src/share/vm/utilities/hashtable.hpp

Mon, 15 May 2017 12:20:15 +0200

author
tschatzl
date
Mon, 15 May 2017 12:20:15 +0200
changeset 8766
ce9a710b0f63
parent 7207
152cf4afc11f
child 8856
ac27a9c85bea
permissions
-rw-r--r--

8180048: Interned string and symbol table leak memory during parallel unlinking
Summary: Make appending found dead BasicHashtableEntrys to the free list atomic.
Reviewed-by: ehelin, shade

     1 /*
     2  * Copyright (c) 2003, 2014, Oracle and/or its affiliates. All rights reserved.
     3  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
     4  *
     5  * This code is free software; you can redistribute it and/or modify it
     6  * under the terms of the GNU General Public License version 2 only, as
     7  * published by the Free Software Foundation.
     8  *
     9  * This code is distributed in the hope that it will be useful, but WITHOUT
    10  * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
    11  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
    12  * version 2 for more details (a copy is included in the LICENSE file that
    13  * accompanied this code).
    14  *
    15  * You should have received a copy of the GNU General Public License version
    16  * 2 along with this work; if not, write to the Free Software Foundation,
    17  * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
    18  *
    19  * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
    20  * or visit www.oracle.com if you need additional information or have any
    21  * questions.
    22  *
    23  */
    25 #ifndef SHARE_VM_UTILITIES_HASHTABLE_HPP
    26 #define SHARE_VM_UTILITIES_HASHTABLE_HPP
    28 #include "classfile/classLoaderData.hpp"
    29 #include "memory/allocation.hpp"
    30 #include "oops/oop.hpp"
    31 #include "oops/symbol.hpp"
    32 #include "runtime/handles.hpp"
    34 // This is a generic hashtable, designed to be used for the symbol
    35 // and string tables.
    36 //
    37 // It is implemented as an open hash table with a fixed number of buckets.
    38 //
    39 // %note:
    40 //  - TableEntrys are allocated in blocks to reduce the space overhead.
    44 template <MEMFLAGS F> class BasicHashtableEntry : public CHeapObj<F> {
    45   friend class VMStructs;
    46 private:
    47   unsigned int         _hash;           // 32-bit hash for item
    49   // Link to next element in the linked list for this bucket.  EXCEPT
    50   // bit 0 set indicates that this entry is shared and must not be
    51   // unlinked from the table. Bit 0 is set during the dumping of the
    52   // archive. Since shared entries are immutable, _next fields in the
    53   // shared entries will not change.  New entries will always be
    54   // unshared and since pointers are align, bit 0 will always remain 0
    55   // with no extra effort.
    56   BasicHashtableEntry<F>* _next;
    58   // Windows IA64 compiler requires subclasses to be able to access these
    59 protected:
    60   // Entry objects should not be created, they should be taken from the
    61   // free list with BasicHashtable.new_entry().
    62   BasicHashtableEntry() { ShouldNotReachHere(); }
    63   // Entry objects should not be destroyed.  They should be placed on
    64   // the free list instead with BasicHashtable.free_entry().
    65   ~BasicHashtableEntry() { ShouldNotReachHere(); }
    67 public:
    69   unsigned int hash() const             { return _hash; }
    70   void set_hash(unsigned int hash)      { _hash = hash; }
    71   unsigned int* hash_addr()             { return &_hash; }
    73   static BasicHashtableEntry<F>* make_ptr(BasicHashtableEntry<F>* p) {
    74     return (BasicHashtableEntry*)((intptr_t)p & -2);
    75   }
    77   BasicHashtableEntry<F>* next() const {
    78     return make_ptr(_next);
    79   }
    81   void set_next(BasicHashtableEntry<F>* next) {
    82     _next = next;
    83   }
    85   BasicHashtableEntry<F>** next_addr() {
    86     return &_next;
    87   }
    89   bool is_shared() const {
    90     return ((intptr_t)_next & 1) != 0;
    91   }
    93   void set_shared() {
    94     _next = (BasicHashtableEntry<F>*)((intptr_t)_next | 1);
    95   }
    96 };
   100 template <class T, MEMFLAGS F> class HashtableEntry : public BasicHashtableEntry<F> {
   101   friend class VMStructs;
   102 private:
   103   T               _literal;          // ref to item in table.
   105 public:
   106   // Literal
   107   T literal() const                   { return _literal; }
   108   T* literal_addr()                   { return &_literal; }
   109   void set_literal(T s)               { _literal = s; }
