Skip to contentSkip to conversation

drivers/thermal/tegra/soctherm: Use devm_add_action_or_reset() for clock disable

PATCH In review Daniel Lezcano 2 participants 6 messages Apr 24 – May 6

Subsystems (2)
Tegra architecture support Thermal
lore
Share & tools
lore

Daniel Lezcano

author
#0 lore raw

PATCH 1/2 drivers/thermal/tegra/soctherm: Use devm_add_action_or_reset() for clock disable

Replace the manual error handling paths disabling the clocks with devm_add_action_or_reset(). This ensures the clocks are properly disabled on probe failure and driver removal, while simplifying the code by removing the explicit error paths.

Signed-off-by Daniel Lezcano <[email protected]>
---
 drivers/thermal/tegra/soctherm.c | 28 +++++++++++++++-------------
 1 file changed, 15 insertions(+), 13 deletions(-)
drivers/thermal/tegra/soctherm.c +15 -13
index 5d26b52beaba..40c3715e84c5 100644
@@ -1499,6 +1499,13 @@ static int soctherm_clk_enable(struct platform_device *pdev, bool enable)
 	return 0;
 }
 
+static void soctherm_clk_disable(void *data)
+{
+	struct platform_device *pdev = data;
+
+	soctherm_clk_enable(pdev, false);
+}
+
 static int throt_get_cdev_max_state(struct thermal_cooling_device *cdev,
 				    unsigned long *max_state)
 {

@@ -2175,6 +2182,10 @@ static int tegra_soctherm_probe(struct platform_device *pdev)
 	if (err)
 		return err;
 
+	err = devm_add_action_or_reset(&pdev->dev, soctherm_clk_disable, pdev);
+	if (err)
+		return err;
+
 	soctherm_thermtrips_parse(pdev);
 
 	soctherm_init_hw_throt_cdev(pdev);

@@ -2184,10 +2195,8 @@ static int tegra_soctherm_probe(struct platform_device *pdev)
 	for (i = 0; i < soc->num_ttgs; ++i) {
 		struct tegra_thermctl_zone *zone =
 			devm_kzalloc(&pdev->dev, sizeof(*zone), GFP_KERNEL);
-		if (!zone) {
-			err = -ENOMEM;
-			goto disable_clocks;
-		}
+		if (!zone)
+			return -ENOMEM;
 
 		zone->reg = tegra->regs + soc->ttgs[i]->sensor_temp_offset;
 		zone->dev = &pdev->dev;

@@ -2201,7 +2210,7 @@ static int tegra_soctherm_probe(struct platform_device *pdev)
 			err = PTR_ERR(z);
 			dev_err(&pdev->dev, "failed to register sensor: %d\n",
 				err);
-			goto disable_clocks;
+			return err;
 		}
 
 		zone->tz = z;

@@ -2210,7 +2219,7 @@ static int tegra_soctherm_probe(struct platform_device *pdev)
 		/* Configure hw trip points */
 		err = tegra_soctherm_set_hwtrips(&pdev->dev, soc->ttgs[i], z);
 		if (err)
-			goto disable_clocks;
+			return err;
 	}
 
 	err = soctherm_interrupts_init(pdev, tegra);

@@ -2218,11 +2227,6 @@ static int tegra_soctherm_probe(struct platform_device *pdev)
 	soctherm_debug_init(pdev);
 
 	return 0;
-
-disable_clocks:
-	soctherm_clk_enable(pdev, false);
-
-	return err;
 }
 
 static void tegra_soctherm_remove(struct platform_device *pdev)

@@ -2230,8 +2234,6 @@ static void tegra_soctherm_remove(struct platform_device *pdev)
 	struct tegra_soctherm *tegra = platform_get_drvdata(pdev);
 
 	debugfs_remove_recursive(tegra->debugfs_dir);
-
-	soctherm_clk_enable(pdev, false);
 }
 
 static int __maybe_unused soctherm_suspend(struct device *dev)
Signature · 1 line
2.43.0
author
#1 lore raw

PATCH 2/2 thermal/drivers/tegra/soctherma: Switch to devm cooling device registration

Use devm_thermal_of_cooling_device_register() to simplify resource management and avoid manual cleanup in error paths.

