log.c 20.3 KB
Newer Older
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319
/*
 * This file is part of UBIFS.
 *
 * Copyright (C) 2006-2008 Nokia Corporation.
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License version 2 as published by
 * the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
 * more details.
 *
 * You should have received a copy of the GNU General Public License along with
 * this program; if not, write to the Free Software Foundation, Inc., 51
 * Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
 *
 * Authors: Artem Bityutskiy (Битюцкий Артём)
 *          Adrian Hunter
 */

/*
 * This file is a part of UBIFS journal implementation and contains various
 * functions which manipulate the log. The log is a fixed area on the flash
 * which does not contain any data but refers to buds. The log is a part of the
 * journal.
 */

#include "ubifs.h"

#ifdef CONFIG_UBIFS_FS_DEBUG
static int dbg_check_bud_bytes(struct ubifs_info *c);
#else
#define dbg_check_bud_bytes(c) 0
#endif

/**
 * ubifs_search_bud - search bud LEB.
 * @c: UBIFS file-system description object
 * @lnum: logical eraseblock number to search
 *
 * This function searches bud LEB @lnum. Returns bud description object in case
 * of success and %NULL if there is no bud with this LEB number.
 */
struct ubifs_bud *ubifs_search_bud(struct ubifs_info *c, int lnum)
{
	struct rb_node *p;
	struct ubifs_bud *bud;

	spin_lock(&c->buds_lock);
	p = c->buds.rb_node;
	while (p) {
		bud = rb_entry(p, struct ubifs_bud, rb);
		if (lnum < bud->lnum)
			p = p->rb_left;
		else if (lnum > bud->lnum)
			p = p->rb_right;
		else {
			spin_unlock(&c->buds_lock);
			return bud;
		}
	}
	spin_unlock(&c->buds_lock);
	return NULL;
}

/**
 * ubifs_get_wbuf - get the wbuf associated with a LEB, if there is one.
 * @c: UBIFS file-system description object
 * @lnum: logical eraseblock number to search
 *
 * This functions returns the wbuf for @lnum or %NULL if there is not one.
 */
struct ubifs_wbuf *ubifs_get_wbuf(struct ubifs_info *c, int lnum)
{
	struct rb_node *p;
	struct ubifs_bud *bud;
	int jhead;

	if (!c->jheads)
		return NULL;

	spin_lock(&c->buds_lock);
	p = c->buds.rb_node;
	while (p) {
		bud = rb_entry(p, struct ubifs_bud, rb);
		if (lnum < bud->lnum)
			p = p->rb_left;
		else if (lnum > bud->lnum)
			p = p->rb_right;
		else {
			jhead = bud->jhead;
			spin_unlock(&c->buds_lock);
			return &c->jheads[jhead].wbuf;
		}
	}
	spin_unlock(&c->buds_lock);
	return NULL;
}

/**
 * next_log_lnum - switch to the next log LEB.
 * @c: UBIFS file-system description object
 * @lnum: current log LEB
 */
static inline int next_log_lnum(const struct ubifs_info *c, int lnum)
{
	lnum += 1;
	if (lnum > c->log_last)
		lnum = UBIFS_LOG_LNUM;

	return lnum;
}

/**
 * empty_log_bytes - calculate amount of empty space in the log.
 * @c: UBIFS file-system description object
 */
static inline long long empty_log_bytes(const struct ubifs_info *c)
{
	long long h, t;

	h = (long long)c->lhead_lnum * c->leb_size + c->lhead_offs;
	t = (long long)c->ltail_lnum * c->leb_size;

	if (h >= t)
		return c->log_bytes - h + t;
	else
		return t - h;
}

/**
 * ubifs_add_bud - add bud LEB to the tree of buds and its journal head list.
 * @c: UBIFS file-system description object
 * @bud: the bud to add
 */
void ubifs_add_bud(struct ubifs_info *c, struct ubifs_bud *bud)
{
	struct rb_node **p, *parent = NULL;
	struct ubifs_bud *b;
	struct ubifs_jhead *jhead;

	spin_lock(&c->buds_lock);
	p = &c->buds.rb_node;
	while (*p) {
		parent = *p;
		b = rb_entry(parent, struct ubifs_bud, rb);
		ubifs_assert(bud->lnum != b->lnum);
		if (bud->lnum < b->lnum)
			p = &(*p)->rb_left;
		else
			p = &(*p)->rb_right;
	}

	rb_link_node(&bud->rb, parent, p);
	rb_insert_color(&bud->rb, &c->buds);
	if (c->jheads) {
		jhead = &c->jheads[bud->jhead];
		list_add_tail(&bud->list, &jhead->buds_list);
	} else
		ubifs_assert(c->replaying && (c->vfs_sb->s_flags & MS_RDONLY));

