tree-log.c 109.2 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
/*
 * Copyright (C) 2008 Oracle.  All rights reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public
 * License v2 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., 59 Temple Place - Suite 330,
 * Boston, MA 021110-1307, USA.
 */

#include <linux/sched.h>
20
#include <linux/slab.h>
21
#include <linux/blkdev.h>
J
Josef Bacik 已提交
22
#include <linux/list_sort.h>
23 24 25 26 27
#include "ctree.h"
#include "transaction.h"
#include "disk-io.h"
#include "locking.h"
#include "print-tree.h"
M
Mark Fasheh 已提交
28
#include "backref.h"
29
#include "tree-log.h"
M
Mark Fasheh 已提交
30
#include "hash.h"
31 32 33 34 35 36 37 38 39 40

/* magic values for the inode_only field in btrfs_log_inode:
 *
 * LOG_INODE_ALL means to log everything
 * LOG_INODE_EXISTS means to log just enough to recreate the inode
 * during log replay
 */
#define LOG_INODE_ALL 0
#define LOG_INODE_EXISTS 1

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
/*
 * directory trouble cases
 *
 * 1) on rename or unlink, if the inode being unlinked isn't in the fsync
 * log, we must force a full commit before doing an fsync of the directory
 * where the unlink was done.
 * ---> record transid of last unlink/rename per directory
 *
 * mkdir foo/some_dir
 * normal commit
 * rename foo/some_dir foo2/some_dir
 * mkdir foo/some_dir
 * fsync foo/some_dir/some_file
 *
 * The fsync above will unlink the original some_dir without recording
 * it in its new location (foo2).  After a crash, some_dir will be gone
 * unless the fsync of some_file forces a full commit
 *
 * 2) we must log any new names for any file or dir that is in the fsync
 * log. ---> check inode while renaming/linking.
 *
 * 2a) we must log any new names for any file or dir during rename
 * when the directory they are being removed from was logged.
 * ---> check inode and old parent dir during rename
 *
 *  2a is actually the more important variant.  With the extra logging
 *  a crash might unlink the old name without recreating the new one
 *
 * 3) after a crash, we must go through any directories with a link count
 * of zero and redo the rm -rf
 *
 * mkdir f1/foo
 * normal commit
 * rm -rf f1/foo
 * fsync(f1)
 *
 * The directory f1 was fully removed from the FS, but fsync was never
 * called on f1, only its parent dir.  After a crash the rm -rf must
 * be replayed.  This must be able to recurse down the entire
 * directory tree.  The inode link count fixup code takes care of the
 * ugly details.
 */

84 85 86 87 88 89 90 91 92 93 94
/*
 * stages for the tree walking.  The first
 * stage (0) is to only pin down the blocks we find
 * the second stage (1) is to make sure that all the inodes
 * we find in the log are created in the subvolume.
 *
 * The last stage is to deal with directories and links and extents
 * and all the other fun semantics
 */
#define LOG_WALK_PIN_ONLY 0
#define LOG_WALK_REPLAY_INODES 1
95 96
#define LOG_WALK_REPLAY_DIR_INDEX 2
#define LOG_WALK_REPLAY_ALL 3
97

98
static int btrfs_log_inode(struct btrfs_trans_handle *trans,
99 100
			     struct btrfs_root *root, struct inode *inode,
			     int inode_only);
101 102 103
static int link_to_fixup_dir(struct btrfs_trans_handle *trans,
			     struct btrfs_root *root,
			     struct btrfs_path *path, u64 objectid);
104 105 106 107 108
static noinline int replay_dir_deletes(struct btrfs_trans_handle *trans,
				       struct btrfs_root *root,
				       struct btrfs_root *log,
				       struct btrfs_path *path,
				       u64 dirid, int del_all);
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

/*
 * tree logging is a special write ahead log used to make sure that
 * fsyncs and O_SYNCs can happen without doing full tree commits.
 *
 * Full tree commits are expensive because they require commonly
 * modified blocks to be recowed, creating many dirty pages in the
 * extent tree an 4x-6x higher write load than ext3.
 *
 * Instead of doing a tree commit on every fsync, we use the
 * key ranges and transaction ids to find items for a given file or directory
 * that have changed in this transaction.  Those items are copied into
 * a special tree (one per subvolume root), that tree is written to disk
 * and then the fsync is considered complete.
 *
 * After a crash, items are copied out of the log-tree back into the
 * subvolume tree.  Any file data extents found are recorded in the extent
 * allocation tree, and the log-tree freed.
 *
 * The log tree is read three times, once to pin down all the extents it is
 * using in ram and once, once to create all the inodes logged in the tree
 * and once to do all the other items.
 */

/*
 * start a sub transaction and setup the log tree
 * this increments the log tree writer count to make the people
 * syncing the tree wait for us to finish
 */
static int start_log_trans(struct btrfs_trans_handle *trans,
			   struct btrfs_root *root)
{
	int ret;
142
	int err = 0;
Y
Yan Zheng 已提交
143 144 145

	mutex_lock(&root->log_mutex);
	if (root->log_root) {
146 147 148 149 150 151 152
		if (!root->log_start_pid) {
			root->log_start_pid = current->pid;
			root->log_multiple_pids = false;
		} else if (root->log_start_pid != current->pid) {
			root->log_multiple_pids = true;
		}

M
Miao Xie 已提交
153
		atomic_inc(&root->log_batch);
Y
Yan Zheng 已提交
154 155 156 157
		atomic_inc(&root->log_writers);
		mutex_unlock(&root->log_mutex);
		return 0;
	}
158 159
	root->log_multiple_pids = false;
	root->log_start_pid = current->pid;
160 161 162
	mutex_lock(&root->fs_info->tree_log_mutex);
	if (!root->fs_info->log_root_tree) {
		ret = btrfs_init_log_root_tree(trans, root->fs_info);
163 164
		if (ret)
			err = ret;
165
	}
166
	if (err == 0 && !root->log_root) {
167
		ret = btrfs_add_log_tree(trans, root);
168 169
		if (ret)
			err = ret;
170 171
	}
	mutex_unlock(&root->fs_info->tree_log_mutex);
M
Miao Xie 已提交
172
	atomic_inc(&root->log_batch);
Y
Yan Zheng 已提交
173 174
	atomic_inc(&root->log_writers);
	mutex_unlock(&root->log_mutex);
175
	return err;
176 177 178 179 180 181 182 183 184 185 186 187 188 189 190
}

/*
 * returns 0 if there was a log transaction running and we were able
 * to join, or returns -ENOENT if there were not transactions
 * in progress
 */
static int join_running_log_trans(struct btrfs_root *root)
{
	int ret = -ENOENT;

	smp_mb();
	if (!root->log_root)
		return -ENOENT;

Y
Yan Zheng 已提交
191
	mutex_lock(&root->log_mutex);
192 193
	if (root->log_root) {
		ret = 0;
Y
Yan Zheng 已提交
194
		atomic_inc(&root->log_writers);
195
	}
Y
Yan Zheng 已提交
196
	mutex_unlock(&root->log_mutex);
197 198 199
	return ret;
}

200 201 202 203 204 205 206 207 208 209 210 211 212 213 214
/*
 * This either makes the current running log transaction wait
 * until you call btrfs_end_log_trans() or it makes any future
 * log transactions wait until you call btrfs_end_log_trans()
 */
int btrfs_pin_log_trans(struct btrfs_root *root)
{
	int ret = -ENOENT;

	mutex_lock(&root->log_mutex);
	atomic_inc(&root->log_writers);
	mutex_unlock(&root->log_mutex);
	return ret;
}

215 216 217 218
/*
 * indicate we're done making changes to the log tree
 * and wake up anyone waiting to do a sync
 */
219
void btrfs_end_log_trans(struct btrfs_root *root)
220
{
Y
Yan Zheng 已提交
221 222 223 224 225
	if (atomic_dec_and_test(&root->log_writers)) {
		smp_mb();
		if (waitqueue_active(&root->log_writer_wait))
			wake_up(&root->log_writer_wait);
	}
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
}


/*
 * the walk control struct is used to pass state down the chain when
 * processing the log tree.  The stage field tells us which part
 * of the log tree processing we are currently doing.  The others
 * are state fields used for that specific part
 */
struct walk_control {
	/* should we free the extent on disk when done?  This is used
	 * at transaction commit time while freeing a log tree
	 */
	int free;

	/* should we write out the extent buffer?  This is used
	 * while flushing the log tree to disk during a sync
	 */
	int write;

	/* should we wait for the extent buffer io to finish?  Also used
	 * while flushing the log tree to disk for a sync
	 */
	int wait;

	/* pin only walk, we record which extents on disk belong to the
	 * log trees
	 */
	int pin;

	/* what stage of the replay code we're currently in */
	int stage;

	/* the root we are currently replaying */
	struct btrfs_root *replay_dest;

	/* the trans handle for the current replay */
	struct btrfs_trans_handle *trans;

	/* the function that gets used to process blocks we find in the
	 * tree.  Note the extent_buffer might not be up to date when it is
	 * passed in, and it must be checked or read if you need the data
	 * inside it
	 */
	int (*process_func)(struct btrfs_root *log, struct extent_buffer *eb,
			    struct walk_control *wc, u64 gen);
};

/*
 * process_func used to pin down extents, write them or wait on them
 */
static int process_one_buffer(struct btrfs_root *log,
			      struct extent_buffer *eb,
			      struct walk_control *wc, u64 gen)
{
281 282
	int ret = 0;

283 284 285 286 287 288 289 290 291 292
	/*
	 * If this fs is mixed then we need to be able to process the leaves to
	 * pin down any logged extents, so we have to read the block.
	 */
	if (btrfs_fs_incompat(log->fs_info, MIXED_GROUPS)) {
		ret = btrfs_read_buffer(eb, gen);
		if (ret)
			return ret;
	}

J
Josef Bacik 已提交
293
	if (wc->pin)
294 295
		ret = btrfs_pin_extent_for_log_replay(log->fs_info->extent_root,
						      eb->start, eb->len);
296

297
	if (!ret && btrfs_buffer_uptodate(eb, gen, 0)) {
298 299
		if (wc->pin && btrfs_header_level(eb) == 0)
			ret = btrfs_exclude_logged_extents(log, eb);
300 301 302 303 304
		if (wc->write)
			btrfs_write_tree_block(eb);
		if (wc->wait)
			btrfs_wait_tree_block_writeback(eb);
	}
305
	return ret;
306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334
}

/*
 * Item overwrite used by replay and tree logging.  eb, slot and key all refer
 * to the src data we are copying out.
 *
 * root is the tree we are copying into, and path is a scratch
 * path for use in this function (it should be released on entry and
 * will be released on exit).
 *
 * If the key is already in the destination tree the existing item is
 * overwritten.  If the existing item isn't big enough, it is extended.
 * If it is too large, it is truncated.
 *
 * If the key isn't in the destination yet, a new item is inserted.
 */
static noinline int overwrite_item(struct btrfs_trans_handle *trans,
				   struct btrfs_root *root,
				   struct btrfs_path *path,
				   struct extent_buffer *eb, int slot,
				   struct btrfs_key *key)
{
	int ret;
	u32 item_size;
	u64 saved_i_size = 0;
	int save_old_i_size = 0;
	unsigned long src_ptr;
	unsigned long dst_ptr;
	int overwrite_root = 0;
335
	bool inode_item = key->type == BTRFS_INODE_ITEM_KEY;
336 337 338 339 340 341 342 343 344

	if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID)
		overwrite_root = 1;

	item_size = btrfs_item_size_nr(eb, slot);
	src_ptr = btrfs_item_ptr_offset(eb, slot);

	/* look for the key in the destination tree */
	ret = btrfs_search_slot(NULL, root, key, path, 0, 0);
345 346 347
	if (ret < 0)
		return ret;

348 349 350 351 352 353 354 355 356
	if (ret == 0) {
		char *src_copy;
		char *dst_copy;
		u32 dst_size = btrfs_item_size_nr(path->nodes[0],
						  path->slots[0]);
		if (dst_size != item_size)
			goto insert;

		if (item_size == 0) {
357
			btrfs_release_path(path);
358 359 360 361
			return 0;
		}
		dst_copy = kmalloc(item_size, GFP_NOFS);
		src_copy = kmalloc(item_size, GFP_NOFS);
362
		if (!dst_copy || !src_copy) {
363
			btrfs_release_path(path);
364 365 366 367
			kfree(dst_copy);
			kfree(src_copy);
			return -ENOMEM;
		}
368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384

		read_extent_buffer(eb, src_copy, src_ptr, item_size);

		dst_ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]);
		read_extent_buffer(path->nodes[0], dst_copy, dst_ptr,
				   item_size);
		ret = memcmp(dst_copy, src_copy, item_size);

		kfree(dst_copy);
		kfree(src_copy);
		/*
		 * they have the same contents, just return, this saves
		 * us from cowing blocks in the destination tree and doing
		 * extra writes that may not have been done by a previous
		 * sync
		 */
		if (ret == 0) {
385
			btrfs_release_path(path);
386 387 388
			return 0;
		}

389 390 391 392 393 394 395
		/*
		 * We need to load the old nbytes into the inode so when we
		 * replay the extents we've logged we get the right nbytes.
		 */
		if (inode_item) {
			struct btrfs_inode_item *item;
			u64 nbytes;
396
			u32 mode;
397 398 399 400 401 402 403

			item = btrfs_item_ptr(path->nodes[0], path->slots[0],
					      struct btrfs_inode_item);
			nbytes = btrfs_inode_nbytes(path->nodes[0], item);
			item = btrfs_item_ptr(eb, slot,
					      struct btrfs_inode_item);
			btrfs_set_inode_nbytes(eb, item, nbytes);
404 405 406 407 408 409 410 411 412

			/*
			 * If this is a directory we need to reset the i_size to
			 * 0 so that we can set it up properly when replaying
			 * the rest of the items in this log.
			 */
			mode = btrfs_inode_mode(eb, item);
			if (S_ISDIR(mode))
				btrfs_set_inode_size(eb, item, 0);
413 414 415
		}
	} else if (inode_item) {
		struct btrfs_inode_item *item;
416
		u32 mode;
417 418 419 420 421 422 423

		/*
		 * New inode, set nbytes to 0 so that the nbytes comes out
		 * properly when we replay the extents.
		 */
		item = btrfs_item_ptr(eb, slot, struct btrfs_inode_item);
		btrfs_set_inode_nbytes(eb, item, 0);
424 425 426 427 428 429 430 431 432

		/*
		 * If this is a directory we need to reset the i_size to 0 so
		 * that we can set it up properly when replaying the rest of
		 * the items in this log.
		 */
		mode = btrfs_inode_mode(eb, item);
		if (S_ISDIR(mode))
			btrfs_set_inode_size(eb, item, 0);
433 434
	}
insert:
435
	btrfs_release_path(path);
436 437 438 439 440 441 442 443 444
	/* try to insert the key into the destination tree */
	ret = btrfs_insert_empty_item(trans, root, path,
				      key, item_size);

	/* make sure any existing item is the correct size */
	if (ret == -EEXIST) {
		u32 found_size;
		found_size = btrfs_item_size_nr(path->nodes[0],
						path->slots[0]);
445
		if (found_size > item_size)
446
			btrfs_truncate_item(root, path, item_size, 1);
447
		else if (found_size < item_size)
448
			btrfs_extend_item(root, path,
449
					  item_size - found_size);
450
	} else if (ret) {
451
		return ret;
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
	}
	dst_ptr = btrfs_item_ptr_offset(path->nodes[0],
					path->slots[0]);

	/* don't overwrite an existing inode if the generation number
	 * was logged as zero.  This is done when the tree logging code
	 * is just logging an inode to make sure it exists after recovery.
	 *
	 * Also, don't overwrite i_size on directories during replay.
	 * log replay inserts and removes directory items based on the
	 * state of the tree found in the subvolume, and i_size is modified
	 * as it goes
	 */
	if (key->type == BTRFS_INODE_ITEM_KEY && ret == -EEXIST) {
		struct btrfs_inode_item *src_item;
		struct btrfs_inode_item *dst_item;

		src_item = (struct btrfs_inode_item *)src_ptr;
		dst_item = (struct btrfs_inode_item *)dst_ptr;

		if (btrfs_inode_generation(eb, src_item) == 0)
			goto no_copy;

		if (overwrite_root &&
		    S_ISDIR(btrfs_inode_mode(eb, src_item)) &&
		    S_ISDIR(btrfs_inode_mode(path->nodes[0], dst_item))) {
			save_old_i_size = 1;
			saved_i_size = btrfs_inode_size(path->nodes[0],
							dst_item);
		}
	}

	copy_extent_buffer(path->nodes[0], eb, dst_ptr,
			   src_ptr, item_size);

	if (save_old_i_size) {
		struct btrfs_inode_item *dst_item;
		dst_item = (struct btrfs_inode_item *)dst_ptr;
		btrfs_set_inode_size(path->nodes[0], dst_item, saved_i_size);
	}

	/* make sure the generation is filled in */
	if (key->type == BTRFS_INODE_ITEM_KEY) {
		struct btrfs_inode_item *dst_item;
		dst_item = (struct btrfs_inode_item *)dst_ptr;
		if (btrfs_inode_generation(path->nodes[0], dst_item) == 0) {
			btrfs_set_inode_generation(path->nodes[0], dst_item,
						   trans->transid);
		}
	}
no_copy:
	btrfs_mark_buffer_dirty(path->nodes[0]);
504
	btrfs_release_path(path);
505 506 507 508 509 510 511 512 513 514
	return 0;
}

/*
 * simple helper to read an inode off the disk from a given root
 * This can only be called for subvolume roots and not for the log
 */
static noinline struct inode *read_one_inode(struct btrfs_root *root,
					     u64 objectid)
{
515
	struct btrfs_key key;
516 517
	struct inode *inode;

518 519 520
	key.objectid = objectid;
	key.type = BTRFS_INODE_ITEM_KEY;
	key.offset = 0;
521
	inode = btrfs_iget(root->fs_info->sb, &key, root, NULL);
522 523 524
	if (IS_ERR(inode)) {
		inode = NULL;
	} else if (is_bad_inode(inode)) {
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
		iput(inode);
		inode = NULL;
	}
	return inode;
}

/* replays a single extent in 'eb' at 'slot' with 'key' into the
 * subvolume 'root'.  path is released on entry and should be released
 * on exit.
 *
 * extents in the log tree have not been allocated out of the extent
 * tree yet.  So, this completes the allocation, taking a reference
 * as required if the extent already exists or creating a new extent
 * if it isn't in the extent allocation tree yet.
 *
 * The extent is inserted into the file, dropping any existing extents
 * from the file that overlap the new one.
 */
static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
				      struct btrfs_root *root,
				      struct btrfs_path *path,
				      struct extent_buffer *eb, int slot,
				      struct btrfs_key *key)
{
	int found_type;
	u64 extent_end;
	u64 start = key->offset;
552
	u64 nbytes = 0;
553 554 555 556 557 558 559 560
	struct btrfs_file_extent_item *item;
	struct inode *inode = NULL;
	unsigned long size;
	int ret = 0;

	item = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
	found_type = btrfs_file_extent_type(eb, item);

