source: rtems-libbsd/freebsd/lib/libc/db/btree/bt_open.c @ 2404264

4.1155-freebsd-126-freebsd-12freebsd-9.3
Last change on this file since 2404264 was 2404264, checked in by Sebastian Huber <sebastian.huber@…>, on 10/30/13 at 12:23:59

Revert superfluous changes

  • Property mode set to 100644
File size: 11.1 KB
Line 
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 * Mike Olson.
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 * 4. 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
33#if defined(LIBC_SCCS) && !defined(lint)
34static char sccsid[] = "@(#)bt_open.c   8.10 (Berkeley) 8/17/94";
35#endif /* LIBC_SCCS and not lint */
36#include <sys/cdefs.h>
37__FBSDID("$FreeBSD$");
38
39/*
40 * Implementation of btree access method for 4.4BSD.
41 *
42 * The design here was originally based on that of the btree access method
43 * used in the Postgres database system at UC Berkeley.  This implementation
44 * is wholly independent of the Postgres code.
45 */
46
47#include "namespace.h"
48#include <rtems/bsd/sys/param.h>
49#include <sys/stat.h>
50
51#include <errno.h>
52#include <fcntl.h>
53#include <limits.h>
54#include <signal.h>
55#include <stdio.h>
56#include <stdlib.h>
57#include <string.h>
58#include <unistd.h>
59#include "un-namespace.h"
60
61#include <db.h>
62#include "btree.h"
63
64#ifdef DEBUG
65#undef  MINPSIZE
66#define MINPSIZE        128
67#endif
68
69static int byteorder(void);
70static int nroot(BTREE *);
71static int tmp(void);
72
73/*
74 * __BT_OPEN -- Open a btree.
75 *
76 * Creates and fills a DB struct, and calls the routine that actually
77 * opens the btree.
78 *
79 * Parameters:
80 *      fname:  filename (NULL for in-memory trees)
81 *      flags:  open flag bits
82 *      mode:   open permission bits
83 *      b:      BTREEINFO pointer
84 *
85 * Returns:
86 *      NULL on failure, pointer to DB on success.
87 *
88 */
89DB *
90__bt_open(const char *fname, int flags, int mode, const BTREEINFO *openinfo, int dflags)
91{
92        struct stat sb;
93        BTMETA m;
94        BTREE *t;
95        BTREEINFO b;
96        DB *dbp;
97        pgno_t ncache;
98        ssize_t nr;
99        int machine_lorder, saved_errno;
100
101        t = NULL;
102
103        /*
104         * Intention is to make sure all of the user's selections are okay
105         * here and then use them without checking.  Can't be complete, since
106         * we don't know the right page size, lorder or flags until the backing
107         * file is opened.  Also, the file's page size can cause the cachesize
108         * to change.
109         */
110        machine_lorder = byteorder();
111        if (openinfo) {
112                b = *openinfo;
113
114                /* Flags: R_DUP. */
115                if (b.flags & ~(R_DUP))
116                        goto einval;
117
118                /*
119                 * Page size must be indx_t aligned and >= MINPSIZE.  Default
120                 * page size is set farther on, based on the underlying file
121                 * transfer size.
122                 */
123                if (b.psize &&
124                    (b.psize < MINPSIZE || b.psize > MAX_PAGE_OFFSET + 1 ||
125                    b.psize & (sizeof(indx_t) - 1) ))
126                        goto einval;
127
128                /* Minimum number of keys per page; absolute minimum is 2. */
129                if (b.minkeypage) {
130                        if (b.minkeypage < 2)
131                                goto einval;
132                } else
133                        b.minkeypage = DEFMINKEYPAGE;
134
135                /* If no comparison, use default comparison and prefix. */
136                if (b.compare == NULL) {
137                        b.compare = __bt_defcmp;
138                        if (b.prefix == NULL)
139                                b.prefix = __bt_defpfx;
140                }
141
142                if (b.lorder == 0)
143                        b.lorder = machine_lorder;
144        } else {
145                b.compare = __bt_defcmp;
146                b.cachesize = 0;
147                b.flags = 0;
148                b.lorder = machine_lorder;
149                b.minkeypage = DEFMINKEYPAGE;
150                b.prefix = __bt_defpfx;
151                b.psize = 0;
152        }
153
154        /* Check for the ubiquitous PDP-11. */
155        if (b.lorder != BIG_ENDIAN && b.lorder != LITTLE_ENDIAN)
156                goto einval;
157
158        /* Allocate and initialize DB and BTREE structures. */
159        if ((t = (BTREE *)calloc(1, sizeof(BTREE))) == NULL)
160                goto err;
161        t->bt_fd = -1;                  /* Don't close unopened fd on error. */
162        t->bt_lorder = b.lorder;
163        t->bt_order = NOT;
164        t->bt_cmp = b.compare;
165        t->bt_pfx = b.prefix;
166        t->bt_rfd = -1;
167
168        if ((t->bt_dbp = dbp = (DB *)calloc(1, sizeof(DB))) == NULL)
169                goto err;
170        if (t->bt_lorder != machine_lorder)
171                F_SET(t, B_NEEDSWAP);
172
173        dbp->type = DB_BTREE;
174        dbp->internal = t;
175        dbp->close = __bt_close;
176        dbp->del = __bt_delete;
177        dbp->fd = __bt_fd;
178        dbp->get = __bt_get;
179        dbp->put = __bt_put;
180        dbp->seq = __bt_seq;
181        dbp->sync = __bt_sync;
182
183        /*
184         * If no file name was supplied, this is an in-memory btree and we
185         * open a backing temporary file.  Otherwise, it's a disk-based tree.
