Mercurial > hg > truffle
annotate src/share/vm/libadt/dict.cpp @ 7090:05ce1defa4f9
Common out some parts of UnsafeLoad/Store in UnsafeAccess
author | Gilles Duboscq <duboscq@ssw.jku.at> |
---|---|
date | Thu, 29 Nov 2012 13:24:08 +0100 |
parents | f95d63e2154a |
children | 78bbf4d43a14 |
rev | line source |
---|---|
0 | 1 /* |
1972 | 2 * Copyright (c) 1997, 2010, Oracle and/or its affiliates. All rights reserved. |
0 | 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 * | |
1552
c18cbe5936b8
6941466: Oracle rebranding changes for Hotspot repositories
trims
parents:
628
diff
changeset
|
19 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA |
c18cbe5936b8
6941466: Oracle rebranding changes for Hotspot repositories
trims
parents:
628
diff
changeset
|
20 * or visit www.oracle.com if you need additional information or have any |
c18cbe5936b8
6941466: Oracle rebranding changes for Hotspot repositories
trims
parents:
628
diff
changeset
|
21 * questions. |
0 | 22 * |
23 */ | |
24 | |
1972 | 25 #include "precompiled.hpp" |
26 #include "libadt/dict.hpp" | |
27 #include "memory/allocation.inline.hpp" | |
28 #include "memory/resourceArea.hpp" | |
29 #include "runtime/thread.hpp" | |
0 | 30 |
1972 | 31 // Dictionaries - An Abstract Data Type |
0 | 32 |
33 // %%%%% includes not needed with AVM framework - Ungar | |
34 | |
35 // #include "port.hpp" | |
36 //IMPLEMENTATION | |
37 // #include "dict.hpp" | |
38 | |
39 #include <assert.h> | |
40 | |
41 // The iostream is not needed and it gets confused for gcc by the | |
42 // define of bool. | |
43 // | |
44 // #include <iostream.h> | |
45 | |
46 //------------------------------data----------------------------------------- | |
47 // String hash tables | |
48 #define MAXID 20 | |
49 static byte initflag = 0; // True after 1st initialization | |
50 static const char shft[MAXID] = {1,2,3,4,5,6,7,1,2,3,4,5,6,7,1,2,3,4,5,6}; | |
51 static short xsum[MAXID]; | |
52 | |
53 //------------------------------bucket--------------------------------------- | |
54 class bucket : public ResourceObj { | |
55 public: | |
56 uint _cnt, _max; // Size of bucket | |
57 void **_keyvals; // Array of keys and values | |
58 }; | |
59 | |
60 //------------------------------Dict----------------------------------------- | |
61 // The dictionary is kept has a hash table. The hash table is a even power | |
62 // of two, for nice modulo operations. Each bucket in the hash table points | |
63 // to a linear list of key-value pairs; each key & value is just a (void *). | |
64 // The list starts with a count. A hash lookup finds the list head, then a | |
65 // simple linear scan finds the key. If the table gets too full, it's | |
66 // doubled in size; the total amount of EXTRA times all hash functions are | |
67 // computed for the doubling is no more than the current size - thus the | |
68 // doubling in size costs no more than a constant factor in speed. | |
69 Dict::Dict(CmpKey initcmp, Hash inithash) : _hash(inithash), _cmp(initcmp), | |
70 _arena(Thread::current()->resource_area()) { | |
71 int i; | |
72 | |
73 // Precompute table of null character hashes | |
74 if( !initflag ) { // Not initializated yet? | |
75 xsum[0] = (1<<shft[0])+1; // Initialize | |
76 for(i=1; i<MAXID; i++) { | |
77 xsum[i] = (1<<shft[i])+1+xsum[i-1]; | |
78 } | |
79 initflag = 1; // Never again | |
80 } | |
81 | |
82 _size = 16; // Size is a power of 2 | |
83 _cnt = 0; // Dictionary is empty | |
84 _bin = (bucket*)_arena->Amalloc_4(sizeof(bucket)*_size); | |
