x86/kprobes: Fix crash when probing CS CALL instructionsgrafted

When using eBPF to probe CS CALL instructions within a function,
a crash can be triggered.

The eBPF tool probes offset 257 of the __hrtimer_run_queues()
function:

<__hrtimer_run_queues+249>:  nopl   0x0(%rax,%rax,1)
<__hrtimer_run_queues+254>:  mov    %r14,%rdi
<__hrtimer_run_queues+257>:  cs call <__x86_indirect_thunk_r12>
<__hrtimer_run_queues+263>:  mov    %eax,%r12d
<__hrtimer_run_queues+266>:  xchg   %ax,%ax
<__hrtimer_run_queues+268>:  mov    %r13,%rdi

Which triggers this crash:

  BUG: unable to handle page fault for address: 00000000000f41c9
  #PF: supervisor write access in kernel mode
  #PF: error_code(0x0002) - not-present page
  PGD 0 P4D 0
  Oops: 0002 [#1] SMP NOPTI
  CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P
  RIP: 0010:__hrtimer_run_queues+0x106/0x230

Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is
at the 6th byte of the above CS CALL instruction. Since the CS CALL
instruction occupies 6 bytes, the exception occurred in the middle of that
call instruction.

The root cause is that when using eBPF tools to probe in the middle of a
function, a kprobe with INT3 is used as the underlying implementation.

During single-step emulation of the original CALL instruction,
int3_emulate_call() assumes that the probed CALL instruction is 5 bytes
long. However, the actual CS-prefixed CALL instruction occupies 6 bytes,
so it constructs an incorrect exception return address. When the CPU
returns from the kprobe handler, the next instruction to be executed is at
the address of the last byte of that CS CALL instruction. Coincidentally,
starting from that address, the CPU fetches and decodes a completely
different instruction, which ultimately triggers a kernel crash.

Fix the issue by using the actual instruction length obtained from
the instruction decoder when constructing the exception return
address, rather than relying on the hardcoded CALL_INSN_SIZE macro.

[ mingo: Refined the changelog ]

Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step")
Suggested-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Jinke Han <jinkehan@didiglobal.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Acked-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Borislav Petkov <bp@alien8.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000

61 files changed, 5383 insertions, 0 deletions

diff --git a/drivers/gpu/host1x/hw/cdma_hw.c b/drivers/gpu/host1x/hw/cdma_hw.c
new file mode 100644
index 000000000..08a3cf2b1
--- /dev/null
+++ b/drivers/gpu/host1x/hw/cdma_hw.c
@@ -0,0 +1,379 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Tegra host1x Command DMA
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ */
+
+#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 memory
+ */
+static void push_buffer_init(struct push_buffer *pb)
+{
+	*(u32 *)(pb->mapped + pb->size) = 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)
+{
+	unsigned int i;
+
+	for (i = 0; i < syncpt_incrs; i++)
+		host1x_syncpt_incr(cdma->timeout.syncpt);
+
+	/* after CPU incr, ensure shadow is up to date */
+	host1x_syncpt_load(cdma->timeout.syncpt);
+}
+
+/*
+ * Start channel DMA
+ */
+static void cdma_start(struct host1x_cdma *cdma)
+{
+	struct host1x_channel *ch = cdma_to_channel(cdma);
+	u64 start, end;
+
+	if (cdma->running)
+		return;
+
+	cdma->last_pos = cdma->push_buffer.pos;
+	start = cdma->push_buffer.dma;
+	end = cdma->push_buffer.size + 4;
+
+	host1x_ch_writel(ch, HOST1X_CHANNEL_DMACTRL_DMASTOP,
+			 HOST1X_CHANNEL_DMACTRL);
+
+	/* set base, put and end pointer */
+	host1x_ch_writel(ch, lower_32_bits(start), HOST1X_CHANNEL_DMASTART);
+#if HOST1X_HW >= 6
+	host1x_ch_writel(ch, upper_32_bits(start), HOST1X_CHANNEL_DMASTART_HI);
+#endif
+	host1x_ch_writel(ch, cdma->push_buffer.pos, HOST1X_CHANNEL_DMAPUT);
+#if HOST1X_HW >= 6
+	host1x_ch_writel(ch, 0, HOST1X_CHANNEL_DMAPUT_HI);
+#endif
+	host1x_ch_writel(ch, lower_32_bits(end), HOST1X_CHANNEL_DMAEND);
+#if HOST1X_HW >= 6
+	host1x_ch_writel(ch, upper_32_bits(end), HOST1X_CHANNEL_DMAEND_HI);
+#endif
+
+	/* 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);
+	u64 start, end;
+
+	if (cdma->running)
+		return;
+
+	cdma->last_pos = cdma->push_buffer.pos;
+
+	host1x_ch_writel(ch, HOST1X_CHANNEL_DMACTRL_DMASTOP,
+			 HOST1X_CHANNEL_DMACTRL);
+
+	start = cdma->push_buffer.dma;
+	end = cdma->push_buffer.size + 4;
+
+	/* set base, end pointer (all of memory) */
+	host1x_ch_writel(ch, lower_32_bits(start), HOST1X_CHANNEL_DMASTART);
+#if HOST1X_HW >= 6
+	host1x_ch_writel(ch, upper_32_bits(start), HOST1X_CHANNEL_DMASTART_HI);
+#endif
+	host1x_ch_writel(ch, lower_32_bits(end), HOST1X_CHANNEL_DMAEND);
+#if HOST1X_HW >= 6
+	host1x_ch_writel(ch, upper_32_bits(end), HOST1X_CHANNEL_DMAEND_HI);
+#endif
+
+	/* 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);
+}
+
+static void cdma_hw_cmdproc_stop(struct host1x *host, struct host1x_channel *ch,
+				 bool stop)
+{
+#if HOST1X_HW >= 6
+	host1x_ch_writel(ch, stop ? 0x1 : 0x0, HOST1X_CHANNEL_CMDPROC_STOP);
+#else
+	u32 cmdproc_stop = host1x_sync_readl(host, HOST1X_SYNC_CMDPROC_STOP);
+	if (stop)
+		cmdproc_stop |= BIT(ch->id);
+	else
+		cmdproc_stop &= ~BIT(ch->id);
+	host1x_sync_writel(host, cmdproc_stop, HOST1X_SYNC_CMDPROC_STOP);
+#endif
+}
+
+static void cdma_hw_teardown(struct host1x *host, struct host1x_channel *ch)
+{
+#if HOST1X_HW >= 6
+	host1x_ch_writel(ch, 0x1, HOST1X_CHANNEL_TEARDOWN);
+#else
+	host1x_sync_writel(host, BIT(ch->id), HOST1X_SYNC_CH_TEARDOWN);
+#endif
+}
+
+/*
+ * 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);
+
+	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);
+
+	cdma_hw_cmdproc_stop(host, ch, true);
+
+	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);
+
+	cdma_hw_teardown(host, ch);
+
+	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);
+
+	dev_dbg(host1x->dev,
+		"resuming channel (id %u, DMAGET restart = 0x%x)\n",
+		ch->id, getptr);
+
+	cdma_hw_cmdproc_stop(host1x, ch, false);
+
+	cdma->torndown = false;
+	cdma_timeout_restart(cdma, getptr);
+}
+
+static void timeout_release_mlock(struct host1x_cdma *cdma)
+{
+#if HOST1X_HW >= 8
+	/* Tegra186 and Tegra194 require a more complicated MLOCK release
+	 * sequence. Furthermore, those chips by default don't enforce MLOCKs,
+	 * so it turns out that if we don't /actually/ need MLOCKs, we can just
+	 * ignore them.
+	 *
+	 * As such, for now just implement this on Tegra234 and above where things
+	 * are stricter but also easy to implement.
+	 */
+	struct host1x_channel *ch = cdma_to_channel(cdma);
+	struct host1x *host1x = cdma_to_host1x(cdma);
+	u32 offset;
+
+	switch (ch->client->class) {
+	case HOST1X_CLASS_VIC:
+		offset = HOST1X_COMMON_VIC_MLOCK;
+		break;
+#if HOST1X_HW == 8
+	case HOST1X_CLASS_NVJPG1:
+		offset = HOST1X_COMMON_NVJPG1_MLOCK;
+		break;
+	case HOST1X_CLASS_NVENC:
+		offset = HOST1X_COMMON_NVENC_MLOCK;
+		break;
+	case HOST1X_CLASS_NVJPG:
+		offset = HOST1X_COMMON_NVJPG_MLOCK;
+		break;
+	case HOST1X_CLASS_NVDEC:
+		offset = HOST1X_COMMON_NVDEC_MLOCK;
+		break;
+	case HOST1X_CLASS_OFA:
+		offset = HOST1X_COMMON_OFA_MLOCK;
+		break;
+#endif
+	default:
+		WARN(1, "%s was not updated for class %u", __func__, ch->client->class);
+		return;
+	}
+
+	host1x_common_writel(host1x, 0x0, offset);
+#endif
+}
+
+/*
+ * 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)
+{
+	u32 syncpt_val;
+	struct host1x_cdma *cdma;
+	struct host1x *host1x;
+	struct host1x_channel *ch;
+
+	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 */
+	cdma_hw_cmdproc_stop(host1x, ch, true);
+
+	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 */
+		cdma_hw_cmdproc_stop(host1x, ch, false);
+		mutex_unlock(&cdma->lock);
+		return;
+	}
+
+	dev_warn(host1x->dev, "%s: timeout: %u (%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);
+
+	/* release any held MLOCK */
+	timeout_release_mlock(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)
+{
+	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_sync(&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 000000000..a8251ec08
--- /dev/null
+++ b/drivers/gpu/host1x/hw/channel_hw.c
@@ -0,0 +1,392 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Tegra host1x Channel
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ */
+
+#include <linux/host1x.h>
+#include <linux/iommu.h>
+#include <linux/overflow.h>
+#include <linux/slab.h>
+
+#include <trace/events/host1x.h>
+
+#include "../channel.h"
+#include "../dev.h"
+#include "../intr.h"
+#include "../job.h"
+
+#define TRACE_MAX_LENGTH 128U
+
+static void trace_write_gather(struct host1x_cdma *cdma, struct host1x_bo *bo,
+			       u32 offset, u32 words)
+{
+	struct device *dev = cdma_to_channel(cdma)->dev;
+	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) {
+			u32 num_words = min(words - i, TRACE_MAX_LENGTH);
+
+			trace_host1x_cdma_push_gather(dev_name(dev), bo,
+						      num_words,
+						      offset + i * sizeof(u32),
+						      mem);
+		}
+
+		host1x_bo_munmap(bo, mem);
+	}
+}
+
+static void submit_wait(struct host1x_job *job, u32 id, u32 threshold)
+{
+	struct host1x_cdma *cdma = &job->channel->cdma;
+
+#if HOST1X_HW >= 2
+	host1x_cdma_push_wide(cdma,
+		host1x_opcode_setclass(
+			HOST1X_CLASS_HOST1X,
+			HOST1X_UCLASS_LOAD_SYNCPT_PAYLOAD_32,
+			/* WAIT_SYNCPT_32 is at SYNCPT_PAYLOAD_32+2 */
+			BIT(0) | BIT(2)
+		),
+		threshold,
+		id,
+		HOST1X_OPCODE_NOP
+	);
+#else
+	/* TODO add waitchk or use waitbases or other mitigation */
+	host1x_cdma_push(cdma,
+		host1x_opcode_setclass(
+			HOST1X_CLASS_HOST1X,
+			host1x_uclass_wait_syncpt_r(),
+			BIT(0)
+		),
+		host1x_class_host_wait_syncpt(id, threshold)
+	);
+#endif
+}
+
+static void submit_setclass(struct host1x_job *job, u32 next_class)
+{
+	struct host1x_cdma *cdma = &job->channel->cdma;
+
+#if HOST1X_HW >= 6
+	u32 stream_id;
+
+	/*
+	 * If a memory context has been set, use it. Otherwise
+	 * (if context isolation is disabled) use the engine's
+	 * firmware stream ID.
