1. Linked Lists: Locking, Lock- Free, and Beyond … Companion slides for The Art of Multiprocessor Programming by Maurice Herlihy & Nir Shavit

2. Art of Multiprocessor Programming2 This Lecture Introduce four “patterns” –Bag of tricks … –Methods that work more than once …

3. Art of Multiprocessor Programming3 This Lecture Introduce four “patterns” –Bag of tricks … –Methods that work more than once … For highly-concurrent objects –Concurrent access –More threads, more throughput

4. Art of Multiprocessor Programming4 First: Fine-Grained Synchronization Instead of using a single lock … Split object into –Independently-synchronized components Methods conflict when they access –The same component … –At the same time

5. Art of Multiprocessor Programming5 Second: Optimistic Synchronization Search without locking … If you find it, lock and check … –OK: we are done –Oops: start over Evaluation –Usually cheaper than locking, but –Mistakes are expensive

6. Art of Multiprocessor Programming6 Third: Lazy Synchronization Postpone hard work Removing components is tricky –Logical removal Mark component to be deleted –Physical removal Do what needs to be done

7. Art of Multiprocessor Programming7 Fourth: Lock-Free Synchronization Don’t use locks at all –Use compareAndSet() & relatives … Advantages –No Scheduler Assumptions/Support Disadvantages –Complex –Sometimes high overhead

8. Art of Multiprocessor Programming8 Linked List Illustrate these patterns … Using a list-based Set –Common application –Building block for other apps

9. Art of Multiprocessor Programming9 Set Interface Unordered collection of items No duplicates Methods –add(x) put x in set –remove(x) take x out of set –contains(x) tests if x in set

10. Art of Multiprocessor Programming10 List-Based Sets public interface Set { public boolean add(T x); public boolean remove(T x); public boolean contains(T x); } Add item to set

11. Art of Multiprocessor Programming11 List-Based Sets public interface Set { public boolean add(T x); public boolean remove(T x); public boolean contains(Tt x); } Remove item from set

12. Art of Multiprocessor Programming12 List-Based Sets public interface Set { public boolean add(T x); public boolean remove(T x); public boolean contains(T x); } Is item in set?

13. Art of Multiprocessor Programming13 List Node public class Node { public T item; public int key; public Node next; }

14. Art of Multiprocessor Programming14 List Node public class Node { public T item; public int key; public Node next; } item of interest

15. Art of Multiprocessor Programming15 List Node public class Node { public T item; public int key; public Node next; } Usually hash code

16. Art of Multiprocessor Programming16 List Node public class Node { public T item; public int key; public Node next; } Reference to next node

17. Art of Multiprocessor Programming17 The List-Based Set abc Sorted with Sentinel nodes (min & max possible keys) -∞ +∞

18. Art of Multiprocessor Programming18 Reasoning about Concurrent Objects Invariant –Property that always holds Established because –True when object is created –Truth preserved by each method Each step of each method

19. Art of Multiprocessor Programming19 Specifically … Invariants preserved by –add() –remove() –contains() Most steps are trivial –Usually one step tricky –Often linearization point

20. Art of Multiprocessor Programming20 Interference Invariants make sense only if –methods considered –are the only modifiers

21. Art of Multiprocessor Programming21 Interference Invariants make sense only if –methods considered –are the only modifiers Language encapsulation helps –List nodes not visible outside class

22. Art of Multiprocessor Programming22 Interference Freedom from interference needed even for removed nodes –Some algorithms traverse removed nodes –Careful with malloc() & free() ! Garbage collection helps here

23. Art of Multiprocessor Programming23 Abstract Data Types Concrete representation: Abstract Type: –{a, b} ab

24. Art of Multiprocessor Programming24 Abstract Data Types Meaning of rep given by abstraction map –S( ) = {a,b} a b

25. Art of Multiprocessor Programming25 Rep Invariant Which concrete values meaningful? –Sorted? –Duplicates? Rep invariant –Characterizes legal concrete reps –Preserved by methods –Relied on by methods

26. Art of Multiprocessor Programming26 Blame Game Rep invariant is a contract Suppose –add() leaves behind 2 copies of x –remove() removes only 1 Which is incorrect?

27. Art of Multiprocessor Programming27 Blame Game Suppose –add() leaves behind 2 copies of x –remove() removes only 1 Which is incorrect? –If rep invariant says no duplicates add() is incorrect –Otherwise remove() is incorrect

28. Art of Multiprocessor Programming28 Rep Invariant (partly) Sentinel nodes –tail reachable from head Sorted No duplicates

29. Art of Multiprocessor Programming29 Abstraction Map S(head) = –{ x | there exists a such that a reachable from head and a.item = x –}

30. Art of Multiprocessor Programming30 Sequential List Based Set a c d a b c Add() Remove()

31. Art of Multiprocessor Programming31 Sequential List Based Set a c d b a b c Add() Remove()

32. Art of Multiprocessor Programming32 Course Grained Locking a b d

33. Art of Multiprocessor Programming33 Course Grained Locking a b d c

34. Art of Multiprocessor Programming34 honk! Course Grained Locking a b d c Simple but hotspot + bottleneck honk!

35. Art of Multiprocessor Programming35 Coarse-Grained Locking Easy, same as synchronized methods –“One lock to rule them all …”

36. Art of Multiprocessor Programming36 Coarse-Grained Locking Easy, same as synchronized methods –“One lock to rule them all …” Simple, clearly correct –Deserves respect! Works poorly with contention –Queue locks help –But bottleneck still an issue

37. Art of Multiprocessor Programming37 Fine-grained Locking Requires careful thought –“Do not meddle in the affairs of wizards, for they are subtle and quick to anger”

38. Art of Multiprocessor Programming38 Fine-grained Locking Requires careful thought –“Do not meddle in the affairs of wizards, for they are subtle and quick to anger” Split object into pieces –Each piece has own lock –Methods that work on disjoint pieces need not exclude each other

