newlib/libc/search/hash.h

   1 /*-
   2  * Copyright (c) 1990, 1993, 1994
   3  *      The Regents of the University of California.  All rights reserved.
   4  *
   5  * This code is derived from software contributed to Berkeley by
   6  * Margo Seltzer.
   7  *
   8  * Redistribution and use in source and binary forms, with or without
   9  * modification, are permitted provided that the following conditions
  10  * are met:
  11  * 1. Redistributions of source code must retain the above copyright
  12  *    notice, this list of conditions and the following disclaimer.
  13  * 2. Redistributions in binary form must reproduce the above copyright
  14  *    notice, this list of conditions and the following disclaimer in the
  15  *    documentation and/or other materials provided with the distribution.
  16  * 3. Neither the name of the University nor the names of its contributors
  17  *    may be used to endorse or promote products derived from this software
  18  *    without specific prior written permission.
  19  *
  20  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  21  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  22  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  23  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  24  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  25  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  26  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  27  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  28  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  29  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  30  * SUCH DAMAGE.
  31  *
  32  *      @(#)hash.h      8.3 (Berkeley) 5/31/94
  33  * $FreeBSD: src/lib/libc/db/hash/hash.h,v 1.6 2002/03/21 22:46:26 obrien Exp $
  34  */
  35
  36 #include <sys/param.h>
  37 #define __need_size_t
  38 #include <stddef.h>
  39 #include <stdint.h>
  40
  41 /* Check that newlib understands the byte order of its target system.  */
  42 #ifndef BYTE_ORDER
  43 #error BYTE_ORDER not defined by sys/param.h
  44 #endif
  45
  46 /* Define DB endianness constants based on target endianness.  */
  47 #define DB_LITTLE_ENDIAN 1234
  48 #define DB_BIG_ENDIAN 4321
  49 #if (BYTE_ORDER == LITTLE_ENDIAN)
  50 #define DB_BYTE_ORDER DB_LITTLE_ENDIAN
  51 #else
  52 #define DB_BYTE_ORDER DB_BIG_ENDIAN
  53 #endif
  54
  55 /* Operations */
  56 typedef enum {
  57         HASH_GET, HASH_PUT, HASH_PUTNEW, HASH_DELETE, HASH_FIRST, HASH_NEXT
  58 } ACTION;
  59
  60 /* Buffer Management structures */
  61 typedef struct _bufhead BUFHEAD;
  62
  63 struct _bufhead {
  64         BUFHEAD         *prev;          /* LRU links */
  65         BUFHEAD         *next;          /* LRU links */
  66         BUFHEAD         *ovfl;          /* Overflow page buffer header */
  67         __uint32_t       addr;          /* Address of this page */
  68         char            *page;          /* Actual page data */
  69         char            flags;
  70 #define BUF_MOD         0x0001
  71 #define BUF_DISK        0x0002
  72 #define BUF_BUCKET      0x0004
  73 #define BUF_PIN         0x0008
  74 };
  75
  76 #define IS_BUCKET(X)    ((X) & BUF_BUCKET)
  77
  78 typedef BUFHEAD **SEGMENT;
  79
  80 /* Hash Table Information */
  81 typedef struct hashhdr {                /* Disk resident portion */
  82         int32_t         magic;          /* Magic NO for hash tables */
  83         int32_t         version;        /* Version ID */
  84         __uint32_t      lorder;         /* Byte Order */
  85         int32_t         bsize;          /* Bucket/Page Size */
  86         int32_t         bshift;         /* Bucket shift */
  87         int32_t         dsize;          /* Directory Size */
  88         int32_t         ssize;          /* Segment Size */
  89         int32_t         sshift;         /* Segment shift */
  90         int32_t         ovfl_point;     /* Where overflow pages are being
  91                                          * allocated */
  92         int32_t         last_freed;     /* Last overflow page freed */
  93         int32_t         max_bucket;     /* ID of Maximum bucket in use */
  94         int32_t         high_mask;      /* Mask to modulo into entire table */
  95         int32_t         low_mask;       /* Mask to modulo into lower half of
  96                                          * table */
  97         int32_t         ffactor;        /* Fill factor */
  98         int32_t         nkeys;          /* Number of keys in hash table */
  99         int32_t         hdrpages;       /* Size of table header */
 100         int32_t         h_charkey;      /* value of hash(CHARKEY) */
 101 #define NCACHED 32                      /* number of bit maps and spare
 102                                          * points */
 103         int32_t         spares[NCACHED];/* spare pages for overflow */
