- author = "Aidan Dakhama and W. B. Langdon and Hector D. Menendez and Karine Even-Mendoza",
- title = "{GreenMalloc}: Allocator Optimisation for Industrial Workloads",
- booktitle = "Search-Based Software Engineering 2025 - Challenge Track",
- year = "2025",
- editor = "Carol Hanna and Mijung Kim and Vincenzo Riccio",
- series = "Lecture Notes in Computer Science",
- address = "Seoul, South Korea",
- month = "16 " # nov,
- publisher = "Springer Nature",
- keywords = "genetic algorithms, genetic programming, genetic improvement, SBSE, software optimisation, gnu runtime library, malloc, TCmalloc, gem5, rand_malloc",
-
URL = "
http://www.cs.ucl.ac.uk/staff/W.Langdon/ftp/papers/Dakhama_2025_SSBSE.pdf",
-
URL = "
https://conf.researchr.org/track/ssbse-2025/ssbse-2025-papers",
-
URL = "https://arxiv.org/abs/2510.21405",
-
code_url = "https://zenodo.org/records/17182847",
- size = "6 pages",
- abstract = "We present GreenMalloc, a multi-objective search-based framework for automatically configuring memory allocators. Our approach uses NSGA-II and rand_malloc \cite{Langdon:2025:UKCI} as a lightweight proxy bench-marking tool, we efficiently explore allocator parameters from execution traces and transfer the best configurations to gem5, a large system simulator, in a case study on two allocators: the GNU C/C++ compiler's glibc malloc and Google's TCMalloc. Across diverse workloads, our empirical results show up to 4.1 percent reduction in average heap usage without loss of runtime efficiency, indeed we get a 0.25 percent reduction.",
-
notes = "
Search used to optimise existing programs. https://conf.researchr.org/track/ssbse-2025/ssbse-2025-challenge",
