3 #ifndef CDSLIB_INTRUSIVE_SPLIT_LIST_RCU_H
4 #define CDSLIB_INTRUSIVE_SPLIT_LIST_RCU_H
8 #include <cds/intrusive/details/split_list_base.h>
9 #include <cds/details/binary_functor_wrapper.h>
11 namespace cds { namespace intrusive {
13 /// Split-ordered list RCU specialization
14 /** @ingroup cds_intrusive_map
15 \anchor cds_intrusive_SplitListSet_rcu
17 Hash table implementation based on split-ordered list algorithm discovered by Ori Shalev and Nir Shavit, see
18 - [2003] Ori Shalev, Nir Shavit "Split-Ordered Lists - Lock-free Resizable Hash Tables"
19 - [2008] Nir Shavit "The Art of Multiprocessor Programming"
21 The split-ordered list is a lock-free implementation of an extensible unbounded hash table. It uses original
22 recursive split-ordering algorithm discovered by Ori Shalev and Nir Shavit that allows to split buckets
23 without moving an item on resizing, see \ref cds_SplitList_algo_desc "short algo description".
27 Template parameters are:
28 - \p RCU - one of \ref cds_urcu_gc "RCU type"
29 - \p OrderedList - ordered list implementation used as bucket for hash set, for example, MichaelList, LazyList.
30 The intrusive ordered list implementation specifies the type \p T stored in the hash-set,
31 the comparing functor for the type \p T and other features specific for the ordered list.
32 - \p Traits - set traits, default isd \p split_list::traits.
33 Instead of defining \p Traits struct you may use option-based syntax with \p split_list::make_traits metafunction.
35 @note About reqired features of hash functor see \ref cds_SplitList_hash_functor "SplitList general description".
38 Before including <tt><cds/intrusive/split_list_rcu.h></tt> you should include appropriate RCU header file,
39 see \ref cds_urcu_gc "RCU type" for list of existing RCU class and corresponding header files.
40 For example, for \ref cds_urcu_general_buffered_gc "general-purpose buffered RCU" and
41 MichaelList-based split-list you should include:
43 #include <cds/urcu/general_buffered.h>
44 #include <cds/intrusive/michael_list_rcu.h>
45 #include <cds/intrusive/split_list_rcu.h>
47 // Declare Michael's list for type Foo and default traits:
48 typedef cds::intrusive::MichaelList< cds::urcu::gc< cds::urcu::general_buffered<> >, Foo > rcu_michael_list;
50 // Declare split-list based on rcu_michael_list
51 typedef cds::intrusive::SplitListSet< cds::urcu::gc< cds::urcu::general_buffered<> >, rcu_michael_list > rcu_split_list;
58 # ifdef CDS_DOXYGEN_INVOKED
59 class Traits = split_list::traits
64 class SplitListSet< cds::urcu::gc< RCU >, OrderedList, Traits >
67 typedef cds::urcu::gc< RCU > gc; ///< RCU garbage collector
68 typedef Traits traits; ///< Traits template parameters
70 /// Hash functor for \ref value_type and all its derivatives that you use
71 typedef typename cds::opt::v::hash_selector< typename traits::hash >::type hash;
74 typedef cds::intrusive::split_list::implementation_tag implementation_tag;
79 typedef split_list::details::rebind_list_traits<OrderedList, traits> wrapped_ordered_list;
83 # ifdef CDS_DOXYGEN_INVOKED
84 typedef OrderedList ordered_list; ///< type of ordered list used as base for split-list
86 typedef typename wrapped_ordered_list::result ordered_list;
88 typedef typename ordered_list::value_type value_type; ///< type of value stored in the split-list
89 typedef typename ordered_list::key_comparator key_comparator; ///< key compare functor
90 typedef typename ordered_list::disposer disposer; ///< Node disposer functor
91 typedef typename ordered_list::rcu_lock rcu_lock; ///< RCU scoped lock