	/*
	 * Note, although this is a new bud, we anyway account this space now,
	 * before any data has been written to it, because this is about to
	 * guarantee fixed mount time, and this bud will anyway be read and
	 * scanned.
	 */
	c->bud_bytes += c->leb_size - bud->start;

	dbg_log("LEB %d:%d, jhead %d, bud_bytes %lld", bud->lnum,
		bud->start, bud->jhead, c->bud_bytes);
	spin_unlock(&c->buds_lock);
}

/**
 * ubifs_create_buds_lists - create journal head buds lists for remount rw.
 * @c: UBIFS file-system description object
 */
void ubifs_create_buds_lists(struct ubifs_info *c)
{
	struct rb_node *p;

	spin_lock(&c->buds_lock);
	p = rb_first(&c->buds);
	while (p) {
		struct ubifs_bud *bud = rb_entry(p, struct ubifs_bud, rb);
		struct ubifs_jhead *jhead = &c->jheads[bud->jhead];

		list_add_tail(&bud->list, &jhead->buds_list);
		p = rb_next(p);
	}
	spin_unlock(&c->buds_lock);
}

/**
 * ubifs_add_bud_to_log - add a new bud to the log.
 * @c: UBIFS file-system description object
 * @jhead: journal head the bud belongs to
 * @lnum: LEB number of the bud
 * @offs: starting offset of the bud
 *
 * This function writes reference node for the new bud LEB @lnum it to the log,
 * and adds it to the buds tress. It also makes sure that log size does not
 * exceed the 'c->max_bud_bytes' limit. Returns zero in case of success,
 * %-EAGAIN if commit is required, and a negative error codes in case of
 * failure.
 */
int ubifs_add_bud_to_log(struct ubifs_info *c, int jhead, int lnum, int offs)
{
	int err;
	struct ubifs_bud *bud;
	struct ubifs_ref_node *ref;

	bud = kmalloc(sizeof(struct ubifs_bud), GFP_NOFS);
	if (!bud)
		return -ENOMEM;
	ref = kzalloc(c->ref_node_alsz, GFP_NOFS);
	if (!ref) {
		kfree(bud);
		return -ENOMEM;
	}

	mutex_lock(&c->log_mutex);

	if (c->ro_media) {
		err = -EROFS;
		goto out_unlock;
	}

	/* Make sure we have enough space in the log */
	if (empty_log_bytes(c) - c->ref_node_alsz < c->min_log_bytes) {
		dbg_log("not enough log space - %lld, required %d",
			empty_log_bytes(c), c->min_log_bytes);
		ubifs_commit_required(c);
		err = -EAGAIN;
		goto out_unlock;
	}

	/*
	 * Make sure the the amount of space in buds will not exceed
	 * 'c->max_bud_bytes' limit, because we want to guarantee mount time
	 * limits.
	 *
	 * It is not necessary to hold @c->buds_lock when reading @c->bud_bytes
	 * because we are holding @c->log_mutex. All @c->bud_bytes take place
	 * when both @c->log_mutex and @c->bud_bytes are locked.
	 */
	if (c->bud_bytes + c->leb_size - offs > c->max_bud_bytes) {
		dbg_log("bud bytes %lld (%lld max), require commit",
			c->bud_bytes, c->max_bud_bytes);
		ubifs_commit_required(c);
		err = -EAGAIN;
		goto out_unlock;
	}

