Merge tag 'misc-habanalabs-next-2021-12-27' of...

Merge tag 'misc-habanalabs-next-2021-12-27' of https://git.kernel.org/pub/scm/linux/kernel/git/ogabbay/linux into char-misc-next

Oded writes:

This tag contains habanalabs driver changes for v5.17:

- Support reset-during-reset. In case the f/w notifies the driver
  that the f/w is going to reset the device, the driver should
  support that even if it is in the middle of doing another
  reset

- Support events from f/w that arrive during device resets.
  These events would be ignored which is bad as critical errors
  would not be reported and treated by the driver.

- Don't kill processes that hold the control device open during
  hard-reset of the device. The control device operations can't
  crash if done during hard-reset. And usually, only monitoring
  applications are using the control device, so killing them
  defies their purpose.

- Fix handling of hwmon nodes when working with legacy f/w

- Change the compute context pointer to be boolean. This pointer
  was abused by multiple code paths that wanted fast access to
  the compute context structure.

- Add uapi to fetch historical errors. This is necessary as errors
  sometimes result in hard-reset where the user application is
  being terminated.

- Optimize GAUDI's MMU cache invalidation.

- Add support for loading the latest f/w.

- Add uapi to fetch HBM replacement and pending rows information.

- Multiple bug fixes to the reset code.

- Multiple bug fixes for Multi-CS ioctl code.

- Multiple bug fixes for wait-for-interrupt ioctl code.

- Many small bug fixes and cleanups.

* tag 'misc-habanalabs-next-2021-12-27' of https://git.kernel.org/pub/scm/linux/kernel/git/ogabbay/linux: (70 commits)
  habanalabs: support hard-reset scheduling during soft-reset
  habanalabs: add a lock to protect multiple reset variables
  habanalabs: refactor reset information variables
  habanalabs: handle skip multi-CS if handling not done
  habanalabs: add CPU-CP packet for engine core ASID cfg
  habanalabs: replace some -ENOTTY with -EINVAL
  habanalabs: fix comments according to kernel-doc
  habanalabs: fix endianness when reading cpld version
  habanalabs: change wait_for_interrupt implementation
  habanalabs: prevent wait if CS in multi-CS list completed
  habanalabs: modify cpu boot status error print
  habanalabs: clean MMU headers definitions
  habanalabs: expose soft reset sysfs nodes for inference ASIC
  habanalabs: sysfs support for two infineon versions
  habanalabs: keep control device alive during hard reset
  habanalabs: fix hwmon handling for legacy f/w
  habanalabs: add current PI value to cpu packets
  habanalabs: remove in_debug check in device open
  habanalabs: return correct clock throttling period
  habanalabs: wait again for multi-CS if no CS completed
  ...

Documentation/ABI/testing/debugfs-driver-habanalabs

@@ -21,11 +21,11 @@ Description:    Allow the root user to disable/enable in runtime the clock
                 a different engine to disable/enable its clock gating feature.
                 The bitmask is composed of 20 bits:
 
-		=======   ============
+                =======   ============
                 0  -  7   DMA channels
                 8  - 11   MME engines
                 12 - 19   TPC engines
-		=======   ============
+                =======   ============
 
                 The bit's location of a specific engine can be determined
                 using (1 << GAUDI_ENGINE_ID_*). GAUDI_ENGINE_ID_* values
@@ -155,6 +155,13 @@ Description:    Triggers an I2C transaction that is generated by the device's
                 CPU. Writing to this file generates a write transaction while
                 reading from the file generates a read transaction
 
+What:           /sys/kernel/debug/habanalabs/hl<n>/i2c_len
+Date:           Dec 2021
+KernelVersion:  5.17
+Contact:        obitton@habana.ai
+Description:    Sets I2C length in bytes for I2C transaction that is generated by
+                the device's CPU
+
 What:           /sys/kernel/debug/habanalabs/hl<n>/i2c_reg
 Date:           Jan 2019
 KernelVersion:  5.1
@@ -226,12 +233,6 @@ Description:    Gets the state dump occurring on a CS timeout or failure.
                 Writing an integer X discards X state dumps, so that the
                 next read would return X+1-st newest state dump.
 
-What:           /sys/kernel/debug/habanalabs/hl<n>/timeout_locked
-Date:           Sep 2021
-KernelVersion:  5.16
-Contact:        obitton@habana.ai
-Description:    Sets the command submission timeout value in seconds.
-
 What:           /sys/kernel/debug/habanalabs/hl<n>/stop_on_err
 Date:           Mar 2020
 KernelVersion:  5.6
@@ -239,6 +240,12 @@ Contact:        ogabbay@kernel.org
 Description:    Sets the stop-on_error option for the device engines. Value of
                 "0" is for disable, otherwise enable.
 
+What:           /sys/kernel/debug/habanalabs/hl<n>/timeout_locked
+Date:           Sep 2021
+KernelVersion:  5.16
+Contact:        obitton@habana.ai
+Description:    Sets the command submission timeout value in seconds.
+
 What:           /sys/kernel/debug/habanalabs/hl<n>/userptr
 Date:           Jan 2019
 KernelVersion:  5.1

drivers/misc/habanalabs/common/command_buffer.c

@@ -57,7 +57,7 @@ static int cb_map_mem(struct hl_ctx *ctx, struct hl_cb *cb)
 		}
 
 		va_block->start = virt_addr;
-		va_block->end = virt_addr + page_size;
+		va_block->end = virt_addr + page_size - 1;
 		va_block->size = page_size;
 		list_add_tail(&va_block->node, &cb->va_block_list);
 	}
@@ -80,13 +80,13 @@ static int cb_map_mem(struct hl_ctx *ctx, struct hl_cb *cb)
 		offset += va_block->size;
 	}
 
-	hdev->asic_funcs->mmu_invalidate_cache(hdev, false, VM_TYPE_USERPTR);
+	rc = hl_mmu_invalidate_cache(hdev, false, MMU_OP_USERPTR | MMU_OP_SKIP_LOW_CACHE_INV);
 
 	mutex_unlock(&ctx->mmu_lock);
 
 	cb->is_mmu_mapped = true;
 
-	return 0;
+	return rc;
 
 err_va_umap:
 	list_for_each_entry(va_block, &cb->va_block_list, node) {
@@ -97,7 +97,7 @@ static int cb_map_mem(struct hl_ctx *ctx, struct hl_cb *cb)
 		offset -= va_block->size;
 	}
 
-	hdev->asic_funcs->mmu_invalidate_cache(hdev, true, VM_TYPE_USERPTR);
+	rc = hl_mmu_invalidate_cache(hdev, true, MMU_OP_USERPTR);
 
 	mutex_unlock(&ctx->mmu_lock);
 
@@ -126,7 +126,7 @@ static void cb_unmap_mem(struct hl_ctx *ctx, struct hl_cb *cb)
 					"Failed to unmap CB's va 0x%llx\n",
 					va_block->start);
 
-	hdev->asic_funcs->mmu_invalidate_cache(hdev, true, VM_TYPE_USERPTR);
+	hl_mmu_invalidate_cache(hdev, true, MMU_OP_USERPTR);
 
 	mutex_unlock(&ctx->mmu_lock);
 
@@ -250,8 +250,7 @@ int hl_cb_create(struct hl_device *hdev, struct hl_cb_mgr *mgr,
 	 * Can't use generic function to check this because of special case
 	 * where we create a CB as part of the reset process
 	 */
-	if ((hdev->disabled) || ((atomic_read(&hdev->in_reset)) &&
-					(ctx_id != HL_KERNEL_ASID_ID))) {
+	if ((hdev->disabled) || (hdev->reset_info.in_reset && (ctx_id != HL_KERNEL_ASID_ID))) {
 		dev_warn_ratelimited(hdev->dev,
 			"Device is disabled or in reset. Can't create new CBs\n");
 		rc = -EBUSY;
@@ -380,8 +379,9 @@ int hl_cb_destroy(struct hl_device *hdev, struct hl_cb_mgr *mgr, u64 cb_handle)
 }
 
 static int hl_cb_info(struct hl_device *hdev, struct hl_cb_mgr *mgr,
-			u64 cb_handle, u32 *usage_cnt)
+			u64 cb_handle, u32 flags, u32 *usage_cnt, u64 *device_va)
 {
+	struct hl_vm_va_block *va_block;
 	struct hl_cb *cb;
 	u32 handle;
 	int rc = 0;
@@ -402,7 +402,18 @@ static int hl_cb_info(struct hl_device *hdev, struct hl_cb_mgr *mgr,
 		goto out;
 	}
 
-	*usage_cnt = atomic_read(&cb->cs_cnt);
+	if (flags & HL_CB_FLAGS_GET_DEVICE_VA) {
+		va_block = list_first_entry(&cb->va_block_list, struct hl_vm_va_block, node);
+		if (va_block) {
+			*device_va = va_block->start;
+		} else {
+			dev_err(hdev->dev, "CB is not mapped to the device's MMU\n");
+			rc = -EINVAL;
+			goto out;
+		}
+	} else {
+		*usage_cnt = atomic_read(&cb->cs_cnt);
+	}
 
 out:
 	spin_unlock(&mgr->cb_lock);
@@ -414,7 +425,7 @@ int hl_cb_ioctl(struct hl_fpriv *hpriv, void *data)
 	union hl_cb_args *args = data;
 	struct hl_device *hdev = hpriv->hdev;
 	enum hl_device_status status;
-	u64 handle = 0;
+	u64 handle = 0, device_va;
 	u32 usage_cnt = 0;
 	int rc;
 
@@ -450,13 +461,20 @@ int hl_cb_ioctl(struct hl_fpriv *hpriv, void *data)
 
 	case HL_CB_OP_INFO:
 		rc = hl_cb_info(hdev, &hpriv->cb_mgr, args->in.cb_handle,
-				&usage_cnt);
-		memset(args, 0, sizeof(*args));
-		args->out.usage_cnt = usage_cnt;
+				args->in.flags,
+				&usage_cnt,
+				&device_va);
+
+		memset(&args->out, 0, sizeof(args->out));
+
+		if (args->in.flags & HL_CB_FLAGS_GET_DEVICE_VA)
+			args->out.device_va = device_va;
+		else
+			args->out.usage_cnt = usage_cnt;
 		break;
 
 	default:
-		rc = -ENOTTY;
+		rc = -EINVAL;
 		break;
 	}
 

drivers/misc/habanalabs/common/context.c

 // SPDX-License-Identifier: GPL-2.0
 
 /*
- * Copyright 2016-2019 HabanaLabs, Ltd.
+ * Copyright 2016-2021 HabanaLabs, Ltd.
  * All Rights Reserved.
  */
 
