-//$$CDS-header$$
+/*
+ This file is a part of libcds - Concurrent Data Structures library
+
+ (C) Copyright Maxim Khizhinsky (libcds.dev@gmail.com) 2006-2017
+
+ Source code repo: http://github.com/khizmax/libcds/
+ Download: http://sourceforge.net/projects/libcds/files/
+
+ Redistribution and use in source and binary forms, with or without
+ modification, are permitted provided that the following conditions are met:
+
+ * Redistributions of source code must retain the above copyright notice, this
+ list of conditions and the following disclaimer.
+
+ * Redistributions in binary form must reproduce the above copyright notice,
+ this list of conditions and the following disclaimer in the documentation
+ and/or other materials provided with the distribution.
+
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
+ FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+ OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*/
#ifndef CDSLIB_INTRUSIVE_DETAILS_FELDMAN_HASHSET_BASE_H
#define CDSLIB_INTRUSIVE_DETAILS_FELDMAN_HASHSET_BASE_H
//@endcond
};
+ /// Hash size option
+ /**
+ @copydetails traits::hash_size
+ */
+ template <size_t Size>
+ struct hash_size {
+ //@cond
+ template <typename Base> struct pack: public Base
+ {
+ enum: size_t {
+ hash_size = Size
+ };
+ };
+ //@endcond
+ };
+
+ /// Hash splitter option
+ /**
+ @copydetails traits::hash_splitter
+ */
+ template <typename Splitter>
+ struct hash_splitter {
+ //@cond
+ template <typename Base> struct pack: public Base
+ {
+ typedef Splitter hash_splitter;
+ };
+ //@endcond
+ };
+
+
/// \p FeldmanHashSet internal statistics
template <typename EventCounter = cds::atomicity::event_counter>
struct stat {
*/
typedef cds::opt::none hash_accessor;
+ /// The size of hash value in bytes
+ /**
+ By default, the size of hash value is <tt>sizeof( hash_type )</tt>.
+ Sometimes it is not correct, for example, for that 6-byte struct \p static_assert will be thrown:
+ \code
+ struct key_type {
+ uint32_t key1;
+ uint16_t subkey;
+ };
+
+ static_assert( sizeof( key_type ) == 6, "Key type size mismatch" );
+ \endcode
+ For that case you can specify \p hash_size explicitly.
+
+ Value \p 0 means <tt>sizeof( hash_type )</tt>.
+ */
+ static CDS_CONSTEXPR size_t const hash_size = 0;
+
+ /// Hash splitter
+ /**
+ This trait specifies hash bit-string splitter algorithm.
+ By default, \p cds::algo::number_splitter is used if \p HashType is a number,
+ \p cds::algo::split_bitstring otherwise.
+ */
+ typedef cds::opt::none hash_splitter;
+
/// Disposer for removing data nodes
typedef cds::intrusive::opt::v::empty_disposer disposer;
/// Array node allocator
/**
- Allocator for array nodes. That allocator is used for creating \p headNode and \p arrayNode when the set grows.
+ Allocator for array nodes. The allocator is used for creating \p headNode and \p arrayNode when the set grows.
Default is \ref CDS_DEFAULT_ALLOCATOR
*/
typedef CDS_DEFAULT_ALLOCATOR node_allocator;
Supported \p Options are:
- \p feldman_hashset::hash_accessor - mandatory option, hash accessor functor.
@copydetails traits::hash_accessor
+ - \p feldman_hashset::hash_size - the size of hash value in bytes.
+ @copydetails traits::hash_size
+ - \p feldman_hashset::hash_splitter - a hash splitter algorithm
+ @copydetails traits::hash_splitter
- \p opt::node_allocator - array node allocator.
@copydetails traits::node_allocator
- \p opt::compare - hash comparison functor. No default functor is provided.
