Merge tag 'misc-habanalabs-next-2019-11-21' of git://people.freedesktop.org/~gabbayo/linux into char-misc-next

Oded writes:

This tag contains the following changes for kernel 5.5:

- MMU code improvements that includes:
  - Distinguish between "normal" unmapping and unmapping that is done as
    part of the tear-down of a user process. This improves performance of
    unmapping during reset of the device.
  - Add future ASIC support in generic MMU code.

- Improve device reset code by adding more protection around accessing the
  device during the reset process.

- Add new H/W queue type for future ASIC support

- Add more information to be retrieved by users through INFO IOCTL:
  - clock rate
  - board name
  - reset counters

- Small bug fixes and minor improvements to code.

* tag 'misc-habanalabs-next-2019-11-21' of git://people.freedesktop.org/~gabbayo/linux: (31 commits)
  habanalabs: add more protection of device during reset
  habanalabs: flush EQ workers in hard reset
  habanalabs: make the reset code more consistent
  habanalabs: expose reset counters via existing INFO IOCTL
  habanalabs: make code more concise
  habanalabs: use defines for F/W files
  habanalabs: remove prints on successful device initialization
  habanalabs: remove unnecessary checks
  habanalabs: invalidate MMU cache only once
  habanalabs: skip VA block list update in reset flow
  habanalabs: optimize MMU unmap
  habanalabs: prevent read/write from/to the device during hard reset
  habanalabs: split MMU properties to PCI/DRAM
  habanalabs: re-factor MMU masks and documentation
  habanalabs: type specific MMU cache invalidation
  habanalabs: re-factor memory module code
  habanalabs: export uapi defines to user-space
  habanalabs: don't print error when queues are full
  habanalabs: increase max jobs number to 512
  habanalabs: set ETR as non-secured
  ...

20 files changed, 1159 insertions, 542 deletions

diff --git a/drivers/misc/habanalabs/command_submission.c b/drivers/misc/habanalabs/command_submission.c
index a9ac045dcfde..8850f475a413 100644
--- a/drivers/misc/habanalabs/command_submission.c
+++ b/drivers/misc/habanalabs/command_submission.c
@@ -65,6 +65,18 @@ static void cs_put(struct hl_cs *cs)
 	kref_put(&cs->refcount, cs_do_release);
 }
 
+static bool is_cb_patched(struct hl_device *hdev, struct hl_cs_job *job)
+{
+	/*
+	 * Patched CB is created for external queues jobs, and for H/W queues
+	 * jobs if the user CB was allocated by driver and MMU is disabled.
+	 */
+	return (job->queue_type == QUEUE_TYPE_EXT ||
+			(job->queue_type == QUEUE_TYPE_HW &&
+					job->is_kernel_allocated_cb &&
+					!hdev->mmu_enable));
+}
+
 /*
  * cs_parser - parse the user command submission
  *
@@ -91,11 +103,13 @@ static int cs_parser(struct hl_fpriv *hpriv, struct hl_cs_job *job)
 	parser.patched_cb = NULL;
 	parser.user_cb = job->user_cb;
 	parser.user_cb_size = job->user_cb_size;
-	parser.ext_queue = job->ext_queue;
+	parser.queue_type = job->queue_type;
+	parser.is_kernel_allocated_cb = job->is_kernel_allocated_cb;
 	job->patched_cb = NULL;
 
 	rc = hdev->asic_funcs->cs_parser(hdev, &parser);
-	if (job->ext_queue) {
+
+	if (is_cb_patched(hdev, job)) {
 		if (!rc) {
 			job->patched_cb = parser.patched_cb;
 			job->job_cb_size = parser.patched_cb_size;
@@ -124,7 +138,7 @@ static void free_job(struct hl_device *hdev, struct hl_cs_job *job)
 {
 	struct hl_cs *cs = job->cs;
 
-	if (job->ext_queue) {
+	if (is_cb_patched(hdev, job)) {
 		hl_userptr_delete_list(hdev, &job->userptr_list);
 
 		/*
@@ -140,6 +154,19 @@ static void free_job(struct hl_device *hdev, struct hl_cs_job *job)
 		}
 	}
 
+	/* For H/W queue jobs, if a user CB was allocated by driver and MMU is
+	 * enabled, the user CB isn't released in cs_parser() and thus should be
+	 * released here.
+	 */
+	if (job->queue_type == QUEUE_TYPE_HW &&
+			job->is_kernel_allocated_cb && hdev->mmu_enable) {
+		spin_lock(&job->user_cb->lock);
+		job->user_cb->cs_cnt--;
+		spin_unlock(&job->user_cb->lock);
+
+		hl_cb_put(job->user_cb);
+	}
+
 	/*
 	 * This is the only place where there can be multiple threads
 	 * modifying the list at the same time
@@ -150,7 +177,8 @@ static void free_job(struct hl_device *hdev, struct hl_cs_job *job)
 
 	hl_debugfs_remove_job(hdev, job);
 
-	if (job->ext_queue)
+	if (job->queue_type == QUEUE_TYPE_EXT ||
+			job->queue_type == QUEUE_TYPE_HW)
 		cs_put(cs);
 
 	kfree(job);
@@ -387,18 +415,13 @@ static void job_wq_completion(struct work_struct *work)
 	free_job(hdev, job);
 }
 
-static struct hl_cb *validate_queue_index(struct hl_device *hdev,
-					struct hl_cb_mgr *cb_mgr,
-					struct hl_cs_chunk *chunk,
-					bool *ext_queue)
+static int validate_queue_index(struct hl_device *hdev,
+				struct hl_cs_chunk *chunk,
+				enum hl_queue_type *queue_type,
+				bool *is_kernel_allocated_cb)
 {
 	struct asic_fixed_properties *asic = &hdev->asic_prop;
 	struct hw_queue_properties *hw_queue_prop;
-	u32 cb_handle;
-	struct hl_cb *cb;
-
-	/* Assume external queue */
-	*ext_queue = true;
 
 	hw_queue_prop = &asic->hw_queues_props[chunk->queue_index];
 
@@ -406,20 +429,29 @@ static struct hl_cb *validate_queue_index(struct hl_device *hdev,
 			(hw_queue_prop->type == QUEUE_TYPE_NA)) {
 		dev_err(hdev->dev, "Queue index %d is invalid\n",
 			chunk->queue_index);
-		return NULL;
+		return -EINVAL;
 	}
 
 	if (hw_queue_prop->driver_only) {
 		dev_err(hdev->dev,
 			"Queue index %d is restricted for the kernel driver\n",
 			chunk->queue_index);
-		return NULL;
-	} else if (hw_queue_prop->type == QUEUE_TYPE_INT) {
-		*ext_queue = false;
-		return (struct hl_cb *) (uintptr_t) chunk->cb_handle;
+		return -EINVAL;
 	}
 
-	/* Retrieve CB object */
+	*queue_type = hw_queue_prop->type;
+	*is_kernel_allocated_cb = !!hw_queue_prop->requires_kernel_cb;
+
+	return 0;
+}
+
+static struct hl_cb *get_cb_from_cs_chunk(struct hl_device *hdev,
+					struct hl_cb_mgr *cb_mgr,
+					struct hl_cs_chunk *chunk)
+{
+	struct hl_cb *cb;
+	u32 cb_handle;
+
 	cb_handle = (u32) (chunk->cb_handle >> PAGE_SHIFT);
 
 	cb = hl_cb_get(hdev, cb_mgr, cb_handle);
@@ -444,7 +476,8 @@ release_cb:
 	return NULL;
 }
 
-struct hl_cs_job *hl_cs_allocate_job(struct hl_device *hdev, bool ext_queue)
+struct hl_cs_job *hl_cs_allocate_job(struct hl_device *hdev,
+		enum hl_queue_type queue_type, bool is_kernel_allocated_cb)
 {
 	struct hl_cs_job *job;
 
@@ -452,12 +485,14 @@ struct hl_cs_job *hl_cs_allocate_job(struct hl_device *hdev, bool ext_queue)
 	if (!job)
 		return NULL;
 
-	job->ext_queue = ext_queue;
+	job->queue_type = queue_type;
+	job->is_kernel_allocated_cb = is_kernel_allocated_cb;
 
-	if (job->ext_queue) {
+	if (is_cb_patched(hdev, job))
 		INIT_LIST_HEAD(&job->userptr_list);
+
+	if (job->queue_type == QUEUE_TYPE_EXT)
 		INIT_WORK(&job->finish_work, job_wq_completion);
-	}
 
 	return job;
 }
@@ -470,7 +505,7 @@ static int _hl_cs_ioctl(struct hl_fpriv *hpriv, void __user *chunks,
 	struct hl_cs_job *job;
 	struct hl_cs *cs;
 	struct hl_cb *cb;
-	bool ext_queue_present = false;
+	bool int_queues_only = true;
 	u32 size_to_copy;
 	int rc, i, parse_cnt;
 
@@ -514,23 +549,33 @@ static int _hl_cs_ioctl(struct hl_fpriv *hpriv, void __user *chunks,
 	/* Validate ALL the CS chunks before submitting the CS */
 	for (i = 0, parse_cnt = 0 ; i < num_chunks ; i++, parse_cnt++) {
 		struct hl_cs_chunk *chunk = &cs_chunk_array[i];
-		bool ext_queue;
+		enum hl_queue_type queue_type;
+		bool is_kernel_allocated_cb;
+
+		rc = validate_queue_index(hdev, chunk, &queue_type,
+						&is_kernel_allocated_cb);
+		if (rc)
+			goto free_cs_object;
 
-		cb = validate_queue_index(hdev, &hpriv->cb_mgr, chunk,
-					&ext_queue);
-		if (ext_queue) {
-			ext_queue_present = true;
+		if (is_kernel_allocated_cb) {
+			cb = get_cb_from_cs_chunk(hdev, &hpriv->cb_mgr, chunk);
 			if (!cb) {
 				rc = -EINVAL;
 				goto free_cs_object;
 			}
+		} else {
+			cb = (struct hl_cb *) (uintptr_t) chunk->cb_handle;
 		}
 
-		job = hl_cs_allocate_job(hdev, ext_queue);
+		if (queue_type == QUEUE_TYPE_EXT || queue_type == QUEUE_TYPE_HW)
+			int_queues_only = false;
+
+		job = hl_cs_allocate_job(hdev, queue_type,
+						is_kernel_allocated_cb);
 		if (!job) {
 			dev_err(hdev->dev, "Failed to allocate a new job\n");
 			rc = -ENOMEM;
-			if (ext_queue)
+			if (is_kernel_allocated_cb)
 				goto release_cb;
 			else
 				goto free_cs_object;
@@ -540,7 +585,7 @@ static int _hl_cs_ioctl(struct hl_fpriv *hpriv, void __user *chunks,
 		job->cs = cs;
 		job->user_cb = cb;
 		job->user_cb_size = chunk->cb_size;
-		if (job->ext_queue)
+		if (is_kernel_allocated_cb)
 			job->job_cb_size = cb->size;
 		else
 			job->job_cb_size = chunk->cb_size;
@@ -553,10 +598,11 @@ static int _hl_cs_ioctl(struct hl_fpriv *hpriv, void __user *chunks,
 		/*
 		 * Increment CS reference. When CS reference is 0, CS is
 		 * done and can be signaled to user and free all its resources
-		 * Only increment for JOB on external queues, because only
-		 * for those JOBs we get completion
+		 * Only increment for JOB on external or H/W queues, because
+		 * only for those JOBs we get completion
 		 */
-		if (job->ext_queue)
+		if (job->queue_type == QUEUE_TYPE_EXT ||
+				job->queue_type == QUEUE_TYPE_HW)
 			cs_get(cs);
 
 		hl_debugfs_add_job(hdev, job);
@@ -570,9 +616,9 @@ static int _hl_cs_ioctl(struct hl_fpriv *hpriv, void __user *chunks,
 		}
 	}
 
-	if (!ext_queue_present) {
+	if (int_queues_only) {
 		dev_err(hdev->dev,
-			"Reject CS %d.%llu because no external queues jobs\n",
+			"Reject CS %d.%llu because only internal queues jobs are present\n",
 			cs->ctx->asid, cs->sequence);
 		rc = -EINVAL;
 		goto free_cs_object;
@@ -580,9 +626,10 @@ static int _hl_cs_ioctl(struct hl_fpriv *hpriv, void __user *chunks,
 
 	rc = hl_hw_queue_schedule_cs(cs);
 	if (rc) {
-		dev_err(hdev->dev,
-			"Failed to submit CS %d.%llu to H/W queues, error %d\n",
-			cs->ctx->asid, cs->sequence, rc);
+		if (rc != -EAGAIN)
+			dev_err(hdev->dev,
+				"Failed to submit CS %d.%llu to H/W queues, error %d\n",
+				cs->ctx->asid, cs->sequence, rc);
 		goto free_cs_object;
 	}
 
diff --git a/drivers/misc/habanalabs/debugfs.c b/drivers/misc/habanalabs/debugfs.c
index 87f37ac31ccd..20413e350343 100644
--- a/drivers/misc/habanalabs/debugfs.c
+++ b/drivers/misc/habanalabs/debugfs.c
@@ -307,45 +307,57 @@ static inline u64 get_hop0_addr(struct hl_ctx *ctx)
 			(ctx->asid * ctx->hdev->asic_prop.mmu_hop_table_size);
 }
 
-static inline u64 get_hop0_pte_addr(struct hl_ctx *ctx, u64 hop_addr,
-		u64 virt_addr)
+static inline u64 get_hopN_pte_addr(struct hl_ctx *ctx, u64 hop_addr,
+					u64 virt_addr, u64 mask, u64 shift)
 {
 	return hop_addr + ctx->hdev->asic_prop.mmu_pte_size *
-			((virt_addr & HOP0_MASK) >> HOP0_SHIFT);
+			((virt_addr & mask) >> shift);
 }
 
-static inline u64 get_hop1_pte_addr(struct hl_ctx *ctx, u64 hop_addr,
-		u64 virt_addr)
+static inline u64 get_hop0_pte_addr(struct hl_ctx *ctx,
+					struct hl_mmu_properties *mmu_specs,
+					u64 hop_addr, u64 vaddr)
 {
-	return hop_addr + ctx->hdev->asic_prop.mmu_pte_size *
-			((virt_addr & HOP1_MASK) >> HOP1_SHIFT);
+	return get_hopN_pte_addr(ctx, hop_addr, vaddr, mmu_specs->hop0_mask,
+					mmu_specs->hop0_shift);
 }
 
-static inline u64 get_hop2_pte_addr(struct hl_ctx *ctx, u64 hop_addr,
-		u64 virt_addr)
+static inline u64 get_hop1_pte_addr(struct hl_ctx *ctx,
+					struct hl_mmu_properties *mmu_specs,
+					u64 hop_addr, u64 vaddr)
 {
-	return hop_addr + ctx->hdev->asic_prop.mmu_pte_size *
-			((virt_addr & HOP2_MASK) >> HOP2_SHIFT);
+	return get_hopN_pte_addr(ctx, hop_addr, vaddr, mmu_specs->hop1_mask,
+					mmu_specs->hop1_shift);
 }
 
-static inline u64 get_hop3_pte_addr(struct hl_ctx *ctx, u64 hop_addr,
-		u64 virt_addr)
+static inline u64 get_hop2_pte_addr(struct hl_ctx *ctx,
+					struct hl_mmu_properties *mmu_specs,
+					u64 hop_addr, u64 vaddr)
 {
-	return hop_addr + ctx->hdev->asic_prop.mmu_pte_size *
-			((virt_addr & HOP3_MASK) >> HOP3_SHIFT);
+	return get_hopN_pte_addr(ctx, hop_addr, vaddr, mmu_specs->hop2_mask,
+					mmu_specs->hop2_shift);
 }
 
-static inline u64 get_hop4_pte_addr(struct hl_ctx *ctx, u64 hop_addr,
-		u64 virt_addr)
+static inline u64 get_hop3_pte_addr(struct hl_ctx *ctx,
+					struct hl_mmu_properties *mmu_specs,
+					u64 hop_addr, u64 vaddr)
 {
-	return hop_addr + ctx->hdev->asic_prop.mmu_pte_size *
-			((virt_addr & HOP4_MASK) >> HOP4_SHIFT);
+	return get_hopN_pte_addr(ctx, hop_addr, vaddr, mmu_specs->hop3_mask,
+					mmu_specs->hop3_shift);
+}
+
+static inline u64 get_hop4_pte_addr(struct hl_ctx *ctx,
+					struct hl_mmu_properties *mmu_specs,
+					u64 hop_addr, u64 vaddr)
+{
+	return get_hopN_pte_addr(ctx, hop_addr, vaddr, mmu_specs->hop4_mask,
+					mmu_specs->hop4_shift);
 }
 
 static inline u64 get_next_hop_addr(u64 curr_pte)
 {
 	if (curr_pte & PAGE_PRESENT_MASK)
-		return curr_pte & PHYS_ADDR_MASK;
+		return curr_pte & HOP_PHYS_ADDR_MASK;
 	else
 		return ULLONG_MAX;
 }
@@ -355,7 +367,10 @@ static int mmu_show(struct seq_file *s, void *data)
 	struct hl_debugfs_entry *entry = s->private;
 	struct hl_dbg_device_entry *dev_entry = entry->dev_entry;
 	struct hl_device *hdev = dev_entry->hdev;
+	struct asic_fixed_properties *prop = &hdev->asic_prop;
+	struct hl_mmu_properties *mmu_prop;
 	struct hl_ctx *ctx;
+	bool is_dram_addr;
 
 	u64 hop0_addr = 0, hop0_pte_addr = 0, hop0_pte = 0,
 		hop1_addr = 0, hop1_pte_addr = 0, hop1_pte = 0,
@@ -377,33 +392,39 @@ static int mmu_show(struct seq_file *s, void *data)
 		return 0;
 	}
 
+	is_dram_addr = hl_mem_area_inside_range(virt_addr, prop->dmmu.page_size,
+				prop->va_space_dram_start_address,
+				prop->va_space_dram_end_address);
+
+	mmu_prop = is_dram_addr ? &prop->dmmu : &prop->pmmu;
+
 	mutex_lock(&ctx->mmu_lock);
 
 	/* the following lookup is copied from unmap() in mmu.c */
 
 	hop0_addr = get_hop0_addr(ctx);
-	hop0_pte_addr = get_hop0_pte_addr(ctx, hop0_addr, virt_addr);
+	hop0_pte_addr = get_hop0_pte_addr(ctx, mmu_prop, hop0_addr, virt_addr);
 	hop0_pte = hdev->asic_funcs->read_pte(hdev, hop0_pte_addr);
 	hop1_addr = get_next_hop_addr(hop0_pte);
 
 	if (hop1_addr == ULLONG_MAX)
 		goto not_mapped;
 
-	hop1_pte_addr = get_hop1_pte_addr(ctx, hop1_addr, virt_addr);
+	hop1_pte_addr = get_hop1_pte_addr(ctx, mmu_prop, hop1_addr, virt_addr);
 	hop1_pte = hdev->asic_funcs->read_pte(hdev, hop1_pte_addr);
 	hop2_addr = get_next_hop_addr(hop1_pte);
 
 	if (hop2_addr == ULLONG_MAX)
 		goto not_mapped;
 
-	hop2_pte_addr = get_hop2_pte_addr(ctx, hop2_addr, virt_addr);
+	hop2_pte_addr = get_hop2_pte_addr(ctx, mmu_prop, hop2_addr, virt_addr);
 	hop2_pte = hdev->asic_funcs->read_pte(hdev, hop2_pte_addr);
 	hop3_addr = get_next_hop_addr(hop2_pte);
 
 	if (hop3_addr == ULLONG_MAX)
 		goto not_mapped;
 