Y
Yan Zheng 已提交
561
	if (found_type == BTRFS_FILE_EXTENT_REG ||
562 563 564 565 566 567 568 569 570 571 572
	    found_type == BTRFS_FILE_EXTENT_PREALLOC) {
		nbytes = btrfs_file_extent_num_bytes(eb, item);
		extent_end = start + nbytes;

		/*
		 * We don't add to the inodes nbytes if we are prealloc or a
		 * hole.
		 */
		if (btrfs_file_extent_disk_bytenr(eb, item) == 0)
			nbytes = 0;
	} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
C
Chris Mason 已提交
573
		size = btrfs_file_extent_inline_len(eb, item);
574
		nbytes = btrfs_file_extent_ram_bytes(eb, item);
575
		extent_end = ALIGN(start + size, root->sectorsize);
576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591
	} else {
		ret = 0;
		goto out;
	}

	inode = read_one_inode(root, key->objectid);
	if (!inode) {
		ret = -EIO;
		goto out;
	}

	/*
	 * first check to see if we already have this extent in the
	 * file.  This must be done before the btrfs_drop_extents run
	 * so we don't try to drop this extent.
	 */
L
Li Zefan 已提交
592
	ret = btrfs_lookup_file_extent(trans, root, path, btrfs_ino(inode),
593 594
				       start, 0);

Y
Yan Zheng 已提交
595 596 597
	if (ret == 0 &&
	    (found_type == BTRFS_FILE_EXTENT_REG ||
	     found_type == BTRFS_FILE_EXTENT_PREALLOC)) {
598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616
		struct btrfs_file_extent_item cmp1;
		struct btrfs_file_extent_item cmp2;
		struct btrfs_file_extent_item *existing;
		struct extent_buffer *leaf;

		leaf = path->nodes[0];
		existing = btrfs_item_ptr(leaf, path->slots[0],
					  struct btrfs_file_extent_item);

		read_extent_buffer(eb, &cmp1, (unsigned long)item,
				   sizeof(cmp1));
		read_extent_buffer(leaf, &cmp2, (unsigned long)existing,
				   sizeof(cmp2));

		/*
		 * we already have a pointer to this exact extent,
		 * we don't have to do anything
		 */
		if (memcmp(&cmp1, &cmp2, sizeof(cmp1)) == 0) {
617
			btrfs_release_path(path);
618 619 620
			goto out;
		}
	}
621
	btrfs_release_path(path);
622 623

	/* drop any overlapping extents */
624
	ret = btrfs_drop_extents(trans, root, inode, start, extent_end, 1);
625 626
	if (ret)
		goto out;
627

Y
Yan Zheng 已提交
628 629
	if (found_type == BTRFS_FILE_EXTENT_REG ||
	    found_type == BTRFS_FILE_EXTENT_PREALLOC) {
630
		u64 offset;
Y
Yan Zheng 已提交
631 632 633 634 635
		unsigned long dest_offset;
		struct btrfs_key ins;

		ret = btrfs_insert_empty_item(trans, root, path, key,
					      sizeof(*item));
636 637
		if (ret)
			goto out;
Y
Yan Zheng 已提交
638 639 640 641 642 643 644 645
		dest_offset = btrfs_item_ptr_offset(path->nodes[0],
						    path->slots[0]);
		copy_extent_buffer(path->nodes[0], eb, dest_offset,
				(unsigned long)item,  sizeof(*item));

		ins.objectid = btrfs_file_extent_disk_bytenr(eb, item);
		ins.offset = btrfs_file_extent_disk_num_bytes(eb, item);
		ins.type = BTRFS_EXTENT_ITEM_KEY;
646
		offset = key->offset - btrfs_file_extent_offset(eb, item);
Y
Yan Zheng 已提交
647 648 649 650 651 652 653 654 655 656 657 658 659 660

		if (ins.objectid > 0) {
			u64 csum_start;
			u64 csum_end;
			LIST_HEAD(ordered_sums);
			/*
			 * is this extent already allocated in the extent
			 * allocation tree?  If so, just add a reference
			 */
			ret = btrfs_lookup_extent(root, ins.objectid,
						ins.offset);
			if (ret == 0) {
				ret = btrfs_inc_extent_ref(trans, root,
						ins.objectid, ins.offset,
661
						0, root->root_key.objectid,
A
Arne Jansen 已提交
662
						key->objectid, offset, 0);
663 664
				if (ret)
					goto out;
Y
Yan Zheng 已提交
665 666 667 668 669
			} else {
				/*
				 * insert the extent pointer in the extent
				 * allocation tree
				 */
670 671 672
				ret = btrfs_alloc_logged_file_extent(trans,
						root, root->root_key.objectid,
						key->objectid, offset, &ins);
673 674
				if (ret)
					goto out;
Y
Yan Zheng 已提交
675
			}
676
			btrfs_release_path(path);
Y
Yan Zheng 已提交
677 678 679 680 681 682 683 684 685 686 687 688 689

			if (btrfs_file_extent_compression(eb, item)) {
				csum_start = ins.objectid;
				csum_end = csum_start + ins.offset;
			} else {
				csum_start = ins.objectid +
					btrfs_file_extent_offset(eb, item);
				csum_end = csum_start +
					btrfs_file_extent_num_bytes(eb, item);
			}

			ret = btrfs_lookup_csums_range(root->log_root,
						csum_start, csum_end - 1,
A
Arne Jansen 已提交
690
						&ordered_sums, 0);
691 692
			if (ret)
				goto out;
Y
Yan Zheng 已提交
693 694 695 696 697
			while (!list_empty(&ordered_sums)) {
				struct btrfs_ordered_sum *sums;
				sums = list_entry(ordered_sums.next,
						struct btrfs_ordered_sum,
						list);
698 699
				if (!ret)
					ret = btrfs_csum_file_blocks(trans,
Y
Yan Zheng 已提交
700 701 702 703 704
						root->fs_info->csum_root,
						sums);
				list_del(&sums->list);
				kfree(sums);
			}
705 706
			if (ret)
				goto out;
Y
Yan Zheng 已提交
707
		} else {
708
			btrfs_release_path(path);
Y
Yan Zheng 已提交
709 710 711 712
		}
	} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
		/* inline extents are easy, we just overwrite them */
		ret = overwrite_item(trans, root, path, eb, slot, key);
713 714
		if (ret)
			goto out;
Y
Yan Zheng 已提交
715
	}
716

717
	inode_add_bytes(inode, nbytes);
718
	ret = btrfs_update_inode(trans, root, inode);
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
out:
	if (inode)
		iput(inode);
	return ret;
}

/*
 * when cleaning up conflicts between the directory names in the
 * subvolume, directory names in the log and directory names in the
 * inode back references, we may have to unlink inodes from directories.
 *
 * This is a helper function to do the unlink of a specific directory
 * item
 */
static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans,
				      struct btrfs_root *root,
				      struct btrfs_path *path,
				      struct inode *dir,
				      struct btrfs_dir_item *di)
{
	struct inode *inode;
	char *name;
	int name_len;
	struct extent_buffer *leaf;
	struct btrfs_key location;
	int ret;

	leaf = path->nodes[0];

	btrfs_dir_item_key_to_cpu(leaf, di, &location);
	name_len = btrfs_dir_name_len(leaf, di);
	name = kmalloc(name_len, GFP_NOFS);
751 752 753
	if (!name)
		return -ENOMEM;

754
	read_extent_buffer(leaf, name, (unsigned long)(di + 1), name_len);
755
	btrfs_release_path(path);
756 757

	inode = read_one_inode(root, location.objectid);
758
	if (!inode) {
759 760
		ret = -EIO;
		goto out;
761
	}
762

763
	ret = link_to_fixup_dir(trans, root, path, location.objectid);
764 765
	if (ret)
		goto out;
766

767
	ret = btrfs_unlink_inode(trans, root, dir, inode, name, name_len);
768 769
	if (ret)
		goto out;
770 771
	else
		ret = btrfs_run_delayed_items(trans, root);
772
out:
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
	kfree(name);
	iput(inode);
	return ret;
}

/*
 * helper function to see if a given name and sequence number found
 * in an inode back reference are already in a directory and correctly
 * point to this inode
 */
static noinline int inode_in_dir(struct btrfs_root *root,
				 struct btrfs_path *path,
				 u64 dirid, u64 objectid, u64 index,
				 const char *name, int name_len)
{
	struct btrfs_dir_item *di;
	struct btrfs_key location;
	int match = 0;

	di = btrfs_lookup_dir_index_item(NULL, root, path, dirid,
					 index, name, name_len, 0);
	if (di && !IS_ERR(di)) {
		btrfs_dir_item_key_to_cpu(path->nodes[0], di, &location);
		if (location.objectid != objectid)
			goto out;
	} else
		goto out;
800
	btrfs_release_path(path);
801 802 803 804 805 806 807 808 809 810

	di = btrfs_lookup_dir_item(NULL, root, path, dirid, name, name_len, 0);
	if (di && !IS_ERR(di)) {
		btrfs_dir_item_key_to_cpu(path->nodes[0], di, &location);
		if (location.objectid != objectid)
			goto out;
	} else
		goto out;
	match = 1;
out:
811
	btrfs_release_path(path);
812 813 814 815 816 817 818 819 820 821 822 823 824 825 826
	return match;
}

/*
 * helper function to check a log tree for a named back reference in
 * an inode.  This is used to decide if a back reference that is
 * found in the subvolume conflicts with what we find in the log.
 *
 * inode backreferences may have multiple refs in a single item,
 * during replay we process one reference at a time, and we don't
 * want to delete valid links to a file from the subvolume if that
 * link is also in the log.
 */
static noinline int backref_in_log(struct btrfs_root *log,
				   struct btrfs_key *key,
M
Mark Fasheh 已提交
827
				   u64 ref_objectid,
828 829 830 831 832 833 834 835 836 837 838 839 840
				   char *name, int namelen)
{
	struct btrfs_path *path;
	struct btrfs_inode_ref *ref;
	unsigned long ptr;
	unsigned long ptr_end;
	unsigned long name_ptr;
	int found_name_len;
	int item_size;
	int ret;
	int match = 0;

	path = btrfs_alloc_path();
841 842 843
	if (!path)
		return -ENOMEM;

844 845 846 847 848
	ret = btrfs_search_slot(NULL, log, key, path, 0, 0);
	if (ret != 0)
		goto out;

	ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]);
M
Mark Fasheh 已提交
849 850 851 852 853 854 855 856 857 858

	if (key->type == BTRFS_INODE_EXTREF_KEY) {
		if (btrfs_find_name_in_ext_backref(path, ref_objectid,
						   name, namelen, NULL))
			match = 1;

		goto out;
	}

	item_size = btrfs_item_size_nr(path->nodes[0], path->slots[0]);
859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878
	ptr_end = ptr + item_size;
	while (ptr < ptr_end) {
		ref = (struct btrfs_inode_ref *)ptr;
		found_name_len = btrfs_inode_ref_name_len(path->nodes[0], ref);
		if (found_name_len == namelen) {
			name_ptr = (unsigned long)(ref + 1);
			ret = memcmp_extent_buffer(path->nodes[0], name,
						   name_ptr, namelen);
			if (ret == 0) {
				match = 1;
				goto out;
			}
		}
		ptr = (unsigned long)(ref + 1) + found_name_len;
	}
out:
	btrfs_free_path(path);
	return match;
}

879
static inline int __add_inode_ref(struct btrfs_trans_handle *trans,
880 881
				  struct btrfs_root *root,
				  struct btrfs_path *path,
882 883 884
				  struct btrfs_root *log_root,
				  struct inode *dir, struct inode *inode,
				  struct extent_buffer *eb,
M
Mark Fasheh 已提交
885 886 887
				  u64 inode_objectid, u64 parent_objectid,
				  u64 ref_index, char *name, int namelen,
				  int *search_done)
888
{
L
liubo 已提交
889
	int ret;
M
Mark Fasheh 已提交
890 891 892
	char *victim_name;
	int victim_name_len;
	struct extent_buffer *leaf;
893
	struct btrfs_dir_item *di;
M
Mark Fasheh 已提交
894 895
	struct btrfs_key search_key;
	struct btrfs_inode_extref *extref;
896

M
Mark Fasheh 已提交
897 898 899 900 901 902
again:
	/* Search old style refs */
	search_key.objectid = inode_objectid;
	search_key.type = BTRFS_INODE_REF_KEY;
	search_key.offset = parent_objectid;
	ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0);
903 904 905 906
	if (ret == 0) {
		struct btrfs_inode_ref *victim_ref;
		unsigned long ptr;
		unsigned long ptr_end;
M
Mark Fasheh 已提交
907 908

		leaf = path->nodes[0];
909 910 911 912

		/* are we trying to overwrite a back ref for the root directory
		 * if so, just jump out, we're done
		 */
M
Mark Fasheh 已提交
913
		if (search_key.objectid == search_key.offset)
914
			return 1;
915 916 917 918 919 920 921

		/* check all the names in this back reference to see
		 * if they are in the log.  if so, we allow them to stay
		 * otherwise they must be unlinked as a conflict
		 */
		ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
		ptr_end = ptr + btrfs_item_size_nr(leaf, path->slots[0]);
C
Chris Mason 已提交
922
		while (ptr < ptr_end) {
923 924 925 926
			victim_ref = (struct btrfs_inode_ref *)ptr;
			victim_name_len = btrfs_inode_ref_name_len(leaf,
								   victim_ref);
			victim_name = kmalloc(victim_name_len, GFP_NOFS);
927 928
			if (!victim_name)
				return -ENOMEM;
929 930 931 932 933

			read_extent_buffer(leaf, victim_name,
					   (unsigned long)(victim_ref + 1),
					   victim_name_len);

M
Mark Fasheh 已提交
934 935 936
			if (!backref_in_log(log_root, &search_key,
					    parent_objectid,
					    victim_name,
937
					    victim_name_len)) {
Z
Zach Brown 已提交
938
				inc_nlink(inode);
939
				btrfs_release_path(path);
940

941 942 943
				ret = btrfs_unlink_inode(trans, root, dir,
							 inode, victim_name,
							 victim_name_len);
M
Mark Fasheh 已提交
944
				kfree(victim_name);
945 946
				if (ret)
					return ret;
947 948 949
				ret = btrfs_run_delayed_items(trans, root);
				if (ret)
					return ret;
M
Mark Fasheh 已提交
950 951
				*search_done = 1;
				goto again;
952 953
			}
			kfree(victim_name);
M
Mark Fasheh 已提交
954

955 956 957
			ptr = (unsigned long)(victim_ref + 1) + victim_name_len;
		}

958 959 960 961
		/*
		 * NOTE: we have searched root tree and checked the
		 * coresponding ref, it does not need to check again.
		 */
962
		*search_done = 1;
963
	}
964
	btrfs_release_path(path);
965

M
Mark Fasheh 已提交
966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989
	/* Same search but for extended refs */
	extref = btrfs_lookup_inode_extref(NULL, root, path, name, namelen,
					   inode_objectid, parent_objectid, 0,
					   0);
	if (!IS_ERR_OR_NULL(extref)) {
		u32 item_size;
		u32 cur_offset = 0;
		unsigned long base;
		struct inode *victim_parent;

		leaf = path->nodes[0];

		item_size = btrfs_item_size_nr(leaf, path->slots[0]);
		base = btrfs_item_ptr_offset(leaf, path->slots[0]);

		while (cur_offset < item_size) {
			extref = (struct btrfs_inode_extref *)base + cur_offset;

			victim_name_len = btrfs_inode_extref_name_len(leaf, extref);

			if (btrfs_inode_extref_parent(leaf, extref) != parent_objectid)
				goto next;

			victim_name = kmalloc(victim_name_len, GFP_NOFS);
990 991
			if (!victim_name)
				return -ENOMEM;
M
Mark Fasheh 已提交
992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007
			read_extent_buffer(leaf, victim_name, (unsigned long)&extref->name,
					   victim_name_len);

			search_key.objectid = inode_objectid;
			search_key.type = BTRFS_INODE_EXTREF_KEY;
			search_key.offset = btrfs_extref_hash(parent_objectid,
							      victim_name,
							      victim_name_len);
			ret = 0;
			if (!backref_in_log(log_root, &search_key,
					    parent_objectid, victim_name,
					    victim_name_len)) {
				ret = -ENOENT;
				victim_parent = read_one_inode(root,
							       parent_objectid);
				if (victim_parent) {
Z
Zach Brown 已提交
1008
					inc_nlink(inode);
M
Mark Fasheh 已提交
1009 1010 1011 1012 1013 1014 1015
					btrfs_release_path(path);

					ret = btrfs_unlink_inode(trans, root,
								 victim_parent,
								 inode,
								 victim_name,
								 victim_name_len);
1016 1017 1018
					if (!ret)
						ret = btrfs_run_delayed_items(
								  trans, root);
M
Mark Fasheh 已提交
1019 1020 1021
				}
				iput(victim_parent);
				kfree(victim_name);
1022 1023
				if (ret)
					return ret;
M
Mark Fasheh 已提交
1024 1025 1026 1027
				*search_done = 1;
				goto again;
			}
			kfree(victim_name);
1028 1029
			if (ret)
				return ret;
M
Mark Fasheh 已提交
1030 1031 1032 1033 1034 1035 1036
next:
			cur_offset += victim_name_len + sizeof(*extref);
		}
		*search_done = 1;
	}
	btrfs_release_path(path);

L
liubo 已提交
1037 1038
	/* look for a conflicting sequence number */
	di = btrfs_lookup_dir_index_item(trans, root, path, btrfs_ino(dir),
M
Mark Fasheh 已提交
1039
					 ref_index, name, namelen, 0);
L
liubo 已提交
1040 1041
	if (di && !IS_ERR(di)) {
		ret = drop_one_dir_item(trans, root, path, dir, di);
1042 1043
		if (ret)
			return ret;
L
liubo 已提交
1044 1045 1046 1047 1048 1049 1050 1051
	}
	btrfs_release_path(path);

	/* look for a conflicing name */
	di = btrfs_lookup_dir_item(trans, root, path, btrfs_ino(dir),
				   name, namelen, 0);
	if (di && !IS_ERR(di)) {
		ret = drop_one_dir_item(trans, root, path, dir, di);
1052 1053
		if (ret)
			return ret;
L
liubo 已提交
1054 1055 1056
	}
	btrfs_release_path(path);

1057 1058
	return 0;
}
1059

M
Mark Fasheh 已提交
1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101
static int extref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr,
			     u32 *namelen, char **name, u64 *index,
			     u64 *parent_objectid)
{
	struct btrfs_inode_extref *extref;

	extref = (struct btrfs_inode_extref *)ref_ptr;

	*namelen = btrfs_inode_extref_name_len(eb, extref);
	*name = kmalloc(*namelen, GFP_NOFS);
	if (*name == NULL)
		return -ENOMEM;

	read_extent_buffer(eb, *name, (unsigned long)&extref->name,
			   *namelen);

