Class NonBlockingHashMap<TypeK,TypeV>

java.lang.Object
java.util.AbstractMap<TypeK,TypeV>
org.jctools.maps.NonBlockingHashMap<TypeK,TypeV>
Type Parameters:
TypeK - the type of keys maintained by this map
TypeV - the type of mapped values
All Implemented Interfaces:
Serializable, Cloneable, ConcurrentMap<TypeK,TypeV>, Map<TypeK,TypeV>

public class NonBlockingHashMap<TypeK,TypeV> extends AbstractMap<TypeK,TypeV> implements ConcurrentMap<TypeK,TypeV>, Cloneable, Serializable
A lock-free alternate implementation of ConcurrentHashMap with better scaling properties and generally lower costs to mutate the Map. It provides identical correctness properties as ConcurrentHashMap. All operations are non-blocking and multi-thread safe, including all update operations. NonBlockingHashMap scales substantially better than ConcurrentHashMap for high update rates, even with a large concurrency factor. Scaling is linear up to 768 CPUs on a 768-CPU Azul box, even with 100% updates or 100% reads or any fraction in-between. Linear scaling up to all cpus has been observed on a 32-way Sun US2 box, 32-way Sun Niagara box, 8-way Intel box and a 4-way Power box. This class obeys the same functional specification as Hashtable, and includes versions of methods corresponding to each method of Hashtable. However, even though all operations are thread-safe, operations do not entail locking and there is not any support for locking the entire table in a way that prevents all access. This class is fully interoperable with Hashtable in programs that rely on its thread safety but not on its synchronization details.

Operations (including put) generally do not block, so may overlap with other update operations (including other puts and removes). Retrievals reflect the results of the most recently completed update operations holding upon their onset. For aggregate operations such as putAll, concurrent retrievals may reflect insertion or removal of only some entries. Similarly, Iterators and Enumerations return elements reflecting the state of the hash table at some point at or since the creation of the iterator/enumeration. They do not throw ConcurrentModificationException. However, iterators are designed to be used by only one thread at a time.

Very full tables, or tables with high re-probe rates may trigger an internal resize operation to move into a larger table. Resizing is not terribly expensive, but it is not free either; during resize operations table throughput may drop somewhat. All threads that visit the table during a resize will 'help' the resizing but will still be allowed to complete their operation before the resize is finished (i.e., a simple 'get' operation on a million-entry table undergoing resizing will not need to block until the entire million entries are copied).

This class and its views and iterators implement all of the optional methods of the Map and Iterator interfaces.

Like Hashtable but unlike HashMap, this class does not allow null to be used as a key or value.

Since:
1.5
See Also:
  • Field Details

    • serialVersionUID

      private static final long serialVersionUID
      See Also:
    • REPROBE_LIMIT

      private static final int REPROBE_LIMIT
      See Also:
    • _Obase

      private static final int _Obase
    • _Oscale

      private static final int _Oscale
    • _Olog

      private static final int _Olog
    • _kvs_offset

      private static final long _kvs_offset
    • _kvs

      private transient Object[] _kvs
    • _last_resize_milli

      private transient long _last_resize_milli
    • MIN_SIZE_LOG

      private static final int MIN_SIZE_LOG
      See Also:
    • MIN_SIZE

      private static final int MIN_SIZE
      See Also:
    • NO_MATCH_OLD

      private static final Object NO_MATCH_OLD
    • MATCH_ANY

      private static final Object MATCH_ANY
    • TOMBSTONE

      public static final Object TOMBSTONE
    • TOMBPRIME

      private static final NonBlockingHashMap.Prime TOMBPRIME
    • _reprobes

      private transient ConcurrentAutoTable _reprobes
    • DUMMY_VOLATILE

      static volatile int DUMMY_VOLATILE
  • Constructor Details

    • NonBlockingHashMap

      public NonBlockingHashMap()
      Create a new NonBlockingHashMap with default minimum size (currently set to 8 K/V pairs or roughly 84 bytes on a standard 32-bit JVM).
    • NonBlockingHashMap

      public NonBlockingHashMap(int initial_sz)
      Create a new NonBlockingHashMap with initial room for the given number of elements, thus avoiding internal resizing operations to reach an appropriate size. Large numbers here when used with a small count of elements will sacrifice space for a small amount of time gained. The initial size will be rounded up internally to the next larger power of 2.
  • Method Details

