During a round corner the L1 distance calculation
was only done in 2D and the z-axis coordinate
was directly coming from the next waypoint.
This lead to an unpredictable altitude profile
between two waypoints. Sometimes almost all
altitude difference was already covered during
the turn instead of going diagonally.
This sets the src/drivers/gps/devices submodule to the latest
release/1.14 branch. This fixes a potential issue with the Unicore M10
GPS driver, making sure the AGRICA message is requested.
Signed-off-by: Julian Oes <julian@oes.ch>
This switches from the horribly intertwined ringbuffer implementation to
the new VariableLengthRingbuffer implementation.
By ditching the previous implementation, we fix MAVLink message
forwarding, which didn't work reliably. The reason it didn't work is
that multiple mavlink messages could be added but only one of them was
sent out because the buffer didn't keep track of the messages properly
and only read the first one.
Signed-off-by: Julian Oes <julian@oes.ch>
This adds a reusable class for a FIFO ringbuffer that accepts variable
length packets. It is using the Ringbuffer class internally.
Signed-off-by: Julian Oes <julian@oes.ch>
During transition to FF, only allow update thrust axis up to 45° as with a high tilt angle
the effectiveness of the thrust axis in z is apporaching 0, and by that is increasing
the motor output to max.
Transition to HF: disable thrust axis tilting, and assume motors are vertical. This is to avoid
a thrust spike when the transition is initiated (as then the tilt is fully forward).
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
The length check for unknown packets did not include PACKET_SIZE_TYPE_SIZE
and CRC_SIZE, and hence working_index could overflow CRSF_MAX_PACKET_LEN,
triggering invalid memory access further down in QueueBuffer_PeekBuffer.
Also the working_segment_size was wrong for unknown packets.
Credits for finding this go to @Pwn9uin.
A setpoint of type IDLE can be published by Navigator without a valid position,
and should be handled in the auto function in FW_POSCTRL_MODE_AUTO. If it wouldn't be
handled there the controller would switch to mode FW_POSCTRL_MODE_OTHER, in which case
no attitude setpoint is published and the lower controllers would be stuck with the
last published value (incl. thrust).
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
* FW Positon Controller: set altitude_ref to 0 if not provided by GPS
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
* FW Positon Controller: set lat/lon reference to 0 if not provided in local_position
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
---------
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
- force initialize takeoff airspeed setpoint at start of takeoff modes
- force set airspeed constraints if slewed value is out of bounds
- always slew airspeed setpoints as long as inside constraints
- move target airspeed setpoint calculation into mode specific logic regions (hand vs runway)
when we're in a takeoff situation, we only want to adapt the airspeed to
avoid accelerated stall due to load factor changes. Disable othre logic
like minimum ground speed, wind based adaption and airspeed slew rating.
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
If the measured voltage is more than 65v we need to split the voltage
over multiple cells in order to avoid overflowing the uint16. This is
according to the MAVLink spec.
Signed-off-by: Julian Oes <julian@oes.ch>
The extension fields need to be 0 by default according to the MAVLink
spec. This is because extensions are 0 by default and need to be 0 when
unknown/unused for backwards compatibility.
The patch also simplifies the flow slightly in that it doesn't create a
temporary array but just fills in the cell voltages directly.
Signed-off-by: Julian Oes <julian@oes.ch>
From looking at the history the BMM150 rotation was initially 0. Then,
this was changed to 6 when the intent was to only change it for Skynode.
A bit later, the rotation was changed back to 0, but only for Skynode.
This tells me that rotation 0 was correct for all 6X including Skynode
all along.
Signed-off-by: Julian Oes <julian@oes.ch>
This is the same jump off point as main and release/1.14
with several NXP arch backports, 2 Commit fixing bad DEBUGASSERT
logic and a Soket Can change.