+	 */
+	if (job->memory_context)
+		stream_id = job->memory_context->stream_id;
+	else
+		stream_id = job->engine_fallback_streamid;
+
+	host1x_cdma_push_wide(cdma,
+		host1x_opcode_setclass(next_class, 0, 0),
+		host1x_opcode_setpayload(stream_id),
+		host1x_opcode_setstreamid(job->engine_streamid_offset / 4),
+		HOST1X_OPCODE_NOP);
+#else
+	host1x_cdma_push(cdma,
+		host1x_opcode_setclass(next_class, 0, 0),
+		HOST1X_OPCODE_NOP
+	);
+#endif
+}
+
+static void submit_gathers(struct host1x_job *job, struct host1x_job_cmd *cmds, u32 num_cmds,
+			   u32 job_syncpt_base)
+{
+	struct host1x_cdma *cdma = &job->channel->cdma;
+#if HOST1X_HW < 6
+	struct device *dev = job->channel->dev;
+#endif
+	unsigned int i;
+	u32 threshold;
+
+	for (i = 0; i < num_cmds; i++) {
+		struct host1x_job_cmd *cmd = &cmds[i];
+
+		if (cmd->is_wait) {
+			if (cmd->wait.relative)
+				threshold = wrapping_add(u32, job_syncpt_base,
+							 cmd->wait.threshold);
+			else
+				threshold = cmd->wait.threshold;
+
+			submit_wait(job, cmd->wait.id, threshold);
+			submit_setclass(job, cmd->wait.next_class);
+		} else {
+			struct host1x_job_gather *g = &cmd->gather;
+
+			dma_addr_t addr = g->base + g->offset;
+			u32 op2, op3;
+
+			op2 = lower_32_bits(addr);
+			op3 = upper_32_bits(addr);
+
+			trace_write_gather(cdma, g->bo, g->offset, g->words);
+
+			if (op3 != 0) {
+#if HOST1X_HW >= 6
+				u32 op1 = host1x_opcode_gather_wide(g->words);
+				u32 op4 = HOST1X_OPCODE_NOP;
+
+				host1x_cdma_push_wide(cdma, op1, op2, op3, op4);
+#else
+				dev_err(dev, "invalid gather for push buffer %pad\n",
+					&addr);
+				continue;
+#endif
+			} else {
+				u32 op1 = host1x_opcode_gather(g->words);
+
+				host1x_cdma_push(cdma, op1, op2);
+			}
+		}
+	}
+}
+
+static inline void synchronize_syncpt_base(struct host1x_job *job)
+{
+	struct host1x_syncpt *sp = job->syncpt;
+	unsigned int id;
+	u32 value;
+
+	value = host1x_syncpt_read_max(sp);
+	id = sp->base->id;
+
+	host1x_cdma_push(&job->channel->cdma,
+			 host1x_opcode_setclass(HOST1X_CLASS_HOST1X,
+				HOST1X_UCLASS_LOAD_SYNCPT_BASE, 1),
+			 HOST1X_UCLASS_LOAD_SYNCPT_BASE_BASE_INDX_F(id) |
+			 HOST1X_UCLASS_LOAD_SYNCPT_BASE_VALUE_F(value));
+}
+
+static void host1x_channel_set_streamid(struct host1x_channel *channel)
+{
+#if HOST1X_HW >= 6
+	u32 stream_id;
+
+	if (!tegra_dev_iommu_get_stream_id(channel->dev->parent, &stream_id))
+		stream_id = TEGRA_STREAM_ID_BYPASS;
+
+	host1x_ch_writel(channel, stream_id, HOST1X_CHANNEL_SMMU_STREAMID);
+#endif
+}
+
+static void host1x_enable_gather_filter(struct host1x_channel *ch)
+{
+#if HOST1X_HW >= 6
+	struct host1x *host = dev_get_drvdata(ch->dev->parent);
+	u32 val;
+
+	if (!host->hv_regs)
+		return;
+
+	val = host1x_hypervisor_readl(
+		host, HOST1X_HV_CH_KERNEL_FILTER_GBUFFER(ch->id / 32));
+	val |= BIT(ch->id % 32);
+	host1x_hypervisor_writel(
+		host, val, HOST1X_HV_CH_KERNEL_FILTER_GBUFFER(ch->id / 32));
+#elif HOST1X_HW >= 4
+	host1x_ch_writel(ch,
+			 HOST1X_CHANNEL_CHANNELCTRL_KERNEL_FILTER_GBUFFER(1),
+			 HOST1X_CHANNEL_CHANNELCTRL);
+#endif
+}
+
+static void channel_program_cdma(struct host1x_job *job)
+{
+	struct host1x_cdma *cdma = &job->channel->cdma;
+	struct host1x_syncpt *sp = job->syncpt;
+
+#if HOST1X_HW >= 6
+	u32 fence;
+	int i = 0;
+
+	if (job->num_cmds == 0)
+		goto prefences_done;
+	if (!job->cmds[0].is_wait || job->cmds[0].wait.relative)
+		goto prefences_done;
+
+	/* Enter host1x class with invalid stream ID for prefence waits. */
+	host1x_cdma_push_wide(cdma,
+		host1x_opcode_acquire_mlock(1),
+		host1x_opcode_setclass(1, 0, 0),
+		host1x_opcode_setpayload(0),
+		host1x_opcode_setstreamid(0x1fffff));
+
+	for (i = 0; i < job->num_cmds; i++) {
+		struct host1x_job_cmd *cmd = &job->cmds[i];
+
+		if (!cmd->is_wait || cmd->wait.relative)
+			break;
+
+		submit_wait(job, cmd->wait.id, cmd->wait.threshold);
+	}
+
+	host1x_cdma_push(cdma,
+		HOST1X_OPCODE_NOP,
+		host1x_opcode_release_mlock(1));
+
+prefences_done:
+	/* Enter engine class with invalid stream ID. */
+	host1x_cdma_push_wide(cdma,
+		host1x_opcode_acquire_mlock(job->class),
+		host1x_opcode_setclass(job->class, 0, 0),
+		host1x_opcode_setpayload(0),
+		host1x_opcode_setstreamid(job->engine_streamid_offset / 4));
+
+	/* Before switching stream ID to real stream ID, ensure engine is idle. */
+	fence = host1x_syncpt_incr_max(sp, 1);
+	host1x_cdma_push(&job->channel->cdma,
+		host1x_opcode_nonincr(HOST1X_UCLASS_INCR_SYNCPT, 1),
+		HOST1X_UCLASS_INCR_SYNCPT_INDX_F(job->syncpt->id) |
+			HOST1X_UCLASS_INCR_SYNCPT_COND_F(4));
+	submit_wait(job, job->syncpt->id, fence);
+	submit_setclass(job, job->class);
+
+	/* Submit work. */
+	job->syncpt_end = host1x_syncpt_incr_max(sp, job->syncpt_incrs);
+	submit_gathers(job, job->cmds + i, job->num_cmds - i,
+		       wrapping_sub(u32, job->syncpt_end, job->syncpt_incrs));
+
+	/* Before releasing MLOCK, ensure engine is idle again. */
+	fence = host1x_syncpt_incr_max(sp, 1);
+	host1x_cdma_push(&job->channel->cdma,
+		host1x_opcode_nonincr(HOST1X_UCLASS_INCR_SYNCPT, 1),
+		HOST1X_UCLASS_INCR_SYNCPT_INDX_F(job->syncpt->id) |
+			HOST1X_UCLASS_INCR_SYNCPT_COND_F(4));
+	submit_wait(job, job->syncpt->id, fence);
+
+	/* Release MLOCK. */
+	host1x_cdma_push(cdma,
+		HOST1X_OPCODE_NOP, host1x_opcode_release_mlock(job->class));
+#else
+	if (job->serialize) {
+		/*
+		 * Force serialization by inserting a host wait for the
+		 * previous job to finish before this one can commence.
+		 */
+		host1x_cdma_push(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)));
+	}
+
+	/* Synchronize base register to allow using it for relative waiting */
+	if (sp->base)
+		synchronize_syncpt_base(job);
+
+	/* add a setclass for modules that require it */
+	if (job->class)
+		host1x_cdma_push(cdma,
+				 host1x_opcode_setclass(job->class, 0, 0),
+				 HOST1X_OPCODE_NOP);
+
+	job->syncpt_end = host1x_syncpt_incr_max(sp, job->syncpt_incrs);
+
+	submit_gathers(job, job->cmds, job->num_cmds,
+		       wrapping_sub(u32, job->syncpt_end, job->syncpt_incrs));
+#endif
+}
+
+static void job_complete_callback(struct dma_fence *fence, struct dma_fence_cb *cb)
+{
+	struct host1x_job *job = container_of(cb, struct host1x_job, fence_cb);
+
+	/* Schedules CDMA update. */
+	host1x_cdma_update(&job->channel->cdma);
+}
+
+static int channel_submit(struct host1x_job *job)
+{
+	struct host1x_channel *ch = job->channel;
+	struct host1x_syncpt *sp = job->syncpt;
+	u32 prev_max = 0;
+	u32 syncval;
+	int err;
+	struct host1x *host = dev_get_drvdata(ch->dev->parent);
+
+	trace_host1x_channel_submit(dev_name(ch->dev),
+				    job->num_cmds, job->num_relocs,
+				    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)
+		return err;
+
+	host1x_channel_set_streamid(ch);
+	host1x_enable_gather_filter(ch);
+	host1x_hw_syncpt_assign_to_channel(host, sp, ch);
+
+	/* begin a CDMA submit */
+	err = host1x_cdma_begin(&ch->cdma, job);
+	if (err) {
+		mutex_unlock(&ch->submitlock);
+		return err;
+	}
+
+	channel_program_cdma(job);
+	syncval = host1x_syncpt_read_max(sp);
+
+	/*
+	 * Create fence before submitting job to HW to avoid job completing
+	 * before the fence is set up.
+	 */
+	job->fence = host1x_fence_create(sp, syncval, true);
+	if (WARN(IS_ERR(job->fence), "Failed to create submit complete fence")) {
+		job->fence = NULL;
+	} else {
+		err = dma_fence_add_callback(job->fence, &job->fence_cb,
+					     job_complete_callback);
+	}
+
+	/* 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);
+
+	mutex_unlock(&ch->submitlock);
+
+	if (err == -ENOENT)
+		host1x_cdma_update(&ch->cdma);
+	else
+		WARN(err, "Failed to set submit complete interrupt");
+
+	return 0;
+}
+
+static int host1x_channel_init(struct host1x_channel *ch, struct host1x *dev,
+			       unsigned int index)
+{
+#if HOST1X_HW < 6
+	ch->regs = dev->regs + index * 0x4000;
+#else
+	ch->regs = dev->regs + index * 0x100;
+#endif
+	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 000000000..4c32aa1b9
--- /dev/null
+++ b/drivers/gpu/host1x/hw/debug_hw.c
@@ -0,0 +1,258 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2010 Google, Inc.
