There's an upcoming patch that moves all static libraries to
`build/<board>/lib/`. That way, the program directory won't be created
automatically by the build system and that will cause problems for PX4 builds,
since it builds ardupilot programs as static libraries and copies stuff to the
program directory.
A (possible) bug in CMake makes it not to relink firmware_nuttx sometimes when
the following sequence of events happen with very short period of time in
between them:
1) cmake target build_firmware_px4fmu-v2 just finishes
2) px4-extra-files/libap_program.a is replaced with the next library
3) cmake target build_firmware_px4fmu-v2 is run
It's unfortunate to have to reconfigure PX4Firmware with a different path
because of the overhead added.
That problem was found with CMake 3.5.2 and is reproducible with the following
bash script (with git HEAD at ae3cb05 'HAL_PX4: disable "Overtime in task"
msg'):
```
modules/waf/waf-light configure --board px4-v2
modules/waf/waf-light build --targets bin/arduplane,bin/arducopter-quad
fails=0
total=0
while true; do
cp build/px4-v2/bin/libarducopter-quad.a \
build/px4-v2/px4-extra-files/libap_program.a
cmake --build build/px4-v2/modules/PX4Firmware \
--target build_firmware_px4fmu-v2 &> /dev/null
h1=$(sha256sum build/px4-v2/modules/PX4Firmware/src/firmware/nuttx/firmware_nuttx \
| cut -d" " -f1)
cp build/px4-v2/bin/libarduplane.a build/px4-v2/px4-extra-files/libap_program.a
cmake --build build/px4-v2/modules/PX4Firmware \
--target build_firmware_px4fmu-v2 &> /dev/null
h2=$(sha256sum build/px4-v2/modules/PX4Firmware/src/firmware/nuttx/firmware_nuttx \
| cut -d" " -f1)
[[ $h1 == $h2 ]] && ((fails++))
((total++))
printf "\r%d/%d" $fails $total
done
```
Create a new configuration task a configuration in the CMakeConfig object
changes. That allows building targets for different configurations in one waf
build run.
Update ccache to 3.2.5
Set ccache size
Display ccache statistics for each build
Disable ccache for configure phase
Export ccache bin path
travis: remove ccache stats before saving cache
Commit 8774f15 causes fly.CopterAVC test failure.
Here is the failure info below:
>>>> FAILED STEP: fly.CopterAVC at Tue May 17 13:21:56 2016 (fly_CopterAVC() got an unexpected keyword argument 'valgrind')
Traceback (most recent call last):
File "./Tools/autotest/autotest.py", line 406, in run_tests
if not run_step(step):
File "./Tools/autotest/autotest.py", line 232, in run_step
return arducopter.fly_CopterAVC(viewerip=opts.viewerip, map=opts.map, valgrind=opts.valgrind)
TypeError: fly_CopterAVC() got an unexpected keyword argument 'valgrind'
('check step: ', 'fly.CopterAVC')
FAILED 1 tests: ['fly.CopterAVC']
Here is the commit info below:
commit 8774f15b9a
Author: Peter Barker <pbarker@barker.dropbear.id.au>
Date: Mon Apr 11 18:05:39 2016 +1000
Tools: add valgrind option to autotest.py
:040000 040000 a111bc5b18609503f3b1 M Tools
Each library in ardupilot is a folder with the source in. That avoids build
failures when there are files in libraries/. That fixes#4099 ("waf doesn't
seem to like TAGS files") and #4093 ("sim_vehicle.py /w waf /w eclipse project
broken on Cygwin").
That problem can be reproduced by just creating a file in libraries/ and trying
to build with waf.
There should be a better way to confirm a path is an ardupilot library. That
can be done later.
When trying to execute sim_vehicle.py this is the current output:
Warning: Could not find '../Tools/autotest/../../ArduCopter/../build/sitl/bin/arducopter-quad',
starting '/bin/bash' instead. Please check your profile settings.
Fix the path to the binary and use absolute paths to ease the debug.
Add noreturn attribute to field_not_found(). This function always call
abort() if the field was not found. The compiler may not know about it
and may think some variables are used uninitialized:
../../Tools/Replay/LR_MsgHandler.cpp: In member function ‘void LR_MsgHandler_GPA_Base::update_from_msg_gpa(uint8_t, uint8_t*)’:
../../Tools/Replay/LR_MsgHandler.cpp:212:24: error: ‘sacc’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
gps.setHIL_Accuracy(gps_offset, vdop*0.01f, hacc*0.01f, vacc*0.01f, sacc*0.01f, have_vertical_velocity, sample_ms);
^
Instead of just initializing the variable (which is pointless in this case),
teach the compiler that this function never returns so require_field() works
properly.
This is a hackish way to allow waf running on windows. In some
combinations of the tools the python interpreter seems to be crashing on
windows:
Found 180 MAVLink message types in 2 XML files
Generating C implementation in directory /tmp/ArduPlane.build/libraries/GCS_MAVLink/include/mavlink/v1.0/ardupilotmega
Generating C implementation in directory /tmp/ArduPlane.build/libraries/GCS_MAVLink/include/mavlink/v1.0/common
Copying fixed headers
last line in mavgen.py
Aborted (core dumped)
michael@WIN-3NBOUTHN4TN /cygdrive/c/Users/michael/Desktop/DIYDrones/ardupilot/ArduPlane
$ echo $?
134
Here we check the return code to be greater than 128 since that means the
interpreter received a signal. In this case we clear the return code.
sim_vehicle.py: choose default waf_target based on specified vehicle
sim_vehicle.py: add --no-rebuild-on-failure option
sim_vehicle.py: make help clearer by adding option groups