	*index = btrfs_inode_extref_index(eb, extref);
	if (parent_objectid)
		*parent_objectid = btrfs_inode_extref_parent(eb, extref);

	return 0;
}

static int ref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr,
			  u32 *namelen, char **name, u64 *index)
{
	struct btrfs_inode_ref *ref;

	ref = (struct btrfs_inode_ref *)ref_ptr;

	*namelen = btrfs_inode_ref_name_len(eb, ref);
	*name = kmalloc(*namelen, GFP_NOFS);
	if (*name == NULL)
		return -ENOMEM;

	read_extent_buffer(eb, *name, (unsigned long)(ref + 1), *namelen);

	*index = btrfs_inode_ref_index(eb, ref);

	return 0;
}

1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114
/*
 * replay one inode back reference item found in the log tree.
 * eb, slot and key refer to the buffer and key found in the log tree.
 * root is the destination we are replaying into, and path is for temp
 * use by this function.  (it should be released on return).
 */
static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
				  struct btrfs_root *root,
				  struct btrfs_root *log,
				  struct btrfs_path *path,
				  struct extent_buffer *eb, int slot,
				  struct btrfs_key *key)
{
1115 1116
	struct inode *dir = NULL;
	struct inode *inode = NULL;
1117 1118
	unsigned long ref_ptr;
	unsigned long ref_end;
1119
	char *name = NULL;
1120 1121 1122
	int namelen;
	int ret;
	int search_done = 0;
M
Mark Fasheh 已提交
1123 1124 1125
	int log_ref_ver = 0;
	u64 parent_objectid;
	u64 inode_objectid;
1126
	u64 ref_index = 0;
M
Mark Fasheh 已提交
1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143
	int ref_struct_size;

	ref_ptr = btrfs_item_ptr_offset(eb, slot);
	ref_end = ref_ptr + btrfs_item_size_nr(eb, slot);

	if (key->type == BTRFS_INODE_EXTREF_KEY) {
		struct btrfs_inode_extref *r;

		ref_struct_size = sizeof(struct btrfs_inode_extref);
		log_ref_ver = 1;
		r = (struct btrfs_inode_extref *)ref_ptr;
		parent_objectid = btrfs_inode_extref_parent(eb, r);
	} else {
		ref_struct_size = sizeof(struct btrfs_inode_ref);
		parent_objectid = key->offset;
	}
	inode_objectid = key->objectid;
1144

1145 1146 1147 1148 1149 1150
	/*
	 * it is possible that we didn't log all the parent directories
	 * for a given inode.  If we don't find the dir, just don't
	 * copy the back ref in.  The link count fixup code will take
	 * care of the rest
	 */
M
Mark Fasheh 已提交
1151
	dir = read_one_inode(root, parent_objectid);
1152 1153 1154 1155
	if (!dir) {
		ret = -ENOENT;
		goto out;
	}
1156

M
Mark Fasheh 已提交
1157
	inode = read_one_inode(root, inode_objectid);
1158
	if (!inode) {
1159 1160
		ret = -EIO;
		goto out;
1161 1162 1163
	}

	while (ref_ptr < ref_end) {
M
Mark Fasheh 已提交
1164 1165 1166 1167 1168 1169 1170 1171 1172
		if (log_ref_ver) {
			ret = extref_get_fields(eb, ref_ptr, &namelen, &name,
						&ref_index, &parent_objectid);
			/*
			 * parent object can change from one array
			 * item to another.
			 */
			if (!dir)
				dir = read_one_inode(root, parent_objectid);
1173 1174 1175 1176
			if (!dir) {
				ret = -ENOENT;
				goto out;
			}
M
Mark Fasheh 已提交
1177 1178 1179 1180 1181
		} else {
			ret = ref_get_fields(eb, ref_ptr, &namelen, &name,
					     &ref_index);
		}
		if (ret)
1182
			goto out;
1183 1184 1185

		/* if we already have a perfect match, we're done */
		if (!inode_in_dir(root, path, btrfs_ino(dir), btrfs_ino(inode),
M
Mark Fasheh 已提交
1186
				  ref_index, name, namelen)) {
1187 1188 1189 1190 1191 1192 1193 1194 1195 1196
			/*
			 * look for a conflicting back reference in the
			 * metadata. if we find one we have to unlink that name
			 * of the file before we add our new link.  Later on, we
			 * overwrite any existing back reference, and we don't
			 * want to create dangling pointers in the directory.
			 */

			if (!search_done) {
				ret = __add_inode_ref(trans, root, path, log,
M
Mark Fasheh 已提交
1197 1198 1199 1200
						      dir, inode, eb,
						      inode_objectid,
						      parent_objectid,
						      ref_index, name, namelen,
1201
						      &search_done);
1202 1203 1204
				if (ret) {
					if (ret == 1)
						ret = 0;
1205 1206
					goto out;
				}
1207 1208 1209 1210
			}

			/* insert our name */
			ret = btrfs_add_link(trans, dir, inode, name, namelen,
M
Mark Fasheh 已提交
1211
					     0, ref_index);
1212 1213
			if (ret)
				goto out;
1214 1215 1216 1217

			btrfs_update_inode(trans, root, inode);
		}

M
Mark Fasheh 已提交
1218
		ref_ptr = (unsigned long)(ref_ptr + ref_struct_size) + namelen;
1219
		kfree(name);
1220
		name = NULL;
M
Mark Fasheh 已提交
1221 1222 1223 1224
		if (log_ref_ver) {
			iput(dir);
			dir = NULL;
		}
1225
	}
1226 1227 1228

	/* finally write the back reference in the inode */
	ret = overwrite_item(trans, root, path, eb, slot, key);
1229
out:
1230
	btrfs_release_path(path);
1231
	kfree(name);
1232 1233
	iput(dir);
	iput(inode);
1234
	return ret;
1235 1236
}

1237 1238 1239 1240 1241 1242 1243 1244 1245 1246
static int insert_orphan_item(struct btrfs_trans_handle *trans,
			      struct btrfs_root *root, u64 offset)
{
	int ret;
	ret = btrfs_find_orphan_item(root, offset);
	if (ret > 0)
		ret = btrfs_insert_orphan_item(trans, root, offset);
	return ret;
}

M
Mark Fasheh 已提交
1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265
static int count_inode_extrefs(struct btrfs_root *root,
			       struct inode *inode, struct btrfs_path *path)
{
	int ret = 0;
	int name_len;
	unsigned int nlink = 0;
	u32 item_size;
	u32 cur_offset = 0;
	u64 inode_objectid = btrfs_ino(inode);
	u64 offset = 0;
	unsigned long ptr;
	struct btrfs_inode_extref *extref;
	struct extent_buffer *leaf;

	while (1) {
		ret = btrfs_find_one_extref(root, inode_objectid, offset, path,
					    &extref, &offset);
		if (ret)
			break;
1266

M
Mark Fasheh 已提交
1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291
		leaf = path->nodes[0];
		item_size = btrfs_item_size_nr(leaf, path->slots[0]);
		ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);

		while (cur_offset < item_size) {
			extref = (struct btrfs_inode_extref *) (ptr + cur_offset);
			name_len = btrfs_inode_extref_name_len(leaf, extref);

			nlink++;

			cur_offset += name_len + sizeof(*extref);
		}

		offset++;
		btrfs_release_path(path);
	}
	btrfs_release_path(path);

	if (ret < 0)
		return ret;
	return nlink;
}

static int count_inode_refs(struct btrfs_root *root,
			       struct inode *inode, struct btrfs_path *path)
1292 1293 1294
{
	int ret;
	struct btrfs_key key;
M
Mark Fasheh 已提交
1295
	unsigned int nlink = 0;
1296 1297 1298
	unsigned long ptr;
	unsigned long ptr_end;
	int name_len;
L
Li Zefan 已提交
1299
	u64 ino = btrfs_ino(inode);
1300

L
Li Zefan 已提交
1301
	key.objectid = ino;
1302 1303 1304
	key.type = BTRFS_INODE_REF_KEY;
	key.offset = (u64)-1;

C
Chris Mason 已提交
1305
	while (1) {
1306 1307 1308 1309 1310 1311 1312 1313
		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
		if (ret < 0)
			break;
		if (ret > 0) {
			if (path->slots[0] == 0)
				break;
			path->slots[0]--;
		}
1314
process_slot:
1315 1316
		btrfs_item_key_to_cpu(path->nodes[0], &key,
				      path->slots[0]);
L
Li Zefan 已提交
1317
		if (key.objectid != ino ||
1318 1319 1320 1321 1322
		    key.type != BTRFS_INODE_REF_KEY)
			break;
		ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]);
		ptr_end = ptr + btrfs_item_size_nr(path->nodes[0],
						   path->slots[0]);
C
Chris Mason 已提交
1323
		while (ptr < ptr_end) {
1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334
			struct btrfs_inode_ref *ref;

			ref = (struct btrfs_inode_ref *)ptr;
			name_len = btrfs_inode_ref_name_len(path->nodes[0],
							    ref);
			ptr = (unsigned long)(ref + 1) + name_len;
			nlink++;
		}

		if (key.offset == 0)
			break;
1335 1336 1337 1338
		if (path->slots[0] > 0) {
			path->slots[0]--;
			goto process_slot;
		}
1339
		key.offset--;
1340
		btrfs_release_path(path);
1341
	}
1342
	btrfs_release_path(path);
M
Mark Fasheh 已提交
1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386

	return nlink;
}

/*
 * There are a few corners where the link count of the file can't
 * be properly maintained during replay.  So, instead of adding
 * lots of complexity to the log code, we just scan the backrefs
 * for any file that has been through replay.
 *
 * The scan will update the link count on the inode to reflect the
 * number of back refs found.  If it goes down to zero, the iput
 * will free the inode.
 */
static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans,
					   struct btrfs_root *root,
					   struct inode *inode)
{
	struct btrfs_path *path;
	int ret;
	u64 nlink = 0;
	u64 ino = btrfs_ino(inode);

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	ret = count_inode_refs(root, inode, path);
	if (ret < 0)
		goto out;

	nlink = ret;

	ret = count_inode_extrefs(root, inode, path);
	if (ret == -ENOENT)
		ret = 0;

	if (ret < 0)
		goto out;

	nlink += ret;

	ret = 0;

1387
	if (nlink != inode->i_nlink) {
M
Miklos Szeredi 已提交
1388
		set_nlink(inode, nlink);
1389 1390
		btrfs_update_inode(trans, root, inode);
	}
1391
	BTRFS_I(inode)->index_cnt = (u64)-1;
1392

1393 1394 1395
	if (inode->i_nlink == 0) {
		if (S_ISDIR(inode->i_mode)) {
			ret = replay_dir_deletes(trans, root, NULL, path,
L
Li Zefan 已提交
1396
						 ino, 1);
1397 1398
			if (ret)
				goto out;
1399
		}
L
Li Zefan 已提交
1400
		ret = insert_orphan_item(trans, root, ino);
1401 1402
	}

M
Mark Fasheh 已提交
1403 1404 1405
out:
	btrfs_free_path(path);
	return ret;
1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418
}

static noinline int fixup_inode_link_counts(struct btrfs_trans_handle *trans,
					    struct btrfs_root *root,
					    struct btrfs_path *path)
{
	int ret;
	struct btrfs_key key;
	struct inode *inode;

	key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID;
	key.type = BTRFS_ORPHAN_ITEM_KEY;
	key.offset = (u64)-1;
C
Chris Mason 已提交
1419
	while (1) {
1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435
		ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
		if (ret < 0)
			break;

		if (ret == 1) {
			if (path->slots[0] == 0)
				break;
			path->slots[0]--;
		}

		btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
		if (key.objectid != BTRFS_TREE_LOG_FIXUP_OBJECTID ||
		    key.type != BTRFS_ORPHAN_ITEM_KEY)
			break;

		ret = btrfs_del_item(trans, root, path);
1436 1437
		if (ret)
			goto out;
1438

1439
		btrfs_release_path(path);
1440
		inode = read_one_inode(root, key.offset);
1441 1442
		if (!inode)
			return -EIO;
1443 1444 1445

		ret = fixup_inode_link_count(trans, root, inode);
		iput(inode);
1446 1447
		if (ret)
			goto out;
1448

1449 1450 1451 1452 1453 1454
		/*
		 * fixup on a directory may create new entries,
		 * make sure we always look for the highset possible
		 * offset
		 */
		key.offset = (u64)-1;
1455
	}
1456 1457
	ret = 0;
out:
1458
	btrfs_release_path(path);
1459
	return ret;
1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477
}


/*
 * record a given inode in the fixup dir so we can check its link
 * count when replay is done.  The link count is incremented here
 * so the inode won't go away until we check it
 */
static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans,
				      struct btrfs_root *root,
				      struct btrfs_path *path,
				      u64 objectid)
{
	struct btrfs_key key;
	int ret = 0;
	struct inode *inode;

	inode = read_one_inode(root, objectid);
1478 1479
	if (!inode)
		return -EIO;
1480 1481 1482 1483 1484 1485 1486

	key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID;
	btrfs_set_key_type(&key, BTRFS_ORPHAN_ITEM_KEY);
	key.offset = objectid;

	ret = btrfs_insert_empty_item(trans, root, path, &key, 0);

1487
	btrfs_release_path(path);
1488
	if (ret == 0) {
1489 1490 1491
		if (!inode->i_nlink)
			set_nlink(inode, 1);
		else
Z
Zach Brown 已提交
1492
			inc_nlink(inode);
1493
		ret = btrfs_update_inode(trans, root, inode);
1494 1495 1496
	} else if (ret == -EEXIST) {
		ret = 0;
	} else {
1497
		BUG(); /* Logic Error */
1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528
	}
	iput(inode);

	return ret;
}

/*
 * when replaying the log for a directory, we only insert names
 * for inodes that actually exist.  This means an fsync on a directory
 * does not implicitly fsync all the new files in it
 */
static noinline int insert_one_name(struct btrfs_trans_handle *trans,
				    struct btrfs_root *root,
				    struct btrfs_path *path,
				    u64 dirid, u64 index,
				    char *name, int name_len, u8 type,
				    struct btrfs_key *location)
{
	struct inode *inode;
	struct inode *dir;
	int ret;

	inode = read_one_inode(root, location->objectid);
	if (!inode)
		return -ENOENT;

	dir = read_one_inode(root, dirid);
	if (!dir) {
		iput(inode);
		return -EIO;
	}
1529

1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565
	ret = btrfs_add_link(trans, dir, inode, name, name_len, 1, index);

	/* FIXME, put inode into FIXUP list */

	iput(inode);
	iput(dir);
	return ret;
}

/*
 * take a single entry in a log directory item and replay it into
 * the subvolume.
 *
 * if a conflicting item exists in the subdirectory already,
 * the inode it points to is unlinked and put into the link count
 * fix up tree.
 *
 * If a name from the log points to a file or directory that does
 * not exist in the FS, it is skipped.  fsyncs on directories
 * do not force down inodes inside that directory, just changes to the
 * names or unlinks in a directory.
 */
static noinline int replay_one_name(struct btrfs_trans_handle *trans,
				    struct btrfs_root *root,
				    struct btrfs_path *path,
				    struct extent_buffer *eb,
				    struct btrfs_dir_item *di,
				    struct btrfs_key *key)
{
	char *name;
	int name_len;
	struct btrfs_dir_item *dst_di;
	struct btrfs_key found_key;
	struct btrfs_key log_key;
	struct inode *dir;
	u8 log_type;
C
Chris Mason 已提交
1566
	int exists;
1567
	int ret = 0;
1568
	bool update_size = (key->type == BTRFS_DIR_INDEX_KEY);
1569 1570

	dir = read_one_inode(root, key->objectid);
1571 1572
	if (!dir)
		return -EIO;
1573 1574 1575

	name_len = btrfs_dir_name_len(eb, di);
	name = kmalloc(name_len, GFP_NOFS);
1576 1577 1578 1579
	if (!name) {
		ret = -ENOMEM;
		goto out;
	}
1580

1581 1582 1583 1584 1585
	log_type = btrfs_dir_type(eb, di);
	read_extent_buffer(eb, name, (unsigned long)(di + 1),
		   name_len);

	btrfs_dir_item_key_to_cpu(eb, di, &log_key);
C
Chris Mason 已提交
1586 1587 1588 1589 1590
	exists = btrfs_lookup_inode(trans, root, path, &log_key, 0);
	if (exists == 0)
		exists = 1;
	else
		exists = 0;
1591
	btrfs_release_path(path);
C
Chris Mason 已提交
1592

1593 1594 1595
	if (key->type == BTRFS_DIR_ITEM_KEY) {
		dst_di = btrfs_lookup_dir_item(trans, root, path, key->objectid,
				       name, name_len, 1);
C
Chris Mason 已提交
1596
	} else if (key->type == BTRFS_DIR_INDEX_KEY) {
1597 1598 1599 1600 1601
		dst_di = btrfs_lookup_dir_index_item(trans, root, path,
						     key->objectid,
						     key->offset, name,
						     name_len, 1);
	} else {
1602 1603 1604
		/* Corruption */
		ret = -EINVAL;
		goto out;
1605
	}
1606
	if (IS_ERR_OR_NULL(dst_di)) {
1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627
		/* we need a sequence number to insert, so we only
		 * do inserts for the BTRFS_DIR_INDEX_KEY types
		 */
		if (key->type != BTRFS_DIR_INDEX_KEY)
			goto out;
		goto insert;
	}

	btrfs_dir_item_key_to_cpu(path->nodes[0], dst_di, &found_key);
	/* the existing item matches the logged item */
	if (found_key.objectid == log_key.objectid &&
	    found_key.type == log_key.type &&
	    found_key.offset == log_key.offset &&
	    btrfs_dir_type(path->nodes[0], dst_di) == log_type) {
		goto out;
	}

	/*
	 * don't drop the conflicting directory entry if the inode
	 * for the new entry doesn't exist
	 */
C
Chris Mason 已提交
1628
	if (!exists)
1629 1630 1631
		goto out;

	ret = drop_one_dir_item(trans, root, path, dir, dst_di);
1632 1633
	if (ret)
		goto out;
1634 1635 1636 1637

	if (key->type == BTRFS_DIR_INDEX_KEY)
		goto insert;
out:
1638
	btrfs_release_path(path);
1639 1640 1641 1642
	if (!ret && update_size) {
		btrfs_i_size_write(dir, dir->i_size + name_len * 2);
		ret = btrfs_update_inode(trans, root, dir);
	}
1643 1644
	kfree(name);
	iput(dir);
1645
	return ret;
1646 1647

insert:
1648
	btrfs_release_path(path);
1649 1650
	ret = insert_one_name(trans, root, path, key->objectid, key->offset,
			      name, name_len, log_type, &log_key);
1651 1652
	if (ret && ret != -ENOENT)
		goto out;
1653
	update_size = false;
1654
	ret = 0;
1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678
	goto out;
}