+ * Author: Erik Gilling <konkers@android.com>
+ *
+ * Copyright (C) 2011-2013 NVIDIA Corporation
+ */
+
+#include "../dev.h"
+#include "../debug.h"
+#include "../cdma.h"
+#include "../channel.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_SETSTRMID = 0x07,
+	HOST1X_OPCODE_SETAPPID  = 0x08,
+	HOST1X_OPCODE_SETPYLD   = 0x09,
+	HOST1X_OPCODE_INCR_W    = 0x0a,
+	HOST1X_OPCODE_NONINCR_W = 0x0b,
+	HOST1X_OPCODE_GATHER_W  = 0x0c,
+	HOST1X_OPCODE_RESTART_W = 0x0d,
+	HOST1X_OPCODE_EXTEND	= 0x0e,
+};
+
+enum {
+	HOST1X_OPCODE_EXTEND_ACQUIRE_MLOCK	= 0x00,
+	HOST1X_OPCODE_EXTEND_RELEASE_MLOCK	= 0x01,
+};
+
+#define INVALID_PAYLOAD				0xffffffff
+
+static unsigned int show_channel_command(struct output *o, u32 val,
+					 u32 *payload)
+{
+	unsigned int mask, subop, num, opcode;
+
+	opcode = val >> 28;
+
+	switch (opcode) {
+	case HOST1X_OPCODE_SETCLASS:
+		mask = val & 0x3f;
+		if (mask) {
+			host1x_debug_cont(o, "SETCL(class=%03x, offset=%03x, mask=%02x, [",
+					    val >> 6 & 0x3ff,
+					    val >> 16 & 0xfff, mask);
+			return hweight8(mask);
+		}
+
+		host1x_debug_cont(o, "SETCL(class=%03x)\n", val >> 6 & 0x3ff);
+		return 0;
+
+	case HOST1X_OPCODE_INCR:
+		num = val & 0xffff;
+		host1x_debug_cont(o, "INCR(offset=%03x, [",
+				    val >> 16 & 0xfff);
+		if (!num)
+			host1x_debug_cont(o, "])\n");
+
+		return num;
+
+	case HOST1X_OPCODE_NONINCR:
+		num = val & 0xffff;
+		host1x_debug_cont(o, "NONINCR(offset=%03x, [",
+				    val >> 16 & 0xfff);
+		if (!num)
+			host1x_debug_cont(o, "])\n");
+
+		return num;
+
+	case HOST1X_OPCODE_MASK:
+		mask = val & 0xffff;
+		host1x_debug_cont(o, "MASK(offset=%03x, mask=%03x, [",
+				    val >> 16 & 0xfff, mask);
+		if (!mask)
+			host1x_debug_cont(o, "])\n");
+
+		return hweight16(mask);
+
+	case HOST1X_OPCODE_IMM:
+		host1x_debug_cont(o, "IMM(offset=%03x, data=%03x)\n",
+				    val >> 16 & 0xfff, val & 0xffff);
+		return 0;
+
+	case HOST1X_OPCODE_RESTART:
+		host1x_debug_cont(o, "RESTART(offset=%08x)\n", val << 4);
+		return 0;
+
+	case HOST1X_OPCODE_GATHER:
+		host1x_debug_cont(o, "GATHER(offset=%03x, insert=%d, type=%d, count=%04x, addr=[",
+				    val >> 16 & 0xfff, val >> 15 & 0x1,
+				    val >> 14 & 0x1, val & 0x3fff);
+		return 1;
+
+#if HOST1X_HW >= 6
+	case HOST1X_OPCODE_SETSTRMID:
+		host1x_debug_cont(o, "SETSTRMID(offset=%06x)\n",
+				  val & 0x3fffff);
+		return 0;
+
+	case HOST1X_OPCODE_SETAPPID:
+		host1x_debug_cont(o, "SETAPPID(appid=%02x)\n", val & 0xff);
+		return 0;
+
+	case HOST1X_OPCODE_SETPYLD:
+		*payload = val & 0xffff;
+		host1x_debug_cont(o, "SETPYLD(data=%04x)\n", *payload);
+		return 0;
+
+	case HOST1X_OPCODE_INCR_W:
+	case HOST1X_OPCODE_NONINCR_W:
+		host1x_debug_cont(o, "%s(offset=%06x, ",
+				  opcode == HOST1X_OPCODE_INCR_W ?
+					"INCR_W" : "NONINCR_W",
+				  val & 0x3fffff);
+		if (*payload == 0) {
+			host1x_debug_cont(o, "[])\n");
+			return 0;
+		} else if (*payload == INVALID_PAYLOAD) {
+			host1x_debug_cont(o, "unknown)\n");
+			return 0;
+		} else {
+			host1x_debug_cont(o, "[");
+			return *payload;
+		}
+
+	case HOST1X_OPCODE_GATHER_W:
+		host1x_debug_cont(o, "GATHER_W(count=%04x, addr=[",
+				  val & 0x3fff);
+		return 2;
+#endif
+
+	case HOST1X_OPCODE_EXTEND:
+		subop = val >> 24 & 0xf;
+		if (subop == HOST1X_OPCODE_EXTEND_ACQUIRE_MLOCK)
+			host1x_debug_cont(o, "ACQUIRE_MLOCK(index=%d)\n",
+					    val & 0xff);
+		else if (subop == HOST1X_OPCODE_EXTEND_RELEASE_MLOCK)
+			host1x_debug_cont(o, "RELEASE_MLOCK(index=%d)\n",
+					    val & 0xff);
+		else
+			host1x_debug_cont(o, "EXTEND_UNKNOWN(%08x)\n", val);
+		return 0;
+
+	default:
+		host1x_debug_cont(o, "UNKNOWN\n");
+		return 0;
+	}
+}
+
+static void show_gather(struct output *o, dma_addr_t phys_addr,
+			unsigned int words, struct host1x_cdma *cdma,
+			dma_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;
+	u32 payload = INVALID_PAYLOAD;
+
+	/*
+	 * 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++) {
+		dma_addr_t addr = phys_addr + i * 4;
+		u32 voffset = offset + i * 4;
+		u32 val;
+
+		/* If we reach the RESTART opcode, continue at the beginning of pushbuffer */
+		if (cdma && voffset >= cdma->push_buffer.size) {
+			addr -= cdma->push_buffer.size;
+			voffset -= cdma->push_buffer.size;
+		}
+
+		val = *(map_addr + voffset / 4);
+
+		if (!data_count) {
+			host1x_debug_output(o, "    %pad: %08x: ", &addr, val);
+			data_count = show_channel_command(o, val, &payload);
+		} else {
+			host1x_debug_cont(o, "%08x%s", val,
+					    data_count > 1 ? ", " : "])\n");
+			data_count--;
+		}
+	}
+}
+
+static void show_channel_gathers(struct output *o, struct host1x_cdma *cdma)
+{
+	struct push_buffer *pb = &cdma->push_buffer;
+	struct host1x_job *job;
+
+	list_for_each_entry(job, &cdma->sync_queue, list) {
+		unsigned int i;
+
+		host1x_debug_output(o, "JOB, syncpt %u: %u timeout: %u num_slots: %u num_handles: %u\n",
+				    job->syncpt->id, job->syncpt_end, job->timeout,
+				    job->num_slots, job->num_unpins);
+
+		show_gather(o, pb->dma + job->first_get, job->num_slots * 2, cdma,
+			    pb->dma, pb->mapped);
+
+		for (i = 0; i < job->num_cmds; i++) {
+			struct host1x_job_gather *g;
+			u32 *mapped;
+
+			if (job->cmds[i].is_wait)
+				continue;
+
+			g = &job->cmds[i].gather;
+
+			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 %pad+%#x, %d words\n",
+					    &g->base, g->offset, g->words);
+
+			show_gather(o, g->base + g->offset, g->words, NULL,
+				    g->base, mapped);
+
+			if (!job->gather_copy_mapped)
+				host1x_bo_munmap(g->bo, mapped);
+		}
+	}
+}
+
+#if HOST1X_HW >= 6
+#include "debug_hw_1x06.c"
+#else
+#include "debug_hw_1x01.c"
+#endif
+
+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/debug_hw_1x01.c b/drivers/gpu/host1x/hw/debug_hw_1x01.c
new file mode 100644
index 000000000..85242a59f
--- /dev/null
+++ b/drivers/gpu/host1x/hw/debug_hw_1x01.c
@@ -0,0 +1,149 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2010 Google, Inc.
+ * Author: Erik Gilling <konkers@android.com>
+ *
+ * Copyright (C) 2011-2013 NVIDIA Corporation
+ */
+
+#include "../dev.h"
+#include "../debug.h"
+#include "../cdma.h"
+#include "../channel.h"
+
+static void host1x_debug_show_channel_cdma(struct host1x *host,
+					   struct host1x_channel *ch,
+					   struct output *o)
+{
+	struct host1x_cdma *cdma = &ch->cdma;
+	dma_addr_t dmastart, dmaend;
+	u32 dmaput, dmaget, dmactrl;
+	u32 cbstat, cbread;
+	u32 val, base, baseval;
+
+	dmastart = host1x_ch_readl(ch, HOST1X_CHANNEL_DMASTART);
+	dmaend = host1x_ch_readl(ch, HOST1X_CHANNEL_DMAEND);
+	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, "%u-%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, "DMASTART %pad, DMAEND %pad\n", &dmastart, &dmaend);
+	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, "%u: 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, NULL);
+		} else {
+			host1x_debug_cont(o, "%08x%s", val,
+					  data_count > 1 ? ", " : "])\n");
+			data_count--;
+		}
+
+		if (rd_ptr == end)
+			rd_ptr = start;
+		else
+			rd_ptr++;
+	} while (rd_ptr != wr_ptr);
+
+	if (data_count)
+		host1x_debug_cont(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)
+{
+	unsigned 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, "%u: locked by channel %u\n",
+				i, HOST1X_SYNC_MLOCK_OWNER_CHID_V(owner));
+		else if (HOST1X_SYNC_MLOCK_OWNER_CPU_OWNS_V(owner))
+			host1x_debug_output(o, "%u: locked by cpu\n", i);
+		else
+			host1x_debug_output(o, "%u: unlocked\n", i);
+	}
+
+	host1x_debug_output(o, "\n");
+}
diff --git a/drivers/gpu/host1x/hw/debug_hw_1x06.c b/drivers/gpu/host1x/hw/debug_hw_1x06.c
new file mode 100644
index 000000000..9d0667879
--- /dev/null
+++ b/drivers/gpu/host1x/hw/debug_hw_1x06.c
@@ -0,0 +1,145 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2010 Google, Inc.
