Re: [PATCH 6/9] android/bluetooth: Add support for remote device RSSI

[Anderson Lizardo <anderson.lizardo@openbossa.org>]

Hi Szymon,

On Thu, Dec 12, 2013 at 5:58 AM, Szymon Janc <szymon.janc@tieto.com> wrote:
>> On Wed, Dec 11, 2013 at 7:31 AM, Szymon Janc <szymon.janc@tieto.com> wrote:
>> \> @@ -2306,11 +2309,13 @@ static uint8_t
>> get_device_friendly_name(struct device *dev)
>> >
>> >  static uint8_t get_device_rssi(struct device *dev)
>> >  {
>> > -       DBG("Not implemented");
>> > +       if (!dev->rssi)
>> > +               return HAL_STATUS_FAILED;
>>
>> 0 dBm is a valid RSSI value, so I suppose this code needs some other
>> way to identify an uninitialized value (e.g. a second boolean field.)
>
> Kernel sends rssi==0 if rssi info was not present on inquiry so I used this
> value, but maybe it should use 127 for that.. ? (yet, this might cause trouble
> as linux bluetoothd is checking 0 as well)

I remember commenting on this when mgmt API (it was for TX Power,
which is more common to be 0 dBm, but I can't remember what was the
result of the discussion). In any case, I think handling this way will
make devices that use TX Power higher than 0 dBm (and thus can make a
0dBm RSSI possible) not report their RSSI. It is not common IMHO, but
it is possible.

Best Regards,
-- 
Anderson Lizardo
Instituto Nokia de Tecnologia - INdT
Manaus - Brazil