@@ -13,13 +13,13 @@ void hl_encaps_handle_do_release(struct kref *ref)
 {
 	struct hl_cs_encaps_sig_handle *handle =
 		container_of(ref, struct hl_cs_encaps_sig_handle, refcount);
-	struct hl_ctx *ctx = handle->hdev->compute_ctx;
-	struct hl_encaps_signals_mgr *mgr = &ctx->sig_mgr;
+	struct hl_encaps_signals_mgr *mgr = &handle->ctx->sig_mgr;
 
 	spin_lock(&mgr->lock);
 	idr_remove(&mgr->handles, handle->id);
 	spin_unlock(&mgr->lock);
 
+	hl_ctx_put(handle->ctx);
 	kfree(handle);
 }
 
@@ -27,8 +27,7 @@ static void hl_encaps_handle_do_release_sob(struct kref *ref)
 {
 	struct hl_cs_encaps_sig_handle *handle =
 		container_of(ref, struct hl_cs_encaps_sig_handle, refcount);
-	struct hl_ctx *ctx = handle->hdev->compute_ctx;
-	struct hl_encaps_signals_mgr *mgr = &ctx->sig_mgr;
+	struct hl_encaps_signals_mgr *mgr = &handle->ctx->sig_mgr;
 
 	/* if we're here, then there was a signals reservation but cs with
 	 * encaps signals wasn't submitted, so need to put refcount
@@ -40,6 +39,7 @@ static void hl_encaps_handle_do_release_sob(struct kref *ref)
 	idr_remove(&mgr->handles, handle->id);
 	spin_unlock(&mgr->lock);
 
+	hl_ctx_put(handle->ctx);
 	kfree(handle);
 }
 
@@ -97,11 +97,9 @@ static void hl_ctx_fini(struct hl_ctx *ctx)
 		/* The engines are stopped as there is no executing CS, but the
 		 * Coresight might be still working by accessing addresses
 		 * related to the stopped engines. Hence stop it explicitly.
-		 * Stop only if this is the compute context, as there can be
-		 * only one compute context
 		 */
-		if ((hdev->in_debug) && (hdev->compute_ctx == ctx))
-			hl_device_set_debug_mode(hdev, false);
+		if (hdev->in_debug)
+			hl_device_set_debug_mode(hdev, ctx, false);
 
 		hdev->asic_funcs->ctx_fini(ctx);
 		hl_cb_va_pool_fini(ctx);
@@ -167,7 +165,7 @@ int hl_ctx_create(struct hl_device *hdev, struct hl_fpriv *hpriv)
 	hpriv->ctx = ctx;
 
 	/* TODO: remove the following line for multiple process support */
-	hdev->compute_ctx = ctx;
+	hdev->is_compute_ctx_active = true;
 
 	return 0;
 
@@ -274,6 +272,27 @@ int hl_ctx_put(struct hl_ctx *ctx)
 	return kref_put(&ctx->refcount, hl_ctx_do_release);
 }
 
+struct hl_ctx *hl_get_compute_ctx(struct hl_device *hdev)
+{
+	struct hl_ctx *ctx = NULL;
+	struct hl_fpriv *hpriv;
+
+	mutex_lock(&hdev->fpriv_list_lock);
+
+	list_for_each_entry(hpriv, &hdev->fpriv_list, dev_node) {
+		/* There can only be a single user which has opened the compute device, so exit
+		 * immediately once we find him
+		 */
+		ctx = hpriv->ctx;
+		hl_ctx_get(hdev, ctx);
+		break;
+	}
+
+	mutex_unlock(&hdev->fpriv_list_lock);
+
+	return ctx;
+}
+
 /*
  * hl_ctx_get_fence_locked - get CS fence under CS lock
  *

drivers/misc/habanalabs/common/debugfs.c

 // SPDX-License-Identifier: GPL-2.0
 
 /*
- * Copyright 2016-2019 HabanaLabs, Ltd.
+ * Copyright 2016-2021 HabanaLabs, Ltd.
  * All Rights Reserved.
  */
 
@@ -15,19 +15,25 @@
 #define MMU_ADDR_BUF_SIZE	40
 #define MMU_ASID_BUF_SIZE	10
 #define MMU_KBUF_SIZE		(MMU_ADDR_BUF_SIZE + MMU_ASID_BUF_SIZE)
+#define I2C_MAX_TRANSACTION_LEN	8
 
 static struct dentry *hl_debug_root;
 
 static int hl_debugfs_i2c_read(struct hl_device *hdev, u8 i2c_bus, u8 i2c_addr,
-				u8 i2c_reg, long *val)
+				u8 i2c_reg, u8 i2c_len, u64 *val)
 {
 	struct cpucp_packet pkt;
-	u64 result;
 	int rc;
 
 	if (!hl_device_operational(hdev, NULL))
 		return -EBUSY;
 
+	if (i2c_len > I2C_MAX_TRANSACTION_LEN) {
+		dev_err(hdev->dev, "I2C transaction length %u, exceeds maximum of %u\n",
+				i2c_len, I2C_MAX_TRANSACTION_LEN);
+		return -EINVAL;
+	}
+
 	memset(&pkt, 0, sizeof(pkt));
 
 	pkt.ctl = cpu_to_le32(CPUCP_PACKET_I2C_RD <<
@@ -35,12 +41,10 @@ static int hl_debugfs_i2c_read(struct hl_device *hdev, u8 i2c_bus, u8 i2c_addr,
 	pkt.i2c_bus = i2c_bus;
 	pkt.i2c_addr = i2c_addr;
 	pkt.i2c_reg = i2c_reg;
+	pkt.i2c_len = i2c_len;
 
 	rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt),
-						0, &result);
-
-	*val = (long) result;
-
+						0, val);
 	if (rc)
 		dev_err(hdev->dev, "Failed to read from I2C, error %d\n", rc);
 
@@ -48,7 +52,7 @@ static int hl_debugfs_i2c_read(struct hl_device *hdev, u8 i2c_bus, u8 i2c_addr,
 }
 
 static int hl_debugfs_i2c_write(struct hl_device *hdev, u8 i2c_bus, u8 i2c_addr,
-				u8 i2c_reg, u32 val)
+				u8 i2c_reg, u8 i2c_len, u64 val)
 {
 	struct cpucp_packet pkt;
 	int rc;
@@ -56,6 +60,12 @@ static int hl_debugfs_i2c_write(struct hl_device *hdev, u8 i2c_bus, u8 i2c_addr,
 	if (!hl_device_operational(hdev, NULL))
 		return -EBUSY;
 
+	if (i2c_len > I2C_MAX_TRANSACTION_LEN) {
+		dev_err(hdev->dev, "I2C transaction length %u, exceeds maximum of %u\n",
+				i2c_len, I2C_MAX_TRANSACTION_LEN);
+		return -EINVAL;
+	}
+
 	memset(&pkt, 0, sizeof(pkt));
 
 	pkt.ctl = cpu_to_le32(CPUCP_PACKET_I2C_WR <<
@@ -63,6 +73,7 @@ static int hl_debugfs_i2c_write(struct hl_device *hdev, u8 i2c_bus, u8 i2c_addr,
 	pkt.i2c_bus = i2c_bus;
 	pkt.i2c_addr = i2c_addr;
 	pkt.i2c_reg = i2c_reg;
+	pkt.i2c_len = i2c_len;
 	pkt.value = cpu_to_le64(val);
 
 	rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt),
@@ -235,6 +246,8 @@ static int vm_show(struct seq_file *s, void *data)
 	struct hl_vm_hash_node *hnode;
 	struct hl_userptr *userptr;
 	struct hl_vm_phys_pg_pack *phys_pg_pack = NULL;
+	struct hl_va_range *va_range;
+	struct hl_vm_va_block *va_block;
 	enum vm_type *vm_type;
 	bool once = true;
 	u64 j;
@@ -314,6 +327,25 @@ static int vm_show(struct seq_file *s, void *data)
 
 	spin_unlock(&dev_entry->ctx_mem_hash_spinlock);
 
+	ctx = hl_get_compute_ctx(dev_entry->hdev);
+	if (ctx) {
+		seq_puts(s, "\nVA ranges:\n\n");
+		for (i = HL_VA_RANGE_TYPE_HOST ; i < HL_VA_RANGE_TYPE_MAX ; ++i) {
+			va_range = ctx->va_range[i];
+			seq_printf(s, "   va_range %d\n", i);
+			seq_puts(s, "---------------------\n");
+			mutex_lock(&va_range->lock);
+			list_for_each_entry(va_block, &va_range->list, node) {
+				seq_printf(s, "%#16llx - %#16llx (%#llx)\n",
+					   va_block->start, va_block->end,
+					   va_block->size);
+			}
+			mutex_unlock(&va_range->lock);
+			seq_puts(s, "\n");
+		}
+		hl_ctx_put(ctx);
+	}
+
 	if (!once)
 		seq_puts(s, "\n");
 
@@ -407,7 +439,7 @@ static int mmu_show(struct seq_file *s, void *data)
 	if (dev_entry->mmu_asid == HL_KERNEL_ASID_ID)
 		ctx = hdev->kernel_ctx;
 	else
-		ctx = hdev->compute_ctx;
+		ctx = hl_get_compute_ctx(hdev);
 
 	if (!ctx) {
 		dev_err(hdev->dev, "no ctx available\n");
@@ -495,7 +527,7 @@ 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)) {
+	if (hdev->reset_info.in_reset) {
 		dev_warn_ratelimited(hdev->dev,
 				"Can't check device idle during reset\n");
 		return 0;
@@ -560,7 +592,7 @@ static int device_va_to_pa(struct hl_device *hdev, u64 virt_addr, u32 size,
 			u64 *phys_addr)
 {
 	struct hl_vm_phys_pg_pack *phys_pg_pack;
-	struct hl_ctx *ctx = hdev->compute_ctx;
+	struct hl_ctx *ctx;
 	struct hl_vm_hash_node *hnode;
 	u64 end_address, range_size;
 	struct hl_userptr *userptr;
@@ -568,6 +600,8 @@ static int device_va_to_pa(struct hl_device *hdev, u64 virt_addr, u32 size,
 	bool valid = false;
 	int i, rc = 0;
 
