This is a breaking change to the WebSockets layer to align it with
recent specification documentation work.
To support this, HAL SimValue changed readonly to a direction enum.
This allows specifying bidirectional in addition to input and output.
The SimValue change is specifically designed to avoid API and ABI breakage.
This is completely transparent in C++; in Java a new callback class was added,
and the old readonly functions have been marked deprecated.
A new SimValue creation function for enums allows specifying double values
for each enum value, not just strings. This allows mapping enum values to
doubles in the WebSockets layer.
A ":" in the SimDevice name now maps it to different WebSocket types (e.g.
"Accel:Name" becomes type "Accel", device "Name"). The type is hidden
in the GUI.
Other WebSockets changes:
* Implemented match_time and game_data
* Added joystick rumble data
* Added builtin accelerometer support
* SimValue enums are mapped to string and double value on WS interface
* Added WebSockets protocol specification
* Added READMEs
Based on run of include-what-you-use.org to identify unused include files in various .h and .cpp files.
The changes mostly fall into 3 categories:
- Actually unused includes - copy-paste errors, not removing includes after cleaning up code, etc
- A too-broad include used where a more specific (and hopefully smaller) header will do
- Interface .h files including headers only needed by the .cpp implementation - moving from .h to .cpp
will mean that code which uses the .h doesn't pay the price of processing the header file they don't need
Old behavior is available via StepTimingAsync.
This makes it significantly easier to use simulation timing with notifiers.
Also update tests to use simulation framework. This also speeds up the
timing-dependent tests by using simulation timing. ResourceLock is used
in the Java tests to prevent parallel execution.
While we're here, tweak HAL Notifier implementation:
- Use wait_for instead of wait_until in WaitForNotifierAlarm
- Check for triggerTime = UINT64_MAX in UpdateNotifierAlarm
This makes it much more user-friendly to use simulation classes without needing
to ifdef for C++ to avoid linker errors or be very careful about construction
to avoid runtime errors in Java.
This allows high-level library classes to implement enhanced simulation
support even if no low-level corresponding simulation library exists, and
avoids the need for bit-banging complex interfaces like SPI or CAN.
Default behavior is still to run the robot main loop in the main thread.
The ability to run the robot main loop in a separate thread and add a hook
for running a different function in the main thread is needed for simulation
GUI support on some platforms.
std::scoped_lock was introduced in C++17 and is strictly better than
std::lock_guard as it supports locking any number of mutexes safely.
It's also easier to use than std::lock for locking multiple mutexes at
once.
The 2019 FPGA image switched the output of auto SPI from plain bytes to a
sequence of 32-bit words (timestamp, then words with the byte values in the
least significant byte of each word).
In addition to changing the HAL and simulators to reflect this, add piecewise
integration support to wpilibc/wpilibj SPI to take advantage of the timestamps
and use it in the ADXRS450 gyro.
HAL_ReadInterruptRisingTimestamp and HAL_ReadInterruptFallingTimestamp
return time as a double. Instead, keep the raw integer count and move the
double conversion into the C++ and Java code. This enables comparison of the
time with other timers.
The old headers were moved into folders because doing so avoids polluting
the system include directories.
Folder names were also normalized to lowercase.
Also switch eventName and gameSpecificData to fixed 64-byte arrays to avoid mallocs and
extra NetComm calls. This behavior matches 2018 LabView.
The DS caching is kept in Java to avoid JNI and/or massive amounts of allocations.
This does not increase latency because Java still only hits NetComm once.
Moving the DS caching benefits all languages other than Java, because it avoids the need
for individual implementations. If caching is ever added to NetComm, it will then only be
necessary to remove it from the HAL and Java rather than all languages.
During shared library loading, a different libLLVM can be pulled in, causing
llvm symbols from dependent libraries to resolve to that library instead of
this one. This has been seen in the wild with the Mesa OpenGL implementation
in JavaFX applications (see wpilibsuite/shuffleboard#361).
This is clearly a very breaking change. For some level of backwards
compatibility, a namespace alias from llvm to wpi is performed in the "llvm"
headers. Unfortunately, forward declarations of llvm classes will still break,
but compilers seem to generate clear error messages in those cases
("namespace alias 'llvm' not allowed here, assuming 'wpi'").
This change also moves all the wpiutil headers to a single "wpi" subdirectory
from the previously split "llvm", "support", "tcpsockets", and "udpsockets".
Shim headers will be added for backwards compatibility in a later commit.