92 typedef typename ordered_list::exempt_ptr exempt_ptr; ///< pointer to extracted node
93 /// Group of \p extract_xxx functions require external locking if underlying ordered list requires that
94 static CDS_CONSTEXPR const bool c_bExtractLockExternal = ordered_list::c_bExtractLockExternal;
96 typedef typename traits::item_counter item_counter; ///< Item counter type
97 typedef typename traits::back_off back_off; ///< back-off strategy for spinning
98 typedef typename traits::memory_model memory_model; ///< Memory ordering. See cds::opt::memory_model option
99 typedef typename traits::stat stat; ///< Internal statistics
102 typedef typename ordered_list::node_type list_node_type; ///< Node type as declared in ordered list
103 typedef split_list::node<list_node_type> node_type; ///< split-list node type
104 typedef node_type dummy_node_type; ///< dummy node type
106 /// Split-list node traits
108 This traits is intended for converting between underlying ordered list node type \ref list_node_type
109 and split-list node type \ref node_type
111 typedef split_list::node_traits<typename ordered_list::node_traits> node_traits;
114 /// Bucket table implementation
115 typedef typename split_list::details::bucket_table_selector<
116 traits::dynamic_bucket_table
119 , opt::allocator< typename traits::allocator >
120 , opt::memory_model< memory_model >
121 >::type bucket_table;
127 /// Ordered list wrapper to access protected members of OrderedList
128 class ordered_list_wrapper: public ordered_list
130 typedef ordered_list base_class;
131 typedef typename base_class::auxiliary_head bucket_head_type;
134 bool insert_at( dummy_node_type * pHead, value_type& val )
136 assert( pHead != nullptr );
137 bucket_head_type h(pHead);
138 return base_class::insert_at( h, val );
141 template <typename Func>
142 bool insert_at( dummy_node_type * pHead, value_type& val, Func f )
144 assert( pHead != nullptr );
145 bucket_head_type h(pHead);
146 return base_class::insert_at( h, val, f );
149 template <typename Func>
150 std::pair<bool, bool> ensure_at( dummy_node_type * pHead, value_type& val, Func func )
152 assert( pHead != nullptr );
153 bucket_head_type h(pHead);
154 return base_class::ensure_at( h, val, func );
157 bool unlink_at( dummy_node_type * pHead, value_type& val )
159 assert( pHead != nullptr );
160 bucket_head_type h(pHead);
161 return base_class::unlink_at( h, val );
164 template <typename Q, typename Compare, typename Func>
165 bool erase_at( dummy_node_type * pHead, split_list::details::search_value_type<Q> const& val, Compare cmp, Func f )
167 assert( pHead != nullptr );
168 bucket_head_type h(pHead);
169 return base_class::erase_at( h, val, cmp, f );
172 template <typename Q, typename Compare>
173 bool erase_at( dummy_node_type * pHead, split_list::details::search_value_type<Q> const& val, Compare cmp )
175 assert( pHead != nullptr );
176 bucket_head_type h(pHead);
177 return base_class::erase_at( h, val, cmp );
180 template <typename Q, typename Compare>
181 value_type * extract_at( dummy_node_type * pHead, split_list::details::search_value_type<Q>& val, Compare cmp )
183 assert( pHead != nullptr );
184 bucket_head_type h(pHead);
185 return base_class::extract_at( h, val, cmp );
188 template <typename Q, typename Compare, typename Func>
189 bool find_at( dummy_node_type * pHead, split_list::details::search_value_type<Q>& val, Compare cmp, Func f ) const
191 assert( pHead != nullptr );
192 bucket_head_type h(pHead);
193 return base_class::find_at( h, val, cmp, f );
196 template <typename Q, typename Compare>
197 bool find_at( dummy_node_type * pHead, split_list::details::search_value_type<Q> const & val, Compare cmp ) const