+	ctx = hl_get_compute_ctx(hdev);
+
 	if (!ctx) {
 		dev_err(hdev->dev, "no ctx available\n");
 		return -EINVAL;
@@ -624,7 +658,7 @@ static ssize_t hl_data_read32(struct file *f, char __user *buf,
 	ssize_t rc;
 	u32 val;
 
-	if (atomic_read(&hdev->in_reset)) {
+	if (hdev->reset_info.in_reset) {
 		dev_warn_ratelimited(hdev->dev, "Can't read during reset\n");
 		return 0;
 	}
@@ -660,7 +694,7 @@ static ssize_t hl_data_write32(struct file *f, const char __user *buf,
 	u32 value;
 	ssize_t rc;
 
-	if (atomic_read(&hdev->in_reset)) {
+	if (hdev->reset_info.in_reset) {
 		dev_warn_ratelimited(hdev->dev, "Can't write during reset\n");
 		return 0;
 	}
@@ -697,7 +731,7 @@ static ssize_t hl_data_read64(struct file *f, char __user *buf,
 	ssize_t rc;
 	u64 val;
 
-	if (atomic_read(&hdev->in_reset)) {
+	if (hdev->reset_info.in_reset) {
 		dev_warn_ratelimited(hdev->dev, "Can't read during reset\n");
 		return 0;
 	}
@@ -733,7 +767,7 @@ static ssize_t hl_data_write64(struct file *f, const char __user *buf,
 	u64 value;
 	ssize_t rc;
 
-	if (atomic_read(&hdev->in_reset)) {
+	if (hdev->reset_info.in_reset) {
 		dev_warn_ratelimited(hdev->dev, "Can't write during reset\n");
 		return 0;
 	}
@@ -768,7 +802,7 @@ static ssize_t hl_dma_size_write(struct file *f, const char __user *buf,
 	ssize_t rc;
 	u32 size;
 
-	if (atomic_read(&hdev->in_reset)) {
+	if (hdev->reset_info.in_reset) {
 		dev_warn_ratelimited(hdev->dev, "Can't DMA during reset\n");
 		return 0;
 	}
@@ -874,22 +908,22 @@ static ssize_t hl_i2c_data_read(struct file *f, char __user *buf,
 	struct hl_dbg_device_entry *entry = file_inode(f)->i_private;
 	struct hl_device *hdev = entry->hdev;
 	char tmp_buf[32];
-	long val;
+	u64 val;
 	ssize_t rc;
 
 	if (*ppos)
 		return 0;
 
 	rc = hl_debugfs_i2c_read(hdev, entry->i2c_bus, entry->i2c_addr,
-			entry->i2c_reg, &val);
+			entry->i2c_reg, entry->i2c_len, &val);
 	if (rc) {
 		dev_err(hdev->dev,
-			"Failed to read from I2C bus %d, addr %d, reg %d\n",
-			entry->i2c_bus, entry->i2c_addr, entry->i2c_reg);
+			"Failed to read from I2C bus %d, addr %d, reg %d, len %d\n",
+			entry->i2c_bus, entry->i2c_addr, entry->i2c_reg, entry->i2c_len);
 		return rc;
 	}
 
-	sprintf(tmp_buf, "0x%02lx\n", val);
+	sprintf(tmp_buf, "%#02llx\n", val);
 	rc = simple_read_from_buffer(buf, count, ppos, tmp_buf,
 			strlen(tmp_buf));
 
@@ -901,19 +935,19 @@ static ssize_t hl_i2c_data_write(struct file *f, const char __user *buf,
 {
 	struct hl_dbg_device_entry *entry = file_inode(f)->i_private;
 	struct hl_device *hdev = entry->hdev;
-	u32 value;
+	u64 value;
 	ssize_t rc;
 
-	rc = kstrtouint_from_user(buf, count, 16, &value);
+	rc = kstrtou64_from_user(buf, count, 16, &value);
 	if (rc)
 		return rc;
 
 	rc = hl_debugfs_i2c_write(hdev, entry->i2c_bus, entry->i2c_addr,
-			entry->i2c_reg, value);
+			entry->i2c_reg, entry->i2c_len, value);
 	if (rc) {
 		dev_err(hdev->dev,
-			"Failed to write 0x%02x to I2C bus %d, addr %d, reg %d\n",
-			value, entry->i2c_bus, entry->i2c_addr, entry->i2c_reg);
+			"Failed to write %#02llx to I2C bus %d, addr %d, reg %d, len %d\n",
+			value, entry->i2c_bus, entry->i2c_addr, entry->i2c_reg, entry->i2c_len);
 		return rc;
 	}
 
@@ -1043,7 +1077,7 @@ static ssize_t hl_clk_gate_write(struct file *f, const char __user *buf,
 	u64 value;
 	ssize_t rc;
 
-	if (atomic_read(&hdev->in_reset)) {
+	if (hdev->reset_info.in_reset) {
 		dev_warn_ratelimited(hdev->dev,
 				"Can't change clock gating during reset\n");
 		return 0;
@@ -1085,7 +1119,7 @@ static ssize_t hl_stop_on_err_write(struct file *f, const char __user *buf,
 	u32 value;
 	ssize_t rc;
 
-	if (atomic_read(&hdev->in_reset)) {
+	if (hdev->reset_info.in_reset) {
 		dev_warn_ratelimited(hdev->dev,
 				"Can't change stop on error during reset\n");
 		return 0;
@@ -1396,6 +1430,11 @@ void hl_debugfs_add_device(struct hl_device *hdev)
 				dev_entry->root,
 				&dev_entry->i2c_reg);
 
+	debugfs_create_u8("i2c_len",
+				0644,
+				dev_entry->root,
+				&dev_entry->i2c_len);
+
 	debugfs_create_file("i2c_data",
 				0644,
 				dev_entry->root,
@@ -1458,7 +1497,7 @@ void hl_debugfs_add_device(struct hl_device *hdev)
 	debugfs_create_x8("skip_reset_on_timeout",
 				0644,
 				dev_entry->root,
-				&hdev->skip_reset_on_timeout);
+				&hdev->reset_info.skip_reset_on_timeout);
 
 	debugfs_create_file("state_dump",
 				0600,

drivers/misc/habanalabs/common/habanalabs_drv.c

 // SPDX-License-Identifier: GPL-2.0
 
 /*
- * Copyright 2016-2019 HabanaLabs, Ltd.
+ * Copyright 2016-2021 HabanaLabs, Ltd.
  * All Rights Reserved.
  *
  */
@@ -153,15 +153,7 @@ int hl_device_open(struct inode *inode, struct file *filp)
 		goto out_err;
 	}
 
-	if (hdev->in_debug) {
-		dev_err_ratelimited(hdev->dev,
-			"Can't open %s because it is being debugged by another user\n",
-			dev_name(hdev->dev));
-		rc = -EPERM;
-		goto out_err;
-	}
-
-	if (hdev->compute_ctx) {
+	if (hdev->is_compute_ctx_active) {
 		dev_dbg_ratelimited(hdev->dev,
 			"Can't open %s because another user is working on it\n",
 			dev_name(hdev->dev));
@@ -175,20 +167,17 @@ int hl_device_open(struct inode *inode, struct file *filp)
 		goto out_err;
 	}
 
-	/* Device is IDLE at this point so it is legal to change PLLs.
-	 * There is no need to check anything because if the PLL is
-	 * already HIGH, the set function will return without doing
-	 * anything
-	 */
-	hl_device_set_frequency(hdev, PLL_HIGH);
-
 	list_add(&hpriv->dev_node, &hdev->fpriv_list);
 	mutex_unlock(&hdev->fpriv_list_lock);
 
 	hl_debugfs_add_file(hpriv);
 
+	atomic_set(&hdev->last_error.cs_write_disable, 0);
+	atomic_set(&hdev->last_error.razwi_write_disable, 0);
+
 	hdev->open_counter++;
 	hdev->last_successful_open_jif = jiffies;
+	hdev->last_successful_open_ktime = ktime_get();
 
 	return 0;
 
@@ -231,12 +220,11 @@ int hl_device_open_ctrl(struct inode *inode, struct file *filp)
 	hpriv->hdev = hdev;
 	filp->private_data = hpriv;
 	hpriv->filp = filp;
-	hpriv->is_control = true;
 	nonseekable_open(inode, filp);
 
 	hpriv->taskpid = find_get_pid(current->pid);
 
-	mutex_lock(&hdev->fpriv_list_lock);
+	mutex_lock(&hdev->fpriv_ctrl_list_lock);
 
 	if (!hl_device_operational(hdev, NULL)) {
 		dev_err_ratelimited(hdev->dev_ctrl,
@@ -246,13 +234,13 @@ int hl_device_open_ctrl(struct inode *inode, struct file *filp)
 		goto out_err;
 	}
 
-	list_add(&hpriv->dev_node, &hdev->fpriv_list);
-	mutex_unlock(&hdev->fpriv_list_lock);
+	list_add(&hpriv->dev_node, &hdev->fpriv_ctrl_list);
+	mutex_unlock(&hdev->fpriv_ctrl_list_lock);
 
 	return 0;
 
 out_err:
-	mutex_unlock(&hdev->fpriv_list_lock);
+	mutex_unlock(&hdev->fpriv_ctrl_list_lock);
 	filp->private_data = NULL;
 	put_pid(hpriv->taskpid);
 
@@ -263,6 +251,7 @@ int hl_device_open_ctrl(struct inode *inode, struct file *filp)
 
 static void set_driver_behavior_per_device(struct hl_device *hdev)
 {
+	hdev->pldm = 0;
 	hdev->fw_components = FW_TYPE_ALL_TYPES;
 	hdev->cpu_queues_enable = 1;
 	hdev->heartbeat = 1;
@@ -279,23 +268,53 @@ static void set_driver_behavior_per_device(struct hl_device *hdev)
 	hdev->axi_drain = 0;
 }
 