39. Art of Multiprocessor Programming39 Hand-over-Hand locking abc

40. Art of Multiprocessor Programming40 Hand-over-Hand locking abc

41. Art of Multiprocessor Programming41 Hand-over-Hand locking abc

42. Art of Multiprocessor Programming42 Hand-over-Hand locking abc

43. Art of Multiprocessor Programming43 Hand-over-Hand locking abc

44. Art of Multiprocessor Programming44 Removing a Node abcd remove(b)

45. Art of Multiprocessor Programming45 Removing a Node abcd remove(b)

46. Art of Multiprocessor Programming46 Removing a Node abcd remove(b)

47. Art of Multiprocessor Programming47 Removing a Node abcd remove(b)

48. Art of Multiprocessor Programming48 Removing a Node abcd remove(b)

49. Art of Multiprocessor Programming49 Removing a Node acd remove(b) Why hold 2 locks?

50. Art of Multiprocessor Programming50 Concurrent Removes abcd remove(c) remove(b)

51. Art of Multiprocessor Programming51 Concurrent Removes abcd remove(b) remove(c)

52. Art of Multiprocessor Programming52 Concurrent Removes abcd remove(b) remove(c)

53. Art of Multiprocessor Programming53 Concurrent Removes abcd remove(b) remove(c)

54. Art of Multiprocessor Programming54 Concurrent Removes abcd remove(b) remove(c)

55. Art of Multiprocessor Programming55 Concurrent Removes abcd remove(b) remove(c)

56. Art of Multiprocessor Programming56 Concurrent Removes abcd remove(b) remove(c)

57. Art of Multiprocessor Programming57 Concurrent Removes abcd remove(b) remove(c)

58. Art of Multiprocessor Programming58 Concurrent Removes abcd remove(b) remove(c)

59. Art of Multiprocessor Programming59 Concurrent Removes abcd remove(b) remove(c)

60. Art of Multiprocessor Programming60 Uh, Oh acd remove(b) remove(c)

61. Art of Multiprocessor Programming61 Uh, Oh acd Bad news, c not removed remove(b) remove(c)

62. Art of Multiprocessor Programming62 Problem To delete node c –Swing node b’s next field to d Problem is, –Someone deleting b concurrently could direct a pointer to c ba cbac

63. Art of Multiprocessor Programming63 Insight If a node is locked –No one can delete node’s successor If a thread locks –Node to be deleted –And its predecessor –Then it works

64. Art of Multiprocessor Programming64 Hand-Over-Hand Again abcd remove(b)

65. Art of Multiprocessor Programming65 Hand-Over-Hand Again abcd remove(b)

66. Art of Multiprocessor Programming66 Hand-Over-Hand Again abcd remove(b)

67. Art of Multiprocessor Programming67 Hand-Over-Hand Again abcd remove(b) Found it!

68. Art of Multiprocessor Programming68 Hand-Over-Hand Again abcd remove(b) Found it!

69. Art of Multiprocessor Programming69 Hand-Over-Hand Again acd remove(b)

70. Art of Multiprocessor Programming70 Removing a Node abcd remove(b) remove(c)

71. Art of Multiprocessor Programming71 Removing a Node abcd remove(b) remove(c)

72. Art of Multiprocessor Programming72 Removing a Node abcd remove(b) remove(c)

73. Art of Multiprocessor Programming73 Removing a Node abcd remove(b) remove(c)

74. Art of Multiprocessor Programming74 Removing a Node abcd remove(b) remove(c)

75. Art of Multiprocessor Programming75 Removing a Node abcd remove(b) remove(c)

76. Art of Multiprocessor Programming76 Removing a Node abcd remove(b) remove(c)

77. Art of Multiprocessor Programming77 Removing a Node abcd remove(b) remove(c)

78. Art of Multiprocessor Programming78 Removing a Node abcd Must acquire Lock of b remove(c)

79. Art of Multiprocessor Programming79 Removing a Node abcd Cannot acquire lock of b remove(c)

80. Art of Multiprocessor Programming80 Removing a Node abcd Wait! remove(c)

81. Art of Multiprocessor Programming81 Removing a Node abd Proceed to remove(b)

82. Art of Multiprocessor Programming82 Removing a Node abd remove(b)

83. Art of Multiprocessor Programming83 Removing a Node abd remove(b)

84. Art of Multiprocessor Programming84 Removing a Node ad remove(b)

85. Art of Multiprocessor Programming85 Removing a Node ad

86. Art of Multiprocessor Programming86 Remove method public boolean remove(Item item) { int key = item.hashCode(); Node pred, curr; try { … } finally { curr.unlock(); pred.unlock(); }}

87. Art of Multiprocessor Programming87 Remove method public boolean remove(Item item) { int key = item.hashCode(); Node pred, curr; try { … } finally { curr.unlock(); pred.unlock(); }} Key used to order node

88. Art of Multiprocessor Programming88 Remove method public boolean remove(Item item) { int key = item.hashCode(); Node pred, curr; try { … } finally { currNode.unlock(); predNode.unlock(); }} Predecessor and current nodes

89. Art of Multiprocessor Programming89 Remove method public boolean remove(Item item) { int key = item.hashCode(); Node pred, curr; try { … } finally { curr.unlock(); pred.unlock(); }} Make sure locks released

90. Art of Multiprocessor Programming90 Remove method public boolean remove(Item item) { int key = item.hashCode(); Node pred, curr; try { … } finally { curr.unlock(); pred.unlock(); }} Everything else

91. Art of Multiprocessor Programming91 Remove method try { pred = this.head; pred.lock(); curr = pred.next; curr.lock(); … } finally { … }

92. Art of Multiprocessor Programming92 Remove method try { pred = this.head; pred.lock(); curr = pred.next; curr.lock(); … } finally { … } lock pred == head