   111   HashtableEntry* next() const {
   112     return (HashtableEntry*)BasicHashtableEntry<F>::next();
   113   }
   114   HashtableEntry** next_addr() {
   115     return (HashtableEntry**)BasicHashtableEntry<F>::next_addr();
   116   }
   117 };
   121 template <MEMFLAGS F> class HashtableBucket : public CHeapObj<F> {
   122   friend class VMStructs;
   123 private:
   124   // Instance variable
   125   BasicHashtableEntry<F>*       _entry;
   127 public:
   128   // Accessing
   129   void clear()                        { _entry = NULL; }
   131   // The following methods use order access methods to avoid race
   132   // conditions in multiprocessor systems.
   133   BasicHashtableEntry<F>* get_entry() const;
   134   void set_entry(BasicHashtableEntry<F>* l);
   136   // The following method is not MT-safe and must be done under lock.
   137   BasicHashtableEntry<F>** entry_addr()  { return &_entry; }
   138 };
   141 template <MEMFLAGS F> class BasicHashtable : public CHeapObj<F> {
   142   friend class VMStructs;
   144 public:
   145   BasicHashtable(int table_size, int entry_size);
   146   BasicHashtable(int table_size, int entry_size,
   147                  HashtableBucket<F>* buckets, int number_of_entries);
   149   // Sharing support.
   150   void copy_buckets(char** top, char* end);
   151   void copy_table(char** top, char* end);
   153   // Bucket handling
   154   int hash_to_index(unsigned int full_hash) {
   155     int h = full_hash % _table_size;
   156     assert(h >= 0 && h < _table_size, "Illegal hash value");
   157     return h;
   158   }
   160   // Reverse the order of elements in each of the buckets.
   161   void reverse();
   163 private:
   164   // Instance variables
   165   int               _table_size;
   166   HashtableBucket<F>*     _buckets;
   167   BasicHashtableEntry<F>* volatile _free_list;
   168   char*             _first_free_entry;
   169   char*             _end_block;
   170   int               _entry_size;
   171   volatile int      _number_of_entries;
   173 protected:
   175 #ifdef ASSERT
   176   int               _lookup_count;
   177   int               _lookup_length;
   178   void verify_lookup_length(double load);
   179 #endif
   181   void initialize(int table_size, int entry_size, int number_of_entries);
   183   // Accessor
   184   int entry_size() const { return _entry_size; }
   186   // The following method is MT-safe and may be used with caution.
   187   BasicHashtableEntry<F>* bucket(int i);
   189   // The following method is not MT-safe and must be done under lock.
   190   BasicHashtableEntry<F>** bucket_addr(int i) { return _buckets[i].entry_addr(); }
   192   // Attempt to get an entry from the free list
   193   BasicHashtableEntry<F>* new_entry_free_list();
   195   // Table entry management
   196   BasicHashtableEntry<F>* new_entry(unsigned int hashValue);
   198   // Used when moving the entry to another table
   199   // Clean up links, but do not add to free_list
   200   void unlink_entry(BasicHashtableEntry<F>* entry) {
   201     entry->set_next(NULL);
   202     --_number_of_entries;
   203   }
   205   // Move over freelist and free block for allocation
   206   void copy_freelist(BasicHashtable* src) {
   207     _free_list = src->_free_list;
   208     src->_free_list = NULL;
   209     _first_free_entry = src->_first_free_entry;
   210     src->_first_free_entry = NULL;
   211     _end_block = src->_end_block;
   212     src->_end_block = NULL;
   213   }
   215   // Free the buckets in this hashtable
   216   void free_buckets();
   218   // Helper data structure containing context for the bucket entry unlink process,
   219   // storing the unlinked buckets in a linked list.
   220   // Also avoids the need to pass around these four members as parameters everywhere.
   221   struct BucketUnlinkContext {
   222     int _num_processed;
   223     int _num_removed;
   224     // Head and tail pointers for the linked list of removed entries.
   225     BasicHashtableEntry<F>* _removed_head;
   226     BasicHashtableEntry<F>* _removed_tail;
   228     BucketUnlinkContext() : _num_processed(0), _num_removed(0), _removed_head(NULL), _removed_tail(NULL) {
   229     }
   231     void free_entry(BasicHashtableEntry<F>* entry);
   232   };
   233   // Add of bucket entries linked together in the given context to the global free list. This method
   234   // is mt-safe wrt. to other calls of this method.
   235   void bulk_free_entries(BucketUnlinkContext* context);
   236 public:
   237   int table_size() { return _table_size; }
   238   void set_entry(int index, BasicHashtableEntry<F>* entry);
