Subject

The proposed semaphore implementation is the class Semaphore defined in the files semaphore.hpp and semaphore.cpp.

Features

• Fairness is guaranteed by the ticket lock algorithm (i.e., FIFO order). It prevents a thread from being starved out of execution for a long time due to inability to pass through a semaphore in favor of other threads.

• A queue of condition variables is used for the sake of waking up one certain thread using notify_one() instead of waking up all the threads using notify_all() in case of one condition variable. Benchmarks registered performance boost against the implementation based on one condition variable. Thus it looks like using an extra queue is reasonable.

• The number of waiting threads is implicitly tracked as the size of the queue of condition variables.

Requirements

• CMake 3.14+

• A C++14 compliant compiler

Building and Testing

Use CMakeLists.txt. You may refer to .github/workflows/cmake-multi-platform.yml as an example.