-	hop3_pte_addr = get_hop3_pte_addr(ctx, hop3_addr, virt_addr);
+	hop3_pte_addr = get_hop3_pte_addr(ctx, mmu_prop, hop3_addr, virt_addr);
 	hop3_pte = hdev->asic_funcs->read_pte(hdev, hop3_pte_addr);
 
 	if (!(hop3_pte & LAST_MASK)) {
@@ -412,7 +433,8 @@ static int mmu_show(struct seq_file *s, void *data)
 		if (hop4_addr == ULLONG_MAX)
 			goto not_mapped;
 
-		hop4_pte_addr = get_hop4_pte_addr(ctx, hop4_addr, virt_addr);
+		hop4_pte_addr = get_hop4_pte_addr(ctx, mmu_prop, hop4_addr,
+							virt_addr);
 		hop4_pte = hdev->asic_funcs->read_pte(hdev, hop4_pte_addr);
 		if (!(hop4_pte & PAGE_PRESENT_MASK))
 			goto not_mapped;
@@ -506,6 +528,12 @@ static int engines_show(struct seq_file *s, void *data)
 	struct hl_dbg_device_entry *dev_entry = entry->dev_entry;
 	struct hl_device *hdev = dev_entry->hdev;
 
+	if (atomic_read(&hdev->in_reset)) {
+		dev_warn_ratelimited(hdev->dev,
+				"Can't check device idle during reset\n");
+		return 0;
+	}
+
 	hdev->asic_funcs->is_device_idle(hdev, NULL, s);
 
 	return 0;
@@ -534,41 +562,50 @@ static int device_va_to_pa(struct hl_device *hdev, u64 virt_addr,
 				u64 *phys_addr)
 {
 	struct hl_ctx *ctx = hdev->compute_ctx;
+	struct asic_fixed_properties *prop = &hdev->asic_prop;
+	struct hl_mmu_properties *mmu_prop;
 	u64 hop_addr, hop_pte_addr, hop_pte;
-	u64 offset_mask = HOP4_MASK | OFFSET_MASK;
+	u64 offset_mask = HOP4_MASK | FLAGS_MASK;
 	int rc = 0;
+	bool is_dram_addr;
 
 	if (!ctx) {
 		dev_err(hdev->dev, "no ctx available\n");
 		return -EINVAL;
 	}
 
+	is_dram_addr = hl_mem_area_inside_range(virt_addr, prop->dmmu.page_size,
+				prop->va_space_dram_start_address,
+				prop->va_space_dram_end_address);
+
+	mmu_prop = is_dram_addr ? &prop->dmmu : &prop->pmmu;
+
 	mutex_lock(&ctx->mmu_lock);
 
 	/* hop 0 */
 	hop_addr = get_hop0_addr(ctx);
-	hop_pte_addr = get_hop0_pte_addr(ctx, hop_addr, virt_addr);
+	hop_pte_addr = get_hop0_pte_addr(ctx, mmu_prop, hop_addr, virt_addr);
 	hop_pte = hdev->asic_funcs->read_pte(hdev, hop_pte_addr);
 
 	/* hop 1 */
 	hop_addr = get_next_hop_addr(hop_pte);
 	if (hop_addr == ULLONG_MAX)
 		goto not_mapped;
-	hop_pte_addr = get_hop1_pte_addr(ctx, hop_addr, virt_addr);
+	hop_pte_addr = get_hop1_pte_addr(ctx, mmu_prop, hop_addr, virt_addr);
 	hop_pte = hdev->asic_funcs->read_pte(hdev, hop_pte_addr);
 
 	/* hop 2 */
 	hop_addr = get_next_hop_addr(hop_pte);
 	if (hop_addr == ULLONG_MAX)
 		goto not_mapped;
-	hop_pte_addr = get_hop2_pte_addr(ctx, hop_addr, virt_addr);
+	hop_pte_addr = get_hop2_pte_addr(ctx, mmu_prop, hop_addr, virt_addr);
 	hop_pte = hdev->asic_funcs->read_pte(hdev, hop_pte_addr);
 
 	/* hop 3 */
 	hop_addr = get_next_hop_addr(hop_pte);
 	if (hop_addr == ULLONG_MAX)
 		goto not_mapped;
-	hop_pte_addr = get_hop3_pte_addr(ctx, hop_addr, virt_addr);
+	hop_pte_addr = get_hop3_pte_addr(ctx, mmu_prop, hop_addr, virt_addr);
 	hop_pte = hdev->asic_funcs->read_pte(hdev, hop_pte_addr);
 
 	if (!(hop_pte & LAST_MASK)) {
@@ -576,10 +613,11 @@ static int device_va_to_pa(struct hl_device *hdev, u64 virt_addr,
 		hop_addr = get_next_hop_addr(hop_pte);
 		if (hop_addr == ULLONG_MAX)
 			goto not_mapped;
-		hop_pte_addr = get_hop4_pte_addr(ctx, hop_addr, virt_addr);
+		hop_pte_addr = get_hop4_pte_addr(ctx, mmu_prop, hop_addr,
+							virt_addr);
 		hop_pte = hdev->asic_funcs->read_pte(hdev, hop_pte_addr);
 
-		offset_mask = OFFSET_MASK;
+		offset_mask = FLAGS_MASK;
 	}
 
 	if (!(hop_pte & PAGE_PRESENT_MASK))
@@ -608,6 +646,11 @@ static ssize_t hl_data_read32(struct file *f, char __user *buf,
 	u32 val;
 	ssize_t rc;
 
+	if (atomic_read(&hdev->in_reset)) {
+		dev_warn_ratelimited(hdev->dev, "Can't read during reset\n");
+		return 0;
+	}
+
 	if (*ppos)
 		return 0;
 
@@ -637,6 +680,11 @@ static ssize_t hl_data_write32(struct file *f, const char __user *buf,
 	u32 value;
 	ssize_t rc;
 
+	if (atomic_read(&hdev->in_reset)) {
+		dev_warn_ratelimited(hdev->dev, "Can't write during reset\n");
+		return 0;
+	}
+
 	rc = kstrtouint_from_user(buf, count, 16, &value);
 	if (rc)
 		return rc;
diff --git a/drivers/misc/habanalabs/device.c b/drivers/misc/habanalabs/device.c
index 459fee70a597..b155e9549076 100644
--- a/drivers/misc/habanalabs/device.c
+++ b/drivers/misc/habanalabs/device.c
@@ -42,12 +42,10 @@ static void hpriv_release(struct kref *ref)
 {
 	struct hl_fpriv *hpriv;
 	struct hl_device *hdev;
-	struct hl_ctx *ctx;
 
 	hpriv = container_of(ref, struct hl_fpriv, refcount);
 
 	hdev = hpriv->hdev;
-	ctx = hpriv->ctx;
 
 	put_pid(hpriv->taskpid);
 
@@ -889,13 +887,19 @@ again:
 	/* Go over all the queues, release all CS and their jobs */
 	hl_cs_rollback_all(hdev);
 
-	/* Kill processes here after CS rollback. This is because the process
-	 * can't really exit until all its CSs are done, which is what we
-	 * do in cs rollback
-	 */
-	if (from_hard_reset_thread)
+	if (hard_reset) {
+		/* Kill processes here after CS rollback. This is because the
+		 * process can't really exit until all its CSs are done, which
+		 * is what we do in cs rollback
+		 */
 		device_kill_open_processes(hdev);
 
+		/* Flush the Event queue workers to make sure no other thread is
+		 * reading or writing to registers during the reset
+		 */
+		flush_workqueue(hdev->eq_wq);
+	}
+
 	/* Release kernel context */
 	if ((hard_reset) && (hl_ctx_put(hdev->kernel_ctx) == 1))
 		hdev->kernel_ctx = NULL;
diff --git a/drivers/misc/habanalabs/firmware_if.c b/drivers/misc/habanalabs/firmware_if.c
index ea2ca67fbfbf..f5bd03171dac 100644
--- a/drivers/misc/habanalabs/firmware_if.c
+++ b/drivers/misc/habanalabs/firmware_if.c
@@ -143,10 +143,7 @@ int hl_fw_test_cpu_queue(struct hl_device *hdev)
 			sizeof(test_pkt), HL_DEVICE_TIMEOUT_USEC, &result);
 
 	if (!rc) {
-		if (result == ARMCP_PACKET_FENCE_VAL)
-			dev_info(hdev->dev,
-				"queue test on CPU queue succeeded\n");
-		else
+		if (result != ARMCP_PACKET_FENCE_VAL)
 			dev_err(hdev->dev,
 				"CPU queue test failed (0x%08lX)\n", result);
 	} else {
diff --git a/drivers/misc/habanalabs/goya/goya.c b/drivers/misc/habanalabs/goya/goya.c
index 6fba14b81f90..c8d16aa4382c 100644
--- a/drivers/misc/habanalabs/goya/goya.c
+++ b/drivers/misc/habanalabs/goya/goya.c
@@ -72,6 +72,9 @@
  *
  */
 
+#define GOYA_UBOOT_FW_FILE	"habanalabs/goya/goya-u-boot.bin"
+#define GOYA_LINUX_FW_FILE	"habanalabs/goya/goya-fit.itb"
+
 #define GOYA_MMU_REGS_NUM		63
 
 #define GOYA_DMA_POOL_BLK_SIZE		0x100		/* 256 bytes */
@@ -337,17 +340,20 @@ void goya_get_fixed_properties(struct hl_device *hdev)
 	for (i = 0 ; i < NUMBER_OF_EXT_HW_QUEUES ; i++) {
 		prop->hw_queues_props[i].type = QUEUE_TYPE_EXT;
 		prop->hw_queues_props[i].driver_only = 0;
+		prop->hw_queues_props[i].requires_kernel_cb = 1;
 	}
 
 	for (; i < NUMBER_OF_EXT_HW_QUEUES + NUMBER_OF_CPU_HW_QUEUES ; i++) {
 		prop->hw_queues_props[i].type = QUEUE_TYPE_CPU;
 		prop->hw_queues_props[i].driver_only = 1;
+		prop->hw_queues_props[i].requires_kernel_cb = 0;
 	}
 
 	for (; i < NUMBER_OF_EXT_HW_QUEUES + NUMBER_OF_CPU_HW_QUEUES +
 			NUMBER_OF_INT_HW_QUEUES; i++) {
 		prop->hw_queues_props[i].type = QUEUE_TYPE_INT;
 		prop->hw_queues_props[i].driver_only = 0;
+		prop->hw_queues_props[i].requires_kernel_cb = 0;
 	}
 
 	for (; i < HL_MAX_QUEUES; i++)
@@ -377,6 +383,23 @@ void goya_get_fixed_properties(struct hl_device *hdev)
 	prop->mmu_hop0_tables_total_size = HOP0_TABLES_TOTAL_SIZE;
 	prop->dram_page_size = PAGE_SIZE_2MB;
 
+	prop->dmmu.hop0_shift = HOP0_SHIFT;
+	prop->dmmu.hop1_shift = HOP1_SHIFT;
+	prop->dmmu.hop2_shift = HOP2_SHIFT;
+	prop->dmmu.hop3_shift = HOP3_SHIFT;
+	prop->dmmu.hop4_shift = HOP4_SHIFT;
+	prop->dmmu.hop0_mask = HOP0_MASK;
+	prop->dmmu.hop1_mask = HOP1_MASK;
+	prop->dmmu.hop2_mask = HOP2_MASK;
+	prop->dmmu.hop3_mask = HOP3_MASK;
+	prop->dmmu.hop4_mask = HOP4_MASK;
+	prop->dmmu.huge_page_size = PAGE_SIZE_2MB;
+
+	/* No difference between PMMU and DMMU except of page size */
+	memcpy(&prop->pmmu, &prop->dmmu, sizeof(prop->dmmu));
+	prop->dmmu.page_size = PAGE_SIZE_2MB;
+	prop->pmmu.page_size = PAGE_SIZE_4KB;
+
 	prop->va_space_host_start_address = VA_HOST_SPACE_START;
 	prop->va_space_host_end_address = VA_HOST_SPACE_END;
 	prop->va_space_dram_start_address = VA_DDR_SPACE_START;
@@ -393,6 +416,9 @@ void goya_get_fixed_properties(struct hl_device *hdev)
 	prop->tpc_enabled_mask = TPC_ENABLED_MASK;
 	prop->pcie_dbi_base_address = mmPCIE_DBI_BASE;
 	prop->pcie_aux_dbi_reg_addr = CFG_BASE + mmPCIE_AUX_DBI;
+
+	strncpy(prop->armcp_info.card_name, GOYA_DEFAULT_CARD_NAME,
+		CARD_NAME_MAX_LEN);
 }
 
 /*
@@ -1454,6 +1480,9 @@ static void goya_init_golden_registers(struct hl_device *hdev)
 				1 << TPC0_NRTR_SCRAMB_EN_VAL_SHIFT);
 		WREG32(mmTPC0_NRTR_NON_LIN_SCRAMB + offset,
 				1 << TPC0_NRTR_NON_LIN_SCRAMB_EN_SHIFT);
+
+		WREG32_FIELD(TPC0_CFG_MSS_CONFIG, offset,
+				ICACHE_FETCH_LINE_NUM, 2);
 	}
 
 	WREG32(mmDMA_NRTR_SCRAMB_EN, 1 << DMA_NRTR_SCRAMB_EN_VAL_SHIFT);
@@ -1533,7 +1562,6 @@ static void goya_init_mme_cmdq(struct hl_device *hdev)
 	u32 mtr_base_lo, mtr_base_hi;
 	u32 so_base_lo, so_base_hi;
 	u32 gic_base_lo, gic_base_hi;
-	u64 qman_base_addr;
 
 	mtr_base_lo = lower_32_bits(CFG_BASE + mmSYNC_MNGR_MON_PAY_ADDRL_0);
 	mtr_base_hi = upper_32_bits(CFG_BASE + mmSYNC_MNGR_MON_PAY_ADDRL_0);
@@ -1545,9 +1573,6 @@ static void goya_init_mme_cmdq(struct hl_device *hdev)
 	gic_base_hi =
 		upper_32_bits(CFG_BASE + mmGIC_DISTRIBUTOR__5_GICD_SETSPI_NSR);
 
-	qman_base_addr = hdev->asic_prop.sram_base_address +
-				MME_QMAN_BASE_OFFSET;
-
 	WREG32(mmMME_CMDQ_CP_MSG_BASE0_ADDR_LO, mtr_base_lo);
 	WREG32(mmMME_CMDQ_CP_MSG_BASE0_ADDR_HI, mtr_base_hi);
 	WREG32(mmMME_CMDQ_CP_MSG_BASE1_ADDR_LO,	so_base_lo);
@@ -2141,13 +2166,11 @@ static void goya_halt_engines(struct hl_device *hdev, bool hard_reset)
  */
 static int goya_push_uboot_to_device(struct hl_device *hdev)
 {
-	char fw_name[200];
 	void __iomem *dst;
 
-	snprintf(fw_name, sizeof(fw_name), "habanalabs/goya/goya-u-boot.bin");
 	dst = hdev->pcie_bar[SRAM_CFG_BAR_ID] + UBOOT_FW_OFFSET;
 
-	return hl_fw_push_fw_to_device(hdev, fw_name, dst);
+	return hl_fw_push_fw_to_device(hdev, GOYA_UBOOT_FW_FILE, dst);
 }
 
 /*
@@ -2160,13 +2183,11 @@ static int goya_push_uboot_to_device(struct hl_device *hdev)
  */
 static int goya_push_linux_to_device(struct hl_device *hdev)
 {
-	char fw_name[200];
 	void __iomem *dst;
 
-	snprintf(fw_name, sizeof(fw_name), "habanalabs/goya/goya-fit.itb");
 	dst = hdev->pcie_bar[DDR_BAR_ID] + LINUX_FW_OFFSET;
 
-	return hl_fw_push_fw_to_device(hdev, fw_name, dst);
+	return hl_fw_push_fw_to_device(hdev, GOYA_LINUX_FW_FILE, dst);
 }
 
 static int goya_pldm_init_cpu(struct hl_device *hdev)
@@ -2291,6 +2312,10 @@ static int goya_init_cpu(struct hl_device *hdev, u32 cpu_timeout)
 		10000,
 		cpu_timeout);
 
+	/* Read U-Boot version now in case we will later fail */
+	goya_read_device_fw_version(hdev, FW_COMP_UBOOT);
+	goya_read_device_fw_version(hdev, FW_COMP_PREBOOT);
+
 	if (rc) {
 		dev_err(hdev->dev, "Error in ARM u-boot!");
 		switch (status) {
@@ -2328,6 +2353,11 @@ static int goya_init_cpu(struct hl_device *hdev, u32 cpu_timeout)
 				"ARM status %d - u-boot stopped by user\n",
 				status);
 			break;
+		case CPU_BOOT_STATUS_TS_INIT_FAIL:
+			dev_err(hdev->dev,
+				"ARM status %d - Thermal Sensor initialization failed\n",
+				status);
+			break;
 		default:
 			dev_err(hdev->dev,
 				"ARM status %d - Invalid status code\n",
@@ -2337,10 +2367,6 @@ static int goya_init_cpu(struct hl_device *hdev, u32 cpu_timeout)
 		return -EIO;
 	}
 
-	/* Read U-Boot version now in case we will later fail */
-	goya_read_device_fw_version(hdev, FW_COMP_UBOOT);
-	goya_read_device_fw_version(hdev, FW_COMP_PREBOOT);
-
 	if (!hdev->fw_loading) {
 		dev_info(hdev->dev, "Skip loading FW\n");
 		goto out;
@@ -2453,7 +2479,8 @@ int goya_mmu_init(struct hl_device *hdev)
 	WREG32_AND(mmSTLB_STLB_FEATURE_EN,
 			(~STLB_STLB_FEATURE_EN_FOLLOWER_EN_MASK));
 
-	hdev->asic_funcs->mmu_invalidate_cache(hdev, true);
+	hdev->asic_funcs->mmu_invalidate_cache(hdev, true,
+					VM_TYPE_USERPTR | VM_TYPE_PHYS_PACK);
 
 	WREG32(mmMMU_MMU_ENABLE, 1);
 	WREG32(mmMMU_SPI_MASK, 0xF);
@@ -2978,9 +3005,6 @@ int goya_test_queue(struct hl_device *hdev, u32 hw_queue_id)
 			"H/W queue %d test failed (scratch(0x%08llX) == 0x%08X)\n",
 			hw_queue_id, (unsigned long long) fence_dma_addr, tmp);
 		rc = -EIO;
-	} else {
-		dev_info(hdev->dev, "queue test on H/W queue %d succeeded\n",
-			hw_queue_id);
 	}
 
 free_pkt:
@@ -3925,7 +3949,7 @@ static int goya_parse_cb_no_ext_queue(struct hl_device *hdev,
 		return 0;
 
 	dev_err(hdev->dev,
-		"Internal CB address %px + 0x%x is not in SRAM nor in DRAM\n",
+		"Internal CB address 0x%px + 0x%x is not in SRAM nor in DRAM\n",
 		parser->user_cb, parser->user_cb_size);
 
 	return -EFAULT;
@@ -3935,7 +3959,7 @@ int goya_cs_parser(struct hl_device *hdev, struct hl_cs_parser *parser)
 {
 	struct goya_device *goya = hdev->asic_specific;
 
-	if (!parser->ext_queue)
+	if (parser->queue_type == QUEUE_TYPE_INT)
 		return goya_parse_cb_no_ext_queue(hdev, parser);
 
 	if (goya->hw_cap_initialized & HW_CAP_MMU)
@@ -4606,7 +4630,7 @@ static int goya_memset_device_memory(struct hl_device *hdev, u64 addr, u64 size,
 		lin_dma_pkt++;
 	} while (--lin_dma_pkts_cnt);
 