85 memset(_bin,0,sizeof(bucket)*_size); | |
86 } | |
87 | |
88 Dict::Dict(CmpKey initcmp, Hash inithash, Arena *arena, int size) | |
89 : _hash(inithash), _cmp(initcmp), _arena(arena) { | |
90 int i; | |
91 | |
92 // Precompute table of null character hashes | |
93 if( !initflag ) { // Not initializated yet? | |
94 xsum[0] = (1<<shft[0])+1; // Initialize | |
95 for(i=1; i<MAXID; i++) { | |
96 xsum[i] = (1<<shft[i])+1+xsum[i-1]; | |
97 } | |
98 initflag = 1; // Never again | |
99 } | |
100 | |
101 i=16; | |
102 while( i < size ) i <<= 1; | |
103 _size = i; // Size is a power of 2 | |
104 _cnt = 0; // Dictionary is empty | |
105 _bin = (bucket*)_arena->Amalloc_4(sizeof(bucket)*_size); | |
106 memset(_bin,0,sizeof(bucket)*_size); | |
107 } | |
108 | |
109 //------------------------------~Dict------------------------------------------ | |
110 // Delete an existing dictionary. | |
111 Dict::~Dict() { | |
112 /* | |
113 tty->print("~Dict %d/%d: ",_cnt,_size); | |
114 for( uint i=0; i < _size; i++) // For complete new table do | |
115 tty->print("%d ",_bin[i]._cnt); | |
116 tty->print("\n");*/ | |
117 /*for( uint i=0; i<_size; i++ ) { | |
118 FREE_FAST( _bin[i]._keyvals ); | |
119 } */ | |
120 } | |
121 | |
122 //------------------------------Clear---------------------------------------- | |
123 // Zap to empty; ready for re-use | |
124 void Dict::Clear() { | |
125 _cnt = 0; // Empty contents | |
126 for( uint i=0; i<_size; i++ ) | |
127 _bin[i]._cnt = 0; // Empty buckets, but leave allocated | |
128 // Leave _size & _bin alone, under the assumption that dictionary will | |
129 // grow to this size again. | |
130 } | |
131 | |
132 //------------------------------doubhash--------------------------------------- | |
133 // Double hash table size. If can't do so, just suffer. If can, then run | |
134 // thru old hash table, moving things to new table. Note that since hash | |
135 // table doubled, exactly 1 new bit is exposed in the mask - so everything | |
136 // in the old table ends up on 1 of two lists in the new table; a hi and a | |
137 // lo list depending on the value of the bit. | |
138 void Dict::doubhash(void) { | |
139 uint oldsize = _size; | |
140 _size <<= 1; // Double in size | |
141 _bin = (bucket*)_arena->Arealloc( _bin, sizeof(bucket)*oldsize, sizeof(bucket)*_size ); | |
142 memset( &_bin[oldsize], 0, oldsize*sizeof(bucket) ); | |
143 // Rehash things to spread into new table | |
144 for( uint i=0; i < oldsize; i++) { // For complete OLD table do | |
145 bucket *b = &_bin[i]; // Handy shortcut for _bin[i] | |
146 if( !b->_keyvals ) continue; // Skip empties fast | |
147 | |
148 bucket *nb = &_bin[i+oldsize]; // New bucket shortcut | |
149 uint j = b->_max; // Trim new bucket to nearest power of 2 | |
150 while( j > b->_cnt ) j >>= 1; // above old bucket _cnt | |
151 if( !j ) j = 1; // Handle zero-sized buckets | |
152 nb->_max = j<<1; | |
153 // Allocate worst case space for key-value pairs | |
154 nb->_keyvals = (void**)_arena->Amalloc_4( sizeof(void *)*nb->_max*2 ); | |
155 uint nbcnt = 0; | |
156 | |
157 for( j=0; j<b->_cnt; j++ ) { // Rehash all keys in this bucket | |
158 void *key = b->_keyvals[j+j]; | |
159 if( (_hash( key ) & (_size-1)) != i ) { // Moving to hi bucket? | |
160 nb->_keyvals[nbcnt+nbcnt] = key; | |
161 nb->_keyvals[nbcnt+nbcnt+1] = b->_keyvals[j+j+1]; | |
162 nb->_cnt = nbcnt = nbcnt+1; | |
163 b->_cnt--; // Remove key/value from lo bucket | |
164 b->_keyvals[j+j ] = b->_keyvals[b->_cnt+b->_cnt ]; | |