	/*
	 * If the journal is full enough - start background commit. Note, it is
	 * OK to read 'c->cmt_state' without spinlock because integer reads
	 * are atomic in the kernel.
	 */
	if (c->bud_bytes >= c->bg_bud_bytes &&
	    c->cmt_state == COMMIT_RESTING) {
		dbg_log("bud bytes %lld (%lld max), initiate BG commit",
			c->bud_bytes, c->max_bud_bytes);
		ubifs_request_bg_commit(c);
	}

	bud->lnum = lnum;
	bud->start = offs;
	bud->jhead = jhead;

	ref->ch.node_type = UBIFS_REF_NODE;
	ref->lnum = cpu_to_le32(bud->lnum);
	ref->offs = cpu_to_le32(bud->start);
	ref->jhead = cpu_to_le32(jhead);

	if (c->lhead_offs > c->leb_size - c->ref_node_alsz) {
		c->lhead_lnum = next_log_lnum(c, c->lhead_lnum);
		c->lhead_offs = 0;
	}

	if (c->lhead_offs == 0) {
		/* Must ensure next log LEB has been unmapped */
		err = ubifs_leb_unmap(c, c->lhead_lnum);
		if (err)
			goto out_unlock;
	}

	if (bud->start == 0) {
		/*
		 * Before writing the LEB reference which refers an empty LEB
		 * to the log, we have to make sure it is mapped, because
		 * otherwise we'd risk to refer an LEB with garbage in case of
		 * an unclean reboot, because the target LEB might have been
		 * unmapped, but not yet physically erased.
		 */
		err = ubi_leb_map(c->ubi, bud->lnum, UBI_SHORTTERM);
		if (err)
			goto out_unlock;
	}

	dbg_log("write ref LEB %d:%d",
		c->lhead_lnum, c->lhead_offs);
	err = ubifs_write_node(c, ref, UBIFS_REF_NODE_SZ, c->lhead_lnum,
			       c->lhead_offs, UBI_SHORTTERM);
	if (err)
		goto out_unlock;

	c->lhead_offs += c->ref_node_alsz;

	ubifs_add_bud(c, bud);

	mutex_unlock(&c->log_mutex);
	kfree(ref);
	return 0;

out_unlock:
320 321
	if (err != -EAGAIN)
		ubifs_ro_mode(c, err);
322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376
	mutex_unlock(&c->log_mutex);
	kfree(ref);
	kfree(bud);
	return err;
}

/**
 * remove_buds - remove used buds.
 * @c: UBIFS file-system description object
 *
 * This function removes use buds from the buds tree. It does not remove the
 * buds which are pointed to by journal heads.
 */
static void remove_buds(struct ubifs_info *c)
{
	struct rb_node *p;

	ubifs_assert(list_empty(&c->old_buds));
	c->cmt_bud_bytes = 0;
	spin_lock(&c->buds_lock);
	p = rb_first(&c->buds);
	while (p) {
		struct rb_node *p1 = p;
		struct ubifs_bud *bud;
		struct ubifs_wbuf *wbuf;

		p = rb_next(p);
		bud = rb_entry(p1, struct ubifs_bud, rb);
		wbuf = &c->jheads[bud->jhead].wbuf;

		if (wbuf->lnum == bud->lnum) {
			/*
			 * Do not remove buds which are pointed to by journal
			 * heads (non-closed buds).
			 */
			c->cmt_bud_bytes += wbuf->offs - bud->start;
			dbg_log("preserve %d:%d, jhead %d, bud bytes %d, "
				"cmt_bud_bytes %lld", bud->lnum, bud->start,
				bud->jhead, wbuf->offs - bud->start,
				c->cmt_bud_bytes);
			bud->start = wbuf->offs;
		} else {
			c->cmt_bud_bytes += c->leb_size - bud->start;
			dbg_log("remove %d:%d, jhead %d, bud bytes %d, "
				"cmt_bud_bytes %lld", bud->lnum, bud->start,
				bud->jhead, c->leb_size - bud->start,
				c->cmt_bud_bytes);
			rb_erase(p1, &c->buds);
			/*
			 * If the commit does not finish, the recovery will need
			 * to replay the journal, in which case the old buds
			 * must be unchanged. Do not release them until post
			 * commit i.e. do not allow them to be garbage
			 * collected.
			 */
E
Eric Sesterhenn 已提交
377
			list_move(&bud->list, &c->old_buds);
378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413
		}
	}
	spin_unlock(&c->buds_lock);
}

