D. McNamee, J. Walpole, C. Pu, C. Cowan, C. Krasic, A. Goel, P. Wangle, C. Consel, G. Muller, and R. Marlet. Specialization Tools and Techniques for Systematic Optimization of System Software. Transactions on Computer Systems, 19(2):217--251, 2001.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....accessed by any thread in the system. Perhaps the most straightforward way to achieve a quiescent state would be to require all clients of the Clustered Object instance to acquire a reader writer lock in read mode before any call to the object (as done in the re plugging mechanism described in [7]) Acquiring this external lock in write mode would thus establish that the object is safe for swapping. However, this approach adds overhead for the common case and can cause locality problems, defeating the scalability advantages of Clustered Objects. Further, the lock could not be part of the ....

D. McNamee, J. Walpole, C. Pu, C. Cowan, C. Krasic, A. Goel, P. Wagle, C. Consel, G. Muller, and R. Marlet. Specialization tools and techniques for systematic optimization of system software. ACM Transactions on Computer Systems (TOCS), 19(2):217-251, 2001.

....the mechanism for updating client references. Pu, et al. describe a replugging mechanism for incremental and optimistic specialization [43] but they assume there can be at most one thread executing in a swappable module at a time. In later work, that constraint is relaxed but is non scalable [38]. In addition to the work done in different reconfiguration mechanisms, many groups have applied online reconfiguration to systems and achieved a variety of benefits. Many different adaptive techniques have been implemented to improve system performance [2, 25, 35] Extensible operating systems ....

D. McNamee, J. Walpole, C. Pu, C. Cowan, C. Krasic, A. Goel, P. Wagle, C. Consel, G. Muller, and R. Marlet. Specialization tools and techniques for systematic optimization of system software. ACM Transactions on Computer Systems (TOCS), 19(2):217--251. ACM Press.

....of the type done by Synthesis [24] and Synthetix [23] for operating systems. The primary disadvantage of such systems is the resulting loss of portability and maintainability. While the above systems used manual specializations, automated specialization approaches have also been described [17]. Most of these approaches rely on partial evaluation based on knowledge of input values. In contrast, we use program profiling to identify optimizable parts of the program. Our technique can be thought of as profile directed specialization and could easily be extended to all classes of programs ....

D. McNamee, J. Walpole, C. Pu, C. Cowan, C. Krasic, A. Goel, P. Wagle, C. Consel, G. Muller, and R. Marlet. Specialization tools and techniques for systematic optimization of system software. In ACM Transactions of Computer Systems, Vol 19, No 2, pages 217--251, May 2001.

.... transformation [8, 11, 25] and both practical and theoretical work aimed at generating programs from speci cations [37, 43] Recent years have seen a rapid growth of interest in generative, automatic approaches to program management, updating, adaptation, transformation, and evolution, e.g. [43, 52, 70]. The motivations are well known: persistence of the software crisis and dreams of achieving industrial style automation, automatic adaptation of programs to new contexts, and automatic optimization. The overall goal is to increase eciency of software production by making it possible to write ....

.... Functional languages ( 6, 14, 38, 44, 74] etc. Logic programming languages ( 25, 26, 48, 51] etc. Compiling or other transformation by specializing interpreters ( 5, 6, 28, 41, 40, 53] etc. Optimization of operating systems, e.g. remote procedure calls, device drivers, etc. [16, 52, 57], etc. 2.1 Equational de nition of a parser generator. To get started, and to connect partial evaluation with program generation, we rst exemplify our notation for program runs on an example. Readers familiar with the eld may wish to skip to Section . Parser generation is a familiar ....

[Article contains additional citation context not shown here]

Dylan McNamee, Jonathan Walpole, Calton Pu, Crispin Cowan, Charles Krasic, Ashvin Goel, Perry Wagle, Charles Consel, Gilles Muller, and Renaud Marlet. Specialization tools and techniques for systematic optimization of system software. ACM Transactions on Computer Systems, 19(2):217-251, 2001.

No context found.

D. McNamee, J. Walpole, C. Pu, C. Cowan, C. Krasic, A. Goel, P Wagle, C. Consel, G. Muller, and R. Marlet. Specialization tools and techniques for systematic optimization of system software. ACM Transactions on Computer Systems, 19(2):217--251, May 2001.

No context found.

