Self-optimizing AST interpreters dynamically adapt to the provided input for faster execution. This adaptation includes initial tests of the input, changes to AST nodes, and insertion of guards that ensure assumptions still hold. Such specialization and speculation is essential for the performance of dynamic programming languages such as JavaScript. In traditional procedural and objectoriented programming languages it can be tedious to write selfoptimizing AST interpreters, as those languages fail to provide constructs that would specifically support that. This paper introduces a declarative domain-specific language (DSL) that greatly simplifies writing self-optimizing AST interpreters. The DSL supports specialization of operations based on types of the input and other properties. It can then use these specializations directly or chain them to represent the operation with the minimum amount of code possible. The DSL significantly reduces the complexity of expressing specializations for those interpreters. We use it in our high-performance implementation of JavaScript, where 274 language operations have an average of about 4 and a maximum of 190 specializations. In addition, the DSL is used in implementations of Ruby, Python, R, and Smalltalk.

Creating an interpreter is a simple and fast way to implement a dynamic programming language. With this ease also come major drawbacks. Interpreters are significantly slower than compiled machine code because they have a high dispatch overhead and cannot perform optimizations. To overcome these limitations, interpreters are commonly combined with just-in-time compilers to improve the overall performance. However, this means that a just-in-time compiler has to be implemented for each language. We explore the approach of taking an interpreter of a dynamic language and running it on top of an optimizing trace-based virtual machine, i.e., we run a guest VM on top of a host VM . The host VM uses trace recording to observe the guest VM executing the application program. Each recorded trace represents a sequence of guest VM bytecodes corresponding to a given execution path through the application program. The host VM optimizes and compiles these traces to machine code, thus eliminating the need for a custom just-in-time compiler for the guest VM. The guest VM only needs to provide basic information about its interpreter loop to the host VM.

Optionally typed languages enable direct performance comparisons between untyped and type annotated source code. We present a comprehensive performance evaluation of two different JIT compilers in the context of ActionScript, a production-quality optionally typed language. One JIT compiler is optimized for quick compilation rather than JIT compiled code performance. The second JIT compiler is a more aggressively optimizing compiler, performing both high-level and low-level optimizations.We evaluate both JIT compilers directly on the same benchmark suite, measuring their performance changes across fully typed, partially typed, and untyped code. Such evaluations are especially relevant to dynamically typed languages such as JavaScript, which are currently evaluating the idea of adding optional type annotations. We demonstrate that low-level optimizations rarely accelerate the program enough to pay back the investment into performing them in an optionally typed language. Our experiments and data demonstrate that high-level optimizations are required to improve performance by any significant amount.

An abstract syntax tree (AST) interpreter is a simple and natural way to implement a programming language. However, it is also considered the slowest approach because of the high overhead of virtual method dispatch. Language implementers therefore define bytecodes to speed up interpretation, at the cost of introducing inflexible and hard to maintain bytecode formats. We present a novel approach to implementing AST interpreters in which the AST is modified during interpretation to incorporate type feedback. This tree rewriting is a general and powerful mechanism to optimize many constructs common in dynamic programming languages. Our system is implemented in Java and uses the static typing and primitive data types of Java elegantly to avoid the cost of boxed representations of primitive values in dynamic programming languages.

Dynamic updates to running programs improve development productivity and reduce downtime of long-running applications. This feature is however severely limited in current virtual machines for object-oriented languages. In particular, changes to classes often apply only to methods invoked after a class change, but not to active methods on the call stack of threads. Additionally, adding and removing methods as well as fields is often not supported. We present a novel programming model for safe and atomic code updates of Java programs that also updates methods that are currently executed. We introduce safe update regions and pause threads only there before an update. We automatically convert the stack frames to suit the new versions of the methods. Our implementation is based on a production-quality Java virtual machine. Additionally, we present SafeWeave, a dynamic aspect-oriented programming system that exposes the atomic code updates through a high-level programming model. AspectJ advice can be added to and removed from a running application.Changes are atomic and correctness is guaranteed even though weaving happens in parallel to program execution, and the system fully supports the dynamic class loading of Java. We show that the enhanced evolution features do not incur any performance penalty before and after version changes.