Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.17119 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914574495645696 |
|---|---|
| author | van Kempen, Nicolas Berger, Emery D. |
| author_facet | van Kempen, Nicolas Berger, Emery D. |
| contents | Programmers using native languages such as C, C++, or Rust can implement custom memory allocation strategies to improve execution time. In their paper titled "Reconsidering Custom Memory Allocation" almost 25 years ago, Berger et al. showed that while per-class allocators provide no significant speedups over a state-of-the-art general-purpose allocator, region-based allocators can improve execution time by allocating and freeing objects in bulk. This paper revisits that work on a modern hardware platform with modern general-purpose allocators to evaluate whether their conclusions still hold. It also augments the benchmark suite with two large real-world applications (Clang and Blender), and introduces a methodology to explore the effect of memory fragmentation on locality in general-purpose allocators. Our results support and extend the original conclusions, demonstrating the locality advantages of region-based custom memory allocators. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_17119 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Reconsidering "Reconsidering Custom Memory Allocation" van Kempen, Nicolas Berger, Emery D. Programming Languages Software Engineering Programmers using native languages such as C, C++, or Rust can implement custom memory allocation strategies to improve execution time. In their paper titled "Reconsidering Custom Memory Allocation" almost 25 years ago, Berger et al. showed that while per-class allocators provide no significant speedups over a state-of-the-art general-purpose allocator, region-based allocators can improve execution time by allocating and freeing objects in bulk. This paper revisits that work on a modern hardware platform with modern general-purpose allocators to evaluate whether their conclusions still hold. It also augments the benchmark suite with two large real-world applications (Clang and Blender), and introduces a methodology to explore the effect of memory fragmentation on locality in general-purpose allocators. Our results support and extend the original conclusions, demonstrating the locality advantages of region-based custom memory allocators. |
| title | Reconsidering "Reconsidering Custom Memory Allocation" |
| topic | Programming Languages Software Engineering |
| url | https://arxiv.org/abs/2605.17119 |