Skip to contentSkip to conversation

hwmon: (sht4x) Add missing locks

PATCH New Guenter Roeck 2 participants 4 messages Aug 21 Hardware monitoring

lore
Share & tools
lore

Guenter Roeck

author maintainer
#0 lore raw

PATCH 1/2 hwmon: (sht4x) Add missing locks

Sashiko reports:

Heater sysfs callbacks (heater_enable_store, heater_power_store, and heater_time_store) are exposed to data races without the hwmon lock.

If a user-space process reads hwmon data while another process enables the heater, heater_enable_store() executes without holding hwmon_lock(dev). This can interleave I2C commands and mutate shared state (data->heating_complete and data->data_pending) concurrently with sht4x_read_values(), leading to corrupted I2C sequences.

Fixes 53dfa12299c1 ("hwmon: (sht4x) Rely on subsystem locking")
Cc Alessandro Zini <[email protected]>
---
 drivers/hwmon/sht4x.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/hwmon/sht4x.c +6 -0
index 9cace0e8acda..7a0dc2ed723d 100644
@@ -277,6 +277,8 @@ static ssize_t heater_enable_store(struct device *dev,
 		heating_time_bound = 1100;
 	}
 
+	guard(hwmon_lock)(dev);
+
 	if (time_before(jiffies, data->heating_complete))
 		return -EBUSY;
 
@@ -314,6 +316,8 @@ static ssize_t heater_power_store(struct device *dev,
 	if (power != 20 && power != 110 && power != 200)
 		return -EINVAL;
 
+	guard(hwmon_lock)(dev);
+
 	data->heater_power = power;
 
 	return count;
@@ -344,6 +348,8 @@ static ssize_t heater_time_store(struct device *dev,
 	if (time != 100 && time != 1000)
 		return -EINVAL;
 
+	guard(hwmon_lock)(dev);
+
 	data->heater_time = time;
 
 	return count;
Signature · 1 line
2.45.2
author maintainer
#1 lore raw

PATCH 2/2 hwmon: (sht4x) Fix return value from heater_enable_store()

Sashiko reports:

The return value in heater_enable_store() causes an unexpected write failure in user-space.

When the heater is successfully enabled, the function returns 0 instead of count:

drivers/hwmon/sht4x.c:heater_enable_store() {
    ...
    data->heating_complete = jiffies + msecs_to_jiffies(heating_time_bound);
    data->data_pending = true;
    return 0;
}

Returning 0 signals to VFS that no bytes were processed. Standard user-space tools will retry the write with the remaining bytes. On the retry, time_before(jiffies, data->heating_complete) evaluates to true, and the function immediately fails with -EBUSY.

Return count as expected to fix the problem.

Fixes 0eed6fc3d2b9e ("hwmon: (sht4x): add heater support")
Cc Antoni Pokusinski <[email protected]>
Cc Alessandro Zini <[email protected]>
---
 drivers/hwmon/sht4x.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/hwmon/sht4x.c +1 -1
index 7a0dc2ed723d..a97dda9e92dc 100644
@@ -288,7 +288,7 @@ static ssize_t heater_enable_store(struct device *dev,
 
 	data->heating_complete = jiffies + msecs_to_jiffies(heating_time_bound);
 	data->data_pending = true;
-	return 0;
+	return count;
 }
 
 static ssize_t heater_power_show(struct device *dev,
Signature · 1 line
2.45.2
#2 lore raw

Re: [PATCH 2/2] hwmon: (sht4x) Fix return value from heater_enable_store()

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] Unit mismatch between jiffies and milliseconds in sht4x_read_values timeout calculation --

--- Patch [2]: [PATCH 2/2] hwmon: (sht4x) Fix return value from heater_enable_store() --- commit ffda8f58b43906bc1ea2253c2bf9b7981f601ded
Author: Guenter Roeck <[email protected]>

hwmon: (sht4x) Fix return value from heater_enable_store()

The return value in heater_enable_store() causes an unexpected write failure in user-space because it returns 0 instead of count. This signals to VFS that no bytes were processed, leading to a busy retry loop. Returning count fixes the problem.

drivers/hwmon/sht4x.c static ssize_t heater_enable_store(struct device *dev, · line 289

Quoted 7 lines
drivers/hwmon/sht4x.c +0 -0
index 7a0dc2ed723d8..a97dda9e92dc5 100644
@@ -288,7 +288,7 @@ static ssize_t heater_enable_store(struct device *dev,
 
 	data->heating_complete = jiffies + msecs_to_jiffies(heating_time_bound);

[Severity: High]
This isn't a bug introduced by this patch, but while reviewing how data->heating_complete is used, I noticed a unit mismatch in sht4x_read_values().

In sht4x_read_values(), the code calculates the timeout like this:

	if (data->data_pending &&
	    time_before(jiffies, data->heating_complete + data->update_interval)) {
		data->data_pending = false;

Is data->update_interval expected to be in milliseconds here?

Since data->update_interval is added directly to data->heating_complete (which is in jiffies) without using msecs_to_jiffies(), does this create a distorted timeout window?

Could this cause the driver to return stale, artificially high temperature data from the heater phase instead of taking a new ambient measurement?

Quoted 4 lines
drivers/hwmon/sht4x.c +1 -1
 	data->data_pending = true;
-	return 0;
+	return count;
 }
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=2
#3 lore raw
In reply to Guenter Roeck
Sashiko has reviewed this patch and found no issues. It looks great!
Quoted 11 lines

Sashiko reports:

Heater sysfs callbacks (heater_enable_store, heater_power_store, and heater_time_store) are exposed to data races without the hwmon lock.

If a user-space process reads hwmon data while another process enables the heater, heater_enable_store() executes without holding hwmon_lock(dev). This can interleave I2C commands and mutate shared state (data->heating_complete and data->data_pending) concurrently with sht4x_read_values(), leading to corrupted I2C sequences.

Quoted 3 lines
Fixes 53dfa12299c1 ("hwmon: (sht4x) Rely on subsystem locking")
Cc Alessandro Zini <[email protected]>

Sashiko has reviewed this patch and found no issues. It looks great!

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1