bt_delete.c revision 165903
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_delete.c	8.13 (Berkeley) 7/28/94";
35#endif /* LIBC_SCCS and not lint */
36#include <sys/cdefs.h>
37__FBSDID("$FreeBSD: head/lib/libc/db/btree/bt_delete.c 165903 2007-01-09 00:28:16Z imp $");
38
39#include <sys/types.h>
40
41#include <errno.h>
42#include <stdio.h>
43#include <string.h>
44
45#include <db.h>
46#include "btree.h"
47
48static int __bt_bdelete(BTREE *, const DBT *);
49static int __bt_curdel(BTREE *, const DBT *, PAGE *, u_int);
50static int __bt_pdelete(BTREE *, PAGE *);
51static int __bt_relink(BTREE *, PAGE *);
52static int __bt_stkacq(BTREE *, PAGE **, CURSOR *);
53
54/*
55 * __bt_delete
56 *	Delete the item(s) referenced by a key.
57 *
58 * Return RET_SPECIAL if the key is not found.
59 */
60int
61__bt_delete(dbp, key, flags)
62	const DB *dbp;
63	const DBT *key;
64	u_int flags;
65{
66	BTREE *t;
67	CURSOR *c;
68	PAGE *h;
69	int status;
70
71	t = dbp->internal;
72
73	/* Toss any page pinned across calls. */
74	if (t->bt_pinned != NULL) {
75		mpool_put(t->bt_mp, t->bt_pinned, 0);
76		t->bt_pinned = NULL;
77	}
78
79	/* Check for change to a read-only tree. */
80	if (F_ISSET(t, B_RDONLY)) {
81		errno = EPERM;
82		return (RET_ERROR);
83	}
84
85	switch (flags) {
86	case 0:
87		status = __bt_bdelete(t, key);
88		break;
89	case R_CURSOR:
90		/*
91		 * If flags is R_CURSOR, delete the cursor.  Must already
92		 * have started a scan and not have already deleted it.
93		 */
94		c = &t->bt_cursor;
95		if (F_ISSET(c, CURS_INIT)) {
96			if (F_ISSET(c, CURS_ACQUIRE | CURS_AFTER | CURS_BEFORE))
97				return (RET_SPECIAL);
98			if ((h = mpool_get(t->bt_mp, c->pg.pgno, 0)) == NULL)
99				return (RET_ERROR);
100
101			/*
102			 * If the page is about to be emptied, we'll need to
103			 * delete it, which means we have to acquire a stack.
104			 */
105			if (NEXTINDEX(h) == 1)
106				if (__bt_stkacq(t, &h, &t->bt_cursor))
107					return (RET_ERROR);
108
109			status = __bt_dleaf(t, NULL, h, c->pg.index);
110
111			if (NEXTINDEX(h) == 0 && status == RET_SUCCESS) {
112				if (__bt_pdelete(t, h))
113					return (RET_ERROR);
114			} else
115				mpool_put(t->bt_mp,
116				    h, status == RET_SUCCESS ? MPOOL_DIRTY : 0);
117			break;
118		}
119		/* FALLTHROUGH */
120	default:
121		errno = EINVAL;
122		return (RET_ERROR);
123	}
124	if (status == RET_SUCCESS)
125		F_SET(t, B_MODIFIED);
126	return (status);
127}
128
129/*
130 * __bt_stkacq --
131 *	Acquire a stack so we can delete a cursor entry.
132 *
133 * Parameters:
134 *	  t:	tree
135 *	 hp:	pointer to current, pinned PAGE pointer
136 *	  c:	pointer to the cursor
137 *
138 * Returns:
139 *	0 on success, 1 on failure
140 */
141static int
142__bt_stkacq(t, hp, c)
143	BTREE *t;
144	PAGE **hp;
145	CURSOR *c;
146{
147	BINTERNAL *bi;
148	EPG *e;
149	EPGNO *parent;
150	PAGE *h;
151	indx_t index;
152	pgno_t pgno;
153	recno_t nextpg, prevpg;
154	int exact, level;
155
156	/*
157	 * Find the first occurrence of the key in the tree.  Toss the
158	 * currently locked page so we don't hit an already-locked page.
