This is an old revision of the document!
Microbench
Microbench is a micro-benchmark suite to compare STMs with existing lock-free and lock-based counterparts. The current version includes three kinds of integer set implementations: linked list, skip list and hash table.
Microbench includes the implementation of several algorithms:
Harris T. (2001) Lock-free Linked List: A Pragmatic Implementation of Non-Blocking Linked Lists. Proc. of the 15th Int'l Symposium on Distributed Computing (DISC). p.300-314.
Heller S., Herlihy M., Luchangco V., Moir M., Scherer III, W., Shavit N. (2005) Lock-based Linked List: A Lazy Concurrent List-Based Set Algorithm. Proc. of the 9th Int'l Conf. on Principles of Dist. Systems (OPODIS). p.3-16.
K. Fraser (2003) Lock-based Linked List: A Lazy Concurrent List-Based Set Algorithm. Cambridge University Technical Report UCAM-CL-TR-579. p.3-16.
Herlihy M., Lev Y., Luchangco V., Shavit N. (2007) Lock-based Linked List: A Lazy Concurrent List-Based Set Algorithm. Colloquim on Structural Information and Communication Complexity (SIROCCO). p.124-138.
Microbench can be used with TinySTM, SwissTM and ε-STM.
Microbench
Microbench aims at comparing STM performance against performance of lock-based and lock-free alternatives. It comprises common data structures: linked list, skip list, hashtable…
- It provides lock-free algorithms (e.g., harris-michael, fraser's lock-free skip-list).
- It also features fine-grained locking algorithms (e.g., lazy linked list, optimistic skip list).
- Finally, the TM benchmarks derive from the STM-based linked list test of TinySTM and it includes the SUN STM-based red-black tree benchmark.
The current version can be downloaded here.
