H. Kopetz and J. Reisinge. The non-blocking write protocol NBW: A solution to a realtime synchronisation problem. In Susan Davidson and Insup Lee, editors, Proceedings of the Real-Time Systems Symposium, pages 131-137, Raleigh-Durham, NC, December 1993. IEEE Computer Society Press.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... scheduling analysis since tasks can be delayed because they were either preempted by other more urgent tasks, or because they are blocked before a critical section by another process that can in turn be preempted by another more urgent task and so on, this is also called as the convoy effect) KopR93] and iii) leads to priority inversion in which a high priority task can be blocked for an unbounded time by a lower priority task [ShaR90] Several synchronisation protocols have been introduced to solve the priority inversion problem for uniprocessor [ShaR90] and multiprocessor [Raj90] systems. ....

H. Kopetz and J. Reisinger. The Non-Blocking Write Protocol NBW: A Solution to a Real-Time Synchronization Problem. In Proc. of the lJth Real-Time Systems Syrup., pages 131-137, 1993.

....situations, but they do replace it for most inter task communication requirements in embedded applications. 7. 3 Previous Work State messages were first used in the MARS OS [22] and have also been implemented in ERCOS [40] The state message implementation used in these systems as described in [41] is as follows. The problem with using global variables for passing messages is that a reader may read a half written message since there is no synchronization between readers and writers. This problem is solved by using an N deep circular buffer for each state message. An associated pointer is ....

....and used by readers to retrieve the latest message. With a deep enough buffer, the scheme can guarantee that data will not be corrupted while it is being read by a reader, but a large N can make state messages infeasible for our limited memory target applications. 26 The solution presented in [41] limits N by having readers repeat the read operation until they get uncorrupted data. This saves memory at the cost of increasing the read time by as much as several hundred microseconds, even under the assumption that writers and readers run on separate processors with shared memory. With such ....

H. Kopetz and J. Reisinger, "The non-blocking write protocol NBW: a solution to a real-time synchronization problem," in Proc. Real-Time Systems Symposium, pp. 131--137, 1993.

....algorithms [9, 10] designed to ensure that reader tasks will always access uncorrupted messages. As mentioned earlier, synchronization, particularly with locks, can cause many problems of its own. Therefore, in this paper, we focus on wait free, singlewriter, multiple reader IPC algorithms [7, 8, 17, 24, 31]. However, these algorithms have higher space overheads than the synchronization based algorithms. Even though the worst case time overhead of these algorithms is significantly lower than that of the synchronization based ones, the execution overheads can still be significant. Later in this paper, ....

....the task with largest RMax . Finally, we keep a pointer to the most recently written message. This is updated by the writer, and subsequently used by the readers to retrieve the latest message. This concept was first introduced in [16] and later implemented in the Non Blocking Write (NBW) protocol [17]. This algorithm is very efficient in terms of execution time, i.e. almost as fast as using global variables with no protection. The only overhead associated with this algorithm is the cost of maintaining the pointer for the most recently written message. Therefore, it is easy to see that it has ....

[Article contains additional citation context not shown here]

H. Kopetz and J. Reisinger. The non-blocking write protocol NBW: A solution to a real-time synchronisation problem. In IEEE Real-Time Systems Symposium, pages 131--137, 1993.

....corresponds to ordering failure semantics as defined in section 3.2. One way to guarantee the fault hypothesis in a shared memory system is to make use of hardware memory protection schemes, or during design eliminate or minimize shared resources using wait free and lock free communication [108][52][13] 6.2 Execution order analysis In this section we present a method for identifying all the possible orders of execution for sets of jobs conforming to the task model introduced in section 6.1. We will also show how the model and analysis can be extended to accommodate for interrupt ....

....we make use of overwriting semantics (statebased communication) using just a few buffers, or we record all data that has been sent by the higher frequency task so that it can be consumed by the lower frequency task when it is activated. There are several approaches to solving this problem [26][52][13] 10.2.3 Component contract analysis We are now going to introduce a framework for formal and probabilistic arguments of component reuse in safety critical real time systems. The basic idea is to provide evidence, based on the components contracts and the experience accumulated, that a ....

Kopetz H. and Reisinger J. The Non-Blocking Write Protocol NBW: A Solution to a RealTime Synchronization Problem. In Proceedings of he 14th Real-Time Systems Symposium, pp. 131-137, 1993.