199 assert( pHead != nullptr );
200 bucket_head_type h(pHead);
201 return base_class::find_at( h, val, cmp );
204 template <typename Q, typename Compare>
205 value_type * get_at( dummy_node_type * pHead, split_list::details::search_value_type<Q>& val, Compare cmp ) const
207 assert( pHead != nullptr );
208 bucket_head_type h(pHead);
209 return base_class::get_at( h, val, cmp );
212 bool insert_aux_node( dummy_node_type * pNode )
214 return base_class::insert_aux_node( pNode );
216 bool insert_aux_node( dummy_node_type * pHead, dummy_node_type * pNode )
218 bucket_head_type h(pHead);
219 return base_class::insert_aux_node( h, pNode );
223 template <typename Less>
224 struct less_wrapper: public cds::opt::details::make_comparator_from_less<Less>
226 typedef cds::opt::details::make_comparator_from_less<Less> base_wrapper;
228 template <typename Q1, typename Q2>
229 int operator()( split_list::details::search_value_type<Q1> const& v1, Q2 const& v2 ) const
231 return base_wrapper::operator()( v1.val, v2 );
234 template <typename Q1, typename Q2>
235 int operator()( Q1 const& v1, split_list::details::search_value_type<Q2> const& v2 ) const
237 return base_wrapper::operator()( v1, v2.val );
243 ordered_list_wrapper m_List; ///< Ordered list containing split-list items
244 bucket_table m_Buckets; ///< bucket table
245 atomics::atomic<size_t> m_nBucketCountLog2; ///< log2( current bucket count )
246 atomics::atomic<size_t> m_nMaxItemCount; ///< number of items container can hold, before we have to resize
247 item_counter m_ItemCounter; ///< Item counter
248 hash m_HashFunctor; ///< Hash functor
249 stat m_Stat; ///< Internal stattistics accumulator
253 typedef cds::details::Allocator< dummy_node_type, typename traits::allocator > dummy_node_allocator;
255 dummy_node_type * alloc_dummy_node( size_t nHash )
257 m_Stat.onHeadNodeAllocated();
258 CDS_TSAN_ANNOTATE_IGNORE_WRITES_BEGIN;
259 dummy_node_type * p = dummy_node_allocator().New( nHash );
260 CDS_TSAN_ANNOTATE_IGNORE_WRITES_END;
263 void free_dummy_node( dummy_node_type * p )
265 CDS_TSAN_ANNOTATE_IGNORE_WRITES_BEGIN;
266 dummy_node_allocator().Delete( p );
267 CDS_TSAN_ANNOTATE_IGNORE_WRITES_END;
268 m_Stat.onHeadNodeFreed();
271 /// Calculates hash value of \p key
272 template <typename Q>
273 size_t hash_value( Q const& key ) const
275 return m_HashFunctor( key );
278 size_t bucket_no( size_t nHash ) const
280 return nHash & ( (1 << m_nBucketCountLog2.load(memory_model::memory_order_relaxed)) - 1 );
283 static size_t parent_bucket( size_t nBucket )
285 assert( nBucket > 0 );
286 return nBucket & ~( 1 << bitop::MSBnz( nBucket ) );
289 dummy_node_type * init_bucket( size_t nBucket )
291 assert( nBucket > 0 );
292 size_t nParent = parent_bucket( nBucket );
294 dummy_node_type * pParentBucket = m_Buckets.bucket( nParent );
295 if ( pParentBucket == nullptr ) {
296 pParentBucket = init_bucket( nParent );
297 m_Stat.onRecursiveInitBucket();
300 assert( pParentBucket != nullptr );
302 // Allocate a dummy node for new bucket
304 dummy_node_type * pBucket = alloc_dummy_node( split_list::dummy_hash( nBucket ) );
305 if ( m_List.insert_aux_node( pParentBucket, pBucket ) ) {
306 m_Buckets.bucket( nBucket, pBucket );
307 m_Stat.onNewBucket();
310 free_dummy_node( pBucket );
313 // Another thread set the bucket. Wait while it done
315 // In this point, we must wait while nBucket is empty.
316 // The compiler can decide that waiting loop can be "optimized" (stripped)
317 // To prevent this situation, we use waiting on volatile bucket_head_ptr pointer.
