duke@435: /*
jrose@1100:  * Copyright 1997-2009 Sun Microsystems, Inc.  All Rights Reserved.
duke@435:  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
duke@435:  *
duke@435:  * This code is free software; you can redistribute it and/or modify it
duke@435:  * under the terms of the GNU General Public License version 2 only, as
duke@435:  * published by the Free Software Foundation.
duke@435:  *
duke@435:  * This code is distributed in the hope that it will be useful, but WITHOUT
duke@435:  * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
duke@435:  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
duke@435:  * version 2 for more details (a copy is included in the LICENSE file that
duke@435:  * accompanied this code).
duke@435:  *
duke@435:  * You should have received a copy of the GNU General Public License version
duke@435:  * 2 along with this work; if not, write to the Free Software Foundation,
duke@435:  * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
duke@435:  *
duke@435:  * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
duke@435:  * CA 95054 USA or visit www.sun.com if you need additional information or
duke@435:  * have any questions.
duke@435:  *
duke@435:  */
duke@435: 
duke@435: # include "incls/_precompiled.incl"
duke@435: # include "incls/_symbolTable.cpp.incl"
duke@435: 
duke@435: // --------------------------------------------------------------------------
duke@435: 
duke@435: SymbolTable* SymbolTable::_the_table = NULL;
duke@435: 
duke@435: // Lookup a symbol in a bucket.
duke@435: 
duke@435: symbolOop SymbolTable::lookup(int index, const char* name,
duke@435:                               int len, unsigned int hash) {
duke@435:   for (HashtableEntry* e = bucket(index); e != NULL; e = e->next()) {
duke@435:     if (e->hash() == hash) {
duke@435:       symbolOop sym = symbolOop(e->literal());
duke@435:       if (sym->equals(name, len)) {
duke@435:         return sym;
duke@435:       }
duke@435:     }
duke@435:   }
duke@435:   return NULL;
duke@435: }
duke@435: 
duke@435: 
duke@435: // We take care not to be blocking while holding the
duke@435: // SymbolTable_lock. Otherwise, the system might deadlock, since the
duke@435: // symboltable is used during compilation (VM_thread) The lock free
duke@435: // synchronization is simplified by the fact that we do not delete
duke@435: // entries in the symbol table during normal execution (only during
duke@435: // safepoints).
duke@435: 
duke@435: symbolOop SymbolTable::lookup(const char* name, int len, TRAPS) {
duke@435:   unsigned int hashValue = hash_symbol(name, len);
duke@435:   int index = the_table()->hash_to_index(hashValue);
duke@435: 
duke@435:   symbolOop s = the_table()->lookup(index, name, len, hashValue);
duke@435: 
duke@435:   // Found
duke@435:   if (s != NULL) return s;
duke@435: 
duke@435:   // Otherwise, add to symbol to table
duke@435:   return the_table()->basic_add(index, (u1*)name, len, hashValue, CHECK_NULL);
duke@435: }
duke@435: 
duke@435: symbolOop SymbolTable::lookup(symbolHandle sym, int begin, int end, TRAPS) {
duke@435:   char* buffer;
duke@435:   int index, len;
duke@435:   unsigned int hashValue;
duke@435:   char* name;
duke@435:   {
duke@435:     debug_only(No_Safepoint_Verifier nsv;)
duke@435: 
duke@435:     name = (char*)sym->base() + begin;
duke@435:     len = end - begin;
duke@435:     hashValue = hash_symbol(name, len);
duke@435:     index = the_table()->hash_to_index(hashValue);
duke@435:     symbolOop s = the_table()->lookup(index, name, len, hashValue);
duke@435: 
duke@435:     // Found
duke@435:     if (s != NULL) return s;
duke@435:   }
duke@435: 
duke@435:   // Otherwise, add to symbol to table. Copy to a C string first.