   240   void add_entry(int index, BasicHashtableEntry<F>* entry);
   242   void free_entry(BasicHashtableEntry<F>* entry);
   244   int number_of_entries() { return _number_of_entries; }
   246   void verify() PRODUCT_RETURN;
   247 };
   250 template <class T, MEMFLAGS F> class Hashtable : public BasicHashtable<F> {
   251   friend class VMStructs;
   253 public:
   254   Hashtable(int table_size, int entry_size)
   255     : BasicHashtable<F>(table_size, entry_size) { }
   257   Hashtable(int table_size, int entry_size,
   258                    HashtableBucket<F>* buckets, int number_of_entries)
   259     : BasicHashtable<F>(table_size, entry_size, buckets, number_of_entries) { }
   261   // Debugging
   262   void print()               PRODUCT_RETURN;
   264   // Reverse the order of elements in each of the buckets. Hashtable
   265   // entries which refer to objects at a lower address than 'boundary'
   266   // are separated from those which refer to objects at higher
   267   // addresses, and appear first in the list.
   268   void reverse(void* boundary = NULL);
   270 protected:
   272   unsigned int compute_hash(Symbol* name) {
   273     return (unsigned int) name->identity_hash();
   274   }
   276   int index_for(Symbol* name) {
   277     return this->hash_to_index(compute_hash(name));
   278   }
   280   // Table entry management
   281   HashtableEntry<T, F>* new_entry(unsigned int hashValue, T obj);
   283   // The following method is MT-safe and may be used with caution.
   284   HashtableEntry<T, F>* bucket(int i) {
   285     return (HashtableEntry<T, F>*)BasicHashtable<F>::bucket(i);
   286   }
   288   // The following method is not MT-safe and must be done under lock.
   289   HashtableEntry<T, F>** bucket_addr(int i) {
   290     return (HashtableEntry<T, F>**)BasicHashtable<F>::bucket_addr(i);
   291   }
   293 };
   295 template <class T, MEMFLAGS F> class RehashableHashtable : public Hashtable<T, F> {
   296  protected:
   298   enum {
   299     rehash_count = 100,
   300     rehash_multiple = 60
   301   };
   303   // Check that the table is unbalanced
   304   bool check_rehash_table(int count);
   306  public:
   307   RehashableHashtable(int table_size, int entry_size)
   308     : Hashtable<T, F>(table_size, entry_size) { }
   310   RehashableHashtable(int table_size, int entry_size,
   311                    HashtableBucket<F>* buckets, int number_of_entries)
   312     : Hashtable<T, F>(table_size, entry_size, buckets, number_of_entries) { }
   315   // Function to move these elements into the new table.
   316   void move_to(RehashableHashtable<T, F>* new_table);
   317   static bool use_alternate_hashcode()  { return _seed != 0; }
   318   static juint seed()                    { return _seed; }
   320   static int literal_size(Symbol *symbol);
   321   static int literal_size(oop oop);
   323   // The following two are currently not used, but are needed anyway because some
   324   // C++ compilers (MacOS and Solaris) force the instantiation of
   325   // Hashtable<ConstantPool*, mtClass>::dump_table() even though we never call this function
   326   // in the VM code.
   327   static int literal_size(ConstantPool *cp) {Unimplemented(); return 0;}
   328   static int literal_size(Klass *k)         {Unimplemented(); return 0;}
   330   void dump_table(outputStream* st, const char *table_name);
   332  private:
   333   static juint _seed;
   334 };
   337 //  Verions of hashtable where two handles are used to compute the index.
   339 template <class T, MEMFLAGS F> class TwoOopHashtable : public Hashtable<T, F> {
   340   friend class VMStructs;
   341 protected:
   342   TwoOopHashtable(int table_size, int entry_size)
   343     : Hashtable<T, F>(table_size, entry_size) {}
   345   TwoOopHashtable(int table_size, int entry_size, HashtableBucket<F>* t,
   346                   int number_of_entries)
   347     : Hashtable<T, F>(table_size, entry_size, t, number_of_entries) {}
   349 public:
   350   unsigned int compute_hash(Symbol* name, ClassLoaderData* loader_data) {
   351     unsigned int name_hash = name->identity_hash();
   352     // loader is null with CDS
   353     assert(loader_data != NULL || UseSharedSpaces || DumpSharedSpaces,
   354            "only allowed with shared spaces");
   355     unsigned int loader_hash = loader_data == NULL ? 0 : loader_data->identity_hash();
   356     return name_hash ^ loader_hash;
   357   }
   359   int index_for(Symbol* name, ClassLoaderData* loader_data) {
   360     return this->hash_to_index(compute_hash(name, loader_data));
   361   }
   362 };
   364 #endif // SHARE_VM_UTILITIES_HASHTABLE_HPP

mercurial