    • rawIndex

      private static long rawIndex(Object[] ary, int idx)
    • CAS_kvs

      private final boolean CAS_kvs(Object[] oldkvs, Object[] newkvs)
    • hash

      private static final int hash(Object key)
    • chm

      private static final NonBlockingHashMap.CHM chm(Object[] kvs)
    • hashes

      private static final int[] hashes(Object[] kvs)
    • len

      private static final int len(Object[] kvs)
    • key

      private static final Object key(Object[] kvs, int idx)
    • val

      private static final Object val(Object[] kvs, int idx)
    • CAS_key

      private static final boolean CAS_key(Object[] kvs, int idx, Object old, Object key)
    • CAS_val

      private static final boolean CAS_val(Object[] kvs, int idx, Object old, Object val)
    • print

      public final void print()
      Verbose printout of table internals, useful for debugging.
    • print

      private final void print(Object[] kvs)
    • print2

      private final void print2(Object[] kvs)
    • reprobes

      public long reprobes()
      Get and clear the current count of reprobes. Reprobes happen on key collisions, and a high reprobe rate may indicate a poor hash function or weaknesses in the table resizing function.
      Returns:
      the count of reprobes since the last call to reprobes() or since the table was created.
    • reprobe_limit

      private static int reprobe_limit(int len)
    • initialize

      private final void initialize(int initial_sz)
    • initialize

      protected final void initialize()
    • size

      public int size()
      Returns the number of key-value mappings in this map.
      Specified by:
      size in interface Map<TypeK,TypeV>
      Overrides:
      size in class AbstractMap<TypeK,TypeV>
      Returns:
      the number of key-value mappings in this map
    • isEmpty

      public boolean isEmpty()
      Returns size() == 0.
      Specified by:
      isEmpty in interface Map<TypeK,TypeV>
      Overrides:
      isEmpty in class AbstractMap<TypeK,TypeV>
      Returns:
      size() == 0
    • containsKey

      public boolean containsKey(Object key)
      Tests if the key in the table using the equals method.
      Specified by:
      containsKey in interface Map<TypeK,TypeV>
      Overrides:
      containsKey in class AbstractMap<TypeK,TypeV>
      Returns:
      true if the key is in the table using the equals method
      Throws:
      NullPointerException - if the specified key is null
    • contains

      public boolean contains(Object val)
      Legacy method testing if some key maps into the specified value in this table. This method is identical in functionality to containsValue(java.lang.Object), and exists solely to ensure full compatibility with class Hashtable, which supported this method prior to introduction of the Java Collections framework.
      Parameters:
      val - a value to search for
      Returns:
      true if this map maps one or more keys to the specified value
      Throws:
      NullPointerException - if the specified value is null
    • put

      public TypeV put(TypeK key, TypeV val)
      Maps the specified key to the specified value in the table. Neither key nor value can be null.

      The value can be retrieved by calling get(java.lang.Object) with a key that is equal to the original key.

      Specified by:
      put in interface Map<TypeK,TypeV>
      Overrides:
      put in class AbstractMap<TypeK,TypeV>
      Parameters:
      key - key with which the specified value is to be associated
      val - value to be associated with the specified key
      Returns:
      the previous value associated with key, or null if there was no mapping for key
      Throws:
      NullPointerException - if the specified key or value is null
    • putIfAbsent

      public TypeV putIfAbsent(TypeK key, TypeV val)
      Atomically, do a put(TypeK, TypeV) if-and-only-if the key is not mapped. Useful to ensure that only a single mapping for the key exists, even if many threads are trying to create the mapping in parallel.
      Specified by:
      putIfAbsent in interface ConcurrentMap<TypeK,TypeV>
      Specified by:
      putIfAbsent in interface Map<TypeK,TypeV>
      Returns:
      the previous value associated with the specified key, or null if there was no mapping for the key
      Throws:
      NullPointerException - if the specified key or value is null
    • remove

      public TypeV remove(Object key)
      Removes the key (and its corresponding value) from this map. This method does nothing if the key is not in the map.
      Specified by:
      remove in interface Map<TypeK,TypeV>
      Overrides:
      remove in class AbstractMap<TypeK,TypeV>
      Returns:
      the previous value associated with key, or null if there was no mapping for key
      Throws:
      NullPointerException - if the specified key is null
    • remove