duke@435:   char stack_buf[128];
duke@435:   ResourceMark rm(THREAD);
duke@435:   if (len <= 128) {
duke@435:     buffer = stack_buf;
duke@435:   } else {
duke@435:     buffer = NEW_RESOURCE_ARRAY_IN_THREAD(THREAD, char, len);
duke@435:   }
duke@435:   for (int i=0; i<len; i++) {
duke@435:     buffer[i] = name[i];
duke@435:   }
duke@435:   // Make sure there is no safepoint in the code above since name can't move.
duke@435:   // We can't include the code in No_Safepoint_Verifier because of the
duke@435:   // ResourceMark.
duke@435: 
duke@435:   return the_table()->basic_add(index, (u1*)buffer, len, hashValue, CHECK_NULL);
duke@435: }
duke@435: 
duke@435: symbolOop SymbolTable::lookup_only(const char* name, int len,
duke@435:                                    unsigned int& hash) {
duke@435:   hash = hash_symbol(name, len);
duke@435:   int index = the_table()->hash_to_index(hash);
duke@435: 
duke@435:   return the_table()->lookup(index, name, len, hash);
duke@435: }
duke@435: 
jrose@1100: // Suggestion: Push unicode-based lookup all the way into the hashing
jrose@1100: // and probing logic, so there is no need for convert_to_utf8 until
jrose@1100: // an actual new symbolOop is created.
jrose@1100: symbolOop SymbolTable::lookup_unicode(const jchar* name, int utf16_length, TRAPS) {
jrose@1100:   int utf8_length = UNICODE::utf8_length((jchar*) name, utf16_length);
jrose@1100:   char stack_buf[128];
jrose@1100:   if (utf8_length < (int) sizeof(stack_buf)) {
jrose@1100:     char* chars = stack_buf;
jrose@1100:     UNICODE::convert_to_utf8(name, utf16_length, chars);
jrose@1100:     return lookup(chars, utf8_length, THREAD);
jrose@1100:   } else {
jrose@1100:     ResourceMark rm(THREAD);
jrose@1100:     char* chars = NEW_RESOURCE_ARRAY(char, utf8_length + 1);;
jrose@1100:     UNICODE::convert_to_utf8(name, utf16_length, chars);
jrose@1100:     return lookup(chars, utf8_length, THREAD);
jrose@1100:   }
jrose@1100: }
jrose@1100: 
jrose@1100: symbolOop SymbolTable::lookup_only_unicode(const jchar* name, int utf16_length,
jrose@1100:                                            unsigned int& hash) {
jrose@1100:   int utf8_length = UNICODE::utf8_length((jchar*) name, utf16_length);
jrose@1100:   char stack_buf[128];
jrose@1100:   if (utf8_length < (int) sizeof(stack_buf)) {
jrose@1100:     char* chars = stack_buf;
jrose@1100:     UNICODE::convert_to_utf8(name, utf16_length, chars);
jrose@1100:     return lookup_only(chars, utf8_length, hash);
jrose@1100:   } else {
jrose@1100:     ResourceMark rm;
jrose@1100:     char* chars = NEW_RESOURCE_ARRAY(char, utf8_length + 1);;
jrose@1100:     UNICODE::convert_to_utf8(name, utf16_length, chars);
jrose@1100:     return lookup_only(chars, utf8_length, hash);
jrose@1100:   }
jrose@1100: }
jrose@1100: 
duke@435: void SymbolTable::add(constantPoolHandle cp, int names_count,
duke@435:                       const char** names, int* lengths, int* cp_indices,
duke@435:                       unsigned int* hashValues, TRAPS) {
duke@435:   SymbolTable* table = the_table();
duke@435:   bool added = table->basic_add(cp, names_count, names, lengths,
duke@435:                                 cp_indices, hashValues, CHECK);
duke@435:   if (!added) {
duke@435:     // do it the hard way
duke@435:     for (int i=0; i<names_count; i++) {
duke@435:       int index = table->hash_to_index(hashValues[i]);
duke@435:       symbolOop sym = table->basic_add(index, (u1*)names[i], lengths[i],
duke@435:                                        hashValues[i], CHECK);
duke@435:       cp->symbol_at_put(cp_indices[i], sym);
duke@435:     }
duke@435:   }
duke@435: }
duke@435: 
duke@435: symbolOop SymbolTable::basic_add(int index, u1 *name, int len,
duke@435:                                  unsigned int hashValue, TRAPS) {
duke@435:   assert(!Universe::heap()->is_in_reserved(name) || GC_locker::is_active(),
duke@435:          "proposed name of symbol must be stable");
duke@435: 
duke@435:   // We assume that lookup() has been called already, that it failed,
duke@435:   // and symbol was not found.  We create the symbol here.
