40604 Commits

Author SHA1 Message Date
Daniel Agar
247b975675 Tools/HIL: disable XON/XOFF, consistent timeouts, proper input buffer clear 2021-12-31 19:28:40 -05:00
Daniel Agar
03371f8522 NuttX with bch flush backport 2021-12-31 10:59:56 -08:00
David Sidrane
db159a43cc test_dataman:Set SEM_PRIO_NONE on all semaphores 2021-12-31 10:27:24 -05:00
Daniel Agar
1b6700592b parameters: require valid BSON document size 2021-12-30 16:36:23 -05:00
Daniel Agar
3514458f42 Tools/HIL/test_airframes.sh: dump all parameter storage after bootup to verify contents 2021-12-30 12:43:00 -05:00
Daniel Agar
b06215d038
Tools/HIL/test_airframes.sh: run param dump after save to verify BSON 2021-12-30 11:11:14 -05:00
PX4 BuildBot
5fc7e551ab Update submodule mavlink to latest Wed Dec 29 12:38:13 UTC 2021
- mavlink in PX4/Firmware (395eeb440a2ca7503591f872d4f733fcd9867218): bfaf605bd6
    - mavlink current upstream: 4bc0de9c38
    - Changes: bfaf605bd6...4bc0de9c38

    4bc0de9c 2021-12-27 olliw42 - Component Information Basics: add camera cap flag (#1752)
8035ad4d 2021-12-27 Hamish Willee - GIMBAL_DEVICE_FLAGS_NEUTRAL - note that this can be any angle (#1758)
2021-12-29 17:16:49 -05:00
Daniel Agar
6706d9b434 Jenkins: always fully clean git workspace 2021-12-29 17:14:11 -05:00
Daniel Agar
6d0339ba0c I2CSPIDriverBase: sensor start failure ERROR if internal, WARN if external 2021-12-28 11:05:35 -05:00
Daniel Agar
990d7c159d pwm_out: cleanup and prep for linux compatibility 2021-12-28 11:04:00 -05:00
Daniel Agar
10f4a2e91f boards: omnibus_f4sd_default disable systemcmds to save flash 2021-12-27 12:19:55 -05:00
Daniel Agar
3ef9c2d16c boards: holybro_kakutef7_default disable modules/gyro_calibration to save flash 2021-12-27 12:19:55 -05:00
FengShun
8f8719d33b cmake: use add_definitions instead of add_compile_definitions
Using make tests in docker will fail, because the current basic image used by px4io/px4-dev-simulation-bionic is ubuntu18.04, the default version of cmake is 3.10, and the add_compile_definitions command is only available in cmake 3.12+(ubuntu 20.04).
2021-12-27 12:19:31 -05:00
FengShun
65e976bbc2 docker: update px4-dev-simulation-bionic container versions to 2021-12-11
The current docker image will cause "make tests" to fail: "kconfiglib is not installed or not in PATH"
2021-12-27 12:19:31 -05:00
Daniel Agar
d94767ef88 cmake: NuttX use cygwin friendly path for linker script 2021-12-27 12:19:12 -05:00
honglang
17a99bc827 mavlink: add HYGROMETER_SENSOR stream 2021-12-27 12:13:09 -05:00
honglang
983867f9af uavcan: support uavcan hygrometer 2021-12-27 12:13:09 -05:00
honglang
b60e59d9be msg: new sensor_hygrometer msg 2021-12-27 12:13:09 -05:00
Daniel Agar
3f17acdcc9 boards: mro_ctrl-zerl-h7-oem pin fixes (sync with mro_ctrl-zero-h7) 2021-12-27 09:25:21 -05:00
Daniel Agar
0f4e1dd9f9 boards: mro ctrl-zero-h7 pin fixes 2021-12-27 09:25:21 -05:00
Daniel Agar
236794235a parameters: export empty BSON document even if there are no parameters to save
- otherwise a 'param reset_all' won't be saved until a parameter is changed
2021-12-27 09:17:00 -05:00
Daniel Agar
33d7b42040 parameters: import quietly retry if there's no data (as a precuation) 2021-12-27 09:17:00 -05:00
Daniel Agar
20d3e6f4e1 parameters: add simple BSON verification pass after export 2021-12-27 09:17:00 -05:00
Beat Küng
72065c3d71 fix protocol_splitter: remove timeout, drop buffer if too full instead
The existing implementation had a flaw: when the buffer was getting full,
mavlink started to busy-loop, as the uart has data (poll returns immediately)
but no new data was read from the uart due to the buffer being full.
As rtps is running at lower prio, it never got the chance to read again,
making the problem even worse.
After 1s the timeout triggered and the buffer was cleared, so it recovered.