93. Art of Multiprocessor Programming93 Remove method try { pred = this.head; pred.lock(); curr = pred.next; curr.lock(); … } finally { … } Lock current

94. Art of Multiprocessor Programming94 Remove method try { pred = this.head; pred.lock(); curr = pred.next; curr.lock(); … } finally { … } Traversing list

95. Art of Multiprocessor Programming95 Remove: searching while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false;

96. Art of Multiprocessor Programming96 Remove: searching while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false; Search key range

97. Art of Multiprocessor Programming97 Remove: searching while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false; At start of each loop: curr and pred locked

98. Art of Multiprocessor Programming98 Remove: searching while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false; If item found, remove node

99. Art of Multiprocessor Programming99 Remove: searching while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false; Unlock predecessor

100. Art of Multiprocessor Programming100 Remove: searching while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false; Only one node locked!

101. Art of Multiprocessor Programming101 Remove: searching while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false; demote current

102. Art of Multiprocessor Programming102 Remove: searching while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = currNode; curr = curr.next; curr.lock(); } return false; Find and lock new current

103. Art of Multiprocessor Programming103 Remove: searching while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = currNode; curr = curr.next; curr.lock(); } return false; Lock invariant restored

104. Art of Multiprocessor Programming104 Remove: searching while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false; Otherwise, not present

105. Art of Multiprocessor Programming105 Why does this work? To remove node e –Must lock e –Must lock e’s predecessor Therefore, if you lock a node –It can’t be removed –And neither can its successor

106. Art of Multiprocessor Programming106 while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false; Why remove() is linearizable pred reachable from head curr is pred.next So curr.item is in the set

107. Art of Multiprocessor Programming107 while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false; Why remove() is linearizable Linearization point if item is present

108. Art of Multiprocessor Programming108 while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false; Why remove() is linearizable Node locked, so no other thread can remove it ….

109. Art of Multiprocessor Programming109 while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false; Why remove() is linearizable Item not present

110. Art of Multiprocessor Programming110 while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false; Why remove() is linearizable pred reachable from head curr is pred.next pred.key < key key < curr.key

111. Art of Multiprocessor Programming111 while (curr.key <= key) { if (item == curr.item) { pred.next = curr.next; return true; } pred.unlock(); pred = curr; curr = curr.next; curr.lock(); } return false; Why remove() is linearizable Linearization point

112. Art of Multiprocessor Programming112 Adding Nodes To add node e –Must lock predecessor –Must lock successor Neither can be deleted –(Is successor lock actually required?)

113. Art of Multiprocessor Programming113 Same Abstraction Map S(head) = –{ x | there exists a such that a reachable from head and a.item = x –}

114. Art of Multiprocessor Programming114 Rep Invariant Easy to check that –tail always reachable from head –Nodes sorted, no duplicates

115. Art of Multiprocessor Programming115 Drawbacks Better than coarse-grained lock –Threads can traverse in parallel Still not ideal –Long chain of acquire/release –Inefficient

116. Art of Multiprocessor Programming116 Optimistic Synchronization Find nodes without locking Lock nodes Check that everything is OK

117. Art of Multiprocessor Programming117 Optimistic: Traverse without Locking b d e a add(c) Aha!

118. Art of Multiprocessor Programming118 Optimistic: Lock and Load b d e a add(c)

119. Art of Multiprocessor Programming119 Optimistic: Lock and Load b d e a add(c) c

120. Art of Multiprocessor Programming120 What could go wrong? b d e a add(c) Aha!

121. Art of Multiprocessor Programming121 What could go wrong? b d e a add(c)

122. Art of Multiprocessor Programming122 What could go wrong? b d e a remove(b )

123. Art of Multiprocessor Programming123 What could go wrong? b d e a remove(b )

124. Art of Multiprocessor Programming124 What could go wrong? b d e a add(c)

125. Art of Multiprocessor Programming125 What could go wrong? b d e a add(c) c

126. Art of Multiprocessor Programming126 What could go wrong? d e a add(c) Uh-oh

127. Art of Multiprocessor Programming127 Validate – Part 1 b d e a add(c) Yes, b still reachable from head

128. Art of Multiprocessor Programming128 What Else Could Go Wrong? b d e a add(c) Aha!

129. Art of Multiprocessor Programming129 What Else Coould Go Wrong? b d e a add(c) add(b’)

130. Art of Multiprocessor Programming130 What Else Coould Go Wrong? b d e a add(c) add(b’) b’

131. Art of Multiprocessor Programming131 What Else Could Go Wrong? b d e a add(c) b’

132. Art of Multiprocessor Programming132 What Else Could Go Wrong? b d e a add(c) c

133. Art of Multiprocessor Programming133 Validate Part 2 (while holding locks) b d e a add(c) Yes, b still points to d

134. Art of Multiprocessor Programming134 Optimistic: Linearization Point b d e a add(c) c

135. Art of Multiprocessor Programming135 Same Abstraction Map S(head) = –{ x | there exists a such that a reachable from head and a.item = x –}

136. Art of Multiprocessor Programming136 Invariants Careful: we may traverse deleted nodes But we establish properties by –Validation –After we lock target nodes

137. Art of Multiprocessor Programming137 Correctness If –Nodes b and c both locked –Node b still accessible –Node c still successor to b Then –Neither will be deleted –OK to delete and return true

138. Art of Multiprocessor Programming138 Unsuccessful Remove abde remove(c) Aha!

139. Art of Multiprocessor Programming139 Validate (1) abde Yes, b still reachable from head remove(c)

140. Art of Multiprocessor Programming140 Validate (2) abde remove(c) Yes, b still points to d

141. Art of Multiprocessor Programming141 OK Computer abde remove(c) return false

142. Art of Multiprocessor Programming142 Correctness If –Nodes b and d both locked –Node b still accessible –Node d still successor to b Then –Neither will be deleted –No thread can add c after b –OK to return false

