RE: Problems managing a buffer to track events to the ble stack

Ignacio Lopez — Fri, 13 Jun 2014 16:45:52 GMT

Hi,

Thanks for your reply.

This is the poster_value_meas_handle function:

static uint32_t poster_value_meas_handler(void)
{
	uint32_t        err_code;

	err_code = ble_pos_poster_value_measurement_send(m_pos, poster_value);
	
	return err_code;
}

And the ble_pos_poster_value_measurement_send function. It is our own profile and it based on the heart rate profile with some modifications:

uint32_t ble_pos_poster_value_measurement_send(ble_pos_t * p_pos, uint8_t *poster_value)
{
    uint32_t err_code;
    
    // Send value if connected and notifying
    if (p_pos->conn_handle != BLE_CONN_HANDLE_INVALID)
    {
        uint16_t               len;
        uint16_t               hvx_len;
        ble_gatts_hvx_params_t hvx_params;
     
			
				len = poster_value[0] + 1;	// packet length is poster data size plus length byte
        hvx_len = len;

        memset(&hvx_params, 0, sizeof(hvx_params));
 
        hvx_params.handle   = p_pos->pos_handles.value_handle;
        hvx_params.type     = BLE_GATT_HVX_NOTIFICATION;
        hvx_params.offset   = 0;
        hvx_params.p_len    = &hvx_len;
        hvx_params.p_data   = poster_value;
 
        err_code = sd_ble_gatts_hvx(p_pos->conn_handle, &hvx_params);
        if ((err_code == NRF_SUCCESS) && (hvx_len != len))
        {
            err_code = NRF_ERROR_DATA_SIZE;
        }
    }
    else
    {
        err_code = NRF_ERROR_INVALID_STATE;
    }

    return err_code;
}

Yes, I receive all the packets that I am sending on Master Control Panel.

RE: Problems managing a buffer to track events to the ble stack

Hung Bui — Tue, 10 Jun 2014 14:11:39 GMT

Hi Ignaco,

I am sorry indeed for the late response.

Could you provide poster_value_meas_handler() function ? It's pretty strange to see this behaviour. I had previously tested with BLE_EVT_TX_COMPLETE and p_ble_evt->evt.common_evt.params.tx_complete.count but hadn't seen this issue.

It could be that some packets hadn't buffered correctly that made the buffer count (sd_buffers_available) went wrong.

On the central side, do you receive all the packet you queued ? You can track which packet was missing when the buffer count stuck at 2.