165 b->_keyvals[j+j+1] = b->_keyvals[b->_cnt+b->_cnt+1]; | |
166 j--; // Hash compacted element also | |
167 } | |
168 } // End of for all key-value pairs in bucket | |
169 } // End of for all buckets | |
170 | |
171 | |
172 } | |
173 | |
174 //------------------------------Dict----------------------------------------- | |
175 // Deep copy a dictionary. | |
176 Dict::Dict( const Dict &d ) : _size(d._size), _cnt(d._cnt), _hash(d._hash),_cmp(d._cmp), _arena(d._arena) { | |
177 _bin = (bucket*)_arena->Amalloc_4(sizeof(bucket)*_size); | |
178 memcpy( _bin, d._bin, sizeof(bucket)*_size ); | |
179 for( uint i=0; i<_size; i++ ) { | |
180 if( !_bin[i]._keyvals ) continue; | |
181 _bin[i]._keyvals=(void**)_arena->Amalloc_4( sizeof(void *)*_bin[i]._max*2); | |
182 memcpy( _bin[i]._keyvals, d._bin[i]._keyvals,_bin[i]._cnt*2*sizeof(void*)); | |
183 } | |
184 } | |
185 | |
186 //------------------------------Dict----------------------------------------- | |
187 // Deep copy a dictionary. | |
188 Dict &Dict::operator =( const Dict &d ) { | |
189 if( _size < d._size ) { // If must have more buckets | |
190 _arena = d._arena; | |
191 _bin = (bucket*)_arena->Arealloc( _bin, sizeof(bucket)*_size, sizeof(bucket)*d._size ); | |
192 memset( &_bin[_size], 0, (d._size-_size)*sizeof(bucket) ); | |
193 _size = d._size; | |
194 } | |
195 uint i; | |
196 for( i=0; i<_size; i++ ) // All buckets are empty | |
197 _bin[i]._cnt = 0; // But leave bucket allocations alone | |
198 _cnt = d._cnt; | |
199 *(Hash*)(&_hash) = d._hash; | |
200 *(CmpKey*)(&_cmp) = d._cmp; | |
201 for( i=0; i<_size; i++ ) { | |
202 bucket *b = &d._bin[i]; // Shortcut to source bucket | |
203 for( uint j=0; j<b->_cnt; j++ ) | |
204 Insert( b->_keyvals[j+j], b->_keyvals[j+j+1] ); | |
205 } | |
206 return *this; | |
207 } | |
208 | |
209 //------------------------------Insert---------------------------------------- | |
210 // Insert or replace a key/value pair in the given dictionary. If the | |
211 // dictionary is too full, it's size is doubled. The prior value being | |
212 // replaced is returned (NULL if this is a 1st insertion of that key). If | |
213 // an old value is found, it's swapped with the prior key-value pair on the | |
214 // list. This moves a commonly searched-for value towards the list head. | |
215 void *Dict::Insert(void *key, void *val, bool replace) { | |
216 uint hash = _hash( key ); // Get hash key | |
217 uint i = hash & (_size-1); // Get hash key, corrected for size | |
218 bucket *b = &_bin[i]; // Handy shortcut | |
219 for( uint j=0; j<b->_cnt; j++ ) { | |
220 if( !_cmp(key,b->_keyvals[j+j]) ) { | |
221 if (!replace) { | |
222 return b->_keyvals[j+j+1]; | |
223 } else { | |
224 void *prior = b->_keyvals[j+j+1]; | |
225 b->_keyvals[j+j ] = key; // Insert current key-value | |
226 b->_keyvals[j+j+1] = val; | |
227 return prior; // Return prior | |
228 } | |
229 } | |
230 } | |
231 if( ++_cnt > _size ) { // Hash table is full | |
232 doubhash(); // Grow whole table if too full | |
233 i = hash & (_size-1); // Rehash | |
234 b = &_bin[i]; // Handy shortcut | |
235 } | |
236 if( b->_cnt == b->_max ) { // Must grow bucket? | |
237 if( !b->_keyvals ) { | |
238 b->_max = 2; // Initial bucket size | |
239 b->_keyvals = (void**)_arena->Amalloc_4(sizeof(void*) * b->_max * 2); | |
240 } else { | |
241 b->_keyvals = (void**)_arena->Arealloc(b->_keyvals, sizeof(void*) * b->_max * 2, sizeof(void*) * b->_max * 4); | |
242 b->_max <<= 1; // Double bucket | |
243 } | |
244 } | |
245 b->_keyvals[b->_cnt+b->_cnt ] = key; | |