143. Art of Multiprocessor Programming143 Validation private boolean validate(Node pred, Node curry) { Node node = head; while (node.key <= pred.key) { if (node == pred) return pred.next == curr; node = node.next; } return false; }

144. Art of Multiprocessor Programming144 private boolean validate(Node pred, Node curr) { Node node = head; while (node.key <= pred.key) { if (node == pred) return pred.next == curr; node = node.next; } return false; } Validation Predecessor & current nodes

145. Art of Multiprocessor Programming145 private boolean validate(Node pred, Node curr) { Node node = head; while (node.key <= pred.key) { if (node == pred) return pred.next == curr; node = node.next; } return false; } Validation Begin at the beginning

146. Art of Multiprocessor Programming146 private boolean validate(Node pred, Node curr) { Node node = head; while (node.key <= pred.key) { if (node == pred) return pred.next == curr; node = node.next; } return false; } Validation Search range of keys

147. Art of Multiprocessor Programming147 private boolean validate(Node pred, Node curr) { Node node = head; while (node.key <= pred.key) { if (node == pred) return pred.next == curr; node = node.next; } return false; } Validation Predecessor reachable

148. Art of Multiprocessor Programming148 private boolean validate(Node pred, Node curr) { Node node = head; while (node.key <= pred.key) { if (node == pred) return pred.next == curr; node = node.next; } return false; } Validation Is current node next?

149. Art of Multiprocessor Programming149 private boolean validate(Node pred, Node curr) { Node node = head; while (node.key <= pred.key) { if (node == pred) return pred.next == curr; node = node.next; } return false; } Validation Otherwise move on

150. Art of Multiprocessor Programming150 private boolean validate(Node pred, Node curr) { Node node = head; while (node.key <= pred.key) { if (node == pred) return pred.next == curr; node = node.next; } return false; } Validation Predecessor not reachable

151. Art of Multiprocessor Programming151 Remove: searching public boolean remove(Item item) { int key = item.hashCode(); retry: while (true) { Node pred = this.head; Node curr = pred.next; while (curr.key <= key) { if (item == curr.item) break; pred = curr; curr = curr.next; } …

152. Art of Multiprocessor Programming152 public boolean remove(Item item) { int key = item.hashCode(); retry: while (true) { Node pred = this.head; Node curr = pred.next; while (curr.key <= key) { if (item == curr.item) break; pred = curr; curr = curr.next; } … Remove: searching Search key

153. Art of Multiprocessor Programming153 public boolean remove(Item item) { int key = item.hashCode(); retry: while (true) { Node pred = this.head; Node curr = pred.next; while (curr.key <= key) { if (item == curr.item) break; pred = curr; curr = curr.next; } … Remove: searching Retry on synchronization conflict

154. Art of Multiprocessor Programming154 public boolean remove(Item item) { int key = item.hashCode(); retry: while (true) { Node pred = this.head; Node curr = pred.next; while (curr.key <= key) { if (item == curr.item) break; pred = curr; curr = curr.next; } … Remove: searching Examine predecessor and current nodes

155. Art of Multiprocessor Programming155 public boolean remove(Item item) { int key = item.hashCode(); retry: while (true) { Node pred = this.head; Node curr = pred.next; while (curr.key <= key) { if (item == curr.item) break; pred = curr; curr = curr.next; } … Remove: searching Search by key

156. Art of Multiprocessor Programming156 public boolean remove(Item item) { int key = item.hashCode(); retry: while (true) { Node pred = this.head; Node curr = pred.next; while (curr.key <= key) { if (item == curr.item) break; pred = curr; curr = curr.next; } … Remove: searching Stop if we find item

157. Art of Multiprocessor Programming157 public boolean remove(Item item) { int key = item.hashCode(); retry: while (true) { Node pred = this.head; Node curr = pred.next; while (curr.key <= key) { if (item == curr.item) break; pred = curr; curr = curr.next; } … Remove: searching Move along

158. Art of Multiprocessor Programming158 On Exit from Loop If item is present –curr holds item –pred just before curr If item is absent –curr has first higher key –pred just before curr Assuming no synchronization problems

159. Art of Multiprocessor Programming159 Remove Method try { pred.lock(); curr.lock(); if (validate(pred,curr) { if (curr.item == item) { pred.next = curr.next; return true; } else { return false; }}} finally { pred.unlock(); curr.unlock(); }}}

160. Art of Multiprocessor Programming160 try { pred.lock(); curr.lock(); if (validate(pred,curr) { if (curr.item == item) { pred.next = curr.next; return true; } else { return false; }}} finally { pred.unlock(); curr.unlock(); }}} Remove Method Always unlock

161. Art of Multiprocessor Programming161 try { pred.lock(); curr.lock(); if (validate(pred,curr) { if (curr.item == item) { pred.next = curr.next; return true; } else { return false; }}} finally { pred.unlock(); curr.unlock(); }}} Remove Method Lock both nodes

162. Art of Multiprocessor Programming162 try { pred.lock(); curr.lock(); if (validate(pred,curr) { if (curr.item == item) { pred.next = curr.next; return true; } else { return false; }}} finally { pred.unlock(); curr.unlock(); }}} Remove Method Check for synchronization conflicts

163. Art of Multiprocessor Programming163 try { pred.lock(); curr.lock(); if (validate(pred,curr) { if (curr.item == item) { pred.next = curr.next; return true; } else { return false; }}} finally { pred.unlock(); curr.unlock(); }}} Remove Method target found, remove node

164. Art of Multiprocessor Programming164 try { pred.lock(); curr.lock(); if (validate(pred,curr) { if (curr.item == item) { pred.next = curr.next; return true; } else { return false; }}} finally { pred.unlock(); curr.unlock(); }}} Remove Method target not found