-	job = hl_cs_allocate_job(hdev, true);
+	job = hl_cs_allocate_job(hdev, QUEUE_TYPE_EXT, true);
 	if (!job) {
 		dev_err(hdev->dev, "Failed to allocate a new job\n");
 		rc = -ENOMEM;
@@ -4835,13 +4859,15 @@ static void goya_mmu_prepare(struct hl_device *hdev, u32 asid)
 		goya_mmu_prepare_reg(hdev, goya_mmu_regs[i], asid);
 }
 
-static void goya_mmu_invalidate_cache(struct hl_device *hdev, bool is_hard)
+static void goya_mmu_invalidate_cache(struct hl_device *hdev, bool is_hard,
+					u32 flags)
 {
 	struct goya_device *goya = hdev->asic_specific;
 	u32 status, timeout_usec;
 	int rc;
 
-	if (!(goya->hw_cap_initialized & HW_CAP_MMU))
+	if (!(goya->hw_cap_initialized & HW_CAP_MMU) ||
+		hdev->hard_reset_pending)
 		return;
 
 	/* no need in L1 only invalidation in Goya */
@@ -4880,7 +4906,8 @@ static void goya_mmu_invalidate_cache_range(struct hl_device *hdev,
 	u32 status, timeout_usec, inv_data, pi;
 	int rc;
 
-	if (!(goya->hw_cap_initialized & HW_CAP_MMU))
+	if (!(goya->hw_cap_initialized & HW_CAP_MMU) ||
+		hdev->hard_reset_pending)
 		return;
 
 	/* no need in L1 only invalidation in Goya */
@@ -5137,7 +5164,8 @@ static const struct hl_asic_funcs goya_funcs = {
 	.init_iatu = goya_init_iatu,
 	.rreg = hl_rreg,
 	.wreg = hl_wreg,
-	.halt_coresight = goya_halt_coresight
+	.halt_coresight = goya_halt_coresight,
+	.get_clk_rate = goya_get_clk_rate
 };
 
 /*
diff --git a/drivers/misc/habanalabs/goya/goyaP.h b/drivers/misc/habanalabs/goya/goyaP.h
index 89b6574f8e4f..c3230cb6e25c 100644
--- a/drivers/misc/habanalabs/goya/goyaP.h
+++ b/drivers/misc/habanalabs/goya/goyaP.h
@@ -233,4 +233,6 @@ void goya_cpu_accessible_dma_pool_free(struct hl_device *hdev, size_t size,
 					void *vaddr);
 void goya_mmu_remove_device_cpu_mappings(struct hl_device *hdev);
 
+int goya_get_clk_rate(struct hl_device *hdev, u32 *cur_clk, u32 *max_clk);
+
 #endif /* GOYAP_H_ */
diff --git a/drivers/misc/habanalabs/goya/goya_coresight.c b/drivers/misc/habanalabs/goya/goya_coresight.c
index b4d406af1bed..c1ee6e2b5dff 100644
--- a/drivers/misc/habanalabs/goya/goya_coresight.c
+++ b/drivers/misc/habanalabs/goya/goya_coresight.c
@@ -8,6 +8,7 @@
 #include "goyaP.h"
 #include "include/goya/goya_coresight.h"
 #include "include/goya/asic_reg/goya_regs.h"
+#include "include/goya/asic_reg/goya_masks.h"
 
 #include <uapi/misc/habanalabs.h>
 
@@ -377,33 +378,32 @@ static int goya_config_etr(struct hl_device *hdev,
 		struct hl_debug_params *params)
 {
 	struct hl_debug_params_etr *input;
-	u64 base_reg = mmPSOC_ETR_BASE - CFG_BASE;
 	u32 val;
 	int rc;
 
-	WREG32(base_reg + 0xFB0, CORESIGHT_UNLOCK);
+	WREG32(mmPSOC_ETR_LAR, CORESIGHT_UNLOCK);
 
-	val = RREG32(base_reg + 0x304);
+	val = RREG32(mmPSOC_ETR_FFCR);
 	val |= 0x1000;
-	WREG32(base_reg + 0x304, val);
+	WREG32(mmPSOC_ETR_FFCR, val);
 	val |= 0x40;
-	WREG32(base_reg + 0x304, val);
+	WREG32(mmPSOC_ETR_FFCR, val);
 
-	rc = goya_coresight_timeout(hdev, base_reg + 0x304, 6, false);
+	rc = goya_coresight_timeout(hdev, mmPSOC_ETR_FFCR, 6, false);
 	if (rc) {
 		dev_err(hdev->dev, "Failed to %s ETR on timeout, error %d\n",
 				params->enable ? "enable" : "disable", rc);
 		return rc;
 	}
 
-	rc = goya_coresight_timeout(hdev, base_reg + 0xC, 2, true);
+	rc = goya_coresight_timeout(hdev, mmPSOC_ETR_STS, 2, true);
 	if (rc) {
 		dev_err(hdev->dev, "Failed to %s ETR on timeout, error %d\n",
 				params->enable ? "enable" : "disable", rc);
 		return rc;
 	}
 
-	WREG32(base_reg + 0x20, 0);
+	WREG32(mmPSOC_ETR_CTL, 0);
 
 	if (params->enable) {
 		input = params->input;
@@ -423,25 +423,26 @@ static int goya_config_etr(struct hl_device *hdev,
 			return -EINVAL;
 		}
 
-		WREG32(base_reg + 0x34, 0x3FFC);
-		WREG32(base_reg + 0x4, input->buffer_size);
-		WREG32(base_reg + 0x28, input->sink_mode);
-		WREG32(base_reg + 0x110, 0x700);
-		WREG32(base_reg + 0x118,
+		WREG32(mmPSOC_ETR_BUFWM, 0x3FFC);
+		WREG32(mmPSOC_ETR_RSZ, input->buffer_size);
+		WREG32(mmPSOC_ETR_MODE, input->sink_mode);
+		WREG32(mmPSOC_ETR_AXICTL,
+				0x700 | PSOC_ETR_AXICTL_PROTCTRLBIT1_SHIFT);
+		WREG32(mmPSOC_ETR_DBALO,
 				lower_32_bits(input->buffer_address));
-		WREG32(base_reg + 0x11C,
+		WREG32(mmPSOC_ETR_DBAHI,
 				upper_32_bits(input->buffer_address));
-		WREG32(base_reg + 0x304, 3);
-		WREG32(base_reg + 0x308, 0xA);
-		WREG32(base_reg + 0x20, 1);
+		WREG32(mmPSOC_ETR_FFCR, 3);
+		WREG32(mmPSOC_ETR_PSCR, 0xA);
+		WREG32(mmPSOC_ETR_CTL, 1);
 	} else {
-		WREG32(base_reg + 0x34, 0);
-		WREG32(base_reg + 0x4, 0x400);
-		WREG32(base_reg + 0x118, 0);
-		WREG32(base_reg + 0x11C, 0);
-		WREG32(base_reg + 0x308, 0);
-		WREG32(base_reg + 0x28, 0);
-		WREG32(base_reg + 0x304, 0);
+		WREG32(mmPSOC_ETR_BUFWM, 0);
+		WREG32(mmPSOC_ETR_RSZ, 0x400);
+		WREG32(mmPSOC_ETR_DBALO, 0);
+		WREG32(mmPSOC_ETR_DBAHI, 0);
+		WREG32(mmPSOC_ETR_PSCR, 0);
+		WREG32(mmPSOC_ETR_MODE, 0);
+		WREG32(mmPSOC_ETR_FFCR, 0);
 
 		if (params->output_size >= sizeof(u64)) {
 			u32 rwp, rwphi;
@@ -451,8 +452,8 @@ static int goya_config_etr(struct hl_device *hdev,
 			 * the buffer is set in the RWP register (lower 32
 			 * bits), and in the RWPHI register (upper 8 bits).
 			 */
-			rwp = RREG32(base_reg + 0x18);
-			rwphi = RREG32(base_reg + 0x3c) & 0xff;
+			rwp = RREG32(mmPSOC_ETR_RWP);
+			rwphi = RREG32(mmPSOC_ETR_RWPHI) & 0xff;
 			*(u64 *) params->output = ((u64) rwphi << 32) | rwp;
 		}
 	}
diff --git a/drivers/misc/habanalabs/goya/goya_hwmgr.c b/drivers/misc/habanalabs/goya/goya_hwmgr.c
index a2a700c3d597..b2ebc01e27f4 100644
--- a/drivers/misc/habanalabs/goya/goya_hwmgr.c
+++ b/drivers/misc/habanalabs/goya/goya_hwmgr.c
@@ -32,6 +32,37 @@ void goya_set_pll_profile(struct hl_device *hdev, enum hl_pll_frequency freq)
 	}
 }
 
+int goya_get_clk_rate(struct hl_device *hdev, u32 *cur_clk, u32 *max_clk)
+{
+	long value;
+
+	if (hl_device_disabled_or_in_reset(hdev))
+		return -ENODEV;
+
+	value = hl_get_frequency(hdev, MME_PLL, false);
+
+	if (value < 0) {
+		dev_err(hdev->dev, "Failed to retrieve device max clock %ld\n",
+			value);
+		return value;
+	}
+
+	*max_clk = (value / 1000 / 1000);
+
+	value = hl_get_frequency(hdev, MME_PLL, true);
+
+	if (value < 0) {
+		dev_err(hdev->dev,
+			"Failed to retrieve device current clock %ld\n",
+			value);
+		return value;
+	}
+
+	*cur_clk = (value / 1000 / 1000);
+
+	return 0;
+}
+
 static ssize_t mme_clk_show(struct device *dev, struct device_attribute *attr,
 				char *buf)
 {
diff --git a/drivers/misc/habanalabs/habanalabs.h b/drivers/misc/habanalabs/habanalabs.h
index 75862be53c60..00c949f4ccd1 100644
--- a/drivers/misc/habanalabs/habanalabs.h
+++ b/drivers/misc/habanalabs/habanalabs.h
@@ -40,8 +40,6 @@
 
 #define HL_MAX_QUEUES			128
 
-#define HL_MAX_JOBS_PER_CS		64
-
 /* MUST BE POWER OF 2 and larger than 1 */
 #define HL_MAX_PENDING_CS		64
 
@@ -85,12 +83,15 @@ struct hl_fpriv;
  * @QUEUE_TYPE_INT: internal queue that performs DMA inside the device's
  *			memories and/or operates the compute engines.
  * @QUEUE_TYPE_CPU: S/W queue for communication with the device's CPU.
+ * @QUEUE_TYPE_HW: queue of DMA and compute engines jobs, for which completion
+ *                 notifications are sent by H/W.
  */
 enum hl_queue_type {
 	QUEUE_TYPE_NA,
 	QUEUE_TYPE_EXT,
 	QUEUE_TYPE_INT,
-	QUEUE_TYPE_CPU
+	QUEUE_TYPE_CPU,
+	QUEUE_TYPE_HW
 };
 
 /**
@@ -98,10 +99,13 @@ enum hl_queue_type {
  * @type: queue type.
  * @driver_only: true if only the driver is allowed to send a job to this queue,
  *               false otherwise.
+ * @requires_kernel_cb: true if a CB handle must be provided for jobs on this
+ *                      queue, false otherwise (a CB address must be provided).
  */
 struct hw_queue_properties {
 	enum hl_queue_type	type;
 	u8			driver_only;
+	u8			requires_kernel_cb;
 };
 
 /**
@@ -110,8 +114,8 @@ struct hw_queue_properties {
  * @VM_TYPE_PHYS_PACK: mapping of DRAM memory to device virtual address.
  */
 enum vm_type_t {
-	VM_TYPE_USERPTR,
-	VM_TYPE_PHYS_PACK
+	VM_TYPE_USERPTR = 0x1,
+	VM_TYPE_PHYS_PACK = 0x2
 };
 
 /**
@@ -127,12 +131,44 @@ enum hl_device_hw_state {
 };
 
 /**
+ * struct hl_mmu_properties - ASIC specific MMU address translation properties.
+ * @hop0_shift: shift of hop 0 mask.
+ * @hop1_shift: shift of hop 1 mask.
+ * @hop2_shift: shift of hop 2 mask.
+ * @hop3_shift: shift of hop 3 mask.
+ * @hop4_shift: shift of hop 4 mask.
+ * @hop0_mask: mask to get the PTE address in hop 0.
+ * @hop1_mask: mask to get the PTE address in hop 1.
+ * @hop2_mask: mask to get the PTE address in hop 2.
+ * @hop3_mask: mask to get the PTE address in hop 3.
+ * @hop4_mask: mask to get the PTE address in hop 4.
+ * @page_size: default page size used to allocate memory.
+ * @huge_page_size: page size used to allocate memory with huge pages.
+ */
+struct hl_mmu_properties {
+	u64	hop0_shift;
+	u64	hop1_shift;
+	u64	hop2_shift;
+	u64	hop3_shift;
+	u64	hop4_shift;
+	u64	hop0_mask;
+	u64	hop1_mask;
+	u64	hop2_mask;
+	u64	hop3_mask;
+	u64	hop4_mask;
+	u32	page_size;
+	u32	huge_page_size;
+};
+
+/**
  * struct asic_fixed_properties - ASIC specific immutable properties.
  * @hw_queues_props: H/W queues properties.
  * @armcp_info: received various information from ArmCP regarding the H/W, e.g.
  *		available sensors.
  * @uboot_ver: F/W U-boot version.
  * @preboot_ver: F/W Preboot version.
+ * @dmmu: DRAM MMU address translation properties.
+ * @pmmu: PCI (host) MMU address translation properties.
  * @sram_base_address: SRAM physical start address.
  * @sram_end_address: SRAM physical end address.
  * @sram_user_base_address - SRAM physical start address for user access.
@@ -169,53 +205,55 @@ enum hl_device_hw_state {
  * @psoc_pci_pll_nf: PCI PLL NF value.
  * @psoc_pci_pll_od: PCI PLL OD value.
  * @psoc_pci_pll_div_factor: PCI PLL DIV FACTOR 1 value.
- * @completion_queues_count: number of completion queues.
  * @high_pll: high PLL frequency used by the device.
  * @cb_pool_cb_cnt: number of CBs in the CB pool.
  * @cb_pool_cb_size: size of each CB in the CB pool.
  * @tpc_enabled_mask: which TPCs are enabled.
+ * @completion_queues_count: number of completion queues.
  */
 struct asic_fixed_properties {
 	struct hw_queue_properties	hw_queues_props[HL_MAX_QUEUES];
-	struct armcp_info	armcp_info;
-	char			uboot_ver[VERSION_MAX_LEN];
-	char			preboot_ver[VERSION_MAX_LEN];
-	u64			sram_base_address;
-	u64			sram_end_address;
-	u64			sram_user_base_address;
-	u64			dram_base_address;
-	u64			dram_end_address;
-	u64			dram_user_base_address;
-	u64			dram_size;
-	u64			dram_pci_bar_size;
-	u64			max_power_default;
-	u64			va_space_host_start_address;
-	u64			va_space_host_end_address;
-	u64			va_space_dram_start_address;
-	u64			va_space_dram_end_address;
-	u64			dram_size_for_default_page_mapping;
-	u64			pcie_dbi_base_address;
-	u64			pcie_aux_dbi_reg_addr;
-	u64			mmu_pgt_addr;
-	u64			mmu_dram_default_page_addr;
-	u32			mmu_pgt_size;
-	u32			mmu_pte_size;
-	u32			mmu_hop_table_size;
-	u32			mmu_hop0_tables_total_size;
-	u32			dram_page_size;
-	u32			cfg_size;
-	u32			sram_size;
-	u32			max_asid;
-	u32			num_of_events;
-	u32			psoc_pci_pll_nr;
-	u32			psoc_pci_pll_nf;
-	u32			psoc_pci_pll_od;
-	u32			psoc_pci_pll_div_factor;
-	u32			high_pll;
-	u32			cb_pool_cb_cnt;
-	u32			cb_pool_cb_size;
-	u8			completion_queues_count;
-	u8			tpc_enabled_mask;
+	struct armcp_info		armcp_info;
+	char				uboot_ver[VERSION_MAX_LEN];
+	char				preboot_ver[VERSION_MAX_LEN];
+	struct hl_mmu_properties	dmmu;
+	struct hl_mmu_properties	pmmu;
+	u64				sram_base_address;
+	u64				sram_end_address;
+	u64				sram_user_base_address;
+	u64				dram_base_address;
+	u64				dram_end_address;
+	u64				dram_user_base_address;
+	u64				dram_size;
+	u64				dram_pci_bar_size;
+	u64				max_power_default;
+	u64				va_space_host_start_address;
+	u64				va_space_host_end_address;
+	u64				va_space_dram_start_address;
+	u64				va_space_dram_end_address;
+	u64				dram_size_for_default_page_mapping;
+	u64				pcie_dbi_base_address;
+	u64				pcie_aux_dbi_reg_addr;
+	u64				mmu_pgt_addr;
+	u64				mmu_dram_default_page_addr;
+	u32				mmu_pgt_size;
+	u32				mmu_pte_size;
+	u32				mmu_hop_table_size;
+	u32				mmu_hop0_tables_total_size;
+	u32				dram_page_size;
+	u32				cfg_size;
+	u32				sram_size;
+	u32				max_asid;
+	u32				num_of_events;
+	u32				psoc_pci_pll_nr;
+	u32				psoc_pci_pll_nf;
+	u32				psoc_pci_pll_od;
+	u32				psoc_pci_pll_div_factor;
+	u32				high_pll;
+	u32				cb_pool_cb_cnt;
+	u32				cb_pool_cb_size;
+	u8				tpc_enabled_mask;
+	u8				completion_queues_count;
 };
 
 /**
@@ -236,8 +274,6 @@ struct hl_dma_fence {
  * Command Buffers
  */
 
-#define HL_MAX_CB_SIZE		0x200000	/* 2MB */
-
 /**
  * struct hl_cb_mgr - describes a Command Buffer Manager.
  * @cb_lock: protects cb_handles.
@@ -481,8 +517,8 @@ enum hl_pll_frequency {
  * @get_events_stat: retrieve event queue entries histogram.
  * @read_pte: read MMU page table entry from DRAM.
  * @write_pte: write MMU page table entry to DRAM.
- * @mmu_invalidate_cache: flush MMU STLB cache, either with soft (L1 only) or
- *                        hard (L0 & L1) flush.
+ * @mmu_invalidate_cache: flush MMU STLB host/DRAM cache, either with soft
+ *                        (L1 only) or hard (L0 & L1) flush.
  * @mmu_invalidate_cache_range: flush specific MMU STLB cache lines with
  *                              ASID-VA-size mask.
  * @send_heartbeat: send is-alive packet to ArmCP and verify response.
@@ -502,6 +538,7 @@ enum hl_pll_frequency {
  * @rreg: Read a register. Needed for simulator support.
  * @wreg: Write a register. Needed for simulator support.
  * @halt_coresight: stop the ETF and ETR traces.
+ * @get_clk_rate: Retrieve the ASIC current and maximum clock rate in MHz
  */
 struct hl_asic_funcs {
 	int (*early_init)(struct hl_device *hdev);
@@ -562,7 +599,8 @@ struct hl_asic_funcs {
 				u32 *size);
 	u64 (*read_pte)(struct hl_device *hdev, u64 addr);
 	void (*write_pte)(struct hl_device *hdev, u64 addr, u64 val);
-	void (*mmu_invalidate_cache)(struct hl_device *hdev, bool is_hard);
+	void (*mmu_invalidate_cache)(struct hl_device *hdev, bool is_hard,
+					u32 flags);
 	void (*mmu_invalidate_cache_range)(struct hl_device *hdev, bool is_hard,
 			u32 asid, u64 va, u64 size);
 	int (*send_heartbeat)(struct hl_device *hdev);
@@ -584,6 +622,7 @@ struct hl_asic_funcs {
 	u32 (*rreg)(struct hl_device *hdev, u32 reg);
 	void (*wreg)(struct hl_device *hdev, u32 reg, u32 val);
 	void (*halt_coresight)(struct hl_device *hdev);
+	int (*get_clk_rate)(struct hl_device *hdev, u32 *cur_clk, u32 *max_clk);
 };
 