159	 */
160	h = *hp;
161	mpool_put(t->bt_mp, h, 0);
162	if ((e = __bt_search(t, &c->key, &exact)) == NULL)
163		return (1);
164	h = e->page;
165
166	/* See if we got it in one shot. */
167	if (h->pgno == c->pg.pgno)
168		goto ret;
169
170	/*
171	 * Move right, looking for the page.  At each move we have to move
172	 * up the stack until we don't have to move to the next page.  If
173	 * we have to change pages at an internal level, we have to fix the
174	 * stack back up.
175	 */
176	while (h->pgno != c->pg.pgno) {
177		if ((nextpg = h->nextpg) == P_INVALID)
178			break;
179		mpool_put(t->bt_mp, h, 0);
180
181		/* Move up the stack. */
182		for (level = 0; (parent = BT_POP(t)) != NULL; ++level) {
183			/* Get the parent page. */
184			if ((h = mpool_get(t->bt_mp, parent->pgno, 0)) == NULL)
185				return (1);
186
187			/* Move to the next index. */
188			if (parent->index != NEXTINDEX(h) - 1) {
189				index = parent->index + 1;
190				BT_PUSH(t, h->pgno, index);
191				break;
192			}
193			mpool_put(t->bt_mp, h, 0);
194		}
195
196		/* Restore the stack. */
197		while (level--) {
198			/* Push the next level down onto the stack. */
199			bi = GETBINTERNAL(h, index);
200			pgno = bi->pgno;
201			BT_PUSH(t, pgno, 0);
202
203			/* Lose the currently pinned page. */
204			mpool_put(t->bt_mp, h, 0);
205
206			/* Get the next level down. */
207			if ((h = mpool_get(t->bt_mp, pgno, 0)) == NULL)
208				return (1);
209			index = 0;
210		}
211		mpool_put(t->bt_mp, h, 0);
212		if ((h = mpool_get(t->bt_mp, nextpg, 0)) == NULL)
213			return (1);
214	}
215
216	if (h->pgno == c->pg.pgno)
217		goto ret;
218
219	/* Reacquire the original stack. */
220	mpool_put(t->bt_mp, h, 0);
221	if ((e = __bt_search(t, &c->key, &exact)) == NULL)
222		return (1);
223	h = e->page;
224
225	/*
226	 * Move left, looking for the page.  At each move we have to move
227	 * up the stack until we don't have to change pages to move to the
228	 * next page.  If we have to change pages at an internal level, we
229	 * have to fix the stack back up.
230	 */
231	while (h->pgno != c->pg.pgno) {
232		if ((prevpg = h->prevpg) == P_INVALID)
233			break;
234		mpool_put(t->bt_mp, h, 0);
235
236		/* Move up the stack. */
237		for (level = 0; (parent = BT_POP(t)) != NULL; ++level) {
238			/* Get the parent page. */
239			if ((h = mpool_get(t->bt_mp, parent->pgno, 0)) == NULL)
240				return (1);
241
242			/* Move to the next index. */
243			if (parent->index != 0) {
244				index = parent->index - 1;
245				BT_PUSH(t, h->pgno, index);
246				break;
247			}
248			mpool_put(t->bt_mp, h, 0);
249		}
250
251		/* Restore the stack. */
252		while (level--) {
253			/* Push the next level down onto the stack. */
254			bi = GETBINTERNAL(h, index);
255			pgno = bi->pgno;
256
257			/* Lose the currently pinned page. */
258			mpool_put(t->bt_mp, h, 0);
259
260			/* Get the next level down. */
261			if ((h = mpool_get(t->bt_mp, pgno, 0)) == NULL)
262				return (1);
263
264			index = NEXTINDEX(h) - 1;
265			BT_PUSH(t, pgno, index);
266		}
267		mpool_put(t->bt_mp, h, 0);
268		if ((h = mpool_get(t->bt_mp, prevpg, 0)) == NULL)
269			return (1);
270	}
271
272
273ret:	mpool_put(t->bt_mp, h, 0);
274	return ((*hp = mpool_get(t->bt_mp, c->pg.pgno, 0)) == NULL);
275}
276
277/*
278 * __bt_bdelete --
279 *	Delete all key/data pairs matching the specified key.
280 *
281 * Parameters:
282 *	  t:	tree
283 *	key:	key to delete
284 *
285 * Returns:
286 *	RET_ERROR, RET_SUCCESS and RET_SPECIAL if the key not found.