-/*
+static void copy_kernel_module_params_to_device(struct hl_device *hdev)
+{
+	hdev->major = hl_major;
+	hdev->memory_scrub = memory_scrub;
+	hdev->reset_on_lockup = reset_on_lockup;
+	hdev->boot_error_status_mask = boot_error_status_mask;
+
+	if (timeout_locked)
+		hdev->timeout_jiffies = msecs_to_jiffies(timeout_locked * 1000);
+	else
+		hdev->timeout_jiffies = MAX_SCHEDULE_TIMEOUT;
+
+}
+
+static int fixup_device_params(struct hl_device *hdev)
+{
+	hdev->asic_prop.fw_security_enabled = is_asic_secured(hdev->asic_type);
+
+	hdev->fw_poll_interval_usec = HL_FW_STATUS_POLL_INTERVAL_USEC;
+
+	hdev->stop_on_err = true;
+	hdev->reset_info.curr_reset_cause = HL_RESET_CAUSE_UNKNOWN;
+	hdev->reset_info.prev_reset_trigger = HL_RESET_TRIGGER_DEFAULT;
+
+	/* Enable only after the initialization of the device */
+	hdev->disabled = true;
+
+	/* Set default DMA mask to 32 bits */
+	hdev->dma_mask = 32;
+
+	return 0;
+}
+
+/**
  * create_hdev - create habanalabs device instance
  *
  * @dev: will hold the pointer to the new habanalabs device structure
  * @pdev: pointer to the pci device
- * @asic_type: in case of simulator device, which device is it
- * @minor: in case of simulator device, the minor of the device
  *
  * Allocate memory for habanalabs device and initialize basic fields
  * Identify the ASIC type
  * Allocate ID (minor) for the device (only for real devices)
  */
-int create_hdev(struct hl_device **dev, struct pci_dev *pdev,
-		enum hl_asic_type asic_type, int minor)
+static int create_hdev(struct hl_device **dev, struct pci_dev *pdev)
 {
+	int main_id, ctrl_id = 0, rc = 0;
 	struct hl_device *hdev;
-	int rc, main_id, ctrl_id = 0;
 
 	*dev = NULL;
 
@@ -303,69 +322,39 @@ int create_hdev(struct hl_device **dev, struct pci_dev *pdev,
 	if (!hdev)
 		return -ENOMEM;
 
-	/* First, we must find out which ASIC are we handling. This is needed
-	 * to configure the behavior of the driver (kernel parameters)
-	 */
-	if (pdev) {
-		hdev->asic_type = get_asic_type(pdev->device);
-		if (hdev->asic_type == ASIC_INVALID) {
-			dev_err(&pdev->dev, "Unsupported ASIC\n");
-			rc = -ENODEV;
-			goto free_hdev;
-		}
-	} else {
-		hdev->asic_type = asic_type;
-	}
-
-	if (pdev)
-		hdev->asic_prop.fw_security_enabled =
-					is_asic_secured(hdev->asic_type);
-	else
-		hdev->asic_prop.fw_security_enabled = false;
+	/* can be NULL in case of simulator device */
+	hdev->pdev = pdev;
 
 	/* Assign status description string */
-	strncpy(hdev->status[HL_DEVICE_STATUS_OPERATIONAL],
-					"operational", HL_STR_MAX);
-	strncpy(hdev->status[HL_DEVICE_STATUS_IN_RESET],
-					"in reset", HL_STR_MAX);
-	strncpy(hdev->status[HL_DEVICE_STATUS_MALFUNCTION],
-					"disabled", HL_STR_MAX);
-	strncpy(hdev->status[HL_DEVICE_STATUS_NEEDS_RESET],
-					"needs reset", HL_STR_MAX);
+	strncpy(hdev->status[HL_DEVICE_STATUS_OPERATIONAL], "operational", HL_STR_MAX);
+	strncpy(hdev->status[HL_DEVICE_STATUS_IN_RESET], "in reset", HL_STR_MAX);
+	strncpy(hdev->status[HL_DEVICE_STATUS_MALFUNCTION], "disabled", HL_STR_MAX);
+	strncpy(hdev->status[HL_DEVICE_STATUS_NEEDS_RESET], "needs reset", HL_STR_MAX);
 	strncpy(hdev->status[HL_DEVICE_STATUS_IN_DEVICE_CREATION],
 					"in device creation", HL_STR_MAX);
 
-	hdev->major = hl_major;
-	hdev->reset_on_lockup = reset_on_lockup;
-	hdev->memory_scrub = memory_scrub;
-	hdev->boot_error_status_mask = boot_error_status_mask;
-	hdev->stop_on_err = true;
+	/* First, we must find out which ASIC are we handling. This is needed
+	 * to configure the behavior of the driver (kernel parameters)
+	 */
+	hdev->asic_type = get_asic_type(pdev->device);
+	if (hdev->asic_type == ASIC_INVALID) {
+		dev_err(&pdev->dev, "Unsupported ASIC\n");
+		rc = -ENODEV;
+		goto free_hdev;
+	}
 
-	hdev->pldm = 0;
+	copy_kernel_module_params_to_device(hdev);
 
 	set_driver_behavior_per_device(hdev);
 
-	hdev->curr_reset_cause = HL_RESET_CAUSE_UNKNOWN;
-	hdev->prev_reset_trigger = HL_RESET_TRIGGER_DEFAULT;
-
-	if (timeout_locked)
-		hdev->timeout_jiffies = msecs_to_jiffies(timeout_locked * 1000);
-	else
-		hdev->timeout_jiffies = MAX_SCHEDULE_TIMEOUT;
-
-	hdev->disabled = true;
-	hdev->pdev = pdev; /* can be NULL in case of simulator device */
-
-	/* Set default DMA mask to 32 bits */
-	hdev->dma_mask = 32;
+	fixup_device_params(hdev);
 
 	mutex_lock(&hl_devs_idr_lock);
 
 	/* Always save 2 numbers, 1 for main device and 1 for control.
 	 * They must be consecutive
 	 */
-	main_id = idr_alloc(&hl_devs_idr, hdev, 0, HL_MAX_MINORS,
-				GFP_KERNEL);
+	main_id = idr_alloc(&hl_devs_idr, hdev, 0, HL_MAX_MINORS, GFP_KERNEL);
 
 	if (main_id >= 0)
 		ctrl_id = idr_alloc(&hl_devs_idr, hdev, main_id + 1,
@@ -405,7 +394,7 @@ int create_hdev(struct hl_device **dev, struct pci_dev *pdev,
  * @dev: pointer to the habanalabs device structure
  *
  */
-void destroy_hdev(struct hl_device *hdev)
+static void destroy_hdev(struct hl_device *hdev)
 {
 	/* Remove device from the device list */
 	mutex_lock(&hl_devs_idr_lock);
@@ -444,7 +433,7 @@ static int hl_pmops_resume(struct device *dev)
 	return hl_device_resume(hdev);
 }
 
-/*
+/**
  * hl_pci_probe - probe PCI habanalabs devices
  *
  * @pdev: pointer to pci device
@@ -454,8 +443,7 @@ static int hl_pmops_resume(struct device *dev)
  * Create a new habanalabs device and initialize it according to the
  * device's type
  */
-static int hl_pci_probe(struct pci_dev *pdev,
-				const struct pci_device_id *id)
+static int hl_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 {
 	struct hl_device *hdev;
 	int rc;
@@ -464,7 +452,7 @@ static int hl_pci_probe(struct pci_dev *pdev,
 		 " device found [%04x:%04x] (rev %x)\n",
 		 (int)pdev->vendor, (int)pdev->device, (int)pdev->revision);
 
-	rc = create_hdev(&hdev, pdev, ASIC_INVALID, -1);
+	rc = create_hdev(&hdev, pdev);
 	if (rc)
 		return rc;
 

drivers/misc/habanalabs/common/habanalabs_ioctl.c

@@ -158,7 +158,7 @@ static int hw_idle(struct hl_device *hdev, struct hl_info_args *args)
 		min((size_t) max_size, sizeof(hw_idle))) ? -EFAULT : 0;
 }
 
-static int debug_coresight(struct hl_device *hdev, struct hl_debug_args *args)
+static int debug_coresight(struct hl_device *hdev, struct hl_ctx *ctx, struct hl_debug_args *args)
 {
 	struct hl_debug_params *params;
 	void *input = NULL, *output = NULL;
@@ -200,7 +200,7 @@ static int debug_coresight(struct hl_device *hdev, struct hl_debug_args *args)
 		params->output_size = args->output_size;
 	}
 
-	rc = hdev->asic_funcs->debug_coresight(hdev, params);
+	rc = hdev->asic_funcs->debug_coresight(hdev, ctx, params);
 	if (rc) {
 		dev_err(hdev->dev,
 			"debug coresight operation failed %d\n", rc);
@@ -269,8 +269,8 @@ static int get_reset_count(struct hl_device *hdev, struct hl_info_args *args)
 	if ((!max_size) || (!out))
 		return -EINVAL;
 
-	reset_count.hard_reset_cnt = hdev->hard_reset_cnt;
-	reset_count.soft_reset_cnt = hdev->soft_reset_cnt;
+	reset_count.hard_reset_cnt = hdev->reset_info.hard_reset_cnt;
+	reset_count.soft_reset_cnt = hdev->reset_info.soft_reset_cnt;
 
 	return copy_to_user(out, &reset_count,
 		min((size_t) max_size, sizeof(reset_count))) ? -EFAULT : 0;
@@ -313,15 +313,38 @@ static int pci_counters_info(struct hl_fpriv *hpriv, struct hl_info_args *args)
 
 static int clk_throttle_info(struct hl_fpriv *hpriv, struct hl_info_args *args)
 {
+	void __user *out = (void __user *) (uintptr_t) args->return_pointer;
 	struct hl_device *hdev = hpriv->hdev;
 	struct hl_info_clk_throttle clk_throttle = {0};
+	ktime_t end_time, zero_time = ktime_set(0, 0);
 	u32 max_size = args->return_size;
-	void __user *out = (void __user *) (uintptr_t) args->return_pointer;
+	int i;
 
 	if ((!max_size) || (!out))
 		return -EINVAL;
 
-	clk_throttle.clk_throttling_reason = hdev->clk_throttling_reason;
+	mutex_lock(&hdev->clk_throttling.lock);
+
+	clk_throttle.clk_throttling_reason = hdev->clk_throttling.current_reason;
+
+	for (i = 0 ; i < HL_CLK_THROTTLE_TYPE_MAX ; i++) {
+		if (!(hdev->clk_throttling.aggregated_reason & BIT(i)))
+			continue;
+
+		clk_throttle.clk_throttling_timestamp_us[i] =
+			ktime_to_us(hdev->clk_throttling.timestamp[i].start);
+
+		if (ktime_compare(hdev->clk_throttling.timestamp[i].end, zero_time))
+			end_time = hdev->clk_throttling.timestamp[i].end;
+		else
+			end_time = ktime_get();
+
+		clk_throttle.clk_throttling_duration_ns[i] =
+			ktime_to_ns(ktime_sub(end_time,
+				hdev->clk_throttling.timestamp[i].start));
+
+	}
+	mutex_unlock(&hdev->clk_throttling.lock);
 
 	return copy_to_user(out, &clk_throttle,
 		min((size_t) max_size, sizeof(clk_throttle))) ? -EFAULT : 0;
@@ -480,6 +503,94 @@ static int open_stats_info(struct hl_fpriv *hpriv, struct hl_info_args *args)
 		min((size_t) max_size, sizeof(open_stats_info))) ? -EFAULT : 0;
 }
 