}
};
+ /// One-level statistics, see \p FeldmanHashSet::get_level_statistics
+ struct level_statistics
+ {
+ size_t array_node_count; ///< Count of array node at the level
+ size_t node_capacity; ///< Array capacity
+
+ size_t data_cell_count; ///< The number of data cells in all array node at this level
+ size_t array_cell_count; ///< The number of array cells in all array node at this level
+ size_t empty_cell_count; ///< The number of empty cells in all array node at this level
+
+ //@cond
+ level_statistics()
+ : array_node_count(0)
+ , data_cell_count(0)
+ , array_cell_count(0)
+ , empty_cell_count(0)
+ {}
+ //@endcond
+ };
+
//@cond
namespace details {
- template <typename HashType, typename UInt = size_t >
- using hash_splitter = cds::algo::split_bitstring< HashType, UInt >;
+ template <typename HashType, size_t HashSize >
+ using hash_splitter = cds::algo::split_bitstring< HashType, HashSize >;
struct metrics {
size_t head_node_size; // power-of-two
} // namespace details
//@endcond
+
+ //@cond
+ template <typename T, typename Traits>
+ class multilevel_array
+ {
+ public:
+ typedef T value_type;
+ typedef Traits traits;
+ typedef typename Traits::node_allocator node_allocator;
+ typedef typename traits::memory_model memory_model;
+ typedef typename traits::back_off back_off; ///< Backoff strategy
+ typedef typename traits::stat stat; ///< Internal statistics type
+
+ typedef typename traits::hash_accessor hash_accessor;
+ static_assert(!std::is_same< hash_accessor, cds::opt::none >::value, "hash_accessor functor must be specified");
+
+ /// Hash type deduced from \p hash_accessor return type
+ typedef typename std::decay<
+ typename std::remove_reference<
+ decltype(hash_accessor()(std::declval<value_type>()))
+ >::type
+ >::type hash_type;
+ static_assert(!std::is_pointer<hash_type>::value, "hash_accessor should return a reference to hash value");
+
+ typedef typename cds::opt::details::make_comparator_from<
+ hash_type,
+ traits,
+ feldman_hashset::bitwise_compare< hash_type >
+ >::type hash_comparator;
+
+ /// The size of hash_type in bytes, see \p traits::hash_size for explanation
+ static CDS_CONSTEXPR size_t const c_hash_size = traits::hash_size == 0 ? sizeof( hash_type ) : static_cast<size_t>( traits::hash_size );
+
+ typedef typename std::conditional<
+ std::is_same< typename traits::hash_splitter, cds::opt::none >::value,
+ typename cds::algo::select_splitter< hash_type, c_hash_size >::type,
+ typename traits::hash_splitter
+ >::type hash_splitter;
+
+ enum node_flags {
+ flag_array_converting = 1, ///< the cell is converting from data node to an array node
+ flag_array_node = 2 ///< the cell is a pointer to an array node
+ };
+
+ protected:
+
+ typedef cds::details::marked_ptr< value_type, 3 > node_ptr;
+ typedef atomics::atomic< node_ptr > atomic_node_ptr;
+
+ struct array_node {
+ array_node * const pParent; ///< parent array node
+ size_t const idxParent; ///< index in parent array node
+ atomic_node_ptr nodes[1]; ///< node array
+
+ array_node(array_node * parent, size_t idx)
+ : pParent(parent)
+ , idxParent(idx)
+ {}
+
+ array_node() = delete;
+ array_node(array_node const&) = delete;
+ array_node(array_node&&) = delete;
+ };
+
+ typedef cds::details::Allocator< array_node, node_allocator > cxx_array_node_allocator;
+
+ struct traverse_data {
+ hash_splitter splitter;
+ array_node * pArr;
+ typename hash_splitter::uint_type nSlot;
+ size_t nHeight;
+
+ traverse_data( hash_type const& hash, multilevel_array& arr )
+ : splitter( hash )
+ {
+ reset( arr );
+ }
+
+ void reset( multilevel_array& arr )
+ {
+ splitter.reset();
+ pArr = arr.head();
+ nSlot = splitter.cut( static_cast<unsigned>( arr.metrics().head_node_size_log ));
+ assert( static_cast<size_t>( nSlot ) < arr.metrics().head_node_size );
+ nHeight = 1;
+ }
+ };
+
+ protected:
+ feldman_hashset::details::metrics const m_Metrics;
+ array_node * m_Head;
+ mutable stat m_Stat;
+
+ public:
+ multilevel_array(size_t head_bits, size_t array_bits )
+ : m_Metrics(feldman_hashset::details::metrics::make( head_bits, array_bits, c_hash_size ))
+ , m_Head( alloc_head_node())
+ {
+ assert( hash_splitter::is_correct( static_cast<unsigned>( metrics().head_node_size_log )));
+ assert( hash_splitter::is_correct( static_cast<unsigned>( metrics().array_node_size_log )));
+ }
+
+ ~multilevel_array()
+ {
+ destroy_tree();
+ free_array_node( m_Head, head_size());
+ }
+
+ node_ptr traverse(traverse_data& pos)
+ {
+ back_off bkoff;
+ while (true) {
+ node_ptr slot = pos.pArr->nodes[pos.nSlot].load(memory_model::memory_order_acquire);
+ if ( slot.bits() == flag_array_node ) {
+ // array node, go down the tree
+ assert(slot.ptr() != nullptr);