319 m_Stat.onBucketInitContenton();
322 dummy_node_type volatile * p = m_Buckets.bucket( nBucket );
324 return const_cast<dummy_node_type *>( p );
326 m_Stat.onBusyWaitBucketInit();
330 dummy_node_type * get_bucket( size_t nHash )
332 size_t nBucket = bucket_no( nHash );
334 dummy_node_type * pHead = m_Buckets.bucket( nBucket );
335 if ( pHead == nullptr )
336 pHead = init_bucket( nBucket );
338 assert( pHead->is_dummy() );
345 // GC and OrderedList::gc must be the same
346 static_assert( std::is_same<gc, typename ordered_list::gc>::value, "GC and OrderedList::gc must be the same");
348 // atomicity::empty_item_counter is not allowed as a item counter
349 static_assert( !std::is_same<item_counter, cds::atomicity::empty_item_counter>::value,
350 "cds::atomicity::empty_item_counter is not allowed as a item counter");
352 // Initialize bucket 0
353 dummy_node_type * pNode = alloc_dummy_node( 0 /*split_list::dummy_hash(0)*/ );
355 // insert_aux_node cannot return false for empty list
356 CDS_VERIFY( m_List.insert_aux_node( pNode ));
358 m_Buckets.bucket( 0, pNode );
361 static size_t max_item_count( size_t nBucketCount, size_t nLoadFactor )
363 return nBucketCount * nLoadFactor;
366 void inc_item_count()
368 size_t nMaxCount = m_nMaxItemCount.load(memory_model::memory_order_relaxed);
369 if ( ++m_ItemCounter <= nMaxCount )
372 size_t sz = m_nBucketCountLog2.load(memory_model::memory_order_relaxed);
373 const size_t nBucketCount = static_cast<size_t>(1) << sz;
374 if ( nBucketCount < m_Buckets.capacity() ) {
375 // we may grow the bucket table
376 const size_t nLoadFactor = m_Buckets.load_factor();
377 if ( nMaxCount < max_item_count( nBucketCount, nLoadFactor ))
378 return; // someone already have updated m_nBucketCountLog2, so stop here
380 m_nMaxItemCount.compare_exchange_strong( nMaxCount, max_item_count( nBucketCount << 1, nLoadFactor ),
381 memory_model::memory_order_relaxed, atomics::memory_order_relaxed );
382 m_nBucketCountLog2.compare_exchange_strong( sz, sz + 1, memory_model::memory_order_relaxed, atomics::memory_order_relaxed );
385 m_nMaxItemCount.store( std::numeric_limits<size_t>::max(), memory_model::memory_order_relaxed );
388 template <typename Q, typename Compare, typename Func>
389 bool find_( Q& val, Compare cmp, Func f )
391 size_t nHash = hash_value( val );
392 split_list::details::search_value_type<Q> sv( val, split_list::regular_hash( nHash ));
393 dummy_node_type * pHead = get_bucket( nHash );
394 assert( pHead != nullptr );
396 return m_Stat.onFind( m_List.find_at( pHead, sv, cmp,
397 [&f](value_type& item, split_list::details::search_value_type<Q>& val){ f(item, val.val ); }));
400 template <typename Q, typename Compare>
401 bool find_value( Q const& val, Compare cmp )
403 size_t nHash = hash_value( val );
404 split_list::details::search_value_type<Q const> sv( val, split_list::regular_hash( nHash ));
405 dummy_node_type * pHead = get_bucket( nHash );
406 assert( pHead != nullptr );
408 return m_Stat.onFind( m_List.find_at( pHead, sv, cmp ));
411 template <typename Q, typename Compare>
412 value_type * get_( Q const& val, Compare cmp )
414 size_t nHash = hash_value( val );
415 split_list::details::search_value_type<Q const> sv( val, split_list::regular_hash( nHash ));
416 dummy_node_type * pHead = get_bucket( nHash );
417 assert( pHead != nullptr );
419 value_type * p = m_List.get_at( pHead, sv, cmp );
420 m_Stat.onFind( p != nullptr );
424 template <typename Q, typename Compare>
425 value_type * extract_( Q const& val, Compare cmp )
427 size_t nHash = hash_value( val );
428 split_list::details::search_value_type<Q const> sv( val, split_list::regular_hash( nHash ));
429 dummy_node_type * pHead = get_bucket( nHash );
430 assert( pHead != nullptr );
432 value_type * pNode = m_List.extract_at( pHead, sv, cmp );
435 m_Stat.onExtractSuccess();
438 m_Stat.onExtractFailed();
442 template <typename Q, typename Less>
443 value_type * extract_with_( Q const& val, Less pred )
446 return extract_( val, typename wrapped_ordered_list::template make_compare_from_less<Less>());
449 template <typename Q, typename Compare>
450 bool erase_( const Q& val, Compare cmp )
452 size_t nHash = hash_value( val );
453 split_list::details::search_value_type<Q const> sv( val, split_list::regular_hash( nHash ));
454 dummy_node_type * pHead = get_bucket( nHash );
455 assert( pHead != nullptr );
457 if ( m_List.erase_at( pHead, sv, cmp ) ) {
459 m_Stat.onEraseSuccess();
462 m_Stat.onEraseFailed();
466 template <typename Q, typename Compare, typename Func>
467 bool erase_( Q const& val, Compare cmp, Func f )
469 size_t nHash = hash_value( val );
470 split_list::details::search_value_type<Q const> sv( val, split_list::regular_hash( nHash ));
471 dummy_node_type * pHead = get_bucket( nHash );
472 assert( pHead != nullptr );
474 if ( m_List.erase_at( pHead, sv, cmp, f )) {
476 m_Stat.onEraseSuccess();
479 m_Stat.onEraseFailed();
