- new intrusive linked list container (c++ template) that sorts on insertion
- primarily for convenience inspecting things in the system like uORB or WorkQueues
- uorb status or top sorted alphabetically
- work_queue status threads sorted by priority, then items sorted alphabetically within each
- track and publish validity based on hover thrust variance, innovation test ratio, and hysteresis
- only publish on actual updates or becoming inactive
- fix dt (previous timestamp wasn't being saved)
- use local position timestamp (corresponding) to accel data rather than current time to avoid unnecessary timing jitter
- check local position validity before using
- mc_hover_thrust_estimator: move from wq:lp_default -> wq:nav_and_controllers to ensure the hover thrust estimator runs after the position controller and uses the same vehicle_local_position data
- land_detector: check hover thrust estimate validity and adjust low throttle thresholds if hover thrust is available
- mc_pos_control: only use hover thrust estimate if valid
- if "landed" and "maybe_landed" states are false then both the "hit_ground" and the "low_thrust" condition need to be true in order to detect landing
- ground contact MC NAN setpoint workaround
- ground contact additionally check acceleration setpoint
- schedule with vehicle_local_position updates (most updates require valid local position)
- don't allow LNDMC_Z_VEL_MAX to exceed MPC_LAND_SPEED
- ground contact horizontal movement checks default to failed if estimates aren't available
- always check system field for validity
- reject any data outside of "servo position" valid range from Spektrum specification
- properly support XPlus channels (12+)
- debug message if channel count changes
- log full sensor_combined by default for now
- small decrease to input_rc rate (manual_control_setpoint is no longer filtered)
- sensor_correction can be logged at full rate (low publication rate)
- this is to prevent vtol_att_control running at absurdly high rates (kilohertz) from every single actuator controls publication from both the multicopter and fixed wing controllers
- I don't really like this solution, but it works for now until there are larger VTOL architectural changes in place
- this lowers the load immediately at boot while still allowing each VehicleIMU instance to properly measure the actual accel/gyro data rates
- for lockstep builds this is ignored and VehicleIMU initially runs with every new sensor_accel/sensor_gyro publication
- vehicle_air_data won't necessary have an update every iteration and these adjusted throttle values aren't stored
- this only would have worked in the past because the vast majority of systems were using the ms5611 barometer with a publication rate that's faster than the controller
This fixes the takeoff and land CI test where we do a takeoff and land
without RC connected.
In my opinion it does not make sense to go into RC lost failsafe if
a datalink is still available.
Of course, this can be dangerous if someone has a companion computer
connected which gets detected as "data link" but actually has no
connection to the ground station.
- increase rate
- cpu load calculation grab timestamp atomically
- only check one task per cycle (but cycle at a higher rate)
- decrease available FD threshold
- minor cleanup
bresch