The Art of Multiprocessor Programming Tel （多处理器编程的艺术电话）.pdf

Errata for The Art of Multiprocessor Programming Version of 10 February 2009 In many places, inserted text is highlighted in red. Preface p. xx “all of which are useful in structuring concurrent applications.” Chapter 1 p. 4 In Fig. 1.1, line 4 should say for (int j = (i * block) + 1; j = (i + 1)* block; j++) { p. 11 Second paragraph of Mutual Exclusion bullet: “Initially the can is either up or down. Let us say it was down. Then only the pets can go in, and mutual exclusion holds.” p. 18 Replace question 6 bullet 2 with:  Suppose the method M accounts for 30% of the programs computation time. Let sn be the programs speedup on n processes, assuming the rest of the program is perfectly parallelizable. Your boss tells you to double this speedup: the revised program should have speedup s ’ ≥ s /2. You n n advertize for a programmer to replace M with an improved version, k times faster. What value of k should you require? Chapter 2 p. 23 Figure 2.3 should be amended as shown: public long getAndIncrement() { lock.lock(); try { long temp = value; value = temp + 1; return temp; } finally { lock.unlock(); } } p. 25 Pragma 2.3.1: “We explain the reasons in Chapter 3 and Appendix B.” p. 26 In Fig. 2.5, remove all declarations of variables as volatile. Indeed as stated in Pragma 2.3.1 on page 25, one should use memory barriers when implementing these algorithms. However, in this chapter (in all code) we avoid such issues to keep