246 b->_keyvals[b->_cnt+b->_cnt+1] = val; | |
247 b->_cnt++; | |
248 return NULL; // Nothing found prior | |
249 } | |
250 | |
251 //------------------------------Delete--------------------------------------- | |
252 // Find & remove a value from dictionary. Return old value. | |
253 void *Dict::Delete(void *key) { | |
254 uint i = _hash( key ) & (_size-1); // Get hash key, corrected for size | |
255 bucket *b = &_bin[i]; // Handy shortcut | |
256 for( uint j=0; j<b->_cnt; j++ ) | |
257 if( !_cmp(key,b->_keyvals[j+j]) ) { | |
258 void *prior = b->_keyvals[j+j+1]; | |
259 b->_cnt--; // Remove key/value from lo bucket | |
260 b->_keyvals[j+j ] = b->_keyvals[b->_cnt+b->_cnt ]; | |
261 b->_keyvals[j+j+1] = b->_keyvals[b->_cnt+b->_cnt+1]; | |
262 _cnt--; // One less thing in table | |
263 return prior; | |
264 } | |
265 return NULL; | |
266 } | |
267 | |
268 //------------------------------FindDict------------------------------------- | |
269 // Find a key-value pair in the given dictionary. If not found, return NULL. | |
270 // If found, move key-value pair towards head of list. | |
271 void *Dict::operator [](const void *key) const { | |
272 uint i = _hash( key ) & (_size-1); // Get hash key, corrected for size | |
273 bucket *b = &_bin[i]; // Handy shortcut | |
274 for( uint j=0; j<b->_cnt; j++ ) | |
275 if( !_cmp(key,b->_keyvals[j+j]) ) | |
276 return b->_keyvals[j+j+1]; | |
277 return NULL; | |
278 } | |
279 | |
280 //------------------------------CmpDict-------------------------------------- | |
281 // CmpDict compares two dictionaries; they must have the same keys (their | |
282 // keys must match using CmpKey) and they must have the same values (pointer | |
283 // comparison). If so 1 is returned, if not 0 is returned. | |
284 int32 Dict::operator ==(const Dict &d2) const { | |
285 if( _cnt != d2._cnt ) return 0; | |
286 if( _hash != d2._hash ) return 0; | |
287 if( _cmp != d2._cmp ) return 0; | |
288 for( uint i=0; i < _size; i++) { // For complete hash table do | |
289 bucket *b = &_bin[i]; // Handy shortcut | |
290 if( b->_cnt != d2._bin[i]._cnt ) return 0; | |
291 if( memcmp(b->_keyvals, d2._bin[i]._keyvals, b->_cnt*2*sizeof(void*) ) ) | |
292 return 0; // Key-value pairs must match | |
293 } | |
294 return 1; // All match, is OK | |
295 } | |
296 | |
297 //------------------------------print------------------------------------------ | |
298 // Handier print routine | |
299 void Dict::print() { | |
300 DictI i(this); // Moved definition in iterator here because of g++. | |
301 tty->print("Dict@0x%lx[%d] = {", this, _cnt); | |
302 for( ; i.test(); ++i ) { | |
303 tty->print("(0x%lx,0x%lx),", i._key, i._value); | |
304 } | |
305 tty->print_cr("}"); | |
306 } | |
307 | |
308 //------------------------------Hashing Functions---------------------------- | |
309 // Convert string to hash key. This algorithm implements a universal hash | |
310 // function with the multipliers frozen (ok, so it's not universal). The | |
311 // multipliers (and allowable characters) are all odd, so the resultant sum | |
605 | 312 // is odd - guaranteed not divisible by any power of two, so the hash tables |
0 | 313 // can be any power of two with good results. Also, I choose multipliers |
314 // that have only 2 bits set (the low is always set to be odd) so | |
315 // multiplication requires only shifts and adds. Characters are required to | |
316 // be in the range 0-127 (I double & add 1 to force oddness). Keys are | |
