Progress with Progress Guarantees - PowerPoint PPT Presentation

progress with progress guarantees l.
Skip this Video
Loading SlideShow in 5 Seconds..
Progress with Progress Guarantees PowerPoint Presentation
Progress with Progress Guarantees

play fullscreen
1 / 36
Download Presentation
Views
Download Presentation

Progress with Progress Guarantees

Presentation Transcript

  1. Progress with Progress Guarantees Erez Petrank - Technion Based on joint work with Anastasia Braginsky, Alex Kogan, MadanlalMusuvathi, FilipPizlo, and BjarneSteensgaard.

  2. Responsive Parallelism • Develop responsive parallel systems by developing stronger algorithms and better system support. • Main tool: progress guarantees. • Guaranteed responsiveness, good scalability, avoidingdeadlocks and livelocks. • Particularly important in several domains • real-time systems • Interactive systems (also OS’s) • operating under service level agreement. • Always good to have • If performance is not harmed too much.

  3. A Progress Guarantee • Intuitively: “ No matter which interleaving is scheduled, my program will make progress. ” • “Progress” is something the developer defines. • Specific lines of code

  4. Progress Guarantees Great guarantee, but difficult to achieve. • Wait-Freedom • If you schedule enough steps of any thread, it will make progress. • Lock-Freedom • If you schedule enough steps across all threads, one of them will make progress. The middle way ? • Obstruction-Freedom • If you let any thread run alone for enough steps, it will make progress. Somewhat weak.

  5. This Talk • Advances with progress guarantees. • Some of it is work-in-progress. • Attempts for lock-free garbage collection [PLDI 08, ISMM 07] • Bounded lock-freedom and lock-free system services [PLDI 09] • A lock-free locality-conscious linked-list [ICDCN 11] • A Lock-free balanced tree [submitted] • A wait-free queue [PPOPP 11] • Making wait-freedom fast [submitted]

  6. A Lock-Free Garbage Collector • There is no such thing right now. • A recent line of work on real-time GC, allows moving objects concurrently (while the program accesses them). • See Pizlo, Petrank, and Steensgaard at [ISMM 2007, PLDI 2008]. What does it mean to “support” lock-freedom?

  7. Services Supporting Lock-Freedom

  8. Services Supporting Lock-Freedom • Consider system services: event counters,memory management, micro-code interpreters, caching, paging, etc. • Normally, designs of lock-free algorithms assume that the underlying system operations do not foil lock-freedom.

  9. Can We Trust Services to Support Lock-Freedom ? • Valois’ lock-free linked-list algorithm has a well known C++ implementation, which uses the C++ new operation. • A hardware problem: lock Free algorithms typically use CAS or LL/SC, but LL/SC implementations are typically weak: spurious failures are permitted. • Background threads increase the chaos (if syncing on shared variables). Conclusion: system support matters.

  10. In the Paper • Definition of a lock-free supporting service (including GC), • Definition of a lock-free program that uses services, • A composition theorem for the two. • Also: bounded lock-freedom Musuvathi, Petrank, and Steensgaard. Progress Guarantee via Bounded Lock-Freedom. [PLDI 2009]

  11. Open Questions for Progress Guarantees • Some important lock-free data structures are not known • Can we build a balanced tree? • Wait-free data structures are difficult to design. • Can we design some basic ones? • Wait-free implementations are slow. • Can we make them fast?

  12. A Locality-Conscious Linked-List [ICDCN’11]A Lock-Free B-Tree [Submitted] Anastasia Braginsky & Erez Petrank

  13. A B-Tree • An important instance of balanced trees, suitable for file systems. Typically a large node with many children. • Fast access to the leaves in a very short traversal. • A node is split or merged when it becomes too dense or sparse. 9 52 92 3 7 9 26 31 40 52 63 77 89 92

  14. Lock-Free Locality-Conscious Linked Lists • List of constant size ''chunks", with minimal and maximal bounds on the number of elements contained. • Each chunk consists of a small linked list. • When a chunk gets too sparse or dense, it is split or merged with its preceding chunk. • Lock-free, locality-conscious, fast access, scalable 3 7 9 12 18 25 26 31 40 52 63 77 89 92

  15. Chunk Split or Merge • A chunk that needs splitting or merging is frozen. • No operations are applied on it anymore. • New chunks (with the updated data) replace it .

  16. A list Structure (2 Chunks) NULL HEAD NextChunk NextChunk Chunk B Chunk A EntriesHead EntriesHead Key: 89 Data: M Key: 67 Data: D Key: 25 Data: A Key: 3 Data: G Key: 14 Data: K

  17. When No More Space for Insertion NULL HEAD Freeze NextChunk NextChunk Chunk B Chunk A EntriesHead EntriesHead Key: 89 Data: M Key: 67 Data: D Key: 25 Data: A Key: 3 Data: G Key: 6 Data: B Key: 14 Data: K Key: 9 Data: C Key: 12 Data: H

  18. Split NULL HEAD Freeze NextChunk NextChunk Chunk B Chunk A EntriesHead EntriesHead Key: 89 Data: M Key: 67 Data: D Key: 25 Data: A Key: 3 Data: G Key: 6 Data: B Key: 14 Data: K Key: 9 Data: C Key: 12 Data: H NextChunk NextChunk Chunk C Chunk D EntriesHead EntriesHead Key: 3 Data: G Key: 6 Data: B Key: 9 Data: C Key: 14 Data: K Key: 12 Data: H