/*
 * find all the names in a directory item and reconcile them into
 * the subvolume.  Only BTRFS_DIR_ITEM_KEY types will have more than
 * one name in a directory item, but the same code gets used for
 * both directory index types
 */
static noinline int replay_one_dir_item(struct btrfs_trans_handle *trans,
					struct btrfs_root *root,
					struct btrfs_path *path,
					struct extent_buffer *eb, int slot,
					struct btrfs_key *key)
{
	int ret;
	u32 item_size = btrfs_item_size_nr(eb, slot);
	struct btrfs_dir_item *di;
	int name_len;
	unsigned long ptr;
	unsigned long ptr_end;

	ptr = btrfs_item_ptr_offset(eb, slot);
	ptr_end = ptr + item_size;
C
Chris Mason 已提交
1679
	while (ptr < ptr_end) {
1680
		di = (struct btrfs_dir_item *)ptr;
1681 1682
		if (verify_dir_item(root, eb, di))
			return -EIO;
1683 1684
		name_len = btrfs_dir_name_len(eb, di);
		ret = replay_one_name(trans, root, path, eb, di, key);
1685 1686
		if (ret)
			return ret;
1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771
		ptr = (unsigned long)(di + 1);
		ptr += name_len;
	}
	return 0;
}

/*
 * directory replay has two parts.  There are the standard directory
 * items in the log copied from the subvolume, and range items
 * created in the log while the subvolume was logged.
 *
 * The range items tell us which parts of the key space the log
 * is authoritative for.  During replay, if a key in the subvolume
 * directory is in a logged range item, but not actually in the log
 * that means it was deleted from the directory before the fsync
 * and should be removed.
 */
static noinline int find_dir_range(struct btrfs_root *root,
				   struct btrfs_path *path,
				   u64 dirid, int key_type,
				   u64 *start_ret, u64 *end_ret)
{
	struct btrfs_key key;
	u64 found_end;
	struct btrfs_dir_log_item *item;
	int ret;
	int nritems;

	if (*start_ret == (u64)-1)
		return 1;

	key.objectid = dirid;
	key.type = key_type;
	key.offset = *start_ret;

	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	if (ret < 0)
		goto out;
	if (ret > 0) {
		if (path->slots[0] == 0)
			goto out;
		path->slots[0]--;
	}
	if (ret != 0)
		btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);

	if (key.type != key_type || key.objectid != dirid) {
		ret = 1;
		goto next;
	}
	item = btrfs_item_ptr(path->nodes[0], path->slots[0],
			      struct btrfs_dir_log_item);
	found_end = btrfs_dir_log_end(path->nodes[0], item);

	if (*start_ret >= key.offset && *start_ret <= found_end) {
		ret = 0;
		*start_ret = key.offset;
		*end_ret = found_end;
		goto out;
	}
	ret = 1;
next:
	/* check the next slot in the tree to see if it is a valid item */
	nritems = btrfs_header_nritems(path->nodes[0]);
	if (path->slots[0] >= nritems) {
		ret = btrfs_next_leaf(root, path);
		if (ret)
			goto out;
	} else {
		path->slots[0]++;
	}

	btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);

	if (key.type != key_type || key.objectid != dirid) {
		ret = 1;
		goto out;
	}
	item = btrfs_item_ptr(path->nodes[0], path->slots[0],
			      struct btrfs_dir_log_item);
	found_end = btrfs_dir_log_end(path->nodes[0], item);
	*start_ret = key.offset;
	*end_ret = found_end;
	ret = 0;
out:
1772
	btrfs_release_path(path);
1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807
	return ret;
}

/*
 * this looks for a given directory item in the log.  If the directory
 * item is not in the log, the item is removed and the inode it points
 * to is unlinked
 */
static noinline int check_item_in_log(struct btrfs_trans_handle *trans,
				      struct btrfs_root *root,
				      struct btrfs_root *log,
				      struct btrfs_path *path,
				      struct btrfs_path *log_path,
				      struct inode *dir,
				      struct btrfs_key *dir_key)
{
	int ret;
	struct extent_buffer *eb;
	int slot;
	u32 item_size;
	struct btrfs_dir_item *di;
	struct btrfs_dir_item *log_di;
	int name_len;
	unsigned long ptr;
	unsigned long ptr_end;
	char *name;
	struct inode *inode;
	struct btrfs_key location;

again:
	eb = path->nodes[0];
	slot = path->slots[0];
	item_size = btrfs_item_size_nr(eb, slot);
	ptr = btrfs_item_ptr_offset(eb, slot);
	ptr_end = ptr + item_size;
C
Chris Mason 已提交
1808
	while (ptr < ptr_end) {
1809
		di = (struct btrfs_dir_item *)ptr;
1810 1811 1812 1813 1814
		if (verify_dir_item(root, eb, di)) {
			ret = -EIO;
			goto out;
		}

1815 1816 1817 1818 1819 1820 1821 1822 1823
		name_len = btrfs_dir_name_len(eb, di);
		name = kmalloc(name_len, GFP_NOFS);
		if (!name) {
			ret = -ENOMEM;
			goto out;
		}
		read_extent_buffer(eb, name, (unsigned long)(di + 1),
				  name_len);
		log_di = NULL;
1824
		if (log && dir_key->type == BTRFS_DIR_ITEM_KEY) {
1825 1826 1827
			log_di = btrfs_lookup_dir_item(trans, log, log_path,
						       dir_key->objectid,
						       name, name_len, 0);
1828
		} else if (log && dir_key->type == BTRFS_DIR_INDEX_KEY) {
1829 1830 1831 1832 1833 1834
			log_di = btrfs_lookup_dir_index_item(trans, log,
						     log_path,
						     dir_key->objectid,
						     dir_key->offset,
						     name, name_len, 0);
		}
1835
		if (!log_di || (IS_ERR(log_di) && PTR_ERR(log_di) == -ENOENT)) {
1836
			btrfs_dir_item_key_to_cpu(eb, di, &location);
1837 1838
			btrfs_release_path(path);
			btrfs_release_path(log_path);
1839
			inode = read_one_inode(root, location.objectid);
1840 1841 1842 1843
			if (!inode) {
				kfree(name);
				return -EIO;
			}
1844 1845 1846

			ret = link_to_fixup_dir(trans, root,
						path, location.objectid);
1847 1848 1849 1850 1851 1852
			if (ret) {
				kfree(name);
				iput(inode);
				goto out;
			}

Z
Zach Brown 已提交
1853
			inc_nlink(inode);
1854 1855
			ret = btrfs_unlink_inode(trans, root, dir, inode,
						 name, name_len);
1856
			if (!ret)
1857
				ret = btrfs_run_delayed_items(trans, root);
1858 1859
			kfree(name);
			iput(inode);
1860 1861
			if (ret)
				goto out;
1862 1863 1864 1865 1866 1867 1868 1869 1870 1871

			/* there might still be more names under this key
			 * check and repeat if required
			 */
			ret = btrfs_search_slot(NULL, root, dir_key, path,
						0, 0);
			if (ret == 0)
				goto again;
			ret = 0;
			goto out;
1872 1873 1874
		} else if (IS_ERR(log_di)) {
			kfree(name);
			return PTR_ERR(log_di);
1875
		}
1876
		btrfs_release_path(log_path);
1877 1878 1879 1880 1881 1882 1883
		kfree(name);

		ptr = (unsigned long)(di + 1);
		ptr += name_len;
	}
	ret = 0;
out:
1884 1885
	btrfs_release_path(path);
	btrfs_release_path(log_path);
1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902
	return ret;
}

/*
 * deletion replay happens before we copy any new directory items
 * out of the log or out of backreferences from inodes.  It
 * scans the log to find ranges of keys that log is authoritative for,
 * and then scans the directory to find items in those ranges that are
 * not present in the log.
 *
 * Anything we don't find in the log is unlinked and removed from the
 * directory.
 */
static noinline int replay_dir_deletes(struct btrfs_trans_handle *trans,
				       struct btrfs_root *root,
				       struct btrfs_root *log,
				       struct btrfs_path *path,
1903
				       u64 dirid, int del_all)
1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931
{
	u64 range_start;
	u64 range_end;
	int key_type = BTRFS_DIR_LOG_ITEM_KEY;
	int ret = 0;
	struct btrfs_key dir_key;
	struct btrfs_key found_key;
	struct btrfs_path *log_path;
	struct inode *dir;

	dir_key.objectid = dirid;
	dir_key.type = BTRFS_DIR_ITEM_KEY;
	log_path = btrfs_alloc_path();
	if (!log_path)
		return -ENOMEM;

	dir = read_one_inode(root, dirid);
	/* it isn't an error if the inode isn't there, that can happen
	 * because we replay the deletes before we copy in the inode item
	 * from the log
	 */
	if (!dir) {
		btrfs_free_path(log_path);
		return 0;
	}
again:
	range_start = 0;
	range_end = 0;
C
Chris Mason 已提交
1932
	while (1) {
1933 1934 1935 1936 1937 1938 1939 1940
		if (del_all)
			range_end = (u64)-1;
		else {
			ret = find_dir_range(log, path, dirid, key_type,
					     &range_start, &range_end);
			if (ret != 0)
				break;
		}
1941 1942

		dir_key.offset = range_start;
C
Chris Mason 已提交
1943
		while (1) {
1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
			int nritems;
			ret = btrfs_search_slot(NULL, root, &dir_key, path,
						0, 0);
			if (ret < 0)
				goto out;

			nritems = btrfs_header_nritems(path->nodes[0]);
			if (path->slots[0] >= nritems) {
				ret = btrfs_next_leaf(root, path);
				if (ret)
					break;
			}
			btrfs_item_key_to_cpu(path->nodes[0], &found_key,
					      path->slots[0]);
			if (found_key.objectid != dirid ||
			    found_key.type != dir_key.type)
				goto next_type;

			if (found_key.offset > range_end)
				break;

			ret = check_item_in_log(trans, root, log, path,
1966 1967
						log_path, dir,
						&found_key);
1968 1969
			if (ret)
				goto out;
1970 1971 1972 1973
			if (found_key.offset == (u64)-1)
				break;
			dir_key.offset = found_key.offset + 1;
		}
1974
		btrfs_release_path(path);
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984
		if (range_end == (u64)-1)
			break;
		range_start = range_end + 1;
	}

next_type:
	ret = 0;
	if (key_type == BTRFS_DIR_LOG_ITEM_KEY) {
		key_type = BTRFS_DIR_LOG_INDEX_KEY;
		dir_key.type = BTRFS_DIR_INDEX_KEY;
1985
		btrfs_release_path(path);
1986 1987 1988
		goto again;
	}
out:
1989
	btrfs_release_path(path);
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
	btrfs_free_path(log_path);
	iput(dir);
	return ret;
}

/*
 * the process_func used to replay items from the log tree.  This
 * gets called in two different stages.  The first stage just looks
 * for inodes and makes sure they are all copied into the subvolume.
 *
 * The second stage copies all the other item types from the log into
 * the subvolume.  The two stage approach is slower, but gets rid of
 * lots of complexity around inodes referencing other inodes that exist
 * only in the log (references come from either directory items or inode
 * back refs).
 */
static int replay_one_buffer(struct btrfs_root *log, struct extent_buffer *eb,
			     struct walk_control *wc, u64 gen)
{
	int nritems;
	struct btrfs_path *path;
	struct btrfs_root *root = wc->replay_dest;
	struct btrfs_key key;
	int level;
	int i;
	int ret;

2017 2018 2019
	ret = btrfs_read_buffer(eb, gen);
	if (ret)
		return ret;
2020 2021 2022 2023 2024 2025 2026

	level = btrfs_header_level(eb);

	if (level != 0)
		return 0;

	path = btrfs_alloc_path();
2027 2028
	if (!path)
		return -ENOMEM;
2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044

	nritems = btrfs_header_nritems(eb);
	for (i = 0; i < nritems; i++) {
		btrfs_item_key_to_cpu(eb, &key, i);

		/* inode keys are done during the first stage */
		if (key.type == BTRFS_INODE_ITEM_KEY &&
		    wc->stage == LOG_WALK_REPLAY_INODES) {
			struct btrfs_inode_item *inode_item;
			u32 mode;

			inode_item = btrfs_item_ptr(eb, i,
					    struct btrfs_inode_item);
			mode = btrfs_inode_mode(eb, inode_item);
			if (S_ISDIR(mode)) {
				ret = replay_dir_deletes(wc->trans,
2045
					 root, log, path, key.objectid, 0);
2046 2047
				if (ret)
					break;
2048 2049 2050
			}
			ret = overwrite_item(wc->trans, root, path,
					     eb, i, &key);
2051 2052
			if (ret)
				break;
2053

2054 2055 2056
			/* for regular files, make sure corresponding
			 * orhpan item exist. extents past the new EOF
			 * will be truncated later by orphan cleanup.
2057 2058
			 */
			if (S_ISREG(mode)) {
2059 2060
				ret = insert_orphan_item(wc->trans, root,
							 key.objectid);
2061 2062
				if (ret)
					break;
2063
			}
2064

2065 2066
			ret = link_to_fixup_dir(wc->trans, root,
						path, key.objectid);
2067 2068
			if (ret)
				break;
2069
		}
2070 2071 2072 2073 2074 2075 2076 2077 2078

		if (key.type == BTRFS_DIR_INDEX_KEY &&
		    wc->stage == LOG_WALK_REPLAY_DIR_INDEX) {
			ret = replay_one_dir_item(wc->trans, root, path,
						  eb, i, &key);
			if (ret)
				break;
		}

2079 2080 2081 2082 2083 2084 2085
		if (wc->stage < LOG_WALK_REPLAY_ALL)
			continue;

		/* these keys are simply copied */
		if (key.type == BTRFS_XATTR_ITEM_KEY) {
			ret = overwrite_item(wc->trans, root, path,
					     eb, i, &key);
2086 2087
			if (ret)
				break;
2088 2089
		} else if (key.type == BTRFS_INODE_REF_KEY ||
			   key.type == BTRFS_INODE_EXTREF_KEY) {
M
Mark Fasheh 已提交
2090 2091
			ret = add_inode_ref(wc->trans, root, log, path,
					    eb, i, &key);
2092 2093 2094
			if (ret && ret != -ENOENT)
				break;
			ret = 0;
2095 2096 2097
		} else if (key.type == BTRFS_EXTENT_DATA_KEY) {
			ret = replay_one_extent(wc->trans, root, path,
						eb, i, &key);
2098 2099
			if (ret)
				break;
2100
		} else if (key.type == BTRFS_DIR_ITEM_KEY) {
2101 2102
			ret = replay_one_dir_item(wc->trans, root, path,
						  eb, i, &key);
2103 2104
			if (ret)
				break;
2105 2106 2107
		}
	}
	btrfs_free_path(path);
2108
	return ret;
2109 2110
}

C
Chris Mason 已提交
2111
static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127
				   struct btrfs_root *root,
				   struct btrfs_path *path, int *level,
				   struct walk_control *wc)
{
	u64 root_owner;
	u64 bytenr;
	u64 ptr_gen;
	struct extent_buffer *next;
	struct extent_buffer *cur;
	struct extent_buffer *parent;
	u32 blocksize;
	int ret = 0;

	WARN_ON(*level < 0);
	WARN_ON(*level >= BTRFS_MAX_LEVEL);

C
Chris Mason 已提交
2128
	while (*level > 0) {
2129 2130 2131 2132
		WARN_ON(*level < 0);
		WARN_ON(*level >= BTRFS_MAX_LEVEL);
		cur = path->nodes[*level];

2133
		WARN_ON(btrfs_header_level(cur) != *level);
2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146

		if (path->slots[*level] >=
		    btrfs_header_nritems(cur))
			break;

		bytenr = btrfs_node_blockptr(cur, path->slots[*level]);
		ptr_gen = btrfs_node_ptr_generation(cur, path->slots[*level]);
		blocksize = btrfs_level_size(root, *level - 1);

		parent = path->nodes[*level];
		root_owner = btrfs_header_owner(parent);

		next = btrfs_find_create_tree_block(root, bytenr, blocksize);
2147 2148
		if (!next)
			return -ENOMEM;
2149 2150

		if (*level == 1) {
2151
			ret = wc->process_func(root, next, wc, ptr_gen);
2152 2153
			if (ret) {
				free_extent_buffer(next);
2154
				return ret;
2155
			}
2156

2157 2158
			path->slots[*level]++;
			if (wc->free) {
2159 2160 2161 2162 2163
				ret = btrfs_read_buffer(next, ptr_gen);
				if (ret) {
					free_extent_buffer(next);
					return ret;
				}
2164

2165 2166 2167 2168 2169 2170 2171
				if (trans) {
					btrfs_tree_lock(next);
					btrfs_set_lock_blocking(next);
					clean_tree_block(trans, root, next);
					btrfs_wait_tree_block_writeback(next);
					btrfs_tree_unlock(next);
				}
2172 2173 2174

				WARN_ON(root_owner !=
					BTRFS_TREE_LOG_OBJECTID);
2175
				ret = btrfs_free_and_pin_reserved_extent(root,
2176
							 bytenr, blocksize);
2177 2178 2179 2180
				if (ret) {
					free_extent_buffer(next);
					return ret;
				}
2181 2182 2183 2184
			}
			free_extent_buffer(next);
			continue;
		}
2185 2186 2187 2188 2189
		ret = btrfs_read_buffer(next, ptr_gen);
		if (ret) {
			free_extent_buffer(next);
			return ret;
		}
2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201

		WARN_ON(*level <= 0);
		if (path->nodes[*level-1])
			free_extent_buffer(path->nodes[*level-1]);
		path->nodes[*level-1] = next;
		*level = btrfs_header_level(next);
		path->slots[*level] = 0;
		cond_resched();
	}
	WARN_ON(*level < 0);
	WARN_ON(*level >= BTRFS_MAX_LEVEL);

2202
	path->slots[*level] = btrfs_header_nritems(path->nodes[*level]);
2203 2204 2205 2206 2207

	cond_resched();
	return 0;
}

C
Chris Mason 已提交
2208
static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans,
2209 2210 2211 2212 2213 2214 2215 2216 2217
				 struct btrfs_root *root,
				 struct btrfs_path *path, int *level,
				 struct walk_control *wc)
{
	u64 root_owner;
	int i;
	int slot;
	int ret;

C
Chris Mason 已提交
2218
	for (i = *level; i < BTRFS_MAX_LEVEL - 1 && path->nodes[i]; i++) {
2219
		slot = path->slots[i];
2220
		if (slot + 1 < btrfs_header_nritems(path->nodes[i])) {
2221 2222 2223 2224 2225
			path->slots[i]++;
			*level = i;
			WARN_ON(*level == 0);
			return 0;
		} else {
Z
Zheng Yan 已提交
2226 2227 2228 2229 2230 2231 2232
			struct extent_buffer *parent;
			if (path->nodes[*level] == root->node)
				parent = path->nodes[*level];
			else
				parent = path->nodes[*level + 1];

			root_owner = btrfs_header_owner(parent);
2233
			ret = wc->process_func(root, path->nodes[*level], wc,
2234
				 btrfs_header_generation(path->nodes[*level]));
2235 2236 2237
			if (ret)
				return ret;