      public boolean remove(Object key, Object val)
      Atomically do a remove(Object) if-and-only-if the key is mapped to a value which is equals to the given value.
      Specified by:
      remove in interface ConcurrentMap<TypeK,TypeV>
      Specified by:
      remove in interface Map<TypeK,TypeV>
      Throws:
      NullPointerException - if the specified key or value is null
    • replace

      public TypeV replace(TypeK key, TypeV val)
      Atomically do a put(key,val) if-and-only-if the key is mapped to some value already.
      Specified by:
      replace in interface ConcurrentMap<TypeK,TypeV>
      Specified by:
      replace in interface Map<TypeK,TypeV>
      Throws:
      NullPointerException - if the specified key or value is null
    • replace

      public boolean replace(TypeK key, TypeV oldValue, TypeV newValue)
      Atomically do a put(key,newValue) if-and-only-if the key is mapped a value which is equals to oldValue.
      Specified by:
      replace in interface ConcurrentMap<TypeK,TypeV>
      Specified by:
      replace in interface Map<TypeK,TypeV>
      Throws:
      NullPointerException - if the specified key or value is null
    • objectsEquals

      private static boolean objectsEquals(Object a, Object b)
    • putIfMatchAllowNull

      public final TypeV putIfMatchAllowNull(Object key, Object newVal, Object oldVal)
    • putIfMatch

      public final TypeV putIfMatch(Object key, Object newVal, Object oldVal)
      Atomically replace newVal for oldVal, returning the value that existed there before. If the oldVal matches the returned value, then newVal was inserted, otherwise not.
      Returns:
      the previous value associated with the specified key, or null if there was no mapping for the key
      Throws:
      NullPointerException - if the key or either value is null
    • putAll

      public void putAll(Map<? extends TypeK,? extends TypeV> m)
      Copies all of the mappings from the specified map to this one, replacing any existing mappings.
      Specified by:
      putAll in interface Map<TypeK,TypeV>
      Overrides:
      putAll in class AbstractMap<TypeK,TypeV>
      Parameters:
      m - mappings to be stored in this map
    • clear

      public void clear()
      Removes all of the mappings from this map.
      Specified by:
      clear in interface Map<TypeK,TypeV>
      Overrides:
      clear in class AbstractMap<TypeK,TypeV>
    • containsValue

      public boolean containsValue(Object val)
      Returns true if this Map maps one or more keys to the specified value. Note: This method requires a full internal traversal of the hash table and is much slower than containsKey(java.lang.Object).
      Specified by:
      containsValue in interface Map<TypeK,TypeV>
      Overrides:
      containsValue in class AbstractMap<TypeK,TypeV>
      Parameters:
      val - value whose presence in this map is to be tested
      Returns:
      true if this map maps one or more keys to the specified value
      Throws:
      NullPointerException - if the specified value is null
    • rehash

      protected void rehash()
    • clone

      public Object clone()
      Creates a shallow copy of this hashtable. All the structure of the hashtable itself is copied, but the keys and values are not cloned. This is a relatively expensive operation.
      Overrides:
      clone in class AbstractMap<TypeK,TypeV>
      Returns:
      a clone of the hashtable.
    • toString

      public String toString()
      Returns a string representation of this map. The string representation consists of a list of key-value mappings in the order returned by the map's entrySet view's iterator, enclosed in braces ("{}"). Adjacent mappings are separated by the characters ", " (comma and space). Each key-value mapping is rendered as the key followed by an equals sign ("=") followed by the associated value. Keys and values are converted to strings as by String.valueOf(Object).
      Overrides:
      toString in class AbstractMap<TypeK,TypeV>
      Returns:
      a string representation of this map
    • keyeq

      private static boolean keyeq(Object K, Object key, int[] hashes, int hash, int fullhash)
    • get

      public TypeV get(Object key)
      Returns the value to which the specified key is mapped, or null if this map contains no mapping for the key.

      More formally, if this map contains a mapping from a key k to a value v such that key.equals(k), then this method returns v; otherwise it returns null. (There can be at most one such mapping.)