  19. Split NULL HEAD Freeze NextChunk NextChunk Chunk B Chunk A EntriesHead EntriesHead Key: 89 Data: M Key: 67 Data: D Key: 25 Data: A Key: 3 Data: G Key: 6 Data: B Key: 14 Data: K Key: 9 Data: C Key: 12 Data: H NextChunk NextChunk Chunk C Chunk D EntriesHead EntriesHead Key: 3 Data: G Key: 6 Data: B Key: 9 Data: C Key: 14 Data: K Key: 12 Data: H

  20. When a Chunk Gets Sparse NULL HEAD Freeze slave NextChunk NextChunk Chunk C Chunk B EntriesHead EntriesHead Key: 3 Data: G Key: 6 Data: B Key: 9 Data: C Key: 89 Data: M Key: 67 Data: D Key: 25 Data: A NextChunk Chunk D EntriesHead Key: 14 Data: K Freeze master

  21. Merge NULL HEAD Freeze slave NextChunk NextChunk Chunk C Chunk B EntriesHead EntriesHead Key: 3 Data: G Key: 6 Data: B Key: 9 Data: C Key: 89 Data: M Key: 67 Data: D Key: 25 Data: A NextChunk Chunk E NextChunk Chunk D EntriesHead EntriesHead Key: 3 Data: G Key: 6 Data: B Key: 14 Data: K Key: 9 Data: C Key: 14 Data: K Freeze master

  22. Merge NULL HEAD Freeze slave NextChunk NextChunk Chunk C Chunk B EntriesHead EntriesHead Key: 3 Data: G Key: 6 Data: B Key: 9 Data: C Key: 89 Data: M Key: 67 Data: D Key: 25 Data: A NextChunk Chunk E NextChunk Chunk D EntriesHead EntriesHead Key: 3 Data: G Key: 6 Data: B Key: 14 Data: K Key: 9 Data: C Key: 14 Data: K Freeze master

  23. Extending to a Lock-Free B-Tree • Each node holds a chunk, handling splits and merges. • Design simplified by a stringent methodology. • Monotonic life span of a node: infant, normal, frozen, & reclaimed, reduces the variety of possible states. • Care with replacement of old nodes with new ones (while data in both old and new node simultaneously). • Care with search for a neighboring node to merge with and with coordinating the merge.

  24. Having built a lock-free B-Tree, let’s look at a … Wait-Free Queue Kogan and Petrank PPOPP’10

  25. FIFO queues enqueue dequeue 3 2 9 A fundamental and ubiquitous data structures

  26. Existing wait-free queues • There exist universal constructions • but are too costly in time & space. • There exist constructions with limited concurrency • [Lamport 83] one enqueuer and one dequeuer. • [David 04] multiple enqueuers, one concurrent dequeuer. • [Jayanti & Petrovic 05] multiple dequeuers, one concurrent enqueuer.

  27. A Wait-Free Queue “Wait-Free Queues With Multiple Enqueuers and Dequeuers”. Kogan & Petrank [PPOPP 2011] • First wait-free queue (which is dynamic, parallel, and based on CASes). • We extend the lock-free design of Michael & Scott. • part of Java Concurrency package

  28. Building a Wait-Free Data Structure The first makes it inefficient. The second makes it hard to design. • Universal construction skeleton: • Publish an operation before starting execution. • Next, help delayed threads and then start executing operation. • Eventually a stuck thread succeeds because all threads help it. • Solve for each specific data structures: • The interaction between threads running the same operation. • In particular, apply each operation exactly once and obtain a consistent return code.

  29. Results • The naïve wait-free queue is 3x slower than the lock-free one. • Optimizations can reduce the ratio to 2x. Can we eliminate the overhead for the wait-free guarantee?

  30. Wait-free algorithms are typically slower than lock-free (but guarantee more). Can we eliminate the overhead completely ? Kogan and Petrank [submitted]

  31. Reducing the Help Overhead • Standard method: Help all previously registered operations before executing own operation. • But this goes over all threads. • Alternative: help only one thread (cyclically). • This is still wait-free ! An operation can only be interrupted a limited number of times. • Trade-off between efficiency and guarantee. But this is still costly !

  32. Why is Wait-Free Slow ? • Because it needs to spend a lot of time on helping others. • Main idea: typically help is not needed. • Ask for help when you need it; • Provide help infrequently. • Teacher: why are you late for school again? • Boy: I helped an old lady cross the street. • Teacher: why did it take so long? • Boy: because she didn’t want to cross!

  33. Fast-Path Slow-Path • When executing an operation, start by running the fast lock-free implementation. • Upon several failures “switch” to the wait-free implementation. • Ask for help, • Keep trying. • Once in a while, threads on the fast path check if help is needed and provide help. Fast path Slow path The ability to switch between modes

  34. Do I need to help ? yes Start Help Someone Apply my op fast path (at most n times) no Success? Apply my op using slow path no yes Return

  35. Results for the Queue • There is no observable overhead ! • Implication: it is possible to avoid starvation at no additional cost. • If you can create a fast- and slow- path and can switch between them.

  36. Conclusion • Some advances on lock-free garbage collection. • Lock-free services, bounded lock-freedom. • A few new important data structures with progress guarantees: • Lock-free chunked linked-list • Lock-free B-Tree • Wait-free Queue • We have proposed a methodology to make wait-freedom as fast as lock-freedom.