 
@@ -688,7 +727,7 @@ struct hl_ctx_mgr {
  * @sgt: pointer to the scatter-gather table that holds the pages.
  * @dir: for DMA unmapping, the direction must be supplied, so save it.
  * @debugfs_list: node in debugfs list of command submissions.
- * @addr: user-space virtual pointer to the start of the memory area.
+ * @addr: user-space virtual address of the start of the memory area.
  * @size: size of the memory area to pin & map.
  * @dma_mapped: true if the SG was mapped to DMA addresses, false otherwise.
  */
@@ -752,11 +791,14 @@ struct hl_cs {
  * @userptr_list: linked-list of userptr mappings that belong to this job and
  *			wait for completion.
  * @debugfs_list: node in debugfs list of command submission jobs.
+ * @queue_type: the type of the H/W queue this job is submitted to.
  * @id: the id of this job inside a CS.
  * @hw_queue_id: the id of the H/W queue this job is submitted to.
  * @user_cb_size: the actual size of the CB we got from the user.
  * @job_cb_size: the actual size of the CB that we put on the queue.
- * @ext_queue: whether the job is for external queue or internal queue.
+ * @is_kernel_allocated_cb: true if the CB handle we got from the user holds a
+ *                          handle to a kernel-allocated CB object, false
+ *                          otherwise (SRAM/DRAM/host address).
  */
 struct hl_cs_job {
 	struct list_head	cs_node;
@@ -766,39 +808,44 @@ struct hl_cs_job {
 	struct work_struct	finish_work;
 	struct list_head	userptr_list;
 	struct list_head	debugfs_list;
+	enum hl_queue_type	queue_type;
 	u32			id;
 	u32			hw_queue_id;
 	u32			user_cb_size;
 	u32			job_cb_size;
-	u8			ext_queue;
+	u8			is_kernel_allocated_cb;
 };
 
 /**
- * struct hl_cs_parser - command submission paerser properties.
+ * struct hl_cs_parser - command submission parser properties.
  * @user_cb: the CB we got from the user.
  * @patched_cb: in case of patching, this is internal CB which is submitted on
  *		the queue instead of the CB we got from the IOCTL.
  * @job_userptr_list: linked-list of userptr mappings that belong to the related
  *			job and wait for completion.
  * @cs_sequence: the sequence number of the related CS.
+ * @queue_type: the type of the H/W queue this job is submitted to.
  * @ctx_id: the ID of the context the related CS belongs to.
  * @hw_queue_id: the id of the H/W queue this job is submitted to.
  * @user_cb_size: the actual size of the CB we got from the user.
  * @patched_cb_size: the size of the CB after parsing.
- * @ext_queue: whether the job is for external queue or internal queue.
  * @job_id: the id of the related job inside the related CS.
+ * @is_kernel_allocated_cb: true if the CB handle we got from the user holds a
+ *                          handle to a kernel-allocated CB object, false
+ *                          otherwise (SRAM/DRAM/host address).
  */
 struct hl_cs_parser {
 	struct hl_cb		*user_cb;
 	struct hl_cb		*patched_cb;
 	struct list_head	*job_userptr_list;
 	u64			cs_sequence;
+	enum hl_queue_type	queue_type;
 	u32			ctx_id;
 	u32			hw_queue_id;
 	u32			user_cb_size;
 	u32			patched_cb_size;
-	u8			ext_queue;
 	u8			job_id;
+	u8			is_kernel_allocated_cb;
 };
 
 
@@ -1048,9 +1095,10 @@ void hl_wreg(struct hl_device *hdev, u32 reg, u32 val);
 
 #define REG_FIELD_SHIFT(reg, field) reg##_##field##_SHIFT
 #define REG_FIELD_MASK(reg, field) reg##_##field##_MASK
-#define WREG32_FIELD(reg, field, val)	\
-	WREG32(mm##reg, (RREG32(mm##reg) & ~REG_FIELD_MASK(reg, field)) | \
-			(val) << REG_FIELD_SHIFT(reg, field))
+#define WREG32_FIELD(reg, offset, field, val)	\
+	WREG32(mm##reg + offset, (RREG32(mm##reg + offset) & \
+				~REG_FIELD_MASK(reg, field)) | \
+				(val) << REG_FIELD_SHIFT(reg, field))
 
 /* Timeout should be longer when working with simulator but cap the
  * increased timeout to some maximum
@@ -1501,7 +1549,8 @@ int hl_cb_pool_init(struct hl_device *hdev);
 int hl_cb_pool_fini(struct hl_device *hdev);
 
 void hl_cs_rollback_all(struct hl_device *hdev);
-struct hl_cs_job *hl_cs_allocate_job(struct hl_device *hdev, bool ext_queue);
+struct hl_cs_job *hl_cs_allocate_job(struct hl_device *hdev,
+		enum hl_queue_type queue_type, bool is_kernel_allocated_cb);
 
 void goya_set_asic_funcs(struct hl_device *hdev);
 
@@ -1513,7 +1562,7 @@ void hl_vm_fini(struct hl_device *hdev);
 
 int hl_pin_host_memory(struct hl_device *hdev, u64 addr, u64 size,
 			struct hl_userptr *userptr);
-int hl_unpin_host_memory(struct hl_device *hdev, struct hl_userptr *userptr);
+void hl_unpin_host_memory(struct hl_device *hdev, struct hl_userptr *userptr);
 void hl_userptr_delete_list(struct hl_device *hdev,
 				struct list_head *userptr_list);
 bool hl_userptr_is_pinned(struct hl_device *hdev, u64 addr, u32 size,
diff --git a/drivers/misc/habanalabs/habanalabs_ioctl.c b/drivers/misc/habanalabs/habanalabs_ioctl.c
index 66d9c710073c..6474b868ef27 100644
--- a/drivers/misc/habanalabs/habanalabs_ioctl.c
+++ b/drivers/misc/habanalabs/habanalabs_ioctl.c
@@ -60,11 +60,16 @@ static int hw_ip_info(struct hl_device *hdev, struct hl_info_args *args)
 	hw_ip.tpc_enabled_mask = prop->tpc_enabled_mask;
 	hw_ip.sram_size = prop->sram_size - sram_kmd_size;
 	hw_ip.dram_size = prop->dram_size - dram_kmd_size;
-	if (hw_ip.dram_size > 0)
+	if (hw_ip.dram_size > PAGE_SIZE)
 		hw_ip.dram_enabled = 1;
 	hw_ip.num_of_events = prop->num_of_events;
-	memcpy(hw_ip.armcp_version,
-		prop->armcp_info.armcp_version, VERSION_MAX_LEN);
+
+	memcpy(hw_ip.armcp_version, prop->armcp_info.armcp_version,
+		min(VERSION_MAX_LEN, HL_INFO_VERSION_MAX_LEN));
+
+	memcpy(hw_ip.card_name, prop->armcp_info.card_name,
+		min(CARD_NAME_MAX_LEN, HL_INFO_CARD_NAME_MAX_LEN));
+
 	hw_ip.armcp_cpld_version = le32_to_cpu(prop->armcp_info.cpld_version);
 	hw_ip.psoc_pci_pll_nr = prop->psoc_pci_pll_nr;
 	hw_ip.psoc_pci_pll_nf = prop->psoc_pci_pll_nf;
@@ -179,17 +184,14 @@ static int debug_coresight(struct hl_device *hdev, struct hl_debug_args *args)
 		goto out;
 	}
 
-	if (output) {
-		if (copy_to_user((void __user *) (uintptr_t) args->output_ptr,
-					output,
-					args->output_size)) {
-			dev_err(hdev->dev,
-				"copy to user failed in debug ioctl\n");
-			rc = -EFAULT;
-			goto out;
-		}
+	if (output && copy_to_user((void __user *) (uintptr_t) args->output_ptr,
+					output, args->output_size)) {
+		dev_err(hdev->dev, "copy to user failed in debug ioctl\n");
+		rc = -EFAULT;
+		goto out;
 	}
 
+
 out:
 	kfree(params);
 	kfree(output);
@@ -221,6 +223,41 @@ static int device_utilization(struct hl_device *hdev, struct hl_info_args *args)
 		min((size_t) max_size, sizeof(device_util))) ? -EFAULT : 0;
 }
 
+static int get_clk_rate(struct hl_device *hdev, struct hl_info_args *args)
+{
+	struct hl_info_clk_rate clk_rate = {0};
+	u32 max_size = args->return_size;
+	void __user *out = (void __user *) (uintptr_t) args->return_pointer;
+	int rc;
+
+	if ((!max_size) || (!out))
+		return -EINVAL;
+
+	rc = hdev->asic_funcs->get_clk_rate(hdev, &clk_rate.cur_clk_rate_mhz,
+						&clk_rate.max_clk_rate_mhz);
+	if (rc)
+		return rc;
+
+	return copy_to_user(out, &clk_rate,
+		min((size_t) max_size, sizeof(clk_rate))) ? -EFAULT : 0;
+}
+
+static int get_reset_count(struct hl_device *hdev, struct hl_info_args *args)
+{
+	struct hl_info_reset_count reset_count = {0};
+	u32 max_size = args->return_size;
+	void __user *out = (void __user *) (uintptr_t) args->return_pointer;
+
+	if ((!max_size) || (!out))
+		return -EINVAL;
+
+	reset_count.hard_reset_cnt = hdev->hard_reset_cnt;
+	reset_count.soft_reset_cnt = hdev->soft_reset_cnt;
+
+	return copy_to_user(out, &reset_count,
+		min((size_t) max_size, sizeof(reset_count))) ? -EFAULT : 0;
+}
+
 static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data,
 				struct device *dev)
 {
@@ -239,6 +276,9 @@ static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data,
 	case HL_INFO_DEVICE_STATUS:
 		return device_status_info(hdev, args);
 
+	case HL_INFO_RESET_COUNT:
+		return get_reset_count(hdev, args);
+
 	default:
 		break;
 	}
@@ -271,6 +311,10 @@ static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data,
 		rc = hw_events_info(hdev, true, args);
 		break;
 
+	case HL_INFO_CLK_RATE:
+		rc = get_clk_rate(hdev, args);
+		break;
+
 	default:
 		dev_err(dev, "Invalid request %d\n", args->op);
 		rc = -ENOTTY;
@@ -406,9 +450,8 @@ static long _hl_ioctl(struct file *filep, unsigned int cmd, unsigned long arg,
 
 	retcode = func(hpriv, kdata);
 
-	if (cmd & IOC_OUT)
-		if (copy_to_user((void __user *)arg, kdata, usize))
-			retcode = -EFAULT;
+	if ((cmd & IOC_OUT) && copy_to_user((void __user *)arg, kdata, usize))
+		retcode = -EFAULT;
 
 out_err:
 	if (retcode)
diff --git a/drivers/misc/habanalabs/hw_queue.c b/drivers/misc/habanalabs/hw_queue.c
index 55b383b2a116..91579dde9262 100644
--- a/drivers/misc/habanalabs/hw_queue.c
+++ b/drivers/misc/habanalabs/hw_queue.c
@@ -58,8 +58,8 @@ out:
 }
 
 /*
- * ext_queue_submit_bd - Submit a buffer descriptor to an external queue
- *
+ * ext_and_hw_queue_submit_bd() - Submit a buffer descriptor to an external or a
+ *                                H/W queue.
  * @hdev: pointer to habanalabs device structure
  * @q: pointer to habanalabs queue structure
  * @ctl: BD's control word
@@ -73,8 +73,8 @@ out:
  * This function must be called when the scheduler mutex is taken
  *
  */
-static void ext_queue_submit_bd(struct hl_device *hdev, struct hl_hw_queue *q,
-				u32 ctl, u32 len, u64 ptr)
+static void ext_and_hw_queue_submit_bd(struct hl_device *hdev,
+			struct hl_hw_queue *q, u32 ctl, u32 len, u64 ptr)
 {
 	struct hl_bd *bd;
 
@@ -174,6 +174,45 @@ static int int_queue_sanity_checks(struct hl_device *hdev,
 }
 
 /*
+ * hw_queue_sanity_checks() - Perform some sanity checks on a H/W queue.
+ * @hdev: Pointer to hl_device structure.
+ * @q: Pointer to hl_hw_queue structure.
+ * @num_of_entries: How many entries to check for space.
+ *
+ * Perform the following:
+ * - Make sure we have enough space in the completion queue.
+ *   This check also ensures that there is enough space in the h/w queue, as
+ *   both queues are of the same size.
+ * - Reserve space in the completion queue (needs to be reversed if there
+ *   is a failure down the road before the actual submission of work).
+ *
+ * Both operations are done using the "free_slots_cnt" field of the completion
+ * queue. The CI counters of the queue and the completion queue are not
+ * needed/used for the H/W queue type.
+ */
+static int hw_queue_sanity_checks(struct hl_device *hdev, struct hl_hw_queue *q,
+					int num_of_entries)
+{
+	atomic_t *free_slots =
+			&hdev->completion_queue[q->hw_queue_id].free_slots_cnt;
+
+	/*
+	 * Check we have enough space in the completion queue.
+	 * Add -1 to counter (decrement) unless counter was already 0.
+	 * In that case, CQ is full so we can't submit a new CB.
+	 * atomic_add_unless will return 0 if counter was already 0.
+	 */
+	if (atomic_add_negative(num_of_entries * -1, free_slots)) {
+		dev_dbg(hdev->dev, "No space for %d entries on CQ %d\n",
+			num_of_entries, q->hw_queue_id);
+		atomic_add(num_of_entries, free_slots);
+		return -EAGAIN;
+	}
+
+	return 0;
+}
+
+/*
  * hl_hw_queue_send_cb_no_cmpl - send a single CB (not a JOB) without completion
  *
  * @hdev: pointer to hl_device structure
@@ -188,7 +227,7 @@ int hl_hw_queue_send_cb_no_cmpl(struct hl_device *hdev, u32 hw_queue_id,
 				u32 cb_size, u64 cb_ptr)
 {
 	struct hl_hw_queue *q = &hdev->kernel_queues[hw_queue_id];
-	int rc;
+	int rc = 0;
 
 	/*
 	 * The CPU queue is a synchronous queue with an effective depth of
@@ -206,11 +245,18 @@ int hl_hw_queue_send_cb_no_cmpl(struct hl_device *hdev, u32 hw_queue_id,
 		goto out;
 	}
 
-	rc = ext_queue_sanity_checks(hdev, q, 1, false);
-	if (rc)
-		goto out;
+	/*
+	 * hl_hw_queue_send_cb_no_cmpl() is called for queues of a H/W queue
+	 * type only on init phase, when the queues are empty and being tested,
+	 * so there is no need for sanity checks.
+	 */
+	if (q->queue_type != QUEUE_TYPE_HW) {
+		rc = ext_queue_sanity_checks(hdev, q, 1, false);
+		if (rc)
+			goto out;
+	}
 
-	ext_queue_submit_bd(hdev, q, 0, cb_size, cb_ptr);
+	ext_and_hw_queue_submit_bd(hdev, q, 0, cb_size, cb_ptr);
 
 out:
 	if (q->queue_type != QUEUE_TYPE_CPU)
@@ -220,14 +266,14 @@ out:
 }
 
 /*
- * ext_hw_queue_schedule_job - submit an JOB to an external queue
+ * ext_queue_schedule_job - submit a JOB to an external queue
  *
  * @job: pointer to the job that needs to be submitted to the queue
  *
  * This function must be called when the scheduler mutex is taken
  *
  */
-static void ext_hw_queue_schedule_job(struct hl_cs_job *job)
+static void ext_queue_schedule_job(struct hl_cs_job *job)
 {
 	struct hl_device *hdev = job->cs->ctx->hdev;
 	struct hl_hw_queue *q = &hdev->kernel_queues[job->hw_queue_id];
@@ -260,7 +306,7 @@ static void ext_hw_queue_schedule_job(struct hl_cs_job *job)
 	 * H/W queues is done under the scheduler mutex
 	 *
 	 * No need to check if CQ is full because it was already
-	 * checked in hl_queue_sanity_checks
+	 * checked in ext_queue_sanity_checks
 	 */
 	cq = &hdev->completion_queue[q->hw_queue_id];
 	cq_addr = cq->bus_address + cq->pi * sizeof(struct hl_cq_entry);
@@ -274,18 +320,18 @@ static void ext_hw_queue_schedule_job(struct hl_cs_job *job)
 
 	cq->pi = hl_cq_inc_ptr(cq->pi);
 
-	ext_queue_submit_bd(hdev, q, ctl, len, ptr);
+	ext_and_hw_queue_submit_bd(hdev, q, ctl, len, ptr);
 }
 
 /*
- * int_hw_queue_schedule_job - submit an JOB to an internal queue
+ * int_queue_schedule_job - submit a JOB to an internal queue
  *
  * @job: pointer to the job that needs to be submitted to the queue
  *
  * This function must be called when the scheduler mutex is taken
  *
  */
-static void int_hw_queue_schedule_job(struct hl_cs_job *job)
+static void int_queue_schedule_job(struct hl_cs_job *job)
 {
 	struct hl_device *hdev = job->cs->ctx->hdev;
 	struct hl_hw_queue *q = &hdev->kernel_queues[job->hw_queue_id];
@@ -308,6 +354,60 @@ static void int_hw_queue_schedule_job(struct hl_cs_job *job)
 }
 
 /*
+ * hw_queue_schedule_job - submit a JOB to a H/W queue
+ *
+ * @job: pointer to the job that needs to be submitted to the queue
+ *
+ * This function must be called when the scheduler mutex is taken
+ *
+ */
+static void hw_queue_schedule_job(struct hl_cs_job *job)
+{
+	struct hl_device *hdev = job->cs->ctx->hdev;
+	struct hl_hw_queue *q = &hdev->kernel_queues[job->hw_queue_id];
+	struct hl_cq *cq;
+	u64 ptr;
+	u32 offset, ctl, len;
+
+	/*
+	 * Upon PQE completion, COMP_DATA is used as the write data to the
+	 * completion queue (QMAN HBW message), and COMP_OFFSET is used as the
+	 * write address offset in the SM block (QMAN LBW message).
+	 * The write address offset is calculated as "COMP_OFFSET << 2".
+	 */
+	offset = job->cs->sequence & (HL_MAX_PENDING_CS - 1);
+	ctl = ((offset << BD_CTL_COMP_OFFSET_SHIFT) & BD_CTL_COMP_OFFSET_MASK) |
+		((q->pi << BD_CTL_COMP_DATA_SHIFT) & BD_CTL_COMP_DATA_MASK);
+
+	len = job->job_cb_size;
+
+	/*
+	 * A patched CB is created only if a user CB was allocated by driver and
+	 * MMU is disabled. If MMU is enabled, the user CB should be used
+	 * instead. If the user CB wasn't allocated by driver, assume that it
+	 * holds an address.
+	 */
+	if (job->patched_cb)
+		ptr = job->patched_cb->bus_address;
+	else if (job->is_kernel_allocated_cb)
+		ptr = job->user_cb->bus_address;
+	else
+		ptr = (u64) (uintptr_t) job->user_cb;
+
+	/*
+	 * No need to protect pi_offset because scheduling to the
+	 * H/W queues is done under the scheduler mutex
+	 *
+	 * No need to check if CQ is full because it was already
+	 * checked in hw_queue_sanity_checks
+	 */
+	cq = &hdev->completion_queue[q->hw_queue_id];
+	cq->pi = hl_cq_inc_ptr(cq->pi);
+
+	ext_and_hw_queue_submit_bd(hdev, q, ctl, len, ptr);
+}
+
+/*
  * hl_hw_queue_schedule_cs - schedule a command submission
  *
  * @job        : pointer to the CS
@@ -330,23 +430,34 @@ int hl_hw_queue_schedule_cs(struct hl_cs *cs)
 	}
 
 	q = &hdev->kernel_queues[0];
-	/* This loop assumes all external queues are consecutive */
 	for (i = 0, cq_cnt = 0 ; i < HL_MAX_QUEUES ; i++, q++) {
-		if (q->queue_type == QUEUE_TYPE_EXT) {
-			if (cs->jobs_in_queue_cnt[i]) {
+		if (cs->jobs_in_queue_cnt[i]) {
+			switch (q->queue_type) {
+			case QUEUE_TYPE_EXT:
 				rc = ext_queue_sanity_checks(hdev, q,
-					cs->jobs_in_queue_cnt[i], true);
-				if (rc)
-					goto unroll_cq_resv;
-				cq_cnt++;
-			}
-		} else if (q->queue_type == QUEUE_TYPE_INT) {
-			if (cs->jobs_in_queue_cnt[i]) {
+						cs->jobs_in_queue_cnt[i], true);
+				break;
+			case QUEUE_TYPE_INT:
 				rc = int_queue_sanity_checks(hdev, q,
-					cs->jobs_in_queue_cnt[i]);
-				if (rc)
-					goto unroll_cq_resv;
+						cs->jobs_in_queue_cnt[i]);
+				break;
+			case QUEUE_TYPE_HW:
+				rc = hw_queue_sanity_checks(hdev, q,
+						cs->jobs_in_queue_cnt[i]);
+				break;
+			default:
+				dev_err(hdev->dev, "Queue type %d is invalid\n",
+					q->queue_type);
+				rc = -EINVAL;
+				break;
 			}
+
+			if (rc)
+				goto unroll_cq_resv;
+
+			if (q->queue_type == QUEUE_TYPE_EXT ||
+					q->queue_type == QUEUE_TYPE_HW)
+				cq_cnt++;
 		}
 	}
 