Instead of allowing for CPU spikes, we now immediately clear the buffer
(only as much as we have to), ensuring that new data is read from the uart.
2021-12-24 20:28:43 -05:00
Rui Miguel Carvalho
542ee86bc9
landing_target_estimator: custom irlock frame params (#18884)
Signed-off-by: ruimscarvalho98 <ruimsc98@gmail.com>
2021-12-24 20:20:35 -05:00
Silvan Fuhrer
28e01c3510 FW Position controller: make loiter switching logic robust against pos_sp with loiter_rad=0
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2021-12-24 20:17:15 -05:00
Daniel Agar
98d706772e boards: px4_fmu-v2_rover disable events module to save flash 2021-12-24 20:06:13 -05:00
Beat Küng
d202bf9631 control_allocator: add scope to reduce stack usage by 256B 2021-12-24 20:06:13 -05:00
Silvan Fuhrer
48be55dec8 Tiltrotor: disable tilt/thrust compensation when dynamic allocation is enabled
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2021-12-24 20:06:13 -05:00
Silvan Fuhrer
31d7328f4d Tailsitter: invert roll (make it NED-correct) for filling vehicle_torque_setpoint
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2021-12-24 20:06:13 -05:00
Silvan Fuhrer
1ad66e606b ControlAllocation: introduce convention for control surface deflection sign
- horizontal control surfaces: up=positive deflection
- vertical control surfaces (rudder): right=positive deflection
- mixed (V-Tail): up=positive deflection

Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2021-12-24 20:06:13 -05:00
Beat Küng
5259877b1b control_allocator: update matrix normalization
- only normalize rpy for MC matrices
- for thrust use the 3D vector, so it works for FW and tilt rotors as well

This keeps MC unchanged.
2021-12-24 20:06:13 -05:00
Silvan Fuhrer
460a0df850 tiltrotor: fill out thrust_setpoint[0] also in hover, based on tilt and hover thrust
In the tiltrotor case, beside an F_z thrust setpoint also a F_x setpoint must be passed
to the allocator as the matrix has non-zero thrust-x effectiveness when tilts are applied.

Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2021-12-24 20:06:13 -05:00
Beat Küng
2333bef670 px4airframes doc: handle common outputs individually
preparation for dynamic control allocation, where we won't have MAIN vs
AUX anymore (at least for the generic frames).
2021-12-24 20:06:13 -05:00
Beat Küng
203e253dfe airframes: sitl boat: switch to dynamic control allocation 2021-12-24 20:06:13 -05:00
Beat Küng
8486ca5766 airframes: sitl cloudship: switch to dynamic control allocation 2021-12-24 20:06:13 -05:00
Beat Küng
9bc9169b77 px4/fmu-v5x: add missing board variants to rc.board_mavlink 2021-12-24 20:06:13 -05:00
Beat Küng
09200b994d airship_att_control: publish thrust + torque setpoint 2021-12-24 20:06:13 -05:00
Silvan Fuhrer
478724c9fe EffectivenessTiltrotor: fix logic for matrix update
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2021-12-24 20:06:13 -05:00
Beat Küng
6c8f322dfe px4/fmu-v2/test: disable modules to reduce flash 2021-12-24 20:06:13 -05:00
Beat Küng
082cd74cab holybro/kakutef7: disable IST8310 to reduce flash 2021-12-24 20:06:13 -05:00
Beat Küng
8e04934eab commander: consider COM_MOT_TEST_EN for COMMAND_ACTUATOR_TEST as well 2021-12-24 20:06:13 -05:00
Silvan Fuhrer
39d441317a ActuatorEffectivenssTailsitterVTOL: only enable MC yaw control if more than 3 rotors
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2021-12-24 20:06:13 -05:00
Silvan Fuhrer
426aa76a59 CA/VTOL: change description of VT_ELEV_MC_LOCK as it applies to not just elevons
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2021-12-24 20:06:13 -05:00
Silvan Fuhrer
18008b2de7 CA module.yaml: add motor direction function also to UI for Tailsitter an Tiltrotor
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2021-12-24 20:06:13 -05:00
Beat Küng
8743b78474 px4/fmu-v2/rover: disable modules to reduce flash 2021-12-24 20:06:13 -05:00
Beat Küng
8c782b7cd9 control_allocator: add Custom + MC with tilts effectiveness
The param group is changed to Geometry, as this reflects the naming in the
QGC UI.
2021-12-24 20:06:13 -05:00
Beat Küng
0818bb4be0 bitcraze/crazyflie{,21}: disable modules to reduce flash 2021-12-24 20:06:13 -05:00
Beat Küng
751539304e omnibus/f4sd: disable modules to reduce flash 2021-12-24 20:06:13 -05:00
Beat Küng
627ec0df5d control_allocator: increase STACK_MAIN to 3000
It was exceeding 2000
2021-12-24 20:06:13 -05:00