186         */
187        if (fname) {
188                switch (flags & O_ACCMODE) {
189                case O_RDONLY:
190                        F_SET(t, B_RDONLY);
191                        break;
192                case O_RDWR:
193                        break;
194                case O_WRONLY:
195                default:
196                        goto einval;
197                }
198
199                if ((t->bt_fd = _open(fname, flags, mode)) < 0)
200                        goto err;
201
202        } else {
203                if ((flags & O_ACCMODE) != O_RDWR)
204                        goto einval;
205                if ((t->bt_fd = tmp()) == -1)
206                        goto err;
207                F_SET(t, B_INMEM);
208        }
209
210        if (_fcntl(t->bt_fd, F_SETFD, 1) == -1)
211                goto err;
212
213        if (_fstat(t->bt_fd, &sb))
214                goto err;
215        if (sb.st_size) {
216                if ((nr = _read(t->bt_fd, &m, sizeof(BTMETA))) < 0)
217                        goto err;
218                if (nr != sizeof(BTMETA))
219                        goto eftype;
220
221                /*
222                 * Read in the meta-data.  This can change the notion of what
223                 * the lorder, page size and flags are, and, when the page size
224                 * changes, the cachesize value can change too.  If the user
225                 * specified the wrong byte order for an existing database, we
226                 * don't bother to return an error, we just clear the NEEDSWAP
227                 * bit.
228                 */
229                if (m.magic == BTREEMAGIC)
230                        F_CLR(t, B_NEEDSWAP);
231                else {
232                        F_SET(t, B_NEEDSWAP);
233                        M_32_SWAP(m.magic);
234                        M_32_SWAP(m.version);
235                        M_32_SWAP(m.psize);
236                        M_32_SWAP(m.free);
237                        M_32_SWAP(m.nrecs);
238                        M_32_SWAP(m.flags);
239                }
240                if (m.magic != BTREEMAGIC || m.version != BTREEVERSION)
241                        goto eftype;
242                if (m.psize < MINPSIZE || m.psize > MAX_PAGE_OFFSET + 1 ||
243                    m.psize & (sizeof(indx_t) - 1) )
244                        goto eftype;
245                if (m.flags & ~SAVEMETA)
246                        goto eftype;
247                b.psize = m.psize;
248                F_SET(t, m.flags);
249                t->bt_free = m.free;
250                t->bt_nrecs = m.nrecs;
251        } else {
252                /*
253                 * Set the page size to the best value for I/O to this file.
254                 * Don't overflow the page offset type.
255                 */
256                if (b.psize == 0) {
257                        b.psize = sb.st_blksize;
258                        if (b.psize < MINPSIZE)
259                                b.psize = MINPSIZE;
260                        if (b.psize > MAX_PAGE_OFFSET + 1)
261                                b.psize = MAX_PAGE_OFFSET + 1;
262                }
263
264                /* Set flag if duplicates permitted. */
265                if (!(b.flags & R_DUP))
266                        F_SET(t, B_NODUPS);
267
268                t->bt_free = P_INVALID;
269                t->bt_nrecs = 0;
270                F_SET(t, B_METADIRTY);
271        }
272
273        t->bt_psize = b.psize;
274
275        /* Set the cache size; must be a multiple of the page size. */
276        if (b.cachesize && b.cachesize & (b.psize - 1) )
277                b.cachesize += (~b.cachesize & (b.psize - 1) ) + 1;
278        if (b.cachesize < b.psize * MINCACHE)
279                b.cachesize = b.psize * MINCACHE;
280
281        /* Calculate number of pages to cache. */
282        ncache = (b.cachesize + t->bt_psize - 1) / t->bt_psize;
283
284        /*
285         * The btree data structure requires that at least two keys can fit on
286         * a page, but other than that there's no fixed requirement.  The user
287         * specified a minimum number per page, and we translated that into the
288         * number of bytes a key/data pair can use before being placed on an
289         * overflow page.  This calculation includes the page header, the size
290         * of the index referencing the leaf item and the size of the leaf item
291         * structure.  Also, don't let the user specify a minkeypage such that
292         * a key/data pair won't fit even if both key and data are on overflow
293         * pages.
