Re: [PATCH] iio: imu: inv_mpu6050: advertise max and min freqs

[Martin Kelly <mkelly@xevo.com>]

On 05/16/2018 12:51 AM, Jean-Baptiste Maneyrol wrote:
> 
> 
> On 15/05/2018 23:53, Martin Kelly wrote:
>> CAUTION: This email originated from outside of the organization. 
>> Please make sure the sender is who they say they are and do not click 
>> links or open attachments unless you recognize the sender and know the 
>> content is safe.
>>
>>
>> Currently, we allow a minimum FIFO rate of 4 and a max of 1000, but we
>> advertise down to only 10 and up to 1000. Expand the advertised range to
>> reflect the full available range.
>>
>> Signed-off-by: Martin Kelly <mkelly@xevo.com>
>> ---
>>   drivers/iio/imu/inv_mpu6050/inv_mpu_core.c | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c 
>> b/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c
>> index 43fba5f7532b..3f4862f09db3 100644
>> --- a/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c
>> +++ b/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c
>> @@ -799,7 +799,7 @@ static const struct iio_chan_spec 
>> inv_mpu_channels[] = {
>>   };
>>
>>   /* constant IIO attribute */
>> -static IIO_CONST_ATTR_SAMP_FREQ_AVAIL("10 20 50 100 200 500");
>> +static IIO_CONST_ATTR_SAMP_FREQ_AVAIL("4 10 20 50 100 200 500 1000");
>>   static IIO_CONST_ATTR(in_anglvel_scale_available,
>>                                            "0.000133090 0.000266181 
>> 0.000532362 0.001064724");
>>   static IIO_CONST_ATTR(in_accel_scale_available,
>> -- 
>> 2.11.0
>>
> 
> Hello,
> 
> in fact we advertise only frequencies that are supported by the low-pass 
> filter. Filter bandwidths are: 5Hz, 10Hz, 20Hz, 42Hz (~50), 98Hz (~100), 
> 188Hz (~200), which are corresponding to a sampling rate of (x2): 10Hz, 
> 20Hz, ~50Hz, ~200Hz, ~400Hz. Only 500Hz is a little out of specs, since 
> 400Hz is not a possible frequency.
> 
> You can always set the frequency to any supported values, but you can 
> suffer from aliasing.
> 
> I strongly suggest we just keep it like it is today. Better not 
> advertise frequencies that are not correctly filtered.
> 
> JB

OK, that makes sense and is fine with me. Perhaps I should instead just 
add a comment explaining the situation, since I couldn't tell from the 
code itself.