D. McNamee, J. Walpole, C. Pu, C. Cowan, C. Krasic, A. Goel, P. Wagle, C. Consel, G. Muller, and R. Marlet. Specialization tools and techniques for systematic optimization of system software. ACM TCS, 19(2), 2001.

....of con guration parameters. Computations that depend on these early values are performed and the corresponding results are encoded in a specialized program. Extensive research and experiments have made program specialization a well established approach to reconciling genericity and performance [2,3]. Program specialization has demonstrated its e ectiveness in a variety of areas such as operating systems, networking, graphics, scienti c computing and compiler generation [3 5] Program specialization is typically performed in two stages: preprocessing and processing. The preprocessing phase ....

.... and experiments have made program specialization a well established approach to reconciling genericity and performance [2,3] Program specialization has demonstrated its e ectiveness in a variety of areas such as operating systems, networking, graphics, scienti c computing and compiler generation [3 5]. Program specialization is typically performed in two stages: preprocessing and processing. The preprocessing phase consists of a form of dependency analysis, called a binding time analysis, that takes a program and a description of the LaBRI Research Report 1299 03 con guration parameters ....

[Article contains additional citation context not shown here]

D. McNamee, J. Walpole, C. Pu, C. Cowan, C. Krasic, C. Goel, C. Consel, G. Muller, R. Marlet, Specialization tools and techniques for systematic optimization of system software, ACM Transactions on Computer Systems 19 (2001) 217-251.

....If we were to build Infopipes the traditional way, modularity would have implied high composition and layering overhead. The first building block we use is the specialization technology to reduce execution overhead through systematic program transformations developed in the Synthetix project [7, 9], among others. A relevant example is the automated specialization of SUN RPC code using the Tempo C specializer [8] This is a demonstration of specialization technology on production software, by cutting out layering overhead using program analysis automated in Tempo C. In addition to C and ....

Dylan McNamee, Jonathan Walpole, Calton Pu, Crispin Cowan, Charles Krasic, Ashvin Goel, and Perry Wagle, Charles Consel, Gilles Muller, and Renaud Marlet, "Specialization Tools and Techniques for Systematic Optimization of System Software ", ACM Transactions on Computer Systems. To appear in the May 2001 issue.

.... functionality can exploit these capabilities, we are investigating scientific collaboration, remote sensing, and operational information systems [15] We are building on our own and others previous work on adaptive systems [24, 25, 34] and on dynamic program or system specialization [14], but our new research differs in that we are also trying to develop techniques for the automatic creation of adaptive software services and of adaptation techniques for such services, which we term service morphing. Specifically, rather than requiring applications to be explicitly made ....

.... online monitoring of distributed resources for scientific applications [12] Finally, the compiler techniques used when dynamically specializing code, e.g. to match changes in power availability, are currently under development, in part based on our earlier research on runtime code specialization [14] and on code generation of embedded systems [11, 21, 32] Our research is based on our previous work with distributed, real time, adaptive, and multimedia systems [24, 27, 30] with sensor based embedded applications [23] and with high performance middleware [4, 9] and applications [31] We are ....

D. McNamee, J. Walpole, C. Pu, C. Cowan, C. Krasic, A. Goel, P. Wagle, C. Consel, G. Muller, and R. Marlet. Specialization tools and techniques for systematic optimization of system software. ACM Transactions on Computer Systems, 19(2):217--251, May 2001.

No context found.

D. McNamee, J. Walpole, C. Pu, C. Cowan, C. Krasic, A. Goel, P. Wangle, C. Consel, G. Muller, and R. Marlet. Specialization Tools and Techniques for Systematic Optimization of System Software. Transactions on Computer Systems, 19(2):217--251, 2001.

No context found.

Dylan McNamee, Jonathan Walpole, Calton Pu, Crispin Cowan, Charles Krasic, Ashvin Goel, Perry Wagle, Charles Consel, Gilles Muller, and Renauld Marlet. Specialization tools and techniques for systematic optimization of system software. ACM Transactions on Computer Systems (TOCS), 19(2):217-251, 2001.

No context found.

D. McNamee, J. Walpole, C. Pu, C. Cowan, C. Krasic, A. Goel, P. Wagle, C. Consel, G. Muller, and R. Marlet. Specialization Tools and Techniques for Systematic Optimization of System Software. ACM Transactions on Computer Systems, 19(2):217--251, May 2001.

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