2238 2239 2240 2241 2242
			if (wc->free) {
				struct extent_buffer *next;

				next = path->nodes[*level];

2243 2244 2245 2246 2247 2248 2249
				if (trans) {
					btrfs_tree_lock(next);
					btrfs_set_lock_blocking(next);
					clean_tree_block(trans, root, next);
					btrfs_wait_tree_block_writeback(next);
					btrfs_tree_unlock(next);
				}
2250 2251

				WARN_ON(root_owner != BTRFS_TREE_LOG_OBJECTID);
2252
				ret = btrfs_free_and_pin_reserved_extent(root,
2253
						path->nodes[*level]->start,
2254
						path->nodes[*level]->len);
2255 2256
				if (ret)
					return ret;
2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280
			}
			free_extent_buffer(path->nodes[*level]);
			path->nodes[*level] = NULL;
			*level = i + 1;
		}
	}
	return 1;
}

/*
 * drop the reference count on the tree rooted at 'snap'.  This traverses
 * the tree freeing any blocks that have a ref count of zero after being
 * decremented.
 */
static int walk_log_tree(struct btrfs_trans_handle *trans,
			 struct btrfs_root *log, struct walk_control *wc)
{
	int ret = 0;
	int wret;
	int level;
	struct btrfs_path *path;
	int orig_level;

	path = btrfs_alloc_path();
T
Tsutomu Itoh 已提交
2281 2282
	if (!path)
		return -ENOMEM;
2283 2284 2285 2286 2287 2288 2289

	level = btrfs_header_level(log->node);
	orig_level = level;
	path->nodes[level] = log->node;
	extent_buffer_get(log->node);
	path->slots[level] = 0;

C
Chris Mason 已提交
2290
	while (1) {
2291 2292 2293
		wret = walk_down_log_tree(trans, log, path, &level, wc);
		if (wret > 0)
			break;
2294
		if (wret < 0) {
2295
			ret = wret;
2296 2297
			goto out;
		}
2298 2299 2300 2301

		wret = walk_up_log_tree(trans, log, path, &level, wc);
		if (wret > 0)
			break;
2302
		if (wret < 0) {
2303
			ret = wret;
2304 2305
			goto out;
		}
2306 2307 2308 2309
	}

	/* was the root node processed? if not, catch it here */
	if (path->nodes[orig_level]) {
2310
		ret = wc->process_func(log, path->nodes[orig_level], wc,
2311
			 btrfs_header_generation(path->nodes[orig_level]));
2312 2313
		if (ret)
			goto out;
2314 2315 2316 2317 2318
		if (wc->free) {
			struct extent_buffer *next;

			next = path->nodes[orig_level];

2319 2320 2321 2322 2323 2324 2325
			if (trans) {
				btrfs_tree_lock(next);
				btrfs_set_lock_blocking(next);
				clean_tree_block(trans, log, next);
				btrfs_wait_tree_block_writeback(next);
				btrfs_tree_unlock(next);
			}
2326 2327 2328

			WARN_ON(log->root_key.objectid !=
				BTRFS_TREE_LOG_OBJECTID);
2329
			ret = btrfs_free_and_pin_reserved_extent(log, next->start,
2330
							 next->len);
2331 2332
			if (ret)
				goto out;
2333 2334 2335
		}
	}

2336
out:
2337 2338 2339 2340
	btrfs_free_path(path);
	return ret;
}

Y
Yan Zheng 已提交
2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360
/*
 * helper function to update the item for a given subvolumes log root
 * in the tree of log roots
 */
static int update_log_root(struct btrfs_trans_handle *trans,
			   struct btrfs_root *log)
{
	int ret;

	if (log->log_transid == 1) {
		/* insert root item on the first sync */
		ret = btrfs_insert_root(trans, log->fs_info->log_root_tree,
				&log->root_key, &log->root_item);
	} else {
		ret = btrfs_update_root(trans, log->fs_info->log_root_tree,
				&log->root_key, &log->root_item);
	}
	return ret;
}

2361 2362
static int wait_log_commit(struct btrfs_trans_handle *trans,
			   struct btrfs_root *root, unsigned long transid)
2363 2364
{
	DEFINE_WAIT(wait);
Y
Yan Zheng 已提交
2365
	int index = transid % 2;
2366

Y
Yan Zheng 已提交
2367 2368 2369 2370 2371
	/*
	 * we only allow two pending log transactions at a time,
	 * so we know that if ours is more than 2 older than the
	 * current transaction, we're done
	 */
2372
	do {
Y
Yan Zheng 已提交
2373 2374 2375
		prepare_to_wait(&root->log_commit_wait[index],
				&wait, TASK_UNINTERRUPTIBLE);
		mutex_unlock(&root->log_mutex);
2376 2377 2378

		if (root->fs_info->last_trans_log_full_commit !=
		    trans->transid && root->log_transid < transid + 2 &&
Y
Yan Zheng 已提交
2379 2380
		    atomic_read(&root->log_commit[index]))
			schedule();
2381

Y
Yan Zheng 已提交
2382 2383
		finish_wait(&root->log_commit_wait[index], &wait);
		mutex_lock(&root->log_mutex);
2384 2385
	} while (root->fs_info->last_trans_log_full_commit !=
		 trans->transid && root->log_transid < transid + 2 &&
Y
Yan Zheng 已提交
2386 2387 2388 2389
		 atomic_read(&root->log_commit[index]));
	return 0;
}

2390 2391
static void wait_for_writer(struct btrfs_trans_handle *trans,
			    struct btrfs_root *root)
Y
Yan Zheng 已提交
2392 2393
{
	DEFINE_WAIT(wait);
2394 2395
	while (root->fs_info->last_trans_log_full_commit !=
	       trans->transid && atomic_read(&root->log_writers)) {
Y
Yan Zheng 已提交
2396 2397 2398
		prepare_to_wait(&root->log_writer_wait,
				&wait, TASK_UNINTERRUPTIBLE);
		mutex_unlock(&root->log_mutex);
2399 2400
		if (root->fs_info->last_trans_log_full_commit !=
		    trans->transid && atomic_read(&root->log_writers))
2401
			schedule();
Y
Yan Zheng 已提交
2402 2403 2404
		mutex_lock(&root->log_mutex);
		finish_wait(&root->log_writer_wait, &wait);
	}
2405 2406 2407 2408 2409
}

/*
 * btrfs_sync_log does sends a given tree log down to the disk and
 * updates the super blocks to record it.  When this call is done,
2410 2411 2412 2413 2414 2415 2416 2417
 * you know that any inodes previously logged are safely on disk only
 * if it returns 0.
 *
 * Any other return value means you need to call btrfs_commit_transaction.
 * Some of the edge cases for fsyncing directories that have had unlinks
 * or renames done in the past mean that sometimes the only safe
 * fsync is to commit the whole FS.  When btrfs_sync_log returns -EAGAIN,
 * that has happened.
2418 2419 2420 2421
 */
int btrfs_sync_log(struct btrfs_trans_handle *trans,
		   struct btrfs_root *root)
{
Y
Yan Zheng 已提交
2422 2423
	int index1;
	int index2;
2424
	int mark;
2425 2426
	int ret;
	struct btrfs_root *log = root->log_root;
Y
Yan Zheng 已提交
2427
	struct btrfs_root *log_root_tree = root->fs_info->log_root_tree;
2428
	unsigned long log_transid = 0;
2429
	struct blk_plug plug;
2430

Y
Yan Zheng 已提交
2431
	mutex_lock(&root->log_mutex);
2432
	log_transid = root->log_transid;
Y
Yan Zheng 已提交
2433 2434
	index1 = root->log_transid % 2;
	if (atomic_read(&root->log_commit[index1])) {
2435
		wait_log_commit(trans, root, root->log_transid);
Y
Yan Zheng 已提交
2436 2437
		mutex_unlock(&root->log_mutex);
		return 0;
2438
	}
Y
Yan Zheng 已提交
2439 2440 2441 2442
	atomic_set(&root->log_commit[index1], 1);

	/* wait for previous tree log sync to complete */
	if (atomic_read(&root->log_commit[(index1 + 1) % 2]))
2443
		wait_log_commit(trans, root, root->log_transid - 1);
2444
	while (1) {
M
Miao Xie 已提交
2445
		int batch = atomic_read(&root->log_batch);
2446 2447
		/* when we're on an ssd, just kick the log commit out */
		if (!btrfs_test_opt(root, SSD) && root->log_multiple_pids) {
2448 2449 2450 2451
			mutex_unlock(&root->log_mutex);
			schedule_timeout_uninterruptible(1);
			mutex_lock(&root->log_mutex);
		}
2452
		wait_for_writer(trans, root);
M
Miao Xie 已提交
2453
		if (batch == atomic_read(&root->log_batch))
2454 2455 2456
			break;
	}

2457 2458 2459
	/* bail out if we need to do a full commit */
	if (root->fs_info->last_trans_log_full_commit == trans->transid) {
		ret = -EAGAIN;
2460
		btrfs_free_logged_extents(log, log_transid);
2461 2462 2463 2464
		mutex_unlock(&root->log_mutex);
		goto out;
	}

2465 2466 2467 2468 2469
	if (log_transid % 2 == 0)
		mark = EXTENT_DIRTY;
	else
		mark = EXTENT_NEW;

2470 2471 2472
	/* we start IO on  all the marked extents here, but we don't actually
	 * wait for them until later.
	 */
2473
	blk_start_plug(&plug);
2474
	ret = btrfs_write_marked_extents(log, &log->dirty_log_pages, mark);
2475
	if (ret) {
2476
		blk_finish_plug(&plug);
2477
		btrfs_abort_transaction(trans, root, ret);
2478
		btrfs_free_logged_extents(log, log_transid);
2479 2480 2481
		mutex_unlock(&root->log_mutex);
		goto out;
	}
Y
Yan Zheng 已提交
2482

2483
	btrfs_set_root_node(&log->root_item, log->node);
Y
Yan Zheng 已提交
2484 2485 2486

	root->log_transid++;
	log->log_transid = root->log_transid;
2487
	root->log_start_pid = 0;
Y
Yan Zheng 已提交
2488 2489
	smp_mb();
	/*
2490 2491 2492
	 * IO has been started, blocks of the log tree have WRITTEN flag set
	 * in their headers. new modifications of the log will be written to
	 * new positions. so it's safe to allow log writers to go in.
Y
Yan Zheng 已提交
2493 2494 2495 2496
	 */
	mutex_unlock(&root->log_mutex);

	mutex_lock(&log_root_tree->log_mutex);
M
Miao Xie 已提交
2497
	atomic_inc(&log_root_tree->log_batch);
Y
Yan Zheng 已提交
2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509
	atomic_inc(&log_root_tree->log_writers);
	mutex_unlock(&log_root_tree->log_mutex);

	ret = update_log_root(trans, log);

	mutex_lock(&log_root_tree->log_mutex);
	if (atomic_dec_and_test(&log_root_tree->log_writers)) {
		smp_mb();
		if (waitqueue_active(&log_root_tree->log_writer_wait))
			wake_up(&log_root_tree->log_writer_wait);
	}

2510
	if (ret) {
2511
		blk_finish_plug(&plug);
2512 2513 2514 2515 2516
		if (ret != -ENOSPC) {
			btrfs_abort_transaction(trans, root, ret);
			mutex_unlock(&log_root_tree->log_mutex);
			goto out;
		}
2517 2518
		root->fs_info->last_trans_log_full_commit = trans->transid;
		btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
2519
		btrfs_free_logged_extents(log, log_transid);
2520 2521 2522 2523 2524
		mutex_unlock(&log_root_tree->log_mutex);
		ret = -EAGAIN;
		goto out;
	}

Y
Yan Zheng 已提交
2525 2526
	index2 = log_root_tree->log_transid % 2;
	if (atomic_read(&log_root_tree->log_commit[index2])) {
2527
		blk_finish_plug(&plug);
2528
		btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
2529 2530
		wait_log_commit(trans, log_root_tree,
				log_root_tree->log_transid);
2531
		btrfs_free_logged_extents(log, log_transid);
Y
Yan Zheng 已提交
2532
		mutex_unlock(&log_root_tree->log_mutex);
2533
		ret = 0;
Y
Yan Zheng 已提交
2534 2535 2536 2537
		goto out;
	}
	atomic_set(&log_root_tree->log_commit[index2], 1);

2538 2539 2540 2541 2542 2543
	if (atomic_read(&log_root_tree->log_commit[(index2 + 1) % 2])) {
		wait_log_commit(trans, log_root_tree,
				log_root_tree->log_transid - 1);
	}

	wait_for_writer(trans, log_root_tree);
Y
Yan Zheng 已提交
2544

2545 2546 2547 2548 2549
	/*
	 * now that we've moved on to the tree of log tree roots,
	 * check the full commit flag again
	 */
	if (root->fs_info->last_trans_log_full_commit == trans->transid) {
2550
		blk_finish_plug(&plug);
2551
		btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
2552
		btrfs_free_logged_extents(log, log_transid);
2553 2554 2555 2556
		mutex_unlock(&log_root_tree->log_mutex);
		ret = -EAGAIN;
		goto out_wake_log_root;
	}
Y
Yan Zheng 已提交
2557

2558 2559 2560 2561
	ret = btrfs_write_marked_extents(log_root_tree,
					 &log_root_tree->dirty_log_pages,
					 EXTENT_DIRTY | EXTENT_NEW);
	blk_finish_plug(&plug);
2562 2563
	if (ret) {
		btrfs_abort_transaction(trans, root, ret);
2564
		btrfs_free_logged_extents(log, log_transid);
2565 2566 2567
		mutex_unlock(&log_root_tree->log_mutex);
		goto out_wake_log_root;
	}
2568
	btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
2569 2570 2571
	btrfs_wait_marked_extents(log_root_tree,
				  &log_root_tree->dirty_log_pages,
				  EXTENT_NEW | EXTENT_DIRTY);
2572
	btrfs_wait_logged_extents(log, log_transid);
2573

2574
	btrfs_set_super_log_root(root->fs_info->super_for_commit,
Y
Yan Zheng 已提交
2575
				log_root_tree->node->start);
2576
	btrfs_set_super_log_root_level(root->fs_info->super_for_commit,
Y
Yan Zheng 已提交
2577
				btrfs_header_level(log_root_tree->node));
2578

Y
Yan Zheng 已提交
2579
	log_root_tree->log_transid++;
2580
	smp_mb();
Y
Yan Zheng 已提交
2581 2582 2583 2584 2585 2586 2587 2588 2589 2590

	mutex_unlock(&log_root_tree->log_mutex);

	/*
	 * nobody else is going to jump in and write the the ctree
	 * super here because the log_commit atomic below is protecting
	 * us.  We must be called with a transaction handle pinning
	 * the running transaction open, so a full commit can't hop
	 * in and cause problems either.
	 */
2591 2592 2593 2594 2595
	ret = write_ctree_super(trans, root->fs_info->tree_root, 1);
	if (ret) {
		btrfs_abort_transaction(trans, root, ret);
		goto out_wake_log_root;
	}
Y
Yan Zheng 已提交
2596

2597 2598 2599 2600 2601
	mutex_lock(&root->log_mutex);
	if (root->last_log_commit < log_transid)
		root->last_log_commit = log_transid;
	mutex_unlock(&root->log_mutex);

2602
out_wake_log_root:
Y
Yan Zheng 已提交
2603 2604 2605 2606
	atomic_set(&log_root_tree->log_commit[index2], 0);
	smp_mb();
	if (waitqueue_active(&log_root_tree->log_commit_wait[index2]))
		wake_up(&log_root_tree->log_commit_wait[index2]);
2607
out:
Y
Yan Zheng 已提交
2608 2609 2610 2611
	atomic_set(&root->log_commit[index1], 0);
	smp_mb();
	if (waitqueue_active(&root->log_commit_wait[index1]))
		wake_up(&root->log_commit_wait[index1]);
2612
	return ret;
2613 2614
}

2615 2616
static void free_log_tree(struct btrfs_trans_handle *trans,
			  struct btrfs_root *log)
2617 2618
{
	int ret;
2619 2620
	u64 start;
	u64 end;
2621 2622 2623 2624 2625
	struct walk_control wc = {
		.free = 1,
		.process_func = process_one_buffer
	};

2626 2627 2628 2629
	ret = walk_log_tree(trans, log, &wc);
	/* I don't think this can happen but just in case */
	if (ret)
		btrfs_abort_transaction(trans, log, ret);
2630

C
Chris Mason 已提交
2631
	while (1) {
2632
		ret = find_first_extent_bit(&log->dirty_log_pages,
2633 2634
				0, &start, &end, EXTENT_DIRTY | EXTENT_NEW,
				NULL);
2635 2636 2637
		if (ret)
			break;

2638 2639
		clear_extent_bits(&log->dirty_log_pages, start, end,
				  EXTENT_DIRTY | EXTENT_NEW, GFP_NOFS);
2640 2641
	}

2642 2643 2644 2645 2646 2647 2648 2649
	/*
	 * We may have short-circuited the log tree with the full commit logic
	 * and left ordered extents on our list, so clear these out to keep us
	 * from leaking inodes and memory.
	 */
	btrfs_free_logged_extents(log, 0);
	btrfs_free_logged_extents(log, 1);

Y
Yan Zheng 已提交
2650 2651
	free_extent_buffer(log->node);
	kfree(log);
2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673
}

/*
 * free all the extents used by the tree log.  This should be called
 * at commit time of the full transaction
 */
int btrfs_free_log(struct btrfs_trans_handle *trans, struct btrfs_root *root)
{
	if (root->log_root) {
		free_log_tree(trans, root->log_root);
		root->log_root = NULL;
	}
	return 0;
}

int btrfs_free_log_root_tree(struct btrfs_trans_handle *trans,
			     struct btrfs_fs_info *fs_info)
{
	if (fs_info->log_root_tree) {
		free_log_tree(trans, fs_info->log_root_tree);
		fs_info->log_root_tree = NULL;
	}
2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706
	return 0;
}

/*
 * If both a file and directory are logged, and unlinks or renames are
 * mixed in, we have a few interesting corners:
 *
 * create file X in dir Y
 * link file X to X.link in dir Y
 * fsync file X
 * unlink file X but leave X.link
 * fsync dir Y
 *
 * After a crash we would expect only X.link to exist.  But file X
 * didn't get fsync'd again so the log has back refs for X and X.link.
 *
 * We solve this by removing directory entries and inode backrefs from the
 * log when a file that was logged in the current transaction is
 * unlinked.  Any later fsync will include the updated log entries, and
 * we'll be able to reconstruct the proper directory items from backrefs.
 *
 * This optimizations allows us to avoid relogging the entire inode
 * or the entire directory.
 */
int btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans,
				 struct btrfs_root *root,
				 const char *name, int name_len,
				 struct inode *dir, u64 index)
{
	struct btrfs_root *log;
	struct btrfs_dir_item *di;
	struct btrfs_path *path;
	int ret;
2707
	int err = 0;
2708
	int bytes_del = 0;
L
Li Zefan 已提交
2709
	u64 dir_ino = btrfs_ino(dir);
2710

