Skip to contentSkip to conversation

iio: imu: bmi160: Fix the error handling at bmi160_chip_init()

PATCH In review Zheyu Ma 2 participants 3 messages May 13, 2022 – May 17, 2022 IIO

lore
Share & tools
lore

Zheyu Ma

author
#0 lore raw

PATCH 1/1 iio: imu: bmi160: Fix the error handling at bmi160_chip_init()

When the driver fails in bmi160_chip_init(), it should disable the regulator.

The following log shows it:

[   18.709662] WARNING: CPU: 0 PID: 294 at drivers/regulator/core.c:2257 _regulator_put+0x3ec/0x4e0
[   18.712069] RIP: 0010:_regulator_put+0x3ec/0x4e0
[   18.716395] Call Trace:
[   18.716522]  <TASK>
[   18.716636]  regulator_bulk_free+0x82/0xe0
[   18.717077]  i2c_device_probe+0x6f0/0x800
Fixes 5dea3fb066f0 ("iio: imu: bmi160: added regulator support")
---
 drivers/iio/imu/bmi160/bmi160_core.c | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)
drivers/iio/imu/bmi160/bmi160_core.c +11 -6
index 824b5124a5f5..33abfc2bbd7c 100644
@@ -730,7 +730,7 @@ static int bmi160_chip_init(struct bmi160_data *data, bool use_spi)
 
 	ret = regmap_write(data->regmap, BMI160_REG_CMD, BMI160_CMD_SOFTRESET);
 	if (ret)
-		return ret;
+		goto err_regulator_disable;
 
 	usleep_range(BMI160_SOFTRESET_USLEEP, BMI160_SOFTRESET_USLEEP + 1);
 
@@ -741,29 +741,34 @@ static int bmi160_chip_init(struct bmi160_data *data, bool use_spi)
 	if (use_spi) {
 		ret = regmap_read(data->regmap, BMI160_REG_DUMMY, &val);
 		if (ret)
-			return ret;
+			goto err_regulator_disable;
 	}
 
 	ret = regmap_read(data->regmap, BMI160_REG_CHIP_ID, &val);
 	if (ret) {
 		dev_err(dev, "Error reading chip id\n");
-		return ret;
+		goto err_regulator_disable;
 	}
 	if (val != BMI160_CHIP_ID_VAL) {
 		dev_err(dev, "Wrong chip id, got %x expected %x\n",
 			val, BMI160_CHIP_ID_VAL);
-		return -ENODEV;
+		ret = -ENODEV;
+		goto err_regulator_disable;
 	}
 
 	ret = bmi160_set_mode(data, BMI160_ACCEL, true);
 	if (ret)
-		return ret;
+		goto err_regulator_disable;
 
 	ret = bmi160_set_mode(data, BMI160_GYRO, true);
 	if (ret)
-		return ret;
+		goto err_regulator_disable;
 
 	return 0;
+
+err_regulator_disable:
+	regulator_bulk_disable(ARRAY_SIZE(data->supplies), data->supplies);
+	return ret;
 }
 
 static int bmi160_data_rdy_trigger_set_state(struct iio_trigger *trig,
Signature · 1 line
2.25.1
maintainer
#1 lore raw
Quoted 12 lines · On Fri, 13 May 2022 22:35:13 +0800 Zheyu Ma <[email protected]> wrote:

When the driver fails in bmi160_chip_init(), it should disable the regulator.

The following log shows it:

[   18.709662] WARNING: CPU: 0 PID: 294 at drivers/regulator/core.c:2257 _regulator_put+0x3ec/0x4e0
[   18.712069] RIP: 0010:_regulator_put+0x3ec/0x4e0
[   18.716395] Call Trace:
[   18.716522]  <TASK>
[   18.716636]  regulator_bulk_free+0x82/0xe0
[   18.717077]  i2c_device_probe+0x6f0/0x800
Quoted 2 lines
Fixes 5dea3fb066f0 ("iio: imu: bmi160: added regulator support")

Hi,

Good find, but Tong Zhang got their first and the resulting patch fixes an additional issue missed in this fix. See below.

Jonathan

Quoted 36 lines
---
 drivers/iio/imu/bmi160/bmi160_core.c | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)
drivers/iio/imu/bmi160/bmi160_core.c +4 -4
index 824b5124a5f5..33abfc2bbd7c 100644
@@ -730,7 +730,7 @@ static int bmi160_chip_init(struct bmi160_data *data, bool use_spi)
 
 	ret = regmap_write(data->regmap, BMI160_REG_CMD, BMI160_CMD_SOFTRESET);
 	if (ret)
-		return ret;
+		goto err_regulator_disable;
 
 	usleep_range(BMI160_SOFTRESET_USLEEP, BMI160_SOFTRESET_USLEEP + 1);
 
@@ -741,29 +741,34 @@ static int bmi160_chip_init(struct bmi160_data *data, bool use_spi)
 	if (use_spi) {
 		ret = regmap_read(data->regmap, BMI160_REG_DUMMY, &val);
 		if (ret)
-			return ret;
+			goto err_regulator_disable;
 	}
 
 	ret = regmap_read(data->regmap, BMI160_REG_CHIP_ID, &val);
 	if (ret) {
 		dev_err(dev, "Error reading chip id\n");
-		return ret;
+		goto err_regulator_disable;
 	}
 	if (val != BMI160_CHIP_ID_VAL) {
 		dev_err(dev, "Wrong chip id, got %x expected %x\n",
 			val, BMI160_CHIP_ID_VAL);
-		return -ENODEV;
+		ret = -ENODEV;

drivers/iio/imu/bmi160/bmi160_core.c static int bmi160_chip_init(struct bmi160_data *data, bool use_spi) · line 764

Quoted 10 lines
drivers/iio/imu/bmi160/bmi160_core.c +2 -1
+		goto err_regulator_disable;
 	}
 
 	ret = bmi160_set_mode(data, BMI160_ACCEL, true);
 	if (ret)
-		return ret;
+		goto err_regulator_disable;
 
 	ret = bmi160_set_mode(data, BMI160_GYRO, true);
 	if (ret)

A failure here does not disable the accelerometer (which is done in the uninit() function which we we never end
up running.

Quoted 11 lines
drivers/iio/imu/bmi160/bmi160_core.c +5 -1
-		return ret;
+		goto err_regulator_disable;
 
 	return 0;
+
+err_regulator_disable:
+	regulator_bulk_disable(ARRAY_SIZE(data->supplies), data->supplies);
+	return ret;
 }
 
 static int bmi160_data_rdy_trigger_set_state(struct iio_trigger *trig,
author
#2 lore raw
Quoted 16 lines · Jonathan Cameron
Quoted 12 lines · On Fri, 13 May 2022 22:35:13 +0800 Zheyu Ma <[email protected]> wrote:

When the driver fails in bmi160_chip_init(), it should disable the regulator.

The following log shows it:

[   18.709662] WARNING: CPU: 0 PID: 294 at drivers/regulator/core.c:2257 _regulator_put+0x3ec/0x4e0
[   18.712069] RIP: 0010:_regulator_put+0x3ec/0x4e0
[   18.716395] Call Trace:
[   18.716522]  <TASK>
[   18.716636]  regulator_bulk_free+0x82/0xe0
[   18.717077]  i2c_device_probe+0x6f0/0x800
Quoted 7 lines
Quoted 2 lines
Fixes 5dea3fb066f0 ("iio: imu: bmi160: added regulator support")

Hi,

Good find, but Tong Zhang got their first and the resulting patch fixes an additional issue missed in this fix. See below.

Thanks for your reply! I'll check the linux-next branch next time.

Zheyu Ma