486 /// Initialize split-ordered list of default capacity
488 The default capacity is defined in bucket table constructor.
489 See split_list::expandable_bucket_table, split_list::static_ducket_table
490 which selects by split_list::dynamic_bucket_table option.
493 : m_nBucketCountLog2(1)
494 , m_nMaxItemCount( max_item_count(2, m_Buckets.load_factor()) )
499 /// Initialize split-ordered list
501 size_t nItemCount ///< estimate average of item count
502 , size_t nLoadFactor = 1 ///< load factor - average item count per bucket. Small integer up to 8, default is 1.
504 : m_Buckets( nItemCount, nLoadFactor )
505 , m_nBucketCountLog2(1)
506 , m_nMaxItemCount( max_item_count(2, m_Buckets.load_factor()) )
514 The function inserts \p val in the set if it does not contain
515 an item with key equal to \p val.
517 The function makes RCU lock internally.
519 Returns \p true if \p val is placed into the set, \p false otherwise.
521 bool insert( value_type& val )
523 size_t nHash = hash_value( val );
524 dummy_node_type * pHead = get_bucket( nHash );
525 assert( pHead != nullptr );
527 node_traits::to_node_ptr( val )->m_nHash = split_list::regular_hash( nHash );
529 if ( m_List.insert_at( pHead, val )) {
531 m_Stat.onInsertSuccess();
534 m_Stat.onInsertFailed();
540 This function is intended for derived non-intrusive containers.
542 The function allows to split creating of new item into two part:
543 - create item with key only
544 - insert new item into the set
545 - if inserting is success, calls \p f functor to initialize value-field of \p val.
547 The functor signature is:
549 void func( value_type& val );
551 where \p val is the item inserted.
553 The function makes RCU lock internally.
555 @warning For \ref cds_intrusive_MichaelList_rcu "MichaelList" as the bucket see \ref cds_intrusive_item_creating "insert item troubleshooting".
556 \ref cds_intrusive_LazyList_rcu "LazyList" provides exclusive access to inserted item and does not require any node-level
559 template <typename Func>
560 bool insert( value_type& val, Func f )
562 size_t nHash = hash_value( val );
563 dummy_node_type * pHead = get_bucket( nHash );
564 assert( pHead != nullptr );
566 node_traits::to_node_ptr( val )->m_nHash = split_list::regular_hash( nHash );
568 if ( m_List.insert_at( pHead, val, f )) {
570 m_Stat.onInsertSuccess();
573 m_Stat.onInsertFailed();
577 /// Ensures that the \p val exists in the set
579 The operation performs inserting or changing data with lock-free manner.
581 If the item \p val is not found in the set, then \p val is inserted into the set.
582 Otherwise, the functor \p func is called with item found.
583 The functor signature is:
585 void func( bool bNew, value_type& item, value_type& val );
588 - \p bNew - \p true if the item has been inserted, \p false otherwise
589 - \p item - item of the set
590 - \p val - argument \p val passed into the \p ensure function
591 If new item has been inserted (i.e. \p bNew is \p true) then \p item and \p val arguments
592 refers to the same thing.
594 The function makes RCU lock internally.
596 Returns std::pair<bool, bool> where \p first is \p true if operation is successfull,
597 \p second is \p true if new item has been added or \p false if the item with \p key
598 already is in the set.
600 @warning For \ref cds_intrusive_MichaelList_rcu "MichaelList" as the bucket see \ref cds_intrusive_item_creating "insert item troubleshooting".