+ * Author: Erik Gilling <konkers@android.com>
+ *
+ * Copyright (C) 2011-2017 NVIDIA Corporation
+ */
+
+#include "../dev.h"
+#include "../debug.h"
+#include "../cdma.h"
+#include "../channel.h"
+
+static void host1x_debug_show_channel_cdma(struct host1x *host,
+					   struct host1x_channel *ch,
+					   struct output *o)
+{
+	struct host1x_cdma *cdma = &ch->cdma;
+	dma_addr_t dmastart = 0, dmaend = 0;
+	u32 dmaput, dmaget, dmactrl;
+	u32 offset, class;
+	u32 ch_stat;
+
+#if defined(CONFIG_ARCH_DMA_ADDR_T_64BIT) && HOST1X_HW >= 6
+	dmastart = host1x_ch_readl(ch, HOST1X_CHANNEL_DMASTART_HI);
+	dmastart <<= 32;
+#endif
+	dmastart |= host1x_ch_readl(ch, HOST1X_CHANNEL_DMASTART);
+
+#if defined(CONFIG_ARCH_DMA_ADDR_T_64BIT) && HOST1X_HW >= 6
+	dmaend = host1x_ch_readl(ch, HOST1X_CHANNEL_DMAEND_HI);
+	dmaend <<= 32;
+#endif
+	dmaend |= host1x_ch_readl(ch, HOST1X_CHANNEL_DMAEND);
+
+	dmaput = host1x_ch_readl(ch, HOST1X_CHANNEL_DMAPUT);
+	dmaget = host1x_ch_readl(ch, HOST1X_CHANNEL_DMAGET);
+	dmactrl = host1x_ch_readl(ch, HOST1X_CHANNEL_DMACTRL);
+	offset = host1x_ch_readl(ch, HOST1X_CHANNEL_CMDP_OFFSET);
+	class = host1x_ch_readl(ch, HOST1X_CHANNEL_CMDP_CLASS);
+	ch_stat = host1x_ch_readl(ch, HOST1X_CHANNEL_CHANNELSTAT);
+
+	host1x_debug_output(o, "%u-%s: ", ch->id, dev_name(ch->dev));
+
+	if (dmactrl & HOST1X_CHANNEL_DMACTRL_DMASTOP ||
+	    !ch->cdma.push_buffer.mapped) {
+		host1x_debug_output(o, "inactive\n\n");
+		return;
+	}
+
+	if (class == HOST1X_CLASS_HOST1X && offset == HOST1X_UCLASS_WAIT_SYNCPT)
+		host1x_debug_output(o, "waiting on syncpt\n");
+	else
+		host1x_debug_output(o, "active class %02x, offset %04x\n",
+				    class, offset);
+
+	host1x_debug_output(o, "DMASTART %pad, DMAEND %pad\n", &dmastart, &dmaend);
+	host1x_debug_output(o, "DMAPUT %08x DMAGET %08x DMACTL %08x\n",
+			    dmaput, dmaget, dmactrl);
+	host1x_debug_output(o, "CHANNELSTAT %02x\n", ch_stat);
+
+	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)
+{
+#if HOST1X_HW <= 6
+	u32 rd_ptr, wr_ptr, start, end;
+	u32 payload = INVALID_PAYLOAD;
+	unsigned int data_count = 0;
+#endif
+	u32 val;
+
+	host1x_debug_output(o, "%u: fifo:\n", ch->id);
+
+	val = host1x_ch_readl(ch, HOST1X_CHANNEL_CMDFIFO_STAT);
+	host1x_debug_output(o, "CMDFIFO_STAT %08x\n", val);
+	if (val & HOST1X_CHANNEL_CMDFIFO_STAT_EMPTY) {
+		host1x_debug_output(o, "[empty]\n");
+		return;
+	}
+
+	val = host1x_ch_readl(ch, HOST1X_CHANNEL_CMDFIFO_RDATA);
+	host1x_debug_output(o, "CMDFIFO_RDATA %08x\n", val);
+
+#if HOST1X_HW <= 6
+	/* Peek pointer values are invalid during SLCG, so disable it */
+	host1x_hypervisor_writel(host, 0x1, HOST1X_HV_ICG_EN_OVERRIDE);
+
+	val = 0;
+	val |= HOST1X_HV_CMDFIFO_PEEK_CTRL_ENABLE;
+	val |= HOST1X_HV_CMDFIFO_PEEK_CTRL_CHANNEL(ch->id);
+	host1x_hypervisor_writel(host, val, HOST1X_HV_CMDFIFO_PEEK_CTRL);
+
+	val = host1x_hypervisor_readl(host, HOST1X_HV_CMDFIFO_PEEK_PTRS);
+	rd_ptr = HOST1X_HV_CMDFIFO_PEEK_PTRS_RD_PTR_V(val);
+	wr_ptr = HOST1X_HV_CMDFIFO_PEEK_PTRS_WR_PTR_V(val);
+
+	val = host1x_hypervisor_readl(host, HOST1X_HV_CMDFIFO_SETUP(ch->id));
+	start = HOST1X_HV_CMDFIFO_SETUP_BASE_V(val);
+	end = HOST1X_HV_CMDFIFO_SETUP_LIMIT_V(val);
+
+	do {
+		val = 0;
+		val |= HOST1X_HV_CMDFIFO_PEEK_CTRL_ENABLE;
+		val |= HOST1X_HV_CMDFIFO_PEEK_CTRL_CHANNEL(ch->id);
+		val |= HOST1X_HV_CMDFIFO_PEEK_CTRL_ADDR(rd_ptr);
+		host1x_hypervisor_writel(host, val,
+					 HOST1X_HV_CMDFIFO_PEEK_CTRL);
+
+		val = host1x_hypervisor_readl(host,
+					      HOST1X_HV_CMDFIFO_PEEK_READ);
+
+		if (!data_count) {
+			host1x_debug_output(o, "%03x 0x%08x: ",
+					    rd_ptr - start, val);
+			data_count = show_channel_command(o, val, &payload);
+		} else {
+			host1x_debug_cont(o, "%08x%s", val,
+					  data_count > 1 ? ", " : "])\n");
+			data_count--;
+		}
+
+		if (rd_ptr == end)
+			rd_ptr = start;
+		else
+			rd_ptr++;
+	} while (rd_ptr != wr_ptr);
+
+	if (data_count)
+		host1x_debug_cont(o, ", ...])\n");
+	host1x_debug_output(o, "\n");
+
+	host1x_hypervisor_writel(host, 0x0, HOST1X_HV_CMDFIFO_PEEK_CTRL);
+	host1x_hypervisor_writel(host, 0x0, HOST1X_HV_ICG_EN_OVERRIDE);
+#endif
+}
+
+static void host1x_debug_show_mlocks(struct host1x *host, struct output *o)
+{
+	/* TODO */
+}
diff --git a/drivers/gpu/host1x/hw/host1x01.c b/drivers/gpu/host1x/hw/host1x01.c
new file mode 100644
index 000000000..8d8a117a5
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x01.c
@@ -0,0 +1,33 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Host1x init for T20 and T30 Architecture Chips
+ *
+ * Copyright (c) 2011-2013, NVIDIA Corporation.
+ */
+
+/* include hw specification */
+#include "host1x01.h"
+#include "host1x01_hardware.h"
+
+/* include code */
+#define HOST1X_HW 1
+
+#include "cdma_hw.c"
+#include "channel_hw.c"
+#include "debug_hw.c"
+#include "intr_hw.c"
+#include "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 000000000..6516c3f1e
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x01.h
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Host1x init for T20 and T30 Architecture Chips
+ *
+ * Copyright (c) 2011-2013, NVIDIA Corporation.
+ */
+#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 000000000..cb93d7c18
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x01_hardware.h
@@ -0,0 +1,20 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Tegra host1x Register Offsets for Tegra20 and Tegra30
+ *
+ * Copyright (c) 2010-2013 NVIDIA Corporation.
+ */
+
+#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"
+
+#include "opcodes.h"
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x02.c b/drivers/gpu/host1x/hw/host1x02.c
new file mode 100644
index 000000000..583b33c04
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x02.c
@@ -0,0 +1,33 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Host1x init for Tegra114 SoCs
+ *
+ * Copyright (c) 2013 NVIDIA Corporation.
+ */
+
+/* include hw specification */
+#include "host1x02.h"
+#include "host1x02_hardware.h"
+
+/* include code */
+#define HOST1X_HW 2
+
+#include "cdma_hw.c"
+#include "channel_hw.c"
+#include "debug_hw.c"
+#include "intr_hw.c"
+#include "syncpt_hw.c"
+
+#include "../dev.h"
+
+int host1x02_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/host1x02.h b/drivers/gpu/host1x/hw/host1x02.h
new file mode 100644
index 000000000..7e5c3e470
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x02.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Host1x init for Tegra114 SoCs
+ *
+ * Copyright (c) 2013 NVIDIA Corporation.
+ */
+
+#ifndef HOST1X_HOST1X02_H
+#define HOST1X_HOST1X02_H
+
+struct host1x;
+
+int host1x02_init(struct host1x *host);
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x02_hardware.h b/drivers/gpu/host1x/hw/host1x02_hardware.h
new file mode 100644
index 000000000..2d1282b9b
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x02_hardware.h
@@ -0,0 +1,20 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Tegra host1x Register Offsets for Tegra114
+ *
+ * Copyright (c) 2010-2013 NVIDIA Corporation.
+ */
+
+#ifndef __HOST1X_HOST1X02_HARDWARE_H
+#define __HOST1X_HOST1X02_HARDWARE_H
+
+#include <linux/types.h>
+#include <linux/bitops.h>
+
+#include "hw_host1x02_channel.h"
+#include "hw_host1x02_sync.h"
+#include "hw_host1x02_uclass.h"
+
+#include "opcodes.h"
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x04.c b/drivers/gpu/host1x/hw/host1x04.c
new file mode 100644
index 000000000..26b459eb2
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x04.c
@@ -0,0 +1,33 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Host1x init for Tegra124 SoCs
+ *
+ * Copyright (c) 2013 NVIDIA Corporation.
+ */
+
+/* include hw specification */
+#include "host1x04.h"
+#include "host1x04_hardware.h"
+
+/* include code */
+#define HOST1X_HW 4
+
+#include "cdma_hw.c"
+#include "channel_hw.c"
+#include "debug_hw.c"
+#include "intr_hw.c"
+#include "syncpt_hw.c"
+
+#include "../dev.h"
+
+int host1x04_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/host1x04.h b/drivers/gpu/host1x/hw/host1x04.h
new file mode 100644
index 000000000..2a1e8153c
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x04.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Host1x init for Tegra124 SoCs
+ *
+ * Copyright (c) 2013 NVIDIA Corporation.
+ */
+
+#ifndef HOST1X_HOST1X04_H
+#define HOST1X_HOST1X04_H
+
+struct host1x;
+
+int host1x04_init(struct host1x *host);
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x04_hardware.h b/drivers/gpu/host1x/hw/host1x04_hardware.h
new file mode 100644
index 000000000..84d244e8a
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x04_hardware.h
@@ -0,0 +1,20 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Tegra host1x Register Offsets for Tegra124
+ *
+ * Copyright (c) 2010-2013 NVIDIA Corporation.
+ */
+
+#ifndef __HOST1X_HOST1X04_HARDWARE_H
+#define __HOST1X_HOST1X04_HARDWARE_H
+
+#include <linux/types.h>
+#include <linux/bitops.h>
+
+#include "hw_host1x04_channel.h"
+#include "hw_host1x04_sync.h"
+#include "hw_host1x04_uclass.h"
+
+#include "opcodes.h"
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x05.c b/drivers/gpu/host1x/hw/host1x05.c
new file mode 100644
index 000000000..6d9803343
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x05.c
@@ -0,0 +1,33 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Host1x init for Tegra210 SoCs
+ *
+ * Copyright (c) 2015 NVIDIA Corporation.
+ */
+
+/* include hw specification */
+#include "host1x05.h"
+#include "host1x05_hardware.h"
+
+/* include code */
+#define HOST1X_HW 5
+
+#include "cdma_hw.c"
+#include "channel_hw.c"
+#include "debug_hw.c"
+#include "intr_hw.c"
+#include "syncpt_hw.c"
+
+#include "../dev.h"
+
+int host1x05_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/host1x05.h b/drivers/gpu/host1x/hw/host1x05.h
new file mode 100644
index 000000000..addfd41e7
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x05.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Host1x init for Tegra210 SoCs
+ *
+ * Copyright (c) 2015 NVIDIA Corporation.
+ */
+
+#ifndef HOST1X_HOST1X05_H
+#define HOST1X_HOST1X05_H
+
+struct host1x;
+
+int host1x05_init(struct host1x *host);
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x05_hardware.h b/drivers/gpu/host1x/hw/host1x05_hardware.h
new file mode 100644
index 000000000..1dcde6ec7
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x05_hardware.h
@@ -0,0 +1,20 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Tegra host1x Register Offsets for Tegra210
+ *
+ * Copyright (c) 2015 NVIDIA Corporation.
+ */
+
+#ifndef __HOST1X_HOST1X05_HARDWARE_H
+#define __HOST1X_HOST1X05_HARDWARE_H
+
+#include <linux/types.h>
+#include <linux/bitops.h>
+
+#include "hw_host1x05_channel.h"
+#include "hw_host1x05_sync.h"
+#include "hw_host1x05_uclass.h"
+
+#include "opcodes.h"
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x06.c b/drivers/gpu/host1x/hw/host1x06.c
new file mode 100644
index 000000000..844f81ae2
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x06.c
@@ -0,0 +1,33 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Host1x init for Tegra186 SoCs
+ *
+ * Copyright (c) 2017 NVIDIA Corporation.
