- Uses 32 version of Cygwin (on a 64 bit machine)
- Adding PYTHON_EXECUTABLE override variable for Cygwin builds
* We have to use the system CMake instead of the Cygwin CMake which detects the wrong version of Python
- Update Bootloader build infrastructure to attempt tests on Windows
- Fixing CMake to always force python3 binary (this should always work)
- Adding macOS Travis build (clang and gcc)
- Disabling all clang builds until macOS (lack of packages) and Linux (broken Travis repos) are fixed
- Recently added the BCD_VERSION which uses git rev-list --count HEAD command
- Unfortunately git 1.7.1 doesn't have this, so I've added a fallback to manually counting the commits using some CMake magic
- CMake 3.6 finally adds proper cross-compiler detection
- When detected, use full compiler detection infrastructure
- Still maintains compatibility back to CMake 2.8 (at least for now)
- With new detection, selecting only the C compiler as it speeds up the CMake generaion process
- Added a GCC 4.9 minimum version requirement for Bootloader
- Add _app_rom_end address location that points to the end of flash
- Check when flashing each page whether the end of the page exceeds the end of the flash
- Use commit number as the basis for bcdDevice field (maximum is 2^16)
- Change iSerial from commit date to chip Unique ID
- Update both Firmware and Bootloader to use the same scheme (bcdDevice and iSerial)
- Cleanup of Bootloader code
- Split out build info CMake commands to a new file
- Added clang (i.e. multi-compiler) support to convenience build scripts
- Updated README
- Added Bootloader build scripts for the two main versions currently used
- Added uartOut and usbMuxUart build scripts (these tend to suffer from build rot the most as they're only used in debugging)
- Attempt to get clang support for the Bootloader
* clang is missing compiler extensions, so this may require a large re-write to get working
- Changing default name from defaultMap.kll to scancode_map.kll
- Old name of defaultMap.kll will still work (and is the final fallback)
- Updated all the main projects to use the new name
- Updated BETKB (not entirely finished, but mostly complete)
- Added sub-use of interconnect scancode cache for converters
- Added infinity_led.bash
No behaviour changes, just restructuring.
- clang.c includes necessary functions to make clang compiler work (tested on teensy 3.1)
- Added support code to generate a compile_commands.json for clang-tidy
* Updates the symlink whenever cmake or make is called (Unix OSs only)
- Rx buffers weren't fast enough, had to use DMA :D
- Basic LCD remote capabilities are working, single node
- Multi-node broadcast seems to have a bug still
- DMA ring buffer allowed for significant processing simplification
* There is an overrun risk, but the buffer is large and generally there isn't too much data being sent (just very quickly)
- Split out LCD layer stack capability into itself and an "exact" version used for updating remote nodes
- Media keys tested working on Linux/Windows/Mac (use Consumer control)
- Fixed enumeration delays
- Fixed virtual serial port configuration issues
- Fixed GET_REPORT and SET_REPORT
- Added intial descriptors and endpoints for Mouse and Joystick devices
- Split out the consumer and system control endpoint
- Added more fault debugging messages
- Added interface names to endpoints (visible in Windows Device Manager)
- Added KLL define for keyboard locale
- Supports up to 255 slave nodes (you'll run into ScanCode limitations before then)
- Requires most recent kll compiler update
- Additional debugging output and stats counters
- Noise and parity checking
- Fixed TxFIFO issue when sending buffers larger than the FIFO
- Cleaned up defaultMap.kll
- Added ScanCode caching (reduces interconnect traffic significantly)
- Interconnect module code is conditionally compiled into PartialMap module if required
* Keyboard interconnect using two uarts
* Supports daisy chain addressing using a master/slave direction communication scheme
- Still needs more testing
- Functions have been tested alone to be generally working
Use a consistent standard - Tabs in front for indenting, spaces after for anything else. This way everything stays nice and lined up while also letting users change there prefered indent level. Most of the new files from Haata where already in this format.
Fixes#22
Each (dfu) interface may have more than one a ltsetting each with their own index and name. According to the DFU_1.1 pdf section 4.2.3, "* Alternate settings can be used by an application to access additional memory segments. In this case,
it is suggested that each alternate setting employ a string descriptor to indicate the target memory
segment; e.g., 'EEPROM'." Whether or not we end up using multiple memory segments it is still good to have a descriptive name incase there are other dfu devices connected.
Edit: Fixed previous indentation
Can be specified from any .kll file
DebounceDivThreshold = 0xFFFFFFFF
Sets to max debounce, default is 0xFFFF.
The 0xFFFFFFFF is about a 2x longer debounce
The max requires more ram (as it uses 32 bit variables instead of 16).
Added support for submodule capabilities files.
- When loading a garbage image onto the device, the chip hangs
and refuses to jump to the bootloader and start dfu
Fix
- Check if chip was in locked state and jump to dfu
- Check if watchdog timer fired and jump to dfu