duke@435:   symbolKlass* sk  = (symbolKlass*) Universe::symbolKlassObj()->klass_part();
duke@435:   symbolOop s_oop = sk->allocate_symbol(name, len, CHECK_NULL);
duke@435:   symbolHandle sym (THREAD, s_oop);
duke@435: 
duke@435:   // Allocation must be done before grapping the SymbolTable_lock lock
duke@435:   MutexLocker ml(SymbolTable_lock, THREAD);
duke@435: 
duke@435:   assert(sym->equals((char*)name, len), "symbol must be properly initialized");
duke@435: 
duke@435:   // Since look-up was done lock-free, we need to check if another
duke@435:   // thread beat us in the race to insert the symbol.
duke@435: 
duke@435:   symbolOop test = lookup(index, (char*)name, len, hashValue);
duke@435:   if (test != NULL) {
twisti@1040:     // A race occurred and another thread introduced the symbol, this one
duke@435:     // will be dropped and collected.
duke@435:     return test;
duke@435:   }
duke@435: 
duke@435:   HashtableEntry* entry = new_entry(hashValue, sym());
duke@435:   add_entry(index, entry);
duke@435:   return sym();
duke@435: }
duke@435: 
duke@435: bool SymbolTable::basic_add(constantPoolHandle cp, int names_count,
duke@435:                             const char** names, int* lengths,
duke@435:                             int* cp_indices, unsigned int* hashValues,
duke@435:                             TRAPS) {
duke@435:   symbolKlass* sk  = (symbolKlass*) Universe::symbolKlassObj()->klass_part();
duke@435:   symbolOop sym_oops[symbol_alloc_batch_size];
duke@435:   bool allocated = sk->allocate_symbols(names_count, names, lengths,
duke@435:                                         sym_oops, CHECK_false);
duke@435:   if (!allocated) {
duke@435:     return false;
duke@435:   }
duke@435:   symbolHandle syms[symbol_alloc_batch_size];
duke@435:   int i;
duke@435:   for (i=0; i<names_count; i++) {
duke@435:     syms[i] = symbolHandle(THREAD, sym_oops[i]);
duke@435:   }
duke@435: 
duke@435:   // Allocation must be done before grabbing the SymbolTable_lock lock
duke@435:   MutexLocker ml(SymbolTable_lock, THREAD);
duke@435: 
duke@435:   for (i=0; i<names_count; i++) {
duke@435:     assert(syms[i]->equals(names[i], lengths[i]), "symbol must be properly initialized");
duke@435:     // Since look-up was done lock-free, we need to check if another
duke@435:     // thread beat us in the race to insert the symbol.
duke@435:     int index = hash_to_index(hashValues[i]);
duke@435:     symbolOop test = lookup(index, names[i], lengths[i], hashValues[i]);
duke@435:     if (test != NULL) {
twisti@1040:       // A race occurred and another thread introduced the symbol, this one
duke@435:       // will be dropped and collected. Use test instead.
