John Plevyak and Andrew Chien. Incremental inference of concrete types. Technical Report UIUCDCSR -93-1829, Department of Computer Science, University of Illinois, Urbana, Illinois, June 1993.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....an interpreter which supports incremental program development and debugging, a source level symbolic debugger, and performance tools to evaluate parallel program performance. These tools are described in greater detail in the next section. More information on the Concert System can be found in [7, 18, 14, 19, 5]. 1 3.3 Concert Tools for Program Development and Tuning The Concert System supports the development of irregular parallel applications. The challenges for program development are much greater for explicitly parallel programs than for models which retain sequential program semantics; ....

John Plevyak and Andrew Chien. Incremental inference of concrete types. Technical Report UIUCDCSR -93-1829, Department of Computer Science, University of Illinois, Urbana, Illinois, June 1993.

....exact, to serve this purpose and allow customization to proceed. A variable of an exact class must refer to an object of exactly that class, and not of any subclass. Knowledge about exact types can be propagated using dataflow techniques similar to those in the Illinois Concert compiler [19]. The source of knowledge about exact types is objects produced by constructors, where the precise implementation returned is known. However, this knowledge is lost as soon as control passes a method dispatch boundary. Knowledge about exact types can be preserved only by customizing called ....

John Plevyak and Andrew A. Chien. Incremental inference of concrete types. Technical Report UIUCDCS-R-93-1829, Dept. of Computer Science, University of Illinois Urbana-Champagne, 1993.

....inference. In general, an 4 Grouping Objects 1 Image, grouped primitiveMaker Full Image primitiveMaker Figure 2. Step 1 groups the objects in the image according to their structure. adaptive approach to grouping may be necessary, analogous to the adaptive type inference algorithms described in [Plevyak Chien 1993, Agesen 1994] The extractor only groups objects as it encounters them. This procrastination saves the effort of grouping objects that are not reached during extraction, and most objects are never grouped: extracting PrimitiveMaker from an image with 100,000 objects, only 5000 of these are ....

.... could discover that R never returns hash tables, there would be no need to extract the size method for hash tables (unless, of course, it was needed elsewhere) In a prior publication, we report on a type inference algorithm that can compute this kind of information [Agesen et al. 1993] See also [Plevyak Chien 1993] for more recent results, and [Agesen 1994] for a comparison of different variations of the type inference algorithm) The second step in the extraction of PrimitiveMaker invokes our type inference algorithm on PrimitiveMaker 1 . After type inference, every expression and slot in PrimitiveMaker is ....

Plevyak, J., Chien, A. A. Incremental Inference of Concrete Types, Technical Report R-93-1829, Department of Computer Science, University of Illinois Urbana-Champaign, 1993.

....highest pay off. We define a general class of algorithms that do this: the adaptive algorithms. The two most powerful of the five algorithms we study fall in this class. 1 Introduction Static analysis of programs is gaining importance: it is the core of any optimizing or parallelizing compiler [20]. Other programming tools based on static analysis are also emerging: 3] describes a tool that can extract compact applications from integrated programming environments by analyzing an image to identify a small but sufficient set of objects for a given application. Constraint based analysis is a ....

....variables will constitute the nodes in a directed Table 1. The five algorithms we analyze. Name Section Reference Adaptive Basic algorithm 3.1 [16] No 1 level expansion algorithm 3.2 [16] No p level expansion algorithm 3.3 [19] No Hash function algorithm 3.5 [2] Yes Iterative algorithm 3. 6 [20] Yes 5 graph. The edges will be constraints. A constraint is the type inference time equivalent of a run time data flow. For example, if the target program executes the assignment x: y there is a data flow from y to x. Intuitively, the significance of the data flow is that any object that can ....

[Article contains additional citation context not shown here]

Plevyak, J. and A.A. Chien, Incremental Inference of Concrete Types, Department of Computer Science, University of Illinois Urbana-Champaign, Technical Report UIUCDCS-R-931829, 1993.

....range of irregular applications. Our application study is done in the context of the Illinois Concert system, a high performance compiler and runtime for parallel computers which has been the vehicle for extensive research on compiler optimization and runtime techniques over the past five years [9, 32, 29, 30, 33, 28, 31, 12, 44, 23, 22]. While no system contains all known optimizations, the Concert system contains a wide range of aggressive optimizations, and has been used to demonstrate high performance in absolute terms on a wide range of applications [23, 45, 10] In effect, the Concert system automatically addresses many of ....

....) 2. 3 Illinois Concert System Our application study is done in the context of the Illinois Concert system, a high performance compiler and runtime for parallel computers which has been the vehicle for extensive research on compiler optimization and runtime techniques over the past five years [9, 32, 29, 30, 33, 28, 31, 12, 44, 23, 22]. While no system contains all known optimizations, the Concert system contains a wide range of them, and has been used to demonstrate high performance in absolute terms on a wide range of applications [23, 45, 10] In effect, the Concert system automatically addresses many of the concerns which ....

John Plevyak and Andrew Chien. Incremental inference of concrete types. Technical Report UIUCDCS-R-93-1829, Department of Computer Science, University of Illinois, Urbana, Illinois, June 1993.

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