W. Halang and A. Stoyenko, Constructing Predictable Real-Time Systems. BostonDordrecht -London: Kluwer Academic Publishers, 1991.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....based on control data flow diagrams (Section II contains a discussion of its relation to previous work) Thus far, we have generated C code in Step 3, though any target language is possible. Step 4 uses the software performance estimation package and classical real time scheduling algorithms [24] [18] to schedule the CFSM s while meeting the given timing constraints. The compilation in Step 5 is done using existing C compilers for the target embedded processors. The paper is organized as follows. Section II contains background information and a summary of the CFSM network model. Section III ....

W. A. Halang and A. D. Stoyenko, Constructing Predictable Real Time Systems. Norwell, MA: Kluwer Academic, 1991.

....available within a predefined deadline. It has turned out that a major progress in order to guarantee the timeliness of real time systems can only be achieved if the scheduling problem is solved accordingly. Most scheduling algorithms assume that the runtime of a task is known a priori (cf. e.g. [1, 2, 3]) Thus the worst case performance of a task plays a crucial role. The most difficult task in estimating the timing behavior of a program is to determine the number of iterations of a certain loop. Ordinary programming languages support two different forms of loop statements: for loops: A loop ....

....loop for i3 in 1. i2 loop for ir in 1. i r 1 loop innermost loop body end loop; is performed exactly i11 il il irl General loops, however, represent a very difficult task. In order to estimate the worst case performance of general loops many methods and tools have been developed, e.g. [2, 5, 6, 7]. In the following we will discuss some of them: In [5] language constructs have been introduced in order to let the program mer integrate knowledge about the actual behavior of algorithms which can not be expressed using standard programming language features. These constructs are scopes, ....

[Article contains additional citation context not shown here]

W. A. Halang and A.D. Stoyenko. Constructing predictable real time systems. Kluwer Academic Publishers, Boston, 1991.

....synthesis methods perform some sort of scheduling sequencing the execution of a set of originally concurrent tasks. Concurrent tasks are an excellent specification mechanism, but cannot be implemented as such on a standard CPU. The scheduling problem (reviewed e.g. by Halang and Stoyenko [127]) amounts to finding a linear execution order for the elementary operations composing the tasks, so that all the EDWARDS et al. DESIGN OF EMBEDDED SYSTEMS: FORMAL MODELS, VALIDATION, AND SYNTHESIS 383 timing constraints are satisfied. Depending on how and when this linearization is performed, ....

....or quasi static scheduling, or at least for minimizing preemptive scheduling in order to minimize scheduling overhead and to improve reliability and predictability. There are, of course, cases in which preemption cannot be avoided, because it is the only feasible solution to the problem instance ([127]) but such cases should be carefully analyzed to limit preemption to a minimum. Many static scheduling methods have been developed. Most somehow construct a precedence graph and then apply or adapt classical methods. We refer the reader to Bhattacharyya et al. 32] and Sih and Lee [128] 129] ....

W.A. Halang and A.D. Stoyenko, Constructing predictable real time systems, Kluwer Academic Publishers, 1991.

....One solution to increase performance is to decrease the administration, e.g. scheduling, clock tick administration etc. by utilising a co processor. There are several examples of utilising special purpose RTOS co processors [11] 12] 13] 14] 16] 8] or standard processor RTOS co processors [9] [10] The benefit of utilising special purpose hardware compared to utilising a standard processor is that it can be designed to be more predictable and to have greater performance, due to utilisation of parallel hardware. Additionally the flexibility of utilising a standard processor is not ....

W. A. Halang, A. D. Stoyenko, "Constructing Predictable Real Time Systems", Kluwer Academic Publisher 1991.

....design of software system is to plan the budgets for the resource demands of all the parts and operations then perform a validation check for the required performance. The real time system requirements such as timeliness, predictibility, dependendibility, and deadline scheduling are discussed in [Halang91]. The idea of validation of resource budgets is discussed in [Woodside99] Performance budgeting allocates resources based on achieving specific measurable resource demands (i.e. CPU execution demand, disk operation demand and network demand etc) rather than the required resources selection and ....

....change. In previous research, there has been frequent mention of budgeting of time in software, particularly in designing for hard real time deadlines. In these systems, schedule analysis plays the role of the model; to verify that timing constraints can be met deterministically (see, e.g. [Halang91]) For soft real time, the discussion has been rather general and solid techniques are scarce. This work is different in that it starts from designer intentions expressed in the UCMs, and that it addresses budgeting issues in soft real time systems. 3.3 Budget Analysis Road Map Figure 3.1 ....