601 \ref cds_intrusive_LazyList_rcu "LazyList" provides exclusive access to inserted item and does not require any node-level
604 template <typename Func>
605 std::pair<bool, bool> ensure( value_type& val, Func func )
607 size_t nHash = hash_value( val );
608 dummy_node_type * pHead = get_bucket( nHash );
609 assert( pHead != nullptr );
611 node_traits::to_node_ptr( val )->m_nHash = split_list::regular_hash( nHash );
613 std::pair<bool, bool> bRet = m_List.ensure_at( pHead, val, func );
614 if ( bRet.first && bRet.second ) {
616 m_Stat.onEnsureNew();
619 m_Stat.onEnsureExist();
623 /// Unlinks the item \p val from the set
625 The function searches the item \p val in the set and unlinks it from the set
626 if it is found and is equal to \p val.
628 Difference between \ref erase and \p unlink functions: \p erase finds <i>a key</i>
629 and deletes the item found. \p unlink finds an item by key and deletes it
630 only if \p val is an item of that set, i.e. the pointer to item found
631 is equal to <tt> &val </tt>.
633 RCU \p synchronize method can be called, therefore, RCU should not be locked.
635 The function returns \p true if success and \p false otherwise.
637 bool unlink( value_type& val )
639 size_t nHash = hash_value( val );
640 dummy_node_type * pHead = get_bucket( nHash );
641 assert( pHead != nullptr );
643 if ( m_List.unlink_at( pHead, val ) ) {
645 m_Stat.onEraseSuccess();
648 m_Stat.onEraseFailed();
652 /// Deletes the item from the set
653 /** \anchor cds_intrusive_SplitListSet_rcu_erase
654 The function searches an item with key equal to \p key in the set,
655 unlinks it from the set, and returns \p true.
656 If the item with key equal to \p key is not found the function return \p false.
658 Difference between \ref erase and \p unlink functions: \p erase finds <i>a key</i>
659 and deletes the item found. \p unlink finds an item by key and deletes it
660 only if \p key is an item of that set, i.e. the pointer to item found
661 is equal to <tt> &key </tt>.
663 RCU \p synchronize method can be called, therefore, RCU should not be locked.
665 Note the hash functor should accept a parameter of type \p Q that can be not the same as \p value_type.
667 template <typename Q>
668 bool erase( Q const& key )
670 return erase_( key, key_comparator() );
673 /// Deletes the item from the set using \p pred for searching
675 The function is an analog of \ref cds_intrusive_SplitListSet_rcu_erase "erase(Q const&)"
676 but \p cmp is used for key compare.
677 \p Less has the interface like \p std::less.
678 \p pred must imply the same element order as the comparator used for building the set.
680 template <typename Q, typename Less>
681 bool erase_with( Q const& key, Less pred )
684 return erase_( key, typename wrapped_ordered_list::template make_compare_from_less<Less>() );
687 /// Deletes the item from the set
688 /** \anchor cds_intrusive_SplitListSet_rcu_erase_func
689 The function searches an item with key equal to \p key in the set,
690 call \p f functor with item found, unlinks it from the set, and returns \p true.
691 The \ref disposer specified by \p OrderedList class template parameter is called
692 by garbage collector \p GC asynchronously.
694 The \p Func interface is
697 void operator()( value_type const& item );
701 If the item with key equal to \p key is not found the function return \p false.
703 RCU \p synchronize method can be called, therefore, RCU should not be locked.
705 Note the hash functor should accept a parameter of type \p Q that can be not the same as \p value_type.
707 template <typename Q, typename Func>
708 bool erase( Q const& key, Func f )
710 return erase_( key, key_comparator(), f );
713 /// Deletes the item from the set using \p pred for searching
715 The function is an analog of \ref cds_intrusive_SplitListSet_rcu_erase_func "erase(Q const&, Func)"
716 but \p cmp is used for key compare.
717 \p Less has the interface like \p std::less.
718 \p pred must imply the same element order as the comparator used for building the set.
720 template <typename Q, typename Less, typename Func>
721 bool erase_with( Q const& key, Less pred, Func f )
724 return erase_( key, typename wrapped_ordered_list::template make_compare_from_less<Less>(), f );
727 /// Extracts an item from the set
728 /** \anchor cds_intrusive_SplitListSet_rcu_extract
729 The function searches an item with key equal to \p key in the set,
730 unlinks it, and returns \ref cds::urcu::exempt_ptr "exempt_ptr" pointer to the item found.
731 If the item with the key equal to \p key is not found the function returns an empty \p exempt_ptr.
733 @note The function does NOT call RCU read-side lock or synchronization,
734 and does NOT dispose the item found. It just excludes the item from the set
735 and returns a pointer to item found.