+static int dram_pending_rows_info(struct hl_fpriv *hpriv, struct hl_info_args *args)
+{
+	struct hl_device *hdev = hpriv->hdev;
+	u32 max_size = args->return_size;
+	u32 pend_rows_num = 0;
+	void __user *out = (void __user *) (uintptr_t) args->return_pointer;
+	int rc;
+
+	if ((!max_size) || (!out))
+		return -EINVAL;
+
+	rc = hl_fw_dram_pending_row_get(hdev, &pend_rows_num);
+	if (rc)
+		return rc;
+
+	return copy_to_user(out, &pend_rows_num,
+			min_t(size_t, max_size, sizeof(pend_rows_num))) ? -EFAULT : 0;
+}
+
+static int dram_replaced_rows_info(struct hl_fpriv *hpriv, struct hl_info_args *args)
+{
+	struct hl_device *hdev = hpriv->hdev;
+	u32 max_size = args->return_size;
+	struct cpucp_hbm_row_info info = {0};
+	void __user *out = (void __user *) (uintptr_t) args->return_pointer;
+	int rc;
+
+	if ((!max_size) || (!out))
+		return -EINVAL;
+
+	rc = hl_fw_dram_replaced_row_get(hdev, &info);
+	if (rc)
+		return rc;
+
+	return copy_to_user(out, &info, min_t(size_t, max_size, sizeof(info))) ? -EFAULT : 0;
+}
+
+static int last_err_open_dev_info(struct hl_fpriv *hpriv, struct hl_info_args *args)
+{
+	struct hl_info_last_err_open_dev_time info = {0};
+	struct hl_device *hdev = hpriv->hdev;
+	u32 max_size = args->return_size;
+	void __user *out = (void __user *) (uintptr_t) args->return_pointer;
+
+	if ((!max_size) || (!out))
+		return -EINVAL;
+
+	info.timestamp = ktime_to_ns(hdev->last_error.open_dev_timestamp);
+
+	return copy_to_user(out, &info, min_t(size_t, max_size, sizeof(info))) ? -EFAULT : 0;
+}
+
+static int cs_timeout_info(struct hl_fpriv *hpriv, struct hl_info_args *args)
+{
+	struct hl_info_cs_timeout_event info = {0};
+	struct hl_device *hdev = hpriv->hdev;
+	u32 max_size = args->return_size;
+	void __user *out = (void __user *) (uintptr_t) args->return_pointer;
+
+	if ((!max_size) || (!out))
+		return -EINVAL;
+
+	info.seq = hdev->last_error.cs_timeout_seq;
+	info.timestamp = ktime_to_ns(hdev->last_error.cs_timeout_timestamp);
+
+	return copy_to_user(out, &info, min_t(size_t, max_size, sizeof(info))) ? -EFAULT : 0;
+}
+
+static int razwi_info(struct hl_fpriv *hpriv, struct hl_info_args *args)
+{
+	struct hl_device *hdev = hpriv->hdev;
+	u32 max_size = args->return_size;
+	struct hl_info_razwi_event info = {0};
+	void __user *out = (void __user *) (uintptr_t) args->return_pointer;
+
+	if ((!max_size) || (!out))
+		return -EINVAL;
+
+	info.timestamp = ktime_to_ns(hdev->last_error.razwi_timestamp);
+	info.addr = hdev->last_error.razwi_addr;
+	info.engine_id_1 = hdev->last_error.razwi_engine_id_1;
+	info.engine_id_2 = hdev->last_error.razwi_engine_id_2;
+	info.no_engine_id = hdev->last_error.razwi_non_engine_initiator;
+	info.error_type = hdev->last_error.razwi_type;
+
+	return copy_to_user(out, &info, min_t(size_t, max_size, sizeof(info))) ? -EFAULT : 0;
+}
+
 static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data,
 				struct device *dev)
 {
@@ -503,6 +614,33 @@ static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data,
 	case HL_INFO_RESET_COUNT:
 		return get_reset_count(hdev, args);
 
+	case HL_INFO_HW_EVENTS:
+		return hw_events_info(hdev, false, args);
+
+	case HL_INFO_HW_EVENTS_AGGREGATE:
+		return hw_events_info(hdev, true, args);
+
+	case HL_INFO_CS_COUNTERS:
+		return cs_counters_info(hpriv, args);
+
+	case HL_INFO_CLK_THROTTLE_REASON:
+		return clk_throttle_info(hpriv, args);
+
+	case HL_INFO_SYNC_MANAGER:
+		return sync_manager_info(hpriv, args);
+
+	case HL_INFO_OPEN_STATS:
+		return open_stats_info(hpriv, args);
+
+	case HL_INFO_LAST_ERR_OPEN_DEV_TIME:
+		return last_err_open_dev_info(hpriv, args);
+
+	case HL_INFO_CS_TIMEOUT_EVENT:
+		return cs_timeout_info(hpriv, args);
+
+	case HL_INFO_RAZWI_EVENT:
+		return razwi_info(hpriv, args);
+
 	default:
 		break;
 	}
@@ -515,10 +653,6 @@ static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data,
 	}
 
 	switch (args->op) {
-	case HL_INFO_HW_EVENTS:
-		rc = hw_events_info(hdev, false, args);
-		break;
-
 	case HL_INFO_DRAM_USAGE:
 		rc = dram_usage_info(hpriv, args);
 		break;
@@ -531,10 +665,6 @@ static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data,
 		rc = device_utilization(hdev, args);
 		break;
 
-	case HL_INFO_HW_EVENTS_AGGREGATE:
-		rc = hw_events_info(hdev, true, args);
-		break;
-
 	case HL_INFO_CLK_RATE:
 		rc = get_clk_rate(hdev, args);
 		break;
@@ -542,18 +672,9 @@ static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data,
 	case HL_INFO_TIME_SYNC:
 		return time_sync_info(hdev, args);
 
-	case HL_INFO_CS_COUNTERS:
-		return cs_counters_info(hpriv, args);
-
 	case HL_INFO_PCI_COUNTERS:
 		return pci_counters_info(hpriv, args);
 
-	case HL_INFO_CLK_THROTTLE_REASON:
-		return clk_throttle_info(hpriv, args);
-
-	case HL_INFO_SYNC_MANAGER:
-		return sync_manager_info(hpriv, args);
-
 	case HL_INFO_TOTAL_ENERGY:
 		return total_energy_consumption_info(hpriv, args);
 
@@ -563,12 +684,16 @@ static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data,
 	case HL_INFO_POWER:
 		return power_info(hpriv, args);
 
-	case HL_INFO_OPEN_STATS:
-		return open_stats_info(hpriv, args);
+
+	case HL_INFO_DRAM_REPLACED_ROWS:
+		return dram_replaced_rows_info(hpriv, args);
+
+	case HL_INFO_DRAM_PENDING_ROWS:
+		return dram_pending_rows_info(hpriv, args);
 
 	default:
 		dev_err(dev, "Invalid request %d\n", args->op);
-		rc = -ENOTTY;
+		rc = -EINVAL;
 		break;
 	}
 
@@ -613,16 +738,17 @@ static int hl_debug_ioctl(struct hl_fpriv *hpriv, void *data)
 				"Rejecting debug configuration request because device not in debug mode\n");
 			return -EFAULT;
 		}
-		args->input_size =
-			min(args->input_size, hl_debug_struct_size[args->op]);
-		rc = debug_coresight(hdev, args);
+		args->input_size = min(args->input_size, hl_debug_struct_size[args->op]);
+		rc = debug_coresight(hdev, hpriv->ctx, args);
 		break;
+
 	case HL_DEBUG_OP_SET_MODE:
-		rc = hl_device_set_debug_mode(hdev, (bool) args->enable);
+		rc = hl_device_set_debug_mode(hdev, hpriv->ctx, (bool) args->enable);
 		break;
+
 	default:
 		dev_err(hdev->dev, "Invalid request %d\n", args->op);
-		rc = -ENOTTY;
+		rc = -EINVAL;
 		break;
 	}
 
@@ -649,7 +775,6 @@ static long _hl_ioctl(struct file *filep, unsigned int cmd, unsigned long arg,
 		const struct hl_ioctl_desc *ioctl, struct device *dev)
 {
 	struct hl_fpriv *hpriv = filep->private_data;
-	struct hl_device *hdev = hpriv->hdev;
 	unsigned int nr = _IOC_NR(cmd);
 	char stack_kdata[128] = {0};
 	char *kdata = NULL;
@@ -658,12 +783,6 @@ static long _hl_ioctl(struct file *filep, unsigned int cmd, unsigned long arg,
 	u32 hl_size;
 	int retcode;
 
-	if (hdev->hard_reset_pending) {
-		dev_crit_ratelimited(dev,
-			"Device HARD reset pending! Please close FD\n");
-		return -ENODEV;
-	}
-
 	/* Do not trust userspace, use our own definition */
 	func = ioctl->func;
 

drivers/misc/habanalabs/common/hw_queue.c

@@ -429,6 +429,9 @@ static int init_signal_cs(struct hl_device *hdev,
 	rc = hl_cs_signal_sob_wraparound_handler(hdev, q_idx, &hw_sob, 1,
 								false);
 
+	job->cs->sob_addr_offset = hw_sob->sob_addr;
+	job->cs->initial_sob_count = prop->next_sob_val - 1;
+
 	return rc;
 }
 
@@ -571,7 +574,7 @@ static int encaps_sig_first_staged_cs_handler
 	struct hl_encaps_signals_mgr *mgr;
 	int rc = 0;
 
