Reducing Feedback Pollution
Open Access
- 17 October 2024
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 65-74
Abstract
Just-in-time compilers enhance the performance of future invocations of a function by generating code tailored to past behavior. To achieve this, compilers use a data structure, often referred to as a feedback vector, to record information about each function’s invocations. However, over time, feedback vectors tend to become less precise, leading to lower-quality code – a phenomenon known as feedback vector pollution. This paper examines feedback vector pollution within the context of a compiler for the R language. We provide data, discuss an approach to reduce pollution in practice, and implement a proof-of-concept implementation of this approach. The preliminary results of the implementation indicate ∼30% decrease in polluted compilations and ∼37% decrease in function pollution throughout our corpus.Keywords
Funding Information
- Czech Science Foundation (23-07580X)
- Czech Ministry of Education, Youth and Sports under program ERC-CZ (LL2325)
- NSF (CCF-1910850)
- NSF (National Science Foundation) (CNS-1925644)
- Student Summer Research Program 2024 of FIT CTU in Prague ()
This publication has 6 references indexed in Scilit:
- Contextual dispatch for function specializationProceedings of the ACM on Programming Languages, 2020
- R melts brains: an IR for first-class environments and lazy effectful arguments2019
- Self-optimizing AST interpreters2012
- Evaluating the Design of the R Language2012
- Adaptive optimization in the Jalapeno JVMACM SIGPLAN Notices, 2011
- An analysis of the dynamic behavior of JavaScript programs2010