@@ -373,21 +484,30 @@ int hl_hw_queue_schedule_cs(struct hl_cs *cs)
 	}
 
 	list_for_each_entry_safe(job, tmp, &cs->job_list, cs_node)
-		if (job->ext_queue)
-			ext_hw_queue_schedule_job(job);
-		else
-			int_hw_queue_schedule_job(job);
+		switch (job->queue_type) {
+		case QUEUE_TYPE_EXT:
+			ext_queue_schedule_job(job);
+			break;
+		case QUEUE_TYPE_INT:
+			int_queue_schedule_job(job);
+			break;
+		case QUEUE_TYPE_HW:
+			hw_queue_schedule_job(job);
+			break;
+		default:
+			break;
+		}
 
 	cs->submitted = true;
 
 	goto out;
 
 unroll_cq_resv:
-	/* This loop assumes all external queues are consecutive */
 	q = &hdev->kernel_queues[0];
 	for (i = 0 ; (i < HL_MAX_QUEUES) && (cq_cnt > 0) ; i++, q++) {
-		if ((q->queue_type == QUEUE_TYPE_EXT) &&
-				(cs->jobs_in_queue_cnt[i])) {
+		if ((q->queue_type == QUEUE_TYPE_EXT ||
+				q->queue_type == QUEUE_TYPE_HW) &&
+				cs->jobs_in_queue_cnt[i]) {
 			atomic_t *free_slots =
 				&hdev->completion_queue[i].free_slots_cnt;
 			atomic_add(cs->jobs_in_queue_cnt[i], free_slots);
@@ -414,8 +534,8 @@ void hl_hw_queue_inc_ci_kernel(struct hl_device *hdev, u32 hw_queue_id)
 	q->ci = hl_queue_inc_ptr(q->ci);
 }
 
-static int ext_and_cpu_hw_queue_init(struct hl_device *hdev,
-				struct hl_hw_queue *q, bool is_cpu_queue)
+static int ext_and_cpu_queue_init(struct hl_device *hdev, struct hl_hw_queue *q,
+					bool is_cpu_queue)
 {
 	void *p;
 	int rc;
@@ -465,7 +585,7 @@ free_queue:
 	return rc;
 }
 
-static int int_hw_queue_init(struct hl_device *hdev, struct hl_hw_queue *q)
+static int int_queue_init(struct hl_device *hdev, struct hl_hw_queue *q)
 {
 	void *p;
 
@@ -485,18 +605,38 @@ static int int_hw_queue_init(struct hl_device *hdev, struct hl_hw_queue *q)
 	return 0;
 }
 
-static int cpu_hw_queue_init(struct hl_device *hdev, struct hl_hw_queue *q)
+static int cpu_queue_init(struct hl_device *hdev, struct hl_hw_queue *q)
+{
+	return ext_and_cpu_queue_init(hdev, q, true);
+}
+
+static int ext_queue_init(struct hl_device *hdev, struct hl_hw_queue *q)
 {
-	return ext_and_cpu_hw_queue_init(hdev, q, true);
+	return ext_and_cpu_queue_init(hdev, q, false);
 }
 
-static int ext_hw_queue_init(struct hl_device *hdev, struct hl_hw_queue *q)
+static int hw_queue_init(struct hl_device *hdev, struct hl_hw_queue *q)
 {
-	return ext_and_cpu_hw_queue_init(hdev, q, false);
+	void *p;
+
+	p = hdev->asic_funcs->asic_dma_alloc_coherent(hdev,
+						HL_QUEUE_SIZE_IN_BYTES,
+						&q->bus_address,
+						GFP_KERNEL | __GFP_ZERO);
+	if (!p)
+		return -ENOMEM;
+
+	q->kernel_address = (u64) (uintptr_t) p;
+
+	/* Make sure read/write pointers are initialized to start of queue */
+	q->ci = 0;
+	q->pi = 0;
+
+	return 0;
 }
 
 /*
- * hw_queue_init - main initialization function for H/W queue object
+ * queue_init - main initialization function for H/W queue object
  *
  * @hdev: pointer to hl_device device structure
  * @q: pointer to hl_hw_queue queue structure
@@ -505,7 +645,7 @@ static int ext_hw_queue_init(struct hl_device *hdev, struct hl_hw_queue *q)
  * Allocate dma-able memory for the queue and initialize fields
  * Returns 0 on success
  */
-static int hw_queue_init(struct hl_device *hdev, struct hl_hw_queue *q,
+static int queue_init(struct hl_device *hdev, struct hl_hw_queue *q,
 			u32 hw_queue_id)
 {
 	int rc;
@@ -516,21 +656,20 @@ static int hw_queue_init(struct hl_device *hdev, struct hl_hw_queue *q,
 
 	switch (q->queue_type) {
 	case QUEUE_TYPE_EXT:
-		rc = ext_hw_queue_init(hdev, q);
+		rc = ext_queue_init(hdev, q);
 		break;
-
 	case QUEUE_TYPE_INT:
-		rc = int_hw_queue_init(hdev, q);
+		rc = int_queue_init(hdev, q);
 		break;
-
 	case QUEUE_TYPE_CPU:
-		rc = cpu_hw_queue_init(hdev, q);
+		rc = cpu_queue_init(hdev, q);
+		break;
+	case QUEUE_TYPE_HW:
+		rc = hw_queue_init(hdev, q);
 		break;
-
 	case QUEUE_TYPE_NA:
 		q->valid = 0;
 		return 0;
-
 	default:
 		dev_crit(hdev->dev, "wrong queue type %d during init\n",
 			q->queue_type);
@@ -554,7 +693,7 @@ static int hw_queue_init(struct hl_device *hdev, struct hl_hw_queue *q,
  *
  * Free the queue memory
  */
-static void hw_queue_fini(struct hl_device *hdev, struct hl_hw_queue *q)
+static void queue_fini(struct hl_device *hdev, struct hl_hw_queue *q)
 {
 	if (!q->valid)
 		return;
@@ -612,7 +751,7 @@ int hl_hw_queues_create(struct hl_device *hdev)
 			i < HL_MAX_QUEUES ; i++, q_ready_cnt++, q++) {
 
 		q->queue_type = asic->hw_queues_props[i].type;
-		rc = hw_queue_init(hdev, q, i);
+		rc = queue_init(hdev, q, i);
 		if (rc) {
 			dev_err(hdev->dev,
 				"failed to initialize queue %d\n", i);
@@ -624,7 +763,7 @@ int hl_hw_queues_create(struct hl_device *hdev)
 
 release_queues:
 	for (i = 0, q = hdev->kernel_queues ; i < q_ready_cnt ; i++, q++)
-		hw_queue_fini(hdev, q);
+		queue_fini(hdev, q);
 
 	kfree(hdev->kernel_queues);
 
@@ -637,7 +776,7 @@ void hl_hw_queues_destroy(struct hl_device *hdev)
 	int i;
 
 	for (i = 0, q = hdev->kernel_queues ; i < HL_MAX_QUEUES ; i++, q++)
-		hw_queue_fini(hdev, q);
+		queue_fini(hdev, q);
 
 	kfree(hdev->kernel_queues);
 }
diff --git a/drivers/misc/habanalabs/include/goya/asic_reg/goya_masks.h b/drivers/misc/habanalabs/include/goya/asic_reg/goya_masks.h
index 8618891d5afa..3c44ef3a23ed 100644
--- a/drivers/misc/habanalabs/include/goya/asic_reg/goya_masks.h
+++ b/drivers/misc/habanalabs/include/goya/asic_reg/goya_masks.h
@@ -260,4 +260,6 @@
 #define DMA_QM_3_GLBL_CFG1_DMA_STOP_SHIFT DMA_QM_0_GLBL_CFG1_DMA_STOP_SHIFT
 #define DMA_QM_4_GLBL_CFG1_DMA_STOP_SHIFT DMA_QM_0_GLBL_CFG1_DMA_STOP_SHIFT
 
+#define PSOC_ETR_AXICTL_PROTCTRLBIT1_SHIFT                           1
+
 #endif /* ASIC_REG_GOYA_MASKS_H_ */
diff --git a/drivers/misc/habanalabs/include/goya/asic_reg/goya_regs.h b/drivers/misc/habanalabs/include/goya/asic_reg/goya_regs.h
index 19b0f0ef1d0b..fce490e6a231 100644
--- a/drivers/misc/habanalabs/include/goya/asic_reg/goya_regs.h
+++ b/drivers/misc/habanalabs/include/goya/asic_reg/goya_regs.h
@@ -84,6 +84,7 @@
 #include "tpc6_rtr_regs.h"
 #include "tpc7_nrtr_regs.h"
 #include "tpc0_eml_cfg_regs.h"
+#include "psoc_etr_regs.h"
 
 #include "psoc_global_conf_masks.h"
 #include "dma_macro_masks.h"
diff --git a/drivers/misc/habanalabs/include/goya/asic_reg/psoc_etr_regs.h b/drivers/misc/habanalabs/include/goya/asic_reg/psoc_etr_regs.h
new file mode 100644
index 000000000000..b7c33e025db5
--- /dev/null
+++ b/drivers/misc/habanalabs/include/goya/asic_reg/psoc_etr_regs.h
@@ -0,0 +1,114 @@
+/* SPDX-License-Identifier: GPL-2.0
+ *
+ * Copyright 2016-2018 HabanaLabs, Ltd.
+ * All Rights Reserved.
+ *
+ */
+
+/************************************
+ ** This is an auto-generated file **
+ **       DO NOT EDIT BELOW        **
+ ************************************/
+
+#ifndef ASIC_REG_PSOC_ETR_REGS_H_
+#define ASIC_REG_PSOC_ETR_REGS_H_
+
+/*
+ *****************************************
+ *   PSOC_ETR (Prototype: ETR)
+ *****************************************
+ */
+
+#define mmPSOC_ETR_RSZ                                               0x2C43004
+
+#define mmPSOC_ETR_STS                                               0x2C4300C
+
+#define mmPSOC_ETR_RRD                                               0x2C43010
+
+#define mmPSOC_ETR_RRP                                               0x2C43014
+
+#define mmPSOC_ETR_RWP                                               0x2C43018
+
+#define mmPSOC_ETR_TRG                                               0x2C4301C
+
+#define mmPSOC_ETR_CTL                                               0x2C43020
+
+#define mmPSOC_ETR_RWD                                               0x2C43024
+
+#define mmPSOC_ETR_MODE                                              0x2C43028
+
+#define mmPSOC_ETR_LBUFLEVEL                                         0x2C4302C
+
+#define mmPSOC_ETR_CBUFLEVEL                                         0x2C43030
+
+#define mmPSOC_ETR_BUFWM                                             0x2C43034
+
+#define mmPSOC_ETR_RRPHI                                             0x2C43038
+
+#define mmPSOC_ETR_RWPHI                                             0x2C4303C
+
+#define mmPSOC_ETR_AXICTL                                            0x2C43110
+
+#define mmPSOC_ETR_DBALO                                             0x2C43118
+
+#define mmPSOC_ETR_DBAHI                                             0x2C4311C
+
+#define mmPSOC_ETR_FFSR                                              0x2C43300
+
+#define mmPSOC_ETR_FFCR                                              0x2C43304
+
+#define mmPSOC_ETR_PSCR                                              0x2C43308
+
+#define mmPSOC_ETR_ITMISCOP0                                         0x2C43EE0
+
+#define mmPSOC_ETR_ITTRFLIN                                          0x2C43EE8
+
+#define mmPSOC_ETR_ITATBDATA0                                        0x2C43EEC
+
+#define mmPSOC_ETR_ITATBCTR2                                         0x2C43EF0
+
+#define mmPSOC_ETR_ITATBCTR1                                         0x2C43EF4
+
+#define mmPSOC_ETR_ITATBCTR0                                         0x2C43EF8
+
+#define mmPSOC_ETR_ITCTRL                                            0x2C43F00
+
+#define mmPSOC_ETR_CLAIMSET                                          0x2C43FA0
+
+#define mmPSOC_ETR_CLAIMCLR                                          0x2C43FA4
+
+#define mmPSOC_ETR_LAR                                               0x2C43FB0
+
+#define mmPSOC_ETR_LSR                                               0x2C43FB4
+
+#define mmPSOC_ETR_AUTHSTATUS                                        0x2C43FB8
+
+#define mmPSOC_ETR_DEVID                                             0x2C43FC8
+
+#define mmPSOC_ETR_DEVTYPE                                           0x2C43FCC
+
+#define mmPSOC_ETR_PERIPHID4                                         0x2C43FD0
+
+#define mmPSOC_ETR_PERIPHID5                                         0x2C43FD4
+
+#define mmPSOC_ETR_PERIPHID6                                         0x2C43FD8
+
+#define mmPSOC_ETR_PERIPHID7                                         0x2C43FDC
+
+#define mmPSOC_ETR_PERIPHID0                                         0x2C43FE0
+
+#define mmPSOC_ETR_PERIPHID1                                         0x2C43FE4
+
+#define mmPSOC_ETR_PERIPHID2                                         0x2C43FE8
+
+#define mmPSOC_ETR_PERIPHID3                                         0x2C43FEC
+
+#define mmPSOC_ETR_COMPID0                                           0x2C43FF0
+
+#define mmPSOC_ETR_COMPID1                                           0x2C43FF4
+
+#define mmPSOC_ETR_COMPID2                                           0x2C43FF8
+
+#define mmPSOC_ETR_COMPID3                                           0x2C43FFC
+
+#endif /* ASIC_REG_PSOC_ETR_REGS_H_ */
diff --git a/drivers/misc/habanalabs/include/hl_boot_if.h b/drivers/misc/habanalabs/include/hl_boot_if.h
index 4cd04c090285..2853a2de8cf6 100644
--- a/drivers/misc/habanalabs/include/hl_boot_if.h
+++ b/drivers/misc/habanalabs/include/hl_boot_if.h
@@ -20,6 +20,8 @@ enum cpu_boot_status {
 	CPU_BOOT_STATUS_DRAM_INIT_FAIL,
 	CPU_BOOT_STATUS_FIT_CORRUPTED,
 	CPU_BOOT_STATUS_UBOOT_NOT_READY,
+	CPU_BOOT_STATUS_RESERVED,
+	CPU_BOOT_STATUS_TS_INIT_FAIL,
 };
 
 enum kmd_msg {
diff --git a/drivers/misc/habanalabs/include/hw_ip/mmu/mmu_general.h b/drivers/misc/habanalabs/include/hw_ip/mmu/mmu_general.h
index 71ea3c3e8ba3..a6851a9d3f03 100644
--- a/drivers/misc/habanalabs/include/hw_ip/mmu/mmu_general.h
+++ b/drivers/misc/habanalabs/include/hw_ip/mmu/mmu_general.h
@@ -12,18 +12,16 @@
 #define PAGE_SHIFT_2MB			21
 #define PAGE_SIZE_2MB			(_AC(1, UL) << PAGE_SHIFT_2MB)
 #define PAGE_SIZE_4KB			(_AC(1, UL) << PAGE_SHIFT_4KB)
-#define PAGE_MASK_2MB			(~(PAGE_SIZE_2MB - 1))
 
 #define PAGE_PRESENT_MASK		0x0000000000001ull
 #define SWAP_OUT_MASK			0x0000000000004ull
 #define LAST_MASK			0x0000000000800ull
-#define PHYS_ADDR_MASK			0xFFFFFFFFFFFFF000ull
 #define HOP0_MASK			0x3000000000000ull
 #define HOP1_MASK			0x0FF8000000000ull
 #define HOP2_MASK			0x0007FC0000000ull
 #define HOP3_MASK			0x000003FE00000ull
 #define HOP4_MASK			0x00000001FF000ull
-#define OFFSET_MASK			0x0000000000FFFull
+#define FLAGS_MASK			0x0000000000FFFull
 
 #define HOP0_SHIFT			48
 #define HOP1_SHIFT			39
@@ -31,8 +29,7 @@
 #define HOP3_SHIFT			21
 #define HOP4_SHIFT			12
 
-#define PTE_PHYS_ADDR_SHIFT		12
-#define PTE_PHYS_ADDR_MASK		~OFFSET_MASK
+#define HOP_PHYS_ADDR_MASK		(~FLAGS_MASK)
 
 #define HL_PTE_SIZE			sizeof(u64)
 #define HOP_TABLE_SIZE			PAGE_SIZE_4KB
diff --git a/drivers/misc/habanalabs/include/qman_if.h b/drivers/misc/habanalabs/include/qman_if.h
index bf59bbe27fdc..0fdb49188ed7 100644
--- a/drivers/misc/habanalabs/include/qman_if.h
+++ b/drivers/misc/habanalabs/include/qman_if.h
@@ -23,6 +23,8 @@ struct hl_bd {
 #define HL_BD_SIZE			sizeof(struct hl_bd)
 
 /*
+ * S/W CTL FIELDS.
+ *
  * BD_CTL_REPEAT_VALID tells the CP whether the repeat field in the BD CTL is
  * valid. 1 means the repeat field is valid, 0 means not-valid,
  * i.e. repeat == 1
@@ -34,6 +36,16 @@ struct hl_bd {
 #define BD_CTL_SHADOW_INDEX_MASK	0x00000FFF
 
 /*
+ * H/W CTL FIELDS
+ */
+
+#define BD_CTL_COMP_OFFSET_SHIFT	16
+#define BD_CTL_COMP_OFFSET_MASK		0x00FF0000
+
+#define BD_CTL_COMP_DATA_SHIFT		0
+#define BD_CTL_COMP_DATA_MASK		0x0000FFFF
+
+/*
  * COMPLETION QUEUE
  */
 
diff --git a/drivers/misc/habanalabs/memory.c b/drivers/misc/habanalabs/memory.c
index 365fb0cb8dff..6c72cb4eff54 100644
--- a/drivers/misc/habanalabs/memory.c
+++ b/drivers/misc/habanalabs/memory.c
@@ -13,7 +13,6 @@
 #include <linux/slab.h>
 #include <linux/genalloc.h>
 