+ */
+
+/* include hw specification */
+#include "host1x06.h"
+#include "host1x06_hardware.h"
+
+/* include code */
+#define HOST1X_HW 6
+
+#include "cdma_hw.c"
+#include "channel_hw.c"
+#include "debug_hw.c"
+#include "intr_hw.c"
+#include "syncpt_hw.c"
+
+#include "../dev.h"
+
+int host1x06_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/host1x06.h b/drivers/gpu/host1x/hw/host1x06.h
new file mode 100644
index 000000000..4ea756895
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x06.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Host1x init for Tegra186 SoCs
+ *
+ * Copyright (c) 2017 NVIDIA Corporation.
+ */
+
+#ifndef HOST1X_HOST1X06_H
+#define HOST1X_HOST1X06_H
+
+struct host1x;
+
+int host1x06_init(struct host1x *host);
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x06_hardware.h b/drivers/gpu/host1x/hw/host1x06_hardware.h
new file mode 100644
index 000000000..c05cfa7e3
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x06_hardware.h
@@ -0,0 +1,21 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Tegra host1x Register Offsets for Tegra186
+ *
+ * Copyright (c) 2017 NVIDIA Corporation.
+ */
+
+#ifndef __HOST1X_HOST1X06_HARDWARE_H
+#define __HOST1X_HOST1X06_HARDWARE_H
+
+#include <linux/types.h>
+#include <linux/bitops.h>
+
+#include "hw_host1x06_channel.h"
+#include "hw_host1x06_uclass.h"
+#include "hw_host1x06_vm.h"
+#include "hw_host1x06_hypervisor.h"
+
+#include "opcodes.h"
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x07.c b/drivers/gpu/host1x/hw/host1x07.c
new file mode 100644
index 000000000..0c6f14f7e
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x07.c
@@ -0,0 +1,33 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Host1x init for Tegra194 SoCs
+ *
+ * Copyright (c) 2018 NVIDIA Corporation.
+ */
+
+/* include hw specification */
+#include "host1x07.h"
+#include "host1x07_hardware.h"
+
+/* include code */
+#define HOST1X_HW 7
+
+#include "cdma_hw.c"
+#include "channel_hw.c"
+#include "debug_hw.c"
+#include "intr_hw.c"
+#include "syncpt_hw.c"
+
+#include "../dev.h"
+
+int host1x07_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/host1x07.h b/drivers/gpu/host1x/hw/host1x07.h
new file mode 100644
index 000000000..419b6eaad
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x07.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Host1x init for Tegra194 SoCs
+ *
+ * Copyright (c) 2018 NVIDIA Corporation.
+ */
+
+#ifndef HOST1X_HOST1X07_H
+#define HOST1X_HOST1X07_H
+
+struct host1x;
+
+int host1x07_init(struct host1x *host);
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x07_hardware.h b/drivers/gpu/host1x/hw/host1x07_hardware.h
new file mode 100644
index 000000000..d67364e03
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x07_hardware.h
@@ -0,0 +1,21 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Tegra host1x Register Offsets for Tegra194
+ *
+ * Copyright (c) 2018 NVIDIA Corporation.
+ */
+
+#ifndef __HOST1X_HOST1X07_HARDWARE_H
+#define __HOST1X_HOST1X07_HARDWARE_H
+
+#include <linux/types.h>
+#include <linux/bitops.h>
+
+#include "hw_host1x07_channel.h"
+#include "hw_host1x07_uclass.h"
+#include "hw_host1x07_vm.h"
+#include "hw_host1x07_hypervisor.h"
+
+#include "opcodes.h"
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x08.c b/drivers/gpu/host1x/hw/host1x08.c
new file mode 100644
index 000000000..754890c34
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x08.c
@@ -0,0 +1,33 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Host1x init for Tegra234 SoCs
+ *
+ * Copyright (c) 2022 NVIDIA Corporation.
+ */
+
+/* include hw specification */
+#include "host1x08.h"
+#include "host1x08_hardware.h"
+
+/* include code */
+#define HOST1X_HW 8
+
+#include "cdma_hw.c"
+#include "channel_hw.c"
+#include "debug_hw.c"
+#include "intr_hw.c"
+#include "syncpt_hw.c"
+
+#include "../dev.h"
+
+int host1x08_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/host1x08.h b/drivers/gpu/host1x/hw/host1x08.h
new file mode 100644
index 000000000..a6bad56e4
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x08.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Host1x init for Tegra234 SoCs
+ *
+ * Copyright (c) 2018 NVIDIA Corporation.
+ */
+
+#ifndef HOST1X_HOST1X08_H
+#define HOST1X_HOST1X08_H
+
+struct host1x;
+
+int host1x08_init(struct host1x *host);
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x08_hardware.h b/drivers/gpu/host1x/hw/host1x08_hardware.h
new file mode 100644
index 000000000..936243060
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x08_hardware.h
@@ -0,0 +1,21 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Tegra host1x Register Offsets for Tegra234
+ *
+ * Copyright (c) 2022 NVIDIA Corporation.
+ */
+
+#ifndef __HOST1X_HOST1X08_HARDWARE_H
+#define __HOST1X_HOST1X08_HARDWARE_H
+
+#include <linux/types.h>
+#include <linux/bitops.h>
+
+#include "hw_host1x08_uclass.h"
+#include "hw_host1x08_vm.h"
+#include "hw_host1x08_hypervisor.h"
+#include "hw_host1x08_common.h"
+
+#include "opcodes.h"
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x10.c b/drivers/gpu/host1x/hw/host1x10.c
new file mode 100644
index 000000000..2800f309b
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x10.c
@@ -0,0 +1,33 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Host1x init for Tegra264 SoCs
+ *
+ * Copyright (c) 2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+ */
+
+/* include hw specification */
+#include "host1x10.h"
+#include "host1x10_hardware.h"
+
+/* include code */
+#define HOST1X_HW 10
+
+#include "cdma_hw.c"
+#include "channel_hw.c"
+#include "debug_hw.c"
+#include "intr_hw.c"
+#include "syncpt_hw.c"
+
+#include "../dev.h"
+
+int host1x10_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/host1x10.h b/drivers/gpu/host1x/hw/host1x10.h
new file mode 100644
index 000000000..577f6ff3d
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x10.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Host1x init for Tegra264 SoCs
+ *
+ * Copyright (c) 2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+ */
+
+#ifndef HOST1X_HOST1X10_H
+#define HOST1X_HOST1X10_H
+
+struct host1x;
+
+int host1x10_init(struct host1x *host);
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x10_hardware.h b/drivers/gpu/host1x/hw/host1x10_hardware.h
new file mode 100644
index 000000000..abbead819
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x10_hardware.h
@@ -0,0 +1,21 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Tegra host1x Register Offsets for Tegra264
+ *
+ * Copyright (c) 2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+ */
+
+#ifndef __HOST1X_HOST1X10_HARDWARE_H
+#define __HOST1X_HOST1X10_HARDWARE_H
+
+#include <linux/types.h>
+#include <linux/bitops.h>
+
+#include "hw_host1x10_uclass.h"
+#include "hw_host1x10_vm.h"
+#include "hw_host1x10_hypervisor.h"
+#include "hw_host1x10_common.h"
+
+#include "opcodes.h"
+
+#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 000000000..8da43eaba
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x01_channel.h
@@ -0,0 +1,108 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2012-2013, NVIDIA Corporation.
+ */
+
+ /*
+  * 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 000000000..ec95e7ae7
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x01_sync.h
@@ -0,0 +1,231 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2012-2013, NVIDIA Corporation.
+ */
+
+ /*
+  * 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_v(u32 v)
+{
+	return (v >> 8) & 0xf;
+}
+#define HOST1X_SYNC_MLOCK_OWNER_CHID_V(v) \
+	host1x_sync_mlock_owner_chid_v(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 000000000..1239bfd46
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x01_uclass.h
@@ -0,0 +1,168 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2012-2013, NVIDIA Corporation.
+ */
+
+ /*
+  * 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_r(void)
+{
+	return 0xb;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE \
+	host1x_uclass_load_syncpt_base_r()
+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/hw_host1x02_channel.h b/drivers/gpu/host1x/hw/hw_host1x02_channel.h
new file mode 100644
index 000000000..210d317ad
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x02_channel.h
@@ -0,0 +1,109 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2013 NVIDIA Corporation.
+ */
+
+ /*
+  * 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 HOST1X_HW_HOST1X02_CHANNEL_H
+#define HOST1X_HW_HOST1X02_CHANNEL_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 >> 11) & 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_host1x02_sync.h b/drivers/gpu/host1x/hw/hw_host1x02_sync.h
new file mode 100644
index 000000000..44b4f8379
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x02_sync.h
@@ -0,0 +1,231 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2013 NVIDIA Corporation.
+ */
+
+ /*
+  * 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 HOST1X_HW_HOST1X02_SYNC_H
+#define HOST1X_HW_HOST1X02_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) & 0x3ff;
+}
+#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) & 0x3ff;
+}
+#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_v(u32 v)
+{
+	return (v >> 8) & 0xf;
+}
+#define HOST1X_SYNC_MLOCK_OWNER_CHID_V(v) \
+	host1x_sync_mlock_owner_chid_v(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 & 0x3ff) << 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 & 0xf) << 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) & 0x3ff;
+}
+#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) & 0x3ff;
+}
+#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
diff --git a/drivers/gpu/host1x/hw/hw_host1x02_uclass.h b/drivers/gpu/host1x/hw/hw_host1x02_uclass.h
new file mode 100644
index 000000000..0a2ab8f1d
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x02_uclass.h
@@ -0,0 +1,181 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2013 NVIDIA Corporation.
+ */
+
+ /*
+  * 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 HOST1X_HW_HOST1X02_UCLASS_H
+#define HOST1X_HW_HOST1X02_UCLASS_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_r(void)
+{
+	return 0xb;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE \
+	host1x_uclass_load_syncpt_base_r()
+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)
+static inline u32 host1x_uclass_load_syncpt_payload_32_r(void)
+{
+	return 0x4e;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_PAYLOAD_32 \
+	host1x_uclass_load_syncpt_payload_32_r()
+static inline u32 host1x_uclass_wait_syncpt_32_r(void)
+{
+	return 0x50;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_32 \
+	host1x_uclass_wait_syncpt_32_r()
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x04_channel.h b/drivers/gpu/host1x/hw/hw_host1x04_channel.h
new file mode 100644
index 000000000..38d110645
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x04_channel.h
@@ -0,0 +1,121 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2013 NVIDIA Corporation.
+ */
+
+ /*
+  * 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 HOST1X_HW_HOST1X04_CHANNEL_H
+#define HOST1X_HW_HOST1X04_CHANNEL_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 >> 11) & 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()
+static inline u32 host1x_channel_channelctrl_r(void)
+{
+	return 0x98;
+}
+#define HOST1X_CHANNEL_CHANNELCTRL \
+	host1x_channel_channelctrl_r()
+static inline u32 host1x_channel_channelctrl_kernel_filter_gbuffer_f(u32 v)
+{
+	return (v & 0x1) << 2;
+}
+#define HOST1X_CHANNEL_CHANNELCTRL_KERNEL_FILTER_GBUFFER(v) \
+	host1x_channel_channelctrl_kernel_filter_gbuffer_f(v)
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x04_sync.h b/drivers/gpu/host1x/hw/hw_host1x04_sync.h
new file mode 100644
index 000000000..0be98562c
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x04_sync.h
@@ -0,0 +1,231 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2013 NVIDIA Corporation.
