[PATCH 0/2] bmc150 magnetometer cleanup

[PATCH 0/2] bmc150 magnetometer cleanup

[Irina Tirdea]

A couple of small cleanup patches starting from Hartmut's series.
There are no conflicts with the rest of Hartmut's patches, so they
can be applied in any order.

Irina Tirdea (2):
  iio: magn: bmc150_magn: add locking comment for runtime resume
  iio: magn: bmc150_magn: do not set power state twice when setting
    trigger state

 drivers/iio/magnetometer/bmc150_magn.c | 13 +++++--------
 1 file changed, 5 insertions(+), 8 deletions(-)

-- 
1.9.1

[PATCH 1/2] iio: magn: bmc150_magn: add locking comment for runtime resume

[Irina Tirdea]

Runtime resume function is called with the data->mutex lock held.
Add a comment to indicate this.

Signed-off-by: Irina Tirdea <irina.tirdea@intel.com>
---
 drivers/iio/magnetometer/bmc150_magn.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/iio/magnetometer/bmc150_magn.c b/drivers/iio/magnetometer/bmc150_magn.c
index 074a99d..d66f8cf 100644
--- a/drivers/iio/magnetometer/bmc150_magn.c
+++ b/drivers/iio/magnetometer/bmc150_magn.c
@@ -1043,6 +1043,9 @@ static int bmc150_magn_runtime_suspend(struct device *dev)
 	return 0;
 }
 
+/*
+ * Should be called with data->mutex held.
+ */
 static int bmc150_magn_runtime_resume(struct device *dev)
 {
 	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
-- 
1.9.1

[PATCH 2/2] iio: magn: bmc150_magn: do not set power state twice when setting trigger state

[Irina Tirdea]

When setting the trigger state, the device power state is set through
buffer preenable and postdisable hooks. There is no need
to also set it in the trigger set state call.

Remove duplicate set power state from the trigger set
state call.

Signed-off-by: Irina Tirdea <irina.tirdea@intel.com>
---
 drivers/iio/magnetometer/bmc150_magn.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)

diff --git a/drivers/iio/magnetometer/bmc150_magn.c b/drivers/iio/magnetometer/bmc150_magn.c
index d66f8cf..38c2eb64 100644
--- a/drivers/iio/magnetometer/bmc150_magn.c
+++ b/drivers/iio/magnetometer/bmc150_magn.c
@@ -783,29 +783,23 @@ static int bmc150_magn_data_rdy_trigger_set_state(struct iio_trigger *trig,
 	if (state == data->dready_trigger_on)
 		goto err_unlock;
 
-	ret = bmc150_magn_set_power_state(data, state);
-	if (ret < 0)
-		goto err_unlock;
-
 	ret = regmap_update_bits(data->regmap, BMC150_MAGN_REG_INT_DRDY,
 				 BMC150_MAGN_MASK_DRDY_EN,
 				 state << BMC150_MAGN_SHIFT_DRDY_EN);
 	if (ret < 0)
-		goto err_poweroff;
+		goto err_unlock;
 
 	data->dready_trigger_on = state;
 
 	if (state) {
 		ret = bmc150_magn_reset_intr(data);
 		if (ret < 0)
-			goto err_poweroff;
+			goto err_unlock;
 	}
 	mutex_unlock(&data->mutex);
 
 	return 0;
 
-err_poweroff:
-	bmc150_magn_set_power_state(data, false);
 err_unlock:
 	mutex_unlock(&data->mutex);
 	return ret;
-- 
1.9.1

Re: [PATCH 1/2] iio: magn: bmc150_magn: add locking comment for runtime resume

[Jonathan Cameron]

On 23/07/15 18:15, Irina Tirdea wrote:
> Runtime resume function is called with the data->mutex lock held.
> Add a comment to indicate this.
> 
> Signed-off-by: Irina Tirdea <irina.tirdea@intel.com>
Sensible follow up to Hartmut's patches.  

Applied to the togreg branch of iio.git

Thanks,

Jonathan
> ---
>  drivers/iio/magnetometer/bmc150_magn.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/iio/magnetometer/bmc150_magn.c b/drivers/iio/magnetometer/bmc150_magn.c
> index 074a99d..d66f8cf 100644
> --- a/drivers/iio/magnetometer/bmc150_magn.c
> +++ b/drivers/iio/magnetometer/bmc150_magn.c
> @@ -1043,6 +1043,9 @@ static int bmc150_magn_runtime_suspend(struct device *dev)
>  	return 0;
>  }
>  
> +/*
> + * Should be called with data->mutex held.
> + */
>  static int bmc150_magn_runtime_resume(struct device *dev)
>  {
>  	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
> 

Re: [PATCH 2/2] iio: magn: bmc150_magn: do not set power state twice when setting trigger state

[Jonathan Cameron]

On 23/07/15 18:15, Irina Tirdea wrote:
> When setting the trigger state, the device power state is set through
> buffer preenable and postdisable hooks. There is no need
> to also set it in the trigger set state call.
> 
> Remove duplicate set power state from the trigger set
> state call.
> 
> Signed-off-by: Irina Tirdea <irina.tirdea@intel.com>
Applied to the togreg branch of iio.git

Thanks,

Jonathan
> ---
>  drivers/iio/magnetometer/bmc150_magn.c | 10 ++--------
>  1 file changed, 2 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/iio/magnetometer/bmc150_magn.c b/drivers/iio/magnetometer/bmc150_magn.c
> index d66f8cf..38c2eb64 100644
> --- a/drivers/iio/magnetometer/bmc150_magn.c
> +++ b/drivers/iio/magnetometer/bmc150_magn.c
> @@ -783,29 +783,23 @@ static int bmc150_magn_data_rdy_trigger_set_state(struct iio_trigger *trig,
>  	if (state == data->dready_trigger_on)
>  		goto err_unlock;
>  
> -	ret = bmc150_magn_set_power_state(data, state);
> -	if (ret < 0)
> -		goto err_unlock;
> -
>  	ret = regmap_update_bits(data->regmap, BMC150_MAGN_REG_INT_DRDY,
>  				 BMC150_MAGN_MASK_DRDY_EN,
>  				 state << BMC150_MAGN_SHIFT_DRDY_EN);
>  	if (ret < 0)
> -		goto err_poweroff;
> +		goto err_unlock;
>  
>  	data->dready_trigger_on = state;
>  
>  	if (state) {
>  		ret = bmc150_magn_reset_intr(data);
>  		if (ret < 0)
> -			goto err_poweroff;
> +			goto err_unlock;
>  	}
>  	mutex_unlock(&data->mutex);
>  
>  	return 0;
>  
> -err_poweroff:
> -	bmc150_magn_set_power_state(data, false);
>  err_unlock:
>  	mutex_unlock(&data->mutex);
>  	return ret;
>