Abstract
Chapter 7, p. 313 on threads and locks ended with the readers and writer problem, which is an important concurrent problem. However, it is clear that the solutions to the readers and writer problem were becoming complex; too complex to feel comfortable about the correctness of the solution without extensive analysis. While coding conventions like split-binary semaphores and baton passing give a formal design approach, the resulting programs are still complex, both to manage and maintain; basically, relying on programmers following coding conventions to assure correctness is a poor approach. As well, there is the subtle problem of properly handling staleness and writing because of the “window” for interrupt ion between releasing the entry semaphore and blocking. In essence, we have arrived at the same impasse that occurred with software solutions for mutual exclusion, that is, the complexity and inefficiency of the locking solution is increasing for more complex critical-section. Therefore, it is worthwhile to search for a different approach to reduce the complexity, as for hardware solutions for mutual exclusion. However, in this case, the programming language, not the hardware, provides the mechanism to simplify the solution.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Aliasing problems may preclude complete checking for nested deadlock situations.
- 2.
No shared variables were supported; hence, the region statement had only a conditional expression so only one critical region could be active in a program.
- 3.
Simula67 [9] was the first object-oriented programming language.
- 4.
In fact, a counter is unnecessary because the state of the buffer, i.e., full or empty, can be determined by appropriate comparisons between front and back. However, these comparisons can be subtle, whereas the counter is straightforward.
- 5.
This section is a revision of “Monitor Classification” in [8] ©1995 ACM. Portions reprinted by permission.
- 6.
For all tables of this form, the queue head is on the left. Also, for a line with a “wait” comment, the wait operation has already been performed by the task in the monitor on the previous line, and the “wait” line shows that task already blocked on the specified condition variable.
- 7.
Again, no dynamic allocation is performed by this algorithm, as the list node is created as a local variable on the task’s stack, which eliminates the need for dynamic storage-allocation.
- 8.
Modula-3 takes this one step further by defining the signal routine to wake up at least one task, which implies that it may wake up more than one task. This semantics is modelled in the taxonomy as a no-priority non-blocking monitor with a loop around each signal that executes a random number of times.
References
Andrews, G.R.: Concurrent Programming: Principles and Practice. Benjamin/Cummings Publishing, Redwood City (1991)
Birrell, A., Brown, M.R., Cardelli, L., Donahue, J., Glassman, L., Gutag, J., Harning, J., Kalsow, B., Levin, R., Nelson, G.: Systems Programming with Modula-3. Prentice Hall Series in Innovative Technology. Prentice-Hall, Englewood Cliffs (1991)
Brinch Hansen, P.: Structured multiprogramming. Commun. ACM 15(7), 574–578 (1972)
Brinch Hansen, P.: Concurrent programming concepts. Softw. Pract. Exp. 5(4), 223–245 (1973)
Brinch Hansen, P.: The programming language concurrent pascal. IEEE Trans. Softw. Eng. 2, 199–206 (1975)
Brinch Hansen, P.: The design of Edison. Softw. Pract. Exp. 11(4), 363–396 (1981)
Buhr, P.A.: Are safe concurrency libraries possible? Commun. ACM 38(2), 117–120 (1995)
Buhr, P.A., Fortier, M., Coffin, M.H.: Monitor classification. ACM Comput. Surv. 27(1), 63–107 (1995)
Dahl, O.J., Myhrhaug, B., Nygaard, K.: Simula67 Common Base Language. Norwegian Computing Center, Oslo Norway (1970)
Dijkstra, E.W.: Hierarchical ordering of sequential processes. Acta Inf. 1, 115–138 (1971)
Hoare, C.A.R.: Towards a theory of parallel programming. In: Hoare, C.A.R., Perott, R.H. (eds.) Operating Systems Techniques, pp. 61–71. Academic Press, New York (1972)
Hoare, C.A.R.: Monitors: An operating system structuring concept. Commun. ACM 17(10), 549–557 (1974)
Howard, J.H.: Signaling in monitors. In: Proceedings Second International Conference Software Engineering, pp. 47–52. IEEE Computer Society, San Francisco, U.S.A (1976)
Kessels, J.L.W.: An alternative to event queues for synchronization in monitors. Commun. ACM 20(7), 500–503 (1977)
Lampson, B.W., Redell, D.D.: Experience with processes and monitors in mesa. Commun. ACM 23(2), 105–117 (1980)
Lea, D.: Concurrent Programming in Java: Design Principles and Patterns, 1st edn. Addison-Wesley, Boston (1997)
Mitchell, J.G., Maybury, W., Sweet, R.: Mesa language manual. Tech. Rep. CSL–79–3, Xerox Palo Alto Research Center (1979)
Trono, J.A.: A new exercise in concurrency. SIGCSE Bull. 26(3), 8–10 (1994)
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing Switzerland
About this chapter
Cite this chapter
Buhr, P.A. (2016). High-Level Concurrency Constructs. In: Understanding Control Flow. Springer, Cham. https://doi.org/10.1007/978-3-319-25703-7_9
Download citation
DOI: https://doi.org/10.1007/978-3-319-25703-7_9
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-25701-3
Online ISBN: 978-3-319-25703-7
eBook Packages: Computer ScienceComputer Science (R0)