duke@435:       cp->symbol_at_put(cp_indices[i], test);
duke@435:     } else {
duke@435:       symbolOop sym = syms[i]();
duke@435:       HashtableEntry* entry = new_entry(hashValues[i], sym);
duke@435:       add_entry(index, entry);
duke@435:       cp->symbol_at_put(cp_indices[i], sym);
duke@435:     }
duke@435:   }
duke@435: 
duke@435:   return true;
duke@435: }
duke@435: 
duke@435: 
duke@435: void SymbolTable::verify() {
duke@435:   for (int i = 0; i < the_table()->table_size(); ++i) {
duke@435:     HashtableEntry* p = the_table()->bucket(i);
duke@435:     for ( ; p != NULL; p = p->next()) {
duke@435:       symbolOop s = symbolOop(p->literal());
duke@435:       guarantee(s != NULL, "symbol is NULL");
duke@435:       s->verify();
duke@435:       guarantee(s->is_perm(), "symbol not in permspace");
duke@435:       unsigned int h = hash_symbol((char*)s->bytes(), s->utf8_length());
duke@435:       guarantee(p->hash() == h, "broken hash in symbol table entry");
duke@435:       guarantee(the_table()->hash_to_index(h) == i,
duke@435:                 "wrong index in symbol table");
duke@435:     }
duke@435:   }
duke@435: }
duke@435: 
duke@435: 
duke@435: //---------------------------------------------------------------------------
duke@435: // Non-product code
duke@435: 
duke@435: #ifndef PRODUCT
duke@435: 
duke@435: void SymbolTable::print_histogram() {
duke@435:   MutexLocker ml(SymbolTable_lock);
duke@435:   const int results_length = 100;
duke@435:   int results[results_length];
duke@435:   int i,j;
duke@435: 
duke@435:   // initialize results to zero
duke@435:   for (j = 0; j < results_length; j++) {
duke@435:     results[j] = 0;
duke@435:   }
duke@435: 
duke@435:   int total = 0;
duke@435:   int max_symbols = 0;
duke@435:   int out_of_range = 0;
duke@435:   for (i = 0; i < the_table()->table_size(); i++) {
duke@435:     HashtableEntry* p = the_table()->bucket(i);
duke@435:     for ( ; p != NULL; p = p->next()) {
duke@435:       int counter = symbolOop(p->literal())->utf8_length();
duke@435:       total += counter;
duke@435:       if (counter < results_length) {
duke@435:         results[counter]++;
duke@435:       } else {
duke@435:         out_of_range++;
duke@435:       }
duke@435:       max_symbols = MAX2(max_symbols, counter);
duke@435:     }
duke@435:   }
duke@435:   tty->print_cr("Symbol Table:");
duke@435:   tty->print_cr("%8s %5d", "Total  ", total);
duke@435:   tty->print_cr("%8s %5d", "Maximum", max_symbols);
duke@435:   tty->print_cr("%8s %3.2f", "Average",
duke@435:           ((float) total / (float) the_table()->table_size()));
duke@435:   tty->print_cr("%s", "Histogram:");
duke@435:   tty->print_cr(" %s %29s", "Length", "Number chains that length");
duke@435:   for (i = 0; i < results_length; i++) {
duke@435:     if (results[i] > 0) {
duke@435:       tty->print_cr("%6d %10d", i, results[i]);
duke@435:     }
duke@435:   }
duke@435:   int line_length = 70;
duke@435:   tty->print_cr("%s %30s", " Length", "Number chains that length");
duke@435:   for (i = 0; i < results_length; i++) {
duke@435:     if (results[i] > 0) {
duke@435:       tty->print("%4d", i);
duke@435:       for (j = 0; (j < results[i]) && (j < line_length);  j++) {
duke@435:         tty->print("%1s", "*");
duke@435:       }
duke@435:       if (j == line_length) {
duke@435:         tty->print("%1s", "+");
duke@435:       }
duke@435:       tty->cr();
duke@435:     }
duke@435:   }
duke@435:   tty->print_cr(" %s %d: %d\n", "Number chains longer than",
duke@435:                     results_length, out_of_range);
duke@435: }
duke@435: 
duke@435: #endif // PRODUCT
duke@435: 
duke@435: // --------------------------------------------------------------------------
duke@435: 
duke@435: #ifdef ASSERT
duke@435: class StableMemoryChecker : public StackObj {
duke@435:   enum { _bufsize = wordSize*4 };
duke@435: 
duke@435:   address _region;
duke@435:   jint    _size;
duke@435:   u1      _save_buf[_bufsize];