/**
 * ubifs_log_start_commit - start commit.
 * @c: UBIFS file-system description object
 * @ltail_lnum: return new log tail LEB number
 *
 * The commit operation starts with writing "commit start" node to the log and
 * reference nodes for all journal heads which will define new journal after
 * the commit has been finished. The commit start and reference nodes are
 * written in one go to the nearest empty log LEB (hence, when commit is
 * finished UBIFS may safely unmap all the previous log LEBs). This function
 * returns zero in case of success and a negative error code in case of
 * failure.
 */
int ubifs_log_start_commit(struct ubifs_info *c, int *ltail_lnum)
{
	void *buf;
	struct ubifs_cs_node *cs;
	struct ubifs_ref_node *ref;
	int err, i, max_len, len;

	err = dbg_check_bud_bytes(c);
	if (err)
		return err;

	max_len = UBIFS_CS_NODE_SZ + c->jhead_cnt * UBIFS_REF_NODE_SZ;
	max_len = ALIGN(max_len, c->min_io_size);
	buf = cs = kmalloc(max_len, GFP_NOFS);
	if (!buf)
		return -ENOMEM;

	cs->ch.node_type = UBIFS_CS_NODE;
414
	cs->cmt_no = cpu_to_le64(c->cmt_no);
415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806
	ubifs_prepare_node(c, cs, UBIFS_CS_NODE_SZ, 0);

	/*
	 * Note, we do not lock 'c->log_mutex' because this is the commit start
	 * phase and we are exclusively using the log. And we do not lock
	 * write-buffer because nobody can write to the file-system at this
	 * phase.
	 */

	len = UBIFS_CS_NODE_SZ;
	for (i = 0; i < c->jhead_cnt; i++) {
		int lnum = c->jheads[i].wbuf.lnum;
		int offs = c->jheads[i].wbuf.offs;

		if (lnum == -1 || offs == c->leb_size)
			continue;

		dbg_log("add ref to LEB %d:%d for jhead %d", lnum, offs, i);
		ref = buf + len;
		ref->ch.node_type = UBIFS_REF_NODE;
		ref->lnum = cpu_to_le32(lnum);
		ref->offs = cpu_to_le32(offs);
		ref->jhead = cpu_to_le32(i);

		ubifs_prepare_node(c, ref, UBIFS_REF_NODE_SZ, 0);
		len += UBIFS_REF_NODE_SZ;
	}

	ubifs_pad(c, buf + len, ALIGN(len, c->min_io_size) - len);

	/* Switch to the next log LEB */
	if (c->lhead_offs) {
		c->lhead_lnum = next_log_lnum(c, c->lhead_lnum);
		c->lhead_offs = 0;
	}

	if (c->lhead_offs == 0) {
		/* Must ensure next LEB has been unmapped */
		err = ubifs_leb_unmap(c, c->lhead_lnum);
		if (err)
			goto out;
	}

	len = ALIGN(len, c->min_io_size);
	dbg_log("writing commit start at LEB %d:0, len %d", c->lhead_lnum, len);
	err = ubifs_leb_write(c, c->lhead_lnum, cs, 0, len, UBI_SHORTTERM);
	if (err)
		goto out;

	*ltail_lnum = c->lhead_lnum;

	c->lhead_offs += len;
	if (c->lhead_offs == c->leb_size) {
		c->lhead_lnum = next_log_lnum(c, c->lhead_lnum);
		c->lhead_offs = 0;
	}

	remove_buds(c);

	/*
	 * We have started the commit and now users may use the rest of the log
	 * for new writes.
	 */
	c->min_log_bytes = 0;

out:
	kfree(buf);
	return err;
}

/**
 * ubifs_log_end_commit - end commit.
 * @c: UBIFS file-system description object
 * @ltail_lnum: new log tail LEB number
 *
 * This function is called on when the commit operation was finished. It
 * moves log tail to new position and unmaps LEBs which contain obsolete data.
 * Returns zero in case of success and a negative error code in case of
 * failure.
 */
int ubifs_log_end_commit(struct ubifs_info *c, int ltail_lnum)
{
	int err;

	/*
	 * At this phase we have to lock 'c->log_mutex' because UBIFS allows FS
	 * writes during commit. Its only short "commit" start phase when
	 * writers are blocked.
	 */
	mutex_lock(&c->log_mutex);

	dbg_log("old tail was LEB %d:0, new tail is LEB %d:0",
		c->ltail_lnum, ltail_lnum);

	c->ltail_lnum = ltail_lnum;
	/*
	 * The commit is finished and from now on it must be guaranteed that
	 * there is always enough space for the next commit.
	 */
	c->min_log_bytes = c->leb_size;

	spin_lock(&c->buds_lock);
	c->bud_bytes -= c->cmt_bud_bytes;
	spin_unlock(&c->buds_lock);