294         */
295        t->bt_ovflsize = (t->bt_psize - BTDATAOFF) / b.minkeypage -
296            (sizeof(indx_t) + NBLEAFDBT(0, 0));
297        if (t->bt_ovflsize < NBLEAFDBT(NOVFLSIZE, NOVFLSIZE) + sizeof(indx_t))
298                t->bt_ovflsize =
299                    NBLEAFDBT(NOVFLSIZE, NOVFLSIZE) + sizeof(indx_t);
300
301        /* Initialize the buffer pool. */
302        if ((t->bt_mp =
303            mpool_open(NULL, t->bt_fd, t->bt_psize, ncache)) == NULL)
304                goto err;
305        if (!F_ISSET(t, B_INMEM))
306                mpool_filter(t->bt_mp, __bt_pgin, __bt_pgout, t);
307
308        /* Create a root page if new tree. */
309        if (nroot(t) == RET_ERROR)
310                goto err;
311
312        /* Global flags. */
313        if (dflags & DB_LOCK)
314                F_SET(t, B_DB_LOCK);
315        if (dflags & DB_SHMEM)
316                F_SET(t, B_DB_SHMEM);
317        if (dflags & DB_TXN)
318                F_SET(t, B_DB_TXN);
319
320        return (dbp);
321
322einval: errno = EINVAL;
323        goto err;
324
325eftype: errno = EFTYPE;
326        goto err;
327
328err:    saved_errno = errno;
329        if (t) {
330                if (t->bt_dbp)
331                        free(t->bt_dbp);
332                if (t->bt_fd != -1)
333                        (void)_close(t->bt_fd);
334                free(t);
335        }
336        errno = saved_errno;
337        return (NULL);
338}
339
340/*
341 * NROOT -- Create the root of a new tree.
342 *
343 * Parameters:
344 *      t:      tree
345 *
346 * Returns:
347 *      RET_ERROR, RET_SUCCESS
348 */
349static int
350nroot(BTREE *t)
351{
352        PAGE *meta, *root;
353        pgno_t npg;
354
355        if ((root = mpool_get(t->bt_mp, 1, 0)) != NULL) {
356                if (root->lower == 0 &&
357                    root->pgno == 0 &&
358                    root->linp[0] == 0) {
359                        mpool_delete(t->bt_mp, root);
360                        errno = EINVAL;
361                } else {
362                        mpool_put(t->bt_mp, root, 0);
363                        return (RET_SUCCESS);
364                }
365        }
366        if (errno != EINVAL)            /* It's OK to not exist. */
367                return (RET_ERROR);
368        errno = 0;
369
370        if ((meta = mpool_new(t->bt_mp, &npg, MPOOL_PAGE_NEXT)) == NULL)
371                return (RET_ERROR);
372
373        if ((root = mpool_new(t->bt_mp, &npg, MPOOL_PAGE_NEXT)) == NULL)
374                return (RET_ERROR);
375
376        if (npg != P_ROOT)
377                return (RET_ERROR);
378        root->pgno = npg;
379        root->prevpg = root->nextpg = P_INVALID;
380        root->lower = BTDATAOFF;
381        root->upper = t->bt_psize;
382        root->flags = P_BLEAF;
383        memset(meta, 0, t->bt_psize);
384        mpool_put(t->bt_mp, meta, MPOOL_DIRTY);
385        mpool_put(t->bt_mp, root, MPOOL_DIRTY);
386        return (RET_SUCCESS);
387}
388
389static int
390tmp(void)
391{
392        sigset_t set, oset;
393        int fd, len;
394        char *envtmp = NULL;
395        char path[MAXPATHLEN];
396
397        if (issetugid() == 0)
398                envtmp = getenv("TMPDIR");
399        len = snprintf(path,
400            sizeof(path), "%s/bt.XXXXXXXXXX", envtmp ? envtmp : "/tmp");
401        if (len < 0 || len >= (int)sizeof(path)) {
402                errno = ENAMETOOLONG;
403                return(-1);
404        }
405
406        (void)sigfillset(&set);
407        (void)_sigprocmask(SIG_BLOCK, &set, &oset);
408        if ((fd = mkstemp(path)) != -1)
409                (void)unlink(path);
410        (void)_sigprocmask(SIG_SETMASK, &oset, NULL);
411        return(fd);
412}
413
414static int
415byteorder(void)
416{
417        u_int32_t x;
418        u_char *p;
419
420        x = 0x01020304;
421        p = (u_char *)&x;
422        switch (*p) {
423        case 1:
424                return (BIG_ENDIAN);
425        case 4:
426                return (LITTLE_ENDIAN);
427        default:
428                return (0);
429        }
430}
431
432int
433__bt_fd(const DB *dbp)
434{
435        BTREE *t;
436
437        t = dbp->internal;
438
439        /* Toss any page pinned across calls. */
440        if (t->bt_pinned != NULL) {
441                mpool_put(t->bt_mp, t->bt_pinned, 0);
442                t->bt_pinned = NULL;
443        }
444
445        /* In-memory database can't have a file descriptor. */
446        if (F_ISSET(t, B_INMEM)) {
447                errno = ENOENT;
448                return (-1);
449        }
450        return (t->bt_fd);
451}
Note: See TracBrowser for help on using the repository browser.