
+
+/*
+ * Updated sync point form hardware, and returns true if syncpoint is expired,
+ * false if we may need to wait
+ */
+static bool syncpt_load_min_is_expired(struct host1x_syncpt *sp, u32 thresh)
+{
+	host1x_hw_syncpt_load(sp->host, sp);
+	return host1x_syncpt_is_expired(sp, thresh);
+}
+
+/*
+ * Main entrypoint for syncpoint value waits.
+ */
+int host1x_syncpt_wait(struct host1x_syncpt *sp, u32 thresh, long timeout,
+			u32 *value)
+{
+	DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq);
+	void *ref;
+	struct host1x_waitlist *waiter;
+	int err = 0, check_count = 0;
+	u32 val;
+
+	if (value)
+		*value = 0;
+
+	/* first check cache */
+	if (host1x_syncpt_is_expired(sp, thresh)) {
+		if (value)
+			*value = host1x_syncpt_load(sp);
+		return 0;
+	}
+
+	/* try to read from register */
+	val = host1x_hw_syncpt_load(sp->host, sp);
+	if (host1x_syncpt_is_expired(sp, thresh)) {
+		if (value)
+			*value = val;
+		goto done;
+	}
+
+	if (!timeout) {
+		err = -EAGAIN;
+		goto done;
+	}
+
+	/* allocate a waiter */
+	waiter = kzalloc(sizeof(*waiter), GFP_KERNEL);
+	if (!waiter) {
+		err = -ENOMEM;
+		goto done;
+	}
+
+	/* schedule a wakeup when the syncpoint value is reached */
+	err = host1x_intr_add_action(sp->host, sp->id, thresh,
+				     HOST1X_INTR_ACTION_WAKEUP_INTERRUPTIBLE,
+				     &wq, waiter, &ref);
+	if (err)
+		goto done;
+
+	err = -EAGAIN;
+	/* Caller-specified timeout may be impractically low */
+	if (timeout < 0)
+		timeout = LONG_MAX;
+
+	/* wait for the syncpoint, or timeout, or signal */
+	while (timeout) {
+		long check = min_t(long, SYNCPT_CHECK_PERIOD, timeout);
+		int remain = wait_event_interruptible_timeout(wq,
+				syncpt_load_min_is_expired(sp, thresh),
+				check);
+		if (remain > 0 || host1x_syncpt_is_expired(sp, thresh)) {
+			if (value)
+				*value = host1x_syncpt_load(sp);
+			err = 0;
+			break;
+		}
+		if (remain < 0) {
+			err = remain;
+			break;
+		}
+		timeout -= check;
+		if (timeout && check_count <= MAX_STUCK_CHECK_COUNT) {
+			dev_warn(sp->host->dev,
+				"%s: syncpoint id %d (%s) stuck waiting %d, timeout=%ld\n",
+				 current->comm, sp->id, sp->name,
+				 thresh, timeout);
+
+			host1x_debug_dump_syncpts(sp->host);
+			if (check_count == MAX_STUCK_CHECK_COUNT)
+				host1x_debug_dump(sp->host);
+			check_count++;
+		}
+	}
+	host1x_intr_put_ref(sp->host, sp->id, ref);
+
+done:
+	return err;
+}
+EXPORT_SYMBOL(host1x_syncpt_wait);
+
+/*
+ * Returns true if syncpoint is expired, false if we may need to wait
+ */
+bool host1x_syncpt_is_expired(struct host1x_syncpt *sp, u32 thresh)
+{
+	u32 current_val;
+	u32 future_val;
+	smp_rmb();
+	current_val = (u32)atomic_read(&sp->min_val);
+	future_val = (u32)atomic_read(&sp->max_val);
+
+	/* Note the use of unsigned arithmetic here (mod 1<<32).
+	 *
+	 * c = current_val = min_val	= the current value of the syncpoint.
+	 * t = thresh			= the value we are checking
+	 * f = future_val  = max_val	= the value c will reach when all
+	 *				  outstanding increments have completed.
+	 *
+	 * Note that c always chases f until it reaches f.
