view src/share/vm/utilities/hashtable.hpp @ 1994:6cd6d394f280

7001033: assert(gch->gc_cause() == GCCause::_scavenge_alot || !gch->incremental_collection_failed()) 7002546: regression on SpecJbb2005 on 7b118 comparing to 7b117 on small heaps Summary: Relaxed assertion checking related to incremental_collection_failed flag to allow for ExplicitGCInvokesConcurrent behaviour where we do not want a failing scavenge to bail to a stop-world collection. Parameterized incremental_collection_will_fail() so we can selectively use, or not use, as appropriate, the statistical prediction at specific use sites. This essentially reverts the scavenge bail-out logic to what it was prior to some recent changes that had inadvertently started using the statistical prediction which can be noisy in the presence of bursty loads. Added some associated verbose non-product debugging messages. Reviewed-by: johnc, tonyp
author ysr
date Tue, 07 Dec 2010 21:55:53 -0800
parents f95d63e2154a
children 3582bf76420e
line wrap: on
line source

/*
 * Copyright (c) 2003, 2010, Oracle and/or its affiliates. All rights reserved.
 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
 *
 * This code is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License version 2 only, as
 * published by the Free Software Foundation.
 *
 * This code is distributed in the hope that it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
 * version 2 for more details (a copy is included in the LICENSE file that
 * accompanied this code).
 *
 * You should have received a copy of the GNU General Public License version
 * 2 along with this work; if not, write to the Free Software Foundation,
 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
 *
 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
 * or visit www.oracle.com if you need additional information or have any
 * questions.
 *
 */

#ifndef SHARE_VM_UTILITIES_HASHTABLE_HPP
#define SHARE_VM_UTILITIES_HASHTABLE_HPP

#include "memory/allocation.hpp"
#include "oops/oop.hpp"
#include "oops/symbolOop.hpp"
#include "runtime/handles.hpp"

// This is a generic hashtable, designed to be used for the symbol
// and string tables.
//
// It is implemented as an open hash table with a fixed number of buckets.
//
// %note:
//  - TableEntrys are allocated in blocks to reduce the space overhead.



class BasicHashtableEntry : public CHeapObj {
  friend class VMStructs;
private:
  unsigned int         _hash;           // 32-bit hash for item

  // Link to next element in the linked list for this bucket.  EXCEPT
  // bit 0 set indicates that this entry is shared and must not be
  // unlinked from the table. Bit 0 is set during the dumping of the
  // archive. Since shared entries are immutable, _next fields in the
  // shared entries will not change.  New entries will always be
  // unshared and since pointers are align, bit 0 will always remain 0
  // with no extra effort.
  BasicHashtableEntry* _next;

  // Windows IA64 compiler requires subclasses to be able to access these
protected:
  // Entry objects should not be created, they should be taken from the
  // free list with BasicHashtable.new_entry().
  BasicHashtableEntry() { ShouldNotReachHere(); }
  // Entry objects should not be destroyed.  They should be placed on
  // the free list instead with BasicHashtable.free_entry().
  ~BasicHashtableEntry() { ShouldNotReachHere(); }

public:

  unsigned int hash() const             { return _hash; }
  void set_hash(unsigned int hash)      { _hash = hash; }
  unsigned int* hash_addr()             { return &_hash; }

  static BasicHashtableEntry* make_ptr(BasicHashtableEntry* p) {
    return (BasicHashtableEntry*)((intptr_t)p & -2);
  }

  BasicHashtableEntry* next() const {
    return make_ptr(_next);
  }

  void set_next(BasicHashtableEntry* next) {
    _next = next;
  }

  BasicHashtableEntry** next_addr() {
    return &_next;
  }

  bool is_shared() const {
    return ((intptr_t)_next & 1) != 0;
  }

  void set_shared() {
    _next = (BasicHashtableEntry*)((intptr_t)_next | 1);
  }
};



class HashtableEntry : public BasicHashtableEntry {
  friend class VMStructs;
private:
  oop               _literal;          // ref to item in table.

public:
  // Literal
  oop literal() const                   { return _literal; }
  oop* literal_addr()                   { return &_literal; }
  void set_literal(oop s)               { _literal = s; }

  HashtableEntry* next() const {
    return (HashtableEntry*)BasicHashtableEntry::next();
  }
  HashtableEntry** next_addr() {
    return (HashtableEntry**)BasicHashtableEntry::next_addr();
  }
};



class HashtableBucket : public CHeapObj {
  friend class VMStructs;
private:
  // Instance variable
  BasicHashtableEntry*       _entry;

public:
  // Accessing
  void clear()                        { _entry = NULL; }

  // The following methods use order access methods to avoid race
  // conditions in multiprocessor systems.
  BasicHashtableEntry* get_entry() const;
  void set_entry(BasicHashtableEntry* l);