      Specified by:
      get in interface Map<TypeK,TypeV>
      Overrides:
      get in class AbstractMap<TypeK,TypeV>
      Throws:
      NullPointerException - if the specified key is null
    • get_impl

      private static final Object get_impl(NonBlockingHashMap topmap, Object[] kvs, Object key)
    • getk

      public TypeK getk(TypeK key)
      Returns the Key to which the specified key is mapped, or null if this map contains no mapping for the key.
      Throws:
      NullPointerException - if the specified key is null
    • getk_impl

      private static final Object getk_impl(NonBlockingHashMap topmap, Object[] kvs, Object key)
    • putIfMatch0

      private static final Object putIfMatch0(NonBlockingHashMap topmap, Object[] kvs, Object key, Object putval, Object expVal)
      Put, Remove, PutIfAbsent, etc. Return the old value. If the returned value is equal to expVal (or expVal is NO_MATCH_OLD) then the put can be assumed to work (although might have been immediately overwritten). Only the path through copy_slot passes in an expected value of null, and putIfMatch only returns a null if passed in an expected null.
      Parameters:
      topmap - the map to act on
      kvs - the KV table snapshot we act on
      key - not null (will result in NullPointerException)
      putval - the new value to use. Not null. TOMBSTONE will result in deleting the entry.
      expVal - expected old value. Can be null. NO_MATCH_OLD for an unconditional put/remove. TOMBSTONE if we expect old entry to not exist(null/TOMBSTONE value). MATCH_ANY will ignore the current value, but only if an entry exists. A null expVal is used internally to perform a strict insert-if-never-been-seen-before operation.
      Returns:
      TOMBSTONE if key does not exist or match has failed. null if expVal is null AND old value was null. Otherwise the old entry value (not null).
    • help_copy

      private final Object[] help_copy(Object[] helper)
    • raw_array

      public Object[] raw_array()
    • elements

      public Enumeration<TypeV> elements()
      Returns an enumeration of the values in this table.
      Returns:
      an enumeration of the values in this table
      See Also:
    • values

      public Collection<TypeV> values()
      Returns a Collection view of the values contained in this map. The collection is backed by the map, so changes to the map are reflected in the collection, and vice-versa. The collection supports element removal, which removes the corresponding mapping from this map, via the Iterator.remove, Collection.remove, removeAll, retainAll, and clear operations. It does not support the add or addAll operations.

      The view's iterator is a "weakly consistent" iterator that will never throw ConcurrentModificationException, and guarantees to traverse elements as they existed upon construction of the iterator, and may (but is not guaranteed to) reflect any modifications subsequent to construction.

      Specified by:
      values in interface Map<TypeK,TypeV>
      Overrides:
      values in class AbstractMap<TypeK,TypeV>
    • keys

      public Enumeration<TypeK> keys()
      Returns an enumeration of the keys in this table.
      Returns:
      an enumeration of the keys in this table
      See Also:
    • keySet

      public Set<TypeK> keySet()
      Returns a Set view of the keys contained in this map. The set is backed by the map, so changes to the map are reflected in the set, and vice-versa. The set supports element removal, which removes the corresponding mapping from this map, via the Iterator.remove, Set.remove, removeAll, retainAll, and clear operations. It does not support the add or addAll operations.

      The view's iterator is a "weakly consistent" iterator that will never throw ConcurrentModificationException, and guarantees to traverse elements as they existed upon construction of the iterator, and may (but is not guaranteed to) reflect any modifications subsequent to construction.

      Specified by:
      keySet in interface Map<TypeK,TypeV>
      Overrides:
      keySet in class AbstractMap<TypeK,TypeV>
    • entrySet

      public Set<Map.Entry<TypeK,TypeV>> entrySet()
      Returns a Set view of the mappings contained in this map. The set is backed by the map, so changes to the map are reflected in the set, and vice-versa. The set supports element removal, which removes the corresponding mapping from the map, via the Iterator.remove, Set.remove, removeAll, retainAll, and clear operations. It does not support the add or addAll operations.

      The view's iterator is a "weakly consistent" iterator that will never throw ConcurrentModificationException, and guarantees to traverse elements as they existed upon construction of the iterator, and may (but is not guaranteed to) reflect any modifications subsequent to construction.

      Warning: the iterator associated with this Set requires the creation of Map.Entry objects with each iteration. The NonBlockingHashMap does not normally create or using Map.Entry objects so they will be created soley to support this iteration. Iterating using Map.keySet() will be more efficient.

      Specified by:
      entrySet in interface Map<TypeK,TypeV>
      Specified by:
      entrySet in class AbstractMap<TypeK,TypeV>
    • writeObject

      private void writeObject(ObjectOutputStream s) throws IOException
      Throws:
      IOException
    • readObject

      private void readObject(ObjectInputStream s) throws IOException, ClassNotFoundException
      Throws:
      IOException
      ClassNotFoundException