pc-ble-driver 1.0 - bonding

[Mike]

Hello,

I work on an application which use the pc-ble-driver 1.0 as BLE central for handling external peripheral devices.

Now, I am working on bonding procedure. To make it simpler I started from 'Just Works' bonding and followed this chart. So, after established connection with peripheral device I call sd_ble_gap_authenticate as follows:

ble_gap_sec_params_t ble_gap_sec_params;
memset(&ble_gap_sec_params, 0, sizeof(ble_gap_sec_params));
ble_gap_sec_params.bond         = 1;
ble_gap_sec_params.mitm         = 0;
ble_gap_sec_params.lesc         = 0;
ble_gap_sec_params.keypress     = 0;
ble_gap_sec_params.io_caps      = BLE_GAP_IO_CAPS_NONE;
ble_gap_sec_params.oob          = 0;
ble_gap_sec_params.min_key_size = 7;
ble_gap_sec_params.max_key_size = 16;
ble_gap_sec_params.kdist_peer.enc = 1;
ble_gap_sec_params.kdist_peer.id  = 1;

err_code = sd_ble_gap_authenticate(p_adapter, p_ble_gattc_evt->conn_handle, &ble_gap_sec_params);

It works well (returns NRF_SUCCESS) and after that I get the BLE_GAP_EVT_SEC_PARAMS_REQUEST event. So, it seems to be OK to this point. My problem starts on calling sd_ble_gap_sec_params_reply. I do it as follows:

switch (p_ble_evt->header.evt_id)
{
...
case BLE_GAP_EVT_SEC_PARAMS_REQUEST:
    {
        ble_gap_sec_keyset_t keys_exchanged;
 
        keys_exchanged.keys_own.p_enc_key  = NULL;
        keys_exchanged.keys_own.p_id_key   = NULL;
        keys_exchanged.keys_own.p_sign_key = NULL;
        keys_exchanged.keys_own.p_pk       = NULL;
        keys_exchanged.keys_peer.p_enc_key = &bond_table[0].peer_enc_key;
        keys_exchanged.keys_peer.p_id_key  = &bond_table[0].peer_id;
        keys_exchanged.keys_peer.p_sign_key  = NULL;
        keys_exchanged.keys_peer.p_pk        = NULL;

        err_code = sd_ble_gap_sec_params_reply(p_adapter,
                                               p_ble_evt->evt.gap_evt.conn_handle,
                                               BLE_GAP_SEC_STATUS_SUCCESS,
                                               NULL,
                                               &keys_exchanged);
        
        printf("sd_ble_gap_sec_params_reply err_code: 0x%X\n", err_code);
    }
    break;
...
}

The main problem is that the sd_ble_gap_sec_params_reply doesn't return anything. The application just hangs on this call.

Could you tell me why it happens? Do I something wrong? If so, why the API call doesn't return an error?

---

Best regards,

Mike

[Jørgen Holmefjord]

This is on your peripheral? Have you tried performing a sniffer trace to see what is happening on-air?

[Mike]

@Jørgen Holmefjord This code is from my central device. I did not sniff on-air packets but I tracked what kind of events my peripheral receive during bonding attempt (the peripheral device is also my custom device nrf51-based, so I can debug it).

I found that the peripheral gets only BLE_GAP_EVT_SEC_PARAMS_REQUEST event, nothing more. So. it seems that the sd_ble_gap_sec_params_reply call on the central has no effect.

I just add, that the peripheral device bonds fine with nRF Connect app on the phone.

[Jørgen Holmefjord]

Could you upload your central code? If you don't want to share it in public, you can also create a case in MyPage and link to this case.

[Rick S.]

Hi Mike, I know it's been a while since you asked this... :)

I got tripped up at the same point, sd_ble_gap_sec_params_reply was causing an assertion failure due to a recursive mutex lock.

I got around this by calling sd_ble_gap_sec_params_reply from the main thread instead of the event handler. (Main function calls Win32 WaitForSingleObject, handler calls SetEvent)

This seems like a bug to me, I haven't had a chance to make an official report.

Best of luck, Rick

[Rick S.]

P.S. This is the pc-ble-driver from July, don't know the version number