summaryrefslogtreecommitdiff
path: root/tools/build/src/engine/hash.c
blob: 36f8366685805aa3f32e0af88a3fe11677ee8137 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
/*
 * Copyright 1993, 1995 Christopher Seiwald.
 *
 * This file is part of Jam - see jam.c for Copyright information.
 */

/*
 * hash.c - simple in-memory hashing routines
 *
 * External routines:
 *     hashinit() - initialize a hash table, returning a handle
 *     hashitem() - find a record in the table, and optionally enter a new one
 *     hashdone() - free a hash table, given its handle
 *
 * Internal routines:
 *     hashrehash() - resize and rebuild hp->tab, the hash table
 */

#include "jam.h"
#include "hash.h"

#include "compile.h"

#include <assert.h>

/* */
#define HASH_DEBUG_PROFILE 1
/* */

/* Header attached to all hash table data items. */

typedef struct item ITEM;
struct item
{
    ITEM * next;
};

#define MAX_LISTS 32

struct hash
{
    /*
     * the hash table, just an array of item pointers
     */
    struct
    {
        int nel;
        ITEM * * base;
    } tab;

    int bloat;  /* tab.nel / items.nel */
    int inel;   /* initial number of elements */

    /*
     * the array of records, maintained by these routines - essentially a
     * microallocator
     */
    struct
    {
        int more;     /* how many more ITEMs fit in lists[ list ] */
        ITEM * free;  /* free list of items */
        char * next;  /* where to put more ITEMs in lists[ list ] */
        int size;     /* sizeof( ITEM ) + aligned datalen */
        int nel;      /* total ITEMs held by all lists[] */
        int list;     /* index into lists[] */

        struct
        {
            int nel;      /* total ITEMs held by this list */
            char * base;  /* base of ITEMs array */
        } lists[ MAX_LISTS ];
    } items;

    char const * name;  /* just for hashstats() */
};

static void hashrehash( struct hash * );
static void hashstat( struct hash * );

static unsigned int hash_keyval( OBJECT * key )
{
    return object_hash( key );
}

#define hash_bucket(hp, keyval) ((hp)->tab.base + ((keyval) % (hp)->tab.nel))

#define hash_data_key(data) (*(OBJECT * *)(data))
#define hash_item_data(item) ((HASHDATA *)((char *)item + sizeof(ITEM)))
#define hash_item_key(item) (hash_data_key(hash_item_data(item)))


#define ALIGNED(x) ((x + sizeof(ITEM) - 1) & ~(sizeof(ITEM) - 1))

/*
 * hashinit() - initialize a hash table, returning a handle
 */

struct hash * hashinit( int datalen, char const * name )
{
    struct hash * hp = (struct hash *)BJAM_MALLOC( sizeof( *hp ) );

    hp->bloat = 3;
    hp->tab.nel = 0;
    hp->tab.base = 0;
    hp->items.more = 0;
    hp->items.free = 0;
    hp->items.size = sizeof( ITEM ) + ALIGNED( datalen );
    hp->items.list = -1;
    hp->items.nel = 0;
    hp->inel = 11;  /* 47 */
    hp->name = name;

    return hp;
}


/*
 * hash_search() - Find the hash item for the given data.
 *
 * Returns a pointer to a hashed item with the given key. If given a 'previous'
 * pointer, makes it point to the item prior to the found item in the same
 * bucket or to 0 if our item is the first item in its bucket.
 */

static ITEM * hash_search( struct hash * hp, unsigned int keyval,
    OBJECT * keydata, ITEM * * previous )
{
    ITEM * i = *hash_bucket( hp, keyval );
    ITEM * p = 0;
    for ( ; i; i = i->next )
    {
        if ( object_equal( hash_item_key( i ), keydata ) )
        {
            if ( previous )
                *previous = p;
            return i;
        }
        p = i;
    }
    return 0;
}


/*
 * hash_insert() - insert a record in the table or return the existing one
 */

HASHDATA * hash_insert( struct hash * hp, OBJECT * key, int * found )
{
    ITEM * i;
    unsigned int keyval = hash_keyval( key );

    #ifdef HASH_DEBUG_PROFILE
    profile_frame prof[ 1 ];
    if ( DEBUG_PROFILE )
        profile_enter( 0, prof );
    #endif

    if ( !hp->items.more )
        hashrehash( hp );

    i = hash_search( hp, keyval, key, 0 );
    if ( i )
        *found = 1;
    else
    {
        ITEM * * base = hash_bucket( hp, keyval );

        /* Try to grab one from the free list. */
        if ( hp->items.free )
        {
            i = hp->items.free;
            hp->items.free = i->next;
            assert( !hash_item_key( i ) );
        }
        else
        {
            i = (ITEM *)hp->items.next;
            hp->items.next += hp->items.size;
        }
        --hp->items.more;
        i->next = *base;
        *base = i;
        *found = 0;
    }

    #ifdef HASH_DEBUG_PROFILE
    if ( DEBUG_PROFILE )
        profile_exit( prof );
    #endif

    return hash_item_data( i );
}


/*
 * hash_find() - find a record in the table or NULL if none exists
 */

HASHDATA * hash_find( struct hash * hp, OBJECT * key )
{
    ITEM * i;
    unsigned int keyval = hash_keyval( key );