 104         __uint16_t      bitmaps[NCACHED];       /* address of overflow page
 105                                                  * bitmaps */
 106 } HASHHDR;
 107
 108 typedef struct htab      {              /* Memory resident data structure */
 109         HASHHDR         hdr;            /* Header */
 110         int             nsegs;          /* Number of allocated segments */
 111         int             exsegs;         /* Number of extra allocated
 112                                          * segments */
 113         __uint32_t                      /* Hash function */
 114             (*hash)(const void *, size_t);
 115         int             flags;          /* Flag values */
 116         int             fp;             /* File pointer */
 117         char            *tmp_buf;       /* Temporary Buffer for BIG data */
 118         char            *tmp_key;       /* Temporary Buffer for BIG keys */
 119         BUFHEAD         *cpage;         /* Current page */
 120         int32_t         cbucket;        /* Current bucket */
 121         int             cndx;           /* Index of next item on cpage */
 122         int             error;          /* Error Number -- for DBM
 123                                          * compatibility */
 124         int             new_file;       /* Indicates if fd is backing store
 125                                          * or no */
 126         int             save_file;      /* Indicates whether we need to flush
 127                                          * file at
 128                                          * exit */
 129         __uint32_t      *mapp[NCACHED]; /* Pointers to page maps */
 130         int             nmaps;          /* Initial number of bitmaps */
 131         int             nbufs;          /* Number of buffers left to
 132                                          * allocate */
 133         BUFHEAD         bufhead;        /* Header of buffer lru list */
 134         SEGMENT         *dir;           /* Hash Bucket directory */
 135 } HTAB;
 136
 137 /*
 138  * Constants
 139  */
 140 #if INT_MAX == 32767
 141 #define MAX_BSIZE               4096
 142 #else
 143 #define MAX_BSIZE               65536           /* 2^16 */
 144 #endif
 145 #define MIN_BUFFERS             6
 146 #define MINHDRSIZE              512
 147 #if INT_MAX == 32767
 148 #define DEF_BUFSIZE             4096
 149 #else
 150 #define DEF_BUFSIZE             65536           /* 64 K */
 151 #endif
 152 #define DEF_BUCKET_SIZE         4096
 153 #define DEF_BUCKET_SHIFT        12              /* log2(BUCKET) */
 154 #define DEF_SEGSIZE             256
 155 #define DEF_SEGSIZE_SHIFT       8               /* log2(SEGSIZE)         */
 156 #define DEF_DIRSIZE             256
 157 #define DEF_FFACTOR             65536
 158 #define MIN_FFACTOR             4
 159 #define SPLTMAX                 8
 160 #define CHARKEY                 "%$sniglet^&"
 161 #define NUMKEY                  1038583
 162 #define BYTE_SHIFT              3
 163 #define INT_TO_BYTE             2
 164 #define INT_BYTE_SHIFT          5
 165 #define ALL_SET                 ((__uint32_t)0xFFFFFFFF)
 166 #define ALL_CLEAR               0
 167
 168 #define PTROF(X)        ((BUFHEAD *)((ptrdiff_t)(X)&~0x3))
 169 #define ISMOD(X)        ((__uint32_t)(ptrdiff_t)(X)&0x1)
 170 #define DOMOD(X)        ((X) = (char *)((ptrdiff_t)(X)|0x1))
 171 #define ISDISK(X)       ((__uint32_t)(ptrdiff_t)(X)&0x2)
 172 #define DODISK(X)       ((X) = (char *)((ptrdiff_t)(X)|0x2))
 173
 174 #define BITS_PER_MAP    32
 175
 176 /* Given the address of the beginning of a big map, clear/set the nth bit */
 177 #define CLRBIT(A, N)    ((A)[(N)/BITS_PER_MAP] &= ~(1<<((N)%BITS_PER_MAP)))
 178 #define SETBIT(A, N)    ((A)[(N)/BITS_PER_MAP] |= (1<<((N)%BITS_PER_MAP)))
 179 #define ISSET(A, N)     ((A)[(N)/BITS_PER_MAP] & (1<<((N)%BITS_PER_MAP)))
 180
 181 /* Overflow management */
 182 /*
 183  * Overflow page numbers are allocated per split point.  At each doubling of
 184  * the table, we can allocate extra pages.  So, an overflow page number has
 185  * the top 5 bits indicate which split point and the lower 11 bits indicate
 186  * which page at that split point is indicated (pages within split points are
 187  * numberered starting with 1).
 188  */
 189
 190 #define SPLITSHIFT      11
 191 #define SPLITMASK       0x7FF
 192 #define SPLITNUM(N)     (((__uint32_t)(N)) >> SPLITSHIFT)
 193 #define OPAGENUM(N)     ((N) & SPLITMASK)
 194 #define OADDR_OF(S,O)   ((__uint32_t)((__uint32_t)(S) << SPLITSHIFT) + (O))
 195
 196 #define BUCKET_TO_PAGE(B) \
 197         (B) + hashp->HDRPAGES + ((B) ? hashp->SPARES[__log2((B)+1)-1] : 0)
 198 #define OADDR_TO_PAGE(B)        \
 199         BUCKET_TO_PAGE ( (1 << SPLITNUM((B))) -1 ) + OPAGENUM((B));
 200
 201 /*
 202  * page.h contains a detailed description of the page format.