+ assert( !pos.splitter.eos());
+ pos.nSlot = pos.splitter.cut( static_cast<unsigned>( metrics().array_node_size_log ));
+ assert( pos.nSlot < metrics().array_node_size );
+ pos.pArr = to_array(slot.ptr());
+ ++pos.nHeight;
+ }
+ else if (slot.bits() == flag_array_converting) {
+ // the slot is converting to array node right now
+ bkoff();
+ stats().onSlotConverting();
+ }
+ else {
+ // data node
+ assert(slot.bits() == 0);
+ return slot;
+ }
+ } // while
+ }
+
+ size_t head_size() const
+ {
+ return m_Metrics.head_node_size;
+ }
+
+ size_t array_node_size() const
+ {
+ return m_Metrics.array_node_size;
+ }
+
+ void get_level_statistics(std::vector< feldman_hashset::level_statistics>& stat) const
+ {
+ stat.clear();
+ gather_level_statistics(stat, 0, m_Head, head_size());
+ }
+
+ protected:
+ array_node * head() const
+ {
+ return m_Head;
+ }
+
+ stat& stats() const
+ {
+ return m_Stat;
+ }
+
+ feldman_hashset::details::metrics const& metrics() const
+ {
+ return m_Metrics;
+ }
+
+ void destroy_tree()
+ {
+ // The function is not thread-safe. For use in dtor only
+ // Destroy all array nodes
+ destroy_array_nodes(m_Head, head_size());
+ }
+
+ void destroy_array_nodes(array_node * pArr, size_t nSize)
+ {
+ for (atomic_node_ptr * p = pArr->nodes, *pLast = p + nSize; p != pLast; ++p) {
+ node_ptr slot = p->load(memory_model::memory_order_relaxed);
+ if (slot.bits() == flag_array_node) {
+ destroy_array_nodes( to_array(slot.ptr()), array_node_size());
+ free_array_node( to_array( slot.ptr()), array_node_size());
+ p->store(node_ptr(), memory_model::memory_order_relaxed);
+ }
+ }
+ }
+
+ static array_node * alloc_array_node(size_t nSize, array_node * pParent, size_t idxParent)
+ {
+ array_node * pNode = cxx_array_node_allocator().NewBlock(sizeof(array_node) + sizeof(atomic_node_ptr) * (nSize - 1), pParent, idxParent);
+ new (pNode->nodes) atomic_node_ptr[nSize];
+ return pNode;
+ }
+
+ array_node * alloc_head_node() const
+ {
+ return alloc_array_node(head_size(), nullptr, 0);
+ }
+
+ array_node * alloc_array_node(array_node * pParent, size_t idxParent) const
+ {
+ return alloc_array_node(array_node_size(), pParent, idxParent);
+ }
+
+ static void free_array_node( array_node * parr, size_t /*nSize*/ )
+ {
+ cxx_array_node_allocator().Delete( parr, 1 );
+ }
+
+ union converter {
+ value_type * pData;
+ array_node * pArr;
+
+ converter(value_type * p)
+ : pData(p)
+ {}
+
+ converter(array_node * p)
+ : pArr(p)
+ {}
+ };
+
+ static array_node * to_array(value_type * p)
+ {
+ return converter(p).pArr;
+ }
+ static value_type * to_node(array_node * p)
+ {
+ return converter(p).pData;
+ }
+
+ void gather_level_statistics(std::vector<feldman_hashset::level_statistics>& stat, size_t nLevel, array_node * pArr, size_t nSize) const
+ {
+ if (stat.size() <= nLevel) {
+ stat.resize(nLevel + 1);
+ stat[nLevel].node_capacity = nSize;
+ }
+
+ ++stat[nLevel].array_node_count;
+ for (atomic_node_ptr * p = pArr->nodes, *pLast = p + nSize; p != pLast; ++p) {
+ node_ptr slot = p->load(memory_model::memory_order_relaxed);
+ if (slot.bits()) {
+ ++stat[nLevel].array_cell_count;
+ if (slot.bits() == flag_array_node)
+ gather_level_statistics(stat, nLevel + 1, to_array(slot.ptr()), array_node_size());
+ }
+ else if (slot.ptr())
+ ++stat[nLevel].data_cell_count;
+ else
+ ++stat[nLevel].empty_cell_count;
+ }
+ }
+
+ bool expand_slot( traverse_data& pos, node_ptr current)
+ {
+ assert( !pos.splitter.eos() );
+ return expand_slot( pos.pArr, pos.nSlot, current, pos.splitter.bit_offset());
+ }
+
+ private:
+ bool expand_slot(array_node * pParent, size_t idxParent, node_ptr current, size_t nOffset)
+ {
+ assert(current.bits() == 0);
+ assert(current.ptr());
+
+ array_node * pArr = alloc_array_node(pParent, idxParent);
+
+ node_ptr cur(current.ptr());
+ atomic_node_ptr& slot = pParent->nodes[idxParent];
+ if (!slot.compare_exchange_strong(cur, cur | flag_array_converting, memory_model::memory_order_release, atomics::memory_order_relaxed))
+ {
+ stats().onExpandNodeFailed();
+ free_array_node( pArr, array_node_size());
+ return false;
+ }
+
+ typename hash_splitter::uint_type idx = hash_splitter( hash_accessor()(*current.ptr()), nOffset ).cut(
+ static_cast<unsigned>( m_Metrics.array_node_size_log ));
+ pArr->nodes[idx].store(current, memory_model::memory_order_release);
+
+ cur = cur | flag_array_converting;
+ CDS_VERIFY(
+ slot.compare_exchange_strong(cur, node_ptr(to_node(pArr), flag_array_node), memory_model::memory_order_release, atomics::memory_order_relaxed)
+ );
+
+ stats().onExpandNodeSuccess();
+ stats().onArrayNodeCreated();
+ return true;
+ }
+ };
+ //@endcond
} // namespace feldman_hashset
//@cond
projects / libcds.git / blobdiff