+ */
+
+ /*
+  * 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 HOST1X_HW_HOST1X04_SYNC_H
+#define HOST1X_HW_HOST1X04_SYNC_H
+
+#define REGISTER_STRIDE	4
+
+static inline u32 host1x_sync_syncpt_r(unsigned int id)
+{
+	return 0xf80 + 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 0xe80 + 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 0xf00 + 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 0xf20 + 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 0xc00 + 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) & 0x3ff;
+}
+#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) & 0x3ff;
+}
+#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_v(u32 v)
+{
+	return (v >> 8) & 0xf;
+}
+#define HOST1X_SYNC_MLOCK_OWNER_CHID_V(v) \
+	host1x_sync_mlock_owner_chid_v(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 0x1380 + 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 0xf60 + 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 0xc80 + 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 & 0x3ff) << 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 & 0xf) << 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) & 0x3ff;
+}
+#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) & 0x3ff;
+}
+#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 0xcc0 + 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
diff --git a/drivers/gpu/host1x/hw/hw_host1x04_uclass.h b/drivers/gpu/host1x/hw/hw_host1x04_uclass.h
new file mode 100644
index 000000000..60c692b92
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x04_uclass.h
@@ -0,0 +1,181 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2013 NVIDIA Corporation.
+ */
+
+ /*
+  * 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 HOST1X_HW_HOST1X04_UCLASS_H
+#define HOST1X_HW_HOST1X04_UCLASS_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_r(void)
+{
+	return 0xb;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE \
+	host1x_uclass_load_syncpt_base_r()
+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)
+static inline u32 host1x_uclass_load_syncpt_payload_32_r(void)
+{
+	return 0x4e;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_PAYLOAD_32 \
+	host1x_uclass_load_syncpt_payload_32_r()
+static inline u32 host1x_uclass_wait_syncpt_32_r(void)
+{
+	return 0x50;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_32 \
+	host1x_uclass_wait_syncpt_32_r()
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x05_channel.h b/drivers/gpu/host1x/hw/hw_host1x05_channel.h
new file mode 100644
index 000000000..7e628ef58
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x05_channel.h
@@ -0,0 +1,121 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2015 NVIDIA Corporation.
+ */
+
+ /*
+  * 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 HOST1X_HW_HOST1X05_CHANNEL_H
+#define HOST1X_HW_HOST1X05_CHANNEL_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 >> 11) & 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()
+static inline u32 host1x_channel_channelctrl_r(void)
+{
+	return 0x98;
+}
+#define HOST1X_CHANNEL_CHANNELCTRL \
+	host1x_channel_channelctrl_r()
+static inline u32 host1x_channel_channelctrl_kernel_filter_gbuffer_f(u32 v)
+{
+	return (v & 0x1) << 2;
+}
+#define HOST1X_CHANNEL_CHANNELCTRL_KERNEL_FILTER_GBUFFER(v) \
+	host1x_channel_channelctrl_kernel_filter_gbuffer_f(v)
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x05_sync.h b/drivers/gpu/host1x/hw/hw_host1x05_sync.h
new file mode 100644
index 000000000..1a85c793b
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x05_sync.h
@@ -0,0 +1,231 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2015 NVIDIA Corporation.
+ */
+
+ /*
+  * 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 HOST1X_HW_HOST1X05_SYNC_H
+#define HOST1X_HW_HOST1X05_SYNC_H
+
+#define REGISTER_STRIDE	4
+
+static inline u32 host1x_sync_syncpt_r(unsigned int id)
+{
+	return 0xf80 + 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 0xe80 + 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 0xf00 + 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 0xf20 + 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 0xc00 + 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) & 0x3ff;
+}
+#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) & 0x3ff;
+}
+#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_v(u32 r)
+{
+	return (r >> 8) & 0xf;
+}
+#define HOST1X_SYNC_MLOCK_OWNER_CHID_V(v) \
+	host1x_sync_mlock_owner_chid_v(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 0x1380 + 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 0xf60 + 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 0xc80 + 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 & 0x3ff) << 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 & 0xf) << 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) & 0x3ff;
+}
+#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) & 0x3ff;
+}
+#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 0xcc0 + 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
diff --git a/drivers/gpu/host1x/hw/hw_host1x05_uclass.h b/drivers/gpu/host1x/hw/hw_host1x05_uclass.h
new file mode 100644
index 000000000..2fcc9a2ad
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x05_uclass.h
@@ -0,0 +1,181 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2015 NVIDIA Corporation.
+ */
+
+ /*
+  * 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 HOST1X_HW_HOST1X05_UCLASS_H
+#define HOST1X_HW_HOST1X05_UCLASS_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_r(void)
+{
+	return 0xb;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE \
+	host1x_uclass_load_syncpt_base_r()
+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)
+static inline u32 host1x_uclass_load_syncpt_payload_32_r(void)
+{
+	return 0x4e;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_PAYLOAD_32 \
+	host1x_uclass_load_syncpt_payload_32_r()
+static inline u32 host1x_uclass_wait_syncpt_32_r(void)
+{
+	return 0x50;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_32 \
+	host1x_uclass_wait_syncpt_32_r()
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x06_channel.h b/drivers/gpu/host1x/hw/hw_host1x06_channel.h
new file mode 100644
index 000000000..18ae1c57b
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x06_channel.h
@@ -0,0 +1,11 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2019 NVIDIA Corporation.
+ */
+
+#ifndef HOST1X_HW_HOST1X06_CHANNEL_H
+#define HOST1X_HW_HOST1X06_CHANNEL_H
+
+#define HOST1X_CHANNEL_SMMU_STREAMID 0x084
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x06_hypervisor.h b/drivers/gpu/host1x/hw/hw_host1x06_hypervisor.h
new file mode 100644
index 000000000..a7fc9ec4b
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x06_hypervisor.h
@@ -0,0 +1,20 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2017 NVIDIA Corporation.
+ */
+
+#define HOST1X_HV_SYNCPT_PROT_EN			0x1ac4
+#define HOST1X_HV_SYNCPT_PROT_EN_CH_EN			BIT(1)
+#define HOST1X_HV_CH_KERNEL_FILTER_GBUFFER(x)		(0x2020 + (x * 4))
+#define HOST1X_HV_CMDFIFO_PEEK_CTRL			0x233c
+#define HOST1X_HV_CMDFIFO_PEEK_CTRL_ADDR(x)		(x)
+#define HOST1X_HV_CMDFIFO_PEEK_CTRL_CHANNEL(x)		((x) << 16)
+#define HOST1X_HV_CMDFIFO_PEEK_CTRL_ENABLE		BIT(31)
+#define HOST1X_HV_CMDFIFO_PEEK_READ			0x2340
+#define HOST1X_HV_CMDFIFO_PEEK_PTRS			0x2344
+#define HOST1X_HV_CMDFIFO_PEEK_PTRS_WR_PTR_V(x)		(((x) >> 16) & 0xfff)
+#define HOST1X_HV_CMDFIFO_PEEK_PTRS_RD_PTR_V(x)		((x) & 0xfff)
+#define HOST1X_HV_CMDFIFO_SETUP(x)			(0x2588 + (x * 4))
+#define HOST1X_HV_CMDFIFO_SETUP_LIMIT_V(x)		(((x) >> 16) & 0xfff)
+#define HOST1X_HV_CMDFIFO_SETUP_BASE_V(x)		((x) & 0xfff)
+#define HOST1X_HV_ICG_EN_OVERRIDE			0x2aa8
diff --git a/drivers/gpu/host1x/hw/hw_host1x06_uclass.h b/drivers/gpu/host1x/hw/hw_host1x06_uclass.h
new file mode 100644
index 000000000..50c32de45
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x06_uclass.h
@@ -0,0 +1,181 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2017 NVIDIA Corporation.
+ */
+
+ /*
+  * 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 HOST1X_HW_HOST1X06_UCLASS_H
+#define HOST1X_HW_HOST1X06_UCLASS_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) << 10;
+}
+#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 & 0x3ff) << 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_r(void)
+{
+	return 0xb;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE \
+	host1x_uclass_load_syncpt_base_r()
+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)
+static inline u32 host1x_uclass_load_syncpt_payload_32_r(void)
+{
+	return 0x4e;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_PAYLOAD_32 \
+	host1x_uclass_load_syncpt_payload_32_r()
+static inline u32 host1x_uclass_wait_syncpt_32_r(void)
+{
+	return 0x50;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_32 \
+	host1x_uclass_wait_syncpt_32_r()
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x06_vm.h b/drivers/gpu/host1x/hw/hw_host1x06_vm.h
new file mode 100644
index 000000000..818564a76
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x06_vm.h
@@ -0,0 +1,35 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2017 NVIDIA Corporation.
+ */
+
+#define HOST1X_CHANNEL_DMASTART				0x0000
+#define HOST1X_CHANNEL_DMASTART_HI			0x0004
+#define HOST1X_CHANNEL_DMAPUT				0x0008
+#define HOST1X_CHANNEL_DMAPUT_HI			0x000c
+#define HOST1X_CHANNEL_DMAGET				0x0010
+#define HOST1X_CHANNEL_DMAGET_HI			0x0014
+#define HOST1X_CHANNEL_DMAEND				0x0018
+#define HOST1X_CHANNEL_DMAEND_HI			0x001c
+#define HOST1X_CHANNEL_DMACTRL				0x0020
+#define HOST1X_CHANNEL_DMACTRL_DMASTOP			BIT(0)
+#define HOST1X_CHANNEL_DMACTRL_DMAGETRST		BIT(1)
+#define HOST1X_CHANNEL_DMACTRL_DMAINITGET		BIT(2)
+#define HOST1X_CHANNEL_CMDFIFO_STAT			0x0024
+#define HOST1X_CHANNEL_CMDFIFO_STAT_EMPTY		BIT(13)
+#define HOST1X_CHANNEL_CMDFIFO_RDATA			0x0028
+#define HOST1X_CHANNEL_CMDP_OFFSET			0x0030
+#define HOST1X_CHANNEL_CMDP_CLASS			0x0034
+#define HOST1X_CHANNEL_CHANNELSTAT			0x0038
+#define HOST1X_CHANNEL_CMDPROC_STOP			0x0048
+#define HOST1X_CHANNEL_TEARDOWN				0x004c
+
+#define HOST1X_SYNC_SYNCPT_CPU_INCR(x)			(0x6400 + 4*(x))
+#define HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(x)	(0x6464 + 4*(x))
+#define HOST1X_SYNC_SYNCPT_THRESH_INT_ENABLE_CPU0(x)	(0x652c + 4*(x))
+#define HOST1X_SYNC_SYNCPT_THRESH_INT_DISABLE(x)	(0x6590 + 4*(x))
+#define HOST1X_SYNC_SYNCPT_BASE(x)			(0x8000 + 4*(x))
+#define HOST1X_SYNC_SYNCPT(x)				(0x8080 + 4*(x))
+#define HOST1X_SYNC_SYNCPT_INT_THRESH(x)		(0x8a00 + 4*(x))
+#define HOST1X_SYNC_SYNCPT_CH_APP(x)			(0x9384 + 4*(x))
+#define HOST1X_SYNC_SYNCPT_CH_APP_CH(v)			(((v) & 0x3f) << 8)
diff --git a/drivers/gpu/host1x/hw/hw_host1x07_channel.h b/drivers/gpu/host1x/hw/hw_host1x07_channel.h
new file mode 100644
index 000000000..96fa72bbd
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x07_channel.h
@@ -0,0 +1,11 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2019 NVIDIA Corporation.
+ */
+
+#ifndef HOST1X_HW_HOST1X07_CHANNEL_H
+#define HOST1X_HW_HOST1X07_CHANNEL_H
+
+#define HOST1X_CHANNEL_SMMU_STREAMID 0x084
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x07_hypervisor.h b/drivers/gpu/host1x/hw/hw_host1x07_hypervisor.h
new file mode 100644
index 000000000..52141d539
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x07_hypervisor.h
@@ -0,0 +1,20 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2018 NVIDIA Corporation.
