Concurrent Programming

Abstract

This chapter covers concurrent programming. It introduces the concept of parallel computing and points out its importance. It compares sequential algorithms with parallel algorithms, and parallelism vs. concurrency. It explains the principles of threads and their advantages over processes. It covers threads operations in Pthreads by examples. These include threads management functions, threads synchronization tools of mutex, join, condition variables and barriers. It demonstrates concurrent programming using threads by detailed examples. These include matrix computation, quicksort and solving systems of linear equations by concurrent threads. It explains the deadlock problem and shows how to prevent deadlocks in concurrent programs. It covers semaphores and demonstrates their advantages over condition variables. It also explains the unique way of supporting threads in Linux. The programming project is to implement user-level threads. It presents a base system to help the reader get started. The base system supports dynamic creation, execution and termination of concurrent tasks, which are equivalent to threads executing in the same address space of a process. The project is for the reader to implement threads join, mutex and semaphores for threads synchronization and demonstrate their usage in concurrent programs. The programming project should allow the reader to have a deeper understanding of the principles and techniques of multitasking, threads synchronization and concurrent programming.