Re: [PATCH v3 3/3] Bluetooth: Add BT_MODE socket option

[Marcel Holtmann <marcel@holtmann.org>]

Hi Luiz,

>>> This adds BT_MODE socket option which can be used to set L2CAP modes,
>>> including modes only supported over LE which were not supported using
>>> the L2CAP_OPTIONS.
>>> 
>>> Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
>>> ---
>>> include/net/bluetooth/bluetooth.h | 11 ++++
>>> net/bluetooth/l2cap_sock.c        | 96 +++++++++++++++++++++++++++++++
>>> 2 files changed, 107 insertions(+)
>>> 
>>> diff --git a/include/net/bluetooth/bluetooth.h b/include/net/bluetooth/bluetooth.h
>>> index 1576353a2773..34191e34bfdc 100644
>>> --- a/include/net/bluetooth/bluetooth.h
>>> +++ b/include/net/bluetooth/bluetooth.h
>>> @@ -139,6 +139,17 @@ struct bt_voice {
>>> #define BT_PHY_LE_CODED_TX    0x00002000
>>> #define BT_PHY_LE_CODED_RX    0x00004000
>>> 
>>> +#define BT_MODE                      15
>>> +
>>> +#define BT_MODE_BASIC                0x00
>>> +#define BT_MODE_RETRANS              0x01
>>> +#define BT_MODE_FLOWCTL              0x02
>>> +#define BT_MODE_ERTM         0x03
>>> +#define BT_MODE_STREAMING    0x04
>>> +#define BT_MODE_EXT_FLOWCTL  0x05
>>> +
>>> +#define BT_MODE_LE_FLOWCTL   0x80
>>> +
>> 
>> what I would do is just this:
>> 
>>        BASIC           0x00
>>        ERTM            0x01
>>        STREAMING       0x02
>>        LE_FLOWCTL      0x03
>>        EXT_FLOWCTL     0x04
>> 
>> Trying to cling onto some old L2CAP definition from the 2.1 days is not helpful. I would really make a clean cut here.
> 
> Just to confirm, that means we will not going to support the old
> flowctl and retransmit modes from BR/EDR with BT_MODE?

we currently don’t support these two modes, nor are there any specs that require it.

>> This way we can also cleanly check the available modes per selected socket and have either setsockopt or connect fail appropriately.
>> 
>>> __printf(1, 2)
>>> void bt_info(const char *fmt, ...);
>>> __printf(1, 2)
>>> diff --git a/net/bluetooth/l2cap_sock.c b/net/bluetooth/l2cap_sock.c
>>> index 117ba20ea194..f2bb376c699f 100644
>>> --- a/net/bluetooth/l2cap_sock.c
>>> +++ b/net/bluetooth/l2cap_sock.c
>>> @@ -500,6 +500,25 @@ static int l2cap_sock_getsockopt_old(struct socket *sock, int optname,
>>>      return err;
>>> }
>>> 
>>> +static u8 l2cap_get_mode(struct l2cap_chan *chan)
>>> +{
>>> +     switch (chan->mode) {
>>> +     case L2CAP_MODE_BASIC:
>>> +     case L2CAP_MODE_RETRANS:
>>> +     case L2CAP_MODE_FLOWCTL:
>>> +     case L2CAP_MODE_ERTM:
>>> +     case L2CAP_MODE_STREAMING:
>>> +             /* Mode above are the same on both old and new sockopt */
>>> +             return chan->mode;
>>> +     case L2CAP_MODE_LE_FLOWCTL:
>>> +             return BT_MODE_FLOWCTL;
>>> +     case L2CAP_MODE_EXT_FLOWCTL:
>>> +             return BT_MODE_EXT_FLOWCTL;
>>> +     }
>>> +
>>> +     return chan->mode;
>>> +}
>>> +
>> 
>> Don’t bother with this. Keep the old socket and new socket independent code. I also want to add Kconfig option later that will allow us to disable the old socket options once we have SOL_L2CAP requirement eradicated.
> 
> The reason I had the defines intermixed was that application would be
> able to use the old sockopt to set it and then use BT_MODE to read it,
> in which case I may need to store the actual socket mode separately
> from the chan->mode and then perhaps fail if application attempt to
> read it with BT_MODE if was not set using it.

If someone used L2CAP_OTIONS, then just lets fail BT_MODE read and write.

Regards

Marcel