Re: [PATCH BlueZ 4/5] core: device: Add device_is_bonded_for_gatt

[Arman Uguray <armansito@chromium.org>]

Hi Luiz,

> On Thu, Jan 29, 2015 at 5:52 AM, Luiz Augusto von Dentz <luiz.dentz@gmail.com> wrote:
> Hi Arman,
>
> On Thu, Jan 29, 2015 at 4:11 AM, Arman Uguray <armansito@chromium.org> wrote:
>> Added the device_is_bonded_for_gatt function, which returns true if
>> the bonding state is true for the bearer over which the GATT db was
>> populated. For now, this only returns the LE state but added for
>> better layering, so that future changes to the internal logic can be
>> simply performed here to avoid regressions.
>>
>> Change-Id: I6ebd16a1111aa38645a0a072536f4171d246b931
>> ---
>>  src/device.c | 14 +++++++++++++-
>>  src/device.h |  1 +
>>  2 files changed, 14 insertions(+), 1 deletion(-)
>>
>> diff --git a/src/device.c b/src/device.c
>> index 9cb5acb..2ef2b6a 100644
>> --- a/src/device.c
>> +++ b/src/device.c
>> @@ -506,8 +506,10 @@ static void gatt_client_cleanup(struct btd_device *device)
>>         bt_gatt_client_unref(device->client);
>>         device->client = NULL;
>>
>> -       if (!device->le_state.bonded)
>> +       if (!device_is_bonded_for_gatt(device)) {
>> +               DBG("Device is not bonded, clearing client attribute cache");
>>                 gatt_db_clear(device->db);
>> +       }
>>  }
>>
>>  static void attio_cleanup(struct btd_device *device)
>> @@ -638,6 +640,16 @@ bool device_is_bonded(struct btd_device *device, uint8_t bdaddr_type)
>>         return state->bonded;
>>  }
>>
>> +bool device_is_bonded_for_gatt(struct btd_device *device)
>> +{
>> +       /*
>> +        * TODO: Once we properly support GATT over BR/EDR, we should remember
>> +        * whether the GATT client  database was populated over LE or BR/EDR and
>> +        * check the correct bonding state.
>> +        */
>> +       return device->le_state.bonded;
>> +}
>> +
>>  gboolean device_is_trusted(struct btd_device *device)
>>  {
>>         return device->trusted;
>> diff --git a/src/device.h b/src/device.h
>> index a7fefee..b4aead1 100644
>> --- a/src/device.h
>> +++ b/src/device.h
>> @@ -82,6 +82,7 @@ const char *device_get_path(const struct btd_device *device);
>>  gboolean device_is_temporary(struct btd_device *device);
>>  bool device_is_paired(struct btd_device *device, uint8_t bdaddr_type);
>>  bool device_is_bonded(struct btd_device *device, uint8_t bdaddr_type);
>> +bool device_is_bonded_for_gatt(struct btd_device *device);
>
> Isn't this equivalent to device_is_bonded or perhaps we don't the
> type? I guess it is possible to retrieve in what transport we are
> connected and then check if we it is bonded.
>

It's pretty much the same thing, except it would return the bonding
state based on the transport that did the GATT discovery. I figured it
would be better to have this function in one place so that we only
have one place to change when we properly support GATT over BR/EDR in
the future. Just using device_is_bonded and passing BDADDR_LE_PUBLIC
for the bdaddr works too for now, we just need to remember to update
them later. I'll go ahead and leave this out for now.

>>  gboolean device_is_trusted(struct btd_device *device);
>>  void device_set_paired(struct btd_device *dev, uint8_t bdaddr_type);
>>  void device_set_unpaired(struct btd_device *dev, uint8_t bdaddr_type);
>> --
>> 2.2.0.rc0.207.ga3a616c
>>
>
>
>
> --
> Luiz Augusto von Dentz

Arman