317 // limited to MAXID characters in length. Experimental evidence on 150K of | |
318 // C text shows excellent spreading of values for any size hash table. | |
319 int hashstr(const void *t) { | |
320 register char c, k = 0; | |
321 register int32 sum = 0; | |
322 register const char *s = (const char *)t; | |
323 | |
324 while( ((c = *s++) != '\0') && (k < MAXID-1) ) { // Get characters till null or MAXID-1 | |
325 c = (c<<1)+1; // Characters are always odd! | |
326 sum += c + (c<<shft[k++]); // Universal hash function | |
327 } | |
328 return (int)((sum+xsum[k]) >> 1); // Hash key, un-modulo'd table size | |
329 } | |
330 | |
331 //------------------------------hashptr-------------------------------------- | |
605 | 332 // Slimey cheap hash function; no guaranteed performance. Better than the |
0 | 333 // default for pointers, especially on MS-DOS machines. |
334 int hashptr(const void *key) { | |
335 #ifdef __TURBOC__ | |
336 return ((intptr_t)key >> 16); | |
337 #else // __TURBOC__ | |
338 return ((intptr_t)key >> 2); | |
339 #endif | |
340 } | |
341 | |
605 | 342 // Slimey cheap hash function; no guaranteed performance. |
0 | 343 int hashkey(const void *key) { |
344 return (intptr_t)key; | |
345 } | |
346 | |
347 //------------------------------Key Comparator Functions--------------------- | |
348 int32 cmpstr(const void *k1, const void *k2) { | |
349 return strcmp((const char *)k1,(const char *)k2); | |
350 } | |
351 | |
562
1580954e694c
6798785: Crash in OopFlow::build_oop_map: incorrect comparison of 64bit pointers
never
parents:
0
diff
changeset
|
352 // Cheap key comparator. |
0 | 353 int32 cmpkey(const void *key1, const void *key2) { |
562
1580954e694c
6798785: Crash in OopFlow::build_oop_map: incorrect comparison of 64bit pointers
never
parents:
0
diff
changeset
|
354 if (key1 == key2) return 0; |
1580954e694c
6798785: Crash in OopFlow::build_oop_map: incorrect comparison of 64bit pointers
never
parents:
0
diff
changeset
|
355 intptr_t delta = (intptr_t)key1 - (intptr_t)key2; |
1580954e694c
6798785: Crash in OopFlow::build_oop_map: incorrect comparison of 64bit pointers
never
parents:
0
diff
changeset
|
356 if (delta > 0) return 1; |
1580954e694c
6798785: Crash in OopFlow::build_oop_map: incorrect comparison of 64bit pointers
never
parents:
0
diff
changeset
|
357 return -1; |
0 | 358 } |
359 | |
360 //============================================================================= | |
361 //------------------------------reset------------------------------------------ | |
362 // Create an iterator and initialize the first variables. | |
363 void DictI::reset( const Dict *dict ) { | |
364 _d = dict; // The dictionary | |
365 _i = (uint)-1; // Before the first bin | |
366 _j = 0; // Nothing left in the current bin | |
367 ++(*this); // Step to first real value | |
368 } | |
369 | |
370 //------------------------------next------------------------------------------- | |
371 // Find the next key-value pair in the dictionary, or return a NULL key and | |
372 // value. | |
373 void DictI::operator ++(void) { | |
374 if( _j-- ) { // Still working in current bin? | |
375 _key = _d->_bin[_i]._keyvals[_j+_j]; | |
376 _value = _d->_bin[_i]._keyvals[_j+_j+1]; | |
377 return; | |
378 } | |
379 | |
380 while( ++_i < _d->_size ) { // Else scan for non-zero bucket | |
381 _j = _d->_bin[_i]._cnt; | |
382 if( !_j ) continue; | |
383 _j--; | |
384 _key = _d->_bin[_i]._keyvals[_j+_j]; | |
385 _value = _d->_bin[_i]._keyvals[_j+_j+1]; | |
386 return; | |
387 } | |
388 _key = _value = NULL; | |
389 } |