-#define PGS_IN_2MB_PAGE	(PAGE_SIZE_2MB >> PAGE_SHIFT)
 #define HL_MMU_DEBUG	0
 
 /*
@@ -159,20 +158,19 @@ pages_pack_err:
 }
 
 /*
- * get_userptr_from_host_va - initialize userptr structure from given host
- *                            virtual address
- *
- * @hdev                : habanalabs device structure
- * @args                : parameters containing the virtual address and size
- * @p_userptr           : pointer to result userptr structure
+ * dma_map_host_va - DMA mapping of the given host virtual address.
+ * @hdev: habanalabs device structure
+ * @addr: the host virtual address of the memory area
+ * @size: the size of the memory area
+ * @p_userptr: pointer to result userptr structure
  *
  * This function does the following:
  * - Allocate userptr structure
  * - Pin the given host memory using the userptr structure
  * - Perform DMA mapping to have the DMA addresses of the pages
  */
-static int get_userptr_from_host_va(struct hl_device *hdev,
-		struct hl_mem_in *args, struct hl_userptr **p_userptr)
+static int dma_map_host_va(struct hl_device *hdev, u64 addr, u64 size,
+				struct hl_userptr **p_userptr)
 {
 	struct hl_userptr *userptr;
 	int rc;
@@ -183,8 +181,7 @@ static int get_userptr_from_host_va(struct hl_device *hdev,
 		goto userptr_err;
 	}
 
-	rc = hl_pin_host_memory(hdev, args->map_host.host_virt_addr,
-			args->map_host.mem_size, userptr);
+	rc = hl_pin_host_memory(hdev, addr, size, userptr);
 	if (rc) {
 		dev_err(hdev->dev, "Failed to pin host memory\n");
 		goto pin_err;
@@ -215,16 +212,16 @@ userptr_err:
 }
 
 /*
- * free_userptr - free userptr structure
- *
- * @hdev                : habanalabs device structure
- * @userptr             : userptr to free
+ * dma_unmap_host_va - DMA unmapping of the given host virtual address.
+ * @hdev: habanalabs device structure
+ * @userptr: userptr to free
  *
  * This function does the following:
  * - Unpins the physical pages
  * - Frees the userptr structure
  */
-static void free_userptr(struct hl_device *hdev, struct hl_userptr *userptr)
+static void dma_unmap_host_va(struct hl_device *hdev,
+				struct hl_userptr *userptr)
 {
 	hl_unpin_host_memory(hdev, userptr);
 	kfree(userptr);
@@ -253,10 +250,9 @@ static void dram_pg_pool_do_release(struct kref *ref)
 }
 
 /*
- * free_phys_pg_pack   - free physical page pack
- *
- * @hdev               : habanalabs device structure
- * @phys_pg_pack       : physical page pack to free
+ * free_phys_pg_pack - free physical page pack
+ * @hdev: habanalabs device structure
+ * @phys_pg_pack: physical page pack to free
  *
  * This function does the following:
  * - For DRAM memory only, iterate over the pack and free each physical block
@@ -264,7 +260,7 @@ static void dram_pg_pool_do_release(struct kref *ref)
  * - Free the hl_vm_phys_pg_pack structure
  */
 static void free_phys_pg_pack(struct hl_device *hdev,
-		struct hl_vm_phys_pg_pack *phys_pg_pack)
+				struct hl_vm_phys_pg_pack *phys_pg_pack)
 {
 	struct hl_vm *vm = &hdev->vm;
 	u64 i;
@@ -519,8 +515,8 @@ static inline int add_va_block(struct hl_device *hdev,
  * - Return the start address of the virtual block
  */
 static u64 get_va_block(struct hl_device *hdev,
-		struct hl_va_range *va_range, u64 size, u64 hint_addr,
-		bool is_userptr)
+			struct hl_va_range *va_range, u64 size, u64 hint_addr,
+			bool is_userptr)
 {
 	struct hl_vm_va_block *va_block, *new_va_block = NULL;
 	u64 valid_start, valid_size, prev_start, prev_end, page_mask,
@@ -528,18 +524,17 @@ static u64 get_va_block(struct hl_device *hdev,
 	u32 page_size;
 	bool add_prev = false;
 
-	if (is_userptr) {
+	if (is_userptr)
 		/*
 		 * We cannot know if the user allocated memory with huge pages
 		 * or not, hence we continue with the biggest possible
 		 * granularity.
 		 */
-		page_size = PAGE_SIZE_2MB;
-		page_mask = PAGE_MASK_2MB;
-	} else {
-		page_size = hdev->asic_prop.dram_page_size;
-		page_mask = ~((u64)page_size - 1);
-	}
+		page_size = hdev->asic_prop.pmmu.huge_page_size;
+	else
+		page_size = hdev->asic_prop.dmmu.page_size;
+
+	page_mask = ~((u64)page_size - 1);
 
 	mutex_lock(&va_range->lock);
 
@@ -549,7 +544,6 @@ static u64 get_va_block(struct hl_device *hdev,
 		/* calc the first possible aligned addr */
 		valid_start = va_block->start;
 
-
 		if (valid_start & (page_size - 1)) {
 			valid_start &= page_mask;
 			valid_start += page_size;
@@ -561,7 +555,6 @@ static u64 get_va_block(struct hl_device *hdev,
 
 		if (valid_size >= size &&
 			(!new_va_block || valid_size < res_valid_size)) {
-
 			new_va_block = va_block;
 			res_valid_start = valid_start;
 			res_valid_size = valid_size;
@@ -631,11 +624,10 @@ static u32 get_sg_info(struct scatterlist *sg, dma_addr_t *dma_addr)
 
 /*
  * init_phys_pg_pack_from_userptr - initialize physical page pack from host
- *                                   memory
- *
- * @ctx                : current context
- * @userptr            : userptr to initialize from
- * @pphys_pg_pack      : res pointer
+ *                                  memory
+ * @ctx: current context
+ * @userptr: userptr to initialize from
+ * @pphys_pg_pack: result pointer
  *
  * This function does the following:
  * - Pin the physical pages related to the given virtual block
@@ -643,16 +635,19 @@ static u32 get_sg_info(struct scatterlist *sg, dma_addr_t *dma_addr)
  *   virtual block
  */
 static int init_phys_pg_pack_from_userptr(struct hl_ctx *ctx,
-		struct hl_userptr *userptr,
-		struct hl_vm_phys_pg_pack **pphys_pg_pack)
+				struct hl_userptr *userptr,
+				struct hl_vm_phys_pg_pack **pphys_pg_pack)
 {
+	struct hl_mmu_properties *mmu_prop = &ctx->hdev->asic_prop.pmmu;
 	struct hl_vm_phys_pg_pack *phys_pg_pack;
 	struct scatterlist *sg;
 	dma_addr_t dma_addr;
 	u64 page_mask, total_npages;
-	u32 npages, page_size = PAGE_SIZE;
+	u32 npages, page_size = PAGE_SIZE,
+		huge_page_size = mmu_prop->huge_page_size;
 	bool first = true, is_huge_page_opt = true;
 	int rc, i, j;
+	u32 pgs_in_huge_page = huge_page_size >> __ffs(page_size);
 
 	phys_pg_pack = kzalloc(sizeof(*phys_pg_pack), GFP_KERNEL);
 	if (!phys_pg_pack)
@@ -675,14 +670,14 @@ static int init_phys_pg_pack_from_userptr(struct hl_ctx *ctx,
 
 		total_npages += npages;
 
-		if ((npages % PGS_IN_2MB_PAGE) ||
-					(dma_addr & (PAGE_SIZE_2MB - 1)))
+		if ((npages % pgs_in_huge_page) ||
+					(dma_addr & (huge_page_size - 1)))
 			is_huge_page_opt = false;
 	}
 
 	if (is_huge_page_opt) {
-		page_size = PAGE_SIZE_2MB;
-		total_npages /= PGS_IN_2MB_PAGE;
+		page_size = huge_page_size;
+		do_div(total_npages, pgs_in_huge_page);
 	}
 
 	page_mask = ~(((u64) page_size) - 1);
@@ -714,7 +709,7 @@ static int init_phys_pg_pack_from_userptr(struct hl_ctx *ctx,
 			dma_addr += page_size;
 
 			if (is_huge_page_opt)
-				npages -= PGS_IN_2MB_PAGE;
+				npages -= pgs_in_huge_page;
 			else
 				npages--;
 		}
@@ -731,19 +726,18 @@ page_pack_arr_mem_err:
 }
 
 /*
- * map_phys_page_pack - maps the physical page pack
- *
- * @ctx                : current context
- * @vaddr              : start address of the virtual area to map from
- * @phys_pg_pack       : the pack of physical pages to map to
+ * map_phys_pg_pack - maps the physical page pack.
+ * @ctx: current context
+ * @vaddr: start address of the virtual area to map from
+ * @phys_pg_pack: the pack of physical pages to map to
  *
  * This function does the following:
  * - Maps each chunk of virtual memory to matching physical chunk
  * - Stores number of successful mappings in the given argument
- * - Returns 0 on success, error code otherwise.
+ * - Returns 0 on success, error code otherwise
  */
-static int map_phys_page_pack(struct hl_ctx *ctx, u64 vaddr,
-		struct hl_vm_phys_pg_pack *phys_pg_pack)
+static int map_phys_pg_pack(struct hl_ctx *ctx, u64 vaddr,
+				struct hl_vm_phys_pg_pack *phys_pg_pack)
 {
 	struct hl_device *hdev = ctx->hdev;
 	u64 next_vaddr = vaddr, paddr, mapped_pg_cnt = 0, i;
@@ -783,6 +777,36 @@ err:
 	return rc;
 }
 
+/*
+ * unmap_phys_pg_pack - unmaps the physical page pack
+ * @ctx: current context
+ * @vaddr: start address of the virtual area to unmap
+ * @phys_pg_pack: the pack of physical pages to unmap
+ */
+static void unmap_phys_pg_pack(struct hl_ctx *ctx, u64 vaddr,
+				struct hl_vm_phys_pg_pack *phys_pg_pack)
+{
+	struct hl_device *hdev = ctx->hdev;
+	u64 next_vaddr, i;
+	u32 page_size;
+
+	page_size = phys_pg_pack->page_size;
+	next_vaddr = vaddr;
+
+	for (i = 0 ; i < phys_pg_pack->npages ; i++, next_vaddr += page_size) {
+		if (hl_mmu_unmap(ctx, next_vaddr, page_size))
+			dev_warn_ratelimited(hdev->dev,
+			"unmap failed for vaddr: 0x%llx\n", next_vaddr);
+
+		/*
+		 * unmapping on Palladium can be really long, so avoid a CPU
+		 * soft lockup bug by sleeping a little between unmapping pages
+		 */
+		if (hdev->pldm)
+			usleep_range(500, 1000);
+	}
+}
+
 static int get_paddr_from_handle(struct hl_ctx *ctx, struct hl_mem_in *args,
 				u64 *paddr)
 {
@@ -839,7 +863,10 @@ static int map_device_va(struct hl_ctx *ctx, struct hl_mem_in *args,
 	*device_addr = 0;
 
 	if (is_userptr) {
-		rc = get_userptr_from_host_va(hdev, args, &userptr);
+		u64 addr = args->map_host.host_virt_addr,
+			size = args->map_host.mem_size;
+
+		rc = dma_map_host_va(hdev, addr, size, &userptr);
 		if (rc) {
 			dev_err(hdev->dev, "failed to get userptr from va\n");
 			return rc;
@@ -850,7 +877,7 @@ static int map_device_va(struct hl_ctx *ctx, struct hl_mem_in *args,
 		if (rc) {
 			dev_err(hdev->dev,
 				"unable to init page pack for vaddr 0x%llx\n",
-				args->map_host.host_virt_addr);
+				addr);
 			goto init_page_pack_err;
 		}
 
@@ -909,7 +936,7 @@ static int map_device_va(struct hl_ctx *ctx, struct hl_mem_in *args,
 
 	mutex_lock(&ctx->mmu_lock);
 
-	rc = map_phys_page_pack(ctx, ret_vaddr, phys_pg_pack);
+	rc = map_phys_pg_pack(ctx, ret_vaddr, phys_pg_pack);
 	if (rc) {
 		mutex_unlock(&ctx->mmu_lock);
 		dev_err(hdev->dev, "mapping page pack failed for handle %u\n",
@@ -917,7 +944,7 @@ static int map_device_va(struct hl_ctx *ctx, struct hl_mem_in *args,
 		goto map_err;
 	}
 
-	hdev->asic_funcs->mmu_invalidate_cache(hdev, false);
+	hdev->asic_funcs->mmu_invalidate_cache(hdev, false, *vm_type);
 
 	mutex_unlock(&ctx->mmu_lock);
 
@@ -955,7 +982,7 @@ shared_err:
 		free_phys_pg_pack(hdev, phys_pg_pack);
 init_page_pack_err:
 	if (is_userptr)
-		free_userptr(hdev, userptr);
+		dma_unmap_host_va(hdev, userptr);
 
 	return rc;
 }
@@ -965,20 +992,20 @@ init_page_pack_err:
  *
  * @ctx                 : current context
  * @vaddr               : device virtual address to unmap
+ * @ctx_free            : true if in context free flow, false otherwise.
  *
  * This function does the following:
  * - Unmap the physical pages related to the given virtual address
  * - return the device virtual block to the virtual block list
  */
-static int unmap_device_va(struct hl_ctx *ctx, u64 vaddr)
+static int unmap_device_va(struct hl_ctx *ctx, u64 vaddr, bool ctx_free)
 {
 	struct hl_device *hdev = ctx->hdev;
 	struct hl_vm_phys_pg_pack *phys_pg_pack = NULL;
 	struct hl_vm_hash_node *hnode = NULL;
 	struct hl_userptr *userptr = NULL;
+	struct hl_va_range *va_range;
 	enum vm_type_t *vm_type;
-	u64 next_vaddr, i;
-	u32 page_size;
 	bool is_userptr;
 	int rc;
 
@@ -1003,9 +1030,10 @@ static int unmap_device_va(struct hl_ctx *ctx, u64 vaddr)
 
 	if (*vm_type == VM_TYPE_USERPTR) {
 		is_userptr = true;
+		va_range = &ctx->host_va_range;
 		userptr = hnode->ptr;
 		rc = init_phys_pg_pack_from_userptr(ctx, userptr,
-				&phys_pg_pack);
+							&phys_pg_pack);
 		if (rc) {
 			dev_err(hdev->dev,
 				"unable to init page pack for vaddr 0x%llx\n",
@@ -1014,6 +1042,7 @@ static int unmap_device_va(struct hl_ctx *ctx, u64 vaddr)
 		}
 	} else if (*vm_type == VM_TYPE_PHYS_PACK) {
 		is_userptr = false;
+		va_range = &ctx->dram_va_range;
 		phys_pg_pack = hnode->ptr;
 	} else {
 		dev_warn(hdev->dev,
@@ -1029,42 +1058,41 @@ static int unmap_device_va(struct hl_ctx *ctx, u64 vaddr)
 		goto mapping_cnt_err;
 	}
 
-	page_size = phys_pg_pack->page_size;
-	vaddr &= ~(((u64) page_size) - 1);
-
-	next_vaddr = vaddr;
+	vaddr &= ~(((u64) phys_pg_pack->page_size) - 1);
 
 	mutex_lock(&ctx->mmu_lock);
 
-	for (i = 0 ; i < phys_pg_pack->npages ; i++, next_vaddr += page_size) {
-		if (hl_mmu_unmap(ctx, next_vaddr, page_size))
-			dev_warn_ratelimited(hdev->dev,
-			"unmap failed for vaddr: 0x%llx\n", next_vaddr);
-
-		/* unmapping on Palladium can be really long, so avoid a CPU
-		 * soft lockup bug by sleeping a little between unmapping pages
-		 */
-		if (hdev->pldm)
-			usleep_range(500, 1000);
-	}
+	unmap_phys_pg_pack(ctx, vaddr, phys_pg_pack);
 
-	hdev->asic_funcs->mmu_invalidate_cache(hdev, true);
+	/*
+	 * During context free this function is called in a loop to clean all
+	 * the context mappings. Hence the cache invalidation can be called once
+	 * at the loop end rather than for each iteration
+	 */
+	if (!ctx_free)
+		hdev->asic_funcs->mmu_invalidate_cache(hdev, true, *vm_type);
 
 	mutex_unlock(&ctx->mmu_lock);
 
-	if (add_va_block(hdev,
-			is_userptr ? &ctx->host_va_range : &ctx->dram_va_range,
-			vaddr,
-			vaddr + phys_pg_pack->total_size - 1))
-		dev_warn(hdev->dev, "add va block failed for vaddr: 0x%llx\n",
-				vaddr);
+	/*
+	 * No point in maintaining the free VA block list if the context is
+	 * closing as the list will be freed anyway
+	 */
+	if (!ctx_free) {
+		rc = add_va_block(hdev, va_range, vaddr,
+					vaddr + phys_pg_pack->total_size - 1);
+		if (rc)
+			dev_warn(hdev->dev,
+					"add va block failed for vaddr: 0x%llx\n",
+					vaddr);
+	}
 
 	atomic_dec(&phys_pg_pack->mapping_cnt);
 	kfree(hnode);
 
 	if (is_userptr) {
 		free_phys_pg_pack(hdev, phys_pg_pack);
-		free_userptr(hdev, userptr);
+		dma_unmap_host_va(hdev, userptr);
 	}
 
 	return 0;
@@ -1189,8 +1217,8 @@ int hl_mem_ioctl(struct hl_fpriv *hpriv, void *data)
 		break;
 
 	case HL_MEM_OP_UNMAP:
-		rc = unmap_device_va(ctx,
-				args->in.unmap.device_virt_addr);
+		rc = unmap_device_va(ctx, args->in.unmap.device_virt_addr,
+					false);
 		break;
 
 	default:
@@ -1203,20 +1231,72 @@ out:
 	return rc;
 }
 
+static int get_user_memory(struct hl_device *hdev, u64 addr, u64 size,
+				u32 npages, u64 start, u32 offset,
+				struct hl_userptr *userptr)
+{
+	int rc;
+
+	if (!access_ok((void __user *) (uintptr_t) addr, size)) {
+		dev_err(hdev->dev, "user pointer is invalid - 0x%llx\n", addr);
+		return -EFAULT;
+	}
+
+	userptr->vec = frame_vector_create(npages);
+	if (!userptr->vec) {
+		dev_err(hdev->dev, "Failed to create frame vector\n");
+		return -ENOMEM;
+	}
+
+	rc = get_vaddr_frames(start, npages, FOLL_FORCE | FOLL_WRITE,
+				userptr->vec);
+
+	if (rc != npages) {
+		dev_err(hdev->dev,
+			"Failed to map host memory, user ptr probably wrong\n");
+		if (rc < 0)
+			goto destroy_framevec;
+		rc = -EFAULT;
+		goto put_framevec;
+	}
+
+	if (frame_vector_to_pages(userptr->vec) < 0) {
+		dev_err(hdev->dev,
+			"Failed to translate frame vector to pages\n");
+		rc = -EFAULT;
+		goto put_framevec;
+	}
+
+	rc = sg_alloc_table_from_pages(userptr->sgt,
+					frame_vector_pages(userptr->vec),
+					npages, offset, size, GFP_ATOMIC);
+	if (rc < 0) {
+		dev_err(hdev->dev, "failed to create SG table from pages\n");
+		goto put_framevec;
+	}
+
+	return 0;
+
+put_framevec:
+	put_vaddr_frames(userptr->vec);
+destroy_framevec:
+	frame_vector_destroy(userptr->vec);
+	return rc;
+}
+
 /*
- * hl_pin_host_memory - pins a chunk of host memory
- *
- * @hdev                : pointer to the habanalabs device structure
- * @addr                : the user-space virtual address of the memory area
- * @size                : the size of the memory area
- * @userptr	        : pointer to hl_userptr structure
+ * hl_pin_host_memory - pins a chunk of host memory.
+ * @hdev: pointer to the habanalabs device structure
+ * @addr: the host virtual address of the memory area
+ * @size: the size of the memory area
+ * @userptr: pointer to hl_userptr structure
  *
  * This function does the following:
  * - Pins the physical pages
- * - Create a SG list from those pages
+ * - Create an SG list from those pages
  */
 int hl_pin_host_memory(struct hl_device *hdev, u64 addr, u64 size,
-			struct hl_userptr *userptr)
+					struct hl_userptr *userptr)
 {
 	u64 start, end;
 	u32 npages, offset;
@@ -1227,11 +1307,6 @@ int hl_pin_host_memory(struct hl_device *hdev, u64 addr, u64 size,
 		return -EINVAL;
 	}
 