W. Halang and A. Stoyenko, Constructing Predictable Real-Time Systems, Kluwer Academic Publishers, Boston, 1991

....In previous research, there has been frequent mention of budgeting of time in software, particularly in designing for hard real time deadlines. In these systems, schedulability analysis plays the role of the model; to verify that timing constraints can be met deterministically (see, e.g. [6]) For soft real time requirements, the discussion has been rather general and solid techniques are scarce. Previously, the creation of a performance model from a design specification was addressed in [1] based on prototyping in a CASE tool (Objectime) and tracing scenarios executed by the ....

W. Halang and A. Stoyenko, Constructing Predictable Real-Time Systems. Amsterdam: Kluwer Academic Publishers, 1991.

....compiler are determining factors in the logical correctness which is highly implementation independent. On the contrary, the temporal correctness is implementation dependent: the actual performances of the processor must be taken into account. Recently, a book has been dedicated to predictability [1]. To guarantee predictable behaviour is a difficult task because it is a matter of both logical and temporal correctness. In this paper, we address the issue of predictability for a restricted 276 C. ANDR E AND M.A. P ERALDI class of real time system: Our applications are almost pure reactive ....

W. Halang and A. Stoyenko, Constructing Predictable Real-Time Systems. Amsterdam: Kluwer Academic Publishers, 1991.

....has been derived (cf. e.g. KU76, KU77] and a large number of algorithms has been developed (cf. e.g. AC76, GW76, HU77, Sre95, SGL98, Tar81a, Tar81b] See [MR90, RP86] for an overview. Worst Case Execution Time (WCET) analysis does not have such a long standing tradition (cf. e.g. CBW96, HS91, ITM90, NP93, Par93, PK89, PS97, Sha89] Designers of real time programming languages usually restrict language features in order to make it possible to guarantee time bounds and introduce new language features to let the programmer add extraneous information on the algorithms which cannot be ....

....and introduce new language features to let the programmer add extraneous information on the algorithms which cannot be determined from the source code. Several di#erent approaches to WCET analysis have been pursued. The most important ones are described shortly in the following: 1. In [KS86, HS91] Real Time Euclid, a language for implementing real time systems, is presented. Real Time Euclid prohibits the use of recursions and goto statements. Loops are restricted to time bounded loops and simple forloops. An algorithm for calculating an upper bound of the WCET of RealTime Euclid programs ....

[Article contains additional citation context not shown here]

Wolfgang A. Halang and Alexander D. Stoyenko, Constructing predictable real time systems, Kluwer Academic Publishers, Boston, 1991. 1, 1, 1, 7, 21

....must respond to criteria of great simplicity and effectiveness, considering the small dimensions of the system. In this area, much of the knowledge acquired in the field of operating systems, especially RealTime Operating Systems (RTOS) has been applied. For an overview of scheduling methods see [Halang and Stoyenko 1991]. As we said above, in [Berry et al. 1991] the authors propose an approach that takes a single FSM that solves the problem of communication and concurrency between the modules starting from the specification in Esterel of the system as a set of concurrent modules, which do not require the ....

Halang, W. and Stoyenko, A. 1991. Constructing predictable real time systems. Kluwer Academic Publishers.

....missing it leads to a system failure, and must be absolutely avoided. A deadline is soft if missing it has just an associated cost, that must generally be minimized over time. 26 Polis provides both conservative analysis techniques (based on classical real time theory results, summarized e.g. in [17]) and the estimated timing simulation technique described above, in order to choose the scheduling policy for a given system. The Real Time Operating System synthesis command in Polis is called gen os. Its purpose is three fold: 1. It assigns I O ports (using port based or memory mapped I O) to ....

W.A. Halang and A.D. Stoyenko. Constructing predictable real time systems. Kluwer Academic Publishers, 1991.

....and session layer service interfaces [topolcic90] anderson90] QuAL language constructs for the expression of processing constraints target handling of real time demands in a generic and robust way, that enables graceful recovery from degradation. A survey of real time languages can be found in [halang91] and an analysis of the approach used in QuAL to handle real time features can be found in [florissi94] 6 Conclusions This paper describes the Quality Assurance Language (QuAL) that eases the management of Quality of Service (QoS) for distributed multimedia computing and communication ....

Halang, W. A. and Stoyenko, A. D., Constructing Predictable Real Time Systems. Boston/Dordrecht/London: Kluwer Academic Publishers, 1991.

....be available within a predefined deadline. It has turned out that major progress in order to guarantee the timeliness of real time systems can only be achieved if the scheduling problem is solved properly. Most scheduling algorithms assume that the runtime of a task is known a priori (cf. e.g. [8, 5, 10]) Thus the worst case performance of a task plays a crucial role. The most difficult tasks in estimating the timing behavior of a program are to determine the number of iterations of a certain loop and to handle problems originating from the use of recursion. A solution to the first problem has ....

