Comprehension of Thread Scheduling for the C++ Programming Language

Writing parallel programs – whether it is a multithreaded or multi-processed – can often be a daunting task. Developers need to pay close attention to a lot of things related with threads and processes. It is possible that one of these things will be ignored against their will. If this happens the program easily can go into a faulty state. In many cases, this means all that is left is to terminate the program manually. There may also be cases where a poorly written parallel program, despite the calculations run on multiple cores or on multiple computers such as social network servers for example, the program itself proves to be slower than its sequential counterpart. This paper investigates to find out whether there is an “actual” parallelism between threads or processes or where there are cases when one of them has to wait for the others to finish their execution. This getting to the point that the program could have been solved with a much simpler sequential program. We propose an approach and its implementation that results in the comprehension of thread schedule. We deal with C++ threads therefore we do not mind the different hardware elements, the language constructs provide abstraction over CPUs and operating systems. This approach is based on static analysis and its implementation takes advantage of the Clang compiler infrastructure.