duke@435: 
duke@435:   int sample(u1* save_buf) {
duke@435:     if (_size <= _bufsize) {
duke@435:       memcpy(save_buf, _region, _size);
duke@435:       return _size;
duke@435:     } else {
duke@435:       // copy head and tail
duke@435:       memcpy(&save_buf[0],          _region,                      _bufsize/2);
duke@435:       memcpy(&save_buf[_bufsize/2], _region + _size - _bufsize/2, _bufsize/2);
duke@435:       return (_bufsize/2)*2;
duke@435:     }
duke@435:   }
duke@435: 
duke@435:  public:
duke@435:   StableMemoryChecker(const void* region, jint size) {
duke@435:     _region = (address) region;
duke@435:     _size   = size;
duke@435:     sample(_save_buf);
duke@435:   }
duke@435: 
duke@435:   bool verify() {
duke@435:     u1 check_buf[sizeof(_save_buf)];
duke@435:     int check_size = sample(check_buf);
duke@435:     return (0 == memcmp(_save_buf, check_buf, check_size));
duke@435:   }
duke@435: 
duke@435:   void set_region(const void* region) { _region = (address) region; }
duke@435: };
duke@435: #endif
duke@435: 
duke@435: 
duke@435: // --------------------------------------------------------------------------
duke@435: 
duke@435: 
duke@435: // Compute the hash value for a java.lang.String object which would
duke@435: // contain the characters passed in. This hash value is used for at
duke@435: // least two purposes.
duke@435: //
duke@435: // (a) As the hash value used by the StringTable for bucket selection
duke@435: //     and comparison (stored in the HashtableEntry structures).  This
duke@435: //     is used in the String.intern() method.
duke@435: //
duke@435: // (b) As the hash value used by the String object itself, in
duke@435: //     String.hashCode().  This value is normally calculate in Java code
duke@435: //     in the String.hashCode method(), but is precomputed for String
duke@435: //     objects in the shared archive file.
duke@435: //
duke@435: //     For this reason, THIS ALGORITHM MUST MATCH String.hashCode().
duke@435: 
duke@435: int StringTable::hash_string(jchar* s, int len) {
duke@435:   unsigned h = 0;
duke@435:   while (len-- > 0) {
duke@435:     h = 31*h + (unsigned) *s;
duke@435:     s++;
duke@435:   }
duke@435:   return h;
duke@435: }
duke@435: 
duke@435: 
duke@435: StringTable* StringTable::_the_table = NULL;
duke@435: 
duke@435: oop StringTable::lookup(int index, jchar* name,
duke@435:                         int len, unsigned int hash) {
duke@435:   for (HashtableEntry* l = bucket(index); l != NULL; l = l->next()) {
duke@435:     if (l->hash() == hash) {
duke@435:       if (java_lang_String::equals(l->literal(), name, len)) {
duke@435:         return l->literal();
duke@435:       }
duke@435:     }
duke@435:   }
duke@435:   return NULL;
duke@435: }
duke@435: 
duke@435: 
duke@435: oop StringTable::basic_add(int index, Handle string_or_null, jchar* name,
duke@435:                            int len, unsigned int hashValue, TRAPS) {
duke@435:   debug_only(StableMemoryChecker smc(name, len * sizeof(name[0])));
duke@435:   assert(!Universe::heap()->is_in_reserved(name) || GC_locker::is_active(),
duke@435:          "proposed name of symbol must be stable");
duke@435: 
duke@435:   Handle string;
duke@435:   // try to reuse the string if possible
duke@435:   if (!string_or_null.is_null() && string_or_null()->is_perm()) {
duke@435:     string = string_or_null;
duke@435:   } else {
duke@435:     string = java_lang_String::create_tenured_from_unicode(name, len, CHECK_NULL);
duke@435:   }
duke@435: 
duke@435:   // Allocation must be done before grapping the SymbolTable_lock lock
duke@435:   MutexLocker ml(StringTable_lock, THREAD);
duke@435: 
duke@435:   assert(java_lang_String::equals(string(), name, len),
duke@435:          "string must be properly initialized");
duke@435: 
duke@435:   // Since look-up was done lock-free, we need to check if another
duke@435:   // thread beat us in the race to insert the symbol.
