about summary refs log tree commit diff
path: root/db/hash/hash.h
blob: 1c599c112c54f26a5965b11eb975597365f72245 (plain) (blame)
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
/*-
 * Copyright (c) 1990, 1993, 1994
 *	The Regents of the University of California.  All rights reserved.
 *
 * This code is derived from software contributed to Berkeley by
 * Margo Seltzer.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. All advertising materials mentioning features or use of this software
 *    must display the following acknowledgement:
 *	This product includes software developed by the University of
 *	California, Berkeley and its contributors.
 * 4. Neither the name of the University nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 *
 *	@(#)hash.h	8.3 (Berkeley) 5/31/94
 */

/* Operations */
typedef enum {
	HASH_GET, HASH_PUT, HASH_PUTNEW, HASH_DELETE, HASH_FIRST, HASH_NEXT
} ACTION;

/* Buffer Management structures */
typedef struct _bufhead BUFHEAD;

struct _bufhead {
	BUFHEAD		*prev;		/* LRU links */
	BUFHEAD		*next;		/* LRU links */
	BUFHEAD		*ovfl;		/* Overflow page buffer header */
	u_int32_t	 addr;		/* Address of this page */
	char		*page;		/* Actual page data */
	char	 	flags;
#define	BUF_MOD		0x0001
#define BUF_DISK	0x0002
#define	BUF_BUCKET	0x0004
#define	BUF_PIN		0x0008
};

#define IS_BUCKET(X)	((X) & BUF_BUCKET)

typedef BUFHEAD **SEGMENT;

/* Hash Table Information */
typedef struct hashhdr {		/* Disk resident portion */
	int		magic;		/* Magic NO for hash tables */
	int		version;	/* Version ID */
	u_int32_t	lorder;		/* Byte Order */
	int		bsize;		/* Bucket/Page Size */
	int		bshift;		/* Bucket shift */
	int		dsize;		/* Directory Size */
	int		ssize;		/* Segment Size */
	int		sshift;		/* Segment shift */
	int		ovfl_point;	/* Where overflow pages are being
					 * allocated */
	int		last_freed;	/* Last overflow page freed */
	int		max_bucket;	/* ID of Maximum bucket in use */
	int		high_mask;	/* Mask to modulo into entire table */
	int		low_mask;	/* Mask to modulo into lower half of
					 * table */
	int		ffactor;	/* Fill factor */
	int		nkeys;		/* Number of keys in hash table */
	int		hdrpages;	/* Size of table header */
	int		h_charkey;	/* value of hash(CHARKEY) */
#define NCACHED	32			/* number of bit maps and spare
					 * points */
	int		spares[NCACHED];/* spare pages for overflow */
	u_int16_t	bitmaps[NCACHED];	/* address of overflow page
						 * bitmaps */
} HASHHDR;

typedef struct htab	 {		/* Memory resident data structure */
	HASHHDR 	hdr;		/* Header */
	int		nsegs;		/* Number of allocated segments */
	int		exsegs;		/* Number of extra allocated
					 * segments */
	u_int32_t			/* Hash function */
	    (*hash)__P((const void *, size_t));
	int		flags;		/* Flag values */
	int		fp;		/* File pointer */
	char		*tmp_buf;	/* Temporary Buffer for BIG data */
	char		*tmp_key;	/* Temporary Buffer for BIG keys */
	BUFHEAD 	*cpage;		/* Current page */
	int		cbucket;	/* Current bucket */
	int		cndx;		/* Index of next item on cpage */
	int		errnum;		/* Error Number -- for DBM
					 * compatability */
	int		new_file;	/* Indicates if fd is backing store
					 * or no */
	int		save_file;	/* Indicates whether we need to flush
					 * file at
					 * exit */
	u_int32_t	*mapp[NCACHED];	/* Pointers to page maps */
	int		nmaps;		/* Initial number of bitmaps */
	int		nbufs;		/* Number of buffers left to
					 * allocate */
	BUFHEAD 	bufhead;	/* Header of buffer lru list */
	SEGMENT 	*dir;		/* Hash Bucket directory */
} HTAB;

/*
 * Constants
 */
#define	MAX_BSIZE		65536		/* 2^16 */
#define MIN_BUFFERS		6
#define MINHDRSIZE		512
#define DEF_BUFSIZE		65536		/* 64 K */
#define DEF_BUCKET_SIZE		4096
#define DEF_BUCKET_SHIFT	12		/* log2(BUCKET) */
#define DEF_SEGSIZE		256
#define DEF_SEGSIZE_SHIFT	8		/* log2(SEGSIZE)	 */
#define DEF_DIRSIZE		256
#define DEF_FFACTOR		65536
#define MIN_FFACTOR		4
#define SPLTMAX			8
#define CHARKEY			"%$sniglet^&"
#define NUMKEY			1038583
#define BYTE_SHIFT		3
#define INT_TO_BYTE		2
#define INT_BYTE_SHIFT		5
#define ALL_SET			((u_int32_t)0xFFFFFFFF)
#define ALL_CLEAR		0

#define PTROF(X)	((BUFHEAD *)((ptrdiff_t)(X)&~0x3))
#define ISMOD(X)	((u_int32_t)(ptrdiff_t)(X)&0x1)
#define DOMOD(X)	((X) = (char *)((ptrdiff_t)(X)|0x1))
#define ISDISK(X)	((u_int32_t)(ptrdiff_t)(X)&0x2)
#define DODISK(X)	((X) = (char *)((ptrdiff_t)(X)|0x2))