-	mgr = &hdev->compute_ctx->sig_mgr;
+	mgr = &cs->ctx->sig_mgr;
 
 	spin_lock(&mgr->lock);
 	encaps_sig_hdl = idr_find(&mgr->handles, cs->encaps_sig_hdl_id);

drivers/misc/habanalabs/common/irq.c

@@ -145,8 +145,12 @@ static void handle_user_cq(struct hl_device *hdev,
 
 	spin_lock(&user_cq->wait_list_lock);
 	list_for_each_entry(pend, &user_cq->wait_list_head, wait_list_node) {
-		pend->fence.timestamp = now;
-		complete_all(&pend->fence.completion);
+		if ((pend->cq_kernel_addr &&
+				*(pend->cq_kernel_addr) >= pend->cq_target_value) ||
+				!pend->cq_kernel_addr) {
+			pend->fence.timestamp = now;
+			complete_all(&pend->fence.completion);
+		}
 	}
 	spin_unlock(&user_cq->wait_list_lock);
 }
@@ -245,10 +249,8 @@ irqreturn_t hl_irq_handler_eq(int irq, void *arg)
 		 */
 		dma_rmb();
 
-		if (hdev->disabled) {
-			dev_warn(hdev->dev,
-				"Device disabled but received IRQ %d for EQ\n",
-					irq);
+		if (hdev->disabled && !hdev->reset_info.is_in_soft_reset) {
+			dev_warn(hdev->dev, "Device disabled but received an EQ event\n");
 			goto skip_irq;
 		}
 

drivers/misc/habanalabs/common/memory.c

@@ -316,7 +316,7 @@ static int free_phys_pg_pack(struct hl_device *hdev,
 	}
 
 	if (rc && !hdev->disabled)
-		hl_device_reset(hdev, HL_RESET_HARD);
+		hl_device_reset(hdev, HL_DRV_RESET_HARD);
 
 end:
 	kvfree(phys_pg_pack->pages);
@@ -477,7 +477,7 @@ static int add_va_block_locked(struct hl_device *hdev,
 		struct list_head *va_list, u64 start, u64 end)
 {
 	struct hl_vm_va_block *va_block, *res = NULL;
-	u64 size = end - start;
+	u64 size = end - start + 1;
 
 	print_va_list_locked(hdev, va_list);
 
@@ -518,7 +518,7 @@ static int add_va_block_locked(struct hl_device *hdev,
 /**
  * add_va_block() - wrapper for add_va_block_locked.
  * @hdev: pointer to the habanalabs device structure.
- * @va_list: pointer to the virtual addresses block list.
+ * @va_range: pointer to the virtual addresses range object.
  * @start: start virtual address.
  * @end: end virtual address.
  *
@@ -538,8 +538,11 @@ static inline int add_va_block(struct hl_device *hdev,
 }
 
 /**
- * is_hint_crossing_range() - check if hint address crossing specified reserved
- * range.
+ * is_hint_crossing_range() - check if hint address crossing specified reserved.
+ * @range_type: virtual space range type.
+ * @start_addr: start virtual address.
+ * @size: block size.
+ * @prop: asic properties structure to retrieve reserved ranges from.
  */
 static inline bool is_hint_crossing_range(enum hl_va_range_type range_type,
 		u64 start_addr, u32 size, struct asic_fixed_properties *prop) {
@@ -644,7 +647,7 @@ static u64 get_va_block(struct hl_device *hdev,
 				continue;
 		}
 
-		valid_size = va_block->end - valid_start;
+		valid_size = va_block->end - valid_start + 1;
 		if (valid_size < size)
 			continue;
 
@@ -707,7 +710,7 @@ static u64 get_va_block(struct hl_device *hdev,
 
 	if (new_va_block->size > size) {
 		new_va_block->start += size;
-		new_va_block->size = new_va_block->end - new_va_block->start;
+		new_va_block->size = new_va_block->end - new_va_block->start + 1;
 	} else {
 		list_del(&new_va_block->node);
 		kfree(new_va_block);
@@ -749,6 +752,7 @@ u64 hl_reserve_va_block(struct hl_device *hdev, struct hl_ctx *ctx,
 
 /**
  * hl_get_va_range_type() - get va_range type for the given address and size.
+ * @ctx: context to fetch va_range from.
  * @address: the start address of the area we want to validate.
  * @size: the size in bytes of the area we want to validate.
  * @type: returned va_range type.
@@ -776,8 +780,8 @@ static int hl_get_va_range_type(struct hl_ctx *ctx, u64 address, u64 size,
  * hl_unreserve_va_block() - wrapper for add_va_block to unreserve a va block.
  * @hdev: pointer to the habanalabs device structure
  * @ctx: pointer to the context structure.
- * @start: start virtual address.
- * @end: end virtual address.
+ * @start_addr: start virtual address.
+ * @size: number of bytes to unreserve.
  *
  * This function does the following:
  * - Takes the list lock and calls add_va_block_locked.
@@ -1201,17 +1205,13 @@ static int map_device_va(struct hl_ctx *ctx, struct hl_mem_in *args,
 		goto map_err;
 	}
 
-	rc = hdev->asic_funcs->mmu_invalidate_cache_range(hdev, false,
-		*vm_type, ctx->asid, ret_vaddr, phys_pg_pack->total_size);
+	rc = hl_mmu_invalidate_cache_range(hdev, false, *vm_type | MMU_OP_SKIP_LOW_CACHE_INV,
+				ctx->asid, ret_vaddr, phys_pg_pack->total_size);
 
 	mutex_unlock(&ctx->mmu_lock);
 
-	if (rc) {
-		dev_err(hdev->dev,
-			"mapping handle %u failed due to MMU cache invalidation\n",
-			handle);
+	if (rc)
 		goto map_err;
-	}
 
 	ret_vaddr += phys_pg_pack->offset;
 
@@ -1349,9 +1349,8 @@ static int unmap_device_va(struct hl_ctx *ctx, struct hl_mem_in *args,
 	 * at the loop end rather than for each iteration
 	 */
 	if (!ctx_free)
-		rc = hdev->asic_funcs->mmu_invalidate_cache_range(hdev, true,
-				*vm_type, ctx->asid, vaddr,
-				phys_pg_pack->total_size);
+		rc = hl_mmu_invalidate_cache_range(hdev, true, *vm_type, ctx->asid, vaddr,
+							phys_pg_pack->total_size);
 
 	mutex_unlock(&ctx->mmu_lock);
 
@@ -1364,11 +1363,6 @@ static int unmap_device_va(struct hl_ctx *ctx, struct hl_mem_in *args,
 	if (!ctx_free) {
 		int tmp_rc;
 
-		if (rc)
-			dev_err(hdev->dev,
-				"unmapping vaddr 0x%llx failed due to MMU cache invalidation\n",
-				vaddr);
-
 		tmp_rc = add_va_block(hdev, va_range, vaddr,
 					vaddr + phys_pg_pack->total_size - 1);
 		if (tmp_rc) {
@@ -2037,7 +2031,7 @@ static int mem_ioctl_no_mmu(struct hl_fpriv *hpriv, union hl_mem_args *args)
 
 	default:
 		dev_err(hdev->dev, "Unknown opcode for memory IOCTL\n");
-		rc = -ENOTTY;
+		rc = -EINVAL;
 		break;
 	}
 
@@ -2162,7 +2156,7 @@ int hl_mem_ioctl(struct hl_fpriv *hpriv, void *data)
 
 	default:
 		dev_err(hdev->dev, "Unknown opcode for memory IOCTL\n");
-		rc = -ENOTTY;
+		rc = -EINVAL;
 		break;
 	}
 
@@ -2339,6 +2333,8 @@ void hl_userptr_delete_list(struct hl_device *hdev,
 /**
  * hl_userptr_is_pinned() - returns whether the given userptr is pinned.
  * @hdev: pointer to the habanalabs device structure.
+ * @addr: user address to check.
+ * @size: user block size to check.
  * @userptr_list: pointer to the list to clear.
  * @userptr: pointer to userptr to check.
  *
@@ -2361,9 +2357,10 @@ bool hl_userptr_is_pinned(struct hl_device *hdev, u64 addr,
 /**
  * va_range_init() - initialize virtual addresses range.
  * @hdev: pointer to the habanalabs device structure.
- * @va_range: pointer to the range to initialize.
+ * @va_ranges: pointer to va_ranges array.
  * @start: range start address.
  * @end: range end address.
+ * @page_size: page size for this va_range.
  *
  * This function does the following:
  * - Initializes the virtual addresses list of the given range with the given
@@ -2388,8 +2385,14 @@ static int va_range_init(struct hl_device *hdev, struct hl_va_range *va_range,
 			start += PAGE_SIZE;
 		}
 
-		if (end & (PAGE_SIZE - 1))
-			end &= PAGE_MASK;
+		/*
+		 * The end of the range is inclusive, hence we need to align it
+		 * to the end of the last full page in the range. For example if
+		 * end = 0x3ff5 with page size 0x1000, we need to align it to
+		 * 0x2fff. The remainig 0xff5 bytes do not form a full page.
+		 */
+		if ((end + 1) & (PAGE_SIZE - 1))
+			end = ((end + 1) & PAGE_MASK) - 1;
 	}
 
 	if (start >= end) {
@@ -2414,7 +2417,7 @@ static int va_range_init(struct hl_device *hdev, struct hl_va_range *va_range,
 /**
  * va_range_fini() - clear a virtual addresses range.
  * @hdev: pointer to the habanalabs structure.
- * va_range: pointer to virtual addresses rang.e
+ * @va_range: pointer to virtual addresses range.
  *
  * This function does the following:
  * - Frees the virtual addresses block list and its lock.
@@ -2434,12 +2437,15 @@ static void va_range_fini(struct hl_device *hdev, struct hl_va_range *va_range)
  * @ctx: pointer to the habanalabs context structure.
  * @host_range_start: host virtual addresses range start.
  * @host_range_end: host virtual addresses range end.
+ * @host_page_size: host page size.
  * @host_huge_range_start: host virtual addresses range start for memory
  *                         allocated with huge pages.
  * @host_huge_range_end: host virtual addresses range end for memory allocated
  *                        with huge pages.
+ * @host_huge_page_size: host huge page size.
  * @dram_range_start: dram virtual addresses range start.
  * @dram_range_end: dram virtual addresses range end.
+ * @dram_page_size: dram page size.
  *
  * This function initializes the following:
  * - MMU for context.
@@ -2564,14 +2570,14 @@ int hl_vm_ctx_init(struct hl_ctx *ctx)
 		return 0;
 
 	dram_range_start = prop->dmmu.start_addr;
-	dram_range_end = prop->dmmu.end_addr;
+	dram_range_end = prop->dmmu.end_addr - 1;
 	dram_page_size = prop->dram_page_size ?
 				prop->dram_page_size : prop->dmmu.page_size;
 	host_range_start = prop->pmmu.start_addr;
-	host_range_end = prop->pmmu.end_addr;
+	host_range_end = prop->pmmu.end_addr - 1;
 	host_page_size = prop->pmmu.page_size;
 	host_huge_range_start = prop->pmmu_huge.start_addr;
-	host_huge_range_end = prop->pmmu_huge.end_addr;
+	host_huge_range_end = prop->pmmu_huge.end_addr - 1;
 	host_huge_page_size = prop->pmmu_huge.page_size;
 
 	return vm_ctx_init_with_ranges(ctx, host_range_start, host_range_end,
@@ -2618,7 +2624,7 @@ void hl_vm_ctx_fini(struct hl_ctx *ctx)
 	 * 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))
+	if (!hdev->reset_info.hard_reset_pending && !hash_empty(ctx->mem_hash))
 		dev_dbg(hdev->dev,
 			"user released device without removing its memory mappings\n");
 