+	 *
+	 * Dtf = (f - t)
+	 * Dtc = (c - t)
+	 *
+	 *  Consider all cases:
+	 *
+	 *	A) .....c..t..f.....	Dtf < Dtc	need to wait
+	 *	B) .....c.....f..t..	Dtf > Dtc	expired
+	 *	C) ..t..c.....f.....	Dtf > Dtc	expired	   (Dct very large)
+	 *
+	 *  Any case where f==c: always expired (for any t).	Dtf == Dcf
+	 *  Any case where t==c: always expired (for any f).	Dtf >= Dtc (because Dtc==0)
+	 *  Any case where t==f!=c: always wait.		Dtf <  Dtc (because Dtf==0,
+	 *							Dtc!=0)
+	 *
+	 *  Other cases:
+	 *
+	 *	A) .....t..f..c.....	Dtf < Dtc	need to wait
+	 *	A) .....f..c..t.....	Dtf < Dtc	need to wait
+	 *	A) .....f..t..c.....	Dtf > Dtc	expired
+	 *
+	 *   So:
+	 *	   Dtf >= Dtc implies EXPIRED	(return true)
+	 *	   Dtf <  Dtc implies WAIT	(return false)
+	 *
+	 * Note: If t is expired then we *cannot* wait on it. We would wait
+	 * forever (hang the system).
+	 *
+	 * Note: do NOT get clever and remove the -thresh from both sides. It
+	 * is NOT the same.
+	 *
+	 * If future valueis zero, we have a client managed sync point. In that
+	 * case we do a direct comparison.
+	 */
+	if (!host1x_syncpt_client_managed(sp))
+		return future_val - thresh >= current_val - thresh;
+	else
+		return (s32)(current_val - thresh) >= 0;
+}
+
+/* remove a wait pointed to by patch_addr */
+int host1x_syncpt_patch_wait(struct host1x_syncpt *sp, void *patch_addr)
+{
+	return host1x_hw_syncpt_patch_wait(sp->host, sp, patch_addr);
+}
+
+int host1x_syncpt_init(struct host1x *host)
+{
+	struct host1x_syncpt *syncpt;
+	int i;
+
+	syncpt = devm_kzalloc(host->dev, sizeof(*syncpt) * host->info->nb_pts,
+		GFP_KERNEL);
+	if (!syncpt)
+		return -ENOMEM;
+
+	for (i = 0; i < host->info->nb_pts; ++i) {
+		syncpt[i].id = i;
+		syncpt[i].host = host;
+	}
+
+	host->syncpt = syncpt;
+
+	host1x_syncpt_restore(host);
+
+	/* Allocate sync point to use for clearing waits for expired fences */
+	host->nop_sp = _host1x_syncpt_alloc(host, NULL, 0);
+	if (!host->nop_sp)
+		return -ENOMEM;
+
+	return 0;
+}
+
+struct host1x_syncpt *host1x_syncpt_request(struct device *dev,
+					    int client_managed)
+{
+	struct host1x *host = dev_get_drvdata(dev->parent);
+	return _host1x_syncpt_alloc(host, dev, client_managed);
+}
+
+void host1x_syncpt_free(struct host1x_syncpt *sp)
+{
+	if (!sp)
+		return;
+
+	kfree(sp->name);
+	sp->dev = NULL;
+	sp->name = NULL;
+	sp->client_managed = 0;
+}
+
+void host1x_syncpt_deinit(struct host1x *host)
+{
+	int i;
+	struct host1x_syncpt *sp = host->syncpt;
+	for (i = 0; i < host->info->nb_pts; i++, sp++)
+		kfree(sp->name);
+}
+
+int host1x_syncpt_nb_pts(struct host1x *host)
+{
+	return host->info->nb_pts;
+}
+
+int host1x_syncpt_nb_bases(struct host1x *host)
+{
+	return host->info->nb_bases;
+}
+
+int host1x_syncpt_nb_mlocks(struct host1x *host)
+{
+	return host->info->nb_mlocks;
+}
+
+struct host1x_syncpt *host1x_syncpt_get(struct host1x *host, u32 id)
+{
+	if (host->info->nb_pts < id)
+		return NULL;
+	return host->syncpt + id;
+}
diff --git a/drivers/gpu/host1x/syncpt.h b/drivers/gpu/host1x/syncpt.h
new file mode 100644
index 00000000000..c99806130f2
--- /dev/null
+++ b/drivers/gpu/host1x/syncpt.h
@@ -0,0 +1,165 @@
+/*
+ * Tegra host1x Syncpoints
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __HOST1X_SYNCPT_H
+#define __HOST1X_SYNCPT_H
+
+#include <linux/atomic.h>
+#include <linux/kernel.h>
+#include <linux/sched.h>
+
+#include "intr.h"
+
+struct host1x;
+
+/* Reserved for replacing an expired wait with a NOP */
+#define HOST1X_SYNCPT_RESERVED			0
+
+struct host1x_syncpt {
+	int id;
+	atomic_t min_val;
+	atomic_t max_val;
+	u32 base_val;
+	const char *name;
+	int client_managed;
+	struct host1x *host;
+	struct device *dev;
+
+	/* interrupt data */
+	struct host1x_syncpt_intr intr;
+};
+
+/* Initialize sync point array  */
+int host1x_syncpt_init(struct host1x *host);
+
+/*  Free sync point array */
+void host1x_syncpt_deinit(struct host1x *host);
+
+/*
+ * Read max. It indicates how many operations there are in queue, either in
+ * channel or in a software thread.