2711 2712 2713
	if (BTRFS_I(dir)->logged_trans < trans->transid)
		return 0;

2714 2715 2716 2717 2718 2719 2720 2721
	ret = join_running_log_trans(root);
	if (ret)
		return 0;

	mutex_lock(&BTRFS_I(dir)->log_mutex);

	log = root->log_root;
	path = btrfs_alloc_path();
2722 2723 2724 2725
	if (!path) {
		err = -ENOMEM;
		goto out_unlock;
	}
2726

L
Li Zefan 已提交
2727
	di = btrfs_lookup_dir_item(trans, log, path, dir_ino,
2728
				   name, name_len, -1);
2729 2730 2731 2732 2733
	if (IS_ERR(di)) {
		err = PTR_ERR(di);
		goto fail;
	}
	if (di) {
2734 2735
		ret = btrfs_delete_one_dir_name(trans, log, path, di);
		bytes_del += name_len;
2736 2737 2738 2739
		if (ret) {
			err = ret;
			goto fail;
		}
2740
	}
2741
	btrfs_release_path(path);
L
Li Zefan 已提交
2742
	di = btrfs_lookup_dir_index_item(trans, log, path, dir_ino,
2743
					 index, name, name_len, -1);
2744 2745 2746 2747 2748
	if (IS_ERR(di)) {
		err = PTR_ERR(di);
		goto fail;
	}
	if (di) {
2749 2750
		ret = btrfs_delete_one_dir_name(trans, log, path, di);
		bytes_del += name_len;
2751 2752 2753 2754
		if (ret) {
			err = ret;
			goto fail;
		}
2755 2756 2757 2758 2759 2760 2761 2762
	}

	/* update the directory size in the log to reflect the names
	 * we have removed
	 */
	if (bytes_del) {
		struct btrfs_key key;

L
Li Zefan 已提交
2763
		key.objectid = dir_ino;
2764 2765
		key.offset = 0;
		key.type = BTRFS_INODE_ITEM_KEY;
2766
		btrfs_release_path(path);
2767 2768

		ret = btrfs_search_slot(trans, log, &key, path, 0, 1);
2769 2770 2771 2772
		if (ret < 0) {
			err = ret;
			goto fail;
		}
2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787
		if (ret == 0) {
			struct btrfs_inode_item *item;
			u64 i_size;

			item = btrfs_item_ptr(path->nodes[0], path->slots[0],
					      struct btrfs_inode_item);
			i_size = btrfs_inode_size(path->nodes[0], item);
			if (i_size > bytes_del)
				i_size -= bytes_del;
			else
				i_size = 0;
			btrfs_set_inode_size(path->nodes[0], item, i_size);
			btrfs_mark_buffer_dirty(path->nodes[0]);
		} else
			ret = 0;
2788
		btrfs_release_path(path);
2789
	}
2790
fail:
2791
	btrfs_free_path(path);
2792
out_unlock:
2793
	mutex_unlock(&BTRFS_I(dir)->log_mutex);
2794 2795 2796
	if (ret == -ENOSPC) {
		root->fs_info->last_trans_log_full_commit = trans->transid;
		ret = 0;
2797 2798 2799
	} else if (ret < 0)
		btrfs_abort_transaction(trans, root, ret);

2800
	btrfs_end_log_trans(root);
2801

2802
	return err;
2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814
}

/* see comments for btrfs_del_dir_entries_in_log */
int btrfs_del_inode_ref_in_log(struct btrfs_trans_handle *trans,
			       struct btrfs_root *root,
			       const char *name, int name_len,
			       struct inode *inode, u64 dirid)
{
	struct btrfs_root *log;
	u64 index;
	int ret;

2815 2816 2817
	if (BTRFS_I(inode)->logged_trans < trans->transid)
		return 0;

2818 2819 2820 2821 2822 2823
	ret = join_running_log_trans(root);
	if (ret)
		return 0;
	log = root->log_root;
	mutex_lock(&BTRFS_I(inode)->log_mutex);

L
Li Zefan 已提交
2824
	ret = btrfs_del_inode_ref(trans, log, name, name_len, btrfs_ino(inode),
2825 2826
				  dirid, &index);
	mutex_unlock(&BTRFS_I(inode)->log_mutex);
2827 2828 2829
	if (ret == -ENOSPC) {
		root->fs_info->last_trans_log_full_commit = trans->transid;
		ret = 0;
2830 2831
	} else if (ret < 0 && ret != -ENOENT)
		btrfs_abort_transaction(trans, root, ret);
2832
	btrfs_end_log_trans(root);
2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858

	return ret;
}

/*
 * creates a range item in the log for 'dirid'.  first_offset and
 * last_offset tell us which parts of the key space the log should
 * be considered authoritative for.
 */
static noinline int insert_dir_log_key(struct btrfs_trans_handle *trans,
				       struct btrfs_root *log,
				       struct btrfs_path *path,
				       int key_type, u64 dirid,
				       u64 first_offset, u64 last_offset)
{
	int ret;
	struct btrfs_key key;
	struct btrfs_dir_log_item *item;

	key.objectid = dirid;
	key.offset = first_offset;
	if (key_type == BTRFS_DIR_ITEM_KEY)
		key.type = BTRFS_DIR_LOG_ITEM_KEY;
	else
		key.type = BTRFS_DIR_LOG_INDEX_KEY;
	ret = btrfs_insert_empty_item(trans, log, path, &key, sizeof(*item));
2859 2860
	if (ret)
		return ret;
2861 2862 2863 2864 2865

	item = btrfs_item_ptr(path->nodes[0], path->slots[0],
			      struct btrfs_dir_log_item);
	btrfs_set_dir_log_end(path->nodes[0], item, last_offset);
	btrfs_mark_buffer_dirty(path->nodes[0]);
2866
	btrfs_release_path(path);
2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883
	return 0;
}

/*
 * log all the items included in the current transaction for a given
 * directory.  This also creates the range items in the log tree required
 * to replay anything deleted before the fsync
 */
static noinline int log_dir_items(struct btrfs_trans_handle *trans,
			  struct btrfs_root *root, struct inode *inode,
			  struct btrfs_path *path,
			  struct btrfs_path *dst_path, int key_type,
			  u64 min_offset, u64 *last_offset_ret)
{
	struct btrfs_key min_key;
	struct btrfs_root *log = root->log_root;
	struct extent_buffer *src;
2884
	int err = 0;
2885 2886 2887 2888 2889
	int ret;
	int i;
	int nritems;
	u64 first_offset = min_offset;
	u64 last_offset = (u64)-1;
L
Li Zefan 已提交
2890
	u64 ino = btrfs_ino(inode);
2891 2892 2893

	log = root->log_root;

L
Li Zefan 已提交
2894
	min_key.objectid = ino;
2895 2896 2897 2898 2899
	min_key.type = key_type;
	min_key.offset = min_offset;

	path->keep_locks = 1;

2900
	ret = btrfs_search_forward(root, &min_key, path, trans->transid);
2901 2902 2903 2904 2905

	/*
	 * we didn't find anything from this transaction, see if there
	 * is anything at all
	 */
L
Li Zefan 已提交
2906 2907
	if (ret != 0 || min_key.objectid != ino || min_key.type != key_type) {
		min_key.objectid = ino;
2908 2909
		min_key.type = key_type;
		min_key.offset = (u64)-1;
2910
		btrfs_release_path(path);
2911 2912
		ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
		if (ret < 0) {
2913
			btrfs_release_path(path);
2914 2915
			return ret;
		}
L
Li Zefan 已提交
2916
		ret = btrfs_previous_item(root, path, ino, key_type);
2917 2918 2919 2920 2921 2922 2923 2924 2925 2926

		/* if ret == 0 there are items for this type,
		 * create a range to tell us the last key of this type.
		 * otherwise, there are no items in this directory after
		 * *min_offset, and we create a range to indicate that.
		 */
		if (ret == 0) {
			struct btrfs_key tmp;
			btrfs_item_key_to_cpu(path->nodes[0], &tmp,
					      path->slots[0]);
C
Chris Mason 已提交
2927
			if (key_type == tmp.type)
2928 2929 2930 2931 2932 2933
				first_offset = max(min_offset, tmp.offset) + 1;
		}
		goto done;
	}

	/* go backward to find any previous key */
L
Li Zefan 已提交
2934
	ret = btrfs_previous_item(root, path, ino, key_type);
2935 2936 2937 2938 2939 2940 2941 2942
	if (ret == 0) {
		struct btrfs_key tmp;
		btrfs_item_key_to_cpu(path->nodes[0], &tmp, path->slots[0]);
		if (key_type == tmp.type) {
			first_offset = tmp.offset;
			ret = overwrite_item(trans, log, dst_path,
					     path->nodes[0], path->slots[0],
					     &tmp);
2943 2944 2945 2946
			if (ret) {
				err = ret;
				goto done;
			}
2947 2948
		}
	}
2949
	btrfs_release_path(path);
2950 2951 2952

	/* find the first key from this transaction again */
	ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
2953
	if (WARN_ON(ret != 0))
2954 2955 2956 2957 2958 2959
		goto done;

	/*
	 * we have a block from this transaction, log every item in it
	 * from our directory
	 */
C
Chris Mason 已提交
2960
	while (1) {
2961 2962 2963 2964 2965 2966
		struct btrfs_key tmp;
		src = path->nodes[0];
		nritems = btrfs_header_nritems(src);
		for (i = path->slots[0]; i < nritems; i++) {
			btrfs_item_key_to_cpu(src, &min_key, i);

L
Li Zefan 已提交
2967
			if (min_key.objectid != ino || min_key.type != key_type)
2968 2969 2970
				goto done;
			ret = overwrite_item(trans, log, dst_path, src, i,
					     &min_key);
2971 2972 2973 2974
			if (ret) {
				err = ret;
				goto done;
			}
2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987
		}
		path->slots[0] = nritems;

		/*
		 * look ahead to the next item and see if it is also
		 * from this directory and from this transaction
		 */
		ret = btrfs_next_leaf(root, path);
		if (ret == 1) {
			last_offset = (u64)-1;
			goto done;
		}
		btrfs_item_key_to_cpu(path->nodes[0], &tmp, path->slots[0]);
L
Li Zefan 已提交
2988
		if (tmp.objectid != ino || tmp.type != key_type) {
2989 2990 2991 2992 2993 2994 2995
			last_offset = (u64)-1;
			goto done;
		}
		if (btrfs_header_generation(path->nodes[0]) != trans->transid) {
			ret = overwrite_item(trans, log, dst_path,
					     path->nodes[0], path->slots[0],
					     &tmp);
2996 2997 2998 2999
			if (ret)
				err = ret;
			else
				last_offset = tmp.offset;
3000 3001 3002 3003
			goto done;
		}
	}
done:
3004 3005
	btrfs_release_path(path);
	btrfs_release_path(dst_path);
3006

3007 3008 3009 3010 3011 3012 3013
	if (err == 0) {
		*last_offset_ret = last_offset;
		/*
		 * insert the log range keys to indicate where the log
		 * is valid
		 */
		ret = insert_dir_log_key(trans, log, path, key_type,
L
Li Zefan 已提交
3014
					 ino, first_offset, last_offset);
3015 3016 3017 3018
		if (ret)
			err = ret;
	}
	return err;
3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045
}

/*
 * logging directories is very similar to logging inodes, We find all the items
 * from the current transaction and write them to the log.
 *
 * The recovery code scans the directory in the subvolume, and if it finds a
 * key in the range logged that is not present in the log tree, then it means
 * that dir entry was unlinked during the transaction.
 *
 * In order for that scan to work, we must include one key smaller than
 * the smallest logged by this transaction and one key larger than the largest
 * key logged by this transaction.
 */
static noinline int log_directory_changes(struct btrfs_trans_handle *trans,
			  struct btrfs_root *root, struct inode *inode,
			  struct btrfs_path *path,
			  struct btrfs_path *dst_path)
{
	u64 min_key;
	u64 max_key;
	int ret;
	int key_type = BTRFS_DIR_ITEM_KEY;

again:
	min_key = 0;
	max_key = 0;
C
Chris Mason 已提交
3046
	while (1) {
3047 3048 3049
		ret = log_dir_items(trans, root, inode, path,
				    dst_path, key_type, min_key,
				    &max_key);
3050 3051
		if (ret)
			return ret;
3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077
		if (max_key == (u64)-1)
			break;
		min_key = max_key + 1;
	}

	if (key_type == BTRFS_DIR_ITEM_KEY) {
		key_type = BTRFS_DIR_INDEX_KEY;
		goto again;
	}
	return 0;
}

/*
 * a helper function to drop items from the log before we relog an
 * inode.  max_key_type indicates the highest item type to remove.
 * This cannot be run for file data extents because it does not
 * free the extents they point to.
 */
static int drop_objectid_items(struct btrfs_trans_handle *trans,
				  struct btrfs_root *log,
				  struct btrfs_path *path,
				  u64 objectid, int max_key_type)
{
	int ret;
	struct btrfs_key key;
	struct btrfs_key found_key;
3078
	int start_slot;
3079 3080 3081 3082 3083

	key.objectid = objectid;
	key.type = max_key_type;
	key.offset = (u64)-1;

C
Chris Mason 已提交
3084
	while (1) {
3085
		ret = btrfs_search_slot(trans, log, &key, path, -1, 1);
3086
		BUG_ON(ret == 0); /* Logic error */
3087
		if (ret < 0)
3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099
			break;

		if (path->slots[0] == 0)
			break;

		path->slots[0]--;
		btrfs_item_key_to_cpu(path->nodes[0], &found_key,
				      path->slots[0]);

		if (found_key.objectid != objectid)
			break;

3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111
		found_key.offset = 0;
		found_key.type = 0;
		ret = btrfs_bin_search(path->nodes[0], &found_key, 0,
				       &start_slot);

		ret = btrfs_del_items(trans, log, path, start_slot,
				      path->slots[0] - start_slot + 1);
		/*
		 * If start slot isn't 0 then we don't need to re-search, we've
		 * found the last guy with the objectid in this tree.
		 */
		if (ret || start_slot != 0)
3112
			break;
3113
		btrfs_release_path(path);
3114
	}
3115
	btrfs_release_path(path);
3116 3117
	if (ret > 0)
		ret = 0;
3118
	return ret;
3119 3120
}

3121 3122 3123 3124 3125
static void fill_inode_item(struct btrfs_trans_handle *trans,
			    struct extent_buffer *leaf,
			    struct btrfs_inode_item *item,
			    struct inode *inode, int log_inode_only)
{
3126 3127 3128
	struct btrfs_map_token token;

	btrfs_init_map_token(&token);
3129 3130 3131 3132 3133 3134 3135

	if (log_inode_only) {
		/* set the generation to zero so the recover code
		 * can tell the difference between an logging
		 * just to say 'this inode exists' and a logging
		 * to say 'update this inode with these values'
		 */
3136 3137
		btrfs_set_token_inode_generation(leaf, item, 0, &token);
		btrfs_set_token_inode_size(leaf, item, 0, &token);
3138
	} else {
3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172
		btrfs_set_token_inode_generation(leaf, item,
						 BTRFS_I(inode)->generation,
						 &token);
		btrfs_set_token_inode_size(leaf, item, inode->i_size, &token);
	}

	btrfs_set_token_inode_uid(leaf, item, i_uid_read(inode), &token);
	btrfs_set_token_inode_gid(leaf, item, i_gid_read(inode), &token);
	btrfs_set_token_inode_mode(leaf, item, inode->i_mode, &token);
	btrfs_set_token_inode_nlink(leaf, item, inode->i_nlink, &token);

	btrfs_set_token_timespec_sec(leaf, btrfs_inode_atime(item),
				     inode->i_atime.tv_sec, &token);
	btrfs_set_token_timespec_nsec(leaf, btrfs_inode_atime(item),
				      inode->i_atime.tv_nsec, &token);

	btrfs_set_token_timespec_sec(leaf, btrfs_inode_mtime(item),
				     inode->i_mtime.tv_sec, &token);
	btrfs_set_token_timespec_nsec(leaf, btrfs_inode_mtime(item),
				      inode->i_mtime.tv_nsec, &token);

	btrfs_set_token_timespec_sec(leaf, btrfs_inode_ctime(item),
				     inode->i_ctime.tv_sec, &token);
	btrfs_set_token_timespec_nsec(leaf, btrfs_inode_ctime(item),
				      inode->i_ctime.tv_nsec, &token);

	btrfs_set_token_inode_nbytes(leaf, item, inode_get_bytes(inode),
				     &token);

	btrfs_set_token_inode_sequence(leaf, item, inode->i_version, &token);
	btrfs_set_token_inode_transid(leaf, item, trans->transid, &token);
	btrfs_set_token_inode_rdev(leaf, item, inode->i_rdev, &token);
	btrfs_set_token_inode_flags(leaf, item, BTRFS_I(inode)->flags, &token);
	btrfs_set_token_inode_block_group(leaf, item, 0, &token);
3173 3174
}

3175 3176 3177 3178 3179 3180 3181
static int log_inode_item(struct btrfs_trans_handle *trans,
			  struct btrfs_root *log, struct btrfs_path *path,
			  struct inode *inode)
{
	struct btrfs_inode_item *inode_item;
	int ret;

3182 3183
	ret = btrfs_insert_empty_item(trans, log, path,
				      &BTRFS_I(inode)->location,
3184 3185 3186 3187 3188 3189 3190 3191 3192 3193
				      sizeof(*inode_item));
	if (ret && ret != -EEXIST)
		return ret;
	inode_item = btrfs_item_ptr(path->nodes[0], path->slots[0],
				    struct btrfs_inode_item);
	fill_inode_item(trans, path->nodes[0], inode_item, inode, 0);
	btrfs_release_path(path);
	return 0;
}

3194
static noinline int copy_items(struct btrfs_trans_handle *trans,
3195
			       struct inode *inode,
3196 3197 3198 3199 3200 3201
			       struct btrfs_path *dst_path,
			       struct extent_buffer *src,
			       int start_slot, int nr, int inode_only)
{
	unsigned long src_offset;
	unsigned long dst_offset;
3202
	struct btrfs_root *log = BTRFS_I(inode)->root->log_root;
3203 3204 3205 3206 3207 3208 3209
	struct btrfs_file_extent_item *extent;
	struct btrfs_inode_item *inode_item;
	int ret;
	struct btrfs_key *ins_keys;
	u32 *ins_sizes;
	char *ins_data;
	int i;
3210
	struct list_head ordered_sums;
3211
	int skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
3212 3213

	INIT_LIST_HEAD(&ordered_sums);
3214 3215 3216

	ins_data = kmalloc(nr * sizeof(struct btrfs_key) +
			   nr * sizeof(u32), GFP_NOFS);
3217 3218 3219
	if (!ins_data)
		return -ENOMEM;

