staging: iio: adc: ad7192: use private driver lock to protect hardware changes

The IIO subsystem is redefining iio_dev->mlock to be used by the IIO core only for protecting device operating mode changes. ie. Changes between INDIO_DIRECT_MODE, INDIO_BUFFER_* modes.

These patches
- replace mlock with private driver lock
- add lock where there are hardware state changes

Changes in v2:
- more descriptive commit messages

Aastha Gupta (2):
staging: iio: adc: ad7192: Replace mlock with driver private lock
staging: iio: adc: ad7192: add lock to protect hardware state changes

 drivers/staging/iio/adc/ad7192.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

2.7.4

The IIO subsystem is redefining iio_dev->mlock to be used by the IIO core only for protecting device operating mode changes. ie. Changes between INDIO_DIRECT_MODE, INDIO_BUFFER_* modes.

In this driver, mlock was being used to protect hardware state changes. Replace it with a lock in the devices global data.

---
 drivers/staging/iio/adc/ad7192.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

index d11c6de..096d29d 100644
@@ -162,6 +162,7 @@ struct ad7192_state {
 	u32				scale_avail[8][2];
 	u8				gpocon;
 	u8				devid;
+	struct mutex			lock;
 
 	struct ad_sigma_delta		sd;
 };

@@ -463,10 +464,10 @@ static int ad7192_read_raw(struct iio_dev *indio_dev,
 	case IIO_CHAN_INFO_SCALE:
 		switch (chan->type) {
 		case IIO_VOLTAGE:
-			mutex_lock(&indio_dev->mlock);
+			mutex_lock(&st->lock);
 			*val = st->scale_avail[AD7192_CONF_GAIN(st->conf)][0];
 			*val2 = st->scale_avail[AD7192_CONF_GAIN(st->conf)][1];
-			mutex_unlock(&indio_dev->mlock);
+			mutex_unlock(&st->lock);
 			return IIO_VAL_INT_PLUS_NANO;
 		case IIO_TEMP:
 			*val = 0;

@@ -634,6 +635,8 @@ static int ad7192_probe(struct spi_device *spi)
 
 	st = iio_priv(indio_dev);
 
+	mutex_init(&st->lock);
+
 	st->avdd = devm_regulator_get(&spi->dev, "avdd");
 	if (IS_ERR(st->avdd))
 		return PTR_ERR(st->avdd);

2.7.4

The IIO subsystem is redefining iio_dev->mlock to be used by the IIO core only for protecting device operating mode changes. ie. Changes between INDIO_DIRECT_MODE, INDIO_BUFFER_* modes.

In this driver, private driver lock is being used to protect hardware state changes as here state variables are being changed.

---
 drivers/staging/iio/adc/ad7192.c | 2 ++
 1 file changed, 2 insertions(+)

index 096d29d..609a649 100644
@@ -511,6 +511,7 @@ static int ad7192_write_raw(struct iio_dev *indio_dev,
 	switch (mask) {
 	case IIO_CHAN_INFO_SCALE:
 		ret = -EINVAL;
+		mutex_lock(&st->lock);
 		for (i = 0; i < ARRAY_SIZE(st->scale_avail); i++)
 			if (val2 == st->scale_avail[i][1]) {
 				ret = 0;

@@ -524,6 +525,7 @@ static int ad7192_write_raw(struct iio_dev *indio_dev,
 				ad7192_calibrate_all(st);
 				break;
 			}
+		mutex_unlock(&st->lock);
 		break;
 	case IIO_CHAN_INFO_SAMP_FREQ:
 		if (!val) {

2.7.4

For this to make sense both changes need to be in the same patch. You are removing the protection in patch 1 and then adding it back in patch 2. A patch series shouldn't leave the driver in a broken intermediate state if only part of the series is applied.