 203  *
 204  * Normally, keys and data are accessed from offset tables in the top of
 205  * each page which point to the beginning of the key and data.  There are
 206  * four flag values which may be stored in these offset tables which indicate
 207  * the following:
 208  *
 209  *
 210  * OVFLPAGE     Rather than a key data pair, this pair contains
 211  *              the address of an overflow page.  The format of
 212  *              the pair is:
 213  *                  OVERFLOW_PAGE_NUMBER OVFLPAGE
 214  *
 215  * PARTIAL_KEY  This must be the first key/data pair on a page
 216  *              and implies that page contains only a partial key.
 217  *              That is, the key is too big to fit on a single page
 218  *              so it starts on this page and continues on the next.
 219  *              The format of the page is:
 220  *                  KEY_OFF PARTIAL_KEY OVFL_PAGENO OVFLPAGE
 221  *
 222  *                  KEY_OFF -- offset of the beginning of the key
 223  *                  PARTIAL_KEY -- 1
 224  *                  OVFL_PAGENO - page number of the next overflow page
 225  *                  OVFLPAGE -- 0
 226  *
 227  * FULL_KEY     This must be the first key/data pair on the page.  It
 228  *              is used in two cases.
 229  *
 230  *              Case 1:
 231  *                  There is a complete key on the page but no data
 232  *                  (because it wouldn't fit).  The next page contains
 233  *                  the data.
 234  *
 235  *                  Page format it:
 236  *                  KEY_OFF FULL_KEY OVFL_PAGENO OVFL_PAGE
 237  *
 238  *                  KEY_OFF -- offset of the beginning of the key
 239  *                  FULL_KEY -- 2
 240  *                  OVFL_PAGENO - page number of the next overflow page
 241  *                  OVFLPAGE -- 0
 242  *
 243  *              Case 2:
 244  *                  This page contains no key, but part of a large
 245  *                  data field, which is continued on the next page.
 246  *
 247  *                  Page format it:
 248  *                  DATA_OFF FULL_KEY OVFL_PAGENO OVFL_PAGE
 249  *
 250  *                  KEY_OFF -- offset of the beginning of the data on
 251  *                              this page
 252  *                  FULL_KEY -- 2
 253  *                  OVFL_PAGENO - page number of the next overflow page
 254  *                  OVFLPAGE -- 0
 255  *
 256  * FULL_KEY_DATA
 257  *              This must be the first key/data pair on the page.
 258  *              There are two cases:
 259  *
 260  *              Case 1:
 261  *                  This page contains a key and the beginning of the
 262  *                  data field, but the data field is continued on the
 263  *                  next page.
 264  *
 265  *                  Page format is:
 266  *                  KEY_OFF FULL_KEY_DATA OVFL_PAGENO DATA_OFF
 267  *
 268  *                  KEY_OFF -- offset of the beginning of the key
 269  *                  FULL_KEY_DATA -- 3
 270  *                  OVFL_PAGENO - page number of the next overflow page
 271  *                  DATA_OFF -- offset of the beginning of the data
 272  *
 273  *              Case 2:
 274  *                  This page contains the last page of a big data pair.
 275  *                  There is no key, only the  tail end of the data
 276  *                  on this page.
 277  *
 278  *                  Page format is:
 279  *                  DATA_OFF FULL_KEY_DATA <OVFL_PAGENO> <OVFLPAGE>
 280  *
 281  *                  DATA_OFF -- offset of the beginning of the data on
 282  *                              this page
 283  *                  FULL_KEY_DATA -- 3
 284  *                  OVFL_PAGENO - page number of the next overflow page
 285  *                  OVFLPAGE -- 0
 286  *
 287  *                  OVFL_PAGENO and OVFLPAGE are optional (they are
 288  *                  not present if there is no next page).
 289  */
 290
 291 #define OVFLPAGE        0
 292 #define PARTIAL_KEY     1
 293 #define FULL_KEY        2
 294 #define FULL_KEY_DATA   3
 295 #define REAL_KEY        4
 296
 297 /* Short hands for accessing structure */
 298 #define BSIZE           hdr.bsize
 299 #define BSHIFT          hdr.bshift
 300 #define DSIZE           hdr.dsize
 301 #define SGSIZE          hdr.ssize
 302 #define SSHIFT          hdr.sshift
 303 #define LORDER          hdr.lorder
 304 #define OVFL_POINT      hdr.ovfl_point
 305 #define LAST_FREED      hdr.last_freed
 306 #define MAX_BUCKET      hdr.max_bucket
 307 #define FFACTOR         hdr.ffactor
 308 #define HIGH_MASK       hdr.high_mask
 309 #define LOW_MASK        hdr.low_mask
 310 #define NKEYS           hdr.nkeys
 311 #define HDRPAGES        hdr.hdrpages
 312 #define SPARES          hdr.spares
 313 #define BITMAPS         hdr.bitmaps
 314 #define HASH_VERSION    hdr.version
 315 #define MAGIC           hdr.magic
 316 #define NEXT_FREE       hdr.next_free
 317 #define H_CHARKEY       hdr.h_charkey