.... where the work of Lamport is particularly significant, not only in the development of protocols, but in defining the problem and introducing methods for its analysis [Lam86] The Time Triggered Architecture (TTA) TTT01, KG94] uses the Nonblocking Write (NBW) protocol of Kopetz and Reisinger [KR93] Anderson [And01] observes that these authors were apparently unaware that Lamport had invented this algorithm 16 years earlier (referring to the algorithm of [Lam77] In fact, there are two parts to NBW: the first part is lock free (and is the same as Lamport s algorithm) but it is not ....

Hermann Kopetz and Johannes Reisinger. The non-blocking write protocol NBW: A solution to a real-time synchronization problem. In Real Time Systems Symposium, pages 131--137, Raleigh-Durham, NC, December 1993. IEEE Computer Society.

....errors, i.e. jobs can only interfere via specified interactions. One way to guarantee the fault hypothesis in a shared memory system is to make use of hardware memory protection schemes, or to during design eliminate or minimize shared resources using waitfree and lock free communication [23][11][3] 3 EXECUTION ORDER ANALYSIS In this section we present a method for identifying all the possible orders of execution for sets of jobs conforming to the task model introduced in section 2. We will also show how the model and analysis can be extended to accommodate for interrupt interference ....

Kopetz H. and Reisinger J. The Non-Blocking Write Protocol NBW: A Solution to a Real- Time Synchronization Problem. In Proceedings of he 14th Real-Time Systems Symposium, pp. 131-137, 1993.

....developed by Clark [10] using Petri nets) by Rushby [59] using model checking) and by Henderson and Paynter [18] using PVS) Hesselink [21] have verified some atomic register constructions from the computer science literature using ACL2. TTA uses a protocol called NBW (nonblocking write) [29] whose wait free element was inspired by Simpson s algorithm, and whose lock free construction is that of Lamport [33] It would be useful to undertake a formal examination of NBW (which is used in the Communication Network Interface (CNI) that provides communication between hosts and their TTA ....

Hermann Kopetz and Johannes Reisinger. The non-blocking write protocol NBW: A solution to a real-time synchronization problem. In Real Time Systems Symposium, pages 131--137, IEEE Computer Society, Raleigh-Durham, NC, December 1993. 15

.... for example, provides a completely nonblocking, asynchronous communication mechanism that nonetheless ensures timely transmission and mutual exclusion (i.e. no simultaneous reading and writing of the same buffer) A generalization of this protocol, called NBW (nonblocking write) is used in TTA [KR93]. There is a rich opportunity to codify and analyze the principles and requirements that underlie algorithms such as these. Codification would be undertaken in the context of the assume guarantee approach to modular certification outlined above. That approach itself requires further elaboration ....

Hermann Kopetz and Johannes Reisinger. The non-blocking write protocol NBW: A solution to a real-time synchronization problem. In Real Time Systems Symposium, pages 131--137, Raleigh-Durham, NC, December 1993. IEEE Computer Society.

.... scheduling analysis since tasks can be delayed because they were either preempted by other more urgent tasks, or because they are blocked before a critical section by another process that can in turn be preempted by another more urgent task and so on, this is also called as the convoy effect) KopR93] and iii) leads to priority inversion in which a high priority task can be blocked for an unbounded time by a lower priority task [ShaR90] Several synchronisation protocols have been introduced to solve the priority inversion problem for uniprocessor [ShaR90] and multiprocessor [Raj90] systems. ....

H. Kopetz and J. Reisinger. The Non-Blocking Write Protocol NBW: A Solution to a Real-Time Synchronization Problem. In Proc. of the 1Jth Real-Time Systems Symp., pages 131-137, 1993.

....# Wait and Lock free communication. The kernel supports a communication type, using buffers, that decreases the need for explicit synchronization using e.g. semaphores. As a consequence blocking times can be reduced as well as increasing the analyzability of the entire real time system [3][6]. # Execution time jitter reduction. The kernel provides a mechanism for minimizing the execution time jitter of individual tasks as well as the jitter originating from the kernel it self. This has been shown to increase the testability of the application enormously, since all executions of the ....

....task available in ready and Executing. 5 When a periodically executing task has terminated 6 When the executing task is preempted. 7 When an executing task is suspended until a signal occurred 8 When an aperiodic task is triggered by a signal. asynchronous multi rate tasks occur [3][6]. In addition WLFC gives decreased blocking times and reduced scheduling complexity. All data transfer is also performed by the tasks themselves, not by the kernel, which decreases kernel overhead and jitter. The data transfer overhead is debited to the tasks involved in the transaction, and ....

Kopetz H. and Reisinger J. The Non-Blocking Write Protocol NBW: A Solution to a Real- Time Synchronization Problem. In Proceedings of he 14th Real-Time Systems Symposium, pp. 131137, 1993.

