1/*-
2 * See the file LICENSE for redistribution information.
3 *
4 * Copyright (c) 1996, 1997, 1998
5 *	Sleepycat Software.  All rights reserved.
6 */
7/*
8 * Copyright (c) 1990, 1993, 1994, 1995, 1996
9 *	Keith Bostic.  All rights reserved.
10 */
11/*
12 * Copyright (c) 1990, 1993, 1994, 1995
13 *	The Regents of the University of California.  All rights reserved.
14 *
15 * This code is derived from software contributed to Berkeley by
16 * Mike Olson.
17 *
18 * Redistribution and use in source and binary forms, with or without
19 * modification, are permitted provided that the following conditions
20 * are met:
21 * 1. Redistributions of source code must retain the above copyright
22 *    notice, this list of conditions and the following disclaimer.
23 * 2. Redistributions in binary form must reproduce the above copyright
24 *    notice, this list of conditions and the following disclaimer in the
25 *    documentation and/or other materials provided with the distribution.
26 * 3. All advertising materials mentioning features or use of this software
27 *    must display the following acknowledgement:
28 *	This product includes software developed by the University of
29 *	California, Berkeley and its contributors.
30 * 4. Neither the name of the University nor the names of its contributors
31 *    may be used to endorse or promote products derived from this software
32 *    without specific prior written permission.
33 *
34 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
35 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
36 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
37 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
38 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
39 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
40 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
41 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
42 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
43 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
44 * SUCH DAMAGE.
45 */
46
47#include "config.h"
48
49#ifndef lint
50static const char sccsid[] = "@(#)bt_open.c	10.39 (Sleepycat) 11/21/98";
51#endif /* not lint */
52
53#ifndef NO_SYSTEM_INCLUDES
54#include <sys/types.h>
55
56#include <errno.h>
57#include <limits.h>
58#include <string.h>
59#endif
60
61#include "db_int.h"
62#include "db_page.h"
63#include "btree.h"
64
65/*
66 * __bam_open --
67 *	Open a btree.
68 *
69 * PUBLIC: int __bam_open __P((DB *, DB_INFO *));
70 */
71int
72__bam_open(dbp, dbinfo)
73	DB *dbp;
74	DB_INFO *dbinfo;
75{
76	BTREE *t;
77	int ret;
78
79	/* Allocate and initialize the private btree structure. */
80	if ((ret = __os_calloc(1, sizeof(BTREE), &t)) != 0)
81		return (ret);
82	dbp->internal = t;
83
84	/*
85	 * Intention is to make sure all of the user's selections are okay
86	 * here and then use them without checking.
87	 */
88	if (dbinfo == NULL) {
89		t->bt_minkey = DEFMINKEYPAGE;
90		t->bt_compare = __bam_defcmp;
91		t->bt_prefix = __bam_defpfx;
92	} else {
93		/* Minimum number of keys per page. */
94		if (dbinfo->bt_minkey == 0)
95			t->bt_minkey = DEFMINKEYPAGE;
96		else {
97			if (dbinfo->bt_minkey < 2)
98				goto einval;
99			t->bt_minkey = dbinfo->bt_minkey;
100		}
101
102		/* Maximum number of keys per page. */
103		if (dbinfo->bt_maxkey == 0)
104			t->bt_maxkey = 0;
105		else {
106			if (dbinfo->bt_maxkey < 1)
107				goto einval;
108			t->bt_maxkey = dbinfo->bt_maxkey;
109		}
110
111		/*
112		 * If no comparison, use default comparison.  If no comparison
113		 * and no prefix, use default prefix.  (We can't default the
114		 * prefix if the user supplies a comparison routine; shortening
115		 * the keys may break their comparison algorithm.  We don't
116		 * permit the user to specify a prefix routine if they didn't
117		 * also specify a comparison routine, they can't know enough
118		 * about our comparison routine to get it right.)
119		 */
120		if ((t->bt_compare = dbinfo->bt_compare) == NULL) {
121			if (dbinfo->bt_prefix != NULL)
122				goto einval;
123			t->bt_compare = __bam_defcmp;
124			t->bt_prefix = __bam_defpfx;
125		} else
126			t->bt_prefix = dbinfo->bt_prefix;
127	}
128
129	/* Initialize the remaining fields/methods of the DB. */
130	dbp->am_close = __bam_close;
131	dbp->del = __bam_delete;
132	dbp->stat = __bam_stat;
133
134	/* Start up the tree. */
135	if ((ret = __bam_read_root(dbp)) != 0)
136		goto err;
137
138	/* Set the overflow page size. */
139	__bam_setovflsize(dbp);
140
141	return (0);
142
143einval:	ret = EINVAL;
144
145err:	__os_free(t, sizeof(BTREE));
146	return (ret);
147}
148
149/*
150 * __bam_close --
151 *	Close a btree.
152 *
153 * PUBLIC: int __bam_close __P((DB *));
154 */
155int
156__bam_close(dbp)
157	DB *dbp;
158{
159	__os_free(dbp->internal, sizeof(BTREE));
160	dbp->internal = NULL;
161
162	return (0);
163}
164
165/*
166 * __bam_setovflsize --
167 *
168 * PUBLIC: void __bam_setovflsize __P((DB *));
169 */
170void
171__bam_setovflsize(dbp)
172	DB *dbp;
173{
174	BTREE *t;
175
176	t = dbp->internal;
177
178	/*
179	 * !!!
180	 * Correction for recno, which doesn't know anything about minimum
181	 * keys per page.