736 You should lock RCU before calling of the function, and you should synchronize RCU
737 outside the RCU lock before reusing returned pointer.
740 typedef cds::urcu::gc< general_buffered<> > rcu;
741 typedef cds::intrusive::MichaelList< rcu, Foo > rcu_michael_list;
742 typedef cds::intrusive::SplitListSet< rcu, rcu_michael_list, foo_traits > rcu_splitlist_set;
744 rcu_splitlist_set theSet;
747 rcu_splitlist_set::exempt_ptr p;
749 // first, we should lock RCU
750 rcu_splitlist_set::rcu_lock lock;
752 // Now, you can apply extract function
753 // Note that you must not delete the item found inside the RCU lock
754 p = theList.extract( 10 );
756 // do something with p
761 // We may safely release p here
762 // release() passes the pointer to RCU reclamation cycle:
763 // it invokes RCU retire_ptr function with the disposer you provided for rcu_michael_list.
767 template <typename Q>
768 exempt_ptr extract( Q const& key )
770 return exempt_ptr(extract_( key, key_comparator() ));
773 /// Extracts an item from the set using \p pred for searching
775 The function is an analog of \p extract(Q const&) but \p pred is used for key compare.
776 \p Less functor has the interface like \p std::less.
777 \p pred must imply the same element order as the comparator used for building the set.
779 template <typename Q, typename Less>
780 exempt_ptr extract_with( Q const& key, Less pred )
782 return exempt_ptr( extract_with_( key, pred ));
785 /// Finds the key \p key
786 /** \anchor cds_intrusive_SplitListSet_rcu_find_func
787 The function searches the item with key equal to \p key and calls the functor \p f for item found.
788 The interface of \p Func functor is:
791 void operator()( value_type& item, Q& key );
794 where \p item is the item found, \p key is the <tt>find</tt> function argument.
796 The functor can change non-key fields of \p item. Note that the functor is only guarantee
797 that \p item cannot be disposed during functor is executing.
798 The functor does not serialize simultaneous access to the set \p item. If such access is
799 possible you must provide your own synchronization schema on item level to exclude unsafe item modifications.
801 The \p key argument is non-const since it can be used as \p f functor destination i.e., the functor
802 can modify both arguments.
804 Note the hash functor specified for class \p Traits template parameter
805 should accept a parameter of type \p Q that can be not the same as \p value_type.
807 The function applies RCU lock internally.
809 The function returns \p true if \p key is found, \p false otherwise.
811 template <typename Q, typename Func>
812 bool find( Q& key, Func f )
814 return find_( key, key_comparator(), f );
817 template <typename Q, typename Func>
818 bool find( Q const& key, Func f )
820 return find_( key, key_comparator(), f );
824 /// Finds the key \p key with \p pred predicate for comparing
826 The function is an analog of \ref cds_intrusive_SplitListSet_rcu_find_func "find(Q&, Func)"
827 but \p cmp is used for key compare.
828 \p Less has the interface like \p std::less.
829 \p cmp must imply the same element order as the comparator used for building the set.
831 template <typename Q, typename Less, typename Func>
832 bool find_with( Q& key, Less pred, Func f )
835 return find_( key, typename wrapped_ordered_list::template make_compare_from_less<Less>(), f );
838 template <typename Q, typename Less, typename Func>
839 bool find_with( Q const& key, Less pred, Func f )
842 return find_( key, typename wrapped_ordered_list::template make_compare_from_less<Less>(), f );
846 /// Finds the key \p key
847 /** \anchor cds_intrusive_SplitListSet_rcu_find_val
848 The function searches the item with key equal to \p key
849 and returns \p true if \p key found or \p false otherwise.
851 template <typename Q>
852 bool find( Q const& key )
854 return find_value( key, key_comparator() );
857 /// Finds the key \p key with \p pred predicate for comparing
859 The function is an analog of \ref cds_intrusive_SplitListSet_rcu_find_val "find(Q const&)"
860 but \p cmp is used for key compare.
861 \p Less has the interface like \p std::less.
862 \p pred must imply the same element order as the comparator used for building the set.