As a side effect this change has the benefit of solving an existing issue. Before, the function tegra_soctherm_remove() only called debugfs_remove_recursive() and never called thermal_cooling_device_unregister() for any of the cooling devices registered here.

After the driver removal, the thermal framework's cdev list would still hold references to thermal_cooling_device objects whose devdata pointer (ts) pointed to memory already freed by the platform device's devm cleanup.

With this change, the cooling device is unregistered when the driver is removed, thus fixing the issue above.

Signed-off-by Daniel Lezcano <[email protected]>
---
 drivers/thermal/tegra/soctherm.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/thermal/tegra/soctherm.c +3 -3
index 40c3715e84c5..6a56638c98f1 100644
@@ -1707,9 +1707,9 @@ static void soctherm_init_hw_throt_cdev(struct platform_device *pdev)
 			stc->init = true;
 		} else {
 
-			tcd = thermal_of_cooling_device_register(np_stcc,
-							 (char *)name, ts,
-							 &throt_cooling_ops);
+			tcd = devm_thermal_of_cooling_device_register(dev, np_stcc,
+								      (char *)name, ts,
+								      &throt_cooling_ops);
 			if (IS_ERR_OR_NULL(tcd)) {
 				dev_err(dev,
 					"throttle-cfg: %s: failed to register cooling device\n",
Signature · 1 line
2.43.0
author
#2 lore raw
In reply to Daniel Lezcano
Quoted 7 lines · Daniel Lezcano

Replace the manual error handling paths disabling the clocks with devm_add_action_or_reset(). This ensures the clocks are properly disabled on probe failure and driver removal, while simplifying the code by removing the explicit error paths.

Signed-off-by Daniel Lezcano <[email protected]>

---

If nobody is against these two changes, I'll apply them

Thanks

-- Daniel

author Applied
#3 lore raw
Quoted 6 lines · Daniel Lezcano
Quoted 5 lines · Daniel Lezcano

Replace the manual error handling paths disabling the clocks with devm_add_action_or_reset(). This ensures the clocks are properly disabled on probe failure and driver removal, while simplifying the code by removing the explicit error paths.

Quoted 9 lines
Quoted 2 lines
Signed-off-by Daniel Lezcano <[email protected]>

---

If nobody is against these two changes, I'll apply them

Thanks

  -- Daniel

Applied, thanks

reviewer Reviewed-by
#4 lore raw

Hi Daniel,

Quoted 10 lines · Daniel Lezcano
Quoted 9 lines · Daniel Lezcano
Quoted 7 lines · Daniel Lezcano

Replace the manual error handling paths disabling the clocks with devm_add_action_or_reset(). This ensures the clocks are properly disabled on probe failure and driver removal, while simplifying the code by removing the explicit error paths.

Signed-off-by Daniel Lezcano <[email protected]>

---

Quoted 8 lines
Quoted 6 lines

If nobody is against these two changes, I'll apply them

Thanks

   -- Daniel

Applied, thanks

My apologies for being late. The changes make perfect sense. Feel free to add to both patches:

Reviewed-by Lukasz Luba <[email protected]>
author
#5 lore raw
Quoted 20 lines · Lukasz Luba

Hi Daniel,

Quoted 17 lines · Daniel Lezcano
Quoted 15 lines · Daniel Lezcano
Quoted 7 lines · Daniel Lezcano

Replace the manual error handling paths disabling the clocks with devm_add_action_or_reset(). This ensures the clocks are properly disabled on probe failure and driver removal, while simplifying the code by removing the explicit error paths.

Signed-off-by Daniel Lezcano <[email protected]>

---

If nobody is against these two changes, I'll apply them

Thanks

   -- Daniel

Quoted 6 lines
Quoted 1 line

Applied, thanks

My apologies for being late. The changes make perfect sense. Feel free to add to both patches:

Reviewed-by Lukasz Luba <[email protected]>

Thanks for the review