182	 */
183	if (t->bt_minkey == 0)
184		t->bt_minkey = DEFMINKEYPAGE;
185
186	/*
187	 * The btree data structure requires that at least two key/data pairs
188	 * can fit on a page, but other than that there's no fixed requirement.
189	 * Translate the minimum number of items into the bytes a key/data pair
190	 * can use before being placed on an overflow page.  We calculate for
191	 * the worst possible alignment by assuming every item requires the
192	 * maximum alignment for padding.
193	 *
194	 * Recno uses the btree bt_ovflsize value -- it's close enough.
195	 */
196	t->bt_ovflsize = (dbp->pgsize - P_OVERHEAD) / (t->bt_minkey * P_INDX)
197	    - (BKEYDATA_PSIZE(0) + ALIGN(1, 4));
198}
199
200/*
201 * __bam_read_root --
202 *	Check (and optionally create) a tree.
203 *
204 * PUBLIC: int __bam_read_root __P((DB *));
205 */
206int
207__bam_read_root(dbp)
208	DB *dbp;
209{
210	BTMETA *meta;
211	BTREE *t;
212	DBC *dbc;
213	DB_LOCK metalock, rootlock;
214	PAGE *root;
215	db_pgno_t pgno;
216	int ret, t_ret;
217
218	ret = 0;
219	t = dbp->internal;
220
221	/* Get a cursor. */
222	if ((ret = dbp->cursor(dbp, NULL, &dbc, 0)) != 0)
223		return (ret);
224
225	/* Get, and optionally create the metadata page. */
226	pgno = PGNO_METADATA;
227	if ((ret =
228	    __bam_lget(dbc, 0, PGNO_METADATA, DB_LOCK_WRITE, &metalock)) != 0)
229		goto err;
230	if ((ret =
231	    memp_fget(dbp->mpf, &pgno, DB_MPOOL_CREATE, (PAGE **)&meta)) != 0) {
232		(void)__BT_LPUT(dbc, metalock);
233		goto err;
234	}
235
236	/*
237	 * If the magic number is correct, we're not creating the tree.
238	 * Correct any fields that may not be right.  Note, all of the
239	 * local flags were set by db_open(3).
240	 */
241	if (meta->magic != 0) {
242		t->bt_maxkey = meta->maxkey;
243		t->bt_minkey = meta->minkey;
244
245		(void)memp_fput(dbp->mpf, (PAGE *)meta, 0);
246		(void)__BT_LPUT(dbc, metalock);
247		goto done;
248	}
249
250	/* Initialize the tree structure metadata information. */
251	memset(meta, 0, sizeof(BTMETA));
252	ZERO_LSN(meta->lsn);
253	meta->pgno = PGNO_METADATA;
254	meta->magic = DB_BTREEMAGIC;
255	meta->version = DB_BTREEVERSION;
256	meta->pagesize = dbp->pgsize;
257	meta->maxkey = t->bt_maxkey;
258	meta->minkey = t->bt_minkey;
259	meta->free = PGNO_INVALID;
260	if (dbp->type == DB_RECNO)
261		F_SET(meta, BTM_RECNO);
262	if (F_ISSET(dbp, DB_AM_DUP))
263		F_SET(meta, BTM_DUP);
264	if (F_ISSET(dbp, DB_RE_FIXEDLEN))
265		F_SET(meta, BTM_FIXEDLEN);
266	if (F_ISSET(dbp, DB_BT_RECNUM))
267		F_SET(meta, BTM_RECNUM);
268	if (F_ISSET(dbp, DB_RE_RENUMBER))
269		F_SET(meta, BTM_RENUMBER);
270	memcpy(meta->uid, dbp->fileid, DB_FILE_ID_LEN);
271
272	/* Create and initialize a root page. */
273	pgno = PGNO_ROOT;
274	if ((ret =
275	    __bam_lget(dbc, 0, PGNO_ROOT, DB_LOCK_WRITE, &rootlock)) != 0)
276		goto err;
277	if ((ret = memp_fget(dbp->mpf, &pgno, DB_MPOOL_CREATE, &root)) != 0) {
278		(void)__BT_LPUT(dbc, rootlock);
279		goto err;
280	}
281	P_INIT(root, dbp->pgsize, PGNO_ROOT, PGNO_INVALID,
282	    PGNO_INVALID, 1, dbp->type == DB_RECNO ? P_LRECNO : P_LBTREE);
283	ZERO_LSN(root->lsn);
284
285	/* Release the metadata and root pages. */
286	if ((ret = memp_fput(dbp->mpf, (PAGE *)meta, DB_MPOOL_DIRTY)) != 0)
287		goto err;
288	if ((ret = memp_fput(dbp->mpf, root, DB_MPOOL_DIRTY)) != 0)
289		goto err;
290
291	/*
292	 * Flush the metadata and root pages to disk -- since the user can't
293	 * transaction protect open, the pages have to exist during recovery.
294	 *
295	 * XXX
296	 * It's not useful to return not-yet-flushed here -- convert it to
297	 * an error.
298	 */
299	if ((ret = memp_fsync(dbp->mpf)) == DB_INCOMPLETE)
300		ret = EINVAL;
301
302	/* Release the locks. */
303	(void)__BT_LPUT(dbc, metalock);
304	(void)__BT_LPUT(dbc, rootlock);
305
306err:
307done:	if ((t_ret = dbc->c_close(dbc)) != 0 && ret == 0)
308		ret = t_ret;
309	return (ret);
310}
311