864 template <typename Q, typename Less>
865 bool find_with( Q const& key, Less pred )
868 return find_value( key, typename wrapped_ordered_list::template make_compare_from_less<Less>() );
871 /// Finds the key \p key and return the item found
872 /** \anchor cds_intrusive_SplitListSet_rcu_get
873 The function searches the item with key equal to \p key and returns the pointer to item found.
874 If \p key is not found it returns \p nullptr.
876 Note the compare functor should accept a parameter of type \p Q that can be not the same as \p value_type.
878 RCU should be locked before call of this function.
879 Returned item is valid only while RCU is locked:
881 cds::intrusive::SplitListSet< your_template_parameters > theSet;
885 hash_set::rcu_lock lock;
887 foo * pVal = theSet.get( 5 );
892 // Unlock RCU by rcu_lock destructor
893 // pVal can be retired by disposer at any time after RCU has been unlocked
897 template <typename Q>
898 value_type * get( Q const& key )
900 return get_( key, key_comparator() );
903 /// Finds the key \p key and return the item found
905 The function is an analog of \ref cds_intrusive_SplitListSet_rcu_get "get(Q const&)"
906 but \p pred is used for comparing the keys.
908 \p Less functor has the semantics like \p std::less but should take arguments of type \ref value_type and \p Q
910 \p pred must imply the same element order as the comparator used for building the set.
912 template <typename Q, typename Less>
913 value_type * get_with( Q const& key, Less pred )
916 return get_( key, typename wrapped_ordered_list::template make_compare_from_less<Less>());
920 /// Returns item count in the set
923 return m_ItemCounter;
926 /// Checks if the set is empty
928 Emptiness is checked by item counting: if item count is zero then the set is empty.
929 Thus, the correct item counting feature is an important part of split-list set implementation.
936 /// Clears the set (not atomic)
939 iterator it = begin();
940 while ( it != end() ) {
948 /// Returns internal statistics
949 stat const& statistics() const
956 template <bool IsConst>
958 :public split_list::details::iterator_type<node_traits, ordered_list, IsConst>
960 typedef split_list::details::iterator_type<node_traits, ordered_list, IsConst> iterator_base_class;
961 typedef typename iterator_base_class::list_iterator list_iterator;
964 : iterator_base_class()
967 iterator_type( iterator_type const& src )
968 : iterator_base_class( src )
971 // This ctor should be protected...
972 iterator_type( list_iterator itCur, list_iterator itEnd )
973 : iterator_base_class( itCur, itEnd )
980 The forward iterator for a split-list has some features:
981 - it has no post-increment operator
982 - it depends on iterator of underlying \p OrderedList
983 - The iterator cannot be moved across thread boundary since it may contain GC's guard that is thread-private GC data.
984 - Iterator ensures thread-safety even if you delete the item that iterator points to. However, in case of concurrent
985 deleting operations it is no guarantee that you iterate all item in the split-list.
987 Therefore, the use of iterators in concurrent environment is not good idea. Use the iterator on the concurrent container
988 for debug purpose only.
990 typedef iterator_type<false> iterator;
991 /// Const forward iterator
993 For iterator's features and requirements see \ref iterator
995 typedef iterator_type<true> const_iterator;
997 /// Returns a forward iterator addressing the first element in a split-list
999 For empty list \code begin() == end() \endcode
1003 return iterator( m_List.begin(), m_List.end() );
1006 /// Returns an iterator that addresses the location succeeding the last element in a split-list
1008 Do not use the value returned by <tt>end</tt> function to access any item.
1010 The returned value can be used only to control reaching the end of the split-list.
1011 For empty list \code begin() == end() \endcode
1015 return iterator( m_List.end(), m_List.end() );
1018 /// Returns a forward const iterator addressing the first element in a split-list
1019 const_iterator begin() const
1023 /// Returns a forward const iterator addressing the first element in a split-list
1024 const_iterator cbegin() const
1026 return const_iterator( m_List.cbegin(), m_List.cend() );
1029 /// Returns an const iterator that addresses the location succeeding the last element in a split-list
1030 const_iterator end() const
1034 /// Returns an const iterator that addresses the location succeeding the last element in a split-list
1035 const_iterator cend() const
1037 return const_iterator( m_List.cend(), m_List.cend() );
1042 }} // namespace cds::intrusive
1044 #endif // #ifndef CDSLIB_INTRUSIVE_SPLIT_LIST_RCU_H