* logger: disable some default topics, which are most likely not used
This is also to safe CPU and lower the amount of file descriptors used.
* logger: use the hrt timer for more accurate scheduling
Under NuttX with the default rate of 285Hz, the actual measured rate was
only 200Hz while on Linux it was ~280Hz. The reason is that NuttX only
uses a usleep() granularity of 1ms, so that the typical sleep time is
longer than what we set.
Now the logger waits on a semaphore, which gets activated periodically
with a hrt timer. With this the measured rate is exactly the expected one,
285Hz.
This fixes a compile error with GCC 5.2.0 and cross-compilation:
/usr/arm-none-eabi/include/c++/5.2.0/arm-none-eabi/armv7e-m/fpu/bits/error_constants.h:122:27: error: 'ENOTSUP' was not declared in this scope
not_supported = ENOTSUP,
The problem was that the #include <iomanip> was taken from the host system,
which conflicts with includes from NuttX (NuttX has only basic C++ STL
support)
This fixes a bug with following setup:
- two (or N > 1) connected gyros
- ekf2 enabled
In this case, sensors would publish with the combined rate of the gyros,
but with N following messages having the same gyro data & timestamp.
Apparently ekf2 cannot handle this, the other estimators can.
We may want to rethink what the proper solution is here.
The MavlinkOrbSubscription only had an interface to either always copy
or copy based on a timestamp. This commit adds a copy interface if the
topic has been updated.
since pure manual control for fixed wings in handled on the io side
the scale parameters for roll, pitch and yaw had to be introduced there
as well.
Signed-off-by: Roman <bapstr@ethz.ch>
These will only start replay if there's an environment variable 'replay',
otherwise they do nothing.
We apply the parameters even before loading the modules, so that the
modules 'see' the same state as the recorded system had. We will have to
see how well this works in practice.
Julian Oes