Practical lock-freedom
By Keir Fraser. 2003
643 citations recorded by the source corpus.
View Paper: https://www.semanticscholar.org/paper/88085676d6f9c1903d25eac49d4eb2a9e324d71f
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
- Håkan Grahn
- T. Harris
- J. Larus
- Ravi Rajwar
- Wenting Zheng
- Simon David Marlow
- Stephen Tu
- Samuel Madden
Citations in the corpus
- Composable memory transactions (2005, background): “There are no lock-induced deadlocks (because there are no locks); there is no priority inversion; and there is no painful tension between granularity and concurrency.”
- Hybrid transactional memory (2006, unknown): “[7] and variants on it [3, 10].”
- Transactional Memory, 2nd edition (2010, background): “, Fraser reports a problem with a transaction looping while traversing a particular kind of red-black tree implementation [105].”
- Parallel Programmability and the Chapel Language (2007, background): “Atomic sections have been of recent interest to the software transactional memory community due to their intuitiveness and ability to replace explicit lock management in user code [33, 32, 1, 26].”
- Transactional memory (2010, background): “Object-Based Software Transactional Memory (OSTM) [36] is the first software transactional memory combining lock-free synchronization and object-based conflict detection granularity.”
- Speedy transactions in multicore in-memory databases (2013, background): “Rather than reference count these objects (which would require that all accesses write to shared memory), Silo leverages its epochs into an epoch-based reclamation scheme à la read-copy update (RCU) [11, 13].”
- Cache craftiness for fast multicore key-value storage (2012, background): “It then writes the next _ layer pointer, and finally marks the key as a LAYER .”
- Advanced contention management for dynamic software transactional memory (2005, background): “The tradeoff between visible and invisible reads remains application-speci.c: visible reads reduce the overhead for incremental validation when opening new objects, but the requisite bookkeeping exacerbates contention for the memory interconnect.”