- Feb 19, 2016
-
-
jwilson authored
-
jwilson authored
-
jwilson authored
-
jwilson authored
-
Julian Oes authored
-
Julian Oes authored
Not sure why they got removed in the first place in the merge. I also don't know how they are used and what they are for, comments welcome.
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
There is no such thing as set_param_no_autosave on QURT, therefore just save it anyway. On the Snapdragon the overhead should not be a problem.
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
Otherwise SITL doesn't work anymore, at least on Linux
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
This mainly fixes the NuttX build.
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
-
Julian Oes authored
Don't try to open the mavlink_fd on QURT because the px4_ioctl leads to timeouts of the attitude_estimator_q loop.
-
Mark Charlebois authored
mainapp was not restoring the original terminal setttings on exit Signed-off-by:
Mark Charlebois <charlebm@gmail.com>
-
Julian Oes authored
-
Julian Oes authored
Like this the timeouts are going through instead of hanging everything.
-
Julian Oes authored
It was wrong to subtract seconds and nanoseconds from each other. This will wrap, therefore use ts_to_abstime and do the subtraction afterwards.
-
Julian Oes authored
The timeout was triggered using absolute time instead of a delay in usec. This lead to the system hanging. With the fix it continues after the timeout, however, the rates still don't seem right.
-