165. Art of Multiprocessor Programming165 Optimistic List Limited hot-spots –Targets of add(), remove(), contains() –No contention on traversals Moreover –Traversals are wait-free –Food for thought …

166. Art of Multiprocessor Programming166 So Far, So Good Much less lock acquisition/release –Performance –Concurrency Problems –Need to traverse list twice –contains() method acquires locks

167. Art of Multiprocessor Programming167 Evaluation Optimistic is effective if –cost of scanning twice without locks is less than –cost of scanning once with locks Drawback –contains() acquires locks –90% of calls in many apps

168. Art of Multiprocessor Programming168 Lazy List Like optimistic, except –Scan once –contains(x) never locks … Key insight –Removing nodes causes trouble –Do it “lazily”

169. Art of Multiprocessor Programming169 Lazy List remove() –Scans list (as before) –Locks predecessor & current (as before) Logical delete –Marks current node as removed (new!) Physical delete –Redirects predecessor’s next (as before)

170. Art of Multiprocessor Programming170 Lazy Removal aa b c d

171. c Art of Multiprocessor Programming171 Lazy Removal aa b d Present in list

172. c Art of Multiprocessor Programming172 Lazy Removal aa b d Logically deleted

173. Art of Multiprocessor Programming173 Lazy Removal aa b c d Physically deleted

174. Art of Multiprocessor Programming174 Lazy Removal aa b d Physically deleted

175. Art of Multiprocessor Programming175 Lazy List All Methods –Scan through locked and marked nodes –Removing a node doesn’t slow down other method calls … Must still lock pred and curr nodes.

176. Art of Multiprocessor Programming176 Validation No need to rescan list! Check that pred is not marked Check that curr is not marked Check that pred points to curr

177. Art of Multiprocessor Programming177 Business as Usual abc

178. Art of Multiprocessor Programming178 Business as Usual abc

179. Art of Multiprocessor Programming179 Business as Usual abc

180. Art of Multiprocessor Programming180 Business as Usual abc remove(b)

181. Art of Multiprocessor Programming181 Business as Usual abc a not marked

182. Art of Multiprocessor Programming182 Business as Usual abc a still points to b

183. Art of Multiprocessor Programming183 Business as Usual a bc Logical delete

184. Art of Multiprocessor Programming184 Business as Usual a bc physical delete

185. Art of Multiprocessor Programming185 Business as Usual a bc

186. Art of Multiprocessor Programming186 New Abstraction Map S(head) = –{ x | there exists node a such that a reachable from head and a.item = x and a is unmarked –}

187. Art of Multiprocessor Programming187 Invariant If not marked then item in the set and reachable from head and if not yet traversed it is reachable from pred

188. Art of Multiprocessor Programming188 Validation private boolean validate(Node pred, Node curr) { return !pred.marked && !curr.marked && pred.next == curr); }

189. Art of Multiprocessor Programming189 private boolean validate(Node pred, Node curr) { return !pred.marked && !curr.marked && pred.next == curr); } List Validate Method Predecessor not Logically removed

190. Art of Multiprocessor Programming190 private boolean validate(Node pred, Node curr) { return !pred.marked && !curr.marked && pred.next == curr); } List Validate Method Current not Logically removed

191. Art of Multiprocessor Programming191 private boolean validate(Node pred, Node curr) { return !pred.marked && !curr.marked && pred.next == curr); } List Validate Method Predecessor still Points to current

192. Art of Multiprocessor Programming192 Remove try { pred.lock(); curr.lock(); if (validate(pred,curr) { if (curr.key == key) { curr.marked = true; pred.next = curr.next; return true; } else { return false; }}} finally { pred.unlock(); curr.unlock(); }}}

193. Art of Multiprocessor Programming193 Remove try { pred.lock(); curr.lock(); if (validate(pred,curr) { if (curr.key == key) { curr.marked = true; pred.next = curr.next; return true; } else { return false; }}} finally { pred.unlock(); curr.unlock(); }}} Validate as before

194. Art of Multiprocessor Programming194 Remove try { pred.lock(); curr.lock(); if (validate(pred,curr) { if (curr.key == key) { curr.marked = true; pred.next = curr.next; return true; } else { return false; }}} finally { pred.unlock(); curr.unlock(); }}} Key found

195. Art of Multiprocessor Programming195 Remove try { pred.lock(); curr.lock(); if (validate(pred,curr) { if (curr.key == key) { curr.marked = true; pred.next = curr.next; return true; } else { return false; }}} finally { pred.unlock(); curr.unlock(); }}} Logical remove

196. Art of Multiprocessor Programming196 Remove try { pred.lock(); curr.lock(); if (validate(pred,curr) { if (curr.key == key) { curr.marked = true; pred.next = curr.next; return true; } else { return false; }}} finally { pred.unlock(); curr.unlock(); }}} physical remove

197. Art of Multiprocessor Programming197 Contains public boolean contains(Item item) { int key = item.hashCode(); Node curr = this.head; while (curr.key < key) { curr = curr.next; } return curr.key == key && !curr.marked; }

198. Art of Multiprocessor Programming198 Contains public boolean contains(Item item) { int key = item.hashCode(); Node curr = this.head; while (curr.key < key) { curr = curr.next; } return curr.key == key && !curr.marked; } Start at the head

199. Art of Multiprocessor Programming199 Contains public boolean contains(Item item) { int key = item.hashCode(); Node curr = this.head; while (curr.key < key) { curr = curr.next; } return curr.key == key && !curr.marked; } Search key range

200. Art of Multiprocessor Programming200 Contains public boolean contains(Item item) { int key = item.hashCode(); Node curr = this.head; while (curr.key < key) { curr = curr.next; } return curr.key == key && !curr.marked; } Traverse without locking (nodes may have been removed)

201. Art of Multiprocessor Programming201 Contains public boolean contains(Item item) { int key = item.hashCode(); Node curr = this.head; while (curr.key < key) { curr = curr.next; } return curr.key == key && !curr.marked; } Present and undeleted?