	err = dbg_check_bud_bytes(c);

	mutex_unlock(&c->log_mutex);
	return err;
}

/**
 * ubifs_log_post_commit - things to do after commit is completed.
 * @c: UBIFS file-system description object
 * @old_ltail_lnum: old log tail LEB number
 *
 * Release buds only after commit is completed, because they must be unchanged
 * if recovery is needed.
 *
 * Unmap log LEBs only after commit is completed, because they may be needed for
 * recovery.
 *
 * This function returns %0 on success and a negative error code on failure.
 */
int ubifs_log_post_commit(struct ubifs_info *c, int old_ltail_lnum)
{
	int lnum, err = 0;

	while (!list_empty(&c->old_buds)) {
		struct ubifs_bud *bud;

		bud = list_entry(c->old_buds.next, struct ubifs_bud, list);
		err = ubifs_return_leb(c, bud->lnum);
		if (err)
			return err;
		list_del(&bud->list);
		kfree(bud);
	}
	mutex_lock(&c->log_mutex);
	for (lnum = old_ltail_lnum; lnum != c->ltail_lnum;
	     lnum = next_log_lnum(c, lnum)) {
		dbg_log("unmap log LEB %d", lnum);
		err = ubifs_leb_unmap(c, lnum);
		if (err)
			goto out;
	}
out:
	mutex_unlock(&c->log_mutex);
	return err;
}

/**
 * struct done_ref - references that have been done.
 * @rb: rb-tree node
 * @lnum: LEB number
 */
struct done_ref {
	struct rb_node rb;
	int lnum;
};

/**
 * done_already - determine if a reference has been done already.
 * @done_tree: rb-tree to store references that have been done
 * @lnum: LEB number of reference
 *
 * This function returns %1 if the reference has been done, %0 if not, otherwise
 * a negative error code is returned.
 */
static int done_already(struct rb_root *done_tree, int lnum)
{
	struct rb_node **p = &done_tree->rb_node, *parent = NULL;
	struct done_ref *dr;

	while (*p) {
		parent = *p;
		dr = rb_entry(parent, struct done_ref, rb);
		if (lnum < dr->lnum)
			p = &(*p)->rb_left;
		else if (lnum > dr->lnum)
			p = &(*p)->rb_right;
		else
			return 1;
	}

	dr = kzalloc(sizeof(struct done_ref), GFP_NOFS);
	if (!dr)
		return -ENOMEM;

	dr->lnum = lnum;

	rb_link_node(&dr->rb, parent, p);
	rb_insert_color(&dr->rb, done_tree);

	return 0;
}

/**
 * destroy_done_tree - destroy the done tree.
 * @done_tree: done tree to destroy
 */
static void destroy_done_tree(struct rb_root *done_tree)
{
	struct rb_node *this = done_tree->rb_node;
	struct done_ref *dr;

	while (this) {
		if (this->rb_left) {
			this = this->rb_left;
			continue;
		} else if (this->rb_right) {
			this = this->rb_right;
			continue;
		}
		dr = rb_entry(this, struct done_ref, rb);
		this = rb_parent(this);
		if (this) {
			if (this->rb_left == &dr->rb)
				this->rb_left = NULL;
			else
				this->rb_right = NULL;
		}
		kfree(dr);
	}
}

/**
 * add_node - add a node to the consolidated log.
 * @c: UBIFS file-system description object
 * @buf: buffer to which to add
 * @lnum: LEB number to which to write is passed and returned here
 * @offs: offset to where to write is passed and returned here
 * @node: node to add
 *
 * This function returns %0 on success and a negative error code on failure.
 */
static int add_node(struct ubifs_info *c, void *buf, int *lnum, int *offs,
		    void *node)
{
	struct ubifs_ch *ch = node;
	int len = le32_to_cpu(ch->len), remains = c->leb_size - *offs;

	if (len > remains) {
		int sz = ALIGN(*offs, c->min_io_size), err;

		ubifs_pad(c, buf + *offs, sz - *offs);
		err = ubifs_leb_change(c, *lnum, buf, sz, UBI_SHORTTERM);
		if (err)
			return err;
		*lnum = next_log_lnum(c, *lnum);
		*offs = 0;
	}
	memcpy(buf + *offs, node, len);
	*offs += ALIGN(len, 8);
	return 0;
}