287 */
288static int
289__bt_bdelete(t, key)
290	BTREE *t;
291	const DBT *key;
292{
293	EPG *e;
294	PAGE *h;
295	int deleted, exact, redo;
296
297	deleted = 0;
298
299	/* Find any matching record; __bt_search pins the page. */
300loop:	if ((e = __bt_search(t, key, &exact)) == NULL)
301		return (deleted ? RET_SUCCESS : RET_ERROR);
302	if (!exact) {
303		mpool_put(t->bt_mp, e->page, 0);
304		return (deleted ? RET_SUCCESS : RET_SPECIAL);
305	}
306
307	/*
308	 * Delete forward, then delete backward, from the found key.  If
309	 * there are duplicates and we reach either side of the page, do
310	 * the key search again, so that we get them all.
311	 */
312	redo = 0;
313	h = e->page;
314	do {
315		if (__bt_dleaf(t, key, h, e->index)) {
316			mpool_put(t->bt_mp, h, 0);
317			return (RET_ERROR);
318		}
319		if (F_ISSET(t, B_NODUPS)) {
320			if (NEXTINDEX(h) == 0) {
321				if (__bt_pdelete(t, h))
322					return (RET_ERROR);
323			} else
324				mpool_put(t->bt_mp, h, MPOOL_DIRTY);
325			return (RET_SUCCESS);
326		}
327		deleted = 1;
328	} while (e->index < NEXTINDEX(h) && __bt_cmp(t, key, e) == 0);
329
330	/* Check for right-hand edge of the page. */
331	if (e->index == NEXTINDEX(h))
332		redo = 1;
333
334	/* Delete from the key to the beginning of the page. */
335	while (e->index-- > 0) {
336		if (__bt_cmp(t, key, e) != 0)
337			break;
338		if (__bt_dleaf(t, key, h, e->index) == RET_ERROR) {
339			mpool_put(t->bt_mp, h, 0);
340			return (RET_ERROR);
341		}
342		if (e->index == 0)
343			redo = 1;
344	}
345
346	/* Check for an empty page. */
347	if (NEXTINDEX(h) == 0) {
348		if (__bt_pdelete(t, h))
349			return (RET_ERROR);
350		goto loop;
351	}
352
353	/* Put the page. */
354	mpool_put(t->bt_mp, h, MPOOL_DIRTY);
355
356	if (redo)
357		goto loop;
358	return (RET_SUCCESS);
359}
360
361/*
362 * __bt_pdelete --
363 *	Delete a single page from the tree.
364 *
365 * Parameters:
366 *	t:	tree
367 *	h:	leaf page
368 *
369 * Returns:
370 *	RET_SUCCESS, RET_ERROR.
371 *
372 * Side-effects:
373 *	mpool_put's the page
374 */
375static int
376__bt_pdelete(t, h)
377	BTREE *t;
378	PAGE *h;
379{
380	BINTERNAL *bi;
381	PAGE *pg;
382	EPGNO *parent;
383	indx_t cnt, index, *ip, offset;
384	u_int32_t nksize;
385	char *from;
386
387	/*
388	 * Walk the parent page stack -- a LIFO stack of the pages that were
389	 * traversed when we searched for the page where the delete occurred.
390	 * Each stack entry is a page number and a page index offset.  The
391	 * offset is for the page traversed on the search.  We've just deleted
392	 * a page, so we have to delete the key from the parent page.
393	 *
394	 * If the delete from the parent page makes it empty, this process may
395	 * continue all the way up the tree.  We stop if we reach the root page
396	 * (which is never deleted, it's just not worth the effort) or if the
397	 * delete does not empty the page.
398	 */
399	while ((parent = BT_POP(t)) != NULL) {
400		/* Get the parent page. */
401		if ((pg = mpool_get(t->bt_mp, parent->pgno, 0)) == NULL)
402			return (RET_ERROR);
403
404		index = parent->index;
405		bi = GETBINTERNAL(pg, index);
406
407		/* Free any overflow pages. */
408		if (bi->flags & P_BIGKEY &&
409		    __ovfl_delete(t, bi->bytes) == RET_ERROR) {
410			mpool_put(t->bt_mp, pg, 0);
411			return (RET_ERROR);
412		}
413
414		/*
415		 * Free the parent if it has only the one key and it's not the
416		 * root page. If it's the rootpage, turn it back into an empty
417		 * leaf page.