202. Art of Multiprocessor Programming202 Summary: Wait-free Contains a 0 0 0 a b c 0 e 1 d Use Mark bit + list ordering 1.Not marked  in the set 2.Marked or missing  not in the set

203. Art of Multiprocessor Programming203 Lazy List a 0 0 0 a b c 0 e 1 d Lazy add() and remove() + Wait-free contains()

204. Art of Multiprocessor Programming204 Evaluation Good: –contains() doesn’t lock –In fact, its wait-free! –Good because typically high % contains() –Uncontended calls don’t re-traverse Bad –Contended add() and remove() calls do re-traverse –Traffic jam if one thread delays

205. Art of Multiprocessor Programming205 Traffic Jam Any concurrent data structure based on mutual exclusion has a weakness If one thread –Enters critical section –And “eats the big muffin” Cache miss, page fault, descheduled … –Everyone else using that lock is stuck! –Need to trust the scheduler….

206. Art of Multiprocessor Programming206 Reminder: Lock-Free Data Structures No matter what … –Guarantees minimal progress in any execution –i.e. Some thread will always complete a method call –Even if others halt at malicious times –Implies that implementation can’t use locks

207. Art of Multiprocessor Programming207 Lock-free Lists Next logical step –Wait-free contains() –lock-free add() and remove() Use only compareAndSet() –What could go wrong?

208. Art of Multiprocessor Programming208 a 0 0 0 a b c 0 e 1 c Logical Removal = Set Mark Bit Physical Removal CAS pointer Use CAS to verify pointer is correct Not enough! Lock-free Lists

209. Art of Multiprocessor Programming209 Problem… a 0 0 0 a b c 0 e 1 c Logical Removal = Set Mark Bit Physical Removal CAS 0 d Problem: d not added to list… Must Prevent manipulation of removed node’s pointer Node added Before Physical Removal CAS

210. Art of Multiprocessor Programming210 The Solution: Combine Bit and Pointer a 0 0 0 a b c 0 e 1 c Logical Removal = Set Mark Bit Physical Removal CAS 0 d Mark-Bit and Pointer are CASed together (AtomicMarkableReference) Fail CAS: Node not added after logical Removal

211. Art of Multiprocessor Programming211 Solution Use AtomicMarkableReference Atomically –Swing reference and –Update flag Remove in two steps –Set mark bit in next field –Redirect predecessor’s pointer

212. Art of Multiprocessor Programming212 Marking a Node AtomicMarkableReference class –Java.util.concurrent.atomic package address F mark bit Reference

213. Art of Multiprocessor Programming213 Extracting Reference & Mark Public Object get(boolean[] marked);

214. Art of Multiprocessor Programming214 Extracting Reference & Mark Public Object get(boolean[] marked); Returns reference Returns mark at array index 0!

215. Art of Multiprocessor Programming215 Extracting Reference Only public object getReference(); Value of reference

216. Art of Multiprocessor Programming216 Extracting Mark Only public boolean isMarked(); Value of mark

217. Art of Multiprocessor Programming217 Changing State Public boolean compareAndSet( Object expectedRef, Object updateRef, boolean expectedMark, boolean updateMark);

218. Art of Multiprocessor Programming218 Changing State Public boolean compareAndSet( Object expectedRef, Object updateRef, boolean expectedMark, boolean updateMark); If this is the current reference … And this is the current mark …

219. Art of Multiprocessor Programming219 Changing State Public boolean compareAndSet( Object expectedRef, Object updateRef, boolean expectedMark, boolean updateMark); …then change to this new reference … … and this new mark

220. Art of Multiprocessor Programming220 Changing State public boolean attemptMark( Object expectedRef, boolean updateMark);

221. Art of Multiprocessor Programming221 Changing State public boolean attemptMark( Object expectedRef, boolean updateMark); If this is the current reference …

222. Art of Multiprocessor Programming222 Changing State public boolean attemptMark( Object expectedRef, boolean updateMark);.. then change to this new mark.

223. b CAS Art of Multiprocessor Programming223 Removing a Node acd remove c

224. Art of Multiprocessor Programming224 Removing a Node abd remove b remove c c failed CAS

225. Art of Multiprocessor Programming225 Removing a Node abd remove b remove c c

226. Art of Multiprocessor Programming226 Removing a Node ad remove b remove c

227. Art of Multiprocessor Programming227 Traversing the List Q: what do you do when you find a “logically” deleted node in your path? A: finish the job. –CAS the predecessor’s next field –Proceed (repeat as needed)

228. Art of Multiprocessor Programming228 Lock-Free Traversal (only Add and Remove) abcd CAS Uh-oh pred curr pred curr

229. Art of Multiprocessor Programming229 The Window Class class Window { public Node pred; public Node curr; Window(Node pred, Node curr) { this.pred = pred; this.curr = curr; }

230. Art of Multiprocessor Programming230 The Window Class class Window { public Node pred; public Node curr; Window(Node pred, Node curr) { this.pred = pred; this.curr = curr; } A container for pred and current values

231. Art of Multiprocessor Programming231 Using the Find Method Window window = find(head, key); Node pred = window.pred; curr = window.curr;

232. Art of Multiprocessor Programming232 Using the Find Method Window window = find(head, key); Node pred = window.pred; curr = window.curr; Find returns window

233. Art of Multiprocessor Programming233 Using the Find Method Window window = find(head, key); Node pred = window.pred; curr = window.curr; Extract pred and curr

234. Art of Multiprocessor Programming© Herlihy-Shavit 2007 234 The Find Method Window window = find(item); At some instant, predcurrsucc item or …

