Skip to content
Snippets Groups Projects
  1. Jan 25, 2017
  2. Jan 16, 2017
  3. May 27, 2016
  4. May 24, 2016
  5. May 23, 2016
  6. Jan 02, 2016
  7. Oct 22, 2015
  8. May 06, 2015
  9. Apr 07, 2015
  10. Feb 25, 2015
    • Kevin Hester's avatar
      params: Fix handling of param_index==65535 · 65efeb76
      Kevin Hester authored
      Hi @tridge.  @eliao noticed an interesting bug.  Some other GCS was talking
      to the vehile and using the PARAM_REQUEST_READ variant which uses -1 for
      the param_index (so just searching by string).  Due to a bug (IMO) in ardupilot
      the vehicle was sending back param_values that looked like:
      
      2015-02-22 16:54:48.04: PARAM_VALUE {param_id : WPNAV_ACCEL_Z, param_value : 100.0, param_type : 9, param_count : 462, param_index : 65535}
      2015-02-22 16:54:48.08: PARAM_VALUE {param_id : WPNAV_SPEED_DN, param_value : 250.0, param_type : 9, param_count : 462, param_index : 65535}
      2015-02-22 16:54:48.08: PARAM_VALUE {param_id : WPNAV_SPEED_UP, param_value : 150.0, param_type : 9, param_count : 462, param_index : 65535}
      
      Notice in the mavlink xml that the param_index is a int16_t in the
      request, but a uint16_t in the response.
      
      These 65535s were confusing mavproxy (it was mistakenly adding a 65535 entry
      mav_param_set).
      
      * In ardupilot GCS_MAVLINK::handle_param_request_read it is sending back the param_index that was included
      in the PARAM_REQUEST_READ, which is not correct.  It it should send back the actual index of the param we found (this is why mavproxy is seeing 65535).  I'll send in a PR to master to fix this.
      * mavproxy gets confused when it sees 65535 - I'll make it treat it the same as the request
      65efeb76
  11. Dec 21, 2014
  12. Jul 29, 2014
  13. Jul 28, 2014
  14. May 16, 2014
  15. Apr 30, 2014
  16. Apr 10, 2014
  17. Apr 09, 2014
  18. Apr 06, 2014
  19. Mar 27, 2014
  20. Mar 20, 2014
  21. Mar 13, 2014
  22. Mar 12, 2014
Loading