418		 */
419		if (NEXTINDEX(pg) == 1)
420			if (pg->pgno == P_ROOT) {
421				pg->lower = BTDATAOFF;
422				pg->upper = t->bt_psize;
423				pg->flags = P_BLEAF;
424			} else {
425				if (__bt_relink(t, pg) || __bt_free(t, pg))
426					return (RET_ERROR);
427				continue;
428			}
429		else {
430			/* Pack remaining key items at the end of the page. */
431			nksize = NBINTERNAL(bi->ksize);
432			from = (char *)pg + pg->upper;
433			memmove(from + nksize, from, (char *)bi - from);
434			pg->upper += nksize;
435
436			/* Adjust indices' offsets, shift the indices down. */
437			offset = pg->linp[index];
438			for (cnt = index, ip = &pg->linp[0]; cnt--; ++ip)
439				if (ip[0] < offset)
440					ip[0] += nksize;
441			for (cnt = NEXTINDEX(pg) - index; --cnt; ++ip)
442				ip[0] = ip[1] < offset ? ip[1] + nksize : ip[1];
443			pg->lower -= sizeof(indx_t);
444		}
445
446		mpool_put(t->bt_mp, pg, MPOOL_DIRTY);
447		break;
448	}
449
450	/* Free the leaf page, as long as it wasn't the root. */
451	if (h->pgno == P_ROOT) {
452		mpool_put(t->bt_mp, h, MPOOL_DIRTY);
453		return (RET_SUCCESS);
454	}
455	return (__bt_relink(t, h) || __bt_free(t, h));
456}
457
458/*
459 * __bt_dleaf --
460 *	Delete a single record from a leaf page.
461 *
462 * Parameters:
463 *	t:	tree
464 *    key:	referenced key
465 *	h:	page
466 *	index:	index on page to delete
467 *
468 * Returns:
469 *	RET_SUCCESS, RET_ERROR.
470 */
471int
472__bt_dleaf(t, key, h, index)
473	BTREE *t;
474	const DBT *key;
475	PAGE *h;
476	u_int index;
477{
478	BLEAF *bl;
479	indx_t cnt, *ip, offset;
480	u_int32_t nbytes;
481	void *to;
482	char *from;
483
484	/* If this record is referenced by the cursor, delete the cursor. */
485	if (F_ISSET(&t->bt_cursor, CURS_INIT) &&
486	    !F_ISSET(&t->bt_cursor, CURS_ACQUIRE) &&
487	    t->bt_cursor.pg.pgno == h->pgno && t->bt_cursor.pg.index == index &&
488	    __bt_curdel(t, key, h, index))
489		return (RET_ERROR);
490
491	/* If the entry uses overflow pages, make them available for reuse. */
492	to = bl = GETBLEAF(h, index);
493	if (bl->flags & P_BIGKEY && __ovfl_delete(t, bl->bytes) == RET_ERROR)
494		return (RET_ERROR);
495	if (bl->flags & P_BIGDATA &&
496	    __ovfl_delete(t, bl->bytes + bl->ksize) == RET_ERROR)
497		return (RET_ERROR);
498
499	/* Pack the remaining key/data items at the end of the page. */
500	nbytes = NBLEAF(bl);
501	from = (char *)h + h->upper;
502	memmove(from + nbytes, from, (char *)to - from);
503	h->upper += nbytes;
504
505	/* Adjust the indices' offsets, shift the indices down. */
506	offset = h->linp[index];
507	for (cnt = index, ip = &h->linp[0]; cnt--; ++ip)
508		if (ip[0] < offset)
509			ip[0] += nbytes;
510	for (cnt = NEXTINDEX(h) - index; --cnt; ++ip)
511		ip[0] = ip[1] < offset ? ip[1] + nbytes : ip[1];
512	h->lower -= sizeof(indx_t);
513
514	/* If the cursor is on this page, adjust it as necessary. */
515	if (F_ISSET(&t->bt_cursor, CURS_INIT) &&
516	    !F_ISSET(&t->bt_cursor, CURS_ACQUIRE) &&
517	    t->bt_cursor.pg.pgno == h->pgno && t->bt_cursor.pg.index > index)
518		--t->bt_cursor.pg.index;
519
520	return (RET_SUCCESS);
521}
522
523/*
524 * __bt_curdel --
525 *	Delete the cursor.
526 *
527 * Parameters:
528 *	t:	tree
529 *    key:	referenced key (or NULL)
530 *	h:	page
531 *  index:	index on page to delete
532 *
533 * Returns:
534 *	RET_SUCCESS, RET_ERROR.