-	if (!access_ok((void __user *) (uintptr_t) addr, size)) {
-		dev_err(hdev->dev, "user pointer is invalid - 0x%llx\n", addr);
-		return -EFAULT;
-	}
-
 	/*
 	 * If the combination of the address and size requested for this memory
 	 * region causes an integer overflow, return error.
@@ -1244,6 +1319,14 @@ int hl_pin_host_memory(struct hl_device *hdev, u64 addr, u64 size,
 		return -EINVAL;
 	}
 
+	/*
+	 * This function can be called also from data path, hence use atomic
+	 * always as it is not a big allocation.
+	 */
+	userptr->sgt = kzalloc(sizeof(*userptr->sgt), GFP_ATOMIC);
+	if (!userptr->sgt)
+		return -ENOMEM;
+
 	start = addr & PAGE_MASK;
 	offset = addr & ~PAGE_MASK;
 	end = PAGE_ALIGN(addr + size);
@@ -1254,42 +1337,12 @@ int hl_pin_host_memory(struct hl_device *hdev, u64 addr, u64 size,
 	userptr->dma_mapped = false;
 	INIT_LIST_HEAD(&userptr->job_node);
 
-	userptr->vec = frame_vector_create(npages);
-	if (!userptr->vec) {
-		dev_err(hdev->dev, "Failed to create frame vector\n");
-		return -ENOMEM;
-	}
-
-	rc = get_vaddr_frames(start, npages, FOLL_FORCE | FOLL_WRITE,
-				userptr->vec);
-
-	if (rc != npages) {
-		dev_err(hdev->dev,
-			"Failed to map host memory, user ptr probably wrong\n");
-		if (rc < 0)
-			goto destroy_framevec;
-		rc = -EFAULT;
-		goto put_framevec;
-	}
-
-	if (frame_vector_to_pages(userptr->vec) < 0) {
+	rc = get_user_memory(hdev, addr, size, npages, start, offset,
+				userptr);
+	if (rc) {
 		dev_err(hdev->dev,
-			"Failed to translate frame vector to pages\n");
-		rc = -EFAULT;
-		goto put_framevec;
-	}
-
-	userptr->sgt = kzalloc(sizeof(*userptr->sgt), GFP_ATOMIC);
-	if (!userptr->sgt) {
-		rc = -ENOMEM;
-		goto put_framevec;
-	}
-
-	rc = sg_alloc_table_from_pages(userptr->sgt,
-					frame_vector_pages(userptr->vec),
-					npages, offset, size, GFP_ATOMIC);
-	if (rc < 0) {
-		dev_err(hdev->dev, "failed to create SG table from pages\n");
+			"failed to get user memory for address 0x%llx\n",
+			addr);
 		goto free_sgt;
 	}
 
@@ -1299,34 +1352,28 @@ int hl_pin_host_memory(struct hl_device *hdev, u64 addr, u64 size,
 
 free_sgt:
 	kfree(userptr->sgt);
-put_framevec:
-	put_vaddr_frames(userptr->vec);
-destroy_framevec:
-	frame_vector_destroy(userptr->vec);
 	return rc;
 }
 
 /*
- * hl_unpin_host_memory - unpins a chunk of host memory
- *
- * @hdev                : pointer to the habanalabs device structure
- * @userptr             : pointer to hl_userptr structure
+ * hl_unpin_host_memory - unpins a chunk of host memory.
+ * @hdev: pointer to the habanalabs device structure
+ * @userptr: pointer to hl_userptr structure
  *
  * This function does the following:
  * - Unpins the physical pages related to the host memory
  * - Free the SG list
  */
-int hl_unpin_host_memory(struct hl_device *hdev, struct hl_userptr *userptr)
+void hl_unpin_host_memory(struct hl_device *hdev, struct hl_userptr *userptr)
 {
 	struct page **pages;
 
 	hl_debugfs_remove_userptr(hdev, userptr);
 
 	if (userptr->dma_mapped)
-		hdev->asic_funcs->hl_dma_unmap_sg(hdev,
-				userptr->sgt->sgl,
-				userptr->sgt->nents,
-				userptr->dir);
+		hdev->asic_funcs->hl_dma_unmap_sg(hdev, userptr->sgt->sgl,
+							userptr->sgt->nents,
+							userptr->dir);
 
 	pages = frame_vector_pages(userptr->vec);
 	if (!IS_ERR(pages)) {
@@ -1342,8 +1389,6 @@ int hl_unpin_host_memory(struct hl_device *hdev, struct hl_userptr *userptr)
 
 	sg_free_table(userptr->sgt);
 	kfree(userptr->sgt);
-
-	return 0;
 }
 
 /*
@@ -1542,43 +1587,16 @@ int hl_vm_ctx_init(struct hl_ctx *ctx)
  * @hdev                : pointer to the habanalabs structure
  * va_range             : pointer to virtual addresses range
  *
- * This function initializes the following:
- * - Checks that the given range contains the whole initial range
+ * This function does the following:
  * - Frees the virtual addresses block list and its lock
  */
 static void hl_va_range_fini(struct hl_device *hdev,
 		struct hl_va_range *va_range)
 {
-	struct hl_vm_va_block *va_block;
-
-	if (list_empty(&va_range->list)) {
-		dev_warn(hdev->dev,
-				"va list should not be empty on cleanup!\n");
-		goto out;
-	}
-
-	if (!list_is_singular(&va_range->list)) {
-		dev_warn(hdev->dev,
-			"va list should not contain multiple blocks on cleanup!\n");
-		goto free_va_list;
-	}
-
-	va_block = list_first_entry(&va_range->list, typeof(*va_block), node);
-
-	if (va_block->start != va_range->start_addr ||
-		va_block->end != va_range->end_addr) {
-		dev_warn(hdev->dev,
-			"wrong va block on cleanup, from 0x%llx to 0x%llx\n",
-				va_block->start, va_block->end);
-		goto free_va_list;
-	}
-
-free_va_list:
 	mutex_lock(&va_range->lock);
 	clear_va_list_locked(hdev, &va_range->list);
 	mutex_unlock(&va_range->lock);
 
-out:
 	mutex_destroy(&va_range->lock);
 }
 
@@ -1613,21 +1631,31 @@ void hl_vm_ctx_fini(struct hl_ctx *ctx)
 
 	hl_debugfs_remove_ctx_mem_hash(hdev, ctx);
 
-	if (!hash_empty(ctx->mem_hash))
-		dev_notice(hdev->dev, "ctx is freed while it has va in use\n");
+	/*
+	 * Clearly something went wrong on hard reset so no point in printing
+	 * another side effect error
+	 */
+	if (!hdev->hard_reset_pending && !hash_empty(ctx->mem_hash))
+		dev_notice(hdev->dev,
+				"ctx %d is freed while it has va in use\n",
+				ctx->asid);
 
 	hash_for_each_safe(ctx->mem_hash, i, tmp_node, hnode, node) {
 		dev_dbg(hdev->dev,
 			"hl_mem_hash_node of vaddr 0x%llx of asid %d is still alive\n",
 			hnode->vaddr, ctx->asid);
-		unmap_device_va(ctx, hnode->vaddr);
+		unmap_device_va(ctx, hnode->vaddr, true);
 	}
 
+	/* invalidate the cache once after the unmapping loop */
+	hdev->asic_funcs->mmu_invalidate_cache(hdev, true, VM_TYPE_USERPTR);
+	hdev->asic_funcs->mmu_invalidate_cache(hdev, true, VM_TYPE_PHYS_PACK);
+
 	spin_lock(&vm->idr_lock);
 	idr_for_each_entry(&vm->phys_pg_pack_handles, phys_pg_list, i)
 		if (phys_pg_list->asid == ctx->asid) {
 			dev_dbg(hdev->dev,
-				"page list 0x%p of asid %d is still alive\n",
+				"page list 0x%px of asid %d is still alive\n",
 				phys_pg_list, ctx->asid);
 			atomic64_sub(phys_pg_list->total_size,
 					&hdev->dram_used_mem);
diff --git a/drivers/misc/habanalabs/mmu.c b/drivers/misc/habanalabs/mmu.c
index 176c315836f1..6262b26e2086 100644
--- a/drivers/misc/habanalabs/mmu.c
+++ b/drivers/misc/habanalabs/mmu.c
@@ -25,10 +25,9 @@ static struct pgt_info *get_pgt_info(struct hl_ctx *ctx, u64 hop_addr)
 	return pgt_info;
 }
 
-static void free_hop(struct hl_ctx *ctx, u64 hop_addr)
+static void _free_hop(struct hl_ctx *ctx, struct pgt_info *pgt_info)
 {
 	struct hl_device *hdev = ctx->hdev;
-	struct pgt_info *pgt_info = get_pgt_info(ctx, hop_addr);
 
 	gen_pool_free(hdev->mmu_pgt_pool, pgt_info->phys_addr,
 			hdev->asic_prop.mmu_hop_table_size);
@@ -37,6 +36,13 @@ static void free_hop(struct hl_ctx *ctx, u64 hop_addr)
 	kfree(pgt_info);
 }
 
+static void free_hop(struct hl_ctx *ctx, u64 hop_addr)
+{
+	struct pgt_info *pgt_info = get_pgt_info(ctx, hop_addr);
+
+	_free_hop(ctx, pgt_info);
+}
+
 static u64 alloc_hop(struct hl_ctx *ctx)
 {
 	struct hl_device *hdev = ctx->hdev;
@@ -105,8 +111,8 @@ static inline void write_pte(struct hl_ctx *ctx, u64 shadow_pte_addr, u64 val)
 	 * clear the 12 LSBs and translate the shadow hop to its associated
 	 * physical hop, and add back the original 12 LSBs.
 	 */
-	u64 phys_val = get_phys_addr(ctx, val & PTE_PHYS_ADDR_MASK) |
-				(val & OFFSET_MASK);
+	u64 phys_val = get_phys_addr(ctx, val & HOP_PHYS_ADDR_MASK) |
+				(val & FLAGS_MASK);
 
 	ctx->hdev->asic_funcs->write_pte(ctx->hdev,
 					get_phys_addr(ctx, shadow_pte_addr),
@@ -159,7 +165,7 @@ static inline int put_pte(struct hl_ctx *ctx, u64 hop_addr)
 	 */
 	num_of_ptes_left = pgt_info->num_of_ptes;
 	if (!num_of_ptes_left)
-		free_hop(ctx, hop_addr);
+		_free_hop(ctx, pgt_info);
 
 	return num_of_ptes_left;
 }
@@ -171,35 +177,50 @@ static inline u64 get_hopN_pte_addr(struct hl_ctx *ctx, u64 hop_addr,
 			((virt_addr & mask) >> shift);
 }
 
-static inline u64 get_hop0_pte_addr(struct hl_ctx *ctx, u64 hop_addr, u64 vaddr)
+static inline u64 get_hop0_pte_addr(struct hl_ctx *ctx,
+					struct hl_mmu_properties *mmu_prop,
+					u64 hop_addr, u64 vaddr)
 {
-	return get_hopN_pte_addr(ctx, hop_addr, vaddr, HOP0_MASK, HOP0_SHIFT);
+	return get_hopN_pte_addr(ctx, hop_addr, vaddr, mmu_prop->hop0_mask,
+					mmu_prop->hop0_shift);
 }
 
-static inline u64 get_hop1_pte_addr(struct hl_ctx *ctx, u64 hop_addr, u64 vaddr)
+static inline u64 get_hop1_pte_addr(struct hl_ctx *ctx,
+					struct hl_mmu_properties *mmu_prop,
+					u64 hop_addr, u64 vaddr)
 {
-	return get_hopN_pte_addr(ctx, hop_addr, vaddr, HOP1_MASK, HOP1_SHIFT);
+	return get_hopN_pte_addr(ctx, hop_addr, vaddr, mmu_prop->hop1_mask,
+					mmu_prop->hop1_shift);
 }
 
-static inline u64 get_hop2_pte_addr(struct hl_ctx *ctx, u64 hop_addr, u64 vaddr)
+static inline u64 get_hop2_pte_addr(struct hl_ctx *ctx,
+					struct hl_mmu_properties *mmu_prop,
+					u64 hop_addr, u64 vaddr)
 {
-	return get_hopN_pte_addr(ctx, hop_addr, vaddr, HOP2_MASK, HOP2_SHIFT);
+	return get_hopN_pte_addr(ctx, hop_addr, vaddr, mmu_prop->hop2_mask,
+					mmu_prop->hop2_shift);
 }
 
-static inline u64 get_hop3_pte_addr(struct hl_ctx *ctx, u64 hop_addr, u64 vaddr)
+static inline u64 get_hop3_pte_addr(struct hl_ctx *ctx,
+					struct hl_mmu_properties *mmu_prop,
+					u64 hop_addr, u64 vaddr)
 {
-	return get_hopN_pte_addr(ctx, hop_addr, vaddr, HOP3_MASK, HOP3_SHIFT);
+	return get_hopN_pte_addr(ctx, hop_addr, vaddr, mmu_prop->hop3_mask,
+					mmu_prop->hop3_shift);
 }
 
-static inline u64 get_hop4_pte_addr(struct hl_ctx *ctx, u64 hop_addr, u64 vaddr)
+static inline u64 get_hop4_pte_addr(struct hl_ctx *ctx,
+					struct hl_mmu_properties *mmu_prop,
+					u64 hop_addr, u64 vaddr)
 {
-	return get_hopN_pte_addr(ctx, hop_addr, vaddr, HOP4_MASK, HOP4_SHIFT);
+	return get_hopN_pte_addr(ctx, hop_addr, vaddr, mmu_prop->hop4_mask,
+					mmu_prop->hop4_shift);
 }
 
 static inline u64 get_next_hop_addr(struct hl_ctx *ctx, u64 curr_pte)
 {
 	if (curr_pte & PAGE_PRESENT_MASK)
-		return curr_pte & PHYS_ADDR_MASK;
+		return curr_pte & HOP_PHYS_ADDR_MASK;
 	else
 		return ULLONG_MAX;
 }
@@ -288,23 +309,23 @@ static int dram_default_mapping_init(struct hl_ctx *ctx)
 	}
 
 	/* need only pte 0 in hops 0 and 1 */
-	pte_val = (hop1_addr & PTE_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK;
+	pte_val = (hop1_addr & HOP_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK;
 	write_pte(ctx, hop0_addr, pte_val);
 
-	pte_val = (hop2_addr & PTE_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK;
+	pte_val = (hop2_addr & HOP_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK;
 	write_pte(ctx, hop1_addr, pte_val);
 	get_pte(ctx, hop1_addr);
 
 	hop2_pte_addr = hop2_addr;
 	for (i = 0 ; i < num_of_hop3 ; i++) {
-		pte_val = (ctx->dram_default_hops[i] & PTE_PHYS_ADDR_MASK) |
+		pte_val = (ctx->dram_default_hops[i] & HOP_PHYS_ADDR_MASK) |
 				PAGE_PRESENT_MASK;
 		write_pte(ctx, hop2_pte_addr, pte_val);
 		get_pte(ctx, hop2_addr);
 		hop2_pte_addr += HL_PTE_SIZE;
 	}
 
-	pte_val = (prop->mmu_dram_default_page_addr & PTE_PHYS_ADDR_MASK) |
+	pte_val = (prop->mmu_dram_default_page_addr & HOP_PHYS_ADDR_MASK) |
 			LAST_MASK | PAGE_PRESENT_MASK;
 
 	for (i = 0 ; i < num_of_hop3 ; i++) {
@@ -400,8 +421,6 @@ int hl_mmu_init(struct hl_device *hdev)
 	if (!hdev->mmu_enable)
 		return 0;
 
-	/* MMU H/W init was already done in device hw_init() */
-
 	hdev->mmu_pgt_pool =
 			gen_pool_create(__ffs(prop->mmu_hop_table_size), -1);
 
@@ -427,6 +446,8 @@ int hl_mmu_init(struct hl_device *hdev)
 		goto err_pool_add;
 	}
 
+	/* MMU H/W init will be done in device hw_init() */
+
 	return 0;
 
 err_pool_add:
@@ -450,10 +471,10 @@ void hl_mmu_fini(struct hl_device *hdev)
 	if (!hdev->mmu_enable)
 		return;
 
+	/* MMU H/W fini was already done in device hw_fini() */
+
 	kvfree(hdev->mmu_shadow_hop0);
 	gen_pool_destroy(hdev->mmu_pgt_pool);
-
-	/* MMU H/W fini will be done in device hw_fini() */
 }
 
 /**
@@ -501,36 +522,36 @@ void hl_mmu_ctx_fini(struct hl_ctx *ctx)
 	dram_default_mapping_fini(ctx);
 
 	if (!hash_empty(ctx->mmu_shadow_hash))
-		dev_err(hdev->dev, "ctx is freed while it has pgts in use\n");
+		dev_err(hdev->dev, "ctx %d is freed while it has pgts in use\n",
+			ctx->asid);
 
 	hash_for_each_safe(ctx->mmu_shadow_hash, i, tmp, pgt_info, node) {
-		dev_err(hdev->dev,
+		dev_err_ratelimited(hdev->dev,
 			"pgt_info of addr 0x%llx of asid %d was not destroyed, num_ptes: %d\n",
 			pgt_info->phys_addr, ctx->asid, pgt_info->num_of_ptes);
-		free_hop(ctx, pgt_info->shadow_addr);
+		_free_hop(ctx, pgt_info);
 	}
 
 	mutex_destroy(&ctx->mmu_lock);
 }
 
-static int _hl_mmu_unmap(struct hl_ctx *ctx, u64 virt_addr)
+static int _hl_mmu_unmap(struct hl_ctx *ctx, u64 virt_addr, bool is_dram_addr)
 {
 	struct hl_device *hdev = ctx->hdev;
 	struct asic_fixed_properties *prop = &hdev->asic_prop;
+	struct hl_mmu_properties *mmu_prop;
 	u64 hop0_addr = 0, hop0_pte_addr = 0,
 		hop1_addr = 0, hop1_pte_addr = 0,
 		hop2_addr = 0, hop2_pte_addr = 0,
 		hop3_addr = 0, hop3_pte_addr = 0,
 		hop4_addr = 0, hop4_pte_addr = 0,
 		curr_pte;
-	bool is_dram_addr, is_huge, clear_hop3 = true;
+	bool is_huge, clear_hop3 = true;
 
-	is_dram_addr = hl_mem_area_inside_range(virt_addr, PAGE_SIZE_2MB,
-				prop->va_space_dram_start_address,
-				prop->va_space_dram_end_address);
+	mmu_prop = is_dram_addr ? &prop->dmmu : &prop->pmmu;
 
 	hop0_addr = get_hop0_addr(ctx);
-	hop0_pte_addr = get_hop0_pte_addr(ctx, hop0_addr, virt_addr);
+	hop0_pte_addr = get_hop0_pte_addr(ctx, mmu_prop, hop0_addr, virt_addr);
 
 	curr_pte = *(u64 *) (uintptr_t) hop0_pte_addr;
 
