Performance implications of dynamic memory allocators on transactional memory systems
Alexandro Baldassin, Edson Borin, Guido Costa Souza De Araujo
Abstract
Alexandro Baldassin, Edson Borin, Guido Costa Souza De Araujo
Abstract
Although dynamic memory management accounts for a significant part of the execution time on many modern software systems, its impact on the performance of transactional memory systems has been mostly overlooked. In order to shed some light into this subject, this paper conducts a thorough investigation of the interplay between memory allocators and software transactional memory (STM) systems. We show that allocators can interfere with the way memory addresses are mapped to versioned locks on state-of-the-art software transactional memory implementations. Moreover, we observed that key aspects of allocators such as false sharing avoidance, scalability, and locality have a drastic impact on the final performance. For instance, we have detected performance differences of up to 171% in the STAMP applications when using distinct allocators. Moreover, we show that optimizations at the STM-level (such as caching transactional objects) are not effective when a modern allocator is already in use. All in all, our study highlights the importance of reporting the allocator utilized in the performance evaluation of transactional memory systems.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Although dynamic memory management accounts for a significant part of the execution time on many modern software systems, its impact on the performance of transactional memory systems has been mostly overlooked. In order to shed some light into this subject, this paper conducts a thorough investigation of the interplay between memory allocators and software transactional memory (STM) systems. We show that allocators can interfere with the way memory addresses are mapped to versioned locks on state-of-the-art software transactional memory implementations. Moreover, we observed that key aspects of allocators such as false sharing avoidance, scalability, and locality have a drastic impact on the final performance. For instance, we have detected performance differences of up to 171% in the STAMP applications when using distinct allocators. Moreover, we show that optimizations at the STM-level (such as caching transactional objects) are not effective when a modern allocator is already in use. All in all, our study highlights the importance of reporting the allocator utilized in the performance evaluation of transactional memory systems.
Key concepts: Allocator, Computer science, Transactional memory, Atomicity, Software transactional memory, C dynamic memory allocation, Memory management, Scalability