535 */
536static int
537__bt_curdel(t, key, h, index)
538	BTREE *t;
539	const DBT *key;
540	PAGE *h;
541	u_int index;
542{
543	CURSOR *c;
544	EPG e;
545	PAGE *pg;
546	int curcopy, status;
547
548	/*
549	 * If there are duplicates, move forward or backward to one.
550	 * Otherwise, copy the key into the cursor area.
551	 */
552	c = &t->bt_cursor;
553	F_CLR(c, CURS_AFTER | CURS_BEFORE | CURS_ACQUIRE);
554
555	curcopy = 0;
556	if (!F_ISSET(t, B_NODUPS)) {
557		/*
558		 * We're going to have to do comparisons.  If we weren't
559		 * provided a copy of the key, i.e. the user is deleting
560		 * the current cursor position, get one.
561		 */
562		if (key == NULL) {
563			e.page = h;
564			e.index = index;
565			if ((status = __bt_ret(t, &e,
566			    &c->key, &c->key, NULL, NULL, 1)) != RET_SUCCESS)
567				return (status);
568			curcopy = 1;
569			key = &c->key;
570		}
571		/* Check previous key, if not at the beginning of the page. */
572		if (index > 0) {
573			e.page = h;
574			e.index = index - 1;
575			if (__bt_cmp(t, key, &e) == 0) {
576				F_SET(c, CURS_BEFORE);
577				goto dup2;
578			}
579		}
580		/* Check next key, if not at the end of the page. */
581		if (index < NEXTINDEX(h) - 1) {
582			e.page = h;
583			e.index = index + 1;
584			if (__bt_cmp(t, key, &e) == 0) {
585				F_SET(c, CURS_AFTER);
586				goto dup2;
587			}
588		}
589		/* Check previous key if at the beginning of the page. */
590		if (index == 0 && h->prevpg != P_INVALID) {
591			if ((pg = mpool_get(t->bt_mp, h->prevpg, 0)) == NULL)
592				return (RET_ERROR);
593			e.page = pg;
594			e.index = NEXTINDEX(pg) - 1;
595			if (__bt_cmp(t, key, &e) == 0) {
596				F_SET(c, CURS_BEFORE);
597				goto dup1;
598			}
599			mpool_put(t->bt_mp, pg, 0);
600		}
601		/* Check next key if at the end of the page. */
602		if (index == NEXTINDEX(h) - 1 && h->nextpg != P_INVALID) {
603			if ((pg = mpool_get(t->bt_mp, h->nextpg, 0)) == NULL)
604				return (RET_ERROR);
605			e.page = pg;
606			e.index = 0;
607			if (__bt_cmp(t, key, &e) == 0) {
608				F_SET(c, CURS_AFTER);
609dup1:				mpool_put(t->bt_mp, pg, 0);
610dup2:				c->pg.pgno = e.page->pgno;
611				c->pg.index = e.index;
612				return (RET_SUCCESS);
613			}
614			mpool_put(t->bt_mp, pg, 0);
615		}
616	}
617	e.page = h;
618	e.index = index;
619	if (curcopy || (status =
620	    __bt_ret(t, &e, &c->key, &c->key, NULL, NULL, 1)) == RET_SUCCESS) {
621		F_SET(c, CURS_ACQUIRE);
622		return (RET_SUCCESS);
623	}
624	return (status);
625}
626
627/*
628 * __bt_relink --
629 *	Link around a deleted page.
630 *
631 * Parameters:
632 *	t:	tree
633 *	h:	page to be deleted
634 */
635static int
636__bt_relink(t, h)
637	BTREE *t;
638	PAGE *h;
639{
640	PAGE *pg;
641
642	if (h->nextpg != P_INVALID) {
643		if ((pg = mpool_get(t->bt_mp, h->nextpg, 0)) == NULL)
644			return (RET_ERROR);
645		pg->prevpg = h->prevpg;
646		mpool_put(t->bt_mp, pg, MPOOL_DIRTY);
647	}
648	if (h->prevpg != P_INVALID) {
649		if ((pg = mpool_get(t->bt_mp, h->prevpg, 0)) == NULL)
650			return (RET_ERROR);
651		pg->nextpg = h->nextpg;
652		mpool_put(t->bt_mp, pg, MPOOL_DIRTY);
653	}
654	return (0);
655}
656