duke@435: 
duke@435:   oop test = lookup(index, name, len, hashValue); // calls lookup(u1*, int)
duke@435:   if (test != NULL) {
duke@435:     // Entry already added
duke@435:     return test;
duke@435:   }
duke@435: 
duke@435:   HashtableEntry* entry = new_entry(hashValue, string());
duke@435:   add_entry(index, entry);
duke@435:   return string();
duke@435: }
duke@435: 
duke@435: 
duke@435: oop StringTable::lookup(symbolOop symbol) {
duke@435:   ResourceMark rm;
duke@435:   int length;
duke@435:   jchar* chars = symbol->as_unicode(length);
duke@435:   unsigned int hashValue = hash_string(chars, length);
duke@435:   int index = the_table()->hash_to_index(hashValue);
duke@435:   return the_table()->lookup(index, chars, length, hashValue);
duke@435: }
duke@435: 
duke@435: 
duke@435: oop StringTable::intern(Handle string_or_null, jchar* name,
duke@435:                         int len, TRAPS) {
duke@435:   unsigned int hashValue = hash_string(name, len);
duke@435:   int index = the_table()->hash_to_index(hashValue);
duke@435:   oop string = the_table()->lookup(index, name, len, hashValue);
duke@435: 
duke@435:   // Found
duke@435:   if (string != NULL) return string;
duke@435: 
duke@435:   // Otherwise, add to symbol to table
duke@435:   return the_table()->basic_add(index, string_or_null, name, len,
duke@435:                                 hashValue, CHECK_NULL);
duke@435: }
duke@435: 
duke@435: oop StringTable::intern(symbolOop symbol, TRAPS) {
duke@435:   if (symbol == NULL) return NULL;
duke@435:   ResourceMark rm(THREAD);
duke@435:   int length;
duke@435:   jchar* chars = symbol->as_unicode(length);
duke@435:   Handle string;
duke@435:   oop result = intern(string, chars, length, CHECK_NULL);
duke@435:   return result;
duke@435: }
duke@435: 
duke@435: 
duke@435: oop StringTable::intern(oop string, TRAPS)
duke@435: {
duke@435:   if (string == NULL) return NULL;
duke@435:   ResourceMark rm(THREAD);
duke@435:   int length;
duke@435:   Handle h_string (THREAD, string);
duke@435:   jchar* chars = java_lang_String::as_unicode_string(string, length);
duke@435:   oop result = intern(h_string, chars, length, CHECK_NULL);
duke@435:   return result;
duke@435: }
duke@435: 
duke@435: 
duke@435: oop StringTable::intern(const char* utf8_string, TRAPS) {
duke@435:   if (utf8_string == NULL) return NULL;
duke@435:   ResourceMark rm(THREAD);
duke@435:   int length = UTF8::unicode_length(utf8_string);
duke@435:   jchar* chars = NEW_RESOURCE_ARRAY(jchar, length);
duke@435:   UTF8::convert_to_unicode(utf8_string, chars, length);
duke@435:   Handle string;
duke@435:   oop result = intern(string, chars, length, CHECK_NULL);
duke@435:   return result;
duke@435: }
duke@435: 
duke@435: void StringTable::verify() {
duke@435:   for (int i = 0; i < the_table()->table_size(); ++i) {
duke@435:     HashtableEntry* p = the_table()->bucket(i);
duke@435:     for ( ; p != NULL; p = p->next()) {
duke@435:       oop s = p->literal();
duke@435:       guarantee(s != NULL, "interned string is NULL");
duke@435:       guarantee(s->is_perm(), "interned string not in permspace");
duke@435: 
duke@435:       int length;
duke@435:       jchar* chars = java_lang_String::as_unicode_string(s, length);
duke@435:       unsigned int h = hash_string(chars, length);
duke@435:       guarantee(p->hash() == h, "broken hash in string table entry");
duke@435:       guarantee(the_table()->hash_to_index(h) == i,
duke@435:                 "wrong index in string table");
duke@435:     }
duke@435:   }
duke@435: }