235. Art of Multiprocessor Programming© Herlihy-Shavit 2007 235 The Find Method Window window = find(item); At some instant, pred curr= null succ item not in list

236. Art of Multiprocessor Programming236 Remove public boolean remove(T item) { Boolean snip; while (true) { Window window = find(head, key); Node pred = window.pred, curr = window.curr; if (curr.key != key) { return false; } else { Node succ = curr.next.getReference(); snip = curr.next.compareAndSet(succ, succ, false true); if (!snip) continue; pred.next.compareAndSet(curr, succ, false, false); return true; }}}

237. Art of Multiprocessor Programming237 Remove public boolean remove(T item) { Boolean snip; while (true) { Window window = find(head, key); Node pred = window.pred, curr = window.curr; if (curr.key != key) { return false; } else { Node succ = curr.next.getReference(); snip = curr.next.compareAndSet (succ, succ, false, true); if (!snip) continue; pred.next.compareAndSet(curr, succ, false, false); return true; }}} Keep trying

238. Art of Multiprocessor Programming238 Remove public boolean remove(T item) { Boolean snip; while (true) { Window window = find(head, key); Node pred = window.pred, curr = window.curr; if (curr.key != key) { return false; } else { Node succ = curr.next.getReference(); snip = curr.next.compareAndSet (succ, succ, false, true); if (!snip) continue; pred.next.compareAndSet(curr, succ, false, false); return true; }}} Find neighbors

239. Art of Multiprocessor Programming239 Remove public boolean remove(T item) { Boolean snip; while (true) { Window window = find(head, key); Node pred = window.pred, curr = window.curr; if (curr.key != key) { return false; } else { Node succ = curr.next.getReference(); snip = curr.next.compareAndSet(succ, succ, false, true); if (!snip) continue; pred.next.compareAndSet(curr, succ, false, false); return true; }}} She’s not there …

240. Art of Multiprocessor Programming240 Remove public boolean remove(T item) { Boolean snip; while (true) { Window window = find(head, key); Node pred = window.pred, curr = window.curr; if (curr.key != key) { return false; } else { Node succ = curr.next.getReference(); snip = curr.next.compareAndSet(succ, succ, false, true); if (!snip) continue; pred.next.compareAndSet(curr, succ, false, false); return true; }}} Try to mark node as deleted

241. Art of Multiprocessor Programming241 Remove public boolean remove(T item) { Boolean snip; while (true) { Window window = find(head, key); Node pred = window.pred, curr = window.curr; if (curr.key != key) { return false; } else { Node succ = curr.next.getReference(); snip = curr.next.compareAndSet(succ, succ, false, true); if (!snip) continue; pred.next.compareAndSet(curr, succ, false, false); return true; }}} If it doesn’t work, just retry, if it does, job essentially done

242. Art of Multiprocessor Programming242 Remove public boolean remove(T item) { Boolean snip; while (true) { Window window = find(head, key); Node pred = window.pred, curr = window.curr; if (curr.key != key) { return false; } else { Node succ = curr.next.getReference(); snip = curr.next.compareAndSet(succ, succ, false, true); if (!snip) continue; pred.next.compareAndSet(curr, succ, false, false); return true; }}} Try to advance reference (if we don’t succeed, someone else did or will). a

243. Art of Multiprocessor Programming243 Add public boolean add(T item) { boolean splice; while (true) { Window window = find(head, key); Node pred = window.pred, curr = window.curr; if (curr.key == key) { return false; } else { Node node = new Node(item); node.next = new AtomicMarkableRef(curr, false); if (pred.next.compareAndSet(curr, node, false, false)) {return true;} }}}

244. Art of Multiprocessor Programming244 Add public boolean add(T item) { boolean splice; while (true) { Window window = find(head, key); Node pred = window.pred, curr = window.curr; if (curr.key == key) { return false; } else { Node node = new Node(item); node.next = new AtomicMarkableRef(curr, false); if (pred.next.compareAndSet(curr, node, false, false)) {return true;} }}} Item already there.

245. Art of Multiprocessor Programming245 Add public boolean add(T item) { boolean splice; while (true) { Window window = find(head, key); Node pred = window.pred, curr = window.curr; if (curr.key == key) { return false; } else { Node node = new Node(item); node.next = new AtomicMarkableRef(curr, false); if (pred.next.compareAndSet(curr, node, false, false)) {return true;} }}} create new node

246. Art of Multiprocessor Programming246 Add public boolean add(T item) { boolean splice; while (true) { Window window = find(head, key); Node pred = window.pred, curr = window.curr; if (curr.key == key) { return false; } else { Node node = new Node(item); node.next = new AtomicMarkableRef(curr, false); if (pred.next.compareAndSet(curr, node, false, false)) {return true;} }}} Install new node, else retry loop

247. Art of Multiprocessor Programming247 Wait-free Contains public boolean contains(T item) { boolean marked; int key = item.hashCode(); Node curr = this.head; while (curr.key < key) curr = curr.next; Node succ = curr.next.get(marked); return (curr.key == key && !marked[0]) }

248. Art of Multiprocessor Programming248 Wait-free Contains public boolean contains(T item) { boolean marked; int key = item.hashCode(); Node curr = this.head; while (curr.key < key) curr = curr.next; Node succ = curr.next.get(marked); return (curr.key == key && !marked[0]) } Only diff is that we get and check marked

249. Art of Multiprocessor Programming249 Lock-free Find public Window find(Node head, int key) { Node pred = null, curr = null, succ = null; boolean[] marked = {false}; boolean snip; retry: while (true) { pred = head; curr = pred.next.getReference(); while (true) { succ = curr.next.get(marked); while (marked[0]) { … } if (curr.key >= key) return new Window(pred, curr); pred = curr; curr = succ; } }}