    #ifdef HASH_DEBUG_PROFILE
    profile_frame prof[ 1 ];
    if ( DEBUG_PROFILE )
        profile_enter( 0, prof );
    #endif

    if ( !hp->items.nel )
    {
        #ifdef HASH_DEBUG_PROFILE
        if ( DEBUG_PROFILE )
            profile_exit( prof );
        #endif
        return 0;
    }

    i = hash_search( hp, keyval, key, 0 );

    #ifdef HASH_DEBUG_PROFILE
    if ( DEBUG_PROFILE )
        profile_exit( prof );
    #endif

    return i ? hash_item_data( i ) : 0;
}


/*
 * hashrehash() - resize and rebuild hp->tab, the hash table
 */

static void hashrehash( struct hash * hp )
{
    int i = ++hp->items.list;
    hp->items.more = i ? 2 * hp->items.nel : hp->inel;
    hp->items.next = (char *)BJAM_MALLOC( hp->items.more * hp->items.size );
    hp->items.free = 0;

    hp->items.lists[ i ].nel = hp->items.more;
    hp->items.lists[ i ].base = hp->items.next;
    hp->items.nel += hp->items.more;

    if ( hp->tab.base )
        BJAM_FREE( (char *)hp->tab.base );

    hp->tab.nel = hp->items.nel * hp->bloat;
    hp->tab.base = (ITEM * *)BJAM_MALLOC( hp->tab.nel * sizeof( ITEM * * ) );

    memset( (char *)hp->tab.base, '\0', hp->tab.nel * sizeof( ITEM * ) );

    for ( i = 0; i < hp->items.list; ++i )
    {
        int nel = hp->items.lists[ i ].nel;
        char * next = hp->items.lists[ i ].base;

        for ( ; nel--; next += hp->items.size )
        {
            ITEM * i = (ITEM *)next;
            ITEM * * ip = hp->tab.base + object_hash( hash_item_key( i ) ) %
                hp->tab.nel;
            /* code currently assumes rehashing only when there are no free
             * items
             */
            assert( hash_item_key( i ) );

            i->next = *ip;
            *ip = i;
        }
    }
}


void hashenumerate( struct hash * hp, void (* f)( void *, void * ), void * data
    )
{
    int i;
    for ( i = 0; i <= hp->items.list; ++i )
    {
        char * next = hp->items.lists[ i ].base;
        int nel = hp->items.lists[ i ].nel;
        if ( i == hp->items.list )
            nel -= hp->items.more;

        for ( ; nel--; next += hp->items.size )
        {
            ITEM * const i = (ITEM *)next;
            if ( hash_item_key( i ) != 0 )  /* Do not enumerate freed items. */
                f( hash_item_data( i ), data );
        }
    }
}


/*
 * hash_free() - free a hash table, given its handle
 */

void hash_free( struct hash * hp )
{
    int i;
    if ( !hp )
        return;
    if ( hp->tab.base )
        BJAM_FREE( (char *)hp->tab.base );
    for ( i = 0; i <= hp->items.list; ++i )
        BJAM_FREE( hp->items.lists[ i ].base );
    BJAM_FREE( (char *)hp );
}


static void hashstat( struct hash * hp )
{
    struct hashstats stats[ 1 ];
    hashstats_init( stats );
    hashstats_add( stats, hp );
    hashstats_print( stats, hp->name );
}


void hashstats_init( struct hashstats * stats )
{
    stats->count = 0;
    stats->num_items = 0;
    stats->tab_size = 0;
    stats->item_size = 0;
    stats->sets = 0;
    stats->num_hashes = 0;
}


void hashstats_add( struct hashstats * stats, struct hash * hp )
{
    if ( hp )
    {
        ITEM * * tab = hp->tab.base;
        int nel = hp->tab.nel;
        int count = 0;
        int sets = 0;
        int i;

        for ( i = 0; i < nel; ++i )
        {
            ITEM * item;
            int here = 0;
            for ( item = tab[ i ]; item; item = item->next )
                ++here;

            count += here;
            if ( here > 0 )
                ++sets;
        }

        stats->count += count;
        stats->sets += sets;
        stats->num_items += hp->items.nel;
        stats->tab_size += hp->tab.nel;
        stats->item_size = hp->items.size;
        ++stats->num_hashes;
    }
}


void hashstats_print( struct hashstats * stats, char const * name )
{
    printf( "%s table: %d+%d+%d (%dK+%luK+%luK) items+table+hash, %f density\n",
        name,
        stats->count,
        stats->num_items,
        stats->tab_size,
        stats->num_items * stats->item_size / 1024,
        (long unsigned)stats->tab_size * sizeof( ITEM * * ) / 1024,
        (long unsigned)stats->num_hashes * sizeof( struct hash ) / 1024,
        (float)stats->count / (float)stats->sets );
}


void hashdone( struct hash * hp )
{
    if ( !hp )
        return;
    if ( DEBUG_MEM || DEBUG_PROFILE )
        hashstat( hp );
    hash_free( hp );
}