The Jalapeño dynamic optimizing compiler for Java

Burke, Michael G.; Choi, Jong-Deok; Fink, Stephen J.; Grove, David; Hind, Michael; Sarkar, Vivek; Serrano, Mauricio J.; Sreedhar, Vugranam C.; Srinivasan, Harini; Whaley, John

doi:10.1145/304065.304113

Cited by 196 publications

(137 citation statements)

References 23 publications

Supporting

Mentioning

129

Contrasting

Unclassified

Order By: Relevance

“…Choi et al [8] use an intermediate representation from the Jalapeño compiler [7] to extract CFGs with exceptional flows. The authors introduce a stackless representation, using assertions to mark the possibility of an instruction raising an exception.…”

Section: Related Workmentioning

confidence: 99%

Sound Control-Flow Graph Extraction for Java Programs with Exceptions

Amighi

Gomes

Gurov

et al. 2012

Software Engineering and Formal Methods

View full text Add to dashboard Cite

Abstract. We present an algorithm to extract control-flow graphs from Java bytecode, considering exceptional flows. We then establish its correctness: the behavior of the extracted graphs is shown to be a sound over-approximation of the behavior of the original programs. Thus, any temporal safety property that holds for the extracted control-flow graph also holds for the original program. This makes the extracted graphs suitable for performing various static analyses, in particular model checking. The extraction proceeds in two phases. First, we translate Java bytecode into BIR, a stack-less intermediate representation. The BIR transformation is developed as a module of Sawja, a novel static analysis framework for Java bytecode. Besides Sawja's efficiency, the resulting intermediate representation is more compact than the original bytecode and provides an explicit representation of exceptions. These features make BIR a natural starting point for sound control-flow graph extraction. Next, we formally define the transformation from BIR to control-flow graphs, which (among other features) considers the propagation of uncaught exceptions within method calls. We prove the correctness of the two-phase extraction by suitably combining the properties of the two transformations with those of an idealized control-flow graph extraction algorithm, whose correctness has been proved directly. The control-flow graph extraction algorithm is implemented in the ConFlEx tool. A number of test-cases show the efficiency and the utility of the implementation.

show abstract

Section: Related Workmentioning

confidence: 99%

Sound Control-Flow Graph Extraction for Java Programs with Exceptions

Amighi

Gomes

Gurov

et al. 2012

Software Engineering and Formal Methods

View full text Add to dashboard Cite

show abstract

“…Once the instrumentation indicates that a given method is hot (i.e. number of times the method is executed is above a threshold), the AOS decides whether to compile it using an optimizing compiler [4]. The OTF hijacks this decision, so that any hot method is also considered for parallelizing optimizations.…”

Section: Online Tuning Frameworkmentioning

confidence: 99%

Optimizing Chip Multiprocessor Work Distribution Using Dynamic Compilation

Zhao

Horsnell

Rogers

et al. 2007

Euro-Par 2007 Parallel Processing

View full text Add to dashboard Cite

Abstract. How can sequential applications benefit from the ubiquitous next generation of chip multiprocessors (CMP)? Part of the answer may be a dynamic execution environment that automatically parallelizes programs and adaptively tunes the work distribution. Experiments using the Jamaica CMP show how a runtime environment is capable of parallelizing standard benchmarks and achieving performance improvements over traditional work distributions.

show abstract

“…The baseline compiler provides a near-direct translation of Java class files thereby compiling very quickly and producing code with execution speeds similar to that of interpreted code. The second compiler is highly optimizing and builds upon extensive compiler technology to perform various levels of optimization [8]. The compilation time using the optimizing compiler is much slower than the baseline (50 times on average in the programs studied), but produces code that executes 3-4 times faster.…”

Section: Jalapeñomentioning

confidence: 99%

“…The primary challenge in using dynamic compilation is to enable high performance execution with minimal compilation overhead. Unfortunately, a common practice thus far in the evaluation of dynamic compilers for Java, has been to omit measurements of compilation overhead and to report only execution time [26,9,24,8]. Hence, it is difficult for users to evaluate the tradeoff between compilation overhead and execution speedup.…”

Section: Introductionmentioning

confidence: 99%