The Art of Multiprocessor Programming
The Art of Multiprocessor Programming
Author : Maurice Herlihy, Nir Shavit
Publisher : Morgan Kaufmann
Page : 529
Overview
The first third covers the principles of concurrent programming, showing how to think like a concurrent programmer. Like many other skills such as driving a car, cooking a meal, or appreciating caviar, thinking concurrently requires cultivation, but it can be learned with moderate effort. Readers who want to start programming right away may skip most of this section, but should still read Chapters 2 and 3 which cover the basic ideas necessary to understand the rest of the book.
We first look at the classic mutual exclusion problem (Chapter 2). This chapter is essential for understanding why concurrent programming is a challenge. It covers basic concepts such as fairness and deadlock. We then ask what it means for a concurrent program to be correct (Chapter 3). We consider several alternative conditions, and the circumstances one might want to use each one. We examine the properties of shared memory essential to concurrent computation (Chapter 4), and we look at the kinds of synchronization primitives needed to implement highly concurrent data structures (Chapters 5 and 6).
Download :
ebook on rapidshare
