Bibliographic record
Abstract
Java programs are deployed in a bytecode format that is executed by a Java virtual machine (JVM). JVM performance is determined by several major components: execution engine, garbage collector, and threading system. The static compilation and optimization approach, such as taken in C/C++ compilers, does not fit in Java's execution model very well because Java allows dynamic class loading, lazy resolution, just-in-time (JIT) compilation, and garbage collection. These dynamic features presents new challenges to JVM designers. In this thesis, we study both the challenges and opportunities of dynamic optimizations in Java virtual machines. Our contributions include a new garbage collector using dynamic techniques and dynamic interprocedural program analyses for speculative optimizations in JIT compilers. We present a novel approach for reducing garbage collection frequencies. Instead of relying on an ahead-of-time escape analysis or a region-based type system, our approach adapts regions based on the runtime history of an application. By freeing regions with associated stack frames, the system can reduce the frequency of garbage collections. We present the overall idea and provide details of a specific design and implementation. Dynamic class loading is a two-edged sword. A JIT compiler can speculatively optimize methods base on loaded classes only. However, newly loaded classes may invalidate previous optimization assumptions. We review existing techniques supporting speculative optimizations, including runtime guards, code patching, and on-stack replacement. We present an improvement and implementation of an on-stack replacement mechanism. A call graph is necessary for developing interprocedural program analyses. Call graph construction in a Java virtual machine needs to overcome the difficulties of dynamic class loading and lazy reference resolution. We show a general approach to adapt static type analyses to dynamic versions suitable for building call graphs in a JIT environment. We also introduce a new call graph profiling mechanism using code stubs. Having dynamic call graphs, we study reachability-based interprocedural analysis. We describe a general type analysis framework for supporting speculative method inlining in a JIT environment. Several popular type analyses were implemented in the framework, including an interprocedural one, VTA [SHR+00]. Using the framework, we present the results of a limit study of method inlining and report our findings and experience. In each chapter we discuss the related work for that chapter's topic. At the end of the thesis, we point out future research directions.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".