+ */
+
+#define HOST1X_HV_SYNCPT_PROT_EN			0x1ac4
+#define HOST1X_HV_SYNCPT_PROT_EN_CH_EN			BIT(1)
+#define HOST1X_HV_CH_KERNEL_FILTER_GBUFFER(x)		(0x2020 + (x * 4))
+#define HOST1X_HV_CMDFIFO_PEEK_CTRL			0x233c
+#define HOST1X_HV_CMDFIFO_PEEK_CTRL_ADDR(x)		(x)
+#define HOST1X_HV_CMDFIFO_PEEK_CTRL_CHANNEL(x)		((x) << 16)
+#define HOST1X_HV_CMDFIFO_PEEK_CTRL_ENABLE		BIT(31)
+#define HOST1X_HV_CMDFIFO_PEEK_READ			0x2340
+#define HOST1X_HV_CMDFIFO_PEEK_PTRS			0x2344
+#define HOST1X_HV_CMDFIFO_PEEK_PTRS_WR_PTR_V(x)		(((x) >> 16) & 0xfff)
+#define HOST1X_HV_CMDFIFO_PEEK_PTRS_RD_PTR_V(x)		((x) & 0xfff)
+#define HOST1X_HV_CMDFIFO_SETUP(x)			(0x2588 + (x * 4))
+#define HOST1X_HV_CMDFIFO_SETUP_LIMIT_V(x)		(((x) >> 16) & 0xfff)
+#define HOST1X_HV_CMDFIFO_SETUP_BASE_V(x)		((x) & 0xfff)
+#define HOST1X_HV_ICG_EN_OVERRIDE			0x2aa8
diff --git a/drivers/gpu/host1x/hw/hw_host1x07_uclass.h b/drivers/gpu/host1x/hw/hw_host1x07_uclass.h
new file mode 100644
index 000000000..887b878f9
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x07_uclass.h
@@ -0,0 +1,181 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2018 NVIDIA Corporation.
+ */
+
+ /*
+  * 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 HOST1X_HW_HOST1X07_UCLASS_H
+#define HOST1X_HW_HOST1X07_UCLASS_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) << 10;
+}
+#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 & 0x3ff) << 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_r(void)
+{
+	return 0xb;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE \
+	host1x_uclass_load_syncpt_base_r()
+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)
+static inline u32 host1x_uclass_load_syncpt_payload_32_r(void)
+{
+	return 0x4e;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_PAYLOAD_32 \
+	host1x_uclass_load_syncpt_payload_32_r()
+static inline u32 host1x_uclass_wait_syncpt_32_r(void)
+{
+	return 0x50;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_32 \
+	host1x_uclass_wait_syncpt_32_r()
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x07_vm.h b/drivers/gpu/host1x/hw/hw_host1x07_vm.h
new file mode 100644
index 000000000..b766851d5
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x07_vm.h
@@ -0,0 +1,34 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2018 NVIDIA Corporation.
+ */
+
+#define HOST1X_CHANNEL_DMASTART				0x0000
+#define HOST1X_CHANNEL_DMASTART_HI			0x0004
+#define HOST1X_CHANNEL_DMAPUT				0x0008
+#define HOST1X_CHANNEL_DMAPUT_HI			0x000c
+#define HOST1X_CHANNEL_DMAGET				0x0010
+#define HOST1X_CHANNEL_DMAGET_HI			0x0014
+#define HOST1X_CHANNEL_DMAEND				0x0018
+#define HOST1X_CHANNEL_DMAEND_HI			0x001c
+#define HOST1X_CHANNEL_DMACTRL				0x0020
+#define HOST1X_CHANNEL_DMACTRL_DMASTOP			BIT(0)
+#define HOST1X_CHANNEL_DMACTRL_DMAGETRST		BIT(1)
+#define HOST1X_CHANNEL_DMACTRL_DMAINITGET		BIT(2)
+#define HOST1X_CHANNEL_CMDFIFO_STAT			0x0024
+#define HOST1X_CHANNEL_CMDFIFO_STAT_EMPTY		BIT(13)
+#define HOST1X_CHANNEL_CMDFIFO_RDATA			0x0028
+#define HOST1X_CHANNEL_CMDP_OFFSET			0x0030
+#define HOST1X_CHANNEL_CMDP_CLASS			0x0034
+#define HOST1X_CHANNEL_CHANNELSTAT			0x0038
+#define HOST1X_CHANNEL_CMDPROC_STOP			0x0048
+#define HOST1X_CHANNEL_TEARDOWN				0x004c
+
+#define HOST1X_SYNC_SYNCPT_CPU_INCR(x)			(0x6400 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(x)	(0x6464 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_THRESH_INT_ENABLE_CPU0(x)	(0x652c + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_THRESH_INT_DISABLE(x)	(0x6590 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT(x)				(0x8080 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_INT_THRESH(x)		(0x9980 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_CH_APP(x)			(0xa604 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_CH_APP_CH(v)			(((v) & 0x3f) << 8)
diff --git a/drivers/gpu/host1x/hw/hw_host1x08_channel.h b/drivers/gpu/host1x/hw/hw_host1x08_channel.h
new file mode 100644
index 000000000..c9272d2ab
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x08_channel.h
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2022 NVIDIA Corporation.
+ */
+
+#ifndef HOST1X_HW_HOST1X08_CHANNEL_H
+#define HOST1X_HW_HOST1X08_CHANNEL_H
+
+#define HOST1X_CHANNEL_SMMU_STREAMID 0x084
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x08_common.h b/drivers/gpu/host1x/hw/hw_host1x08_common.h
new file mode 100644
index 000000000..8e0c99150
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x08_common.h
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2022 NVIDIA Corporation.
+ */
+
+#define HOST1X_COMMON_OFA_MLOCK			0x4050
+#define HOST1X_COMMON_NVJPG1_MLOCK		0x4070
+#define HOST1X_COMMON_VIC_MLOCK			0x4078
+#define HOST1X_COMMON_NVENC_MLOCK		0x407c
+#define HOST1X_COMMON_NVDEC_MLOCK		0x4080
+#define HOST1X_COMMON_NVJPG_MLOCK		0x4084
diff --git a/drivers/gpu/host1x/hw/hw_host1x08_hypervisor.h b/drivers/gpu/host1x/hw/hw_host1x08_hypervisor.h
new file mode 100644
index 000000000..22964324c
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x08_hypervisor.h
@@ -0,0 +1,9 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2022 NVIDIA Corporation.
+ */
+
+#define HOST1X_HV_SYNCPT_PROT_EN			0x1724
+#define HOST1X_HV_SYNCPT_PROT_EN_CH_EN			BIT(1)
+#define HOST1X_HV_CH_MLOCK_EN(x)			(0x1700 + (x * 4))
+#define HOST1X_HV_CH_KERNEL_FILTER_GBUFFER(x)		(0x1710 + (x * 4))
diff --git a/drivers/gpu/host1x/hw/hw_host1x08_uclass.h b/drivers/gpu/host1x/hw/hw_host1x08_uclass.h
new file mode 100644
index 000000000..4fb1d090e
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x08_uclass.h
@@ -0,0 +1,181 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2018 NVIDIA Corporation.
+ */
+
+ /*
+  * 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 HOST1X_HW_HOST1X08_UCLASS_H
+#define HOST1X_HW_HOST1X08_UCLASS_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) << 10;
+}
+#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 & 0x3ff) << 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_r(void)
+{
+	return 0xb;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE \
+	host1x_uclass_load_syncpt_base_r()
+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)
+static inline u32 host1x_uclass_load_syncpt_payload_32_r(void)
+{
+	return 0x4e;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_PAYLOAD_32 \
+	host1x_uclass_load_syncpt_payload_32_r()
+static inline u32 host1x_uclass_wait_syncpt_32_r(void)
+{
+	return 0x50;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_32 \
+	host1x_uclass_wait_syncpt_32_r()
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x08_vm.h b/drivers/gpu/host1x/hw/hw_host1x08_vm.h
new file mode 100644
index 000000000..1455a4670
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x08_vm.h
@@ -0,0 +1,36 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2022 NVIDIA Corporation.
+ */
+
+#define HOST1X_CHANNEL_DMASTART				0x0000
+#define HOST1X_CHANNEL_DMASTART_HI			0x0004
+#define HOST1X_CHANNEL_DMAPUT				0x0008
+#define HOST1X_CHANNEL_DMAPUT_HI			0x000c
+#define HOST1X_CHANNEL_DMAGET				0x0010
+#define HOST1X_CHANNEL_DMAGET_HI			0x0014
+#define HOST1X_CHANNEL_DMAEND				0x0018
+#define HOST1X_CHANNEL_DMAEND_HI			0x001c
+#define HOST1X_CHANNEL_DMACTRL				0x0020
+#define HOST1X_CHANNEL_DMACTRL_DMASTOP			BIT(0)
+#define HOST1X_CHANNEL_DMACTRL_DMAGETRST		BIT(1)
+#define HOST1X_CHANNEL_DMACTRL_DMAINITGET		BIT(2)
+#define HOST1X_CHANNEL_CMDFIFO_STAT			0x0024
+#define HOST1X_CHANNEL_CMDFIFO_STAT_EMPTY		BIT(13)
+#define HOST1X_CHANNEL_CMDFIFO_RDATA			0x0028
+#define HOST1X_CHANNEL_CMDP_OFFSET			0x0030
+#define HOST1X_CHANNEL_CMDP_CLASS			0x0034
+#define HOST1X_CHANNEL_CHANNELSTAT			0x0038
+#define HOST1X_CHANNEL_CMDPROC_STOP			0x0048
+#define HOST1X_CHANNEL_TEARDOWN				0x004c
+#define HOST1X_CHANNEL_SMMU_STREAMID			0x0084
+
+#define HOST1X_SYNC_SYNCPT_CPU_INCR(x)			(0x6400 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(x)	(0x6600 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_INTR_DEST(x)			(0x6684 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_THRESH_INT_ENABLE_CPU0(x)	(0x770c + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_THRESH_INT_DISABLE(x)	(0x7790 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT(x)				(0x8080 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_INT_THRESH(x)		(0xa088 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_CH_APP(x)			(0xb090 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_CH_APP_CH(v)			(((v) & 0x3f) << 8)
diff --git a/drivers/gpu/host1x/hw/hw_host1x10_common.h b/drivers/gpu/host1x/hw/hw_host1x10_common.h
new file mode 100644
index 000000000..48a632672
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x10_common.h
@@ -0,0 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+ */
+
+#define HOST1X_COMMON_VIC_MLOCK			0x4060
diff --git a/drivers/gpu/host1x/hw/hw_host1x10_hypervisor.h b/drivers/gpu/host1x/hw/hw_host1x10_hypervisor.h
new file mode 100644
index 000000000..8c9069caf
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x10_hypervisor.h
@@ -0,0 +1,10 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+ */
+
+#define HOST1X_HV_SYNCPT_PROT_EN			0x172c
+#define HOST1X_HV_SYNCPT_PROT_EN_CH_EN			BIT(1)
+#define HOST1X_HV_CH_MLOCK_EN(x)			(0x1708 + (x * 4))
+#define HOST1X_HV_CH_KERNEL_FILTER_GBUFFER(x)		(0x1718 + (x * 4))
+#define HOST1X_HV_SYNCPT_VM(x)				(0x0 + 4 * (x))
diff --git a/drivers/gpu/host1x/hw/hw_host1x10_uclass.h b/drivers/gpu/host1x/hw/hw_host1x10_uclass.h
new file mode 100644
index 000000000..abe83e67f
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x10_uclass.h
@@ -0,0 +1,181 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+ */
+
+ /*
+  * 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 HOST1X_HW_HOST1X10_UCLASS_H
+#define HOST1X_HW_HOST1X10_UCLASS_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) << 10;
+}
+#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 & 0x3ff) << 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_r(void)
+{
+	return 0xb;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE \
+	host1x_uclass_load_syncpt_base_r()
+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)
+static inline u32 host1x_uclass_load_syncpt_payload_32_r(void)
+{
+	return 0x4e;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_PAYLOAD_32 \
+	host1x_uclass_load_syncpt_payload_32_r()
+static inline u32 host1x_uclass_wait_syncpt_32_r(void)
+{
+	return 0x50;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_32 \
+	host1x_uclass_wait_syncpt_32_r()
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x10_vm.h b/drivers/gpu/host1x/hw/hw_host1x10_vm.h
new file mode 100644
index 000000000..75f5b881c
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x10_vm.h
@@ -0,0 +1,36 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+ */
+
+#define HOST1X_CHANNEL_DMASTART				0x0000
+#define HOST1X_CHANNEL_DMASTART_HI			0x0004
+#define HOST1X_CHANNEL_DMAPUT				0x0008
+#define HOST1X_CHANNEL_DMAPUT_HI			0x000c
+#define HOST1X_CHANNEL_DMAGET				0x0010
+#define HOST1X_CHANNEL_DMAGET_HI			0x0014
+#define HOST1X_CHANNEL_DMAEND				0x0018
+#define HOST1X_CHANNEL_DMAEND_HI			0x001c
+#define HOST1X_CHANNEL_DMACTRL				0x0020
+#define HOST1X_CHANNEL_DMACTRL_DMASTOP			BIT(0)
+#define HOST1X_CHANNEL_DMACTRL_DMAGETRST		BIT(1)
+#define HOST1X_CHANNEL_DMACTRL_DMAINITGET		BIT(2)
+#define HOST1X_CHANNEL_CMDFIFO_STAT			0x0024
+#define HOST1X_CHANNEL_CMDFIFO_STAT_EMPTY		BIT(13)
+#define HOST1X_CHANNEL_CMDFIFO_RDATA			0x0028
+#define HOST1X_CHANNEL_CMDP_OFFSET			0x0030
+#define HOST1X_CHANNEL_CMDP_CLASS			0x0034
+#define HOST1X_CHANNEL_CHANNELSTAT			0x0038
+#define HOST1X_CHANNEL_CMDPROC_STOP			0x0048
+#define HOST1X_CHANNEL_TEARDOWN				0x004c
+#define HOST1X_CHANNEL_SMMU_STREAMID			0x0084
+
+#define HOST1X_SYNC_SYNCPT_CPU_INCR(x)			(0x6400 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(x)	(0x6600 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_INTR_DEST(x)			(0x6684 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_THRESH_INT_ENABLE_CPU0(x)	(0x770c + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_THRESH_INT_DISABLE(x)	(0x7790 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT(x)				(0x8080 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_INT_THRESH(x)		(0xa088 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_CH_APP(x)			(0xb090 + 4 * (x))
+#define HOST1X_SYNC_SYNCPT_CH_APP_CH(v)			(((v) & 0x3f) << 8)
diff --git a/drivers/gpu/host1x/hw/intr_hw.c b/drivers/gpu/host1x/hw/intr_hw.c
new file mode 100644
index 000000000..bd5b5ef62
--- /dev/null
+++ b/drivers/gpu/host1x/hw/intr_hw.c
@@ -0,0 +1,154 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Tegra host1x Interrupt Management
+ *
+ * Copyright (C) 2010 Google, Inc.
