Use `-DSPACEMOUSE=YES` CMake option to activate on Horus color targets.
Co-authored-by: Niel Horne (elecpower), Raphael Coeffic (raphaelcoeffic) and Peter Feerick (pfeerick).
Lug BG/FG is never used. Only LUA. Removed extra
+ Change my oopsie (model_select.cpp) compiler warning that didn't show up until I started using docker
* Missing debug timer
* USB driver cleanup
* Re #4479: Joystick update called from mixer thread every 10ms
* Cosmetics
* Compilation fix
* Cosmetics
* Another compilation fix
* Flash savings: getFileExtension() de-templated, saves 160 bytes
* Flash savings: several variables moved to rodata, RGB macro optimization (speed and can now be used to initialize rodata)
Result for Horus:
text data bss dec
1261182 808603 42020 2111805 before
1267149 803335 42124 2112608 after (data size also counts .sram section)
5kB saved in .data section (RAM and FLASH)
* Removed compiler warnings
* RGB() and ARGB() gtests added
* ARGB() optimized
* USB fixes:
* USB initialized when USB cable is inserted and de-initialized when disconnected (this was the original way, which proved itself unreliable in the past. Now with the new USB library it appears to work without any problems. Used in this way result in safer operation - USB controller is disabled while unconnected)
Works without force device mod
* Serial: initialization of global variables
* Reduced RAM usage - data moved to FLASH
* Detailed USB interrupt counters (+ debug print fix)
* * Switches audio filenames were wrong [Horus]
* sdAvailableLogicalSwitchAudioFiles was too small (we now have 64 LSW), fixed with BitFiled class
* Enabled Travis Gtests on Horus
* SD card version mismatch trace
* Re #3815: Audio debug timers added
Based on projectkk2glider work, with a lot of bugs I am sure. Still not
tested, this is why I preferred working on my own branch. I push it now
so that Jenkins do the first tests and you can review if I am on the
right track.
I am really wondering if a merge with the Lua api would be possible /
nice, what do you think?