3220 3221 3222 3223 3224 3225 3226 3227 3228
	ins_sizes = (u32 *)ins_data;
	ins_keys = (struct btrfs_key *)(ins_data + nr * sizeof(u32));

	for (i = 0; i < nr; i++) {
		ins_sizes[i] = btrfs_item_size_nr(src, i + start_slot);
		btrfs_item_key_to_cpu(src, ins_keys + i, i + start_slot);
	}
	ret = btrfs_insert_empty_items(trans, log, dst_path,
				       ins_keys, ins_sizes, nr);
3229 3230 3231 3232
	if (ret) {
		kfree(ins_data);
		return ret;
	}
3233

3234
	for (i = 0; i < nr; i++, dst_path->slots[0]++) {
3235 3236 3237 3238 3239
		dst_offset = btrfs_item_ptr_offset(dst_path->nodes[0],
						   dst_path->slots[0]);

		src_offset = btrfs_item_ptr_offset(src, start_slot + i);

3240
		if (ins_keys[i].type == BTRFS_INODE_ITEM_KEY) {
3241 3242 3243
			inode_item = btrfs_item_ptr(dst_path->nodes[0],
						    dst_path->slots[0],
						    struct btrfs_inode_item);
3244 3245 3246 3247 3248
			fill_inode_item(trans, dst_path->nodes[0], inode_item,
					inode, inode_only == LOG_INODE_EXISTS);
		} else {
			copy_extent_buffer(dst_path->nodes[0], src, dst_offset,
					   src_offset, ins_sizes[i]);
3249
		}
3250

3251 3252 3253 3254
		/* take a reference on file data extents so that truncates
		 * or deletes of this inode don't have to relog the inode
		 * again
		 */
3255 3256
		if (btrfs_key_type(ins_keys + i) == BTRFS_EXTENT_DATA_KEY &&
		    !skip_csum) {
3257 3258 3259 3260
			int found_type;
			extent = btrfs_item_ptr(src, start_slot + i,
						struct btrfs_file_extent_item);

3261 3262 3263
			if (btrfs_file_extent_generation(src, extent) < trans->transid)
				continue;

3264
			found_type = btrfs_file_extent_type(src, extent);
3265
			if (found_type == BTRFS_FILE_EXTENT_REG) {
3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276
				u64 ds, dl, cs, cl;
				ds = btrfs_file_extent_disk_bytenr(src,
								extent);
				/* ds == 0 is a hole */
				if (ds == 0)
					continue;

				dl = btrfs_file_extent_disk_num_bytes(src,
								extent);
				cs = btrfs_file_extent_offset(src, extent);
				cl = btrfs_file_extent_num_bytes(src,
3277
								extent);
3278 3279 3280 3281 3282
				if (btrfs_file_extent_compression(src,
								  extent)) {
					cs = 0;
					cl = dl;
				}
3283 3284 3285 3286

				ret = btrfs_lookup_csums_range(
						log->fs_info->csum_root,
						ds + cs, ds + cs + cl - 1,
A
Arne Jansen 已提交
3287
						&ordered_sums, 0);
3288 3289 3290 3291 3292
				if (ret) {
					btrfs_release_path(dst_path);
					kfree(ins_data);
					return ret;
				}
3293 3294 3295 3296 3297
			}
		}
	}

	btrfs_mark_buffer_dirty(dst_path->nodes[0]);
3298
	btrfs_release_path(dst_path);
3299
	kfree(ins_data);
3300 3301 3302 3303 3304

	/*
	 * we have to do this after the loop above to avoid changing the
	 * log tree while trying to change the log tree.
	 */
3305
	ret = 0;
C
Chris Mason 已提交
3306
	while (!list_empty(&ordered_sums)) {
3307 3308 3309
		struct btrfs_ordered_sum *sums = list_entry(ordered_sums.next,
						   struct btrfs_ordered_sum,
						   list);
3310 3311
		if (!ret)
			ret = btrfs_csum_file_blocks(trans, log, sums);
3312 3313 3314
		list_del(&sums->list);
		kfree(sums);
	}
3315
	return ret;
3316 3317
}

J
Josef Bacik 已提交
3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333
static int extent_cmp(void *priv, struct list_head *a, struct list_head *b)
{
	struct extent_map *em1, *em2;

	em1 = list_entry(a, struct extent_map, list);
	em2 = list_entry(b, struct extent_map, list);

	if (em1->start < em2->start)
		return -1;
	else if (em1->start > em2->start)
		return 1;
	return 0;
}

static int log_one_extent(struct btrfs_trans_handle *trans,
			  struct inode *inode, struct btrfs_root *root,
3334
			  struct extent_map *em, struct btrfs_path *path)
J
Josef Bacik 已提交
3335 3336
{
	struct btrfs_root *log = root->log_root;
3337 3338
	struct btrfs_file_extent_item *fi;
	struct extent_buffer *leaf;
3339
	struct btrfs_ordered_extent *ordered;
3340
	struct list_head ordered_sums;
3341
	struct btrfs_map_token token;
J
Josef Bacik 已提交
3342
	struct btrfs_key key;
3343 3344 3345 3346
	u64 mod_start = em->mod_start;
	u64 mod_len = em->mod_len;
	u64 csum_offset;
	u64 csum_len;
3347 3348
	u64 extent_offset = em->start - em->orig_start;
	u64 block_len;
J
Josef Bacik 已提交
3349
	int ret;
3350
	int index = log->log_transid % 2;
3351
	bool skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
J
Josef Bacik 已提交
3352

J
Josef Bacik 已提交
3353 3354 3355 3356 3357
	ret = __btrfs_drop_extents(trans, log, inode, path, em->start,
				   em->start + em->len, NULL, 0);
	if (ret)
		return ret;

3358
	INIT_LIST_HEAD(&ordered_sums);
3359
	btrfs_init_map_token(&token);
3360 3361 3362 3363 3364
	key.objectid = btrfs_ino(inode);
	key.type = BTRFS_EXTENT_DATA_KEY;
	key.offset = em->start;

	ret = btrfs_insert_empty_item(trans, log, path, &key, sizeof(*fi));
J
Josef Bacik 已提交
3365
	if (ret)
3366 3367 3368 3369
		return ret;
	leaf = path->nodes[0];
	fi = btrfs_item_ptr(leaf, path->slots[0],
			    struct btrfs_file_extent_item);
3370

3371 3372
	btrfs_set_token_file_extent_generation(leaf, fi, em->generation,
					       &token);
3373 3374
	if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) {
		skip_csum = true;
3375 3376 3377
		btrfs_set_token_file_extent_type(leaf, fi,
						 BTRFS_FILE_EXTENT_PREALLOC,
						 &token);
3378
	} else {
3379 3380 3381
		btrfs_set_token_file_extent_type(leaf, fi,
						 BTRFS_FILE_EXTENT_REG,
						 &token);
3382
		if (em->block_start == EXTENT_MAP_HOLE)
3383 3384 3385 3386 3387
			skip_csum = true;
	}

	block_len = max(em->block_len, em->orig_block_len);
	if (em->compress_type != BTRFS_COMPRESS_NONE) {
3388 3389 3390 3391 3392
		btrfs_set_token_file_extent_disk_bytenr(leaf, fi,
							em->block_start,
							&token);
		btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, block_len,
							   &token);
3393
	} else if (em->block_start < EXTENT_MAP_LAST_BYTE) {
3394 3395 3396 3397 3398
		btrfs_set_token_file_extent_disk_bytenr(leaf, fi,
							em->block_start -
							extent_offset, &token);
		btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, block_len,
							   &token);
3399
	} else {
3400 3401 3402 3403 3404 3405 3406 3407 3408
		btrfs_set_token_file_extent_disk_bytenr(leaf, fi, 0, &token);
		btrfs_set_token_file_extent_disk_num_bytes(leaf, fi, 0,
							   &token);
	}

	btrfs_set_token_file_extent_offset(leaf, fi,
					   em->start - em->orig_start,
					   &token);
	btrfs_set_token_file_extent_num_bytes(leaf, fi, em->len, &token);
J
Josef Bacik 已提交
3409
	btrfs_set_token_file_extent_ram_bytes(leaf, fi, em->ram_bytes, &token);
3410 3411 3412 3413
	btrfs_set_token_file_extent_compression(leaf, fi, em->compress_type,
						&token);
	btrfs_set_token_file_extent_encryption(leaf, fi, 0, &token);
	btrfs_set_token_file_extent_other_encoding(leaf, fi, 0, &token);
3414
	btrfs_mark_buffer_dirty(leaf);
3415

3416 3417 3418 3419
	btrfs_release_path(path);
	if (ret) {
		return ret;
	}
3420

3421 3422
	if (skip_csum)
		return 0;
J
Josef Bacik 已提交
3423

3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506
	/*
	 * First check and see if our csums are on our outstanding ordered
	 * extents.
	 */
again:
	spin_lock_irq(&log->log_extents_lock[index]);
	list_for_each_entry(ordered, &log->logged_list[index], log_list) {
		struct btrfs_ordered_sum *sum;

		if (!mod_len)
			break;

		if (ordered->inode != inode)
			continue;

		if (ordered->file_offset + ordered->len <= mod_start ||
		    mod_start + mod_len <= ordered->file_offset)
			continue;

		/*
		 * We are going to copy all the csums on this ordered extent, so
		 * go ahead and adjust mod_start and mod_len in case this
		 * ordered extent has already been logged.
		 */
		if (ordered->file_offset > mod_start) {
			if (ordered->file_offset + ordered->len >=
			    mod_start + mod_len)
				mod_len = ordered->file_offset - mod_start;
			/*
			 * If we have this case
			 *
			 * |--------- logged extent ---------|
			 *       |----- ordered extent ----|
			 *
			 * Just don't mess with mod_start and mod_len, we'll
			 * just end up logging more csums than we need and it
			 * will be ok.
			 */
		} else {
			if (ordered->file_offset + ordered->len <
			    mod_start + mod_len) {
				mod_len = (mod_start + mod_len) -
					(ordered->file_offset + ordered->len);
				mod_start = ordered->file_offset +
					ordered->len;
			} else {
				mod_len = 0;
			}
		}

		/*
		 * To keep us from looping for the above case of an ordered
		 * extent that falls inside of the logged extent.
		 */
		if (test_and_set_bit(BTRFS_ORDERED_LOGGED_CSUM,
				     &ordered->flags))
			continue;
		atomic_inc(&ordered->refs);
		spin_unlock_irq(&log->log_extents_lock[index]);
		/*
		 * we've dropped the lock, we must either break or
		 * start over after this.
		 */

		wait_event(ordered->wait, ordered->csum_bytes_left == 0);

		list_for_each_entry(sum, &ordered->list, list) {
			ret = btrfs_csum_file_blocks(trans, log, sum);
			if (ret) {
				btrfs_put_ordered_extent(ordered);
				goto unlocked;
			}
		}
		btrfs_put_ordered_extent(ordered);
		goto again;

	}
	spin_unlock_irq(&log->log_extents_lock[index]);
unlocked:

	if (!mod_len || ret)
		return ret;

3507 3508 3509 3510 3511 3512 3513
	if (em->compress_type) {
		csum_offset = 0;
		csum_len = block_len;
	} else {
		csum_offset = mod_start - em->start;
		csum_len = mod_len;
	}
3514

3515 3516 3517 3518 3519 3520 3521
	/* block start is already adjusted for the file extent offset. */
	ret = btrfs_lookup_csums_range(log->fs_info->csum_root,
				       em->block_start + csum_offset,
				       em->block_start + csum_offset +
				       csum_len - 1, &ordered_sums, 0);
	if (ret)
		return ret;
J
Josef Bacik 已提交
3522

3523 3524 3525 3526 3527 3528 3529 3530
	while (!list_empty(&ordered_sums)) {
		struct btrfs_ordered_sum *sums = list_entry(ordered_sums.next,
						   struct btrfs_ordered_sum,
						   list);
		if (!ret)
			ret = btrfs_csum_file_blocks(trans, log, sums);
		list_del(&sums->list);
		kfree(sums);
J
Josef Bacik 已提交
3531 3532
	}

3533
	return ret;
J
Josef Bacik 已提交
3534 3535 3536 3537 3538
}

static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans,
				     struct btrfs_root *root,
				     struct inode *inode,
3539
				     struct btrfs_path *path)
J
Josef Bacik 已提交
3540 3541 3542 3543 3544 3545
{
	struct extent_map *em, *n;
	struct list_head extents;
	struct extent_map_tree *tree = &BTRFS_I(inode)->extent_tree;
	u64 test_gen;
	int ret = 0;
3546
	int num = 0;
J
Josef Bacik 已提交
3547 3548 3549 3550 3551 3552 3553 3554

	INIT_LIST_HEAD(&extents);

	write_lock(&tree->lock);
	test_gen = root->fs_info->last_trans_committed;

	list_for_each_entry_safe(em, n, &tree->modified_extents, list) {
		list_del_init(&em->list);
3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567

		/*
		 * Just an arbitrary number, this can be really CPU intensive
		 * once we start getting a lot of extents, and really once we
		 * have a bunch of extents we just want to commit since it will
		 * be faster.
		 */
		if (++num > 32768) {
			list_del_init(&tree->modified_extents);
			ret = -EFBIG;
			goto process;
		}

J
Josef Bacik 已提交
3568 3569
		if (em->generation <= test_gen)
			continue;
3570 3571 3572
		/* Need a ref to keep it from getting evicted from cache */
		atomic_inc(&em->refs);
		set_bit(EXTENT_FLAG_LOGGING, &em->flags);
J
Josef Bacik 已提交
3573
		list_add_tail(&em->list, &extents);
3574
		num++;
J
Josef Bacik 已提交
3575 3576 3577 3578
	}

	list_sort(NULL, &extents, extent_cmp);

3579
process:
J
Josef Bacik 已提交
3580 3581 3582 3583 3584 3585 3586 3587 3588
	while (!list_empty(&extents)) {
		em = list_entry(extents.next, struct extent_map, list);

		list_del_init(&em->list);

		/*
		 * If we had an error we just need to delete everybody from our
		 * private list.
		 */
3589
		if (ret) {
3590
			clear_em_logging(tree, em);
3591
			free_extent_map(em);
J
Josef Bacik 已提交
3592
			continue;
3593 3594 3595
		}

		write_unlock(&tree->lock);
J
Josef Bacik 已提交
3596

3597
		ret = log_one_extent(trans, inode, root, em, path);
3598
		write_lock(&tree->lock);
3599 3600
		clear_em_logging(tree, em);
		free_extent_map(em);
J
Josef Bacik 已提交
3601
	}
3602 3603
	WARN_ON(!list_empty(&extents));
	write_unlock(&tree->lock);
J
Josef Bacik 已提交
3604 3605 3606 3607 3608

	btrfs_release_path(path);
	return ret;
}

3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622
/* log a single inode in the tree log.
 * At least one parent directory for this inode must exist in the tree
 * or be logged already.
 *
 * Any items from this inode changed by the current transaction are copied
 * to the log tree.  An extra reference is taken on any extents in this
 * file, allowing us to avoid a whole pile of corner cases around logging
 * blocks that have been removed from the tree.
 *
 * See LOG_INODE_ALL and related defines for a description of what inode_only
 * does.
 *
 * This handles both files and directories.
 */
3623
static int btrfs_log_inode(struct btrfs_trans_handle *trans,
3624 3625 3626 3627 3628 3629 3630 3631
			     struct btrfs_root *root, struct inode *inode,
			     int inode_only)
{
	struct btrfs_path *path;
	struct btrfs_path *dst_path;
	struct btrfs_key min_key;
	struct btrfs_key max_key;
	struct btrfs_root *log = root->log_root;
3632
	struct extent_buffer *src = NULL;
3633
	int err = 0;
3634
	int ret;
3635
	int nritems;
3636 3637
	int ins_start_slot = 0;
	int ins_nr;
J
Josef Bacik 已提交
3638
	bool fast_search = false;
L
Li Zefan 已提交
3639
	u64 ino = btrfs_ino(inode);
3640 3641

	path = btrfs_alloc_path();
3642 3643
	if (!path)
		return -ENOMEM;
3644
	dst_path = btrfs_alloc_path();
3645 3646 3647 3648
	if (!dst_path) {
		btrfs_free_path(path);
		return -ENOMEM;
	}
3649

L
Li Zefan 已提交
3650
	min_key.objectid = ino;
3651 3652 3653
	min_key.type = BTRFS_INODE_ITEM_KEY;
	min_key.offset = 0;

L
Li Zefan 已提交
3654
	max_key.objectid = ino;
3655 3656


J
Josef Bacik 已提交
3657
	/* today the code can only do partial logging of directories */
3658 3659 3660 3661
	if (S_ISDIR(inode->i_mode) ||
	    (!test_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
		       &BTRFS_I(inode)->runtime_flags) &&
	     inode_only == LOG_INODE_EXISTS))
3662 3663 3664 3665 3666
		max_key.type = BTRFS_XATTR_ITEM_KEY;
	else
		max_key.type = (u8)-1;
	max_key.offset = (u64)-1;

3667 3668 3669 3670 3671 3672 3673 3674 3675
	/* Only run delayed items if we are a dir or a new file */
	if (S_ISDIR(inode->i_mode) ||
	    BTRFS_I(inode)->generation > root->fs_info->last_trans_committed) {
		ret = btrfs_commit_inode_delayed_items(trans, inode);
		if (ret) {
			btrfs_free_path(path);
			btrfs_free_path(dst_path);
			return ret;
		}
3676 3677
	}

3678 3679
	mutex_lock(&BTRFS_I(inode)->log_mutex);

3680 3681
	btrfs_get_logged_extents(log, inode);

3682 3683 3684 3685 3686 3687 3688 3689 3690
	/*
	 * a brute force approach to making sure we get the most uptodate
	 * copies of everything.
	 */
	if (S_ISDIR(inode->i_mode)) {
		int max_key_type = BTRFS_DIR_LOG_INDEX_KEY;

		if (inode_only == LOG_INODE_EXISTS)
			max_key_type = BTRFS_XATTR_ITEM_KEY;
L
Li Zefan 已提交
3691
		ret = drop_objectid_items(trans, log, path, ino, max_key_type);
3692
	} else {
J
Josef Bacik 已提交
3693 3694
		if (test_and_clear_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
				       &BTRFS_I(inode)->runtime_flags)) {
3695 3696
			clear_bit(BTRFS_INODE_COPY_EVERYTHING,
				  &BTRFS_I(inode)->runtime_flags);
J
Josef Bacik 已提交
3697 3698
			ret = btrfs_truncate_inode_items(trans, log,
							 inode, 0, 0);
3699
		} else if (test_and_clear_bit(BTRFS_INODE_COPY_EVERYTHING,
3700 3701
					      &BTRFS_I(inode)->runtime_flags) ||
			   inode_only == LOG_INODE_EXISTS) {
3702 3703
			if (inode_only == LOG_INODE_ALL)
				fast_search = true;
3704
			max_key.type = BTRFS_XATTR_ITEM_KEY;
J
Josef Bacik 已提交
3705
			ret = drop_objectid_items(trans, log, path, ino,
3706
						  max_key.type);
3707 3708 3709 3710 3711 3712 3713 3714 3715
		} else {
			if (inode_only == LOG_INODE_ALL)
				fast_search = true;
			ret = log_inode_item(trans, log, dst_path, inode);
			if (ret) {
				err = ret;
				goto out_unlock;
			}
			goto log_extents;
J
Josef Bacik 已提交
3716
		}
3717

