habanalabs: support clock gating enable/disable

In Gaudi there is a feature of clock gating certain engines.
Therefore, add this property to the device structure.

In addition, due to a limitation of this feature, the driver needs to
dynamically enable or disable this feature during run-time. Therefore, add
ASIC interface functions to enable/disable this function from the common
code.

Moreover, this feature must be turned off when the user wishes to debug the
ASIC by reading/writing registers and/or memory through the driver's
debugfs. Therefore, add an option to enable/disable clock gating via the
debugfs interface.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

Documentation/ABI/testing/debugfs-driver-habanalabs

@@ -8,6 +8,16 @@ Description:    Sets the device address to be used for read or write through
                 only when the IOMMU is disabled.
                 The acceptable value is a string that starts with "0x"
 
+What:           /sys/kernel/debug/habanalabs/hl<n>/clk_gate
+Date:           May 2020
+KernelVersion:  5.8
+Contact:        oded.gabbay@gmail.com
+Description:    Allow the root user to disable/enable in runtime the clock
+                gating mechanism in Gaudi. Due to how Gaudi is built, the
+                clock gating needs to be disabled in order to access the
+                registers of the TPC and MME engines. This is sometimes needed
+                during debug by the user and hence the user needs this option
+
 What:           /sys/kernel/debug/habanalabs/hl<n>/command_buffers
 Date:           Jan 2019
 KernelVersion:  5.1

drivers/misc/habanalabs/debugfs.c

@@ -970,6 +970,55 @@ static ssize_t hl_device_write(struct file *f, const char __user *buf,
 	return count;
 }
 
+static ssize_t hl_clk_gate_read(struct file *f, char __user *buf,
+					size_t count, loff_t *ppos)
+{
+	struct hl_dbg_device_entry *entry = file_inode(f)->i_private;
+	struct hl_device *hdev = entry->hdev;
+	char tmp_buf[200];
+	ssize_t rc;
+
+	if (*ppos)
+		return 0;
+
+	sprintf(tmp_buf, "%d\n", hdev->clock_gating);
+	rc = simple_read_from_buffer(buf, strlen(tmp_buf) + 1, ppos, tmp_buf,
+			strlen(tmp_buf) + 1);
+
+	return rc;
+}
+
+static ssize_t hl_clk_gate_write(struct file *f, const char __user *buf,
+				     size_t count, loff_t *ppos)
+{
+	struct hl_dbg_device_entry *entry = file_inode(f)->i_private;
+	struct hl_device *hdev = entry->hdev;
+	u32 value;
+	ssize_t rc;
+
+	if (atomic_read(&hdev->in_reset)) {
+		dev_warn_ratelimited(hdev->dev,
+				"Can't change clock gating during reset\n");
+		return 0;
+	}
+
+	rc = kstrtouint_from_user(buf, count, 10, &value);
+	if (rc)
+		return rc;
+
+	if (value) {
+		hdev->clock_gating = 1;
+		if (hdev->asic_funcs->enable_clock_gating)
+			hdev->asic_funcs->enable_clock_gating(hdev);
+	} else {
+		if (hdev->asic_funcs->disable_clock_gating)
+			hdev->asic_funcs->disable_clock_gating(hdev);
+		hdev->clock_gating = 0;
+	}
+
+	return count;
+}
+
 static ssize_t hl_stop_on_err_read(struct file *f, char __user *buf,
 					size_t count, loff_t *ppos)
 {
@@ -1058,6 +1107,12 @@ static const struct file_operations hl_device_fops = {
 	.write = hl_device_write
 };
 
+static const struct file_operations hl_clk_gate_fops = {
+	.owner = THIS_MODULE,
+	.read = hl_clk_gate_read,
+	.write = hl_clk_gate_write
+};
+
 static const struct file_operations hl_stop_on_err_fops = {
 	.owner = THIS_MODULE,
 	.read = hl_stop_on_err_read,
@@ -1201,6 +1256,12 @@ void hl_debugfs_add_device(struct hl_device *hdev)
 				dev_entry,
 				&hl_device_fops);
 