@@ -2633,8 +2639,8 @@ void hl_vm_ctx_fini(struct hl_ctx *ctx)
 	mutex_lock(&ctx->mmu_lock);
 
 	/* 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);
+	hl_mmu_invalidate_cache(hdev, true, MMU_OP_USERPTR);
+	hl_mmu_invalidate_cache(hdev, true, MMU_OP_PHYS_PACK);
 
 	mutex_unlock(&ctx->mmu_lock);
 

drivers/misc/habanalabs/common/mmu/mmu.c

@@ -637,3 +637,28 @@ u64 hl_mmu_descramble_addr(struct hl_device *hdev, u64 addr)
 {
 	return addr;
 }
+
+int hl_mmu_invalidate_cache(struct hl_device *hdev, bool is_hard, u32 flags)
+{
+	int rc;
+
+	rc = hdev->asic_funcs->mmu_invalidate_cache(hdev, is_hard, flags);
+	if (rc)
+		dev_err_ratelimited(hdev->dev, "MMU cache invalidation failed\n");
+
+	return rc;
+}
+
+int hl_mmu_invalidate_cache_range(struct hl_device *hdev, bool is_hard,
+					u32 flags, u32 asid, u64 va, u64 size)
+{
+	int rc;
+
+	rc = hdev->asic_funcs->mmu_invalidate_cache_range(hdev, is_hard, flags,
+								asid, va, size);
+	if (rc)
+		dev_err_ratelimited(hdev->dev, "MMU cache range invalidation failed\n");
+
+	return rc;
+}
+

drivers/misc/habanalabs/common/mmu/mmu_v1.c

@@ -269,7 +269,7 @@ static int dram_default_mapping_init(struct hl_ctx *ctx)
 
 	num_of_hop3 = prop->dram_size_for_default_page_mapping;
 	do_div(num_of_hop3, prop->dram_page_size);
-	do_div(num_of_hop3, PTE_ENTRIES_IN_HOP);
+	do_div(num_of_hop3, HOP_PTE_ENTRIES_512);
 
 	/* add hop1 and hop2 */
 	total_hops = num_of_hop3 + 2;
@@ -330,7 +330,7 @@ static int dram_default_mapping_init(struct hl_ctx *ctx)
 
 	for (i = 0 ; i < num_of_hop3 ; i++) {
 		hop3_pte_addr = ctx->dram_default_hops[i];
-		for (j = 0 ; j < PTE_ENTRIES_IN_HOP ; j++) {
+		for (j = 0 ; j < HOP_PTE_ENTRIES_512 ; j++) {
 			write_final_pte(ctx, hop3_pte_addr, pte_val);
 			get_pte(ctx, ctx->dram_default_hops[i]);
 			hop3_pte_addr += HL_PTE_SIZE;
@@ -369,7 +369,7 @@ static void dram_default_mapping_fini(struct hl_ctx *ctx)
 
 	num_of_hop3 = prop->dram_size_for_default_page_mapping;
 	do_div(num_of_hop3, prop->dram_page_size);
-	do_div(num_of_hop3, PTE_ENTRIES_IN_HOP);
+	do_div(num_of_hop3, HOP_PTE_ENTRIES_512);
 
 	hop0_addr = get_hop0_addr(ctx);
 	/* add hop1 and hop2 */
@@ -379,7 +379,7 @@ static void dram_default_mapping_fini(struct hl_ctx *ctx)
 
 	for (i = 0 ; i < num_of_hop3 ; i++) {
 		hop3_pte_addr = ctx->dram_default_hops[i];
-		for (j = 0 ; j < PTE_ENTRIES_IN_HOP ; j++) {
+		for (j = 0 ; j < HOP_PTE_ENTRIES_512 ; j++) {
 			clear_pte(ctx, hop3_pte_addr);
 			put_pte(ctx, ctx->dram_default_hops[i]);
 			hop3_pte_addr += HL_PTE_SIZE;
@@ -573,7 +573,7 @@ static int _hl_mmu_v1_unmap(struct hl_ctx *ctx,
 
 	curr_pte = *(u64 *) (uintptr_t) hop3_pte_addr;
 
-	is_huge = curr_pte & LAST_MASK;
+	is_huge = curr_pte & mmu_prop->last_mask;
 
 	if (is_dram_addr && !is_huge) {
 		dev_err(hdev->dev,
@@ -597,7 +597,7 @@ static int _hl_mmu_v1_unmap(struct hl_ctx *ctx,
 
 	if (hdev->dram_default_page_mapping && is_dram_addr) {
 		u64 default_pte = (prop->mmu_dram_default_page_addr &
-				HOP_PHYS_ADDR_MASK) | LAST_MASK |
+				HOP_PHYS_ADDR_MASK) | mmu_prop->last_mask |
 					PAGE_PRESENT_MASK;
 		if (curr_pte == default_pte) {
 			dev_err(hdev->dev,
@@ -729,7 +729,7 @@ static int _hl_mmu_v1_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr,
 
 	if (hdev->dram_default_page_mapping && is_dram_addr) {
 		u64 default_pte = (prop->mmu_dram_default_page_addr &
-					HOP_PHYS_ADDR_MASK) | LAST_MASK |
+					HOP_PHYS_ADDR_MASK) | mmu_prop->last_mask |
 						PAGE_PRESENT_MASK;
 
 		if (curr_pte != default_pte) {
@@ -769,7 +769,7 @@ static int _hl_mmu_v1_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr,
 		goto err;
 	}
 
-	curr_pte = (phys_addr & HOP_PHYS_ADDR_MASK) | LAST_MASK
+	curr_pte = (phys_addr & HOP_PHYS_ADDR_MASK) | mmu_prop->last_mask
 			| PAGE_PRESENT_MASK;
 
 	if (is_huge)
@@ -930,7 +930,7 @@ static int hl_mmu_v1_get_tlb_info(struct hl_ctx *ctx, u64 virt_addr,
 		if (!(hops->hop_info[i].hop_pte_val & PAGE_PRESENT_MASK))
 			return -EFAULT;
 
-		if (hops->hop_info[i].hop_pte_val & LAST_MASK)
+		if (hops->hop_info[i].hop_pte_val & mmu_prop->last_mask)
 			break;
 	}
 

drivers/misc/habanalabs/common/sysfs.c

@@ -139,7 +139,7 @@ static ssize_t cpld_ver_show(struct device *dev, struct device_attribute *attr,
 	struct hl_device *hdev = dev_get_drvdata(dev);
 
 	return sprintf(buf, "0x%08x\n",
-			hdev->asic_prop.cpucp_info.cpld_version);
+			le32_to_cpu(hdev->asic_prop.cpucp_info.cpld_version));
 }
 
 static ssize_t cpucp_kernel_ver_show(struct device *dev,
@@ -163,8 +163,13 @@ static ssize_t infineon_ver_show(struct device *dev,
 {
 	struct hl_device *hdev = dev_get_drvdata(dev);
 
-	return sprintf(buf, "0x%04x\n",
-			hdev->asic_prop.cpucp_info.infineon_version);
+	if (hdev->asic_prop.cpucp_info.infineon_second_stage_version)
+		return sprintf(buf, "%#04x %#04x\n",
+			le32_to_cpu(hdev->asic_prop.cpucp_info.infineon_version),
+			le32_to_cpu(hdev->asic_prop.cpucp_info.infineon_second_stage_version));
+	else
+		return sprintf(buf, "%#04x\n",
+			le32_to_cpu(hdev->asic_prop.cpucp_info.infineon_version));
 }
 
 static ssize_t fuse_ver_show(struct device *dev, struct device_attribute *attr,
@@ -206,7 +211,7 @@ static ssize_t soft_reset_store(struct device *dev,
 		goto out;
 	}
 
-	if (!hdev->allow_inference_soft_reset) {
+	if (!hdev->asic_prop.allow_inference_soft_reset) {
 		dev_err(hdev->dev, "Device does not support inference soft-reset\n");
 		goto out;
 	}
@@ -236,7 +241,7 @@ static ssize_t hard_reset_store(struct device *dev,
 
 	dev_warn(hdev->dev, "Hard-Reset requested through sysfs\n");
 
-	hl_device_reset(hdev, HL_RESET_HARD);
+	hl_device_reset(hdev, HL_DRV_RESET_HARD);
 
 out:
 	return count;
@@ -298,7 +303,7 @@ static ssize_t soft_reset_cnt_show(struct device *dev,
 {
 	struct hl_device *hdev = dev_get_drvdata(dev);
 
-	return sprintf(buf, "%d\n", hdev->soft_reset_cnt);
+	return sprintf(buf, "%d\n", hdev->reset_info.soft_reset_cnt);
 }
 
 static ssize_t hard_reset_cnt_show(struct device *dev,
@@ -306,7 +311,7 @@ static ssize_t hard_reset_cnt_show(struct device *dev,
 {
 	struct hl_device *hdev = dev_get_drvdata(dev);
 
-	return sprintf(buf, "%d\n", hdev->hard_reset_cnt);
+	return sprintf(buf, "%d\n", hdev->reset_info.hard_reset_cnt);
 }
 
 static ssize_t max_power_show(struct device *dev, struct device_attribute *attr,
@@ -419,8 +424,6 @@ static struct attribute *hl_dev_attrs[] = {
 	&dev_attr_max_power.attr,
 	&dev_attr_pci_addr.attr,
 	&dev_attr_preboot_btl_ver.attr,
-	&dev_attr_soft_reset.attr,
-	&dev_attr_soft_reset_cnt.attr,
 	&dev_attr_status.attr,
 	&dev_attr_thermal_ver.attr,
 	&dev_attr_uboot_ver.attr,
@@ -445,15 +448,25 @@ static const struct attribute_group *hl_dev_attr_groups[] = {
 	NULL,
 };
 