3718
	}
3719 3720 3721 3722
	if (ret) {
		err = ret;
		goto out_unlock;
	}
3723 3724
	path->keep_locks = 1;

C
Chris Mason 已提交
3725
	while (1) {
3726
		ins_nr = 0;
3727
		ret = btrfs_search_forward(root, &min_key,
3728
					   path, trans->transid);
3729 3730
		if (ret != 0)
			break;
3731
again:
3732
		/* note, ins_nr might be > 0 here, cleanup outside the loop */
L
Li Zefan 已提交
3733
		if (min_key.objectid != ino)
3734 3735 3736
			break;
		if (min_key.type > max_key.type)
			break;
3737

3738
		src = path->nodes[0];
3739 3740 3741 3742 3743 3744 3745
		if (ins_nr && ins_start_slot + ins_nr == path->slots[0]) {
			ins_nr++;
			goto next_slot;
		} else if (!ins_nr) {
			ins_start_slot = path->slots[0];
			ins_nr = 1;
			goto next_slot;
3746 3747
		}

3748
		ret = copy_items(trans, inode, dst_path, src, ins_start_slot,
3749
				 ins_nr, inode_only);
3750 3751 3752 3753
		if (ret) {
			err = ret;
			goto out_unlock;
		}
3754 3755 3756
		ins_nr = 1;
		ins_start_slot = path->slots[0];
next_slot:
3757

3758 3759 3760 3761 3762 3763 3764
		nritems = btrfs_header_nritems(path->nodes[0]);
		path->slots[0]++;
		if (path->slots[0] < nritems) {
			btrfs_item_key_to_cpu(path->nodes[0], &min_key,
					      path->slots[0]);
			goto again;
		}
3765
		if (ins_nr) {
3766
			ret = copy_items(trans, inode, dst_path, src,
3767 3768
					 ins_start_slot,
					 ins_nr, inode_only);
3769 3770 3771 3772
			if (ret) {
				err = ret;
				goto out_unlock;
			}
3773 3774
			ins_nr = 0;
		}
3775
		btrfs_release_path(path);
3776

3777
		if (min_key.offset < (u64)-1) {
3778
			min_key.offset++;
3779
		} else if (min_key.type < max_key.type) {
3780
			min_key.type++;
3781 3782
			min_key.offset = 0;
		} else {
3783
			break;
3784
		}
3785
	}
3786
	if (ins_nr) {
3787
		ret = copy_items(trans, inode, dst_path, src, ins_start_slot,
3788
				 ins_nr, inode_only);
3789 3790 3791 3792
		if (ret) {
			err = ret;
			goto out_unlock;
		}
3793 3794
		ins_nr = 0;
	}
J
Josef Bacik 已提交
3795

3796
log_extents:
3797 3798
	btrfs_release_path(path);
	btrfs_release_path(dst_path);
J
Josef Bacik 已提交
3799
	if (fast_search) {
3800
		ret = btrfs_log_changed_extents(trans, root, inode, dst_path);
J
Josef Bacik 已提交
3801 3802 3803 3804
		if (ret) {
			err = ret;
			goto out_unlock;
		}
3805
	} else if (inode_only == LOG_INODE_ALL) {
3806 3807 3808
		struct extent_map_tree *tree = &BTRFS_I(inode)->extent_tree;
		struct extent_map *em, *n;

3809
		write_lock(&tree->lock);
3810 3811
		list_for_each_entry_safe(em, n, &tree->modified_extents, list)
			list_del_init(&em->list);
3812
		write_unlock(&tree->lock);
J
Josef Bacik 已提交
3813 3814
	}

3815
	if (inode_only == LOG_INODE_ALL && S_ISDIR(inode->i_mode)) {
3816
		ret = log_directory_changes(trans, root, inode, path, dst_path);
3817 3818 3819 3820
		if (ret) {
			err = ret;
			goto out_unlock;
		}
3821
	}
3822
	BTRFS_I(inode)->logged_trans = trans->transid;
3823
	BTRFS_I(inode)->last_log_commit = BTRFS_I(inode)->last_sub_trans;
3824
out_unlock:
3825 3826
	if (err)
		btrfs_free_logged_extents(log, log->log_transid);
3827 3828 3829 3830
	mutex_unlock(&BTRFS_I(inode)->log_mutex);

	btrfs_free_path(path);
	btrfs_free_path(dst_path);
3831
	return err;
3832 3833
}

3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844
/*
 * follow the dentry parent pointers up the chain and see if any
 * of the directories in it require a full commit before they can
 * be logged.  Returns zero if nothing special needs to be done or 1 if
 * a full commit is required.
 */
static noinline int check_parent_dirs_for_sync(struct btrfs_trans_handle *trans,
					       struct inode *inode,
					       struct dentry *parent,
					       struct super_block *sb,
					       u64 last_committed)
3845
{
3846 3847
	int ret = 0;
	struct btrfs_root *root;
3848
	struct dentry *old_parent = NULL;
3849
	struct inode *orig_inode = inode;
3850

3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861
	/*
	 * for regular files, if its inode is already on disk, we don't
	 * have to worry about the parents at all.  This is because
	 * we can use the last_unlink_trans field to record renames
	 * and other fun in this file.
	 */
	if (S_ISREG(inode->i_mode) &&
	    BTRFS_I(inode)->generation <= last_committed &&
	    BTRFS_I(inode)->last_unlink_trans <= last_committed)
			goto out;

3862 3863 3864 3865 3866 3867 3868
	if (!S_ISDIR(inode->i_mode)) {
		if (!parent || !parent->d_inode || sb != parent->d_inode->i_sb)
			goto out;
		inode = parent->d_inode;
	}

	while (1) {
3869 3870 3871 3872 3873 3874 3875 3876
		/*
		 * If we are logging a directory then we start with our inode,
		 * not our parents inode, so we need to skipp setting the
		 * logged_trans so that further down in the log code we don't
		 * think this inode has already been logged.
		 */
		if (inode != orig_inode)
			BTRFS_I(inode)->logged_trans = trans->transid;
3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894
		smp_mb();

		if (BTRFS_I(inode)->last_unlink_trans > last_committed) {
			root = BTRFS_I(inode)->root;

			/*
			 * make sure any commits to the log are forced
			 * to be full commits
			 */
			root->fs_info->last_trans_log_full_commit =
				trans->transid;
			ret = 1;
			break;
		}

		if (!parent || !parent->d_inode || sb != parent->d_inode->i_sb)
			break;

3895
		if (IS_ROOT(parent))
3896 3897
			break;

3898 3899 3900
		parent = dget_parent(parent);
		dput(old_parent);
		old_parent = parent;
3901 3902 3903
		inode = parent->d_inode;

	}
3904
	dput(old_parent);
3905
out:
3906 3907 3908 3909 3910 3911 3912 3913 3914
	return ret;
}

/*
 * helper function around btrfs_log_inode to make sure newly created
 * parent directories also end up in the log.  A minimal inode and backref
 * only logging is done of any parent directories that are older than
 * the last committed transaction
 */
3915 3916 3917
static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
			    	  struct btrfs_root *root, struct inode *inode,
			    	  struct dentry *parent, int exists_only)
3918
{
3919
	int inode_only = exists_only ? LOG_INODE_EXISTS : LOG_INODE_ALL;
3920
	struct super_block *sb;
3921
	struct dentry *old_parent = NULL;
3922 3923 3924 3925 3926
	int ret = 0;
	u64 last_committed = root->fs_info->last_trans_committed;

	sb = inode->i_sb;

S
Sage Weil 已提交
3927 3928 3929 3930 3931
	if (btrfs_test_opt(root, NOTREELOG)) {
		ret = 1;
		goto end_no_trans;
	}

3932 3933 3934 3935 3936 3937
	if (root->fs_info->last_trans_log_full_commit >
	    root->fs_info->last_trans_committed) {
		ret = 1;
		goto end_no_trans;
	}

3938 3939 3940 3941 3942 3943
	if (root != BTRFS_I(inode)->root ||
	    btrfs_root_refs(&root->root_item) == 0) {
		ret = 1;
		goto end_no_trans;
	}

3944 3945 3946 3947
	ret = check_parent_dirs_for_sync(trans, inode, parent,
					 sb, last_committed);
	if (ret)
		goto end_no_trans;
3948

3949
	if (btrfs_inode_in_log(inode, trans->transid)) {
3950 3951 3952 3953
		ret = BTRFS_NO_LOG_SYNC;
		goto end_no_trans;
	}

3954 3955 3956
	ret = start_log_trans(trans, root);
	if (ret)
		goto end_trans;
3957

3958
	ret = btrfs_log_inode(trans, root, inode, inode_only);
3959 3960
	if (ret)
		goto end_trans;
3961

3962 3963 3964 3965 3966 3967 3968 3969
	/*
	 * for regular files, if its inode is already on disk, we don't
	 * have to worry about the parents at all.  This is because
	 * we can use the last_unlink_trans field to record renames
	 * and other fun in this file.
	 */
	if (S_ISREG(inode->i_mode) &&
	    BTRFS_I(inode)->generation <= last_committed &&
3970 3971 3972 3973
	    BTRFS_I(inode)->last_unlink_trans <= last_committed) {
		ret = 0;
		goto end_trans;
	}
3974 3975

	inode_only = LOG_INODE_EXISTS;
3976 3977
	while (1) {
		if (!parent || !parent->d_inode || sb != parent->d_inode->i_sb)
3978 3979
			break;

3980
		inode = parent->d_inode;
3981 3982 3983
		if (root != BTRFS_I(inode)->root)
			break;

3984 3985 3986
		if (BTRFS_I(inode)->generation >
		    root->fs_info->last_trans_committed) {
			ret = btrfs_log_inode(trans, root, inode, inode_only);
3987 3988
			if (ret)
				goto end_trans;
3989
		}
3990
		if (IS_ROOT(parent))
3991
			break;
3992

3993 3994 3995
		parent = dget_parent(parent);
		dput(old_parent);
		old_parent = parent;
3996
	}
3997
	ret = 0;
3998
end_trans:
3999
	dput(old_parent);
4000 4001 4002 4003
	if (ret < 0) {
		root->fs_info->last_trans_log_full_commit = trans->transid;
		ret = 1;
	}
4004 4005 4006
	btrfs_end_log_trans(root);
end_no_trans:
	return ret;
4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017
}

/*
 * it is not safe to log dentry if the chunk root has added new
 * chunks.  This returns 0 if the dentry was logged, and 1 otherwise.
 * If this returns 1, you must commit the transaction to safely get your
 * data on disk.
 */
int btrfs_log_dentry_safe(struct btrfs_trans_handle *trans,
			  struct btrfs_root *root, struct dentry *dentry)
{
4018 4019 4020 4021 4022 4023 4024
	struct dentry *parent = dget_parent(dentry);
	int ret;

	ret = btrfs_log_inode_parent(trans, root, dentry->d_inode, parent, 0);
	dput(parent);

	return ret;
4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046
}

/*
 * should be called during mount to recover any replay any log trees
 * from the FS
 */
int btrfs_recover_log_trees(struct btrfs_root *log_root_tree)
{
	int ret;
	struct btrfs_path *path;
	struct btrfs_trans_handle *trans;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct btrfs_key tmp_key;
	struct btrfs_root *log;
	struct btrfs_fs_info *fs_info = log_root_tree->fs_info;
	struct walk_control wc = {
		.process_func = process_one_buffer,
		.stage = 0,
	};

	path = btrfs_alloc_path();
T
Tsutomu Itoh 已提交
4047 4048 4049 4050
	if (!path)
		return -ENOMEM;

	fs_info->log_root_recovering = 1;
4051

4052
	trans = btrfs_start_transaction(fs_info->tree_root, 0);
4053 4054 4055 4056
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto error;
	}
4057 4058 4059 4060

	wc.trans = trans;
	wc.pin = 1;

T
Tsutomu Itoh 已提交
4061
	ret = walk_log_tree(trans, log_root_tree, &wc);
4062 4063 4064 4065 4066
	if (ret) {
		btrfs_error(fs_info, ret, "Failed to pin buffers while "
			    "recovering log root tree.");
		goto error;
	}
4067 4068 4069 4070 4071 4072

again:
	key.objectid = BTRFS_TREE_LOG_OBJECTID;
	key.offset = (u64)-1;
	btrfs_set_key_type(&key, BTRFS_ROOT_ITEM_KEY);

C
Chris Mason 已提交
4073
	while (1) {
4074
		ret = btrfs_search_slot(NULL, log_root_tree, &key, path, 0, 0);
4075 4076 4077 4078 4079 4080

		if (ret < 0) {
			btrfs_error(fs_info, ret,
				    "Couldn't find tree log root.");
			goto error;
		}
4081 4082 4083 4084 4085 4086 4087
		if (ret > 0) {
			if (path->slots[0] == 0)
				break;
			path->slots[0]--;
		}
		btrfs_item_key_to_cpu(path->nodes[0], &found_key,
				      path->slots[0]);
4088
		btrfs_release_path(path);
4089 4090 4091
		if (found_key.objectid != BTRFS_TREE_LOG_OBJECTID)
			break;

4092
		log = btrfs_read_fs_root(log_root_tree, &found_key);
4093 4094 4095 4096 4097 4098
		if (IS_ERR(log)) {
			ret = PTR_ERR(log);
			btrfs_error(fs_info, ret,
				    "Couldn't read tree log root.");
			goto error;
		}
4099 4100 4101 4102 4103 4104

		tmp_key.objectid = found_key.offset;
		tmp_key.type = BTRFS_ROOT_ITEM_KEY;
		tmp_key.offset = (u64)-1;

		wc.replay_dest = btrfs_read_fs_root_no_name(fs_info, &tmp_key);
4105 4106
		if (IS_ERR(wc.replay_dest)) {
			ret = PTR_ERR(wc.replay_dest);
4107 4108 4109
			free_extent_buffer(log->node);
			free_extent_buffer(log->commit_root);
			kfree(log);
4110 4111 4112 4113
			btrfs_error(fs_info, ret, "Couldn't read target root "
				    "for tree log recovery.");
			goto error;
		}
4114

Y
Yan Zheng 已提交
4115
		wc.replay_dest->log_root = log;
4116
		btrfs_record_root_in_trans(trans, wc.replay_dest);
4117 4118
		ret = walk_log_tree(trans, log, &wc);

4119
		if (!ret && wc.stage == LOG_WALK_REPLAY_ALL) {
4120 4121 4122 4123 4124
			ret = fixup_inode_link_counts(trans, wc.replay_dest,
						      path);
		}

		key.offset = found_key.offset - 1;
Y
Yan Zheng 已提交
4125
		wc.replay_dest->log_root = NULL;
4126
		free_extent_buffer(log->node);
4127
		free_extent_buffer(log->commit_root);
4128 4129
		kfree(log);

4130 4131 4132
		if (ret)
			goto error;

4133 4134 4135
		if (found_key.offset == 0)
			break;
	}
4136
	btrfs_release_path(path);
4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148 4149 4150 4151 4152

	/* step one is to pin it all, step two is to replay just inodes */
	if (wc.pin) {
		wc.pin = 0;
		wc.process_func = replay_one_buffer;
		wc.stage = LOG_WALK_REPLAY_INODES;
		goto again;
	}
	/* step three is to replay everything */
	if (wc.stage < LOG_WALK_REPLAY_ALL) {
		wc.stage++;
		goto again;
	}

	btrfs_free_path(path);

4153 4154 4155 4156 4157
	/* step 4: commit the transaction, which also unpins the blocks */
	ret = btrfs_commit_transaction(trans, fs_info->tree_root);
	if (ret)
		return ret;

4158 4159 4160 4161
	free_extent_buffer(log_root_tree->node);
	log_root_tree->log_root = NULL;
	fs_info->log_root_recovering = 0;
	kfree(log_root_tree);
4162

4163
	return 0;
4164
error:
4165 4166
	if (wc.trans)
		btrfs_end_transaction(wc.trans, fs_info->tree_root);
4167 4168
	btrfs_free_path(path);
	return ret;
4169
}
4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182

/*
 * there are some corner cases where we want to force a full
 * commit instead of allowing a directory to be logged.
 *
 * They revolve around files there were unlinked from the directory, and
 * this function updates the parent directory so that a full commit is
 * properly done if it is fsync'd later after the unlinks are done.
 */
void btrfs_record_unlink_dir(struct btrfs_trans_handle *trans,
			     struct inode *dir, struct inode *inode,
			     int for_rename)
{
4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195
	/*
	 * when we're logging a file, if it hasn't been renamed
	 * or unlinked, and its inode is fully committed on disk,
	 * we don't have to worry about walking up the directory chain
	 * to log its parents.
	 *
	 * So, we use the last_unlink_trans field to put this transid
	 * into the file.  When the file is logged we check it and
	 * don't log the parents if the file is fully on disk.
	 */
	if (S_ISREG(inode->i_mode))
		BTRFS_I(inode)->last_unlink_trans = trans->transid;

4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240
	/*
	 * if this directory was already logged any new
	 * names for this file/dir will get recorded
	 */
	smp_mb();
	if (BTRFS_I(dir)->logged_trans == trans->transid)
		return;

	/*
	 * if the inode we're about to unlink was logged,
	 * the log will be properly updated for any new names
	 */
	if (BTRFS_I(inode)->logged_trans == trans->transid)
		return;

	/*
	 * when renaming files across directories, if the directory
	 * there we're unlinking from gets fsync'd later on, there's
	 * no way to find the destination directory later and fsync it
	 * properly.  So, we have to be conservative and force commits
	 * so the new name gets discovered.
	 */
	if (for_rename)
		goto record;

	/* we can safely do the unlink without any special recording */
	return;

record:
	BTRFS_I(dir)->last_unlink_trans = trans->transid;
}

/*
 * Call this after adding a new name for a file and it will properly
 * update the log to reflect the new name.
 *
 * It will return zero if all goes well, and it will return 1 if a
 * full transaction commit is required.
 */
int btrfs_log_new_name(struct btrfs_trans_handle *trans,
			struct inode *inode, struct inode *old_dir,
			struct dentry *parent)
{
	struct btrfs_root * root = BTRFS_I(inode)->root;

4241 4242 4243 4244 4245 4246 4247
	/*
	 * this will force the logging code to walk the dentry chain
	 * up for the file
	 */
	if (S_ISREG(inode->i_mode))
		BTRFS_I(inode)->last_unlink_trans = trans->transid;

4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260
	/*
	 * if this inode hasn't been logged and directory we're renaming it
	 * from hasn't been logged, we don't need to log it
	 */
	if (BTRFS_I(inode)->logged_trans <=
	    root->fs_info->last_trans_committed &&
	    (!old_dir || BTRFS_I(old_dir)->logged_trans <=
		    root->fs_info->last_trans_committed))
		return 0;

	return btrfs_log_inode_parent(trans, root, inode, parent, 1);
}