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ */
+
+#include <linux/io.h>
+
+#include "../intr.h"
+#include "../dev.h"
+
+static void process_32_syncpts(struct host1x *host, unsigned long val, u32 reg_offset)
+{
+	unsigned int id;
+
+	if (!val)
+		return;
+
+	host1x_sync_writel(host, val, HOST1X_SYNC_SYNCPT_THRESH_INT_DISABLE(reg_offset));
+	host1x_sync_writel(host, val, HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(reg_offset));
+
+	for_each_set_bit(id, &val, 32)
+		host1x_intr_handle_interrupt(host, reg_offset * 32 + id);
+}
+
+static irqreturn_t syncpt_thresh_isr(int irq, void *dev_id)
+{
+	struct host1x_intr_irq_data *irq_data = dev_id;
+	struct host1x *host = irq_data->host;
+	unsigned long reg;
+	unsigned int i;
+
+#if !defined(CONFIG_64BIT)
+	for (i = irq_data->offset; i < DIV_ROUND_UP(host->info->nb_pts, 32);
+	     i += host->num_syncpt_irqs) {
+		reg = host1x_sync_readl(host,
+			HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(i));
+
+		process_32_syncpts(host, reg, i);
+	}
+#elif HOST1X_HW == 6 || HOST1X_HW == 7
+	/*
+	 * Tegra186 and Tegra194 have the first INT_STATUS register not 64-bit aligned,
+	 * and only have one interrupt line.
+	 */
+	reg = host1x_sync_readl(host, HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(0));
+	process_32_syncpts(host, reg, 0);
+
+	for (i = 1; i < (host->info->nb_pts / 32) - 1; i += 2) {
+		reg = host1x_sync_readq(host,
+			HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(i));
+
+		process_32_syncpts(host, lower_32_bits(reg), i);
+		process_32_syncpts(host, upper_32_bits(reg), i + 1);
+	}
+
+	reg = host1x_sync_readl(host, HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(i));
+	process_32_syncpts(host, reg, i);
+#else
+	/* All 64-bit capable SoCs have number of syncpoints divisible by 64 */
+	for (i = irq_data->offset; i < DIV_ROUND_UP(host->info->nb_pts, 64);
+	     i += host->num_syncpt_irqs) {
+		reg = host1x_sync_readq(host,
+			HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(i * 2));
+
+		process_32_syncpts(host, lower_32_bits(reg), i * 2 + 0);
+		process_32_syncpts(host, upper_32_bits(reg), i * 2 + 1);
+	}
+#endif
+
+	return IRQ_HANDLED;
+}
+
+static void host1x_intr_disable_all_syncpt_intrs(struct host1x *host)
+{
+	unsigned int i;
+
+	for (i = 0; i < DIV_ROUND_UP(host->info->nb_pts, 32); ++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)
+{
+#if HOST1X_HW < 6
+	/* 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);
+#endif
+#if HOST1X_HW >= 8
+	u32 id;
+
+	/*
+	 * Program threshold interrupt destination among 8 lines per VM,
+	 * per syncpoint. For each group of 64 syncpoints (corresponding to two
+	 * interrupt status registers), direct to one interrupt line, going
+	 * around in a round robin fashion.
+	 */
+	for (id = 0; id < host->info->nb_pts; id++) {
+		u32 reg_offset = id / 64;
+		u32 irq_index = reg_offset % host->num_syncpt_irqs;
+
+		host1x_sync_writel(host, irq_index, HOST1X_SYNC_SYNCPT_INTR_DEST(id));
+	}
+#endif
+
+	return 0;
+}
+
+static void host1x_intr_set_syncpt_threshold(struct host1x *host,
+					      unsigned int id,
+					      u32 thresh)
+{
+	host1x_sync_writel(host, thresh, HOST1X_SYNC_SYNCPT_INT_THRESH(id));
+}
+
+static void host1x_intr_enable_syncpt_intr(struct host1x *host,
+					    unsigned int id)
+{
+	host1x_sync_writel(host, BIT(id % 32),
+		HOST1X_SYNC_SYNCPT_THRESH_INT_ENABLE_CPU0(id / 32));
+}
+
+static void host1x_intr_disable_syncpt_intr(struct host1x *host,
+					     unsigned int id)
+{
+	host1x_sync_writel(host, BIT(id % 32),
+		HOST1X_SYNC_SYNCPT_THRESH_INT_DISABLE(id / 32));
+	host1x_sync_writel(host, BIT(id % 32),
+		HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(id / 32));
+}
+
+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,
+	.isr = syncpt_thresh_isr,
+};
diff --git a/drivers/gpu/host1x/hw/opcodes.h b/drivers/gpu/host1x/hw/opcodes.h
new file mode 100644
index 000000000..649614499
--- /dev/null
+++ b/drivers/gpu/host1x/hw/opcodes.h
@@ -0,0 +1,150 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Tegra host1x opcodes
+ *
+ * Copyright (c) 2022 NVIDIA Corporation.
+ */
+
+#ifndef __HOST1X_OPCODES_H
+#define __HOST1X_OPCODES_H
+
+#include <linux/types.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;
+}
+
+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;
+}
+
+static inline u32 host1x_opcode_setstreamid(unsigned streamid)
+{
+	return (7 << 28) | streamid;
+}
+
+static inline u32 host1x_opcode_setpayload(unsigned payload)
+{
+	return (9 << 28) | payload;
+}
+
+static inline u32 host1x_opcode_gather_wide(unsigned count)
+{
+	return (12 << 28) | count;
+}
+
+static inline u32 host1x_opcode_acquire_mlock(unsigned mlock)
+{
+	return (14 << 28) | (0 << 24) | mlock;
+}
+
+static inline u32 host1x_opcode_release_mlock(unsigned mlock)
+{
+	return (14 << 28) | (1 << 24) | mlock;
+}
+
+#define HOST1X_OPCODE_NOP host1x_opcode_nonincr(0, 0)
+
+#endif
diff --git a/drivers/gpu/host1x/hw/syncpt_hw.c b/drivers/gpu/host1x/hw/syncpt_hw.c
new file mode 100644
index 000000000..8cf35b2ef
--- /dev/null
+++ b/drivers/gpu/host1x/hw/syncpt_hw.c
@@ -0,0 +1,141 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Tegra host1x Syncpoints
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ */
+
+#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)
+{
+	u32 min = host1x_syncpt_read_min(sp);
+	struct host1x *host = sp->host;
+
+	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)
+{
+#if HOST1X_HW < 7
+	struct host1x *host = sp->host;
+
+	host1x_sync_writel(host, sp->base_val,
+			   HOST1X_SYNC_SYNCPT_BASE(sp->id));
+#endif
+}
+
+/*
+ * Read waitbase value from hw.
+ */
+static void syncpt_read_wait_base(struct host1x_syncpt *sp)
+{
+#if HOST1X_HW < 7
+	struct host1x *host = sp->host;
+
+	sp->base_val =
+		host1x_sync_readl(host, HOST1X_SYNC_SYNCPT_BASE(sp->id));
+#endif
+}
+
+/*
+ * 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 int 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))
+		return -EINVAL;
+
+	host1x_sync_writel(host, BIT(sp->id % 32),
+			   HOST1X_SYNC_SYNCPT_CPU_INCR(reg_offset));
+	wmb();
+
+	return 0;
+}
+
+/**
+ * syncpt_assign_to_channel() - Assign syncpoint to channel
+ * @sp: syncpoint
+ * @ch: channel
+ *
+ * On chips with the syncpoint protection feature (Tegra186+), assign @sp to
+ * @ch, preventing other channels from incrementing the syncpoints. If @ch is
+ * NULL, unassigns the syncpoint.
+ *
+ * On older chips, do nothing.
+ */
+static void syncpt_assign_to_channel(struct host1x_syncpt *sp,
+				  struct host1x_channel *ch)
+{
+#if HOST1X_HW >= 6
+	struct host1x *host = sp->host;
+
+	host1x_sync_writel(host,
+			   HOST1X_SYNC_SYNCPT_CH_APP_CH(ch ? ch->id : 0xff),
+			   HOST1X_SYNC_SYNCPT_CH_APP(sp->id));
+#endif
+}
+
+/**
+ * syncpt_enable_protection() - Enable syncpoint protection
+ * @host: host1x instance
+ *
+ * On chips with the syncpoint protection feature (Tegra186+), enable this
+ * feature. On older chips, do nothing.
+ */
+static void syncpt_enable_protection(struct host1x *host)
+{
+#if HOST1X_HW >= 6
+	if (!host->hv_regs)
+		return;
+
+	host1x_hypervisor_writel(host, HOST1X_HV_SYNCPT_PROT_EN_CH_EN,
+				 HOST1X_HV_SYNCPT_PROT_EN);
+#endif
+}
+
+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,
+	.assign_to_channel = syncpt_assign_to_channel,
+	.enable_protection = syncpt_enable_protection,
+};