Interflow: interprocedural flow-sensitive type inference and method duplication

Scala heavily relies on a number of object-oriented abstractions to support its feature-rich collections library. There are known techniques that optimize those abstractions away in just-in-time (JIT) compilers, but applying them in the ahead-of-time (AOT) setting is problematic. Profile-guided optimization (PGO) lets AOT compilers apply some of the same optimizations that JIT compilers employ, but it comes at a high complexity cost. In this paper, we introduce Interflow, an alternative approach towards ahead-of-time optimization of Scala programs which relies on interprocedural flow-sensitive type inference and method duplication. Our evaluation shows that an Interflow-based optimizing compiler built on top of the Scala Native toolchain outperforms existing PGO-based optimizing compilers for Scala. Moreover, we demonstrate that Interflow and PGO can be combined to achieve further improvements. On our benchmarks, with both Interflow and PGO enabled, the Scala Native toolchain approaches the performance of the HotSpot JVM.

Just-in-time performance without warm-up

Denys Shabalin

Scala has been developed as a language that deeply integrates with the Java ecosystem. It offers seamless interoperability with existing Java libraries. Since the Scala compiler targets Java bytecode, Scala programs have access to high-performance runtimes including the HotSpot virtual machine. HotSpot provides impressive performance results achieved via just-in-time compilation. It starts program execution in interpreter mode, collecting profile feedback about called methods. This information allows HotSpot to identify hot spots in the program, which are then compiled on the fly to native code. This compilation scheme enables high peak performance at the cost of warmup time required to collect the profile data and perform just-in-time compilation. This is a good example of the traditional tradeoff between ahead-of-time (AOT) and just-in-time (JIT) compilation. With AOT, compilers have less information, but the runtime story is reasonably straightforward. With JIT, compilers have more information, which enables advanced optimizations, but the runtime story becomes complicated. In this dissertation, we present the design and implementation of Scala Native, an optimizing compiler for Scala. With Scala Native, Scala programs are compiled ahead of time, which avoids runtime compilation and enables instant startup times. On the other hand, Scala Native is able to match and supersede the peak performance of HotSpot on our benchmarks. In addition to that, Scala Native is a general-purpose Scala compiler - programs compiled by Scala Native closely match the behavior of programs compiled by the Scala compiler. First, we introduce NIR, an intermediate representation designed with ahead-of-time compilation in mind. NIR represents programs in the single-static assignment form and has support for object-oriented features such as virtual dispatch and multiple inheritance. This representation is a key enabler of our compilation and optimization pipeline. Secondly, we present Interflow, a link-time optimizer that takes advantage of the closed-world assumption to optimize the whole program at once. Our optimizer employs a number of techniques including partial evaluation, allocation sinking, and method duplication. The combination of these techniques allows Scala Native to outperform HotSpot on the majority of our benchmarks. Finally, we describe how to improve runtime performance even further based on profile feedback. We propose a technique that splits methods apart isolating key hot paths that are then optimized more aggressively than the cold parts of the program. This provides a further performance advantage over HotSpot.

EPFL2020

Embedded Domain-Specific Languages using Libraries and Dynamic Metaprogramming

Gilles Dubochet

Domain-specific programming is bound to play a central role in improving the productivity of writing software. This is because modern, intricate, general-purpose abstractions are rarely applicable directly to everyday programming. However, such abstractions can be used to encode domain-specific abstractions, so that the general-purpose language serves as a host in which domain-specific fragments are embedded. The benefit of traditional software engineering is maintained, while allowing writing software that integrates multiple domains. This thesis describes two such techniques in Scala –a modern language with functional and object-oriented features. The first allows embedding fragments of code with domain-specific syntax, which is weaved within the host language's own syntax so that it appears as being different. The second technique uses structural types to encode domain-specific correctness properties that object-oriented types cannot represent. An implementation of structural types in Scala is also described. The common characteristic of both techniques is that domain-specific properties are provided by a library relying only on general-purpose abstractions. However, the techniques also exemplify how certain embedding tasks rely on data that is available in the compiler but is then lost in statically-typed languages. General-purpose abstractions do not give access to this data, nor do traditional runtime reflection techniques; keeping all data is also impractical. This thesis describes two metaprogramming techniques whereby compiler data on code (code lifting) and types (manifests) can be requested by embedding libraries when needed. Compiler data is then available as literal values in the program. It further describes a metaprogramming framework that combine compiler data available in code with reflection. Its design is mirror-based, and allows a high degree of code sharing between the metaprogramming framework and the compiler.

EPFL2011

Surgical Precision JIT Compilers

Hassan Chafi, Tiark Rompf

Just-in-time (JIT) compilation of running programs provides more optimization opportunities than offline compilation. Modern JIT compilers, such as those in virtual machines like Oracle's HotSpot for Java or Google's V8 for JavaScript, rely on dynamic profiling as their key mechanism to guide optimizations. While these JIT compilers offer good average performance, their behavior is a black box and the achieved performance is highly unpredictable. In this paper, we propose to turn JIT compilation into a precision tool by adding two essential and generic metaprogramming facilities: First, allow programs to invoke JIT compilation explicitly. This enables controlled specialization of arbitrary code at run-time, in the style of partial evaluation. It also enables the JIT compiler to report warnings and errors to the program when it is unable to compile a code path in the demanded way. Second, allow the JIT compiler to call back into the program to perform compile-time computation. This lets the program itself define the translation strategy for certain constructs on the fly and gives rise to a powerful JIT macro facility that enables "smart" libraries to supply domain-specific compiler optimizations or safety checks. We present Lancet, a JIT compiler framework for Java bytecode that enables such a tight, two-way integration with the running program. Lancet itself was derived from a high-level Java bytecode interpreter: staging the interpreter using LMS (Lightweight Modular Staging) produced a simple bytecode compiler. Adding abstract interpretation turned the simple compiler into an optimizing compiler. This fact provides compelling evidence for the scalability of the staged-interpreter approach to compiler construction. In the case of Lancet, JIT macros also provide a natural interface to existing LMS-based toolchains such as the Delite parallelism and DSL framework, which can now serve as accelerator macros for arbitrary JVM bytecode.

Assoc Computing Machinery2014