Re: [PATCH 07/11] staging: iio: ad2s1200: Ensure udelay(1) in all necessary code paths

[David Julian Veenstra <davidjulianveenstra@gmail.com>]

On 23, March 2018 14:20, Jonathan Cameron wrote:

> On Sun, 18 Mar 2018 14:36:15 +0100
> David Veenstra <davidjulianveenstra@gmail.com> wrote:
>
>> After a successful spi transaction, a udelay(1) is needed.
>> This doesn't happen for the default case of the switch statement
>> in ad2s1200_read_raw. This patch makes sure that it does.
>> 
>> Signed-off-by: David Veenstra <davidjulianveenstra@gmail.com>
> Given this one can't actually happen as the chan->type
> is always one of the two options, I can't see the point
> in making sure we sleep.
>
> The default is really just there to catch bugs and to suppress
> the build warning we would otherwise get.
>
> So I am unconvinced on this patch.

I wasn't sure about this patch, but tried to be as critical as possible.
I'll leave it out for V2.

Best regards,
David Veenstra

>
> Jonathan
>> ---
>>  drivers/staging/iio/resolver/ad2s1200.c | 8 +++++---
>>  1 file changed, 5 insertions(+), 3 deletions(-)
>> 
>> diff --git a/drivers/staging/iio/resolver/ad2s1200.c b/drivers/staging/iio/resolver/ad2s1200.c
>> index e0e7c88368ed..6ce9ca13094a 100644
>> --- a/drivers/staging/iio/resolver/ad2s1200.c
>> +++ b/drivers/staging/iio/resolver/ad2s1200.c
>> @@ -78,20 +78,22 @@ static int ad2s1200_read_raw(struct iio_dev *indio_dev,
>>  	switch (chan->type) {
>>  	case IIO_ANGL:
>>  		*val = vel >> 4;
>> +		ret = IIO_VAL_INT;
>>  		break;
>>  	case IIO_ANGL_VEL:
>>  		*val = sign_extend32((s16)vel >> 4, 11);
>> +		ret = IIO_VAL_INT;
>>  		break;
>>  	default:
>> -		mutex_unlock(&st->lock);
>> -		return -EINVAL;
>> +		ret = -EINVAL;
>> +		break;
>>  	}
>>  
>>  	/* delay (2 * AD2S1200_TSCLK + 20) ns for sample pulse */
>>  	udelay(1);
>>  	mutex_unlock(&st->lock);
>>  
>> -	return IIO_VAL_INT;
>> +	return ret;
>>  }
>>  
>>  static const struct iio_chan_spec ad2s1200_channels[] = {