  // The following method is not MT-safe and must be done under lock.
  BasicHashtableEntry** entry_addr()  { return &_entry; }
};


class BasicHashtable : public CHeapObj {
  friend class VMStructs;

public:
  BasicHashtable(int table_size, int entry_size);
  BasicHashtable(int table_size, int entry_size,
                 HashtableBucket* buckets, int number_of_entries);

  // Sharing support.
  void copy_buckets(char** top, char* end);
  void copy_table(char** top, char* end);

  // Bucket handling
  int hash_to_index(unsigned int full_hash) {
    int h = full_hash % _table_size;
    assert(h >= 0 && h < _table_size, "Illegal hash value");
    return h;
  }

  // Reverse the order of elements in each of the buckets.
  void reverse();

private:
  // Instance variables
  int               _table_size;
  HashtableBucket*  _buckets;
  BasicHashtableEntry* _free_list;
  char*             _first_free_entry;
  char*             _end_block;
  int               _entry_size;
  int               _number_of_entries;

protected:

#ifdef ASSERT
  int               _lookup_count;
  int               _lookup_length;
  void verify_lookup_length(double load);
#endif

  void initialize(int table_size, int entry_size, int number_of_entries);

  // Accessor
  int entry_size() const { return _entry_size; }
  int table_size() { return _table_size; }

  // The following method is MT-safe and may be used with caution.
  BasicHashtableEntry* bucket(int i);

  // The following method is not MT-safe and must be done under lock.
  BasicHashtableEntry** bucket_addr(int i) { return _buckets[i].entry_addr(); }

  // Table entry management
  BasicHashtableEntry* new_entry(unsigned int hashValue);

public:
  void set_entry(int index, BasicHashtableEntry* entry);

  void add_entry(int index, BasicHashtableEntry* entry);

  void free_entry(BasicHashtableEntry* entry);

  int number_of_entries() { return _number_of_entries; }

  void verify() PRODUCT_RETURN;
};


class Hashtable : public BasicHashtable {
  friend class VMStructs;

public:
  Hashtable(int table_size, int entry_size)
    : BasicHashtable(table_size, entry_size) { }

  Hashtable(int table_size, int entry_size,
                   HashtableBucket* buckets, int number_of_entries)
    : BasicHashtable(table_size, entry_size, buckets, number_of_entries) { }

  // Invoke "f->do_oop" on the locations of all oops in the table.
  void oops_do(OopClosure* f);

  // Debugging
  void print()               PRODUCT_RETURN;

  // GC support
  //   Delete pointers to otherwise-unreachable objects.
  void unlink(BoolObjectClosure* cl);

  // Reverse the order of elements in each of the buckets. Hashtable
  // entries which refer to objects at a lower address than 'boundary'
  // are separated from those which refer to objects at higher
  // addresses, and appear first in the list.
  void reverse(void* boundary = NULL);

protected:

  static unsigned int hash_symbol(const char* s, int len);

  unsigned int compute_hash(symbolHandle name) {
    return (unsigned int) name->identity_hash();
  }

  int index_for(symbolHandle name) {
    return hash_to_index(compute_hash(name));
  }

  // Table entry management
  HashtableEntry* new_entry(unsigned int hashValue, oop obj);

  // The following method is MT-safe and may be used with caution.
  HashtableEntry* bucket(int i) {
    return (HashtableEntry*)BasicHashtable::bucket(i);
  }

  // The following method is not MT-safe and must be done under lock.
  HashtableEntry** bucket_addr(int i) {
    return (HashtableEntry**)BasicHashtable::bucket_addr(i);
  }
};


//  Verions of hashtable where two handles are used to compute the index.

class TwoOopHashtable : public Hashtable {
  friend class VMStructs;
protected:
  TwoOopHashtable(int table_size, int entry_size)
    : Hashtable(table_size, entry_size) {}

  TwoOopHashtable(int table_size, int entry_size, HashtableBucket* t,
                  int number_of_entries)
    : Hashtable(table_size, entry_size, t, number_of_entries) {}

public:
  unsigned int compute_hash(symbolHandle name, Handle loader) {
    // Be careful with identity_hash(), it can safepoint and if this
    // were one expression, the compiler could choose to unhandle each
    // oop before calling identity_hash() for either of them.  If the first
    // causes a GC, the next would fail.
    unsigned int name_hash = name->identity_hash();
    unsigned int loader_hash = loader.is_null() ? 0 : loader->identity_hash();
    return name_hash ^ loader_hash;
  }

  int index_for(symbolHandle name, Handle loader) {
    return hash_to_index(compute_hash(name, loader));
  }
};

#endif // SHARE_VM_UTILITIES_HASHTABLE_HPP