@@ -539,7 +560,7 @@ static int _hl_mmu_unmap(struct hl_ctx *ctx, u64 virt_addr)
 	if (hop1_addr == ULLONG_MAX)
 		goto not_mapped;
 
-	hop1_pte_addr = get_hop1_pte_addr(ctx, hop1_addr, virt_addr);
+	hop1_pte_addr = get_hop1_pte_addr(ctx, mmu_prop, hop1_addr, virt_addr);
 
 	curr_pte = *(u64 *) (uintptr_t) hop1_pte_addr;
 
@@ -548,7 +569,7 @@ static int _hl_mmu_unmap(struct hl_ctx *ctx, u64 virt_addr)
 	if (hop2_addr == ULLONG_MAX)
 		goto not_mapped;
 
-	hop2_pte_addr = get_hop2_pte_addr(ctx, hop2_addr, virt_addr);
+	hop2_pte_addr = get_hop2_pte_addr(ctx, mmu_prop, hop2_addr, virt_addr);
 
 	curr_pte = *(u64 *) (uintptr_t) hop2_pte_addr;
 
@@ -557,7 +578,7 @@ static int _hl_mmu_unmap(struct hl_ctx *ctx, u64 virt_addr)
 	if (hop3_addr == ULLONG_MAX)
 		goto not_mapped;
 
-	hop3_pte_addr = get_hop3_pte_addr(ctx, hop3_addr, virt_addr);
+	hop3_pte_addr = get_hop3_pte_addr(ctx, mmu_prop, hop3_addr, virt_addr);
 
 	curr_pte = *(u64 *) (uintptr_t) hop3_pte_addr;
 
@@ -575,7 +596,8 @@ static int _hl_mmu_unmap(struct hl_ctx *ctx, u64 virt_addr)
 		if (hop4_addr == ULLONG_MAX)
 			goto not_mapped;
 
-		hop4_pte_addr = get_hop4_pte_addr(ctx, hop4_addr, virt_addr);
+		hop4_pte_addr = get_hop4_pte_addr(ctx, mmu_prop, hop4_addr,
+							virt_addr);
 
 		curr_pte = *(u64 *) (uintptr_t) hop4_pte_addr;
 
@@ -584,7 +606,7 @@ static int _hl_mmu_unmap(struct hl_ctx *ctx, u64 virt_addr)
 
 	if (hdev->dram_default_page_mapping && is_dram_addr) {
 		u64 default_pte = (prop->mmu_dram_default_page_addr &
-				PTE_PHYS_ADDR_MASK) | LAST_MASK |
+				HOP_PHYS_ADDR_MASK) | LAST_MASK |
 					PAGE_PRESENT_MASK;
 		if (curr_pte == default_pte) {
 			dev_err(hdev->dev,
@@ -667,25 +689,36 @@ not_mapped:
 int hl_mmu_unmap(struct hl_ctx *ctx, u64 virt_addr, u32 page_size)
 {
 	struct hl_device *hdev = ctx->hdev;
+	struct asic_fixed_properties *prop = &hdev->asic_prop;
+	struct hl_mmu_properties *mmu_prop;
 	u64 real_virt_addr;
 	u32 real_page_size, npages;
 	int i, rc;
+	bool is_dram_addr;
 
 	if (!hdev->mmu_enable)
 		return 0;
 
+	is_dram_addr = hl_mem_area_inside_range(virt_addr, prop->dmmu.page_size,
+				prop->va_space_dram_start_address,
+				prop->va_space_dram_end_address);
+
+	mmu_prop = is_dram_addr ? &prop->dmmu : &prop->pmmu;
+
 	/*
-	 * The H/W handles mapping of 4KB/2MB page. Hence if the host page size
-	 * is bigger, we break it to sub-pages and unmap them separately.
+	 * The H/W handles mapping of specific page sizes. Hence if the page
+	 * size is bigger, we break it to sub-pages and unmap them separately.
 	 */
-	if ((page_size % PAGE_SIZE_2MB) == 0) {
-		real_page_size = PAGE_SIZE_2MB;
-	} else if ((page_size % PAGE_SIZE_4KB) == 0) {
-		real_page_size = PAGE_SIZE_4KB;
+	if ((page_size % mmu_prop->huge_page_size) == 0) {
+		real_page_size = mmu_prop->huge_page_size;
+	} else if ((page_size % mmu_prop->page_size) == 0) {
+		real_page_size = mmu_prop->page_size;
 	} else {
 		dev_err(hdev->dev,
-			"page size of %u is not 4KB nor 2MB aligned, can't unmap\n",
-				page_size);
+			"page size of %u is not %uKB nor %uMB aligned, can't unmap\n",
+			page_size,
+			mmu_prop->page_size >> 10,
+			mmu_prop->huge_page_size >> 20);
 
 		return -EFAULT;
 	}
@@ -694,7 +727,7 @@ int hl_mmu_unmap(struct hl_ctx *ctx, u64 virt_addr, u32 page_size)
 	real_virt_addr = virt_addr;
 
 	for (i = 0 ; i < npages ; i++) {
-		rc = _hl_mmu_unmap(ctx, real_virt_addr);
+		rc = _hl_mmu_unmap(ctx, real_virt_addr, is_dram_addr);
 		if (rc)
 			return rc;
 
@@ -705,10 +738,11 @@ int hl_mmu_unmap(struct hl_ctx *ctx, u64 virt_addr, u32 page_size)
 }
 
 static int _hl_mmu_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr,
-		u32 page_size)
+			u32 page_size, bool is_dram_addr)
 {
 	struct hl_device *hdev = ctx->hdev;
 	struct asic_fixed_properties *prop = &hdev->asic_prop;
+	struct hl_mmu_properties *mmu_prop;
 	u64 hop0_addr = 0, hop0_pte_addr = 0,
 		hop1_addr = 0, hop1_pte_addr = 0,
 		hop2_addr = 0, hop2_pte_addr = 0,
@@ -716,21 +750,19 @@ static int _hl_mmu_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr,
 		hop4_addr = 0, hop4_pte_addr = 0,
 		curr_pte = 0;
 	bool hop1_new = false, hop2_new = false, hop3_new = false,
-		hop4_new = false, is_huge, is_dram_addr;
+		hop4_new = false, is_huge;
 	int rc = -ENOMEM;
 
+	mmu_prop = is_dram_addr ? &prop->dmmu : &prop->pmmu;
+
 	/*
-	 * This mapping function can map a 4KB/2MB page. For 2MB page there are
-	 * only 3 hops rather than 4. Currently the DRAM allocation uses 2MB
-	 * pages only but user memory could have been allocated with one of the
-	 * two page sizes. Since this is a common code for all the three cases,
-	 * we need this hugs page check.
+	 * This mapping function can map a page or a huge page. For huge page
+	 * there are only 3 hops rather than 4. Currently the DRAM allocation
+	 * uses huge pages only but user memory could have been allocated with
+	 * one of the two page sizes. Since this is a common code for all the
+	 * three cases, we need this hugs page check.
 	 */
-	is_huge = page_size == PAGE_SIZE_2MB;
-
-	is_dram_addr = hl_mem_area_inside_range(virt_addr, page_size,
-				prop->va_space_dram_start_address,
-				prop->va_space_dram_end_address);
+	is_huge = page_size == mmu_prop->huge_page_size;
 
 	if (is_dram_addr && !is_huge) {
 		dev_err(hdev->dev, "DRAM mapping should use huge pages only\n");
@@ -738,28 +770,28 @@ static int _hl_mmu_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr,
 	}
 
 	hop0_addr = get_hop0_addr(ctx);
-	hop0_pte_addr = get_hop0_pte_addr(ctx, hop0_addr, virt_addr);
+	hop0_pte_addr = get_hop0_pte_addr(ctx, mmu_prop, hop0_addr, virt_addr);
 	curr_pte = *(u64 *) (uintptr_t) hop0_pte_addr;
 
 	hop1_addr = get_alloc_next_hop_addr(ctx, curr_pte, &hop1_new);
 	if (hop1_addr == ULLONG_MAX)
 		goto err;
 
-	hop1_pte_addr = get_hop1_pte_addr(ctx, hop1_addr, virt_addr);
+	hop1_pte_addr = get_hop1_pte_addr(ctx, mmu_prop, hop1_addr, virt_addr);
 	curr_pte = *(u64 *) (uintptr_t) hop1_pte_addr;
 
 	hop2_addr = get_alloc_next_hop_addr(ctx, curr_pte, &hop2_new);
 	if (hop2_addr == ULLONG_MAX)
 		goto err;
 
-	hop2_pte_addr = get_hop2_pte_addr(ctx, hop2_addr, virt_addr);
+	hop2_pte_addr = get_hop2_pte_addr(ctx, mmu_prop, hop2_addr, virt_addr);
 	curr_pte = *(u64 *) (uintptr_t) hop2_pte_addr;
 
 	hop3_addr = get_alloc_next_hop_addr(ctx, curr_pte, &hop3_new);
 	if (hop3_addr == ULLONG_MAX)
 		goto err;
 
-	hop3_pte_addr = get_hop3_pte_addr(ctx, hop3_addr, virt_addr);
+	hop3_pte_addr = get_hop3_pte_addr(ctx, mmu_prop, hop3_addr, virt_addr);
 	curr_pte = *(u64 *) (uintptr_t) hop3_pte_addr;
 
 	if (!is_huge) {
@@ -767,13 +799,14 @@ static int _hl_mmu_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr,
 		if (hop4_addr == ULLONG_MAX)
 			goto err;
 
-		hop4_pte_addr = get_hop4_pte_addr(ctx, hop4_addr, virt_addr);
+		hop4_pte_addr = get_hop4_pte_addr(ctx, mmu_prop, hop4_addr,
+							virt_addr);
 		curr_pte = *(u64 *) (uintptr_t) hop4_pte_addr;
 	}
 
 	if (hdev->dram_default_page_mapping && is_dram_addr) {
 		u64 default_pte = (prop->mmu_dram_default_page_addr &
-					PTE_PHYS_ADDR_MASK) | LAST_MASK |
+					HOP_PHYS_ADDR_MASK) | LAST_MASK |
 						PAGE_PRESENT_MASK;
 
 		if (curr_pte != default_pte) {
@@ -813,7 +846,7 @@ static int _hl_mmu_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr,
 		goto err;
 	}
 
-	curr_pte = (phys_addr & PTE_PHYS_ADDR_MASK) | LAST_MASK
+	curr_pte = (phys_addr & HOP_PHYS_ADDR_MASK) | LAST_MASK
 			| PAGE_PRESENT_MASK;
 
 	if (is_huge)
@@ -823,25 +856,25 @@ static int _hl_mmu_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr,
 
 	if (hop1_new) {
 		curr_pte =
-			(hop1_addr & PTE_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK;
+			(hop1_addr & HOP_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK;
 		write_pte(ctx, hop0_pte_addr, curr_pte);
 	}
 	if (hop2_new) {
 		curr_pte =
-			(hop2_addr & PTE_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK;
+			(hop2_addr & HOP_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK;
 		write_pte(ctx, hop1_pte_addr, curr_pte);
 		get_pte(ctx, hop1_addr);
 	}
 	if (hop3_new) {
 		curr_pte =
-			(hop3_addr & PTE_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK;
+			(hop3_addr & HOP_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK;
 		write_pte(ctx, hop2_pte_addr, curr_pte);
 		get_pte(ctx, hop2_addr);
 	}
 
 	if (!is_huge) {
 		if (hop4_new) {
-			curr_pte = (hop4_addr & PTE_PHYS_ADDR_MASK) |
+			curr_pte = (hop4_addr & HOP_PHYS_ADDR_MASK) |
 					PAGE_PRESENT_MASK;
 			write_pte(ctx, hop3_pte_addr, curr_pte);
 			get_pte(ctx, hop3_addr);
@@ -890,25 +923,36 @@ err:
 int hl_mmu_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr, u32 page_size)
 {
 	struct hl_device *hdev = ctx->hdev;
+	struct asic_fixed_properties *prop = &hdev->asic_prop;
+	struct hl_mmu_properties *mmu_prop;
 	u64 real_virt_addr, real_phys_addr;
 	u32 real_page_size, npages;
 	int i, rc, mapped_cnt = 0;
+	bool is_dram_addr;
 
 	if (!hdev->mmu_enable)
 		return 0;
 
+	is_dram_addr = hl_mem_area_inside_range(virt_addr, prop->dmmu.page_size,
+				prop->va_space_dram_start_address,
+				prop->va_space_dram_end_address);
+
+	mmu_prop = is_dram_addr ? &prop->dmmu : &prop->pmmu;
+
 	/*
-	 * The H/W handles mapping of 4KB/2MB page. Hence if the host page size
-	 * is bigger, we break it to sub-pages and map them separately.
+	 * The H/W handles mapping of specific page sizes. Hence if the page
+	 * size is bigger, we break it to sub-pages and map them separately.
 	 */
-	if ((page_size % PAGE_SIZE_2MB) == 0) {
-		real_page_size = PAGE_SIZE_2MB;
-	} else if ((page_size % PAGE_SIZE_4KB) == 0) {
-		real_page_size = PAGE_SIZE_4KB;
+	if ((page_size % mmu_prop->huge_page_size) == 0) {
+		real_page_size = mmu_prop->huge_page_size;
+	} else if ((page_size % mmu_prop->page_size) == 0) {
+		real_page_size = mmu_prop->page_size;
 	} else {
 		dev_err(hdev->dev,
-			"page size of %u is not 4KB nor 2MB aligned, can't map\n",
-				page_size);
+			"page size of %u is not %dKB nor %dMB aligned, can't unmap\n",
+			page_size,
+			mmu_prop->page_size >> 10,
+			mmu_prop->huge_page_size >> 20);
 
 		return -EFAULT;
 	}
@@ -923,7 +967,7 @@ int hl_mmu_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr, u32 page_size)
 
 	for (i = 0 ; i < npages ; i++) {
 		rc = _hl_mmu_map(ctx, real_virt_addr, real_phys_addr,
-				real_page_size);
+				real_page_size, is_dram_addr);
 		if (rc)
 			goto err;
 
@@ -937,7 +981,7 @@ int hl_mmu_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr, u32 page_size)
 err:
 	real_virt_addr = virt_addr;
 	for (i = 0 ; i < mapped_cnt ; i++) {
-		if (_hl_mmu_unmap(ctx, real_virt_addr))
+		if (_hl_mmu_unmap(ctx, real_virt_addr, is_dram_addr))
 			dev_warn_ratelimited(hdev->dev,
 				"failed to unmap va: 0x%llx\n", real_virt_addr);
 
diff --git a/include/uapi/misc/habanalabs.h b/include/uapi/misc/habanalabs.h
index 39c4ea51a719..4faa2c9767e5 100644
--- a/include/uapi/misc/habanalabs.h
+++ b/include/uapi/misc/habanalabs.h
@@ -88,13 +88,19 @@ enum hl_device_status {
  *                         internal engine.
  * HL_INFO_DEVICE_STATUS - Retrieve the device's status. This opcode doesn't
  *                         require an open context.
- * HL_INFO_DEVICE_UTILIZATION - Retrieve the total utilization of the device
- *                              over the last period specified by the user.
- *                              The period can be between 100ms to 1s, in
- *                              resolution of 100ms. The return value is a
- *                              percentage of the utilization rate.
+ * HL_INFO_DEVICE_UTILIZATION  - Retrieve the total utilization of the device
+ *                               over the last period specified by the user.
+ *                               The period can be between 100ms to 1s, in
+ *                               resolution of 100ms. The return value is a
+ *                               percentage of the utilization rate.
  * HL_INFO_HW_EVENTS_AGGREGATE - Receive an array describing how many times each
  *                               event occurred since the driver was loaded.
+ * HL_INFO_CLK_RATE            - Retrieve the current and maximum clock rate
+ *                               of the device in MHz. The maximum clock rate is
+ *                               configurable via sysfs parameter
+ * HL_INFO_RESET_COUNT   - Retrieve the counts of the soft and hard reset
+ *                         operations performed on the device since the last
+ *                         time the driver was loaded.
  */
 #define HL_INFO_HW_IP_INFO		0
 #define HL_INFO_HW_EVENTS		1
@@ -103,8 +109,11 @@ enum hl_device_status {
 #define HL_INFO_DEVICE_STATUS		4
 #define HL_INFO_DEVICE_UTILIZATION	6
 #define HL_INFO_HW_EVENTS_AGGREGATE	7
+#define HL_INFO_CLK_RATE		8
+#define HL_INFO_RESET_COUNT		9
 
 #define HL_INFO_VERSION_MAX_LEN	128
+#define HL_INFO_CARD_NAME_MAX_LEN	16
 
 struct hl_info_hw_ip_info {
 	__u64 sram_base_address;
@@ -123,6 +132,7 @@ struct hl_info_hw_ip_info {
 	__u8 dram_enabled;
 	__u8 pad[2];
 	__u8 armcp_version[HL_INFO_VERSION_MAX_LEN];
+	__u8 card_name[HL_INFO_CARD_NAME_MAX_LEN];
 };
 
 struct hl_info_dram_usage {
@@ -149,6 +159,16 @@ struct hl_info_device_utilization {
 	__u32 pad;
 };
 
+struct hl_info_clk_rate {
+	__u32 cur_clk_rate_mhz;
+	__u32 max_clk_rate_mhz;
+};
+
+struct hl_info_reset_count {
+	__u32 hard_reset_cnt;
+	__u32 soft_reset_cnt;
+};
+
 struct hl_info_args {
 	/* Location of relevant struct in userspace */
 	__u64 return_pointer;
@@ -181,13 +201,15 @@ struct hl_info_args {
 /* Opcode to destroy previously created command buffer */
 #define HL_CB_OP_DESTROY	1
 
+#define HL_MAX_CB_SIZE		0x200000	/* 2MB */
+
 struct hl_cb_in {
 	/* Handle of CB or 0 if we want to create one */
 	__u64 cb_handle;
 	/* HL_CB_OP_* */
 	__u32 op;
-	/* Size of CB. Maximum size is 2MB. The minimum size that will be
-	 * allocated, regardless of this parameter's value, is PAGE_SIZE
+	/* Size of CB. Maximum size is HL_MAX_CB_SIZE. The minimum size that
+	 * will be allocated, regardless of this parameter's value, is PAGE_SIZE
 	 */
 	__u32 cb_size;
 	/* Context ID - Currently not in use */
@@ -233,6 +255,8 @@ struct hl_cs_chunk {
 
 #define HL_CS_STATUS_SUCCESS		0
 
+#define HL_MAX_JOBS_PER_CS		512
+
 struct hl_cs_in {
 	/* this holds address of array of hl_cs_chunk for restore phase */
 	__u64 chunks_restore;
@@ -242,9 +266,13 @@ struct hl_cs_in {
 	 * Currently not in use
 	 */
 	__u64 chunks_store;
-	/* Number of chunks in restore phase array */
+	/* Number of chunks in restore phase array. Maximum number is
+	 * HL_MAX_JOBS_PER_CS
+	 */
 	__u32 num_chunks_restore;
-	/* Number of chunks in execution array */
+	/* Number of chunks in execution array. Maximum number is
+	 * HL_MAX_JOBS_PER_CS
+	 */
 	__u32 num_chunks_execute;
 	/* Number of chunks in restore phase array - Currently not in use */
 	__u32 num_chunks_store;
@@ -589,7 +617,7 @@ struct hl_debug_args {
  *
  * The user can call this IOCTL with a handle it received from the CS IOCTL
  * to wait until the handle's CS has finished executing. The user will wait
- * inside the kernel until the CS has finished or until the user-requeusted
+ * inside the kernel until the CS has finished or until the user-requested
  * timeout has expired.
  *
  * The return value of the IOCTL is a standard Linux error code. The possible