/**
 * ubifs_consolidate_log - consolidate the log.
 * @c: UBIFS file-system description object
 *
 * Repeated failed commits could cause the log to be full, but at least 1 LEB is
 * needed for commit. This function rewrites the reference nodes in the log
 * omitting duplicates, and failed CS nodes, and leaving no gaps.
 *
 * This function returns %0 on success and a negative error code on failure.
 */
int ubifs_consolidate_log(struct ubifs_info *c)
{
	struct ubifs_scan_leb *sleb;
	struct ubifs_scan_node *snod;
	struct rb_root done_tree = RB_ROOT;
	int lnum, err, first = 1, write_lnum, offs = 0;
	void *buf;

	dbg_rcvry("log tail LEB %d, log head LEB %d", c->ltail_lnum,
		  c->lhead_lnum);
	buf = vmalloc(c->leb_size);
	if (!buf)
		return -ENOMEM;
	lnum = c->ltail_lnum;
	write_lnum = lnum;
	while (1) {
		sleb = ubifs_scan(c, lnum, 0, c->sbuf);
		if (IS_ERR(sleb)) {
			err = PTR_ERR(sleb);
			goto out_free;
		}
		list_for_each_entry(snod, &sleb->nodes, list) {
			switch (snod->type) {
			case UBIFS_REF_NODE: {
				struct ubifs_ref_node *ref = snod->node;
				int ref_lnum = le32_to_cpu(ref->lnum);

				err = done_already(&done_tree, ref_lnum);
				if (err < 0)
					goto out_scan;
				if (err != 1) {
					err = add_node(c, buf, &write_lnum,
						       &offs, snod->node);
					if (err)
						goto out_scan;
				}
				break;
			}
			case UBIFS_CS_NODE:
				if (!first)
					break;
				err = add_node(c, buf, &write_lnum, &offs,
					       snod->node);
				if (err)
					goto out_scan;
				first = 0;
				break;
			}
		}
		ubifs_scan_destroy(sleb);
		if (lnum == c->lhead_lnum)
			break;
		lnum = next_log_lnum(c, lnum);
	}
	if (offs) {
		int sz = ALIGN(offs, c->min_io_size);

		ubifs_pad(c, buf + offs, sz - offs);
		err = ubifs_leb_change(c, write_lnum, buf, sz, UBI_SHORTTERM);
		if (err)
			goto out_free;
		offs = ALIGN(offs, c->min_io_size);
	}
	destroy_done_tree(&done_tree);
	vfree(buf);
	if (write_lnum == c->lhead_lnum) {
		ubifs_err("log is too full");
		return -EINVAL;
	}
	/* Unmap remaining LEBs */
	lnum = write_lnum;
	do {
		lnum = next_log_lnum(c, lnum);
		err = ubifs_leb_unmap(c, lnum);
		if (err)
			return err;
	} while (lnum != c->lhead_lnum);
	c->lhead_lnum = write_lnum;
	c->lhead_offs = offs;
	dbg_rcvry("new log head at %d:%d", c->lhead_lnum, c->lhead_offs);
	return 0;

out_scan:
	ubifs_scan_destroy(sleb);
out_free:
	destroy_done_tree(&done_tree);
	vfree(buf);
	return err;
}

#ifdef CONFIG_UBIFS_FS_DEBUG

/**
 * dbg_check_bud_bytes - make sure bud bytes calculation are all right.
 * @c: UBIFS file-system description object
 *
 * This function makes sure the amount of flash space used by closed buds
 * ('c->bud_bytes' is correct). Returns zero in case of success and %-EINVAL in
 * case of failure.
 */
static int dbg_check_bud_bytes(struct ubifs_info *c)
{
	int i, err = 0;
	struct ubifs_bud *bud;
	long long bud_bytes = 0;

	if (!(ubifs_chk_flags & UBIFS_CHK_GEN))
		return 0;

	spin_lock(&c->buds_lock);
	for (i = 0; i < c->jhead_cnt; i++)
		list_for_each_entry(bud, &c->jheads[i].buds_list, list)
			bud_bytes += c->leb_size - bud->start;

	if (c->bud_bytes != bud_bytes) {
		ubifs_err("bad bud_bytes %lld, calculated %lld",
			  c->bud_bytes, bud_bytes);
		err = -EINVAL;
	}
	spin_unlock(&c->buds_lock);

	return err;
}

#endif /* CONFIG_UBIFS_FS_DEBUG */