+static struct attribute *hl_dev_inference_attrs[] = {
+	&dev_attr_soft_reset.attr,
+	&dev_attr_soft_reset_cnt.attr,
+	NULL,
+};
+
+static struct attribute_group hl_dev_inference_attr_group = {
+	.attrs = hl_dev_inference_attrs,
+};
+
+static const struct attribute_group *hl_dev_inference_attr_groups[] = {
+	&hl_dev_inference_attr_group,
+	NULL,
+};
+
 int hl_sysfs_init(struct hl_device *hdev)
 {
 	int rc;
 
-	if (hdev->asic_type == ASIC_GOYA)
-		hdev->pm_mng_profile = PM_AUTO;
-	else
-		hdev->pm_mng_profile = PM_MANUAL;
-
 	hdev->max_power = hdev->asic_prop.max_power_default;
 
 	hdev->asic_funcs->add_device_attr(hdev, &hl_dev_clks_attr_group);
@@ -465,10 +478,25 @@ int hl_sysfs_init(struct hl_device *hdev)
 		return rc;
 	}
 
+	if (!hdev->asic_prop.allow_inference_soft_reset)
+		return 0;
+
+	rc = device_add_groups(hdev->dev, hl_dev_inference_attr_groups);
+	if (rc) {
+		dev_err(hdev->dev,
+			"Failed to add groups to device, error %d\n", rc);
+		return rc;
+	}
+
 	return 0;
 }
 
 void hl_sysfs_fini(struct hl_device *hdev)
 {
 	device_remove_groups(hdev->dev, hl_dev_attr_groups);
+
+	if (!hdev->asic_prop.allow_inference_soft_reset)
+		return;
+
+	device_remove_groups(hdev->dev, hl_dev_inference_attr_groups);
 }

drivers/misc/habanalabs/gaudi/gaudiP.h

@@ -357,8 +357,8 @@ void gaudi_init_security(struct hl_device *hdev);
 void gaudi_ack_protection_bits_errors(struct hl_device *hdev);
 void gaudi_add_device_attr(struct hl_device *hdev,
 			struct attribute_group *dev_attr_grp);
-int gaudi_debug_coresight(struct hl_device *hdev, void *data);
-void gaudi_halt_coresight(struct hl_device *hdev);
+int gaudi_debug_coresight(struct hl_device *hdev, struct hl_ctx *ctx, void *data);
+void gaudi_halt_coresight(struct hl_device *hdev, struct hl_ctx *ctx);
 void gaudi_mmu_prepare_reg(struct hl_device *hdev, u64 reg, u32 asid);
 
 #endif /* GAUDIP_H_ */

drivers/misc/habanalabs/gaudi/gaudi_coresight.c

@@ -848,7 +848,7 @@ static int gaudi_config_spmu(struct hl_device *hdev,
 	return 0;
 }
 
-int gaudi_debug_coresight(struct hl_device *hdev, void *data)
+int gaudi_debug_coresight(struct hl_device *hdev, struct hl_ctx *ctx, void *data)
 {
 	struct hl_debug_params *params = data;
 	int rc = 0;
@@ -887,7 +887,7 @@ int gaudi_debug_coresight(struct hl_device *hdev, void *data)
 	return rc;
 }
 
-void gaudi_halt_coresight(struct hl_device *hdev)
+void gaudi_halt_coresight(struct hl_device *hdev, struct hl_ctx *ctx)
 {
 	struct hl_debug_params params = {};
 	int i, rc;

drivers/misc/habanalabs/goya/goyaP.h

@@ -153,9 +153,15 @@
 #define HW_CAP_GOLDEN		0x00000400
 #define HW_CAP_TPC		0x00000800
 
+struct goya_work_freq {
+	struct hl_device *hdev;
+	struct delayed_work work_freq;
+};
+
 struct goya_device {
 	/* TODO: remove hw_queues_lock after moving to scheduler code */
 	spinlock_t	hw_queues_lock;
+	struct goya_work_freq	*goya_work;
 
 	u64		mme_clk;
 	u64		tpc_clk;
@@ -166,6 +172,9 @@ struct goya_device {
 	u32		events_stat_aggregate[GOYA_ASYNC_EVENT_ID_SIZE];
 	u32		hw_cap_initialized;
 	u8		device_cpu_mmu_mappings_done;
+
+	enum hl_pll_frequency		curr_pll_profile;
+	enum hl_pm_mng_profile		pm_mng_profile;
 };
 
 int goya_set_fixed_properties(struct hl_device *hdev);
@@ -211,8 +220,8 @@ void goya_set_pll_profile(struct hl_device *hdev, enum hl_pll_frequency freq);
 void goya_add_device_attr(struct hl_device *hdev,
 			struct attribute_group *dev_attr_grp);
 int goya_cpucp_info_get(struct hl_device *hdev);
-int goya_debug_coresight(struct hl_device *hdev, void *data);
-void goya_halt_coresight(struct hl_device *hdev);
+int goya_debug_coresight(struct hl_device *hdev, struct hl_ctx *ctx, void *data);
+void goya_halt_coresight(struct hl_device *hdev, struct hl_ctx *ctx);
 
 int goya_suspend(struct hl_device *hdev);
 int goya_resume(struct hl_device *hdev);
@@ -237,5 +246,6 @@ void goya_mmu_remove_device_cpu_mappings(struct hl_device *hdev);
 
 u32 goya_get_queue_id_for_cq(struct hl_device *hdev, u32 cq_idx);
 u64 goya_get_device_time(struct hl_device *hdev);
+int goya_set_frequency(struct hl_device *hdev, enum hl_pll_frequency freq);
 
 #endif /* GOYAP_H_ */

drivers/misc/habanalabs/goya/goya_coresight.c

@@ -652,7 +652,7 @@ static int goya_config_spmu(struct hl_device *hdev,
 	return 0;
 }
 
-int goya_debug_coresight(struct hl_device *hdev, void *data)
+int goya_debug_coresight(struct hl_device *hdev, struct hl_ctx *ctx, void *data)
 {
 	struct hl_debug_params *params = data;
 	int rc = 0;
@@ -691,7 +691,7 @@ int goya_debug_coresight(struct hl_device *hdev, void *data)
 	return rc;
 }
 
-void goya_halt_coresight(struct hl_device *hdev)
+void goya_halt_coresight(struct hl_device *hdev, struct hl_ctx *ctx)
 {
 	struct hl_debug_params params = {};
 	int i, rc;

drivers/misc/habanalabs/include/common/hl_boot_if.h

@@ -32,6 +32,7 @@ enum cpu_boot_err {
 	CPU_BOOT_ERR_DEVICE_UNUSABLE_FAIL = 13,
 	CPU_BOOT_ERR_BOOT_FW_CRIT_ERR = 18,
 	CPU_BOOT_ERR_BINNING_FAIL = 19,
+	CPU_BOOT_ERR_TPM_FAIL = 20,
 	CPU_BOOT_ERR_ENABLED = 31,
 	CPU_BOOT_ERR_SCND_EN = 63,
 	CPU_BOOT_ERR_LAST = 64 /* we have 2 registers of 32 bits */
@@ -108,6 +109,8 @@ enum cpu_boot_err {
  *					malfunctioning components might still be
  *					in use.
  *
+ * CPU_BOOT_ERR0_TPM_FAIL		TPM verification flow failed.
+ *
  * CPU_BOOT_ERR0_ENABLED		Error registers enabled.
  *					This is a main indication that the
  *					running FW populates the error
@@ -130,6 +133,7 @@ enum cpu_boot_err {
 #define CPU_BOOT_ERR0_DEVICE_UNUSABLE_FAIL	(1 << CPU_BOOT_ERR_DEVICE_UNUSABLE_FAIL)
 #define CPU_BOOT_ERR0_BOOT_FW_CRIT_ERR		(1 << CPU_BOOT_ERR_BOOT_FW_CRIT_ERR)
 #define CPU_BOOT_ERR0_BINNING_FAIL		(1 << CPU_BOOT_ERR_BINNING_FAIL)
+#define CPU_BOOT_ERR0_TPM_FAIL			(1 << CPU_BOOT_ERR_TPM_FAIL)
 #define CPU_BOOT_ERR0_ENABLED			(1 << CPU_BOOT_ERR_ENABLED)
 #define CPU_BOOT_ERR1_ENABLED			(1 << CPU_BOOT_ERR_ENABLED)
 

drivers/misc/habanalabs/include/hw_ip/mmu/mmu_general.h

@@ -16,27 +16,18 @@
 #define PAGE_PRESENT_MASK		0x0000000000001ull
 #define SWAP_OUT_MASK			0x0000000000004ull
 #define LAST_MASK			0x0000000000800ull
-#define HOP0_MASK			0x3000000000000ull
-#define HOP1_MASK			0x0FF8000000000ull
-#define HOP2_MASK			0x0007FC0000000ull
-#define HOP3_MASK			0x000003FE00000ull
-#define HOP4_MASK			0x00000001FF000ull
 #define FLAGS_MASK			0x0000000000FFFull
 
-#define HOP0_SHIFT			48
-#define HOP1_SHIFT			39
-#define HOP2_SHIFT			30
-#define HOP3_SHIFT			21
-#define HOP4_SHIFT			12
-
 #define MMU_ARCH_5_HOPS			5
 
 #define HOP_PHYS_ADDR_MASK		(~FLAGS_MASK)
 
 #define HL_PTE_SIZE			sizeof(u64)
-#define HOP_TABLE_SIZE			PAGE_SIZE_4KB
-#define PTE_ENTRIES_IN_HOP		(HOP_TABLE_SIZE / HL_PTE_SIZE)
-#define HOP0_TABLES_TOTAL_SIZE		(HOP_TABLE_SIZE * MAX_ASID)
+
+/* definitions for HOP with 512 PTE entries */
+#define HOP_PTE_ENTRIES_512		512
+#define HOP_TABLE_SIZE_512_PTE		(HOP_PTE_ENTRIES_512 * HL_PTE_SIZE)
+#define HOP0_512_PTE_TABLES_TOTAL_SIZE	(HOP_TABLE_SIZE_512_PTE * MAX_ASID)
 
 #define MMU_HOP0_PA43_12_SHIFT		12
 #define MMU_HOP0_PA49_44_SHIFT		(12 + 32)