....systems are embedded systems with limited storage space, the result of a recursive procedure must be computed using a limited amount of stack space. In view of these problems most designers of realtime programming languages decide to forbid recursion in their languages, e.g. RT Euclid (cf. [6, 5]) Supported by the Austrian Science Foundation (FWF) under grant P10188 MAT. PEARL (cf. 3] Real Time Concurrent C (cf. 4] and the MARS approach (cf. 7, 11] Our approach is different in that we do not forbid recursion, but instead constrain recursive procedures such that their space and ....

W. A. Halang and A. D. Stoyenko. Constructing predictable real time systems. Kluwer Academic Publishers, Boston, 1991.

....synthesis methods perform some sort of scheduling sequencing the execution of a set of originally concurrent tasks. Concurrent tasks are an excellent specification mechanism, but cannot be implemented as such on a standard CPU. The scheduling problem (reviewed e.g. by Halang and Stoyenko [HS91] amounts to finding a linear execution order for the elementary operations composing the tasks, so that all the timing constraints are satisfied. Depending on how and when this linearization is performed, scheduling algorithms can be classified as: Static, where all scheduling decisions are ....

....or quasi static scheduling, or at least for minimizing preemptive scheduling in order to minimize scheduling overhead and to improve reliability and predictability. There are, of course, cases in which preemption cannot be avoided, because it is the only feasible solution to the problem instance ( HS91] but such cases should be carefully analyzed to limit preemption to a minimum. 47 paper model interface constraint scheduling granularity algorithm Cochran [Coc92] task list none task RMA (runtime) Chou [CWB94] task list synthesized task heuristic (static) operation Gupta [GJM94] CDFG ....

W.A. Halang and A.D. Stoyenko. Constructing predictable real time systems. Kluwer Academic Publishers, 1991.

....One solution to increase performance is to decrease the administration, e.g. scheduling, clock tick administration etc. by utilising a co processor. There are several examples of utilising special purpose RTOS co processors [11] 12] 13] 14] 16] 8] or standard processor RTOS coprocessors [9] [10] The benefit of utilising special purpose hardware compared to utilising a standard processor is that it can be designed to be more predictable and to have greater performance, due to utilisation of parallel hardware. Additionally the flexibility of utilising a standard processor is not as ....

W. A. Halang, A. D. Stoyenko, "Constructing Predictable Real Time Systems", Kluwer Academic Publisher 1991.

....1976; Hecht and Ullman, 1977; Sreedhar, 1995; Sreedhar et al. 1998; Tarjan, 1981a; Tarjan, 1981b) See (Marlowe and Ryder, 1990; Ryder and Paull, 1986) for an overview. Worst Case Execution Time (WCET) analysis does not have such a long standing tradition (cf. e.g. Chapman et al. 1996; Halang and Stoyenko, 1991; Ishikawa et al. 1990; Nirkhe and Pugh, 1993; Park, 1993; Puschner and Koza, 1989; Puschner and Schedl, 1997; Shaw, 1989) Designers of real time programming languages usually restrict language features in order to make it possible to guarantee time bounds and introduce new language features to ....

....code. c fl 2000 Kluwer Academic Publishers. Printed in the Netherlands. dfwcetklu.tex; 14 04 2000; 9:10; p.1 2 Johann Blieberger Several different approaches to WCET analysis have been pursued. The most important ones are described shortly in the following: 1. In (Kligerman and Stoyenko, 1986; Halang and Stoyenko, 1991) Real Time Euclid, a language for implementing real time systems, is presented. Real Time Euclid prohibits the use of recursions and goto statements. Loops are restricted to time bounded loops and simple for loops. An algorithm for calculating an upper bound of the WCET of Real Time Euclid ....

[Article contains additional citation context not shown here]

Halang, W. A. and A. D. Stoyenko: 1991, Constructing predictable real time systems. Boston: Kluwer Academic Publishers.

....worst cases, deadlines, maximum run times, and maximum delays need to be considered. For the realisation of predictable and dependable real time systems, reasoning in static terms and the acceptance of physical constraints is a must all dynamic and virtual features are considered harmful. [7] They further advocate that features such as . caching, paging and swapping must either be disallowed or restricted. 7] These arguments do not necessarily rule out the use of DRAM and other high latency memories entirely. However, in their view, attempts at reducing the memory latency, for ....

....dependable real time systems, reasoning in static terms and the acceptance of physical constraints is a must all dynamic and virtual features are considered harmful. 7] They further advocate that features such as . caching, paging and swapping must either be disallowed or restricted. [7] These arguments do not necessarily rule out the use of DRAM and other high latency memories entirely. However, in their view, attempts at reducing the memory latency, for the purposes of determining worst case execution times, Mwave is a trademark of International Business Machines ....

[Article contains additional citation context not shown here]

Wolfgang A. Halang and Alexander D. Stoyenko. Constructing Predictable Real Time Systems, Boston: Kluwer Academic Publishers, 1991.

