unittest.txt

@Igorshp cool!

Pretty much all compilation errors resolved.

having trouble with sensor_gyro_unittest.cc, it complains about missing arm_math.h, I belive it could be because i'm compiling it on x86 architecture. Not sure.

Some tests are also failing, expected as the code has diverged since.

@Igorshp the file is at ./lib/main/CMSIS/DSP/Include/arm_math.h but i think it has sth to do with the target supporting SMT or not.

@Igorshp the file is at ./lib/main/CMSIS/DSP/Include/arm_math.h but i think it has sth to do with the target supporting SMT or not.

hmm, intresting point. Thanks. I'll check it can be toggled with flags

@Igorshp , PR435_da337b270_unittests.txt

Status update

All but one tests compile (sensor_gyro_unittest requires arm_math.h library and I haven't yet figured out how to provide it)

3 tests compile but are failing since code base have moved on.

The github action step fails to run the tests. Looks like environment setup issue.

@Igorshp thanks!

The github action step fails to run the tests. Looks like environment setup issue.

it's a compiler issue - the action will be fine. the current state of the affair is like this:

• the tests need to be compiled with clang-10 at least
• or the arm sdk like for the cross compiles but the component we use is not updated yet Added arm-gcc-10-2020-q4 revision fiam/arm-none-eabi-gcc#12
• the betaflight folks have worked on this but the discussion is horrifying
• i'm working on getting your changes and the above merged but not done yet

we should just wait a couple of days for the upstreams to settle. 🍵

ok got the test to run to completion. credits to @mikeller 😗

@Igorshp now, i can either PR this up as new or you rebase my branch on your open PR.. i'm fine both ways.

@Quick-Flash the tests show some interesting results including the pid controller 😼

@gretel interesting, thanks for the insight!

I've had a fair share of issues with the compiler myself and considering building a docker container to run the tests in. It'll help normalize the testing environment for everyone and from what I understand github actions can use containers.
any thoughts?

Then there's matter of broken tests. crsf ones dont look to bad. pid controller stuff though is a bit out of my league at the moment.

I've had a fair share of issues with the compiler myself and considering building a docker container to run the tests in. It'll help normalize the testing environment for everyone and from what I understand github actions can use containers.
any thoughts?

no more docker please 🙈 i think it's working fine now.

Then there's matter of broken tests. crsf ones dont look to bad. pid controller stuff though is a bit out of my league at the moment.

there is also the matter of broken implemenations 👯

Merged in your changes, let's wait for the build

EDIT: hmm, couple of other commits got pulled in. I'll rebase on master when we're done

i think the Take the trash out out step is also "required" for the test run.

raw logs:
logs_1229.zip
test: Error: Process completed with exit code 2.

Tried getting this to work with clang-11 and/or clang-12 for modern OS.
easy to change src/test/Makefile, but DWARF/ld errors occur throughout.

[  PASSED  ] 8 tests.
running test_alignsensor_unittest: PASS
linking ../../obj/test/arming_prevention_unittest/arming_prevention_unittest 
/usr/bin/ld: /usr/bin/ld: DWARF error: could not find variable specification at offset 38fb
/usr/bin/ld: DWARF error: could not find variable specification at offset 5979
/usr/bin/ld: DWARF error: could not find variable specification at offset 5a1e
/usr/bin/ld: DWARF error: could not find variable specification at offset 5ac4
/usr/bin/ld: DWARF error: could not find variable specification at offset 5b6f
/usr/bin/ld: DWARF error: could not find variable specification at offset 5c1a
/usr/bin/ld: DWARF error: could not find variable specification at offset 5cc5
/usr/bin/ld: DWARF error: could not find variable specification at offset 5d70
/usr/bin/ld: DWARF error: could not find variable specification at offset 5e1b
../../obj/test/arming_prevention_unittest/arming_prevention_unittest.o:./src/test/../main/flight/servos.h:44: multiple definition of `inputSource_e'; ../../obj/test/arming_prevention_unittest/fc/fc_core.c.o:./src/test/../main/flight/servos.h:44: first defined here
/usr/bin/ld: ../../obj/test/arming_prevention_unittest/arming_prevention_unittest.o:./src/test/unit/arming_prevention_unittest.cc:58: multiple definition of `rcCommand'; ../../obj/test/arming_prevention_unittest/fc/rc_controls.c.o:./src/test/../main/fc/rc_controls.c:73: first defined here
clang: error: linker command failed with exit code 1 (use -v to see invocation)
make[1]: *** [Makefile:664: ../../obj/test/arming_prevention_unittest/arming_prevention_unittest] Error 1
make[1]: Leaving directory './src/test'
make: *** [Makefile:540: test] Error 2

if this can be fixed, then master merged in, it will need an #ifdef for SmithPredictor

--- a/src/main/sensors/gyro.h
+++ b/src/main/sensors/gyro.h
@@ -195,7 +195,9 @@ bool gyroOverflowDetected(void);
 bool gyroYawSpinDetected(void);
 uint16_t gyroAbsRateDps(int axis);
 uint8_t gyroReadRegister(uint8_t whichSensor, uint8_t reg);
+#ifdef USE_SMITH_PREDICTOR
 float applySmithPredictor(smithPredictor_t *smithPredictor, float gyroFiltered);
+#endif