+ * */
+static inline u32 host1x_syncpt_read_max(struct host1x_syncpt *sp)
+{
+	smp_rmb();
+	return (u32)atomic_read(&sp->max_val);
+}
+
+/*
+ * Read min, which is a shadow of the current sync point value in hardware.
+ */
+static inline u32 host1x_syncpt_read_min(struct host1x_syncpt *sp)
+{
+	smp_rmb();
+	return (u32)atomic_read(&sp->min_val);
+}
+
+/* Return number of sync point supported. */
+int host1x_syncpt_nb_pts(struct host1x *host);
+
+/* Return number of wait bases supported. */
+int host1x_syncpt_nb_bases(struct host1x *host);
+
+/* Return number of mlocks supported. */
+int host1x_syncpt_nb_mlocks(struct host1x *host);
+
+/*
+ * Check sync point sanity. If max is larger than min, there have too many
+ * sync point increments.
+ *
+ * Client managed sync point are not tracked.
+ * */
+static inline bool host1x_syncpt_check_max(struct host1x_syncpt *sp, u32 real)
+{
+	u32 max;
+	if (sp->client_managed)
+		return true;
+	max = host1x_syncpt_read_max(sp);
+	return (s32)(max - real) >= 0;
+}
+
+/* Return true if sync point is client managed. */
+static inline int host1x_syncpt_client_managed(struct host1x_syncpt *sp)
+{
+	return sp->client_managed;
+}
+
+/*
+ * Returns true if syncpoint min == max, which means that there are no
+ * outstanding operations.
+ */
+static inline bool host1x_syncpt_idle(struct host1x_syncpt *sp)
+{
+	int min, max;
+	smp_rmb();
+	min = atomic_read(&sp->min_val);
+	max = atomic_read(&sp->max_val);
+	return (min == max);
+}
+
+/* Return pointer to struct denoting sync point id. */
+struct host1x_syncpt *host1x_syncpt_get(struct host1x *host, u32 id);
+
+/* Request incrementing a sync point. */
+void host1x_syncpt_cpu_incr(struct host1x_syncpt *sp);
+
+/* Load current value from hardware to the shadow register. */
+u32 host1x_syncpt_load(struct host1x_syncpt *sp);
+
+/* Check if the given syncpoint value has already passed */
+bool host1x_syncpt_is_expired(struct host1x_syncpt *sp, u32 thresh);
+
+/* Save host1x sync point state into shadow registers. */
+void host1x_syncpt_save(struct host1x *host);
+
+/* Reset host1x sync point state from shadow registers. */
+void host1x_syncpt_restore(struct host1x *host);
+
+/* Read current wait base value into shadow register and return it. */
+u32 host1x_syncpt_load_wait_base(struct host1x_syncpt *sp);
+
+/* Increment sync point and its max. */
+void host1x_syncpt_incr(struct host1x_syncpt *sp);
+
+/* Indicate future operations by incrementing the sync point max. */
+u32 host1x_syncpt_incr_max(struct host1x_syncpt *sp, u32 incrs);
+
+/* Wait until sync point reaches a threshold value, or a timeout. */
+int host1x_syncpt_wait(struct host1x_syncpt *sp, u32 thresh,
+			long timeout, u32 *value);
+
+/* Check if sync point id is valid. */
+static inline int host1x_syncpt_is_valid(struct host1x_syncpt *sp)
+{
+	return sp->id < host1x_syncpt_nb_pts(sp->host);
+}
+
+/* Patch a wait by replacing it with a wait for syncpt 0 value 0 */
+int host1x_syncpt_patch_wait(struct host1x_syncpt *sp, void *patch_addr);
+
+/* Return id of the sync point */
+u32 host1x_syncpt_id(struct host1x_syncpt *sp);
+
+/* Allocate a sync point for a device. */
+struct host1x_syncpt *host1x_syncpt_request(struct device *dev,
+		int client_managed);
+
+/* Free a sync point. */
+void host1x_syncpt_free(struct host1x_syncpt *sp);
+
+#endif