8080939: ForkJoinPool and Phaser deadlock
8044616: Clients of Unsafe.compareAndSwapLong need to beware of using direct stores to the same field
8071638: [JAVADOC] Buggy example in javadoc for afterExecute to access a submitted job's Throwable
8043743: Data missed in java.util.concurrent.LinkedTransferQueue
8054446: Repeated offer and remove on ConcurrentLinkedQueue lead to an OutOfMemoryError
8031374: TEST_BUG: java/util/concurrent/ConcurrentQueues/OfferRemoveLoops.java fails Intermittently
8034208: Cleanup to test/java/util/concurrent/BlockingQueue/Interrupt.java
8035661: Test fix java/util/concurrent/ConcurrentQueues/OfferRemoveLoops.java from jsr166 CVS
8062841: ConcurrentHashMap.computeIfAbsent stuck in an endless loop
8073208: javadoc typo in java.util.concurrent.Executor
8073704: FutureTask.isDone returns true when task has not yet completed
8037093: java/util/concurrent/locks/Lock/TimedAcquireLeak.java fails intermittently
8022642: ScheduledThreadPoolExecutor with zero corePoolSize create endlessly threads
8065320: Busy loop in ThreadPoolExecutor.getTask for ScheduledThreadPoolExecutor
8129861: High processor load for ScheduledThreadPoolExecutor with 0 core threads
8051859: ScheduledExecutorService.scheduleWithFixedDelay fails with max delay
7146994: example afterExecute for ScheduledThreadPoolExecutor hangs
Reviewed-by: martin, psandoz, chegar
All of the hashing based Map implementations: HashMap, Hashtable, LinkedHashMap, WeakHashMap and ConcurrentHashMap are modified to use an enhanced hashing algorithm for string keys when the capacity of the hash table has ever grown beyond 512 entries. The enhanced hashing implementation uses the murmur3 hashing algorithm along with random hash seeds and index masks. These enhancements mitigate cases where colliding String hash values could result in a performance bottleneck.
Reviewed-by: alanb, forax, dl
Extend techniques developed for ConcurrentLinkedQueue and LinkedTransferQueue to implement a non-blocking concurrent Deque with interior removes.
Reviewed-by: martin, dholmes, chegar
LinkedList of size N creates N+1 instead of N+2 objects. Comparing against null is faster than comparing against sentinel node
Reviewed-by: dl, jjb, forax
6815766: LinkedBlockingQueue's iterator can return null if drainTo(c) executes concurrently
Faster, more correct. Use self-linking trick to avoid gc retention
Reviewed-by: martin, dholmes