250. Art of Multiprocessor Programming250 Lock-free Find public Window find(Node head, int key) { Node pred = null, curr = null, succ = null; boolean[] marked = {false}; boolean snip; retry: while (true) { pred = head; curr = pred.next.getReference(); while (true) { succ = curr.next.get(marked); while (marked[0]) { … } if (curr.key >= key) return new Window(pred, curr); pred = curr; curr = succ; } }} If list changes while traversed, start over. Lock-Free because we start over only if someone else makes progress

251. Art of Multiprocessor Programming251 public Window find(Node head, int key) { Node pred = null, curr = null, succ = null; boolean[] marked = {false}; boolean snip; retry: while (true) { pred = head; curr = pred.next.getReference(); while (true) { succ = curr.next.get(marked); while (marked[0]) { … } if (curr.key >= key) return new Window(pred, curr); pred = curr; curr = succ; } }} Lock-free Find Start looking from head

252. Art of Multiprocessor Programming252 public Window find(Node head, int key) { Node pred = null, curr = null, succ = null; boolean[] marked = {false}; boolean snip; retry: while (true) { pred = head; curr = pred.next.getReference(); while (true) { succ = curr.next.get(marked); while (marked[0]) { … } if (curr.key >= key) return new Window(pred, curr); pred = curr; curr = succ; } }} Lock-free Find Move down the list

253. Art of Multiprocessor Programming253 public Window find(Node head, int key) { Node pred = null, curr = null, succ = null; boolean[] marked = {false}; boolean snip; retry: while (true) { pred = head; curr = pred.next.getReference(); while (true) { succ = curr.next.get(marked); while (marked[0]) { … } if (curr.key >= key) return new Window(pred, curr); pred = curr; curr = succ; } }} Lock-free Find Get ref to successor and current deleted bit

254. Art of Multiprocessor Programming254 public Window find(Node head, int key) { Node pred = null, curr = null, succ = null; boolean[] marked = {false}; boolean snip; retry: while (true) { pred = head; curr = pred.next.getReference(); while (true) { succ = curr.next.get(marked); while (marked[0]) { … } if (curr.key >= key) return new Window(pred, curr); pred = curr; curr = succ; } }} Lock-free Find Try to remove deleted nodes in path…code details soon

255. Art of Multiprocessor Programming255 public Window find(Node head, int key) { Node pred = null, curr = null, succ = null; boolean[] marked = {false}; boolean snip; retry: while (true) { pred = head; curr = pred.next.getReference(); while (true) { succ = curr.next.get(marked); while (marked[0]) { … } if (curr.key >= key) return new Window(pred, curr); pred = curr; curr = succ; } }} Lock-free Find If curr key that is greater or equal, return pred and curr

256. Art of Multiprocessor Programming256 public Window find(Node head, int key) { Node pred = null, curr = null, succ = null; boolean[] marked = {false}; boolean snip; retry: while (true) { pred = head; curr = pred.next.getReference(); while (true) { succ = curr.next.get(marked); while (marked[0]) { … } if (curr.key >= key) return new Window(pred, curr); pred = curr; curr = succ; } }} Lock-free Find Otherwise advance window and loop again

257. Art of Multiprocessor Programming257 Lock-free Find retry: while (true) { … while (marked[0]) { snip = pred.next.compareAndSet(curr, succ, false, false); if (!snip) continue retry; curr = succ; succ = curr.next.get(marked); } …

258. Art of Multiprocessor Programming258 Lock-free Find retry: while (true) { … while (marked[0]) { snip = pred.next.compareAndSet(curr, succ, false, false); if (!snip) continue retry; curr = succ; succ = curr.next.get(marked); } … Try to snip out node

259. Art of Multiprocessor Programming259 Lock-free Find retry: while (true) { … while (marked[0]) { snip = pred.next.compareAndSet(curr, succ, false, false); if (!snip) continue retry; curr = succ; succ = curr.next.get(marked); } … if predecessor’s next field changed must retry whole traversal

260. Art of Multiprocessor Programming260 Lock-free Find retry: while (true) { … while (marked[0]) { snip = pred.next.compareAndSet(curr, succ, false, false); if (!snip) continue retry; curr = succ; succ = curr.next.get(marked); } … Otherwise move on to check if next node deleted

261. Art of Multiprocessor Programming261 Performance On 16 node shared memory machine Benchmark throughput of Java List-based Set algs. Vary % of Contains() method Calls.

262. Art of Multiprocessor Programming262 High Contains Ratio Lock-free Lazy list Coarse Grained Fine Lock-coupling

263. Art of Multiprocessor Programming263 Low Contains Ratio

264. Art of Multiprocessor Programming264 As Contains Ratio Increases % Contains()

265. Art of Multiprocessor Programming265 Summary Coarse-grained locking Fine-grained locking Optimistic synchronization Lock-free synchronization

266. Art of Multiprocessor Programming266 “To Lock or Not to Lock” Locking vs. Non-blocking: Extremist views on both sides The answer: nobler to compromise, combine locking and non-blocking –Example: Lazy list combines blocking add() and remove() and a wait-free contains() –Remember: Blocking/non-blocking is a property of a method

267. Art of Multiprocessor Programming267 This work is licensed under a Creative Commons Attribution- ShareAlike 2.5 License.Creative Commons Attribution- ShareAlike 2.5 License You are free: –to Share — to copy, distribute and transmit the work –to Remix — to adapt the work Under the following conditions: –Attribution. You must attribute the work to “The Art of Multiprocessor Programming” (but not in any way that suggests that the authors endorse you or your use of the work). –Share Alike. If you alter, transform, or build upon this work, you may distribute the resulting work only under the same, similar or a compatible license. For any reuse or distribution, you must make clear to others the license terms of this work. The best way to do this is with a link to –http://creativecommons.org/licenses/by-sa/3.0/. Any of the above conditions can be waived if you get permission from the copyright holder. Nothing in this license impairs or restricts the author's moral rights.