....explicit support for interaction among the specification, software development, compilation, and operating system layers, which many system try to decouple, rather than coordinate. Halang and Stoyenko point out that an integrated approach is required to construct a predictable real time system [9]. Systems addressing some aspects of integration have existed for some time, but do not span the software development cycle [1] In most system, functional abstraction is supported by some off theshelf design language followed by implementation with a real time programming language such as ....

Halang, W. and Stoyenko, A. Constructing Predictable Real-Time Systems. Kluwer Academic Publishers, 1991.

....expressions[CH74] For real time multiprocessor systems, there is an ongoing debate as to what is the appropriate mode of synchronization and communication. One view is to have time as a parameter, in the synchronization and communication routines both at the operating system and language levels[HS91]. For example: seize(resource id, first instant of a schedule, increment between the instants, last instant) This routine sets a timetable of when a task shall try to grab a resource. The critics holding the dissenting view [Geb92] Tur89] discount the use of time as a parameter for real time ....

....mailboxes for inter task communication. Our scheme is unique from the above mentioned schemes, as we are not just handling synchronization and communication, but also providing hints to the run time environment improving the overall predictability of the system. Time parameters are proposed in [HS91]. But those propositions were at language level as extensions to PEARL, an industrial real time programming language. Our design is at the operating system level. Doing it at language level is less effective unless the underlying operating system also supports it. Otherwise, at most, the compiler ....

W. A. Halang, A.D. Stoyenko, "Constructing Predictable Real-Time Systems", Kluwer Academic Publishers, Boston, MA, USA, 1993

....available within a predefined deadline. It has turned out that a major progress in order to guarantee the timeliness of real time systems can only be achieved if the scheduling problem is solved accordingly. Most scheduling algorithms assume that the runtime of a task is known a priori (cf. e.g. [LL73, HS91, Mok84]) Thus the worst case performance of a task plays a crucial role. Supported by the Austrian Science Foundation (FWF) under grant P10188 MAT. 1 The most difficult task in estimating the timing behavior of a program is to determine the number of iterations of a certain loop. Determining the ....

....is trivial. For example the loop body of the loop for i in 1. N loop loop body end loop; is performed exactly N times. General loops, however, represent a very difficult task. In order to estimate the worst case performance of general loops many methods and tools have been developed, e.g. [HS91, PK89, NP93, Par93]. Most researchers, however, try to ease the task of estimating the number of general loop iterations by forbidding general loops, i.e. by forcing the user to supply constant upper bounds for the number of iterations. Another approach is to let the user specify a time bound within the loop has to ....

[Article contains additional citation context not shown here]

W. A. Halang and A. D. Stoyenko. Constructing predictable real time systems. Kluwer Academic Publishers, Boston, 1991.

....available within a predefined deadline. It has turned out that a major progress in order to guarantee the timeliness of real time systems can only be achieved if the scheduling problem is solved accordingly. Most scheduling algorithms assume that the runtime of a task is known a priori (cf. e.g. [1, 2, 3]) Thus the worst case performance of a task plays a crucial role. The most difficult task in estimating the timing behavior of a program is to determine the number of iterations of a certain loop. Ordinary programming languages support two different forms of loop statements: for loops: A loop ....

....X i 1 =1 i 1 X i 2 =1 i 2 X i 3 =1 Delta Delta Delta i r Gamma1 X i r =1 1 = N r Gamma 1 r times. General loops, however, represent a very difficult task. In order to estimate the worst case performance of general loops many methods and tools have been developed, e.g. [2, 5, 6, 7]. In the following we will discuss some of them: ffl In [5] language constructs have been introduced in order to let the programmer integrate knowledge about the actual behavior of algorithms which can not be expressed using standard programming language features. These constructs are scopes, ....

[Article contains additional citation context not shown here]

W. A. Halang and A. D. Stoyenko. Constructing predictable real time systems. Kluwer Academic Publishers, Boston, 1991.

No context found.

W. Halang and A. Stoyenko, Constructing Predictable Real-Time Systems. BostonDordrecht -London: Kluwer Academic Publishers, 1991.

No context found.

W.A. Halang and A.D. Stoyenko. Constructing predictable real time systems. Kluwer Academic Publishers, 1991.

No context found.

W. Halang and A. Stoyenko, Constructing Predictable Real-Time Systems. Amsterdam: Kluwer Academic Publishers, 1991.

No context found.

W.A. Halang and K. M. Sacha. Constructing Predictable Real-Time Systems. Kluwer Academic Publishers, Boston, 1991.

No context found.

HS91. Wolfgang A. Halang and Alexander D. Stoyenko, Constructing predictable real time systems, Kluwer Academic Publishers, Boston, 1991.

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