#define BITS_PER_MAP	32

/* Given the address of the beginning of a big map, clear/set the nth bit */
#define CLRBIT(A, N)	((A)[(N)/BITS_PER_MAP] &= ~(1<<((N)%BITS_PER_MAP)))
#define SETBIT(A, N)	((A)[(N)/BITS_PER_MAP] |= (1<<((N)%BITS_PER_MAP)))
#define ISSET(A, N)	((A)[(N)/BITS_PER_MAP] & (1<<((N)%BITS_PER_MAP)))

/* Overflow management */
/*
 * Overflow page numbers are allocated per split point.  At each doubling of
 * the table, we can allocate extra pages.  So, an overflow page number has
 * the top 5 bits indicate which split point and the lower 11 bits indicate
 * which page at that split point is indicated (pages within split points are
 * numberered starting with 1).
 */

#define SPLITSHIFT	11
#define SPLITMASK	0x7FF
#define SPLITNUM(N)	(((u_int32_t)(N)) >> SPLITSHIFT)
#define OPAGENUM(N)	((N) & SPLITMASK)
#define	OADDR_OF(S,O)	((u_int32_t)((u_int32_t)(S) << SPLITSHIFT) + (O))

#define BUCKET_TO_PAGE(B) \
	(B) + hashp->HDRPAGES + ((B) ? hashp->SPARES[__log2((B)+1)-1] : 0)
#define OADDR_TO_PAGE(B) 	\
	BUCKET_TO_PAGE ( (1 << SPLITNUM((B))) -1 ) + OPAGENUM((B));

/*
 * page.h contains a detailed description of the page format.
 *
 * Normally, keys and data are accessed from offset tables in the top of
 * each page which point to the beginning of the key and data.  There are
 * four flag values which may be stored in these offset tables which indicate
 * the following:
 *
 *
 * OVFLPAGE	Rather than a key data pair, this pair contains
 *		the address of an overflow page.  The format of
 *		the pair is:
 *		    OVERFLOW_PAGE_NUMBER OVFLPAGE
 *
 * PARTIAL_KEY	This must be the first key/data pair on a page
 *		and implies that page contains only a partial key.
 *		That is, the key is too big to fit on a single page
 *		so it starts on this page and continues on the next.
 *		The format of the page is:
 *		    KEY_OFF PARTIAL_KEY OVFL_PAGENO OVFLPAGE
 *
 *		    KEY_OFF -- offset of the beginning of the key
 *		    PARTIAL_KEY -- 1
 *		    OVFL_PAGENO - page number of the next overflow page
 *		    OVFLPAGE -- 0
 *
 * FULL_KEY	This must be the first key/data pair on the page.  It
 *		is used in two cases.
 *
 *		Case 1:
 *		    There is a complete key on the page but no data
 *		    (because it wouldn't fit).  The next page contains
 *		    the data.
 *
 *		    Page format it:
 *		    KEY_OFF FULL_KEY OVFL_PAGENO OVFL_PAGE
 *
 *		    KEY_OFF -- offset of the beginning of the key
 *		    FULL_KEY -- 2
 *		    OVFL_PAGENO - page number of the next overflow page
 *		    OVFLPAGE -- 0
 *
 *		Case 2:
 *		    This page contains no key, but part of a large
 *		    data field, which is continued on the next page.
 *
 *		    Page format it:
 *		    DATA_OFF FULL_KEY OVFL_PAGENO OVFL_PAGE
 *
 *		    KEY_OFF -- offset of the beginning of the data on
 *				this page
 *		    FULL_KEY -- 2
 *		    OVFL_PAGENO - page number of the next overflow page
 *		    OVFLPAGE -- 0
 *
 * FULL_KEY_DATA
 *		This must be the first key/data pair on the page.
 *		There are two cases:
 *
 *		Case 1:
 *		    This page contains a key and the beginning of the
 *		    data field, but the data field is continued on the
 *		    next page.
 *
 *		    Page format is:
 *		    KEY_OFF FULL_KEY_DATA OVFL_PAGENO DATA_OFF
 *
 *		    KEY_OFF -- offset of the beginning of the key
 *		    FULL_KEY_DATA -- 3
 *		    OVFL_PAGENO - page number of the next overflow page
 *		    DATA_OFF -- offset of the beginning of the data
 *
 *		Case 2:
 *		    This page contains the last page of a big data pair.
 *		    There is no key, only the  tail end of the data
 *		    on this page.
 *
 *		    Page format is:
 *		    DATA_OFF FULL_KEY_DATA <OVFL_PAGENO> <OVFLPAGE>
 *
 *		    DATA_OFF -- offset of the beginning of the data on
 *				this page
 *		    FULL_KEY_DATA -- 3
 *		    OVFL_PAGENO - page number of the next overflow page
 *		    OVFLPAGE -- 0
 *
 *		    OVFL_PAGENO and OVFLPAGE are optional (they are
 *		    not present if there is no next page).
 */

#define OVFLPAGE	0
#define PARTIAL_KEY	1
#define FULL_KEY	2
#define FULL_KEY_DATA	3
#define	REAL_KEY	4

/* Short hands for accessing structure */
#define BSIZE		hdr.bsize
#define BSHIFT		hdr.bshift
#define DSIZE		hdr.dsize
#define SGSIZE		hdr.ssize
#define SSHIFT		hdr.sshift
#define LORDER		hdr.lorder
#define OVFL_POINT	hdr.ovfl_point
#define	LAST_FREED	hdr.last_freed
#define MAX_BUCKET	hdr.max_bucket
#define FFACTOR		hdr.ffactor
#define HIGH_MASK	hdr.high_mask
#define LOW_MASK	hdr.low_mask
#define NKEYS		hdr.nkeys
#define HDRPAGES	hdr.hdrpages
#define SPARES		hdr.spares
#define BITMAPS		hdr.bitmaps
#define VERSION		hdr.version
#define MAGIC		hdr.magic
#define NEXT_FREE	hdr.next_free
#define H_CHARKEY	hdr.h_charkey