* added general settings tests
* fixed pre-219 switchUnlockState (CPN)
* X7 radio settings tests
* updated X10 model
* patched TESTS_PATH to point directly at the complete tests path
* fixed horus general settings switch conversion
* cpn x10 general settings tests
* fixed horus cpn import
* use x10 model from otx file to ease editing with companion
* added make targets "gtests" and "tests"
gtests builds all the tests (radio & companion)
tests builds and executes all the tests (radio & companion)
* fixed companion tests cmake
* fixed gtests-radio target
* fixed commit tests
* fixed converting LS & RS on X10
* fixed converting global functions on X10
* fixed source conversion on X10
* fixed throttle source on X10
* fixed throttle source on radio and companion
* more checks and re-enabled Horus screenData copy
* more tests (X10)
* [X10] restore correct direction for RS & LS
* [horus] convert "Value" widget source
* fixed conversion of ModelHeader::bitmap length
* added missing X10 radio tests to companion tests
* fixed conversion of telemetry sources in logical switches
* [horus] fixed potential crash
* fixed "AND switch" conversion in logical switches
* Basics to allow Horus BT module on X9E
* wip
* Temporary hack to prevent X9E to enter reset loop
* The wait game starts :)
* Temporary compil fixes
* More sensibles defines
* [TravisCI] Set up for parallel Travis builds and use Qt5.7 (with option of selecting version). commit-tests.sh script can now optionally build one, all, or sets of targets.
* [TravisCI] Move "COMPANION" build to first slot and rename to "DEFAULT".
* [TravisCI] Fix "ALL" build flavor, remove TARANIS flavor, rename HORUS to X12, streamline some code, add comments, fix indents.
* [build] Add comprehensive firmware size reporting scripts for ARM MCUs.
* [build] Add firmware size reports to commit-tests.sh.
* [build] Set permissions on new scripts.
* [build] Add -Wno-error option to commit-tests.sh.
* [build] Revert firmware size report to using `arm-none-eabi-size` if bash not found (eg. on Windows).
* [git] Add .gitattributes to force line endings on UNIX shell scripts.
* [TravisCI] Build using avr-gcc 4.9.2 (up from 4.8.2). Relies on https://github.com/opentx/opentx/pull/4169 . Fixes all linker warnings, makes clean build.
* [build] Add ARM/AVR compiler version check/display and a workaround for avr-gcc linker warnings when using WARNINGS_AS_ERRORS=true.
* [TravisCI] Enable -Werror on all build targets (WARNINGS_AS_ERRORS=true).
* [build] CMake script updates:
Enable WARNINGS_AS_ERRORS option for all targets (not just firmware);
Consolidate some C/CXX flag settings for firmware target, avoid duplication/redundancy;
All warning flags are now passed to linker (GCC only);
Add a "build report" for each main target showing compiler, flags, and defines, with verbosity controlled by new VERBOSE_CMAKELISTS option;
Added properties to PCB, TRANSLATIONS, and a few other options (makes option selector lists in cmake-gui);
Add macro for conditionally adding C++11 flag;
Silence CMake >3.0 warnings about CMP0054 policy not being set;
Reverts AVR linker -Werror workaround;
Misc. cleanup/consolidation.
* [build] More CMakeLists updates:
Silence superfluous "CRT_SECURE" series MSVC warnings;
Formalize finding pthread lib/dll on MSVC builds and fix some issues with install target script;
Add property string lists to more options;
Quotes some paths with possible spaces, & other minor cosmetics.
* [build] firmware/bootloader: Remove remaining CMake default compiler flags and add FIRMWARE_C[XX}_FLAGS[_DEBUG] options for user to specify additional flags. Also move ASM language call to silence CMake warning w/MSVC.
* [build] Workaround bogus "uninitialized" warnings from AVR linker.
* [tests] Add custom printer for less verbose gtest output (use --verbose option to revert to gtest default).
* [build] Use less verbose CMake output for TravisCI builds. Formalize search for gtest code and enable gtests target on WIN32-GCC. Minor language fix.
* Various fixes for gtests that were shown when running tests in random sequence. Enabled random gtests shuffling in commit tests.
* 9X gtests fixes
* more gtests fixes
* even more gtests fixes
* Another round of gtests fixes
* Code deduplication with test fixtures
* More fixtures