....only regular registers, but the editor was afraid that the concept of nonatomic operations on individual digits might be considered heretical and insisted that it be removed from the paper. In 1993, Kopetz and Reisinger applied Lamport s lockfree bu er algorithm in a real time control system [20]. They were apparently unaware that Lamport had invented this algorithm 16 years earlier, because they did not credit him for it. However, their main focus was not the algorithm itself, but scheduling analysis techniques for accounting for synchronization overheads when the algorithm is used. 4.5 ....

H. Kopetz and J. Reisinger. The non-blocking write protocol nbw: A solution to a real-time synchronization problem. In Proceedings of the 14 th IEEE Symposium on Real-Time Systems, pp. 131-137. December 1993.

....accompanied by a simple, concise proof, would be relevant for practical systems, and also serve as an interesting case study of the methods developed here. Operating systems, especially distributed operating systems, provide another rich domain for problems and protocols, such as synchronization [KR93] and scheduling [Jef92, Zho92] with hard real time constraints, that might be analyzed using this methodology. For these, and other problems, it is important to characterize not only correctness but also timeliness. Perhaps the most useful application of these techniques lies in the analysis of ....

Hermann Kopetz and Johannes Reisinge. The non-blocking write protocol nbw: A solution to a real-time synchronization problem. In Proceedings of the Fourteenth Real-Time Systems Symposium, pages 131--137, Raleigh-Durham, NC, December 1993. IEEE Computer Society Press.

....7. 3 Previous work on state messages Theoretical work on data sharing without synchronization was first presented by Lamport [16] State messages were first used in the MARS OS [14] and have also been implemented in ERCOS [24] The state message implementation used in these systems as described in [15] is as follows. The problem with using global variables for passing messages is that a reader may read a half written message since there is no synchronization between readers and writers. This problemissolvedbyusinganN deep circular buffer for each state message. An associated pointer is used ....

....and used by readers to retrieve the latest message. With a deep enough buffer, the scheme can guarantee that data will not be corrupted while it is being read by a reader, but a large N can make state messages infeasible for our limited memory target applications. The solution presented in [15] limits N by having readers repeat the read operation until they get uncorrupted data. This saves memory at the cost of increasing the read time by as much as several hundred microseconds, even under the assumption that writers and readers run on separate processors with shared memory. With such ....

KOPETZ,H .,AND REISINGER, J. The non-blocking write protocol NBW: a solution to a real-time synchronization problem. In Proc. Real-Time Systems Symposium (1993), pp. 131--137.

....from many of the same shortcomings as conventional lock based approaches. Unfortunately, the thread of researchonwait free and lock free communication begun by Sorenson and Hamacher was lost in the real time systems community for manyyears. Recently,however, this thread of research resurfaced in #Kopetz and Reisinger 1993# and in #Johnson and Harathi 1994#. In the former paper, a simple lock free, one writer, read#write bu#er is presented, and scheduling conditions are given for tasks sharing the bu#er. In the latter paper, the primary focus is implementations of lock free algorithms rather than scheduling. Our ....

Kopetz, H. and Reisinger, J. 1993. The non-blocking write protocol NBW: A solution to a real-time synchronization problem. In Proceedings of the 14 th IEEE Symposium on Real-Time Systems #December 1993#, pp. 131#137. IEEE.

....7. 3 Previous work on state messages Theoretical work on data sharing without synchronization was first presented by Lamport [16] State messages were first used in the MARS OS [14] and have also been implemented in ERCOS [24] The state message implementation used in these systems as described in [15] is as follows. The problem with using global variables for passing messages is that a reader may read a half written message since there is no synchronization between readers and writers. This problem is solved by using an N deep circular buffer for each state message. An associated pointer is ....

....and used by readers to retrieve the latest message. With a deep enough buffer, the scheme can guarantee that data will not be corrupted while it is being read by a reader, but a large N can make state messages infeasible for our limited memory target applications. The solution presented in [15] limits N by having readers repeat the read operation until they get uncorrupted data. This saves memory at the cost of increasing the read time by as much as several hundred microseconds, even under the assumption that writers and readers run on separate processors with shared memory. With such ....

KOPETZ, H., AND REISINGER, J. The non-blocking write protocol NBW: a solution to a real-time synchronization problem. In Proc. Real-Time Systems Symposium (1993), pp. 131--137.

....is reduced especially within multiprocessor environments. As discussed in the literature, there are other negative effects of employing lock based mutual exclusion [23, 18, 39] An alternative solution is to avoid mutual exclusion, allowing concurrent reading while shared data is being updated [21, 23, 31]. Such algorithms are referred to as being lock free or non blocking because shared data may be accessed, without being locked, by the reader and the writer at any time. Nevertheless, the consistency as well as the coherence of data values must be guaranteed. This is usually achieved by ....