+	debugfs_create_file("clk_gate",
+				0200,
+				dev_entry->root,
+				dev_entry,
+				&hl_clk_gate_fops);
+
 	debugfs_create_file("stop_on_err",
 				0644,
 				dev_entry->root,

drivers/misc/habanalabs/device.c

@@ -603,6 +603,9 @@ int hl_device_set_debug_mode(struct hl_device *hdev, bool enable)
 
 		hdev->in_debug = 0;
 
+		if (!hdev->hard_reset_pending)
+			hdev->asic_funcs->enable_clock_gating(hdev);
+
 		goto out;
 	}
 
@@ -613,6 +616,7 @@ int hl_device_set_debug_mode(struct hl_device *hdev, bool enable)
 		goto out;
 	}
 
+	hdev->asic_funcs->disable_clock_gating(hdev);
 	hdev->in_debug = 1;
 
 out:

drivers/misc/habanalabs/goya/goya.c

@@ -5018,6 +5018,16 @@ int goya_armcp_info_get(struct hl_device *hdev)
 	return 0;
 }
 
+static void goya_enable_clock_gating(struct hl_device *hdev)
+{
+
+}
+
+static void goya_disable_clock_gating(struct hl_device *hdev)
+{
+
+}
+
 static bool goya_is_device_idle(struct hl_device *hdev, u32 *mask,
 				struct seq_file *s)
 {
@@ -5239,6 +5249,8 @@ static const struct hl_asic_funcs goya_funcs = {
 	.mmu_invalidate_cache = goya_mmu_invalidate_cache,
 	.mmu_invalidate_cache_range = goya_mmu_invalidate_cache_range,
 	.send_heartbeat = goya_send_heartbeat,
+	.enable_clock_gating = goya_enable_clock_gating,
+	.disable_clock_gating = goya_disable_clock_gating,
 	.debug_coresight = goya_debug_coresight,
 	.is_device_idle = goya_is_device_idle,
 	.soft_reset_late_init = goya_soft_reset_late_init,

drivers/misc/habanalabs/habanalabs.h

@@ -578,6 +578,8 @@ enum hl_pll_frequency {
  * @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.
+ * @enable_clock_gating: enable clock gating for reducing power consumption.
+ * @disable_clock_gating: disable clock for accessing registers on HBW.
  * @debug_coresight: perform certain actions on Coresight for debugging.
  * @is_device_idle: return true if device is idle, false otherwise.
  * @soft_reset_late_init: perform certain actions needed after soft reset.
@@ -678,6 +680,8 @@ struct hl_asic_funcs {
 	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);
+	void (*enable_clock_gating)(struct hl_device *hdev);
+	void (*disable_clock_gating)(struct hl_device *hdev);
 	int (*debug_coresight)(struct hl_device *hdev, void *data);
 	bool (*is_device_idle)(struct hl_device *hdev, u32 *mask,
 				struct seq_file *s);
@@ -1408,6 +1412,7 @@ struct hl_device_idle_busy_ts {
  *                   huge pages.
  * @init_done: is the initialization of the device done.
  * @mmu_enable: is MMU enabled.
+ * @clock_gating: is clock gating enabled.
  * @device_cpu_disabled: is the device CPU disabled (due to timeouts)
  * @dma_mask: the dma mask that was set for this device
  * @in_debug: is device under debug. This, together with fpriv_list, enforces
@@ -1494,6 +1499,7 @@ struct hl_device {
 	u8				dram_default_page_mapping;
 	u8				pmmu_huge_range;
 	u8				init_done;
+	u8				clock_gating;
 	u8				device_cpu_disabled;
 	u8				dma_mask;
 	u8				in_debug;

drivers/misc/habanalabs/habanalabs_drv.c

@@ -231,6 +231,7 @@ static void set_driver_behavior_per_device(struct hl_device *hdev)
 	hdev->fw_loading = 1;
 	hdev->cpu_queues_enable = 1;
 	hdev->heartbeat = 1;
+	hdev->clock_gating = 1;
 
 	hdev->reset_pcilink = 0;
 }