....last reading, it must repeat the read operation. Repeated read and check operations might be necessary before a successful reading is completed thereby increasing the response time of the reader. This timing overheads can be reduced or eliminated by using multiple buffers to hold the shared data [31, 21, 34]. A trade off between time and space overheads, to satisfy application requirements, is thus necessary. In real time systems, using a lock based synchronization protocol is the traditional approach to implementing data objects shared among tasks. There have been many effective synchronization ....

[Article contains additional citation context not shown here]

Kopetz, H., and Reisinger, J. The Non-Blocking Write Protocol NBW: A Solution to a Real-Time Synchronization Problem. In Proc. of the 14th IEEE Real-Time Systems Symposium (1993), pp. 131--137.

....of short to medium duration allowed (the case we most expect and optimize for) but also calls that are quite long, approaching the length of an entire quantum. Our work builds upon recent research by us and others on using lock free and wait free shared object algorithms in real time systems [3, 4, 5, 6, 15, 19]. Operations on lockfree objects are optimistically performed using a user level retry loop. Such an operation is atomically validated and committed by invoking a synchronization primitive such as compare and swap (CAS) The retry loop is executed repeatedly until this validation step succeeds. ....

H. Kopetz and J. Reisinger. The non-blocking write protocol nbw: A solution to a real-time synchronization problem. In Proceedings of the 14th IEEE Symp. on Real-Time Systems, pp. 131--137. 1993.

No context found.

H. Kopetz and J. Reisinger. The Non-Blocking Write Protocol NBW: A Solution to a Real-Time Synchronisation Problem. In Proc. 14th Real-Time Systems Symposium, Raleigh-Durham, North Carolina, USA, Dec. 1993.

No context found.

H. Kopetz and J. Reisinger. The Non-Blocking Write Protocol NBW: A Solution to a Real-Time Synchronisation Problem. In Proc. 14th Real-Time Systems Symposium, Raleigh-Durham, North Carolina, USA, Dec. 1993.

....cf. Section II E) propagation of control errors is prohibited by design. We call an interface that prevents propagation of control errors by design a temporal firewall [46] The integrity of the data in the temporal firewall is assured by the non blocking write (NBW) concurrency control protocol [47]. Fig. 8. Data Flow and Control Flow at a TTA Interface From the point of view of complexity management and composability, it is useful to distinguish between three different types of interfaces of a node: the real time service (RS) interface, the diagnostic and maintenance (DM) interface, and ....

H. Kopetz and J. Reisinger. The Non-Blocking Write Protocol NBW: A Solution to a Real-Time Synchronization Problem. In Proceedings of the 14th Real-Time Systems Symposium, pages 131--137, 1993.

....of a file record is atomic. This implies that after a file write the horizontal check byte and the check record must be updated atomically. If a user needs a level of atomicity beyond the single file record, he she has to design his her own concurrency control protocol, for example an NBW protocol [11] that uses one record as a concurrency control field. RODL Files A RODL file is a distributed file that contains the specification of a RODL for a particular round. A RODL file consists of a collection of (sub)files, one in each participating node. The RODL file name is also the round name. The ....

Kopetz, H. and J. Reisinger (1993). The Non-Blocking Write Protocol NBW: A Solution to a Real-Time Synchronisation Problem. Proc. 14th Real-Time Systems Symposium, Raleigh-Durham, North Carolina. pp.131-137

No context found.

H. Kopetz and J. Reisinge. The non-blocking write protocol NBW: A solution to a realtime synchronisation problem. In Susan Davidson and Insup Lee, editors, Proceedings of the Real-Time Systems Symposium, pages 131-137, Raleigh-Durham, NC, December 1993. IEEE Computer Society Press.

No context found.

Hermann Kopetz and Johannes Reisinger. The non-blocking write protocol NBW: A solution to a real-time synchronization problem. In Real Time Systems Symposium, pages 131--137, Raleigh-Durham, NC, December 1993. IEEE Computer Society.

No context found.

Hermann Kopetz and Johannes Reisinger. The non-blocking write protocol NBW: A solution to a real-time synchronization problem. In Real Time Systems Symposium, pages 131--137, Raleigh-Durham, NC, December 1993. IEEE Computer Society.

No context found.

Kopetz H. and Reisinger J. The Non-Blocking Write Protocol NBW: A Solution to a RealTime Synchronization Problem. In Proceedings of he 14th Real-Time Systems Symposium, pp. 131-137, 1993.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC